In this study, we use an anaerobic-aerobic integrated denitrification (Fe/C-ZACID) device with an iron-carbon-activated carbon and zeolite composite filter to remove nitrogen from simulated low carbon-nitrogen ratio (C/N) sewage. The impacts of dissolved oxygen (DO) level, hydraulic retention time (HRT), C/N and nitrate recirculation ratio on denitrification performance were studied. The results show that when HRT was 6 h, DO was 3 ± 0.1 mg/L, influent C/N was 3, and nitrate recirculation ratio was 100%, and removal rates of 95% for ammonia and 85% for total nitrogen (TN) were achieved. A beaker comparison test demonstrated that this synergistic denitrification system included heterotrophic denitrification, physicochemical denitrification, iron autotrophic denitrification and hydrogen autotrophic denitrification, etc. The Fe/C-ZACID device has a high-efficiency nitrogen removal effect for low C/N ratio sewage and strong shock resistance, which provides technical support and a theoretical basis for advanced denitrification of rural domestic sewage.
In recent years, with the acceleration of China's rural urbanization, the consumption of water resources has increased dramatically along with the discharge of rural sewage. At the same time, due to shortages of treatment funds, residents' weak awareness of environmental protection, the imperfect drainage pipe network and centralized treatment facilities, untreated sewage is discharged via point and non-point sources, seriously polluting various waterways (Chen et al. 2016). To address the characteristics and demands of rural sewage management, such as the quantity, wide area of production, and dispersion, research has been conducted on decentralized sewage treatment technology.
Conventional treatment technologies, such as artificial wetlands, land infiltration, advanced algae pond technology, aerobic biological treatment units and membrane bioreactors, have achieved certain effects in decentralized sewage treatment (Wu et al. 2013; Nie et al. 2015). However, with the improvement of the quality of life of people in rural areas of China and the changes in people's lifestyles, the amount of N in wastewater has increased year by year, causing the C/N ratio of its water quality to decline (Gong et al. 2012). In the biological treatment process, the lack of a carbon source is very unfavourable for nitrification and denitrification. For low C/N ratio sewage with small water volumes, conventional treatment technologies struggle to achieve advanced nitrogen removal, which limits their promotion in rural areas (Cui 2013). To improve the denitrification step, traditional processes often need to add carbon sources, but the cost of adding carbon sources is high and it is easy to cause secondary pollution. In four new processes for low C/N ratio sewage treatment – the simultaneous nitrification and denitrification (SND) process (Chiu et al. 2007), the anammox process, the sulphur autotrophic denitrification process and the iron autotrophic denitrification process – the culture of anaerobic ammonium oxidizing bacteria is difficult, and additional processes need to be combined to remove chemical oxygen demand (COD) in sewage (Ali et al. 2015); sulphur autotrophic denitrification requires sulphide or sulphur as a substrate and is generally suitable only for the treatment of sulphur-containing wastewater (Chen et al. 2010). Therefore, in the face of the denitrification problem of dispersed domestic sewage with low C/N ratios, there is an urgent need for an efficient, low-cost, simple and stable effluent technology.
Therefore, this study addresses the low C/N ratio sewage prevalent in most rural areas. Based on the iron carbon internal electrolytic filter (Fe/C) previously developed by our group (Pan et al. 2010), a prototype of a Fe/C-ZAC- A/OID (iron carbon-activated carbon and zeolite composite filter material, anaerobic-aerobic integrated, denitrification) device was developed. This device has composite filter materials (iron carbon internal electrolytic filter, natural zeolite and biologically activated carbon) as the core and an A/O biological filter as the main body, and integrates aerobic and anaerobic technologies to keep wastewater in an alternating aerobic and anaerobic environment. The new [H] and Fe2+ produced by the internal electrolysis of iron-carbon filters provide electron donors for biological denitrification, thus removing the dependence of the denitrification system on carbon sources (Zhang et al. 2016), to achieve high-efficiency denitrification of sewage with a low C/N ratio under the condition of no additional carbon source. Natural zeolite (Z) has a strong selective ion exchange capacity for ammonia. Zeolite is often used for chemical treatment of ammonia in sewage. Most pollutants, such as organics and ammonia, can be degraded by microorganisms, which stably grow on the surface of AC (AC can be regenerated to some degree); other pollutants can be adsorbed by AC in filter material. AC is an efficient material for adsorption, degradation, and recycling (Daud & Houshamnd 2010; Derylo-Marczewska et al. 2011; Wang et al. 2011). Other studies found that the advanced pore structure of the two filter materials is suitable for microbial growth under the condition of aeration (Ye et al. 2017). In this study, the physicochemical effect of Fe/C is coupled with the microbial degradation of a biofilm to establish a joint denitrification system and further combined with zeolite activated carbon (ZAC), with different filter media acting in a coupled system at their highest efficiencies. The main part of the device uses the A/O process, for which the total investment cost and operation cost are relatively low (Jiang et al. 2012). Moreover, bio-filters have high biomass and good purification effects, which can improve the treatment efficiency of the reactor, thereby reducing the volume of the reactor required to treat a given area.
We have investigated and optimized various important operational parameters for the above Fe/C-ZACID device including the dissolved oxygen (DO), hydraulic retention time (HRT), C/N ratio of influent and nitrate recirculation ratio to achieve a high N-removal capacity. The device provides a promising, efficient denitrification treatment method for dispersed sewage with a low C/N ratio.
MATERIALS AND METHODS
As shown in Figure 1, the Fe/C-ZACID device combines an anaerobic zone (A), aerobic zone (O), backflow grooves and outlet grooves. The water is driven by a peristaltic pump into the lower end of the anaerobic zone, and an up-flow mode was adopted. The effluent from the upper triangular weir enters the aerobic zone, and part of the aerobic zone effluent returns to the anaerobic zone inlet. The device is made of high-density polyethylene material, 150 mm wide, 200 mm high, wall thickness 5 mm, which is composed of a left part and a right part separated by a baffle. The left side is a 200 mm long anaerobic zone, and its effective volume from the bottom to the outlet of the triangular weir is approximately 4.7 L. The right side is a 140 mm long aerobic zone, and its effective volume from the bottom to the outlet is approximately 2.8 L. Fe/C developed by our group were used along with Z and AC to make composite filters in A and O internally (according to the volume ratio of 2:1:1 and 1:1:1, respectively). The materials were filled in a packing frame and integrated, making a filter that can be easily extracted from the device for maintenance and cleaning. The distance between the packing frame and the bottom of the device is adjustable, and the height of the frame is also adjustable. In the experiment, the volume of the packing is changed by adjusting the distance and height of the filled packing layer, so that the volume of the composite packing accounts for approximately 70% to 80% of the effective volume of the device. The effective volumes of A and O after placement of the composite filter are approximately 2.0 L and 1.0 L, respectively.
Experimental water quality
The simulated wastewater samples used in this experiment were prepared artificially, with C6H12O6, NH4Cl and KNO3 plus KH2PO4 as the carbon, nitrogen and phosphorus sources of the influent, respectively. The concentration of COD and nitrogen compounds in the influent was configured according to different test design requirements. The concentrations of CODcr, NH4+-H, NO3–-N and total nitrogen (TN) of the original water were 50 ± 5, 5 ± 1, 20 ± 1 and 25 ± 2 mg/L, respectively, and the pH was 7.5 ± 0.1. To ensure the presence of trace elements necessary to microorganism growth and reproduction, 1 mL nutrient solution was added to each litre of simulated wastewater influent (Xing et al. 2017). The chemicals were all analytical grade, and the test water was ultra-pure water.
COD, TN, ammonia, nitrate and nitrite levels were determined according to the Standard Methods for the Examination of Water and Wastewater (SEPA 2002). PH was evaluated using a REX, PHS-25. The ORP and temperature were evaluated using a Ray E - 201 - C portable redox potentiometer.
RESULTS AND DISCUSSION
Impact of DO on the denitrification performance of integrated devices
According to the low C/N ratio of rural domestic sewage, the CODcr concentration in the influent water was controlled to 50 mg/L, the NH4+-N concentration was 5 mg/L, the NO3–-N concentration was 20 mg/L, and the resulting C/N was 2. The device was operated with continuous influent under normal temperature conditions, and the DO concentration in A was controlled to 1 ± 0.2 mg/L, 2 ± 0.2 mg/L, 3 ± 0.2 mg/L, 4 ± 0.2 mg/L and 5 ± 0.2 mg/L by adjusting the aeration rate. The inflow rate was 0.33 L/h, total HRT was 6 h, nitrate recirculation ratio was 100%, operating temperature was 25–30 °C, and influent pH was 7.5 ± 0.1. The device was trained by an inoculation training acclimation method to cultivate the microbial community for treating sewage and realizes combined denitrification by controlling the DO, the supply of the carbon source and the mechanical cycle. The operating parameters were changed and COD, NH4+-N, NO3–-N, NO2–N, and DO were monitored every day, then each parameter was monitored for 5 days after stable operation. The impacts of DO on the denitrification performance using low C/N ratio domestic sewage were studied.
Figure 2(a) shows that when DO < 4 ± 0.2 mg/L, ammonia consumption increases with the increase in DO. When DO was 4 ± 0.2 mg/L, ammonia consumption increased to 96.7%; at this time, the effluent ammonia concentration was 0.16 mg/L. When DO was continuously increased to 5 ± 0.2 mg/L, ammonia consumption increased as well (93.6%), indicating that DO has a great influence on the nitrification efficiency.
As seen from Figure 2(b), when DO was 1 mg/L, the average removal rate of TN was 50.8%, and the TN of the effluent was 12.56 mg/L. Combined with the nitrogen conversion in Figure 3, it can be seen that when DO was low the nitrification was very weak, the ammonia in the nitrified liquid occupied a large proportion, and the nitrified liquid reflux to the anaerobic zone had an effect on the denitrifying bacteria. When DO increased to 3 mg/L, the average removal of TN increased rapidly to 81.0%, and the TN of the effluent decreased to 4.72 mg/L, which was attributed to the internal electrolysis reaction providing electron donors for denitrifying bacteria. Oxygen was consumed because the iron carbon filter removes organic matter, which reduces the interference of DO on the anaerobic environment.
Continuing to increase DO to 4 mg/L, the improvement of TN removal was not significant, only increasing by 3.7%. When DO was 5 mg/L, the removal of TN dropped to 73.3%, and the concentration of TN rose to 6.52 mg/L. The main reason is that excess DO causes nitrate to be inferior in competing with DO as an electron acceptor. Only a small part of the iron can supply electrons to nitrates, which weakens autotrophic denitrification. In contrast, a high oxygen content in the reflux nitrifying liquid destroys the anaerobic environment, thus decreasing denitrifying bacterial activity. The aerobic effluent nitrogen transformation also indicates that the proportion of nitrite-N increases, and nitrite-N would impact the respiration and cellular proliferation of microorganisms if it accumulates to a sufficient concentration. Therefore, the optimum DO of the system is 3–4 mg/L.
Impact of HRT on the denitrification performance of integrated devices
To study the influence of different HRT on the denitrification performance of the system, the test maintained the nitrate recirculation ratio at 100%, DO at 3 ± 0.2 mg/L, operating temperature at 25–30 °C, and pH at 7.5 ± 0.1. By adjusting the influent flow, the total HRT of the reactor was 3 h, 4 h, 5 h, 6 h and 7 h. In the simulated sewage, NH4+-N = 5 mg/L, and NO3–-N = 20 mg/L. The experimental results are shown in Figure 4.
Figure 4 shows that the TN removal rate appears to first increase and then stabilise when HRT was prolonged. When HRT was 3 h, denitrification could not be fully carried out and the TN removal was just 51.7%, but the system still maintained approximately half of the removal rate of TN due to electrochemical reduction of nitrate nitrogen and autotrophic denitrification of iron and hydrogen, which permit physicochemical effects even at very short HRTs. When HRT increased to 4 h, 5 h, and 6 h, the TN removal increased to 64.5%, 77.0%, and 87.8%, respectively. During this process, the TN removal rate increased sharply. Appropriate extension of the HRT is beneficial to the growth of nitrifying bacteria, which need a longer generation time. NH4+-N is fully in contact with the biofilm and is completely oxidized, thereby prolonging the provision of a more stable hypoxic microenvironment for microorganisms on the biofilm of the filter and improving the denitrification effect of NO3–-N in the anoxic zone. When HRT increased from 6 h to 7 h, a large amount of organics consumption caused insufficiency of a carbon source for the reaction, but the system TN removal rate fluctuated only slightly, decreasing from 87.8% to 87.6%, and the effluent TN increased from 3.04 mg/L to 3.06 mg/L. In general, HRT = 6 h is the optimal operating state of the device.
Impact of influent C/N on the denitrification performance of integrated devices
With C/N ratios of 0, 1, 2 and 3, the effect of C/N ratio on the denitrification performance of the device was studied. The nitrate recirculation ratio was controlled to 100%, DO in O was 3 ± 0.2 mg/L, HRT was 6 h, and pH was 7.5 ± 0.1. The test kept the influent TN value (NH4+-N = 5 mg/L, NO3–-N = 20 mg/L) unchanged and controlled the influent COD to adjust the influent C/N by adjusting the amount of glucose added. The impact of influent C/N on the denitrification performance of the devices is shown in Figure 5.
Figure 5 shows that the TN removal rate increased as influent C/N increased. When C/N = 0, the TN removal rate was only 32.8%; when C/N was increased to 1 and 2, the TN removal rate increased to 62.5% and 85.3%, respectively, and the effluent TN values were 6 mg/L and 5 mg/L, respectively. When C/N increased to 3, the TN removal rate continued to increase, and the average TN removal rate reached 90.3%. At this time, the effluent TN was 4 mg/L.
Past studies have shown that when the carbon source is sufficient, a large number of aerobic denitrifying bacteria exist, and the denitrification rate is fast. When the carbon source is insufficient, autotrophic denitrifying bacteria are mainly present, and these autotrophic denitrifying bacteria grow slowly, affecting denitrification. The denitrification effect is not good. The traditional sewage treatment technology requires influent C/N in excess of 6 to 7 to meet the carbon source demand (Virdis et al. 2010). In the Fe/C-ZACID device, using the composite filter, the C/N only needs to reach 2 to achieve a good TN removal effect because the denitrifying bacteria in the system first use organics for heterotrophic denitrification to denitrification. When the organic carbon source is insufficient, Fe2+ and H2 produced by elemental iron or microelectrolysis are used as electron donors for iron or hydrogen autotrophic denitrification. In addition, iron ions can promote the growth of microorganisms, and zeolite can accelerate the progress of nitrification and increase the efficiency of denitrification, thereby achieving the purpose of effectively removing nitrogen from low C/N ratio sewage.
Impact of the nitrate recirculation ratio on the denitrification performance of integrated devices
To study the influence of the nitrate recirculation ratio on the denitrification performance of the system, the test kept DO at 3 ± 0.2 mg/L with an HRT of 6 h, operating temperature of 25–30 °C, and pH of 7.5 ± 0.1. The recirculation ratios were 0%, 50%, 100%, 150%, and 200%. In the simulated sewage, NH4+-N = 5 mg/L and NO3–-N = 20 mg/L. The experimental results are shown in Figure 6.
Figure 6(a) shows that the recirculation ratio has little effect on the ammonia removal rate of the whole system. When the recirculation ratio increased from 0% to 200%, the ammonia removal rate exceeded 90%.
It can be seen from Figure 6(b) that the nitrate recirculation ratio had a significant effect on the TN removal rate. The TN removal rate appears to first increase and then decrease with the increase in the nitrate recirculation ratio. When the recirculation ratio was increased from 0% to 200%, the corresponding TN removal rates of the system were 41.55%, 61.18%, 81.21%, 82.81%, and 69.85% at each studied stage. When the recirculation ratio was 100% and 150%, the system TN removal rate reached the highest level; the average effluent TN was 4.82 mg/L and 4.32 mg/L, respectively.
This pattern can be attributed to the fact that with a recirculation ratio below 150%, the nitrifying liquid flows back to the anaerobic zone as the recirculation ratio increases, resulting in the increase of nitrate concentrations in the influent mixture, which improves the TN removal rate. When the recirculation ratio is increased to the critical value, the nitrate concentration exceeds the denitrification capacity of the anaerobic zone, and a continuous increase in the recirculation ratio does not affect the nitrate removal effect. However, the increase in the recirculation ratio would shorten the HRT of the aerobic unit, making the contact time between ammonia and the biofilm in the sewage insufficient, which is not conducive to nitrification. At the same time, this increased ratio would inhibit the progress of denitrification. More DO would be brought into the anaerobic zone due to the increase of the recirculation ratio. Additionally, the actual influent water concentration (especially the organics concentration) is diluted by reflux, which is not conducive to heterotrophic denitrification. When the recirculation ratio increased from 100% to 150%, the TN removal rate increased slightly. With an analysis of the process operation and removal efficiency, the preferred recirculation ratio of the device was 100%.
Verification of the biofilm coupling effect of the composite filter
After the system was stably operated for a period of time under the optimal operating parameters obtained in the above test, the filter with the biofilm growing well in the device was taken out, and 3 aliquots (500 mL each) were placed in a 1000 mL beaker, which were recorded as group 1, group 2, and group 3. Group 1 bio-filter was not irradiated, as a positive control. The biofilm on the group 2 bio-filter was stripped, placed in the beaker and removed, and the biofilm on the group 3 bio-filter was stripped and placed in a high-pressure steam sterilization pot. The autoclave was sterilized at 105 °C and 1.5 MPa for 30 min to kill the microorganisms remaining on the surface of the filter.
The same amount of a simulated water sample was added to the beaker, the oxygen pump was adjusted to control the DO level, and the shaker was sealed and started. The tested running times were 0 min, 10 min, 30 min, 60 min, 120 min, and 240 min, with continuous monitoring of nitrate nitrogen and TN.
Comparing the changes of nitrate nitrogen and TN concentration in the three groups of beakers under different treatment conditions, it can be seen from the figure that the nitrate nitrogen and TN concentrations in group 1, group 2, and group 3 gradually decreased as time passed (Figure 7). The basic tendency is to decrease first and then stabilize. The TN removal by group 1 was better than that of group 2 and group 3. When the test lasted for 300 min, the TN removal rates of group 1, group 2, and group 3 were 79.4%, 36.5%, and 29.1%, respectively, and the nitrate removal rates were 84.2%, 44.3%, and 79.4%, respectively.
The experimental results show that the TN removal rate by the composite filter-coupled biofilm system is higher than that of the single nitrogen-removal system (13.69 > 8.37 > 5.25), which is also higher than the simple superposition of biochemical action and the filter physicochemical effect on TN removal (13.69 > 10.62). The device is a combined denitrification system, including heterotrophic denitrification, physicochemical denitrification, iron autotrophic denitrification and hydrogen autotrophic denitrification. In addition, some studies have pointed out that smaller n is associated with a lower dependence of the reaction on the substrate concentration and greater shock resistance (Wang & Hu 2006). The calculation results show that the shock resistance capacity of the device is significantly higher than that of the single-filter physicochemical denitrification and biochemical denitrification, which also indicates synergy between physicochemical and biochemical effects in this system.
In summary, the integrated device can obtain a high nitrogen-removal efficiency at a low C/N ratio and is suitable for efficient nitrogen removal from low C/N ratio sewage.
In this paper, a Fe/C-ZACID device was developed for high-efficiency nitrogen removal from wastewater with a low C/N ratio and small water volumes. The development of this device provides technical support and a theoretical basis for advanced denitrification of rural domestic sewage in China. In the study of the device, we found the following:
The optimal control parameters for this device are a HRT of 6 h, a DO of 3 ± 0.1 mg/L, an influent C/N of 3, and a nitrate recirculation ratio of 100%. When the influent ammonia is 5 mg/L, TN is 25 mg, and the system is operating in the optimal operation mode, the effluent ammonia concentration is basically maintained at approximately 0.2 mg/L, TN is reduced to 5 mg/L, and the ammonia and TN removal rates can reach 95% and 84%, respectively.
The beaker comparison test demonstrates that the device is a combined denitrification system including heterotrophic denitrification, physicochemical denitrification, iron autotrophic denitrification, and hydrogen autotrophic denitrification, etc. There are synergistic effects between physicochemical and biochemical actions.
This device has the advantages of good water quality, strong shock resistance, simple cleaning and maintenance, reduced space and operation costs, and convenient installation and removal. This research provides technical support and a theoretical basis for advanced denitrification of rural domestic sewage in China.
Finally, we found that HRT is essential for simultaneous nitrification and denitrification and the guarantee of efficient nitrogen removal. However, in this study, only the total HRT of the device was tested. The volume ratio of the anaerobic filter and the aerobic filter directly affects the HRT of each zone, which in turn affects the nitrification and denitrification reactions in the nitrogen-removal process. The effective volume ratio of anaerobic and aerobic filters should be altered to further optimize the process parameters of the device.
The authors acknowledge the financial support of the National Science Foundation of China (No. 20676079) and the National Science and Technological Foundation of China (No. 2006BAJO4A09).