- Email: [email protected]
Reuse of drinking water treatment residuals as a substrate in constructed wetlands for sewage tertiary treatment
Ecological Engineering 70 (2014) 295–303
Contents lists available at ScienceDirect
Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng
Reuse of drinking water treatment residuals as a substrate in constructed wetlands for sewage tertiary treatment Leilei Bai a , Changhui Wang a , Caihong Huang b , Liansheng He b , Yuansheng Pei a,∗ a b
The Key Laboratory of Water and Sediment Sciences, Ministry of Education, School of Environment, Beijing Normal University, Beijing 100875, PR China Water Environment System Project Laboratory, Chinese Research Academy of Environmental Sciences, Beijing 100012, PR China
a r t i c l e
i n f o
Article history: Received 22 September 2013 Received in revised form 26 April 2014 Accepted 23 June 2014 Keywords: Constructed wetlands Drinking water treatment residuals Nitrogen Phosphorus Tertiary treatment
a b s t r a c t This work investigated the feasibility of reusing drinking water treatment residuals (WTR) as a substrate in constructed wetlands (WTR-CW) to treat secondary effluent at short hydraulic retention times (HRTs) (1–3 d). The results indicated that both continuous flow operation (CFCW) and tidal flow operation (TFCW) can achieve satisfactory removal of total nitrogen (TN), total phosphorus (TP), chemical oxygen demand and suspended solids although the ammonia nitrogen concentration of the CFCW effluent increased slightly. The WTR was found to be beneficial for denitrification, and the mean nitrate removal rates of CFCW and TFCW were 3.45 and 2.47 g N/m3 d, respectively. The TP removal efficiency of the two WTR-CWs still remained at 98% after 260 d of operation, and the lifetime regarding P saturation was estimated to be longer than 10 years. The HRT played a more significant role in TN removal, and the most optimal and stable TN removal (>76%) was obtained at 3 d HRT. Moreover, the leaching of Fe and Al from the two WTR-CWs was minor. Based on regulations, it is feasible to reuse the WTR as a substrate in constructed wetlands for the purpose of sewage tertiary treatment and waste recycling. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Nutrients, including nitrogen (N) and phosphorus (P), from municipal wastewater can lead to eutrophication of the receiving water bodies, including fresh-water and some seawater (Conley et al., 2009; Oleszkiewicz and Barnard, 2006). Normally, it is difficult to maintain the effluent of secondary biological treatment at a cleanliness level that satisfies the environmental standard to enable it to be discharged into the receiving water bodies (SEPA, 2003). Various processes, such as coagulation, ion exchange and adsorption, have been used as tertiary treatments to deal with the remaining contaminants after secondary treatment (Ganrot et al., 2007; Gupta et al., 2012; Wang et al., 2005). Among these tertiary treatment methods, constructed wetlands (CW) have
Abbreviations: CFCW, continuous flow operated WTR-CW; CODcr , chemical oxygen demand; CW, constructed wetlands; DO, dissolved oxygen; HRT, hydraulic retention time; ICP-AES, inductively coupled plasma atomic emission spectroscopy; N, nitrogen; NH4 -N, ammonia nitrogen; NO2 -N, nitrite; NO3 -N, nitrate; P, phosphorus; SS, suspend solids; TFCW, tidal-flow operated WTR-CW; TN, total N; TP, total P; WTR, drinking water treatment residuals; WTR-CW, CW based on WTR. ∗ Corresponding author. Tel.: +86 10 5880 1830; fax: +86 10 5880 1830. E-mail address: [email protected] (Y. Pei). http://dx.doi.org/10.1016/j.ecoleng.2014.06.015 0925-8574/© 2014 Elsevier B.V. All rights reserved.
demonstrated advantages in terms of aesthetics, lower energy consumption and more economical construction and operation. The performance of CW is generally good in terms of the removal of organic matter and suspended solids (SS). Effective N removal can also be achieved by different operation schemes, such as continuous feeding, intermittent feeding, step-feeding and artificial aeration (Albuquerque et al., 2012; Hu et al., 2012a; Vymazal, 2001). Traditionally, CW has a limited capacity for the removal of P unless materials with high P adsorption capacity are used (Vymazal, 2007). Therefore, many researchers have used some industry by-products and wastes as substrates for enhancing P removal (Drizo et al., 2002; Wendling et al., 2012). Recently, significant research effort has been focused on the reuse of drinking water treatment residuals (WTR) in CW (WTR-CW) to facilitate water reuse and waste recycling. Water treatment residuals, a non-hazardous inevitable byproduct generated from drinking water treatment plants, has been demonstrated to exhibit high P retention capabilities (Yang et al., 2006). Babatunde et al. (2010) demonstrated that a pilot fieldscale WTR-CW operated in tidal-flow mode exhibited satisfactory performance of the removal of P and organic matter. In addition, Hu et al. (2012a,b) constructed a four-stage WTR-CW and a single-stage WTR-CW with intermittent aeration to treat dairy wastewater, each of which exhibited a high rate of N removal.
296
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Table 1 Wastewater characteristics (mean ± SD) and average loading rates.
a
HRT (d) HLR (m3 /m3 d)a CODcr (mg/L) NH4 -N (mg/L) NO2 -N (mg/L) NO3 -N (mg/L) TN (mg/L) TP (mg/L) SS (mg/L) pH OLR (g CODcr /m3 d)a NLR (g N/m3 d)a PLR (g P/m3 d)a CODcr /TN a b c
Period 1 (15–100 d)
Period 2 (100–200 d)
Period 3 (200–260 d)
1 0.45 45.03 ± 20.23 0.65 ± 0.64 0.24 ± 0.17 16.70 ± 2.28 17.88 ± 2.54 1.88 ± 0.45 2.10 ± 0.24 8.29 ± 0.18 20.25 8.04 0.84 2.52
2 0.22 53.72 ± 18.87 0.28 ± 0.21 0.04 ± 0.02 15.80 ± 4.15 16.61 ± 4.15 1.83 ± 0.68 2.47 ± 0.44 8.36 ± 0.19 11.83 3.66 0.40 3.23
3 0.15 71.10 ± 45.00 0.69 ± 0.67 0.39 ± 0.29 11.10 ± 1.93 12.96 ± 2.74 1.15 ± 0.50 2.73 ± 0.31 8.18 ± 0.40 10.67 1.95 0.18 5.48
Dairy wastewaterb – – 3491 ± 368 ± – – 442 ± 39 ± 1739 ± 7.87 ± – – – –
2045 114
114 22 1687 0.3
Discharge standardc – – 50 5 – – 15 0.5 10 6–9 – – – –
HRT: hydraulic retention time; HLR: hydraulic loading rate; OLR: organic loading rate; NLR: nitrogen loading rate; PLR: phosphorus loading rate. Refer to Hu et al. (2012b). The first-class requirements of Discharge Standard of Pollutants for Municipal Wastewater Treatment Plant (GB 18918-2002) (SEPA, 2003).
Overall, the WTR-CW has been successful in the treatment of highstrength wastewater (high SS, chemical oxygen demand (CODcr ), total N (TN) and total P (TP)), demonstrating a high removal efficiency for pollutants. However, there have been limited data reported regarding the application of WTR-CW for sewage tertiary treatment. Unlike dairy wastewater, secondary effluent is normally characterised by low CODcr and SS levels and a relatively lower nutrient level. Therefore, there is a significant need to determine the feasibility of using WTR-CW to purify secondary effluents. In this study, two lab-scale single-stage WTR-CWs were constructed to treat nitrate (NO3 -N)-dominated secondary effluent from a municipal wastewater treatment plant. The performance of the two CW operation schemes, including continuous flow (CFCW) and tidal flow (TFCW), was monitored. The objectives of the study were (1) to investigate the effectiveness of WTR-CW under two different operation schemes, (2) to determine the mechanisms of nutrients removal, and (3) to estimate the applicability of WTRCW as a tertiary treatment technology. The results of this work provide support for the application of WTR-CW for sewage tertiary treatment.
2. Materials and methods 2.1. Sample preparation Dewatered WTR was collected from Beijing City No. 9 Waterworks in China. The particle sizes of the raw cake of WTR were in the range of 1–3 cm, and the moisture content was 55%. The characteristics of the WTR were available in former studies (Wang et al., 2012a, 2013a): Fe of 101.56 mg/g, Al of 50.36 mg/g, P of 0.61 mg/g, pH of 7.04, organic matter of 57.65 mg/g and surface area of 78.83 m2 /g. Furthermore, a maximum P adsorption capacity of 7.42 mg/g was calculated at pH of 7 using Langmuir isotherm (Fig. S1 in Supplementary Material). The secondary effluent was collected after an A2 /O process from the Xiaojiahe Sewage Treatment Plant of Beijing. The A2 /O process was a single-sludge suspended growth system incorporating anaerobic, anoxic and aerobic zones in sequence (Wang et al., 2006). After collection, the wastewater was stored at 4 ◦ C. The characteristics and loading rates during the entire waste treatment operation are summarised in Table 1. Compared with the dairy wastewater studied in previous work (Hu et al., 2012b), the secondary wastewater had significantly lower CODcr , SS, TN and TP levels. However, the secondary wastewater (CODcr , TN and TP) still failed to meet the requirements of the standard (SEPA, 2003). In
the secondary wastewater, the TN primarily existed in the form of NO3 -N. Overall, tertiary treatment was required to improve the water quality of the secondary wastewater. 2.2. System description and operation The two lab-scale single-stage WTR-CW systems were implemented with two Plexiglas columns (diameter of 9.3 cm and depth of 90 cm) (Fig. 1). Gravel (depth of 10 cm) was added at the bottom of the columns as the support medium, and a 60-cm depth of dewatered WTR (dry weight of 1.2 kg) was added as the main wetland medium layer, which resulted in a total volume of 5.09 L and an effective volume of 2.0 L (an initial porosity of 39%). Common reeds, Phragmitesaustralis, were planted on the top of the two columns. For the CFCW, the influent was continuously introduced into the column from the bottom by a peristaltic pump, while the effluent overflowed from the top. The desired hydraulic retention times (HRTs) were adjusted by controlling the flowrate of the influent. For the TFCW, the influent was introduced into the column from the top, while the effluent was drained from the bottom by peristaltic pumps. The feeding dosage and the discharge dosage were both half of the total volume per cycle, thus maintaining a stationary volume of 1.0 L during the experiments. The wet/dry (h) ratio was maintained at 22:2 by using pre-set programmable timers. The desired HRTs were adjusted by control of the flood and drain stages sequentially. The two WTR-CWs were operated for 260 d at room temperature, and this time was divided into three periods according to the theoretical HRTs (Table 1). Samples were collected from the influent and effluent of the two WTR-CWs 1–2 times per week and were analysed for CODcr , SS, ammonia nitrogen (NH4 -N), nitrite (NO2 -N), NO3 -N, TN, TP, and pH. In addition, the samples for Fe and Al analysis were collected from the effluent once each month. 2.3. Sample analysis The CODcr level was measured using the potassium dichromate method (Maria et al., 2004), and the SS level was measured using the gravimetric method (Clescerl et al., 1995). The NH4 -N, NO2 -N and NO3 -N levels were detected using a spectrophotometer (UV-2000, Unico) according to the standard methods (Clescerl et al., 1995). The TN level was measured by alkaline potassium persulphate oxidation followed by NO3 -N analysis (Clescerl et al., 1995). The TP was measured by persulphate digestion followed by orthophosphate analysis (Clescerl et al., 1995). The pH was measured using a pH metre (pH-10, Sartorius). Dissolved Fe and
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
297
Fig. 1. Two different operation schemes of the laboratory WTR-CW: (a) CFCW and (b) TFCW.
Al were analysed by inductively coupled plasma atomic emission spectroscopy (ICP-AES, ULTIMA, JY, France) (Clescerl et al., 1995). The analytical detection limits were 5 mg/L for CODcr , 0.5 mg/L for SS, 0.01 mg/L for NH4 -N, 0.015 mg/L for NO2 -N, 0.1 mg/L for NO3 -N, 0.1 mg/L for TN, 0.02 mg/L for TP, 0.01 for pH, 0.01 mg/L for Fe and 0.01 mg/L for Al. The presented tests were repeated twice, and the average values were reported. The standard error deviation was within 10%. 2.4. Statistical analysis To investigate the difference in treatment performance of the two WTR-CWs under different HRTs, statistical analyses were performed using PASW Statistics 20 (SPSS Inc.). One-way ANOVA multiple comparisons for the mean removal efficiencies were computed using Tamhane’s T2 test at a 95% confidence level (˛ = 0.05). The method was selected because equal variance between groups was not assumed. 3. Results and discussion 3.1. Variation of TN and TP Both the CFCW and the TFCW demonstrated a satisfactory performance in TN and TP removal during the entire operation period (Fig. 2). The mean effluent TN concentrations of the CFCW and the TFCW were 5.27 mg/L and 6.63 mg/L, respectively, resulting in an average TN removal rate of 2.95 g N/m3 d for the CFCW and 2.36 g N/m3 d for the TFCW. The most stable TN removal efficiency of the two WTR-CWs was obtained in period 3, with an observed efficiency of greater than 75%. Compared with the TN removal, the TP removal was more efficient and stable, regardless of the operation conditions. The TP concentrations of the two WTR-CW effluents were both nearly zero, which is significantly lower than the discharge standard (SEPA, 2003). After 260 d of operation, the TP removal efficiency still remained above 98% under a P loading
rate of 0.18 g P/m3 d. Overall, the favourable TN and TP removal under short HRTs (1–3 d) indicated that the WTR-CW was suitable for tertiary nutrient removal. The TN removal rates of the two WTR-CWs were comparable to those reported in a previous study by Lin et al. (2008). Although the N transfer was complex, biological function was the main N removal mechanism (Vymazal, 2007). Wang et al. (2012a) reported that WTR can increase the total abundance of bacteria due to its high surface area, thus providing an adequate surface for biofilm attachment. Furthermore, Wang et al. (2013a) found that WTR can promote aggregation and strengthen the activity of anammox bacteria in lake sediments. Therefore, the WTR was found to be beneficial to biological N removal. Adsorption was the main P removal pathway in the CW (Babatunde et al., 2010). The WTR had a large potential for P adsorption due to its large amount of Fe and Al. In addition, Wang et al. (2013b) demonstrated that organic matter, ion strength and anaerobic conditions did not affect the stability of the P adsorbed by WTR, suggesting that the retained P cannot be desorbed easily. In a former study, Xu et al. (2006) reported that the lifetime of a CW could be calculated using the maximum P sorption capacity of substrate and area of CW. In this study, the amounts of TP removal by the CFCW and the TFCW during the 260 d were 454.16 mg and 452.73 mg (calculated by the difference in between the influent and effluent P concentrations), respectively, resulting in a P adsorption quantity of 0.38 mg/g for the WTR, which is far from the maximum adsorption capacity of 7.42 mg/g. Consequently, considering the presence of other functions, such as physical trapping, filtration and sedimentation, the lifetime of the WTR-CW was estimated to be longer than ten years. There was no significant difference in the TN and TP removal efficiencies between CFCW and TFCW (P > 0.05) according to multiple comparisons (one-way ANOVA) that were performed to analyse the mean removal efficiencies. The main difference between the two operation schemes was oxygen transfer. For the CFCW, the oxygen was mainly derived from the dissolved oxygen (DO) in
298
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Influent and effluent TP (mg/L)
Influent and effluent TN (mg/L)
Period 1
Period 3 Inf Eff of CFCW Eff of TFCW
Period 2
a
24 20 16 12 8 4 0
0
40
80
120
160
200
240
280
80
120
160
200
240
280
4
b 3
2
1
0 0
40
Time (d) Fig. 2. The TN and TP removal of the two WTR-CWs during the entire operation period: (a) TN removal and (b) TP removal.
the influent and from surface diffusion, while both pathways were limited, and the system was always anoxic. In contrast, the internal environment of the TFCW changed from aerobic to anaerobic in one cycle (Wu et al., 2011). During the drain and rest stage, air quickly diffused into the substrate, and the system was aerobic. During the work stage, oxygen was consumed by biological reactions, causing the system to become anaerobic. In this study, the anaerobic stage was dominant according to the high wet/dry ratio of 22:2. Hence, both of the WTR-CWs had a long anaerobic period that favoured the elimination of NO3 -N, which was the dominant species of the TN; thus, the TN removal was similar and effective for the two WTR-CWs. The same trend was found by Lin et al. (2008) using free-water surface wetlands and subsurface-flow wetlands to treat highly nitrated groundwater. Moreover, Wang et al. (2013b) demonstrated that DO had no significant influence on the P adsorption of the WTR. Thus, the P removal was the similar under the different operation schemes; even the anoxic condition of the WTR-CW performed well. Overall, the removal of TN was more complex and exhibited more fluctuations than that of TP, which makes it necessary to determine the N transformation over the entire operation period. 3.2. Variation of N forms The N transformation of the CFCW and the TFCW are shown in Fig. 3. For the CFCW, the NH4 -N concentration of the effluent was greater than 6 mg/L at the beginning; it then decreased and finally became stable at 1.35 ± 1.15 mg/L. The effluent NH4 -N levels during the three periods were all significantly higher than the influent levels (P < 0.05). In contrast, there was a non-significant difference between the NH4 -N concentrations in the influent and the effluent of TFCW (P > 0.05). The stable NH4 -N concentration of the effluent
was 0.75 ± 0.75 mg/L. Moreover, the denitrification rate was calculated by the difference between the influent (NO2 -N + NO3 -N) and effluent levels. For the CFCW, the NO2 -N in the effluent was nearly zero, and the NO3 -N level was in the range of 0–6.78 mg/L, resulting in an average removal efficiency of (NO2 -N + NO3 -N) of 80%. For the TFCW effluent, the (NO2 -N + NO3 -N) concentration was significantly higher than that of the CFCW by 2.53 mg/L (P < 0.05), and the average removal efficiency was 65%. Comparatively, the CFCW had a higher (NO2 -N + NO3 -N) loss than did the TFCW, while the effluent NH4 -N increased slightly. The new production of NH4 -N in the CFCW may result from the ammonification of organic nitrogen in the secondary effluent and the WTR. However, the latter was dominant because the influent organic nitrogen was minimal (<0.50 mg/L), which was also confirmed by the drastic decrease in NH4 -N during the first 100 d when the system was under a steady-state condition. Gray et al. (2000) found a similar trend when using marl as a substrate to treat wastewater with a low NH4 -N influent. In addition, the lower NH4 N of the TFCW effluent suggested that the better oxygen supply promoted the conversion of NH4 -N into NO2 -N and NO3 -N. Overall, note that although the ammonification was clearly weaker than the TN removal, pre-treatment or aeration operation schemes, including artificial aeration, tidal-flow and hybrid systems, were required to eliminate the increase in NH4 -N. In addition to the ammonification-nitrification of the organic nitrogen in the WTR, the weaker denitrification was another reason for the higher (NO2 -N + NO3 -N) level of the TFCW effluent. According to the low ratio of the influent CODcr and N (Table 1), denitrification was the primary mechanism for NO3 N removal (Akunna et al., 1992). Thus the mean denitrification rates of the CFCW and the TFCW were determined to be 3.45 g N/m3 d and 2.47 g N/m3 d, respectively, over the entire period. The
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Inf NH4-N Eff NH4-N Inf NO2-N+NO3-N Eff NO2-N+NO3-N
15 12
28 24 20 16
9
12
6
8
3
4
0 0
40
80
120
160
200
240
0 280
NO2-N+NO3-N concentration (mg/L)
a
18
NH4-N concentration (mg/L)
Period 3
Period 2
Period 1
299
NH4-N concentration (mg/L)
Inf NH4-N Eff NH4-N Inf NO2-N+NO3-N Eff NO2-N+NO3-N
15 12
24 20 16
9
12
6
8
3
4
0 0
40
80
120
160
200
240
0 280
NO2-N+NO3-N concentration (mg/L)
28
b
18
Time (d) Fig. 3. The transformation of the N forms of the two WTR-CWs during the entire period: (a) CFCW and (b) TFCW.
denitrification was controlled by a variety of factors, including the characteristics of wetland substrate, the types and species of macrophyte, microbial communities, oxygen supply, and the NO3 N and organic matter concentrations in the influent (Gale et al., 1993; Lin et al., 2002). Since the average oxygen supply of the TFCW was generally much higher than that of the conventional CW, the denitrification was hindered and the denitrification rate was lower than that of CFCW (Wu et al., 2011). Furthermore, the lack of external organic carbon arising from the secondary effluent did not seem to have a significant influence on the denitrification rate. This may be due to the organic matter of WTR was higher than that of common soil or gravel and could provide carbon source for denitrification. Saeed and Sun (2011) reported that denitrifiers could use not only organic carbon from the influent but also the organic matter of the substrates. In addition, although Campos et al. (2008) indicated the denitrification with sulphur compounds as electron donors is an alternative to heterotrophic denitrification for wastewaters with high nitrate concentration and low organic matter content, in this study, the contribution of sulphide oxidation was minor because the sulfide in the WTR was limited and stable (Wang and Pei, 2012b). Overall, the WTR-CW was appropriate for denitrification, even though the influent organic matter level was low. 3.3. Variation of SS, CODcr and pH The performance of SS, CODcr and pH of the two WTR-CWs is shown in Fig. 4. The two WTR-CWs exhibited an average SS removal rate of 90% under all the tested HRTs (Fig. 4a). Even the influent SS levels were significantly higher during period 3 than during the early days, the effluent was still low. With regards to the organic matter, the average CODcr removals of the CFCW and the TFCW
were 54% and 47%, respectively, with effluent concentrations of less than 50 mg/L (Fig. 4b). The effluent pH of the CFCW was lower than the influent pH by an average of 0.37, and the corresponding pH reduction of the TFCW was 0.76 (Fig. 4c). Overall, the WTR-CW exhibited excellent removal capacities of SS and CODcr , and the change of pH was limited. The SS was removed entirely by physical interception by the substrate, which can be enhanced by the new generation of biofilms. In contrast, the CODcr removal from the CW system was accomplished by cooperation between physical and microbial processes (Lee et al., 2004). The physical process separated the organic solids, thus allowing better hydrolysis for biodegradation, while the substrate allowed the accumulation of immense amounts of attached bacteria, which helps in rapidly catalysing the chemical reactions. Thus, the WTR exhibited good physical characteristics for SS and CODcr removal. Normally, P adsorption of WTR and denitrification lead to an increase of pH (Vymazal, 2007; Yang et al., 2006). The pH could decrease in the presence of organic acid release from the substrate and nitrification (Alburquerque et al., 2006; Vymazal, 2007). In this study, the small decrease of pH as the wastewater flowed through the two WTR-CWs suggested that the organic acid release of the substrate was dominant. In addition, the stronger nitrification of the TFCW resulted in a lower effluent pH than CFCW. However, the variation of pH was in the normal range of the discharge standard. 3.4. The influence of HRT The HRT has a slight influence on the SS removal in the CFCW, according to the average removal efficiencies in the range of 79–92% at the three HRTs (Table 2). This result was also confirmed by the similar slope and R2 of the correlations between the SS mass
300
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Period 2
Influent and effluent SS (mg/L)
Period 1
a
18
Inf Eff of CFCW Eff of TFCW
15 12 9 6 3 0
Influent and effluent COD (mg/L)
0 160
40
80
120
160
200
240
280
40
80
120
160
200
240
280
40
80
120
160
b
140 120 100 80 60 40 20 0 0
10.0
c
9.5
Influent and effluent pH
Period 3
9.0 8.5 8.0 7.5 7.0 6.5 6.0
0
200
240
280
Time (d) Fig. 4. The performance of the removal of SS and CODcr and the pH level of the two WTR-CWs over the operation period: (a) SS, (b) CODcr , and (c) pH.
loading and removal (Fig. 5a). The same trend was found in the TFCW, with an average removal efficiency in the range of 88–93% (Fig. 5e). Although the CODcr removal efficiencies were similar at the three HRTs, the mass removal of the CFCW and the TFCW both exhibited the strongest relationship to mass loading, with an R2 greater than 0.73 at 3 d HRT (Fig. 5b and f). The HRT played a more significant role in the removal of TN. For CFCW, the TN mass
reduction efficiency increased from 54% to 77%, with the increase of HRT occurring for a value of HRT from 1 d to 2 d. However, there was no significant difference between the TN reduction efficiency at a HRT of 2 d and that at a HRT of 3 d. The correlation between TN mass loading rates and removal rates followed the pattern similar to that of the TN reduction efficiency (Fig. 5c). For the TFCW, the TN mass reduction efficiency was approximately two times more at a
Table 2 The average pollutants removal rate (%) of the two WTR-CWs for different HRTs. Parameters
SS CODcr TN TP
CFCW
TFCW
Period 1
Period 2
Period 3
Period 1
Period 2
Period 3
79 54 54 97
87 52 77 98
92 56 77 98
88 45 34 95
89 48 70 96
93 49 79 98
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
5
5
a
3
2
2
0
0
40
1
2 3 3 SS loading (g/m d)
4
5
0
1
2
f
2
HRT=1d R =0.6753 2 HRT=2d R =0.4443 2 HRT=3d R =0.8614
5
2
3
CODcr removal (g/m d)
20
10
20
10
0 0
10
20 30 3 CODcr loading (g/m d)
0
40
10
20 30 3 CODcr loading (g/m d)
40
8
8
c
g
2
HRT=1d R =0.5975 2 HRT=2d R =0.8571 2 HRT=3d R =0.7461
2
HRT=1d R =0.2300 2 HRT=2d R =0.5904 2 HRT=3d R =0.6267
6
3
3
TN removal (g/m d)
6
4
2
0
0
2
4
6
8
4
2
0
10
0
2
3
6
8
10
TN loading (g/m d)
1.5
1.5
d
2
h
HRT=1d R =0.8882 2 HRT=2d R =0.9803 2 HRT=3d R =0.9862
HRT=1d R2=0.8011 HRT=2d R2=0.9898 HRT=3d R2=0.9585
1.2 TP removal (g/m d)
1.2
3
0.9
0.6
0.9
0.6
0.3
0.3
0.0 0.0
4
3
TN loading (g/m d)
3
4
HRT=1d R =0.7291 2 HRT=2d R =0.4670 2 HRT=3d R =0.7263
30
0
TP removal (g/m d)
3
SS loading (g/m3d)
40
b
30 CODcr removal (g/m3 d)
3
1
1
TN removal (g/m d)
2
HRT=1d R =0.9943 2 HRT=2d R =0.9744 2 HRT=3d R =0.9767
4 SS removal (g/m3 d)
SS removal (g/m3 d)
e
2
HRT=1d R =0.8781 2 HRT=2d R =0.9721 2 HRT=3d R =0.9591
4
0
301
0.3
0.6
0.9 3
TP loading (g/m d)
1.2
1.5
0.0 0.0
0.3
0.6
0.9
1.2
1.5
TP loading (g/m3 d)
Fig. 5. Correlations between the influent loading and the removals of SS, CODcr , TN and TP of the two WTR-CWs at different HRTs: (a)–(d), the correlations of the CFCW; (e)–(h), the correlations of the TFCW.
302
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
0.08
0.08
0.07
Fe concentration (mg/L)
0.06
0.07 0.06
0.05
0.05
0.04
0.04
0.03
0.03
0.02
0.02
0.01
0.01
0.00
0.00 0
40
80
120
160
Al concentration (mg/L)
Eff Fe of CFCW Eff Fe of TFCW Eff Al of CFCW Eff Al of TFCW
200
Time (d)
below the general regulatory guideline limit of 0.30 mg/L (Fe) and 0.20 mg/L (Al) for the drinking water quality standard and the effluent discharge. The Al released was only 0.01% of the total Al content of the WTR. Wang and Pei (2013a) also suggested that WTR present a low Al-leaching risk in lake water. This low Al-leaching risk was due to Al speciation in the WTR, which is dominated by hydrolysis and organically complexed Al forms rather than free Al3+ (Ippolito et al., 2011). In addition, the release of other metals (As, Ba, Be, Ca, Cd, Co, Cr, Cu, Mg, Mn, Mo, Ni, Pb, Sr, V and Zn) was reported to be stable at a relatively low level at the common pH levels of secondary effluent (7.0–7.5) (Wang et al., 2013c; Elphick et al., 2005). Although the chemical extractability and bioaccessibility of many metals increased in the WTR after being leached, the leached WTR could still be considered non-hazardous. Based on these findings, the metal release from the WTR-CW was inferred to present no risk. 3.6. Implications for practical application
Fig. 6. Fe and Al leaching from the two WTR-CWs during the first 171 d.
HRT of 3 d than at a HRT of 1 d, and the strongest linear relationship (R2 = 0.63) was achieved at a HRT of 3 d (Fig. 5g). In contrast, the mass reduction efficiency of TP was above 94% at all HRTs, and there was no apparent difference among the correlations between the TP mass loading and the removal of the two WTR-CWs. The R2 values were all above 0.80, and the slope was nearly 1 (Fig. 5d and h). Overall, a longer HRT not only helped to achieve the maximum removal of pollutants but also maintained the stability of the treatment efficiency. Several studies have demonstrated the effect of the HRT in subsurface flow wetlands, revealing that higher HRTs improve the quality of the effluent. Although Toet et al. (2005) reported that HRT did not significantly influence the organic matter removal, Akratos and Tsihrintzis (2007) found that a HRT of more than 8 d was adequate for a relatively high removal of organic matter (92%). In this study, the highest HRT of 3 d resulted in the maximum observed removal of SS and CODcr while maintaining the stability of the treatment efficiency. This result was expected because a higher HRT implies a lower loading and more contact time, thus improving the stability in the efficiency. Although the increase of HRT from 1 d to 2 d led to a considerable increase in the average TN removal efficiency of the two WTR-CWs, there was no significant difference when the HRT increased from 2 d to 3 d (Table 2). A similar trend was observed by Ghosh and Gopal (2010), who reported that the concentration of pollutants decreased sharply with an increase in the HRT from 1 d to 3 d, but it only increased slightly more with a further increase in HRT to 4 d. Decreasing the HRT supported the development of aerobic conditions and hindered the denitrification processes. Nevertheless, the highest HRT with lower oxygen contents promotes denitrification and thus resulted in the highest (NO2 -N + NO3 -N) removal. Unlike the research of Akratos and Tsihrintzis (2007), who reported a maximum of 88% P removal at the highest HRT of 20 d, the P removal was the same at all HRTs in this study. Yang et al. (2006) demonstrated that a retention time of 48 h was sufficient for equilibrium of P adsorption by WTR. Therefore, even for the relatively short HRTs used in this study, the P removal was still significant. 3.5. Metal leaching Because the metal concentrations in the effluent were too low to be detected after 171 d, the Fe and Al leaching of the two WTR-CWs during the first 171 d are shown in Fig. 6. The effluent concentrations of both Fe and Al under different operation schemes were all
In this study, the two WTR-CWs with different operation schemes exhibited similar TN and TP removal characteristics from secondary wastewater at short HRTs (1–3 d). A development of anaerobic conditions favoured the TN removal, while DO had no effect on TP removal. The removal of pollutants can be further enhanced with an increase in the HRT, and the most optimal and stable removal was achieved at a HRT of 3 d. In addition, the WTR was found to present a low leaching risk for Fe and Al. Both of the WTR-CWs effluents met the Sewage Discharge Standard (SEPA, 2003). In general, the results of this work demonstrated that both continuous flow operation and tidal flow operation of CW can achieve satisfactory nutrient removal for NO3 -N-dominated secondary effluents. Normally, the continuous flow is easily maintained, and the cost of tidal flow is higher due to the requirement for more pumps. However, with regard to the control of the ammonification of organic nitrogen in the WTR, the N removal of the TFCW was more effective due to the low NH4 -N effluent. Moreover, the ammonification of organic nitrogen in WTR was not presented by Hu et al. (2012a,b), who used WTR-CW to treat agricultural wastewater. There are two reasons that may account for the difference between that study and this work. One reason is that the organic nitrogen content in the WTR was significantly lower than the NH4 -N in the agricultural wastewater of that study, so the ammonification was not apparent in that prior study. The other reason is that the WTR exhibited a wide range of characteristics according to the producing area of the respective WTR. Normally, the organic nitrogen came from raw drinking water; thus, different water sources and sample seasons may affect the nitrogen content of the WTR. Therefore, choosing WTR with low N content and conducting specific operation schemes, such as aeration or hybrid CW systems, can be used to eliminate the ammonification of organic nitrogen. 4. Conclusions This study evaluated the potential of two WTR-CWs with the same size but using different operation schemes for tertiary treatment of wastewater at short HRTs (1–3 d). The results indicated that both the CFCW and the TFCW could remove nutrients effectively, even though there was a slight increase in the NH4 -N concentration of the CFCW. The concentration of pollutants in both effluents met the discharge standard, and the quality of the effluents can be further enhanced with the increase of the HRT. At a HRT of 3 d, the highest removal efficiency of TN (>76%), TP (>98%), CODcr (>49%) and SS (>92%) were achieved. Moreover, the WTR exhibited a low
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
leaching risk for Fe and Al. Overall, the application of WTR as a substrate in CW to treat the secondary effluent is feasible. Acknowledgements This research was supported by the National Key Technology R&D Program (2012BAJ21B08) and the National Natural Science Foundation of China (51278055; 51179008). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.ecoleng.2014.06.015. References Akratos, C.S., Tsihrintzis, V.A., 2007. Effect of temperature, HRT, vegetation and porous media on removal efficiency of pilot-scale horizontal subsurface flow constructed wetlands. Ecol. Eng. 29, 173–191. Akunna, J.C., Bizeau, C., Moletta, R., 1992. Denitrification in anaerobic digesters: possibilities and influence of wastewater COD/N-NOX ratio. Environ. Technol. 13, 825–836. Albuquerque, A., Makinia, J., Pagilla, K., 2012. Impact of aeration conditions on the removal of low concentrations of nitrogen in a tertiary partially aerated biological filter. Ecol. Eng. 44, 44–52. Alburquerque, J.A., Gonzálvez, J., García, D., Cegarra, J., 2006. Effects of bulking agent on the composting of “alperujo”, the solid by-product of the two-phase centrifugation method for olive oil extraction. Process Biochem. 41, 127–132. Babatunde, A.O., Zhao, Y.Q., Zhao, X.H., 2010. Alum sludge-based constructed wetland system for enhanced removal of P and OM from wastewater: concept, design and performance analysis. Bioresour. Technol. 101, 6576–6579. Campos, J.L., Carvalho, S., Portela, R., Mosquera-Corral, A., Méndez, R., 2008. Kinetics of denitrification using sulphur compounds: effects of S/N ratio, endogenous and exogenous compounds. Bioresour. Technol. 99, 1293–1299. Clescerl, L.S., Greenberg, A.E., Eaton, A.D., 1995. Standard Methods for Examination of Water and Wastewater, 20th ed. APHA-AWWA-WEF, Washington, DC. Conley, D.J., Paerl, H.W., Howarth, R.W., Boesch, D.F., Seitzinger, S.P., Havens, K.E., Lancelot, C., Likens, G.E., 2009. Controlling eutrophication: nitrogen and phosphorus. Science 323, 1014–1015. Drizo, A., Comeau, Y., Forget, C., Chapuis, R.P., 2002. Phosphorus saturation potential: a parameter for estimating the longevity of constructed wetland systems. Environ. Sci. Technol. 36, 4642–4648. Elphick, J.R., Bailey, H.C., Hindle, A., Bertold, S.E., 2005. Aeration with carbon dioxidesupplemented air as a method to control pH drift in toxicity tests with effluents from wastewater treatment plants. Environ. Toxicol. Chem. 24, 2222–2225. Gale, P.M., Dévai, I., Reddy, K.R., Graetz, D.A., 1993. Denitrification potential of soils from constructed and natural wetlands. Ecol. Eng. 2, 119–130. Ganrot, Z., Dave, G., Nilsson, E., 2007. Recovery of N and P from human urine by freezing, struvite precipitation and adsorption to zeolite and active carbon. Bioresour. Technol. 98, 3112–3121. Ghosh, D., Gopal, B., 2010. Effect of hydraulic retention time on the treatment of secondary effluent in a subsurface flow constructed wetland. Ecol. Eng. 36, 1044–1051. Gray, S., Kinross, J., Read, P., Marland, A., 2000. The nutrient assimilative capacity of maerl as a substrate in constructed wetland systems for waste treatment. Water Res. 34, 2183–2190. Gupta, M.D., Loganathan, P., Vigneswaran, S., 2012. Adsorptive removal of nitrate and phosphate from water by a purolite ion exchange resin and hydrous ferric oxide columns in series. Sep. Sci. Technol. 47, 1785–1792. Hu, Y.S., Zhao, Y.Q., Zhao, X.H., Kumar, J.L.G., 2012a. Comprehensive analysis of stepfeeding strategy to enhance biological nitrogen removal in alum sludge-based tidal flow constructed wetlands. Bioresour. Technol. 111, 27–35.
303
Hu, Y.S., Zhao, Y.Q., Zhao, X.H., Kumar, J.L.G., 2012b. High rate nitrogen removal in an alum sludge-based intermittent aeration constructed wetland. Environ. Sci. Technol. 46, 4583–4590. Ippolito, J.A., Barbarick, K.A., Elliott, H.A., 2011. Drinking water treatment residuals: a review of recent uses. J. Environ. Qual. 40, 1–12. Lee, C.Y., Lee, C.C., Lee, F.Y., Tseng, S.K., Liao, C.J., 2004. Performance of subsurface flow constructed wetland taking pretreated swine effluent under heavy loads. Bioresour. Technol. 92, 173–179. Lin, Y.F., Jing, S.R., Wang, T.W., Lee, D.Y., 2002. Effects of macrophytes and external carbon sources on nitrate removal from groundwater in constructed wetlands. Environ. Pollut. 119, 413–420. Lin, Y.F., Jing, S.R., Lee, D.Y., Chang, Y.F., Shih, K.C., 2008. Nitrate removal from groundwater using constructed wetlands under various hydraulic loading rates. Bioresour. Technol. 99, 7504–7513. Maria, P., Richard, J.E., Mohd, A.A.H., Alex, D.B., Ian, S.M., Dionissios, M., 2004. Sonocatalytic oxidation processes for the removal of contaminants containing aromatic rings from aqueous effluents. Sep. Purif. Technol. 34, 35–42. Oleszkiewicz, J.A., Barnard, J.L., 2006. Nutrient removal technology in North America and the European Union: a review. Water Qual. Res. J. Can. 41, 14. Saeed, T., Sun, G.Z., 2011. A comparative study on the removal of nutrients and organic matter in wetland reactors employing organic media. Chem. Eng. J. 171, 439–447. State Environmental Protection Administration (SEPA), 2003. Discharge Standard of Pollutants for Municipal Wastewater Treatment Plant (GB 18918-2002). Standard Press of China, Beijing, China (in Chinese). Toet, S., Logtestijn, R.P., Kampf, R., Schreijer, M., Verhoeven, J.A., 2005. The effect of hydraulic retention time on the removal of pollutants from sewage treatment plant effluent in a surface-flow wetland system. Wetlands 25, 375–391. Vymazal, J., 2001. Types of constructed wetlands for wastewater treatment: their potential for nutrient removal. In: Vymazal, J. (Ed.), Transformations on Nutrients in Natural and Constructed Wetlands. Backhuys Publishers, Leiden, The Netherlands, pp. 1–93. Vymazal, J., 2007. Removal of nutrients in various types of constructed wetlands. Sci. Total Environ. 380, 48–65. Wang, C.H., Qi, Y., Pei, Y.S., 2012a. Laboratory investigation of phosphorus immobilization in lake sediments using water treatment residuals. Chem. Eng. J. 209, 379–385. Wang, C.H., Pei, Y.S., 2012b. The removal of hydrogen sulfide in solution by ferric and alum water treatment residuals. Chemosphere 88, 1178–1183. Wang, C.H., Pei, Y.S., 2013a. A comparison of the phosphorus immobilization capabilities of water treatment residuals before and after settling from lake water. Sep. Purif. Technol. 117, 83–88. Wang, C.H., Bai, L.L., Pei, Y.S., 2013b. Assessing the stability of phosphorus in lake sediments amended with water treatment residuals. J. Environ. Manage. 122, 31–36. Wang, C.H., Yuan, N.N., Pei, Y.S., 2013c. Effect of pH on metal lability in drinking water treatment residuals. J. Environ. Qual. 43, 389–397. Wang, X.L., Peng, Y.Z., Wang, S.Y., Fan, J., Cao, X.M., 2006. Influent of wastewater composition on nitrogen and phosphorus removal and process control in A2 /O process. Bioprocess Biosyst. Eng. 28, 397–404. Wang, Y.Q., Han, T.W., Xu, Z., Bao, G.Q., Zhu, T., 2005. Optimization of phosphorus removal from secondary effluent using simplex method in Tianjin, China. J. Hazard. Mater. 121, 183–186. Wang, Z.Y., Wang, C.H., Wang, Z.X., Pei, Y.S., 2013a. Enhancement of anaerobic ammonium oxidation in lake sediment by applying drinking water treatment residuals. Bioresour. Technol. 142, 745–749. Wendling, L.A., Douglas, G.B., Coleman, S., Yuan, Z., 2012. Nutrient and dissolved organic carbon removal from water using mining and metallurgical by-products. Water Res. 46, 2705–2717. Wu, S.B., Zhang, D.X., Austin, D., Dong, R.J., Pang, C.L., 2011. Evaluation of a lab-scale tidal flow constructed wetland performance: oxygen transfer capacity, organic matter and ammonium removal. Ecol. Eng. 37, 1789–1795. Xu, D.F., Xu, J.M., Wu, J.J., Muhammad, A., 2006. Studies on the phosphorus sorption capacity of substrates used in constructed wetland systems. Chemosphere 63, 344–352. Yang, Y., Zhao, Y.Q., Babatunde, A.O., Wang, L., Ren, Y.X., Han, Y., 2006. Characteristics and mechanisms of phosphate adsorption on dewatered alum sludge. Sep. Purif. Technol. 51, 193–200.
Contents lists available at ScienceDirect
Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng
Reuse of drinking water treatment residuals as a substrate in constructed wetlands for sewage tertiary treatment Leilei Bai a , Changhui Wang a , Caihong Huang b , Liansheng He b , Yuansheng Pei a,∗ a b
The Key Laboratory of Water and Sediment Sciences, Ministry of Education, School of Environment, Beijing Normal University, Beijing 100875, PR China Water Environment System Project Laboratory, Chinese Research Academy of Environmental Sciences, Beijing 100012, PR China
a r t i c l e
i n f o
Article history: Received 22 September 2013 Received in revised form 26 April 2014 Accepted 23 June 2014 Keywords: Constructed wetlands Drinking water treatment residuals Nitrogen Phosphorus Tertiary treatment
a b s t r a c t This work investigated the feasibility of reusing drinking water treatment residuals (WTR) as a substrate in constructed wetlands (WTR-CW) to treat secondary effluent at short hydraulic retention times (HRTs) (1–3 d). The results indicated that both continuous flow operation (CFCW) and tidal flow operation (TFCW) can achieve satisfactory removal of total nitrogen (TN), total phosphorus (TP), chemical oxygen demand and suspended solids although the ammonia nitrogen concentration of the CFCW effluent increased slightly. The WTR was found to be beneficial for denitrification, and the mean nitrate removal rates of CFCW and TFCW were 3.45 and 2.47 g N/m3 d, respectively. The TP removal efficiency of the two WTR-CWs still remained at 98% after 260 d of operation, and the lifetime regarding P saturation was estimated to be longer than 10 years. The HRT played a more significant role in TN removal, and the most optimal and stable TN removal (>76%) was obtained at 3 d HRT. Moreover, the leaching of Fe and Al from the two WTR-CWs was minor. Based on regulations, it is feasible to reuse the WTR as a substrate in constructed wetlands for the purpose of sewage tertiary treatment and waste recycling. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Nutrients, including nitrogen (N) and phosphorus (P), from municipal wastewater can lead to eutrophication of the receiving water bodies, including fresh-water and some seawater (Conley et al., 2009; Oleszkiewicz and Barnard, 2006). Normally, it is difficult to maintain the effluent of secondary biological treatment at a cleanliness level that satisfies the environmental standard to enable it to be discharged into the receiving water bodies (SEPA, 2003). Various processes, such as coagulation, ion exchange and adsorption, have been used as tertiary treatments to deal with the remaining contaminants after secondary treatment (Ganrot et al., 2007; Gupta et al., 2012; Wang et al., 2005). Among these tertiary treatment methods, constructed wetlands (CW) have
Abbreviations: CFCW, continuous flow operated WTR-CW; CODcr , chemical oxygen demand; CW, constructed wetlands; DO, dissolved oxygen; HRT, hydraulic retention time; ICP-AES, inductively coupled plasma atomic emission spectroscopy; N, nitrogen; NH4 -N, ammonia nitrogen; NO2 -N, nitrite; NO3 -N, nitrate; P, phosphorus; SS, suspend solids; TFCW, tidal-flow operated WTR-CW; TN, total N; TP, total P; WTR, drinking water treatment residuals; WTR-CW, CW based on WTR. ∗ Corresponding author. Tel.: +86 10 5880 1830; fax: +86 10 5880 1830. E-mail address: [email protected] (Y. Pei). http://dx.doi.org/10.1016/j.ecoleng.2014.06.015 0925-8574/© 2014 Elsevier B.V. All rights reserved.
demonstrated advantages in terms of aesthetics, lower energy consumption and more economical construction and operation. The performance of CW is generally good in terms of the removal of organic matter and suspended solids (SS). Effective N removal can also be achieved by different operation schemes, such as continuous feeding, intermittent feeding, step-feeding and artificial aeration (Albuquerque et al., 2012; Hu et al., 2012a; Vymazal, 2001). Traditionally, CW has a limited capacity for the removal of P unless materials with high P adsorption capacity are used (Vymazal, 2007). Therefore, many researchers have used some industry by-products and wastes as substrates for enhancing P removal (Drizo et al., 2002; Wendling et al., 2012). Recently, significant research effort has been focused on the reuse of drinking water treatment residuals (WTR) in CW (WTR-CW) to facilitate water reuse and waste recycling. Water treatment residuals, a non-hazardous inevitable byproduct generated from drinking water treatment plants, has been demonstrated to exhibit high P retention capabilities (Yang et al., 2006). Babatunde et al. (2010) demonstrated that a pilot fieldscale WTR-CW operated in tidal-flow mode exhibited satisfactory performance of the removal of P and organic matter. In addition, Hu et al. (2012a,b) constructed a four-stage WTR-CW and a single-stage WTR-CW with intermittent aeration to treat dairy wastewater, each of which exhibited a high rate of N removal.
296
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Table 1 Wastewater characteristics (mean ± SD) and average loading rates.
a
HRT (d) HLR (m3 /m3 d)a CODcr (mg/L) NH4 -N (mg/L) NO2 -N (mg/L) NO3 -N (mg/L) TN (mg/L) TP (mg/L) SS (mg/L) pH OLR (g CODcr /m3 d)a NLR (g N/m3 d)a PLR (g P/m3 d)a CODcr /TN a b c
Period 1 (15–100 d)
Period 2 (100–200 d)
Period 3 (200–260 d)
1 0.45 45.03 ± 20.23 0.65 ± 0.64 0.24 ± 0.17 16.70 ± 2.28 17.88 ± 2.54 1.88 ± 0.45 2.10 ± 0.24 8.29 ± 0.18 20.25 8.04 0.84 2.52
2 0.22 53.72 ± 18.87 0.28 ± 0.21 0.04 ± 0.02 15.80 ± 4.15 16.61 ± 4.15 1.83 ± 0.68 2.47 ± 0.44 8.36 ± 0.19 11.83 3.66 0.40 3.23
3 0.15 71.10 ± 45.00 0.69 ± 0.67 0.39 ± 0.29 11.10 ± 1.93 12.96 ± 2.74 1.15 ± 0.50 2.73 ± 0.31 8.18 ± 0.40 10.67 1.95 0.18 5.48
Dairy wastewaterb – – 3491 ± 368 ± – – 442 ± 39 ± 1739 ± 7.87 ± – – – –
2045 114
114 22 1687 0.3
Discharge standardc – – 50 5 – – 15 0.5 10 6–9 – – – –
HRT: hydraulic retention time; HLR: hydraulic loading rate; OLR: organic loading rate; NLR: nitrogen loading rate; PLR: phosphorus loading rate. Refer to Hu et al. (2012b). The first-class requirements of Discharge Standard of Pollutants for Municipal Wastewater Treatment Plant (GB 18918-2002) (SEPA, 2003).
Overall, the WTR-CW has been successful in the treatment of highstrength wastewater (high SS, chemical oxygen demand (CODcr ), total N (TN) and total P (TP)), demonstrating a high removal efficiency for pollutants. However, there have been limited data reported regarding the application of WTR-CW for sewage tertiary treatment. Unlike dairy wastewater, secondary effluent is normally characterised by low CODcr and SS levels and a relatively lower nutrient level. Therefore, there is a significant need to determine the feasibility of using WTR-CW to purify secondary effluents. In this study, two lab-scale single-stage WTR-CWs were constructed to treat nitrate (NO3 -N)-dominated secondary effluent from a municipal wastewater treatment plant. The performance of the two CW operation schemes, including continuous flow (CFCW) and tidal flow (TFCW), was monitored. The objectives of the study were (1) to investigate the effectiveness of WTR-CW under two different operation schemes, (2) to determine the mechanisms of nutrients removal, and (3) to estimate the applicability of WTRCW as a tertiary treatment technology. The results of this work provide support for the application of WTR-CW for sewage tertiary treatment.
2. Materials and methods 2.1. Sample preparation Dewatered WTR was collected from Beijing City No. 9 Waterworks in China. The particle sizes of the raw cake of WTR were in the range of 1–3 cm, and the moisture content was 55%. The characteristics of the WTR were available in former studies (Wang et al., 2012a, 2013a): Fe of 101.56 mg/g, Al of 50.36 mg/g, P of 0.61 mg/g, pH of 7.04, organic matter of 57.65 mg/g and surface area of 78.83 m2 /g. Furthermore, a maximum P adsorption capacity of 7.42 mg/g was calculated at pH of 7 using Langmuir isotherm (Fig. S1 in Supplementary Material). The secondary effluent was collected after an A2 /O process from the Xiaojiahe Sewage Treatment Plant of Beijing. The A2 /O process was a single-sludge suspended growth system incorporating anaerobic, anoxic and aerobic zones in sequence (Wang et al., 2006). After collection, the wastewater was stored at 4 ◦ C. The characteristics and loading rates during the entire waste treatment operation are summarised in Table 1. Compared with the dairy wastewater studied in previous work (Hu et al., 2012b), the secondary wastewater had significantly lower CODcr , SS, TN and TP levels. However, the secondary wastewater (CODcr , TN and TP) still failed to meet the requirements of the standard (SEPA, 2003). In
the secondary wastewater, the TN primarily existed in the form of NO3 -N. Overall, tertiary treatment was required to improve the water quality of the secondary wastewater. 2.2. System description and operation The two lab-scale single-stage WTR-CW systems were implemented with two Plexiglas columns (diameter of 9.3 cm and depth of 90 cm) (Fig. 1). Gravel (depth of 10 cm) was added at the bottom of the columns as the support medium, and a 60-cm depth of dewatered WTR (dry weight of 1.2 kg) was added as the main wetland medium layer, which resulted in a total volume of 5.09 L and an effective volume of 2.0 L (an initial porosity of 39%). Common reeds, Phragmitesaustralis, were planted on the top of the two columns. For the CFCW, the influent was continuously introduced into the column from the bottom by a peristaltic pump, while the effluent overflowed from the top. The desired hydraulic retention times (HRTs) were adjusted by controlling the flowrate of the influent. For the TFCW, the influent was introduced into the column from the top, while the effluent was drained from the bottom by peristaltic pumps. The feeding dosage and the discharge dosage were both half of the total volume per cycle, thus maintaining a stationary volume of 1.0 L during the experiments. The wet/dry (h) ratio was maintained at 22:2 by using pre-set programmable timers. The desired HRTs were adjusted by control of the flood and drain stages sequentially. The two WTR-CWs were operated for 260 d at room temperature, and this time was divided into three periods according to the theoretical HRTs (Table 1). Samples were collected from the influent and effluent of the two WTR-CWs 1–2 times per week and were analysed for CODcr , SS, ammonia nitrogen (NH4 -N), nitrite (NO2 -N), NO3 -N, TN, TP, and pH. In addition, the samples for Fe and Al analysis were collected from the effluent once each month. 2.3. Sample analysis The CODcr level was measured using the potassium dichromate method (Maria et al., 2004), and the SS level was measured using the gravimetric method (Clescerl et al., 1995). The NH4 -N, NO2 -N and NO3 -N levels were detected using a spectrophotometer (UV-2000, Unico) according to the standard methods (Clescerl et al., 1995). The TN level was measured by alkaline potassium persulphate oxidation followed by NO3 -N analysis (Clescerl et al., 1995). The TP was measured by persulphate digestion followed by orthophosphate analysis (Clescerl et al., 1995). The pH was measured using a pH metre (pH-10, Sartorius). Dissolved Fe and
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
297
Fig. 1. Two different operation schemes of the laboratory WTR-CW: (a) CFCW and (b) TFCW.
Al were analysed by inductively coupled plasma atomic emission spectroscopy (ICP-AES, ULTIMA, JY, France) (Clescerl et al., 1995). The analytical detection limits were 5 mg/L for CODcr , 0.5 mg/L for SS, 0.01 mg/L for NH4 -N, 0.015 mg/L for NO2 -N, 0.1 mg/L for NO3 -N, 0.1 mg/L for TN, 0.02 mg/L for TP, 0.01 for pH, 0.01 mg/L for Fe and 0.01 mg/L for Al. The presented tests were repeated twice, and the average values were reported. The standard error deviation was within 10%. 2.4. Statistical analysis To investigate the difference in treatment performance of the two WTR-CWs under different HRTs, statistical analyses were performed using PASW Statistics 20 (SPSS Inc.). One-way ANOVA multiple comparisons for the mean removal efficiencies were computed using Tamhane’s T2 test at a 95% confidence level (˛ = 0.05). The method was selected because equal variance between groups was not assumed. 3. Results and discussion 3.1. Variation of TN and TP Both the CFCW and the TFCW demonstrated a satisfactory performance in TN and TP removal during the entire operation period (Fig. 2). The mean effluent TN concentrations of the CFCW and the TFCW were 5.27 mg/L and 6.63 mg/L, respectively, resulting in an average TN removal rate of 2.95 g N/m3 d for the CFCW and 2.36 g N/m3 d for the TFCW. The most stable TN removal efficiency of the two WTR-CWs was obtained in period 3, with an observed efficiency of greater than 75%. Compared with the TN removal, the TP removal was more efficient and stable, regardless of the operation conditions. The TP concentrations of the two WTR-CW effluents were both nearly zero, which is significantly lower than the discharge standard (SEPA, 2003). After 260 d of operation, the TP removal efficiency still remained above 98% under a P loading
rate of 0.18 g P/m3 d. Overall, the favourable TN and TP removal under short HRTs (1–3 d) indicated that the WTR-CW was suitable for tertiary nutrient removal. The TN removal rates of the two WTR-CWs were comparable to those reported in a previous study by Lin et al. (2008). Although the N transfer was complex, biological function was the main N removal mechanism (Vymazal, 2007). Wang et al. (2012a) reported that WTR can increase the total abundance of bacteria due to its high surface area, thus providing an adequate surface for biofilm attachment. Furthermore, Wang et al. (2013a) found that WTR can promote aggregation and strengthen the activity of anammox bacteria in lake sediments. Therefore, the WTR was found to be beneficial to biological N removal. Adsorption was the main P removal pathway in the CW (Babatunde et al., 2010). The WTR had a large potential for P adsorption due to its large amount of Fe and Al. In addition, Wang et al. (2013b) demonstrated that organic matter, ion strength and anaerobic conditions did not affect the stability of the P adsorbed by WTR, suggesting that the retained P cannot be desorbed easily. In a former study, Xu et al. (2006) reported that the lifetime of a CW could be calculated using the maximum P sorption capacity of substrate and area of CW. In this study, the amounts of TP removal by the CFCW and the TFCW during the 260 d were 454.16 mg and 452.73 mg (calculated by the difference in between the influent and effluent P concentrations), respectively, resulting in a P adsorption quantity of 0.38 mg/g for the WTR, which is far from the maximum adsorption capacity of 7.42 mg/g. Consequently, considering the presence of other functions, such as physical trapping, filtration and sedimentation, the lifetime of the WTR-CW was estimated to be longer than ten years. There was no significant difference in the TN and TP removal efficiencies between CFCW and TFCW (P > 0.05) according to multiple comparisons (one-way ANOVA) that were performed to analyse the mean removal efficiencies. The main difference between the two operation schemes was oxygen transfer. For the CFCW, the oxygen was mainly derived from the dissolved oxygen (DO) in
298
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Influent and effluent TP (mg/L)
Influent and effluent TN (mg/L)
Period 1
Period 3 Inf Eff of CFCW Eff of TFCW
Period 2
a
24 20 16 12 8 4 0
0
40
80
120
160
200
240
280
80
120
160
200
240
280
4
b 3
2
1
0 0
40
Time (d) Fig. 2. The TN and TP removal of the two WTR-CWs during the entire operation period: (a) TN removal and (b) TP removal.
the influent and from surface diffusion, while both pathways were limited, and the system was always anoxic. In contrast, the internal environment of the TFCW changed from aerobic to anaerobic in one cycle (Wu et al., 2011). During the drain and rest stage, air quickly diffused into the substrate, and the system was aerobic. During the work stage, oxygen was consumed by biological reactions, causing the system to become anaerobic. In this study, the anaerobic stage was dominant according to the high wet/dry ratio of 22:2. Hence, both of the WTR-CWs had a long anaerobic period that favoured the elimination of NO3 -N, which was the dominant species of the TN; thus, the TN removal was similar and effective for the two WTR-CWs. The same trend was found by Lin et al. (2008) using free-water surface wetlands and subsurface-flow wetlands to treat highly nitrated groundwater. Moreover, Wang et al. (2013b) demonstrated that DO had no significant influence on the P adsorption of the WTR. Thus, the P removal was the similar under the different operation schemes; even the anoxic condition of the WTR-CW performed well. Overall, the removal of TN was more complex and exhibited more fluctuations than that of TP, which makes it necessary to determine the N transformation over the entire operation period. 3.2. Variation of N forms The N transformation of the CFCW and the TFCW are shown in Fig. 3. For the CFCW, the NH4 -N concentration of the effluent was greater than 6 mg/L at the beginning; it then decreased and finally became stable at 1.35 ± 1.15 mg/L. The effluent NH4 -N levels during the three periods were all significantly higher than the influent levels (P < 0.05). In contrast, there was a non-significant difference between the NH4 -N concentrations in the influent and the effluent of TFCW (P > 0.05). The stable NH4 -N concentration of the effluent
was 0.75 ± 0.75 mg/L. Moreover, the denitrification rate was calculated by the difference between the influent (NO2 -N + NO3 -N) and effluent levels. For the CFCW, the NO2 -N in the effluent was nearly zero, and the NO3 -N level was in the range of 0–6.78 mg/L, resulting in an average removal efficiency of (NO2 -N + NO3 -N) of 80%. For the TFCW effluent, the (NO2 -N + NO3 -N) concentration was significantly higher than that of the CFCW by 2.53 mg/L (P < 0.05), and the average removal efficiency was 65%. Comparatively, the CFCW had a higher (NO2 -N + NO3 -N) loss than did the TFCW, while the effluent NH4 -N increased slightly. The new production of NH4 -N in the CFCW may result from the ammonification of organic nitrogen in the secondary effluent and the WTR. However, the latter was dominant because the influent organic nitrogen was minimal (<0.50 mg/L), which was also confirmed by the drastic decrease in NH4 -N during the first 100 d when the system was under a steady-state condition. Gray et al. (2000) found a similar trend when using marl as a substrate to treat wastewater with a low NH4 -N influent. In addition, the lower NH4 N of the TFCW effluent suggested that the better oxygen supply promoted the conversion of NH4 -N into NO2 -N and NO3 -N. Overall, note that although the ammonification was clearly weaker than the TN removal, pre-treatment or aeration operation schemes, including artificial aeration, tidal-flow and hybrid systems, were required to eliminate the increase in NH4 -N. In addition to the ammonification-nitrification of the organic nitrogen in the WTR, the weaker denitrification was another reason for the higher (NO2 -N + NO3 -N) level of the TFCW effluent. According to the low ratio of the influent CODcr and N (Table 1), denitrification was the primary mechanism for NO3 N removal (Akunna et al., 1992). Thus the mean denitrification rates of the CFCW and the TFCW were determined to be 3.45 g N/m3 d and 2.47 g N/m3 d, respectively, over the entire period. The
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Inf NH4-N Eff NH4-N Inf NO2-N+NO3-N Eff NO2-N+NO3-N
15 12
28 24 20 16
9
12
6
8
3
4
0 0
40
80
120
160
200
240
0 280
NO2-N+NO3-N concentration (mg/L)
a
18
NH4-N concentration (mg/L)
Period 3
Period 2
Period 1
299
NH4-N concentration (mg/L)
Inf NH4-N Eff NH4-N Inf NO2-N+NO3-N Eff NO2-N+NO3-N
15 12
24 20 16
9
12
6
8
3
4
0 0
40
80
120
160
200
240
0 280
NO2-N+NO3-N concentration (mg/L)
28
b
18
Time (d) Fig. 3. The transformation of the N forms of the two WTR-CWs during the entire period: (a) CFCW and (b) TFCW.
denitrification was controlled by a variety of factors, including the characteristics of wetland substrate, the types and species of macrophyte, microbial communities, oxygen supply, and the NO3 N and organic matter concentrations in the influent (Gale et al., 1993; Lin et al., 2002). Since the average oxygen supply of the TFCW was generally much higher than that of the conventional CW, the denitrification was hindered and the denitrification rate was lower than that of CFCW (Wu et al., 2011). Furthermore, the lack of external organic carbon arising from the secondary effluent did not seem to have a significant influence on the denitrification rate. This may be due to the organic matter of WTR was higher than that of common soil or gravel and could provide carbon source for denitrification. Saeed and Sun (2011) reported that denitrifiers could use not only organic carbon from the influent but also the organic matter of the substrates. In addition, although Campos et al. (2008) indicated the denitrification with sulphur compounds as electron donors is an alternative to heterotrophic denitrification for wastewaters with high nitrate concentration and low organic matter content, in this study, the contribution of sulphide oxidation was minor because the sulfide in the WTR was limited and stable (Wang and Pei, 2012b). Overall, the WTR-CW was appropriate for denitrification, even though the influent organic matter level was low. 3.3. Variation of SS, CODcr and pH The performance of SS, CODcr and pH of the two WTR-CWs is shown in Fig. 4. The two WTR-CWs exhibited an average SS removal rate of 90% under all the tested HRTs (Fig. 4a). Even the influent SS levels were significantly higher during period 3 than during the early days, the effluent was still low. With regards to the organic matter, the average CODcr removals of the CFCW and the TFCW
were 54% and 47%, respectively, with effluent concentrations of less than 50 mg/L (Fig. 4b). The effluent pH of the CFCW was lower than the influent pH by an average of 0.37, and the corresponding pH reduction of the TFCW was 0.76 (Fig. 4c). Overall, the WTR-CW exhibited excellent removal capacities of SS and CODcr , and the change of pH was limited. The SS was removed entirely by physical interception by the substrate, which can be enhanced by the new generation of biofilms. In contrast, the CODcr removal from the CW system was accomplished by cooperation between physical and microbial processes (Lee et al., 2004). The physical process separated the organic solids, thus allowing better hydrolysis for biodegradation, while the substrate allowed the accumulation of immense amounts of attached bacteria, which helps in rapidly catalysing the chemical reactions. Thus, the WTR exhibited good physical characteristics for SS and CODcr removal. Normally, P adsorption of WTR and denitrification lead to an increase of pH (Vymazal, 2007; Yang et al., 2006). The pH could decrease in the presence of organic acid release from the substrate and nitrification (Alburquerque et al., 2006; Vymazal, 2007). In this study, the small decrease of pH as the wastewater flowed through the two WTR-CWs suggested that the organic acid release of the substrate was dominant. In addition, the stronger nitrification of the TFCW resulted in a lower effluent pH than CFCW. However, the variation of pH was in the normal range of the discharge standard. 3.4. The influence of HRT The HRT has a slight influence on the SS removal in the CFCW, according to the average removal efficiencies in the range of 79–92% at the three HRTs (Table 2). This result was also confirmed by the similar slope and R2 of the correlations between the SS mass
300
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
Period 2
Influent and effluent SS (mg/L)
Period 1
a
18
Inf Eff of CFCW Eff of TFCW
15 12 9 6 3 0
Influent and effluent COD (mg/L)
0 160
40
80
120
160
200
240
280
40
80
120
160
200
240
280
40
80
120
160
b
140 120 100 80 60 40 20 0 0
10.0
c
9.5
Influent and effluent pH
Period 3
9.0 8.5 8.0 7.5 7.0 6.5 6.0
0
200
240
280
Time (d) Fig. 4. The performance of the removal of SS and CODcr and the pH level of the two WTR-CWs over the operation period: (a) SS, (b) CODcr , and (c) pH.
loading and removal (Fig. 5a). The same trend was found in the TFCW, with an average removal efficiency in the range of 88–93% (Fig. 5e). Although the CODcr removal efficiencies were similar at the three HRTs, the mass removal of the CFCW and the TFCW both exhibited the strongest relationship to mass loading, with an R2 greater than 0.73 at 3 d HRT (Fig. 5b and f). The HRT played a more significant role in the removal of TN. For CFCW, the TN mass
reduction efficiency increased from 54% to 77%, with the increase of HRT occurring for a value of HRT from 1 d to 2 d. However, there was no significant difference between the TN reduction efficiency at a HRT of 2 d and that at a HRT of 3 d. The correlation between TN mass loading rates and removal rates followed the pattern similar to that of the TN reduction efficiency (Fig. 5c). For the TFCW, the TN mass reduction efficiency was approximately two times more at a
Table 2 The average pollutants removal rate (%) of the two WTR-CWs for different HRTs. Parameters
SS CODcr TN TP
CFCW
TFCW
Period 1
Period 2
Period 3
Period 1
Period 2
Period 3
79 54 54 97
87 52 77 98
92 56 77 98
88 45 34 95
89 48 70 96
93 49 79 98
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
5
5
a
3
2
2
0
0
40
1
2 3 3 SS loading (g/m d)
4
5
0
1
2
f
2
HRT=1d R =0.6753 2 HRT=2d R =0.4443 2 HRT=3d R =0.8614
5
2
3
CODcr removal (g/m d)
20
10
20
10
0 0
10
20 30 3 CODcr loading (g/m d)
0
40
10
20 30 3 CODcr loading (g/m d)
40
8
8
c
g
2
HRT=1d R =0.5975 2 HRT=2d R =0.8571 2 HRT=3d R =0.7461
2
HRT=1d R =0.2300 2 HRT=2d R =0.5904 2 HRT=3d R =0.6267
6
3
3
TN removal (g/m d)
6
4
2
0
0
2
4
6
8
4
2
0
10
0
2
3
6
8
10
TN loading (g/m d)
1.5
1.5
d
2
h
HRT=1d R =0.8882 2 HRT=2d R =0.9803 2 HRT=3d R =0.9862
HRT=1d R2=0.8011 HRT=2d R2=0.9898 HRT=3d R2=0.9585
1.2 TP removal (g/m d)
1.2
3
0.9
0.6
0.9
0.6
0.3
0.3
0.0 0.0
4
3
TN loading (g/m d)
3
4
HRT=1d R =0.7291 2 HRT=2d R =0.4670 2 HRT=3d R =0.7263
30
0
TP removal (g/m d)
3
SS loading (g/m3d)
40
b
30 CODcr removal (g/m3 d)
3
1
1
TN removal (g/m d)
2
HRT=1d R =0.9943 2 HRT=2d R =0.9744 2 HRT=3d R =0.9767
4 SS removal (g/m3 d)
SS removal (g/m3 d)
e
2
HRT=1d R =0.8781 2 HRT=2d R =0.9721 2 HRT=3d R =0.9591
4
0
301
0.3
0.6
0.9 3
TP loading (g/m d)
1.2
1.5
0.0 0.0
0.3
0.6
0.9
1.2
1.5
TP loading (g/m3 d)
Fig. 5. Correlations between the influent loading and the removals of SS, CODcr , TN and TP of the two WTR-CWs at different HRTs: (a)–(d), the correlations of the CFCW; (e)–(h), the correlations of the TFCW.
302
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
0.08
0.08
0.07
Fe concentration (mg/L)
0.06
0.07 0.06
0.05
0.05
0.04
0.04
0.03
0.03
0.02
0.02
0.01
0.01
0.00
0.00 0
40
80
120
160
Al concentration (mg/L)
Eff Fe of CFCW Eff Fe of TFCW Eff Al of CFCW Eff Al of TFCW
200
Time (d)
below the general regulatory guideline limit of 0.30 mg/L (Fe) and 0.20 mg/L (Al) for the drinking water quality standard and the effluent discharge. The Al released was only 0.01% of the total Al content of the WTR. Wang and Pei (2013a) also suggested that WTR present a low Al-leaching risk in lake water. This low Al-leaching risk was due to Al speciation in the WTR, which is dominated by hydrolysis and organically complexed Al forms rather than free Al3+ (Ippolito et al., 2011). In addition, the release of other metals (As, Ba, Be, Ca, Cd, Co, Cr, Cu, Mg, Mn, Mo, Ni, Pb, Sr, V and Zn) was reported to be stable at a relatively low level at the common pH levels of secondary effluent (7.0–7.5) (Wang et al., 2013c; Elphick et al., 2005). Although the chemical extractability and bioaccessibility of many metals increased in the WTR after being leached, the leached WTR could still be considered non-hazardous. Based on these findings, the metal release from the WTR-CW was inferred to present no risk. 3.6. Implications for practical application
Fig. 6. Fe and Al leaching from the two WTR-CWs during the first 171 d.
HRT of 3 d than at a HRT of 1 d, and the strongest linear relationship (R2 = 0.63) was achieved at a HRT of 3 d (Fig. 5g). In contrast, the mass reduction efficiency of TP was above 94% at all HRTs, and there was no apparent difference among the correlations between the TP mass loading and the removal of the two WTR-CWs. The R2 values were all above 0.80, and the slope was nearly 1 (Fig. 5d and h). Overall, a longer HRT not only helped to achieve the maximum removal of pollutants but also maintained the stability of the treatment efficiency. Several studies have demonstrated the effect of the HRT in subsurface flow wetlands, revealing that higher HRTs improve the quality of the effluent. Although Toet et al. (2005) reported that HRT did not significantly influence the organic matter removal, Akratos and Tsihrintzis (2007) found that a HRT of more than 8 d was adequate for a relatively high removal of organic matter (92%). In this study, the highest HRT of 3 d resulted in the maximum observed removal of SS and CODcr while maintaining the stability of the treatment efficiency. This result was expected because a higher HRT implies a lower loading and more contact time, thus improving the stability in the efficiency. Although the increase of HRT from 1 d to 2 d led to a considerable increase in the average TN removal efficiency of the two WTR-CWs, there was no significant difference when the HRT increased from 2 d to 3 d (Table 2). A similar trend was observed by Ghosh and Gopal (2010), who reported that the concentration of pollutants decreased sharply with an increase in the HRT from 1 d to 3 d, but it only increased slightly more with a further increase in HRT to 4 d. Decreasing the HRT supported the development of aerobic conditions and hindered the denitrification processes. Nevertheless, the highest HRT with lower oxygen contents promotes denitrification and thus resulted in the highest (NO2 -N + NO3 -N) removal. Unlike the research of Akratos and Tsihrintzis (2007), who reported a maximum of 88% P removal at the highest HRT of 20 d, the P removal was the same at all HRTs in this study. Yang et al. (2006) demonstrated that a retention time of 48 h was sufficient for equilibrium of P adsorption by WTR. Therefore, even for the relatively short HRTs used in this study, the P removal was still significant. 3.5. Metal leaching Because the metal concentrations in the effluent were too low to be detected after 171 d, the Fe and Al leaching of the two WTR-CWs during the first 171 d are shown in Fig. 6. The effluent concentrations of both Fe and Al under different operation schemes were all
In this study, the two WTR-CWs with different operation schemes exhibited similar TN and TP removal characteristics from secondary wastewater at short HRTs (1–3 d). A development of anaerobic conditions favoured the TN removal, while DO had no effect on TP removal. The removal of pollutants can be further enhanced with an increase in the HRT, and the most optimal and stable removal was achieved at a HRT of 3 d. In addition, the WTR was found to present a low leaching risk for Fe and Al. Both of the WTR-CWs effluents met the Sewage Discharge Standard (SEPA, 2003). In general, the results of this work demonstrated that both continuous flow operation and tidal flow operation of CW can achieve satisfactory nutrient removal for NO3 -N-dominated secondary effluents. Normally, the continuous flow is easily maintained, and the cost of tidal flow is higher due to the requirement for more pumps. However, with regard to the control of the ammonification of organic nitrogen in the WTR, the N removal of the TFCW was more effective due to the low NH4 -N effluent. Moreover, the ammonification of organic nitrogen in WTR was not presented by Hu et al. (2012a,b), who used WTR-CW to treat agricultural wastewater. There are two reasons that may account for the difference between that study and this work. One reason is that the organic nitrogen content in the WTR was significantly lower than the NH4 -N in the agricultural wastewater of that study, so the ammonification was not apparent in that prior study. The other reason is that the WTR exhibited a wide range of characteristics according to the producing area of the respective WTR. Normally, the organic nitrogen came from raw drinking water; thus, different water sources and sample seasons may affect the nitrogen content of the WTR. Therefore, choosing WTR with low N content and conducting specific operation schemes, such as aeration or hybrid CW systems, can be used to eliminate the ammonification of organic nitrogen. 4. Conclusions This study evaluated the potential of two WTR-CWs with the same size but using different operation schemes for tertiary treatment of wastewater at short HRTs (1–3 d). The results indicated that both the CFCW and the TFCW could remove nutrients effectively, even though there was a slight increase in the NH4 -N concentration of the CFCW. The concentration of pollutants in both effluents met the discharge standard, and the quality of the effluents can be further enhanced with the increase of the HRT. At a HRT of 3 d, the highest removal efficiency of TN (>76%), TP (>98%), CODcr (>49%) and SS (>92%) were achieved. Moreover, the WTR exhibited a low
L. Bai et al. / Ecological Engineering 70 (2014) 295–303
leaching risk for Fe and Al. Overall, the application of WTR as a substrate in CW to treat the secondary effluent is feasible. Acknowledgements This research was supported by the National Key Technology R&D Program (2012BAJ21B08) and the National Natural Science Foundation of China (51278055; 51179008). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.ecoleng.2014.06.015. References Akratos, C.S., Tsihrintzis, V.A., 2007. Effect of temperature, HRT, vegetation and porous media on removal efficiency of pilot-scale horizontal subsurface flow constructed wetlands. Ecol. Eng. 29, 173–191. Akunna, J.C., Bizeau, C., Moletta, R., 1992. Denitrification in anaerobic digesters: possibilities and influence of wastewater COD/N-NOX ratio. Environ. Technol. 13, 825–836. Albuquerque, A., Makinia, J., Pagilla, K., 2012. Impact of aeration conditions on the removal of low concentrations of nitrogen in a tertiary partially aerated biological filter. Ecol. Eng. 44, 44–52. Alburquerque, J.A., Gonzálvez, J., García, D., Cegarra, J., 2006. Effects of bulking agent on the composting of “alperujo”, the solid by-product of the two-phase centrifugation method for olive oil extraction. Process Biochem. 41, 127–132. Babatunde, A.O., Zhao, Y.Q., Zhao, X.H., 2010. Alum sludge-based constructed wetland system for enhanced removal of P and OM from wastewater: concept, design and performance analysis. Bioresour. Technol. 101, 6576–6579. Campos, J.L., Carvalho, S., Portela, R., Mosquera-Corral, A., Méndez, R., 2008. Kinetics of denitrification using sulphur compounds: effects of S/N ratio, endogenous and exogenous compounds. Bioresour. Technol. 99, 1293–1299. Clescerl, L.S., Greenberg, A.E., Eaton, A.D., 1995. Standard Methods for Examination of Water and Wastewater, 20th ed. APHA-AWWA-WEF, Washington, DC. Conley, D.J., Paerl, H.W., Howarth, R.W., Boesch, D.F., Seitzinger, S.P., Havens, K.E., Lancelot, C., Likens, G.E., 2009. Controlling eutrophication: nitrogen and phosphorus. Science 323, 1014–1015. Drizo, A., Comeau, Y., Forget, C., Chapuis, R.P., 2002. Phosphorus saturation potential: a parameter for estimating the longevity of constructed wetland systems. Environ. Sci. Technol. 36, 4642–4648. Elphick, J.R., Bailey, H.C., Hindle, A., Bertold, S.E., 2005. Aeration with carbon dioxidesupplemented air as a method to control pH drift in toxicity tests with effluents from wastewater treatment plants. Environ. Toxicol. Chem. 24, 2222–2225. Gale, P.M., Dévai, I., Reddy, K.R., Graetz, D.A., 1993. Denitrification potential of soils from constructed and natural wetlands. Ecol. Eng. 2, 119–130. Ganrot, Z., Dave, G., Nilsson, E., 2007. Recovery of N and P from human urine by freezing, struvite precipitation and adsorption to zeolite and active carbon. Bioresour. Technol. 98, 3112–3121. Ghosh, D., Gopal, B., 2010. Effect of hydraulic retention time on the treatment of secondary effluent in a subsurface flow constructed wetland. Ecol. Eng. 36, 1044–1051. Gray, S., Kinross, J., Read, P., Marland, A., 2000. The nutrient assimilative capacity of maerl as a substrate in constructed wetland systems for waste treatment. Water Res. 34, 2183–2190. Gupta, M.D., Loganathan, P., Vigneswaran, S., 2012. Adsorptive removal of nitrate and phosphate from water by a purolite ion exchange resin and hydrous ferric oxide columns in series. Sep. Sci. Technol. 47, 1785–1792. Hu, Y.S., Zhao, Y.Q., Zhao, X.H., Kumar, J.L.G., 2012a. Comprehensive analysis of stepfeeding strategy to enhance biological nitrogen removal in alum sludge-based tidal flow constructed wetlands. Bioresour. Technol. 111, 27–35.
303
Hu, Y.S., Zhao, Y.Q., Zhao, X.H., Kumar, J.L.G., 2012b. High rate nitrogen removal in an alum sludge-based intermittent aeration constructed wetland. Environ. Sci. Technol. 46, 4583–4590. Ippolito, J.A., Barbarick, K.A., Elliott, H.A., 2011. Drinking water treatment residuals: a review of recent uses. J. Environ. Qual. 40, 1–12. Lee, C.Y., Lee, C.C., Lee, F.Y., Tseng, S.K., Liao, C.J., 2004. Performance of subsurface flow constructed wetland taking pretreated swine effluent under heavy loads. Bioresour. Technol. 92, 173–179. Lin, Y.F., Jing, S.R., Wang, T.W., Lee, D.Y., 2002. Effects of macrophytes and external carbon sources on nitrate removal from groundwater in constructed wetlands. Environ. Pollut. 119, 413–420. Lin, Y.F., Jing, S.R., Lee, D.Y., Chang, Y.F., Shih, K.C., 2008. Nitrate removal from groundwater using constructed wetlands under various hydraulic loading rates. Bioresour. Technol. 99, 7504–7513. Maria, P., Richard, J.E., Mohd, A.A.H., Alex, D.B., Ian, S.M., Dionissios, M., 2004. Sonocatalytic oxidation processes for the removal of contaminants containing aromatic rings from aqueous effluents. Sep. Purif. Technol. 34, 35–42. Oleszkiewicz, J.A., Barnard, J.L., 2006. Nutrient removal technology in North America and the European Union: a review. Water Qual. Res. J. Can. 41, 14. Saeed, T., Sun, G.Z., 2011. A comparative study on the removal of nutrients and organic matter in wetland reactors employing organic media. Chem. Eng. J. 171, 439–447. State Environmental Protection Administration (SEPA), 2003. Discharge Standard of Pollutants for Municipal Wastewater Treatment Plant (GB 18918-2002). Standard Press of China, Beijing, China (in Chinese). Toet, S., Logtestijn, R.P., Kampf, R., Schreijer, M., Verhoeven, J.A., 2005. The effect of hydraulic retention time on the removal of pollutants from sewage treatment plant effluent in a surface-flow wetland system. Wetlands 25, 375–391. Vymazal, J., 2001. Types of constructed wetlands for wastewater treatment: their potential for nutrient removal. In: Vymazal, J. (Ed.), Transformations on Nutrients in Natural and Constructed Wetlands. Backhuys Publishers, Leiden, The Netherlands, pp. 1–93. Vymazal, J., 2007. Removal of nutrients in various types of constructed wetlands. Sci. Total Environ. 380, 48–65. Wang, C.H., Qi, Y., Pei, Y.S., 2012a. Laboratory investigation of phosphorus immobilization in lake sediments using water treatment residuals. Chem. Eng. J. 209, 379–385. Wang, C.H., Pei, Y.S., 2012b. The removal of hydrogen sulfide in solution by ferric and alum water treatment residuals. Chemosphere 88, 1178–1183. Wang, C.H., Pei, Y.S., 2013a. A comparison of the phosphorus immobilization capabilities of water treatment residuals before and after settling from lake water. Sep. Purif. Technol. 117, 83–88. Wang, C.H., Bai, L.L., Pei, Y.S., 2013b. Assessing the stability of phosphorus in lake sediments amended with water treatment residuals. J. Environ. Manage. 122, 31–36. Wang, C.H., Yuan, N.N., Pei, Y.S., 2013c. Effect of pH on metal lability in drinking water treatment residuals. J. Environ. Qual. 43, 389–397. Wang, X.L., Peng, Y.Z., Wang, S.Y., Fan, J., Cao, X.M., 2006. Influent of wastewater composition on nitrogen and phosphorus removal and process control in A2 /O process. Bioprocess Biosyst. Eng. 28, 397–404. Wang, Y.Q., Han, T.W., Xu, Z., Bao, G.Q., Zhu, T., 2005. Optimization of phosphorus removal from secondary effluent using simplex method in Tianjin, China. J. Hazard. Mater. 121, 183–186. Wang, Z.Y., Wang, C.H., Wang, Z.X., Pei, Y.S., 2013a. Enhancement of anaerobic ammonium oxidation in lake sediment by applying drinking water treatment residuals. Bioresour. Technol. 142, 745–749. Wendling, L.A., Douglas, G.B., Coleman, S., Yuan, Z., 2012. Nutrient and dissolved organic carbon removal from water using mining and metallurgical by-products. Water Res. 46, 2705–2717. Wu, S.B., Zhang, D.X., Austin, D., Dong, R.J., Pang, C.L., 2011. Evaluation of a lab-scale tidal flow constructed wetland performance: oxygen transfer capacity, organic matter and ammonium removal. Ecol. Eng. 37, 1789–1795. Xu, D.F., Xu, J.M., Wu, J.J., Muhammad, A., 2006. Studies on the phosphorus sorption capacity of substrates used in constructed wetland systems. Chemosphere 63, 344–352. Yang, Y., Zhao, Y.Q., Babatunde, A.O., Wang, L., Ren, Y.X., Han, Y., 2006. Characteristics and mechanisms of phosphate adsorption on dewatered alum sludge. Sep. Purif. Technol. 51, 193–200.