The influence of alkalinity on the two stages of S0-based autotrophic denitrification

In this study, two S0 particle packed reactors were constructed to explore the influence of alkalinity content on S0-based autotrophic denitrification (SAD) with nitrate and nitrite as electron acceptors respectively. As a result of the comparative experiment, when the alkalinity of influent is sufficient (HCO3 -: NO3 - mole ratio > 1.5:1), NO3 --N and NO2 --N reactor both showed high removal efficiency. But when the HCO3 -: NO3 - mole ratio decreased to 0.2:1, the average removal rate of NO3 --N decreased to 31%, and NO2 --N average removal rate still maintained at 95%. Based on the results, the reduction of alkalinity can significantly reduce the nitrate removal efficiency, but has little effect on the nitrite removal efficiency. This study further explained the characteristics of alkalinity consumption of the two stage of SAD and may provide a reference for future research.


Introduction
Nitrogen pollution is one of the main problems in water environment, and also one of the main factors causing eutrophication of water body. The effluent from wastewater treatment plant is one of the main sources of nitrogen pollutants in water environment system, and nitrate is the main component of nitrogen pollutants. Therefore, the removal of nitrate in wastewater treatment plant is very important for the improvement of water environment quality Heterotrophic denitrification is the main technology for nitrate removal in wastewater treatment, but its effect is affected by the content of organic carbon, especially the wastewater with low C/N ratio need to add methanol, ethanol and acetate, etc. as electron donors [1]. It often has problems of expensive cost, overdoing concern and high sludge yields. Autotrophic denitrification use inorganic matter (such as S 0 , S 2-, S 2 O 3 2-, SO 3 2-, H 2 , Fe 2-) as electron donor instead of organic matter to reduce the nitrate to nitrogen can effectively solve this problem [2,3]. Elemental sulfur (S 0 ) based autotrophic denitrification is favored because it is nontoxic, odorless, and does not have excessive dosage problems as a solid. SAD was usually used in the form of fixed bed for nitrogen removal of low-carbon industrial wastewater and deep denitrification treatment of tail water in municipal sewage treatment plant. The stoichiometry of SAD is as shown in the following equation [ It can be seen from the above that removal of 1mol NO 3 will produce 1.28mol H + , which means that the process requires a lot of alkalinity. It is also expensive. To solve this problem, limestone was usually used to provide alkalinity mixed with S 0 particle in packed bed reactor, because of its low cost and availability [5,6]. However, the addition of limestone occupied a part of the reactor volume, which reduced the denitrification efficiency. In addition, the dissolution of limestone would lead to an increase in the concentration of calcium ions, resulted in scaling of the tank walls and pipes. In view of this, some studies constructed the mixotrophic denitrification system by adding amount of organic carbon source such as liquid organic matter, wood-chip and corn core in the SAD system to reduce alkalinity consumption and sulfate production [7][8][9]. In the mixed system of autotrophic and heterotrophic denitrification, no matter what form of organic carbon source was added, the biomass yield would be higher than that of the elemental sulfur autotrophic denitrification system alone. At the same time, it might lead to the conversion of sulfate to hydrogen sulfide gas under anaerobic environment, causing the odor problem due to the existence of organic matter. SAD process consists of two stages, nitrate is first reduced to nitrite, and then reduced to nitrogen [10]. These studies usually focused on the overall effect of the system and the microbial composition and rarely studied the alkalinity consumption characteristics of the two stages of the SAD separately.
In this study, we constructed two S 0 particle packed reactor to explore the influence of alkalinity dosage on the two stages of denitrification process respectively. The results can further explain the characteristics of alkalinity consumption and provide a reference for future research.

Set-up of SAD reactor
The experimental set-up of SAD is shown in Fig.1. Two continuous upflow fixed bed cylindrical reactors (R1, R2) packed with 3-5mm diameter elemental sulfur particles were used in this study. The effective volume of the reactor was 1L with 50mm diameter and 500mm high. Each reactor was inoculated with 100mL activated sludge from a municipal sewage treatment plant at a concentration of about 3500 mg/L, and the empty bed contact time (EBCT) was set to 1h. The reactor was operated at room temperature (24~28℃) without additional insulation. 5 sampling points were uniformly set along the direction of height on two reactors.

Wastewater composition
Artificial synthetic wastewater was used in this study, which was composed of NO 3 --N (R1, 50mg/L) and

Analytical methods
The NO 3 --N (Ultraviolet spectrophotometry method, HITACHI U-5100) and NO 2 --N (Ultraviolet spectrophotometry method, HITACHI U-5100) concentrations in the influent and effluent were measured every two days. SO 4 2was measured using ion chromatography method (Thermo Fisher AQ-1100), and pH was measured using pH meter (METTLER, S2). In addition, samples along the reactor were obtained three times through the sampling point on the reactor at 79~100d.

The influence of alkalinity on SAD with nitrate as electron acceptor
In Fig.1, NO 3 --N removal effect decreases significantly when the alkalinity of influent decreases. The initial average NO 3 --N removal rate is 90% when HCO 3 -: NO 3 mole ratio in influent is 1.5:1. Then as the HCO 3 -: NO 3 mole ratio decreases to 0.2:1, average removal rate of NO 3 --N decreases to 31%. However, it is worth noting that the nitrate removal rate recovered immediately when HCO 3 -: NO 3 mole ratio is increased from 0.2:1 to 1.5:1 straightly again. The decrease in alkalinity of influent does not result in significant accumulation of nitrite, except during the initial start-up of reactor. It can be seen that in the process of nitrate reduction, whether the alkalinity is excessive or not may not cause the accumulation of nitrite. And the complete removal of nitrate of reactor requires a minimum HCO 3 -: NO 3 mole ratio of 1.5:1 in influent.

The influence of alkalinity on SAD with nitrite as electron acceptor
The relationship between the removal effect of NO 2 --N and the change of alkalinity in influent is shown in Fig.2. It can be seen that the reduction of alkalinity in influent within a certain range has little effect on NO 2 --N removal effect. When the mole ratios of HCO 3 -: NO 3 are 1.5:1 and 1:1, NO 2 --N removal rate reaches to nearly 100%. Although the removal effect of NO 2 --N has a slight reduction as the HCO 3 -: NO 3 mole ratio decreases to 0.5:1 and 0 removal rate recovers to almost 100% immediately when HCO 3 -: NO 3 mole ratio increases from 0.2:1 to 1.5:1 straightly. During the whole operation, the lowest nitrite removal rate is 83% occurs when HCO 3 -: NO 3 mole ratio changes from 0.5:1 to 0.2:1, and then increases rapidly. It can also be seen that complete removal of nitrite requires little or no alkalinity as long as the appropriate pH is maintained. Fig. 3. Nitrite concentration of R2 (SAD with nitrite as electron acceptor)

The characteristics of SAD process with excess alkalinity
When the HCO 3 -: NO 3 mole ratio in influent is 1.5:1, the concentrations of nitrate, nitrite and sulfate change along R1 is shown in Fig. 4. Nitrate content decreases gradually with the increase of time along the reactor, while the sulfate concentration increases. The mass ratio of SO 4 2--S increment to NO 3 --N decrement is 1.94 nearly to the theoretical value. Only a small amount of nitrite appears at the first sampling point in the entire reactor and then reduces to nearly zero. And it also can be seen that the reaction rate of nitrate decreases with the increase of reactor height, which may be reasons for the decrease of substrate concentration, alkalinity and pH [11]. And Nitrate is mainly removed in the first 40 minutes of the reactor, and the removal rate is about 90%. 2--S increase to NO 2 --N decrease is 1.18 nearly to the theoretical value. It's worth noting that the reduction rate of NO 2 --N in the first half reactor is much higher than in the second half, so that 95% nitrite was removed in the first half an hour. Combined with the results of Fig.4 and Fig.5, it can be seen that the removal rate of nitrite of R2 is significantly higher than that of nitrate in R1. And removing the same amount of NO 3 --N produces 1.6 times as much sulfate as removing the same amount of NO 2 --N by comparing two reactors.

Conclusions
The effect of alkalinity content in influent on the two stages of SAD is very different. The reduction of alkalinity can significantly reduce the nitrate removal efficiency, but has little effect on the nitrite removal efficiency. When the HCO 3 -: NO 3 mole ratio decreases to 0.2:1, the average removal rate of NO 3 --N decreases to 31%, but the average removal rate of NO 2 --N can still maintain at 95%. And the decrease of alkalinity may also reduce the reaction rate significantly.