Journal of Environmental Chemical Engineering

Performance of a combined process of reflux deodorization and tidal flow constructed

wetland for sewage treatment
--Manuscript Draft--

Manuscript Number: JECE-D-21-06927

Article Type: Research Paper

Keywords: Rural sewage, Constructed wetland, Tidal flown, Denitrification and deodorization,
Reflux ratio

Abstract: A combined process of reflux deodorization and tidal flow constructed wetland was

developed, and its removal efficiencies of chemical oxygen demand (COD), ammonia
nitrogen (NH4+-N), total phosphorus (TP), sulfide (S2-), and odor (threshold odor
number, TON) were evaluated. The effect of the reflux ratio (R) on the performance of
the combined process was also assessed. The removal efficiencies of COD, NH4+-N,
S2-, and TON increased as R increased. When R=150%, the average removal
rates of COD, NH4+-N, S2-, and TON were 76.6%, 75.0%, 100%, and 88.0%,
respectively. ρ(NO3--N)/ρ(S2-) and ρ(COD)/ρ(NO3--N) had an important effect on the
denitrification and deodorization of the deodorizing pool (DP). When the influent
ρ(NO3--N)/ρ(S2-) was 0.8, the effluent of the DP had the following properties: TON<37,
S2-<1.5 mg/L, and NO3--N<4.6 mg/L. When the influent ρ(COD)/ρ(NO3--N) was 4, the
effluent of the DP had the following properties: TON<30, S2-<1.7 mg/L, and NO3--
N<5.9 mg/L. Thus, denitrification and deodorization were enhanced under both influent
conditions. The microbial community and functional groups of microbes of the system
were analyzed. The main taxa involved in desulfurization in the DP were Thiobacillus,
Thiocapsa, Desulfomicrobium, Limnobacter, and Sulfuritalea. Thiobacillus played a
major role in denitrifying sulfide removal, which explained the mechanism of
deodorization of the DP.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Graphical Abstract
Highlights (for review)

Highlights:

 A combined process was developed to simultaneously remove pollutants and

odors from effluent

 At a R of 150%, the combined process has good removal effect on pollutants and

odors.

 N/S and C/N are important factors affecting denitrifying sulfide removal

 Thiobacillus in the deodorizing pool was involved in denitrifying sulfide removal

Manuscript Click here to access/download;Manuscript;Manuscript.docx

1 1 Performance of a combined process of reflux deodorization

2
3
4 2 and tidal flow constructed wetland for sewage treatment
5
6 3 Hao Zhenga, b, Lei Jianga, b, Qian Zhanga, b, Qiushi,Shena, b, Yuanxiang Maoa, b,
7
8
9 4 Fangying Jia, b*
10
11
12 5
13
14
15 6 a Key Laboratory of Three Gorges Reservoir Region's Eco-Environment, Ministry of
16
17
18 7 Education, Chongqing University, Chongqing 400045, China
19
20
21 8 b College of Environment and Ecology, Chongqing University, Chongqing 400045,
22
23
24 9 China
25
26
27
28 10
29
30
31 11 *Corresponding author:
32
33
34
35
12 Fangying Ji, E-mail Address: jfy@cqu.edu.cn
36
37
38 13 Address: College of Environment and Ecology, Chongqing University, No 174
39
40
41 14 Shazheng Street, Chongqing 400045, China.
42
43 15
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 16 Abstract
2
3
4
5 17 A combined process of reflux deodorization and tidal flow constructed wetland
6
7 18 was developed, and its removal efficiencies of chemical oxygen demand (COD),
8
9
10 19 ammonia nitrogen (NH4+-N), total phosphorus (TP), sulfide (S2-), and odor (threshold
11
12
13 20 odor number, TON) were evaluated. The effect of the reflux ratio (R) on the
14
15
16 21 performance of the combined process was also assessed. The removal efficiencies of
17
18 22 COD, NH4+-N, S2-, and TON increased as R increased. When R=150%, the average
19
20
21 23 removal rates of COD, NH4+-N, S2-, and TON were 76.6%, 75.0%, 100%, and 88.0%,
22
23
24 24 respectively. ρ(NO3--N)/ρ(S2-) and ρ(COD)/ρ(NO3--N) had an important effect on the
25
26
27 25 denitrification and deodorization of the deodorizing pool (DP). When the influent
28
29
26 ρ(NO3--N)/ρ(S2-) was 0.8, the effluent of the DP had the following properties:
30
31
32 27 TON<37, S2-<1.5 mg/L, and NO3--N<4.6 mg/L. When the influent ρ(COD)/ρ(NO3--N)
33
34
35 28 was 4, the effluent of the DP had the following properties: TON<30, S2-<1.7 mg/L,
36
37
38 29 and NO3--N<5.9 mg/L. Thus, denitrification and deodorization were enhanced under
39
40
41
30 both influent conditions. The microbial community and functional groups of microbes
42
43 31 of the system were analyzed. The main taxa involved in desulfurization in the DP
44
45
46 32 were Thiobacillus, Thiocapsa, Desulfomicrobium, Limnobacter, and Sulfuritalea.
47
48
49 33 Thiobacillus played a major role in denitrifying sulfide removal, which explained the
50
51
52
34 mechanism of deodorization of the DP.
53
54 35 Keywords: Rural sewage, Constructed wetland, Tidal flown, Denitrification and
55
56
57 36 deodorization, Reflux ratio
58
59
60
61
62
63
64
65
 1 37 1. Introduction
2
3
4
5 38 The pollution of China's rural water resources has become increasingly severe in
6
7 39 recent years, and this has been accompanied by a decrease in the quality of these
8
9
10 40 water resources. Indeed, ammonia nitrogen (NH4+-N), chemical oxygen demand
11
12
13 41 (COD), and total phosphorus (TP) indicators in most rural water sources have been
14
15
16 42 shown to exceed water quality standards (Liu et al., 2011). Rural sewage is mainly
17
18 43 divided into rural domestic sewage, agricultural planting sewage, livestock sewage,
19
20
21 44 and poultry breeding sewage (Deng et al., 2016). To improve the rural living
22
23
24 45 environment, the Chinese government has advocated for the construction and
25
26
27 46 renovation of hygienic toilets and septic tanks in rural areas. However, the residence
28
29
47 time of the effluent in the septic tanks in rural areas is generally greater than 60 d; the
30
31
32 48 effluent COD, NH4+-N, and TP concentrations are high; and the odor is strong. As
33
34
35 49 many farmers have no use for the manure, the treatment of septic tank effluent in rural
36
37
38 50 areas has become a major challenge.
39
40
41
51 Ecological treatment technology is often used for sewage treatment in rural areas
42
43 52 (Yi et al., 2020). Constructed wetlands (CWs) (Ma et al., 2021) are ecological
44
45
46 53 treatment systems that have been widely used for their high pollutant removal
47
48
49 54 efficiency, convenient operation and management, low investment, and low operating
50
51
52
55 costs. In traditional horizontal subsurface flow CWs, the environment in the pond
53
54 56 tends to be anaerobic because of the long-term submergence of the substrate, and the
55
56
57 57 reoxygenation effect is poor (Lin et al., 2002). In vertical flow CWs, sewage flows in
58
59
60 58 the matrix via gravity, and the reoxygenation capacity is improved. However, the
61
62
63
64
65
 1 59 ability of these CWs to remove organic matter and NH4+-N is insufficient for meeting
2
3
4
60 water quality standards (Reddy et al., 1997). Tidal flow constructed wetlands (TFCWs)
5
6 61 use the interstitial suction generated by changes in the infiltration surface of the bed
7
8
9 62 during tidal operation to draw atmospheric oxygen into the wetland matrix or soil
10
11
12 63 voids, thereby increasing the oxygen transmission in the wetland pool and improving
13
14
15
64 its ability to remove organic matter and NH4+-N (Zhao et al., 2004). A TFCW used to
16
17 65 treat rural sewage has been shown to have average removal rates of BOD5, NH4+-N,
18
19
20 66 and TN of 79%, 95%, and 47%, respectively (Du et al., 2014). A four-stage vertical
21
22
23 67 subsurface flow CW operating in a tidal flow mode for treating domestic sewage has
24
25
26 68 been shown to have average removal rates of COD, NH4+-N, TN, and TP of 84.5%,
27
28 69 95.4%, 56.5%, and 99.4%, respectively (Yang et al., 2012).
29
30
31 70 Although TFCWs have been shown to be effective for pollutant removal,
32
33
34 71 TFCWs alone are insufficient for treating sewage with high COD, NH4+-N, and TP.
35
36
37 72 Another outstanding issue facing the treatment of domestic sewage in scattered rural
38
39 73 areas is odor. Following facultative anaerobic treatment of fecal sewage in septic
40
41
42 74 tanks, small molecular compounds containing sulfur, nitrogen, organic acids,
43
44
45 75 aldehydes, and ketones are produced (Mahmood et al., 2007; Capelli et al., 2009),
46
47
48 76 which are responsible for sewage odor. The simultaneous removal and spill control of
49
50 77 odorous compounds in septic tank sewage by current rural sewage treatment
51
52
53 78 technology require improvement. Deodorization technology currently used for the
54
55
56 79 treatment of malodorous gas in urban sewage treatment plants primarily employs the
57
58
59 80 following techniques: activated carbon adsorption, chemical absorption, biological
60
61
62
63
64
65
 1 81 filter, and combustion methods (Wysocka et al., 2019). These deodorization
2
3
4
82 technologies have high infrastructure costs, and they are difficult to operate and
5
6 83 manage because of their complexity, which precludes them from being used for the
7
8
9 84 decentralized treatment of rural sewage.
10
11
12 85 The aim of this study was to develop a combined process of reflux deodorization
13
14
15
86 and TFCW to improve the water quality characteristics (COD, NH4+-N, TP, and odor)
16
17 87 of effluent derived from rural septic tanks. This process reduces the concentration of
18
19
20 88 influent water and odorous substances in the CWs through backflow dilution and
21
22
23 89 promotes the conversion of odorous substances through the oxidation of
24
25
26 90 nitrate-nitrogen (NO3--N) in the backflow liquid. The effect of the reflux ratio (R) on
27
28 91 the removal of pollutants and the deodorization effect of the combined process were
29
30
31 92 examined, and the patterns of denitrification and deodorization in the deodorizing
32
33
34 93 pool (DP) and their underlying mechanisms were evaluated. Finally, the microbial
35
36
37 94 community and functional groups of bacteria in the DP and CWs were analyzed.
38
39 95 Generally, the results of our study provide new insights that could be used to aid the
40
41
42 96 development of decentralized treatment technologies to treat rural domestic sewage.
43
44 97
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 98 2. Materials and methods
2
3
4
5 99 2.1. Experimental equipment and materials
6
7
8
9 100 The combined process and process flow of reflux deodorization and TFCW are
10
11 101 shown in Fig. 1. The effluent from the septic tank and the backflow from the tailwater
12
13
14 102 regulating tank enter the DP. After passing through the combined elastic packing area,
15
16
17 103 the effluent enters the iron-carbon micro-electrolysis packing area to complete the
18
19
20 104 dilution and conversion of odorous substances. The sewage is then pumped into the
21
22 105 threaded water distributor under the action of the lift pump and enters the primary CW
23
24
25 106 (PCW) by falling water, followed by reoxygenation. Sewage flows through the filling
26
27
28 107 area in a vertical flow model in the PCW and then into the water outlet trough through
29
30
31 108 the perforated bottom plate, where it naturally overflows; it then enters the secondary
32
33 109 CW (SCW) by falling water, followed by reoxygenation. In the SCW, the sewage
34
35
36 110 flows through the fill area in a vertical flow model and is discharged in tidal flow
37
38
39 111 mode. The effluent then enters the tailwater adjustment tank, and part of the tailwater
40
41
42 112 is returned for dilution of the septic tank sewage. The other part naturally overflows
43
44 113 out of the system.
45
46
47 114 The DP is made of an organic glass plate, and the device is cuboid (600 mm ×
48
49
50 115 300 mm × 400 mm) with a volume of 48 L. The pool is filled with polyethylene
51
52
53 116 elastic composite filler (diameter 150 mm) and iron-carbon micro-electrolytic filler
54
55 117 (particle diameter 30–40 mm) from left to right. TFCWs are divided into two levels.
56
57
58 118 The PCWs are made of an organic glass column (250 mm diameter, 450 mm height),
59
60
61
62
63
64
65
 1 119 with a circular water outlet (width 50 mm). The device is filled with 50 mm of gravel
2
3
4
120 (particle size: 10–15 mm), 300 mm of zeolite (particle size 3–8 mm), and 50 mm of
5
6 121 gravel (particle size 10–15 mm) from top to bottom. The inner volume of the pond is
7
8
9 122 50 L, the porosity is 30%, and canna is planted. The SCWs are made of an organic
10
11
12 123 glass column (500 mm diameter, 450 mm height), and the device is filled with 50 mm
13
14
15
124 of gravel (particle size 10–15 mm), 300 mm of sea oyster shells (particle size 20–30
16
17 125 mm), and 50 mm of gravel (grain size 10–15 mm) from top to bottom. The inner
18
19
20 126 volume of the pond is 250 L, the porosity is 30%, and canna is planted.
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51 127
52
53 128 Fig.1. Process flow diagram
54
55
56
57
58
59
60
61
62
63
64
65
 1 129 2.2 Experimental method
2
3
4
5 130 The activated sludge was obtained from a sewage treatment plant in Chongqing,
6
7 131 China. The sludge was inoculated into the DP and TFCWs; after soaking for 7 d, the
8
9
10 132 operation of the TFCW system was initiated. The film stabilized after the system ran
11
12
13 133 for 40 d. The intermittent recirculation method was used to study the effect of
14
15
16 134 recirculation on the removal of COD, NH4+-N, and TP in the combined process and
17
18 135 the deodorization effect of the DP. In the first set of tests, Rs of 0, 50%, 100%, 150%,
19
20
21 136 and 200% were examined. To characterize patterns of deodorization and
22
23
24 137 denitrification in the DP and their underlying mechanisms, the second set of tests
25
26
27 138 consisted of DP inlet water ρ(NO3--N)/ρ(S2-) with nitrogen-sulfur ratios (N/S) of 0.2,
28
29
139 0.4, 0.6, 0.8, and 1.0 was used. In the third set of tests, the system was run using inlet
30
31
32 140 water ρ(COD)/ρ(NO3--N) with C/N ratios of 2, 4, 6, 8, and 10. In each working
33
34
35 141 condition for each test, the system was run for 30 d.
36
37
38
39 142 2.3 Test sewage
40
41
42
43 143 The test sewage was the effluent from a septic tank of a rural village in
44
45
46
144 Chongqing, China. The daily output of sewage was approximately 80 L/d. The
47
48 145 primary pollutant concentrations are shown in Table 1.
49
50 146
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 147 Table 1. Main water quality indicators of test sewage
2
3 Index scope
4 pH 6.6~7.1
5
6 COD (mg/L) 208~253
7 NH4+-N (mg/L) 59.6~66.0
8 TP (mg/L) 11.9~13.7
9
10 S2- (mg/L) 14.2~16.6
11 Odor threshold (TON) 158~177
12
13 148 Sampling and monitoring were started after 10 d of stable operation in each
14
15
16 149 working condition. Samples were taken every 2 d for 20 d from the water inlet of the
17
18
19 150 system, the water outlet of the DP, and the water outlet of the SCWs. The water
20
21
22
151 samples were filtered through a 0.45-μm membrane and then analyzed and tested. The
23
24 152 NH4+-N, NO3−-N, TP, and sulfide (S2-) concentrations and threshold odor number
25
26
27 153 (TON) were determined according to Chinese SEPA standard methods (SEPA, 2002).
28
29
30 154 COD was determined by the fast digestion spectrophotometric method using a
31
32
33 155 portable CODCr instrument (DR1010, Hach, USA). The pH and temperature were
34
35 156 determined using a pH meter (FE28, Mettler Toledo, USA).
36
37
38 157 The biofilms of the DP and CWs were obtained after the first phase of the
39
40
41 158 experiment. The biological samples of the inoculated sludge were obtained at the
42
43
44 159 beginning of the experiment. The biofilms were sent to Shanghai Meiji Biomedical
45
46 160 Technology Co., Ltd. for analysis.
47
48
49 161
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 162 3. Results and Discussion
2
3
4
5 163 3.1 Effect of R on the pollutant removal effect
6
7
8
9 164 3.1.1 Effect of R on the removal of COD, NH4+-N, TP, and NO3--N
10
11 165 The changes in pollutants and removal rates under different Rs in the system
12
13
14 166 inlet water, DP outlet, and TFCW outlet water are shown in Fig. 2.
15
16
17 167 The average influent COD of the system was 243±5 mg/L, the average effluent
18
19
20 168 COD of the DP was 162±46 mg/L, the average effluent COD of the SCWs was 66±11
21
22 169 mg/L, and the average removal rate of system COD was 72.8 ±5.0% (Fig. 2a). When
23
24
25 170 R was 150% or 200%, the COD of the system effluent was less than 60 mg/L. The
26
27
28 171 COD removal efficiency of the system increased as R increased. The system COD
29
30
31 172 removal rate increased significantly when R changed from 50% to 150%; when R
32
33 173 increased to 200%, the system COD removal rate increased slowly. The decrease in
34
35
36 174 the COD concentration in the system's effluent stems from the dilution effect of the
37
38
39 175 reflux and the degradation effect of the CWs. As R increased, the reflux dilution effect
40
41
42 176 increased, and the hydraulic retention time (HRT) of the CWs gradually decreased,
43
44 177 which in turn caused the COD removal efficiency of the CWs to decrease. However,
45
46
47 178 the increase in the reduction of COD via the reflux dilution effect was always greater
48
49
50 179 than that of the CWs; consequently, the COD removal rate of the system increased.
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 180
27
28 181 Fig. 2. Effect of R on the removal of COD, NH4+-N, TP, and NO3--N
29
30
31 182 The average influent NH4+-N of the system was 63.3±2.0 mg/L, the average
32
33
34 183 effluent NH4+-N of the DP was 45.2±13.5 mg/L, the average NH4+-N effluent of the
35
36
37 184 SCWs was 16.8±1.1 mg/L, and the average removal rate of NH4+-N in the system was
38
39 185 73.5±1.2% (Fig. 2b). As R increased, the NH4+-N removal efficiency of the system
40
41
42 186 increased. When R changed from 0% to 150%, the removal efficiency of NH4+-N
43
44
45 187 increased; when R was 200%, the removal efficiency of NH4+-N decreased. The
46
47
48 188 decrease in the concentration of NH4+-N in the effluent of the system decreased
49
50 189 because of the dilution effect of reflux and the degradation of the CWs. As R
51
52
53 190 increased from 0 to 200%, the HRT of the CWs decreased from 29.9 h to 9.7 h, and
54
55
56 191 the submerging and emptying time of the bed of the SCWs gradually decreased.
57
58
59 192 Previous studies have shown that NH4+-N absorption and denitrification occur in the
60
61
62
63
64
65
 1 193 submerged stage, and nitrification occurs in the emptying stage in TFCWs (Chang et
2
3
4
194 al., 2014). Therefore, as R increased, the dilution effect of reflux increased, but the
5
6 195 removal rate of NH4+-N in the CWs decreased, as it was affected by the shortening of
7
8
9 196 HRT and emptying time. When R changed from 0% to 150%, the reduction of
10
11
12 197 NH4+-N stemming from reflux dilution was greater than that from the degradation
13
14
15
198 effect of the CWs, and the removal rate of NH4+-N in the system increased. When R
16
17 199 was 200%, the reduction of NH4+-N stemming from reflux dilution was smaller than
18
19
20 200 that stemming from the degradation effect of the CWs, and the removal efficiency of
21
22
23 201 NH4+-N in the system decreased.
24
25
26 202 The average TP of the system was 12.9±0.4 mg/L, the average TP of the DP was
27
28 203 9.6±2.4 mg/L, the average TP of the SCWs was 3.4±0.4 mg/L, and the average
29
30
31 204 removal rate of system TP was 73.7±3.4% (Fig. 2c). As R increased, the TP removal
32
33
34 205 efficiency of the system decreased. The concentration of TP in the effluent of the
35
36
37 206 system decreased because of the dilution effect of reflux and the degradation effect of
38
39 207 the CWs. As R increased, the HRT of the CWs gradually shortened, which reduced
40
41
42 208 the contact time of the sewage with the substrate and microorganisms and thus the TP
43
44
45 209 removal efficiency of the CWs. As R increased, the reduction in effluent TP stemming
46
47
48 210 from reflux dilution was smaller than that stemming from the degradation effect of the
49
50 211 CWs, and the TP removal efficiency of the system decreased.
51
52
53 212 The average influent NO3--N of the system was 0.8±0.1 mg/L, the average
54
55
56 213 effluent NO3--N of the DP was 1.7±1.3 mg/L, and the average effluent NO3--N of the
57
58
59 214 SCWs was 6.5±1.0 mg/L (Fig. 2d). As R increased, the concentration of NO3--N in the
60
61
62
63
64
65
 1 215 effluent from the DP and CWs increased. This can be explained by the fact that
2
3
4
216 denitrifying bacteria have a limited ability to remove nitrate under certain conditions.
5
6 217 As R increased, the NO3--N concentration in the DP, CWs, and influent increased.
7
8
9 218 The results indicate that the combined process has a strong removal effect on
10
11
12 219 COD, NH4+-N, and TP, and R has a significant effect on the removal efficiency of
13
14
15
220 COD, NH4+-N, and TP (P < 0.001). As R increased, the removal efficiency of COD
16
17 221 and NH4+-N increased, and the removal efficiency of TP decreased.
18
19
20 222 3.1.2 Effect of R on deodorization
21
22
23 223 The malodorous substances in sewage are mainly divided into three types:
24
25
26 224 sulfur-containing compounds, nitrogen-containing compounds, and carbon hydroxides
27
28 225 (Choi et al., 2012; Lebreroi et al., 2011). The fluid state of the sewage continually
29
30
31 226 changes during the treatment process, which causes the release of these odorous gases.
32
33
34 227 Previous studies have shown that H2S and NH3 are the main components of odor
35
36
37 228 (Gostelow et al., 2000). H2S has been shown to be the most odorous gas, and it has
38
39 229 the most significant effect on the human body (Gostelow, P et al., 2001, Latos et al.,
40
41
42 230 2011). The ratio of the dissolved H2S content to the total S2- content in water is related
43
44
45 231 to the pH of the water. When the pH is 6.0, the H2S content makes up 90% of the total
46
47
48 232 S2- content; when the pH is 7.0, the H2S content in the S2- makes up 43.5% of the total
49
50 233 S2- content. Therefore, the amount of H2S released can be controlled by reducing S2-
51
52
53 234 in the water. The odor threshold number (TON) indicates the strength of the odor
54
55
56 235 emitted by the water body.
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25 236
26
27 237 Fig. 3. Effect of different reflux ratios on deodorization effect
28
29
30 238 The average influent S2- of the system was 15.4±0.7 mg/L, the average effluent
31
32
239 S2- in the DP was 3.0±4.6 mg/L, the average effluent S2- in the SCWs was 2.2±3.4
33
34
35 240 mg/L, and the average removal rate of system S2- was 85.8±22.2% (Fig. 3a, c). As R
36
37
38 241 increased, the system S2- removal efficiency increased. When R was 100%, S2- in the
39
40
41 242 water was completely removed. Previous studies have shown that Thiobacillus can
42
43
243 use O2 and NO3--N as electron acceptors to oxidize S2- to elemental sulfur or sulfate,
44
45
46 244 resulting in autotrophic sulfur denitrification (Hao et al., 2000). As R increased, the
47
48
49 245 reflux dilution effect increased. This also introduces more dissolved oxygen (DO) and
50
51
52 246 NO3--N to the DP, which promotes simultaneous desulfurization and denitrification
53
54
55
247 and reduces the S2- content in the water body.
56
57 248 The average inflow TON of the system was 166±6, the average outflow TON of
58
59
60 249 the DP was 43±33, and the average removal rate of system TON was 73.5±21.4% (Fig.
61
62
63
64
65
 1 250 3b, c). As R increased, the system TON removal efficiency increased. When R was
2
3
4
251 increased from 0 to 100%, the TON removal rate greatly increased; when R was
5
6 252 increased from 100% to 200%, the TON removal rate was only increased by 4.15%. A
7
8
9 253 previous study has shown that most malodorous substances in sewage are formed
10
11
12 254 under extremely low redox potential conditions (Aethur et al., 1998). Increasing the
13
14
15
255 water redox potential can effectively inhibit the generation of malodorous substances.
16
17 256 As R increased, the dilution effect of the reflux was enhanced, and the reflux
18
19
20 257 introduced more NO3--N and DO, which increased the redox potential of the water
21
22
23 258 body and reduced its TON. When R was between 100% and 200%, the S2- in the
24
25
26 259 water body was removed, and the TON of the water body did not change. This
27
28 260 suggests that the concentration of S2- (H2S) in the water body can have a substantial
29
30
31 261 effect on the odor.
32
33
34 262 The results indicate that the combined process has a strong deodorizing effect. R
35
36
37 263 has a strong effect on the removal efficiency of S2- and the TON of the system
38
39 264 (P<0.001). The removal efficiency of S2- and TON increased as R increased. The
40
41
42 265 removal efficiency was optimal when R was 100%, as the average removal rates of
43
44
45 266 system S2- and TON at this R were 99.9% and 85.1%, respectively.
46
47 267
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 268 3.2 Microbial community
2
3
4
5 269 The main microbial phyla in the CWs were Proteobacteria, Actinobacteria,
6
7 270 Patescibacteria, Bacteroidetes, and Chloroflexi, which made up over 85% of all taxa
8
9
10 271 (Fig. 4a). Proteobacteria and Patescibacteria are involved in denitrification.
11
12
13 272 Actinobacteria are involved in the removal of organic matter in sewage. Bacteroidetes
14
15
16 273 are primarily hydrolytic acidifying bacteria involved in the degradation of organic
17
18 274 matter. The main microbial phyla in the DP were Firmicutes, Proteobacteria,
19
20
21 275 Chloroflexi, Actinobacteria, and Bacteroidetes, which made up over 85% of all taxa
22
23
24 276 (Fig. 4b). With the exception of Firmicutes, the main microbial phyla from the middle
25
26
27 277 door of the DP were similar to those of the CWs. Firmicutes degrade organic matter
28
29
278 through the fermentation and decomposition of carbohydrates, proteins, and amino
30
31
32 279 acids.
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
280
58
59
60 281 Fig. 4. Microbial community structure and functional bacteria. (a) Relative abundance
61
62
63
64
65
 1 282 of phyla in the CWs; (b) relative abundance of phyla in the DP; (c) horizontal heat
2
3
4
283 map of genera in the CWs; (d) horizontal heat map of genera in the DP; and (e)
5
6 284 relative abundance of different functional groups of bacteria in the DP and CWs.
7
8
9 285 Next, we analyzed the abundances of different functional groups of
10
11
12 286 microorganisms, including organic degraders, ammonia oxidizers, nitrate reducers,
13
14
15
287 and hydrogen sulfide oxidizers. The distribution of genera in the two CWs differed
16
17 288 (Fig. 4c), which might be caused by their different operating modes. The organic
18
19
20 289 degraders in the CWs mainly included Nakamurella, Ferruginibacter, Conexibacter,
21
22
23 290 norank_f__norank_o__Saccharimonadales, and norank_f__Xanthomonadaceae;
24
25
26 291 ammonia oxidizers included Chujaibacter, Nitrosomonas, Nitrospira, and
27
28 292 Ellinobacter 6067; and nitrate reducers included Rhodanobacter, Ottowia,
29
30
31 293 Micropruina, Thermomonas, Iamia, Bradyrhizobium, and Nitrosomonas. The
32
33
34 294 dominant functional group in the PCWs was organic degraders (51.8%) (Fig. 4e). The
35
36
37 295 abundance of ammonia oxidizers (3.6%) and nitrate reducers (26.9%) in the SCWs
38
39 296 and the PCWs significantly differed. This might be explained by the high organic
40
41
42 297 matter load of influent in the PCWs, which favors the growth of organic degraders.
43
44
45 298 By contrast, the organic matter load of the influent in the SCWs is low, and the tidal
46
47
48 299 flow method is used to discharge the water, which enhances the growth of nitrate
49
50 300 reducers.
51
52
53 301 In the DP, the main organic degraders were Trichococcus, Propionicimonas,
54
55
56 302 Denitrattisoma, Desulfomicrobium, and unclassified_f__Rhodocyclaceae; nitrate
57
58
59 303 reducers were Thiobacillus, Denitratisoma, Rhodobacter,
60
61
62
63
64
65
 1 304 unclassified_f__Rhodocyclaceae, and Hydrogenophaga; and H2S oxidizers were
2
3
4
305 Thiocapsa, Thiobacillus, Sulfuritalea, Desulfoprunum, and Rhodopseudomonas (Fig.
5
6 306 4d). The abundance of organic degraders, nitrate reducers, and H2S oxidizers in the
7
8
9 307 DP was 40.1%, 6.9%, and 8.0%, respectively (Fig. 4e).
10
11
12 308 The tidal flow operation mode significantly altered the microbial community and
13
14
15
309 the distribution of functional groups in the CWs and favored the growth of ammonia
16
17 310 oxidizers and nitrate reducers in the CWs. Thiobacillus performed simultaneous
18
19
20 311 denitrification and deodorization reactions in the DP.
21
22 312
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 313 3.3 Simultaneous denitrification and deodorization and its
2
3
4 314 underlying mechanism
5
6
7 315 Analysis of the microbial community and functional groups in the DP revealed
8
9
10 316 that the removal of S2- in the DP stemmed from the dilution of the reflux liquid and
11
12
13 317 the denitrification and desulfurization of Thiobacillus. In denitrification and
14
15
16 318 desulfurization, Thiobacillus uses nitrate in the reflux solution as an electron acceptor
17
18 319 and S2- in the water as an electron donor; denitrification and desulfurization reactions
19
20
21 320 occur under anoxic or anaerobic conditions, and the specific reaction equation is as
22
23
24 321 follows (Huang et al., 2015):
25
26 322 12𝐻 + + 2𝑁𝑂3− + 5𝑆 2− → 𝑁2 + 5𝑆 + 6𝐻2 𝑂
27
28 323 ∆𝐺 = −1151.38 𝐾𝐽/𝑚𝑜𝑙 (1)
29
30
31
324 5𝑆 + 6𝑁𝑂3− + 8𝐻2 𝑂 → 5𝐻2 𝑆𝑂4 + 6𝑂𝐻 − + 3𝑁2
32
33
325 ∆𝐺 = −1833.96 𝐾𝐽/𝑚𝑜𝑙 (2)
34
35 326 In the simultaneous desulfurization and denitrification reaction, S2- is first
36
37
38 327 oxidized to elemental sulfur, and the N/S ratio required for the reaction is 2/5. The
39
40
41
328 elemental sulfur is then continuously oxidized to sulfate by nitrate. Many studies have
42
43 329 examined the factors affecting the denitrification and desulfurization reactions, and
44
45
46 330 this work has shown that N/S(ρ(NO3--N)/ρ(S2-)) and C/N(ρ(COD)/ρ(NO3--N)) greatly
47
48
49 331 affect the simultaneous desulfurization and denitrification reaction (Huang et al., 2021;
50
51
52
332 Xu et al., 2016; Wen et al., 2019).
53
54 333 3.3.1 Effect of N/S on denitrification and deodorization
55
56
57 334 Changes in TON, S2-, and NO3--N of the inlet and outlet water of the DP and
58
59
60 335 their removal efficiencies under different N/S are shown in Fig. 5. As N/S increased,
61
62
63
64
65
 1 336 the removal rate of TON and S2- increased, and the removal rate of NO3--N decreased
2
3
4
337 (Fig. 5). When N/S was 0.8–1.0, the average removal rates of TON, S2-, and NO3--N
5
6 338 were 80.3±15.5%, 92.3±6.8%, and 54.9±12.5%, respectively.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32 339
33 340 Fig. 5. Effect of N/S on the denitrification and deodorization of the DP
34
35 341 Previous studies have shown that N/S and the S2- concentration are key factors
36
37
38 342 affecting the simultaneous desulfurization and denitrification process. When N/S is
39
40
41 343 3/5 and S2- is less than 300 mg/L, the conversion rate of elemental sulfur reaches 94%
42
43
344 (Wang et al., 2005). The N/S optimal for desulfurization and denitrification differs;
44
45
46 345 specifically, the optimal N/S for desulfurization is 0.4, whereas that for denitrification
47
48
49 346 is 1.0 (Xu et al., 2019). The DP is in an oxygen-deficient environment, and the S2-
50
51
52 347 removal pathway might include the autocatalytic reaction of polysulfide in addition to
53
54
55
348 nitrate oxidation. Therefore, as the N/S ratio increases, the polysulfur autocatalytic
56
57 349 reaction proceeds more readily than autotrophic denitrification and desulfurization,
58
59
60 350 thereby increasing the removal rate of S2- and decreasing the removal rate of nitrate.
61
62
63
64
65
 1 351 N/S significantly affected the removal efficiency of system TON, S2-, and
2
3
4
352 NO3--N (P<0.001). As N/S increased, the removal rate of TON and S2- increased, and
5
6 353 the removal rate of NO3--N decreased. When N/S was 0.8, the average removal rates
7
8
9 354 of TON, S2-, and NO3--N in the DP were 78.8%, 91.4%, and 62.3%, respectively
10
11
12 355 (effluent TON<37, S2-<1.54 mg/L, and NO3--N<4.63 mg/L), which enhanced
13
14
15
356 deodorization and denitrification.
16
17 357 3.3.2 Effect of C/N on denitrification and deodorization
18
19
20 358 Changes in TON, S2-, and NO3--N of the inlet and outlet water of the DP and the
21
22
23 359 removal rate under different C/N and N/S of 0.8 are shown in Fig. 6. As C/N
24
25
26 360 increased, the removal rate of TON and S2- decreased, and the removal rate of NO3--N
27
28 361 increased (Fig. 6). When C/N was 2–4, the average removal rates of TON, S2-, and
29
30
31 362 NO3--N were 85.6±12.5%, 89.5±4.8%, and 48.2±10.0%, respectively.
32
33
34 363 The simultaneous removal of carbon, nitrogen, and sulfur was mainly
35
36
37 364 accomplished by heterotrophic and autotrophic denitrifiers, which differ in their
38
39 365 ability to use organic carbon sources (Chen et al., 2017). As C/N increased, organic
40
41
42 366 carbon sources became limiting, and the activity of heterotrophic denitrifying
43
44
45 367 microorganisms increased. This increased the metabolism of nitrates, which
46
47
48 368 compromised the ability of Thiobacillus denitrificans to obtain nitrate, resulting in a
49
50 369 decrease in the removal rate of S2- and TON.
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
370
28
29 371 Fig. 6. Effect of C/N on the denitrification and deodorization of the DP
30
31 372 C/N significantly affected the removal efficiency of TON, S2-, and NO3--N
32
33
34 373 (P<0.001). As C/N increased, the removal rates of TON and S2- decreased, and the
35
36
37 374 removal rate of NO3--N increased. When C/N was 0.4, the average removal rates of
38
39 375 TON, S2-, and NO3--N in the DP were 82.6%, 88.7%, and 51.8%, respectively
40
41
42 376 (effluent TON<30, S2-<1.71 mg/L, NO3--N<5.92 mg/L), which enhanced
43
44
45 377 deodorization and denitrification.
46
47 378
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 379 4. Conclusion
2
3
4
5 380 When the combined process of reflux deodorization and TFCW was used to treat
6
7 381 septic tank effluent, the removal efficiency of COD, NH4+-N, S2-, and TON increased
8
9
10 382 and the removal rate of TP decreased as R increased. When R=150%, the average
11
12
13 383 removal rates of COD, NH4+-N, TP, S2-, and TON by the combined process were
14
15
16 384 76.6%, 75.0%, 71.1%, 100%, and 88.0%, respectively. The combined process had a
17
18 385 strong removal effect on COD, NH4+-N, and odor. The tidal flow operation mode
19
20
21 386 significantly altered the microbial community and the distribution of functional
22
23
24 387 groups of bacteria in the CWs. The abundance of ammonia oxidizers and nitrate
25
26
27 388 reducers in the CWs increased. Thiobacillus were the main bacteria involved in
28
29
389 denitrification and desulfurization in the DP. N/S and C/N were both important factors
30
31
32 390 affecting denitrification and deodorization. As N/S increased, the removal rate of
33
34
35 391 TON and S2- increased, and the removal rate of NO3--N decreased. When N/S was 0.8,
36
37
38 392 the effluent of the DP had the following properties: TON<37, S2-<1.54 mg/L, and
39
40
41
393 NO3--N<4.63 mg/L. As C/N increased, the removal rate of TON and S2- decreased,
42
43 394 and the removal rate of NO3--N increased. When C/N was 4, the effluent of the DP
44
45
46 395 had the following properties: TON<30, S2-<1.71 mg/L, and NO3--N<5.92 mg/L.
47
48
49 396 Under both influent conditions, strong deodorizing and denitrification effects were
50
51
52
397 achieved.
53
54 398
55
56
57
58
59
60
61
62
63
64
65
 1 399 Acknowledgments
2
3
4
400 This study was financially supported by National Key Research and
5
6 401 Development Program [grant number: 2018YFD1100501], and Chongqing Science
7
8
9 402 and Technology Commission [grant number: cstc2018jszx-zdyfxmX0014].
10
11 403
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1 404 References
2
3
4
5 405 Aethur, G., et al. (1998). "Septicity in sewage and sludge." Water＆Waste Treatment
6
7 406 41(9):24－26．
8
9
10 407 Chen, C., et al. (2017). "Enhanced performance of denitrifying sulfide removal
11
12
13 408 process at high carbon to nitrogen ratios under micro-aerobic condition." Bioresource
14
15
16 409 Technology 232: 417-422.
17
18 410 Capelli, L., et al. (2009). "Predicting odour emissions from wastewater treatment
19
20
21 411 plants by means of odour emission factors." Water Research 43(7): 1977-1985.
22
23
24 412 Chang, Y. J., et al. (2014). "Dynamics of nitrogen transformation depending on
25
26
27 413 different operational strategies in laboratory-scale tidal flow constructed wetlands."
28
29
414 Science Of the Total Environment 487: 49-56.
30
31
32 415 Choi, I., et al. (2012). "Evaluation of the Effectiveness of Five Odor Reducing Agents
33
34
35 416 for Sewer System Odors Using an On-Line Total Reduced Sulfur Analyzer." Sensors
36
37
38 417 12(12): 16892-16906.
39
40
41
418 Deng, Y. H. and A. Wheatley (2016). "Wastewater Treatment in Chinese Rural Areas."
42
43 419 Asian Journal Of Water Environment And Pollution 13(4): 1-11.
44
45
46 420 Du, X. L., et al. (2014). "ENHANCED NITROGEN REMOVAL IN A
47
48
49 421 NON-PLANTED MODIFIED EXTENDED TIDAL FLOW CONSTRUCTED
50
51
52
422 WETLAND." Fresenius Environmental Bulletin 23(1A): 285-289.
53
54 423 Gostelow, P. and S. A. Parsons (2000). "Sewage treatment works odour
55
56
57 424 measurement." Water Science And Technology 41(6): 33-40.
58
59
60 425 Gostelow, P., et al. (2001). "Odour measurements for sewage treatment works." Water
61
62
63
64
65
 1 426 Research 35(3): 579-597.
2
3
4
427 Hao, W., et al. (2020). "Organic carbon coupling with sulfur reducer boosts sulfur
5
6 428 based denitrification by Thiobacillus denitrificans." Science Of the Total Environment
7
8
9 429 748.
10
11
12 430 Huang, C., et al. (2015). "Enhanced elementary sulfur recovery with sequential
13
14
15
431 sulfate-reducing, denitrifying sulfide-oxidizing processes in a cylindrical-type
16
17 432 anaerobic baffled reactor." Bioresource Technology 192: 478-485.
18
19
20 433 Huang, S., et al. (2021). "Realization of nitrite accumulation in a sulfide-driven
21
22
23 434 autotrophic denitrification process: Simultaneous nitrate and sulfur removal."
24
25
26 435 Chemosphere 278.
27
28 436 Lin, Y. F., et al. (2002). "Effects of macrophytes and external carbon sources on
29
30
31 437 nitrate removal from groundwater in constructed wetlands." Environmental Pollution
32
33
34 438 119(3): 413-420.
35
36
37 439 Liu, Z. H. (2011). Current Situation and Main Pollution Sources of Rural Water
38
39 440 Environment in China. Environment Materials And Environment Management. Z. Y.
40
41
42 441 Du and C. B. Li. 281: 113-116.
43
44
45 442 Lebrero, R., et al. (2011). "Odor Assessment and Management in Wastewater
46
47
48 443 Treatment Plants: A Review." Critical Reviews In Environmental Science And
49
50 444 Technology 41(10): 915-950.
51
52
53 445 Ma, Y. H., et al. (2021). "Nitrification, denitrification and anammox process coupled
54
55
56 446 to iron redox in wetlands for domestic wastewater treatment." Journal Of Cleaner
57
58
59 447 Production 300.
60
61
62
63
64
65
 1 448 Ma, Y. H., et al. (2021). "Nitrification, denitrification and anammox process coupled
2
3
4
449 to iron redox in wetlands for domestic wastewater treatment." Journal Of Cleaner
5
6 450 Production 300.
7
8
9 451 Latos, M., et al. (2011). "Dispersion of Odorous Gaseous Compounds Emitted from
10
11
12 452 Wastewater Treatment Plants." Water Air And Soil Pollution 215(1-4): 667-677.
13
14
15
453 Reddy, K. R. and E. M. Dangelo (1997). "Biogeochemical indicators to evaluate
16
17 454 pollutant removal efficiency in constructed wetlands." Water Science And Technology
18
19
20 455 35(5): 1-10.
21
22
23 456 Wysocka, I., et al. (2019). "Technologies for deodorization of malodorous gases."
24
25
26 457 Environmental Science And Pollution Research 26(10): 9409-9434.
27
28 458 Wang, A. J., et al. (2005). "An innovative process of simultaneous desulfurization and
29
30
31 459 denitrification by Thiobacillus denitrificans." Journal Of Environmental Science And
32
33
34 460 Health Part a-Toxic/Hazardous Substances & Environmental Engineering 40(10):
35
36
37 461 1939-1949.
38
39 462 Wen, S. L., et al. (2019). "The effects of Fe2+ on sulfur-oxidizing bacteria (SOB)
40
41
42 463 driven autotrophic denitrification." Journal Of Hazardous Materials 373: 359-366.
43
44
45 464 Xu, X. J., et al. (2016). "Characterization of a newly isolated strain Pseudomonas sp
46
47
48 465 C27 for sulfide oxidation: Reaction kinetics and stoichiometry." Scientific Reports 6.
49
50 466 Xu, J. L., et al. (2019). "Regulation of influent sulfide concentration on anaerobic
51
52
53 467 denitrifying sulfide removal." Environmental Technology 40(11): 1392-1400.
54
55
56 468 Yang, Y., et al. (2012). "Enhanced P, N and C removal from domestic wastewater
57
58
59 469 using constructed wetland employing construction solid waste (CSW) as main
60
61
62
63
64
65
 1 470 substrate." Water Science And Technology 66(5): 1022-1028.
2
3
4
471 Yi, X. S., et al. (2020). "Ecological treatment technology for agricultural non-point
5
6 472 source pollution in remote rural areas of China." Environmental Science And
7
8
9 473 Pollution Research: 13.
10
11
12 474 Zhao, Y. Q., et al. (2004). "Purification capacity of a highly loaded laboratory scale
13
14
15
475 tidal flow reed bed system with effluent recirculation." Science Of the Total
16
17 476 Environment 330(1-3): 1-8.
18
19
20 477
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65