You are on page 1of 10

Feasibility of a constructed wetland and wastewater stabiiisation pond system as a sewage reciamation system for agricultural reuse in a decentralised

rural area
J.H. Ham*, CG. Yoon*, J.H. Jeon" and H.C. Kim"
"Department of Environment Science, Konkuk University, 1 Hwayang-dong, Kwangjin-gu, Seoul, Korea (E-mail: "Research Division, Korea Environment Institute, 613 2 Bulgwang-dong, Eunpyeong-gu, Seoul, Korea Abstract The perfomiance of a constructed wetland (CW) and wastewater stabilisation pond (WSP) system for sewage rclamation and paddy rice irrigation in a decentralised rural area was examined using a feasibiiity study. The CW was satisfactory for sewage treatment, with good removal efficiency even in the winter period, but the effluent concentration was relatively high in the winter period owing to the high influent concentration. The CW effluent was further treated in a WSP and the WSP effluent was considered safe for crop irrigation with respect to sewage-bcrne pathogens. Reclaimed water irngation did not adversely affect the yield of rice; on the contrary, it resulted in an approximately 50% greater yield Ihan in controls. Tbe chemical characteristics of the soil did not change significantly during the experimental period of irrigation with reclaimed water, In the winter, CW effluent could be stored and treated in a WSP until the spring; the water could then be discharged or reused for supplemental irrigation dunng the typical Korean spring drought. Overall, sewage treatment and agronomic reuse using a CW-WSP system could be a practical integrated sewage management measure for protecting receiving water bodies and overcoming water shortages in decentralised ruraf areas. Keywords Agncultural reuse; CW-WSP system; decentralised rural area; reclaimed water irrigation; sewage reclamation

| | s

1ig 9.

Z. M ^ g i 2 9 S Tj z 3T *^ g

Introduction Increases in population density and economic development in Korea have resulted in increased deinand.s for water and competition among the water demands of agricultural, residential and indu.strial Several regions in Korea suffer water stress; while the average precipitation in Korea is 1,310 mm/year, it is not evenly distributed spatially or temporally. Furthermore, available freshwater is not always suitable for its intended use because of water quality problems assiiciated with increased pollutant discharges. Increased sewage discharge associated with the growing human population threatens the water quality of receiving water bodies and rural areas often experience water quality prtiblems. Because of waler shortages and poor water quality of receiving water bodies, decentralised wastewater management is of great importance in developing long-term strategies for the management of rural environments. Decentralised wastewater management involves the collection, treatment, disposal and reuse of wastewater from individual homes, clusters of homes and isolated communities at or near the point of generation (Crites and Tchobanoglous, 1998). Centralised wastewater treatment plants are not feasible in rural areas of Korea because of scattered housing patterns and geographical constraints. Small low-technology wastewater treatment plants are preferred in rural areas and one of the practical alternatives under review is constructed wetlands (CW) in combination with wastewater stabilisation ponds (WSP).
doi: 10.2168/WSI.2007.014

Paddy rice culture irrigation is responsible for more than 50% of the total water consumption in Korea. Therefore, irrigation of paddy rice fields could be a practical candidate for treated sewage reuse in terms of both quantity and quality of water. Agronotnic reuse of reclaimed water has been practised for centuries in many parts of the world. The use of reclaimed water for irrigation can provide a vital resource to enhance agricultural productivity and reduce water pollution problems in receiving water bodies. However, reclaimed water differs from regular irrigation water in many aspects. It contains organic and inorganic components that may affect plant growth, different levels of macronutrients that should be taken into consideration as plant nutrients, trace elements that may be present in excessive levels and residual pathogenic microorganisms that can cause public health problems (Metcalf and Eddy, 2003). Therefore, the feasibility of inigation with reclaimed water should be evaluated based on several factors, including toxic chemicals, microorganisms, salinity, and nutrients. The effects of reclaitned water irrigation on rice culture and soil characteristics in experimental plots should be examined. In this study, we examined the feasibility of using a CW-WSP system as a sewage reclamation system for agricultural of water in a decentralised rural area. Materials and methods
Constructed wetiand (CW)

An experimental CW was installed and has been working continuously since 1997 on the campus of Konkuk University in Seoul, Korea. The mean annual rainfall and temperature from 1975-2004 were 1.344mm and 12.2T, respectively. Precipitation primarily occurs during the four summer months, June, July, August and September. The average ambient air temperature in the winter period (December to February) for the last 30 years was - 0 . 2 C and it occasionally dropped below - 10C. The treatment basin (8 m long, 2 m wide, 1 ni deep) was filled with sand, and planted with common reed {Phraf-miies australis Cav.). The walls and bottom of ihe basin were made of concrete, and the bottom slope wa.s 1% towards the outlet (Figure 1). The void ratio and specific gravity of the sand used to fill the basin were 0.36 and 2.64, respectively. The hydraulic loading rate and hydraulic residence time of the system were 6.3 cm/day and 3.5 days, respectively. The CW system operated in a continuous flow without artificial thermal insulation. The .sewage used in the experiment originated from a school building, with toilets being the main source of pollutant discharge. A septic tank had been designed to collect ihe sewiige and discharge it to the public sewer system after pretreatment. The tank had three

Paddy rice Piots

^ ^ ^ J Pond 1


Figure 1 Schematic ot the natural treatment system




compartments and the wastewater in the last compartment was pumped into the storage tank and then allowed to flow into the CW system. The average water quality of the sewage was pH 7.91. 18.80C, 0.40mg/L dis.solved oxygen (DO), 121.96mg/L biological oxygen demand over a 5-day period (BOD5), 68.95 mg/L total suspended solids (TSS), 13.36mg/L total phosphorus (T-P) and 121.23mg/L total nitrogen (T-N).
Wastewater stabilisation pond (WSP)

To investigate lhe feasibility of using a WSP for further polishing of CW effluent, two experimental WSPs were created from concrete along with the CW (Figure I). Each pond measured 2.0 m long. 2.0 m wide and 2.0 m deep. The pond bottoms were 1.5 m below the ground surface to minimise temperature effects on the WSP water colimin, and lhe bottom 0.5 m was filled with sand. The iniUient was introduced up to the 1.9 m level from the bottom of the WSP (water column 1.4 m and sand 0.5 m). During 1 (December-April), the WSP was operated with intermittent discharge; it was filled with CW efiluent in winter for further polishing and was dischiirged in the spring. During phase-2 (May-November), the WSP was operated with continuous flow; the CW effluent was introduced into the WSP and the hydraulic loading rate and hydraulic retention time were 12.5cm/day and II days, respectively. The WSP effluent was used as irrigation water for the paddy rice culture experiment.
Paddy rice culture with recialmed water irrigation

Paddy rice was grown on plots located next to the WSP and each plot was 1.0 m long. 1.0 m wide and 1.0 m deep, with a surface area of about I m~. lUpumbyeo (a Korean rice cultivar) was planted at a rate of one plant per hill and 22 hills per plot on 25 May. Three treatment plots and one control plot were replicated three times, for a total of 12 plots. The three treatments were (I) irrigation with WSP effluent and no fertilisation (WENF); (2) irrigation with WSP effluent and half of the conventional fertilisation (WEHF); and (3) irrigation with WSP effluent and conventional fertilisation (WECF). The control plots (CONTROL) were irrigated with potable water, and conventional fertilisation was applied. A conventional fertilisation rate of 110-70-80 kg/ha N-P2O3-K2O and other standard cultivation practices of the local district were followed. The l-m" plots were irrigated with a totai volume of 430 L of water, in addition to 1.2(X)L of natural precipitation (first year 1,401 L, second year 980 L), to maintain the water level between 5 and 10 cm for the entire crop season. The paddy rice was harvested on 20 October and the yield was measured. Water quality of the irrigation water was analysed using Stattdard Melbods for tbe E.xarnina!ion <if Water and Wastewater (1998) and soil samples were analysed using the Methods of Soil Analysis (1982). Results and discussion
CW performance

The CW worked stably in winter without freezing even at temperatures below lO'^C, and waste removal was still significant. The effluent DO was greater than the influent DO and was generally above 2.0 mg/L. even without auxiliary aeration. Reduction of BOD5 and TSS was apparent in ihe CW system over the entire siudy period. Although the effluent BODs level was higher (p < 0.05) in winter (December-February) than in the growing season (Marc h-November), there was no difference ( p > 0.05) in mass retention of BOD5 between winter and the growing season (Table I ). The high effluent TSS concentrations in winter may be attributed to relatively higher loading rather than to lower removal efficiency related to temperature dependence. T-P removal was relaliveiy poor compared to BODs and TSS, but better than T-N removal. The influent T-P concentrations were in the

Table 1 Summary of constnjcted wetland (CW) performance

Constituents CW influent CW effluent Removai rate

Growing season DO (mg/L) BOD5 (mg/L) TSS {mg/U T-P (mg/L) T-N (mg/L) FCMMPN/IOOmL) Winter DO (mg/U BOD5 (mg/L) TSS (mg/L) T-P (mg/L) T-N (mg/L) FC^(MPN/100mL)

0.3 116.2 62.2 13.4 121.4 693,800 1.2 150.9 102.7 13.0 120.2 564,910

i + i

0.61 59.51 36.05 4.03 39.28 207,299 1.04 75.37 44.61 4.91 46.84 192,486

2.4 20.9 14.0 7.1 93.9 7,667 2.3 62.2 32.8 8.5 108.0 5,256

0.97 17.98 11.40 3.58 35.47 3,250 0.63 48.63 19.14 2,81 36.18 2,158

81.0 71.6 44.0 20,0 98.9

12.78 23.34 33.20 28.00 38.34

61.8 64.8 26.8 7.7 99.1

15.13 20.19 27.15 12.91 40.58

DO. dissolved oxygen; BOD5, biological oxygen demand over a 5-day period; TSS, total suspended solids; T-P, total phosphorus; T-N, total nitrogen; FC, fecal coiiform "Fecal coliform measured as the most probable number (MPN)

same range for both winter and the growing season, but the effluent concentraiion in the growing season was lower than in winter {p < 0.05). The removal of T-N was poor, and the removal efficiency of T-N in winter was even lower than in the growing season because the nitrogen removaJ processes (ammonification, niirificalion and denitrification) were temperature-dependent in Ihe CW system (Kadlec and Knight. 1996). The low T-N removal efficiency in this sludy may have been related to the high influent concen tration (over lCK) mg/L) and the limited hydraulic residence time (3.5 days) of the experimental CW system. Overall, the CW system produced satisfactory effluent concentrations and performed stably in ihe growing season, The average removal rates of constituents in winter were lower than in the growing season, but winter removal performance was still significant. Generally, the CW system was considered to be adequate for decentralised sewage treatment. However, winter performance was a cause for concern and measures for subsequent management of CW effluent should be considered further.

Further polishing by WSP for crop irrigation


C W effluent was further polished using a W S P for crop irrigation in the irrigation season ( M a y - O c t o b e r ) . There w a s n o difference {p > 0.05) in B O D 5 and T S S between C W effluent and W S P effluent because BOD5 and T S S concentrations in the C W effluent were low. T h e average BOD5 and T S S levels in the W S P effluent were 10.8 mg/L and 9.9 mg/L, respectively, and were within an acceptable range. T h e average T - P and T-N removal rate by W S P w a s approximately 28 and 2 5 % . respectively (Table 2). In the s a m e period, the T - P and T - N removal rale by the C W w a s 2 3 and 3 1 % and the total T - P and T - N removal rate of the c o m b i n e d treatment system ( C W -f- W S P ) w a s about 4 6 and 5 1 % , respectively. Fecal coliform ( F C ) concenirations in C W or W S P effluent are coinmonly reduced by at least o n e log unit, but levels below 5 0 0 MPN/IOOniL are difficult to achieve consistently (Tanner and Sukias, 2003). Although the average F C removal rate w a s substantial in the C W , the average concentration remained high, at about 7 . 9 0 0 M P N / l O O m L (Table 2), After paiisage through the W S P , however, F C concentrations dropped another order of magnitude (about 9 6 % ) , averaging 4 3 0 M P N / I O O m L .

Table 2 Summary of the wastewater stabilisation pond (WSP) performance during the irrigation season
Constituents pH CW eHluent WSP effluent Removal rate
_ -

EC (dS/m) DO (mg/L) BOD5 (mg/L) TSS (mg/L) T-P (mg/L) T-N (mg/L) F C (MPN/100 mL)

7.2 1.18 3.8 8.9 10.8 5.7 53.0 7,883,3

0.30 0.365 0.47 2.19 4.44 0.19 2.97 949,90

7.2 1,13 5.3 10.8 9.9 4.1 38.9 431.7

0.81 0,350 0,78 2,11 1.80 0.34 3,27 65.55

28.0 6.57 25.1 6.80 96.1 1.44 %

CW, constructed wetland; constiluents are as in Table 1 ^Fecal coliform measured as the most probable number (MPN)

Assessment of redaimed water irrigation

Although irrigation has been practised ihroughout the world for .iseveral millennia, only in the century has the importance of the quality of irrigation water been recognised. The feasibility of irrigation with reclaimed water was evaluated based on several factors, including microorganism concenlrations, salinity and nutrient levels. in rural areas, toxic chemicals normally occur in low concentrations in reclaimed water and may not cause great health risks. There has been considerable emphasis on the control of microbiological healrh risks, especially regarding municipal wastewater. Based on the World Health Organisation (WHO) microbiological guidelines tor treated wastewater (WHO. 1989, 2000), paddy rice culture irrigation practices may belong to category B2, i.e. flood irrigation with restricted application to non-edible crops and exposure to adult workers only. FC counts should be less than 1,000 MPN/ lOOmL under these condilions. As indicated in Table 2, the FC levels of CW effluent were still too high for reuse as irrigation water. FC levels were further reduced to less than 500MPN/iOOinL on average, showing over 95% removal, in the subsequent WSP treatment, and these levels meet the WHO recommended guidelines (WHO. 1989. 2000). Excess salinity results in salt accumulation in the crop root zone and poor crop yields. The average salinity of WSP effluent was l.l8dS/ni. Based on the Food and Agriculture Organisation (FAO) irrigation water quality guidelines (FAO. 1985), the degree of use restriction for this WSP effluent was slight to moderate (0.7-3.0dS/m). The nutrients in reclaimed water can provide fertiliser value for crop production. The most beneficial nutrient, often present in high concentrations in reclaimed municipal water, is nitrogen. However, excessive nitrogen may be detrimental to many crops, causing excessive vegetative growth, delayed or uneven maturity, or reduced crop quality. The average T-N concentration of WSP effluent was 53 mg/L, More than 30 mg/L is considered excessive, even for crops tolerant of excess nitrogen (FAO, 1985). Problems caused by excessive salinity and nitrogen are unlikely to occur in Ihis case, however, because the reclaimed irrigation water may be blended with about 900 mm of precipitation water, a volume about two or three times greater than that of the irrigation water during the irrigation season (May-October). To reuse WSP effluent as crop irrigation water, blending WSP effluent witb other irrigation water keeps salinity under control (<0.7dS/m), and its supplemental use is recommended. The CW and WSP systems were determined to be an effective and feasible alternative for sewage reclamation in decentralised rural areas, considering their stable perfomiance, effectiveness in the removai of Indicator bacteria, as well as adequate levels of other water quality parameters, low maintenance and cost-effectiveness.

Crop yield

s_ I S ^

The average rice yield in the WECF treatment, which received reclaimed water irrigation and conventional fertilisation, was about 50% greater than that of the CONTROL (Table 3), indicating that reclaimed water irrigation can increase paddy rice yields. The yield in plots with reduced fertilisation (WEHF) was also significantly higher than that of the CONTROL, and the yield in plots without fertilisation (WENF) was within the same range as that of the CONTROL (Table 3). This implies that nutrients in the reclaimed water helped to increase the yield, and made up for part of the conventional fertilisation deficit. The amount of niiirients applied hy fertilisation and reclaimed water irrigation may partly explain the difference in crop yield. The average crop yield for the cxperimental period approximately matched the quantity of nutrient input. As long as the constituent concentrations are not extremely high, reclaimed water irrigation of paddy rice cultures may not be harmful and may even enhance crop yields.
Soli characteristics

The soil characteristics of the experimental plots were analysed to examine the effects of reclaimed water irrigation (Figure 2). The results are expressed as averages of three replicate samples per plot, taken from the root zone after clearing the suriace organic layer (the upper 3-2Ocm of soil). As paddy culture continued, the soil pH values slightly increased in reclaimed water irrigation plots (WENF, WEHF. and WECF). VazqueaMontiel et a. (1996) reported significant increases in soil pH as a result of wastewater application, Where conditions for nitrification are adequate, plants take up mainly nitrate, even when ammonium fertilisers are applied (Menge! and Kirkby, 1987) and when plants take up nitrate ions the growth medium becomes more alkaline (Vazquea-Montiel et ai, 1996). In this study, more nilrate may have been absorbed in the reclaimed water irrigation plots than in the CONTROL plots owing to the high nitrogen input, and this may partially explain the increase in soil pH. Electrical conductivity (EC) increased after the first year of the experiment compared to the initial conditions, but it decreased after the second year (Figure 2). Continuous irrigation with reclaimed water could cause salt accumulation in the soil, but this was not observed in our study. The possibility of sait accumulation may be lower in paddy rice fields than in uplands because of different irrigation water management practices. Water depth was maintained at 5 - lOcm during the irrigation season in (he paddy rice culture and all of the water in paddy rice fields is usually drained before the harvest. As paddy culture continued, the soil SAR values slightly increased in reclaimed water irrigation plots, but the values were not high enough to cause serious physical soil problems. When the SAR rises above 12-15, permeability and aeration problems arise and plants have difficulty absorbing water (Munshower. 1994). Monitoring of soil structural stability and/or permeability should be carried out periodically to detect deterioration that may adversely affect continued reclaimed water irrigation or future land use. The cation exchange capacity (CEC) increased in all plots, including the CONTROL, as rice culture continued
Table 3 Comparison of grain yields in irrigation and fertilisation treatments

Yield (kg/10 a)

1 st year 2nd year

588.6 558.7

672.8 657.9

799.6 752.8

524.4 505.9

120.0 144.6

"Least significant difference WENF, irrigation with WSP effluent and no fertilisation; WEHF, irrigation with WSP effluent and half of the conventional fertilisation; WECF, irrigation with WSP effluent and conventional fertilisation; CONTROL, irrigation with potable water, and conventional fertilisation

1 5

1 5

1 5

1 5

9 1

1 5

1^'year 2"''year Month WENF

2""^ year Month WEHF WECF

9 1 5 9 1 1^' year 2"'^year Month CONTROL

Rgure 2 Analysis of soil samples from experimental plots. EC, electrical conductivity; SAR, sodium adsorption ratio; CEC, cation exchange capacity; T-P, total phosphorus; T-N, total nitrogen

during the study period. There was no difference in CEC between the CONTROL atid olhcr treatments (Figure 2). Total phosphoru.s (T-P) and total nitrogen (T-N) concentrations remained fairly constant during the study period. The soil nutrient concentrations may be little affected by reclaimed water, as long as the water is adequately treated.

Sewage management tn decentralised rural areas

CW-WSP systems can be properly used for the pnxluction of microbiological!y safe reclaimed water for crop irrigation in decentralised rural areas. During the irrigation season, the sewage can first be treated in a CW. and its effluent discharged inlo a WSP designed to achieve the required water quality. CWs can function as the primary and secondary treatment and WSPs as the reclamation system. Irrigation with reclaimed water can not only provide a vital resource for enhancing agricultural productivity, but it can also help increase crop yields, decrease chemical feriiliser use. enhance water qualily in receiving water bodies via zero discharge, and allow a greater allocation of freshwater supplies to urban use (Asano and Levine, 1998). However, CW effluent can only be used for crop irrigation during the irrigation season. At other times of the year, CW effluent must be discharged into receiving water bodies. Relatively poor quality CW effiuent may occur during winter (Table I ) and this CW effluent would require further treatment to prevent water quality degradation of the receiving water bodies. WSPs could be used as intennittent discharge ponds during the winter period to solve this problem. The CW effluent could be stored and treated in WSPs during the winter, and then discharged to the receiving water bodies in the spring. In a previous study (Ham et ai. 2004). 3 months of storage in an intermittent discharge pond provided substantial water quality improvement. WSP water could be discharged to receiving water bodies in April, and used for crop irrigation during the drought period (April and May) in Korea. After WSP discharge in the spring, the WSP could be operated in a continuous flow regime to remove pathogens and nutrients, as described above, and to provide water for crop irrigation. The CW-WSP system can work as a novel

year-round method for sewage treatment and agronomic reuse in decentralised rural area.s. The CW-WSP system can be located in decentralised rural areas and become a part of the rural ecosystem. Theretbre. the ultimate disposal of sewage for agricultural purposes and subsequent land treatment may also be available. ^ X ^ Conclusions A CW was demonstrated to be highly effective at treating sewage in decentralised rural areas during the growing season, but its effiuent was not clean enough for discharge to receiving water bodies. The CW effluent was reclaimed for crop irrigation using a WSP during the growing season. After reclamation with a WSP. the level of FC (<500MPN/100tnL) met WHO guidelines ( 1,000 MPN/lOOniL) for unrestricted irrigation. Therefore, WSP effluent may be used for crop irrigation without much concern for infection by sewage-bome pathogens. Salt and nitrogen concentrations were high in WSP reclaimed water, but were not a limiting factor in this study. Supplemental use of reclaimed water combined with existing sources of irrigation water is recommended, rather than irrigation using only a single water source. Irrigation of reclaimed water on to paddy rice cultures did not adversely affect the yield of rice. On the contrary, experimental rice plots treated with reclaimed water demonstrated about a 507c greater yield on average than control plots. The chemical characteristics of irrigated soil did not vary during the experimental period. Overall, the CW-WSP was found to be an effective sewage reclamation system for decentralised rural areas, and reclaimed water could be reused as a supplemental source of irrigation water for paddy rice culture with little concern for adverse effects. Sewage treatment and agronomic reuse via CW-WSP systems could be practical integrated sewage management measures in decentralised rural areas for protecting the water quality of receiving water bodies and solving water shortages. Acknowledgements This research wa.s supported by a grant (code# 4-5-2) from the Sustainable Water Resources Research Center of the 21st Century Fronlier Research Program in Korea.

Asano, T. anJ Levine. D. ( 19^8). Waslewaier reclamation, recycling, antl reuse: an inirodiiclion. In Wastewater Reclamation and Reuse, Asanii. T. (ed.). Technomic Publishing, Lancaster, PA, USA. pp. 1-56Crites. R. and Tchobanoglous. G. ( 1998). Small and Decentralized Wastewater Management Sy.>tems, McGraw-Hill. Boston Burr Ridge. IL. USA. FAO ( 1985). Water Quality for Agriculture: irrigation and Drainage, Paper No 29 Rev. 1. Frnxl and Agriculture Organisation, Rome. Italy. Ham. J.H.. Yoon. C.G.. Hwang. S.J. and Jung. K.W. (2004). Seawnal performance of constructed wetland and winter storage pond for sewage treatment in Korea. J. Environ. Sei. Health. A39(5). 1325-1339. Kuilkv. R.H. and Knighl. R.L. (1996). Trentment Wetlands. Lewis Publishers. Boca Raton, FT-, USA. Metealf and Bddy. Inc.. (2003). Wastev,-ater Engineering. Treatmeni and 4th edn. McGraw-Hill, Boston BurT Ridge, IL. USA. Methods of Soil Analysis. Part 2: Chemical and Microbiological Properties (1982). 2nd edn.. American Society of Agn>nomy, and Soil Science Society of America, Madison, Wi. USA. Mengel. K. and Kirkby. E.A. (197t. Principies of P tant Nutrition. 4th edn., International Potash Institute, Beme. Switzerland. Munshower. F.F. (1994). Practical Handbook of Disturbed Lund Revegetation, Lewis Publishers. BtKa Raton. FL, USA. Standard Methods for the E.xaminaiion of Water and Wastewater {1998). 2()th edn. American Puhlic Healih Asstwi ation/American Water Works AssociationAVater Environment Federation, Washingltm DC, USA.

Tanner, CC. and Sukias, J.P.S. (2003). Linking pond and wetland treatment: perfonnance of domtistic and farm systems in New Zealand. V/tit. Sei, Tech.. 48(2). 331-3.39. Vazquea-Montiel. O.. Horan, N.J. and Mara. D.D. (1996). Management of domestic wasiewater for reuse in irrigation. Wat. Sd. Tech., 33(ltH I), 355-362. WHO ( 1989). Heaith Guieline.i for the Use nf Wa.itewater in Agriculture and Aquaculture, Technical Report Series No. 778. Worid Health Organization, Geneva. Switzerland. WHO (2000). Guiiicliiies for the Microbiological Quality of Treated Wasiewater Used in Agriculture: Recommendaiion.s for Revising WHO Guidelines, WHO Bulletin 78(9). World Health Organisation. Getieva, Switzerland.

i~ '

Copyright of Water Science & Technology is the property of IWA Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.