- Email: [email protected]
A novel strategy for simultaneous removal of nitrogen and organic matter using anaerobic granular sludge in anammox hybrid reactor
Accepted Manuscript A novel strategy for simultaneous removal of nitrogen and organic matter using anaerobic granular sludge in anammox hybrid reactor Swati Tomar, Sunil Kumar Gupta, Brijesh Kumar Mishra PII: DOI: Reference:
S0960-8524(15)01163-3 http://dx.doi.org/10.1016/j.biortech.2015.08.057 BITE 15410
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
27 June 2015 10 August 2015 12 August 2015
Please cite this article as: Tomar, S., Gupta, S.K., Mishra, B.K., A novel strategy for simultaneous removal of nitrogen and organic matter using anaerobic granular sludge in anammox hybrid reactor, Bioresource Technology (2015), doi: http://dx.doi.org/10.1016/j.biortech.2015.08.057
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A novel strategy for simultaneous removal of nitrogen and organic matter using anaerobic granular sludge in anammox hybrid reactor Swati Tomar1, Sunil Kumar Gupta2*, Brijesh Kumar Mishra3 1
Senior Research Fellow, Department of Environmental Science & Engineering, Indian School of
Mines, Dhanbad- 826004, Email: [email protected], Phone No.: +91-326-2235474, Fax: +91326-2296624 2*
Associate Professor, Department of Environmental Science & Engineering, Indian School of Mines,
Dhanbad- 826004, Email: [email protected], Phone No.: +91-326-2235474, Fax: +91-3262296624 (Corresponding Author) 3
Assistant Professor, Department of Environmental Science & Engineering, Indian School of Mines,
Dhanbad- 826004, Email: [email protected], Phone No.: +91-326-2235752, Fax: +91-3262296624
Abstract The coexistence of organic matter (OM) and nitrogen in industrial effluent is the major bottleneck in field application of anammox process. The present study emphasized on investigating the role of seeding anaerobic granular sludge towards simultaneous removal of ammonium and OM in anammox hybrid reactor (AHR). The study delineated simultaneous reduction of both OM (94.8%) and nitrogen (96.8%) at optimal COD/N ratio (0.54). Pearson correlation matrix showed positive and strong correlation of ARE (ammonium removal efficiency) and CRE (COD removal efficiency) with NRE (nitrogen removal efficiency). The negative correlation of OLR and COD/TN ratio with NRE indicated that increase in organic loadings may suppress anammox activity. The process inhibition was evaluated using Haldane model considering free ammonia, OM and nitrite as inhibitors. The strategy of using anaerobic granular sludge not only augmented endurance of bacterial communities against OM inhibition but also facilitated simultaneous removal of OM and nitrogen.
Keywords: Anammox, AHR, organic matter, nitrogen removal, modelling, inhibition kinetics.
INTRODUCTION 1
Over the last decade, nitrogen removal from wastewater has become a worldwide emerging concern pertaining to its toxic nature to aquatic species and eutrophication in receiving water body. The conventional method of nitrogen removal involves nitrification and denitrification, which is characterized by high energy demand, operational cost and problems of sludge disposal. In addition, the application of ammonia stripper is limited as a result of significantly lower stripping efficiency (50-65%) and the associated problem of air pollution. Anammox is a novel microbiological approach that has changed the traditional concept of biological nitrogen removal (Fernandez et al. 2012). The process facilitates direct oxidation of ammonium nitrogen into nitrogen gas under anoxic conditions with nitrite as an electron acceptor. Significant reduction in aeration costs, exogenous electron donor saving and low sludge production makes the process techno-economically feasible over existing conventional treatment technologies. While, the newly discovered anammox process opens up the new possibilities for nitrogen removal from wastewater, the industrial applications are constrained due to presence of organic matter (OM) in the industrial effluent (Magri et al. 2013), which has been reported to cause deterioration in anammox activity (Ni et al. 2012). Industrial effluents, such as coke ovens (Toh and Ashbolt 2002), fertiliser (Keluskar et al. 2013), antibiotics (Tang et al. 2011) contain high concentration of OM varying from 46.6-2200 mg/l. Hence, there is an acute need to evolve a novel technique which can facilitate simultaneous removal of ammonium and OM from the industrial effluent, if anammox is applied as a treatment alternative. Several researchers (Ni et al. 2012; Molinuevo et al. 2009) have investigated the effects of OM on anammox process and revealed that it poses severe threat to anammox bacteria. Guven et al. (2005) revealed that even 0.5 mM of methanol resulted in immediate and complete inactivation of anammox activity. Similarly, Molinuevo et al. (2009) also reported that COD concentrations upto 292 mg/l led to complete inhibition of anammox process. The inhibition is mainly caused due to higher affinity of nitrite to denitrifiers than anammox bacteria leading to upset in substrate co-substrate ratio which is essentially required for smooth functioning of the process. Extensive literature review suggests that process inhibition is also caused by FA (free ammonia) and nitrite in addition to COD (Fernandez et al. 2012; Kimura et al. 2010). FA concentration of 57-187 mg/l is sufficient to induce toxicity in anammox process (Tang et al. 2010a). Kimura et al. (2010) investigated the effects of nitrite on anammox bacteria entrapped in gel carriers using batch and continuous feeding
2
tests and reported that nitrite concentration beyond 274 mg/l was found inhibitory to anammox bacteria. Several researchers advocated that the effect of inhibition can significantly be minimised by extensive acclimation and use of granular anammox sludge in addition to intermittent seeding of the reactor. Ni et al. (2012) revealed that granular anammox sludge demonstrated higher tolerance to OM as compared to flocculant sludge and was capable of delivering ammonium removal of 80% even at a higher COD to N ratio of 3:1. Seeding of granular sludge significantly contributed towards enhanced COD removal efficiency (CRE) in addition to its capability to sustain high organic loads (Molinuevo et al. 2009). Abbasi and Abbasi (2012) realized that granular sludge of appropriate particle size, particle density, and microfilm characteristics enhances the reactor efficiency in addition to reduction in start-up time. No such study has been reported in literature to investigate the simultaneous removal of OM and ammonical nitrogen by establishing the proper symbiosis amongst anammox, denitrifiers and anaerobic bacteria. The symbiotic association of these bacteria may help in simultaneous removal of both OM and ammonical nitrogen from the industrial effluent. Hence, the present study was undertaken to explore the feasibility of addition of anaerobic granular sludge to make the process viable for its industrial applications.
MATERIAL AND METHODS
Experimental Set-up of AHR
The experimental setup of AHR is shown in Fig. 1. The reactor was fabricated of transparent acrylic plastic with an internal diameter of 10 cm and height 65 cm. The total working volume was 5 litres. Corrugated polyvinyl chloride (PVC) pipes of length 2.25 cm and diameter 2.25 cm were used as filter media. Total 55 nos. of PVC carriers were added to the reactor to constitute an attached growth system. The sludge blanket in the lower half of the reactor constitutes suspended growth system while filter media in the upper part provides attached growth for the microorganisms. The reactor was also completely covered with black cloth to avoid the growth of phototrophic organisms and oxygen production. The reactor was fed with the synthetic wastewater (van de Graaf et al. 1996) using a peristaltic pump to maintain a constant flow rate.
3
Fig. 1(A) Schematic diagram of AHR
Fig. 1(B) Experimental set-up of AHR
Strategy of operation
The reactors were started with initial influent ammonium and nitrite concentrations of 100 mg/l each to maintain an optimal NH4/NO2 ratio of 1:1 at an HRT of 1d. The reactors were operated for a period of 30 days till the attainment of pseudo-steady state condition. Then, the ammonium and nitrite concentrations were gradually increased to 600 mg/l each, and maintained constant to assess the performance of reactor during variable hydraulic retention times (HRTs). During this study, the efficiency of the reactor in terms of ammonium, nitrite removal was observed to devise the optimal HRT required for the efficient functioning of the reactor. After attainment of optimal HRT, the reactors were seeded with anaerobic granular sludge (1:5 v/v) to establish the mixed consortia of anammox and anaerobic bacteria. The anaerobic granular sludge was collected from high rate UASB (upflow anaerobic sludge blanket) reactor treating coke oven wastewater. The mixture was then homogenised and acclimated to higher organic COD by gradual addition of glucose in synthetic wastewater. The optimal COD/TN ratio is very essential for developing a symbiotic relation between different communities of bacteria for its efficient functioning. The COD/TN ratio in the study was varied from 0.04-0.83 by varying the influent COD concentration from 50 to 1000 mg/L, till process inhibition was observed. The process inhibition was studied using Haldane model considering FA, nitrite and COD as inhibitors to explore the major cause of inhibition.
Analytical Methods
The analysis of pH, alkalinity, COD, NH4-N, NO2-N, NO3-N, VSS and TSS were carried out as per the standard methods (APHA 2012). The volume of gas produced in the reactors was measured by water displacement method and the composition was analysed through Gas Chromatograph (GC) equipped with thermal conductivity detector (TCD) using DPK Spherocarp column. The GC was operated at injector and detector temperatures of 120ºC and 150ºC, respectively. H2 @ 20 ml/min was used as a carrier gas.
Statistical Analysis 4
Pearson correlation matrix (significance levels α = 0.05, 0.01) was used to determine the correlation between nitrogen removal efficiency (NRE), and operational parameters i.e., pH, Alkalinity, COD/TN ratio, OLR (organic loading rate), CRE, NO3-N, etc. using SPSS 16.0. The biasness of the kinetic model was tested using t test. The values of bio-kinetic coefficients were determined using Graph pad prism software (Version-5.03) for non-linear regression method.
Quality control/quality assurance procedure (QA/QC)
The laboratory reagent blanks were prepared and analyzed to determine if any interference was present in the effluent samples. The precision of the measurements were estimated using triplicate sample analysis. The relative percentage difference (RPD) between two parallel samples was calculated and cross verified. In case the RPD exceeded > 5%, the samples were recollected and analyzed. The average of the triplicate readings was reported as the final value. Continuous calibration checks were performed during GC analysis after injection of every 10 samples. If the RPD between the response of the initial calibration and the calibration check standard was > 10%, the instrument was considered as out of calibration, and was recalibrated. The high precision gas standard procured from Chemtron Science Laboratories Pvt. Ltd., Mumbai was used for calibration.
Haldane Model
A number of substrates serve as nutrients at low concentrations, but act as inhibitors at high concentrations (Edwards 1970). High substrate concentration inhibits bacterial growth and deteriorates anammox performance. This was evaluated using Haldane model which can be represented by Eq. (1):
=
(1)
1+ +
5
where, q, qmax is specific substrate conversion rate and maximum specific conversion rate, respectively (1/d); Ks is half saturation constant (mg/L); Ki is substrate inhibition constant (mg/L) and S is the substrate concentration (mg/l).
RESULTS AND DISCUSSION
Effect of HRT on the performance of AHR
Optimisation of HRT is very essential to evolve most economical design of bioreactor (Chaganti et al. 2013; Ndegwa et al. 2005). Shorter HRT assists smaller reactor volume, and hence, reduces the construction costs (Wang et al. 2009). Results indicated that the NRE increased with increase in HRT and was maximum (97.5%) at HRT of 2.25 d (Fig. 2). However, beyond HRT of 1d, the rate of increase was not significant. Further increase in HRT beyond 2.25 d, resulted in substantial decline in NRE, which may be attributed to lower availability of the substrate. Hence, the HRT of 1.0 day was considered optimal corresponding to substantial nitrogen removal of 95.1%. Ammonium removal also depicted similar profile. However, nitrite removal efficiency (NiRE) was not affected significantly. A possible reason could be the consumption of nitrite by heterotrophic denitrification pathway. The efficiency of the reactor observed in our study is significantly higher than reported (8089.9%) in the literature (Ni et al. 2010; Duan et al. 2012). The higher efficiency observed in our system may be attributed to the higher biomass retention and attached filter media which might have enhanced the NRE. Duan et al. (2012) reported that the use of non-woven carrier as AGM effectively increased the biomass retention by 2.8% and overall NRR by 8.1 % in the hybrid reactor. Grandhi et al. (2011) also investigated the performance of UASB and AHR for the treatment of distillery spent wash and reported that the hybrid reactor configuration contributed an additional 5% COD removal and 25% reduction in sludge washout rate. It was also observed that at lower HRT < 1 day, the pH of the reactor was 8.8±0.2, and FA concentration varied from 9.4 mg/l to 15.7 mg/l. Jung et al. (2007) reported that FA concentration of 1.7 mg/l is sufficient to induce toxicity in anammox process. While, Waki et al. (2007) and Tang et al. (2010a) reported comparatively higher toxicity range of FA varying from 13–187 mg/l. This might be cause for upset in reactor performance at lower HRT
6
However, the effect of process inhibition due to FA was not substantial in our study, as NRE never dropped below 78% even at minimum HRT of 0.25d. The nitrate production which is considered as one of the bottlenecks of anammox process did not vary significantly and found comparatively lower (37.2 to 52.3 mg/l) than reported in the literature (Jin et al. 2013; Jin and Zheng 2009). The lower concentration of nitrate observed in our study may be attributed to lower availability of nitrite nitrogen due to consistently higher nitrite removal (95-98%) mediated by anammox pathway. Strous et al. (1999) reported that excess nitrite in the system results in formation of nitrate and hydroxylamine in the process. The larger fluctuation of nitrite is also unfavorable for the steady operation of the system.
Fig. 2 Performance of AHR at different HRTs
Effect of seeding of anaerobic granular sludge towards simultaneous removal of OM and nitrogen Both COD concentrations and COD/TN ratio play an instrumental role and affect the overall efficiency of anammox process. The analysis of data revealed that NRE did not fluctuate much with increase in COD/TN ratio from 0.04 to 0.54 (Fig. 3). However, the CRE increased consistently upto COD/TN ratio of 0.54. The simultaneous removal of both COD and nitrogen observed in our study indicated establishment of symbiosis between anammox and anaerobic bacteria. But when the ratio was further increased, both CRE and NRE declined at elevated COD concentration (> 650 mg/l). The, optimal COD/TN ratio corresponding to simultaneous removal of COD and nitrogen was 0.54. Decline in effluent nitrate coupled with increase in alkalinity at high organic load, also dictated deterioration in anammox performance (Fig. 4). The process might have been inhibited due to increase in COD concentration at higher COD/TN ratio. Several researchers have reported that OM poses adverse effects on the activity of anammox bacteria. Chamchoi et al. (2008) studied the effect of OM on anammox performance and reported that COD concentration over 300 mg/l could inactivate or even eradicate anammox communities. Molinuevo et al. (2009) also reported that COD concentrations upto 292 mg/l led to complete inhibition of anammox process. However, the mechanism of anammox suppression in our study outcompeted others. This may be due to seeding of anaerobic granular sludge which might have enhanced the endurance of the reactor towards OM, as is evident from comparatively higher CRE (84.57
94.8%) observed in our study than reported (30-56.5%) in the literature (Nhat et al. 2014; Chen et al. 2013). The robustness of AHR to sustain high organic loads (650mgCOD/L) in our study might be attributed to the seeding of granular anaerobic sludge which resulted in establishing harmonious symbiotic association between different communities of anammox bacteria, anaerobes and heterotrophic denitrifiers.
Fig. 3 Effect of COD/TN ratio on the performance of AHR
Fig. 4 Effluent Nitrate and alkalinity profile of AHR subjected to organic loads
Correlation of process parameters with NRE
Pearson correlation analysis revealed that NRE is significantly influenced by various operating parameters (Table 1). The nitrate concentration and pH showed strong and positive correlation with NRE. Many researchers (Li et al. 2012; Jaroszynski et al. 2011) also indicated that increase in effluent nitrate and pH are the indicators of successful anammox process. ARE and CRE both showed positive correlation with NRE indicating simultaneous reduction of ammonium and COD in AHR. A possible reason could be coexistence of anammox autotrophs, anaerobic consortia and denitrifiers which simultaneously use ammonical nitrogen, OM and nitrite nitrogen as a substrate in their removal pathways. However, the higher correlation of NRE with ARE than CRE, suggested predominance of anammox pathway. The parameters alkalinity, OLR and COD/N ratio showed strong but negative correlation. The negative correlation of OLR and COD/TN ratio with NRE indicated that increased organic loading rates may suppress the anammox activity leading to deterioration of reactor performance. At higher COD/TN ratio, denitrifiers are more prevalent and outcompeted anammox bacteria due to their higher affinity to nitrite nitrogen. This results in upset in the performance of anammox consortia due to less availability of nitrite nitrogen as a co-substrate. Alkalinity also showed strong but negative correlation. This may be attributed to progressive consumption of bicarbonate as a part of enrichment medium for the growth and development of anammox populations (Suneethi and Joseph 2011)
Table 1 Correlation analysis of NRE and various operational parameters in AHR 8
Assessment of process inhibition using Haldane model
Process kinetics and modelling are the important tools for designing and predicting the performance of bioreactor. Process inhibition kinetics was evaluated considering COD, nitrite and FA as major inhibitors. The plots of Haldane model showed higher correlation for COD followed by ammonium and nitrite (Fig. 5). The maximum specific conversion rate for ammonium and nitrite was comparatively lower than reported values (Tang et al., 2009; Chen et al., 2011) as shown in Table 2. This may be attributed to the process inhibition caused due to OM and low availability of substrate in our study. Half saturation constant for ammonium was in line with the reported values but, it was significantly higher for nitrite nitrogen (Zu et al. 2008). This may be due to excess utilisation of nitrite nitrogen by denitrifiers and anammox bacteria during process inhibition conditions. The Ki values of ammonium, nitrite and COD dictated that the reactor exhibited highest tolerance for ammonical nitrogen followed by COD and nitrite. The values of Ki with OM inhibition were still comparable with those reported in literature (Zu et al. 2008; Chen et al. 2011; Molinuevo et al. 2009; Ni et al. 2012) indicating enhanced endurance of AHR towards OM. Accordingly, the Haldane model fit for NH4-N, NO2-N and COD can be represented by Eqs. (2-4): 17.49 (2) 45.81
1 + + 930.2 22.56 = (3) 73.55
1+ +
220.8 32.92 = (4) 184.2
1 + + 354.9 =
Table 2 Summary of kinetic constants calculated from the Haldane model
Fig. 5 Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
Validation of Model The models were validated by predicting the effluent ammonium, nitrite and COD concentration for a new set of values excluding the ones which were used for determination of kinetics. The plot between observed and predicted values demonstrated strong correlation 9
with COD followed by ammonium, but it was found poor in case of nitrite (Fig. 6). The standard error of prediction was also minimum in case of COD (Table 3). This indicates that COD is the most influential parameter which governs process inhibition in AHR. The chances of inhibition due to FA was minimised due to to the absence of excessive alkaline conditions which promotes the chance of FA inhibition. The FA concentration throughout the study varied from 3.4-45 mg/l at corresponding pH of 8.1-8.6. The possibility of inhibition due to nitrite can also be curtailed as is evident from consistently higher nitrite removal throughout the study. It can also be seen that when, nitrite was considered as inhibitory substrate, the observed and predicted effluent nitrite fell close to each other till a concentration of 800 mg/l, beyond which the model failed and predicted significantly lower nitrite concentration. Thus, Haldane model can suitably be applied to assess the fate and process inhibition caused due to COD. To determine the biasness of the model, t test was also performed. The tstats value was much lower than tcrit indicating that the model is unbiased. A hypothesis was proposed in this study which declares that process inhibition can be minimised by optimising the ratio of seed sludge in the bioreactor. The fine tuning of ratio of seed sludge and anoxic sludge might result in proficient symbiotic association between bacterial communities in addition to eradication of process inhibition caused by OM. The study on the proposed hypothesis is ongoing by varying ratio of seed sludge/anammox sludge to arrive at the optimal ratio.
Fig. 6 Validation of Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
Table 3 Validation statistics of Haldane model
CONCLUSION
AHR dictated enormous potential towards nitrogen removal (95.1%) at optimum HRT of 1 day. Proficient symbiosis of anammox bacteria with anaerobes facilitated simultaneous removal of OM and nitrogen. The optimal performance was observed at COD/N ratio of 0.54 beyond which process inhibition was instrumental. Haldane model successfully evaluated process inhibition with COD as the major inhibitor than FA and nitrite. Addition of anaerobic 10
granular sludge (1:5v/v) promoted appreciable bacterial endurance towards OM inhibition. The study opens the door for researchers and scientists for optimizing simultaneous removal of OM and nitrogen which is a major bottleneck for field-scale application of anammox.
ACKNOWLEDGEMENT The authors thank the financial support from Indian School of Mines, Dhanbad under Junior Research Fellowship scheme funded by Ministry of Human Resource Development (MHRD), Government of India, New Delhi, for carrying out the this study.
CONFLICT OF INTEREST
Authors declare that they have no conflict of interest.
REFERENCES
1. Abbasi T, Abbasi SA (2012) Formation and impact of granules in fostering clean energy production and wastewater treatment in upflow anaerobic sludge blanket (UASB) reactors. Renewable and Sustainable Energy Reviews 16:1696– 1708 2. APHA, AWWA, WEF (2012) Standard Methods for Examination of Water and Wastewater, 22nd ed. United Book Press, USA 3. Chaganti SR, Pendyala B, Lalman JA, Veeravalli SS, Heath DD (2013) Influence of linoleic acid, pH and HRT on anaerobic microbial populations and metabolic shifts in ASBRs during dark hydrogen fermentation of lignocellulosic sugars. International Journal of Hydrogen Energy 38:2212-2220 4. Chamchoi N, Nitisorvut S, Schmidt JE (2008) Inactivation of ANAMMOX communities under concurrent operation of anaerobic ammonium oxidation (ANAMMOX) and denitrification. Bioresour Technol 99:3331–3336 5. Chen C, Huang X, Lei C, Zhang TC, Wu W (2013) Effect of organic matter strength on anammox for modified greenhouse turtle breeding wastewater treatment plant. Bioresour Technol 148:172-179 6. Chen T, Zheng P, Shen L, Ding S, Mahmood Q (2011) Kinetic characteristics and microbial community of anammox-ESGB reactor. J Hazard Mater 190:28-35 11
7. Duan X, Zhou J, Qiao S, Yin X, Tian T, Xu F (2012) Start-up of the anammox process from the conventional activated sludge in a hybrid bioreactor. J Environ Sci 24:1083-1090 8. Edwards VH (1970) Influence of high substrate concentrations on microbial kinetics. Biotechnol Bioeng 12:679–712 9. Fernandez I, Dosta J, Fajardo C, Campos JL, Mosquera-Corral A, Méndez R (2012) Short- and long-term effects of ammonium and nitrite on the Anammox process. J Environ Manage 95:S170-S174 10. Grandhi SC, Pandey LMS, Gupta SK, Singh G (2011) Comparative evaluation of high rate anaerobic processes for treatment of distillery spent wash. J Indus Res Technol 1(1):17-23 11. Güven D, Dapena A, Kartal B, Schmid MC, Maas B, van de Pas-Schoonen K, Sozen S, Mendez R, Op den Camp HJM, Jetten MSM (2005) Propionate oxidation by and methanol inhibition of anaerobic ammonium-oxidizing bacteria. Appl Environ Microbiol 71(2):1066–1071 12. Jaroszynski LW, Cicek N, Sparling R, Oleszkiewicz JA (2011) Importance of the operating pH in maintaining the stability of anoxic ammonium oxidation (anammox) activity in moving bed biofilm reactors. Bioresour Technol 102: 7051-7056 13. Jin RC, Xing BS, Yu JJ, Qin TY, Chen SX (2013) The importance of the substrate ratio in the operation of the Anammox process in upflow biofilter. Ecol Eng 53:130– 137 14. Jin RC, Zheng P (2009) Kinetics of nitrogen removal in high rate anammox upflow filter. J Hazard Mater 170:652–656 15. Jung JY, Kang SH, Chung YC, Ahn DH (2007) Factors affecting the activity of anammox bacteria during start up in the continuous culture reactor. Water Sci Technol 55(1):459–468 16. Keluskar R, Nerurkar A, Desai A (2013) Development of a simultaneous partial nitrification, anaerobic ammonia oxidation and denitrification (SNAD) bench scale process for removal of ammonia from effluent of a fertilizer industry. Bioresour Technol 130: 390–397 17. Kimura Y, Isaka K, Kazama F, Sumino T (2010) Effect of nitrite inhibition on anaerobic ammonium oxidation. Appl Microbiol Biotechnol 86:359–365
12
18. Li H, Zhou S, Ma W, Huang G, Xu B (2012) Fast start-up of ANAMMOX reactor: operational strategy and some characteristics as indicators of reactor performance. Desalination 286:436–441 19. Magrí A, Béline F, Dabert P (2013) Feasibility and interest of the anammox process as treatment alternative for anaerobic digester supernatants in manure processing-An overview. J Environ Manage 131:170-184 20. Molinuevo B, Garcia MC, Karakashev D, Angelidaki I (2009) Anammox for ammonia removal from pig manure effluents: effect of organic matter content on process performance. Bioresour Technol 100:2171–2175 21. Ndegwa PM, Hamilton DW, Lalman JA, Cumba HJ (2005) Optimization of anaerobic sequencing batch reactors treating dilute swine slurries. Transactions of the ASAE 48(4): 1575−1583 22. Nhat PT, Biec HN, Mai NTT, Thanh BX, Dan NP (2014) Application of a partial nitritation and anammox system for the old landfill leachate treatment. International Biodeterioration & Biodegradation 95:144-150 23. Ni SQ, Ni JY, Hu DL, Sung S (2012) Effect of organic matter on the performance of granular anammox process. Bioresour Technol 110:701–705 24. Ni SQ, Lee PH, Sung S (2010) The kinetics of nitrogen removal and biogas production in an anammox non-woven membrane reactor. Bioresour Technol 101: 5767–5773 25. Strous M, Kuenen JG, Jetten MSM (1999) Key physiology of anaerobic ammonium oxidation. Appl Environ Microbiol 65:3248–3250 26. Suneethi S, Joseph K (2011) ANAMMOX process start up and stabilization with an anaerobic seed in Anaerobic Membrane Bioreactor (AnMBR). Bioresour Technol 102: 8860-8867 27. Tang CJ, Ping Z, Chai LY, Min XB (2013) Thermodynamic and kinetic investigation of anaerobic bioprocesses on ANAMMOX under high organic conditions. Chemical Engineering Journal 230:149–157 28. Tang CJ, Zheng P, Chen TT, Mahmood Q, Zhang JQ, Chen XG, Ding S, Chen JW, Wu DT (2011) Enhanced nitrogen removal from pharmaceutical wastewater using SBA-ANAMMOX process. Water Res. 45:201–210 29. Tang CJ, Zheng P, Mahmood Q, Chen JW (2010a) Effect of substrate concentration on stability of anammox biofilm reactors. J Cent South Univ Technol 17: 79–84 13
30. Tang CJ, Zheng P, Mahmood Q, Chen JW (2009) Start-up and inhibition analysis of the Anammox process seeded with anaerobic granular sludge. J Ind Microbiol Biotechnol 36:1093–1100 31. Toh SK, Ashbolt NJ (2002) Adaptation of anaerobic ammonium-oxidising consortium to synthetic coke-ovens wastewater. Appl Microbiol Biotechnol 59:344–352 32. Van de Graaf AA, de Bruijn P, Robertson LA, Jetten MSM, Kuenen JG (1996) Autotrophic growth of anaerobic ammonium-oxidizing micro-organisms in a fluidized bed reactor. Microbiology 142(8):2187–2196 33. Wang Z, Wang W, Zhang X, Zhang G (2009) Digestion of thermally hydrolyzed sewage sludge by anaerobic sequencing batch reactor. J Hazard Material 162(23):799-803 34. Waki M, Tokutomi T, Yokoyama H, Tanaka Y (2007) Nitrogen removal from animal waste treatment water by anammox enrichment. Bioresour Technol 98:2775–2780 35. Zu B, Zhang DJ, Yan Q (2008) Effect of traceNO2 and kinetic characteristics for anaerobic ammonium oxidation of granular sludge. Environ Sci 29:683–687
14
Figure Captions
Fig. 1(A) Schematic diagram of AHR
Fig. 1(B) Experimental set-up of AHR
Fig. 2 Performance of AHR at different HRTs
Fig. 3 Effect of COD/TN ratio on the performance of AHR
Fig. 4 Effluent Nitrate and alkalinity profile of AHR subjected to organic loads
Fig. 5 Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
Fig. 6 Validation of Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2N (C) COD
15
Table 1 Correlation analysis of NRE and various operational parameters in AHR Correlation Operational Parameters COD/TN ratio ARE NRE CRE Nitrate Alkalinity pH OLR
COD/TN ratio ARE NRE CRE Nitrate Alkalinity pH OLR * ** * 1 -.825 -.884 -.284 -.601 .725 -.347 1.000** 1 .992** .714* .886** -.966** .773* -.827* 1 .651 .852** -.941** .711* -.887** 1 .888** -.817* .982** -.289 1 -.964** .928** -.605 1 -.871** .728* 1 -.353 1
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).
Table 2 Summary of kinetic constants calculated from the Haldane model Kinetic Constants qmax (mgN/gVSS .d) Ks (mg/l) Ki (mg/l)
NH4-N
NO2-N
COD
References
Obsd 17.49
Rptd 38.65297.2
Obsd 22.56
Rptd 202.9304.7
Obsd 32.92
Rptd -
45.81
48.41– 87.1 887.11123
73.55
6.552– 15.39 159.5720.6
184.2
-
354.9
292400
930.2
220.8
Tang et al., 2009; Chen et al., 2011;Molinuevo et al. 2009; Ni et al. 2012; Zu et al. 2008
Table 3 Validation statistics of Haldane model Parameters Degree of freedom, df
NH4-N 8
NO2-N 8
COD 8
Standard deviation
101.7
49.8
21.7
Standard error
33.9
16.6
7.23
t statistic
1.50
1.11
0.74
t critical
2.30
2.30
2.30
Fig. 1A Schematic diagram of AHR
Fig. 1B Experimental set-up of AHR
ARE
NiRE
NRE
Removal Efficiency (%)
100 95 90 85 80 75 70 0
0.5
1
1.5
HRT (days)
Fig. 2 Performance of AHR at different HRTs
2
2.5
3
NRE
NiRE
% COD Removal 130 120
95
110 90
100
85
90 80
80
70
75
60 0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
COD/TN ratio
Fig. 3 Effect of COD/TN ratio on the performance of AHR
0.80
COD Removal (%)
Nitrogen Removal Efficiency (%)
100
Effluent Alkalinity
Nitrate Concentration (mg/l)
60
450 400 350 300 250 200 150 100 50 0
50 40 30 20 10 0 0
Alkalinity (mg/l)
NO3-N
100 200 300 400 500 600 700 800 900 1000 Influent COD (mg/l)
Fig. 4 Effluent Nitrate and alkalinity profile of AHR subjected to organic loads
Haldane Kinetics for NO2-N
A
15
2
R2=0.970, qmax =22.56, Ks =73.55, Ki=220.8
q (mg N/gVSS d)
q (mg N/g VSSd)
R =0.972, qmax =17.49, Ks =45.81, Ki=930.2 10
5
0
10
5
100
200
SA (mg/L)
300
400
R2=0.984, qmax =32.92, Ks =184.2, Ki=354.9
15
10
5
0
0 0
Haldane Kinetics for COD
B q (mg N/g VSS d)
Haldane Kinetics for NH4-N 15
0
50
100
150
SA (mg/L)
200
250
0
50
100
150
SA (mg/L)
Fig. 5 Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
200
250
C
A
350 300 250 200 150 100 50 0 0
500
1000
1500
Influent concentration (mg/l)
300
B
200 150 100 50 0
Effluent concentration (mg/l)
400
Effluent concenteration (mg/l)
Effluent concentration (mg/l)
250
450
C
250 200 150 100 50 0
0
500
1000
1500
Influent concenteration (mg/l)
0
500 1000 1500 Influent concentration (mg/l)
Fig. 6 Validation of Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
NiRE
% COD Removal 130
100.0
120
95.0
110
90.0
100 90
85.0
80
80.0
70
75.0
Operational COD/T Parameters N ratio ARE
COD Removal (%)
COD/TN ratio ARE
1
CRE
5
0
*
.773* -.827*
.651
.852**
-.941**
.711*
1
.888**
-.817*
.982** -.289
1
-.964**
.928** -.605
1
.728* .871** 1
.887**
-.353 1
Pearson correlation analysis of various operational parameters in AHR
1.02N2 + 0.26NO3- + 0.066CH2O0.5N0.15 + 2.3H2O
400
Haldane Kinetics for COD
B
R2=0.970, qmax =22.56, Ks=73.55, Ki=220.8 10
5
R2=0.984, qmax =32.92, Ks =184.2, Ki=354.9
15
10
5
0
0
SA (mg/L)
-.347
-.966**
Alkalinity
15
q (mg N/gVSS d)
10
300
.725*
.714* .886**
Nitrate
Haldane Kinetics for NO2-N
A
R2=0.972, qmax =17.49, Ks =45.81, Ki=930.2
200
-.601
OLR 1.000*
OLR
NH4+ + 1.32 NO2- + 0.066HCO3- + 0.13H+
100
.992**
-.284
pH
ANAMMOX
Experimental Set-up of AHR
q (mg N/g VSSd)
1
NRE
Effect of COD/TN ratio on the performance of AHR
0
-.825* -.884**
CRE Nitrate Alkalinity
pH
COD/TN ratio
15
NRE
60 0.000.100.200.300.400.500.600.700.80
Haldane Kinetics for NH4-N
1
q (mg N/g VSS d)
Nitrogen Removal Efficiency (%)
Correlation
NRE
0
50
100
150
SA (mg/L)
200
250
0
50
100
150
200
SA (mg/L)
Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
250
C
HIGHLIGHTS
• • • • • •
AHR demonstrated high nitrogen removal (95.1%) at optimum HRT of 1 d Addition of anaerobic granular sludge (1:5v/v) enhanced bacterial tolerance to OM Symbiosis of anammox and anaerobes resulted 96.8% NH4-N & 94.8% COD removal Haldane model successfully evaluated process inhibtion in AHR Inhibition modelling demonstrated COD as the major inhibitor to anammox process. The optimal removal of OM and nitrogen was observed at COD/N ratio of 0.54
16
S0960-8524(15)01163-3 http://dx.doi.org/10.1016/j.biortech.2015.08.057 BITE 15410
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
27 June 2015 10 August 2015 12 August 2015
Please cite this article as: Tomar, S., Gupta, S.K., Mishra, B.K., A novel strategy for simultaneous removal of nitrogen and organic matter using anaerobic granular sludge in anammox hybrid reactor, Bioresource Technology (2015), doi: http://dx.doi.org/10.1016/j.biortech.2015.08.057
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A novel strategy for simultaneous removal of nitrogen and organic matter using anaerobic granular sludge in anammox hybrid reactor Swati Tomar1, Sunil Kumar Gupta2*, Brijesh Kumar Mishra3 1
Senior Research Fellow, Department of Environmental Science & Engineering, Indian School of
Mines, Dhanbad- 826004, Email: [email protected], Phone No.: +91-326-2235474, Fax: +91326-2296624 2*
Associate Professor, Department of Environmental Science & Engineering, Indian School of Mines,
Dhanbad- 826004, Email: [email protected], Phone No.: +91-326-2235474, Fax: +91-3262296624 (Corresponding Author) 3
Assistant Professor, Department of Environmental Science & Engineering, Indian School of Mines,
Dhanbad- 826004, Email: [email protected], Phone No.: +91-326-2235752, Fax: +91-3262296624
Abstract The coexistence of organic matter (OM) and nitrogen in industrial effluent is the major bottleneck in field application of anammox process. The present study emphasized on investigating the role of seeding anaerobic granular sludge towards simultaneous removal of ammonium and OM in anammox hybrid reactor (AHR). The study delineated simultaneous reduction of both OM (94.8%) and nitrogen (96.8%) at optimal COD/N ratio (0.54). Pearson correlation matrix showed positive and strong correlation of ARE (ammonium removal efficiency) and CRE (COD removal efficiency) with NRE (nitrogen removal efficiency). The negative correlation of OLR and COD/TN ratio with NRE indicated that increase in organic loadings may suppress anammox activity. The process inhibition was evaluated using Haldane model considering free ammonia, OM and nitrite as inhibitors. The strategy of using anaerobic granular sludge not only augmented endurance of bacterial communities against OM inhibition but also facilitated simultaneous removal of OM and nitrogen.
Keywords: Anammox, AHR, organic matter, nitrogen removal, modelling, inhibition kinetics.
INTRODUCTION 1
Over the last decade, nitrogen removal from wastewater has become a worldwide emerging concern pertaining to its toxic nature to aquatic species and eutrophication in receiving water body. The conventional method of nitrogen removal involves nitrification and denitrification, which is characterized by high energy demand, operational cost and problems of sludge disposal. In addition, the application of ammonia stripper is limited as a result of significantly lower stripping efficiency (50-65%) and the associated problem of air pollution. Anammox is a novel microbiological approach that has changed the traditional concept of biological nitrogen removal (Fernandez et al. 2012). The process facilitates direct oxidation of ammonium nitrogen into nitrogen gas under anoxic conditions with nitrite as an electron acceptor. Significant reduction in aeration costs, exogenous electron donor saving and low sludge production makes the process techno-economically feasible over existing conventional treatment technologies. While, the newly discovered anammox process opens up the new possibilities for nitrogen removal from wastewater, the industrial applications are constrained due to presence of organic matter (OM) in the industrial effluent (Magri et al. 2013), which has been reported to cause deterioration in anammox activity (Ni et al. 2012). Industrial effluents, such as coke ovens (Toh and Ashbolt 2002), fertiliser (Keluskar et al. 2013), antibiotics (Tang et al. 2011) contain high concentration of OM varying from 46.6-2200 mg/l. Hence, there is an acute need to evolve a novel technique which can facilitate simultaneous removal of ammonium and OM from the industrial effluent, if anammox is applied as a treatment alternative. Several researchers (Ni et al. 2012; Molinuevo et al. 2009) have investigated the effects of OM on anammox process and revealed that it poses severe threat to anammox bacteria. Guven et al. (2005) revealed that even 0.5 mM of methanol resulted in immediate and complete inactivation of anammox activity. Similarly, Molinuevo et al. (2009) also reported that COD concentrations upto 292 mg/l led to complete inhibition of anammox process. The inhibition is mainly caused due to higher affinity of nitrite to denitrifiers than anammox bacteria leading to upset in substrate co-substrate ratio which is essentially required for smooth functioning of the process. Extensive literature review suggests that process inhibition is also caused by FA (free ammonia) and nitrite in addition to COD (Fernandez et al. 2012; Kimura et al. 2010). FA concentration of 57-187 mg/l is sufficient to induce toxicity in anammox process (Tang et al. 2010a). Kimura et al. (2010) investigated the effects of nitrite on anammox bacteria entrapped in gel carriers using batch and continuous feeding
2
tests and reported that nitrite concentration beyond 274 mg/l was found inhibitory to anammox bacteria. Several researchers advocated that the effect of inhibition can significantly be minimised by extensive acclimation and use of granular anammox sludge in addition to intermittent seeding of the reactor. Ni et al. (2012) revealed that granular anammox sludge demonstrated higher tolerance to OM as compared to flocculant sludge and was capable of delivering ammonium removal of 80% even at a higher COD to N ratio of 3:1. Seeding of granular sludge significantly contributed towards enhanced COD removal efficiency (CRE) in addition to its capability to sustain high organic loads (Molinuevo et al. 2009). Abbasi and Abbasi (2012) realized that granular sludge of appropriate particle size, particle density, and microfilm characteristics enhances the reactor efficiency in addition to reduction in start-up time. No such study has been reported in literature to investigate the simultaneous removal of OM and ammonical nitrogen by establishing the proper symbiosis amongst anammox, denitrifiers and anaerobic bacteria. The symbiotic association of these bacteria may help in simultaneous removal of both OM and ammonical nitrogen from the industrial effluent. Hence, the present study was undertaken to explore the feasibility of addition of anaerobic granular sludge to make the process viable for its industrial applications.
MATERIAL AND METHODS
Experimental Set-up of AHR
The experimental setup of AHR is shown in Fig. 1. The reactor was fabricated of transparent acrylic plastic with an internal diameter of 10 cm and height 65 cm. The total working volume was 5 litres. Corrugated polyvinyl chloride (PVC) pipes of length 2.25 cm and diameter 2.25 cm were used as filter media. Total 55 nos. of PVC carriers were added to the reactor to constitute an attached growth system. The sludge blanket in the lower half of the reactor constitutes suspended growth system while filter media in the upper part provides attached growth for the microorganisms. The reactor was also completely covered with black cloth to avoid the growth of phototrophic organisms and oxygen production. The reactor was fed with the synthetic wastewater (van de Graaf et al. 1996) using a peristaltic pump to maintain a constant flow rate.
3
Fig. 1(A) Schematic diagram of AHR
Fig. 1(B) Experimental set-up of AHR
Strategy of operation
The reactors were started with initial influent ammonium and nitrite concentrations of 100 mg/l each to maintain an optimal NH4/NO2 ratio of 1:1 at an HRT of 1d. The reactors were operated for a period of 30 days till the attainment of pseudo-steady state condition. Then, the ammonium and nitrite concentrations were gradually increased to 600 mg/l each, and maintained constant to assess the performance of reactor during variable hydraulic retention times (HRTs). During this study, the efficiency of the reactor in terms of ammonium, nitrite removal was observed to devise the optimal HRT required for the efficient functioning of the reactor. After attainment of optimal HRT, the reactors were seeded with anaerobic granular sludge (1:5 v/v) to establish the mixed consortia of anammox and anaerobic bacteria. The anaerobic granular sludge was collected from high rate UASB (upflow anaerobic sludge blanket) reactor treating coke oven wastewater. The mixture was then homogenised and acclimated to higher organic COD by gradual addition of glucose in synthetic wastewater. The optimal COD/TN ratio is very essential for developing a symbiotic relation between different communities of bacteria for its efficient functioning. The COD/TN ratio in the study was varied from 0.04-0.83 by varying the influent COD concentration from 50 to 1000 mg/L, till process inhibition was observed. The process inhibition was studied using Haldane model considering FA, nitrite and COD as inhibitors to explore the major cause of inhibition.
Analytical Methods
The analysis of pH, alkalinity, COD, NH4-N, NO2-N, NO3-N, VSS and TSS were carried out as per the standard methods (APHA 2012). The volume of gas produced in the reactors was measured by water displacement method and the composition was analysed through Gas Chromatograph (GC) equipped with thermal conductivity detector (TCD) using DPK Spherocarp column. The GC was operated at injector and detector temperatures of 120ºC and 150ºC, respectively. H2 @ 20 ml/min was used as a carrier gas.
Statistical Analysis 4
Pearson correlation matrix (significance levels α = 0.05, 0.01) was used to determine the correlation between nitrogen removal efficiency (NRE), and operational parameters i.e., pH, Alkalinity, COD/TN ratio, OLR (organic loading rate), CRE, NO3-N, etc. using SPSS 16.0. The biasness of the kinetic model was tested using t test. The values of bio-kinetic coefficients were determined using Graph pad prism software (Version-5.03) for non-linear regression method.
Quality control/quality assurance procedure (QA/QC)
The laboratory reagent blanks were prepared and analyzed to determine if any interference was present in the effluent samples. The precision of the measurements were estimated using triplicate sample analysis. The relative percentage difference (RPD) between two parallel samples was calculated and cross verified. In case the RPD exceeded > 5%, the samples were recollected and analyzed. The average of the triplicate readings was reported as the final value. Continuous calibration checks were performed during GC analysis after injection of every 10 samples. If the RPD between the response of the initial calibration and the calibration check standard was > 10%, the instrument was considered as out of calibration, and was recalibrated. The high precision gas standard procured from Chemtron Science Laboratories Pvt. Ltd., Mumbai was used for calibration.
Haldane Model
A number of substrates serve as nutrients at low concentrations, but act as inhibitors at high concentrations (Edwards 1970). High substrate concentration inhibits bacterial growth and deteriorates anammox performance. This was evaluated using Haldane model which can be represented by Eq. (1):
=
(1)
1+ +
5
where, q, qmax is specific substrate conversion rate and maximum specific conversion rate, respectively (1/d); Ks is half saturation constant (mg/L); Ki is substrate inhibition constant (mg/L) and S is the substrate concentration (mg/l).
RESULTS AND DISCUSSION
Effect of HRT on the performance of AHR
Optimisation of HRT is very essential to evolve most economical design of bioreactor (Chaganti et al. 2013; Ndegwa et al. 2005). Shorter HRT assists smaller reactor volume, and hence, reduces the construction costs (Wang et al. 2009). Results indicated that the NRE increased with increase in HRT and was maximum (97.5%) at HRT of 2.25 d (Fig. 2). However, beyond HRT of 1d, the rate of increase was not significant. Further increase in HRT beyond 2.25 d, resulted in substantial decline in NRE, which may be attributed to lower availability of the substrate. Hence, the HRT of 1.0 day was considered optimal corresponding to substantial nitrogen removal of 95.1%. Ammonium removal also depicted similar profile. However, nitrite removal efficiency (NiRE) was not affected significantly. A possible reason could be the consumption of nitrite by heterotrophic denitrification pathway. The efficiency of the reactor observed in our study is significantly higher than reported (8089.9%) in the literature (Ni et al. 2010; Duan et al. 2012). The higher efficiency observed in our system may be attributed to the higher biomass retention and attached filter media which might have enhanced the NRE. Duan et al. (2012) reported that the use of non-woven carrier as AGM effectively increased the biomass retention by 2.8% and overall NRR by 8.1 % in the hybrid reactor. Grandhi et al. (2011) also investigated the performance of UASB and AHR for the treatment of distillery spent wash and reported that the hybrid reactor configuration contributed an additional 5% COD removal and 25% reduction in sludge washout rate. It was also observed that at lower HRT < 1 day, the pH of the reactor was 8.8±0.2, and FA concentration varied from 9.4 mg/l to 15.7 mg/l. Jung et al. (2007) reported that FA concentration of 1.7 mg/l is sufficient to induce toxicity in anammox process. While, Waki et al. (2007) and Tang et al. (2010a) reported comparatively higher toxicity range of FA varying from 13–187 mg/l. This might be cause for upset in reactor performance at lower HRT
6
However, the effect of process inhibition due to FA was not substantial in our study, as NRE never dropped below 78% even at minimum HRT of 0.25d. The nitrate production which is considered as one of the bottlenecks of anammox process did not vary significantly and found comparatively lower (37.2 to 52.3 mg/l) than reported in the literature (Jin et al. 2013; Jin and Zheng 2009). The lower concentration of nitrate observed in our study may be attributed to lower availability of nitrite nitrogen due to consistently higher nitrite removal (95-98%) mediated by anammox pathway. Strous et al. (1999) reported that excess nitrite in the system results in formation of nitrate and hydroxylamine in the process. The larger fluctuation of nitrite is also unfavorable for the steady operation of the system.
Fig. 2 Performance of AHR at different HRTs
Effect of seeding of anaerobic granular sludge towards simultaneous removal of OM and nitrogen Both COD concentrations and COD/TN ratio play an instrumental role and affect the overall efficiency of anammox process. The analysis of data revealed that NRE did not fluctuate much with increase in COD/TN ratio from 0.04 to 0.54 (Fig. 3). However, the CRE increased consistently upto COD/TN ratio of 0.54. The simultaneous removal of both COD and nitrogen observed in our study indicated establishment of symbiosis between anammox and anaerobic bacteria. But when the ratio was further increased, both CRE and NRE declined at elevated COD concentration (> 650 mg/l). The, optimal COD/TN ratio corresponding to simultaneous removal of COD and nitrogen was 0.54. Decline in effluent nitrate coupled with increase in alkalinity at high organic load, also dictated deterioration in anammox performance (Fig. 4). The process might have been inhibited due to increase in COD concentration at higher COD/TN ratio. Several researchers have reported that OM poses adverse effects on the activity of anammox bacteria. Chamchoi et al. (2008) studied the effect of OM on anammox performance and reported that COD concentration over 300 mg/l could inactivate or even eradicate anammox communities. Molinuevo et al. (2009) also reported that COD concentrations upto 292 mg/l led to complete inhibition of anammox process. However, the mechanism of anammox suppression in our study outcompeted others. This may be due to seeding of anaerobic granular sludge which might have enhanced the endurance of the reactor towards OM, as is evident from comparatively higher CRE (84.57
94.8%) observed in our study than reported (30-56.5%) in the literature (Nhat et al. 2014; Chen et al. 2013). The robustness of AHR to sustain high organic loads (650mgCOD/L) in our study might be attributed to the seeding of granular anaerobic sludge which resulted in establishing harmonious symbiotic association between different communities of anammox bacteria, anaerobes and heterotrophic denitrifiers.
Fig. 3 Effect of COD/TN ratio on the performance of AHR
Fig. 4 Effluent Nitrate and alkalinity profile of AHR subjected to organic loads
Correlation of process parameters with NRE
Pearson correlation analysis revealed that NRE is significantly influenced by various operating parameters (Table 1). The nitrate concentration and pH showed strong and positive correlation with NRE. Many researchers (Li et al. 2012; Jaroszynski et al. 2011) also indicated that increase in effluent nitrate and pH are the indicators of successful anammox process. ARE and CRE both showed positive correlation with NRE indicating simultaneous reduction of ammonium and COD in AHR. A possible reason could be coexistence of anammox autotrophs, anaerobic consortia and denitrifiers which simultaneously use ammonical nitrogen, OM and nitrite nitrogen as a substrate in their removal pathways. However, the higher correlation of NRE with ARE than CRE, suggested predominance of anammox pathway. The parameters alkalinity, OLR and COD/N ratio showed strong but negative correlation. The negative correlation of OLR and COD/TN ratio with NRE indicated that increased organic loading rates may suppress the anammox activity leading to deterioration of reactor performance. At higher COD/TN ratio, denitrifiers are more prevalent and outcompeted anammox bacteria due to their higher affinity to nitrite nitrogen. This results in upset in the performance of anammox consortia due to less availability of nitrite nitrogen as a co-substrate. Alkalinity also showed strong but negative correlation. This may be attributed to progressive consumption of bicarbonate as a part of enrichment medium for the growth and development of anammox populations (Suneethi and Joseph 2011)
Table 1 Correlation analysis of NRE and various operational parameters in AHR 8
Assessment of process inhibition using Haldane model
Process kinetics and modelling are the important tools for designing and predicting the performance of bioreactor. Process inhibition kinetics was evaluated considering COD, nitrite and FA as major inhibitors. The plots of Haldane model showed higher correlation for COD followed by ammonium and nitrite (Fig. 5). The maximum specific conversion rate for ammonium and nitrite was comparatively lower than reported values (Tang et al., 2009; Chen et al., 2011) as shown in Table 2. This may be attributed to the process inhibition caused due to OM and low availability of substrate in our study. Half saturation constant for ammonium was in line with the reported values but, it was significantly higher for nitrite nitrogen (Zu et al. 2008). This may be due to excess utilisation of nitrite nitrogen by denitrifiers and anammox bacteria during process inhibition conditions. The Ki values of ammonium, nitrite and COD dictated that the reactor exhibited highest tolerance for ammonical nitrogen followed by COD and nitrite. The values of Ki with OM inhibition were still comparable with those reported in literature (Zu et al. 2008; Chen et al. 2011; Molinuevo et al. 2009; Ni et al. 2012) indicating enhanced endurance of AHR towards OM. Accordingly, the Haldane model fit for NH4-N, NO2-N and COD can be represented by Eqs. (2-4): 17.49 (2) 45.81
1 + + 930.2 22.56 = (3) 73.55
1+ +
220.8 32.92 = (4) 184.2
1 + + 354.9 =
Table 2 Summary of kinetic constants calculated from the Haldane model
Fig. 5 Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
Validation of Model The models were validated by predicting the effluent ammonium, nitrite and COD concentration for a new set of values excluding the ones which were used for determination of kinetics. The plot between observed and predicted values demonstrated strong correlation 9
with COD followed by ammonium, but it was found poor in case of nitrite (Fig. 6). The standard error of prediction was also minimum in case of COD (Table 3). This indicates that COD is the most influential parameter which governs process inhibition in AHR. The chances of inhibition due to FA was minimised due to to the absence of excessive alkaline conditions which promotes the chance of FA inhibition. The FA concentration throughout the study varied from 3.4-45 mg/l at corresponding pH of 8.1-8.6. The possibility of inhibition due to nitrite can also be curtailed as is evident from consistently higher nitrite removal throughout the study. It can also be seen that when, nitrite was considered as inhibitory substrate, the observed and predicted effluent nitrite fell close to each other till a concentration of 800 mg/l, beyond which the model failed and predicted significantly lower nitrite concentration. Thus, Haldane model can suitably be applied to assess the fate and process inhibition caused due to COD. To determine the biasness of the model, t test was also performed. The tstats value was much lower than tcrit indicating that the model is unbiased. A hypothesis was proposed in this study which declares that process inhibition can be minimised by optimising the ratio of seed sludge in the bioreactor. The fine tuning of ratio of seed sludge and anoxic sludge might result in proficient symbiotic association between bacterial communities in addition to eradication of process inhibition caused by OM. The study on the proposed hypothesis is ongoing by varying ratio of seed sludge/anammox sludge to arrive at the optimal ratio.
Fig. 6 Validation of Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
Table 3 Validation statistics of Haldane model
CONCLUSION
AHR dictated enormous potential towards nitrogen removal (95.1%) at optimum HRT of 1 day. Proficient symbiosis of anammox bacteria with anaerobes facilitated simultaneous removal of OM and nitrogen. The optimal performance was observed at COD/N ratio of 0.54 beyond which process inhibition was instrumental. Haldane model successfully evaluated process inhibition with COD as the major inhibitor than FA and nitrite. Addition of anaerobic 10
granular sludge (1:5v/v) promoted appreciable bacterial endurance towards OM inhibition. The study opens the door for researchers and scientists for optimizing simultaneous removal of OM and nitrogen which is a major bottleneck for field-scale application of anammox.
ACKNOWLEDGEMENT The authors thank the financial support from Indian School of Mines, Dhanbad under Junior Research Fellowship scheme funded by Ministry of Human Resource Development (MHRD), Government of India, New Delhi, for carrying out the this study.
CONFLICT OF INTEREST
Authors declare that they have no conflict of interest.
REFERENCES
1. Abbasi T, Abbasi SA (2012) Formation and impact of granules in fostering clean energy production and wastewater treatment in upflow anaerobic sludge blanket (UASB) reactors. Renewable and Sustainable Energy Reviews 16:1696– 1708 2. APHA, AWWA, WEF (2012) Standard Methods for Examination of Water and Wastewater, 22nd ed. United Book Press, USA 3. Chaganti SR, Pendyala B, Lalman JA, Veeravalli SS, Heath DD (2013) Influence of linoleic acid, pH and HRT on anaerobic microbial populations and metabolic shifts in ASBRs during dark hydrogen fermentation of lignocellulosic sugars. International Journal of Hydrogen Energy 38:2212-2220 4. Chamchoi N, Nitisorvut S, Schmidt JE (2008) Inactivation of ANAMMOX communities under concurrent operation of anaerobic ammonium oxidation (ANAMMOX) and denitrification. Bioresour Technol 99:3331–3336 5. Chen C, Huang X, Lei C, Zhang TC, Wu W (2013) Effect of organic matter strength on anammox for modified greenhouse turtle breeding wastewater treatment plant. Bioresour Technol 148:172-179 6. Chen T, Zheng P, Shen L, Ding S, Mahmood Q (2011) Kinetic characteristics and microbial community of anammox-ESGB reactor. J Hazard Mater 190:28-35 11
7. Duan X, Zhou J, Qiao S, Yin X, Tian T, Xu F (2012) Start-up of the anammox process from the conventional activated sludge in a hybrid bioreactor. J Environ Sci 24:1083-1090 8. Edwards VH (1970) Influence of high substrate concentrations on microbial kinetics. Biotechnol Bioeng 12:679–712 9. Fernandez I, Dosta J, Fajardo C, Campos JL, Mosquera-Corral A, Méndez R (2012) Short- and long-term effects of ammonium and nitrite on the Anammox process. J Environ Manage 95:S170-S174 10. Grandhi SC, Pandey LMS, Gupta SK, Singh G (2011) Comparative evaluation of high rate anaerobic processes for treatment of distillery spent wash. J Indus Res Technol 1(1):17-23 11. Güven D, Dapena A, Kartal B, Schmid MC, Maas B, van de Pas-Schoonen K, Sozen S, Mendez R, Op den Camp HJM, Jetten MSM (2005) Propionate oxidation by and methanol inhibition of anaerobic ammonium-oxidizing bacteria. Appl Environ Microbiol 71(2):1066–1071 12. Jaroszynski LW, Cicek N, Sparling R, Oleszkiewicz JA (2011) Importance of the operating pH in maintaining the stability of anoxic ammonium oxidation (anammox) activity in moving bed biofilm reactors. Bioresour Technol 102: 7051-7056 13. Jin RC, Xing BS, Yu JJ, Qin TY, Chen SX (2013) The importance of the substrate ratio in the operation of the Anammox process in upflow biofilter. Ecol Eng 53:130– 137 14. Jin RC, Zheng P (2009) Kinetics of nitrogen removal in high rate anammox upflow filter. J Hazard Mater 170:652–656 15. Jung JY, Kang SH, Chung YC, Ahn DH (2007) Factors affecting the activity of anammox bacteria during start up in the continuous culture reactor. Water Sci Technol 55(1):459–468 16. Keluskar R, Nerurkar A, Desai A (2013) Development of a simultaneous partial nitrification, anaerobic ammonia oxidation and denitrification (SNAD) bench scale process for removal of ammonia from effluent of a fertilizer industry. Bioresour Technol 130: 390–397 17. Kimura Y, Isaka K, Kazama F, Sumino T (2010) Effect of nitrite inhibition on anaerobic ammonium oxidation. Appl Microbiol Biotechnol 86:359–365
12
18. Li H, Zhou S, Ma W, Huang G, Xu B (2012) Fast start-up of ANAMMOX reactor: operational strategy and some characteristics as indicators of reactor performance. Desalination 286:436–441 19. Magrí A, Béline F, Dabert P (2013) Feasibility and interest of the anammox process as treatment alternative for anaerobic digester supernatants in manure processing-An overview. J Environ Manage 131:170-184 20. Molinuevo B, Garcia MC, Karakashev D, Angelidaki I (2009) Anammox for ammonia removal from pig manure effluents: effect of organic matter content on process performance. Bioresour Technol 100:2171–2175 21. Ndegwa PM, Hamilton DW, Lalman JA, Cumba HJ (2005) Optimization of anaerobic sequencing batch reactors treating dilute swine slurries. Transactions of the ASAE 48(4): 1575−1583 22. Nhat PT, Biec HN, Mai NTT, Thanh BX, Dan NP (2014) Application of a partial nitritation and anammox system for the old landfill leachate treatment. International Biodeterioration & Biodegradation 95:144-150 23. Ni SQ, Ni JY, Hu DL, Sung S (2012) Effect of organic matter on the performance of granular anammox process. Bioresour Technol 110:701–705 24. Ni SQ, Lee PH, Sung S (2010) The kinetics of nitrogen removal and biogas production in an anammox non-woven membrane reactor. Bioresour Technol 101: 5767–5773 25. Strous M, Kuenen JG, Jetten MSM (1999) Key physiology of anaerobic ammonium oxidation. Appl Environ Microbiol 65:3248–3250 26. Suneethi S, Joseph K (2011) ANAMMOX process start up and stabilization with an anaerobic seed in Anaerobic Membrane Bioreactor (AnMBR). Bioresour Technol 102: 8860-8867 27. Tang CJ, Ping Z, Chai LY, Min XB (2013) Thermodynamic and kinetic investigation of anaerobic bioprocesses on ANAMMOX under high organic conditions. Chemical Engineering Journal 230:149–157 28. Tang CJ, Zheng P, Chen TT, Mahmood Q, Zhang JQ, Chen XG, Ding S, Chen JW, Wu DT (2011) Enhanced nitrogen removal from pharmaceutical wastewater using SBA-ANAMMOX process. Water Res. 45:201–210 29. Tang CJ, Zheng P, Mahmood Q, Chen JW (2010a) Effect of substrate concentration on stability of anammox biofilm reactors. J Cent South Univ Technol 17: 79–84 13
30. Tang CJ, Zheng P, Mahmood Q, Chen JW (2009) Start-up and inhibition analysis of the Anammox process seeded with anaerobic granular sludge. J Ind Microbiol Biotechnol 36:1093–1100 31. Toh SK, Ashbolt NJ (2002) Adaptation of anaerobic ammonium-oxidising consortium to synthetic coke-ovens wastewater. Appl Microbiol Biotechnol 59:344–352 32. Van de Graaf AA, de Bruijn P, Robertson LA, Jetten MSM, Kuenen JG (1996) Autotrophic growth of anaerobic ammonium-oxidizing micro-organisms in a fluidized bed reactor. Microbiology 142(8):2187–2196 33. Wang Z, Wang W, Zhang X, Zhang G (2009) Digestion of thermally hydrolyzed sewage sludge by anaerobic sequencing batch reactor. J Hazard Material 162(23):799-803 34. Waki M, Tokutomi T, Yokoyama H, Tanaka Y (2007) Nitrogen removal from animal waste treatment water by anammox enrichment. Bioresour Technol 98:2775–2780 35. Zu B, Zhang DJ, Yan Q (2008) Effect of traceNO2 and kinetic characteristics for anaerobic ammonium oxidation of granular sludge. Environ Sci 29:683–687
14
Figure Captions
Fig. 1(A) Schematic diagram of AHR
Fig. 1(B) Experimental set-up of AHR
Fig. 2 Performance of AHR at different HRTs
Fig. 3 Effect of COD/TN ratio on the performance of AHR
Fig. 4 Effluent Nitrate and alkalinity profile of AHR subjected to organic loads
Fig. 5 Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
Fig. 6 Validation of Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2N (C) COD
15
Table 1 Correlation analysis of NRE and various operational parameters in AHR Correlation Operational Parameters COD/TN ratio ARE NRE CRE Nitrate Alkalinity pH OLR
COD/TN ratio ARE NRE CRE Nitrate Alkalinity pH OLR * ** * 1 -.825 -.884 -.284 -.601 .725 -.347 1.000** 1 .992** .714* .886** -.966** .773* -.827* 1 .651 .852** -.941** .711* -.887** 1 .888** -.817* .982** -.289 1 -.964** .928** -.605 1 -.871** .728* 1 -.353 1
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).
Table 2 Summary of kinetic constants calculated from the Haldane model Kinetic Constants qmax (mgN/gVSS .d) Ks (mg/l) Ki (mg/l)
NH4-N
NO2-N
COD
References
Obsd 17.49
Rptd 38.65297.2
Obsd 22.56
Rptd 202.9304.7
Obsd 32.92
Rptd -
45.81
48.41– 87.1 887.11123
73.55
6.552– 15.39 159.5720.6
184.2
-
354.9
292400
930.2
220.8
Tang et al., 2009; Chen et al., 2011;Molinuevo et al. 2009; Ni et al. 2012; Zu et al. 2008
Table 3 Validation statistics of Haldane model Parameters Degree of freedom, df
NH4-N 8
NO2-N 8
COD 8
Standard deviation
101.7
49.8
21.7
Standard error
33.9
16.6
7.23
t statistic
1.50
1.11
0.74
t critical
2.30
2.30
2.30
Fig. 1A Schematic diagram of AHR
Fig. 1B Experimental set-up of AHR
ARE
NiRE
NRE
Removal Efficiency (%)
100 95 90 85 80 75 70 0
0.5
1
1.5
HRT (days)
Fig. 2 Performance of AHR at different HRTs
2
2.5
3
NRE
NiRE
% COD Removal 130 120
95
110 90
100
85
90 80
80
70
75
60 0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
COD/TN ratio
Fig. 3 Effect of COD/TN ratio on the performance of AHR
0.80
COD Removal (%)
Nitrogen Removal Efficiency (%)
100
Effluent Alkalinity
Nitrate Concentration (mg/l)
60
450 400 350 300 250 200 150 100 50 0
50 40 30 20 10 0 0
Alkalinity (mg/l)
NO3-N
100 200 300 400 500 600 700 800 900 1000 Influent COD (mg/l)
Fig. 4 Effluent Nitrate and alkalinity profile of AHR subjected to organic loads
Haldane Kinetics for NO2-N
A
15
2
R2=0.970, qmax =22.56, Ks =73.55, Ki=220.8
q (mg N/gVSS d)
q (mg N/g VSSd)
R =0.972, qmax =17.49, Ks =45.81, Ki=930.2 10
5
0
10
5
100
200
SA (mg/L)
300
400
R2=0.984, qmax =32.92, Ks =184.2, Ki=354.9
15
10
5
0
0 0
Haldane Kinetics for COD
B q (mg N/g VSS d)
Haldane Kinetics for NH4-N 15
0
50
100
150
SA (mg/L)
200
250
0
50
100
150
SA (mg/L)
Fig. 5 Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
200
250
C
A
350 300 250 200 150 100 50 0 0
500
1000
1500
Influent concentration (mg/l)
300
B
200 150 100 50 0
Effluent concentration (mg/l)
400
Effluent concenteration (mg/l)
Effluent concentration (mg/l)
250
450
C
250 200 150 100 50 0
0
500
1000
1500
Influent concenteration (mg/l)
0
500 1000 1500 Influent concentration (mg/l)
Fig. 6 Validation of Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
NiRE
% COD Removal 130
100.0
120
95.0
110
90.0
100 90
85.0
80
80.0
70
75.0
Operational COD/T Parameters N ratio ARE
COD Removal (%)
COD/TN ratio ARE
1
CRE
5
0
*
.773* -.827*
.651
.852**
-.941**
.711*
1
.888**
-.817*
.982** -.289
1
-.964**
.928** -.605
1
.728* .871** 1
.887**
-.353 1
Pearson correlation analysis of various operational parameters in AHR
1.02N2 + 0.26NO3- + 0.066CH2O0.5N0.15 + 2.3H2O
400
Haldane Kinetics for COD
B
R2=0.970, qmax =22.56, Ks=73.55, Ki=220.8 10
5
R2=0.984, qmax =32.92, Ks =184.2, Ki=354.9
15
10
5
0
0
SA (mg/L)
-.347
-.966**
Alkalinity
15
q (mg N/gVSS d)
10
300
.725*
.714* .886**
Nitrate
Haldane Kinetics for NO2-N
A
R2=0.972, qmax =17.49, Ks =45.81, Ki=930.2
200
-.601
OLR 1.000*
OLR
NH4+ + 1.32 NO2- + 0.066HCO3- + 0.13H+
100
.992**
-.284
pH
ANAMMOX
Experimental Set-up of AHR
q (mg N/g VSSd)
1
NRE
Effect of COD/TN ratio on the performance of AHR
0
-.825* -.884**
CRE Nitrate Alkalinity
pH
COD/TN ratio
15
NRE
60 0.000.100.200.300.400.500.600.700.80
Haldane Kinetics for NH4-N
1
q (mg N/g VSS d)
Nitrogen Removal Efficiency (%)
Correlation
NRE
0
50
100
150
SA (mg/L)
200
250
0
50
100
150
200
SA (mg/L)
Haldane model for substrate inhibition kinetics (A) NH4-N (B) NO2-N (C) COD
250
C
HIGHLIGHTS
• • • • • •
AHR demonstrated high nitrogen removal (95.1%) at optimum HRT of 1 d Addition of anaerobic granular sludge (1:5v/v) enhanced bacterial tolerance to OM Symbiosis of anammox and anaerobes resulted 96.8% NH4-N & 94.8% COD removal Haldane model successfully evaluated process inhibtion in AHR Inhibition modelling demonstrated COD as the major inhibitor to anammox process. The optimal removal of OM and nitrogen was observed at COD/N ratio of 0.54
16