- Email: [email protected]
Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier
G Model JIEC 4078 No. of Pages 7
Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier N.V. Tuyena , J.H. Ryua , J.B. Yaeb , H.G. Kimb , S.W. Honga,b , D.H. Ahna,b,* a b
Department of Environmental Engineering and Energy, Myongji University, Yongin, Gyeonggi-do 17058, Republic of Korea Bluebank Co., Ltd., Myongji University, #18109-1, Business Incubator Center 116, Myongji-ro, Cheoin-gu, Yongin-si, Gyeonggi-do 17058, Republic of Korea
A R T I C L E I N F O
Article history: Received 1 February 2018 Received in revised form 25 June 2018 Accepted 2 July 2018 Available online xxx Keywords: Polyvinyl alcohol Anaerobic ammonium oxidation Sodium sulfate Immobilization Nitrogen removal
A B S T R A C T
In this study, the result shows that polyvinyl alcohol–sodium alginate (PVA–SA) gel bead crosslinked with sodium sulfate are better among the different methods by comparing the relative mechanical strength, mechanical strength swelling and expansion coefficient of beads in water. Subsequently, anammox biomass entrapment by PVA–SA gel was introduced into continuous stirred tank reactor (CSTR). After 24 operation days, the nitrogen removal efficiency achieved 60%, while the nitrogen loading rate (NLR) was 0.14 kg N/m3/d and the experiment data indicated that PVA–SA gel bead crosslinked with sodium sulfate can be used to initiate anammox process. Furthermore, it is an alternative for culturing anammox in a long-term operation. © 2018 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Since the anammox process was identified in a denitrifying fluidized bed reactor in deft, evidence for anammox activity has been obtained in a variety of laboratory and engineered systems, pilot plants for ammonium removal, surface and subsurface sediments of deep ocean, temperate estuarine, coastal and offshore sediments, lakes, freshwaters, polar region sediments and multiyear sea ice, oil reservoirs, deep sea hydrothermal vent and the sub-oxic zone of the Black Sea [1–4]. The anammox process is considered having enormous potential in the treatment of high nitrogen content wastewater such as landfill leachate, wastewater from seafood processing industries, etc. In this process, ammonium is oxidized with nitrite serving as the electron acceptor under anaerobic condition, producing nitrogen gas [1] with CO2/ bicarbonate as the sole carbon source like the following formula: 1NH4+ + 1,32NO2 + 0,066HCO3 + 0,13H+ → 1,02N2 + 0,26NO3 + 0,066CH2O0,5N0,15 + 2,03H2O
(1)
However, there are still some important challenges in practical application of anammox because of its slow growth rate and unstable start-up period. Therefore, the maintenance of a sufficient
* Corresponding author at: Department of Environmental Engineering and Energy, Myongji University, Yongin, Gyeonggi-do, 17058, Republic of Korea. E-mail address: [email protected] (D.H. Ahn).
amount of anammox bacteria during the start-up is one of the most essential step with an anammox reactor. Immobilization of microorganism in spherical polymeric matrices is promising for improvement of the efficiency of bioprocess, especially in the production of metabolites and biological treatment of wastewater. Advantages of the immobilized cells compared with free cells include: protection from harsh environment conditions such as pH, temperature, organic solvent and toxic compound, relative ease of product separation, reusability, increased cell density, and reduced susceptibility to contamination by foreign microorganism. Recently polyvinyl alcohol (PVA) has been used widely in the field of wastewater treatment, these advantages include lower cost, high durability and chemical stability, non-toxicity to viable cells [5–7], nonbiodegradable, high mechanical strength and long-life span of gel. Moreover, with porosity, PVA gel has a large amount of water and allows for effective transition of oxygen and nutritional values inside the gel beads. With a porous microstructure, PVA gels bead can be an efficiency way to anammox reactor, it allows the microorganism to live inside the polymeric matrix, protects the anammox bacteria from inhibition factor in surrounding environment like oxygen and substrates still can diffuse to the bacteria, prevents anammox biomass from being washout through the effluent. Furthermore, PVA has several advantages including low cost, high durability and chemical stability, and especially non-toxic to microorganism [8,9]. However, the boric acid technique is often used to produce PVA gel bead and it also has drawbacks
https://doi.org/10.1016/j.jiec.2018.07.004 1226-086X/© 2018 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
2
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
that the bacteria in the PVA matrix can be damaged by boric acid during the process and has a tendency to agglomerate [10]. To eliminate these problems, this study focused on the enrichment and performance of anammox bacteria in a continuous stirred tank reactor (CSTR) with the gel beads which are made by boric acid method with sodium sulfate. The combination of PVA and sodium alginate (SA) was used to improve the surface properties of the beads, reducing the tendency to agglomerate, while the PVA–SA gel bead crosslinked with sodium sulfate has lower toxicity to bacteria than boric acid, higher stability in water and greater biological activity than conventional PVA–boric acid hydrogels bead [11]. After that, the characteristic of nitrogen removal with anammox biomass entrapped in PVA–SA gel bead and anammox performance in the CSTR are evaluated to develop the anammox process in conventional CSTR for further application in wastewater treatment. Material and methods Seed sludge For experiment, the active sludge from a wastewater treatment plant operated under anoxic/anaerobic and aerobic conditions is used. (MLSS = 20–25 g/L, MLVSS = 15–20 g/L). The sludge is washed with distilled water and PBS solution to remove the residual substrate before using. Anammox sludge used in this study was obtained from a lab scale reactor. Sludge was washed three time with phosphate buffered saline (PBS, pH = 7.4, 0.1 M) solution to remove any residual substrate on the sludge surface. Afterwards, the sludge was centrifuged for 5 min at 2000 rpm. Immobilization methods The seed sludge prepared from the previous step was encapsulated in polymer gel by different methods. 200 mL of solution containing 12% PVA (w/v) and 2% sodium alginate (SA) (w/v) was autoclaved at 121 C for 20 min. After cooling to 37 C, this solution is mixed with 200 mL of seed sludge completely. The mixture is dropped into a solution of saturated B(OH)3 and CaCl2 2% (w/v) to form spherical beads. Then, with the method A, the beads are cursed in B(OH)3/CaCl2 solution for 1 h, transferred to 0.5 M Na2SO4 solution and cursed for 1 h [12]. With method B, the beads are cursed in B(OH)3/CaCl2 solution for 3.5 h, transferred to 0.5 M KH2PO4 solution and cursed for 5.6 h. In method C, the mixture is dropped into CaCl2 4% (w/v) solution and cursed in the same solution for 12 h [11]. Finally, all of beads are removed, washed and stored at 4 C with distilled water before using.
Relative strength 20 immobilized granules of similar size of each material type were selected and placed in a 100 mL beaker with 80 mL of DI water. The beads were stirred by Thermo ScientificTM CimarecTM Digital Stirring Hotplates while the speed is controlled from 500 to 2500 rpm and the intact gel beads is counted daily [10]. Swelling of bead in water 20 immobilized granules of similar size of each material type were selected and placed in a 100-mL beaker with 80 mL of DI water. The beaker was mixed for 7 days by magnetic stirrer machine and the size of gel beads were checked daily. Expansion coefficient 20 immobilized granules of similar size of each material type were selected and placed in a 100-mL beaker with 80 mL of DI water. The beaker was slowly shaken at 35 C for 72 h (exactly same with reactor operation in the next step) then the diameter of the gel beads before and after were measured. The ratio of the mean diameter of the gel beads after 72 h of treatment to the mean diameter of the original gel bead was the expansion coefficient [13]. Reactor and experiment method A continuous stirred tank reactor with total working volume of 3 L and 30% packing ratio of gel beads is used with synthetic wastewater. The temperature is maintained at36 C by using water jacket and pH is controlled around 7.5–8 by HCL 1 M. Stirring speed is set to 100 rpm. Stirring was essentially required to mix the influent and remove nitrogen gas bubble which formed on the surface of the gel beads [14]. The influent was supplied by an inflow peristaltic pump from the synthetic wastewater tank through a feed line tubing (Figs. 1 and 2). The CSTR reactor was operated with a hydraulic retention time ranged between 12–24 h, with the nitrogen loading rate from 0,09 to 0,134 kg N/m3/d. The synthetic wastewater used through the experiments contained (per lit): NH4+-N 50–60 mg, NO2 -N 40– 50 mg, KHCO3 500 mg, KH2PO4 27,2 mg, 0.3 g of CaCl22H2O, 0.2 g of MgSO47H2O, 0.00625 g of FeSO4, 0.00625 g of EDTA and 1 mL/L of trace element solution. Trace element solution was composed of ZnSO47H2O, 0.43 g; CoCl26H2O, 0.24 g; MnC124H2O, 0.99 g; CuSO45H2O, 0.25 g; NaMoO42H2O, 0.22 g; NiCl26H2O, 0.19 g; NaSeO410H2O, 0.21 g; H3BO40,014 g (1 mL/L). The effluent sample was collected daily to evaluate the treatment performance. According to the standard method, NO2 -N, NH4+-N, NO3 -N were measured by kit HACH and standard method. The pH was measured using a pH-meter and DO was measured by a portable meter (Figs. 1 and 2) Results and discussion
Characterization of the PVA–SA gel beads Characterization of the PVA–SA gel bead Different methods were applied to make PVA–SA gel beads. In order to determine the better method, relative mechanical strength, mechanical strength swelling of beads in water and expansion coefficient were used following the most suitable gel beads is chosen to culture anammox bacteria in CSTR. Mechanical strength 20 immobilized granules of similar size of each material type were selected and placed in a 100 mL beaker. Then, 80 mL of deionized water (DI water) was added to the beaker. The mixture was magnetically stirred at 500 r/m for 24 h by using Thermo ScientificTM CimarecTM Digital Stirring Hotplates, after which the ratio of intact gel bead to the original number of immobilized gel beads was determined [13].
The Table 1 lists the data of the mechanical strength and expansion coefficient of 4 types of gel bead in this study. Following this table, the mechanical strength of method C was lowest, after 24 h of experiment and only 47% of gel beads were intact. With method A, PVA–SA gel beads crosslinked with sodium sulfate, the mechanical strength of these beads were 0.87, many times higher than that of method C, and close to that of method B, 0.73. Gel bead of method A has highest mechanical strength, however it’s expansion coefficient is lowest, only 1.17. The expansion coefficient of method C was 2, and the size of gel bead increases from 3 mm to 6 mm in day under specific condition of water, resulting in gel beads becoming very soft and easy to dissolve in the water following the raising of the rate of stirring machine. Takei et al. also
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
3
Fig. 1. Anammox reactor schematic diagram.
compared the mechanical strength and expansion coefficient of SA and PVA–SA gel beads made by method C [12]. The result revealed that PVA–SA are stronger and better than SA gel bead. Furthermore, in this experiment, 2 other methods for making PVA–SA gel bead were used, and they both have better characteristic compared to PVA–SA made by method C. Bao-e and Yong-you also reported that the strength of PVA–SA bead is better than SA, however the saturated wet density of SA is higher than that of PVA–SA [15], resulting in a much higher expansion coefficient of PVA–SA beads in comparison with SA bead. With the addition of SA to PVA solution, the hydroxyl on the PVA molecular chain and the carboxyl on the SA molecular chain can form a strong hydrogen. Moreover, the SA macromolecules can diffuse the charge density of PVA macromolecules, so that electrostatic interaction between different PVA molecular chain is attenuated. Thus, the PVA–SA bead is stronger and more stable. Beside mechanical strength, relative mechanical strengths are also checked by increasing the rate of stirring machine. Following
Fig. 3, feeble mechanical strength of the bead of method C was identified, when 50% of beads were broken with 700 rpm of mixing speed and even no intact bead was collected with 1200 rpm. On the contrary, beads of methods A and B were much stronger than that of method C. There was no broken bead even though the mixing speed was 1200 rpm. However, 80% of the latter was broken with the speed over 2500 rpm opposed to 60% of the former. Subsequently, the swelling of bead incubated in water was investigated (Table 2) because polymer matrix for bioremediation of wastewater is required to have high stability in water. With the bead of method C, about 60% of bead was broken during incubation in water for 12 days. On the other hand, all the beads of methods A and B maintained their spherical shapes for 7 days, although the beads of method B became soft. In the first 3 days, the dimension of method C beads increased quickly from 3 mm to 5 mm while the method B increased from 3 mm to 4 mm and 3.5 mm was the size of method A bead. Until day 7, size of bead of methods B and C was bigger than bead of methods A which was be swelling little and did
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
4
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx Table 2 Swelling of bead in water.
Fig. 2. PVA–SA gel beads in the reactor with 30% packing ratio.
Table 1 Mechanical strength and expansion coefficient. Methods
Mechanical strength
Expansion coefficient
Method A Method B Method C
0.87 0.73 0.47
1.17 1.3 2
not be broken in water. Although, the method B bead did not solute in water, but they were swelling in the water. Swelling in water make structure of bead of method C become very weak. The differences between the stability and mechanical strength of the gel beads in the water were due to the differences of the
Days
Method A
Method B
Method C
0 1 2 3 4 5 6 7
3 mm 3 mm 3 mm 3.5 mm 4 mm 4 mm 4 mm 4 mm
3 mm 3.5 mm 3.5 mm 4 mm 4 mm 4 mm 5 mm 5 mm
3 mm 4 mm 4.5 mm 5 mm 5.5 mm 5.5 mm 6 mm 6 mm
making procedures. In method C, the mixture of PVA–SA solution and seed sludge was dropped into the CaCl2 solution for making gel bead. Following this way, the macro-network structure of SA can improve the mass transfer property of PVA and the conglutination behaviour of PVA gel bead [13,16,17]. However, because of presence of SA, the gel bead will become soft, swollen and easy to be broke, while Ca2+ only can improve the surface properties of the bead and reduce the tendency to agglomerate [18]. The advantage of this method is that it is not harmful to the immobilized cell, however, for long time operation, the gel bead need to be strong enough to keep the cell inside polymer matrix. To avoid the problem of boric acid, other types of PVA–SA beads crosslinked with potassium orthorphosphate or sodium sulfate have been reported and these compounds were used in methods B and A. After being dropped into the boric acid and CaCl2 solution, the bead was immersed in sodium sulfate solution. Then, the sulfate ions penetrated into the PVA bead and destabilized hydrogen bonding between hydroxyl groups of PVA and water molecules by polarizing the water molecules. As a result, sulfate ions possess the ability to form linkages among the crosslinked PVA [5], resulting in PVA crystallite formation. In addition, Takei et al. also suggested that [19] sulfate induces a higher degree of crosslinking of PVA due to its higher hydration force compared with phosphate. These information are suitable with the data of this experiment. The swelling of bead, mechanical strength and expansion coefficient of method A bead are better than that of bead of method B. In addition, the biological activity of conventional PVA–boric acid beads drastically decreasing was that surviving cells after immersion in saturated boric acid solution had more severe injuries, which required time for recovery, than the cells incubated in sodium sulfate solution [12]. For these reasons, sodium sulfate can be more useful than sodium phosphate for preparation of PVA–SA gel bead and is an efficient way to immobilize anammox biomass for applying in a CSTR reactor. Therefore, in the next step of experiment, PVA–SA gel beads crosslinked with sodium sulfate was chosen to apply in the CSTR reactor. Anammox start-up and stoichiometric ratios
Fig. 3. Relative mechanical strengths of PVA–SA gel beads made by methods A–C.
The start-up of anammox is a process with the anammox reaction gradually being dominant in the reactor, which can be reflected by some indicators such as molar ratios of consumed NO2 -N to consumed NH4+-N and produced NO3 -N to consumed NH4+-N [20]. This study was inoculated with a mixture of active sludge and anammox sludge immobilized in PVA–SA gel beads with an NLR of 0.1 kg N/m3/d (NO2 -N/NH4+-N feeding ratio equal to 0.8) (Fig. 4). In the first 15 days of experiment, the nitrogen removal rate (NRR) was 0.03 kg N/m3 d, while the NO2 -N/NH4+-N ratio and NO3 -N/NH4+-N ratio were much lower than the theoretical stoichiometric ratios which were reported to be 1.32 and 0.26 respectively [21]. The ratio of produced NO3 -N/ consumed NH4+-N was lower than the theoretical value, indicating a low nitrate generation (Fig. 5). This might be attributed to the continuous happening of the heterotrophic denitrification process
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
Fig. 4. Evaluation of nitrogen loading rate (NLR) and nitrogen removal rate (NRR) during 80 operation days.
relying on the lysis products in the reactor. Otherwise, anammox has very slow growth rate compared with the other types of bacteria such as AOB, NOB. Hence, other bacteria were dominant with anammox bacteria in the first 15 days. Until day 20, when the NLR was 0.11 kg N/m3/d, the NRR increased to 0.07 kg N/m3/d. Moreover, the NO2 -N/NH4+-N ratio and NO3 -N/NH4+-N ratio were 1.1 and 0.2. This indicated that the transformations are in agreement with the anammox reaction, implying the dominance of the anammox process. The stoichiometric ratios of anammox obtained were quite stable during the next 10 days. At day 30, the NLR was increased to 0.18 kg N/m3/d in order to make anammox bacteria grow faster. Although the NRR was 0.07 kg N/m3/d, same as previous days, however, the NO2 -N/NH4+-N ratio decreased sharply to 0.4. This data reveals that the NOB could grow very fast in this period of experiment and it would affect anammox bacteria if there had been no adjustment. Thereby, the NLR was decreased to initial value and kept unchange. However, there had not been change until the HRT was decreased from 24 h to 12 h. As a result, the NLR rose to 0.25 kg N/m3/d (NO2 -N/NH4+-N feeding ratio equal to 0.8) and NRR rose to 0.12 kg N/m3/d. Especially, the NO2 N/NH4+-N ratio rose fast to 0.8. This trend implies that the growth of NOB was inhibited and anammox bacteria was becoming dominant in the reactor.
Fig. 5. Stoichiometric ratios of NO2 -N/NH4+-N and NO3 -N/NH4+-N of CSTR during 80 operation days.
5
Fig. 6. Influent, effluent concentration and removal efficiency of NO2 -N during 80 operation days.
Nitrogen removal performance Nitrite concentration Fig. 6 shows the influent and effluent of nitrite during the experiment, which lasted 80 days. The concentration of nitrite during the startup has crucial importance for growth of anammox: a too low amount of it will result in substrate limitation and thus slower growth, while concentration above 50–150 mg/L can already lead to inhibition [22–24]. From the beginning of the experiment, the influent concentration of nitrite was about 40 mg/ L to prevent the inhibition of nitrite with anammox bacteria in the bead. There is no day that the effluent of Nitrite is higher than the influent, it indicates that nitrite oxidation bacteria presented in the gels beads and converted NO2 -N to NO3 -N. Afterwards, the influent concentration of NO2 -N was increased to 50–57 mg/L for encouraging the development of anammox inside the bead. At this time, the effluent of NO2 -N reduced from about 30 mg/L at the beginning to 12.5 mg/L with 78% of NO2 -N removed from the water in day 18. Thus, the influent NO2 -N was increased to 78 mg/ L. However, the effluent NO2 -N was very high and the removal efficiency was only 20%. In this case, the anammox bacteria could be inhibited by high concentration of nitrite in wastewater while they were not strong enough. Therefore, the influent NO2 -N was reduced to about 40 mg/L to recovery the activity of anammox bacteria in the reactor. At day 59, the effluent was about 15 mg/L and the removal efficiency was achieved 70%. The influent was slightly increased and maintained at 50 mg/L. Although in this period, there was a decrease of HRT from 24 h to 12 h, from day 70 to 80, the efficiency removal was stable around 50–60%. This result show that, the anammox bacteria could adapt with synthetic wastewater, and was retained inside the gel beads. Ammonium and nitrate concentration Figs. 7 and 8 give data about changes in concentration of NH4+N and NO3 -N during the experiment. In the start-up period of the experiment, the influent nitrite-to-ammonium molar ratio was maintained at 0.76 mol NO2 -N/NH4+-N (lower than the theoretical stoichiometric ratio) to avoid possible toxicity caused by nitrite accumulation that it could inhibit the anammox process [25]. Hence, the influent concentration of NO2 -N was 40–50 mg/L, along with 50–60 mg/L of NH4+-N. In addition, it is important that the synthetic wastewater contained sufficient nitrate concentration to prevent sulphate reduction to sulphite which has been demonstrated to be toxic to anammox bacteria [25–28]. For this reason, 6–7 mg/L of the nitrate influent concentration was observed. In the first 10 days of experiment, the effluent
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
6
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
Fig. 7. Influent and effluent concentration of NO3 -N during 80 operation days.
constant oxygen saturation for AOB is 0.3 mg/L lower than NOB (1 mg/L) [31–33]. Other one has revealed that many AOB species are able to adapt their metabolism even in acute absence of oxygen [34]. These data explained the change of NH4+-N concentration in the experiment beside anammox bacteria. Similar with influent of NO2 -N, from day 30 to 40, the influent NH4+-N was increased to 90 mg/L. Then, the effluent NH4+-N was 30 mgN/L and removal efficiency was 60%. There were not many differences of these data from previous days while NO3 -N effluent rose dramatically from 10 to 20 mg/L. On the other hand, NOB bacteria also can grow under oxygen limitation, along with the increase of NO2 -N which is electron donor, NOB developed in these days. From day 40, for anammox bacteria reactivity after being inhibited under high concentration of NO2 -N, the influent NH4+-N was maintained at 60 mg/L. There was a sharp drop in the effluent NH4+-N, which was lower than 10 mg/L and the removal efficiency was 80%, along with an increase of 25 mg/L of NO3 -N in the effluent. The reason of this trend is that the AOB and NOB predominated over the anammox bacteria. From day 66, the HRT was adjusted to 12 h for keeping the optimal ratio of influent NO2 N/NH4+-N and preventing the accumulation of NO2 -N which can inhibit the anammox bacteria. Following this, from day 68, the effluent NO3 -N was 15 mg/L, the effluent NH4+-N rose to 20 mg/L. This result indicated that NOB growth was controlled which can lead to the development of anammox bacteria. Conclusion
Fig. 8. Influent, effluent concentration and removal efficiency of NH4+-N during 80 operation days.
concentration of NH4+-N was relatively high at about 50 mg/L and close to influent and the removal efficiency was about 20%. This data has been due to the anammox bacteria and other bacteria have not adapted with the synthetic wastewater yet, whereas the concentration of NO3 -N also indicated that there was problem with the bacteria in the reactor. While the influent NO3 -N was 6 mg/L, the effluent was 5 mg N/L. There was a fall in the NO3 -N concentration because of endogenous heterotrophic denitrification. Although, the organic carbon source was not introduced in the synthetic wastewater which is important for heterotrophic bacteria. However, at the beginning of experiment, there was a small number of broken gel beads. As a result, the SS of effluent increased, the bacteria in the broken bead could die when they contact with the new environment of wastewater [29,30]. After that, the cell lysis happened and released the organic compound which could be later consumed by heterotrophic bacteria as carbon source for denitrification process. From day 12 to 30 there was no more broken bead, the effluent NO3 -N became higher than influent and was around 10–15 mg/L. At this time, the ammonium removal efficiency rose gradually and achieved 66% at day 30, the effluent NH4+-N was 23 mg/L. Without DO supply, the DO concentration in reactor maintained lower than 0.5 mg O2/L. But, the removal efficiency of NH4+-N was still high compared with NO2 -N removal efficiency, even though AOB bacteria requires oxygen as substrate. Some researcher reported that the Ks value of
In this study, different type of method had been used to make PVA–SA gel bead and the properties of these gel beads were also checked. Lastly, PVA–SA gel beads crosslinked with Sodium sulfate is high mechanic strength, non-solute in the water, short immersion time for solidification compared with other methods. It can be suitable for anammox immobilization and applying in the CSTR reactor. After 80 days of this experiment, the volume of gel bead still occupied nearly 30% of reactor and the beads were strong. The total nitrogen removal was achieved 50–60% after 25 operation days with. Stoichiometric ratios of consumed NO2 -N to consumed NH4+-N and produced NO3 -N to consumed NH4+-N in the experiment are close to the theoretical stoichiometric ratios of anammox reaction. In conclusion, this data indicated that PVA–SA gel beads crosslinked with Sodium sulfate can be used to fast startup of anammox process in a short time. However, further research about reactor configuration and sufficient amount of seed anammox biomass for better nitrogen removal performance of anammox bacteria immobilized with PVA–SA gel beads crosslinked with sodium sulfate need to be done. Acknowledgement This subject is supported by Korea Ministry of Environment as “Global Top Project” (Project No.: 2016002190006). References [1] A. Mulder, A.A. van de Graaf, L.A. Robertson, et al., FEMS Microbiol. Ecol. 16 (1995) 177. [2] H. Wu, G. Zeng, J. Liang, J. Chen, J. Xu, et al., Int. J. Appl. Earth Obs. Geoinf. 56 (2017) 36. [3] H. Wu, G. Zeng, J. Liang, et al., Ecol. Eng. 57 (2013) 72. [4] H. Wu, G. Zeng, J. Liang, et al., Ecol. Indic. 53 (2015) 129. [5] N.A. Mohd Zain, M.S. Suhaimi, A. Idris, Process Biochem. 46 (2011) 2122. [6] K.W. Jung, M.J. Hwang, T.U. Jeong, D.M. Chau, J. Ind. Eng. Chem. (2017) 1. [7] G. Zeng, H. Wu, J. Liang, et al., RSC Adv. 5 (2015) 34541. [8] R. Surudzic, A. Jankovic, M. Mitric, I. Matic, Z.D. Juranic, L. Zivkovic, V. Miskovic-Stankovic, K.Y. Rhee, S.J. Park, D. Hui, J. Ind. Eng. Chem. (2015). [9] M. Shafiq, A. Sabir, A. Islam, S.M.M. Khan, S.N. Hussain, M.T.Z.Z. Butt, T. Jamil, J. Ind. Eng. Chem. (2016) 1. [10] P.V. Dinh, L.T. Bach, Int. J. Sci. Eng. 7 (2014) 41.
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx [11] M. Ali, M. Oshiki, L. Rathnayake, S. Ishii, H. Satoh, S. Okabe, Water Res. 79 (2015) 147. [12] T. Takei, K. Ikeda, H. Ijima, K. Kawakami, Process Biochem. 46 (2011) 566. [13] G. Chen, J. Li, S. Tabassum, Z. Zhang, RSC Adv. 5 (2015) 25372. [14] L.M. Quan, D.P. Khanh, D. Hira, T. Fujii, K. Furukawa, Biodegradation 22 (2011) 1155. [15] W. Bao-e, H. Yong-you, J. Environ. Sci. 19 (2007) 451. [16] G.P. Lim, M.S. Ahmad, J. Ind. Eng. Chem 56 (2017) 382. [17] H. Wu, G. Zeng, J. Liang, et al., Appl. Microbiol. Biotechnol. 100 (2016) 8583. [18 Kuo-Ying Amanda Wu, Keith D. Wisecarver, Biotechnol. Bioeng. (1992) 447. [19] T. Takei, K. Ikeda, H. Ijima, M. Yoshida, K. Kawakami, Polym. Bull. 69 (2012) 363. [20] Y.F. Lu, L.J. Ma, L. Ma, B. Shan, J.J. Chang, Environ. Technol. 39 (2017) 59. [21] M. Strous, J.J. Heijnen, J.G. Kuenen, M.S.M. Jetten, Appl. Microbiol. Biotechnol. 50 (1998) 589. [22] S. Suneethi, S. Shalini, K. Joseph, Int. J. Waste Resour. 4 (2014). [23] S. Uyanik, Ozan K. Bekmezci, A. Yurtsever, Clean – Soil Air Water 39 (2011) 653.
7
[24] H. Wu, C. Lai, G. Zeng, et al., Crit. Rev. Biotechnol. 37 (6) (2017) 754. [25] H. Lopez, S. Puig, R. Ganigue, M. Ruscalleda, M.D. Balaguer, J. Colprim, J. Chem. Technol. Biotechnol. 83 (2008) 233. [26] X. Ma, Y. Wang, S. Zhou, Y. Yan, X. Lin, M. Wu, Chem. Eng. J. 313 (2016) 1233. [27] S.E. Vlaeminck, J. Geets, H. Vervaeren, N. Boon, Willy Verstraete, Environ. Biotechnol. 74 (2007) 1376. [28] L.G.J.M.V. Dongen, M.S.M. Jetten, M.C.M.V The Combined Sharon/Anammox Process Loosdrecht, IWA Pulishing (2001). [29] S. Jenni, S.E. Vlaeminck, E. Morgenroth, K.M. Udert, Water Res. 49 (2014) 316. [30] H. Bae, M. Choi, C. Lee, et al., Chem. Eng. J. 48 (2015) 667. [31] A.N. Nozhevnikova, M.V. Simankova, Yu. V. Litti, Appl. Biochem. Microbiol. 5 (2011) 8. [32] T.N. Viet, S.K. Behera, J.W. Kim, H.S. Park, Environ. Eng. Res. 13 (2008) 210. [33] D.W. Pak, J.Y. Jung, Y.C. Jung, Environ. Eng. Res. 1 (1996) 37. [34] S. Bagchi, R. Biswas, T. Nandy, Microbiol. Biotechnol. 37 (2010) 943.
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier N.V. Tuyena , J.H. Ryua , J.B. Yaeb , H.G. Kimb , S.W. Honga,b , D.H. Ahna,b,* a b
Department of Environmental Engineering and Energy, Myongji University, Yongin, Gyeonggi-do 17058, Republic of Korea Bluebank Co., Ltd., Myongji University, #18109-1, Business Incubator Center 116, Myongji-ro, Cheoin-gu, Yongin-si, Gyeonggi-do 17058, Republic of Korea
A R T I C L E I N F O
Article history: Received 1 February 2018 Received in revised form 25 June 2018 Accepted 2 July 2018 Available online xxx Keywords: Polyvinyl alcohol Anaerobic ammonium oxidation Sodium sulfate Immobilization Nitrogen removal
A B S T R A C T
In this study, the result shows that polyvinyl alcohol–sodium alginate (PVA–SA) gel bead crosslinked with sodium sulfate are better among the different methods by comparing the relative mechanical strength, mechanical strength swelling and expansion coefficient of beads in water. Subsequently, anammox biomass entrapment by PVA–SA gel was introduced into continuous stirred tank reactor (CSTR). After 24 operation days, the nitrogen removal efficiency achieved 60%, while the nitrogen loading rate (NLR) was 0.14 kg N/m3/d and the experiment data indicated that PVA–SA gel bead crosslinked with sodium sulfate can be used to initiate anammox process. Furthermore, it is an alternative for culturing anammox in a long-term operation. © 2018 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Since the anammox process was identified in a denitrifying fluidized bed reactor in deft, evidence for anammox activity has been obtained in a variety of laboratory and engineered systems, pilot plants for ammonium removal, surface and subsurface sediments of deep ocean, temperate estuarine, coastal and offshore sediments, lakes, freshwaters, polar region sediments and multiyear sea ice, oil reservoirs, deep sea hydrothermal vent and the sub-oxic zone of the Black Sea [1–4]. The anammox process is considered having enormous potential in the treatment of high nitrogen content wastewater such as landfill leachate, wastewater from seafood processing industries, etc. In this process, ammonium is oxidized with nitrite serving as the electron acceptor under anaerobic condition, producing nitrogen gas [1] with CO2/ bicarbonate as the sole carbon source like the following formula: 1NH4+ + 1,32NO2 + 0,066HCO3 + 0,13H+ → 1,02N2 + 0,26NO3 + 0,066CH2O0,5N0,15 + 2,03H2O
(1)
However, there are still some important challenges in practical application of anammox because of its slow growth rate and unstable start-up period. Therefore, the maintenance of a sufficient
* Corresponding author at: Department of Environmental Engineering and Energy, Myongji University, Yongin, Gyeonggi-do, 17058, Republic of Korea. E-mail address: [email protected] (D.H. Ahn).
amount of anammox bacteria during the start-up is one of the most essential step with an anammox reactor. Immobilization of microorganism in spherical polymeric matrices is promising for improvement of the efficiency of bioprocess, especially in the production of metabolites and biological treatment of wastewater. Advantages of the immobilized cells compared with free cells include: protection from harsh environment conditions such as pH, temperature, organic solvent and toxic compound, relative ease of product separation, reusability, increased cell density, and reduced susceptibility to contamination by foreign microorganism. Recently polyvinyl alcohol (PVA) has been used widely in the field of wastewater treatment, these advantages include lower cost, high durability and chemical stability, non-toxicity to viable cells [5–7], nonbiodegradable, high mechanical strength and long-life span of gel. Moreover, with porosity, PVA gel has a large amount of water and allows for effective transition of oxygen and nutritional values inside the gel beads. With a porous microstructure, PVA gels bead can be an efficiency way to anammox reactor, it allows the microorganism to live inside the polymeric matrix, protects the anammox bacteria from inhibition factor in surrounding environment like oxygen and substrates still can diffuse to the bacteria, prevents anammox biomass from being washout through the effluent. Furthermore, PVA has several advantages including low cost, high durability and chemical stability, and especially non-toxic to microorganism [8,9]. However, the boric acid technique is often used to produce PVA gel bead and it also has drawbacks
https://doi.org/10.1016/j.jiec.2018.07.004 1226-086X/© 2018 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
2
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
that the bacteria in the PVA matrix can be damaged by boric acid during the process and has a tendency to agglomerate [10]. To eliminate these problems, this study focused on the enrichment and performance of anammox bacteria in a continuous stirred tank reactor (CSTR) with the gel beads which are made by boric acid method with sodium sulfate. The combination of PVA and sodium alginate (SA) was used to improve the surface properties of the beads, reducing the tendency to agglomerate, while the PVA–SA gel bead crosslinked with sodium sulfate has lower toxicity to bacteria than boric acid, higher stability in water and greater biological activity than conventional PVA–boric acid hydrogels bead [11]. After that, the characteristic of nitrogen removal with anammox biomass entrapped in PVA–SA gel bead and anammox performance in the CSTR are evaluated to develop the anammox process in conventional CSTR for further application in wastewater treatment. Material and methods Seed sludge For experiment, the active sludge from a wastewater treatment plant operated under anoxic/anaerobic and aerobic conditions is used. (MLSS = 20–25 g/L, MLVSS = 15–20 g/L). The sludge is washed with distilled water and PBS solution to remove the residual substrate before using. Anammox sludge used in this study was obtained from a lab scale reactor. Sludge was washed three time with phosphate buffered saline (PBS, pH = 7.4, 0.1 M) solution to remove any residual substrate on the sludge surface. Afterwards, the sludge was centrifuged for 5 min at 2000 rpm. Immobilization methods The seed sludge prepared from the previous step was encapsulated in polymer gel by different methods. 200 mL of solution containing 12% PVA (w/v) and 2% sodium alginate (SA) (w/v) was autoclaved at 121 C for 20 min. After cooling to 37 C, this solution is mixed with 200 mL of seed sludge completely. The mixture is dropped into a solution of saturated B(OH)3 and CaCl2 2% (w/v) to form spherical beads. Then, with the method A, the beads are cursed in B(OH)3/CaCl2 solution for 1 h, transferred to 0.5 M Na2SO4 solution and cursed for 1 h [12]. With method B, the beads are cursed in B(OH)3/CaCl2 solution for 3.5 h, transferred to 0.5 M KH2PO4 solution and cursed for 5.6 h. In method C, the mixture is dropped into CaCl2 4% (w/v) solution and cursed in the same solution for 12 h [11]. Finally, all of beads are removed, washed and stored at 4 C with distilled water before using.
Relative strength 20 immobilized granules of similar size of each material type were selected and placed in a 100 mL beaker with 80 mL of DI water. The beads were stirred by Thermo ScientificTM CimarecTM Digital Stirring Hotplates while the speed is controlled from 500 to 2500 rpm and the intact gel beads is counted daily [10]. Swelling of bead in water 20 immobilized granules of similar size of each material type were selected and placed in a 100-mL beaker with 80 mL of DI water. The beaker was mixed for 7 days by magnetic stirrer machine and the size of gel beads were checked daily. Expansion coefficient 20 immobilized granules of similar size of each material type were selected and placed in a 100-mL beaker with 80 mL of DI water. The beaker was slowly shaken at 35 C for 72 h (exactly same with reactor operation in the next step) then the diameter of the gel beads before and after were measured. The ratio of the mean diameter of the gel beads after 72 h of treatment to the mean diameter of the original gel bead was the expansion coefficient [13]. Reactor and experiment method A continuous stirred tank reactor with total working volume of 3 L and 30% packing ratio of gel beads is used with synthetic wastewater. The temperature is maintained at36 C by using water jacket and pH is controlled around 7.5–8 by HCL 1 M. Stirring speed is set to 100 rpm. Stirring was essentially required to mix the influent and remove nitrogen gas bubble which formed on the surface of the gel beads [14]. The influent was supplied by an inflow peristaltic pump from the synthetic wastewater tank through a feed line tubing (Figs. 1 and 2). The CSTR reactor was operated with a hydraulic retention time ranged between 12–24 h, with the nitrogen loading rate from 0,09 to 0,134 kg N/m3/d. The synthetic wastewater used through the experiments contained (per lit): NH4+-N 50–60 mg, NO2 -N 40– 50 mg, KHCO3 500 mg, KH2PO4 27,2 mg, 0.3 g of CaCl22H2O, 0.2 g of MgSO47H2O, 0.00625 g of FeSO4, 0.00625 g of EDTA and 1 mL/L of trace element solution. Trace element solution was composed of ZnSO47H2O, 0.43 g; CoCl26H2O, 0.24 g; MnC124H2O, 0.99 g; CuSO45H2O, 0.25 g; NaMoO42H2O, 0.22 g; NiCl26H2O, 0.19 g; NaSeO410H2O, 0.21 g; H3BO40,014 g (1 mL/L). The effluent sample was collected daily to evaluate the treatment performance. According to the standard method, NO2 -N, NH4+-N, NO3 -N were measured by kit HACH and standard method. The pH was measured using a pH-meter and DO was measured by a portable meter (Figs. 1 and 2) Results and discussion
Characterization of the PVA–SA gel beads Characterization of the PVA–SA gel bead Different methods were applied to make PVA–SA gel beads. In order to determine the better method, relative mechanical strength, mechanical strength swelling of beads in water and expansion coefficient were used following the most suitable gel beads is chosen to culture anammox bacteria in CSTR. Mechanical strength 20 immobilized granules of similar size of each material type were selected and placed in a 100 mL beaker. Then, 80 mL of deionized water (DI water) was added to the beaker. The mixture was magnetically stirred at 500 r/m for 24 h by using Thermo ScientificTM CimarecTM Digital Stirring Hotplates, after which the ratio of intact gel bead to the original number of immobilized gel beads was determined [13].
The Table 1 lists the data of the mechanical strength and expansion coefficient of 4 types of gel bead in this study. Following this table, the mechanical strength of method C was lowest, after 24 h of experiment and only 47% of gel beads were intact. With method A, PVA–SA gel beads crosslinked with sodium sulfate, the mechanical strength of these beads were 0.87, many times higher than that of method C, and close to that of method B, 0.73. Gel bead of method A has highest mechanical strength, however it’s expansion coefficient is lowest, only 1.17. The expansion coefficient of method C was 2, and the size of gel bead increases from 3 mm to 6 mm in day under specific condition of water, resulting in gel beads becoming very soft and easy to dissolve in the water following the raising of the rate of stirring machine. Takei et al. also
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
3
Fig. 1. Anammox reactor schematic diagram.
compared the mechanical strength and expansion coefficient of SA and PVA–SA gel beads made by method C [12]. The result revealed that PVA–SA are stronger and better than SA gel bead. Furthermore, in this experiment, 2 other methods for making PVA–SA gel bead were used, and they both have better characteristic compared to PVA–SA made by method C. Bao-e and Yong-you also reported that the strength of PVA–SA bead is better than SA, however the saturated wet density of SA is higher than that of PVA–SA [15], resulting in a much higher expansion coefficient of PVA–SA beads in comparison with SA bead. With the addition of SA to PVA solution, the hydroxyl on the PVA molecular chain and the carboxyl on the SA molecular chain can form a strong hydrogen. Moreover, the SA macromolecules can diffuse the charge density of PVA macromolecules, so that electrostatic interaction between different PVA molecular chain is attenuated. Thus, the PVA–SA bead is stronger and more stable. Beside mechanical strength, relative mechanical strengths are also checked by increasing the rate of stirring machine. Following
Fig. 3, feeble mechanical strength of the bead of method C was identified, when 50% of beads were broken with 700 rpm of mixing speed and even no intact bead was collected with 1200 rpm. On the contrary, beads of methods A and B were much stronger than that of method C. There was no broken bead even though the mixing speed was 1200 rpm. However, 80% of the latter was broken with the speed over 2500 rpm opposed to 60% of the former. Subsequently, the swelling of bead incubated in water was investigated (Table 2) because polymer matrix for bioremediation of wastewater is required to have high stability in water. With the bead of method C, about 60% of bead was broken during incubation in water for 12 days. On the other hand, all the beads of methods A and B maintained their spherical shapes for 7 days, although the beads of method B became soft. In the first 3 days, the dimension of method C beads increased quickly from 3 mm to 5 mm while the method B increased from 3 mm to 4 mm and 3.5 mm was the size of method A bead. Until day 7, size of bead of methods B and C was bigger than bead of methods A which was be swelling little and did
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
4
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx Table 2 Swelling of bead in water.
Fig. 2. PVA–SA gel beads in the reactor with 30% packing ratio.
Table 1 Mechanical strength and expansion coefficient. Methods
Mechanical strength
Expansion coefficient
Method A Method B Method C
0.87 0.73 0.47
1.17 1.3 2
not be broken in water. Although, the method B bead did not solute in water, but they were swelling in the water. Swelling in water make structure of bead of method C become very weak. The differences between the stability and mechanical strength of the gel beads in the water were due to the differences of the
Days
Method A
Method B
Method C
0 1 2 3 4 5 6 7
3 mm 3 mm 3 mm 3.5 mm 4 mm 4 mm 4 mm 4 mm
3 mm 3.5 mm 3.5 mm 4 mm 4 mm 4 mm 5 mm 5 mm
3 mm 4 mm 4.5 mm 5 mm 5.5 mm 5.5 mm 6 mm 6 mm
making procedures. In method C, the mixture of PVA–SA solution and seed sludge was dropped into the CaCl2 solution for making gel bead. Following this way, the macro-network structure of SA can improve the mass transfer property of PVA and the conglutination behaviour of PVA gel bead [13,16,17]. However, because of presence of SA, the gel bead will become soft, swollen and easy to be broke, while Ca2+ only can improve the surface properties of the bead and reduce the tendency to agglomerate [18]. The advantage of this method is that it is not harmful to the immobilized cell, however, for long time operation, the gel bead need to be strong enough to keep the cell inside polymer matrix. To avoid the problem of boric acid, other types of PVA–SA beads crosslinked with potassium orthorphosphate or sodium sulfate have been reported and these compounds were used in methods B and A. After being dropped into the boric acid and CaCl2 solution, the bead was immersed in sodium sulfate solution. Then, the sulfate ions penetrated into the PVA bead and destabilized hydrogen bonding between hydroxyl groups of PVA and water molecules by polarizing the water molecules. As a result, sulfate ions possess the ability to form linkages among the crosslinked PVA [5], resulting in PVA crystallite formation. In addition, Takei et al. also suggested that [19] sulfate induces a higher degree of crosslinking of PVA due to its higher hydration force compared with phosphate. These information are suitable with the data of this experiment. The swelling of bead, mechanical strength and expansion coefficient of method A bead are better than that of bead of method B. In addition, the biological activity of conventional PVA–boric acid beads drastically decreasing was that surviving cells after immersion in saturated boric acid solution had more severe injuries, which required time for recovery, than the cells incubated in sodium sulfate solution [12]. For these reasons, sodium sulfate can be more useful than sodium phosphate for preparation of PVA–SA gel bead and is an efficient way to immobilize anammox biomass for applying in a CSTR reactor. Therefore, in the next step of experiment, PVA–SA gel beads crosslinked with sodium sulfate was chosen to apply in the CSTR reactor. Anammox start-up and stoichiometric ratios
Fig. 3. Relative mechanical strengths of PVA–SA gel beads made by methods A–C.
The start-up of anammox is a process with the anammox reaction gradually being dominant in the reactor, which can be reflected by some indicators such as molar ratios of consumed NO2 -N to consumed NH4+-N and produced NO3 -N to consumed NH4+-N [20]. This study was inoculated with a mixture of active sludge and anammox sludge immobilized in PVA–SA gel beads with an NLR of 0.1 kg N/m3/d (NO2 -N/NH4+-N feeding ratio equal to 0.8) (Fig. 4). In the first 15 days of experiment, the nitrogen removal rate (NRR) was 0.03 kg N/m3 d, while the NO2 -N/NH4+-N ratio and NO3 -N/NH4+-N ratio were much lower than the theoretical stoichiometric ratios which were reported to be 1.32 and 0.26 respectively [21]. The ratio of produced NO3 -N/ consumed NH4+-N was lower than the theoretical value, indicating a low nitrate generation (Fig. 5). This might be attributed to the continuous happening of the heterotrophic denitrification process
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
Fig. 4. Evaluation of nitrogen loading rate (NLR) and nitrogen removal rate (NRR) during 80 operation days.
relying on the lysis products in the reactor. Otherwise, anammox has very slow growth rate compared with the other types of bacteria such as AOB, NOB. Hence, other bacteria were dominant with anammox bacteria in the first 15 days. Until day 20, when the NLR was 0.11 kg N/m3/d, the NRR increased to 0.07 kg N/m3/d. Moreover, the NO2 -N/NH4+-N ratio and NO3 -N/NH4+-N ratio were 1.1 and 0.2. This indicated that the transformations are in agreement with the anammox reaction, implying the dominance of the anammox process. The stoichiometric ratios of anammox obtained were quite stable during the next 10 days. At day 30, the NLR was increased to 0.18 kg N/m3/d in order to make anammox bacteria grow faster. Although the NRR was 0.07 kg N/m3/d, same as previous days, however, the NO2 -N/NH4+-N ratio decreased sharply to 0.4. This data reveals that the NOB could grow very fast in this period of experiment and it would affect anammox bacteria if there had been no adjustment. Thereby, the NLR was decreased to initial value and kept unchange. However, there had not been change until the HRT was decreased from 24 h to 12 h. As a result, the NLR rose to 0.25 kg N/m3/d (NO2 -N/NH4+-N feeding ratio equal to 0.8) and NRR rose to 0.12 kg N/m3/d. Especially, the NO2 N/NH4+-N ratio rose fast to 0.8. This trend implies that the growth of NOB was inhibited and anammox bacteria was becoming dominant in the reactor.
Fig. 5. Stoichiometric ratios of NO2 -N/NH4+-N and NO3 -N/NH4+-N of CSTR during 80 operation days.
5
Fig. 6. Influent, effluent concentration and removal efficiency of NO2 -N during 80 operation days.
Nitrogen removal performance Nitrite concentration Fig. 6 shows the influent and effluent of nitrite during the experiment, which lasted 80 days. The concentration of nitrite during the startup has crucial importance for growth of anammox: a too low amount of it will result in substrate limitation and thus slower growth, while concentration above 50–150 mg/L can already lead to inhibition [22–24]. From the beginning of the experiment, the influent concentration of nitrite was about 40 mg/ L to prevent the inhibition of nitrite with anammox bacteria in the bead. There is no day that the effluent of Nitrite is higher than the influent, it indicates that nitrite oxidation bacteria presented in the gels beads and converted NO2 -N to NO3 -N. Afterwards, the influent concentration of NO2 -N was increased to 50–57 mg/L for encouraging the development of anammox inside the bead. At this time, the effluent of NO2 -N reduced from about 30 mg/L at the beginning to 12.5 mg/L with 78% of NO2 -N removed from the water in day 18. Thus, the influent NO2 -N was increased to 78 mg/ L. However, the effluent NO2 -N was very high and the removal efficiency was only 20%. In this case, the anammox bacteria could be inhibited by high concentration of nitrite in wastewater while they were not strong enough. Therefore, the influent NO2 -N was reduced to about 40 mg/L to recovery the activity of anammox bacteria in the reactor. At day 59, the effluent was about 15 mg/L and the removal efficiency was achieved 70%. The influent was slightly increased and maintained at 50 mg/L. Although in this period, there was a decrease of HRT from 24 h to 12 h, from day 70 to 80, the efficiency removal was stable around 50–60%. This result show that, the anammox bacteria could adapt with synthetic wastewater, and was retained inside the gel beads. Ammonium and nitrate concentration Figs. 7 and 8 give data about changes in concentration of NH4+N and NO3 -N during the experiment. In the start-up period of the experiment, the influent nitrite-to-ammonium molar ratio was maintained at 0.76 mol NO2 -N/NH4+-N (lower than the theoretical stoichiometric ratio) to avoid possible toxicity caused by nitrite accumulation that it could inhibit the anammox process [25]. Hence, the influent concentration of NO2 -N was 40–50 mg/L, along with 50–60 mg/L of NH4+-N. In addition, it is important that the synthetic wastewater contained sufficient nitrate concentration to prevent sulphate reduction to sulphite which has been demonstrated to be toxic to anammox bacteria [25–28]. For this reason, 6–7 mg/L of the nitrate influent concentration was observed. In the first 10 days of experiment, the effluent
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
6
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx
Fig. 7. Influent and effluent concentration of NO3 -N during 80 operation days.
constant oxygen saturation for AOB is 0.3 mg/L lower than NOB (1 mg/L) [31–33]. Other one has revealed that many AOB species are able to adapt their metabolism even in acute absence of oxygen [34]. These data explained the change of NH4+-N concentration in the experiment beside anammox bacteria. Similar with influent of NO2 -N, from day 30 to 40, the influent NH4+-N was increased to 90 mg/L. Then, the effluent NH4+-N was 30 mgN/L and removal efficiency was 60%. There were not many differences of these data from previous days while NO3 -N effluent rose dramatically from 10 to 20 mg/L. On the other hand, NOB bacteria also can grow under oxygen limitation, along with the increase of NO2 -N which is electron donor, NOB developed in these days. From day 40, for anammox bacteria reactivity after being inhibited under high concentration of NO2 -N, the influent NH4+-N was maintained at 60 mg/L. There was a sharp drop in the effluent NH4+-N, which was lower than 10 mg/L and the removal efficiency was 80%, along with an increase of 25 mg/L of NO3 -N in the effluent. The reason of this trend is that the AOB and NOB predominated over the anammox bacteria. From day 66, the HRT was adjusted to 12 h for keeping the optimal ratio of influent NO2 N/NH4+-N and preventing the accumulation of NO2 -N which can inhibit the anammox bacteria. Following this, from day 68, the effluent NO3 -N was 15 mg/L, the effluent NH4+-N rose to 20 mg/L. This result indicated that NOB growth was controlled which can lead to the development of anammox bacteria. Conclusion
Fig. 8. Influent, effluent concentration and removal efficiency of NH4+-N during 80 operation days.
concentration of NH4+-N was relatively high at about 50 mg/L and close to influent and the removal efficiency was about 20%. This data has been due to the anammox bacteria and other bacteria have not adapted with the synthetic wastewater yet, whereas the concentration of NO3 -N also indicated that there was problem with the bacteria in the reactor. While the influent NO3 -N was 6 mg/L, the effluent was 5 mg N/L. There was a fall in the NO3 -N concentration because of endogenous heterotrophic denitrification. Although, the organic carbon source was not introduced in the synthetic wastewater which is important for heterotrophic bacteria. However, at the beginning of experiment, there was a small number of broken gel beads. As a result, the SS of effluent increased, the bacteria in the broken bead could die when they contact with the new environment of wastewater [29,30]. After that, the cell lysis happened and released the organic compound which could be later consumed by heterotrophic bacteria as carbon source for denitrification process. From day 12 to 30 there was no more broken bead, the effluent NO3 -N became higher than influent and was around 10–15 mg/L. At this time, the ammonium removal efficiency rose gradually and achieved 66% at day 30, the effluent NH4+-N was 23 mg/L. Without DO supply, the DO concentration in reactor maintained lower than 0.5 mg O2/L. But, the removal efficiency of NH4+-N was still high compared with NO2 -N removal efficiency, even though AOB bacteria requires oxygen as substrate. Some researcher reported that the Ks value of
In this study, different type of method had been used to make PVA–SA gel bead and the properties of these gel beads were also checked. Lastly, PVA–SA gel beads crosslinked with Sodium sulfate is high mechanic strength, non-solute in the water, short immersion time for solidification compared with other methods. It can be suitable for anammox immobilization and applying in the CSTR reactor. After 80 days of this experiment, the volume of gel bead still occupied nearly 30% of reactor and the beads were strong. The total nitrogen removal was achieved 50–60% after 25 operation days with. Stoichiometric ratios of consumed NO2 -N to consumed NH4+-N and produced NO3 -N to consumed NH4+-N in the experiment are close to the theoretical stoichiometric ratios of anammox reaction. In conclusion, this data indicated that PVA–SA gel beads crosslinked with Sodium sulfate can be used to fast startup of anammox process in a short time. However, further research about reactor configuration and sufficient amount of seed anammox biomass for better nitrogen removal performance of anammox bacteria immobilized with PVA–SA gel beads crosslinked with sodium sulfate need to be done. Acknowledgement This subject is supported by Korea Ministry of Environment as “Global Top Project” (Project No.: 2016002190006). References [1] A. Mulder, A.A. van de Graaf, L.A. Robertson, et al., FEMS Microbiol. Ecol. 16 (1995) 177. [2] H. Wu, G. Zeng, J. Liang, J. Chen, J. Xu, et al., Int. J. Appl. Earth Obs. Geoinf. 56 (2017) 36. [3] H. Wu, G. Zeng, J. Liang, et al., Ecol. Eng. 57 (2013) 72. [4] H. Wu, G. Zeng, J. Liang, et al., Ecol. Indic. 53 (2015) 129. [5] N.A. Mohd Zain, M.S. Suhaimi, A. Idris, Process Biochem. 46 (2011) 2122. [6] K.W. Jung, M.J. Hwang, T.U. Jeong, D.M. Chau, J. Ind. Eng. Chem. (2017) 1. [7] G. Zeng, H. Wu, J. Liang, et al., RSC Adv. 5 (2015) 34541. [8] R. Surudzic, A. Jankovic, M. Mitric, I. Matic, Z.D. Juranic, L. Zivkovic, V. Miskovic-Stankovic, K.Y. Rhee, S.J. Park, D. Hui, J. Ind. Eng. Chem. (2015). [9] M. Shafiq, A. Sabir, A. Islam, S.M.M. Khan, S.N. Hussain, M.T.Z.Z. Butt, T. Jamil, J. Ind. Eng. Chem. (2016) 1. [10] P.V. Dinh, L.T. Bach, Int. J. Sci. Eng. 7 (2014) 41.
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004
G Model JIEC 4078 No. of Pages 7
N.V. Tuyen et al. / Journal of Industrial and Engineering Chemistry xxx (2018) xxx–xxx [11] M. Ali, M. Oshiki, L. Rathnayake, S. Ishii, H. Satoh, S. Okabe, Water Res. 79 (2015) 147. [12] T. Takei, K. Ikeda, H. Ijima, K. Kawakami, Process Biochem. 46 (2011) 566. [13] G. Chen, J. Li, S. Tabassum, Z. Zhang, RSC Adv. 5 (2015) 25372. [14] L.M. Quan, D.P. Khanh, D. Hira, T. Fujii, K. Furukawa, Biodegradation 22 (2011) 1155. [15] W. Bao-e, H. Yong-you, J. Environ. Sci. 19 (2007) 451. [16] G.P. Lim, M.S. Ahmad, J. Ind. Eng. Chem 56 (2017) 382. [17] H. Wu, G. Zeng, J. Liang, et al., Appl. Microbiol. Biotechnol. 100 (2016) 8583. [18 Kuo-Ying Amanda Wu, Keith D. Wisecarver, Biotechnol. Bioeng. (1992) 447. [19] T. Takei, K. Ikeda, H. Ijima, M. Yoshida, K. Kawakami, Polym. Bull. 69 (2012) 363. [20] Y.F. Lu, L.J. Ma, L. Ma, B. Shan, J.J. Chang, Environ. Technol. 39 (2017) 59. [21] M. Strous, J.J. Heijnen, J.G. Kuenen, M.S.M. Jetten, Appl. Microbiol. Biotechnol. 50 (1998) 589. [22] S. Suneethi, S. Shalini, K. Joseph, Int. J. Waste Resour. 4 (2014). [23] S. Uyanik, Ozan K. Bekmezci, A. Yurtsever, Clean – Soil Air Water 39 (2011) 653.
7
[24] H. Wu, C. Lai, G. Zeng, et al., Crit. Rev. Biotechnol. 37 (6) (2017) 754. [25] H. Lopez, S. Puig, R. Ganigue, M. Ruscalleda, M.D. Balaguer, J. Colprim, J. Chem. Technol. Biotechnol. 83 (2008) 233. [26] X. Ma, Y. Wang, S. Zhou, Y. Yan, X. Lin, M. Wu, Chem. Eng. J. 313 (2016) 1233. [27] S.E. Vlaeminck, J. Geets, H. Vervaeren, N. Boon, Willy Verstraete, Environ. Biotechnol. 74 (2007) 1376. [28] L.G.J.M.V. Dongen, M.S.M. Jetten, M.C.M.V The Combined Sharon/Anammox Process Loosdrecht, IWA Pulishing (2001). [29] S. Jenni, S.E. Vlaeminck, E. Morgenroth, K.M. Udert, Water Res. 49 (2014) 316. [30] H. Bae, M. Choi, C. Lee, et al., Chem. Eng. J. 48 (2015) 667. [31] A.N. Nozhevnikova, M.V. Simankova, Yu. V. Litti, Appl. Biochem. Microbiol. 5 (2011) 8. [32] T.N. Viet, S.K. Behera, J.W. Kim, H.S. Park, Environ. Eng. Res. 13 (2008) 210. [33] D.W. Pak, J.Y. Jung, Y.C. Jung, Environ. Eng. Res. 1 (1996) 37. [34] S. Bagchi, R. Biswas, T. Nandy, Microbiol. Biotechnol. 37 (2010) 943.
Please cite this article in press as: N.V. Tuyen, et al., Nitrogen removal performance of anammox process with PVA–SA gel bead crosslinked with sodium sulfate as a biomass carrier, J. Ind. Eng. Chem. (2018), https://doi.org/10.1016/j.jiec.2018.07.004