The effect of chemical oxygen demand (COD)/N ratio and sludge retention time (SRT) on the simultaneous nitrification and denitrification (SND) process in aerated moving-bed sequencing batch reactor (A-MBSBR), for treating synthetic municipal wastewater, was studied. Three lab-scale reactors with SRTs of 5, 10, and 15 days were operated under COD/N ratios of 10 (phase1) and 20 (phase2). The high correlation coefficients between nitrification and denitrification efficiencies in both phases demonstrated that the denitrification efficiency in A-MBSBRs was strongly affected by nitrification efficiency (nitrate concentration). The high COD/N ratio of 20 in phase2 led to weaker nitrification (29.6–39.1%) and consequently weaker denitrification (15.8–27.9%) under all SRTs, compared with nitrification and denitrification efficiencies (44.4–62.7% and 26.3–42.8%) in phase1 with COD/N ratio of 10. This was most probably due to a better balance between the population of heterotrophs and autotrophs in phase1. In phase2, increasing SRT from 5 to 15 days was beneficial for both nitrification and denitrification reactions. Therefore, the optimal SND efficiency in phase2 was expected to be achieved under SRTs higher than 15 days. In phase1, the highest and the lowest nitrification and denitrification efficiencies were obtained at SRTs of 10 and 5 days, respectively.

  • Study of interaction effect of chemical oxygen demand (COD)/N ratio (10:20) and sludge retention time (SRT) (5, 10, and 15 days) on simultaneous nitrification and denitrification (SND) in aerated moving-bed sequencing batch reactor.

  • Under both COD/N ratios, denitrification was mainly controlled by nitrification.

  • COD/N of 10 yielded higher nitrification and denitrification values than COD/N of 20.

  • Under COD/N of 10, SRT of around 10 days yielded the highest SND efficiency.

  • Under COD/N of 20, optimal SND was expected to occur at SRT of higher than 15 days.

The municipal and many industrial wastewaters contain a variety of nitrogen compounds, such as organic nitrogen and ammonia. The adverse impacts associated with increasing amounts of nitrogen compounds in water bodies are leading to severe environmental problems and health effects (Liu et al. 2020; Sasani & Ghasemi 2021). As the most important ones, promotion of eutrophication, toxicity to aquatic organisms, and depletion of dissolved oxygen (DO) in receiving water bodies because of bacterial oxidation of ammonia to nitrate can be noted (Rahimi et al. 2011; Song et al. 2019; Zhang et al. 2019). Consequently, considerable attention has been attracted to remove nitrogen from domestic and industrial wastewaters to meet the required wastewater discharge standards (Liu et al. 2020).

Unlike traditional nitrogen removal processes such as anoxic-aerobic process (A/O), which are too costly and require multiple units and large constructions (Kim et al. 2015; Liu et al. 2020), simultaneous nitrification and denitrification (SND) can be considered as a well-established method with lower operating costs (Seifi & Fazaelipoor 2012). In this process, due to gradient of DO concentration, anoxic denitrifiers and aerobic nitrifiers can simultaneously exist in the inner and the outer parts of biofilm or granular sludge, respectively (Chen et al. 2020; Layer et al. 2020). In aerobic zones, ammonia is converted to nitrite and nitrate via ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB), respectively. Then, the nitrate is reduced to nitrogen (N2) by heterotrophic denitrifying bacteria in anoxic zones in the same reactor (Zhao et al. 2017).

SND has been applied and investigated in variety of treatment systems such as submerged biofilm membrane bioreactor (Han et al. 2018), aerobic granular sequencing batch reactor (SBR) (He et al. 2018), sequencing batch biofilm reactor (Wang et al. 2014), moving-bed sequencing batch reactor (MBSBR) (Cao et al. 2017), moving-bed biofilm reactor (Shitu et al. 2020), and SBR (Goh et al. 2009). Among these systems, MBSBR, which incorporates both attached-growth and suspended-growth biomass processes, can be considered a suitable SND reactor (Lim et al. 2012; Sun 2014; Maslon & Tomaszek 2015). The system benefits from the advantages of both attached growth and suspended growth and as well is profited by the characteristics of SBR system including operation flexibility, stable performance, no sludge return, strong resistance to possibly harmful compounds, high tolerance to shock loads, and less space requirement (Sun 2014; Faridnasr et al. 2016). Jing et al. (2009), who studied removal of COD (chemical oxygen demand) from coking-plant wastewater, explained that MBSBR can be considered as a suitable alternative for SBR. As they described, the attached biofilm contributed about 60% to the total COD removal and showed higher activity than the suspended sludge (Jing et al. 2009).

One of the most important contributing factors to the nitrogen removal efficiency in biological treatment systems is the influent COD/N ratio. The COD/N ratio directly governs the distribution of the population of autotrophic and heterotrophic bacteria in a nitrogen removal system (Santos et al. 2016; Pelaz et al. 2018). Although a low COD/N ratio is favorable for the growth and enrichment of nitrifier bacteria and improves the nitrification, it may be an obstacle for the denitrification stage (Meng et al. 2008; Sun et al. 2010; Yang et al. 2014). As reported by Meng et al. (2008), who investigated the SND process in an airlift internal circulation membrane bioreactor (AIC-MBR), among three COD/N ratios of about 5, 10, and 15, the highest total nitrogen (TN) removal efficiency was obtained under COD/N ratio of 10. In another case, in a modified membrane bioreactor (MBR), nitrogen was optimally removed under COD/N ratio of 9.3 (Fu et al. 2009). Maslon & Tomaszek (2015) used a lab-scale MBSBR system with volumetric organic loading rates between 0.84 and 0.978 g COD L−1 d−1. They explained that the amounts of nitrogen removal positively correlated with the influent COD/N ratios, and the maximum nitrogen removal efficiencies were achieved at COD/N ratios of 10–11. Generally, based on several studies, COD/N values of around 6–11 could allow a proper nitrogen removal (Beylier et al. 2011). However, the amount of COD/N ratio for obtaining the highest nitrogen removal efficiency depends on operating parameters such as sludge retention time (SRT).

SRT is a very important operating parameter, which indicates the mean residence time of microorganisms in the reactor and is commonly used for design of wastewater treatment plants (WWTPs). Only organisms that are able to reproduce themselves during this time can be detained and enriched in the system (Clara et al. 2005). Therefore, SRT is another effective factor on nitrogen removal reactions (Wu et al. 2011). Since AOB and NOB have different growth rates (Blackburne et al. 2008), SRT may be used to change the microbial population distribution in biological nitrogen removal systems (Wu et al. 2011). Autotrophic microorganisms have a lower growth rate and require longer retention time compared with aerobic heterotrophs. Hence, more sludge age could be favorable for nitrification (Li & Wu 2014). Choi et al. (2008) explained that for SRTs of 20, 25, 30, and 40 days in an intermittently aerated MBR, there was little difference in the removal of COD and TN. Simsek et al. (2016) examined nitrogen removal efficiency under SRTs from 0.3 to 13 days. As they explained, nitrification was limited at SRTs of 0.3 and 0.7 days, while full nitrification occurred at SRTs of 2 days and longer. In another recent study, Li & Stenstrom (2018) showed that with increasing sludge age from 4 to 13.3 days in a lab-scale modified Ludzak–Ettinger reactor, the ammonium removal increased from less than 50% to about 80%.

In WWTPs, the influent COD/N ratio depends on wastewater characteristics, and its control and adjustment are costly. For this reason, COD/N ratio is usually considered as one of the input variables for design of WWTPs. Conversely, SRT is an important design variable and can be selected based on the treatment aims. Therefore, when encountering different COD/N ratios, choosing appropriate SRT can help achieve higher nitrogen removal efficiencies. Although discussion of the interaction effect of SRT and COD/N ratio has been reported for systems such as full-scale step-feed reactor (SF) (Phanwilai et al. 2020) and SBR (Najartabar Bisheh et al. 2021; Sharma & Bhatti 2022), there is a significant lack of research addressing this interaction in the context of nitrogen removal through SND process, particularly in aerated moving-bed sequencing batch reactor (A-MBSBR) systems. Accordingly, the main aim of the present study was to investigate the interaction effect of SRT and COD/N ratio on the SND process in A-MBSBR and to determine the SND pattern in this system, as well.

Experimental setup and operating conditions

The experiments were performed using three A-MBSBRs (reactors R1, R2, and R3) made from plexiglass with working volume of 12.9 L (total volume of 15.26 L), inner diameter of 18 cm, and height of 60 cm. Kaldnes-3 media was used as the biofilm carrier. About 40% of the apparent working volume of each reactor was filled with the media. Each reactor was aerated using an air pump (HAILEA, ACO-5505, China) with the capacity of 5 L/min and a ring shape diffuser at the bottom of the reactor. The aeration flowrate was kept constant at about 4 L/min in three phases of the work. The schematic of the used A-MBSBRs and the images of media after the growth of the biofilm are shown in Figure 1. SRTs lower than 15 days were tested in the present study. Accordingly, the reactors R1, R2, and R3 with the sludge ages of 5, 10, and 15 days, respectively, were operated in two phases with the applied COD/N ratios of 10 for phase 1 and 20 for phase 2. The adaptation time before beginning of each phase was more than 3 weeks.
Figure 1

Schematic of the used A-MBSBRs and images of media after the growth of biofilm.

Figure 1

Schematic of the used A-MBSBRs and images of media after the growth of biofilm.

Close modal

The reactors were inoculated with biological sludge taken from the sludge return line of Sahebgharanieh WWTP, a local municipal wastewater treatment plant in Tehran, Iran. The operating time schedule of each reactor consisted of 0.5 h feeding, 8 h reaction (aeration), 1.5 h settling, 0.25 h decanting, and 1.75 h idling. Therefore, the reactors operated under two treatment cycles per day with a total time of 12 h per cycle. The SRT adjustment was performed only based on suspended sludge concentration. Accordingly, at the end of each reaction step, a certain volume of mixed liquor was discharged manually under complete mixing. The discharge volume per cycle for the reactors R1, R2, and R3 were about 1.25, 0.65, and 0.43 L, respectively.

Then the reactors were allowed to settle. A volume exchange ratio (VER) of 0.5 was implemented in both phases. Therefore, after settling, the clarified supernatant was withdrawn up to 50% of the rectors' volume using valves mounted on the wall of the reactors. At the beginning of each new cycle, the reactors were fed and filled with the prepared synthetic wastewater.

The synthetic wastewater was prepared using tap water and composed of glucose, sucrose, and sodium acetate as the sources of organic matter, ammonium sulfate as the source of nitrogen, and potassium phosphate as the source of phosphorus. All materials were of industrial grade. The synthetic municipal wastewater characteristics are presented in Table 1. As seen in the table, to double the COD/N ratio in phase 2, the nitrogen concentration in the feed was kept constant, while the COD concentration was increased from 600 mg L−1 in phase 1 to 1,200 mg L−1 in phase 2.

Table 1

Constituents of the used synthetic wastewater

CompoundsConcentration (mg L−1)
Phase 1Phase 2
Sodium acetate (CH3COONa) 200 400 
Glucose (C6H12O6200 400 
Sucrose (C12H22O11200 400 
Ammonium sulfate ((NH4)2SO4283 283 
Potassium phosphate (KH2PO426 26 
Ammonium nitrogen (NH4-N) 60 60 
Total COD 600 1,200 
COD/N 10 20 
CompoundsConcentration (mg L−1)
Phase 1Phase 2
Sodium acetate (CH3COONa) 200 400 
Glucose (C6H12O6200 400 
Sucrose (C12H22O11200 400 
Ammonium sulfate ((NH4)2SO4283 283 
Potassium phosphate (KH2PO426 26 
Ammonium nitrogen (NH4-N) 60 60 
Total COD 600 1,200 
COD/N 10 20 

Analytical methods

A Lutron DO-5510 oxygen-meter was used for measuring the DO concentration. The pH was measured by means of a CyberScane PC300 pH-meter. The concentrations of COD, (ammonium nitrogen), (nitrite nitrogen), (nitrate nitrogen), and TN were measured using HACH test kits and HACH spectrophotometer DR/4000 (USA). For evaluating the N-removal efficiency, the study considered the end of the reaction period, which lasted for 8 h of aeration. At this point, the concentrations of ammonium, nitrate, and nitrite in the effluent as well as the organic nitrogen content of the waste sludge were carefully measured for both phases. Measurement of mixed liquor suspended solids (MLSS) and mixed liquor volatile suspended solids (MLVSS) were performed according to the procedures presented in standard methods (APHA 1998). The mass of biofilm (mg media−1) was calculated from the difference between the mass of the dried media with the attached biofilm and the mass of the dried media after the biofilm was removed from its surface.

Formulas and calculations

Statistical measurements

The values of the standard error (error bars in the figures) are calculated using Equation (1) (Altman et al. 2013):
formula
(1)
where SE represents the standard error, s is the standard deviation, and n is the number of observations.

COD removal efficiency

The COD removal efficiency was calculated using the following equation:
formula
(2)
where RCOD is the COD removal efficiency, C0 is the COD concentration in the influent, and C1 is the COD concentration in the effluent.

Organic nitrogen content

To estimate the nitrogen content of the biomass (organic nitrogen), mixed liquor samples were taken from the reactors at the end of the reaction time. Then the TN concentration of the mixed liquor () as well as the concentrations of the ammonium nitrogen (), nitrate nitrogen (), and nitrite nitrogen () in the clarified effluent were measured in the laboratory. The samples were taken at the end of the reaction time. , , and were measured using the samples' filtrate passed through filter paper, while was measured using samples without filtering. The following equations were then used to calculate the organic nitrogen concentration () and the organic nitrogen content () of the waste sludge:
formula
(3)
formula
(4)
where is the volume of wasted sludge per cycle, which depended on the SRT values of each reactor. Accordingly, was equal to 1.25, 0.65, and 0.425 L cycle−1 for reactors 1, 2, and 3, respectively.

Nitrification efficiency

The nitrification efficiency (), i.e., the conversion rate of ammonium to nitrite and nitrate, was calculated using the following equations:
formula
(5)
formula
(6)
formula
(7)
where is the mass of the ammonium injected into the reactors at the beginning of each cycle with a constant ammonium concentration () of 60 mg N L−1. is the ammonium content of the clarified effluent and Q is the influent/effluent flow per cycle, equivalent to 50% of the working volume of the reactor ( = 6.45 L cycle−1). It should be noted that the TN content of the influent () was solely related to the input ammonium ().

Mass balance equations

For the TN mass balance of the whole treatment system, the mass of the TN entering the system () must be equal to the mass of the TN leaving the system () through effluent (), denitrification (), and wasted sludge (). Accordingly, the following equation can be written for the TN mass balance of the whole treatment system from which and denitrification efficiency () can be calculated:
formula
(8)
formula
(9)
formula
(10)
where and are determined based on Equations (4) and (6), respectively, and is calculated using the following equation:
formula
(11)

Total nitrogen removal

The TN removal efficiency (nitrogen removal from water) is the sum of the nitrogen removed through denitrification and the nitrogen removed through wasted sludge. Therefore, it is calculated using Equation (12):
formula
(12)

MLSS, MLVSS, and biofilm mass

As summarized in Table 2, in both phases, with increasing the sludge age from 5 to 15 days, the average MLSS and MLVSS concentrations increased considerably. According to the table, the biomass concentration was directly related to the SRT values. With increasing the SRT from 5 to 15 days, the average MLSS concentration increased from 835 to 1,858 mg L−1 in phase 1 and from 1,142 to 2,920 mg L−1 in phase 2. Besides, shifting the COD/N ratio from 10 in phase 1 to 20 in phase 2 increased the MLSS concentration in all reactors. However, the MLSS concentration in reactor 2 (SRT of 10 days) was increased to a lesser extent in comparison with the reactors 1 and 3 (SRTs of 5 and 15 days).

Table 2

Average values of MLSS and MLVSS concentration in A-MBSBRs

Phase 1
Phase 2
ParameterR1R2R3R1R2R3
MLSS (mg L−1835 ± 26 1,496 ± 83 1,858 ± 125 1,142 ± 108 1,587 ± 78 2,920 ± 142 
MLVSS (mg L−1778 ± 31 1,320 ± 71 1,660 ± 110 1,105 ± 105 1,491 ± 84 2,694 ± 130 
Phase 1
Phase 2
ParameterR1R2R3R1R2R3
MLSS (mg L−1835 ± 26 1,496 ± 83 1,858 ± 125 1,142 ± 108 1,587 ± 78 2,920 ± 142 
MLVSS (mg L−1778 ± 31 1,320 ± 71 1,660 ± 110 1,105 ± 105 1,491 ± 84 2,694 ± 130 

Figure 2(a) shows the average biofilm mass of the reactors. Generally, with doubling the COD/N ratio in phase 2, the biofilm mass increased significantly in all three reactors. It can also be found from the figure that in phase 1 the SRT change from 5 to 15 days did not have much effect on the biofilm mass (13.9–15.6 mg per media). However, in phase 2, the reactors showed different biofilm masses so that the highest and the lowest values belonged to reactors 1 and 2 with 39.75 and 25.55 mg per media, respectively.
Figure 2

The average values of biofilm mass per media (a), COD concentration inside the reactors and in the reactors' effluent (b) and DO concentrations in phase 1 (c) and phase 2 (d).

Figure 2

The average values of biofilm mass per media (a), COD concentration inside the reactors and in the reactors' effluent (b) and DO concentrations in phase 1 (c) and phase 2 (d).

Close modal

DO concentration

As depicted in Figure 2(c) and 2(d), the mean DO concentration in the early minutes of the reaction time (about first 30 and 120 min for phases 1 and 2, respectively) was in the range of 0.8–2.3 mg L−1 in phase 1 and 0.5–1.7 mg L−1 in phase 2. The DO profile vs. time is presented in Figure 3. As seen in the figure, the oxygen concentration in the reactors reached relative stability after about 60 min in phase 1 and 180 min in phase 2. The mean DO concentration during the stability period was in the ranges of 2.5–3.0 and 2.9–4.1 mg L−1 in phases 1 and 2, respectively, which were higher than at the beginning of the reaction time.
Figure 3

DO profile vs. time in MBSBRs: (a) phase 1 and (b) phase 2.

Figure 3

DO profile vs. time in MBSBRs: (a) phase 1 and (b) phase 2.

Close modal

It should also be noted that the DO concentration in the early minutes of the reaction time was higher in phase 1 compared with phase 2, while during the rest of the cycle, higher DO concentrations were observed in phase 2 compared with phase 1.

COD removal

The average amounts of COD removed in different time intervals of the reaction step as well as the COD concentration of the reactors' effluents are shown in Figure 2(b). Due to the VER of 0.5, the influent COD concentration was reduced by half inside the reactors at the beginning of each cycle (Figure 2(b)). As seen in the figure, in phase 1, the COD removal efficiency of the A-MBSBRs was not influenced much by the variation of SRT within the test range. About 265–270 mg L−1 (88–90%) of the initial COD content of the reactors was instantly disappeared after feeding the reactors with new wastewater, while only 2–6 mg L−1 COD reduction was occurred over the rest of the reaction time (450 min). The effluent COD concentrations of the reactors were in the range of 27–31 mg L−1 in phase 1. However, A-MBSBRs performed differently in phase 2. Based on Figure 2(b), in phase 2, 140–227 mg L−1 (23–38%) of the initial COD content of the reactors was removed in the early minutes of the cycle and 270–335 mg L−1 (45–56%) was gradually decreased during the rest of the reaction time. In phase 2, the effluent COD concentrations of the reactors were about 96, 190, and 38 mg L−1 for A-MBSBRs 1, 2, and 3, respectively. The effluent COD concentration of reactor 2 was clearly higher than that of the two other reactors.

Nitrification

The average concentration of ammonium, nitrite, and nitrate in the effluent of the reactors is presented in Figure 4(a) and 4(b). According to the figures, the effluent ammonium concentrations were considerably higher in phase 2 compared with phase 1. To calculate the exact amounts of the nitrification, the organic nitrogen content of the wasted sludge (Figure 4(c)) and the effluent ammonium nitrogen were subtracted from the total influent ammonium nitrogen (Equations (5)–(7)). The results are depicted in Figure 4(d). As seen, both SRT and COD/N ratio remarkably affected the nitrification efficiency in A-MBSBRs. In phase 1, as the SRT increased from 5 to 10 and 15 days, the nitrification efficiency first increased from 44.35 to 62.71% and then decreased slightly to 56.03%. In phase 2, much lower nitrification efficiencies were obtained in all reactors compared with phase 1. The average nitrification efficiencies in this phase for reactors 1–3 were about 29.57, 32.46, and 39.12%, respectively.
Figure 4

The effluent concentration of ammonium, nitrate, and nitrite in phase 1 (a) and phase 2 (b), the organic nitrogen content of waste sludge (c), and the average nitrification efficiency (d).

Figure 4

The effluent concentration of ammonium, nitrate, and nitrite in phase 1 (a) and phase 2 (b), the organic nitrogen content of waste sludge (c), and the average nitrification efficiency (d).

Close modal

Denitrification and total nitrogen removal

The average denitrification efficiencies are calculated using Equations (8)–(10). The results are shown in Figure 5(a). According to the figure, the COD/N ratio had a significant effect on the denitrification performance of the A-MBSBRs so that much higher denitrification efficiencies were obtained in phase 1 than in phase 2. In addition, the TN removal efficiencies () were calculated using Equation (12), and the results are presented in Figure 6, in which the values of the nitrogen removed through denitrification and the nitrogen removed through wasted sludge are shown separately. According to the figure, the TN removal efficiencies of the reactors in phase 1 were higher than those in phase 2, mainly due to the higher denitrification efficiencies in phase 1.
Figure 5

The average denitrification efficiency under three SRTs (a) and the correlation of denitrification efficiencies with nitrification and COD removal efficiencies (b) in the two phases.

Figure 5

The average denitrification efficiency under three SRTs (a) and the correlation of denitrification efficiencies with nitrification and COD removal efficiencies (b) in the two phases.

Close modal
Figure 6

The average values of the total nitrogen removal in the MBSBRs.

Figure 6

The average values of the total nitrogen removal in the MBSBRs.

Close modal

General performance of A-MBSBRs

Biomass

The observed increase of the MLSS concentration with the SRT in both phases can be explained using the MLSS mass balance equations, which were well represented by Metcalf and Eddy (2003). Kargi & Uygur (2002) investigated the nutrient removal in a SBR reactor and reported that the biomass concentration increased with increasing SRT up to 20 days. Besides, an increase of the MLSS concentration with doubling the influent COD concentration in phase 2 was simply due to the augmented growth rate of heterotrophic microorganisms under much more available organic matter in phase 2. The observed increase in biofilm mass in phase 2 occurred due to two essential reasons: (i) the increase in the organic matter available for the heterotrophic bacteria of the biofilm and (ii) the tendency of the suspended bacteria to settle on and adhere to the biofilm to stay longer in the system and completely consume the excess inlet COD. This phenomenon was more obvious in reactor 1 (SRT: 5 days), where, due to more MLSS wasted daily, bacteria intelligently increased their retention time in the system by joining biofilm layers so that they could consume more organic matter.

Due to the fact that the autotrophic bacteria do not need organic matter, the increase in the biofilm mass in phase 2 was mostly related to the heterotrophic bacteria. This seems to be beneficial for improving the anoxic zones in biofilm layer and subsequently for enhancing the denitrification. Liu et al. (2010), who investigated the effect of the COD/N ratio on microbial community structure in membrane aerated biofilm reactor, reported that with increasing this ratio, the population of heterotrophic bacteria in biofilm increased compared to the population of autotrophic bacteria.

DO concentration

Due to the abundance of the organic matter at the beginning of the reaction step, which needed a high oxygen uptake by heterotrophic bacteria to be degraded, the DO concentration at the early minutes of the reaction step was lower than in the rest of it, in both phases. By consuming a portion of the organic matter early in the cycle, both the activity of the heterotrophic bacteria and the concentration of impurities (organic matter) decreased, resulting in an increase in the DO concentration for the rest of the reaction step. Moreover, as the oxygen uptake rate of heterotrophic bacteria and the concentration of impurities (MLSS and organic matter) were higher in phase 2 than in phase 1, the DO concentration at the early minutes of the reaction step was lower in phase 2. In contrast, the higher DO concentration during the rest of the reaction step in phase 2 could be attributed to the lower nitrification rate in this phase compared with phase 1 (refer to Section 5.2). Meng et al. (2008) who used an AIC-MBR system for the SND process explained that very high or very low DO concentrations suppressed the process. They found 0.75–1.0 mg L−1 as the optimum range of the DO concentration for the SND process in their system. The optimum DO range reported by Meng et al. (2008) is very narrow and much lower than the DO concentrations in the present study. In fact, owing to the presence of the anoxic region in interior parts of biofilm in the A-MBSBRs, controlling and adjusting the DO concentration was not necessary in our work.

COD removal

Although in phase 1 the COD removal efficiency was more or less the same in the three reactors, with doubling the input organic load in phase 2, significant differences were observed between the reactors' COD removal performances. The highest COD removal efficiency at the sludge age of 15 days in phase 2 was most probably due to the highest MLSS concentration and the existence of more appropriate enzymes for degrading all types of organic matters in the feed. In reactor 1 with the SRT of 5 days, higher metabolic rate of young cells and more biofilm mass were likely effective on higher COD removal efficiency in this reactor compared with reactor 2. It is noteworthy that the mentioned differences were not observed in phase 1, but the twofold increase of the COD load in phase 2 made them appeared. Generally, it can be said that an increase in the organic load in reactors 1 and 3 led to an increase in MLSS concentration and biofilm mass, while in reactor 2, it caused an increase in the biofilm mass and the effluent COD concentration.

In phase 1, the COD concentration decreased rapidly at the early minutes of the cycle coincided with the drop in DO concentration in this phase. The rapid decrease of the COD concentration, as described by Zuriaga-Agusti et al. (2010) who reported similar observations in their SBR system, was likely due to the high-speed adsorption of organic matters by the heterotrophic bacteria after a long starvation period during the previous cycle. The abundance of the organic matter in phase 2 led to a shorter starvation period at the end of the cycle in this phase and alleviated the intense adsorption of the inlet COD by the heterotrophs.

As described by Phanwilai et al. (2020), insufficient carbon sources at the full-scale step-feed municipal WWTP with a COD/N ratio of 4 adversely affected nitrogen removal efficiency. Therefore, in the present work, the gradual uptake of the organic matter in phase 2 (COD/N ratio of 20) and remaining part of it until the end of the cycle seem to be beneficial for denitrification process in this phase. This is discussed more in Section 5.2.

SND process in A-MBSBRs

Nitrification

In phase 1, the nitrification efficiency of the three A-MBSBRs was in the range of 44.35–62.71%. The lowest efficiency in this phase was obtained for reactor 1 (Figure 4(d)). This was probably because the autotrophic bacteria, due to their low growth rate (Metcalf & Eddy 2003), could not grow much under a very low sludge age of 5 days in this reactor. It was also observed that with increasing the sludge age from 10 to 15 days, despite the conditions becoming more suitable for increasing the population of autotrophs, the nitrification efficiency slightly decreased. Lower nitrification in reactor 3 was probably due to the relatively lower DO concentration in this reactor compared to reactor 2, particularly over the first 30 min of the cycle. According to a research performed by Huang et al. (2001) in comparison to the process of organic matter degradation, the nitrification process was much more dependent on the presence of oxygen. Therefore, even small changes in DO concentration could affect the rate of nitrification (Huang et al. 2001). Also, Wang et al. (2020) explained that the DO concentrations of less than 1.5 mg L−1 considerably limited the nitrification process in their MBSBR system. Accordingly, it can be said that the DO concentration below 1 mg L−1 in the early minutes of the cycle in reactor 3 has limited the growth and activity of autotrophic bacteria in this reactor.

In phase 2, very lower nitrification efficiencies between 29.57 and 39.12% were obtained for the reactors, compared with phase 1. Doubling the concentration of organic matter in phase 2 provided suitable conditions for the increase and dominance of the population of heterotrophic bacteria, which have a higher growth rate than the autotrophs (Metcalf & Eddy 2003). This is in line with those reported by Gong et al. (2012) and Liu et al. (2004) who demonstrated the enrichment of autotrophic nitrifiers over heterotrophic bacteria under low COD/N ratios (below 10). Accordingly, in phase 2, a significant portion of the DO was consumed by the heterotrophs for degrading the excess inlet COD. The severe oxygen uptake by the heterotrophs during the early minutes of the cycle in phase 2 (Figure 2(d)) suppressed the autotrophic bacteria and prevented them from growth and continuing nitrification. Overall, the inability of the autotrophic bacteria to compete with the high population of heterotrophs for oxygen uptake caused poor nitrification in all reactors in phase 2.

Another important point is that in phase 2, with an increase of the sludge age, the nitrification efficiency increased steadily (Figure 4(d)). This observation is consistent with the increase of the autotrophs population with the sludge age and suggests that under higher COD/N ratio of 20, the role of the authotrophs' population in the nitrification process became more important and obvious than the role of the DO concentration.

Higher nitrification rate in phase 1 caused higher nitrite and nitrate contents in this phase compared with phase 2, under all applied SRTs. The nitrite concentration in phase 1 was between 0.5 and 1.25 mg N L−1, while in phase 2, it reduced to below 0.1 mg N L−1 due to lower nitrification efficiency. In a similar pattern, all reactors in phase 2 showed a lower nitrate concentration compared with phase 1. In a recent study conducted by Xu et al. (2019), the accumulated nitrite and nitrate concentrations as well as the ammonia oxidation efficiency in lab-scale cyclic activated sludge system under COD/N ratios between 4 and 10 did not changed much. Comparing the results of the present study with the results reported by Xu et al. (2019), on the one hand, it can show the importance of the investigated range of the COD/N ratio, and on the other hand, it can show the sensitivity of the biofilm treatment system to changes in the COD/N ratio, which was not observed in a conventional system.

Denitrification and total nitrogen removal based on mass balance

Table 3 shows the mass balance results of the TN flux for the whole treatment system through which, and were calculated. The results are also presented in Figures 5(a) and 6. As seen in the figures, the denitrification efficiency was in the ranges of 26.32–42.81% and 15.79–27.94% for phases 1 and 2, respectively. Accordingly, in phase 2 with the doubled COD/N ratio of 20, despite the increase in the biofilm mass and the consequent increase in its anoxic zones, as well as the fact that part of the organic matter entering the system remained until the end of the cycle (factors that make the conditions more suitable for denitrification), the denitrification efficiency and the TN removal decreased considerably in all reactors, compared with phase 1.

Table 3

The mass balance results of the total nitrogen flux for the whole treatment system

PhaseReactorParameters
Measured
Calculated
Phase 1 A-MBSBR1 107.00 18.30 9.57 1.25 77.88 97.36 29.12 187.82 387 101.82 26.31 
A-MBSBR2 144.50 10.03 11.07 0.87 122.53 79.65 21.97 141.70 387 165.65 42.80 
A-MBSBR3 170.83 16.87 9.00 0.64 144.32 61.34 26.51 170.99 387 154.67 39.97 
Phase 2 A-MBSBR1 113.15 26.75 6.35 0.05 80.00 99.99 33.15 213.82 387 73.19 18.91 
A-MBSBR2 144.19 30.04 10.04 0.06 104.05 67.63 40.14 258.90 387 60.47 15.63 
A-MBSBR3 187.30 26.37 6.62 0.08 154.22 65.55 33.07 213.30 387 108.15 27.95 
Unit mg NL−1 mg NL−1 mg NL−1 mg NL−1 mg NL−1 mg Ncycle−1 mg Ncycle−1 mg Ncycle−1 mg Ncycle−1 mg Ncycle−1 
PhaseReactorParameters
Measured
Calculated
Phase 1 A-MBSBR1 107.00 18.30 9.57 1.25 77.88 97.36 29.12 187.82 387 101.82 26.31 
A-MBSBR2 144.50 10.03 11.07 0.87 122.53 79.65 21.97 141.70 387 165.65 42.80 
A-MBSBR3 170.83 16.87 9.00 0.64 144.32 61.34 26.51 170.99 387 154.67 39.97 
Phase 2 A-MBSBR1 113.15 26.75 6.35 0.05 80.00 99.99 33.15 213.82 387 73.19 18.91 
A-MBSBR2 144.19 30.04 10.04 0.06 104.05 67.63 40.14 258.90 387 60.47 15.63 
A-MBSBR3 187.30 26.37 6.62 0.08 154.22 65.55 33.07 213.30 387 108.15 27.95 
Unit mg NL−1 mg NL−1 mg NL−1 mg NL−1 mg NL−1 mg Ncycle−1 mg Ncycle−1 mg Ncycle−1 mg Ncycle−1 mg Ncycle−1 

In a recent study performed by Kim et al. (2021), the authors declared that with an increase in COD/N ration from 5 to 20 in an aerobic granular sludge SBR, the TN removal efficiency increased steadily. They pointed out that the enough supply of organic matter needed for denitrification reaction and the domination of aerobic denitrifiers improved SND by aerobic granules in their system. However, Kim et al. (2021) did not analyze and discuss nitrification and denitrification processes, separately. In other words, they did not clarify whether the nitrification process is affected when the denitrifiers predominate in the granules. In general, it can be concluded that the effect of increasing the COD/N ratio on the SND process is different in aerobic granular sludge SBR and A-MBSBR.

Figure 5(b) shows the correlation between nitrification and denitrification efficiencies of A-MBSBRs for the two applied COD/N ratios. The high correlation coefficients confirm that in both phases, the denitrification efficiency was strongly controlled by the nitrification efficiency or in other words, by the nitrate concentration.

According to the aforementioned explanations, it can be stated that although doubling the input COD in phase 2 first provided the organic matter needed by heterotrophs for denitrification until the end of the reaction time, second, increased heterotrophic growth within the biofilm, and third, strengthened the anoxic zones for denitrification, the nitrate deficiency in the medium led to a weak denitrification in this phase. A similar trend was reported by Meng et al. (2008) who evaluated the SND process in an AIC-MBR system. They explained that the SND efficiency was higher than 70% under COD/N ratios of 4.77 and 10.04, while it decreased to less than 50% as the COD/N ratio shifted to 15.11 (Meng et al. 2008). Under the COD/N ratios of about 10 and 15, the efficiency of the TN removal mainly depended on the degree of the nitrification in the system (Meng et al. 2008).

It is also illustrated in Figure 5(b) that under the COD/N ratio of 10 in phase 1, no meaningful relation between the denitrification efficiency and the COD removal was observed since the COD removal was very high and almost equal in all three reactors. However, under higher COD/N ratio of 20 in phase 2, the values of the denitrification efficiency correlated well with the COD removal. The former observation indicates insufficient organic matter in phase 1 as almost all the COD content of the reactors was consumed by heterotrophs. In contrast, the latter observation reveals the importance of the COD content and its influence on denitrification efficiency, considering that the COD content should not be so high as to inhibit the growth and activity of autotrophic bacteria. Putting these two observations together, it can be inferred that there is an optimal COD/N ratio between 10 and 20, under which neither the population of autotrophic bacteria will decrease nor organic matter deficiency will occur for heterotrophs, resulting in enhancement of both nitrification and denitrification efficiencies under all SRTs in A-MBSBRs.

It is also worth mentioning that the rapid adsorption of the organic matter by the heterotrophic bacteria at the early minutes of the cycle in phase 1 did not prevent the denitrification process in this phase. This suggests that the consumption of the COD adsorbed by the heterotrophs was more or less gradual so that the denitrification process was carried out almost simultaneously with the nitrification process, over the reaction time.

SND pattern in the A-MBSBRs

According to Figure 7(a), in phase 1, the SRT leading to optimal nitrification was about 10 days. In this phase, the SRT for the optimal denitrification was also about 10 days (Figure 7(b)). Besides, the COD removal for all three reactors was more than 90% in this phase and did not change much with the SRT. As a result, it can be said that for treating conventional municipal wastewaters with a normal COD/N ratio of about 10 using the A-MBSBR system, SRT of around 10 days is likely to yield the highest SND efficiency.
Figure 7

The SND pattern in the A-MBSBRs under the effect of SRT and COD/N ratio: nitrification (a) and denitrification (b).

Figure 7

The SND pattern in the A-MBSBRs under the effect of SRT and COD/N ratio: nitrification (a) and denitrification (b).

Close modal

In phase 2, increasing the SRT in the range of 5–15 days was beneficial for both nitrification and denitrification processes (Figure 7) and for the COD removal (Figure 2(b)), as well. Therefore, it can be inferred that for wastewaters with a high COD/N ratio of about 20, the optimal efficiencies of SND and COD removal will be achieved under an SRT of higher than 15 days.

According to the obtained results, with increasing the COD/N ratio from 10 in phase 1 to 20 in phase 2, nitrification decreased significantly in all three A-MBSBR due to the providing appropriate conditions for the increase and dominance of the population of heterotrophic bacteria over autotrophic nitrifiers. Also in phase 2, lower denitrification efficiencies were observed mainly due to nitrate deficiency in the reactors. Overall, when using an A-MBSBR for treating wastewater with a COD/N ratio of about 10 (common COD/N ratio for municipal wastewater), applying sludge ages of around 10 days is likely to yield the highest SND efficiency. While under a high COD/N ratio of 20, the highest SND efficiency seems to be achieved under sludge ages of more than 15 days.

The authors gratefully acknowledge the Iran University of Science and Technology (IUST) for its financial supports and providing the research materials and equipment.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Altman
D.
,
Machin
D.
,
Bryant
T.
&
Gardner
M.
2013
Statistics with confidence: confidence intervals and statistical guidelines
,
2nd edn. BMJ Books, London, UK
.
APHA, AWWA, WPCF
1998
Standard Methods for the Examination of Water and Wastewater
, 20th edn.
American Public Health Association, American Water Work Association, Water Environment Federation
,
Washington, DC, USA
.
Beylier
M. R.
,
Balaguer
M. D.
,
Colprim
J.
,
Pellicer-Nàcher
C.
,
Ni
B. J.
,
Smets
B. F.
,
Sun
S. P.
&
Wang
R. C.
2011
Biological nitrogen removal from domestic wastewate
. In:
Comprehensive Biotechnology
, 2nd edn. Elsevier, Amsterdam, The Netherlands.
Kim
H. G.
,
Kim
J. T.
&
Ahn
D. H.
2021
Effects of carbon to nitrogen ratio on the performance and stability of aerobic granular sludge
.
Environ. Eng. Res.
26
(
1
),
190284
.
Metcalf and Eddy
2003
Wastewater Engineering: Treatment and Reuse
, 4th edn.
McGraw-Hill
,
New York, NY, USA
.
Najartabar Bisheh
F.
,
Amini
M.
,
Abyar
H.
,
Attenborough
N. K.
,
Ling
I.
,
Salamatinia
B.
,
Younesi
H.
&
Zinatizadeh
A. A.
2021
Response surface methodology approach for simultaneous carbon, nitrogen, and phosphorus removal from industrial wastewater in a sequencing batch reactor
.
Adv. Environ. Technol.
7
(
2
),
119
136
.
Pelaz
L.
,
Gómez
A.
,
Letona
A.
,
Garralón
G.
&
Fdz-Polanco
M.
2018
Nitrogen removal in domestic wastewater. Effect of nitrate recycling and COD/N ratio
.
Chemosphere
212
,
8
14
.
Rahimi
Y.
,
Torabian
A.
,
Mehrdadi
N.
&
Shahmoradi
B.
2011
Simultaneous nitrification–denitrification and phosphorus removal in a fixed bed sequencing batch reactor (FBSBR)
.
J. Hazard. Mater.
185
(
2–3
),
852
857
.
Sasani
H.
&
Ghasemi
M.
2021
A comparative study on the use of membrane bioreactor (MBR) and activated sludge followed by ultrafiltration (CAS/UF) processes for advanced treatment of industrial wastewater
.
J. Hydraul. Struct.
7
(
2
),
100
115
.
Song
Z.
,
Zhang
X.
,
Ngo
H. H.
,
Guo
W.
,
Song
P.
,
Zhang
Y.
,
Wen
H.
&
Guo
J.
2019
Zeolite powder based polyurethane sponges as biocarriers in moving bed biofilm reactor for improving nitrogen removal of municipal wastewater
.
Sci. Total Environ.
651
,
1078
1086
.
Sun
Z.
2014
Attached-growth System for Nitrification at low Temperature
.
Doctoral Dissertation
,
Purdue University, West Lafayette, IN, USA
.
Sun
S. P.
,
Nàcher
C. P. I.
,
Merkey
B.
,
Zhou
Q.
,
Xia
S. Q.
,
Yang
D. H.
,
Sun
J. H.
&
Smets
B. F.
2010
Effective biological nitrogen removal treatment processes for domestic wastewaters with low C/N ratios: A review
.
Environ. Eng. Sci.
27
(
2
),
111
126
.
Wu
C. Y.
,
Peng
Y. Z.
,
Wan
C. L.
&
Wang
S. Y.
2011
Performance and microbial population variation in a plug-flow A2O process treating domestic wastewater with low C/N ratio
.
J. Chem. Technol. Biotechnol.
86
(
3
),
461
467
.
Yang
Y.
,
Lesage
G.
,
Barret
M.
,
Bernet
N.
,
Grasmick
A.
,
Hamelin
J.
&
Héran
M.
2014
New urban wastewater treatment with autotrophic membrane bioreactor at low chemical oxygen demand/N substrate ratio
.
Water Sci. Technol.
69
(
5
),
960
965
.
Zuriaga-Agusti
E.
,
Iborra-Clar
M. I.
,
Mendoza-Roca
J. A.
,
Tancredi
M.
,
Alcaina-Miranda
M. I.
&
Iborra-Clar
A.
2010
Sequencing batch reactor technology coupled with nanofiltration for textile wastewater reclamation
.
Chem. Eng. J.
161
(
1–2
),
122
128
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).