Study on Nitrogen Removal Performance of MBBR for Quasi-IV Upgrading of Wastewater Treatment Plants in Southwest China During Winter

Study on Nitrogen Removal Performance of MBBR for Quasi-IV Upgrading of a Wastewater Treatment Plant in Southwest China During Winter

The southwestern region of China is vast, with significant differences in climate and geological conditions. The wide sources of wastewater quality lead to large fluctuations in water quality. Although this region is located in the south, due to the overall high altitude and terrain, the wastewater temperature can drop below 10 ℃ during the cold winter season. Combined with the complex terrain of the region, with interwoven mountains and deep valleys, hills and plains, and large temperature differences between day and night, the daily fluctuation of wastewater temperature is also obvious. The complex influent characteristics and excessively low water temperature in winter impose higher requirements on the design and operation of wastewater treatment plants in this region. Against the background of upgrading and reconstruction of domestic wastewater treatment plants, it is even more important to select an appropriate treatment process for upgrading in southwestern plants to ensure stable compliance during the low-temperature season.

The Moving Bed Biofilm Reactor (MBBR) process, by adding suspended carriers in a fixed area, improves the enrichment of functional bacteria and enhances low-temperature resistance. It has been widely used in upgrading projects of wastewater treatment plants. Currently, the application scale of the MBBR process in China has exceeded 2.5×10⁷ m³·d⁻¹, and MBBR has been included as a "widely used and safe and reliable process" in the newly revised "Outdoor Drainage Design Standard" (GB 50014-2021). The widespread application of the process has also promoted rapid follow-up research. Research hotspots on the MBBR process mainly focus on process design analysis, laboratory mechanism research, and engineering effect tracking. Relevant conclusions mainly include MBBR enhancing biochemical efficiency, biofilm improving impact resistance and enhancing low-temperature resistance, MBBR enhancing functional bacteria enrichment, and biofilm SND enhancing nitrogen removal. For example, Huang Qing et al. found that after upgrading with the MBBR process, the impact resistance of the wastewater treatment plant was significantly improved, and the effluent could still meet the standards stably even under long-term influent BOD and TN exceedance. Liu Qiang et al. found that the MBBR process significantly enhanced the treatment capacity of the plant at low temperatures, and the effluent could stably meet the design standards at low temperatures of 9–11 ℃. Luan Zhixiang et al. found that after upgrading a northern wastewater treatment plant with the MBBR process, the aerobic MBBR zone could remove 20.12% of TN through SND of suspended carriers. Han Wenjie et al. compared five municipal wastewater treatment plants with hybrid MBBR processes in the Yangtze River Delta region under low winter temperatures and found that the addition of suspended carriers significantly increased the nitrifying bacteria biomass in the system, with the suspended carriers enriching 86.12%–95.36% of the nitrifying bacteria. However, existing research still has certain limitations. For example, a large number of studies separate MBBR design from operational performance, lacking comparative verification of design and performance for the same project. Regarding the impact resistance of MBBR, only "qualitative but not quantitative" conclusions are available, without quantitative inference of the maximum tolerance load. Regarding the SND phenomenon, it is only "measurable but not controllable," lacking direct verification of the source of nitrogen removal through designed macro experiments, nor rational operation suggestions from the perspective of engineering control.

Based on the limitations of existing research, this study takes the upgrading of a CASS wastewater treatment plant in Southwest China as an example to explain the design principles and logic of the MBBR process. Combined with the actual system performance under the most unfavorable winter conditions, the rationality of the MBBR design was analyzed. The system's treatment potential was quantitatively analyzed through changes in water quality along the biological tank and high-substrate laboratory tests. A denitrification laboratory test using aerobic biofilm was designed to directly verify the source of the SND phenomenon in the hybrid MBBR system. The influence of engineering operating parameters on the SND effect of the MBBR system was studied, and combined with data fitting, optimization operation suggestions for SND were proposed. The differences in microbial community distribution between sludge and biofilm were analyzed from the perspective of bacterial composition, and the microscopic influencing factors of SND were analyzed from the types and relative abundance of aerobic denitrifying bacteria. It is hoped that this research can provide a reference for the MBBR process design of other similar wastewater treatment plants, provide basic data for the stable operation of MBBR plants, and provide operation suggestions for the efficient performance of SND in MBBR systems.

1 Materials and Methods

1.1 Project Overview

A wastewater treatment plant in Southwest China, with a design capacity of 1×10⁵ m³·d⁻¹, the effluent was required to meet the Class A standard of the "Discharge Standard of Pollutants for Municipal Wastewater Treatment Plant" (GB 18918-2002). Due to land use restrictions during early construction, a CASS process without traditional secondary sedimentation tanks was adopted. However, during operation, problems such as insufficient nitrification at low winter temperatures, poor sludge settleability, and decreased treatment efficiency occurred. In 2019, the plant was upgraded to quasi-Class IV standards. The upgrading had to consider both the increase in influent nitrogen concentration and the improvement of effluent standards. The design influent and effluent quality before and after reconstruction are shown in Table 1. On the basis that the influent NH₃-N and TN loads increased by 66.67% and 62.16% respectively, the effluent carbon, nitrogen, phosphorus, etc., needed to be upgraded from Class A of GB 18918-2002 to Class IV of the "Environmental Quality Standards for Surface Water" (GB 3838-2002), which is equivalent to increasing the nitrification and denitrification loads by 127.27% and 94%, respectively. This project adopted the Bardenpho-MBBR process, converting the original CASS tank into a biological tank and a sedimentation tank.

 After reconstruction, for nitrogen removal, activated sludge continued to play the main role. The denitrification retention time of the original CASS tank was extended from 3.78 h to 6.35 h, and the denitrification volumetric load of the activated sludge increased from 0.140 to 0.189 kg·(m³·d)⁻¹. To enhance activated sludge denitrification, pre-anoxic and post-anoxic zones were set up with retention times of 4.23 h and 2.12 h, respectively. The post-anoxic zone used external carbon source addition to enhance denitrification to solve the problem of insufficient carbon source in the raw water. For nitrification, the nitrification retention time of the original CASS tank was reduced from 7.55 h to 4.23 h, while the nitrification load needed to be greatly increased from 0.079 to 0.275 kg·(m³·d)⁻¹. To enhance the nitrification effect, suspended carriers were added in the aerobic zone to form an aerobic MBBR zone. The carriers used were SPR-III type, with a diameter of (25±0.5) mm and a thickness of (10±1) mm. After biofilm attachment, the density was close to that of water, and the effective specific surface area was greater than 800 m²·m⁻³, meeting the industry standard "High-density polyethylene suspended carrier for water treatment" (CJ/T461-2014). The biofilm on the suspended carriers appeared dark gray after attachment (Figure 1). The MBBR zone used only perforated aeration to provide hydraulic shear, controlling the biofilm thickness at (208.82±37.57) μm, providing good conditions for mass transfer and oxygen transfer.

1.2 Determination of Nitrogen Removal in the Biological Tank

To understand the nitrogen removal pathway in each functional zone of the biological tank under low winter temperatures after reconstruction, the water quality along the biological tank was monitored daily from December 15, 2020, to December 22, 2020. During this period, the water temperature in the biological tank was 9.5–10.2 ℃. The samples included influent of the biological tank, effluent of the anaerobic zone, effluent of the anoxic zone, effluent of the MBBR zone, effluent of the post-anoxic zone, effluent of the post-aerobic zone, and effluent of the sedimentation tank. Samples were taken every 2–3 h. After collection, the samples were pretreated promptly. After sedimentation, the supernatant was collected, and three samples were mixed in equal amounts to determine NH₃-N, NO₃-N, TN, and COD, verifying the pollutant removal pathway and nitrification/denitrification load of each functional zone.

1.3 Determination of Nitrification Potential of MBBR System and Contribution Rate of Suspended Carriers

To verify the maximum tolerance load of the existing MBBR system, a nitrification laboratory test was conducted on February 27 and 28, 2021. Suspended carriers and activated sludge from the aerobic zone of the plant were collected daily to simulate the impact resistance of the MBBR system when the influent load increased in winter. The laboratory test simulated the actual influent of the MBBR zone along the flow path. The supernatant of the anoxic zone effluent after sedimentation was used, and NH₃-N and alkalinity were supplemented by adding ammonium chloride and sodium bicarbonate as the influent for the test. The nitrification test device is shown in Figure 2. Pure-biofilm, pure-sludge, and hybrid systems were set up in sequence. The sludge concentration and carrier filling ratio were consistent with the actual conditions of the biological tank. To fully utilize the nitrification potential of the MBBR system and enhance mass transfer and oxygen transfer, the influent NH₃-N concentration was increased from about 10 mg·L⁻¹ along the flow path to about 25 mg·L⁻¹, and the DO was increased from 2–4 mg·L⁻¹ along the flow path to about 6–7 mg·L⁻¹. During the test, the water temperature was 9.7–10.4 ℃. Aeration pumps and wave makers were used to maintain system oxygenation and carrier fluidization. Samples were taken every 30 min and pretreated to determine the change in ammonia nitrogen concentration. According to reported methods, the biofilm load calculation method in the hybrid MBBR system is shown in equations (1) and (2).

Where: N₂ is the nitrification volumetric load of the biofilm in the hybrid system, kg·(m³·d)⁻¹; N₁ is the overall nitrification volumetric load of the MBBR hybrid system, kg·(m³·d)⁻¹; N₃ is the nitrification volumetric load of the activated sludge, kg·(m³·d)⁻¹; η is the nitrification contribution rate of the suspended carriers, %.

1.4 Verification and Optimization Control Analysis of SND in MBBR System

For the simultaneous nitrification and denitrification (SND) phenomenon in the MBBR system, to directly verify the source of the SND effect, a denitrification laboratory test using pure MBBR biofilm was conducted on February 28, 2021. Suspended carriers from the aerobic zone were collected. The carrier filling ratio was consistent with the actual conditions of the biological tank. The influent for the test was prepared by adding potassium nitrate to the raw water influent of the plant. Samples were taken every 20 min and pretreated to determine nitrate nitrogen concentration.

From December 1, 2020, to February 28, 2021, the water quality and operating parameters of the MBBR zone after reconstruction were tracked daily throughout the winter. The samples included effluent of the anoxic zone and effluent of the MBBR zone. Samples were taken every 2–3 h. After collection, the samples were pretreated promptly. After sedimentation, the supernatant was collected, and three samples were mixed in equal amounts to determine TN, NH₃-N, NO₃-N, and COD. At the same time, the DO level in the aerobic MBBR zone was measured. The measured data were subjected to principal component analysis (PCA) to verify the main response parameters affecting the SND effect in the aerobic zone. The optimal control parameters were determined through fitting analysis to provide a basis for the scale-up control of SND.

1.5 High-Throughput Sequencing of Microorganisms in the Wastewater Treatment Plant

To analyze the differences between the sludge and biofilm phases in the MBBR system in terms of microbial diversity, dominant species, and functional bacteria enrichment efficiency, and to microscopically verify the source of SND, activated sludge and suspended carriers from the MBBR system were collected on February 28, 2021, with one set of parallel samples, for high-throughput sequencing of 16S rDNA amplicons. At this time, the system had experienced the entire winter, and the microbial community composition could better reflect the actual situation under low-temperature conditions. The specific methods for high-throughput determination and analysis refer to previously reported literature.

1.6 Conventional Water Quality Determination Methods

All water quality parameters involved in this study were determined using standard methods. COD was determined by the national standard method; NH₃-N was determined by the Nessler's reagent spectrophotometric method; NO₃-N was determined by the ultraviolet spectrophotometric method; TN was determined by the alkaline potassium persulfate digestion-UV spectrophotometric method; TP was determined by the ammonium molybdate spectrophotometric method; DO and water temperature were measured using a portable dissolved oxygen meter.

2 Results and Discussion

2.1 Overall Treatment Effect Under Low-Temperature Conditions Before and After Reconstruction

First, the overall treatment effect of the system under low winter temperatures before and after reconstruction (2018-12-01 to 2019-02-28, 2020-12-01 to 2021-02-28) was compared. Before and after reconstruction, the winter water temperature was basically maintained at 9–13 ℃. Before reconstruction, the winter flow was (9.12±0.58)×10⁴ m³·d⁻¹, and after reconstruction, it increased to (9.42±0.44)×10⁴ m³·d⁻¹. As shown in Figure 3(a)-(d), after reconstruction, the average influent COD and BOD₅ loads increased by 16.82% and 26.97%, respectively, compared to before reconstruction, but the effluent decreased to (20.04±2.38) and (4.47±0.84) mg·L⁻¹, respectively. After reconstruction, the average influent TP load increased by 23.40% compared to before reconstruction, but the effluent TP could still be as low as (0.26±0.02) mg·L⁻¹. After reconstruction, the influent SS load was slightly lower than before, but the effluent SS decreased by more than 17.51% compared to before, to (3.44±1.17) mg·L⁻¹, with a maximum of only 4.75 mg·L⁻¹.

The core highlight of this project after reconstruction was the enhancement of nitrogen removal capacity. After reconstruction, the average influent NH₃-N and TN loads increased by 21.56% and 19.79%, respectively, compared to before reconstruction, but the effluent decreased to (0.94±0.22) and (8.59±0.64) mg·L⁻¹, respectively. The maximum effluent NH₃-N was only 1.33 mg·L⁻¹, and the maximum TN was only 9.7 mg·L⁻¹. For nitrogen removal, since activated sludge still played the core role after reconstruction, based on the significant extension of anoxic retention time, the actual winter denitrification volumetric load of the system before and after reconstruction did not change much. For nitrification, after reconstruction, the nitrification effect was enhanced by adding suspended carriers to form a hybrid MBBR system to strengthen NH₃-N removal. Based on the significant reduction of aerobic retention time, the actual winter nitrification load increased from 0.081 to 0.182 kg·(m³·d)⁻¹ after reconstruction, an increase of 124.69% compared to before reconstruction. The application of the MBBR process in this project achieved a significant improvement in nitrification performance.

2.2 Analysis of System Nitrogen Removal Performance, Potential, and Role of Suspended Carriers

To further determine the nitrogen removal effect of each functional zone in the biological section after reconstruction, the water quality along the biological tank was tracked, and the nitrogen removal pathway of each functional zone after reconstruction was analyzed through the mean water quality parameters of each functional section. As shown in Figure 4, during the measurement period, the mean actual influent TN, NH₃-N, and COD concentrations of the biological tank were 43.59, 38.50, and 278.00 mg·L⁻¹, respectively, and the effluent TN, NH₃-N, and COD decreased to 7.53, 0.51, and 24.46 mg·L⁻¹, respectively. For TN removal, the system TN removal rate calculated from influent and effluent TN concentrations was 78.95±6.01%, and removal mainly occurred in the pre-anoxic zone, with a nitrogen removal amount of 27.51 mg·L⁻¹. The denitrification load and contribution rate were 0.149 kg·(m³·d)⁻¹ and 79.16%±2.24%, respectively. Combined with the change in COD concentration, the COD/TN consumed for denitrification in the pre-anoxic zone was 5.22±0.21. The post-anoxic zone used external sodium acetate for denitrification, with a nitrogen removal amount and load of 4.35 mg·L⁻¹ and 0.024 kg·(m³·d)⁻¹, respectively. The calculated COD/TN consumed for denitrification was 4.66±0.57. Sodium acetate, as a small molecular organic acid salt, is easily utilized by microorganisms and has higher denitrification efficiency than the complex organic carbon source in raw water. After reconstruction, the influent had good biodegradability, and the denitrification requirements could be basically met only through the pre-anoxic zone, thereby reducing the pressure on post-anoxic denitrification and the amount of external carbon source. For NH₃-N removal, the DO in the MBBR zone during actual operation was 2–4 mg·L⁻¹, and the system NH₃-N removal rate calculated from influent and effluent NH₃-N concentrations was 98.67%±3.14%.

In actual operation of wastewater treatment plants, influent quality shocks often occur due to natural factors such as rainfall and snowmelt, industrial factors such as increased industrial enterprises or temporary business abnormalities, and social factors such as holidays and special events. Improving the impact resistance of the biological treatment system has become the key to the safe and stable operation of the plant. To verify the maximum tolerance load of the system, a nitrification laboratory test under high substrate and high DO was conducted. At the same time, to verify the actual role of suspended carriers in the hybrid MBBR system, three groups of experiments (hybrid, pure-biofilm, and pure-sludge) were set up twice. The changes in NH₃-N in each system during the test are shown in Figure 5. With influent NH₃-N at 25 mg·L⁻¹, the hybrid and pure-biofilm systems reduced NH₃-N to below 1.5 mg·L⁻¹ in 2.5 and 3.0 h, respectively, while the pure-sludge system had a much slower decrease in NH₃-N throughout the test period. The calculated mean nitrification volumetric loads of the pure-sludge, pure-biofilm, and hybrid systems were 0.072, 0.215, and 0.286 kg·(m³·d)⁻¹, respectively. According to equation (1), under the presence of activated sludge, the volumetric load of the suspended carriers in the MBBR system was 0.214 kg·(m³·d)⁻¹, which was not significantly different from the measured load of the pure-biofilm group. Compared with other plants using the hybrid MBBR process, with similar carrier filling ratios, the maximum nitrification load of the suspended carriers in this project was also at a relatively high level, which may be related to the further enhancement of mass transfer of the suspended carriers and increased activity of functional bacteria under high influent load conditions. In addition, with the combination of activated sludge, the maximum nitrification volumetric load under high substrate and high DO conditions reached 0.286 kg·(m³·d)⁻¹, meeting and exceeding the design load (0.275 kg·(m³·d)⁻¹), meeting design expectations.

To quantitatively analyze the maximum tolerance load of the biological system in this project, based on the volumetric load of the hybrid system measured in the laboratory test, combined with the influent water quality data along the flow path, the system could withstand a maximum treatment capacity of 1.23×10⁵ m³·d⁻¹ (calculated assuming a positive linear relationship between system nitrification load and hydraulic retention time) while ensuring stable effluent NH₃-N compliance. This is an increase of 30.57% compared to the actual winter treatment capacity of this project, with the system hydraulic retention time further reduced to 3.43 h. Similarly, based on the actual treatment capacity after reconstruction, the tolerable influent NH₃-N concentration of the biological tank was as high as 55 mg·L⁻¹, an increase of 56.96% compared to the mean actual winter influent NH₃-N of this project.

Regarding the role of suspended carriers in the hybrid MBBR system of this project, according to equation (2), the nitrification contribution rate of the suspended carrier biofilm in this project was 74.83%. It can be seen that under low-temperature conditions, the suspended carriers have become the dominant factor for nitrification in the hybrid MBBR system. Combined with literature reports, it is analyzed that under low winter temperatures, the activated sludge system, limited by its loose structure and high specific surface area, has good heat dissipation performance and poor thermal insulation for internal microorganisms, leading to reduced nitrification performance. In contrast, the biofilm on the suspended carriers, benefiting from its dense structure and high EPS protection, provides good thermal insulation for the inner layer of the biofilm, thus ensuring good nitrification performance under low-temperature conditions. In addition, the along-flow measurement results of this project showed a significant TN removal phenomenon in the MBBR zone. Combined with the change in NH₃-N concentration along the aerobic zone and the recirculation ratio, the calculated TN removal was (3.42±0.11) mg·L⁻¹, with a contribution rate of 9.83%±0.67%, indicating the existence of an SND effect in the system.

2.3 Verification of SND Effect and Analysis of Influencing Factors

At present, most wastewater treatment plants applying the MBBR process can observe nitrogen removal in the aerobic zone, generally removing 3–8 mg·L⁻¹ of TN. The reason for this phenomenon is mainly attributed to the SND effect of the biofilm. That is, as the penetration capacity of dissolved oxygen weakens, an anoxic/hypoxic layer forms in the inner layer of the biofilm, providing conditions for the enrichment of denitrifying bacteria, thus laying the foundation for SND nitrogen removal.

To further confirm the source of aerobic SND, a denitrification laboratory test using aerobic biofilm was conducted, as shown in Figure 6. Throughout the test period, the decrease in NO₃-N was relatively gentle and uniform, with the system removing 3.72 mg·L⁻¹ of nitrate nitrogen in 120 min. The average C/N consumed by the system was 3.43±0.05. Theoretically, the C/N required for denitrification is 2.86. Considering the consumption by biological assimilation and DO in the raw water, the required C/N generally reaches above 4. However, the C/N required for nitrogen removal by the suspended carriers in this project was lower than that consumed by general microbial denitrification. It is speculated that when the suspended carriers in the MBBR zone exert SND nitrogen removal, they mainly use internal carbon sources generated from the aging and lysis of the biofilm as electron donors, so the utilization rate of organic matter (external carbon source) in the raw water is generally low. The use of aged biofilm as an internal carbon source for denitrification also suggests that the MBBR process may have lower sludge production. Using the entire test period as a node, the calculated denitrification volumetric load of the aerobic biofilm was (0.045±0.003) kg·(m³·d)⁻¹. The biofilm on the suspended carriers added in the aerobic zone had a denitrification effect, providing evidence for the occurrence of simultaneous nitrification and denitrification in the biofilm.

SND, as a beneficial effect after the application of the MBBR process for upgrading in this project, further improved the system TN removal rate and reduced the dependence on external carbon sources for denitrification to a certain extent. To further amplify the aerobic SND effect, from the perspective of optimized operation, the influence of influent and effluent water quality and control parameters on the SND effect of the MBBR zone was further explored. Through long-term tracking of control parameters such as influent nitrate nitrogen (NO₃-Ninf), effluent nitrate nitrogen (NO₃-Neff), influent organic matter (CODinf), effluent organic matter (CODeff), influent ammonia nitrogen (NH₃-Ninf), effluent ammonia nitrogen (NH₃-Neff), total nitrogen removal (ΔTN), and DO in the MBBR zone throughout the winter, principal component analysis (PCA) was used to explore the correlations between these parameters. The results are shown in Figure 7. The ΔTN in the aerobic zone showed a relatively significant negative correlation with DO, while the correlation with CODinf, CODeff, and NO₃-Ninf was not obvious. Too high DO leads to more thorough mass transfer in the biofilm, making it difficult to form an anoxic zone inside, which is unfavorable for the denitrification process and inhibits the SND effect. In addition, CODinf and NO₃-Ninf in the MBBR zone did not have a significant impact on the SND effect of the suspended carriers. On the one hand, this again verifies that the SND process of the biofilm mainly uses internal carbon sources rather than external carbon sources. On the other hand, it also indicates that the NO₃-N removed by SND in the MBBR zone does not come from influent NO₃-N, but mainly from NO₃-N generated by nitrification in the aerobic tank. NO₃-Neff is negatively correlated with ΔTN in the aerobic zone. As the result of the combined action of nitrification and SND in the aerobic zone, on the one hand, higher DO leads to better nitrification and higher effluent NO₃-Neff; on the other hand, higher DO leads to poorer SND and more inhibited nitrogen removal, further leading to higher effluent NO₃-Neff.

To achieve "measurable and controllable" SND and provide guidance for plant operation control, DO, which can be manually controlled, was selected as the independent variable, and ΔTN was used as the dependent variable for linear regression analysis. The results are shown in Figure 8(a). When the actual DO in the MBBR zone of the plant was 2.06–3.67 mg·L⁻¹, an SND effect existed, with nitrogen removal of 2.36–4.10 mg·L⁻¹, and ΔTN and DO basically satisfied ΔTN = 6.0040 – 0.9579 DO (R² = 0.9734). Furthermore, using the fitted curve, it was calculated that theoretically, when DO is controlled below 6.27 mg·L⁻¹, the MBBR zone can exert an SND nitrogen removal effect. This is consistent with the conclusion of Wang Xuejiang et al. in studying the effect of DO on the SND effect of an MBBR system, where they found that the TN removal rate could still exceed 30% when DO was as high as 6.0 mg·L⁻¹. However, it should be noted that this value is a theoretical value obtained by fitting, and the actual application effect needs further study in subsequent operation. In addition, it should be considered that DO regulation simultaneously affects the nitrification effect of the system, thereby affecting NH₃-N removal efficiency. Therefore, the results of linear regression analysis using DO as the independent variable and ΔNH₃-N as the dependent variable are shown in Figure 8(b). According to the mean influent NH₃-N of the MBBR zone in winter (10.47 mg·L⁻¹) and the design effluent standard (1.50 mg·L⁻¹), it is calculated that ΔNH₃-N should be controlled not less than 8.97 mg·L⁻¹. Combined with the linear regression results, DO should be controlled above 2.47 mg·L⁻¹. In summary, when the DO in this project was 2.06–3.67 mg·L⁻¹, an SND effect existed. However, based on the principle of meeting effluent ammonia nitrogen standards while reducing aeration energy consumption, DO could be controlled around 2.47 mg·L⁻¹, where the SND effect of the system is optimal, removing 3.64 mg·L⁻¹ of TN.

To verify the influence of operating parameters on the SND effect of the biofilm and obtain a general pattern, the SND effects of six wastewater treatment plants applying the MBBR process under different operating conditions were compared. The results are shown in Table 2. First, the DO in the aerobic zone of each plant was mostly controlled at 2–3 mg·L⁻¹, and the sludge concentration was 3.0–4.0 g·L⁻¹. Under these conditions, a relatively obvious SND effect could be achieved. According to the fitting results of ΔTN and DO in this project, when DO was controlled below 6.27 mg·L⁻¹, an SND effect could occur. This project, with a relatively thicker biofilm in winter, had a stronger tolerance to DO. Second, Plant B, with the lowest sludge concentration, had the best SND effect. Plants A, D, and E, with higher carrier filling ratios, had more obvious SND effects, all indicating that in the hybrid MBBR system, the SND phenomenon mainly originates from the biofilm.

In summary, in addition to enhancing the SND effect by controlling DO, this project could also try to gradually reduce the system sludge to about 3.0 g·L⁻¹, providing better conditions for biofilm thickness cultivation and anoxic layer formation through low sludge concentration, thereby providing conditions for further SND performance.

2.4 Analysis of Nitrogen-Removing Bacterial Community Composition

In terms of the differences in species richness and diversity between the sludge and biofilm phases, based on the small difference in the results of two amplicon sequencing runs, one set of data was selected for analysis. As shown in Figure 9, the Chao, Ace, and Simpson indices characterize the richness and distribution uniformity of the microbial community, while the Shannon index represents the diversity of the microbial community. Higher Chao and Ace indices indicate higher OTU richness, a lower Simpson index indicates better species distribution uniformity, and a higher Shannon index indicates higher community diversity. The results showed that both species richness and diversity were significantly higher in the biofilm than in the activated sludge, and the species distribution in the biofilm was also more uniform than in the activated sludge. The total numbers of OTUs in the biofilm and activated sludge were 1,652 and 1,372, respectively, with 770 shared OTUs, accounting for 46.6% and 56.1% of the total OTUs in the biofilm and activated sludge, respectively. The biofilm had 882 unique OTUs, accounting for nearly 40% of the total OTUs. It can be seen that the biofilm enriched a portion of microorganisms that were not present in the original activated sludge. The addition of suspended carriers significantly increased the richness and diversity of the microbial community in the system, and the more complex microbial community composition also laid the foundation for the impact resistance of the MBBR system.

To further understand the differences in dominant species and functional bacterial composition between the biofilm and activated sludge and to provide microscopic support for the SND phenomenon in the biofilm, the microbial composition at the genus level was analyzed. As shown in Figure 10, the dominant species in the biological system of this project included norank_f_Caldilineaceae, norank_f_JG30-KF-CM45, Nitrospira, Pedomicrobium, and SWB02. Among them, norank_f_Caldilineaceae and norank_f_JG30-KF-CM45 could be enriched in both the biofilm and activated sludge, with relative abundances in the biofilm and activated sludge of 3.05, 5.69% and 7.06, 3.08%, respectively. norank_f_Caldilineaceae are mostly filamentous bacteria, and some genera have denitrification ability. They were enriched in the biofilm because the biofilm can secrete more EPS for their proliferation. On the other hand, they can also be enriched in activated sludge as PAOs, promoting phosphorus removal. norank_f_JG30-KF-CM45 can participate in carbon and nitrogen transformation processes, enhancing the removal of organic matter and nitrogen.

According to literature reports, it has been found that the MBBR process can achieve efficient enrichment of nitrifying bacteria such as Nitrospirota, Nitrosomonas, and Nitrolancea at low temperatures. The main nitrifying genus in this project was Nitrospira, which has complete nitrification ability. Its relative abundances in the biofilm and activated sludge were 4.73% and 0.39%, respectively. According to literature reports, the relative abundance of nitrifying bacteria in other plants using activated sludge systems under low winter temperatures is mostly below 1%. After the hybrid MBBR reconstruction in this project, the relative abundance of nitrifying bacteria in the activated sludge was relatively low, providing evidence for the redistribution of nitrifying bacteria in the MBBR system under low-temperature conditions, as mentioned earlier. The relative abundance of Nitrospira in the biofilm of this project was more than 10 times that in the activated sludge, and it was calculated that more than 92% of the nitrifying bacteria in the system were enriched in the biofilm. This is the advantage of the MBBR process under low-temperature conditions, ensuring the stable operation of the system.

Regarding the SND phenomenon in the biofilm, the suspended carriers also showed a stronger enrichment capacity for some denitrifying bacteria than the activated sludge. Studies have shown that both SWB02 and Pedomicrobium can participate in the denitrification process, and SWB02 can secrete C6-HSL, which has a good inhibitory effect on sludge bulking. The relative abundances of SWB02 and Pedomicrobium in the biofilm were 3.68% and 3.78%, respectively, much higher than their relative abundances in the activated sludge (0.21% and 0.22%). Moreover, it was calculated that the proportions of SWB02 and Pedomicrobium in the entire hybrid system were both over 94.5%. Although the suspended carriers were only added in the aerobic zone, the mature biofilm has a certain thickness. Due to the limitations of mass transfer and oxygen transfer, an anoxic/hypoxic environment exists in the inner layer of the biofilm, enriching denitrifying bacteria, thus endowing the biofilm with denitrification ability. This is the microbiological basis for the simultaneous nitrification and denitrification ability of the biofilm. In addition, it was found that Hyphomicrobium can also be enriched in low-temperature environments. This genus can use toxic and refractory organic compounds for denitrification. Its relative abundance in the biofilm was 1.80%, also much higher than its relative abundance in the activated sludge (0.12%).

Comparing the dominant denitrifying bacteria (DNB) and their relative abundances in the biofilm of other plants, the results are shown in Table 3. First, in terms of the composition of denitrifying bacteria in the aerobic biofilm, the types of denitrifying bacteria enriched in the aerobic biofilm of each project were not consistent. The only recurring genera were Terrimonas and Hyphomicrobium. Terrimonas is also a common denitrifying bacterium in activated sludge systems, and similar to Hyphomicrobium, it can degrade anthracene compounds, and some are classified as aerobic denitrifying bacteria. These two bacteria appeared in Plants F, H, and this project. Plant H and this project were frequently impacted by industrial wastewater during actual operation, and the specific water quality of industrial water may be one of the reasons for the enrichment of these two bacterial groups. Therefore, it can be inferred that specific denitrifying genera in the aerobic biofilm are not a necessary condition for SND, and differences in water quality among different projects may be the main reason for differences in denitrifying bacterial groups. Second, in terms of the relative abundance of denitrifying bacteria in the aerobic biofilm, except for Plants F, H, and this project, other projects achieved a relative abundance of denitrifying bacteria in the aerobic biofilm of about 8%, with SND nitrogen removal exceeding 2 mg·L⁻¹. Under the impact of industrial water, although the relative abundance of denitrifying bacteria in the aerobic biofilm of Plant H and this project was relatively high, the denitrification efficiency was not particularly outstanding. It can be seen that the influent water quality has a high priority in selecting the types of denitrifying bacteria in the aerobic biofilm, and the impact of industrial water inhibits the denitrification efficiency of the aerobic biofilm.

In summary, the MBBR process can significantly increase the richness and diversity of the microbial community in the system. The enrichment capacity of the biofilm for the nitrifying genus Nitrospira is more than 10 times that of the activated sludge, greatly enhancing the nitrification effect of the system. Denitrifying genera such as SWB02, Pedomicrobium, and Hyphomicrobium are better enriched in the aerobic biofilm, providing the microscopic basis for SND performance and enhancing TN removal. However, due to the impact of industrial water, the SND denitrification efficiency is not significant.

3 Conclusions

① A wastewater treatment plant in Southwest China upgraded its original CASS tank to a Bardenpho-MBBR biological tank and sedimentation tank to achieve upgrading from Class A to quasi-Class IV standards. After reconstruction, under winter water temperatures of 9.3–13.1 ℃ and with mean influent NH₃-N and TN loads increased by about 20% compared to before reconstruction, the effluent indicators stably met the quasi-Class IV water standards. The maximum nitrification load of the MBBR system reached 0.286 kg·(m³·d)⁻¹, meeting design expectations. The nitrification contribution rate of the suspended carriers was 74.83%, occupying a dominant position. The system still has the potential to increase water quality by 56.96% and flow by 30.57%.

② In the MBBR zone, the suspended carriers use internal carbon sources to exert an SND effect, achieving a nitrogen removal contribution rate of 9.83%±0.67%. To amplify the SND effect, DO should be controlled as much as possible at 2.47 mg·L⁻¹ in actual operation, where SND can remove 3.64 mg·L⁻¹ of TN. Later, attempts can be made to further reduce the sludge concentration to about 3.0 g·L⁻¹ to provide conditions for further amplification of the SND effect of the biofilm.

③ The MBBR process significantly increased the richness and diversity of the microbial community in the system. The enrichment capacity of the biofilm for the nitrifying genus Nitrospira was more than 10 times that of the activated sludge, providing the microbiological basis for enhanced nitrification in this project. The denitrifying genera enriched in the aerobic biofilm, such as SWB02, Pedomicrobium, and Hyphomicrobium, laid the microscopic foundation for the occurrence of the SND phenomenon in the biofilm.