Abstract

This study investigated the influence of the unique internal recirculation characteristics of an oxidation ditch (OD) system, namely, the internal recirculation frequency (IRF) on denitrifying phosphorus removal (DNPR). The ratios of denitrifying polyphosphate-accumulating organisms (DPAOs) to polyphosphate-accumulating organisms (PAOs) under different IRF conditions were measured using a batch experiment. On this basis, the variation of nutrient transformations was studied using the IRF changes by the mass balance method. The results showed that, for the OD system that had an anaerobic zone upstream from the circular corridor and set anoxic and aerobic zones along the circular corridor, when the IRF was between 3.4 h−1 and 7.5 h−1, the DPAOs/PAOs ratio reached about 50%. Approximately 20% of the total phosphorus (TP) was removed and over 11% of the total nitrogen (TN) was transformed into nitrogen gas by the DNPR process, and meanwhile the total removal efficiencies of the TP and TN were over 93% and 80%. When the IRF was greater than 11.5 h−1, the TN removal efficiency decreased significantly, and this was not conducive to simultaneous nitrogen and phosphorus removal. The results indicated that the OD process would possess a better DNPR potential if the IRF were controlled within the proper scope.

INTRODUCTION

Oxidation ditches (ODs) are a type of sewage treatment technology widely used in worldwide wastewater treatment plants (WWTP). By the end of 2014, 4,436 wastewater treatment facilities were put into operation in China, of which 1,161 are using OD, accounting for 26.2% of all treatment facilities (MEE 2015).

Initially, OD systems were only used for organic matter removal; however, due to more strict environmental protection requirements, the simultaneous removal of nitrogen, phosphorus, and organic matter is required nowadays. By setting an anaerobic zone upstream from the circular corridor (Insel et al. 2005), and spatially distributing the aerobic and anoxic zones along the circular corridor, simultaneous nitrogen and phosphorus removal can be achieved in an OD system.

With the development of OD technology, many aspects have been researched to comprehend the biochemical reaction processes and to improve nutrient removals. For example, Insel et al. (2005) showed that the level of nitrate returned into the anaerobic reactor affected the enhanced biological phosphorus removal. Alaya et al. (2010) determined that the influent flow rate and mixed liquor flow velocity affected the dissolved oxygen (DO) and oxygen uptake rates, and the bio-oxidation, nitrification, and denitrification processes were affected by the irregular DO distribution. Hou et al. (2009) installed an anaerobic and anoxic zone upstream from the circular corridor and found there was denitrifying phosphorus removal (DNPR) in the anoxic zone. Chen et al. (2012) presented a dual DO control technology to provide outstanding nutrient removal efficiencies even with large influent fluctuations. Ratanatamskul & Kongwong (2017) established the inclined tube/OD-membrane bioreactor for a building's wastewater recycling. Ammary & Radaideh (2005) and Zhou et al. (2015) found that simultaneous nitrification and denitrification (SND) existed in a full-scale OD plant. Liu et al. (2010) studied the operational conditions of SND in laboratory- and pilot-scale OD systems using real domestic wastewater, and the result showed that the low optimal DO, higher ratio of organic matter to ammonia, and shorter hydraulic retention time (HRT) were propitious to SND. Qiu et al. (2018) proposed an integral energy model including the energy, aeration, and fluidic effects of surface aerators, by which the energy for aeration of each aerator could be estimated using online data. Li et al. (2019) studied the start-up and nitrogen removal performance of CANON (completely autotrophic nitrogen removal over nitrite) and SNAD (simultaneous partial nitrification, anaerobic ammonium oxidation, and denitrification) processes in a pilot-scale OD system.

However, the DNPR process has rarely been studied in an OD system, which can achieve simultaneous nitrogen and phosphorus removal, and save both organic carbon sources and aeration energy demand, and also reduce the sludge production (Kuba et al. 1996; Meinhold et al. 1999; Hou et al. 2009). A mixed liquor circulates in the closed-loop corridor ceaselessly in an OD system, and this is a significant difference from other wastewater treatment processes; however, the effect of the internal recirculation characteristics of OD on DNPR is seldom explored.

The internal recirculation frequency (IRF) and circulatory period (T) can be used to quantify the internal recirculation characteristics of OD systems. T is the time consumed by the mixed liquor flowing during one lap in the circular corridor, and the IRF is the laps of mixed liquor flowing in the circular corridor during 1 hour. Moreover, IRF and T are reciprocals of each other. However, IRF and T have not been obligatory parameters for the design and manipulation of OD systems so far, leading to a dramatic difference in the IRF and T of the existing worldwide ODs. The IRF and T of some full-scale OD systems in different WWTPs are listed in Table 1.

Table 1

IRF and T of some full-scale OD systems

Wastewater treatment plantsIRF (h−1)T (h)
WWTP in Honjo, Akita Prefecture, Japan (Kanazawa & Urushigawa 200717.5 0.06 
Mahres WWTP in Tunisia (Alaya et al. 201014.9 0.07 
Zuoyunxian WWTP in Shanxi province, China (Chen 200410.9 0.09 
Carrousel WWTP for domestic sewage of Xinglong Zhuang coal mine in Shandong province of China (Zhou et al. 20044.71 0.21 
Carrousel WWTP in Rotterdam, the Netherlands (Abusam et al. 20042.7 0.38 
Handan West WWTP in Hebei province of China (Chen 20120.90 1.11 
Wastewater treatment plantsIRF (h−1)T (h)
WWTP in Honjo, Akita Prefecture, Japan (Kanazawa & Urushigawa 200717.5 0.06 
Mahres WWTP in Tunisia (Alaya et al. 201014.9 0.07 
Zuoyunxian WWTP in Shanxi province, China (Chen 200410.9 0.09 
Carrousel WWTP for domestic sewage of Xinglong Zhuang coal mine in Shandong province of China (Zhou et al. 20044.71 0.21 
Carrousel WWTP in Rotterdam, the Netherlands (Abusam et al. 20042.7 0.38 
Handan West WWTP in Hebei province of China (Chen 20120.90 1.11 

Therefore, this study aimed to investigate the influence of the unique internal recirculation characteristics of an OD system, such as the IRF, on the DNPR process.

The ratio of denitrifying polyphosphate-accumulating organisms (DPAOs) to polyphosphate-accumulating organisms (PAOs) was measured under different IRF conditions in a laboratory-scale OD system where the anaerobic zone was installed upstream from the circular corridor, and the anoxic and aerobic zones were set along the circular corridor. Then the contribution of the DNPR to nutrient removal under different IRF conditions was analysed using the mass balance method, and the particular IRF scope for enhancing the DNPR was obtained.

MATERIALS AND METHODS

Apparatus and methods

Laboratory-scale OD system

As shown in Figure 1, the laboratory-scale OD system consisted of a main ditch and a sedimentation tank with working volumes of 100 L and 23.6 L, respectively. There were three anaerobic zones (a total of 15 L) ahead of the circular corridor (85 L) in the main ditch, and anoxic and aerobic zones were set in the circular corridor. Submerged mixers were placed in the anaerobic zones and corners of the circular corridor to impel the mixed liquor, and fine bubble aerators were installed in the aerobic zones. Two flashboards were set up in front of anoxic zone 1 and aerobic zone 1, respectively, in order to manipulate the IRF by changing the opening area of the flashboards. The arrows show the flow orientations of the wastewater, mixed liquor, and returned activated sludge in the system.

Figure 1

Schematic diagram of the laboratory-scale OD system.

Figure 1

Schematic diagram of the laboratory-scale OD system.

Synthetic wastewater and activated sludge

Steady state conditions are necessary for mass balance analyses in activated sludge systems (Mantziaras & Katsiri 2011; Chen et al. 2012). As the raw wastewater quality varies with time in full-scale WWTPs, synthetic wastewater was used in this study.

The synthetic wastewater contained 400 mg COD/L, 40 mg TN/L, and 7 mg TP/L (COD, chemical oxygen demand; TN, total nitrogen; TP, total phosphorus). The composition of the synthetic wastewater was (per L) 102.5 mg CH3COONa (80 mg COD), 135 mg starch (160 mg COD), 75 mg glucose (80 mg COD), 112 mg peptone (80 mg COD, 5 mg N), 133.75 mg NH4Cl (35 mg N), 30.71 mg KH2PO4 (7 mg P), 57.31 mg KCl (30 mg K), 100 mg MgSO4 (20 mg Mg, 26.67 mg S), 55.5 mg CaCl2 (20 mg Ca), 550 mg NaHCO3, and 0.2 mL trace element solution.

The trace element solution contained (per L) 1.5 g H3BO3, 1.5 g CoCl2 · 6H2O, 0.3 g CuSO4 · 5H2O, 16.1 g FeCl3 · 6H2O, 1.8 g KI, 1.2 g MnCl2 · 4H2O, 0.6 g Na2MoO42H2O, 1.5 g NiCl · 6H2O, 1.2 g ZnSO4 · 7H2O, and 15 g EDTA.

The seed sludge was acquired from a municipal WWTP in Tianjin City.

Operation conditions

The continuous influent flow and returned activated sludge flow were automatically controlled at a respective constant flow rate by individual peristaltic pumps and six phases were arranged in the experiment.

The non-variable operational conditions in the six phases were as follows: the influent flow rate (QIn), the external return activated sludge ratio (RS), HRT, the solids retention time (SRT), and the discharge flow of the excess activated sludge (QEx). The mixed liquor suspended solids (MLSS), sludge loading rate (Lθ), temperature of the mixed liquor (Tθ), and the DO concentrations in different zones are listed in Table 2.

Table 2

Operation conditions in the laboratory-scale OD system

PhaseIIIIIIIVVVI
IRF 0.5 1.6 3.4 7.5 11.4 18.7 
Air supply methods Strip aeration Point aeration 
Air supply quantity (L/min) 10 
QIn (L/d) 240 
RS (%) 80 
HRT (h) 10 (in the main ditch) 
QEx (L/d) 8 (discharging mixed liquor from the outflow of main ditch) 
SRT (d) 12.5 
MLSS (mg/L) 3,600–4,000 
Lθ (kg COD/ (kg MLSS·d)) 0.24–0.27 
Tθ (°C) 20–25 
Tθ (°C) 20–25 
DO (mg/L)       
 Anaerobic zones (1–3) 
 Anoxic zones (1–4) 0–0.5 
 Aerobic zone (1) 1.0–1.5 
 Aerobic zone (2) 1.5–2.0 
PhaseIIIIIIIVVVI
IRF 0.5 1.6 3.4 7.5 11.4 18.7 
Air supply methods Strip aeration Point aeration 
Air supply quantity (L/min) 10 
QIn (L/d) 240 
RS (%) 80 
HRT (h) 10 (in the main ditch) 
QEx (L/d) 8 (discharging mixed liquor from the outflow of main ditch) 
SRT (d) 12.5 
MLSS (mg/L) 3,600–4,000 
Lθ (kg COD/ (kg MLSS·d)) 0.24–0.27 
Tθ (°C) 20–25 
Tθ (°C) 20–25 
DO (mg/L)       
 Anaerobic zones (1–3) 
 Anoxic zones (1–4) 0–0.5 
 Aerobic zone (1) 1.0–1.5 
 Aerobic zone (2) 1.5–2.0 

The variable operational condition during the six phases was the IRF. Furthermore, the air supply was adjusted to maintain a nearly constant DO concentration during the six phases. In phases I, II, III, and IV, the air was supplied by uniform micro pore aeration along the aerobic channel. In phases V and VI, air was supplied by point aeration in front of each aerobic channel.

Experimental design

During each phase, after the laboratory-scale OD system was stably operated for 25–30 days (stable operation period), the wastewater and sludge samples were drawn and analysed for 8–12 days (sampling-analysis period).

During the stable operation and sampling-analysis period, the Tθ, DO, and MLSS were detected daily, and the COD, TN and TP of the influent and effluent, and the nitrate nitrogen (NO3-N) of the effluent were measured once every 2 days.

During the sampling-analysis period, the COD, ammonium (NH4+-N), and TP in different zones of the OD system were measured for homogeneity in the wastewater quality. At the same time, the following parameters were measured for the mass balance: the mixed liquor volatile suspended solids (MLVSS) in the main ditch, the concentrations of COD, TN, and TP in the influent wastewater and effluent wastewater (sedimentation tank outflow), and the concentrations of TP in anaerobic zone 3.

Meanwhile, the batch experiments to acquire the proportion of DPAOs activity to PAOs activity (DPAOs/PAOs) were carried out during the sampling-analysis period in each phase.

Batch experiment

To determine the DPAOs/PAOs ratio, Wachtmeister et al. (1997) and Meinhold et al. (1999) used nitrate as the electron acceptor for phosphorus removal. However, in view of the fact that existing nitrate respiration microorganisms can only reduce nitrate to nitrite and cannot produce gaseous nitrogen further, the existing nitrate respiration microorganisms will result in an overestimation of the DPAOs population (Hu et al. 2003). Therefore, according to Hu et al. (2003), nitrite can be used as electron acceptor to evaluate the DPAO contributions only from the denitrifiers and excludes the influence of the nitrate respiring organisms.

Mixed liquors were collected from anaerobic zone 3 where the sludge had completed phosphorus release. After static settling, the supernatant was discarded, and the activated sludge was rinsed three times with ultra-pure water. The washed sludge was divided into three 1.5 L groups that were placed into conical flasks and used for the aerobic phosphorus uptake, anoxic phosphorus uptake with nitrate, and anoxic phosphorus uptake with nitrite. The initial phosphorus concentration was 30 mg P/L in each conical flask. During the aerobic phosphorus uptake test, oxygen was continuously supplied and controlled at DO >4 mg/L. The second group and the third group of anoxic phosphorus uptake experiments with nitrate and with nitrite had initial NO3-N and nitrite nitrogen (NO2-N) concentrations of 15 mg/L. The temperature was 25 ± 1 °C, and the pH value ranged between 7.2 and 7.5. After 10 min of reaction, each group was sampled and the phosphorus concentration was analysed.

The removed phosphorus amounts were measured and then the aerobic phosphorus uptake rate (O-PUR), anoxic phosphorus uptake rate with nitrate (N-PUR), and the anoxic phosphorus uptake rates with nitrite (N-PUR′) were calculated.

During each phase, the batch experiment was repeated three times. The values of O-PUR, N-PUR, and N-PUR′ were averaged from the three experiments during each phase.

Analytical methods

The DO and Tθ were measured in situ using a portable DO meter (WTW340i). The MLSS, MLVSS, COD, NH4+-N, NO3-N, TN, and TP were analysed according to Chinese standard methods (CEPB 2002).

Nutrients mass balance

Hypothesis

It was hypothesised that in the OD system aerobic oxidation of organics, traditional denitrification, aerobic phosphorus removal, and DNPR were occurring. All the above biochemical reaction processes promote microbial growth and produce carbon dioxide (CO2), and, moreover, traditional denitrification and DNPR produce nitrogen gas (N2). The amount of organic matter, nitrogen, and phosphorus in the daily discharged excess activated sludge was equal to the amount used by the daily growth of microorganisms.

It was assumed that the nutrient concentration in the outflow of the circular corridor was equal to that of the sedimentation tank effluent.

The mass, concentration, and flow rates are denoted by the letters M, C, and Q, respectively.

Phosphorus mass balance

The overall phosphorus mass balance is shown in Equation (1) 
formula
(1)
where MTP,In was the quantity of TP in the influent which could be calculated by Equation (2), MTP,Eff was the quantity of TP in the effluent which could be calculated by Equation (3), MCELL_TP,Ex was the quantity of TP used for bacterial growth and discharged in the form of excessive activated sludge which could be calculated by Equation (4), Mdenphosre_TP,Ex was the quantity of TP removed by DNPR and discharged in the form of excessive activated sludge which could be calculated by Equation (5), and Maerophosre_TP,Ex was the quantity of TP removed by aerobic phosphorus removal and discharged in the form of excessive activated sludge which could be calculated by Equation (6). 
formula
(2)
where CTP,In was the measured concentration of TP in the influent presented in Table 3. QIn was the influent flow rate of 240 L/d. 
formula
(3)
where CTP,Eff was the measured concentration of TP in the sedimentation tank effluent presented in Table 3. QEff was the effluent flow rate of 240 L/d. 
formula
(4)
where fP was the amount of phosphorus in each mg of MLVSS using the recommended value of 0.02 mg P/mg VSS (Henze et al. 2008), MLVSS was obtained from Table 3, and QEx was the discharge flow of excess activated sludge at 8 L/d. 
formula
(5)
 
formula
(6)
where γ was the effective DPAOs/PAOs ratio. The values of DPAOs/PAOs ratios for each IRF were measured by batch experiment. However, the DPAOs and PAOs did not work in all regions. The DPAOs only worked in anoxic zones, and the PAOs only worked in aerobic zones. Meanwhile, it was considered that all of the PAOs performed aerobic phosphorus uptake, and only some of the PAOs implemented anoxic phosphorus uptake; therefore, the effective DPAOs/PAOs ratio (γ) was used during the mass balance and was calculated using the following formula: 
formula
(7)
where 0.48 was calculated by VAno/(VAno + VAer), namely VAno/(VAno + VAer) = 0.48, VAno was the total volume of the anoxic zones, and VAer was the total volume of the aerobic zones.
Table 3

IRF and T in the laboratory-scale OD system

PhaseIRF (h−1)T (h)T (min)Taer (min)Tano (min)
0.5 1.93 115.9 60.3 55.6 
II 1.6 0.62 36.9 19.2 17.7 
III 3.4 0.30 17.9 9.3 8.6 
IV 7.5 0.13 7.9 4.1 3.8 
11.4 0.09 5.3 2.8 2.5 
VI 18.7 0.05 3.2 1.7 1.5 
PhaseIRF (h−1)T (h)T (min)Taer (min)Tano (min)
0.5 1.93 115.9 60.3 55.6 
II 1.6 0.62 36.9 19.2 17.7 
III 3.4 0.30 17.9 9.3 8.6 
IV 7.5 0.13 7.9 4.1 3.8 
11.4 0.09 5.3 2.8 2.5 
VI 18.7 0.05 3.2 1.7 1.5 

Tano and Taer are the residence times of the mixed liquor flowing through the anoxic zones and aerobic zones, respectively, during one lap.

Nitrogen mass balance

The overall nitrogen mass balance is shown in Equation (8) 
formula
(8)
where MTN,In was the quantity of TN in the influent which could be calculated by Equation (9), MTN,Eff was the quantity of TN in the effluent which could be calculated by Equation (10), MCELL_TN,Ex was the quantity of TN used for bacterial growth and discharged in the form of excessive activated sludge which could be calculated by Equation (11), Mdenphosre_N2,Air was the quantity of N2 produced by anoxic phosphorus uptake and released into the air which could be calculated by Equation (12), Mtraden_N2,Air was the quantity of N2 produced by traditional denitrification processes and released into the air which could be calculated by Equation (13). 
formula
(9)
where CTN,In was the measured concentration of TN in the influent and was obtained from Table 3. 
formula
(10)
where CTN,Eff was the measured concentration of TN in the sedimentation tank effluent obtained from Table 3. 
formula
(11)
where fN was the amount of nitrogen in each mg of MLVSS using the recommended value of 0.1 mg N/mg VSS (Barker & Dold 1995; Lee et al. 2007). 
formula
(12)
where CTP,Ana was the measured concentration of TP in anaerobic zone 3 which could be obtained from Table 3, Rs was the external return activated sludge ratio of 0.8, α was the amount of N2 produced for each mg of phosphorus uptake by the DNPR according to the stoichiometric coefficient in ASM2D (Henze et al. 2000), which was taken as 0.28 mg N/mg P. 
formula
(13)

COD mass balance

The overall COD mass balance is shown in Equation (14) 
formula
(14)
where MCOD,In was the quantity of COD in the influent which could be calculated by Equation (15), MCOD,Eff was the quantity of COD in the effluent which could be calculated by Equation (16), MCELL_COD,Ex was the quantity of COD used for bacterial growth and discharged in the form of excessive activated sludge which could be calculated by Equation (17), Mdenphosre_CO2,Air was the quantity of CO2 produced by anoxic phosphorus uptake and released into the air which could be calculated by Equation (18), Maerophosre_CO2,Air was the quantity of CO2 produced by aerobic phosphorus uptake and released into the air which could be calculated by Equation (19), Mtraden_CO2,Air was the quantity of CO2 produced by traditional denitrification processes and released into the air which could be calculated by Equation (20), and Maeroxida_CO2,Air was the quantity of CO2 produced by organic aerobic oxidation and released into the air which could be calculated by Equation (21). 
formula
(15)
where CCOD,In was the measured concentration of COD in the influent obtained from Table 3. 
formula
(16)
where CCOD,Eff was the measured concentration of COD in the sedimentation tank effluent obtained from Table 3. 
formula
(17)
where fCV was the amount of COD in each mg of MLVSS using the recommended value of 1.42 mg NCOD/mg VSS (Henze et al. 2008). 
formula
(18)
where β was the amount of CO2 produced (recorded in equivalent oxygen) for each mg of phosphorus removed by the DNPR according to the stoichiometric coefficient in ASM2D (Henze et al. 2000) of 0.8 mg O/mg P. 
formula
(19)
 
formula
(20)
where δ was the amount of CO2 produced (recorded in equivalent oxygen) for each mg of N2 produced from traditional denitrification processes according to the stoichiometric coefficient in ASM2D (Henze et al. 2000) of 2.81 mg O/mg N. 
formula
(21)

RESULTS AND DISCUSSION

Measurement of DPAOs/PAOs ratio

As shown in Figure 2, N-PUR′ was similar to N-PUR, indicating that there were no nitrate respiration processes occurring in the laboratory-scale OD system (Hu et al. 2003). Hence, N-PUR/O-PUR was calculated as the value of the DPAOs/PAOs ratio.

Figure 2

Measurement results of the O-PUR, N-PUR, N-PUR′ and DPAOs/PAOs ratio under different IRF conditions in the laboratory-scale OD system.

Figure 2

Measurement results of the O-PUR, N-PUR, N-PUR′ and DPAOs/PAOs ratio under different IRF conditions in the laboratory-scale OD system.

Different species of PAOs have been reported, such as Betaproteobacteria, Actinobacteria, Alphaproteobacteria and Gammaproteobacteria; in particular, the Rhodocyclus group from subclass 2 of the Betaproteobacteria was represented to a greater extent in PAOs, and which was named as ‘Candidatus Accumulibacter phosphatis’ and often abbreviated to Accumulibacter (Oehmen et al. 2007). He et al. (2008) pointed out that Accumulibacter was widely used as a model organism to study enhanced biological phosphorus removal (EBPR), as Accumulibacter could be repeatedly enriched in laboratory-scale reactors fed with simple carbon sources (e.g., acetate or propionate). Meanwhile, He et al. (2008) found that Accumulibacter accounted for 40%–69% of PAOs among five full-scale EBPR systems. Furthermore, Type I and Type II Candidatus Accumulibacter phosphatis were proved to be capable of denitrification; the Type I was able to denitrify from nitrate and/or nitrite while Type II was able to denitrify from nitrite (Carvalho et al. 2007; Flowers et al. 2009; Acevedo et al. 2012). So, N-PUR′ was similar to N-PUR, also indicating that DPAOs in the laboratory-scale OD system were mainly the Type I Accumulibacter.

As shown in Figure 2, when the IRF was 0.5 h−1, the O-PUR was up to 26.5 mg P/(g VSS.h), and then the O-PUR first decreased and then increased. When the IRF was 3.4 h−1 and 7.5 h−1, the DPAOs/PAOs ratio reached about 50%.

In order to explain the O-PUR, N-PUR, and DPAOs/PAOs ratio changes, the IRF and T values are listed in Table 3.

According to Table 3, under 0.5 h−1 IRF conditions, 60.3 min and 55.6 min of mixed liquor flowed through the aerobic and anoxic zones during one lap. This means that in each lap, the Taer was sufficiently long enough to allow the aerobic phosphorus uptake to fully react and this made the intracellular poly-β-hydroxyalkanoate (PHA) levels insufficient when the mixed liquor flowed into the anoxic zones. Even though Tano was also longer, due to the lack of sufficient PHA, the DNPR reaction could not work well. After long-term operation, the DPAOs/PAOs ratio was low.

When the IRF was increased to 1.6 h−1, 3.4 h−1, and 7.5 h−1, the O-PUR gradually decreased and N-PUR increased. The Taer was shortened to 19.2 min, 9.3 min, and 4.1 min during one lap and the PHA consumption was reduced when the mixed liquor passed through the aerobic zones, which was beneficial to the DNPR reaction in the anoxic zones.

When the IRF increased to 11.4 h−1 and 18.7 h−1, the Taer and Tano were all shortened in one lap. As it was usually considered that all the PAOs could perform aerobic phosphorus uptake and only some of the PAOs could implement anoxic phosphorus uptake, as such, during very short retention times, the aerobic phosphorus uptake reaction went better than the DNPR reaction. After long-term operation, the resulting O-PUR increased and the N-PUR decreased.

Variations in the nutrient removal and homogeneity of the wastewater quality

Nutrient removal efficiencies and effluent nitrogen concentrations during the sampling-analysis period under different IRF conditions are shown in Figure 3. Wastewater quality in the different zones of the laboratory-scale OD system are shown in Figure 4.

Figure 3

Nutrient removal efficiencies and effluent nitrogen concentrations in the lab-scale OD system under different IRF conditions.

Figure 3

Nutrient removal efficiencies and effluent nitrogen concentrations in the lab-scale OD system under different IRF conditions.

Figure 4

Wastewater quality in different zones of the laboratory-scale OD system. (Ana3, Ano1, Ano2, Ano3, Ano4, Aer1, Aer2, and Eff are the sampling position, namely the influent wastewater, the wastewater from anaerobic zone 3, anoxic zones 1–4, aerobic zones 1–2, and the effluent wastewater of the sedimentation tank, respectively.)

Figure 4

Wastewater quality in different zones of the laboratory-scale OD system. (Ana3, Ano1, Ano2, Ano3, Ano4, Aer1, Aer2, and Eff are the sampling position, namely the influent wastewater, the wastewater from anaerobic zone 3, anoxic zones 1–4, aerobic zones 1–2, and the effluent wastewater of the sedimentation tank, respectively.)

As shown in Figure 3, under each IRF condition, the COD removal efficiency barely changed. When the IRF increased to 11.4 h−1 and 18.7 h−1, the residence time of the mixed liquor flowing through the anoxic and aerobic zones in one lap became increasingly shorter, making it impossible to thoroughly implement the nitrification and denitrification. As such, the effluent nitrogen concentration increased, and finally the TN removal efficiency dropped significantly; meanwhile the TP removal efficiency slightly decreased. When the IRF was 3.4 h−1 and 7.53 h−1, the removal efficiencies of the COD, TN, and TP were over 92%, 80%, and 93%, respectively, and all the values were higher than those of the other IRFs.

It can be seen from Figure 4(a) and 4(c) that, under each IRF condition, the COD and NH4+-N decreased dramatically from the influent into anaerobic zone 3. The reasons were as follows: firstly, the influent wastewater and external returned activated sludge were mixed in the anaerobic zones, which had a dilution effect on the COD and NH4+-N concentrations; secondly, the anaerobic phosphorus release consumed the COD from the wastewater and transformed the COD into PHA, which was stored inside PAO cells; thirdly, in view of the small amount of nitrate within the return sludge, the traditional denitrification processes occurred in the anaerobic zones, which also utilised the COD; and finally, some large molecular COD was adsorbed extracellularly by the activated sludge. The first two reasons are the most important, while the latter two reasons have little influence. However, there was no significant difference between the COD concentrations of the anoxic and aerobic zones in the circular corridor.

In contrast, the TP concentration significantly increased in the anaerobic zone because of the anaerobic phosphorus release under each of the IRF conditions as shown in Figure 4(b). When the IRF was 0.5 h−1 and 1.6 h−1, Tano was longer to perform a better traditional denitrification process, so the nitrate concentrations in the effluent and external return sludge were less than the other IRFs (see Table 3). Therefore, the TP increment in the anaerobic zones was higher than that of the other IRFs.

Table 4

Mass balance measurement values

IRF (h−1)
C*,&0.51.63.47.511.418.7
CCOD,In (mg O/L) 398 397 395 386 389 395 
CCOD,Eff mg O/L) 24 36 22 29 33 31 
CTN,In (mg N/L) 39.5 39.0 38.9 39.5 38.1 39.8 
CNO3-N,Eff (mg N/L) 0.00 0.00 2.83 2.09 3.34 4.06 
CNO2-N,Eff (mg N/L) 1.36 0.24 0.14 0.15 0.15 0.14 
CNOx-N,Eff (mg N/L) 0.82 0.14 2.91 2.18 3.43 4.14 
CTN,Eff (mg N/L) 6.03 5.43 6.26 6.06 8.97 14.35 
CTP,In (mg P/L) 7.0 7.0 6.5 6.7 6.8 6.6 
CTP,Ana (mg P/L) 42.2 42.4 30.1 31.4 29.1 26.1 
CTP,Eff (mg P/L) 0.66 0.69 0.35 0.41 0.65 0.77 
MLVSS (mg/L) 3,351 3,168 3,071 3,185 3,259 3,280 
IRF (h−1)
C*,&0.51.63.47.511.418.7
CCOD,In (mg O/L) 398 397 395 386 389 395 
CCOD,Eff mg O/L) 24 36 22 29 33 31 
CTN,In (mg N/L) 39.5 39.0 38.9 39.5 38.1 39.8 
CNO3-N,Eff (mg N/L) 0.00 0.00 2.83 2.09 3.34 4.06 
CNO2-N,Eff (mg N/L) 1.36 0.24 0.14 0.15 0.15 0.14 
CNOx-N,Eff (mg N/L) 0.82 0.14 2.91 2.18 3.43 4.14 
CTN,Eff (mg N/L) 6.03 5.43 6.26 6.06 8.97 14.35 
CTP,In (mg P/L) 7.0 7.0 6.5 6.7 6.8 6.6 
CTP,Ana (mg P/L) 42.2 42.4 30.1 31.4 29.1 26.1 
CTP,Eff (mg P/L) 0.66 0.69 0.35 0.41 0.65 0.77 
MLVSS (mg/L) 3,351 3,168 3,071 3,185 3,259 3,280 

CNOx-N,Eff is the equivalent concentration of nitrate. In denitrification process, it takes 5 mol electrons to convert 1 mol NO3-N into N2, and 3 mol electrons to convert 1 mol NO2-N into N2, so 1 mol NO2-N is equivalent to 0.6 mol NO3-N, so CNOx-N,Eff = CNO3-N,Eff + 0.6 CNO2-N,Eff.

As shown in Figure 4(b) and 4(c), under 0.5 h−1 IRF conditions, the TP and NH4+-N concentrations in aerobic zone 1 and aerobic zone 2 consistently decreased compared with anoxic zone 4 due to the aerobic phosphorus uptake reaction and nitrification reaction. The effluent of anaerobic zone 3 flowed directly into anoxic zone 2, and meanwhile the internal recirculation flow of the mixed liquor from aerobic zone 2 flowed into anoxic zone 1. As such, the TP and NH4+-N concentrations in anoxic zone 1 were significantly lower than those of anaerobic zone 3 and anoxic zones 2, 3, and 4 under the 0.5 h−1 IRF condition. With the increasing of IRF from 1.6 h−1, 3.4 h−1, and 7.5 h−1, 11.4 h−1, to 18.7 h−1, the homogeneity of the wastewater was enhanced, and then the TP and NH4+-N concentration distributions in the circular corridor were more homogenous.

Results of the nutrients mass balance

The mass balance measurement values are listed in Table 4. The nutrients mass balance results are shown in Figure 5.

Figure 5

Nutrient mass balance results of the laboratory-scale OD system.

Figure 5

Nutrient mass balance results of the laboratory-scale OD system.

As shown in Figure 5(a)–5(c), under each IRF condition, the percentages of COD and TP that were discharged by the effluent had little difference (MCOD,Eff/MCOD,In: 5.6%–9.1%; MTP,Eff/MTP,In: 5.4%–11.7%), but the percentage of TN discharged by the effluent (MTN,Eff/MTN,In) increased from 13.9% to 36.0%. The percentages of COD, TN and TP that were transformed into microbial cells had little difference under each IRF condition (MCELL_COD,Ex/MCOD,In: 36.8%–39.8%; MCELL_TN,Ex/MTN,In: 26.3%–28.5%; MCELL_TP,Ex/MTP,In: 30.3%–33.3%). When the IRF was 3.4 h−1 and 7.5 h−1, the percentages of organic matter, nitrogen, and phosphorus removed by the DNPR were higher than under other IRF conditions. When the IRF was 3.4 h−1 and 7.5 h−1, about 20% of the TP was removed (Mdenphosre_TP,Ex/MTP,In) and more than 11% of the TN was discharged into the atmosphere in the form of N2 (Mdenphosre_N2,Air/MTN,In). Additionally, more than 3% of the COD was discharged into the atmosphere in the form of CO2 (Mdenphosre_CO2,Air/MCOD,In) from the DNPR process.

As shown in Figure 5(a), with the increase of the IRF from 0.5 h−1 to 18.7 h−1, the percentage of COD discharged into the atmosphere in the form of CO2 by traditional denitrification (Mtraden_CO2,Air/MCOD,In) decreased from 14.7% to 9.3%. The aerobic phosphorus uptake (Maerophosre_CO2,Air/MCOD,In) was 14.0%, 12.4%, 7.6%, 7.9%, 8.2%, and 8.2% respectively for each incremental increase of the IRF. The organic aerobic oxidation (Maeroxida_CO2,Air/MCOD,In) varied between 24.4% and 34.4%.

As shown in Figure 5(b), the percentage of TN discharged into the atmosphere in the form of N2 by traditional denitrification (Mtraden_N2,Air/MTN,In) decreased from 52.8% to 33.0% when the IRF increased from 0.5 h−1 to 18.7 h−1.

As shown in Figures 2 and 5(c), with the rising of the IRF from 0.5 h−1 to 18.7 h−1, the O-PUR first decreased and then increased, the percentage of TP removed by aerobic phosphorus removal (Maerophosre_TP,Ex/MTP,In) was 54.6%, 48.8%, 44.2%, 42.4%, 45.8%, and 48.9%.

Application discussion

The DNPR process can save both organic carbon sources and aeration energy demand, and also reduce the sludge production (Kuba et al. 1996; Meinhold et al. 1999; Hou et al. 2009), so it is meaningful to strengthen the DNPR process even in ODs where effluent quality has reached the standard.

From the above results, it was known that the DNPR process could be enhanced in an OD system when IRF was between 3.4 h−1 and 7.5 h−1. Meanwhile, IRF is affected by the circular corridor length of the OD system (L) and the flow velocity of mixed liquor (v). Therefore, the optimum IRF scope (3.4 h−1–7.5 h−1) for enhancing DNPR process can be realized with consideration of the following aspects in full-scale OD systems.

For the OD system in design

During the design of the OD system, according to the inflow rate and raw wastewater quality, the loading rate and HRT are usually determined first; then, the volume of the OD system (V) can be confirmed. According to the selected values of IRF and v, L can be confirmed. Then according to V and L, the width and height of the circular corridor are obtained.

For an established OD system

For an established OD system with IRF <3.4 h−1, an increase in v can be tried in order to regulate IRF in the optimum range. For the established OD system with IRF >7.5 h−1, v can be decreased, or a diversion wall can be built to reduce the cross-section area in the circular corridor, and IRF can be regulated in the optimum range.

As v has its own appropriate range, and other factors need to be taken into account during the operation or reformation, not all established OD systems can be regulated to have IRF in the optimum range.

CONCLUSIONS

For the OD system that had an anaerobic zone ahead of the circular corridor, set anoxic and aerobic zones along the circular corridor, and decoupled aeration from the plug-flow, when the IRF was between 3.4 h−1 and 7.5 h−1 the DPAOs/PAOs could reach about 50%. The percentages of organic matter, nitrogen, and phosphorus removed by the DNPR were higher than under other IRF conditions. When the IRF was greater than 11.5 h−1, the TN removal efficiency decreased significantly and this was not conducive to simultaneous nitrogen and phosphorus removal.

ACKNOWLEDGEMENTS

This research work was financially supported by the National Natural Science Foundation of China (No. 51108299 and No. 51678388). The authors would like to thank the help and support of Dan Dan Yuan, Chao Jiao, Miao Qing Kang, and Ying Kai Tang for sample measurements and experimental device operation during the study.

REFERENCES

REFERENCES
Abusam
A.
,
Keesman
K. J.
,
Spanjers
H.
&
Straten
G. v.
2004
Benchmarking procedure for full-scale activated sludge plants
.
Control Engineering Practice
12
(
3
),
315
322
.
Acevedo
B.
,
Oehmen
A.
,
Carvalho
G.
,
Seco
A.
,
Borrás
L.
&
Barat
R.
2012
Metabolic shift of polyphosphate-accumulating organisms with different levels of polyphosphate storage
.
Water Research
46
(
6
),
1889
1900
.
Alaya
S. B.
,
Haouech
L.
,
Cherif
H.
&
Shayeb
H.
2010
Aeration management in an oxidation ditch
.
Desalination
252
(
1–3
),
172
178
.
Ammary
B. Y.
&
Radaideh
J. A.
2005
Simultaneous nitrification and denitrification in an oxidation ditch plant
.
Chemical and Biochemical Engineering Quarterly
19
(
2
),
207
212
.
Barker
P. S.
&
Dold
P. L.
1995
COD and nitrogen mass balances in activated sludge systems
.
Water Research
29
(
2
),
633
643
.
Carvalho
G.
,
Lemos
P. C.
,
Oehmen
A.
&
Reis
M. A.
2007
Denitrifying phosphorus removal: linking the process performance with the microbial community structure
.
Water Research
41
(
19
),
4383
4396
.
CEPB
2002
Analysis Methods for Water and Wastewater
, 4th edn (Supplement Edition).
China Environmental Science Press
,
Beijing, China
(in Chinese)
.
Chen
F.
2004
Design of the oxidation ditch sewage treatment plant of Zuoyun county
.
Sci/Tech Information Development and Economy
14
(
8
),
306
307
(in Chinese)
.
Chen
R.
2012
Optimized operation of improved oxidation ditch process in Handan West wastewater treatment plant
.
Water Industry Market
(
2
),
49
52
(in Chinese)
.
Chen
X.
,
Fujiwara
T.
,
Nakamachi
K.
,
Kawaguchi
Y.
,
Nishimura
F.
&
Ohtoshi
K.
2012
Evaluation of a novel oxidation ditch system with dual DO control technology for biological nutrient removal by mass balance analysis
.
Desalination
286
(
1
),
24
33
.
Flowers
J. J.
,
He
S.
,
Yilmaz
S.
,
Noguera
D. R.
&
McMahon
K. D.
2009
Denitrification capabilities of two biological phosphorus removal sludges dominated by different ‘Candidatus Accumulibacter’ clades
.
Environmental Microbiology Reports
1
(
6
),
583
588
.
Henze
M.
,
Gujer
W.
,
Mino
T.
&
van Loosdrecht
M.
2000
Activated Sludge Models ASM1,ASM2,ASM2d and ASM3
.
IWA Publishing
,
London, UK
.
Henze
M.
,
van Loosdrecht
M. C. M.
,
Ekama
G. A.
&
Brdjanovic
D.
2008
Biological Wastewater Treatment: Principles, Modelling and Design
.
IWA Publishing
,
London
,
UK
.
Hou
H.
,
Wang
S.
,
Peng
Y.
,
Yuan
Z.
,
Yin
F.
&
Gan
W.
2009
Anoxic phosphorus removal in a pilot scale anaerobic-anoxic oxidation ditch process
.
Frontiers of Environmental Science & Engineering in China
3
(
1
),
106
111
.
Insel
G.
,
Artan
N.
&
Orhon
D.
2005
Effect of aeration on nutrient removal performance of oxidation ditch system
.
Environmental Engineering Science
22
(
6
),
802
814
.
Lee
J. K.
,
Lee
K. H.
&
Yim
S. B.
2007
Optimization of nitrogen removal in a sequencing batch reactor system by variation of the time distribution
.
Journal of Environmental Science and Health, Part A
42
,
1655
1663
.
Li
X.
,
Zhang
J.
,
Zhang
X.
,
Li
J.
,
Liu
F.
&
Chen
Y.
2019
Start-up and nitrogen removal performance of CANON and SNAD processes in a pilot-scale oxidation ditch reactor
.
Process Biochemistry
84
,
134
142
.
MEE
2015
List of Urban Sewage Treatment Facilities Put Into Operation Nationwide. Ministry of Ecology and Environment of the People's Republic of China
, pp.
1
214
(in Chinese)
.
Meinhold
J.
,
Filipe
C. D. M.
,
Daigger
G. T.
&
Isaacs
S.
1999
Characterization of the denitrifying fraction of phosphate accumulating organisms in biological phosphate removal
.
Water Science and Technology
39
(
1
),
31
42
.
Oehmen
A.
,
Lemos
P. C.
,
Carvalho
G.
,
Yuan
Z.
,
Keller
J.
,
Blackall
L. L.
&
Reis
M. A. M.
2007
Advances in enhanced biological phosphorus removal: from micro to macro scale
.
Water Research
41
(
11
),
2271
2300
.
Qiu
Y.
,
Zhang
C.
,
Li
B.
,
Li
J.
,
Zhang
X.
,
Liu
Y.
,
Liang
P.
&
Huang
X.
2018
Optimal surface aeration control in full-scale oxidation ditches through energy consumption analysis
.
Water
10
(
7
),
1
15
.
Wachtmeister
A.
,
Kuba
T.
,
van Loosdrecht
M. C. M.
&
Heijnen
J. J.
1997
A sludge characterization assay for aerobic and denitrifying phosphorus removing sludge
.
Water Research
31
(
3
),
471
478
.
Zhou
R. L.
,
Sun
Y.
&
Ma
G. T.
2004
Application of Carrousel oxidation ditch to treat domestic sewage in Xinglong Zhuang coal mine
.
Energy Environmental Protection
18
(
4
),
24
26
.
(in Chinese)
.