Anaerobic/Oxic/Anoxic Mode Sequencing Batch Reactor: Treatment Performance and Development of Aerobic Granular Sludge

Objective: The objective of the present work was to reduce the start-up time of aerobic granular sludge (AGS) formation using medium-strength sewage and to find the most effective anaerobic, aerobic (oxic), and anoxic (A/O/A) cyclic time distribution over the 6 h cycle time for the removal of nutrients. Methods: A rectangular sequencing batch reactor (SBR) operated systematically in A/O/A cycle was designed for AGS formation. In order to achieve complete nutrient removal, the anoxic phase (5% of the total cycle duration) was introduced in the SBR cycle. Results: Growth in the biomass profile was observed after 20 days of operation, and complete granulation was achieved within 35 days of operational studies. Regular-shaped, compact granules with visible outlines and clear boundaries predominate inside the system, with an average particle size of 220.19 µm. The A/O/A system achieved average removal efficiencies of 70.02±5.98%, 78.21±18.66%, 60.62±9.69%, and 53.49±8.14% for COD, NH 4+ -N, TN, and PO 43--P, respectively. The reasons for poor COD and PO 43--P removal performance were the short anaerobic phase (25% of total cycle duration of 6 h) and higher biomass concentration. However, the long aerobic phase (60% of total cycle duration of 6 h) with the anoxic phase achieved outstanding (100%) NH 4+ -N removal efficiency after the system reached a steady state. SEM results inferred that Rod-shaped bacteria , Cocci , and Filamentous fungi were the dominant microorganisms found within the cultivated granules. Conclusions: Sewage with a high COD loading rate accelerates AGS formation and development within an A/O/A cycle SBR reactor. The rapid granulation and simultaneous removal of organics and nutrients by the A/O/A mode SBR without any support materials provide an alternative for wastewater treatment.


INTRODUCTION
Aerobic granular sludge (AGS) is an excellent alternative technology for treating both low-strength and high-strength wastewater. Up to now, various biological wastewater treatment processes have been implemented at full scale in the field and at lab scale for industrial and municipal wastewater treatment [1][2][3][4][5] . The activated sludge process (ASP) is considered one of the best conventional biological treatment approaches as it utilizes microorganisms to eliminate biodegradable contaminants found in different types of wastewaters. However, AGS offers several advantages over conventional activated sludge (CAS) systems. These advantages include outstanding settling properties, a compact microbial structure, the potential for nutrient recovery, the ability to handle high organic loading rates (OLRs), high efficiency in removing organics and ammonia, reduced electricity requirements (23-40%), lower operation costs (20-25%), and decreased space requirements (50-75%) compared to CAS systems 4,6,7 .
However, the long start-up periods and instability during long-term operation are some of the major factors that hindered the wide application of AGS based wastewater treatment plants (WWTPs). The previous studies reported that a minimum of 25-35 days were required to achieve complete granulation from CAS if no support material was applied 4 . Moreover, the high costs associated with operation and maintenance pose further obstacles to the widespread growth of AGS technology. Therefore, it is crucial to explore methods that can accelerate the granulation system while focusing on cost savings and resource consumption 1 .
The AGS was commonly used in sequencing batch reactors (SBRs) for the treatment of lowstrength and high-strength wastewater. The typical operation regime of an SBR involves several steps, including filling, reaction (consisting of aerobic, anaerobic, and oxic phases), and decanting or removing the effluent from the SBR 8 . The cycle types are very important factors in determining treatment performance, stability, and the generation of AGS in SBRs [8][9][10] . Anaerobic, aerobic (oxic), and anoxic (A/O/A) cycles were introduced a few decades ago with the aim of enriching microbial communities for the simultaneous removal of organics and nutrients throughout the granulation process 11 .
In A/O/A cycles, the oxic phase that follows the anoxic phase promotes the proliferation of denitrifying glycogen-accumulating organisms (DGAOs) and denitrifying phosphate-accumulating organisms (DPAOs) 8,12 . It has been observed that in A/O/A cycles, the residual oxygen in the anoxic phase can effectively remove NH 4 + -N from the wastewater 13 . However, researchers have also noted that the competition between DGAOs and DPAOs for carbon sources can lead to low phosphorus removal and the accumulation of nitrite (NO 2 --N) and nitrate (NO 3 --N) in the effluents, as both groups of organisms require carbon sources 12,13 .
Anaerobic slow feeding, also known as "filling," is typically beneficial for slow-growing microorganisms such as glycogen-accumulating organisms (GAOs) and polyphosphate-accumulating organisms (PAOs) 11 . The granulation process, involving the development of nitrifying bacteria, PAOs, and GAOs, can benefit from alternating anaerobic and aerobic conditions, as the growth of heterotrophic bacteria is hindered in the aerobic phase due to the lack of a carbon source 11 .
In anaerobic and oxic (A/O) cycles, which involve an anaerobic phase along with an aerobic or oxic phase, a higher percentage of the oxic phase (accounting for 65% of the overall cycle length) has been found to promote the formation of granules with good settling ability and resistance 8 . Conversely, long anaerobic or anoxic phases in A/O/A cycles resulted in the formation of puffy and unstable granules, despite their effectiveness in nutrient removal 8 . Previous studies have also reported that including medium aerobic phase duration (55% of the total cycle time) and short anoxic phase duration (10% of the total cycle time) in A/O/A cycles led to superior treatment performance and stability of granules in SBRs 7,8 .
Therefore, the objective of this study was to address the issue of long start-up times for AGS formation and optimize the distribution of cycle times as a cost-effective and environmentally friendly approach for organic and nutrient removal in wastewater treatment. In this work, we designed a novel strategy A/O/A mode SBR reactor in which the aeration period is fixed at 60% of the total cycle duration of 6 h for complete nitrification and phosphorous sequestration. The remaining times were adjusted in the anaerobic and anoxic phases. This strategy reduced operational costs, specifically energy costs, while improving the removal efficiency of nutrients and organics matter.

Reactor configuration
A rectangular reactor was set up for the cultivation of AGS with a total volume of 20 litres and a working volume of 14 litres. The volume exchange ratio (VER) was 40%, and the hydraulic retention time (HRT) was fixed at 12 hours. The reactor configuration is depicted in Figs. 1(a) and 2. The current study lasted for 60 days, whereby an alternating A/O/A mode SBR reactor was fed with sewage collected from the partition chamber of sewage treatment plant (STP) Bahadurgarh, Haryana, India. The cycle time was adjusted to 6 h, which included the filling phase (20 min or 5% of the total cycle duration), the anaerobic phase (90 min or 25% of the total cycle duration), the oxic phase, or aerobic phase (216 min or 60% of the total cycle duration), and the anoxic phase (18 min or 5% of the total cycle duration), which is followed by settling and decanting for 10 min and 6 min, respectively.
The influent (sewage) was pumped into the A/O/A SBR reactor using a peristaltic pump (PP-274-6A NEOLAB). Aeration was introduced into the reactor with an aquarium air pump (SB-348A, SEBO) and dispersed by the four diffusers located at the base of the reactor. During the aerobic phase, mixing was carried out using an agitator at a constant speed of 40rpm. The influent filling phase, anaerobic phase, and anoxic phase were considered static, without aeration and stirring. The whole automatic cyclic operation was regulated by the laptron timers. An effluent discharge port was located at an 8-cm height, through which treated effluent was discharged with the help of a solenoid valve (2W-160-15 AKARI). The AGS was not discharged throughout the operational studies. The highest lab temperature, DO during the oxic phase, influent pH, and effluent pH were monitored daily, typically in the ranges of 15-24°C, 3-4 mg/L, 6.9-8.1, and 7.4-8.2, respectively.

Wastewater composition and seed sludge
The seed sludge used in this study was obtained from a 36 MLD STP in Bahadurgarh, Haryana, India, based on SBR technology. The seed sludge had a mixed liquor suspended solids (MLSS) concentration of 4490 mg/L and a mixed liquor volatile suspended solids (MLVSS) concentration of 2560 mg/L. On day 0 of the study, the reactor was seeded with 3.8 litres of aerobically cultivated sludge.
The reactor was fed with sewage collected from the partition chamber after the fine screen of the UASBbased full-scale STP Bahadurgarh throughout the experimental studies. The average composition of sewage used as an influent in this experimental work was as follows: COD 818.01±121.98 mg/L, BOD 353.33±103.66 mg/L, PO 4 3--P 15.11±1.54 mg/L, Total Nitrogen (TN) 63.99±9.39 mg/L, SO 4 2-462.99±98.72 mg/L, and S 2-8.71±1.65 mg/L, respectively.

Analytical methods
The influent (sewage) and lab-scale SBR reactor-treated effluent samples were collected twice a week for analysis. Parameters such as COD, TN (sum of NH 4 + -N, NO 2 --N, NO 3 --N), PO 4 3--P, and SO 4 2were measured via spectrophotometer following the standard methods (APHA 2005). Meanwhile, alkalinity, BOD, and S 2concentrations were monitored by titration methods (APHA 2005). DO, pH, and temperature were regularly monitored using WTW Oxical SL probes, a pH meter, and a thermometer, respectively. To assess the biomass concentration in the reactor, MLSS, MLVSS, and the sludge volume index (SVI) were measured using the standard protocol for the examination of water and wastewater. The TSS and VSS were also evaluated according to the standard (APHA 2005) protocol. The settling velocity, formation, and development of granules were examined by particle size analysis, SV 30 , SV 10 , and SV 5 . The AGS size was measured through a Zeta potential analyzer.

Statistical methods
All analysis were carried out in triplicate, and the results were expressed as the mean value of triplicates. The data was statistically analyzed using Two-way analysis of variance (ANOVA) and Pearson correlation coefficient, etc., using Origin 2022.

Scanning electron microscopy (SEM)
SEM was performed at the Centre for Nanoscience and Nanotechnology, Jamia Milia Islamia, New Delhi, India, for the examination of microstructure, microbial communities, and morphology of biomass in the A/O/A cycle SBR reactor as described M. W. Mirza et al., 14 . Two AGB samples were collected, one on day 0 seed sludge and another on day 60 mature granules in the reactor, immobilized onto a poly-L-lysine-coated glass slide, and allowed to air dry. After that, 2.5% glutaraldehyde was used to fix these dried samples, and stored in a freezer at 4 0 C for twenty-four hours. Afterwards, phosphate buffer solution with pH of 7.4 was used to rinse these fixed samples in a centrifuge at a speed of 4000 rotations per minute (rpm) for a minute. The free-moving upper supernatant was discarded, and solid sludge samples were collected. After this step, double distilled water was utilized for the preparation of a graded ethanol series consisting of distinct proportions of 30%, 40%, 70%, and 100%, respectively. The samples were then dehydrated using a succession of increasing concentrations of ethanol, followed by drying in a hot air oven at 110 0 C. Finally, these prepared samples were coated with gold (100 nm) and examined via scanning electron microscopy.

Fourier transform infrared spectroscopy (FT-IR)
An FTIR investigation was carried out to explore the functional groups. The FTIR spectra were acquired using potassium bromide (KBr) pellet techniques. The seed sludge used to start the SBR reactor and the mature AGS obtained after 60 days of operation were collected and dried in an oven. 2 mg of dried AGS sample was mixed with 250 mg of potassium bromide and ground into a fine powder. The mixture was then compressed into a pellet using a hydraulic press (Spectrachrome instrument CAP-15T). The prepared pellet was placed in the holder, and FT-IR spectra were obtained using an FT-IR instrument (Varian 7000 FTIR) at Jawaharlal Nehru University, New Delhi, India.

Formation and development of granules
Sewage or high-strength industrial wastewater contains various carbon sources along with different other pollutants such as particulate matter, endocrine disrupting agents, colloidal species, etc., which are not easily removed by biological treatment processes. The cultivation of AGS using real sewage leads to different morphological structures of granules, such as poorer physical strength and stability, and further effects on EPS composition and the microbial community when compared with AGS grown with synthetic wastewater. In this experimental work, we focused on developing granules in A/O/A cycle SBR reactor fed with real sewage. The formation and development mechanisms of AGS occurred in four stages: acclimating, shaping, developing, and finally maturing.
In the acclimating phase, the microbial communities within the granules were adjusted to the environmental conditions of a lab-scale reactor. After that, the color and shape of the granular sludge changed over time. The seed sludge used in this study was slightly black in color on day 0, as shown in Fig. 3 (a). When sewage was introduced into the reactor, the granules began to change color. On day 15, the entire sludge had turned brown, as shown in Fig. 3 (b), which turned a yellowish-brown color on day 20 of the experimental studies. AGS was mostly observed in yellow, but it was also seen in black 15 . According to F. A. Dahalan et al., 16 the yellow-brown morphology of granular sludge indicates the presence of aerobic microorganisms on the outer surface of the granules. During the initial stages of the study, minor amount of fine silt was present in the reactor along with particulate matter, which had adverse effects on the morphology, structure, shape, and size of the granules.
The average size of AGS was initially measured at 150.10 µm on day 0 using a particle size analyzer. During the initial start-up phase of operational studies (5-15 days), when the aeration phase ended, the entire reactor was filled with floccular sludge due to poor biomass settling velocity. However, after 20 days of operation, a few small, irregular granules were observed inside the reactor, indicating the initiation of granule growth. Afterward, a significant number of irregular granules were visible on day 25 of the study. Moreover, as a result of hydraulic shear force and aeration intensity, many round-shaped (regular) granules were observed in the reactor on day 34. At this stage, granules become dominant inside the reactor. The reactor ran continuously for 60 days. At the end of the study, 90% of the granules were regular, round-shaped, and exhibited a compact structure, with an average particle size of 220.19 µm.
respectively. The abundant biomass concentration was observed because AGS was only discharged through effluent withdrawal, and 50 mL of sludge was taken twice a month during the aeration phase for biomass concentration analysis.
The settling velocity of the AGS was determined by calculating the SVI. The initial SVI value of the seed sludge used to inoculate the reactor was measured at 46.77 mL/g. During the initial start-up phase, the SVI values exhibited an increasing trend. On day 6, the SVI value was 57.64 mL/g, and it continued to increase until day 20, reaching 240.81 mL/g. As mentioned earlier, granules start growing after day 20 of the experimental work. As the granules formed, the SVI value decreased to 62.25 mL/g on day 34 of operation. At this stage, complete granulation was achieved within the system, with regular-shaped granules dominating and exhibiting good settling ability.
However, some of the granules disintegrated inside the reactor, causing a drop in MLSS concentration, and the SVI value increased to 95.04 mL/g on day 47 of the study. After 50 days of experimental work, growth in the biomass profile was observed, and the system reached a stabilized state. At the end of the study, compact, round-shaped granules became dominant within the system, demonstrating good settling ability. The SVI value was measured at 84.66 mL/g, indicating a significant decrease (p < 0.05) and suggesting an improvement in the settling properties of AGS. According to Y. Liu and Q. S. Liu. 11 , an SVI below 80 mL/g in the SBR reactor indicates long-term stability and good settling ability of the AGS.

Biomass profile and settling characteristics
The initial characteristics of seed sludge were determined as MLSS 4490 mg/L and MLVSS 2560 mg/L, with SVI 46.77 mg/L. Fig. 4 shows SVI, MLSS, and MLVSS concentration dynamics. The results reveal that during the initial start-up phase of the operational studies, both MLSS and MLVSS concentrations exhibited a decreasing trend. On day 6 of the experimental work, the MLSS was detected at 3990 mg/L, whereas the MLVSS was determined at 1890 mg/L. Subsequently, on day 20 of the operational study, the MLSS and MLVSS were measured to be 2450 mg/L and 1580 mg/L, respectively. During this period, a significant amount of floccular sludge was washed out with the treated effluent, resulting in a reduction of MLSS and MLVSS concentrations. This decrease was attributed to the collapse of AGS due to hydraulic pressure, causing the breakdown of fewer compact granules. Consequently, granules with low settling abilities were washed out from the reactor. During the AGS start-up phase, the COD removal efficiency decreased from 82.39% on day 7 to 66.17% on day 14. This decrease was attributed to the washout of poor-settling granules. However, after day 20 of operation, the granules began to grow, and the removal percentage resumed to 68.12% on day 28. Afterward, no significant differences occurred in the removal efficiency.
In the stabilization period, after 30 days of the study, the biomass concentration increased, while the COD removal efficiency fluctuated between 66% and 69% on most of the days, as shown in Fig. 4 and 5. At the end of the study, a COD removal percentage of 78.25% was achieved. The average COD removal percentage over the entire study period was 70.02 ± 5.98%.
It should be noted that the results obtained in this study differ from those presented S. L. de S. Rollember et al., 8 where a COD removal efficiency of 90% was achieved in an A/O/A cycle SBR reactor. In the current study, the reactor was fed with medium-strength sewage containing COD of 818.01±121.98 mg/L, including filling+anaerobic phase (30%), and aerobic phase (60%), over a 6-h cycle duration. In contrast S. L. de S. Rollemberg et al., 8 used low-strength synthetic wastewater (COD 550 mg/L) along with filling+anaerobic phase (35%), and aerobic phase (55%), within a 6-h cycle duration. This indicates that the shorter anaerobic phase in this study may have led to poorer COD removal performance.
Another possible reason for the lower organic removal efficiency could be the higher biomass concentration of approximately 7.7 g/L in the reactor. A. Alattabi et al., 17 have reported that an increase in biomass concentration from 4 to 6 g/L resulted in a 10% decrease in COD removal percentage. They also suggested that a higher biomass concentration reduces the food-tomicroorganism (F/M) ratio in SBR reactors, making them less effective in organic removal.

Nutrient removal
The aerobic granular biomass in the system contains ammonium-oxidizing bacteria (AOB) and nitrate-oxidizing bacteria (NOB), which play a crucial role in the nitrification process. These bacteria oxidize NH 4 + -N to NO 2 --N or NO 3 --N while simultaneously removing organic matter from the wastewater. In this study, the average concentrations of NH 4 + -N in the influent and A/O/A SBR treated effluent were found to be 42.5±8.5 mg/L and 10.11±9.11 mg/L, respectively, with corresponding removal percentages of 78.21±18.66%.
During the start-up phase of the operational study, the NH 4 + -N removal percentages ranged from 56% to 58%. After day 17 of operation, there was a continuous increase in NH 4 + -N removal efficiency, indicating improvement within the system. This observation suggested that the system raised NH 4 + -N removal efficiency with the formation and development of granules. At the end of the stabilization period, after 30 days of experimental studies, the system achieved 100% NH 4 + -N removal efficiency.
The higher NH 4 + -N removal indicated the presence of enriched nitrifying bacteria (ammoniumoxidizing bacteria) in the reactor. These results differ from those reported by A. Alattabi et al., 17 , who observed an 11% decrease in NH 4 + -N removal rate with an increase in MLSS concentration from 4 to 6 g/L. However, the present study demonstrates that NH 4 + -N removal efficiency increased by 20% with biomass accumulation from 3 to 7 g/L inside the reactor. These findings suggest that increasing biomass concentration did not affect NH 4 + -N removal performance. These results are in agreement with the results reported by S. Rollemberg et al., 8 , who observed that a minimum of 50% aerobic phase is required for complete nitrification in the treatment of medium-strength wastewater. However, complete denitrification did not occur, as a high amount of NO 3 --N was detected in the treated effluent. In the start-up phase of the study, 5-12 mg/L NO 3 --N was present inside the reactor, and its concentration increased in the treated effluent after the stabilization period, ranging from 15 to 22 mg/L until day 56 of the study. This accumulation of NO 3 --N in the treated effluent may be attributed to a higher level of NO 3 --N in the sewage.
T h e ave rag e r em ova l o f TN ove r the experimental studies was measured at 60.62±9.69%. Overall, poor TN removal efficiency was observed in this experimental work due to the small size of the granules. The results reveal that complete nitrification and partial denitrification occurred after the stabilization period in the A/O/A cycle SBR reactor. The overall poor removal of PO 4 3--P can be attributed to the competition between phosphate-accumulating organisms (PAOs) and glycogen-accumulating organisms (GAOs). While GAOs metabolize carbon compounds and volatile fatty acids similar to PAOs, but GAOs are unable to accumulate PO 4 3--P 7 . Another likely factor is the absence of sludge discharge or withdrawal during the experimental work. According to 7,18 , PO 4 3--P removal efficiency increases with the removal of old sludge from the bottom of the reactor.
The results obtained in this experimental study differ from those reported by S. Rollemberg et al., 8 Fig. 9(a) illustrates the loose, irregular, and fluffy structure of the seed sludge used to start the system. In contrast, the mature granular sludge exhibited a regular shape and compact structure, with clear and visible outline boundaries for 90% of the granules, as shown in Figure 9(b). Furthermore, the SEM results showed the presence of spherical-shaped cocci, coccobacillus, and bacillus rod-like bacteria in the seed sludge. Extracellular polymeric substances (EPS) were also observed within the system. The microbial diversity and morphology of AGS changed with the passage of time. At the end of the experimental period, mature granules were observed, where groups of rod-shaped bacteria compacted with each other to form single colonies. Some cocci and filamentous fungi were also embedded on the surface of the rod-shaped bacteria. These findings align with the results reported by 19 , who observed the dominance of rod-shaped bacteria such as Flavobacterium species, Azoarcus species, and Thauera species in similar systems. The dense microbial population facilitates simultaneous nutrient and organic matter removal. Additionally, it also makes cell surfaces less water-repellent and makes it easier for biomass to settle.

Fourier transform infrared spectroscopy (FT-IR) analysis
FTIR analysis was conducted to determine the functional groups and structure of proteins that were present in the seed sludge used to start the granulation process and analyze the functional groups that were responsible for the AGS development and treatment system. Two samples were taken: sample (I) represented the seed sludge used as inoculum on day 0, while sample (II) was taken at the final stages on day 60, after the completion of the experiments.
According to the findings of the FTIR analysis (Fig. 10), the primary bands representing carbohydrate, carboxylic acid, nucleic acid, and amide groups were located in the range between 500 and 1700 cm -1 . These functional groups were thoroughly investigated. The peak at 1454 cm -1 indicated the existence of carboxylic acid groups in both the seed sludge and the mature granules, suggesting the acidic nature of extracellular polymeric substances (EPS) 20 . The amides (I), (II), and (III) were detected at the following frequencies: 1655 and 1646 cm -1 , 1549 cm -1 , and 1239 cm -1 , respectively 21 . Apart from this, the peaks at 2853 cm -1 and 2924 cm -1 were related to aliphatic methylene groups, and they served to stimulate the formation of lipids and fats 22 . Additionally, the sharp bands within the wavelength range of 2800-3000 cm -1 are indicative of C-H stretching vibrations, which are characteristic of lipid constituents and can contribute to the formation of microbial films 23 .
Lipids are a significant class of substances that are characterized by their low solubility in water and are typically located in sewage sludge. The presence of large bands at 1028 cm -1 indicates the degradation process of organic components in sewage 22 . In addition, the absence of bands between 1300 -1400 cm -1 indicates the absence of functional groups (NO 2 --N and NH 4 + -N), which is beneficial for the nitrogen removal process 14 . The FTIR spectra obtained from both samples show that the same primary components were present (at similar peak positions), but there was a difference in the degree to which they were enriched (the normalized peak absorbance value was different).
increased in sewage, which reduced the alkalinity removal rate. This result shows that the bloom of the nitrification process consumed more alkalinity from the sewage. Similar results were reported by S. Gao et al., Q. Su et al., 26,27 .
Fur thermore, a relationship can be observed between the removal rate of TN and the PO 4 3--P. During the start-up period, the removal rate of TN increased from 48.81% to 63.86%, while the removal rate of PO 4 3--P decreased from 66.67% to 43.75% on days 7 and 21, respectively. Likewise, COD and NH 4 + -N removal rates, TN and PO 4 3--P removal rates were also found in the increasing trend after the development of granules. At the end of the study, the removal rates of TN and PO 4 3--P were observed to be 80.17% and 54.85%, respectively. The statistical analysis (ANOVA) test confirmed that the TN and PO   Figure 11 shows the removal percentages of COD, alkalinity, NH 4 + -N, TN, and PO 4 3--P at different times (days) over the experimental studies. Initially, the removal rates of COD decreased from 82.39% on day 7 to 66.17% on day 14, whereas the removal rates of TN increased from 48.81% to 52.41% on days 7 and 14, respectively. This decrease in COD removal profile was due to the discharge of biomass with poor settleability, which could not resist the hydraulic shear forces in the start-up period 24,25 . In the COD and TN removal rate comparison, results reveal that the removal rates of both parameters increased after granules started growing and remained in the same trend up to the end of the study. The overall removal rates of COD and TN all increased with the passage of time.

Comparison of removal efficiency of different parameters
Similarly, the removal efficiency of NH 4 + -N and alkalinity showed a strong relationship. On day 14, the removal percentages of NH 4 + -N and alkalinity were 56% and 30%, respectively, and they exhibited a continuous increasing trend. On day 42, NH 4 + -N removal reached 100%, and alkalinity removal was found at 46.92%. After that, the NH 4 + -N removal rate remained constant at 100%, but the alkalinity removal rate declined to 37.44% on day 56. In the meantime, the alkalinity influent concentration was

Analysis of variance (ANOVA) test
The ANOVA is a statistical method for comparing the total amount of variation in a set of data.  (Table 1). Likewise, the ANOVA test of COD reduction and PO 4 3--P removal in A/O/A reactor treatment performance is statistically non-significant with p-values>0.05. However, in the case of MLSS-MLVSS dynamics over the experimental study, it was determined to be non-significant with p-values>0.05 (Table 1). Furthermore, the TN-PO 4 3--P assessment of A/O/A reactor treatment performance was found to be highly significant with p-values<0.05. A similar study was conducted on the aerobic and anaerobic treatment of wastewater, and the statistical analysis results were comparable to the findings 28 .