- Email: [email protected]
Pilot-scale comparison of constructed wetlands operated under high hydraulic loading rates and attached biofilm reactors for domestic wastewater treatment
Science of the Total Environment 407 (2009) 2996–3003
Contents lists available at ScienceDirect
Science of the Total Environment j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i t o t e n v
Pilot-scale comparison of constructed wetlands operated under high hydraulic loading rates and attached biofilm reactors for domestic wastewater treatment M.S. Fountoulakis a,⁎, S. Terzakis a,b, A. Chatzinotas c, H. Brix d, N. Kalogerakis b, T. Manios a,e a
School of Agricultural Technology, Technological Educational Institute of Crete, Heraklion, Greece Department of Environmental Engineering, Technical University of Crete, Chania, Greece UFZ, Helmholtz Centre for Environmental Research – UFZ, Department of Environmental Microbiology, Permoserstrasse 15, D-04318 Leipzig, Germany d Department of Biological Sciences, Aarhus University, Denmark e Greek Open University, School of Science and Technology, Patras, Greece b c
a r t i c l e
i n f o
Article history: Received 7 July 2008 Received in revised form 23 December 2008 Accepted 5 January 2009 Available online 30 January 2009 Keywords: Constructed wetland Treatment performance Nutrients Rotating biological contactor Packed bed filter
a b s t r a c t Four different pilot-scale treatment units were constructed to compare the feasibility of treating domestic wastewater in the City of Heraklio, Crete, Greece: (a) a free water surface (FWS) wetland system, (b) a horizontal subsurface flow (HSF) wetland system, (c) a rotating biological contactor (RBC), and (d) a packed bed filter (PBF). All units operated in parallel at various hydraulic loading rates (HLR) ranging from 50% to 175% of designed operating HLR. The study was conducted during an 8 month period and showed that COD removal efficiency of HSF was comparable (N 75%) to that of RBC and PBF, whereas that of the FWS system was only 57%. Average nutrient removal efficiencies for FWS, HSF, RBC and PBF were 6%, 21%, 40% and 43%, respectively for total nitrogen and 21%, 39%, 41% and 42%, respectively for total phosphorus. Removals of total coliforms were lowest in FWS and PBF (1.3 log units) and higher in HSF and RBC (2.3 to 2.6 log units). HSF showed slightly lower but comparable effluent quality to that of RBC and PBF systems, but the construction cost and energy requirements for this system are significantly lower. Overall the final decision for the best non-conventional wastewater treatment system depends on the construction and operation cost, the area demand and the required quality of effluent. © 2009 Published by Elsevier B.V.
1. Introduction There is an urgent need to implement efficient treatment of domestic wastewater from small towns as imposed by the Water Framework Directive 2000/60/EC (EU Parliament and Council, 2000). However, the per capita cost for implementation of the common activated sludge process in small communities is much higher than that for large cities. Therefore non-conventional technologies ranging from simple biological low rate systems such as ponds, constructed wetlands and sand filters to complex high-rate suspended and fixed biomass reactors have to be evaluated according to their treatment performance, foot print, process reliability, investment and operation costs (Boller, 1997; Colmenarejo et al., 2006; Fahd et al., 2007). Constructed wetlands (CW) are used worldwide to treat municipal wastewater (Brix, 1994a; He and Mankin, 2002; Nitisoravut and Klomjek, 2005; Paing and Voisin, 2005; Brix and Arias, 2005; Chung et al., 2008). CW systems are based on the functioning of natural ecosystems and the treatment processes involve complex interactions between soil, water, plants and microorganisms. CWs are generally efficient in removal of organic matter (BOD) and suspended solids (SS), but the removal of nitrogen and phosphorus is often relatively poor ⁎ Corresponding author. Tel.: +30 2810 379456; fax: +30 2810 318204. E-mail address: [email protected] (M.S. Fountoulakis). 0048-9697/$ – see front matter © 2009 Published by Elsevier B.V. doi:10.1016/j.scitotenv.2009.01.005
(Verhoeven and Meuleman, 1999; Tanner et al., 1999; Kuschk et al., 2003; Vymazal, 2007). Constructed wetlands can be classified into two main categories depending on how the water passes through the systems: subsurface-flow or surface-flow design. The most widely used CW design in Europe is designed with horizontal subsurface flow (HSF). However, free water surface (FWS) constructed wetlands are increasingly being favoured because they are cheaper to construct and may have higher wildlife habitat values. There is a concern about the feasibility of wetlands to become a cost effective method because wetlands typically require a low hydraulic loading rate (HLR) and a long hydraulic retention time (HRT) to achieve efficient pollutant removal. That means wetland treatment method may need a large land area. USEPA have recommended that the organic loading rate should not exceed 6 g BOD m− 2 d− 1 in HSF (USEPA, 2000) and 11.2 g BOD m− 2 d− 1 in FWS (USEPA, 1988). These suggestions may not be applicable when land is expensive or limited. Using high HLR to operate the constructed wetland may potentially reduce the required area. Recent years there is an effort to study the performance of constructed wetlands under high HLRs. Caselles-Osorio et al. (2007a) found that there was no significant difference in COD removal between two HSF systems operated at an HLR of 6 g COD m− 2 d− 1 and 23 g COD m− 2 d− 1. Another study in France concluded that overloads up to ten times the dry weather flow in vertical flow constructed wetlands are possible while still complying with the European standards (Molle
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
2997
Fig. 1. View of the pilot scale units, a) free water surface constructed wetland, b) subsurface constructed wetland, c) rotating biological contactor, d) packed bed reactor.
et al., 2006). Furthermore, Lin et al. (2005) shown that a unit consisting of FWS and HSF wetland cells treating intensive shrimp aquaculture wastewater effectively removed TSS, BOD and NH4–N under high hydraulic loading rates (157–195 m d− 1). On the other hand, there are also several compact systems using fixed and movable bed reactors with high removal efficiencies for BOD, SS and nitrogen that potentially can be used for small towns. These systems in general have good process stability, less footprint requirements compared to CWs and high specific removal rates (Helmer and Kunst, 1998; Gebara, 1999; Sirianuntapiboon, 2006). Different types of attached biomass technologies have been widely applied as small-scale plants, including packed bed biofilm reactors (Mann and Stephenson, 1997; Aesoy et al., 1998; Schubert and Wolfgang Günthert, 2001) and rotating biological contactors (Hansford et al., 1978; Ayoub and Saikaly, 2004; Tawfik et al., 2006). These systems generally achieve high removal efficiencies when treating wastewater from 25–500 person equivalents (p.e.), however the capital and operations costs for small towns (1000–5000 p.e.) may be very high. The aim of this work was to assess and compare the performance of different non-conventional wastewater treatment systems in controlled experiments under Mediterranean climatic conditions. Four pilot-scale treatment systems were constructed: a) FWS, b) HSF, c) RBC, and d) PBF. All units operated in parallel receiving primarily treated municipal wastewater. In order to decrease land requirements for constructed wetlands so that are more competitive to compact attached biomass systems the FWS and HSF were designed to operate at high hydraulic loading rates. 2. Materials and methods 2.1. Pilot-scale unit description All pilot systems were constructed during summer 2006 in our open-air laboratory with a total surface area of approximately 360 m2.
The facility is located in Heraklion, Crete, South Greece (N 35°, 19q; E 25°, 10q). The FWS system was constructed with dimensions of 12.4 m long and 3.4 m wide, and with three separated zones, two vegetated zones and one deeper anoxic un-vegetated zone in the centre (Fig. 1a). A soil layer of 40 cm depth was added in the vegetated zones, and was planted with two species of reeds, Phragmites australis and Arundo donax. Plants were transplanted until a total cover of 40% was reached. Then the wetland was filled with tap water to a depth of 50 cm. Wastewater was mixed with tap water at gradually increasing wastewater/tap water ratios until only wastewater was added after 3 weeks. The incoming wastewater entered the wetland through a 40 cm gravel layer to distribute the water across the width of the bed. The unit was designed to treat 6 m3 of domestic wastewater daily (HLR c. 140 mm d− 1). The constructed HSF was constructed with a length of 8.4 m and a width of 5.4 m (Fig. 1b). The average gravel porosity was equal to 0.45 and the depth of the bed was 0.45 m. The front and the effluent end of the bed were established with 60 to 100 mm diameter gravel. The CWs bed consisted of 30 mm diameter gravel with a top layer of 10 mm diameter for supporting the vegetation. Plant species in the HSF system were again P. australis and A. donax. The initial wastewater addition Table 1 Operational data on the system HLR % of designed
FWS OLR (gCOD m− 2 d− 1)
HRT (d− 1)
OLR (gCOD m− 2 d− 1)
HSF HRT (d− 1)
OLR (kgCOD m− 3 d− 1)
RBC HRT (h− 1)
OLR (kgCOD m− 3 d− 1)
PBF HRT (h− 1)
50 75 100 125 150 175
37.8 56.6 56.3 46.0 73.1 142.8
3.84 2.56 1.92 1.54 1.28 1.10
35.1 52.7 52.3 42.8 68.0 132.7
2.88 1.92 1.44 1.15 0.96 0.82
0.53 0.80 0.79 0.65 1.03 2.01
24.0 16.0 12.0 9.6 8.0 6.9
0.37 0.56 0.55 0.45 0.72 1.40
34.3 22.9 17.1 13.7 11.4 9.8
2998
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
media was 1.85 m3 and the volume of the recirculation tank was 2.5 m3. The specific PBF was designed to treat 3.5 m3 of wastewater daily. All units operated in parallel, receiving primarily (sedimentation) treated wastewater from the sewage treatment plant of Heraklion, at various HLRs ranging from 50% to 175% of designed HLR. The study was carried out from February 2007 to September 2007. Table 1 shows the organic loading rates (OLRs) and hydraulic retention time (HRT) for each system during experimental operation. 2.2. Sampling and analysis Water samples were taken as grab samples from the influent and from the effluent of each treatment unit two times every week throughout a period of 8 months, totalling 51 sampling dates. Samples were analyzed for: Chemical Oxygen Demand (COD), Total Suspended Solids (TSS), pH, Electrical Conductivity (EC), Total Nitrogen (TN), Nitrates (NΟ3–N), Ammonium (NH4–N) and Total Phosphorus (TP), according to APHA (1995). Faecal bacteria were analyzed by the membrane filtration technique (APHA, 1995), using sterile 0.2-µm pore-size Millipore filters. Three sample volumes of 0.01, 0.1 and 1 ml were used to increase the probability of obtaining counts within acceptable ranges. Filtered samples were incubated on M-FC agar for 24 h at 44.5 °C to enumerate Faecal coliforms (FC) colonies and incubated on K-FS agar for 48 h at 35 °C to assess Faecal streptococci (FS). Statistical analyses were carried out with MicroCal Origin 7.0 (OriginLab) or Statgraphics Centurion XV (Statpoint, Inc., Virginia, USA). Data were tested for homogeneity of variances using Cochran's C test. The data was analyzed using one and two-way analysis of variance (ANOVA) to compare inlet and outlet concentrations between systems and loading rates and the removal of COD, TSS, TN, NO3–N, NH4–N and TP. For inlet and outlet water quality posthoc comparisons of means were carried out using Scheffes test at the 5% significance level. To detect the statistical significance of differences (P b 0.05) between means of treatments, the Tukey test was performed. 3. Results and discussion 3.1. General
Fig. 2. Variation of COD, TSS, TN, TP, N–NH+4 and N–NO−3 in influent and effluent of different treatment systems during operation at a loading rate of 50%, 75%, 100%, 125%, 150% and 175% of designed loading rate.
was carried out in a similar way to the FWS. The HSF was designed to treat 6 m3 of domestic wastewater daily (HLR c. 130 mm d− 1). The RBC system used in this study (EKOL 4, AquaImpex) was provided by Dialynas S.A (Fig. 1c). The plant consists of a polypropylene tank partitioned into a primary sedimentation section, a storage section, a section for biological treatment with rotating biological contactor and finally a sedimentation section. The RBC consists of a cage filled with small plastic elements with a total surface area of 235 m2. The volume of the sedimentation and storing sections were 6.2 m3 and the volume of the biological contactor tank was 2.0 m3. The unit was designed to treat domestic wastewaters from hotels, schools, and small villages with a flow of 4 m3/d. The packed bed filter (PBF) unit used was the Advantex-AX20 (Orenco) (Fig. 1d). This system is a recirculating filter using a textile material as the treatment media. The overall volume of the filter
Fig. 2 presents the varying concentrations of pollutants and suspended solids in the four units for the entire operation period. In this study, 48–51 samples have been collected for the measurements of COD, TSS, TN, TP, NO3–N and NH4–N for every unit exposed to different HLRs. The composition of the influent wastewater varied throughout the study (Table 2). The characteristics of the influent are typical for primarily treated domestic wastewater (Korkusuz et al., 2007; Caselles-Osorio et al., 2007b).
Table 2 Average composition of the wastewater loaded into the experimental treatment systems throughout the study and results of one-way ANOVA (P-values) testing the mean composition between the six different loading rates (n = number of samples) Parameter
n
Mean (min–max)
P-value⁎
COD (mg L− 1) TSS (mg L− 1) Total-N (mg L− 1) Total-P (mg L− 1) NH4–N (mg L− 1) NO3–N (mg L− 1) Total coliforms (CFU/100 ml) × 105 Fecal coliforms (CFU/100 ml) × 105 Fecal streptococci (CFU/100ml) × 105
48 50 51 51 51 50 7 7 7
465 (105–1089) 129 (66–283) 70 (52–87) 15 (8–21) 11.3 (1.7–32.1) 0.85 (0–3.0) 411 (174–690) 39 (14–73) 46 (10–85)
0.0022 0.0019 0.0034 0.0049 0.0000 0.0442 0.3010 0.1602 0.2285
⁎: Figures in bold indicate statistical significant differences between the mean composition at the different loading rates at the 95% confidence level.
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
Effluent concentrations for COD, TSS, TN, TP, NO3–N and NH4+–N are presented in Table 3 and the removal percentage for the four systems and the six loading rates are presented in Fig. 3. Fig. 4 present charts of pH, conductivity and air temperature during the functioning period of the units. It can be observed that generally for all units the pH value increased about 0.4–0.8 of the influent pH value. The conductivity decreased slightly from 1.60 to 1.40 mS cm− 1 for two attached biofilm units and remained almost stable for two constructed wetlands. EC values in constructed wetlands depends on several parameters as soil or gravel composition, evaportranspiration, treatment performance etc. Normally, during a wastewater treatment plant EC values decreased however in constructed wetlands the evaportranspiration and the interactions between soil and water affect on final EC values. Air temperature for almost all cases was above 12 °C. Water temperature expecting to had a variation from the air temperature of 0–3 °C (Akratos and Tsihrintzis, 2007). 3.2. Organic matter removal The outlet concentrations of COD and TSS differed significantly between systems and depended also on loading rate, but the dependency of loading rate differed between systems as shown by the significant interaction term in the two-way ANOVA (Table 4). Effluent levels of COD and TSS were significantly higher for FWS systems compared with the other systems. COD and TSS concentration values did not significantly differ between the HSF, RBC and PBF unit for all tested loading rates (Table 3). On the other hand, FWS effluent was in most cases found to have statistical significant different concentration from the other three systems. The mean removal efficiency of HSF for COD and TSS was 76.9% and 80.7%, 74.7% and 83.4%, 61.0% and 68.4% for low (50–75% of designed), nominal and high (125–175% of designed) HLR respectively (Fig. 3). This is in accordance with previous studies which focused on
2999
the treatment of domestic wastewater with HSF showing removal efficiencies about 75% for COD and 80% for TSS (Vymazal, 2005, Babatunde et al., 2008). Furthermore, our results suggest that a possible underestimation of designed HLRs may result in a considerable decrease of COD and TSS reduction. RBC and PBF removed 86.7%– 90.4%, 95.1%–95.4% and 74.6%–72.0% of the COD for low, nominal and high HLRs respectively. High removal efficiencies observed in nominal HLRs for these treatment systems agree well with data from literature (Gebara,1999; Ayoub and Saikaly, 2004). Generally, the removal of COD decreased in all systems for HLRs higher than the loading the wetlands were designed to operate. The decreasing COD removal feasibility of treatment systems facing fluctuating HLRs is a remarkable disadvantage because in high tourist regions such as Crete Island there are significant fluctuations of influent quantity between winter (no or low tourist period) and summer (extremely high tourist period). The unit with the lowest removal efficiency for COD was the FWS (57.0%, 32.3% and 20.1%). These COD removal rates were lower than mean ranges reported in literature. For instance, a combined Typha domingensis and P. australis wetland treating urban wastewater in Spain removed 87% of the incoming COD (Gomez Cerezo et al., 2001). Similar results were observed by other researchers in the USA (Verhoeven and Meuleman, 1999) and in Europe (Brix, 1994b; Tsihrintzis et al., 2007). However, there are few other researchers who found COD removal efficiencies relatively close to our observations. A wetland system in a rural area in Spain for sewage treatment with Typha latifolia and Salix atrocinerea removed only 60% of the incoming COD (Ansola et al., 2003) and Rousseau et al. (2004) reported average COD removal efficiencies of 61% for FWS treatment systems in Flanders region. The hydraulic loading rates applied in this study in order to reduce land requirements were abnormally high as compared with most works. The high organic strength of influent resulting in a high organic loading rate was the major reason for poor treatment feasibility of FWS. Hiley (1995) report that most wetlands are oxygen limited and that performance is
Table 3 Mean values of quality parameters for the effluent from each treatment system for different HLRs System FWS HLRd
HSF HLRd
RBC HLRd
PBF HLRd
COD (mg/L)
TSS (mg/L)
TN (mg/L)
NO−3 (mg/L)
NH+4 (mg/L)
TP (mg/L)
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
50% 75% 100% 125% 150% 175%
227.9a 228.3a 230.0a 317.9a 187.9a 294.3a
97.3 114.2 76.5 83.8 101.9 133.2
58.9a 62.0a 75.9a 126.7a 88.8a 70.1a
17.7 35.6 24.4 72.3 40.9 32.5
58.5a 55.8a 65.7a 87.0a 63.6a 65.8a
9.2 16.6 8.4 18.4 3.4 4.9
11.3a 11.3a 11.4a 9.7a 12.4a 11.2a
1.0 1.2 1.0 0.9 0.9 1.5
17.5a 10.4a 6.2a 6.9a 8.4a 15.2a
4.9 4.2 3.0 2.6 2.8 3.8
0.9a 0.7a 2.1a 1.4a 1.0a 0.7a
0.9 0.5 1.9 1.1 0.5 0.4
50% 75% 100% 125% 150% 175%
130.5b 99.6b 85.7b 34.6b 126.3a 341.1a
103.3 49.4 58.9 24.8 65.2 189.1
26.0b 20.7b 29.4b 25.7b 30.8b 39.4a
6.7 5.1 10.2 19.9 12.6 10.8
68.9a 41.2a 49.4b 47.0b,c 51.2a,b 67.4a
25.0 12.3 7.3 7.8 8.8 7.2
8.6b 9.2b 7.8b 7.9b 9.1b 10.0a,b
0.7 0.6 0.6 0.3 0.8 0.8
17.4a 11.4a 7.8b 7.2a 7.8a 14.6a
8.4 1.6 3.6 4.4 1.7 4.5
6.2a 3.1a 2.7a 1.2a 1.3a 0.7a
7.1 2.5 2.2 0.4 0.3 0.6
50% 75% 100% 125% 150% 175%
95.5b 43.4b 16.6b 38.4b 94.0a 178.9a
62.8 18.4 9.7 19.6 63.1 101.8
14.6c 11.9b 14.9b 14.2b 26.0b 61.8a
5.4 4.5 6.3 7.5 8.1 23.1
59.2a 38.4a 25.6c 61.5b 42.2b 41.4b
9.7 10.4 5.1 11.6 11.4 4.5
8.6b,c 7.9c 8.6c 8.0b 6.8b 8.7b,c
1.0 0.9 1.7 1.3 2.4 0.8
2.9b 1.8b 5.5b 8.6a 10.7a 5.2b
1.3 1.1 3.1 3.9 1.0 3.9
25.1b 30.5b 28.1b 2.9a 1.8a 22.0b
23.0 8.0 20.9 2.5 0.7 8.8
50% 75% 100% 125% 150% 175%
51.9b 50.3b 18.3b 32.0b 111.2a 216.0a
43.0 32.2 18.3 16.0 61.0 123.1
10.2c 11.1b 19.3b 19.2b 33.8b 57.5a
8.6 7.0 8.1 3.5 9.8 24.9
55.0a 47.5a 28.0c 35.5c 38.0b 39.0b
10.4 13.5 5.1 15.3 13.3 17.1
7.5c 8.1c 8.6c 7.8b 8.6b 8.1c
0.9 1.1 0.7 0.8 1.1 0.9
1.1b 1.5b 4.4b 5.2a 7.1a 12.5a,b
0.5 1.1 2.0 2.6 1.4 1.0
24.8b 34.6b 27.8b 3.5a 2.6a 5.2a
25.8 10.8 13.3 2.7 1.6 0.5
a, b, c: In each column for each loading rate, mean values followed by a different symbol are significantly different (P b 0.05).
3000
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
Fig. 4. Variation of pH, electrical conductivity and air temperature during operation period. Fig. 3. Removal percentage for each HLR and system.
enhanced if extra aeration is provided. The oxygen release from roots of macrophytes seems to be far less than the amount needed for aerobic degradation of the oxygen consuming substances delivered with sewage (Brix, 1990; Brix and Schierup, 1990). As shown in Fig. 5, there is a correlation between the loading rate and the effluent values of the analyzed parameters. It has been observed that there is a positive linear relationship between increasing load and effluent COD or TSS concentration. In the case of RBC and PBF we observed significant adverse effects to organic matter concentration for increasing load. On the other hand the two constructed wetlands turned out to be less sensitive to variations in HLRs.
suggest that the denitrification process was the rate limiting step for nitrogen biodegradation. On the other hand during high HLRs nitrification occurred only partially as the NO3–N concentrations were the lowest. The inhibition of the nitrification process under high organic loading rates in attached biofilm reactors were in general consistent to several other published reports (Gupta and Gupta, 2001; Sirianuntapiboon, 2006). Relatively poor nitrogen removal performance of constructed wetlands treating domestic wastewater was often reported in literature (Tanner et al., 1999; Rousseau et al., 2004; Brix and Arias, 2005). In this work, TN average removal of both FWS and HSF with each examined loading rate was not above 12.7% and 27.0%, respectively. NH4+–N and NO3–N measured concentrations showed that both ammonification and nitrification processes were low. In fact, the magnitude processes which ultimately remove the total nitrogen
3.3. Nutrients removal The effluent quality with respect to nitrogen was better for the two attached biofilm technologies, while nitrogen values were moderate for HSF and still high for FWS (Table 3.). Nitrate concentrations were significantly higher for the RBC and PBF than for the two constructed wetlands. Even though ammonium mean concentration values for RBC and PBF were lower than for FWS and HSF, statistically there were no significant differences. The average TN removal efficiencies of two attached biofilm systems under nominal HLRs were 62.3% and 58.1%. These results were in accordance with previous reported nitrogen removal values of about 45–65% (Tawfik et al., 2006; Griffin et al., 1999). An interesting point to note was that nitrogen removal of these systems was decreased both for low and for high HLRs operation. According to Table 3, for low (50–75%) HLRs these systems had the lowest NH4–N concentrations; the higher NO3–N concentrations
Table 4 Results of two-way ANOVA (F-ratios) used to test for differences in outlet water quality between the four systems tested (main factor ‘System’) and the six loading rates (main factor ‘Loading’) and their interaction (System × Loading) Parameter
System
Loading
System × Loading
COD TSS Total-N Total-P NH4–N NO3–N Total coliforms Fecal coliforms Fecal streptococci
40.11⁎⁎⁎ 76.87⁎⁎⁎ 37.52⁎⁎⁎ 92.75⁎⁎⁎ 34.76⁎⁎⁎ 30.63⁎⁎⁎ 7.20⁎⁎ 0.77NS 4.55⁎
17.42⁎⁎⁎ 12.03⁎⁎⁎ 12.00⁎⁎⁎ 3.80⁎⁎ 14.65⁎⁎⁎ 11.29⁎⁎⁎ 0.67NS 0.59NS 1.07NS
2.36⁎⁎ 4.16⁎⁎⁎ 5.84⁎⁎⁎ 3.36⁎⁎⁎ 10.89⁎⁎⁎ 3.77⁎⁎⁎ 1.40NS 1.06NS 0.38NS
⁎: P b 0.05; ⁎⁎: P b 0.01; ⁎⁎⁎: P b 0.001; NS: Not Significant.
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
3001
Table 6 Removals rate and costs of each wastewater treatment system Treatment system
FWS HSF RBC PBF
Requirements
Removal rate (gr inhab.− 1 d− 1) COD
TSS
N
P
Land (m2 inhab.− 1)
Power (kWh inhab.− 1 year− 1)
33.4 62.0 75.8 75.6
21.0 30.4 33.2 32.4
0.4 3.8 8.4 8.0
0.7 1.4 1.3 1.3
1.4 1.4 0.3 0.2
~0 ~0 219 43
Construction costs (€ inhab.− 1) 237 268 476 672
from the HLR indicating that the main mechanism for TP removal is the adsorption to the porous media (Knight et al., 2000; Jing et al., 2001, Vymazal, 2002; Akratos and Tsihrintzis, 2007). 3.4. Pathogen removal According to Table 5, average concentrations of pathogenic bacteria in the influent in the range of typical domestic wastewater (Karrathanasis et al., 2003; Decamp and Warren, 2000). Furthermore the average concentration of fungi in the influent was 36 ± 18 × 105 CFU/ 100 ml. Faecal indicators were efficiently removed in all examined systems. The average removal efficiencies for TC were between 1.3 and 2.2 log units. HSF showed a better efficiency to remove pathogens than the other systems with average effluent concentrations of 0.3 × 105 and 0.4 × 105 CFU/100 ml for FC and FS, respectively. These values show that only HSF met the level recommended by the EPA of 1.0 × 105 CFU/ 100 ml. These efficiencies appear to be in the range described in previous studies both for constructed wetlands (Karrathanasis et al., 2003; Ansola et al., 2003; Keffala and Ghrabi, 2005; Ottova et al., 1997; Sleytr et al., 2007) and attached biofilm reactors (Sagy and Kott, 1990; Tawfik et al., 2002; Hua et al., 2003). 3.5. Criteria for optimum treatment system Fig. 5. Correlation charts of effluent concentration and pollutant load for COD, TSS, TN and TP.
from these systems is usually low and therefore removal of TN is commonly low in single-stage constructed wetlands (Vymazal, 2007). The concentration of TP was reduced in HSF, RBC and PBF in relatively satisfactory efficiencies ranging from 40 to 50%. This result is in accordance with previous works observing removal efficiencies of about 25–65% (Chen et al., 2006; Tsihrintzis et al., 2007; Vymazal, 2005; Kivaisi, 2001; Yun et al., 2004). FWS was found to have the lower capacity to remove phosphorus. The low removal rate observed in FWS may be due to the fact that the most important processes involved, occur in the sediment and not in the water column (Verhoeven and Meuleman, 1999; Kadlec, 2006). All examined systems show a potential to decrease phosporus removal as loading rate increase. However this potential seems to be rather independent
A comparison of several critical parameters among the examined non-conventional wastewater treatment systems is presented in Table 6. Results based on per capita values suggested that the average wastewater production is 0.2 m3 per capita. Removal rates were calculated from experimental data during the period which the systems were operated at 100% of designed HLR. Remarked that land requirements and construction costs evaluated based only on this specific experimental process. The value of construction cost for FWS and HSF was 237 and 268 € Inhab− 1 respectively while Rousseau et al. (2004) found significant higher average construction cost of constructed wetlands treating wastewater in Flanders region (392 and 1258 € Inhab− 1 for FWS and HSF, respectively). The overloaded designed constructed wetlands in this work resulting lower construction cost as well as lower land requirement. Furthermore, Rousseau et al. (2004) found a footprint for FWS and HSF of 7.0 and 4.8 m2 Inhab− 1 respectively whereas in this work
Table 5 Influent and effluent concentrations, and removal efficiencies for microbiological parameters for each treatment system Parameter
Total coliforms Mean S.D. Faecal coliforms Mean S.D. Faecal streptococci Mean S.D.
Influent
FWS
(CFU/100 ml)
(CFU/100 ml)
Removal
HSF (CFU/100 ml)
Removal
(CFU/100 ml)
RBC Removal
(CFU/100 ml)
PBF Removal
x 105
x105
log units
x 105
log units
x105
log units
x 105
log units
411.5 217.4
20.4 14.0
1.3 1.2
2.6 2.7
2.2 1.9
2.9 4.2
2.2 1.7
16.8 11.7
1.4 1.3
56.3 38.3
5.8 3.9
1.0 1.0
0.3 0.3
2.3 2.1
4.8 11.1
1.1 0.5
5.0 6.7
1.1 0.8
47.9 30.7
6.0 2.6
0.9 1.1
0.4 0.4
2.1 1.9
1.5 2.3
1.5 1.1
3.4 3.1
1.1 1.0
3002
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
the land requirement was notably lower. Although FWS found to have the lowest construction cost and generally very low requirements, the effluent quality characteristics are significantly poorer than from the other three systems. The effort to operate this system in higher HLRs in order to reduce land requirements had significant negative effect on effluent characteristics. On the other hand RBC and PBF, which require the smallest land area (0.2–0.3 m2 inhab− 1), showed approximately the same effluent values both for organic matter and nutrients concentrations. The quite low power requirements for PBF as compared to RBC balance the relatively higher construction costs. The construction cost for HSF is significant lower than those for the two fixed film media units and additionally found to have almost the same removal rates for COD, TSS and TP. Consequently, the HSF wetland designed and operated under high HLRs showed both effective wastewater treatment and low land requirements. However nitrogen removal efficiency is by far lower. Thus, the choice of the optimal treatment system depends on the overall requirements in terms of treatment efficiency. 4. Conclusions Four non-conventional wastewater treatment systems operated continuously with various HLRs for about 8 months. During this period, with the exception of FWS all other units used in this study removed efficiently the organic matter of domestic wastewater. The RBC and PBF performance showed lower performance efficiency under shock loading conditions. Nitrogen reduction was higher in RBC and PBF than in the two constructed wetlands while phosphorous removal was not significantly different for all examined systems. Additionally, all studied non-conventional units showed considerable potential for removing fecal bacteria from domestic wastewater. Overall, treatment efficiencies of the two attached biofilm technologies are comparable while power requirements are considerable higher for RBC. HSF wetland designed to operate under high HLRs provide comparable effluent quality as attached biofilm technologies whereas the construction and operation costs are lower. However whenever the requirements of nitrogen level in the effluent are crucial, HSF would not successfully meet these criteria. Since overloaded designed FWS wetlands showed the lowest treatment performance, the selection of this technology could only be justified when land is available (lower HLRs) and the effluent quality requirements are not very strict. Therefore, the selection of the optimum unit depends on the demand for treatment efficiency and the total available budget. Acknowledgements This research was funded by the Greek General Secretariat for Research and Technology, and well as Prisma Domi S.A. The authors wish to express their gratitude to Professor Vassilis Manios, Vaggelis Theodorakopoulos and Ioannis Sabathianakis for their continued involvement in the realisation of this project. References Aesoy A, Odegaard H, Bach K, Pujol R, Hamon M. Dentrification in a packed bed biofilm reactor (biofor) — experiments with different carbon sources. Water Res 1998;32:1463–70. Akratos CS, Tsihrintzis VA. Effect of temperature, HRT, vegetation and porous media on removal efficiency of pilot-scale horizontal subsurface flow constructed wetlands. Ecol Eng 2007;29:173–91. Ansola G, Gonzalez JM, Cortijo R, de Luis E. Experimental and full-scale pilot plant constructed wetlands for municipal wastewaters treatment. Ecol Eng 2003;21:43–52. APHA. Standard Methods for the Examination of Water and Wastewater. 19th ed. Washington DC, USA: American Public Health Association; 1995. Ayoub GM, Saikaly P. The combined effect of step-feed and recycling on RBC performance. Water Res 2004;38:3009–16. Boller M. Small wastewater treatment plants-a challenge to wastewater engineers. Water Sci Technol 1997;35:1-12.
Babatunde AO, Zhao YQ, O'Neill M, O'Sullivan B. Constructed wetlands for environmental pollution control: a review of developments, research and practice in Ireland. Environ Int 2008;34:116–26. Brix H. Gas exchange through the soil–atmosphere interphase and through dead culms of Phragmites australis in a constructed reed bed receiving domestic sewage. Water Res 1990;24:259–66. Brix H. Use of constructed wetlands in water pollution control: historical development, present status, and future perspectives. Water Sci Technol 1994a;30:209–23. Brix H. Constructed wetlands for municipal wastewater treatment in Europe. Global wetlands; 1994b. p. 325–33. Brix H, Arias CA. The use of vertical flow constructed wetlands for on-site treatment of domestic wastewater: new Danish guidelines. Ecol Eng 2005;25:491–500. Brix H, Schierup H-H. Soil oxygenation in constructed reed beds: the role of macrophyte and soil–atmosphere interface oxygen transport. In: Cooper PF, Findlater BC, editors. Constructed wetlands in water pollution control. Oxford: Pergamon; 1990. p. 53–66. Caselles-Osorio A, Porta A, Porras M, Garcia J. Effect of high organic loading rates of particulate and dissolved organic matter on the efficiency of shallow experimental horizontal subsurface-flow constructed wetland. Water Air Soil Pollut 2007a;183:367–75. Caselles-Osorio A, Puigagut J, Segu E, Vaello N, Granes F, Garcia D, et al. Solids accumulation in six full-scale subsurface flow constructed wetlands. Water Res 2007b;41:1388–98. Chen TY, Kao CM, Yeh TY, Chien HY, Chao AC. Application of a constructed wetland for industrial wastewater treatment: a pilot-scale study. Chemosphere 2006;64:497–502. Chung AKC, Wu Y, Tam NFY, Wong MH. Nitrogen and phosphate mass balance in a sub-surface flow constructed wetland for treating municipal wastewater. Ecol Eng 2008;32:81–9. Colmenarejo MF, Rubio A, Sanchez E, Vicente J, Garcia MG, Borja R. Evaluation of municipal wastewater treatment plants with different technologies at Las Rozas, Madrid (Spain). J Environ Manag 2006;81:399–404. Decamp O, Warren A. Investigation of Escherichia coli removal in various designs of subsurface flow wetlands used for wastewater treatment. Ecol Eng 2000;14:293–9. European Parliament and Council. Water framework directive (Directive 2000/60/EC of 23 October 2000). Eur J 2000;L327:1-72. Fahd K, Martin I, Salas JJ. The Carrion de los Cespedes Experimental Plant and the technological Transfer Centre: urban wastewater treatment experimental platforms for the small rural communities in the Mediterranean area. Desalination 2007;215:12–21. Gebara F. Activated sludge biofilm wastewater treatment system. Water Res 1999;33:230–8. Gomez Cerezo R, Suarez ML, Vidal-Abarca MR. The performance of a multistage system of a constructed wetlands for urban wastewater treatment in a semiarid region of SE Spain. Ecol Eng 2001;16:501–17. Griffin P, Jennings P, Bowman E. Advanced nitrogen removal by rotating biological contactors recycle and constructed wetlands. Water Sci Technol 1999;40:383–90. Gupta AB, Gupta SK. Simultaneous carbon and nitrogen removal from high strength domestic wastewater in an aerobic RBC biofilm. Water Res 2001;35:1714–22. Hansford GS, Andrews JF, Grieves CG, Carr AD. A steady-state model for the rotating biological disc reactor. Water Res 1978;12:855–68. He Q, Mankin KR. Performance variations of COD and nitrogen removal by vegetated submerged bed wetlands. J Am Water Resour Assoc 2002;38:1679–89. Helmer C, Kunst S. Simultaneous nitrification/denitrification in aerobic bio-film system. Water Sci Technol 1998;37:183–7. Hiley PD. The reality of sewage treatment using wetlands. Water Sci Technol 1995;32:329–38. Hua J, An P, Winter J, Gallert C. Elimination of COD, microorganisms and pharmaceuticals from sewage by trickling through sandy soil below leaking sewers. Water Res 2003;37:4395–404. Jing SR, Lin YF, Lee DY, Wang TW. Nutrient removal from polluted river water by using constructed wetlands. Biores Technol 2001;76:131–5. Kadlec RH. Free surface wetlands for phosphorus removal: the position of the Everglades Nutrient Removal Project. Ecol Eng 2006;27:361–79. Karrathanasis AD, Potter CL, Coyne MS. Vegetation effects on faecal bacteria, BOD, and suspended solid removal in constructed wetlands treating domestic wastewater. Ecol Eng 2003;20:157–69. Keffala C, Ghrabi A. Nitrogen and bacterial removal in constructed wetlands treating domestic wastewater. Desalination 2005;185:383–9. Kivaisi AK. The potential for constructed wetlands for wastewater treatment and reuse in developing countries: a review. Ecol Eng 2001;16:545–60. Knight RL, Payne VWE, Borer RE, Clarke RA, Pries JH. Constructed wetlands for livestock wastewater management. Ecol Eng 2000;15:41–55. Korkusuz EA, Beklioglu M, Demirer GN. Use of blast furnace granulated slag as a substrate in vertical flow reed beds: field application. Biores Technol 2007;98:2089–101. Kuschk P, Wiebner A, Kappelmeyer U, Weibrodt E, Kastner M, Stottmeister U. Annual cycle of nitrogen removal by a pilot-scale subsurface horizontal flow in a constructed wetland under moderate climate. Water Res 2003;37:4236–42. Lin YF, Jing SR, Lee DY, Chang YF, Chen YM, Shih KC. Performance of a constructed wetland treating intensive shrimp aquaculture wastewater under high hydraulic loading rate. Environ Pollut 2005;134:411–21. Mann AT, Stephenson T. Modelling biological aerated filters for wastewater treatment. Water Res 1997;31:2443–8. Molle P, Lienard A, Grasmick A, Iwema A. Effect of reeds and feeding operations on hydraulic behaviour of vertical flow constructed wetlands under hydraulic overloads. Water Res 2006;40:606–12. Nitisoravut S, Klomjek P. Inhibition kinetics of salt-affected wetland for municipal wastewater treatment. Water Res 2005;39:4413–9. Ottova V, Balcarova J, Vymazal J. Microbial characteristics of constructed wetlands. Water Sci Technol 1997;35:117–23.
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003 Paing J, Voisin J. Vertical flow constructed wetlands for municipal wastewater and septage treatment in French rural area. Water Sci Technol 2005;51:145–55. Rousseau DPL, Vanrolleghem PA, De Pauw N. Model-based design of horizontal subsurface flow constructed treatment wetlands: a review. Water Res 2004;38:1484–93. Sagy M, Kott Y. Efficiency of rotating biological contactor in removing pathogenic bacteria from domestic wastewater. Water Res 1990;24:1125–8. Schubert W, Wolfgang Günthert F. Particle size distribution in effluent of trickling filters and in humus tanks. Water Res 2001;35:3993–7. Sirianuntapiboon S. Treatment of wastewater containing Cl2 residue by packed cage rotating biological contactor (RBC) system. Bioresour Technol 2006;97:1735–44. Sleytr K, Tielz A, Langergraber G, Haberl R. Investigation of bacterial removal during the filtration process in constructed wetlands. Sci Total Environ 2007;380:173–80. Tanner CC, D'Eugenio J, McBride GB, Sukias JPS, Thompson K. Effect of water level fluctuation on nitrogen removal from constructed wetland mesocosms. Ecol Eng 1999;12:67–92. Tawfik A, Klapwijk B, El-Gohary F, Lettinga G. Treatment of anaerobically pre-treated domestic sewage by a rotating biological contactor. Water Res 2002;36:147–55. Tawfik A, Temmink H, Zeeman G, Klapwijk B. Sewage treatment in rotating biological contactor (RBC) system. Water Air Soil Pollut 2006;175:275–89.
3003
Tsihrintzis VA, Akratos CS, Gikas GD, Karamouzis D, Angelakis AN. Performance and cost comparison of a FWS and a VSF constructed wetland system. Environ Technol 2007;28:621–8. USEPA. Design manual constructed wetlands and aquatic plant system for municipal wastewater treatment. Office of Research and Development, Cincinnati, Ohio, USA; 1988; 1988. USEPA. Constructed wetlands treatment of municipal wastewaters. Office of Research and Development, Cincinnati, Ohio, USA; 2000; 2000. Verhoeven JTA, Meuleman AFM. Wetlands for wastewater treatment: opportunities and limitations. Ecol Eng 1999;12:5-12. Vymazal J. The use of sub-surface constructed wetlands for wastewater treatment in the Czech Republic: 10 years experience. Ecol Eng 2002;18:633–46. Vymazal J. Horizontal sub-surface flow and hybrid constructed wetlands systems for wastewater treatment. Ecol Eng 2005;25:478–90. Vymazal J. Removal of nutrients in various types of constructed wetlands. Sci Total Environ 2007;380:48–65. Yun Z, Lee H, Choi E. Enhanced biological phosphorus removal in RBC with SBR modification. Water Sci Technol 2004;50:121–30.
Contents lists available at ScienceDirect
Science of the Total Environment j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i t o t e n v
Pilot-scale comparison of constructed wetlands operated under high hydraulic loading rates and attached biofilm reactors for domestic wastewater treatment M.S. Fountoulakis a,⁎, S. Terzakis a,b, A. Chatzinotas c, H. Brix d, N. Kalogerakis b, T. Manios a,e a
School of Agricultural Technology, Technological Educational Institute of Crete, Heraklion, Greece Department of Environmental Engineering, Technical University of Crete, Chania, Greece UFZ, Helmholtz Centre for Environmental Research – UFZ, Department of Environmental Microbiology, Permoserstrasse 15, D-04318 Leipzig, Germany d Department of Biological Sciences, Aarhus University, Denmark e Greek Open University, School of Science and Technology, Patras, Greece b c
a r t i c l e
i n f o
Article history: Received 7 July 2008 Received in revised form 23 December 2008 Accepted 5 January 2009 Available online 30 January 2009 Keywords: Constructed wetland Treatment performance Nutrients Rotating biological contactor Packed bed filter
a b s t r a c t Four different pilot-scale treatment units were constructed to compare the feasibility of treating domestic wastewater in the City of Heraklio, Crete, Greece: (a) a free water surface (FWS) wetland system, (b) a horizontal subsurface flow (HSF) wetland system, (c) a rotating biological contactor (RBC), and (d) a packed bed filter (PBF). All units operated in parallel at various hydraulic loading rates (HLR) ranging from 50% to 175% of designed operating HLR. The study was conducted during an 8 month period and showed that COD removal efficiency of HSF was comparable (N 75%) to that of RBC and PBF, whereas that of the FWS system was only 57%. Average nutrient removal efficiencies for FWS, HSF, RBC and PBF were 6%, 21%, 40% and 43%, respectively for total nitrogen and 21%, 39%, 41% and 42%, respectively for total phosphorus. Removals of total coliforms were lowest in FWS and PBF (1.3 log units) and higher in HSF and RBC (2.3 to 2.6 log units). HSF showed slightly lower but comparable effluent quality to that of RBC and PBF systems, but the construction cost and energy requirements for this system are significantly lower. Overall the final decision for the best non-conventional wastewater treatment system depends on the construction and operation cost, the area demand and the required quality of effluent. © 2009 Published by Elsevier B.V.
1. Introduction There is an urgent need to implement efficient treatment of domestic wastewater from small towns as imposed by the Water Framework Directive 2000/60/EC (EU Parliament and Council, 2000). However, the per capita cost for implementation of the common activated sludge process in small communities is much higher than that for large cities. Therefore non-conventional technologies ranging from simple biological low rate systems such as ponds, constructed wetlands and sand filters to complex high-rate suspended and fixed biomass reactors have to be evaluated according to their treatment performance, foot print, process reliability, investment and operation costs (Boller, 1997; Colmenarejo et al., 2006; Fahd et al., 2007). Constructed wetlands (CW) are used worldwide to treat municipal wastewater (Brix, 1994a; He and Mankin, 2002; Nitisoravut and Klomjek, 2005; Paing and Voisin, 2005; Brix and Arias, 2005; Chung et al., 2008). CW systems are based on the functioning of natural ecosystems and the treatment processes involve complex interactions between soil, water, plants and microorganisms. CWs are generally efficient in removal of organic matter (BOD) and suspended solids (SS), but the removal of nitrogen and phosphorus is often relatively poor ⁎ Corresponding author. Tel.: +30 2810 379456; fax: +30 2810 318204. E-mail address: [email protected] (M.S. Fountoulakis). 0048-9697/$ – see front matter © 2009 Published by Elsevier B.V. doi:10.1016/j.scitotenv.2009.01.005
(Verhoeven and Meuleman, 1999; Tanner et al., 1999; Kuschk et al., 2003; Vymazal, 2007). Constructed wetlands can be classified into two main categories depending on how the water passes through the systems: subsurface-flow or surface-flow design. The most widely used CW design in Europe is designed with horizontal subsurface flow (HSF). However, free water surface (FWS) constructed wetlands are increasingly being favoured because they are cheaper to construct and may have higher wildlife habitat values. There is a concern about the feasibility of wetlands to become a cost effective method because wetlands typically require a low hydraulic loading rate (HLR) and a long hydraulic retention time (HRT) to achieve efficient pollutant removal. That means wetland treatment method may need a large land area. USEPA have recommended that the organic loading rate should not exceed 6 g BOD m− 2 d− 1 in HSF (USEPA, 2000) and 11.2 g BOD m− 2 d− 1 in FWS (USEPA, 1988). These suggestions may not be applicable when land is expensive or limited. Using high HLR to operate the constructed wetland may potentially reduce the required area. Recent years there is an effort to study the performance of constructed wetlands under high HLRs. Caselles-Osorio et al. (2007a) found that there was no significant difference in COD removal between two HSF systems operated at an HLR of 6 g COD m− 2 d− 1 and 23 g COD m− 2 d− 1. Another study in France concluded that overloads up to ten times the dry weather flow in vertical flow constructed wetlands are possible while still complying with the European standards (Molle
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
2997
Fig. 1. View of the pilot scale units, a) free water surface constructed wetland, b) subsurface constructed wetland, c) rotating biological contactor, d) packed bed reactor.
et al., 2006). Furthermore, Lin et al. (2005) shown that a unit consisting of FWS and HSF wetland cells treating intensive shrimp aquaculture wastewater effectively removed TSS, BOD and NH4–N under high hydraulic loading rates (157–195 m d− 1). On the other hand, there are also several compact systems using fixed and movable bed reactors with high removal efficiencies for BOD, SS and nitrogen that potentially can be used for small towns. These systems in general have good process stability, less footprint requirements compared to CWs and high specific removal rates (Helmer and Kunst, 1998; Gebara, 1999; Sirianuntapiboon, 2006). Different types of attached biomass technologies have been widely applied as small-scale plants, including packed bed biofilm reactors (Mann and Stephenson, 1997; Aesoy et al., 1998; Schubert and Wolfgang Günthert, 2001) and rotating biological contactors (Hansford et al., 1978; Ayoub and Saikaly, 2004; Tawfik et al., 2006). These systems generally achieve high removal efficiencies when treating wastewater from 25–500 person equivalents (p.e.), however the capital and operations costs for small towns (1000–5000 p.e.) may be very high. The aim of this work was to assess and compare the performance of different non-conventional wastewater treatment systems in controlled experiments under Mediterranean climatic conditions. Four pilot-scale treatment systems were constructed: a) FWS, b) HSF, c) RBC, and d) PBF. All units operated in parallel receiving primarily treated municipal wastewater. In order to decrease land requirements for constructed wetlands so that are more competitive to compact attached biomass systems the FWS and HSF were designed to operate at high hydraulic loading rates. 2. Materials and methods 2.1. Pilot-scale unit description All pilot systems were constructed during summer 2006 in our open-air laboratory with a total surface area of approximately 360 m2.
The facility is located in Heraklion, Crete, South Greece (N 35°, 19q; E 25°, 10q). The FWS system was constructed with dimensions of 12.4 m long and 3.4 m wide, and with three separated zones, two vegetated zones and one deeper anoxic un-vegetated zone in the centre (Fig. 1a). A soil layer of 40 cm depth was added in the vegetated zones, and was planted with two species of reeds, Phragmites australis and Arundo donax. Plants were transplanted until a total cover of 40% was reached. Then the wetland was filled with tap water to a depth of 50 cm. Wastewater was mixed with tap water at gradually increasing wastewater/tap water ratios until only wastewater was added after 3 weeks. The incoming wastewater entered the wetland through a 40 cm gravel layer to distribute the water across the width of the bed. The unit was designed to treat 6 m3 of domestic wastewater daily (HLR c. 140 mm d− 1). The constructed HSF was constructed with a length of 8.4 m and a width of 5.4 m (Fig. 1b). The average gravel porosity was equal to 0.45 and the depth of the bed was 0.45 m. The front and the effluent end of the bed were established with 60 to 100 mm diameter gravel. The CWs bed consisted of 30 mm diameter gravel with a top layer of 10 mm diameter for supporting the vegetation. Plant species in the HSF system were again P. australis and A. donax. The initial wastewater addition Table 1 Operational data on the system HLR % of designed
FWS OLR (gCOD m− 2 d− 1)
HRT (d− 1)
OLR (gCOD m− 2 d− 1)
HSF HRT (d− 1)
OLR (kgCOD m− 3 d− 1)
RBC HRT (h− 1)
OLR (kgCOD m− 3 d− 1)
PBF HRT (h− 1)
50 75 100 125 150 175
37.8 56.6 56.3 46.0 73.1 142.8
3.84 2.56 1.92 1.54 1.28 1.10
35.1 52.7 52.3 42.8 68.0 132.7
2.88 1.92 1.44 1.15 0.96 0.82
0.53 0.80 0.79 0.65 1.03 2.01
24.0 16.0 12.0 9.6 8.0 6.9
0.37 0.56 0.55 0.45 0.72 1.40
34.3 22.9 17.1 13.7 11.4 9.8
2998
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
media was 1.85 m3 and the volume of the recirculation tank was 2.5 m3. The specific PBF was designed to treat 3.5 m3 of wastewater daily. All units operated in parallel, receiving primarily (sedimentation) treated wastewater from the sewage treatment plant of Heraklion, at various HLRs ranging from 50% to 175% of designed HLR. The study was carried out from February 2007 to September 2007. Table 1 shows the organic loading rates (OLRs) and hydraulic retention time (HRT) for each system during experimental operation. 2.2. Sampling and analysis Water samples were taken as grab samples from the influent and from the effluent of each treatment unit two times every week throughout a period of 8 months, totalling 51 sampling dates. Samples were analyzed for: Chemical Oxygen Demand (COD), Total Suspended Solids (TSS), pH, Electrical Conductivity (EC), Total Nitrogen (TN), Nitrates (NΟ3–N), Ammonium (NH4–N) and Total Phosphorus (TP), according to APHA (1995). Faecal bacteria were analyzed by the membrane filtration technique (APHA, 1995), using sterile 0.2-µm pore-size Millipore filters. Three sample volumes of 0.01, 0.1 and 1 ml were used to increase the probability of obtaining counts within acceptable ranges. Filtered samples were incubated on M-FC agar for 24 h at 44.5 °C to enumerate Faecal coliforms (FC) colonies and incubated on K-FS agar for 48 h at 35 °C to assess Faecal streptococci (FS). Statistical analyses were carried out with MicroCal Origin 7.0 (OriginLab) or Statgraphics Centurion XV (Statpoint, Inc., Virginia, USA). Data were tested for homogeneity of variances using Cochran's C test. The data was analyzed using one and two-way analysis of variance (ANOVA) to compare inlet and outlet concentrations between systems and loading rates and the removal of COD, TSS, TN, NO3–N, NH4–N and TP. For inlet and outlet water quality posthoc comparisons of means were carried out using Scheffes test at the 5% significance level. To detect the statistical significance of differences (P b 0.05) between means of treatments, the Tukey test was performed. 3. Results and discussion 3.1. General
Fig. 2. Variation of COD, TSS, TN, TP, N–NH+4 and N–NO−3 in influent and effluent of different treatment systems during operation at a loading rate of 50%, 75%, 100%, 125%, 150% and 175% of designed loading rate.
was carried out in a similar way to the FWS. The HSF was designed to treat 6 m3 of domestic wastewater daily (HLR c. 130 mm d− 1). The RBC system used in this study (EKOL 4, AquaImpex) was provided by Dialynas S.A (Fig. 1c). The plant consists of a polypropylene tank partitioned into a primary sedimentation section, a storage section, a section for biological treatment with rotating biological contactor and finally a sedimentation section. The RBC consists of a cage filled with small plastic elements with a total surface area of 235 m2. The volume of the sedimentation and storing sections were 6.2 m3 and the volume of the biological contactor tank was 2.0 m3. The unit was designed to treat domestic wastewaters from hotels, schools, and small villages with a flow of 4 m3/d. The packed bed filter (PBF) unit used was the Advantex-AX20 (Orenco) (Fig. 1d). This system is a recirculating filter using a textile material as the treatment media. The overall volume of the filter
Fig. 2 presents the varying concentrations of pollutants and suspended solids in the four units for the entire operation period. In this study, 48–51 samples have been collected for the measurements of COD, TSS, TN, TP, NO3–N and NH4–N for every unit exposed to different HLRs. The composition of the influent wastewater varied throughout the study (Table 2). The characteristics of the influent are typical for primarily treated domestic wastewater (Korkusuz et al., 2007; Caselles-Osorio et al., 2007b).
Table 2 Average composition of the wastewater loaded into the experimental treatment systems throughout the study and results of one-way ANOVA (P-values) testing the mean composition between the six different loading rates (n = number of samples) Parameter
n
Mean (min–max)
P-value⁎
COD (mg L− 1) TSS (mg L− 1) Total-N (mg L− 1) Total-P (mg L− 1) NH4–N (mg L− 1) NO3–N (mg L− 1) Total coliforms (CFU/100 ml) × 105 Fecal coliforms (CFU/100 ml) × 105 Fecal streptococci (CFU/100ml) × 105
48 50 51 51 51 50 7 7 7
465 (105–1089) 129 (66–283) 70 (52–87) 15 (8–21) 11.3 (1.7–32.1) 0.85 (0–3.0) 411 (174–690) 39 (14–73) 46 (10–85)
0.0022 0.0019 0.0034 0.0049 0.0000 0.0442 0.3010 0.1602 0.2285
⁎: Figures in bold indicate statistical significant differences between the mean composition at the different loading rates at the 95% confidence level.
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
Effluent concentrations for COD, TSS, TN, TP, NO3–N and NH4+–N are presented in Table 3 and the removal percentage for the four systems and the six loading rates are presented in Fig. 3. Fig. 4 present charts of pH, conductivity and air temperature during the functioning period of the units. It can be observed that generally for all units the pH value increased about 0.4–0.8 of the influent pH value. The conductivity decreased slightly from 1.60 to 1.40 mS cm− 1 for two attached biofilm units and remained almost stable for two constructed wetlands. EC values in constructed wetlands depends on several parameters as soil or gravel composition, evaportranspiration, treatment performance etc. Normally, during a wastewater treatment plant EC values decreased however in constructed wetlands the evaportranspiration and the interactions between soil and water affect on final EC values. Air temperature for almost all cases was above 12 °C. Water temperature expecting to had a variation from the air temperature of 0–3 °C (Akratos and Tsihrintzis, 2007). 3.2. Organic matter removal The outlet concentrations of COD and TSS differed significantly between systems and depended also on loading rate, but the dependency of loading rate differed between systems as shown by the significant interaction term in the two-way ANOVA (Table 4). Effluent levels of COD and TSS were significantly higher for FWS systems compared with the other systems. COD and TSS concentration values did not significantly differ between the HSF, RBC and PBF unit for all tested loading rates (Table 3). On the other hand, FWS effluent was in most cases found to have statistical significant different concentration from the other three systems. The mean removal efficiency of HSF for COD and TSS was 76.9% and 80.7%, 74.7% and 83.4%, 61.0% and 68.4% for low (50–75% of designed), nominal and high (125–175% of designed) HLR respectively (Fig. 3). This is in accordance with previous studies which focused on
2999
the treatment of domestic wastewater with HSF showing removal efficiencies about 75% for COD and 80% for TSS (Vymazal, 2005, Babatunde et al., 2008). Furthermore, our results suggest that a possible underestimation of designed HLRs may result in a considerable decrease of COD and TSS reduction. RBC and PBF removed 86.7%– 90.4%, 95.1%–95.4% and 74.6%–72.0% of the COD for low, nominal and high HLRs respectively. High removal efficiencies observed in nominal HLRs for these treatment systems agree well with data from literature (Gebara,1999; Ayoub and Saikaly, 2004). Generally, the removal of COD decreased in all systems for HLRs higher than the loading the wetlands were designed to operate. The decreasing COD removal feasibility of treatment systems facing fluctuating HLRs is a remarkable disadvantage because in high tourist regions such as Crete Island there are significant fluctuations of influent quantity between winter (no or low tourist period) and summer (extremely high tourist period). The unit with the lowest removal efficiency for COD was the FWS (57.0%, 32.3% and 20.1%). These COD removal rates were lower than mean ranges reported in literature. For instance, a combined Typha domingensis and P. australis wetland treating urban wastewater in Spain removed 87% of the incoming COD (Gomez Cerezo et al., 2001). Similar results were observed by other researchers in the USA (Verhoeven and Meuleman, 1999) and in Europe (Brix, 1994b; Tsihrintzis et al., 2007). However, there are few other researchers who found COD removal efficiencies relatively close to our observations. A wetland system in a rural area in Spain for sewage treatment with Typha latifolia and Salix atrocinerea removed only 60% of the incoming COD (Ansola et al., 2003) and Rousseau et al. (2004) reported average COD removal efficiencies of 61% for FWS treatment systems in Flanders region. The hydraulic loading rates applied in this study in order to reduce land requirements were abnormally high as compared with most works. The high organic strength of influent resulting in a high organic loading rate was the major reason for poor treatment feasibility of FWS. Hiley (1995) report that most wetlands are oxygen limited and that performance is
Table 3 Mean values of quality parameters for the effluent from each treatment system for different HLRs System FWS HLRd
HSF HLRd
RBC HLRd
PBF HLRd
COD (mg/L)
TSS (mg/L)
TN (mg/L)
NO−3 (mg/L)
NH+4 (mg/L)
TP (mg/L)
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
50% 75% 100% 125% 150% 175%
227.9a 228.3a 230.0a 317.9a 187.9a 294.3a
97.3 114.2 76.5 83.8 101.9 133.2
58.9a 62.0a 75.9a 126.7a 88.8a 70.1a
17.7 35.6 24.4 72.3 40.9 32.5
58.5a 55.8a 65.7a 87.0a 63.6a 65.8a
9.2 16.6 8.4 18.4 3.4 4.9
11.3a 11.3a 11.4a 9.7a 12.4a 11.2a
1.0 1.2 1.0 0.9 0.9 1.5
17.5a 10.4a 6.2a 6.9a 8.4a 15.2a
4.9 4.2 3.0 2.6 2.8 3.8
0.9a 0.7a 2.1a 1.4a 1.0a 0.7a
0.9 0.5 1.9 1.1 0.5 0.4
50% 75% 100% 125% 150% 175%
130.5b 99.6b 85.7b 34.6b 126.3a 341.1a
103.3 49.4 58.9 24.8 65.2 189.1
26.0b 20.7b 29.4b 25.7b 30.8b 39.4a
6.7 5.1 10.2 19.9 12.6 10.8
68.9a 41.2a 49.4b 47.0b,c 51.2a,b 67.4a
25.0 12.3 7.3 7.8 8.8 7.2
8.6b 9.2b 7.8b 7.9b 9.1b 10.0a,b
0.7 0.6 0.6 0.3 0.8 0.8
17.4a 11.4a 7.8b 7.2a 7.8a 14.6a
8.4 1.6 3.6 4.4 1.7 4.5
6.2a 3.1a 2.7a 1.2a 1.3a 0.7a
7.1 2.5 2.2 0.4 0.3 0.6
50% 75% 100% 125% 150% 175%
95.5b 43.4b 16.6b 38.4b 94.0a 178.9a
62.8 18.4 9.7 19.6 63.1 101.8
14.6c 11.9b 14.9b 14.2b 26.0b 61.8a
5.4 4.5 6.3 7.5 8.1 23.1
59.2a 38.4a 25.6c 61.5b 42.2b 41.4b
9.7 10.4 5.1 11.6 11.4 4.5
8.6b,c 7.9c 8.6c 8.0b 6.8b 8.7b,c
1.0 0.9 1.7 1.3 2.4 0.8
2.9b 1.8b 5.5b 8.6a 10.7a 5.2b
1.3 1.1 3.1 3.9 1.0 3.9
25.1b 30.5b 28.1b 2.9a 1.8a 22.0b
23.0 8.0 20.9 2.5 0.7 8.8
50% 75% 100% 125% 150% 175%
51.9b 50.3b 18.3b 32.0b 111.2a 216.0a
43.0 32.2 18.3 16.0 61.0 123.1
10.2c 11.1b 19.3b 19.2b 33.8b 57.5a
8.6 7.0 8.1 3.5 9.8 24.9
55.0a 47.5a 28.0c 35.5c 38.0b 39.0b
10.4 13.5 5.1 15.3 13.3 17.1
7.5c 8.1c 8.6c 7.8b 8.6b 8.1c
0.9 1.1 0.7 0.8 1.1 0.9
1.1b 1.5b 4.4b 5.2a 7.1a 12.5a,b
0.5 1.1 2.0 2.6 1.4 1.0
24.8b 34.6b 27.8b 3.5a 2.6a 5.2a
25.8 10.8 13.3 2.7 1.6 0.5
a, b, c: In each column for each loading rate, mean values followed by a different symbol are significantly different (P b 0.05).
3000
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
Fig. 4. Variation of pH, electrical conductivity and air temperature during operation period. Fig. 3. Removal percentage for each HLR and system.
enhanced if extra aeration is provided. The oxygen release from roots of macrophytes seems to be far less than the amount needed for aerobic degradation of the oxygen consuming substances delivered with sewage (Brix, 1990; Brix and Schierup, 1990). As shown in Fig. 5, there is a correlation between the loading rate and the effluent values of the analyzed parameters. It has been observed that there is a positive linear relationship between increasing load and effluent COD or TSS concentration. In the case of RBC and PBF we observed significant adverse effects to organic matter concentration for increasing load. On the other hand the two constructed wetlands turned out to be less sensitive to variations in HLRs.
suggest that the denitrification process was the rate limiting step for nitrogen biodegradation. On the other hand during high HLRs nitrification occurred only partially as the NO3–N concentrations were the lowest. The inhibition of the nitrification process under high organic loading rates in attached biofilm reactors were in general consistent to several other published reports (Gupta and Gupta, 2001; Sirianuntapiboon, 2006). Relatively poor nitrogen removal performance of constructed wetlands treating domestic wastewater was often reported in literature (Tanner et al., 1999; Rousseau et al., 2004; Brix and Arias, 2005). In this work, TN average removal of both FWS and HSF with each examined loading rate was not above 12.7% and 27.0%, respectively. NH4+–N and NO3–N measured concentrations showed that both ammonification and nitrification processes were low. In fact, the magnitude processes which ultimately remove the total nitrogen
3.3. Nutrients removal The effluent quality with respect to nitrogen was better for the two attached biofilm technologies, while nitrogen values were moderate for HSF and still high for FWS (Table 3.). Nitrate concentrations were significantly higher for the RBC and PBF than for the two constructed wetlands. Even though ammonium mean concentration values for RBC and PBF were lower than for FWS and HSF, statistically there were no significant differences. The average TN removal efficiencies of two attached biofilm systems under nominal HLRs were 62.3% and 58.1%. These results were in accordance with previous reported nitrogen removal values of about 45–65% (Tawfik et al., 2006; Griffin et al., 1999). An interesting point to note was that nitrogen removal of these systems was decreased both for low and for high HLRs operation. According to Table 3, for low (50–75%) HLRs these systems had the lowest NH4–N concentrations; the higher NO3–N concentrations
Table 4 Results of two-way ANOVA (F-ratios) used to test for differences in outlet water quality between the four systems tested (main factor ‘System’) and the six loading rates (main factor ‘Loading’) and their interaction (System × Loading) Parameter
System
Loading
System × Loading
COD TSS Total-N Total-P NH4–N NO3–N Total coliforms Fecal coliforms Fecal streptococci
40.11⁎⁎⁎ 76.87⁎⁎⁎ 37.52⁎⁎⁎ 92.75⁎⁎⁎ 34.76⁎⁎⁎ 30.63⁎⁎⁎ 7.20⁎⁎ 0.77NS 4.55⁎
17.42⁎⁎⁎ 12.03⁎⁎⁎ 12.00⁎⁎⁎ 3.80⁎⁎ 14.65⁎⁎⁎ 11.29⁎⁎⁎ 0.67NS 0.59NS 1.07NS
2.36⁎⁎ 4.16⁎⁎⁎ 5.84⁎⁎⁎ 3.36⁎⁎⁎ 10.89⁎⁎⁎ 3.77⁎⁎⁎ 1.40NS 1.06NS 0.38NS
⁎: P b 0.05; ⁎⁎: P b 0.01; ⁎⁎⁎: P b 0.001; NS: Not Significant.
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
3001
Table 6 Removals rate and costs of each wastewater treatment system Treatment system
FWS HSF RBC PBF
Requirements
Removal rate (gr inhab.− 1 d− 1) COD
TSS
N
P
Land (m2 inhab.− 1)
Power (kWh inhab.− 1 year− 1)
33.4 62.0 75.8 75.6
21.0 30.4 33.2 32.4
0.4 3.8 8.4 8.0
0.7 1.4 1.3 1.3
1.4 1.4 0.3 0.2
~0 ~0 219 43
Construction costs (€ inhab.− 1) 237 268 476 672
from the HLR indicating that the main mechanism for TP removal is the adsorption to the porous media (Knight et al., 2000; Jing et al., 2001, Vymazal, 2002; Akratos and Tsihrintzis, 2007). 3.4. Pathogen removal According to Table 5, average concentrations of pathogenic bacteria in the influent in the range of typical domestic wastewater (Karrathanasis et al., 2003; Decamp and Warren, 2000). Furthermore the average concentration of fungi in the influent was 36 ± 18 × 105 CFU/ 100 ml. Faecal indicators were efficiently removed in all examined systems. The average removal efficiencies for TC were between 1.3 and 2.2 log units. HSF showed a better efficiency to remove pathogens than the other systems with average effluent concentrations of 0.3 × 105 and 0.4 × 105 CFU/100 ml for FC and FS, respectively. These values show that only HSF met the level recommended by the EPA of 1.0 × 105 CFU/ 100 ml. These efficiencies appear to be in the range described in previous studies both for constructed wetlands (Karrathanasis et al., 2003; Ansola et al., 2003; Keffala and Ghrabi, 2005; Ottova et al., 1997; Sleytr et al., 2007) and attached biofilm reactors (Sagy and Kott, 1990; Tawfik et al., 2002; Hua et al., 2003). 3.5. Criteria for optimum treatment system Fig. 5. Correlation charts of effluent concentration and pollutant load for COD, TSS, TN and TP.
from these systems is usually low and therefore removal of TN is commonly low in single-stage constructed wetlands (Vymazal, 2007). The concentration of TP was reduced in HSF, RBC and PBF in relatively satisfactory efficiencies ranging from 40 to 50%. This result is in accordance with previous works observing removal efficiencies of about 25–65% (Chen et al., 2006; Tsihrintzis et al., 2007; Vymazal, 2005; Kivaisi, 2001; Yun et al., 2004). FWS was found to have the lower capacity to remove phosphorus. The low removal rate observed in FWS may be due to the fact that the most important processes involved, occur in the sediment and not in the water column (Verhoeven and Meuleman, 1999; Kadlec, 2006). All examined systems show a potential to decrease phosporus removal as loading rate increase. However this potential seems to be rather independent
A comparison of several critical parameters among the examined non-conventional wastewater treatment systems is presented in Table 6. Results based on per capita values suggested that the average wastewater production is 0.2 m3 per capita. Removal rates were calculated from experimental data during the period which the systems were operated at 100% of designed HLR. Remarked that land requirements and construction costs evaluated based only on this specific experimental process. The value of construction cost for FWS and HSF was 237 and 268 € Inhab− 1 respectively while Rousseau et al. (2004) found significant higher average construction cost of constructed wetlands treating wastewater in Flanders region (392 and 1258 € Inhab− 1 for FWS and HSF, respectively). The overloaded designed constructed wetlands in this work resulting lower construction cost as well as lower land requirement. Furthermore, Rousseau et al. (2004) found a footprint for FWS and HSF of 7.0 and 4.8 m2 Inhab− 1 respectively whereas in this work
Table 5 Influent and effluent concentrations, and removal efficiencies for microbiological parameters for each treatment system Parameter
Total coliforms Mean S.D. Faecal coliforms Mean S.D. Faecal streptococci Mean S.D.
Influent
FWS
(CFU/100 ml)
(CFU/100 ml)
Removal
HSF (CFU/100 ml)
Removal
(CFU/100 ml)
RBC Removal
(CFU/100 ml)
PBF Removal
x 105
x105
log units
x 105
log units
x105
log units
x 105
log units
411.5 217.4
20.4 14.0
1.3 1.2
2.6 2.7
2.2 1.9
2.9 4.2
2.2 1.7
16.8 11.7
1.4 1.3
56.3 38.3
5.8 3.9
1.0 1.0
0.3 0.3
2.3 2.1
4.8 11.1
1.1 0.5
5.0 6.7
1.1 0.8
47.9 30.7
6.0 2.6
0.9 1.1
0.4 0.4
2.1 1.9
1.5 2.3
1.5 1.1
3.4 3.1
1.1 1.0
3002
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003
the land requirement was notably lower. Although FWS found to have the lowest construction cost and generally very low requirements, the effluent quality characteristics are significantly poorer than from the other three systems. The effort to operate this system in higher HLRs in order to reduce land requirements had significant negative effect on effluent characteristics. On the other hand RBC and PBF, which require the smallest land area (0.2–0.3 m2 inhab− 1), showed approximately the same effluent values both for organic matter and nutrients concentrations. The quite low power requirements for PBF as compared to RBC balance the relatively higher construction costs. The construction cost for HSF is significant lower than those for the two fixed film media units and additionally found to have almost the same removal rates for COD, TSS and TP. Consequently, the HSF wetland designed and operated under high HLRs showed both effective wastewater treatment and low land requirements. However nitrogen removal efficiency is by far lower. Thus, the choice of the optimal treatment system depends on the overall requirements in terms of treatment efficiency. 4. Conclusions Four non-conventional wastewater treatment systems operated continuously with various HLRs for about 8 months. During this period, with the exception of FWS all other units used in this study removed efficiently the organic matter of domestic wastewater. The RBC and PBF performance showed lower performance efficiency under shock loading conditions. Nitrogen reduction was higher in RBC and PBF than in the two constructed wetlands while phosphorous removal was not significantly different for all examined systems. Additionally, all studied non-conventional units showed considerable potential for removing fecal bacteria from domestic wastewater. Overall, treatment efficiencies of the two attached biofilm technologies are comparable while power requirements are considerable higher for RBC. HSF wetland designed to operate under high HLRs provide comparable effluent quality as attached biofilm technologies whereas the construction and operation costs are lower. However whenever the requirements of nitrogen level in the effluent are crucial, HSF would not successfully meet these criteria. Since overloaded designed FWS wetlands showed the lowest treatment performance, the selection of this technology could only be justified when land is available (lower HLRs) and the effluent quality requirements are not very strict. Therefore, the selection of the optimum unit depends on the demand for treatment efficiency and the total available budget. Acknowledgements This research was funded by the Greek General Secretariat for Research and Technology, and well as Prisma Domi S.A. The authors wish to express their gratitude to Professor Vassilis Manios, Vaggelis Theodorakopoulos and Ioannis Sabathianakis for their continued involvement in the realisation of this project. References Aesoy A, Odegaard H, Bach K, Pujol R, Hamon M. Dentrification in a packed bed biofilm reactor (biofor) — experiments with different carbon sources. Water Res 1998;32:1463–70. Akratos CS, Tsihrintzis VA. Effect of temperature, HRT, vegetation and porous media on removal efficiency of pilot-scale horizontal subsurface flow constructed wetlands. Ecol Eng 2007;29:173–91. Ansola G, Gonzalez JM, Cortijo R, de Luis E. Experimental and full-scale pilot plant constructed wetlands for municipal wastewaters treatment. Ecol Eng 2003;21:43–52. APHA. Standard Methods for the Examination of Water and Wastewater. 19th ed. Washington DC, USA: American Public Health Association; 1995. Ayoub GM, Saikaly P. The combined effect of step-feed and recycling on RBC performance. Water Res 2004;38:3009–16. Boller M. Small wastewater treatment plants-a challenge to wastewater engineers. Water Sci Technol 1997;35:1-12.
Babatunde AO, Zhao YQ, O'Neill M, O'Sullivan B. Constructed wetlands for environmental pollution control: a review of developments, research and practice in Ireland. Environ Int 2008;34:116–26. Brix H. Gas exchange through the soil–atmosphere interphase and through dead culms of Phragmites australis in a constructed reed bed receiving domestic sewage. Water Res 1990;24:259–66. Brix H. Use of constructed wetlands in water pollution control: historical development, present status, and future perspectives. Water Sci Technol 1994a;30:209–23. Brix H. Constructed wetlands for municipal wastewater treatment in Europe. Global wetlands; 1994b. p. 325–33. Brix H, Arias CA. The use of vertical flow constructed wetlands for on-site treatment of domestic wastewater: new Danish guidelines. Ecol Eng 2005;25:491–500. Brix H, Schierup H-H. Soil oxygenation in constructed reed beds: the role of macrophyte and soil–atmosphere interface oxygen transport. In: Cooper PF, Findlater BC, editors. Constructed wetlands in water pollution control. Oxford: Pergamon; 1990. p. 53–66. Caselles-Osorio A, Porta A, Porras M, Garcia J. Effect of high organic loading rates of particulate and dissolved organic matter on the efficiency of shallow experimental horizontal subsurface-flow constructed wetland. Water Air Soil Pollut 2007a;183:367–75. Caselles-Osorio A, Puigagut J, Segu E, Vaello N, Granes F, Garcia D, et al. Solids accumulation in six full-scale subsurface flow constructed wetlands. Water Res 2007b;41:1388–98. Chen TY, Kao CM, Yeh TY, Chien HY, Chao AC. Application of a constructed wetland for industrial wastewater treatment: a pilot-scale study. Chemosphere 2006;64:497–502. Chung AKC, Wu Y, Tam NFY, Wong MH. Nitrogen and phosphate mass balance in a sub-surface flow constructed wetland for treating municipal wastewater. Ecol Eng 2008;32:81–9. Colmenarejo MF, Rubio A, Sanchez E, Vicente J, Garcia MG, Borja R. Evaluation of municipal wastewater treatment plants with different technologies at Las Rozas, Madrid (Spain). J Environ Manag 2006;81:399–404. Decamp O, Warren A. Investigation of Escherichia coli removal in various designs of subsurface flow wetlands used for wastewater treatment. Ecol Eng 2000;14:293–9. European Parliament and Council. Water framework directive (Directive 2000/60/EC of 23 October 2000). Eur J 2000;L327:1-72. Fahd K, Martin I, Salas JJ. The Carrion de los Cespedes Experimental Plant and the technological Transfer Centre: urban wastewater treatment experimental platforms for the small rural communities in the Mediterranean area. Desalination 2007;215:12–21. Gebara F. Activated sludge biofilm wastewater treatment system. Water Res 1999;33:230–8. Gomez Cerezo R, Suarez ML, Vidal-Abarca MR. The performance of a multistage system of a constructed wetlands for urban wastewater treatment in a semiarid region of SE Spain. Ecol Eng 2001;16:501–17. Griffin P, Jennings P, Bowman E. Advanced nitrogen removal by rotating biological contactors recycle and constructed wetlands. Water Sci Technol 1999;40:383–90. Gupta AB, Gupta SK. Simultaneous carbon and nitrogen removal from high strength domestic wastewater in an aerobic RBC biofilm. Water Res 2001;35:1714–22. Hansford GS, Andrews JF, Grieves CG, Carr AD. A steady-state model for the rotating biological disc reactor. Water Res 1978;12:855–68. He Q, Mankin KR. Performance variations of COD and nitrogen removal by vegetated submerged bed wetlands. J Am Water Resour Assoc 2002;38:1679–89. Helmer C, Kunst S. Simultaneous nitrification/denitrification in aerobic bio-film system. Water Sci Technol 1998;37:183–7. Hiley PD. The reality of sewage treatment using wetlands. Water Sci Technol 1995;32:329–38. Hua J, An P, Winter J, Gallert C. Elimination of COD, microorganisms and pharmaceuticals from sewage by trickling through sandy soil below leaking sewers. Water Res 2003;37:4395–404. Jing SR, Lin YF, Lee DY, Wang TW. Nutrient removal from polluted river water by using constructed wetlands. Biores Technol 2001;76:131–5. Kadlec RH. Free surface wetlands for phosphorus removal: the position of the Everglades Nutrient Removal Project. Ecol Eng 2006;27:361–79. Karrathanasis AD, Potter CL, Coyne MS. Vegetation effects on faecal bacteria, BOD, and suspended solid removal in constructed wetlands treating domestic wastewater. Ecol Eng 2003;20:157–69. Keffala C, Ghrabi A. Nitrogen and bacterial removal in constructed wetlands treating domestic wastewater. Desalination 2005;185:383–9. Kivaisi AK. The potential for constructed wetlands for wastewater treatment and reuse in developing countries: a review. Ecol Eng 2001;16:545–60. Knight RL, Payne VWE, Borer RE, Clarke RA, Pries JH. Constructed wetlands for livestock wastewater management. Ecol Eng 2000;15:41–55. Korkusuz EA, Beklioglu M, Demirer GN. Use of blast furnace granulated slag as a substrate in vertical flow reed beds: field application. Biores Technol 2007;98:2089–101. Kuschk P, Wiebner A, Kappelmeyer U, Weibrodt E, Kastner M, Stottmeister U. Annual cycle of nitrogen removal by a pilot-scale subsurface horizontal flow in a constructed wetland under moderate climate. Water Res 2003;37:4236–42. Lin YF, Jing SR, Lee DY, Chang YF, Chen YM, Shih KC. Performance of a constructed wetland treating intensive shrimp aquaculture wastewater under high hydraulic loading rate. Environ Pollut 2005;134:411–21. Mann AT, Stephenson T. Modelling biological aerated filters for wastewater treatment. Water Res 1997;31:2443–8. Molle P, Lienard A, Grasmick A, Iwema A. Effect of reeds and feeding operations on hydraulic behaviour of vertical flow constructed wetlands under hydraulic overloads. Water Res 2006;40:606–12. Nitisoravut S, Klomjek P. Inhibition kinetics of salt-affected wetland for municipal wastewater treatment. Water Res 2005;39:4413–9. Ottova V, Balcarova J, Vymazal J. Microbial characteristics of constructed wetlands. Water Sci Technol 1997;35:117–23.
M.S. Fountoulakis et al. / Science of the Total Environment 407 (2009) 2996–3003 Paing J, Voisin J. Vertical flow constructed wetlands for municipal wastewater and septage treatment in French rural area. Water Sci Technol 2005;51:145–55. Rousseau DPL, Vanrolleghem PA, De Pauw N. Model-based design of horizontal subsurface flow constructed treatment wetlands: a review. Water Res 2004;38:1484–93. Sagy M, Kott Y. Efficiency of rotating biological contactor in removing pathogenic bacteria from domestic wastewater. Water Res 1990;24:1125–8. Schubert W, Wolfgang Günthert F. Particle size distribution in effluent of trickling filters and in humus tanks. Water Res 2001;35:3993–7. Sirianuntapiboon S. Treatment of wastewater containing Cl2 residue by packed cage rotating biological contactor (RBC) system. Bioresour Technol 2006;97:1735–44. Sleytr K, Tielz A, Langergraber G, Haberl R. Investigation of bacterial removal during the filtration process in constructed wetlands. Sci Total Environ 2007;380:173–80. Tanner CC, D'Eugenio J, McBride GB, Sukias JPS, Thompson K. Effect of water level fluctuation on nitrogen removal from constructed wetland mesocosms. Ecol Eng 1999;12:67–92. Tawfik A, Klapwijk B, El-Gohary F, Lettinga G. Treatment of anaerobically pre-treated domestic sewage by a rotating biological contactor. Water Res 2002;36:147–55. Tawfik A, Temmink H, Zeeman G, Klapwijk B. Sewage treatment in rotating biological contactor (RBC) system. Water Air Soil Pollut 2006;175:275–89.
3003
Tsihrintzis VA, Akratos CS, Gikas GD, Karamouzis D, Angelakis AN. Performance and cost comparison of a FWS and a VSF constructed wetland system. Environ Technol 2007;28:621–8. USEPA. Design manual constructed wetlands and aquatic plant system for municipal wastewater treatment. Office of Research and Development, Cincinnati, Ohio, USA; 1988; 1988. USEPA. Constructed wetlands treatment of municipal wastewaters. Office of Research and Development, Cincinnati, Ohio, USA; 2000; 2000. Verhoeven JTA, Meuleman AFM. Wetlands for wastewater treatment: opportunities and limitations. Ecol Eng 1999;12:5-12. Vymazal J. The use of sub-surface constructed wetlands for wastewater treatment in the Czech Republic: 10 years experience. Ecol Eng 2002;18:633–46. Vymazal J. Horizontal sub-surface flow and hybrid constructed wetlands systems for wastewater treatment. Ecol Eng 2005;25:478–90. Vymazal J. Removal of nutrients in various types of constructed wetlands. Sci Total Environ 2007;380:48–65. Yun Z, Lee H, Choi E. Enhanced biological phosphorus removal in RBC with SBR modification. Water Sci Technol 2004;50:121–30.