Nitrate and COD removal in an upflow biofilter under an aerobic atmosphere

Bioresource Technology 158 (2014) 156–160

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Nitrate and COD removal in an upflow biofilter under an aerobic atmosphere Bin Ji, Hongyu Wang, Kai Yang ⇑ School of Civil Engineering, Wuhan University, Wuhan 430072, China

h i g h l i g h t s  An aerobic biofilter packed with ceramsite was constructed for nitrate removal.  Pseudomonas stutzeri X31, an aerobic denitrifier isolate, was used as an inoculum. 

 The top section of the bioreactor possesses higher COD and NO3 —N removal rates.  Pseudomonas stutzeri and Paracoccus versutus were the most dominant bacteria.

a r t i c l e

i n f o

Article history: Received 6 January 2014 Received in revised form 6 February 2014 Accepted 8 February 2014 Available online 15 February 2014 Keywords: Continuous-upflow biofilter Nitrate removal Aerobic denitrification PCR–DGGE

a b s t r a c t A continuous-upflow submerged biofilter packed with ceramsite was constructed for nitrate removal under an aerobic atmosphere. Pseudomonas stutzeri X31, an aerobic denitrifier isolate, was added to 1 . The best the bioreactor as an inoculum. The influent NO 3 —N concentrations were 63.0–73.8 mg L results were achieved when dissolved oxygen level was 4.6 mg L1 and C/N ratio was 4.5. The maximum removal efficiencies of carbon oxygen demand (COD) and NO 3 —N were 94.04% and 98.48%, respectively at 30 °C, when the hydraulic load was 0.75 m h1. The top section of the bioreactor possessed less biofilm but higher COD and NO 3 —N removal rates than the bottom section. Polymerase chain reaction (PCR)– denaturing gradient gel electrophoresis (DGGE) technique combined with electron microscopic examination indicated P. stutzeri X31 and Paracoccus versutus were the most dominant bacteria. Amoeba sp., Vorticella sp., Philodina sp., and Stephanodiscus sp. were also found in the bioreactor. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Groundwater is an important source of municipal water supply for domestic and industrial use (Showers et al., 2008). The removal of nitrate from surface and underground waters has received increased attention on account of the wastewater discharge, deforestation of riparian zones, and the extensive use of fertilizers, etc. Increased nitrate concentrations in drinking water may be responsible for methemoglobinemia and diverse kinds of cancers in humans (Bhatnagar and Sillanpää, 2011). The nitrate removal technologies mainly include ion exchange, reverse osmosis, adsorption, electrodialysis, and biological denitrification. And the biological nitrogen removal process is one of the most commonly used technologies due to its effectiveness and relatively low cost (Haugen et al., 2002). Biological denitrification is considered to be the most economical strategy to use because it does not require post-treatment or produce by-products (Liu et al., 2012). ⇑ Corresponding author. Tel.: +86 27 61218623; fax: +86 27 68775328. E-mail address: [email protected] (K. Yang). http://dx.doi.org/10.1016/j.biortech.2014.02.025 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.

Denitrification is commonly considered to be achieved under anaerobic or anoxic conditions. However, there have been sporadic reports on aerobic denitrifiers (Kim et al., 2008a; Wang et al., 2013), which are able to conduct an aerobic respiratory process in which nitrate is gradually converted to N2. It is commonly accepted that dissolved oxygen concentration, C/N ratio, pH and temperature are the rate-limiting parameters, and the first two of them are suggested to be the major factors affecting aerobic denitrification (Bernat and Wojnowska-Baryła, 2007). As one of the popular biomembrance processes, biofilter process has gradually been adopted in the small communities, since it is simple to manage and able to remove organic materials and suspended solids simultaneously (Kim et al., 2008b). The upflow bioilter is a plug-flow reactor which follows first-level reaction kinetics. In recent years, the submerged filter has been applied to treat nitrate polluted groundwater under an anoxic atmosphere (Gomez et al., 2009; Moreno et al., 2005). Inoculation of a submerged biofilters inoculated with Hydrogenophaga pseudoflava showed better results in terms of system stability, higher superficial hydraulic loading and superficial nitrate loading rates than

B. Ji et al. / Bioresource Technology 158 (2014) 156–160

activated sludge (Moreno et al., 2005). Optimum C/N ratio was found to range from 1.08 to 2 in the filter (Cortez et al., 2009, 2011; Moreno et al., 2005) for nitrate removal. Although a promising nitrate removal efficiency of over 95% in some cases has been achieved, residual nitrite seemed to be inevitable in the treated water. The main objective of this study was to utilise aerobic denitrification to remove nitrate in an upflow submerged biofilter. This was carried out by using pure cultures of Pseudomonas stutzeri X31, which showed tolerance to high oxygen concentrations (Ji et al., 2013). 2. Methods 2.1. Reactor operation As shown in Fig. 1, the reactor was a 1800 mm high organic glass column, with 100 mm inner diameter, below which was a long handle filter nozzle in the middle of retainer plate with dozens of holes with a diameter of 0.9 mm, whose function mainly was aeration and backwashing. The bioreactor was filled with a depth of 1400 mm ceramsite filter media (3–5 mm diameter, 1730 kg m3 density, 650 m2 m3 specific surface, 52% porosity and non-uniform coefficient K80 = d80/d10 < 2.0). The system was continuously supplied with groundwater of Wuhan, China, which was pumped from the underground zone and supplemented with KNO3 and NaAc. The influent characteristics were as follows: nitrate–N, 63.0–73.8 mg L1; nitrite–N, 0.53–3.07 mg L1; phosphate, 0.4–0.7 mg L1; sulphate, 39– 91 mg L1; chloride, 53–135 mg L1; total dissolved solids, 452– 546 mg L1; turbidity, 0.92–2.65 NTU; and pH 7.0–7.5. In order to get the biofilm formation, the precultured P. stutzeri X31 isolate (7.9  107, 0.5% v/v) were inoculated into the reactor, amended with KNO3 (2 g L1), NaAc (3 g L1) and trace element solution as described (Zhang et al., 2011). Aeration was applied consistently from the bottom of the reactor for 3 days. After the inoculation phase, the reactor was operated at a low hydraulic load of 0.15 m h1 for 7 days, with influent continuously injected into the reactor. Then the hydraulic load was raised gradually to 0.75 m h1 in one week.

157

Different DO values in the reactor were acquired by controlling the aeration intensity, which were 2.7, 4.2, 5.1, 6.0 and 7.4 mg L1, when the C/N ratio was controlled at about 5. While the C/N ratios of the influent were adjusted to 1.8, 2.5, 3.5, 4.2, 4.5, and 5 by changing the amount of the carbon source and by maintaining a constant nitrogen concentration. Each operating condition was maintained for 10 days and average variations of nitrogen and COD of seven samples were determined. 2.2. Analysis methods COD, nitrate nitrogen and nitrite nitrogen of the samples were determined according to the Standard Methods (APHA, 2005). The pH of the influent was measured by an 828 Orion pH meter. Dissolved oxygen (DO) and temperature of the solution were measured with a 52 YSI DO meter. To measure the attached biomass, about 1% of ceramsite were collected from the bioreactor and dried for 2 h at 105 °C. The total weight of the dried ceramsite was measured and the weight of the attached biomass was obtained by subtracting the original weight of the ceramsite. Physiological and biochemical identification (More et al., 2012; Zhang et al., 2011) along with scanning electron microscope examination (VEGA 3 LMU, TESCAN) was used to identify the dominant bacteria preliminarily. For electron microscope examination, biofilm samples were pretreated by fixing with 2.5% pentanediol in a 0.1 M phosphate buffer, then soaked in 1% osmic acid. Afterwards, the samples were washed and dehydrated in a graded series of ethanol solutions (50%, 70%, 80%, 90% and 100%). The samples were dried by the critical point method and coated with gold. 2.3. Assessment of contaminants removal at different heights of the reactor Biomass and biofilm activity were analysed from sampling port A, B and C (Fig. 1), which were at the height of 25 cm, 75 cm and 125 cm, respectively, from the inlet of the bottom. A certain amount of packing was carefully rinsed with double-distilled water three times. Three equal parts of the biofilm were respectively added to three 500 mL flasks containing sterile oxygen-rich medium, which had the same components as the influent. Then the flask was placed on a magnetic stirrer with a medium stirring speed at 30 °C. After cultured for 1 h, the supernatant was sampled and measured in triplicate. COD and nitrate consumption rates were determined accordingly. 2.4. PCR – denaturing gradient gel electrophoresis (DGGE)

Fig. 1. Schematic diagram of the biofilter: (1) ceramsite filter media; (2) electrical heating belts; (3) temperature controller; (4) temperature sensor; (5) air compressor; (6) valve; (7) gas flowmeter; (8) sampling port A; (9) sampling port B; (10) sampling port C.

Total bacterial DNA was extracted with a genomic DNA extraction kit (TianGen, China) following the manufacturer’s instructions. PCR – DGGE of the community was performed on the partial 16S rRNA gene. The primers used for DGGE were BSF338-GC (50 -CGC CCG CCG CGC GCG GCG GGC GGG GCG GGG GCA CGG GGG GAC TCC TAC GGG AGG CAG CAG-30 ) and BSR518 (50 - ATT ACC GCG GCT GCT GG -30 ). The DGGE was performed using the D-Code Universal Detection Mutation System (Bio-Rad). 8% polyacrylamide gels having gradients of 35–70% denaturant were used to run the gel. The 100% denaturing solution strength was defined as 5.6 M urea and 32% (v/v) deionized formamide. The gel was run using TAE buffer for 9 h at 60 °C with 110 V and visualised by modified silver staining. The DGGE band was cut and the DNA was amplified. The amplification products were linked with pMD 18-T Vector and added into competent cells of Escherichia coli TOP10. Positive clones were picked and sequenced, and the 16S rRNA partial sequences were aligned with the same region of the closest relative strains available in the GenBank database by BLAST search.

158

B. Ji et al. / Bioresource Technology 158 (2014) 156–160

3. Results and discussion 3.1. Effect of DO concentration and C/N ratio As shown in Fig. 2A, DO concentrations in a range from 3.3 to 5.2 mg L1 show high nitrate and COD removal efficiencies, which slightly drops when DO concentration increases to 6.1 mg L1. The excessive aeration exacerbates erosion of the biofilm and is not favourable for the growth of the microorganisms. The aerobic nitrate removal is unlike conventional denitrification by maintaining the DO level below 1 mg L1 (Upadhyaya et al., 2010). The stable variations of the contaminants removal efficiencies reveal that this aerobic filter is easy to control for aeration. The nitrite nitrogen is very few (0.2–3.1 mg L1) during the whole phase and not detectable indeed when DO level is 4.6 mg L1. As carbon source is essential for cell growth and nitrate reduction processes, the optimal COD is a key parameter in the denitrification process. Denitrifiers grow slowly at low carbon concentrations due to the insufficient organics as electron donors; whereas their growth may be inhibited at an extremely high C/N ratio (Huang and Tseng, 2001). The results (Fig. 2B) show that NO 3 —N removal efficiency is correlated with C/N ratio, which can be simulated by a sigmoid equation (y = 101.4/(1 + 60.3 exp (1.58x)); R2 = 0.995; Chi2/DoF = 7.94; where y = NO 3 —N removal efficiency in% and x = C/N ratio). The optimum denitrification activity was obtained when the C/N ratio was 4.5, which is similar to Citrobacter diversus (Huang and Tseng, 2001).The range of the C/N ratio is higher than that of traditional filters used for anoxic denitrification (Moreno et al., 2005). The data also suggest that the microorganisms are able to adapt to temperature which ranged from 18 to 30 °C. The highest denitrication efficiencies are 91.5% and 98.4% at 18 °C and 30 °C, respectively. Enzymes related to aerobic denitrification, including nap, nir, nor and nos (Ji et al., 2013), tend to possess highest activity at 30 °C.

3.2. Nitrate and COD removal rates at different heights of the biofilter It can be concluded from Table 1 that the bottom of the reactor has much more biofilm than the top. As with COD and nitrate removal, the contributions of the microorganisms at bottom of the filter are almost four times as many as those at the top. Actually, the height effect detected on microbial diversity mostly affected the microbial richness in biofilters (Cabrol et al., 2012). Although the bottom and middle of the reactor contribute mainly to the

nutrients degradation, the biofilm from the top of the biofilter shows higher removal rate of both COD and nitrate. The possible reason is that the microbial biomass at the bottom of the biofilter includes the trapped suspended solids as well as inert biofilm. Some of the suspended solids are nondegradable and will gradually become part of the biofilm. Moreover, numerous heterotrophic microbes grow at the inlet end and compete for carbon and nitrogen source, which result in the lack of nutrient for the microorganisms in the inner part of the biofilm. Hence, the microorganisms at the top of the biofilter have more possibility to own higher removal rate of both COD and nitrate.

3.3. Microoganisms in the filter The bacteria in the biofilter are mainly bacilli and cocci (Supplementary Fig. 1A). And numerous Stephanodiscus sp. (Supplementary Fig. 1B) can be found at the bottom of the reactor, which indicates the influent is eutrophic and alkalescent. Moreover, Amoeba sp., Vorticella sp. and Philodina sp. can be found by optical microscope observation (Supplementary Fig. 2). Small amoebae are proposed as indicative of high performance N-removal (Pérez-Uz et al., 2010). Vorticella sp. and Philodina sp. make up the complex microbial community in aerobic wastewater treatment (Lee and Welander, 1996). The presence of those protozoa and metazoa, which work as indicator species for participating in ecosystem stability, suggests that the aerobic biofilter works effectively. The bacterial diversity of biofilm samples is evaluated by DGGE analysis of the amplified V3 region 16S rRNA genes. As shown in Fig. 3, about nine bands can be found on the DGGE gel. PCR products of bands A-I are cloned and sequenced. And BLAST sequence results are shown in Table 2. It indicates that nearly all the bacteria identified in the biofilm belong to the proteobacterial group, which predominates in aquatic ecosystems and commonly exists in aquifers (Iribar et al., 2008). Lane S3 shows as many bands as lane S2, which indicates that the bacteria in the bioreactor can permanently exist from 20th day to 60th day despite of slight change in amount for certain bacterium. In addition, Paracoccus versutus and P. stutzeri X31 are the most dominant bacteria, as DGGE fingerprinting can be used for relative comparison and analysis of the dominant phylotypes (Srinandan et al., 2012). It implies they are effective in nitrate removal. P. versutus Strain LYM performs perfect activity in aerobically converting over 95% nitrate (approximate 400 mg L1) to nitrogen gas (Shi et al., 2013). Paracoccus sp. strain YF1 immobilized on bamboo carbon was developed for the denitrification at 30 °C and 150 rpm

Fig. 2. Effect of the factors on the biofilter: (A) DO concentration; (B) C/N ratio.

159

B. Ji et al. / Bioresource Technology 158 (2014) 156–160 Table 1 Nitrate and COD removal rates at different heights of the biofilter. Sampling port

Biomass (g MLSS m3 packing)

COD removal rate (g COD g1 h1)

Nitrate removal rate 1 1 (g NO h ) 3 —N g

A B C

5.13 2.26 1.18

16.86 20.06 19.84

4.01 3.94 4.83

Table 2 Identifications of DGGE bands A-I. Band number

Most related sequence

Class

A

Acidovorax sp. R-25212 (AM084022) Massilia sp. LP01 (HM053474) Uncultured Rhodocyclales bacterium (HM086449) Uncultured Planctomyces sp. (JX575902) Pseudomonas sp. X31 (FJ480211) Asticcacaulis benevestitus strain Z-0023T (AM087199) Uncultured Paracoccus sp. (HQ264096) Aquaspirillum sp. 411 (AY904024) Pseudomonas versutus strain GW1 (GU111570)

Betaproteobacteria

99

Betaproteobacteria

98

Betaproteobacteria

96

Planctomycetia

95

B C D E F G H I

Gammaproteobacteria

Similarity (%)

100

Alphaproteobacteria

96

Alphaproteobacteria

97

Betaproteobacteria

99

Alphaproteobacteria

99

Furthermore, the reactor is easy to operate because only one basin is required and the change of DO values in aerobic zone has no huge impact on the contaminant removal efficiency. Therefore, it can be one of the options for biological nitrate removal. 4. Conclusion The continuous-upflow submerged biofilter inoculated with an aerobic isolate worked effectively to remove NO 3 —N as well as COD under an aerobic atmosphere with few nitrite accumulation. The appropriate DO concentration and C/N ratio were 4.6 mg L1 and 4.5, respectively, with a hydraulic load of 0.75 m h1. PCR– DGGE technique combined with electron microscopic examination showed P. versutus and P. stutzeri X31 were the dominant aerobic denitrifiers. The results of the present study can be used for reference for nitrate removal. Future research should aim to combine several kinds of aerobic dinitrifiers to cope with real nitrate-contaminated groundwater. Acknowledgements

Fig. 3. DGGE profile of the 16S rRNA gene fragments. Lanes: S1, pure culture of X31; S2, biofilm sample on 20th day; S3, biofilm sample on 60th day.

(Liu et al., 2012). P. stutzeri X31 performs nitrate respiration without accumulation of nitrite (less than 2.12 mg L1) as an intermediate and possesses high oxygen tolerance (Ji et al., 2013), and the similar results were reported such as P. stutzeri SU2 (Su et al., 2001) and P. stutzeri YZN-001 (Zhang et al., 2011) as aerobic denitrifiers. The dominant bacteria in mixed cultures, P. versutus and P. stutzeri X31, ensure that the nitrite is very few during the denitification. Considering the high activity of denitrificaton genes of P. versutus and P. stutzeri X31, it can be concluded that the nitrate is mostly converted to nitrogen gas and there is no nitrite buildup. It indicates that this upflow biofilter inoculated with P. stutzeri X31 under an aerobic atmosphere is effective for the growth of aerobic denitrifiers as well as other various microorganisms. It merits our attention that this biofilter is efficiency for nitrate and organics removal. The results show that when influent is of about 70 mg L1 1 NO COD, the reactor works most effectively. 3 —N and 350 mg L

This research was financially supported by the open fund of State Key Lab of Urban Water Resources and Environment (HIT) (No. QA200810; QAK201014), the National Natural Science Foundation of China (NSFC) (No. 51208397; 51378400), and Major Science and Technology Program for Water Pollution Control and Treatment (No. 2009ZX07317-008-003). 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.biortech.2014.02. 025. References Apha, A.W., 2005. Standard Methods for the Examination of Water and Wastewater. American Public Health Association, Washington, DC. Bernat, K., Wojnowska-Baryła, I., 2007. Carbon source in aerobic denitrification. Biochem. Eng. J. 36, 116–122. Bhatnagar, A., Sillanpää, M., 2011. A review of emerging adsorbents for nitrate removal from water. Chem. Eng. J. 168, 493–504. Cabrol, L., Malhautier, L., Poly, F., Lepeuple, A.S., Fanlo, J.L., 2012. Bacterial dynamics in steady-state biofilters: beyond functional stability. FEMS Microbiol. Ecol. 79, 260–271.

160

B. Ji et al. / Bioresource Technology 158 (2014) 156–160

Cortez, S., Teixeira, P., Oliveira, R., Mota, M., 2009. Effect of operating parameters on denitrification in an anoxic rotating biological contactor. Environ. Technol. 30, 1381–1389. Cortez, S., Teixeira, P., Oliveira, R., Mota, M., 2011. Denitrification of a landfill leachate with high nitrate concentration in an anoxic rotating biological contactor. Biodegradation 22, 661–671. Gomez, M.A., Gonzalez-Lopez, J., De La Rua, A., 2009. Effect of temperature over start-up of a groundwater-denitrifying submerged filter inoculated with psychrotolerant bacteria.. J. Environ. Sci. Health A Tox. Hazard. Subst. Environ. Eng. 44, 1298–1305. Haugen, K.S., Semmens, M.J., Novak, P.J., 2002. A novel in situ technology for the treatment of nitrate contaminated groundwater. Water Res. 36, 3497–3506. Huang, H.K., Tseng, S.K., 2001. Nitrate reduction by Citrobacter diversus under aerobic environment. Appl. Microbiol. Biotechnol. 55, 90–94. Iribar, A., Sánchez-Pérez, J.M., Lyautey, E., Garabétian, F., 2008. Differentiated freeliving and sediment-attached bacterial community structure inside and outside denitrification hotspots in the river–groundwater interface. Hydrobiologia 598, 109–121. Ji, B., Wang, H., Yang, K., 2013. Tolerance of an aerobic denitrifier (Pseudomonas stutzeri) to high O2 concentrations. Biotechnol. Lett.. http://dx.doi.org/10.1007/ s10529-013-1417-x. Kim, M., Jeong, S., Yoon, S.J., Cho, S.J., Kim, Y.H., Kim, M.J., Ryu, E.Y., Lee, S., 2008a. Aerobic Denitrification of Pseudomonas putida AD-21 at Different C/N Ratios. J. Biosci. Bioeng. 106, 498–502. Kim, Y., Tanaka, K., Lee, Y., Chung, J., 2008b. Development and application of kinetic model on biological anoxic/aerobic filter. Chemosphere 70, 990–1001. Lee, N.M., Welander, T., 1996. Reducing sludge production in aerobic wastewater treatment through manipulation of the ecosystem. Water Res. 30, 1781– 1790. Liu, Y., Gan, L., Chen, Z., Megharaj, M., Naidu, R., 2012. Removal of nitrate using Paracoccus sp. YF1 immobilized on bamboo carbon. J. Hazard. Mater. 229–230, 419–425.

More, T.T., Yan, S., John, R.P., Tyagi, R.D., Surampalli, R.Y., 2012. Biochemical diversity of the bacterial strains and their biopolymer producing capabilities in wastewater sludge. Bioresour. Technol. 121, 304–311. Moreno, B., Gómez, M.A., González-López, J., Hontoria, E., 2005. Inoculation of a submerged filter for biological denitrification of nitrate polluted groundwater: a comparative study. J. Hazard. Mater. 117, 141–147. Pérez-Uz, B., Arregui, L., Calvo, P., Salvadó, H., Fernández, N., Rodríguez, E., Zornoza, A., Serrano, S., 2010. Assessment of plausible bioindicators for plant performance in advanced wastewater treatment systems. Water Res. 44, 5059–5069. Shi, Z., Zhang, Y., Zhou, J., Chen, M., Wang, X., 2013. Biological removal of nitrate and ammonium under aerobic atmosphere by Paracoccus versutus LYM. Bioresour. Technol. 148, 144–148. Showers, W.J., Genna, B., McDade, T., Bolich, R., Fountain, J.C., 2008. Nitrate contamination in groundwater on an urbanized dairy farm. Environ. Sci. Technol. 42, 4683–4688. Srinandan, C.S., D Souza, G., Srivastava, N., Nayak, B.B., Nerurkar, A.S., 2012. Carbon sources influence the nitrate removal activity, community structure and biofilm architecture. Bioresour. Technol. 117, 292–299. Su, J.J., Liu, B.Y., Liu, C.Y., 2001. Comparison of aerobic denitrification under high oxygen atmosphere by Thiosphaera pantotropha ATCC 35512 and Pseudomonas stutzeri SU2 newly isolated from the activated sludge of a piggery wastewater treatment system. J. Appl. Microbiol. 90, 457–462. Upadhyaya, G., Jackson, J., Clancy, T.M., Hyun, S.P., Brown, J., Hayes, K.F., Raskin, L., 2010. Simultaneous removal of nitrate and arsenic from drinking water sources utilizing a fixed-bed bioreactor system. Water Res. 44, 4958–4969. Wang, P., Yuan, Y., Li, Q., Yang, J., Zheng, Y., He, M., Geng, H., Xiong, L., Liu, D., 2013. Isolation and immobilization of new aerobic denitrifying bacteria. Int. Biodeter. Biodegr. 76, 12–17. Zhang, J., Wu, P., Hao, B., Yu, Z., 2011. Heterotrophic nitrification and aerobic denitrification by the bacterium Pseudomonas stutzeri YZN-001. Bioresour. Technol. 102, 9866–9869.