- Email: [email protected]
Simultaneous biological removal of nitrogen and phosphorus in a vertical bioreactor
Journal of Environmental Chemical Engineering 4 (2016) 130–136
Contents lists available at ScienceDirect
Journal of Environmental Chemical Engineering journal homepage: www.elsevier.com/locate/jece
Simultaneous biological removal of nitrogen and phosphorus in a vertical bioreactor Maryam Rezaa,* , Manuel Alvarez Cuencab a b
University of Waterloo, Canada Ryerson University, Canada
A R T I C L E I N F O
A B S T R A C T
Article history: Received 27 July 2015 Received in revised form 18 September 2015 Accepted 25 October 2015 Available online 30 October 2015
Nutrients pollution has become a global environmental threat. A large fraction of nutrient pollution is caused by point sources like the discharges of untreated domestic and industrial wastewater. As a result, there is a great demand to develop reliable, compact and efficient nutrient removal technologies. In the wastewater industry, most of the bioreactors have a large foot print with a plane configuration. Furthermore, the existing nutrient removal processes are based on well-known microbial species of genus Nitrosomonas and Nitrobacter involved in nitrification/denitrification and Candidatus Accumulibater Phosphatis responsible for biological phosphorous removal (BPR). The present work is unprecedented in two aspects: (1) The biological nitrogen and phosphorus removal from wastewater occurred in a multistage vertical, tubular bioreactor and (2) Two abundant microbial species were responsible for the simultaneous nitrification–denitrification–BPR in this vertical bioreactor. The abundant microbial populations included an unidentified bacteria of Saprospiraceae family and bacteria affiliated with the genus Zoogloea. The composition of the synthetic wastewater used in this study was: total phosphorous (TP) 32.6 0.7 mg/L, total nitrogen (TN) 272 7.5 in which 45 1.8 mg/L was ammonia-nitrogen (NH3-N). The bioreactor was continuously operated for over 350 days at constant flowrate, temperature and pH of 240 (L/day), 22–24 ( C) and 7–7.5. The results showed that simultaneous nitrification–denitrification–BPR was the dominant process by an, as yet not fully classified, microbial species. The effluent TP and TN concentrations were 2.7 0.4 and 4.3 1.2 mg/L respectively. Concentrations of NH3-N, NO2 and NO3-N in the effluent were 0.7 0.1, 0.8 0.5 and 0.3 0.1. The simultaneous nitrification–denitrification–BPR was highly effective delivering above 90% TN and TP removal efficiency. The successful results presented in this paper have led to the construction of a 20,000 L/day demonstration plant in the City of Pickering, Ontario. ã 2015 Elsevier Ltd. All rights reserved.
Keywords: Simultaneous nitrification–denitrification Biological phosphorus removal (BPR) Biological nutrient removal (BNR) Vertical tubular bioreactor
1. Introduction Two design parameters are of paramount importance in the design of wastewater treatment plants. These are: (1) high standards of effluent quality and (2) low construction space or minimal foot print. For example, upgrade/expansion of plants in urban areas is limited by site constraints caused by demographic pressures and severe regulations. Thus, successful technologies in wastewater treatment must meet the above design criteria. Increasing chemical complexity in industrial and domestic wastewater, flow-rates variability, mixing, and emerging contaminants impose further difficulties in wastewater treatment and control. Among current environmental issues, nutrient pollution
* Corresponding author. E-mail address: [email protected] (M. Reza). http://dx.doi.org/10.1016/j.jece.2015.10.035 2213-3437/ ã 2015 Elsevier Ltd. All rights reserved.
has become an overwhelming problem for many countries around the world. Excessive use of fertilizers in the agricultural sector and daily human activities are the main sources of nutrient pollution. This in turn causes extensive economic losses because of the degradation of fisheries, tourist facilities and coastal regions. This global problem can be mitigated with more stringent environmental regulations, and hence advanced bioreactors and processes. 1.1. Research objectives In some wastewater treatment plants, nitrogen removal processes such as nitrification and denitrification are accompanied by biological phosphorus removal (BPR). However, due to hypersensitivity of the phosphorus removing organisms many plants adopt chemical treatments for phosphorus removal. Chemical treatment is both more expensive and less
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
environmentally benign than biological treatment. Below is a list of the most common suspended growth processes which carry out both biological nitrogen and phosphorus removal [1]:
3 Stage pho-redox (A2/O) 5 Stage bardenpho Modified University of Cape Town (modified UCT) Oxidation ditch Sequencing batch reactor (SBR)
The most appropriate reactor configuration for a biological nutrient removal (BNR) process depends on factors such as the target effluent quality, influent quality, process control, biological process complexity, and available foot print. Many of the existing BNR processes take place in plane, horizontal basins with large foot print. As pointed out before, construction space limitations is one of the problems facing municipal wastewater treatment facilities. The prime objective of the research presented herein was to design a pilot scale vertical bioreactor capable of achieving an effective biological nutrient removal process. The advantages of this reactor configuration were: Tubular geometry and vertical configuration provided construction flexibility and lower land requirements than comparable BNR bioreactors of the same flowrate. Since, tubular vertical reactors have lower foot print they can easily adapt to expansion and retrofits. The tubular geometry of this bioreactor provides superior mixing and mass transfer because they avoid the stagnancies which normally develop in rectangular bioreactors. The vertical configuration of this bioreactor provides smooth flow of water from one stage to the other without additional need for pumps. The configuration of the bioreactor permitted the creation of the environmental conditions leading to the formation of ecosystems that favor the growth of new microbial species as suggested by Littleton et al. [2]. The novel multistage, vertical bioreactor allows the nitrification, denitrification and BPR processes to
131
integrate and operate simultaneously under the same environmental condition. As shown in our experimental results simultaneous nitrification–denitrification–BPR has lower organic carbon requirements than conventional BNR systems. The multistage vertical bioreactor was designed, constructed and tested in the Water Technologies Laboratory at Ryerson University in Toronto. 1.2. Background Nitrification is the oxidation of NH4+ to NO3 through a twostep process (Knapp) [3]. Step 1: ammonium oxidizing bacteria (AOBs) are responsible for the first process: NH4+ + 3/2O2 ! NO2 + 2H+ + H2O
(1.1)
Step 2: nitrite oxidizing bacteria (NOBs) complete the second step in nitrification: NO2 + 1/2O2 ! NO3
(1.2)
Denitrification is very similar to the biological oxidation of organic matter except that it occurs in the absence of dissolved oxygen. Heterotrophic bacteria utilize NO3 for oxidation of organic compounds as shown below: NO3 + Organic Matter ! N2 + CO2 + OH + H2O
(1.3)
Biological phosphorus removal (BPR) process requires an anaerobic phase followed by an anoxic or aerobic phase. Under anaerobic condition, phosphorus accumulating organisms (PAOs) uptake volatile fatty acids (VFA) then produce and store intracellularly polyhydroxyalkanoates (PHAs). Under the aerobic or anoxic conditions, PAOs or denitrifying PAOs (DPAOs) break down their internal PHAs which release energy. This energy is immediately used by the PAOs to uptake dissolved phosphorus and produce and store polyphosphates. Haandel van and Lubbe van der [4] reported that PAOs can store polyphosphate up to 38% of their
Fig. 1. Block diagram of the vertical bioreactor aligned with a clarifier and anaerobic lateral unit (ALU).
132
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
cellular mass whereas other microorganisms can store less than 5% of phosphorus in their cell mass.
All the design parameters and operating conditions are summarized in Table 1.
2. Materials and methods
2.3. Composition of the feed
2.1. Experimental set-up
The bioreactor’s influent was a synthetic wastewater composed of the following compounds: NH4Cl (11.25 g), KH2 PO4 (2.77 g), Na2HPO4 (3.125 g), Na2HPO4H2O (2.807 g), Urea CH4N2O (5 g), calcium carbonate (5 g), CaCl2H2O (1.5 g), MgSO47H2O (1.5), Na2SO4 (1.5 g), FeCl3 (1.5 g/L), ZnCl2 (0.12 g/L) and EDTA (7 mg/L). A mixture of acetic acid (10 ml), propionic acid (10 ml), butyric acid (10 ml) and sugar 20(g) was added to the ALU. All of these chemicals were supplied by VWR international. The feed composition is presented in Table 2. The bioreactor was inoculated on November 10, 2012 with activated sludge from the North Toronto Wastewater Treatment Plant. During the start-up period, the biomass was maintained and internally recycled within the bioreactor for approximately three (3) months. During the start-up period, samples were collected regularly from the bioreactor to detect any reduction in nitrogen and phosphorus concentrations.
A vertical bioreactor with three consecutive stages, Anoxic 1, Anoxic 2 and Aerobic, was used to cultivate a mixed culture of heterotrophic/autotrophic nitrifiers, denitrifiers and PAOs. This bioreactor was connected to an anaerobic lateral unit (ALU) of 60 L capacity. The ALU provided a strict anaerobic condition to cultivate and promote the growth of PAOs. The pilot scale bioreactor of cylindrical cross-section was made of high density polyethylene (HDPE) and a working volume of 65 L. The rate of synthetic feed to the bioreactor was maintained at 10 L/h. (240 L/day). The three stages of the bioreactor (Fig. 1) were separated from each other using rigid plastic boards bolted on top of each other. Wastewater flowed by gravity through external pipes from Anoxic 1 to Anoxic 2 and finally to the Aerobic stage. A recycle stream from the Aerobic stage to Anoxic 1 provided mixing and created a uniform composition of microorganisms and nutrients throughout the bioreactor. A 90-L cylindrical clarifier was used to separate the biomass from the treated effluent. The treated effluent was tested for (1) nutrients concentrations, (2) evaluation of the simultaneous biological nutrient removal process and (3) reactor’s performance. The settled biomass flowed by gravity from the clarifier to the ALU and then was pumped continuously (by a metering pump) from the ALU to the first Anoxic stage. The experimental set-up was equipped with pH, temperature, dissolved oxygen (DO) sensors and flow meters. All the sensors and transmitters were connected to a data acquisition system which recorded and monitored the bioreactor continuously. 2.2. Process development methodology To develop biological nitrogen removal, the following design parameters were adopted from studies by other researchers like Jetten et al. [5] and Winkler et al. [6]: Solids residence time (SRT): 50 days Dissolved oxygen (DO) concentration in the Aerobic stage: 2.5–3.5 mg/L DO concentration in Anoxic 1 and Anoxic 2: <0.1 mg/L pH: 7–7.5 To avoid NO2 accumulation, DO in the Aerobic stage was varied to obtain optimum NO2 concentration
2.4. Analytical and microbial procedures To analyze concentrations of NH3-N, NO2, NO3-N and TP a Hach test kits such as a spectrophotometer (DR2700) and an anaerobic digester (DRB200) were used. Samples were taken continuously for over 350 days from all three stages of the bioreactor, the feed and the effluent leaving the clarifier. Samples were filtered using 0.45 mm filter paper prior to the analytical measurements. To analyze the microbial population, 2 and 4 ml biomass samples were centrifuged followed by DNA extraction with the Powersoil DNA Isolation Kit. The protocol involved a harsh cell lysis and rigorous purification. Then, bacteria-specific primers were used for the polymerization change reaction (PCR) analysis. The biomass samples were sequenced using a MiSeq1 system and recovered 8,768,510 pairs in total. UPARSE was used to determine the structure of the microbial community. This software compressed the data into a unique taxa that were 97% similar to one another. The most abundant sequence within each 97% cluster was selected as the representative sequence for each operational taxonomic unit (OTU). 3. Experimental results 3.1. Microbial analysis
Critical parameters in developing the BPR process included: A mixture of acetic acid, propionic acid and butyric acid has been found as the best carbon source to favor the growth of PAOs over glycogen accumulating organisms (GAOs) (the PAOs competitors) [7,8]. COD to TP ratio in the influent was maintained at 30:1 (COD:TP) [9]. Optimum DO concentration of 2.5–3.5 mg/L was used. DO has shown significant effect on the PAO–GAO competition. High DO concentrations (i.e. 4.5–5.0 mg/L) reduce the BPR efficiency [10]. Low pH (below 6.5) can be detrimental for the BPR process since a low pH enviornment can promote the growth of GAOs. Calcium carbonate and sodium hyrdroxide were added to maintain the pH of the proccess within 7–7.5 range. The temperature was maintained at approximately 25 C. Some researchers like Baetens [11] reported that temperatures higher than 30 C can negatively affect BPR.
The taxonomic microbial diversity analysis showed a significant presence of heterotrophic bacteria. High-resolution insight into the composition of the microbial population indicated that the denitrifiers that dominated the reactor belonged to the classes of Table 1 Operating conditions in the three stages of the bioreactor and the ALU. Parameters (ranges)
Three stages of the bioreactor Anoxic 1 & 2 Aerobic
Anaerobic lateral unit (ALU)
DO (mg/L) COD (mg/L)
<0.1 500–300
0 1600
TP (mg/L) pH Temperature HRT SRT
2.5–3.5 250– 100 32 (feed conc.)
50–60 (TP release by PAOs) 7–7.5 22–25 C 4h 50 days
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136 Table 2 Synthetic wastewater composition. Nutrient
Concentration
NH3-N NO3-N a NO2 TP Organic carbon addition to the ALU b COD concentrations
45 1.8 (mg/L) 32 0.9 (mg/L) 33 1.3(mg/L) 32.6 0.7 (mg/L)
only indication of Saprospiraceae in literature was their presence in a biological nutrient removal plant involving a SBR samples by Ginige et al. [13]. Further microbial studies are needed to identify the physiology and function of Saprospiraceae. Fig. 2 shows the taxonomic hierarchy of the microbial communities in the bioreactor. 3.2. Analytical results
1600 (mg/L)
a NO2 concentration in synthetic feed was due to the chemical reaction of urea in water. No nitrite containing compounds were used. b COD concentration formed by adding a mixture of propionic, butyric and acetic acids as well as sugar.
Nutrient removal results were recorded and analyzed to understand the microbial processes in the reactor. Results showed that nitrogen removal began 140 days after start-up. NH3-N, NO2 and NO3-N concentrations in the effluent were unsteady for approximately 240 days of the reactor operation. The nitrogen removal reached steady state conditions as concentrations were stable and consistent from day 240 until the end of the experiments (day 350). Over time, NH3-N trends in the effluent decreased to below 1 mg/L while influent concentrations were kept constant at 45 1.8 (mg/L). Table 3 summarizes the average nutrient concentrations and associated material balances from various sampling locations. The mixed liquor suspended solids (MLSS) concentration at steady state was approximately 5 g/L. To maintain the MLSS concentration of 5 g/L and SRT of 50 days, 0.14 g of biomass was
Zoogloea. An abundant presence of Rhodocyclales was also detected. In a study by Hesselsoe et al. [12], Rhodocyclales were identified in nitrifying–denitrifying, phosphorus-removing activated sludge process with unique substrate-utilization profiles. The results showed that a previously unknown bacterium from the Saprospiraceae family dominated all the samples. Saprospiraceae found in this study were only 96% identical to the nearest Genbank sequence. That is, this dominant bacterium was likely a distinct species with uncertain metabolism, not detected previously. The
Rhodospirillaceae, Caulobacteraceae, mitochondria, Hyphomicrobiaceae Rhodobacteraceae, Sphingomonadaceae, Methylocystaceae, Acetobacteraceae
Proteobacteria
α-proteobacteria β-proteobacteria
Nitrosomonadaceae, Comamonadaceae, Rhodocyclaceae (Zoogloea), Neisseriaceae Chromatiaceae, Moraxellaceae, Xanthomonadaceae, Sinobacteraceae, Methylococcaceae, Coxiellaceae, Pseudomonadaceae, Aeromonadaceae, Ectothiorhodospiraceae
Acidobacteria
γ-proteobacteria δ- proteobacteria
Actinobacteria
Desulfovibrionaceae, Syntrophobacteraceae, Bacteriovoracaceae, Cystobacterineae, Geobacteraceae, Myxococcaceae, Bdellovibrionaceae
Bacteroidetes
Holophagae, Solibacteres, Chloracidobacteria, Acidobacteria-6, Thermoleophilia Actinomycetaceae, Micrococcaceae, Microbacteriaceae
Nitrospirae
Chitinophagaceae, Weeksellaceae, Cytophagaceae, Saprospiraceae, Rikenellaceae, Flavobacteriaceae
Nitrospiraceae
Pirellulaceae, Gemmataceae
Synergistetes
Planctomycetes
133
Synergistaceae, Dethiosulfovibrionaceae, TTA_B6
Fig. 2. Taxonomic hierarchy of the microbial population in the bioreactor.
134
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
Table 3 Summary of the results and material balance for phosphorus, nitrogen and carbon. Parameters (Mean SD)
NH3-N
NO2
NO3-N
TP
COD
TN
DO
Total carbon
Feed Anoxic 1 Anoxic 2 Aerobic Effluent ALU Biomass
45 1.8 2.6 0.56 2.87 1.2 1.0 0.4 0.7 0.5 – –
33 1.3 87 3.9 61 6.3 14.4 0.5 0.8 0.5 – –
32 0.9 0.74 0.14 0.7 0.1 1.1 0.1 0.3 0.1 – –
32.6 0.7 14.8 4 12 3.3 7.4 0.7 2.7 0.4 55 5 142
526 217 153 97 80 1600 –
272 7.5 96.8 4.8 79.4 4.4 23.2 8.1 4.3 1.2 – 240 5
1.25 <0.1 <0.1 2.20.2 <0.1 – –
– – – – – – 230 2
removed from the reactor every day. Chemical analyses showed that phosphorus concentrations in the biomass were 175 and 113 mg/L in the two Anoxic stages and in the Aerobic stage respectively. Total organic nitrogen was 165 mg/L and total carbon was 247 mg/L in the biomass. The microbial analysis showed a high amount of nitrifiers and denitrifiers in the biomass and there was no trace or indication of Anammox bacteria in the samples. Therefore, simultaneous nitrification–denitrification seemed to be the only pathway for conversion of NH3-N–N2. Fig. 3 shows NH3-N, NO3-N and NO2 concentrations in the influent and effluent. The graph shows a high variability period in nitrogen concentrations during the early stages of the experiment (first 240 days). From day 240 until the end of the experimental work concentrations in the effluent showed a stable and consistent trend. Biological phosphorus removal started much later than nitrogen removal. It took almost 230 days (ca. 7 months) from the reactor start-up date to detect the BPR process. The phosphorus removal efficiency increased from 21%, observed in July 2013 to over 90% in December 2013. High NO2 concentration in the bioreactor, during the first seven (7) months was found as the main inhibitor for BPR. This inhibitory effect diminished once NO2 concentration in the reactor was less than 100 mg/L due to utilization by NOBs. This result is consistent with the findings by Saito et al. [14] who reported that NO2 could inhibit the phosphorus uptake by the PAOs. A number of researchers have confirmed the negative effect of NO2, free nitrous acid (FNA) and other protonated species of NO2 on aerobic phosphorus uptake of PAOs [15,16,17]. In this study, when nitrification–denitrification reached steady state, the NO2 fluctuation in the reactor decreased. As a result, PAOs/DPAOs started to uptake phosphorus from
wastewater. TP concentration in the influent was kept constant at 32.6 0.7 (mg/L) while it gradually reduced to 2.7 0.4 (mg/L) in the effluent leaving the clarifier as illustrated in Fig. 4. The action of ALU (see Fig. 1) with 4 h of HRT and addition of volatile organic carbon (approximately 1600 mg/L) were essential in developing the BPR process. TP concentration in ALU was 55 5 mg/L. High TP concentration in the ALU was due to the phosphorus release by PAOs and DPAOs. The experimental data confirmed that TP removal occurred in all three stages of the vertical bioreactor whereas NH3-N removal mainly took place in the first Anoxic stage. TP was equally removed in each stage as shown in Fig. 5. This indicated that PAOs and DPAOs performed equally under the two Anoxic stages and the Aerobic stage. This observation differed from studies by other researchers who claimed that PAOs had higher growth yield and phosphorus uptake rate than DPAOs [18,19]. Substantial reduction of NH3-N in the absence of DO and presence of NO2 and NO3-N in the first and second Anoxic stage showed that simultaneous nitrification–denitrification was successfully achieved. The physical characteristics of the biomass changed during the scope of this study. The settleability of the sludge in the clarifier increased when simultaneous nitrification–denitrification–BPR process reached steady state. According to Grady et al. [20], microbial populations of Zoogloea,Thauera, Alphaproteobacteria, Betaproteobacteria and Gammaproteobacteria readily flocculate in activated sludge processes. As shown in Fig. 2, these floc-forming organisms were abundant in the present bioreactor and were capable of flocculation which played a very important role in biomass settling in the secondary clarifier.
Fig. 3. Average nitrogen concentrations in the influent and effluent throughout the experimental period.
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
135
Fig. 4. TP concentration in the influent, effluent and ALU.
Fig. 5. NH3-N and TP removed in the three stages of the bioreactor during the last 3 months of the experiment.
view of COD concentration curve in the bioreactor with increase in the rate of removal of TN and TP in 390 min (6.5 h of HRT). The average TP and TN removal rates were 0.13 (mg/L min) and 1.6 (mg/L min).
300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0
mg/L of TP Removed
600
mg/L of TN Removed 500
COD Concentration (mg/L)
400
300 200 100
0
20
40
60
80
100
120
140
160
180
Hydraulic Residence Time (min) Fig. 6. TP and TN removal rate relative to COD concentration over time.
200
0
COD Concentration (mg.L)
TN and TP Removed (mg/L)
There were no changes in the nitrogen removal efficiency since the BPR process was observed in the reactor. This indicated that PAOs, DPAOs, AOBs and NOBs coexist under the same environmental conditions even though they have distinctive nutrient removal activities and growth rates. Fig. 6 provides an expanded
136
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
The average COD reduction rate in the bioreactor was found to be 10.7 (mg/L min). As shown in Fig. 6, the COD concentration was mainly used for TN removal via nitrification and denitrification which highly depend on readily biodegradable COD. Whereas, the energy required for phosphorus uptake under anoxic or aerobic condition came from the utilization of intracellular PHAs [11]. Finally, the analytical results indicated that the biological nutrient removal required a COD to TN ratio of 1.9:1. This is much lower than C:N ratio found by Yang et al. [21], Hanki et al. [22] and Pochana and Keller [15]. Low C:N ratio indicated that simultaneous nitrification–denitrification–BPR in this vertical bioreactor required much lower organic carbon than conventional bioreactors. 4. Conclusions The results presented here show the successful performance of a multistage vertical bioreactor in which a new microbial process for nutrient removal has been developed. This process satisfies the design criteria of foot print minimization and high simultaneous removal of both nitrogen and phosphorous based nutrients. The biological process was successful due to the bioreactor’s innovative structure and performance as described by Alvarez-Cuenca and Reza [23]. Simultaneous nitrification–denitrification occurred much earlier than BPR. NH3-N removal started approximately 120 days after the bioreactor start-up and reached a stable condition after day 240. It took approximately 230 days to detect BPR and 270 days to attain steady state BPR. BPR was established much later than it was anticipated due to the inhibition caused by the presence of NO2 and other nitrite compounds. The simultaneous nitrification– denitrification–BPR process was highly effective delivering removal efficiency above 90% for both TP and TN. The configuration and sequence of the different stages of the bioreactor provided a receptive ecosystem for the growth of previously unrecognized species of Saprospiraceae capable of simultaneous nitrification– denitrification and BPR. This method exhibited lower requirements of COD than other comparable nutrient removal methods. Acknowledgements We would like to thank the Chemical Engineering Department at Ryerson University for the financial support in the construction and commissioning of the experimental facility. Also, we would like to extend our thanks to Dr. Josh D. Neufeld and his research team in the Biology Department of the University of Waterloo for microbial analyses of our samples. The successful results presented in this paper have led to the construction of a 20,000 L/day demonstration plant in the Duffin Creek Water Pollution Control Plant (WPCP), located in the City of Pickering, Ontario. The demonstration plant project has been funded and supported by the Regional Municipality of York, Ontario, Canada. References [1] The Cadmus Group Inc., Nutrient Control Design Manual, United States Environmental Protection Agency, (2010) 2–8.
[2] H.X. Littleton, G.T. Daigger, P.F. Strom, R.A. Cowan, Simultaneous biological nutrient removal: evaluation of autotrophic denitrification, heterotrophic nitrification, and biological phosphorus removal in full-scale systems, Water Environ. Res. 75 (2) (2003) 138–150. [3] C.W. Knapp, D.W. Graham, Nitrite-oxidizing bacteria guild ecology associated with nitrification failure in a continuous-flow reactor, FEMS Microbiol. Ecol. 62 (2) (2007) 195–201. [4] A. Haandel van, J. Lubbe van der, Treatment design and optimization of activated sludge systems, Handbook Biological Wastewater, Quist Publishing, Netherlands, 2007, pp. 188. [5] M.S.M. Jetten, S.J. Horn, M.C.M. van Loosdrecht, Towards a more sustainable municipal wastewater treatment system, Water Sci. Technol. 35 (9) (1997) 171–180. [6] K.H.M. Winkler, R. Kleerebezem, J.G. Kuenen, J. Yang, C.M.M. van Loosdrecht, Segregation of biomass in cyclic anaerobic/aerobic granular sludge allows the enrichment of anaerobic ammonium oxidizing bacteria at low temperatures, Environ. Sci. Technol. 45 (2011) 7330–7337. [7] C.W. Randall, D. Stansel, J.L. Barnard, Design and Retrofit of Wastewater Treatment Plants for Biological Nutrient Removal, Technomic Publishing Co. Inc., USA, 1992, pp. 17–18. [8] K.R. Reddy, Nutrient removal by aquatic systems used for wastewater treatment, Nutrient Control Manual, Water Environment Federation, USA, 1998, pp. 305–332 Chapter 10. [9] M. Reza, M. Alvarez-Cuenca, Analysis of a vertical bioreactor for denitrification and biological phosphorus removal from wastewater, Chem. Eng. Process Tech. (2013) 1006. [10] A. Oehmen, C.P. Lemos, G. Carvalho, Z. Yuan, J. Keller, L.L. Blackall, A.M.M. Reis, Advances in enhanced biological phosphorus removal: from micro to macro scale, Water Res. J. 41 (11) (2007) 2271–2300. [11] Baetens, D., Enhanced biological phosphorus removal: modeling and experimental design. Ph.D Thesis, Department of Technical Chemistry and Environmental Engineering and the Department of Applied Mathematics, Biometrics and Process Control (BIOMATH), Ghent University, Belgium, 2001, pp. 32. [12] M. Hesselsoe, S. Fȕreder, M. Schloter, L. Bodrossy, N. Iversen, P. Roslev, P.H. Nielsen, M. Wagner, A. Loy, Isotope array analysis of Rhodocyclales uncovers functional redundancy and versatility in an activated sludge, ISME J. 3 (2009) 1349–1364. [13] M.P. Ginige, P. Hugenholtz, H. Daims, M. Wagner, J. Keller, L.L. Blackall, Use of stable-isotope probing, full-cycle rRNA analysis, and fluorescence in situ hybridization-microautoradiography to study a methanol-fed denitrifying microbial community, Appl. Environ. Microbiol. 70 (1) (2004) 588–596. [14] T. Saito, D. Brdjanovic, M.C.M. van Loosdrecht, Effect of nitrite on phosphate uptake by phosphate accumulating organisms, Water Res. 38 (2004) 3760–3768. [15] K. Pochana, J. Keller, Study of factors affecting simultaneous nitrification and denitrification (SND), Water Sci. Technol. 39 (6) (1999) 61–68. [16] W. Zeng, B. Li, Y. Yang, X. Wang, L. Li, Y. Peng, Impact of nitrite on aerobic phosphorus uptake by poly-phosphate accumulating organisms in enhanced biological phosphorus removal sludges, Bioprocess Biosyst. Eng. 37 (2) (2014) 277–287. [17] Y. Zhou, M. Pijuan, Z.G. Yuan, Free nitrous acid inhibition on anoxic phosphorus uptake and denitrification by poly-phosphate accumulating organisms, Biotechnol. Bioeng. 98 (2007) 903–912. [18] Z. Hu, M.C. Wentzel, G.A. Ekama, Anoxic growth of phosphate-accumulating organisms (PAOs) in biological nutrient removal activated sludge systems, Water Res. 36 (2002) 4927–4937. [19] A. Johwan, D. Tomotaka, S. Tsuneda, A. Hirata, Characterization of denitrifying phosphate-accumulation organisms cultivated under different electron acceptor conditions using polymerase chain reaction-denaturing gradient gel electrophoresis assay, Water Res. 36 (2002) 403–412. [20] L.C.P. Grady, G.T. Daigger, N.G. Love, C.D.M. Filipe, Biological Wastewater Treatment, third edition, IWA, Publishing, UK, 2011, pp. 41–42. [21] S. Yang, J. Tay, Y. Liu, Effect of substrate nitrogen/chemical oxygen demand ratio on the formation of aerobic Granules, J. Environ. Eng. 131 (1) (2005) 86– 92. [22] K. Hanki, C. Wantawin, S. Ohgaki, Nitrification at low levels of dissolved oxygen with and without organic loading in a suspended growth reactor, Water Res. 24 (3) (1990) 297–302. [23] Alvarez-Cuenca M., and Reza M., CUBEN, a Novel bioreactor for the removal of nutrients: a performance evaluation, NATO Science for Peace and Security Series-C: Environmental Security, Springer in collaboration with NATO Public Diplomacy Division, 2013, pp. 63–92.
Contents lists available at ScienceDirect
Journal of Environmental Chemical Engineering journal homepage: www.elsevier.com/locate/jece
Simultaneous biological removal of nitrogen and phosphorus in a vertical bioreactor Maryam Rezaa,* , Manuel Alvarez Cuencab a b
University of Waterloo, Canada Ryerson University, Canada
A R T I C L E I N F O
A B S T R A C T
Article history: Received 27 July 2015 Received in revised form 18 September 2015 Accepted 25 October 2015 Available online 30 October 2015
Nutrients pollution has become a global environmental threat. A large fraction of nutrient pollution is caused by point sources like the discharges of untreated domestic and industrial wastewater. As a result, there is a great demand to develop reliable, compact and efficient nutrient removal technologies. In the wastewater industry, most of the bioreactors have a large foot print with a plane configuration. Furthermore, the existing nutrient removal processes are based on well-known microbial species of genus Nitrosomonas and Nitrobacter involved in nitrification/denitrification and Candidatus Accumulibater Phosphatis responsible for biological phosphorous removal (BPR). The present work is unprecedented in two aspects: (1) The biological nitrogen and phosphorus removal from wastewater occurred in a multistage vertical, tubular bioreactor and (2) Two abundant microbial species were responsible for the simultaneous nitrification–denitrification–BPR in this vertical bioreactor. The abundant microbial populations included an unidentified bacteria of Saprospiraceae family and bacteria affiliated with the genus Zoogloea. The composition of the synthetic wastewater used in this study was: total phosphorous (TP) 32.6 0.7 mg/L, total nitrogen (TN) 272 7.5 in which 45 1.8 mg/L was ammonia-nitrogen (NH3-N). The bioreactor was continuously operated for over 350 days at constant flowrate, temperature and pH of 240 (L/day), 22–24 ( C) and 7–7.5. The results showed that simultaneous nitrification–denitrification–BPR was the dominant process by an, as yet not fully classified, microbial species. The effluent TP and TN concentrations were 2.7 0.4 and 4.3 1.2 mg/L respectively. Concentrations of NH3-N, NO2 and NO3-N in the effluent were 0.7 0.1, 0.8 0.5 and 0.3 0.1. The simultaneous nitrification–denitrification–BPR was highly effective delivering above 90% TN and TP removal efficiency. The successful results presented in this paper have led to the construction of a 20,000 L/day demonstration plant in the City of Pickering, Ontario. ã 2015 Elsevier Ltd. All rights reserved.
Keywords: Simultaneous nitrification–denitrification Biological phosphorus removal (BPR) Biological nutrient removal (BNR) Vertical tubular bioreactor
1. Introduction Two design parameters are of paramount importance in the design of wastewater treatment plants. These are: (1) high standards of effluent quality and (2) low construction space or minimal foot print. For example, upgrade/expansion of plants in urban areas is limited by site constraints caused by demographic pressures and severe regulations. Thus, successful technologies in wastewater treatment must meet the above design criteria. Increasing chemical complexity in industrial and domestic wastewater, flow-rates variability, mixing, and emerging contaminants impose further difficulties in wastewater treatment and control. Among current environmental issues, nutrient pollution
* Corresponding author. E-mail address: [email protected] (M. Reza). http://dx.doi.org/10.1016/j.jece.2015.10.035 2213-3437/ ã 2015 Elsevier Ltd. All rights reserved.
has become an overwhelming problem for many countries around the world. Excessive use of fertilizers in the agricultural sector and daily human activities are the main sources of nutrient pollution. This in turn causes extensive economic losses because of the degradation of fisheries, tourist facilities and coastal regions. This global problem can be mitigated with more stringent environmental regulations, and hence advanced bioreactors and processes. 1.1. Research objectives In some wastewater treatment plants, nitrogen removal processes such as nitrification and denitrification are accompanied by biological phosphorus removal (BPR). However, due to hypersensitivity of the phosphorus removing organisms many plants adopt chemical treatments for phosphorus removal. Chemical treatment is both more expensive and less
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
environmentally benign than biological treatment. Below is a list of the most common suspended growth processes which carry out both biological nitrogen and phosphorus removal [1]:
3 Stage pho-redox (A2/O) 5 Stage bardenpho Modified University of Cape Town (modified UCT) Oxidation ditch Sequencing batch reactor (SBR)
The most appropriate reactor configuration for a biological nutrient removal (BNR) process depends on factors such as the target effluent quality, influent quality, process control, biological process complexity, and available foot print. Many of the existing BNR processes take place in plane, horizontal basins with large foot print. As pointed out before, construction space limitations is one of the problems facing municipal wastewater treatment facilities. The prime objective of the research presented herein was to design a pilot scale vertical bioreactor capable of achieving an effective biological nutrient removal process. The advantages of this reactor configuration were: Tubular geometry and vertical configuration provided construction flexibility and lower land requirements than comparable BNR bioreactors of the same flowrate. Since, tubular vertical reactors have lower foot print they can easily adapt to expansion and retrofits. The tubular geometry of this bioreactor provides superior mixing and mass transfer because they avoid the stagnancies which normally develop in rectangular bioreactors. The vertical configuration of this bioreactor provides smooth flow of water from one stage to the other without additional need for pumps. The configuration of the bioreactor permitted the creation of the environmental conditions leading to the formation of ecosystems that favor the growth of new microbial species as suggested by Littleton et al. [2]. The novel multistage, vertical bioreactor allows the nitrification, denitrification and BPR processes to
131
integrate and operate simultaneously under the same environmental condition. As shown in our experimental results simultaneous nitrification–denitrification–BPR has lower organic carbon requirements than conventional BNR systems. The multistage vertical bioreactor was designed, constructed and tested in the Water Technologies Laboratory at Ryerson University in Toronto. 1.2. Background Nitrification is the oxidation of NH4+ to NO3 through a twostep process (Knapp) [3]. Step 1: ammonium oxidizing bacteria (AOBs) are responsible for the first process: NH4+ + 3/2O2 ! NO2 + 2H+ + H2O
(1.1)
Step 2: nitrite oxidizing bacteria (NOBs) complete the second step in nitrification: NO2 + 1/2O2 ! NO3
(1.2)
Denitrification is very similar to the biological oxidation of organic matter except that it occurs in the absence of dissolved oxygen. Heterotrophic bacteria utilize NO3 for oxidation of organic compounds as shown below: NO3 + Organic Matter ! N2 + CO2 + OH + H2O
(1.3)
Biological phosphorus removal (BPR) process requires an anaerobic phase followed by an anoxic or aerobic phase. Under anaerobic condition, phosphorus accumulating organisms (PAOs) uptake volatile fatty acids (VFA) then produce and store intracellularly polyhydroxyalkanoates (PHAs). Under the aerobic or anoxic conditions, PAOs or denitrifying PAOs (DPAOs) break down their internal PHAs which release energy. This energy is immediately used by the PAOs to uptake dissolved phosphorus and produce and store polyphosphates. Haandel van and Lubbe van der [4] reported that PAOs can store polyphosphate up to 38% of their
Fig. 1. Block diagram of the vertical bioreactor aligned with a clarifier and anaerobic lateral unit (ALU).
132
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
cellular mass whereas other microorganisms can store less than 5% of phosphorus in their cell mass.
All the design parameters and operating conditions are summarized in Table 1.
2. Materials and methods
2.3. Composition of the feed
2.1. Experimental set-up
The bioreactor’s influent was a synthetic wastewater composed of the following compounds: NH4Cl (11.25 g), KH2 PO4 (2.77 g), Na2HPO4 (3.125 g), Na2HPO4H2O (2.807 g), Urea CH4N2O (5 g), calcium carbonate (5 g), CaCl2H2O (1.5 g), MgSO47H2O (1.5), Na2SO4 (1.5 g), FeCl3 (1.5 g/L), ZnCl2 (0.12 g/L) and EDTA (7 mg/L). A mixture of acetic acid (10 ml), propionic acid (10 ml), butyric acid (10 ml) and sugar 20(g) was added to the ALU. All of these chemicals were supplied by VWR international. The feed composition is presented in Table 2. The bioreactor was inoculated on November 10, 2012 with activated sludge from the North Toronto Wastewater Treatment Plant. During the start-up period, the biomass was maintained and internally recycled within the bioreactor for approximately three (3) months. During the start-up period, samples were collected regularly from the bioreactor to detect any reduction in nitrogen and phosphorus concentrations.
A vertical bioreactor with three consecutive stages, Anoxic 1, Anoxic 2 and Aerobic, was used to cultivate a mixed culture of heterotrophic/autotrophic nitrifiers, denitrifiers and PAOs. This bioreactor was connected to an anaerobic lateral unit (ALU) of 60 L capacity. The ALU provided a strict anaerobic condition to cultivate and promote the growth of PAOs. The pilot scale bioreactor of cylindrical cross-section was made of high density polyethylene (HDPE) and a working volume of 65 L. The rate of synthetic feed to the bioreactor was maintained at 10 L/h. (240 L/day). The three stages of the bioreactor (Fig. 1) were separated from each other using rigid plastic boards bolted on top of each other. Wastewater flowed by gravity through external pipes from Anoxic 1 to Anoxic 2 and finally to the Aerobic stage. A recycle stream from the Aerobic stage to Anoxic 1 provided mixing and created a uniform composition of microorganisms and nutrients throughout the bioreactor. A 90-L cylindrical clarifier was used to separate the biomass from the treated effluent. The treated effluent was tested for (1) nutrients concentrations, (2) evaluation of the simultaneous biological nutrient removal process and (3) reactor’s performance. The settled biomass flowed by gravity from the clarifier to the ALU and then was pumped continuously (by a metering pump) from the ALU to the first Anoxic stage. The experimental set-up was equipped with pH, temperature, dissolved oxygen (DO) sensors and flow meters. All the sensors and transmitters were connected to a data acquisition system which recorded and monitored the bioreactor continuously. 2.2. Process development methodology To develop biological nitrogen removal, the following design parameters were adopted from studies by other researchers like Jetten et al. [5] and Winkler et al. [6]: Solids residence time (SRT): 50 days Dissolved oxygen (DO) concentration in the Aerobic stage: 2.5–3.5 mg/L DO concentration in Anoxic 1 and Anoxic 2: <0.1 mg/L pH: 7–7.5 To avoid NO2 accumulation, DO in the Aerobic stage was varied to obtain optimum NO2 concentration
2.4. Analytical and microbial procedures To analyze concentrations of NH3-N, NO2, NO3-N and TP a Hach test kits such as a spectrophotometer (DR2700) and an anaerobic digester (DRB200) were used. Samples were taken continuously for over 350 days from all three stages of the bioreactor, the feed and the effluent leaving the clarifier. Samples were filtered using 0.45 mm filter paper prior to the analytical measurements. To analyze the microbial population, 2 and 4 ml biomass samples were centrifuged followed by DNA extraction with the Powersoil DNA Isolation Kit. The protocol involved a harsh cell lysis and rigorous purification. Then, bacteria-specific primers were used for the polymerization change reaction (PCR) analysis. The biomass samples were sequenced using a MiSeq1 system and recovered 8,768,510 pairs in total. UPARSE was used to determine the structure of the microbial community. This software compressed the data into a unique taxa that were 97% similar to one another. The most abundant sequence within each 97% cluster was selected as the representative sequence for each operational taxonomic unit (OTU). 3. Experimental results 3.1. Microbial analysis
Critical parameters in developing the BPR process included: A mixture of acetic acid, propionic acid and butyric acid has been found as the best carbon source to favor the growth of PAOs over glycogen accumulating organisms (GAOs) (the PAOs competitors) [7,8]. COD to TP ratio in the influent was maintained at 30:1 (COD:TP) [9]. Optimum DO concentration of 2.5–3.5 mg/L was used. DO has shown significant effect on the PAO–GAO competition. High DO concentrations (i.e. 4.5–5.0 mg/L) reduce the BPR efficiency [10]. Low pH (below 6.5) can be detrimental for the BPR process since a low pH enviornment can promote the growth of GAOs. Calcium carbonate and sodium hyrdroxide were added to maintain the pH of the proccess within 7–7.5 range. The temperature was maintained at approximately 25 C. Some researchers like Baetens [11] reported that temperatures higher than 30 C can negatively affect BPR.
The taxonomic microbial diversity analysis showed a significant presence of heterotrophic bacteria. High-resolution insight into the composition of the microbial population indicated that the denitrifiers that dominated the reactor belonged to the classes of Table 1 Operating conditions in the three stages of the bioreactor and the ALU. Parameters (ranges)
Three stages of the bioreactor Anoxic 1 & 2 Aerobic
Anaerobic lateral unit (ALU)
DO (mg/L) COD (mg/L)
<0.1 500–300
0 1600
TP (mg/L) pH Temperature HRT SRT
2.5–3.5 250– 100 32 (feed conc.)
50–60 (TP release by PAOs) 7–7.5 22–25 C 4h 50 days
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136 Table 2 Synthetic wastewater composition. Nutrient
Concentration
NH3-N NO3-N a NO2 TP Organic carbon addition to the ALU b COD concentrations
45 1.8 (mg/L) 32 0.9 (mg/L) 33 1.3(mg/L) 32.6 0.7 (mg/L)
only indication of Saprospiraceae in literature was their presence in a biological nutrient removal plant involving a SBR samples by Ginige et al. [13]. Further microbial studies are needed to identify the physiology and function of Saprospiraceae. Fig. 2 shows the taxonomic hierarchy of the microbial communities in the bioreactor. 3.2. Analytical results
1600 (mg/L)
a NO2 concentration in synthetic feed was due to the chemical reaction of urea in water. No nitrite containing compounds were used. b COD concentration formed by adding a mixture of propionic, butyric and acetic acids as well as sugar.
Nutrient removal results were recorded and analyzed to understand the microbial processes in the reactor. Results showed that nitrogen removal began 140 days after start-up. NH3-N, NO2 and NO3-N concentrations in the effluent were unsteady for approximately 240 days of the reactor operation. The nitrogen removal reached steady state conditions as concentrations were stable and consistent from day 240 until the end of the experiments (day 350). Over time, NH3-N trends in the effluent decreased to below 1 mg/L while influent concentrations were kept constant at 45 1.8 (mg/L). Table 3 summarizes the average nutrient concentrations and associated material balances from various sampling locations. The mixed liquor suspended solids (MLSS) concentration at steady state was approximately 5 g/L. To maintain the MLSS concentration of 5 g/L and SRT of 50 days, 0.14 g of biomass was
Zoogloea. An abundant presence of Rhodocyclales was also detected. In a study by Hesselsoe et al. [12], Rhodocyclales were identified in nitrifying–denitrifying, phosphorus-removing activated sludge process with unique substrate-utilization profiles. The results showed that a previously unknown bacterium from the Saprospiraceae family dominated all the samples. Saprospiraceae found in this study were only 96% identical to the nearest Genbank sequence. That is, this dominant bacterium was likely a distinct species with uncertain metabolism, not detected previously. The
Rhodospirillaceae, Caulobacteraceae, mitochondria, Hyphomicrobiaceae Rhodobacteraceae, Sphingomonadaceae, Methylocystaceae, Acetobacteraceae
Proteobacteria
α-proteobacteria β-proteobacteria
Nitrosomonadaceae, Comamonadaceae, Rhodocyclaceae (Zoogloea), Neisseriaceae Chromatiaceae, Moraxellaceae, Xanthomonadaceae, Sinobacteraceae, Methylococcaceae, Coxiellaceae, Pseudomonadaceae, Aeromonadaceae, Ectothiorhodospiraceae
Acidobacteria
γ-proteobacteria δ- proteobacteria
Actinobacteria
Desulfovibrionaceae, Syntrophobacteraceae, Bacteriovoracaceae, Cystobacterineae, Geobacteraceae, Myxococcaceae, Bdellovibrionaceae
Bacteroidetes
Holophagae, Solibacteres, Chloracidobacteria, Acidobacteria-6, Thermoleophilia Actinomycetaceae, Micrococcaceae, Microbacteriaceae
Nitrospirae
Chitinophagaceae, Weeksellaceae, Cytophagaceae, Saprospiraceae, Rikenellaceae, Flavobacteriaceae
Nitrospiraceae
Pirellulaceae, Gemmataceae
Synergistetes
Planctomycetes
133
Synergistaceae, Dethiosulfovibrionaceae, TTA_B6
Fig. 2. Taxonomic hierarchy of the microbial population in the bioreactor.
134
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
Table 3 Summary of the results and material balance for phosphorus, nitrogen and carbon. Parameters (Mean SD)
NH3-N
NO2
NO3-N
TP
COD
TN
DO
Total carbon
Feed Anoxic 1 Anoxic 2 Aerobic Effluent ALU Biomass
45 1.8 2.6 0.56 2.87 1.2 1.0 0.4 0.7 0.5 – –
33 1.3 87 3.9 61 6.3 14.4 0.5 0.8 0.5 – –
32 0.9 0.74 0.14 0.7 0.1 1.1 0.1 0.3 0.1 – –
32.6 0.7 14.8 4 12 3.3 7.4 0.7 2.7 0.4 55 5 142
526 217 153 97 80 1600 –
272 7.5 96.8 4.8 79.4 4.4 23.2 8.1 4.3 1.2 – 240 5
1.25 <0.1 <0.1 2.20.2 <0.1 – –
– – – – – – 230 2
removed from the reactor every day. Chemical analyses showed that phosphorus concentrations in the biomass were 175 and 113 mg/L in the two Anoxic stages and in the Aerobic stage respectively. Total organic nitrogen was 165 mg/L and total carbon was 247 mg/L in the biomass. The microbial analysis showed a high amount of nitrifiers and denitrifiers in the biomass and there was no trace or indication of Anammox bacteria in the samples. Therefore, simultaneous nitrification–denitrification seemed to be the only pathway for conversion of NH3-N–N2. Fig. 3 shows NH3-N, NO3-N and NO2 concentrations in the influent and effluent. The graph shows a high variability period in nitrogen concentrations during the early stages of the experiment (first 240 days). From day 240 until the end of the experimental work concentrations in the effluent showed a stable and consistent trend. Biological phosphorus removal started much later than nitrogen removal. It took almost 230 days (ca. 7 months) from the reactor start-up date to detect the BPR process. The phosphorus removal efficiency increased from 21%, observed in July 2013 to over 90% in December 2013. High NO2 concentration in the bioreactor, during the first seven (7) months was found as the main inhibitor for BPR. This inhibitory effect diminished once NO2 concentration in the reactor was less than 100 mg/L due to utilization by NOBs. This result is consistent with the findings by Saito et al. [14] who reported that NO2 could inhibit the phosphorus uptake by the PAOs. A number of researchers have confirmed the negative effect of NO2, free nitrous acid (FNA) and other protonated species of NO2 on aerobic phosphorus uptake of PAOs [15,16,17]. In this study, when nitrification–denitrification reached steady state, the NO2 fluctuation in the reactor decreased. As a result, PAOs/DPAOs started to uptake phosphorus from
wastewater. TP concentration in the influent was kept constant at 32.6 0.7 (mg/L) while it gradually reduced to 2.7 0.4 (mg/L) in the effluent leaving the clarifier as illustrated in Fig. 4. The action of ALU (see Fig. 1) with 4 h of HRT and addition of volatile organic carbon (approximately 1600 mg/L) were essential in developing the BPR process. TP concentration in ALU was 55 5 mg/L. High TP concentration in the ALU was due to the phosphorus release by PAOs and DPAOs. The experimental data confirmed that TP removal occurred in all three stages of the vertical bioreactor whereas NH3-N removal mainly took place in the first Anoxic stage. TP was equally removed in each stage as shown in Fig. 5. This indicated that PAOs and DPAOs performed equally under the two Anoxic stages and the Aerobic stage. This observation differed from studies by other researchers who claimed that PAOs had higher growth yield and phosphorus uptake rate than DPAOs [18,19]. Substantial reduction of NH3-N in the absence of DO and presence of NO2 and NO3-N in the first and second Anoxic stage showed that simultaneous nitrification–denitrification was successfully achieved. The physical characteristics of the biomass changed during the scope of this study. The settleability of the sludge in the clarifier increased when simultaneous nitrification–denitrification–BPR process reached steady state. According to Grady et al. [20], microbial populations of Zoogloea,Thauera, Alphaproteobacteria, Betaproteobacteria and Gammaproteobacteria readily flocculate in activated sludge processes. As shown in Fig. 2, these floc-forming organisms were abundant in the present bioreactor and were capable of flocculation which played a very important role in biomass settling in the secondary clarifier.
Fig. 3. Average nitrogen concentrations in the influent and effluent throughout the experimental period.
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
135
Fig. 4. TP concentration in the influent, effluent and ALU.
Fig. 5. NH3-N and TP removed in the three stages of the bioreactor during the last 3 months of the experiment.
view of COD concentration curve in the bioreactor with increase in the rate of removal of TN and TP in 390 min (6.5 h of HRT). The average TP and TN removal rates were 0.13 (mg/L min) and 1.6 (mg/L min).
300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0
mg/L of TP Removed
600
mg/L of TN Removed 500
COD Concentration (mg/L)
400
300 200 100
0
20
40
60
80
100
120
140
160
180
Hydraulic Residence Time (min) Fig. 6. TP and TN removal rate relative to COD concentration over time.
200
0
COD Concentration (mg.L)
TN and TP Removed (mg/L)
There were no changes in the nitrogen removal efficiency since the BPR process was observed in the reactor. This indicated that PAOs, DPAOs, AOBs and NOBs coexist under the same environmental conditions even though they have distinctive nutrient removal activities and growth rates. Fig. 6 provides an expanded
136
M. Reza, M.A. Cuenca / Journal of Environmental Chemical Engineering 4 (2016) 130–136
The average COD reduction rate in the bioreactor was found to be 10.7 (mg/L min). As shown in Fig. 6, the COD concentration was mainly used for TN removal via nitrification and denitrification which highly depend on readily biodegradable COD. Whereas, the energy required for phosphorus uptake under anoxic or aerobic condition came from the utilization of intracellular PHAs [11]. Finally, the analytical results indicated that the biological nutrient removal required a COD to TN ratio of 1.9:1. This is much lower than C:N ratio found by Yang et al. [21], Hanki et al. [22] and Pochana and Keller [15]. Low C:N ratio indicated that simultaneous nitrification–denitrification–BPR in this vertical bioreactor required much lower organic carbon than conventional bioreactors. 4. Conclusions The results presented here show the successful performance of a multistage vertical bioreactor in which a new microbial process for nutrient removal has been developed. This process satisfies the design criteria of foot print minimization and high simultaneous removal of both nitrogen and phosphorous based nutrients. The biological process was successful due to the bioreactor’s innovative structure and performance as described by Alvarez-Cuenca and Reza [23]. Simultaneous nitrification–denitrification occurred much earlier than BPR. NH3-N removal started approximately 120 days after the bioreactor start-up and reached a stable condition after day 240. It took approximately 230 days to detect BPR and 270 days to attain steady state BPR. BPR was established much later than it was anticipated due to the inhibition caused by the presence of NO2 and other nitrite compounds. The simultaneous nitrification– denitrification–BPR process was highly effective delivering removal efficiency above 90% for both TP and TN. The configuration and sequence of the different stages of the bioreactor provided a receptive ecosystem for the growth of previously unrecognized species of Saprospiraceae capable of simultaneous nitrification– denitrification and BPR. This method exhibited lower requirements of COD than other comparable nutrient removal methods. Acknowledgements We would like to thank the Chemical Engineering Department at Ryerson University for the financial support in the construction and commissioning of the experimental facility. Also, we would like to extend our thanks to Dr. Josh D. Neufeld and his research team in the Biology Department of the University of Waterloo for microbial analyses of our samples. The successful results presented in this paper have led to the construction of a 20,000 L/day demonstration plant in the Duffin Creek Water Pollution Control Plant (WPCP), located in the City of Pickering, Ontario. The demonstration plant project has been funded and supported by the Regional Municipality of York, Ontario, Canada. References [1] The Cadmus Group Inc., Nutrient Control Design Manual, United States Environmental Protection Agency, (2010) 2–8.
[2] H.X. Littleton, G.T. Daigger, P.F. Strom, R.A. Cowan, Simultaneous biological nutrient removal: evaluation of autotrophic denitrification, heterotrophic nitrification, and biological phosphorus removal in full-scale systems, Water Environ. Res. 75 (2) (2003) 138–150. [3] C.W. Knapp, D.W. Graham, Nitrite-oxidizing bacteria guild ecology associated with nitrification failure in a continuous-flow reactor, FEMS Microbiol. Ecol. 62 (2) (2007) 195–201. [4] A. Haandel van, J. Lubbe van der, Treatment design and optimization of activated sludge systems, Handbook Biological Wastewater, Quist Publishing, Netherlands, 2007, pp. 188. [5] M.S.M. Jetten, S.J. Horn, M.C.M. van Loosdrecht, Towards a more sustainable municipal wastewater treatment system, Water Sci. Technol. 35 (9) (1997) 171–180. [6] K.H.M. Winkler, R. Kleerebezem, J.G. Kuenen, J. Yang, C.M.M. van Loosdrecht, Segregation of biomass in cyclic anaerobic/aerobic granular sludge allows the enrichment of anaerobic ammonium oxidizing bacteria at low temperatures, Environ. Sci. Technol. 45 (2011) 7330–7337. [7] C.W. Randall, D. Stansel, J.L. Barnard, Design and Retrofit of Wastewater Treatment Plants for Biological Nutrient Removal, Technomic Publishing Co. Inc., USA, 1992, pp. 17–18. [8] K.R. Reddy, Nutrient removal by aquatic systems used for wastewater treatment, Nutrient Control Manual, Water Environment Federation, USA, 1998, pp. 305–332 Chapter 10. [9] M. Reza, M. Alvarez-Cuenca, Analysis of a vertical bioreactor for denitrification and biological phosphorus removal from wastewater, Chem. Eng. Process Tech. (2013) 1006. [10] A. Oehmen, C.P. Lemos, G. Carvalho, Z. Yuan, J. Keller, L.L. Blackall, A.M.M. Reis, Advances in enhanced biological phosphorus removal: from micro to macro scale, Water Res. J. 41 (11) (2007) 2271–2300. [11] Baetens, D., Enhanced biological phosphorus removal: modeling and experimental design. Ph.D Thesis, Department of Technical Chemistry and Environmental Engineering and the Department of Applied Mathematics, Biometrics and Process Control (BIOMATH), Ghent University, Belgium, 2001, pp. 32. [12] M. Hesselsoe, S. Fȕreder, M. Schloter, L. Bodrossy, N. Iversen, P. Roslev, P.H. Nielsen, M. Wagner, A. Loy, Isotope array analysis of Rhodocyclales uncovers functional redundancy and versatility in an activated sludge, ISME J. 3 (2009) 1349–1364. [13] M.P. Ginige, P. Hugenholtz, H. Daims, M. Wagner, J. Keller, L.L. Blackall, Use of stable-isotope probing, full-cycle rRNA analysis, and fluorescence in situ hybridization-microautoradiography to study a methanol-fed denitrifying microbial community, Appl. Environ. Microbiol. 70 (1) (2004) 588–596. [14] T. Saito, D. Brdjanovic, M.C.M. van Loosdrecht, Effect of nitrite on phosphate uptake by phosphate accumulating organisms, Water Res. 38 (2004) 3760–3768. [15] K. Pochana, J. Keller, Study of factors affecting simultaneous nitrification and denitrification (SND), Water Sci. Technol. 39 (6) (1999) 61–68. [16] W. Zeng, B. Li, Y. Yang, X. Wang, L. Li, Y. Peng, Impact of nitrite on aerobic phosphorus uptake by poly-phosphate accumulating organisms in enhanced biological phosphorus removal sludges, Bioprocess Biosyst. Eng. 37 (2) (2014) 277–287. [17] Y. Zhou, M. Pijuan, Z.G. Yuan, Free nitrous acid inhibition on anoxic phosphorus uptake and denitrification by poly-phosphate accumulating organisms, Biotechnol. Bioeng. 98 (2007) 903–912. [18] Z. Hu, M.C. Wentzel, G.A. Ekama, Anoxic growth of phosphate-accumulating organisms (PAOs) in biological nutrient removal activated sludge systems, Water Res. 36 (2002) 4927–4937. [19] A. Johwan, D. Tomotaka, S. Tsuneda, A. Hirata, Characterization of denitrifying phosphate-accumulation organisms cultivated under different electron acceptor conditions using polymerase chain reaction-denaturing gradient gel electrophoresis assay, Water Res. 36 (2002) 403–412. [20] L.C.P. Grady, G.T. Daigger, N.G. Love, C.D.M. Filipe, Biological Wastewater Treatment, third edition, IWA, Publishing, UK, 2011, pp. 41–42. [21] S. Yang, J. Tay, Y. Liu, Effect of substrate nitrogen/chemical oxygen demand ratio on the formation of aerobic Granules, J. Environ. Eng. 131 (1) (2005) 86– 92. [22] K. Hanki, C. Wantawin, S. Ohgaki, Nitrification at low levels of dissolved oxygen with and without organic loading in a suspended growth reactor, Water Res. 24 (3) (1990) 297–302. [23] Alvarez-Cuenca M., and Reza M., CUBEN, a Novel bioreactor for the removal of nutrients: a performance evaluation, NATO Science for Peace and Security Series-C: Environmental Security, Springer in collaboration with NATO Public Diplomacy Division, 2013, pp. 63–92.