Abstract

With the improvement of wastewater discharge standards, wastewater treatment plants (WWTPs) are continually undergoing technological improvements to meet the evolving standards. In this study, a quartz sand deep bed denitrification filter (DBDF) was used to purify WWTP secondary effluent, utilizing high nitrate nitrogen concentrations and a low C/N ratio. Results show that more than 90% of nitrate nitrogen (NO3-N) and 75% of chemical oxygen demand (COD) could be removed by the 20th day of filtration. When the filter layer depth was set to 1,600 mm and the additional carbon source CH3OH was maintained at 30 mg L−1 COD (20 mg L−1 methanol), the total nitrogen (TN) and COD concentrations of DBDF effluent were stabilized below 5 and 30 mg L−1, respectively. Analysis of fluorescence revealed that DBDF had a stronger effect on the removal of dissolved organic matter (DOM), especially of aromatic protein-like substances. High throughput sequencing and qPCR results indicate a distinctly stratified microbial distribution for the main functional species in DBDF, with quartz sand providing a good environment for microbes. The phyla Proteobacteria, Bacteroidetes, and Chloroflexi were found to be the dominant species in DBDF.

INTRODUCTION

Municipal wastewater treatment plants (WWTPs) are a crucial part of modern water pollution control, however many developed regions have recognized the limitations of WWTPs in the removal of various pollutants, resulting in effluents being a potential source of aquatic pollution (Lowenberg et al. 2016). Even in the most stringent Chinese municipal WWTP pollutant discharge standard (GB18918–2002 1A), the concentration limits for nitrogen and phosphorus are higher than those of Class III – IV in the Environmental Quality Standard for Surface Water (GB3838-2002) (MOEP 2002) (Table 1). Nitrogen removal forms an important aspect of WWTP management, with nitrogen from secondary effluents negatively affecting both the aquatic environment and human health (Cao et al. 2016). Restrictions on effluent nitrogen concentrations are critical for the prevention of eutrophication in water bodies and cyanobacterial blooms (Abell et al. 2010). Due to growing concern regarding aquatic environmental health, many cities have established more stringent drainage requirements, resulting in the need for further upgrading of WWTPs to meet standards. However, upgrades are a difficult task and require additional manpower, land, money, and time (Jokela et al. 2002); therefore, establishing cost-effective WWTP development schemes is of significant research interest.

Table 1

Comparisons of Chinese surface water environmental quality standards and the wastewater treatment plant discharge standards

  COD (mg/L) NH4-N (mg/L) TN (mg/L) TP (mg/L) 
Municipal WWTP pollutant discharge standard (GB18918-2002 1A) 50 15 0.5 
Environmental Quality Standard for Surface Water (GB3838-2002 V) 40 0.2 
Environmental Quality Standard for Surface Water (GB3838-2002 IV) 30 1.5 1.5 0.1 
Environmental Quality Standard for Surface Water (GB3838-2002 III) 20 1.0 1.0 0.05 
  COD (mg/L) NH4-N (mg/L) TN (mg/L) TP (mg/L) 
Municipal WWTP pollutant discharge standard (GB18918-2002 1A) 50 15 0.5 
Environmental Quality Standard for Surface Water (GB3838-2002 V) 40 0.2 
Environmental Quality Standard for Surface Water (GB3838-2002 IV) 30 1.5 1.5 0.1 
Environmental Quality Standard for Surface Water (GB3838-2002 III) 20 1.0 1.0 0.05 

In recent years, numerous intensive studies have been conducted on advanced nitrogen removal for WWTP secondary effluent, which can be classified into three types: (1) physical and chemical technologies, such as coagulation (Tang et al. 2013); activated carbon adsorption (Chen et al. 2017) and membrane filtration (Lee et al. 2012); (2) biological technologies, such as bamboo charcoal filters (BCF) (Cao et al. 2016), deep bed denitrification filters (DBDF) (Joseph A 2012), and partial-denitrification (denitrification to nitrite) sequencing batch reactors (PDSBR) (Cao et al. 2017); (3) ecological technologies, such as constructed wetlands (Xiong et al. 2011), oxidation ponds (Zhao et al. 2013), and ecological floating-bed processes (Wang 2012). Among these, physical and chemical technologies have various reported disadvantages such as relatively high operating costs and possible secondary pollution, to name a few. Ecological technologies generally require a large area of land, complex management, and contribute secondary pollution to the environment. In contrast, biological technologies have the advantages of low running costs and high efficiency, with DBDF being particularly well suited to advanced nitrogen removal of the secondary effluent.

Domestic wastewater contains both persistent organics and easily-biodegradable organic matter (OM), which affect denitrification processes (Cao et al. 2017). Although the OM of wastewater plant influents fluctuates significantly, OM concentrations in the secondary effluent remain relatively constant, with especially low bioavailable. Consequently, there is insufficient denitrification of carbon sources during secondary processing (Guo et al. 2010). To provide OM for further denitrification, additional carbon sources are required and commonly, two broad categories of carbon sources have been utilized: traditional and emerging. Traditional carbon sources include methanol, ethanol, acetate, glucose, and other low molecular organic compounds and carbohydrates. Emerging carbon sources include wheat straw, pruned plant matter, corncob, food waste leachate, etc. However, methanol has been the most commonly used additional carbon source due to its low cost.

The aim of the present research was to find a efficient improvement scheme to effectively treat WWTP secondary effluent. Methanol was utilized as the carbon source, with a quartz sand deep bed filter used as the denitrification reactor to investigate microbial dynamics with varying DBDF packing layer depth. The microorganism community in the DBDF denitrification processes was analyzed via high-throughput sequencing, and the obtained results provide a comprehensive basis for the management of WWTP upgrades. In addition, the influence on the denitrification process of additional carbon sources and varying packing depths were studied.

METHODS

Experimental set-up and influent quality

The experimental site was located in a WWTP (north latitude: 31°45′54.60″, east longitude: 117°20′51.80″) in Hefei City (Anhui Province, China), which applied a micro-aeration oxidation ditch treatment process for water treatment. The experiment was conducted for a six-month period, from April 1 2017, to October 1 2017. Methanol was applied as an additional carbon source. The influent supplied to experimental devices was the effluent produced by the WWTP secondary sedimentation tank, with effluent quality characteristics outlined in Figure 1.

Figure 1

Concentrations of NO3-N, NH4-N, NO2-N, COD and BOD5 in effluent from secondary sedimentation tank.

Figure 1

Concentrations of NO3-N, NH4-N, NO2-N, COD and BOD5 in effluent from secondary sedimentation tank.

Design of deep bed denitrification filters (DBDF)

The experimental setup is displayed in Figure 2; the experimental DBDF was constructed of polymethyl methacrylate. WWTP secondary effluent was pumped from the sedimentation tank effluent to the setup influent tank. From there it was pumped to the DBDF using a controlled volume pump. The flow in the DBDF was from top to bottom. The DBDF had an inner diameter of 0.20 m, and a height of 5.05 m. The main layers were a supporting layer, a quartz sand layer and a water layer. The supporting layer consisted of pebbles with a diameter of 3–38 mm, to a total height of 0.45 m; the filling layer consisted of quartz sand with a diameter of 2–3 mm, to a total height of 2.00 m. The hydraulic retention time (HRT) was 2 h. The backwashing system adopted was air and water combined recycle, with a 48 h operation cycle. The first 5 min formed the water backwash stage, followed by 15 min of air and water combined backwash, and finally 5 min of air backwashing. The water backwashing strength was 14.7 m3 (m2•h)−1 and the air backwashing strength was 91.4 m3 (m2•h)−1. The carbon source was added through a tank containing methanol and dosed using a metering pump. There were six sampling openings, spaced 400 mm apart, from the top to the bottom of the filter layer, numbered as S-0, S-400, S-800, S-1200, S-1600, and S-2000.

Figure 2

Scheme of the pilot-scale set-up of the utilized deep bed denitrification filter.

Figure 2

Scheme of the pilot-scale set-up of the utilized deep bed denitrification filter.

Experimental analysis

Wastewater quality analysis

Influent and effluent water samples were analyzed for concentrations of ammonium (NH4-N), nitrate nitrogen (NO3-N), nitrite nitrogen (NO2-N), total nitrogen (TN), chemical oxygen demand (COD), and biochemical oxygen demand at 5 days (BOD5). Parameters were measured according to Standard Methods for the Examination of Water and Wastewater (APHA 2012). Dissolved oxygen (DO) concentration and water temperature were measured immediately, using a DO meter (YSI-550A).

3D-EEM Characterization

Chemical features of DOM were characterized via 3D-EEM spectra, established using a Hitachi F-4600 fluorescence spectrometer (Hitachi, Japan) equipped with a 150-W continuous output xenon arc lamp. The monitored excitation wavelengths (Ex) were in the range of 220–400 nm, the emission wavelengths (Em) ranged from 280 to 500 nm, and sampling intervals were 5 nm for both. The response time was 0.004 s and the scan speed was 2,400 nm min−1. To prevent any internal filter effects, samples were diluted using purified water to provide a UV absorbance at 260 nm of 0.1. Samples were filtered through a 0.45 μm pore size polycarbonate filter before measurements were taken using a fluorescence spectrometer.

DNA extraction and PCR

When the DBDF achieved stable operation, six samples were taken from the sampling port of the filter. Samples contained quartz sand and water. Therefore, they were vibrated strongly to strip microbes from the surface of the quartz sand to the liquid prior to DNA extraction. Extraction of each sample was then performed using an E.Z.N.ATM Mag-Bind Soil DNA Kit for soil extraction (Omega Bio-tek, Inc., USA), according to the manufacturer's protocol. Concentration and purity of the extracted DNA were determined using a spectrophotometer (NanoDrop-2000, Themo Scientific, USA) and confirmed via agarose gel (1%) electrophoresis. 16S rRNA gene segments (16S rDNA) were polymerase chain reaction (PCR)-amplified using the forward primer 341F (CCTACGGGNGG CWGCAG) and the reverse primer 805R (GACTACHVGGGTATCTAATCC), targeting the V3-V4 hypervariable region. The first PCR amplification was performed in a reaction system (30 μL) containing 10–20 ng of genomic DNA, 15 μL 2 × Taq master mix, 1 μL bar-PCR primer F, 1 μL primer R, and H2O. PCR was performed under the following conditions: 94 °C for 3 min; 5 cycles of 94 °C for 30 s, 45 °C for 30 s and 65 °C for 30 s; 20 cycles of 94 °C for 20 s, 55 °C for 20 s, and 72 °C for 30 s; and a final extension at 72 °C for 10 min. The second PCR amplification was also performed in a reaction system (30 μL), introducing Illumina bridge type PCR compatible primers. This system contained 20 ng genomic DNA, 15 μL 2 × Taq master mix, 1 μL primer F, 1 μL primer R, and H2O. The second amplification was performed under the following conditions: 95 °C for 3 min; 5 cycles of 94 °C for 20 s, 55 °C for 20 s, and 72 °C for 30 s; and a final extension at 72 °C for 5 min. PCR products were checked via agarose gel (1%) electrophoresis and the DNA library was constructed and run on the Miseq 2 × 300 bp by Sangon Biotech Co., Ltd (Shanghai, China).

High-throughput sequencing of 16S rRNA genes

The Qubit3.0 DNA detection kit was used to quantify the recovered DNA, allowing convenient sequencing at a 1:1 ratio when the DNA volume of each sample was 10 ng, providing a final sequencing concentration of 20 pmol. Following this, the DNA library was constructed and run on the Miseq 2 × 300 bp by Sangon Biotech Co., Ltd (Shanghai, China). After sequencing, data were processed and the sample sequence was distinguished via barcode, with the sequence of samples then going through QC to remove nonspecific amplification sequences and chimeras.

RESULTS AND DISCUSSION

Quality of secondary effluent

The characteristics of the WWTP secondary sedimentation tank effluent are presented in Figure 2. The concentration of NO2-N was lowest at 0.23 ± 0.06 mg L−1, while NO3-N concentrations accounted for more than 85% of the TN content. The COD levels ranged between 18–32 mg L−1 and BOD5 levels remained below 5 mg L−1. Calculation of the BOD5/COD ratio was 0.16 ± 0.03, significantly below 0.3, which forms the lower limit of biodegradability in wastewater (Ying et al. 2012). This finding shows that the bioavailability of OM in the secondary sedimentation tank effluent is very low and therefore, further denitrification required additional OM. Previous studies have also reported low concentrations of biodegradable organic carbon in WWTP secondary effluent, resulting in the need for external organic carbon sources for further denitrification (Park et al. 2009).

Bioreactor performance

Biofilms are an essential component of quartz sand denitrification filters, with attachment of sufficient microbes to the bed filter being key to an effective DBDF operation. This study utilized continuous flow water to prepare the biofilm, with microbial growth facilitated via addition of a methanol concentration of 30 mg L−1 (COD). At the beginning of the experiment, the COD removal rate was about 50%, while the NO3-N removal rate was only 8%. This may be because the bed filter had a good filtration effect on OM, intercepting and removing part of the OM component. The results also show the removal of NO3-N during the DBDF start-up period. By the 20th day, results show that the removal rate of NO3-N was stable at above 90%, while the COD removal rate also increased to more than 70%, indicating successful biofilm formation.

To investigate the influence of different concentrations of COD on quartz sand DBDF performance, the removal rates of NO3-N, NH4-N, TN, and COD were studied, with the addition of varying concentrations of methanol resulting in additional COD levels of 20 mg L−1, 30 mg L−1, and 40 mg L−1. HRT was set at 2 h and the operational cycle time was set at 48 h. With the increase in COD dosage from 20 mg L−1 to 30 mg L−1, the average removal rates of TN and NO3-N significantly increased from 81.2 ± 6.8% to 87.2 ± 5.7% and 88.1 ± 4.5% to 92.8 ± 5.3%, respectively. Previously reported findings (Hidaka et al. 2003) showed that the degree of denitrification increased with higher C/N ratios and COD concentrations; however, in the present study, the average removal rates of TN and NO3-N decreased to 85.8 ± 5.9% and 91.1 ± 4.6% with further increased COD dosage of 40 mg L−1. This phenomenon may be due to the high concentration of methanol limiting microbial activity, as previously demonstrated by Fernandez-Nava et al. (2010), when effluent COD concentrations exceeded 30 mg L−1. Additionally, the average removal rate of NH4-N showed a similar trend (Figure 3). For the WWTP used in the present study, large fluctuations in influent COD levels are common; therefore, it is important to find the appropriate COD dosage required to stabilize the effluent and meet emission requirements. When COD was supplemented at a maximum of 30 mg L−1, optimal rates of removal of TN and NO3-N were observed, with effluent concentrations of NH4-N, and TN stabilized lower than 1.5 mg L−1 and 5 mg L−1, respectively. Moreover, the ratio of consumed COD to removed NO3-N was 3.65–4.99 mg (mg)−1, which is consistent with previously reported findings (Shi et al. 2015). As a result, the optimal COD concentration was established as 30 mg L−1.

Figure 3

Changes of nitrogen and COD removals at different CH3OH dosages.

Figure 3

Changes of nitrogen and COD removals at different CH3OH dosages.

Six samples were obtained from different sampling points along the direction of water flow in the quartz sand DBDF. Under the conditions of 30 mg L−1 COD, the concentrations of COD, TN, NO3-N, NH4-N, and NO2-N varied according to different packing layer depths (Figure 4). NO2-N is an intermediate product formed in the process of denitrification, accumulating when the carbon source is insufficient (Karanasios et al. 2016). Several previous studies have reported this situation in denitrification experiments (e.g. Cao et al. 2016); however, no NO2-N accumulation was observed in the present study. Therefore, the 30 mg L−1 COD dosage provided sufficient levels of carbon source for the denitrification process. No significant change was observed in the concentration of NH4-N with varying depths of the filter layer, possibly due to the low concentration of NH4-N in the influent and the main NH4-N removal method being absorption by microbes.

Figure 4

Changes of nitrogen and COD concentrations at different DBDF depths (September 10–15).

Figure 4

Changes of nitrogen and COD concentrations at different DBDF depths (September 10–15).

In contrast, the concentrations of COD, TN, and NO3-N significantly changed with varying layer depth, with upper packing layer concentrations of COD, TN, and NO3-N being 54.18, 12.97, and 11.87 mg L−1, respectively. At a packing layer depth of 1,600 mm, concentrations decreased to 22.58, 4.04, and 2.73 mg L−1, respectively. The resulting effluent readily met all requirements of the GB18918-2002 1A discharge standard (MOEP 2002) and when taking the economic cost into account, this is optimal when the depth of filter layer is set to 1,600 mm.

Change in DOM levels

OM is an important component in aquatic ecosystems in which DOM is the most easily released and active substance; therefore, DOM plays a more important role in the ecosystem overall (Liu et al. 2018). Consequently, after treatment with additional carbon sources by DBDF, it is necessary to study the composition and removal of DOM to minimize the effects of DOM on the water environment.

Fluorescence EEM analysis provides a qualitative estimate of the DOM removal performance of the quartz sand DBDF, as shown in Figure 5. Fluorophores in DOM were classified as: Peak A (Ex/Em = 225–237/340–381 nm); Peak B (Ex/Em = 275–285/340–345 nm); Peak C (Ex/Em = 315–350/415–440 nm); Peak D (270–280/415–430 nm); and Peak E (220/310 nm). Based on previously reported literature, peaks A, B, C, D, and E were related to aromatic protein-like, tryptophan protein-like, humic acid-like, fulvic acid-like, and tyrosine protein-like substances, respectively (Liu et al. 2018).

Figure 5

Fluorescence EEMs (RU) of water samples.

Figure 5

Fluorescence EEMs (RU) of water samples.

As shown in Figure 5(a), two main peaks (peak A and B) were observed in the EEM fluorescence spectra of DOM extracted from the secondary sedimentation tank effluent. This implies that oxidation ditches may not be ideal for the removal of aromatic protein-like substances. Figure 5(b) shows the EEM fluorescence spectrum after the addition of methanol (30 mg L−1 based on COD) to the secondary sedimentation tank effluent. In comparison to Figure 5(a), the fluorescence intensity of peaks A and B were enhanced, particularly in peak B.

A strong decolorizing effect can be seen in Figure 5(c) and, compared to Figure 5(a), the fluorescence spectrum of peak A was significantly weaker. Furthermore, peak B was also attenuated, showing that the DBDF with quartz sand has a better removal rate for DOM, especially aromatic protein-like DOM, than the WWTP's oxidation ditch.

Microbial analysis

Microorganisms are a key factor for effective wastewater treatment by biological means. Microbial biochemistry processes occurring in the DBDF are functional in wastewater treatment, with each microbial species present determining which pollutants will be effectively removed. Therefore, microbial analysis can provide a comprehensive understanding of the active treatment mechanisms and aid in characterization of mechanisms and pathways (Hua et al., 2018). To investigate microbial community dynamics during the experimental process, six samples were taken from various filter depths. 16S rRNA gene based high-throughput sequencing technology was utilized to analyze the microbial communities present in these samples. After removing low quality sequences and chimeras, approximately 51,419–68,134 effective sequences were obtained, with an average length ranging from 422 bp to 427 bp (Table 2). These effective sequences were further classified into operational taxonomic units (OTUs) at a 3% distance, to assess the diversity and richness of the bacterial community (Cao et al. 2017).

Table 2

Richness and diversity of the water samples at various filter depths

Sample Mean length OTU counts Shannon index ACE index Chao1 index Coverage Simpson 
S-0 426.98 1,845 2.31 2,831.69 2,538.01 0.988 0.42 
S-400 424.85 2,299 3.81 3,359.32 3,166.01 0.987 0.16 
S-800 422.12 2,255 5.05 3,150.47 3,064.44 0.985 0.038 
S-1200 423.92 2,540 4.49 3,620.08 3,447.36 0.983 0.091 
S-1600 422.23 2,665 5.09 3,395.45 3,287.88 0.989 0.043 
S-2000 424.72 2,110 3.04 2,848.46 2,668.92 0.990 0.33 
Sample Mean length OTU counts Shannon index ACE index Chao1 index Coverage Simpson 
S-0 426.98 1,845 2.31 2,831.69 2,538.01 0.988 0.42 
S-400 424.85 2,299 3.81 3,359.32 3,166.01 0.987 0.16 
S-800 422.12 2,255 5.05 3,150.47 3,064.44 0.985 0.038 
S-1200 423.92 2,540 4.49 3,620.08 3,447.36 0.983 0.091 
S-1600 422.23 2,665 5.09 3,395.45 3,287.88 0.989 0.043 
S-2000 424.72 2,110 3.04 2,848.46 2,668.92 0.990 0.33 

From S-0 to S-1600, the Shannon and Simpson diversity indices showed that the bacterial diversity increased significantly (Table 2), clearly indicating that quartz sand can provide an effective living environment for bacteria. The OTU counts from S-0 to S-2000 showed that the species richness and diversity in S-1600 were noticeably higher than S-2000, which was located between quartz sand and pebbles. Furthermore, the DO level of the S-2000 layer (1.8 ± 0.08 mg L−1) was higher than that of the S-1600 layer (1.2 ± 0.11 mg L−1), which may also be a contributing factor.

The most significantly abundant phylum in the influent S-0 samples was Proteobacteria, accounting for 92.34% of the sequences, while in other samples Proteobacteria were also dominant with an average proportion of 76.71% (Figure 6). A high abundance of Proteobacteria has been previously observed in some wastewater treatment bioreactors (Ma et al. 2015). Proteobacteria dominance decreased, from 92.34% in S-0 samples to 65.23% in S-1600 samples, suggesting that DO had an adverse effect on these microbes. Previous research (Cao et al. 2017) demonstrated that the concentration of OM has a noticeable impact on Proteobacteria. In contrast, in the present study, the phyla Bacteroidetes and Chloroflexi showed a significant increase, from 2.41% and 0.17% in S-0 samples to 7.61% and 8.42% in S-1600 samples, respectively.

Figure 6

Bacterial community distribution in water samples at phylum level (September 25).

Figure 6

Bacterial community distribution in water samples at phylum level (September 25).

To gain a deeper insight into the variations in bacterial genera, a heatmap of hierarchical clustering was prepared for the 50 most abundant genera, as shown in Figure 7. Enrichment of Methyloversatilis, Sideroxydans, Longilinea, Azospira, and Hyphomicrobium was identified due to their increased abundance in the sand layer, in comparison to influent concentrations. Methylophilus and Methyloversatilis were two of the most significantly changing genera. Lu et al. (2012) previously reported that Methyloversatilis induced incomplete denitrification, reducing nitrates or nitrites to N2O instead of N2 (Lu et al. 2012). Methylophilus is a genus of aerobic Gram-negative bacteria, which can use methanol as carbon and energy source (Jenkins et al. 1987).

Figure 7

Heatmaps of bacterial genera relative abundance of water samples at different DBDF depths (25 September).

Figure 7

Heatmaps of bacterial genera relative abundance of water samples at different DBDF depths (25 September).

CONCLUSIONS

The quartz sand DBDF in this experiment showed very good nitrogen and COD removal performance (TN removal >85%, NO3-N removal >90%, and COD removal >75%). The DBDF start-up period was short, with steady running achieved within 20 days. When the influent TN concentration was below 15 mg L−1 and COD levels above 50 mg L−1, the effluent TN and COD could be stabilized to less than 5 and 30 mg L−1, respectively. 3D-EEM results show that the DBDF had a better removal capacity for DOM, especially aromatic protein-like substances, than the WWTP oxidation ditch at which it was operated. High throughput sequencing and qPCR results showed a distinctly stratified distribution of the main functional microbial species in DBDF. These findings demonstrate that quartz sand can provide an effective living environment for microbes.

ACKNOWLEDGEMENTS

This work was supported by the Major Science and Technology Program for Water Pollution Control and Treatment (2012ZX07103-004). We also thank the editor and the anonymous reviewers for their helpful comments.

REFERENCES

REFERENCES
APHA
2012
Standard Methods for the Examination of Water and Wastewater
,
21st edn
.
American Public Health Association
,
Washington, DC
,
USA
.
Cao
,
X.
,
Li
,
Y.
,
Jiang
,
X.
,
Zhou
,
P.
,
Zhang
,
J.
&
Zheng
,
Z.
2016
Treatment of artificial secondary effluent for effective nitrogen removal using a combination of corncob carbon source and bamboo charcoal filter
.
International Biodeterioration & Biodegradation
115
,
164
170
.
Fernández-Nava
,
Y.
,
Marañón
,
E.
,
Soons
,
J.
&
Castrillón
,
L.
2010
Denitrification of high nitrate concentration wastewater using alternative carbon sources
.
Journal of Hazardous Materials
173
(
1–3
),
682
688
.
Hidaka
,
T.
,
Tsuno
,
H.
&
Kishimoto
,
N.
2003
Advanced treatment of sewage by pre-coagulation and biological filtration process
.
Water Research
37
(
17
),
4259
4269
.
Jenkins
,
O.
,
Byrom
,
D.
&
Jones
,
D.
1987
Methylophilus – a new genus of methanol-utilizing bacteria
.
International Journal of Systematic Bacteriology
37
(
4
),
446
448
.
Karanasios
,
K. A.
,
Vasiliadou
,
I. A.
,
Tekerlekopoulou
,
A. G.
,
Akratos
,
C. S.
,
Pavlou
,
S.
&
Vayenas
,
D. V.
2016
Effect of C/N ratio and support material on heterotrophic denitrification of potable water in bio-filters using sugar as carbon source
.
International Biodeterioration & Biodegradation
111
,
62
73
.
Liu
,
S.
,
Zhao
,
T.
,
Zhu
,
Y.
,
Qu
,
X.
,
He
,
Z.
,
Giesy
,
J. P.
&
Meng
,
W.
2018
Molecular characterization of macrophyte-derived dissolved organic matters and their implications for lakes
.
Science of the Total Environment
616–617
,
602
613
.
Lowenberg
,
J.
,
Zenker
,
A.
,
Krahnstover
,
T.
,
Boehler
,
M.
,
Baggenstos
,
M.
,
Koch
,
G.
&
Wintgens
,
T.
2016
Upgrade of deep bed filtration with activated carbon dosage for compact micropollutant removal from wastewater in technical scale
.
Water Research
94
,
246
256
.
Ma
,
Q.
,
Qu
,
Y.
,
Shen
,
W.
,
Zhang
,
Z.
,
Wang
,
J.
,
Liu
,
Z.
,
Li
,
D.
,
Li
,
H.
&
Zhou
,
J.
2015
Bacterial community compositions of coking wastewater treatment plants in steel industry revealed by Illumina high-throughput sequencing
.
Bioresource Technology
179
,
436
443
.
MOEP
2002
Ministry of Environmental Protection of the People's Republic of China (MOEP), 2002
.
Discharge Standard of Pollutants for Municipal Wastewater Treatment Plant.GB18918-2002
.
Beijing
,
China
(in Chinese)
.
Park
,
J. B.
,
Craggs
,
R. J.
&
Sukias
,
J. P.
2009
Removal of nitrate and phosphorus from hydroponic wastewater using a hybrid denitrification filter (HDF)
.
Bioresource Technology
100
(
13
),
3175
3179
.
Tang
,
X.
,
Wu
,
Q. Y.
,
Huang
,
H.
,
Hu
,
H. Y.
&
Li
,
Q.
2013
Removal potential of anti-estrogenic activity in secondary effluents by coagulation
.
Chemosphere
93
(
10
),
2562
2567
.
Xiong
,
J.
,
Guo
,
G.
,
Mahmood
,
Q.
&
Yue
,
M.
2011
Nitrogen removal from secondary effluent by using integrated constructed wetland system
.
Ecological Engineering
37
(
4
),
659
662
.
Zhao
,
H.
,
Xu
,
X.
,
Ke
,
F.
,
Li
,
W.
,
Feng
,
M.
&
Zhang
,
H.
2013
Nitrogen removal from wastewater plant secondary effluent in a compound natural treatment system
.
Ecological Engineering
57
(
8
),
361
365
.