Partial nitrification under limited dissolved oxygen conditions

Partial nitrification to nitrite is technically feasible and economically favourable, especially when wastewaters contained high ammonium concentrations or low C/N ratios. Partial nitrification can be obtained by selectively inhibiting nitrite-oxidizing bacteria (NOB) through appropriate regulation of the pH, temperature and dissolved oxygen (DO) concentrations. The effect of pH, DO levels and temperature on ammonia oxidation rate and nitrite accumulation was investigated in order to determine the optimal conditions for partial nitrification of synthetic wastewater with high ammonia concentration. The experiments performed at low DO levels to lower the total oxygen needed in the nitrification step, which means great saving in aeration. During the start-up stage pH and DO were set at 7.0–7.4 and 0.5 mg/l, respectively. The reactor was operated until complete partial nitrification was achieved. The effect of pH, DO on partial nitrification was studied, as pH was kept at 6.5, 7.5, 8.5, 9.5 and DO at 0.5±0.2, 1.5±0.2 and 2.5±0.2 mg/l, and temperature at 30 °C. The influence of temperature on k_a value was studied by keeping pH=7.5, DO=1.5 mg/l and temperature was controlled at 12, 20 and 30 °C, respectively. The results showed that partial nitrification to nitrite was steadily obtained and the optimal operational parameters were pH=7.5, DO=1.5 mg/l, T=30 °C based on ammonia oxidation rate and nitrite accumulation rate. The maximum k_a was achieved and to be 115.1×10⁻³ mg NH₄⁺–N (mg VSS h)⁻¹ under this condition.     

Dynamic simulation of a WWTP operated at low dissolved oxygen condition by integrating activated sludge model and a floc model

While an aeration tank in an activated sludge process is often operated with high dissolved oxygen (DO) concentration to ensure organic degradation and nitrification, it may be operated at low DO concentration to reduce energy consumption and achieve desired denitrification. The ASM1 (Activated Sludge Model No. 1) can be used to describe the activated sludge process if the nitrification and denitrification occur either during different phases or in different tanks, but it may encounter problems in simulating the denitrification phenomenon caused by low DO concentration in the aeration tank. In the present work, we developed a model integrating the ASM1 kinetics and a biofloc model to account for the actual anoxic and aerobic rates. Oxygen was assumed the only substrate of both bio-kinetically and flux limiting in the flocs and its dispersion coefficient was estimated as 1.2 × 10⁻⁴ m² day⁻¹ by using a set of measured effluent qualities of a full-scale wastewater treatment plant (WWTP) operating at low DO concentration (∼0.80 mg L⁻¹) for 60 days. Simulation studies predicted the optimal DO level of 0.36 mg L⁻¹ which would lead to minimum total nitrogen of 15.7 mg N L⁻¹ and also showed the insignificance of the addition of carbon source for nitrogen removal for the operation under study. The developed model may be helpful for process engineers to predict the plant behaviors under various configurations or operating strategies.     

Wastewater treatment by using kenaf in paddy soil and effect of dissolved oxygen concentration on efficiency

We previously reported that kenaf (Hibiscus cannabinus L.) planted in a zeolite-bed filter-ditch system provided highly effective treatment of wastewater. Here we compared that system with treatment in fallow paddy fields irrigated in different ways in a greenhouse. Paddy soil was a useful alternative to zeolite as the bed filter material. The efficiency of removal of N and P under furrow irrigation and flooding was 82–92% of that of the zeolite system. Most kenaf roots were distributed in water with a high dissolved oxygen (DO) concentration and a high redox potential; few roots grew in reducing soil under water. The roots distributed in the water contributed most to wastewater treatment. A low DO concentration (0.3 mg L⁻¹) decreased the efficiency of N and P removal. However, nightly low DO concentration (near 0 mg L⁻¹) alternating with daily high DO concentration did not seriously restrict the efficiency. An increase of alpha-naphthylamine oxidation activity in kenaf roots at low DO concentration is discussed in regard to induction of an oxygen-protective enzyme.     

Control and optimization of nitrifying communities for nitritation from domestic wastewater at room temperatures

To achieve nitritation from complete-nitrification seed sludge at room temperature of 19 ± 1 °C, a lab-scale sequencing batch reactor (SBR) treating domestic wastewater with low C/N ratios was operated to investigate the control and optimization of nitrifying communities. Ammonia oxidizing bacteria (AOB) dominance was enhanced through the combination of low DO concentrations (<1.0 mg/L) and preset short-cycle control of aeration time. Nitritation was successfully established with NO₂^?-N/NO_x^?-N over 95%. To avoid the adverse impact of low DO concentrations on AOB activities, DO concentrations were increased to 1–2 mg/L. At the normal DO levels and temperatures, on-line control strategy of aerobic durations maintained the stability of nitritation with nitrite accumulation rate over 95% and ammonia removal above 97%. Fluorescence in-situ hybridization (FISH) analysis presented that the maximal percentage of AOB in biomass reached 10.9% and nitrite oxidizing bacteria (NOB) were washed out.     

Operation of partial nitrification to nitrite of landfill leachate and its performance with respect to different oxygen conditions

The coupled system of partial nitrification and anaerobic ammonium oxidation (Anammox) is efficient in nitrogen removal from wastewater. In this study, the effect of different oxygen concentrations on partial nitrification performance with a sequencing batch reactor (SBR) was investigated. Results indicate that, partial nitrification of landfill leachate could be successfully achieved under the 1.0–2.0 mg L⁻¹ dissolved oxygen (DO) condition after 118 d long-term operation, and that the effluent is suitable for an Anammox reactor. Further decreasing or increasing the DO concentration, however, would lead to a decay of nitrification performance. Additionally, the MLSS concentration in the reactor increased with increasing DO concentration. Respirometric assays suggest that low DO conditions (<2 mg L⁻¹) favor the ammonia-oxidizing bacteria (AOB) and significantly inhibit nitrite oxidizing bacteria (NOB) and aerobic heterotrophic bacteria (AHB); whereas high DO conditions (>3 mg L⁻¹) allow AHB to dominate and significantly inhibit AOB. Therefore, the optimal condition for partial nitrification of landfill leachate is 1.0–2.0 mg L⁻¹ DO concentration.     

The investigation of dairy industry wastewater treatment in a biological high performance membrane system

The dairy industry is generally considered to be the largest source of food processing wastewater in many countries. The highly variable nature of dairy wastewaters in terms of volumes and flowrates and in terms of high organic materials contents such as COD 921–9004 mg L⁻¹, BOD 483–6080 mg L⁻¹, TN of 8–230 mg L⁻¹ and SS of 134–804 mg L⁻¹ makes the choice of an effective wastewater treatment regime difficult. A high performance bioreactor, an aerobic jet loop reactor, combined with a ceramic membrane filtration unit, was used to investigate its suitability for the treatment of the dairy processing wastewater. The oxygen transfer rates of the bioreactor were found to be very high (100–285 h⁻¹) on the operating conditions. A loading rate of 53 kg COD m⁻³ d⁻¹ resulted in 97–98% COD removal efficiencies under 3 h hydraulic retention time. The high MLSS concentrations could be retained in the system (up to 38,000 mg L⁻¹) with the contribution of UF (ultrafiltration) unit. During the filtration of activated sludge, the fluxes decreased with increasing MLSS. Cake formation fouling was determined as dominant fouling mechanisms. The results demonstrate that jet loop membrane bioreactor system was a suitable and effective treatment choice for treating dairy industry wastewater.     

Characteristics of nitrous oxide (N2O) emission from intermittently aerated sequencing batch reactors (IASBRs) treating slaughterhouse wastewater at low temperature

This study investigated the characteristics of nitrous oxide (N₂O) emission from intermittently aerated sequencing batch reactors (IASBRs) treating high strength slaughterhouse wastewater at 11 °C, where partial nitrification followed by denitrification (PND) was achieved. N₂O generation and emission was examined at three aeration rates of 0.4, 0.6, and 0.8 L air/min in three IASBRs (SBR1, SBR2, and SBR3, respectively). The slaughterhouse wastewater contained chemical oxygen demand (COD) of 6057 ± 172.6 mg/L, total nitrogen (TN) of 576 ± 15.1 mg/L, total phosphorus (TP) of 52 ± 2.7 mg/L and suspended solids (SS) of 1843 ± 280.5 g/L. In the pseudo-steady state, the amount of N₂O emission was up to 5.7–11.0% of incoming TN. The aeration rate negatively affected N₂O emission and the ratio of N₂O emission to incoming TN was reduced by 48.2% when the aeration rate was increased from 0.4 to 0.8 L air/min. Results showed that more N₂O was generated in non-aeration periods than in aeration periods. Lower DO concentrations enhanced N₂O generation in the aeration periods (probably via nitrifier denitrification) while low DO concentrations (lower than 0.2 mg/L) did not affect N₂O generation in the non-aeration periods (probably via heterotrophic denitrification). When PHB was utilized as the organic substrate for denitrification, there was a high N₂O generation potential. It was estimated that 1.8 mg N₂O-N was generated accompanying per mg PHB consumed.     

Real-time control strategy for nitrogen removal via nitrite in a SHARON reactor using pH and ORP sensors

This paper presents a real-time control strategy for nitrogen removal via nitrite in a continuous flow SHARON reactor using on-line available and industrially feasible sensors (pH and ORP). The developed control strategy optimizes the length of aerobic and anoxic phases as well as the external carbon source addition. This strategy, implemented in a laboratory-scale SHARON reactor fed with synthetic wastewater and real dewatering sludge supernatant, was able to cope with step variations in influent flow rate and ammonium concentration. The main advantages of this control strategy over the traditional operation mode with fixed carbon source dosification and fixed length cycle operation were: better effluent quality (ammonia concentration decreased from 12 to 2 mg NH₄–N L^?1 and nitrogen removal efficiency raised from 95% to 98%) as result of the shorter cycle length: 2.9 h versus 4.0 h, and savings in external carbon addition: 1332 mg COD d^?1 versus 2100 mg COD d^?1.     

Sulfate reduction and microbial community of autotrophic biocathode in response to acidity

Microbial electrolysis cells (MECs) with autotrophic biocathode are a promising technology for removal of pollutants in wastewater. The aim of this study was to investigate the effect of initial acidity of wastewater on performance of sulfate-reducing biocathodes. MECs with biocathodes were operated with initial pH values of catholyte ranged from 3.0 to 7.0. The optimum initial pH value was 6.0 with a maximum sulfate reductive rate and biomass of 57 mg L⁻¹ d⁻¹ and 2.1 ± 0.4 mg g⁻¹, respectively. With initial pH 7.0, the pH value of catholyte increased to 9.8 ± 0.2 after an operation cycle, which resulted in low performance of the biocathode. A considerable sulfate reductive rate of 31 ± 0.85 mg L⁻¹ d⁻¹ was achieved with initial pH 3.0. Desulfovibrio sp. grew dominantly with abundance of 46%–66% in the cathode biofilm with initial pH values from 3.0 to 6.0 and contributed to the sulfate reduction. Clostridium and Parapedobacter also had high abundance in pH 6.0 cathode, indicated that interspecies electron transfer between electrochemical active and sulfate-reducing bacteria could play an important role in sulfate removal. The results suggest that acidity of catholyte is an important factor to be considered to utilize autotrophic biocathode MECs for wastewater treatment.     

Nitrification rates of algal–bacterial biofilms in wastewater stabilization ponds under light and dark conditions

The objective of this study was to investigate nitrification rates in algal–bacterial biofilms of waste stabilization ponds (WSP) under different conditions of light, oxygen and pH. Biofilms were grown on wooden plates of 6.0 cm by 8.0 cm by 0.4 cm in a PVC tray continuously fed with synthetic wastewater with initial NH₄-N and Chemical Oxygen Demand (COD) concentrations of 40 mg l^?1 and 100 mg l^?1, respectively, under light intensity of 85–95 μE m^?2 s^?1. Batch activity tests were carried out by exposure of the plates to light conditions as above (to simulate day time), dim light of 1.8–2.2 μE m^?2 s^?1 (to simulate reduced light as in deeper locations in WSP) and dark conditions (to simulate night time). Dissolved oxygen (DO) concentration and pH were controlled. At some experiments, both parameters were kept constant, and at others they were left to vary as in WSP. Results show biofilm nitrification rates of 945–1817 mg-N m^?2 d^?1 and 1124–1615 mg-N m^?2 d^?1 for light and dark experiments. When the minimum DO was 4.1 mg l^?1, the biofilm nitrification rates under light and dark conditions did not differ significantly at 95% confidence. When the minimum DO in the dim light experiment was 3.2 mg l^?1, the nitrification rates under light and dim light conditions were 945 mg-N m^?2 d^?1 and 563 mg-N m^?2 d^?1 and these significantly differed. Further decrease of DO to 1.1 mg l^?1 under dark conditions resulted in more decrease of the nitrification rates to 156 mg-N m^?2 d^?1. It therefore seems that under these experimental conditions, biofilm nitrification rates are significantly reduced at a certain point when bulk water DO is between 3.2 mg l^?1 and 4.1 mg l^?1. As long as bulk water DO under dark is high, light is not important in influencing the process of nitrification.     

An integrated treatment system for coffee processing wastewater using anaerobic and aerobic process

The experiment was conducted to develop an integrated treatment system for coffee processing wastewater (CPWW) through the combination of biomethanation with aeration and wetland plants treatment. The biomethanation was carried out at different hydraulic retention times (HRTs) using upflow anaerobic hybrid reactor (UAHR) and 18 h of HRT was found to be optimum. The maximum biochemical oxygen demand (BOD), chemical oxygen demand (COD) and total solids (TS) reduction were 66.0%, 61.0% and 58.0%, respectively with organic loading rate of 9.55 kg m^?3 day^?1. The reduction of pollution load of the wastewater by microbial action augmented by aeration resulted in the reduction of electrical conductivity (EC), BOD, COD, and total solids (TS). Continuous aeration of wastewater resulted in maximum reduction of BOD (74.6%), COD (68.6%) and TS (49.3%). The wetland plant, Typha latifolia reduced 85.4% and 78.0% of BOD and COD, respectively in biomethanated cum aerated CPWW.     

Fed-batch fermentation of recombinant Citrobacter freundii with expression of a violacein-synthesizing gene cluster for efficient violacein production from glycerol

The extensive prospects of violacein in the pharmaceutical industry have attracted increasing interest. However, the fermentation levels of violacein are currently inadequate to meet the demands of industrial production. This study was undertaken to develop an efficient process for the production of violacein by recombinant Citrobacter freundii. The effects of dissolved oxygen (DO) and pH on cell growth and violacein production in batch cultures were investigated first. When the DO and pH of the medium were controlled at around 25% and 7.0, respectively, the biomass and concentration of violacein were maximized. Based on the consumption of nutrients in the medium observed during batch culture, a fed-batch fermentation strategy with controlled DO and pH was implemented. By continuously feeding glycerol, NH₄Cl, and l-tryptophan at a constant feeding rate of 16 mL h⁻¹, the final concentration of violacein reached 4.13 g L⁻¹, which was 4.09-fold higher than the corresponding batch culture, and the maximal dry cell weight (DCW) and average violacein productivity obtained for the fed-batch culture were 3.34 g DCW L⁻¹ and 82.6 mg L⁻¹ h⁻¹, respectively. To date, this is the first report on the efficient production of violacein by genetically engineered strains in a fermentor.     

Effects of redox potential control on succinic acid production by engineered Escherichia coli under anaerobic conditions

The effects of oxido-reduction potential (ORP) control on succinic acid production have been investigated in Escherichia coli LL016. In LL016, two CO₂ fixation pathways were achieved and NAD⁺ supply was enhanced by co-expression of heterologous pyruvate carboxylase (PYC) and nicotinic acid phosphoribosyltransferase (NAPRTase). During anaerobic fermentation, cell growth and metabolite distribution were changed with redox potential levels in the range of −200 to −400 mV. From the results, the ORP level of −400 mV was preferable, which resulted in the high succinic acid concentration (28.6 g/L) and high succinic acid productivity (0.33 g/L/h). Meanwhile, the yield of succinic acid at the ORP level of −400 mV was 39% higher than that at the ORP level of −200 mV. In addition, a higher NADH/NAD⁺ ratio and increased enzyme activities were also achieved by regulating the culture to a more reductive environment, which further enhanced the succinic acid production.     

A unified model of ammonium oxidation rate at various initial ammonium strength and active ammonium oxidizer concentrations

This paper attempts to provide insight into the biological ammonium oxidation process applied to high-strength ammonium wastewater treatment. The ammonium oxidation process has been investigated at various ammonium and biomass concentrations. Using the oxygen uptake rate (OUR) method, a proportion of both active ammonium oxidizers (AAO) and nitrite oxidizers to the total suspended solids were separately estimated, and then tested to normalize the ammonium oxidation rate at various ammonium strengths and AAO concentrations. High-ammonium strength showed no significant inhibition to ammonium oxidation due to high-AAO concentration. It was demonstrated that the key factor deciding the specific ammonium oxidation rate was the ratio of ammonium concentration to the active nitrifiers (AN) concentration, but not the sole-variable such as initial ammonium concentration and AN concentration. Contois model was screened to suitably fit the ammonium oxidation kinetics under the high-ammonium loading condition, resulting in a half-saturation constant of 0.028 mg N mg^?1 AAO and a maximum specific ammonium oxidation rate of 3.56 g N g^?1 AAO d^?1.     

Efficient nitrogen removal in a pilot system based on upflow multi-layer bioreactor for treatment of strong nitrogenous swine wastewater

In order to prevent the toxic effect caused by high strength ammonium in a swine wastewater treatment system, a patented upflow multi-layer bioreactor (UMBR) as a pre-anoxic tank was applied to a pilot-scale plant with a treatment capacity of 5 m³/d. This plant was operated for 4 months at a high IR ratio in the range of 10–17, in order to alleviate the toxic effects caused by high strength ammonium. A computational fluid dynamic (CFD) analysis was also conducted to design and configure the rotating distributors in the UMBR. At an IR ratio of about 17, the influent NH₄⁺-N (1169 mg N/L) was diluted to less than 80 mg N/L at the head of the UMBR, and then was completely nitrified (about 98.3%) in the aeration tank, without any inhibition caused by high strength ammonium. The nitrate at a concentration of about 58.2 mg N/L recycled from the aeration tank was completely denitrified in UMBR #1, which was operated at an actual hydraulic retention time (HRT) of 3.5 h.     

Denitrification of nitrified and non-nitrified swine lagoon wastewater in the suspended sludge layer of treatment wetlands

One method for managing livestock-wastewater N is the use of treatment wetlands. The objectives of this study were to (1) assess the magnitude of denitrification enzyme activity (DEA) in the suspended sludge layers of bulrush and cattail treatment wetlands, and (2) evaluate the impact of nitrogen pretreatment on DEA in the suspended sludge layer. The study used four wetland cells (3.6 m × 33.5 m) with two cells connected in series. Each wetland series received either untreated or partially nitrified swine wastewater from a single-cell anaerobic lagoon. The DEA of the suspended sludge layers of the constructed wetlands was measured by the acetylene inhibition method. The control DEA treatment for the sludge layer had a mean rate of 18 μg N₂O-N g^?1 sludge h^?1. Moreover, the potential DEA (nitrate-N and glucose-C added) mean was very large, 121 μg N₂O-N g^?1 sludge h^?1. These DEA rates are consistent with the previously reported high levels of nitrogen removal by denitrification from these wetlands, especially when the wastewater was partially nitrified. Stepwise regression using distance within the wetland, wastewater nitrate, and wastewater ammonia explained much of the variation in DEA rates. In both bulrush and cattail wetlands, there were zones of very high potential DEA.     

Dephenolisation of olive mill wastewater using adapted Trametes versicolor

Olive mill wastewater (OMW) is an effluent of the olive oil extraction process. The large volumes involved, along with the high phenolic content and chemical oxygen demand, cause major environmental problems. The presence of phenolics limits the effectiveness of aerobic or anaerobic treatment of this wastewater. In most of the studies performed on OMW, the concentration of phenolics is reduced by diluting the OMW prior to biological treatment, which leads to an increase in waste volume. Therefore, the aim of this work was to investigate the possibility of reducing the phenolic content without dilution and without any addition of nutrients or pretreatment by using the white-rot fungi Trametes versicolor FPRL 28A INI. Through an adaptation process, the fungus was able to remove 78% of total phenolics in shake flask experiments and 39% in static culture using undiluted OMW medium. In continuously stirred tank reactor (CSTR) conditions, 70% of total phenolics removal was achieved. Analysis with GC–MS showed that all simple phenolics disappeared from the medium after the 8th day of cultivation at an 0.25 vvm aeration rate. The maximum activities of phenol degrading enzymes laccase and manganese peroxidase (MnP) obtained under these conditions were 762.14 ± 42.11 and 97.80 ± 8.11 U l^?1 respectively.     

Effect of the aeration rate and agitation speed on β-carotene production and morphology of Blakeslea trispora in a stirred tank reactor: mathematical modeling

The effect of aeration rate and agitation speed on β-carotene production and morphology of Blakeslea trispora in a stirred tank reactor was investigated. B. trispora formed hyphae, zygophores and zygospores during the fermentation. The zygospores were the morphological form responsible for β-carotene production. Both aeration and agitation significantly affected β-carotene concentration, productivity, biomass and the volumetric mass transfer coefficient (K_La). The highest β-carotene concentration (1.5 kg m⁻³) and the highest productivity (0.08 kg m⁻³ per day) were obtained at low impeller speed (150 rpm) and high aeration rate (1.5 vvm). Also, maximum productivity (0.08 kg m⁻³ per day) and biomass dry weight (26.4 kg m⁻³) were achieved at high agitation speed (500 rpm) and moderate aeration rate (1.0 vvm). Conversely, the highest value of K_La (0.33 s⁻¹) was observed at high agitation speed (500 rpm) and high aeration rate (1.5 vvm). The experiments were arranged according to a central composite statistical design. Response surface methodology was used to describe the effect of impeller speed and aeration rate on the most important fermentation parameters. In all cases, the fit of the model was found to be good. All fermentation parameters (except biomass concentration) were strongly affected by the interactions among the operation variables. β-Carotene concentration and productivity were significantly influenced by the aeration, agitation, and by the positive or negative quadratic effect of the aeration rate. Biomass concentration was principally related to the aeration rate, agitation speed, and the positive or negative quadratic effect of the impeller speed and aeration rate, respectively. Finally, the volumetric mass transfer coefficient was characterized by the significant effect of the agitation speed, while the aeration rate had a small effect on K_La.     

Simultaneous removal of chlorpyrifos and dissolved organic carbon using horizontal sub-surface flow pilot wetlands

Rural areas of developing countries require low-cost treatment systems to purify wastewater which is contaminated with pesticides and organic matter. This work evaluated for six months the simultaneous removal of chlorpyrifos and dissolved organic matter in water using four horizontal sub-surface flow constructed wetlands (SSFCW) at a pilot scale, that were planted with Phragmites australis at 20 ± 2 °C water temperature. In each wetland, three concentrations of chlorpyrifos and three of dissolved organic carbon (DOC) were tested by liquid chromatography and an organic carbon analyzer respectively. The pesticide and DOC were added to the wetlands in synthetic wastewater. For the experiments, four wetlands of equal dimensions were used, with granular material of igneous rocks, 3.9–6.4 mm in diameter and at a depth of 0.3 m with a layer of water 0.2 m deep. For each treatment, regular sampling was carried out for the influent and effluents. As a supporting feature NH₄⁺, NO₃^? and PO₄^3? were quantified and in situ measurements of dissolved oxygen (DO), pH, electrical conductivity, water temperature and redox potential were taken. The overall removal of the chlorpyrifos (92.6%) and DOC (93.2%) was high, as was DOC removal as a function of pesticide concentration in the influent. The minimum magnitude (92.0%) was reached with 425.6 μg L^?1 of chlorpyrifos and, with the highest pesticide removal (96.8%). At lower concentrations of the agrochemical, DOC removal increased. The removals were possibly due to mineralization processes, biological decomposition and sorption in plants. These findings demonstrate that SSFCW are capable of simultaneously removing dissolved organic matter and organophosphate pesticides such as chlorpyrifos, which indicate that chlorpyrifos did not interfere with the removal of organic material.     

Dominance of yeast in activated sludge under acidic pH and high organic loading

Flow cytometry-fluorescent in situ hybridization (FC-FISH) was used to investigate the effect of controlled pH and/or varied organic loading on the content of yeast and bacterial cells in an activated sludge system (AS) individually operating in continuous and batch mode for treatment of high-strength industrial wastewater. Specifically, we attempted to develop a yeast-predominant activated sludge system (Y-AS). For the batch-mode AS, bacteria-dominated AS (B-AS) obtained at pH 6.5–7.5 induced higher chemical oxygen demand (COD) removal than Y-AS obtained at acidic pH (5.0–6.0 and 4.0–5.0). For the continuous-mode AS operating at COD loadings of 2.5–2.8 kg COD m⁻³ d⁻¹, it was difficult to achieve a Y-AS solely by controlling the pH level at 7.0 to 5.1 then to 4.1 because bacteria stably accounted for greater than 98% of the total cells, regardless of the pH levels. Therefore, the effects of varied COD loadings (2.1, 8.7 and 21.0 kg COD m⁻³ d⁻¹) on continuous-mode AS operation at acidic pH (4.5) was investigated. Both acidic pH and high COD loading levels were found to be prerequisites for yeast to dominate the sludge microbial community in the continuous-mode AS.