The main body of the sewage treatment equipment is an integrated MBR reactor. Raw sewage was used to discharge wastewater from a comprehensive railway station, which consists of a railway passenger station, a large warehouse area, and related supporting facilities. The water quality indexes were COD levels of 100–400 mg/L, NH₄ ⁺-N concentrations of 15 –45 mg/L, TN concentrations of 20–60 mg/L, and TP concentrations of 2–6 mg/L.

As shown in Figure 1 1) sewage from the integrated railway station was collected through the pipeline to the grid pool of the sewage treatment plant. Inorganic suspended matter in the sewage was removed through a thick and fine grid to reduce the wear on the subsequent pipelines and equipment. The effluent enters the catchment pool. As shown in Figure 1 2) A large number of facultative aerobic bacteria are contained in the facultative MBR system and can degrade organic matter in sewage by the dual action of microelectrolysis and facultative bacteria metabolism to degrade macromolecular organic pollutants into small molecular organic matter, which are eventually oxidized and decomposed into stable inorganic substances. Such as carbon dioxide and water. Moreover, power consumption is reduced, because the generation of facultative bacteria does not need the guarantee of dissolved oxygen (DO). The main function of aeration in a sewage treatment system is to scour and shock membrane filaments, and the DO produced can be used to oxidize a part of the small molecular organic matter and maintain the DO value of the effluent to ensure the normal microbial metabolism in the combined MBR system with concurrent oxygen microelectrolysis.

The design scale of the combined treatment system is 400 m³/day, the design total sludge age is infinite, the organic residual sludge discharge is nearly “zero” discharge, the mixed liquid concentration (MLSS) is 8000–20000 mg/L, and the sludge load is 0.02 –0.10 kg COD/(kg MLSS·day).

The iron–carbon-based material consisted of 40%–50% elemental iron powder (300 mesh), 35%–42% activated carbon powder (200 mesh), 6%–8% metal catalyst (made of various metals), 5%–8% adhesive, and foaming agent, etc. The materials were mixed evenly to prepare 1 –2 cm balls, which were dried at 105 C for about 2 h in a drying oven, preheated at 600 C for half an hour in a muffle furnace, and heated at 1000 C for 3 h. After annealing and cooling, the balls were crushed and sifted to retain particles below 150 mesh. The time between production and use should not be too long and should be suitable for real-time production before use to ensure the activity of iron powder and avoid excessive oxidation.

The data shown in Figure 2 1) show that the COD of the influent municipal sewage varies at 112–328 mg/L without regularity, and a large change in COD value will have a load impact on the water treatment system (Vleeschauwer et al., 2020). Studies have shown that such a load impact will often have a great impact on the water treatment system (Yadu et al., 2019). Experimental data show that the sewage treatment system formed a larger organic shock load, the stable effluent COD value was between 19 and 43 mg/L, and the water COD and effluent COD values had a significant positive correlation. However, the ratio changed, which allowed the sewage treatment system to have a good performance and strong load impact resistance. Compared with the original MBR system without ICMC-AS, the COD removal rate of the combined system increased from 80% to 92%.

As shown in Figure 2 2), the NH₄ ⁺-N value measured at the inlet in the small test fluctuated greatly between 16 and 42 mg/L, and the TN value was between 2 and 55 mg/L, resulting in the large load impacts of NH₄ ⁺-N and TN on the sewage treatment system. The load impacts of NH₄ ⁺-N and TN loads have great influence on water treatment system (Meng et al., 2018). The NH₄ ⁺-N and TN values measured at the outlet were maintained at about 2–3 mg/L, indicating that the sewage treatment system has a strong ability to resist the load impacts of NH₄ ⁺-N and TN. In addition, the NH₄ ⁺-N and TN removal rates were about 93.5% and 94%, respectively, which are not much different from the removal rates in the laboratory. This finding indicates that the removal capacities for NH₄ ⁺ and TN in the sewage treatment system were strong, and the current concentrations of NH₄ ⁺ and TN did not reach the treatment limits. Additionally, the results show that the sewage treatment system has strong adaptability and adjustment ability and can maintain efficient nitrogen pollutant removal ability under a large load shock. Compared with the original MBR system without ICMC-AS, the NH₄ ⁺-N and TN removal rates increased from 40% and 30% to 93% and 93.5%, respectively, in the combined system.

As shown in Figure 2 3), the sewage treatment system has a good removal effect. TP values oscillated between 2 and 6 mg/L and were maintained at 0.3 mg/L after being treated by the sewage treatment system. This result indicates that the system had a stable treatment capacity for the TP of municipal sewage between 2 and 6 mg/L (most municipal sewage is within this range), and its removal rate was 92.58% on average. Compared with the original MBR system without ICMC-AS, the TP removal rate of the combined system increased from 10% to 92%.

As shown in Figure 3 1), COD was removed by rapid consumption within the first 120 min, and the value decreased from 400 mg/L to 100 mg/L. NH₄ ⁺-N decreased from 20 mg/L to 2.32 mg/L within 270 min, and NH₄ ⁺-N decreased more slowly at 40 –60 and 90 –120 min. The DO values in these two periods oscillated between 0.2 and 0.3 mg/L, indicating that the biochemical reaction of the system was very violent between 0 and 120 min, which made the DO values hover between 0.2 and 0.3 mg/L and inhibited NH₄ ⁺-N decomposition. However, NO₃ ⁻-N concentration decreased from 5 mg/L to 2.81 mg/L between 0 and 40 min, increased from 2.58 mg/L to 2.89 mg/L between 40 and 180 min, and gradually decreased to 0.65 mg/L between 180 and 360 min. This finding is because the concentration change of NO₃ ⁻-N is affected by nitrification and denitrification reactions, as well as the biochemical reactions of other elements (such as the biochemical absorption of iron and carbon) (Xing et al., 2016). Therefore, during the first 0–40 min, NO₃ ⁻-N gradually entered the adsorption plane of various reactions and participated in various reactions. Therefore, its value was rapidly reduced to 2.81 mg/L. When these participating reactions reached saturation or equilibrium state, it reflects the comparison between the rates of nitrification reaction (NO₃ ⁻-N generation) and denitrification reaction (NO₃ ⁻-N transformation and removal) (Deng et al., 2016b). When the rates of nitrification and denitrification reactions reach equilibrium, the value of NO₃ ⁻-N will keep oscillating within a certain numerical range. However, when the denitrification reaction is larger than the nitrification reaction, the value of NO₃ ⁻-N will gradually decrease. In this case, NO₂ ⁻-N is usually seen as an intermediate product in the conversion of NH₄ ⁺-N to NO₃ ⁻-N (Deng et al., 2020b). NO₂ ⁻-N accumulated gradually from 0 mg/L to 2.21 mg/L at 0–60 min, decreased to 0.61 mg/L at 40–90 min, accumulated gradually accumulated from 0.61 m/L to 1.57 mg/L at 90–240 min, and then decreased to 0.35 mg/L at 240–360 min. In the whole process, the value of NO₂ ⁻-N experienced two accumulation processes. More complex changes occurred. This change law does not accord with the characteristics of synchronous nitrification and denitrification or short-cut nitrification (Jiaohui et al., 2021; Wang et al., 2021; Tong et al., 2022a; Tong et al., 2022b). Therefore, based on the analysis of the change values of TN and DO, COD, NH₄ ⁺-N, NO₃ ⁻-N, NO₂ ⁻-N, and DO are co-changing according to a certain correlation. DO showed periodic oscillations and repeated changes during the operation of the system. The changes in DO can be divided into three stages. In the first stage, the operation time of the treatment system was between 0 and 120 min, and the DO value in this period changed between 0.1 and 0.4 mg/L. In the second stage, the operation time of the treatment system was between 120 and 270 min, and the DO value in this period varied between 1.0 and 1.4 mg/L. In the third stage, the operation time of the treatment system was between 270 and 360 min, and the DO value in this period varied between 1.8 and 2.3 mg/L.

As shown in Figure 3 2), the change rate of the NO₃ ⁻-N process value was close to that of COD at 0–40 min, indicating that the process of NO₃ ⁻-N reaction during this period was equally dramatic. On the one hand, the reason is that many biochemical reactions of microorganisms in the system require the participation of NO₃ ⁻-N. On the other hand, heterotrophic denitrification bacteria at this stage have sufficient organic carbon sources and a large amount of oxygen to efficiently carry out heterotrophic denitrification and consume NO₃ ⁻-N. At 40–180 min, the NO₃ ⁻-N process value hardly changed and was in a relative equilibrium state because of the lack of relative oxygen and the indominance of the number of relative denitrifiers in the whole system. However, from the analysis of TN removal rate, the overall TN removal rate was maintained at a relatively high growth level. Moreover, the change rates of the decomposition and process values of NH₄ ⁺-N were also maintained at relatively high levels, which indicates that the nitrification and denitrification reactions were maintained at a relatively high dynamic balance. At 180–360 min, the change rate of the NO₃ ⁻-N process value was relatively large, and the COD value dropped to 57 mg/L, which made the organic carbon source become relatively short, and the NO₃ ⁻-N reaction process was transferred into the ICMC-AS. At this time, the organic carbon source materials (such as extracellular polymers, etc.) stored in ICMC-AS can be combined with the sufficient aerobic environment outside ICMC-AS to carry out heterotrophic denitrification, and the electron supply of ICMC-AS can be combined with the anoxic, facultative, and aerobic regions of MBR to carry out different degrees of autotrophic denitrification (Adav and Lee 2011; Pellicer-Nacher et al., 2013). Thus, an efficient denitrification reaction with a change rate similar to the process value of NH₄ ⁺-N can be achieved.

As shown in Figure 4, the reaction rate constants of each substance were obtained by fitting the COD, NH₄ ⁺-N, NO₃ ⁻-N, and TN removal efficiencies. Among them, the K _COD (0–120 min) was 0.647 ± 0.017, which is the maximum reaction rate constant of each substance, and the minimum value of K _TN was 0.258 ± 0.083. If only the conversion of NH₄ ⁺-N into NO₃ ⁻-N under aerobic condition was considered, the converted K N O 3 − − N will be about 0.416 ± 0.044, but the actual conversion to NO₃ ⁻-N is not only the ammonification reaction of NH₄ ⁺-N. The actual value of K N O 3 − − N will be higher than 0.416. Ammonification reaction is the main way to convert NH₄ ⁺-N to NO₃ ⁻-N; thus, other transformation ways are greatly affected by the reaction conditions. Here, we only calculated the NH₄ ⁺-N ammonification reaction after weighted estimation. According to the size of the reaction rate constant, the reaction rates were in the order: K _COD (0.647 ± 0.017) > K N O 3 − − N (0.416 ± 0.044) > K N H 4 + − N (0.275 ± 0.014) > K _TN (0.258 ± 0.083). Therefore, the combined application system is a wastewater treatment system with high nitrogen removal efficiency dominated by COD removal. The removal rates of various pollutants were in the order of: COD > NO₃ ⁻-N > NH₄ ⁺-N > TN.

Membrane contamination is inevitable during the operation of the combined system. The system stops the sewage treatment operation and starts the membrane cleaning operation when the membrane pressure difference reaches 50 kPa. The membrane cleaning system adopts full automatic control. The membrane pressure difference of the original MBR system reached 50 kPa between 5 and 7 days, whereas that of the combined system reached 50 kPa between 7 and 10 days. This finding indicates that the combined system effectively mitigated the membrane contamination. Compared with the original MBR system, the sludge settling performance of the combined system was remarkably improved, and the settling performance of the combined system can reach the settling effect of the original system (30 min) in 5 min.

The sewage treatment scale of the railway station is 3,000 t/day, including six lifting pumps with a total operating power of 9 kW and six MBR wastewater processors with a total operating power of 99 kW. This scale translates into a daily power consumption of 1,903.2 kWh. According to the TN removal rate, which increased by three times conversion, a total of 1268.8 kW h/day energy saving was achieved, and the annual electricity saving was 463,112 kWh. The amount of membrane-cleaning water was 159.6 t/time, and the drug dosage was 1.92 t/day. The average membrane-cleaning volume was reduced from 61 times per year to 43 times, with a reduction of 18 times, saving 2,872.8 t of water and 34.54 t of drug consumption. According to CO₂ conversion, the annual reduction in electricity was 463,112 kWh, which corresponds to the carbon emission reduction of 363,542.92 kg CO₂e; the water saved was 2,872.8 t, which corresponds to the carbon emission reduction of 482.63 kg CO₂e; and the reduction in pharmaceutical dosage was 34.54 t, which corresponds to the comprehensive carbon emission reduction of 24,178 kg CO₂e. Therefore, the annual carbon reduction of the combined system is at least 388,203.55 kg CO₂e.