• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.12 no.1
• /
• pp.111-122
• /
• 2009

Abatements of TN and ${NO_3}^-$-N in a horizontal subsurface-flow wetland with litter layer on its surface were compared with those without one. The wetland was constructed in 2001 on floodplain of the Gwangju Stream which flows through Gwangju City in Korea. Its dimensions were 29m in length, 9m in width and 0.65m in depth. A bottom layer of 45cm was filled with crushed granites (15~40mm in diameter) and a middle layer of 10cm had pea pebbles. An upper layer of 5cm contained coarse sands. Reeds (Phragmites australis) growing in natural wetlands were transplanted on its surface. Water of the stream was channelled into the wetland by gravity flow and its effluent was discharged back into the stream. Average Litter layer of 12.2cm was formed on its surface in 2007. The layer and above-ground parts of reeds were eliminated in April 2008. Volumes and water quality of influent and effluent of the wetland were analyzed from May to November in 2007 and 2008, respectively. Inflow into the wetland both in 2007 and 2008 averaged approximately 40$m^3$/day and hydraulic residence time both in 2007 and 2008 was about 1.5days. Influent TN concentration in 2007 and 2008 averaged 3.96 and 3.89mg/L, respectively and average influent ${NO_3}^-$-N concentration in 2007 and 2008 was 2.11 and 2.05mg/L, respectively. With a 0.05 significance level, influent concentrations of TN and ${NO_3}^-$-N, temperatures and pH of effluent, and heights and stem numbers of reeds showed no difference between the wetland with litter layer and without one. TN retention in the wetland with litter layer and without one averaged 64,76 and 54.69%, respectively and ${NO_3}^-$-N removal averaged 60.83 and 50.61%, respectively. Both TN and ${NO_3}^-$-N abatement rates in the wetland with litter layer were significantly high (TN abatement: p<0,001, ${NO_3}^-$-N abatement: p=0.001) when compared with those without one. The subsurface-flow wetland having litter layer on its surface was more efficient for TN and ${NO_3}^-$-N removal.

• Journal of Korean Society on Water Environment
• /
• v.23 no.1
• /
• pp.155-160
• /
• 2007

Constructed wetland has been widely used for the treatment of sewage, stormwater runoff, industrial wastewater, agricultural runoff, acid mine drainage and landfill leachate. For the removal of nitrogen and phosphorus, uptake by plants and adsorption to media material are the major processes, and, therefore, the selection of media with specific adsorption capacity is the critical factor for the optimal design of wetland along with the selection of appropriate plant species. In this study, two media materials (loess bead and mixed media) possessing different adsorption characteristics for ammonium and phosphate were selected, and their adsorption characteristics were evaluated. In addition, the performance of subsurface-flow wetland systems employing these media was evaluated in both batch and continuous flow systems. With LB medium, beter phosphorus removal was observed, while better ammonia removal was obtained with MM medium. In addition, enhanced removal efficiencies were observed in the wetland systems employing both media and aquatic plants, mainly due to the better environment for microbial growth. As a result, appropriate selection or combination of media with respect to the inflow water quality maybe important factors for the successful design and operation of wetland systems.

• Proceedings of the Korean Society of Environment and Ecology Conference
• /
• 2003.10a
• /
• pp.132-135
• /
• 2003

• Journal of Wetlands Research
• /
• v.21 no.4
• /
• pp.334-343
• /
• 2019

A lab-scale vertical flow subsurface (VFS) wetland composed of three parallel columns with polypropylene synthetic fiber as main substrate was operated. Piggery stormwater diluted from swine excreta was fed to the wetland on the basis of three different hydraulic regimes or hydraulic retention time (HRT) of 2, 4, and 8 days with daily internal recirculation. Then, monitoring of common water quality parameters was carried out. Unexpectedly, an increase of effluent COD concentration accompanying the appearance of foams was observed during a distinct period in the wetland with HRT 2, 4, and 8 days, successively. Subsequently, a series of experiments was conducted to investigate the origin of the foams. Foams and the increase of COD concentration were found to be induced by the release of organic matter from the synthetic polypropylene fiber which was fed with piggery stormwater. Meanwhile, nitrogen removal was found to be enhanced during a period which overlapped the distinct foaming period signifying that foaming played two important functions in biological nitrogen removal. Foams which form rapidly and then burst easily could hold up and then release oxygen for nitrification. Foams which contain organic surfactants could serve as carbon sources for denitrification as well. Hence, nitrogen removal was enhanced during the foaming stage. After that, COD concentration decreased slowly to a level prior to the foaming stage, and nitrogen removal efficiency declined as well.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.6 no.6
• /
• pp.49-55
• /
• 2003

Total phosphorous removal rate was examined of a subsurface-flow treatment wetland system which was constructed on floodplain in the down reach of the Kwangju Stream in Korea from May to June 2001. Its dimensions were 29 meter in length, 9 meter in width and 0.65 meter in depth. A bottom layer of 45 cm in depth was filled with crushed granite with about 15~30 mm in diameter and a middle layer of 10 cm in depth had pea pebbles with about 10 mm in diameter. An upper layer of 5 cm in depth contained course sand. Reeds(Phragmites australis) were transplanted on the surface of the system. They were dug out of natural wetlands and stems were cut at about 40 cm height from their bottom ends. Water of the Kwangju Stream flowed from a submerged dam into it via a pipe by gravity flow and treated effluent was funneled back into the Stream. The number of reed stems increased from 80 stems/$m^2$ in July 2001 to 136 stems/$m^2$ in September 2001. The hight of stems was 44.2 cm in July 2001 and 75.3 cm in September 2001. The establishment of reeds at early operating stage of the system was good. Volume and water quality of inflow and outflow were investigated from July 2001 through December 2001. The average inflow was 40 $m^3$/day and hydraulic detention time was about 1.5 days. The concentration of total phosphorous n influent and effluent was 0.83 and 0.33 mg/L, respectively. The removal rate of total phosphorous averaged about 60%. The removal efficiency was slightly higher, compared with that of subsurface-flow wetlands operating in North America, whose retention rate of total phosphorous was reported to be about 56%. The good abatement rate could be attributed to sedimentation of particle phosphorous in pores of the media and adsorption of phosphorous to the biofilm developed on the surface of them. Increase of standing density of reeds within a few years will develop root zones which may lead to increment in the phosphorous retention rate.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.10 no.3
• /
• pp.1-7
• /
• 2007

For waste water treatment, artificial wetland was constructed in 1998. The size of artificial wetland is 20m${\times}$200m, with sand and gravel as media and Phragmites japonica was implanted. The removal rate of BOD, TN, and TP were 86%, 33% and 25% from June 2004 to November 2005 respectively, while those were 88%, 38% and 55% in 1999. Organic materials and nitrogen compounds are still effectively removed, after 6 years of construction, but the removal efficiency of phosphorus compounds is reduced. So for sustaining of artificial wetland as waste water treatment system, the removal efficiency of phosphorus compounds must be elevated.

• Journal of Korean Society on Water Environment
• /
• v.16 no.4
• /
• pp.491-499
• /
• 2000

Effluent concentration estimation equations for treatment wetland were reviewed with 3 -year experimental data. Four equations from USEPA, WPCF, Kadlec and Knight, and this study were applied to the over 100 data points of 1996 to 1999 study at the pilot plant in Konkuk University. The system was a subsurface flow type and consisted of 60cm depth of sand and reeds, and it worked continuously including winter with domestic sewage from school building. Generally, all the equations demonstrated reasonable agreement with experimental data and they could be used for design process if selected carefully. Among them, the equation from this study showed the best fit for the data. The reason might be not only the equation was derived from the experimental data, but also it included plant coverage parameter in the equation while others did not Plant coverage was proved to be an important parameter in the prediction of the treatment wetland system, and its inclusion in the estimation equation could improve the accuracy. Although existing equations could be used in the wetland design, pilot plant experiment for the anticipated condition and subsequent equation development can provide more reliable equation. It takes time to obtain meaningful data from wetland system. Therefore, timely onset of well organized study is recommended before large scale application of treatment wetland system to either point or nonpoint source pollution abatement.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.45 no.7
• /
• pp.70-82
• /
• 2003

A pilot study was performed from July 1998 to December 2002, including winter performance, to examine seasonal performance of a constructed wetland and subsequent pond system for treatment of sewage in small communities of Korea. Pond was operated as a intermittent-discharge pond during winter period, and continuous flow system during growing season; its effects was evaluated from December 2001 to April 2003. The subsurface flow (SSF) wetland was satisfactory for treating sewage with good removal efficiency even during the winter period. The wetland effluent concentrations of $BOD_5$ and TSS were often higher in winter than in the growing season, but this was explained by the higher loading rates, rather than lower removal efficiency. The relatively poor-quality wetland effluent was further polished during winter in the pond. The upper layer of the pond water column became remarkably clear immediately after ice melt. In the growing season, ponds could be operated as a continuous flow system to remove nutrients and pathogens, and the effluent of pond could be reused as a supplemental irrigation water without risk of infection by sewage-borne pathogens as well as causing adverse effect on growth and yield. Overall, the wetland system was found to be adequate for treating sewage with stable removal efficiency, and the intermittent-discharge pond was found to be effective for further polishing if necessary. Therefore, the combination of a wetland and subsequent pond system and reuse of effluent as crop irrigation water is recommended as a practical alternative to treat sewage in Korean small communities, and partial discharge of pond water in March is suggested.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.4 no.4
• /
• pp.56-63
• /
• 2001

This paper presents treatment efficiency and plant growth of a subsurface-flow constructed wetland system (23 m in length, 6.5 m in width, 0.65 m in depth) over one year after its establishment on floodplain of a stream in June 2000. An upper layer of 10 cm in depth was filled with course sand and the main biological layer of 50 cm depth with crushed stone with 8 - 15 mm in diameter. The system was planted with common reeds (Phragmites australis) grown on pots. Effluent discharged from a secondary-level treatment plant was funneled into it. Reed stems emerging in April 2001 grew up to 145.9cm until July 2001. The number of reed stems in July 2001 increased by about 11 times compared with that just after planting. The system was inundated seven times by storms over the monitoring period. Reeds were slightly bent after flooding, however they returned to almost upright standing in a couple of weeks. Small portion of inside slope of berm was eroded and the system surface had a sedimentation of 2 - 3 mm in depth. The average removal rates for SS, $BOD_5$, T-N and T-P was 73%, 70%, 53%, and 72%, respectively. The purification efficiencies for SS and $BOD_5$ were fairly good. The reduction rates for T-N was relatively low for the period of late fall through winter until early spring due to lower water temperature which retarded microbial nitrification and denitrification mechanisms. Reduction in the concentration of T-P during fall and winter was relatively higher than that during spring. Leach of phosphorous from plant litters lying on system surface and slight resuspension of precipitated phosphorous in substrates resulted in lower reduction for T-P in spring.

• Journal of Korean Society on Water Environment
• /
• v.22 no.3
• /
• pp.546-556
• /
• 2006

The purpose of this study is to improve the polluted stream water quality by pilot-scale five different constructed wetlands (CWs). Cell 1 to 3 are newly designed 2SFCW (Surface-subsurface flow CW) with 1 to 3 flow shifters (FS) in the middle of the wetland system. Cell 4 and 5 are control CW (CCW), but Cell 5 is the same type as Cell 3. The FS, which converts the route of surface and subsurface flow between two wetlands connected in series, was able to enhance the treatability of TN via nitrification and denitrification and of SS due to filtration and sedimentation. The void fraction and dispersion number of Cell 1, 2 and 3 obtained from the RTD analysis were found to be 0.73 and 0.17, respectively. COD and TP removal efficiencies of Cell 1 to 3 were similar to that of Cell 4 and 5. SS removal efficiencies of Cell 1 to 3 and 5 with FS were 5-10% higher than that of Cell 4 without FS. TN removal efficiencies of Cell 1 to 3 were 3-14% higher than that of Cell 4 and 5. The average $R^2$ values of COD, SS, TN and TP obtained from nonlinear regression analysis were similar to the results of other researchers.