SciELO - Scientific Electronic Library Online

vol.41 issue4Effect of altitude on erosive characteristics of concurrent rainfall events in the northern KwaZulu-Natal Drakensberg author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • On index processSimilars in Google


Water SA

On-line version ISSN 1816-7950
Print version ISSN 0378-4738

Water SA vol.41 n.4 Pretoria Jul. 2015 



Recent developments in biological nutrient removal*



George A Ekama**

Water Research Group, Department of Civil Engineering, University of Cape Town, Rondebosch, 7700, Western Cape, South Africa




Biological nitrogen (N) and phosphorus (P) removal from municipal wastewater with the activated sludge (AS) system has been the preferred technology for the last 40 years. While several questions remain to be answered for more consistent, reliable and stable performance for enhanced biological P removal (EBPR), recent developments in this technology have focused on (i) increasing capacity and reducing the plant space footprint and (ii) improving N removal. To increase capacity and reduce AS system space, (a) integrated fixed-film activated sludge (IFAS), (b) external nitrification, (c) membrane, (d) aerobic granulation BNR systems and (e) more efficient N removal bioprocesses (anammox and nitrite shunt) have been developed. With IFAS, fixed media are added to the aerobic activated sludge reactor to make nitrification independent of the suspended AS sludge age. With external nitrification, nitrification is achieved in a side-stream fixed media reactor, which removes the size-defining nitrification process from the suspended AS system and halves its sludge age, improves sludge settleability and increases capacity. With membranes, secondary settling tanks are replaced with in-reactor membranes for solid-liquid separation. With aerobic granulation, the activated sludge process is controlled to form fast-settling granules comprising heterotrophs, nitrifiers, denitrifiers and phosphorus-accumulating organisms (PAOs) in a sequencing batch (SBR) type reactor - the granules not only settle fast but the SBR-type operation also removes the need for secondary settling tanks allowing higher reactor solids concentrations and hence smaller reactors. To achieve N removal more efficiently, methods are being developed to (i) short-circuit nitrification-denitrification (ND) by preventing nitrate formation and enforcing ND over nitrite - this requires less oxygen and organics than ND over nitrate allowing lower N concentrations to be achieved for the same influent organics concentration and oxygen supply, and (ii) encouraging the growth of anammox bacteria in the activated sludge which remove N autotrophically by combining ammonia and nitrite to form nitrogen gas - this halves oxygen demand for nitrification and requires no organics. These recent developments in BNR technology are briefly reviewed in this paper.




The size, footprint and energy consumption of the activated sludge (AS) system is governed by the requirement of the system to remove nitrogen - if nitrogen does not need to be removed by nitrification-denitrification (ND), for example, when 100% source separation of urine is practised, the AS system could be much smaller and consume much less energy (Ekama et al., 2010). The sludge age of the biological nutrient removal (BNR) AS system is governed by the slowest growing organisms in the system (Ekama, 2011). When nitrogen removal is required, these are the autotrophic nitrifiers, which nitrify ammonia to nitrate. So the sludge age of the BNRAS system needs to be greater than the minimum required for the nitrifiers to be sustained in it. Furthermore, the effluent ammonia concentration, being a dissolved constituent, is strongly affected by influent ammonia cyclic flow and load conditions. The further the sludge age is beyond the minimum for nitrification, the greater is the attenuation in effluent ammonia concentration relative to influent ammonia cyclic flow and load variation (Ekama and Wentzel, 2008).

Once the sludge age is selected to ensure efficient nitrification, the influent organic (COD) and inorganic (ISS) loads fix the mass of sludge (TSS) in the biological reactor and the oxygen demand. The greater the sludge age, the greater the mass of sludge in the reactor and the higher the oxygen demand (Ekama and Wentzel, 2008). The volume of the reactor and the surface area of the secondary settling tanks are then determined by selecting the reactor TSS concentration that minimizes the combined cost of biological reactor and secondary settling tank for a selected sludge settleability (Ekama et al., 1997). While the nitrifiers add to the total oxygen demand of the system, they have a negligible effect on the reactor TSS concentration - the nitrifier biomass makes up less than 2% of the reactor TSS (Ekama and Wentzel, 2008). So the impact of the nitrifiers is that they dictate the sludge age and thereafter the organic load to be removed, and sludge settleability dictates the AS system size. If nitrification can be achieved at lower sludge ages and the solid-liquid separation can be made less sensitive to sludge concentration and settleability, then the BNRAS system can be significantly reduced in size (or capacity increased for an existing system). The developments in BNR technology over the past 30 years have all focused on looking for different ways of getting around these two issues and have resulted in some remarkable discoveries and inventions, some of which are still on-going. For example, (i) integrated fixed-film activated sludge (IFAS), (ii) external nitrification, (iii) membrane solid-liquid separation in membrane bioreactors (MBR), and (iv) aerobic granulation BNR systems have been developed. To improve N removal, methods are being developed to (v) enhance the nitrite shunt, which 'short-circuits' ND by suppressing nitrate formation and forcing ND over nitrite, and (vi) encourage the growth in the BNR reactor of anammox bacteria, which remove N autotrophically by combining ammonia and nitrite to form nitrogen gas. These six inventions and developments have been made for non-saline water 'aerobic' activated sludge systems and are briefly described in this paper.

Other novel inventions and developments, such as the SANI system for saline seawater treatment arising from seawater toilet flushing, which offers major reductions in sludge production and oxygen demand (Lu et al., 2012), are not discussed. No doubt, many more discoveries and inventions for fresh and saline wastewater treatment will still be made and developed in future.



To reduce the sludge age required for nitrification, static or moving fixed media (e.g. solid AccuFAS or Bio-Blok), suspended rope (e.g. Ringlace) or moving bed (e.g. Kaldness) carriers are added to the aerobic reactor (Wanner et al., 1988; Sen et al., 1994; Thomas et al., 2008). Such systems are called integrated fixed-film activated sludge (IFAS). The nitrifiers grow on the fixed media establishing a population permanently resident in the aerobic reactor. These nitrifiers are neither part of the unaerated sludge mass fraction (fraction of sludge mass in reactor not aerated) nor of the suspended mixed liquor. The system sludge age therefore can be reduced. Such a reduction in system sludge age is particularly beneficial for low temperature wastewaters (10-15°C).

The overall objective of IFAS is to increase the treatment capacity and nitrification performance of the existing suspended AS system by adding fixed media to it instead of extending the plant with additional reactors. Fixed film systems are well known for biological treatment of ammonia (and dissolved organics), particularly in cold climates, because the biomass mediating the bioprocesses on the fixed media are retained in the system and not removed via the waste sludge. Hence, during the cold wastewater temperature operation, the majority of the ammonia-oxidizing bacteria are found on the media and good nitrification performance is maintained by the system even though the sludge age of the system (suspended AS) is lower than the minimum required for nitrification.

Generally, the static media are placed above the bubble aeration system so that the bulk liquid can make its way through the media providing contact of the wastewater constituents with the biomass on the media. Free-floating media(e.g. Kaldness) are generally small plastic buoyant media which are placed in a reactor and move freely throughout the entire aeration basin volume. Since these media move freely in the reactor, screens are required to retain them in the reactor so that they do not escape with the effluent. The approach velocity of the media to the screens is important and must be sufficiently low to prevent them accumulating on the screens and to keep them moving around the reactor.

There are several full-scale IFAS systems in operation. The suspended medium solids retention time (SRT or sludge age) versus wastewater temperature of these plants are shown in Fig. 1 (Ødegaard et al., 2014). All of these plants are operating well below the minimum suspended medium SRT for nitrification recommended by the ATV (Abwasser Technischen Vereinigung) 131 guideline (blue line). The red line in Fig. 1 represents the average nitrification performance of these IFAS plants. If these plants were conventional suspended medium AS systems, then the maximum specific growth rate of nitrifiers at 20°C (μAm20) and temperature sensitivity coefficient (θμ) in the minimum sludge age for nitrification (Rsm) equation, Rsm=1/ {μAm20(θμ)(T-20) - bA20(1.03)(T-20)}, that best fits the red lireta Fig. 1 are μAm20= 1.10 /d and θμ = 1.143. This temperature sensitivity is quite close to that used for nitrification in suspended medium AS, i.e. θμ =1.123 (Ekama and Wentzel, 2008), but the μΑm20of 1.10 /d is much higher, at least double that used for nitrification in suspended medium AS (the ATV blue line in Fig. 2a has best μAm20= 0.545/d and θ = 1.148). This indicates that the fixed media have at least halved the minimum system sludge age for nitrification, which makes a significant volume saving for the activated sludge reactor. Although the DO concentration in IFAS reactors is required to be high (5-6 mg O/) for effluent ammonia concentration below 0.5 mg N/, the oxygen transfer rate is increased by the presence of the fixed media which offsets some of the aeration energy required by the high DO.



Some media surface specific nitrification rates (rn) are reported in the literature. Zimmermann et al. (2003) found r n= rnmax[1-exp(-k.L )] where rnmax= 1.30 gFSA-N/(m2.d) and k = 0.93. Lnis the ammonia loading rate in gFSA-N/(m2-d) and ranged between 0.44 and 1.65 gFSA-N/(m2-d). Their rates were measured at a DO concentration of 5 mgO/ and temperature of 15°C. Rusten et al. (2006) give a linear increase in rn with increase in DO concentration, increasing from 0.60 gFSA-N/ (m2.d) at 2 mgO/ to 2.1 gFSA-N/(m2-d) at 8 mgO/ at 15oC, zero organic loading and residual ammonia concentration >2.5 mgN/. Di Trapani et al. (2011) observed rn at 5.0 mgO/, 14°C and 3.4 days sludge of 0.92 gFSA-N/(m2.d). To achieve low effluent ammonia concentrations of around 0.5 mgN/, the ammonia removal rate (rn ), and hence also the ammonialoading rate (Ln), are significantly lower, i.e., rn is around 0.5 gFSA-N/(m2-d). Also, the nitrification rate (rn) decreases with increasing organic loading rate. Rusten et al. (2006) give a value of around 0.30 gFSA-N/(m2.d) at 8 mgO/, 15°C and 5 gBOD/(m2-d).

Placing the media in the middle section of the aerobic reactor has several advantages: Significant organic removal will have already taken place, the ammonia concentration is highest in early stages of the reactor favouring the nitrification capacity of the attached biomass, the DO may be reduced in the last compartment of the aerobic reactor so less DO is recycled back to the anoxic reactor, low intensity of mixing in the last compartment improves flocculation, and the last compartment is seeded with nitrifiers from the media increasing the suspended AS nitrification in the last compartment.



If nitrification can be achieved independently of the BNRAS mixed liquor, the system sludge age can be reduced from the usual 10 to 15 d to less than half, around 5 to 8 d. The reduction in sludge age increases the wastewater (WW) treatment capacity of the system by some 50% or, alternatively, reduces the biological reactor volume requirement per M WW treated by about a third, without negatively impacting either biological N or P removal: In fact, a reduction in sludge age increases both biological N and P removal per mass organic load (Wentzel et al., 1990) and this would be particularly beneficial for low-temperature wastewaters (10-15°C). Because nitrification is no longer required, the aerobic mass fraction is governed by the P uptake process, for which aerobic mass fractions can be smaller than for nitrification.

External nitrification can be achieved at wastewater treatment plants (WWTPs) where existing trickling filter (TF) plants have been extended with a BNRAS system or by adding nitrifying trickling filters to an existing BNRAS plant (Hu et al., 2000). There are many WWTPs with old TFs. Often at these WWTPs, to retain the benefit of the old TF, a proportion of influent WW is passed through the TF and the effluent (see Fig. 2a) is either (i) discharged to the BNRAS system for biological N and P removal (Van Huyssteen et al., 1990) - this removes organics, the 'fuel' for N and P removal and therefore decreases N and P removal, or (ii) is chemically treated to precipitate the P before discharge to the BNRAS system. This not only adds cost, but also reduces the alkalinity of the water and does not decrease the N load on the BNRAS system.

A significantly better system is obtained if the nitrification process is transferred to the TF and all of the WW flow discharged to the BNRAS system (Hu et al., 2000; Muller et al., 2005; Fig. 2b). A side-stream of mixed liquor is taken from the end of the anaerobic zone and passed through internal secondary settling tanks to remove the AS solids. The underflow sludge is discharged to the beginning of the anoxic zone and the overflow is passed onto the TF for nitrification. The nitrified TF effluent is then discharged to the anoxic zone for denitrification. In this way the TF assists the BNRAS system in its weakness, i.e., nitrification, rather than taking away from its strength, i.e., organics-driven biological N and P removal. Furthermore, the oxygen demand in the aerobic reactor is markedly reduced because nitrification no longer takes place there. Indeed, not only is the nitrification oxygen demand obtained 'free' outside the BNRAS system, but also the oxygen equivalent of the nitrate generated in the trickling filter helps to reduce the carbonaceous oxygen demand in the BNRAS system, by about a third. In fact, with external nitrification, the reduction in oxygen demand in the BNRAS system is much greater than when a third of the WW is bypassed to the trickling filter, as in existing TF/BNRAS systems (Fig. 2a). Therefore, by changing the TF to a nitrifying system as in Fig. 2b, the treatment capacity of the BNRAS plant is increased without having to increase aeration capacity and N&P removal are achieved on the full WW flow. If a TF plant is not available, it is possible to include plastic fixed media systems, the cost of which may be offset by the increase in WW treatment capacity.

At short sludge ages and small aerobic mass fractions, nitrifiers would not ordinarily be supported in the BNRAS system. However, nitrifiers are not completely excluded from the BNRAS system because they are seeded into the system from the TF effluent. Therefore, nitrification in the aerobic reactor still takes place and the nitrate concentration in the aerobic reactor is governed by the ammonia concentration that enters it. Provided the TF nitrifies well (Muller et al., 2005), this nitrate concentration is mainly from the ammonia which bypasses the TF via the internal settling tank underflow, and therefore will be relatively low. If the TF does not nitrify well and the residual ammonia concentration from it is high, then, if sufficient nitrifiers are present in the aerobic reactor, the nitrate concentration will be high, with the result that a significant nitrate concentration will be present in the underflow from the final settling tank. To protect the BEPR against this potential nitrate ingress to the anaerobic reactor, a pre-anoxic reactor is placed in the underflow to denitrify the nitrate (Fig. 2b). If sufficient nitrifiers are not present in the aerobic reactor, then the ammonia concentration in the aerobic reactor will only be partially nitrified with the result that return sludge nitrate concentration will be relatively low, but the effluent TKN concentration will be high, the concentration depending on the nitrification efficiency of the TF.

Tertiary nitrifying trickling filters (TNTFs), which are employed for nitrification only, and not for any organic material removal, are fairly common in the USA (Lutz et al., 1990). While certain problems with macro-fauna (snails, worms, larvae and flies), which reduce nitrification rates, have been encountered, high removals of ammonia have been economically achieved in TNTFs (Parker et al. 1989, 1995, 1996). This has also been found to be the case for rock media TFs (Muller et al., 2005). Therefore, implementing the external nitrification scheme (Fig. 2b) is entirely feasible.

Despite the significant differences in technology, it is interesting that the specific surface nitrification rates in ventilated TNTF systems, viz., around 1.0 gFSA-N/(m2-d) for plastic media (Parker et al., 1995) and 0.86 gFSA-N/ (m2-d) for rock media (Muller et al., 2005), are of a similar magnitude to those in IFAS systems (see section on IFAS above). The significant reduction in nitrification rate with increasing organic load in IFAS systems is also observed in NTFs.



Effective solid-liquid separation in suspended medium biological wastewater treatment (WWT) systems is an essential step in the process, because it has a major influence on effluent quality - in fact, secondary settling tanks (SSTs) are expected to achieve a 99.5% suspended solids removal to maintain an effluent suspended solids concentration < 20 mgSS/. While conventionally solid-liquid separation in activated sludge systems has been accomplished in SSTs, membranes are an increasingly attractive alternative due to decreasing costs, improved performance and life, moderate maintenance and control, smaller footprint of WWTP and improved effluent quality for reuse.

Membrane solid-liquid separation in place of sedimentation in SSTs offers several advantages for the activated sludge (AS) system in general, and for biological nutrient removal (BNR) systems in particular (Ramphao et al., 2005). These advantages are:

Insensitivity to sludge settleability and filamentous bulking; this is a significant advantage as BNR systems notoriously produce rather poor-settling sludges (DSVI~150 m /g) when aerobic mass fractions are low (< 60%).

Insensitivity to AS flocculation characteristics and hydraulic shear in the reactor as membranes retain all solids, which may include free-swimming bacteria, depending on pore size.

SSTs are not required, a WWTP footprint reduction.

High reactor concentrations of 8 to 10 gTSS/ (0.8 to 1.0%TSS) resulting in reduced reactor volumes compared with conventional BNR systems with SSTs (further footprint reduction); originally 12 to 18 gTSS/l reactor concentrations were proposed but these have progressively decreased due to the low aeration alpha value at such high TSS resulting in low oxygen transfer rates (kgO/kWh) -such high TSS concentrations are now only for the membrane tanks in separated AS and membrane tank systems.

Variable anaerobic, anoxic and aerobic mass fractions (within a range) by varying the inter-reactor recycle ratios (see below), allowing flexibility of the biological N and P removal as dictated by influent wastewater characteristics and discharge effluent N and P concentrations.

Production of high-quality effluent with reduced tertiary treatment requirements for industrial or horticultural reuse or reduced disinfection requirements for release to surface water.

Possibly obviate waste AS thickening when reactor concentrations are at the high end of the range for optimal membrane operation.

Generally, there are three types of membranes applied in activated sludge: (i) flat sheet (FS, Kubota), which are hollow rectangular panels covered with membrane sheets; (ii) hollow fibre (HF), which are 2 m long, small (1 mm) diameter tubes; and (iii) tubular. The first two are outside to inside membranes and the third is inside to outside. Flat sheet membranes are submerged vertically in the AS reactor in units typically comprising 50, 100, 150 or 200 panels (Fig. 3a). They operate under low trans-membrane pressures (100 to 800 mm water) and do not require pumping. Hollow-fibre membranes are also submerged but usually placed in a separate reactor and are 'vacuum' pumped to create a trans-membrane pressure (Fig. 3b).



Fouling of the membranes is reduced by coarse bubble aeration, which, in the case of the flat sheet membranes, also supplies some or all of the oxygen required for aeration of the AS. For the hollow fibre membranes separate aeration is provided for the activated sludge reactor (fine bubble) and membrane reactor (course bubble). To increase the scour effectiveness of the air and increase the critical flux, the solids (TSS) concentration in the reactor can be high, in the range of 12 to 18 gTSS/. This is advantageous because it reduces reactor volume and increases flux (saves capital cost) but is also disadvantageous because it strongly decreases the oxygen transfer rate in the activated sludge reactor (increases running cost) (Judd et al., 2008). This is the crux of the issue with MBR AS - finding an appropriate reactor concentration which minimizes total cost (capital + running). Flat sheet membrane units can also be placed one above the other in 'double-storey' stacks which increases the membrane surface area per unit reactor volume and air flow.

To size a membrane BNR system, the aerobic reactor volume and aeration requirements of the activated sludge system are imposed on those of the membranes (Ramphao et al., 2005). The biological reactor volume is governed by (i) organic and inorganic loads, (ii) sludge age, (iii) selected reactor MLSS concentration, (iv) wastewater characteristics and (v) incorporation of primary sedimentation or not. The membrane surface area, and hence the volume required to accommodate these membranes in the aerobic reactor, is governed by the peak wet weather flow (PWWF, QPWWF) and the critical flux (m3 effluent per m2 membrane area per day, jcrit) through the membranes. For normal municipal wastewater, flat sheet membranes have short-term (at PWWF) and long-term (balanced flow) critical operational fluxes (jmax) of 1.0 and 0.7 m3/(m2-d), respectively. In order to minimize membrane fouling, coarse bubble aeration is supplied below the membranes, which, together with the high TSS concentration, creates an effective scour over the membrane surfaces. This coarse bubble aeration also supplies some, or all, of the peak biological oxygen requirement. If the oxygen supplied by the membrane aeration system is insufficient to meet the peak biological oxygen demand, the aerobic reactor has to be enlarged to accommodate additional fine bubble aeration to supply the biological oxygen demand deficit. Therefore two criteria govern the volume of the aerobic reactor: (i) at low influent COD concentration, the volume required to accommodate the membranes, and (ii) at high influent COD concentration, the volume required to transfer the peak biological oxygen demand when the scour of the membrane units does not meet this demand. Accurate aeration information is therefore essential to correctly size the reactor. The impact of these two criteria on the sludge age and volume requirements of flat sheet membrane BNR systems with different anaerobic, anoxic and aerobic mass fractions, in comparison with conventional BNR systems with the same mass fractions, was evaluated by Ramphao et al. (2005) for three cases: i.e., diurnal flow and load variation with a PWWF to average dry weather flow (ADWF) ratio (ƒq) of 2:1 with single (Case 1) and double (Case 2) storey membrane layouts in the aerobic reactor, and an influent flow balanced case with a single-storey membrane layout (Case 3). These three cases were evaluated for treatment of normal raw and settled municipal wastewater.

A uniform distribution of sludge mass in membrane BNR systems will not occur, even in systems with a single recycle flow from the aerobic zone to the zone receiving the influent flow. For example, changing an anoxic-aerobic reactor modified Ludzack-Ettinger (MLE) ND system or a UCT system with SSTs to membrane solid-liquid separation systems (Fig. 4) will change the distribution of the sludge mass in the systems, the magnitude of change depending on the magnitude of the recycle ratios. The equations linking the volume and sludge mass fractions of common BNR configurations are given by Ramphao et al. (2005).

A significant advantage of multi-zone BNR systems with membranes in the aerobic reactor and fixed volumes for the anaerobic, anoxic and aerobic zones (i.e. fixed volume fractions) is that the mass fractions can be varied (within a range) by varying the inter-reactor recycle ratios. For example, in a UCT system (Fig. 4b) with anaerobic, anoxic and aerobic zone volume fractions of 0.25, 0.35 and 0.40 and an r-recycle ratio from the anoxic to the anaerobic zones of 1:1, the anaerobic, anoxic and aerobic zone mass fractions can be varied from 0 to 0.131, 0 to 0.366 and 1 to 0.503, respectively, by changing the a-recycle ratio from 0:1 to 5:1 (Fig. 5). Increasing the a-recycle ratio concomitantly increases the nitrate load on the anoxic reactor thereby increasing the denitrification and N removal as the anoxic mass fraction increases. Increasing the r-recycle ratio increases the anaerobic mass fraction (at the expense of the other two zone mass fractions) and increases (not proportionally) the P removal. This zone mass fraction flexibility is a significant advantage of membrane BNR systems over conventional BNR systems with SSTs because it allows changing the mass fractions to optimize biological N and P removal in conformity with influent wastewater characteristics and the effluent N and P concentrations required.



If required, the performance of membrane BNR systems can be simulated with current BNR activated sludge simulation models such as IWA ASM Nos 1, 2 and 2d (Henze et al., 1987) and BEPR (Henze et al., 1994, 1999) by returning the SST underflow into the aerobic zone from which the SST feed flow exits. Such simulations require a-priori information on the reactor and zone volumes and recycle flows, which can be determined with the usual steady-state design procedures (Wentzel et al., 1990, 1997, 2008). The only kinetic rate that was found to be influenced by reactor TSS concentration was the maximum specific growth rate of the nitrifiers (μAmT), which decreased with increasing TSS concentration relative to a parallel BNR system at a lower TSS concentration (4 gTSS/) (Parco et al., 2010). However, unlike conventional BNR systems with SSTs, the maximum specific growth rate of the nitrifiers (μAmT) does not usually define the size of the membrane BNR reactor at high TSS concentration (Ramphao et al., 2005) - sludge ages need to be longer than for conventional systems to generate sufficient TSS mass in the reactor if operated at elevated reactor TSS concentration.

Generally, the longer the sludge age of the membrane system, the greater the biological reactor volume as a percentage of the equivalent conventional system volume at 15 d sludge age and 4 gTSS/ aerobic zone concentration. Although ND EBPR systems generate more sludge (10-20%) and utilize less oxygen (5-10%) than ND systems (Ekama and Wentzel, 1997, 2008) and therefore require shorter sludge ages, the difference is not large enough to make a significant difference to the volume of the membrane ND EBPR system as a percentage of the equivalent conventional system volume. The aerobic mass fraction (ƒmaer) has the greatest impact on the sludge age of the system and therefore also the membrane reactor volume as a percentage of the equivalent conventional system volume - the lower the aerobic mass fraction, the greaterer the sludge age and the greater the percentage volume. For ƒmaer<0.50 and low wastewater strengths, the volume of the membrane BNR system increases above 50% of the equivalent conventional system volume, which erodes some of the volume reduction advantage of using membranes. Influent flow balancing significantly reduces the sludge age of the membrane BNR system and hence also the percentage of the equivalent conventional system volume. Double-storey membrane units (without flow balancing) also reduce the sludge age and the percentage volume of the equivalent conventional system, but not as significantly as influent flow balancing.

Combining membranes and SSTs for solid-liquid separation in the same BNR system is possible in a side-stream membrane reactor with a low sludge return flow (20%) to the reactor to harvest a high-grade effluent from the BNR reactor for re-use. Also, wasting activated sludge directly from such a side-stream reactor obviates waste sludge thickening, but the costs of this would have to be compared with other thickening systems such as dissolved air flotation. Gravity thickening of P-rich waste activated sludge (WAS) is counter-productive because it results in the return to the influent of an overflow with a very high Ρ concentration due to P release - MBR or flotation thickening keeps the WAS aerobic and the P in the phosphorus-accumulating organisms (PAO) during thickening.

Installing membranes for solid-liquid separation into biological nutrient removal (BNR) activated sludge (AS) systems makes a profound difference not only to the design of the BNR system itself, but also to the approach to the design of the whole WWTP.



Sludge settleability remains one of the most unpredictable, capricious and least controllable parameters of the suspended activated sludge system that governs the size of the secondary settling tanks (SST) (Ekama et al., 1997). While significant advances have been made in understanding the causes and control of filamentous organisms, such as control of specific filament organism group proliferation with kinetic and metabolic selection (Jenkins et al., 1993; Ekama et al., 1996; Wanner, 2003; Van Loosdrecht et al., 2008) or with non-specific control with toxicants like chlorine or ozone (Jenkins et al., 1993), the problem of filamentous organism proliferation (bulking) resulting in poor sludge settleability is far from solved. However, other developments such as membrane solid-liquid separation (see section on MBR above) and aerobic granular activated sludge (AGAS) essentially eliminate the sensitivity of the suspended AS system to sludge settleability. Full biological nutrient (N&P) removal is included in AGAS - in fact, the growth of PAO is essential to effective aerobic granulation (De Kreuk and De Bruin, 2004; De Bruin et al., 2004; De Kreuk, 2006).

The physical classification of the aerobic granules is (i) a minimum diameter of 0.21 mm, and (ii) a sludge volume index (SVI) at 5 min closely equal to that at 30 min. A photo of typical aerobic granules is given in Fig. 6.



The AGAS is a sequencing batch (SBR) operated system with 3 cycle steps (Fig. 7):



(i) Fill and draw - due to the fast settling velocity of the granules, wastewater can be fed into the reactor without significantly expanding the concentrated sludge bed on the reactor floor after the settling phase. This feature allows a plug flow feed which displaces the effluent at the top of the reactor like a settling tank weir overflow. Due to the high concentration of the settled sludge bed, any residual nitrate in it is rapidly denitrified creating anaerobic conditions in the bed. The feed into the bottom of the reactor into the 'anaerobic' settled sludge bed on the floor results in conversion of readily biodegradable organics to volatile fatty acids and selects for PAOs.

(ii) Aeration - during this step all the usual BNR biological reactions take place simultaneously. This is possible due to an oxygen gradient inside the granule, resulting in an aerobic outer shell and a core that is anoxic/anaerobic (Fig. 8). Nitrifying bacteria and PAO grow in the outer shell due the 'high' DO concentration from the bulk liquid resulting in nitrification of ammonia to nitrite/nitrate and P uptake. The nitrite/nitrate is denitrified to nitrogen gas in the inner part of the granule by heterotrophic and autotrophic denitrifiers. Due to the deep penetration of substrate into the granule during the feed phase, sufficient stored organics for denitrification are available in the core during aeration. Also, as a result of long-term anoxic conditions in the granule, anammox and nitrite-shunt bioprocesses are able to take place (see Figs 8 and 9 below).





(iii) Settling - to separate the treated wastewater and the granules, only a short settling time is required due to the very fast settling velocity of the granules. The reduced settling time minimises 'idle time' in the reactor (Gademan et al., 2010).

In comparison with the conventional AS, the benefits of the AGAS are:

All the BNR bioprocesses take place in the same reactor, including settling, making wide-ranging process configurations possible by manipulating the process conditions in the reactor - the process configuration is flexible and not 'cast in concrete' because the 'size' of the anaerobic, anoxic and aerobic 'zones' is determined not by concrete divisions but by a PLC controller which controls the cycle times and aeration rate so that these system parameters can be adjusted to obtain optimum system conditions for BNR.

A mixed liquor sludge concentration of 8 000-10 000mg/ (compared with 3 000-6 000 mg/ in conventional AS systems) allowing for smaller reactor size.

A granule settling rate of 8-12 m/h (compared with 0.5-1.5 m/h).

No separate settling tanks required (compared with 1 m2/ (m3-h)) which results in a much reduced plant footprint.

No recycle pumps (compared to a mixed liquor a-recycle of 3 - 5 x ADWF and an underflow s-recycle of 1 x ADWF) which results in energy savings and reduced maintenance.

No mixers in the anaerobic and anoxic reactors, once again reducing energy consumption and a reduction in maintenance.

Due to the granular nature of the waste activated sludge, its dewatering characteristics are better than conventional flocculent-activated sludge (Gademan et al., 2013).



Nitrogen removal is accomplished in wastewater treatment by nitrification and denitrification (ND). Nitrification is mediated by two groups of autotrophic nitrifying organisms: (i) the first, generically called the ammonia-oxidizing bacteria (AOB), utilize ammonia as electron donor and oxygen as electron acceptor and produce nitrite, and (ii) the second, generally called nitrite-oxidizing bacteria (NOB), utilize nitrite as electron donor and oxygen as electron acceptor and produce nitrate. In denitrification, most facultative heterotrophic organisms utilize organics as electron donor and nitrate as electron acceptor and produce nitrogen gas. Some facultative heterotrophs perform only the first step, viz., denitrify nitrate to nitrite, while others perform only the second step, viz., denitrify nitrite to nitrogen gas. Generally very little nitrite accumulation takes place in activated sludge and practically all nitrate formed by nitrifiers is denitrified by facultative heterotrophs to nitrogen gas. This nitrogen removal cycle is shown schematically in Fig. 9.

A generalised complete CHONPS, COD and charge mass balanced stoichiometry for a wide range of bioprocesses has been developed (Ekama, 2009), inter alia, methanogenesis, sulphidogenesis, autotrophic denitrification, and also for the nitrification and denitrification (both in 1 and 2 steps) and anammox bioprocesses relevant to this paper (Fig. 9). This generalized stoichiometry is not given in this paper, but the results relevant to the bioprocesses in Fig. 9 are. In Table 1 are given stoichiometric values for (i) the autotrophic nitrification bioprocesses #3, #3a, #3b, (ii) the heterotrophic denitrification bioprocesses #6, 6a, 6b utilizing methanol (CH3OH) as organic substrate, (iii) the anammox bioprocess #7, and (iv) aerobic heterotrophic bioprocesses #4 utilizing methanol as organic substrate for (i) biomass net yield (E) [i.e. fraction of substrate electron donating capacity (EDC) becoming biomass (anabo-lism) and fraction (1-E) passed to the electron acceptor (catab-olism)], of facultative heterotrophs, nitrifiers (AOB and NOB), anammox and aerobic heterotrophs of 0.40, 0.15, 0.05 and 0.50, respectively, and (ii) a generic biomass composition of all bacterial species of C1H16811O04480N01655P0S0, which is obtained from COD, C, H, O ,N, P and S mass ratios (g/gVSS) of 1.481, 0.518, 0.0726, 0.3094, 0.100, 0.00 and 0.00 respectively.

From this stoichiometry, the oxygen requirements for the two steps of nitrification are 2.92 gO/gN ammonia transformed to nitrite and 0.97 gO/gN nitrite transformed to nitrate. The organics (COD) requirement for the two steps of denitrifica-tion are 1.91 gCOD/gN nitrate transformed to nitrite and 2.86 gCOD/gN nitrite transformed to nitrogen gas. So if all the N removal takes place over nitrite instead of over nitrate, the oxygen saving would be 25% and the methanol saving would be 40%. These are considerable savings, so significant research effort is being made to find ways to discourage the growth of NOB and encourage the growth of anammox bacteria in N-removal activated sludge systems.

Anaerobic ammonia-oxidizing (anammox) bacteria were first discovered about 20 years ago (Mulder et al., 1995) - for a history of its discovery see Kuenen (1995). These bacteria utilize ammonia as electron donor and nitrite as electron acceptor to form nitrogen gas. They have been found in the back anaerobic part of biofilm systems (Siegrist et al., 1998) and old rock media trickling filters (Wilsenach et al., 2014). This process has been called different names in the literature, such as 'oxygen limited aerobic nitrification denitrification' (OLAND), 'deam-monification' (DEMON) and 'completely autotrophic nitrogen removal over nitrite' (CANON) (Van Loosdrecht, 2008). Although anammox bacteria are anaerobic bacteria and very slow growers, they have also been found in activated sludge systems, which is remarkable considering the generally aerobic conditions of activated sludge. Interestingly, these bacteria are slightly denser than activated sludge and so accumulate in the denser and faster settling fraction of activated sludge (Winkler et al., 2011). This has led to the use of hydro-cyclones on waste activated sludge streams, wasting only the light fraction of activated sludge and returning the denser fraction to the reactor (Wett et al., 2013). In this way the anammox bacteria are retained, accumulate in the activated sludge and can make a considerable contribution to the N removal. This saves both oxygen and organics - N removal by anammox bacteria requires only about 1.42 gO/gN removed and zero organics consumption (Fig. 9) allowing N removal wastewater treatment plants to become energy self-sufficient.

Simultaneous nitrification-denitrification (SND) at low aerobic reactor DO has been observed for many years in many WWTPs. In instances where the low DO operation did not compromise nitrification, it was welcomed as additional N removal. It was believed to take place in the usual way over nitrate (Fig. 9). However, increasingly evidence is coming to light that at least some of this N removal is taking place over nitrite, called nitritation/denitritation or nitrite shunt. Exploitation of nitrite shunt is still limited in BNR plants because knowledge of design, control and operational conditions which stimulate it are not well known yet. To achieve nitrite-shunt in the mainstream WWTP requires suppression of the NOB activity. Under 'normal' WWTP conditions, the NOB are faster growers than the AOB and this is the reason why nitrification kinetics in many AS models are based on the AOB maximum specific growth rate (μAm20) converting ammonia to nitrate in one step. Finding the conditions resulting in suppression of NOB is a growing research topic. While conditions for AOB proliferation and NOB suppression are well known for reject water treatment (Hellinga et al., 1998), these conditions cannot easily be replicated in mainstream WWTPs. Compounding the difficulty, there is currently conflicting information on the role of DO in NOB suppression to stimulate nitrite-shunt: Low DO suppresses NOB Blackburne et al., 2008) while high DO was found to favour AOB over NOB in other mainstream studies (Cao et al., 2013; De Clippeleir et al., 2013; Remgi et al., 2014). Jimenez et al. (2014) describes a two-stage anaerobic(25%)-aerobic(75%) Phoredox (or A/O) plant (Southwest WWTP in St Petersburg, Florida, USA) treating an influent wastewater with a COD/TKN ratio of 7:1 and temperature between 23 and 30°C at low aerobic reactor DO (0.4 to 0.1 mgO/). This plant achieves a low effluent total inorganic N (2-4 mgN/) and low ortho-P (0.1 mgOP/). Specific nitrification [mgFSA-N/(gVSS.h)] and denitrification [mgNO3-N/ (gVSS.h)] rate tests revealed that the NOB were significantly suppressed due to the low DO operation and that nitrite-shunt occurs at the plant. Phosphorus release and uptake tests indicated that the low DO operation (and high temperature) did not adversely affect the biological P removal.



The drive to intensify the activated sludge (AS) system so that it requires less space and consumes less energy without compromising delivery of a high-quality treated effluent has led to some remarkable inventions and developments in biological nutrient removal over the past 2 decades. The main focus of these inventions and developments is to (i) maintain nitrifiers in the system at low sludge ages (Type A), (ii) make the system less sensitive to the capricious sludge settleability (Type B), and (iii) remove more nitrogen with less oxygen (energy) and organics (Type C). Six of these inventions and developments have been briefly described in this paper, viz.: (i) the integrated fixed-film activated sludge (IFAS) system (Type A), (ii) external nitrification (Type A), (iii) membrane solid-liquid separation (Type B), (iv) aerobic granulation BNR systems (Type B), (v) short-circuiting nitrification-denitrification (ND) by preventing nitrate formation and enforcing ND over only nitrite (nitrite-shunt) (Type C), and (vi) encouraging the growth of anammox bacteria in the activated sludge (Type C). With IFAS, fixed media are placed in the aerobic reactor (internal) to make nitrification independent of the suspended AS sludge age. With external nitrification, nitrification is achieved in a side-stream fixed media reactor, which removes the size-defining nitrification process from the suspended AS system. Both these systems (Type A) halve the suspended AS system sludge age and so increase wastewater treatment capacity. With membranes, secondary settling tanks are replaced with in-reactor membranes for solid-liquid separation. With aerobic granulation, the activated sludge is controlled to form fast-settling granules comprising heterotrophs, nitrifiers, denitrifiers and phosphorus-accumulating organisms (PAOs) in a sequencing batch (SBR) type reactor. With membranes and granules (Type B), secondary settling tanks are no longer required making a major reduction in wastewater treatment plant (WWTP) size. Short-circuiting ND by suppressing nitrite-oxidizing organisms and enforcing ND over nitrite and/or encouraging the growth of anammox bacteria in the activated sludge, which remove N autotrophically by combining ammonia and nitrite to form nitrogen gas, requires significantly less oxygen and organics (Type C) than ND over nitrate.



BLACKBURNE R, YUAN Z and KELLER J (2008) Partial nitrification to nitrite using low dissolved oxygen concentration as the main selection factor. Biodegradation 19 303-312.         [ Links ]

CAO Y, KWOK BH, YONG WH, CHUA SC, WAH YL and GHANI YA (2013) Mainstream partial nitritation -ANAMMOX nitrogen removal in the largest full-scale activated sludge process in Singapore: process analysis. Proc. WEF/IWA Nutrient Removal and Recovery Conference, 28-31 July 2013, Vancouver.         [ Links ]

DE KREUK MK (2006) Aerobic granular sludge - Scaling-up a new technology. PhD thesis, Technical University Delft, Netherlands. 199 pp.         [ Links ]

DE BRUIN LMM, DE KREUK MK, VAN DER ROEST HFR and VAN LOOSDRECHT MCM (2004) Aerobic granular sludge technology: Alternative for activated sludge technology? Water Sci. Technol. 49 (11/12) 1-9.         [ Links ]

DE KREUK MK and DE BRUIN LMM (2004) Aerobic Granule Reactor Technology. IWA Publishing, London.         [ Links ]

DE CLIPPELEIR H, VLAEMINCK SE, DE WILDE F, DAENINCK K, MOSQUERA M, BOECKX P, VERSTRAETE W and BOON N (2013) One-stage partial nitritation/anammox at 15°C on pretreated sewage: feasibility demonstration at lab-scale. Appl. Microbiol. Biotechnol. 97 10199-10210.         [ Links ]

DI TRAPANI D, CHRISTENSSON M and 0DEGAARD H (2011) Hybrid activated sludge/biofilm process for the treatment of municipal wastewater in a cold climate region: A case study. Water Sci. Technol. 63 (6) 1121-1129.         [ Links ]

EKAMA GA, WENTZEL MC, CASEY TG and MARAIS GvR (1996) Filamentous organism bulking in nutrient removal activated sludge systems. Paper 6: Review, evaluation and consolidation of results. Water SA 22 (2) 147-152.         [ Links ]

EKAMA GA and WENTZEL MC (2008) Nitrogen removal. Chapter 5. In: Henze M, Van Loosdrecht MCM, Ekama GA and Brdjanovic D (eds) Biological Wastewater Treatment: Principles, Modelling and Design. IWA Publishing, London. 528 pp. ISBN - 10:1843391880; 13:9781843391883.         [ Links ]

EKAMA GA (2010) The role and control of sludge age in biological nutrient removal activated sludge systems. Water Sci. Technol. 61 (7) 1645-1652.         [ Links ]

EKAMA GA and WENTZEL MC (2008) Organic material removal. Chapter 4. In: Henze M, Van Loosdrecht MCM, Ekama GA and Brdjanovic D (eds) Biological Wastewater Treatment: Principles, Modelling and Design. IWA Publishing, London. 528 pp. ISBN: 10:1843391880; 13:9781843391883.         [ Links ]

EKAMA GA, BARNARD JL, GUENTHERT FW, KREBS P, MCCORQUODALE JA, PARKER DS and WAHLBERG EJ (1997) Secondary settling tanks: Theory, design, modelling and operation. IAWQ STR No 6. IWA Publishing, London. ISBN: 900222 03 5. ISSN: 1025-0913. 216 pp.         [ Links ]

EKAMA GA, WILSENACH JA and CHEN G-H (2011) Saline sewage treatment and source separation of urine for more sustainable urban water management. Water Sci. Technol. 64 (6) 1307-1316.         [ Links ]

GADEMAN GE, GAYDON PN, DE BRUIN B and EKAMA GA (2013) Commissioning the first Nereda aerobic granular sludge technology wastewater treatment plant in South Africa at Gansbaai. Research Report W140, Dept of Civil Engineering, University of Cape Town, Rondebosch, 7700, Cape, South Africa (IP restrictions apply).         [ Links ]

GADEMAN GE, GAYDON PN and DE BRUIN B (2010) Granular aerobic activated sludge. Proc. 11th Biennial WISA Conference and Exhibition, 18-21 April 2010, Durban.         [ Links ]

HELLINGA C, SCHELLEN AAJC, MULDER JW, VAN LOOSDRECHT MCM and HEINEN JJ (1998) The SHARON process: an innovative method for nitrogen removal from ammonia rich wastewater. Water Sci. Technol. 37 (9) 135-142.         [ Links ]

HENZE M, GUJER W, MINO T, MATSUO T, WENTZEL M C and MARAIS GvR (1994) Activated sludge model No.2. IAWQ Scientific and Technical Report No 3. IAWQ, London. ISSN: 1025-0913, 32 pp.         [ Links ]

HENZE M, GRADY C P L, GUJER W, MARAIS GvR and MATSUO T (1987) Activated sludge model No.1. IAWPRC Scientific and Technical Reports No 1. International Association on Water Pollution Research and Control (IAWPRC), London. ISSN: 1010-707X. 33 pp.         [ Links ]

HENZE M, GUJER W, MINO T, MATSUO T, WENTZEL MC, MARAIS GvR and VAN LOOSDRECHT MCM (1999) Activated sludge model No.2d (ASM2d). Water Sci. Technol. 39 (1) 165-182.         [ Links ]

HU Z-R, WENTZEL MC and EKAMA GA (2000) External nitrification in biological nutrient removal activated sludge systems. Water SA 26 (2) 225-238.         [ Links ]

JENKINS D, RICHARD MG AND DAIGGER GT (1993) Manual for the Causes and Control of Activated Sludge Bulking, Foaming and other Solids Separation Problems (3rd edn). Lewis Publishers Inc., Boca Raton, USA. ISBN: 9781566706476.         [ Links ]

JIMENEZ J, WISE G, BURGER G, DU WW AND DOLD PL (2014) Mainstream nitrite-shunt with biological phosphorus removal at the City of St. Petersburg Southwest WRF. Proc. 87th Annual Water Environment Federation Technical Exhibition and Conference -WEFTEC, 27 Sept-1 Oct 2014, New Orleans.         [ Links ]

JUDD S, KIM B-G AND AMY G (2008) Membrane bioreactors. Chapter 13. In: Henze M, Van Loosdrecht MCM, Ekama GA and Brdjanovic D (eds) Biological Wastewater Treatment: Principles, Modelling and Design. IWA Publishing, London. 528 pp. ISBN: 10:1843391880; 13:9781843391883.         [ Links ]

KUENEN JG (2008) Anammox bacteria: From discovery to application. Nat. Rev. Microbiol. 6 320-326.         [ Links ]

LU H, EKAMA GA, WU D, VAN LOOSDRECHT MCM and CHEN GH (2012) SANI process realizes sustainable saline sewage treatment for saline water supply: Steady-state model-based evaluation of the pilot scale trial of the process. Water Res. 46 475-490.         [ Links ]

LUTZ MP, PRATT AM, PARKER DS and BRISCHKE KV (1990) Full-scale performance of nitrifying trickling filters. Proc. 63rdWPCF Conference, 7-11 Oct 1990, Washington DC.         [ Links ]

MULDER JW, VAN DE GRAAF AA, ROBERTSON LA and KUENEN JG (1995) Anaerobic ammonium oxidation discovered in a denitrifying fluidized bed reactor. FEMS Microb. Ecol. 16 177-184.         [ Links ]

MULLER AW, WENTZEL MC and EKAMA GA (2005) Reviving an old workhorse: Integration of rock media nitrifying filters for external nitrification in BNR activated sludge. Proc. 78th Water Environment Federation Conference and Exhibition, 29 Oct-2 Nov 2005, Washington DC. Vols. CD-ROM WEF, 3939-3948, Alexandria VA USA.         [ Links ]

ØDEGAARD H, CHRISTENSSON M and S0RENSEN K (2014) Chapter 15 - Hybrid systems. In: Jenkins D and Wanner J (eds) Activated Sludge 100 years - and Counting. IWA publishing, London.         [ Links ]

PARCO V, DU TOIT GJG and EKAMA GA (2010) The impact of high MLSS concentration on the kinetics of biological N and P removal in MBR BNR activated sludge systems. Proc. 82nd Annual Water Environment Federation Technical Exhibition and conference - WEFTEC, 10-14 October 2010, Orlando.         [ Links ]

PARKER DS, LUTZ M, DAHL R and BERNKOPF S (1989) Enhancing reaction rates in nitrification trickling filters through biofilm control. J. WPCF 61 618-631.         [ Links ]

PARKER DS, LUTZ M, ANDERSSON B and ASPEGEN H (1995) Effect of operating variables on nitrification rates in trickling filters. Water Environ. Res. 67 1111-1118.         [ Links ]

PARKER DS, JACOBS T, BOWER E, STOWE DW AND FARMER G (1996) Maximizing trickling filter nitrification rates through biofilm control: Research review and full scale application. Proc. 3rd IAWQ Specialized Conference on Biofilm Systems, 27-30 Aug 1996, Copenhagen.         [ Links ]

RAMPHAO MC, WENTZEL MC, MERRITT R, EKAMA GA, YOUNG T and BUCKLEY CA (2005) The impact of membrane solid-liquid separation on the design of biological nutrient removal activated sludge systems. Biotechnol. Bioeng. 89 630-646.         [ Links ]

REGMI P, MILLER MW, HOLGATE B, BUNCE R, PARK HK, CHANDRAN K, WETT B, MURTHY S and BOTT CB (2014) Control of aeration, aerobic SRT and COD input for mainstream nitritation/denitritation. Water Res. 57 (1) 162-171.         [ Links ]

RUSTEN B, EIKEBROKK B, ULGENES Y and LYGREN E (2006) Design and operation of the Kaldness moving bed biofilm reactor. J. Aquacult. Eng. 34 322-331.         [ Links ]

SEN D, MITTA P and RANDALL CW (1994) Performance of fixed film media integrated in activated sludge reactors to enhanced nitrogen removal. Water Sci. Technol. 30 (11) 13-24.         [ Links ]

SIEGRIST H, REITHAAR S, KOCH G and LAIS P (1998) Nitrogen loss in a nitrifying rotating contactor treating ammonia rich wastewater without organic carbon. Water Sci. Technol. 38 (8/9) 241-248.         [ Links ]

THOMAS WA, BOTT CB, REGMI P, SCHAFRAN G, MCQUARRIE J, RUTHERFORD B, BAUMBER R and WALTRIP D (2008) Evaluation of nitrification kinetics for a 2 MGD IFAS process demonstration. Proc.81" Annual Water Environment Federation Conference and Exhibition 4 97-116.         [ Links ]

VAN LOOSDRECHT MCM (2008) Innovative nitrogen removal. Chapter 6. In: Henze M, Van Loosdrecht MCM, Ekama GA and Brdjanovic D (eds) Biological Wastewater Treatment: Principles, Modelling and Design. IWA Publishing, London. 528 pp. ISBN: 10:1843391880; 13:9781843391883.         [ Links ]

VAN LOOSDRECHT MCM, MARTINS AM and EKAMA GA (2008) Bulking sludge. Chapter 11. In: Henze M, Van Loosdrecht MCM, Ekama GA and Brdjanovic D (eds) Biological Wastewater Treatment: Principles, Modelling and Design. IWA Publishing, London. 528 pp. ISBN: 10:1843391880; 13:9781843391883.         [ Links ]

VAN HUYSSTEEN JA, BARNARD JL and HENDRIKSZ J (1990) The Olifantsfontein nutrient removal plant. Water Sci. Technol. 22 (7/8) 1-8.         [ Links ]

WANNER J, KUCMAN K and GRAU P (1988) Activated sludge process with biofilm cultivation. Water Res. 22 207-215.         [ Links ]

WANNER J (2003) Control of filamentous bulking in activated sludge. John Wiley & Sons, Inc., New York. DOI: 10.1002/0471263397.env301.         [ Links ]

WENTZEL MC, COMEAU Y, EKAMA GA, VAN LOOSDRECHT MCM and BRDJANOVIC D (2008) Enhanced biological phosphorus removal. Chapter 7. In: Henze M, Van Loosdrecht MCM, Ekama GA and Brdjanovic D (eds) Biological Wastewater Treatment: Principles, Modelling and Design. IWA Publishing, London. 528 pp. ISBN: 10:1843391880; 13:9781843391883.         [ Links ]

WENTZEL MC AND EKAMA GA (1997) Principles in the design of single sludge activated sludge systems for biological removal of carbon, nitrogen and phosphorus. Water Environ. Res. 69 (7) 1222-1231.         [ Links ]

WENTZEL MC, DOLD PL, EKAMA GA and MARAIS GvR (1990) Biological excess phosphorus removal - steady state process design. Water SA 16 (1) 29-48.         [ Links ]

WETT B, OMARI A, PODMIRSEG SM, HAN M, AKINTAYO O, GÓMEZ BRANDÓN M, MURTHY S, BOTT C, HELL M, TAKÁCS I, NYHUIS G and O'SHAUGHNESSY M (2013) Going for mainstream deammonification from bench to full scale for maximized resource efficiency. Water Sci. Technol. 68 (2) 283-289.         [ Links ]

WILSENACH JA, BURKE L, RADEBE V, MASHEGO M, STONE W, MOUTON M and BOTHA A (2014) Anaerobic ammonium oxidation in the old trickling filters at Daspoort Wastewater Treatment Works. Water SA 40 (1) 81-88.         [ Links ]

WINKLER M-KH, KLEEREBEZEM R, KUENEN, JG, YANG J and VAN LOOSDRECHT MCM (2011) Segregation of biomass in cyclic anaerobic/aerobic granular sludge allows the enrichment of anaerobic ammonium oxidizing bacteria at low temperatures. Environ. Sci. Technol. 45 7330-7337.         [ Links ]

ZIMMERMANN RA, RICHARD D, BRADSHAW AT and CRADOCK PP (2003) Pilot scale evaluations of separate stage nitrification using attached growth moving bed media process. Water Environ. Res. 75 422-433.         [ Links ]



Received 5 January 2015
Accepted in revised form 9 June 2015



* This paper was presented at the 25th Anniversary International Conference of the Hong Kong Drainage Services Department, 11-14 November, 2014.
** To whom all correspondence should be addressed. 021 650 2585; e-mail:

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License