Subscribe Now Subscribe Today
Research Article

The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge

Ali Akbar Azimi and Mirzaman Zamanzadeh
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail

The objective of this pilot plant research was to study the causes of and control strategies for filamentous sludge bulking in the Esfahan South Wastewater Treatment Plant. In this regard three pilot scale activated sludge reactors with different hydraulic regimes including plug flow, completely-mixed and completely-mixed with selector, which their dispersion numbers were 0.06, 0.77 and 0.2, respectively, were constructed. During the research the complete mix pilot faced with filamentous sludge bulking resulted from the predominant growth of Microthrix parvicella and Type 1701 bacteria which were the same in the full scale plant. For the other reactors, which had better sludge settleability, the type and abundance of filamentous bacteria were different according to the operational conditions. On the basis of the research results, Sludge Volume Index (SVI) was 80-160, 25-200, 65-400 mL g-1 for plug flow, completely-mixed with and without selector, respectively. Also, the most appropriate hydraulic detention time for filamentous bulking control in selectors under aerobic and anaerobic conditions was evaluated 15 and 30 min, respectively. With respect to the results, filamentous bulking can be controlled in ESWTP by changing the hydraulic regime of the reactor from completely mixed to plug flow or by constructing selectors prior to completely mixed aeration tank.

Related Articles in ASCI
Search in Google Scholar
View Citation
Report Citation

  How to cite this article:

Ali Akbar Azimi and Mirzaman Zamanzadeh, 2006. The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge. Pakistan Journal of Biological Sciences, 9: 345-349.

DOI: 10.3923/pjbs.2006.345.349



The use of activated-sludge selectors for bulking control and nutrient removal has become widespread in recent years (Fisher and Boyle, 1999).

One of the most common filamentous microorganisms in nutrient removal activated sludge plants is Microthrix parvicella (Andreasen and Nielsen, 2000; Blackbeard et al., 1988). Recent studies (Eikelboom, 1994; Gabb et al., 1996) indicate that selectors were unsuccessful in controlling low F/M (Food to microorganism ratio) bulking. On the other hand there are reports in the literature that claim effective M. parvicella control with the use of anaerobic selectors (Pujol and Canler, 1994; Daigger et al., 1985).

Also, previous works on aerobic selectors (Cha et al., 1992) and anoxic selectors (Cha et al., 1992; Shao, 1987) showed that it was possible to control Nocardioform organisms. However, no Nocardioform organism control has been observed when using anaerobic selectors (Mamais, 1991; Pitt and Jenkins, 1990). Pure–culture experiments on Gordona amarae (Blackall et al., 1991) suggested that anaerobic selectors should control this organism in activated sludge.

Plug flow configuration is another way to control filamentous bulking. Pilot scale experiments indicate that plug flow configuration produces a sludge with significantly better settling characteristics compared with that of complete mix systems, suggested that the best control strategy for suppressing the M. parvicella growth can be achieved by adoption of continuous plug flow reactors (Mamais et al., 1998). The introduction of a substrate gradient in treatment plants by applying selectors or plug flow pattern has generally turned out to reduce the growth of M. parvicella (Heide and Pasveer, 1993).

Esfahan South Wastewater Treatment Plant is a conventional activated sludge plant with 122000 m3/day capacity which serves 720000 population of Esfahan city. Since its start-up in 1982, this plant has occasionally confronted with filamentous sludge bulking. Thus, this research project was carried out to find the bulking reasons in this plant and recommend design parameters for selectors to be built.


This study was done at on-site of Esfahan Wastewater Treatment Plant from September 2002 to March 2003.

Table 1: Influent wastewater characteristics to the aeration tank
Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge

Table 2: Operational parameters of pilot plants
Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge
*Hydraulic retention time of selector

Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge
Fig. 1: Schematic diagram of three pilot-scale activated sludge plants

Three pilot-scale activated sludge plants were constructed to study the effects of aeration tank configuration and selectors on filamentous sludge bulking (Fig. 1).

The pilot plants, which have been operated for 181 days, consisted of one plug flow unit and two completely-mixed flow units with and without selector. Each of the units had separate secondary clarifier and returned activated sludge line. The pilot plants were located in site of ESWTP and raw wastewater conveyed to the units after passing through screens, grit chamber and primary settling tank. Influent wastewater characteristics during the study are given in Table 1.

The plug flow unit was 120 cm long, 50 cm wide and 60 cm deep and had a liquid depth of 40 cm. To provide a plug flow hydraulic regime, ten baffles were put in the aeration tank. The first baffle distance to the influent pipe was 15 cm and distance between each two other baffles was 10.5 cm. Aeration tank dimensions in complete mix pilot plants with and without selector were similar to the plug flow unit, apart from baffles in the aeration tank. Air requirements in aeration tanks were supplied by a blower and using 11 coarse bubble diffusers in each tank.

Selector made of plastic and its volume in different stages of the study was changed in a way that hydraulic detention times of 120, 60, 30 and 15 min could be achieved.

Aeration tanks effluent flowed to the secondary clarifiers by gravity. Secondary clarifiers made of galvanized sheet that had two panes of glasses in two sides in order to control sludge blanket height in the clarifiers. Dimensions of each clarifier were 25 cm long, 25 cm wide and 96 cm deep and had 60 L capacity with 2 h detention time.

Returned activated sludge sent to the aeration tank through plastic hose using a pump. Flow rate to each unit was 0.5 l min-1 which was adjustable by slide valves.

Since abundant filamentous microorganisms existed in the aeration tank of ESWTP, it was attempted to produce initially activated sludge in the pilot units without seeding. Therefore, influent wastewater was used to fill the aeration tanks of pilot units and then exclusively aerated for a period of 72 h. After that, wastewater gradually and steadily sent with low flow rate to the aeration tanks. Different stages of aeration and hydraulic detention times in each pilot plant are shown in Table 2. All experiments were conducted at temperatures between 18 to 27°C and influent wastewater temperatures to the selector were 13 to 23°C.

During the study waste sludge flow rate was determined on the basis of 10 day sludge age and it was continuously drained from the aeration tank. Dispersion number for aeration tank measured by the method described in the UK Water Research Center Technical Report (Tomilinson and Chambers, 1979). All laboratory analyses were performed following Standard Methods (Anonymous, 1992) procedures.


Plug flow reactor: After establishing a fixed level of Mixed Liquor Suspended Solids (MLSS) in the aeration tank, all compartments of the plug flow pilot-scale reactor were aerated in which the initial value of Sludge Volume Index (SVI) was 150 mL g-1, but after day 53, SVI value reached 195 mL g-1, as shown in Fig. 2. However, there was again a decrease in SVI value after 10 days and it reached 160 mL g-1. In general, sludge settleability was desirable during this phase. Microscopic observations indicated that filamentous bacteria relatively developed in this phase in which Type 1701 was the dominant bacteria. It should be stated that for this phase Dissolved Oxygen (DO) concentration and F/M ratio in the first part in the aeration tank were 1.3 mg L-1 and 1.76 kg BOD5 kg-1 MLSS day-1, respectively.

During the next phase, aeration of the first compartment of the tank was ceased, while aeration was continued to the rest 10 compartments. As a result, SVI value considerably reduced and reached 130 mL g-1. Microscopic investigations in this phase indicated that filaments did not developed and there were flocs with good settling characteristics. During the last phase, aeration of the first and second compartments of the tank was ceased and the rest 9 compartments were aerated for a period of 33 days. In this phase SVI value considerably reduced and reached 84 mL g-1.

Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge
Fig. 2: SVI variations in the plug flow reactor (SVI: Sludge Volume Index; MLSS: Mixed Liquor Suspended Solids)

Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge
Fig. 3: SVI variations in completely-mixed reactor with selector (SVI: Sludge Volume Index; MLSS: Mixed Liquor Suspended Solids)

Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge
Fig. 4: SVI variations in completely-mixed reactor (SVI: Sludge Volume Index; MLSS: Mixed Liquor Suspended Solids)

It should be pointed out that during the last phase pin-point flocs developed and resulted in high turbidity and SS in treated effluent. Therefore, due to high turbidity in effluent compared to the previous phases, removal efficiency of BOD5, COD, SS, TP and TN decreased 88.4, 62.2, 88.7, 76.7 and 83.2%, respectively (Table 3).

Completely mixed reactor with selector: To determine the effects of HRT (Hydraulic Residence Time) and DO concentration on selector behavior six stages with different HRTs and types of selector (aerobic or anaerobic) were studied, as indicated in Table 2. When the selector operated in anaerobic conditions in order to determine the effects of HRT of selector on sludge settleability, different HRTs of 120, 60 and 30 min were experienced in which SVI values were 210, 176 and 127 mL g-1, respectively. Nitrate-nitrogen concentrations during the various stages of selector were determined in the recycle line in which the average concentrations for anaerobic and aerobic conditions were 0.19 and 1.7 mg L-1, respectively. However, it was observed that continuing to decrease HRT from 60 to 15 min (in the sixth stage) resulted in SVI increase.

On the other hand, when the selector operated in aerobic condition, reducing HRT from 60 (fourth stage) to 15 min (fifth stage) resulted in SVI decrease from 171 to 146 mL g-1. Microscopic observations is provided in Table 4, which shows filaments type and their abundance. Based on theses observations the dominant filamentous microorganisms in anaerobic conditions were M. parvicella, H. hydrosis and Thiothrix sp. for aerobic conditions were M. parvicella, 1701 and 0041.

Table 3: Removal efficiency for plug flow, complete mix with and without selector (%)
Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge
*No. of samples

Table 4: Predominant filamentous bacteria and their abundance for different stages of completely-mixed reactor with selector
Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge
Image for - The Effect of Selectors and Reactor Configuration on Filamentous Sludge Bulking Control in Activated Sludge

As it is seen, M. parvicella was present in anaerobic and aerobic conditions. For anaerobic selectors (stages 1, 2, 3 and 6) the most appropriate sludge settling characteristics (SVI = 127 mL g-1) occurred in the third stage operated anaerobicly with HRT of 30 min, as depicted in Fig. 3, whilst other stages could not completely control the development and growth of filamentous bacteria.

By reducing HRT in the selector from 2 h (first stage) to 1 h (second stage), Thiothrix growth has been deceased, as seen in Table 3. Microscopy indicated that due probably to low DO concentration (Richard, 1989) in the selector Type 1701 was the dominant filament in the fourth stage Table 4 provides the results obtained in completely–mixed pilot experiments with selector.

Completely–mixed pilot-plant reactor without selector: Steady-state conditions were achieved in the reactor after 34 days; afterwards two other stages were studied as shown in Table 2. The second stage lasted for 124 days after which lime injected to the influent flow to the secondary clarifier tank (third stage) so as to determine its effect on SVI. SVI variations for the reactor presented in Fig. 4. During the second stage SVI value ranged from 150 to 480 mL g-1.

Microscopic observations indicated that M. parvicella, H. hydrosis and Type 0041 were dominant filamentous bacteria in the second stage (Table 3).

In the third stage lime added in a way that its concentration in the influent wastewater to the aeration tank was 10 mg L-1, as a consequent SVI value considerably reduced from 480 to about 290 mL g-1 (Table 3).


Despite reasonable settling characteristics during the second stage of plug flow reactor, Type 1701 growth for a period of 10 days resulted in settling deterioration and SVI increase (SVI= 195 mL g-1). In plug flow configuration it is more likely that high F/M ratio results in low DO concentration in the first part of the aeration tank; consequently, Type 1701 associated strongly with low aeration basin DO concentration (Richard, 1989) found to be the dominant filament. Following lack of aeration to the first and second compartment of plug flow reactor, Type 1701 growth and development significantly reduced as this filamentous organism requires oxygen for growth and does not proliferate under anaerobic conditions (Richard, 1989). During the second stage M. parvicella, which can tolerate anaerobic conditions (Pitt and Jenkins, 1990), found to be dominant bacteria; but its presence was low and the filament abundance degree was 2 (some). Increased anaerobic conditions in the third stage (1.7 h) resulted in almost complete removal of filamentous bacteria in a way that pin-point flocs developed and SS removal efficiency significantly decreased (Table 3). In general, based on the results of this study a plug flow configuration creates sludge with better settling prosperities, which is in good agreement with other studies (Mamais et al., 1998) compared with those of completely-mixed reactors with and without selectors. Since high F/M ratio in the first part of plug flow reactors exists and substrate gradients can be achieved (Andreasen and Nielsen, 2000), they have superiority in filamentous bulking control and in production of good settling characteristics.

The subjective scoring system (Richard, 1989) was used to determine filamentous bacteria abundance in each stage of selector. Comparing these two Figures it shows that filamentous bacteria abundance reduced from 4 (very common) to 2 (some), indicating the role of selector in controlling filaments.

According to Table 2, low F/M in the first and second stage of selector led to the growth of M. parvicella, which is able to develop in low F/M and anaerobic conditions (Jenkins et al., 1993). Since Thiothrix sp. is associated with septic wastes or wastes containing sulfides (Jenkins et al., 1993), the presence of this filament in the first stage (influent SO4 Conc. 45 mg L-1) indicates that due probably to high HRT in this stage (2 h) bacteria started reducing sulfates (SO4 Conc. 13.2 mg L-1) to sulfides stimulating Thiothrix development. On the other hand, by decreasing HRT from 2 h (first stage) to 1 h (second stage) resulted in a decline of Thiothrix. In the last stage low HRT (15 min) caused the influent substrate to leave the selector without being influenced by bacteria; thus low F/M in the aeration tank developed and resulted in reappearance of M. parvicella. In the fourth stage that the selector operated in aerobic condition the dominant filament due to low DO concentration and high F/M ratio (Table 2) was Type 1701.

Based on the results it can be concluded that construction of a selector prior to completely-mixed aeration tank can control filamentous sludge bulking. The results of the pilot-scale study indicate that the optimum HRT for anaerobic and aerobic selectors were 30 and 15 min, respectively. Meanwhile, plug flow configuration shows superiority in filamentous sludge bulking control over complete mix reactors with and without selector. SVI values for plug flow reactor and complete mix reactor with and without selectors ranged 80-160, 125-200 and 260-400 mL g-1, respectively. The dominant filamentous bacteria in the ESWTP were identified as M. parvicella, H. hydrosis, Thiothrix sp. and Beggiatoa. Also, addition of lime into the secondary clarifier influent can be used as a temporary solution in filamentous sludge bulking control. As a result of 10 mg L-1 lime addition, SVI decreased from 480 to 290 mL g-1.


1:  Anonymous, 1992. American Water Works Association and Water Environment Federation: Standard Methods for the Examination of Water and Wastewater. 18th Edn., Washington DC., New Yark

2:  Andreasen, K. and P.H. Nielsen, 2000. Growth of Microthrix parvicella in nutrient removal activated sludge plants: Studies of in situ physiology. Water Res., 34: 1559-1569.
CrossRef  |  

3:  Blackbeard, Jr. D., M.D. Gabb, G. Ekama and G.V.R. Marais, 1988. Identification of filamentous organisms in nutrient removal activated sludge plants in South Africa. Water SA., 14: 29-34.

4:  Blackall, L.L., V. Tandoi and D. Jenkins, 1991. Continuous culture studies with nocardia amarea from activated sludge and their implications for foaming and control. Res. J. Water Pollut. Control Fed., 62: 44-49.

5:  Cha, D.K., D. Jenkins, W.P. Lewis and W.H. Kido, 1992. Process control factors influencing nocardia populations in activated sludge. Water Environ. Res., 64: 37-43.
Direct Link  |  

6:  Daigger, G.T., M.H. Robbins and B.R. Marshall, 1985. The design of a selector to control low-F/M filamentous bulking. J. Water Pollut. Control Fed., 16: 35-53.
Direct Link  |  

7:  Eikelboom, D.H., 1994. The Microthrix parvicella puzzle. Water Sci. Tech., 29: 271-279.

8:  Fisher, M.J. and W.C. Boyle, 1999. Effect of anaerobic and anoxic selectors on oxygen transfer in wastewater. Water Environ. Res., 71: 84-93.
CrossRef  |  

9:  Gabb, D.M.D., G.A. Ekama, D. Jenkins, M.C. Wentzel, T.G. Casey and G.V.R. Marais, 1996. Filamentous organism bulking in nutrient removal activated sludge systems. Paper 4: System configurations and operating conditions to develop low F/M filament bulking sludges at laboratory scale. Water SA., 22: 127-137.

10:  Heide, B. and A. Pasveer, 1993. Oxidation ditch: Prevention and control of filamentous bulking. H2O, 7: 373-377.

11:  Jenkins, D., M.G. Richard and G.T. Daigger, 1993. Manual on the Causes and Control of Activated Sludge Bulking and Foaming. LEWIS Publisher Inc., New York, pp: 123-137

12:  Mamais, D., A. Andreadakis, C. Noutsopoulos and C. Kalergis, 1998. Causes of and control strategies for, Microthrix parvicella bulking and foaming in nutrient removal activated sludge systems. Water Sci. Technol., 37: 9-17.
CrossRef  |  

13:  Mamais, D., 1991. The effects of temperature and mean cell residence time on enhanced biological removal by activated sludge. Ph.D. Thesis, Department Civ. Environment Eng., University California Berkeley, pp: 29-33.

14:  Pitt, P. and D. Jenkins, 1990. Causes and control of nocardia in activated sludge. Res. J. Water Pollut. Control Fed., 62: 143-150.
Direct Link  |  

15:  Pujol, R. and J.P. Canler, 1994. Contact zone: French practice with low F/M bulking control. Water Sci. Tech., 29: 221-228.

16:  Richard, M.G., 1989. Activated Sludge Microbiology. Water Pollution Control Federation, Alexandria, Virginia, USA., pp: 32-43

17:  Shao, Y.J., 1987. The mechanism and design of anoxic selectors for the control of low F/M filamentous bulking. Ph.D. Thesis, Department Civ. Environment Eng., University California Berkeley, pp: 43-47.

18:  Tomilinson, E.J. and B. Chambers, 1979. The effects of longitudinal mixing on the settleability of activated sludge. WRC Technical Report TR 122, Water Research Center, UK, pp: 46.

©  2021 Science Alert. All Rights Reserved