Environ Sci Pollut Res DOI 10.1007/s11356-017-8840-9
RESEARCH ARTICLE
An anoxic-aerobic system for simultaneous biodegradation of phenol and ammonia in a sequencing batch reactor Qifeng Liu 1,2 & Vijay P. Singh 3 & Zhimin Fu 1,2,3
&
Jing Wang 1,2 & La Hu 1,2
Received: 20 December 2016 / Accepted: 15 March 2017 # Springer-Verlag Berlin Heidelberg 2017
Abstract A laboratory-scale sequencing batch reactor (SBR) was investigated to treat artificial pretreated coal gasification wastewater that was mainly contained of ammonia and phenol. The efficiency of SBR fed with increasing phenol concentrations (from 150 to 300 mg l−1) and the relationship among phenol, nitrogen removal, and the microbial community structure were evaluated. When the phenol feeding concentration was increased to about 300 mg l−1, the removal efficiency was above 99.0%, demonstrating the robustness of phenol removal capacity. The study showed that most phenol was degraded in anoxic stage. The average removal efficiencies of ammonia and total nitrogen were 98.4 and 81.9%, respectively, with average NH 4 + -N concentration of 107.5 mg l−1 and COD/N 7.5. Low temperature caused sludge loss that led to the decreased performance. Increasing the temperature could not recover the performance effectively. The data from bacterial analysis revealed that Delftia, Hydrogenophaga, and unclassified Xanthomonadaceae played a significant role in phenol degradation before the temperature increase, while uncultured Syntrophococcus sp. and
Responsible editor: Bingcai Pan * Zhimin Fu [email protected]
1
School of Ecology and Environment, Inner Mongolia University, Hohhot 010021, People’s Republic of China
2
Inner Mongolia Coal Chemical Industry Wastewater Treatment and Reuse Engineering Technology Research Center, Inner Mongolia University, Hohhot 010021, People’s Republic of China
3
Departments of Biological and Agricultural Engineering and Zachry Department of Civil Engineering, Texas A&M University, College Station, TX 77840, USA
unclassified Rhodocyclaceae were responsible for phenol degradation after the temperature increase. These results imply that the SBR holds potential for the simultaneous removal of phenolic compounds and nitrogen through aerobic ammonia oxidation and anoxic denitrification with phenol as the coorganic carbon source. Keywords SBR . Ammonia . Phenol . Denitrification . PCR-DGGE
Introduction Several industrial processes, such as petroleum refinement, coke oven plant, and petrochemical manufacturing paints, discharge complicated wastewaters containing both ammonia and phenolic compounds (Milia et al. 2012; Kim et al. 2013). It is well known that the ammonia discharge deteriorates water ecosystems through stimulation growth of fastgrowing biomass and bloom-forming algae (Moreno-Marín et al. 2016). While phenolic compounds are toxic to humans, animals, and plants (Al-Khalid and El-Naas 2012), phenol has been listed in the water priority control pollutant blacklist by many countries, including China and the USA. Therefore, both ammonia and phenol must be removed before wastewater discharge in order to reduce their adverse environmental impacts and protect waterbody ecosystems. The cost-effective ammonia removal method is the biological nitrogen removal process based on nitrification and denitrification reactions. Typically, the nitrification process is performed in two consecutive steps: ammonia oxidation to nitrite by ammonia-oxidizing bacteria (AOB), and nitrite oxidation to nitrate by nitrite-oxidizing bacteria (NOB) utilizing dissolved oxygen (DO) as electron acceptor. In the denitrification process, nitrate is reduced to nitrogen gas using organic
Environ Sci Pollut Res
compounds as electronic donor. Biological treatment is also considered to be efficient for mitigating the effects of phenol. Phenol can be degraded to harmless compounds by microorganisms under aerobic or anaerobic conditions (Al-Khalid and El-Naas 2012; Rosenkranz et al. 2013). Conventionally, aerobic processes are preferred due to microorganism’s faster growth and its mineralization (Al-Khalid and El-Naas 2012). It has been widely documented that phenolic compounds are inhibitors of nitrifying microorganisms, especially AOB (Lauchnor et al. 2011; Lauchnor and Semprini 2013). However, Amor et al. (2005) reported that in the aerobic activated sludge reactor, there was no inhibition of nitrification by phenol when the phenol concentration was increased from 35 up to 2800 mg l−1. The absence of inhibition can be attributed to the negligible phenol concentration in the unit as a result of fast mixing in the stirred tank reactor as well as its fast removal. During the denitrification process, nitrite and nitrate usually need an anaerobic condition and an organic carbon source for dinitrogen gas production. The denitrification reaction can be restricted by the deficient amount of organic carbon source, which leads to a large excess of nitrogen in effluent (Kartal et al. 2010). It has been suggested to add additional organic carbon source, such as methanol or lactate, for optimal denitrification, but due to increase in operation costs, it is undesirable from the engineering point of view (Torà et al. 2011). Researchers have demonstrated that phenol can be biodegraded under anoxic denitrifying conditions, with phenol mineralized to small-molecule acids and nitrate reduced to nitrogen. Bajaj et al. (2010) investigated the effect of phenol as a toxic co-substrate with glucose towards an anoxic mixed microbial flora in a reactor that removes nitrate. Batch assay results of phenol degradation under anoxic conditions have shown high phenol degradation rates at low concentrations (188 mg l −1 ). Anoxic bioreactors, dominanted with Rhodococcus pyridinivorans, have shown the highest removal efficiency of various organic pollutants (phenolics, aromatic hydrocarbons, and cyanide) commonly found in coke oven wastewater (Sharma and Philip 2015). Recently, denitritation using phenol as the sole organic carbon to reduce nitrite was successfully achieved with granular biomass in an upflow sludge blanket reactor (Ramos et al. 2016). The most abundant genera in the anoxic granules were denitrifying bacteria Ignavibacterium and Denitratisoma. These studies present a possible technological option to achieve simultaneous removal of phenol and ammonia nitrogen using the phenol as the carbon source for the denitrification process. The concept of this article is that nitrate produced from ammonia oxidation is removed by using phenol as electron donor. This can be realized by using a sequencing batch reactor (SBR) with an anoxic stage and an aerobic stage. In anoxic stage, nitrite and nitrate can be removed by heterotrophic denitrification with phenol as co-substrate organic substrates,
while in aerobic stage, the ammonia can be oxidized to nitrite and nitrate. The objective of this study therefore is to demonstrate this concept and underlying mechanisms by bacterial analysis.
Experimental method Experimental setup and operation A Plexiglas cylindrical SBR reactor with an effective height 1.0 m and a diameter of 0.07 m was utilized in this study. A guide tube with a height of 0.6 m and a diameter of 0.04 m was fixed within it. The operation cycle time was 6 h, which included anoxic stage for 120 min (including 10 min of substrate filling), aerobic stage for 330–350 min, and settling stage for 30–10 min (including 5 min of effluent withdrawal). Influent was supplied by a peristaltic pump (BT100-2J, LongerPump/ Baoding Longer constant-flow pump Co. Ltd., China), and air was supplied through a diffuser at the reactor bottom with an airflow rate of 18 l/h, which was measured by a glass gas flow meter (LZB-3WB, Kede/Changzhou Kede thermal instrument Co. Ltd., China). To enhance the mixture of the anoxic stage, aeration was provided two times at 40 and 80 min, lasting each time for 5 min. The volume exchange ratio was 27–28%. The reactor was inoculated with activated sludge from an anoxic-oxic process treating municipal wastewater for simultaneous carbon and nitrogen removal (Xinxinban, Hohhot). Data were measured after 1-month cultivation. The operation period consisted of three different phases (Table 1). In phase I, the settling stage was 30 min, with influent phenol concentration of 150 mg l−1. In phase II, the settling stage was decreased from 30 to 10 min, but the influent composition was the same as phase I. In phase III, the settling stage was 10 min, and influent phenol concentration was increased from 150 to 300 mg l−1. The temperature variation is listed in Table 1 and marked in Fig.2. In phase I, first, the temperature was kept at 25 °C; then, it was decreased to 18 °C. During phase II and the first part of phase III, the temperature was not adjusted but changed as the atmosphere temperature. A heating rod was added into the reactor from 185 days (III) to keep the temperature at 25 °C. Synthetic wastewater and analysis methods The influent synthetic wastewater (SWW) was prepared in tap water containing the following (g l−1): 0.35–0.7 glucose, 0.15–0.30 C6H5OH, 0.418 NH4Cl, 0.044 KH2PO4, and 0.8 NaHCO3. To enhance the setting ability of the sludge, CaCl2 (0.141 g l−1) was added to phase II and phase III. NH4+-N, NO3−-N, NO2−-N, phenol, and COD were analyzed according to the Water and Wastewater Monitoring Analysis Method (CEPA 2002). Total nitrogen (TN) was the
Environ Sci Pollut Res Table 1 Phase
Operation parameters of different phases days
COD (mg l−1)
Phenol (mg l−1)
Ammonia (mg l−1)
I
1–70
786.95 ± 94.91
156.23 ± 26.17
102.6 ± 30.07
II
71–118
793.46 ± 125.45
140.76 ± 23.57
108.38 ± 18.38
III
118–252
968.86 ± 98.87
268.60 ± 75.50
110.51 ± 24.69
sum of NH4+-N, NO3−-N, and NO2−-N. Liquid samples collected at regular time intervals during the cycle period were analyzed to reveal the NH4+-N and phenol transform process. DNA extraction and polymerase chain reaction-denatured gradient gel electrophoresis Active sludge samples were taken out in phase III. The total DNA was extracted with a soil genomic DNA quick extraction kit SK8233 (Sangon Biotech, China). The extracted RNA was digested with RNase A (Sangon Biotech, China) for RNA elimination and extracted with chloroform and alcohol according to the recommendations of the kit protocol. The treatment efficiency was verified by PCR, which showed amplification with the universal primers F338-GC (5′-CGCC CGCCGCGCCCCGCGCCCGGCCCGCCGC CCCCGCCC CCCTACGGGAGGCAGCAG-3′) and R518 (5′-ATTA CCGCGGCT GCTGG-3′). The PCR reaction system contained 5 μl of 10× buffer (containing 2.0 mM MgCl2), 0.5 μl of template DNA, 0.25 μl of Taq enzyme (5 U μl−1), 2 μl of primers (10 μM), 1 μl of dNTP mixture at 10 mM, and 41.25 μl of sterilized distilled water. PCR conditions were as follows: predegeneration at 94 °C for 240 s, denatured at 94 °C for 30 s, primer annealing at 56 °C for 60 s, and chain extension at 72 °C for 30 s with an additional extension time of 7 min on the final cycle for a total of 30 cycles. The total DNAwas stored at −20 °C in a freezer for further application. For DGGE analysis, 400 ng of PCR V3 product was loaded onto 8% (w/v) polyacrylamide gel with a 30–60% denaturant gradient. Electrophoresis was performed at 60 °C at 60 V for 16 h. The middle portion of each selected DGGE band was excised, washed, transferred to 1.5-ml centrifuge tube, and recycled PCR products with kit SK8183 (Sangon Biotech, China). PCR products were reamplified and sequenced.
Results and discussion COD and phenol removal efficiency During the operation (Fig. 1a), the influent COD concentrations varied between 650 and 1100 mg l−1, with an average COD concentration of 842.5 mg l−1. Data showed that the
Temperature (°C)
Sludge
Settling time (min)
25 (1–19 days) >18 (21–70 days) Ambient (71–118 days)
Recycle to SBR
30
No recycle
15 (71–76 days) 10 (76–85 days)
Ambient (118–185 days) 25 (185–255 days)
No recycle
5
average effluent COD concentration was 102.3 mg l−1, which indicated that the COD removal performance had been less affected by the temperature change, phenol concentration increase, and the settling time decrease. The average COD load was 0.91 g l−1 day−1, and the average removal efficiency of COD was 87.6%. These data indicated that the system could provide a consistent high efficiency of COD removal. During the experiment operation period, phenol concentration and removal efficiency variations are shown in Fig. 1b. Although the corresponding average influent phenol load was increased from 164.8 to 284.6 mg l−1 day−1, the removal efficiency of phenol was higher than 99.0%. Most of the effluent concentration of phenol was lower than 0.5 mg l−1, except for the days of temperature decrease (phase I, 22 days; phase III, 150, 162 days) and influent phenol concentration increase (phase III, 118 days). A sharp buildup of phenol was observed in the reactor. However, this buildup quickly dissipated as the sludge adapted rapidly to the change, indicating that the phenol removal capacity of the system was robust. However, the effluent phenol concentration was kept as 1.0 mg l−1 during phase III (190–218 days), which was due to the loss of mixed liquid volatile suspended solid (MLVSS) (Fig. 2b). Amor et al. (2005) reported that high phenol removal efficiencies (above 99.9%) were maintained at all the applied organic loading rates (30–2700 mg COD l−1 day−1), which was increased stepwise by increasing the phenol concentration from 35 up to 2800 mg l−1. Adav et al. (2007) developed aerobic granules that degrade 1000 mg l−1 phenol at a constant rate of 1176 mg phenol g−1 VSS day−1. Furthermore, adding acetate enhanced the phenol degradation rate for the granular system (Ho et al. 2009). Therefore, the low phenol concentration and loading rate as well as the glucose presence assured the SBR systems to degrade phenol effectively. NH4−-N and TN removal efficiency The average ammonia concentration in the influent was maintained constant at 107.5 mg l−1, and the applied ammonia loading rate was 116.1 mg NH4+-N l−1 day−1. The evolution of ammonia in the influent and effluent and ammonia removal efficiency of the system are shown in Fig. 2a. During the stable operation days of phases I (50–71 days), II (92– 118 days), and III (125–155 days), the average NH4+-N
Environ Sci Pollut Res Fig. 1 The COD and phenol concentration variations during the operation. a COD in the influent, effluent, and COD removal efficiency. b Phenol in the influent, effluent, and phenol removal efficiency
removal efficiencies were 99.2, 95.5, and 98.4%, respectively, and effluent NH4+-N concentrations were 0.88, 4.65, and 1.89 mg l−1, respectively. Data showed that the decrease in the settling time from 30 to 10 min led to a slight decline in the nitrification performance, which was speculated due to the sludge washout (Fig. 2b). AOB are chemoautotrophic and slow-growing; accordingly, numerous AOB populations, which cannot become big and dense enough to settle fast, were washed out from the system (Zhang et al. 2011). The increased sludge (72–90 days) concentration was attributed to the heterotrophic organism growth. The 98.4% NH4+-N removal efficiency of phase III indicated that no inhibition of nitrification by phenol, which was consistent with the results reported by Amor et al. (2005). The absence of inhibition can be due to the negligible phenol concentration in the unit as a result of its fast removal in the anoxic stage (Fig. 4).
On day 19 (phase I), the heating rod was suddenly broken and the atmosphere temperature was fluctuated between 3 and 15 °C, and the NH4+-N concentration in the effluent was dramatically increased to about 40–50 mg l−1 from day 20. The heating temperature was set at 18 °C on day 21. The NH4+-N concentration in the effluent was not decreased until day 50. The recovery of the nitrification process required very long times (1 month) owing to the slow growth rates of nitrifying bacteria (Kim et al. 2013). However, the reactor experienced a sharp decrease on day 140 days (4–15 °C, phase III) of the atmosphere temperature, and the NH4+-N removal efficiency was not affected. Meanwhile, during days 160–170, the reactor experienced a low temperature (4–20 °C) and the ammonia removal efficiency decreased to 64%, but it was recovered to 90% on day 171, which was speculated to the higher MLVSS (4765–4180 mg l−1) during the period.
Environ Sci Pollut Res Fig. 2 The NH4+-N, MLVSS, SVI, and temperature variations during the operation. a NH4+-N in the influent, effluent, and NH4+-N removal efficiency. b MLVSS, SVI, and temperature
The MLVSS started to decrease from day 160 as shown in Fig. 2b. Winkler et al. (2012) reported a twofold slower settling velocity for the same granules when the temperature of water decreased from 5 to 40 °C, and the main factor influencing the settling velocity was the change in the viscosity of water. Therefore, the temperature was controlled at 25 °C after day 185 to avoid deterioration of the process performance. However, contrary to expectations, the MLVSS decreased from 3545 to 1190 mg l−1 and the sludge volume index (SVI) increased from 39.5 to 100.8 ml g−1. Data showed that the temperature increase did not stop the loss of sludge (Fig. 2b). The ammonia
concentration of the effluent was about 50 mg l−1 due to the low sludge concentration. The NO2−-N, NO3−-N, and TN concentration variations and TN removal efficiency during the operation are shown in Fig. 3. During the stable operation days of phases I (50– 71 days), II (92–118 days), and III (125–155 days), the average TN removal efficiencies were 82.1, 73.9, and 81.9%, respectively, under the COD/N ratios of 7.0, 7.6, and 7.5. The average TN efficiencies were higher than that reported by Guo et al. (2013). Wan et al. (2009) suggested that alternating anoxic feast-aerobic famine encouraged heterotrophic growth deeply inside the aggregates and promoted the denitrification
Environ Sci Pollut Res Fig. 3 The NO2−-N, NO3−-N, and TN concentration variations and TN removal efficiency during the operation
efficiency. Hence, it was reasonable to conclude that the anoxic-aerobic SBR had higher denitrification ability than aerobic simultaneous nitrification and denitrification (SND) SBR (Zhang et al. 2015). Under anoxic conditions, denitrifiers utilized biological organic substrate as the electron donor to reduce nitrite and nitrate. The COD/N ratio of influent was one of the most critical parameters for wastewater nitrogen removal process. Chiu et al. (2007) reported that the SBR system resulted in the optimal removal of both organic matter and NH4+-N with no accumulation of intermediate by-products, when the COD/N ratio was controlled at 11.1. The low C/N ratio, causing the electron donor deficiency for denitrification process, resulted in a low TN removal efficiency (Roy et al. 2010). However, in the present study, the higher COD/N value did not coincidence with the greater TN removal efficiency. In phase II (Fig. 4b), it was found that the removed ammonia exceeded nitrification products, implying that the aerobic denitrification or SND process contributed to the nitrogen removal. Taking into account that the vast majority of organic matter was consumed in the anoxic stage, it was reasonable to speculate that endogenous carbon source (polyhydroxybutyrate (PHB), glycogen) was utilized for the denitrification process (Winkler et al. 2011). The endogenous denitrification rate was generally lower than that of a pre-anoxic process, because the endogenous carbon degradation rate limited the nitrogen performance rate for the SND process (Zhu et al. 2013). Therefore, the low TN removal efficiency of phase II was attributed to the low
endogenous denitrification rate, and prolonged aeration reaction time might enhance the nitrogen removal performance. During the stable operation days of phases I (50–71 days), II (92–118 days), and III (125–155 days), the average effluent NO3−-N concentrations were 17.1, 15.6, ad 14.4 mg l−1, and the average effluent NO2−-N concentrations were 1.5, 6.8, and 7.0 mg l−1, respectively. Although NO3−-N was the main nitrification product, the effluent NO2−-N concentration was increased obviously in phases II and III. In phase II day 85, the settling time was shortened to 10 min, which caused the concentration of activated sludge to decrease sharply. The nitrification bacteria, AOB and NOB, were washed out from the reactor. Paredes et al. (2007) reported that at the temperature over 15.0 °C, AOB had higher growth rates than NOB. Therefore, NO2−-N was accumulated in phase II. In phase III, the NO3−-N production dropped drastically and NO2−-N accumulated, which suggested that the NO2−-N accumulation was caused by a drastic decline in the NOB activity. In day phase III (125–155 days), the atmosphere temperature decreased gradually and varied between 5 and 26 °C. Both AOB and NOB are sensitive to temperature and suffered from the temperature decrease. The specific ammonia oxidation rate decreased by 1.5 times with the temperature decreasing from 25 to 15 °C (Guo et al. 2010). During the critical temperature 12–14 °C, the rate of nitrate formation controls the overall nitrification process, meaning that nitrite would accumulate (Robinson et al. 2004). Furthermore, Nitrospira and Nitrobacter are the commonly identified NOB in the
Environ Sci Pollut Res
Fig. 4 Typical profiles of cycle study and particle size distribution. a Phase I. b Phase II. c Phase III. d Particle size distribution
nitrogen removal wastewater treatment plants (Huang et al. 2010). Nitrospira would outcome Nitrobacter in the wastewater treatment system, due to that Nitrospira had a lower nitrite half-saturation constant and the NO2−-N concentration in system was low (Blackburne et al. 2007). Gilbert et al. (2014) reported that Nitrospira was affected most by temperatures below 13 °C, showing no recovery. The activity of AOB increased more than that of NOB as temperature increased (Zhang et al. 2014). Therefore, the NO2−-N accumulation in phase III was speculated due to the low temperature. In days 20–50 (phase I), the TN removal efficiency decreased and fluctuated at 45–50; at the same time, the nitrite concentration was higher than nitrate, which was also attributed to the low temperature. Results of cycle experiment Figure 4 shows the typical phenol and nitrogen species profiles in reactors during a steady-state cycle in phases I, II, and III. Throughout the anoxic stage, cycle profile measurements revealed that almost 100% phenol and COD (not given) were
utilized in phases I, II, and III. Simultaneously, nitrite and nitrate were removed during the anoxic stage. Kim et al. (2013, 2015) reported that most phenol was first degraded in the anoxic tanks with any remaining phenol almost completely degraded in the oxic tanks. However, Liu et al. (2015) have reported that the degradation of phenol occurred under the oxic stage in the same reactor. Aburto et al. (2009) reported that oxygen concentrations below 0.03 mg l−1 can support aerobic benzene-degrading communities. Therefore, the discrepancy of phenol degradation might explain by the two times 5-min aeration during the anoxic phase. Subsequently, ammonia was nitrified in the period of aerobic stage. During the aerobic period (120–240 min), the ammonia removal rate of phases I, II, and III, was 20.76, 10.20, and 12.42 mg N l−1 h−1, respectively. Shortening the settling time led to the AOB activity decrease. The influent phenol concentration increase in phase III did not have any inhibition effect on the AOB activity. During this period, the ammonia concentration decreased fast for nitrification and nitrate accumulated. Although there was accumulation of nitrite, it decreased quickly to about zero in phases I and III, except stage
Environ Sci Pollut Res
II. Data showed that the simultaneous denitrification in this period occurred, leading to the removal of TN. The total removed nitrogen during the aerobic stage (120–320 min) was 11.38, 17.79, and 0.18 mg l−1 in phases I, II, and III, respectively. It was obvious that SND and conventional biological nitrification/denitrification contributed to the nitrogen removal in phases I and II, while aerobic SND almost did not played any role in phase III. It was known that the nitrogen removal during the SND process was achieved by the coupled denitrification during the aerobic stage, i.e., aerobic denitrification (Pseudomonas spp., Alcaligenes faecalis, Thiosphaera Pantotropho, Nitrosomonas-like AOB) and heterotrophic denitrification (Chiu et al. 2007; Wan et al. 2009). The average diameters of sludge particles in phase I and phase II were 131.2 and 135.7 μm, respectively. Numerous researches have shown experimentally as well as by mathematical modeling that oxygen penetration is restricted to the outer rim (
RESEARCH ARTICLE
An anoxic-aerobic system for simultaneous biodegradation of phenol and ammonia in a sequencing batch reactor Qifeng Liu 1,2 & Vijay P. Singh 3 & Zhimin Fu 1,2,3
&
Jing Wang 1,2 & La Hu 1,2
Received: 20 December 2016 / Accepted: 15 March 2017 # Springer-Verlag Berlin Heidelberg 2017
Abstract A laboratory-scale sequencing batch reactor (SBR) was investigated to treat artificial pretreated coal gasification wastewater that was mainly contained of ammonia and phenol. The efficiency of SBR fed with increasing phenol concentrations (from 150 to 300 mg l−1) and the relationship among phenol, nitrogen removal, and the microbial community structure were evaluated. When the phenol feeding concentration was increased to about 300 mg l−1, the removal efficiency was above 99.0%, demonstrating the robustness of phenol removal capacity. The study showed that most phenol was degraded in anoxic stage. The average removal efficiencies of ammonia and total nitrogen were 98.4 and 81.9%, respectively, with average NH 4 + -N concentration of 107.5 mg l−1 and COD/N 7.5. Low temperature caused sludge loss that led to the decreased performance. Increasing the temperature could not recover the performance effectively. The data from bacterial analysis revealed that Delftia, Hydrogenophaga, and unclassified Xanthomonadaceae played a significant role in phenol degradation before the temperature increase, while uncultured Syntrophococcus sp. and
Responsible editor: Bingcai Pan * Zhimin Fu [email protected]
1
School of Ecology and Environment, Inner Mongolia University, Hohhot 010021, People’s Republic of China
2
Inner Mongolia Coal Chemical Industry Wastewater Treatment and Reuse Engineering Technology Research Center, Inner Mongolia University, Hohhot 010021, People’s Republic of China
3
Departments of Biological and Agricultural Engineering and Zachry Department of Civil Engineering, Texas A&M University, College Station, TX 77840, USA
unclassified Rhodocyclaceae were responsible for phenol degradation after the temperature increase. These results imply that the SBR holds potential for the simultaneous removal of phenolic compounds and nitrogen through aerobic ammonia oxidation and anoxic denitrification with phenol as the coorganic carbon source. Keywords SBR . Ammonia . Phenol . Denitrification . PCR-DGGE
Introduction Several industrial processes, such as petroleum refinement, coke oven plant, and petrochemical manufacturing paints, discharge complicated wastewaters containing both ammonia and phenolic compounds (Milia et al. 2012; Kim et al. 2013). It is well known that the ammonia discharge deteriorates water ecosystems through stimulation growth of fastgrowing biomass and bloom-forming algae (Moreno-Marín et al. 2016). While phenolic compounds are toxic to humans, animals, and plants (Al-Khalid and El-Naas 2012), phenol has been listed in the water priority control pollutant blacklist by many countries, including China and the USA. Therefore, both ammonia and phenol must be removed before wastewater discharge in order to reduce their adverse environmental impacts and protect waterbody ecosystems. The cost-effective ammonia removal method is the biological nitrogen removal process based on nitrification and denitrification reactions. Typically, the nitrification process is performed in two consecutive steps: ammonia oxidation to nitrite by ammonia-oxidizing bacteria (AOB), and nitrite oxidation to nitrate by nitrite-oxidizing bacteria (NOB) utilizing dissolved oxygen (DO) as electron acceptor. In the denitrification process, nitrate is reduced to nitrogen gas using organic
Environ Sci Pollut Res
compounds as electronic donor. Biological treatment is also considered to be efficient for mitigating the effects of phenol. Phenol can be degraded to harmless compounds by microorganisms under aerobic or anaerobic conditions (Al-Khalid and El-Naas 2012; Rosenkranz et al. 2013). Conventionally, aerobic processes are preferred due to microorganism’s faster growth and its mineralization (Al-Khalid and El-Naas 2012). It has been widely documented that phenolic compounds are inhibitors of nitrifying microorganisms, especially AOB (Lauchnor et al. 2011; Lauchnor and Semprini 2013). However, Amor et al. (2005) reported that in the aerobic activated sludge reactor, there was no inhibition of nitrification by phenol when the phenol concentration was increased from 35 up to 2800 mg l−1. The absence of inhibition can be attributed to the negligible phenol concentration in the unit as a result of fast mixing in the stirred tank reactor as well as its fast removal. During the denitrification process, nitrite and nitrate usually need an anaerobic condition and an organic carbon source for dinitrogen gas production. The denitrification reaction can be restricted by the deficient amount of organic carbon source, which leads to a large excess of nitrogen in effluent (Kartal et al. 2010). It has been suggested to add additional organic carbon source, such as methanol or lactate, for optimal denitrification, but due to increase in operation costs, it is undesirable from the engineering point of view (Torà et al. 2011). Researchers have demonstrated that phenol can be biodegraded under anoxic denitrifying conditions, with phenol mineralized to small-molecule acids and nitrate reduced to nitrogen. Bajaj et al. (2010) investigated the effect of phenol as a toxic co-substrate with glucose towards an anoxic mixed microbial flora in a reactor that removes nitrate. Batch assay results of phenol degradation under anoxic conditions have shown high phenol degradation rates at low concentrations (188 mg l −1 ). Anoxic bioreactors, dominanted with Rhodococcus pyridinivorans, have shown the highest removal efficiency of various organic pollutants (phenolics, aromatic hydrocarbons, and cyanide) commonly found in coke oven wastewater (Sharma and Philip 2015). Recently, denitritation using phenol as the sole organic carbon to reduce nitrite was successfully achieved with granular biomass in an upflow sludge blanket reactor (Ramos et al. 2016). The most abundant genera in the anoxic granules were denitrifying bacteria Ignavibacterium and Denitratisoma. These studies present a possible technological option to achieve simultaneous removal of phenol and ammonia nitrogen using the phenol as the carbon source for the denitrification process. The concept of this article is that nitrate produced from ammonia oxidation is removed by using phenol as electron donor. This can be realized by using a sequencing batch reactor (SBR) with an anoxic stage and an aerobic stage. In anoxic stage, nitrite and nitrate can be removed by heterotrophic denitrification with phenol as co-substrate organic substrates,
while in aerobic stage, the ammonia can be oxidized to nitrite and nitrate. The objective of this study therefore is to demonstrate this concept and underlying mechanisms by bacterial analysis.
Experimental method Experimental setup and operation A Plexiglas cylindrical SBR reactor with an effective height 1.0 m and a diameter of 0.07 m was utilized in this study. A guide tube with a height of 0.6 m and a diameter of 0.04 m was fixed within it. The operation cycle time was 6 h, which included anoxic stage for 120 min (including 10 min of substrate filling), aerobic stage for 330–350 min, and settling stage for 30–10 min (including 5 min of effluent withdrawal). Influent was supplied by a peristaltic pump (BT100-2J, LongerPump/ Baoding Longer constant-flow pump Co. Ltd., China), and air was supplied through a diffuser at the reactor bottom with an airflow rate of 18 l/h, which was measured by a glass gas flow meter (LZB-3WB, Kede/Changzhou Kede thermal instrument Co. Ltd., China). To enhance the mixture of the anoxic stage, aeration was provided two times at 40 and 80 min, lasting each time for 5 min. The volume exchange ratio was 27–28%. The reactor was inoculated with activated sludge from an anoxic-oxic process treating municipal wastewater for simultaneous carbon and nitrogen removal (Xinxinban, Hohhot). Data were measured after 1-month cultivation. The operation period consisted of three different phases (Table 1). In phase I, the settling stage was 30 min, with influent phenol concentration of 150 mg l−1. In phase II, the settling stage was decreased from 30 to 10 min, but the influent composition was the same as phase I. In phase III, the settling stage was 10 min, and influent phenol concentration was increased from 150 to 300 mg l−1. The temperature variation is listed in Table 1 and marked in Fig.2. In phase I, first, the temperature was kept at 25 °C; then, it was decreased to 18 °C. During phase II and the first part of phase III, the temperature was not adjusted but changed as the atmosphere temperature. A heating rod was added into the reactor from 185 days (III) to keep the temperature at 25 °C. Synthetic wastewater and analysis methods The influent synthetic wastewater (SWW) was prepared in tap water containing the following (g l−1): 0.35–0.7 glucose, 0.15–0.30 C6H5OH, 0.418 NH4Cl, 0.044 KH2PO4, and 0.8 NaHCO3. To enhance the setting ability of the sludge, CaCl2 (0.141 g l−1) was added to phase II and phase III. NH4+-N, NO3−-N, NO2−-N, phenol, and COD were analyzed according to the Water and Wastewater Monitoring Analysis Method (CEPA 2002). Total nitrogen (TN) was the
Environ Sci Pollut Res Table 1 Phase
Operation parameters of different phases days
COD (mg l−1)
Phenol (mg l−1)
Ammonia (mg l−1)
I
1–70
786.95 ± 94.91
156.23 ± 26.17
102.6 ± 30.07
II
71–118
793.46 ± 125.45
140.76 ± 23.57
108.38 ± 18.38
III
118–252
968.86 ± 98.87
268.60 ± 75.50
110.51 ± 24.69
sum of NH4+-N, NO3−-N, and NO2−-N. Liquid samples collected at regular time intervals during the cycle period were analyzed to reveal the NH4+-N and phenol transform process. DNA extraction and polymerase chain reaction-denatured gradient gel electrophoresis Active sludge samples were taken out in phase III. The total DNA was extracted with a soil genomic DNA quick extraction kit SK8233 (Sangon Biotech, China). The extracted RNA was digested with RNase A (Sangon Biotech, China) for RNA elimination and extracted with chloroform and alcohol according to the recommendations of the kit protocol. The treatment efficiency was verified by PCR, which showed amplification with the universal primers F338-GC (5′-CGCC CGCCGCGCCCCGCGCCCGGCCCGCCGC CCCCGCCC CCCTACGGGAGGCAGCAG-3′) and R518 (5′-ATTA CCGCGGCT GCTGG-3′). The PCR reaction system contained 5 μl of 10× buffer (containing 2.0 mM MgCl2), 0.5 μl of template DNA, 0.25 μl of Taq enzyme (5 U μl−1), 2 μl of primers (10 μM), 1 μl of dNTP mixture at 10 mM, and 41.25 μl of sterilized distilled water. PCR conditions were as follows: predegeneration at 94 °C for 240 s, denatured at 94 °C for 30 s, primer annealing at 56 °C for 60 s, and chain extension at 72 °C for 30 s with an additional extension time of 7 min on the final cycle for a total of 30 cycles. The total DNAwas stored at −20 °C in a freezer for further application. For DGGE analysis, 400 ng of PCR V3 product was loaded onto 8% (w/v) polyacrylamide gel with a 30–60% denaturant gradient. Electrophoresis was performed at 60 °C at 60 V for 16 h. The middle portion of each selected DGGE band was excised, washed, transferred to 1.5-ml centrifuge tube, and recycled PCR products with kit SK8183 (Sangon Biotech, China). PCR products were reamplified and sequenced.
Results and discussion COD and phenol removal efficiency During the operation (Fig. 1a), the influent COD concentrations varied between 650 and 1100 mg l−1, with an average COD concentration of 842.5 mg l−1. Data showed that the
Temperature (°C)
Sludge
Settling time (min)
25 (1–19 days) >18 (21–70 days) Ambient (71–118 days)
Recycle to SBR
30
No recycle
15 (71–76 days) 10 (76–85 days)
Ambient (118–185 days) 25 (185–255 days)
No recycle
5
average effluent COD concentration was 102.3 mg l−1, which indicated that the COD removal performance had been less affected by the temperature change, phenol concentration increase, and the settling time decrease. The average COD load was 0.91 g l−1 day−1, and the average removal efficiency of COD was 87.6%. These data indicated that the system could provide a consistent high efficiency of COD removal. During the experiment operation period, phenol concentration and removal efficiency variations are shown in Fig. 1b. Although the corresponding average influent phenol load was increased from 164.8 to 284.6 mg l−1 day−1, the removal efficiency of phenol was higher than 99.0%. Most of the effluent concentration of phenol was lower than 0.5 mg l−1, except for the days of temperature decrease (phase I, 22 days; phase III, 150, 162 days) and influent phenol concentration increase (phase III, 118 days). A sharp buildup of phenol was observed in the reactor. However, this buildup quickly dissipated as the sludge adapted rapidly to the change, indicating that the phenol removal capacity of the system was robust. However, the effluent phenol concentration was kept as 1.0 mg l−1 during phase III (190–218 days), which was due to the loss of mixed liquid volatile suspended solid (MLVSS) (Fig. 2b). Amor et al. (2005) reported that high phenol removal efficiencies (above 99.9%) were maintained at all the applied organic loading rates (30–2700 mg COD l−1 day−1), which was increased stepwise by increasing the phenol concentration from 35 up to 2800 mg l−1. Adav et al. (2007) developed aerobic granules that degrade 1000 mg l−1 phenol at a constant rate of 1176 mg phenol g−1 VSS day−1. Furthermore, adding acetate enhanced the phenol degradation rate for the granular system (Ho et al. 2009). Therefore, the low phenol concentration and loading rate as well as the glucose presence assured the SBR systems to degrade phenol effectively. NH4−-N and TN removal efficiency The average ammonia concentration in the influent was maintained constant at 107.5 mg l−1, and the applied ammonia loading rate was 116.1 mg NH4+-N l−1 day−1. The evolution of ammonia in the influent and effluent and ammonia removal efficiency of the system are shown in Fig. 2a. During the stable operation days of phases I (50–71 days), II (92– 118 days), and III (125–155 days), the average NH4+-N
Environ Sci Pollut Res Fig. 1 The COD and phenol concentration variations during the operation. a COD in the influent, effluent, and COD removal efficiency. b Phenol in the influent, effluent, and phenol removal efficiency
removal efficiencies were 99.2, 95.5, and 98.4%, respectively, and effluent NH4+-N concentrations were 0.88, 4.65, and 1.89 mg l−1, respectively. Data showed that the decrease in the settling time from 30 to 10 min led to a slight decline in the nitrification performance, which was speculated due to the sludge washout (Fig. 2b). AOB are chemoautotrophic and slow-growing; accordingly, numerous AOB populations, which cannot become big and dense enough to settle fast, were washed out from the system (Zhang et al. 2011). The increased sludge (72–90 days) concentration was attributed to the heterotrophic organism growth. The 98.4% NH4+-N removal efficiency of phase III indicated that no inhibition of nitrification by phenol, which was consistent with the results reported by Amor et al. (2005). The absence of inhibition can be due to the negligible phenol concentration in the unit as a result of its fast removal in the anoxic stage (Fig. 4).
On day 19 (phase I), the heating rod was suddenly broken and the atmosphere temperature was fluctuated between 3 and 15 °C, and the NH4+-N concentration in the effluent was dramatically increased to about 40–50 mg l−1 from day 20. The heating temperature was set at 18 °C on day 21. The NH4+-N concentration in the effluent was not decreased until day 50. The recovery of the nitrification process required very long times (1 month) owing to the slow growth rates of nitrifying bacteria (Kim et al. 2013). However, the reactor experienced a sharp decrease on day 140 days (4–15 °C, phase III) of the atmosphere temperature, and the NH4+-N removal efficiency was not affected. Meanwhile, during days 160–170, the reactor experienced a low temperature (4–20 °C) and the ammonia removal efficiency decreased to 64%, but it was recovered to 90% on day 171, which was speculated to the higher MLVSS (4765–4180 mg l−1) during the period.
Environ Sci Pollut Res Fig. 2 The NH4+-N, MLVSS, SVI, and temperature variations during the operation. a NH4+-N in the influent, effluent, and NH4+-N removal efficiency. b MLVSS, SVI, and temperature
The MLVSS started to decrease from day 160 as shown in Fig. 2b. Winkler et al. (2012) reported a twofold slower settling velocity for the same granules when the temperature of water decreased from 5 to 40 °C, and the main factor influencing the settling velocity was the change in the viscosity of water. Therefore, the temperature was controlled at 25 °C after day 185 to avoid deterioration of the process performance. However, contrary to expectations, the MLVSS decreased from 3545 to 1190 mg l−1 and the sludge volume index (SVI) increased from 39.5 to 100.8 ml g−1. Data showed that the temperature increase did not stop the loss of sludge (Fig. 2b). The ammonia
concentration of the effluent was about 50 mg l−1 due to the low sludge concentration. The NO2−-N, NO3−-N, and TN concentration variations and TN removal efficiency during the operation are shown in Fig. 3. During the stable operation days of phases I (50– 71 days), II (92–118 days), and III (125–155 days), the average TN removal efficiencies were 82.1, 73.9, and 81.9%, respectively, under the COD/N ratios of 7.0, 7.6, and 7.5. The average TN efficiencies were higher than that reported by Guo et al. (2013). Wan et al. (2009) suggested that alternating anoxic feast-aerobic famine encouraged heterotrophic growth deeply inside the aggregates and promoted the denitrification
Environ Sci Pollut Res Fig. 3 The NO2−-N, NO3−-N, and TN concentration variations and TN removal efficiency during the operation
efficiency. Hence, it was reasonable to conclude that the anoxic-aerobic SBR had higher denitrification ability than aerobic simultaneous nitrification and denitrification (SND) SBR (Zhang et al. 2015). Under anoxic conditions, denitrifiers utilized biological organic substrate as the electron donor to reduce nitrite and nitrate. The COD/N ratio of influent was one of the most critical parameters for wastewater nitrogen removal process. Chiu et al. (2007) reported that the SBR system resulted in the optimal removal of both organic matter and NH4+-N with no accumulation of intermediate by-products, when the COD/N ratio was controlled at 11.1. The low C/N ratio, causing the electron donor deficiency for denitrification process, resulted in a low TN removal efficiency (Roy et al. 2010). However, in the present study, the higher COD/N value did not coincidence with the greater TN removal efficiency. In phase II (Fig. 4b), it was found that the removed ammonia exceeded nitrification products, implying that the aerobic denitrification or SND process contributed to the nitrogen removal. Taking into account that the vast majority of organic matter was consumed in the anoxic stage, it was reasonable to speculate that endogenous carbon source (polyhydroxybutyrate (PHB), glycogen) was utilized for the denitrification process (Winkler et al. 2011). The endogenous denitrification rate was generally lower than that of a pre-anoxic process, because the endogenous carbon degradation rate limited the nitrogen performance rate for the SND process (Zhu et al. 2013). Therefore, the low TN removal efficiency of phase II was attributed to the low
endogenous denitrification rate, and prolonged aeration reaction time might enhance the nitrogen removal performance. During the stable operation days of phases I (50–71 days), II (92–118 days), and III (125–155 days), the average effluent NO3−-N concentrations were 17.1, 15.6, ad 14.4 mg l−1, and the average effluent NO2−-N concentrations were 1.5, 6.8, and 7.0 mg l−1, respectively. Although NO3−-N was the main nitrification product, the effluent NO2−-N concentration was increased obviously in phases II and III. In phase II day 85, the settling time was shortened to 10 min, which caused the concentration of activated sludge to decrease sharply. The nitrification bacteria, AOB and NOB, were washed out from the reactor. Paredes et al. (2007) reported that at the temperature over 15.0 °C, AOB had higher growth rates than NOB. Therefore, NO2−-N was accumulated in phase II. In phase III, the NO3−-N production dropped drastically and NO2−-N accumulated, which suggested that the NO2−-N accumulation was caused by a drastic decline in the NOB activity. In day phase III (125–155 days), the atmosphere temperature decreased gradually and varied between 5 and 26 °C. Both AOB and NOB are sensitive to temperature and suffered from the temperature decrease. The specific ammonia oxidation rate decreased by 1.5 times with the temperature decreasing from 25 to 15 °C (Guo et al. 2010). During the critical temperature 12–14 °C, the rate of nitrate formation controls the overall nitrification process, meaning that nitrite would accumulate (Robinson et al. 2004). Furthermore, Nitrospira and Nitrobacter are the commonly identified NOB in the
Environ Sci Pollut Res
Fig. 4 Typical profiles of cycle study and particle size distribution. a Phase I. b Phase II. c Phase III. d Particle size distribution
nitrogen removal wastewater treatment plants (Huang et al. 2010). Nitrospira would outcome Nitrobacter in the wastewater treatment system, due to that Nitrospira had a lower nitrite half-saturation constant and the NO2−-N concentration in system was low (Blackburne et al. 2007). Gilbert et al. (2014) reported that Nitrospira was affected most by temperatures below 13 °C, showing no recovery. The activity of AOB increased more than that of NOB as temperature increased (Zhang et al. 2014). Therefore, the NO2−-N accumulation in phase III was speculated due to the low temperature. In days 20–50 (phase I), the TN removal efficiency decreased and fluctuated at 45–50; at the same time, the nitrite concentration was higher than nitrate, which was also attributed to the low temperature. Results of cycle experiment Figure 4 shows the typical phenol and nitrogen species profiles in reactors during a steady-state cycle in phases I, II, and III. Throughout the anoxic stage, cycle profile measurements revealed that almost 100% phenol and COD (not given) were
utilized in phases I, II, and III. Simultaneously, nitrite and nitrate were removed during the anoxic stage. Kim et al. (2013, 2015) reported that most phenol was first degraded in the anoxic tanks with any remaining phenol almost completely degraded in the oxic tanks. However, Liu et al. (2015) have reported that the degradation of phenol occurred under the oxic stage in the same reactor. Aburto et al. (2009) reported that oxygen concentrations below 0.03 mg l−1 can support aerobic benzene-degrading communities. Therefore, the discrepancy of phenol degradation might explain by the two times 5-min aeration during the anoxic phase. Subsequently, ammonia was nitrified in the period of aerobic stage. During the aerobic period (120–240 min), the ammonia removal rate of phases I, II, and III, was 20.76, 10.20, and 12.42 mg N l−1 h−1, respectively. Shortening the settling time led to the AOB activity decrease. The influent phenol concentration increase in phase III did not have any inhibition effect on the AOB activity. During this period, the ammonia concentration decreased fast for nitrification and nitrate accumulated. Although there was accumulation of nitrite, it decreased quickly to about zero in phases I and III, except stage
Environ Sci Pollut Res
II. Data showed that the simultaneous denitrification in this period occurred, leading to the removal of TN. The total removed nitrogen during the aerobic stage (120–320 min) was 11.38, 17.79, and 0.18 mg l−1 in phases I, II, and III, respectively. It was obvious that SND and conventional biological nitrification/denitrification contributed to the nitrogen removal in phases I and II, while aerobic SND almost did not played any role in phase III. It was known that the nitrogen removal during the SND process was achieved by the coupled denitrification during the aerobic stage, i.e., aerobic denitrification (Pseudomonas spp., Alcaligenes faecalis, Thiosphaera Pantotropho, Nitrosomonas-like AOB) and heterotrophic denitrification (Chiu et al. 2007; Wan et al. 2009). The average diameters of sludge particles in phase I and phase II were 131.2 and 135.7 μm, respectively. Numerous researches have shown experimentally as well as by mathematical modeling that oxygen penetration is restricted to the outer rim (
