Project Background
- Location: A wastewater treatment plant under a chemical group in Zhejiang
- Difficulty: It used the A2O process to remove high Kjeldahl nitrogen from wastewater bofore. But its influent consisted of various production wastewaters, containing a large amount of Ca²⁺ and S²⁻ ions in the supernatant of carbide slag from acetylene wastewater. These ions seriously affected the microbial nitrification process and the operation of the plant.
- Our measures: 1. Add pretreatment facilities at the initial stage. 2. Convert the secondary aerobic stage to MBBR process.
Experimental Materials and Methods
Raw Water Quality
The influent for the pilot plant was derived from the effluent of the wastewater treatment plant's homogenization tank. Table 1 presents the influent water quality indicators. According to the experimental requirements, the influent ammonia nitrogen concentration was raised, with ammonium sulfate used as a nitrogen supplement to maintain a TKN concentration of 120-220 mg/L.
Process Flow and Main Equipment
The current treatment plant process is the A2O process (Figure 1), with the AO stage using the activated sludge method and the final aerobic tank adding elastic fillers. The design flow rate is 14,400 m³/d. The pilot unit process is shown in Figure 2, with a flow rate of 100 L/h and continuous 24-hour operation.
The pilot unit's main characteristics include the addition of a pretreatment facility and the conversion of the secondary aerobic stage to an MBBR process. This hybrid suspended-attached state design eliminates the need for a secondary sedimentation tank. Table 2 lists the specifications and dimensions of the primary equipment.
Results and Discussion
1. Startup and Operating Conditions
The pilot unit commenced in April 2007, starting with microbial inoculation. The sludge added was sourced from the plant's initial aerobic tank. MBBR's nitrifying bacteria were cultured with domestic sewage and ammonium sulfate-treated tap water, with NaHCO₃ added to regulate alkalinity and gradually increase the ammonia nitrogen load. The influent water volume in the primary aerobic tank was then increased, reaching the desired conditions after approximately one month, allowing for continuous influent testing. During operation, the MLSS in the anoxic and aerobic tanks was 4,832 mg/L, while in the MBBR tank, the suspended MLSS was 5,091 mg/L. DO levels in the anoxic tank were 3 mg/L, and in the MBBR tank, they were between 3-4 mg/L, with a pH of 7.4-7.5, promoting favorable conditions for nitrifying bacteria growth.
2. Pretreatment Effectiveness
FeSO₄ and NaHCO₃ were added to the pre-aeration tank, adjusting the pH to approximately 7.7, with monitoring of the effluent Ca²⁺ and S²⁻ concentrations. Effluent Ca²⁺ concentration was around 300 mg/L, while S²⁻ was reduced to levels that would not inhibit microbial activity in the biological treatment stage. However, the pretreatment was less effective for Ca²⁺ removal, leaving a relatively high concentration.
3. COD Removal Efficiency
The influent COD concentration was raised to 1,000 mg/L as per test requirements. Due to the two-stage aerobic setup with pre-denitrification, COD consumption was particularly high during denitrification. According to reports, a COD:TKN ratio above 6.6:1 is needed for complete denitrification; however, the experimental ratio was between 4.5-8.3, resulting in an average denitrification rate of 69%. Although influent COD was relatively high, the effluent COD concentration remained below 100 mg/L. Figure 3 illustrates the effluent COD concentration during the testing period from August to September 2007, showing concentrations between 40-80 mg/L and an average removal rate of 93.3%, meeting China's "Comprehensive Discharge Standard of Wastewater" (GB 8978-1996) Class I discharge standard.
4. Ammonia Nitrogen Removal Efficiency
During the test period from August to September 2007, the influent TKN concentration was between 120-220 mg/L, with a removal rate exceeding 95%. This process effectively removed Kjeldahl nitrogen due to the MBBR process used in the latter stage, which included both suspended and attached forms of sludge, thus increasing sludge concentration and enhancing the system's resistance to shock loads. The ammonia nitrogen load was 0.018 kg/(kg·d). However, as nighttime temperatures in September began to vary significantly compared to August, there was a slight decrease in the overall TKN removal efficiency.
5. Calcium Ion Analysis in MBBR Tank Fillers
Reports indicate that Ca²⁺ deposition inhibits nitrification. In the treatment plant's operation, calcium ion deposition on the flexible fillers in the aerobic tank inhibited microbial growth, reducing nitrification efficiency in the final aerobic tank. Since the pretreatment was ineffective for Ca²⁺ removal, necessary monitoring of Ca²⁺ in the MBBR process was conducted. Calcium content measurements were 2.13% in May, 1.89% in July, and 1.04% in September, indicating Ca²⁺ deposition on the fillers. However, due to the movable nature of MBBR fillers, deposited Ca²⁺ would automatically shed under aeration impact, preventing adverse effects on nitrification.
Conclusions
This pilot study by AquaSust retrofitted the plant by adding pre-treatment facilities in the initial stage and specializing in the MBBR process in the aerobic stage. The final data showed positive results as follows:
1. Following effective pretreatment, the effluent S²⁻ concentration was low, though Ca²⁺ removal efficiency remained low. Overall process stability was maintained, benefiting downstream biological treatment.
2. When the influent COD concentration reached 1,000 mg/L, the effluent COD remained below 80 mg/L, with an average COD removal rate of 93.3%, meeting the requirements.
3. The MBBR process consistently achieved a high Kjeldahl nitrogen removal rate, averaging over 95% with an ammonia nitrogen load of 0.018 kg/(kg·d).
4. Calcium ion monitoring in the MBBR tank fillers showed that significant deposition was prevented, avoiding adverse effects on nitrification.