Available online at www.sciencedirect.

com

Journal of Hazardous Materials 154 (2008) 550–557

Wastewater treatment in a hybrid activated sludge baffled reactor

Mostafa Tizghadam, Christophe Dagot ∗ , Michel Baudu
Laboratoire des Sciences de l’Eau et de l’Environnement, Université de Limoges, ENSIL, Parc ESTER,
16 Rue Atlantis, F-87068 Limoges Cedex, France
Received 17 August 2006; received in revised form 17 October 2007; accepted 19 October 2007
Available online 30 October 2007

Abstract
A novel hybrid activated sludge baffled reactor (HASBR), which contained both suspended and attached-growth biomass perfect mixing cells in
series, was developed by installing standing and hanging baffles and introducing plastic brushes into a conventional activated sludge (CAS) reactor.
It was used for the treatment of domestic wastewater. The effects on the operational performance of developing the suspended and attached-growth
biomass and reactor configuration were investigated.
The change of the flow regime from complete-mix to plug-flow, and the addition of plastic brushes as a support for biofilm, resulted in considerable
improvements in the COD, nitrogen removal efficiency of domestic wastewater and sludge settling properties. In steady state, approximately
98 ± 2% of the total COD and 98 ± 2% of the ammonia of the influent were removed in the HASBR, when the influent wastewater concentration
was 593 ± 11 mg COD/L and 43 ± 5 mg N/L, respectively, at a HRT of 10 h. These results were 93 ± 3 and 6 ± 3% for the CAS reactor, respectively.
Approximately 90 ± 7% of the total COD was removed in the HASBR, when the influent wastewater concentration was 654 ± 16 mg COD/L at a
3 h HRT, and in the organic loading rate (OLR) of 5.36 kg COD m−3 day−1 . The result for the CAS reactor was 60 ± 3%.
Existing CAS plants can be upgraded by changing the reactor configuration and introducing biofilm support media into the aeration tank.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Wastewater treatment; Activated sludge; Hybrid activated sludge baffled reactor; Biofilm

1. Introduction Upgrading overloaded conventional activated sludge (CAS)

treatment plants is a promising solution, particularly when they
More than 80% of biological wastewater plants are based on have space limitations or need modifications that will require
the principle of activated sludge process, in which suspended large investment. Upgrading activated sludge treatment plants
bacteria oxidise the carbonaceous and nitrogen compounds to for the enhancement of the COD removal, nitrification, den-
produce an effluent that is in accordance with legal standards, itrification and sludge settling properties, can be achieved by
and that corresponds to a minimal environmental impact. The optimising the existing system by changing the bioreactor con-
increase of organic and hydraulic loads, related to the improve- figuration, or/and changing to a higher biological capacity
ment of wastewater collection, and the implementation of new process, with a combination of suspended and attached-growth
European directives and national regulations, often leads to dis- processes.
charges which do not comply with the standards. A new deal In the first case, it has been definitely shown that the hydro-
for wastewater engineers is now to stretch the performance of dynamics of the process have an effect on rates of pollution
existing infrastructure, which represents one of the most sig- removal, on the global kinetics of organic matter degradation
nificant challenges to the practice of wastewater engineering and on the sludge settling properties. The hydrodynamics of the
[1]. Currently, innovative processes are on the market, like the process are primarily dependent on the geometric characteris-
sequencing batch reactor (SBR) [2], membrane processes [3] tics of the reactor system, such as the configuration and the size
and attached or hybrid growth processes [4]. of the reactor, and on the characteristics of aeration dynamics
[5–7]. Filamentous bulking can be controlled by changing the
hydraulic regime of the reactor from complete-mix to plug-flow.
∗ Corresponding author. Tel.: +33 55 542 3670; fax: +33 55 542 3662. Pilot scale experiments indicate that a plug-flow configuration
E-mail address: dagot@ensil.unilim.fr (C. Dagot). produces sludge with significantly better settling characteris-

0304-3894/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jhazmat.2007.10.092
 M. Tizghadam et al. / Journal of Hazardous Materials 154 (2008) 550–557 551

tics than a conventional system [8]. Residence time distribution (HAS) reactor and to an activated sludge baffled reactor
(RTD) experiments make it possible to characterise the hydro- (ASBR).
dynamics for different flow rates and geometrical parameters. The following outcomes were expected:
Modelling the RTD may be carried out in a plug-flow reactor
with an axial distribution; or in perfect mixing cells in series, 1. Reduction in the aeration tank volume upon introduction of
with a back-mixing model [6]. biofilm support media and change in the flow regime in the
The second way also widely accepts that CAS process, com- aeration tank to meet a certain treatment objective.
bined with biofilm support media in the aerobic zone, has 2. An increase in the treatment system stability and performance
been demonstrated as an alternative design for biological nitro- in existing or newly developed CAS processes.
gen removal, as well as a cost-effective option for retrofitting
wastewater treatment plants [9–12]. The performance of an inte- 2. Materials and methods
grated fixed-film activated sludge (IFAS) system was compared
to a CAS system, in the field of biological nutrient removal 2.1. Experimental set-up
(BNR) [10].
A hybrid aerated submerged fixed film (HASFF) process sys- Two sets of activated sludge reactors were operated concur-
tem could be a viable option for upgrading activated sludge rently. The experimental design used in this study is shown in
treatment plants, because the HASFF process achieved over- Fig. 1. Table 1 presents the design details for two reactors (see
all BOD5 removal efficiencies of more than 94% at an HRT of
2 h with a four-fold increase of the organic loading rate, which
indicated the robustness of hybrid reactors [11].
An activated sludge biofilm wastewater treatment system
(ASBWTS), which is a CAS process combined with plastic
nets fitted inside an aeration tank, presents the positive effect
of the plastic nets on BOD5 removal efficiency and sludge set-
tling properties [13]. BOD5 removal increases from 17.5 to 25%,
according to the number of nets. Accordingly, wastewater treat-
ment with an activated sludge system contained both suspended
and attached growth and it can be called hybrid biological reac-
tor (HBR) [14]. This technology has many advantages, like
small area requirements, high biomass concentrations, high SRT
and low sludge production rates, etc. There are, however, some
disadvantages, such as biofilm control difficulties.
As far as the bibliography is concerned, the use of hybrid
reactor systems for treating organic matter and nutrients in
wastewater is an innovation that could be suitable for increasing
the capacity and the efficiency of wastewater treatment plants
(WWTP) [15].
To combine hydrodynamics effects and the adding of a carrier
in a CAS reactor, a novel hybrid system, which uses suspended
as well as attached biomass, has been developed and exam-
ined in this study. The hybrid activated sludge baffled reactor
(HASBR) was constructed by installing the standing and hang-
ing baffles and introducing a plastic support into a CAS reactor.
The HASBR, with a novel configuration and biofilm support in
the aerobic zone, may be an alternative design for upgrading
the CAS process without increasing its physical tankage and to
sustain nitrification, without significantly increasing the sludge
quantity to be treated.
The objective was to investigate the system’s feasibility as an
upgrading tool to any existing activated sludge facilities for the
anticipated increase in hydraulic and organic loading, without
additional tank construction. It also was to examine the perfor-
mance of the HASBR in terms of organic removal, under normal
operating conditions, as well as under excessive hydraulic load-
ing rates.
The performance of the HASBR was investigated by com- Fig. 1. Schematic diagram of experimental treatment of the suspended growth
paring this reactor to a CAS reactor, a hybrid activated sludge and the hybrid system.
552 M. Tizghadam et al. / Journal of Hazardous Materials 154 (2008) 550–557

Table 1 ing a total surface area of 0.40 m2 during the first run in the
Characteristics of laboratory biological reactors
HAS and HASBR reactors, and 0.80 m2 in the HASBR during
Biological reactor HASBR CAS the other runs. The brushes were completely immersed in the
External dimension (cm) biological reactor. Organisms grow in the form of biofilm on the
Length 58 58 surfaces of the brush structures and rod surfaces. The biofilm
Width 25 25 area, per unit wastewater volume, was 14.8 m2 m−3 during the
Height 28 28 first run in the HASBR and HAS reactors and 29.6 m2 m−3 dur-
Working volume (L) 27.5 27.5
No. of compartments 4 1
ing the other runs in the HASBR. The biomass concentration in
Compartment volume (L) 6.75 27.5 steady state on the rod and brush surface in form of biofilm was
Liquid depth (cm) 20 20 approximately 40 ± 2 g m−2 . The volume fraction occupied by
Settling tank volume (L) 8.3 8.3 the plastic brushes without biofilm, varied from 3 to 6% of the
effective volume of the tank.
The reactors were fed continuously with settled municipal
Fig. 1a). Two Plexiglas tanks with a 27.5 L effective volume were wastewater by a Masterflex peristaltic pump (model 7529-90,
used as the aeration tank. The basic design for the CAS reactor Cole Parmer, USA) to the wastewater holding (100 cylindrical
was a rectangular reactor. The ASBR was a rectangular reactor plastic tank) and then to the reactors, by two Masterflex peri-
with internal vertical baffles alternately hanging and standing. staltic pumps (model 7518-10, Cole Parmer, USA). The effluent
These baffles divided the ASBR into four compartments. Baffle from the reactors was withdrawn by gravity. Activated sludge
spacing was determined by making the compartments equal in was recycled by two Masterflex pumps (model 7518-10, Cole
size, and keeping the same width to upward and downward flow Parmer, USA).
in each compartment. In the hybrid case, plastic brushes were
fitted into the CAS and ASBR reactors. 2.2. Operation conditions
The aeration tank effluent was carried through an outlet con-
nected by a 1.9 cm diameter pipe to the clarifier. Air was supplied The experimental study was carried out at the Limoges
through porous ceramic diffusers, located at the base of the tank. municipal WWTP (Limoges, France). The characteristics of the
Adequate mixing was provided by the effect of the rising air bub- settled municipal wastewater during the different experimental
bles. The secondary settling tank had an overall liquid depth of runs are shown in Table 2.
50 cm and consisted of a 10 cm internal diameter cylinder joined Activated sludge taken from Limoges (France) municipal
to a conical bottom (10 cm in height). A wire, bent to conform to WWTP was inoculated into the reactors. For each start-up,
the inside of the cone, was rotated by 5 rpm. The clarifier volume approximately 10 L of recycled activated sludge was collected
was maintained constant at 8.3 L. from the secondary clarifier recycling line and was inoculated
The experimental program consisted of two phases. In phase on the same day to every reactor. A solid test was performed on
1, an ASBR was compared to a CAS reactor (Fig. 1a). In phase the sludge to determine its concentration (7–9 g L−1 ), and the
2, a HASBR was compared to HAS and CAS reactors (Fig. 1b). sludge was diluted with water to a desired concentration about
The attached biomass was developed on plastic brushes, 12 cm 3 g L−1 . The system was operated under batch mode for nearly
in length and 4 cm in diameter (Bioblock-France). Each brush 12 h without influent to aeration of the sludge. In phase 2, the
had 5250 bristles 2.0 cm long, with a diameter of 0.3 mm, yield- reactors were operated under continuous aerobic conditions for

Table 2
Operating parameters of the reactors
Parameters Phase 1 Phase 2 Phase 2 Phase 2 Phase 2 Phase 2
Run 1 Run 2 Run 3 Run 4 Run 5

Time (day) 50 1–64 65–90 90–115 116–142 143–174

Influent Organic loading (mg COD/L) 545 ± 73 443 ± 134 593 ± 11 590 ± 106 674.5 ± 56 654 ± 16
NH4 + -N influent loading (mg L−1 ) – 37 ± 9 43 ± 5 46 ± 9 48 ± 5 42 ± 7
HRT (h) 10 10 10 7 5 3
SRT (day) 13 13 13 13 13 13
TSS (g L−1 )
HASBR 2.4–3.5 2.2–3.3 2.2–3.3 2.2–3.3 2.2–3.3 2.2–3.3
CAS 2.2–3.4 2.2–3.3 2.2–3.3 2.2–3.3 2.2–3.3 2.2–3.3
Total surface area of plastic brushes (m2 )
HASBR 0 0.4 0.8 0.8 0.8 0.8
CAS 0 0.4 0 0 0 0
Biofilm area per unit wastewater volume (m2 m−3 )
HASBR 0 14.8 29.6 29.6 29.6 29.6
CAS 0 14.8 0 0 0 0
Average OLR (kg COD m−3 day−1 ) 1.6 0.98 1.42 2.04 3.24 5.36
 M. Tizghadam et al. / Journal of Hazardous Materials 154 (2008) 550–557 553

1 month, in order to grow biofilm on continuously fed wastew- 3. Results and discussion
ater. Sludge wasting was manually performed and the wasted
volume was recorded on a daily basis. The experiments were Under the same influent wastewater concentrations, the effect
conducted indoors, at an operation temperature of 18–22 ◦ C. The of the reactor configuration and fixed-film biomass on COD
steady-state condition was considered reached when repetitive removal efficiency, nitrification performance and sludge set-
analytical results were obtained. tling properties were studied by comparing the performance the
ASBR to the CAS reactor, the performance of the HASBR to
2.3. Operation procedures the CAS and to the HAS reactors. These treatment regimes were
tested in parallel in the reactors.
Detailed operation parameters are shown in Table 2. Each
experimental run lasted in 3–4 weeks. We aimed to take the 3.1. Reactor hydrodynamics studies
grab samples at the same time of each experimental day. Hydro-
dynamics model of every reactor was determined by RTD First of all, the hydrodynamics behaviour of the reactors treat-
techniques. Tracer tests with NaCl pulse addition were used to ing municipal wastewater was characterized by means of a RTD
investigate the hydraulics of the tank. The conductivity in the technique, using NaCl as a tracer, to monitor hydraulic flow
effluent was detected and RTD curves were generated. For the regime in the ASBR and the CAS reactors and the presence
RTD analysis, the following procedure was used: after calculat- of dead spaces, or short cuts, in the reactors. During the RTD
ing RTD curves or E(t) values, the RTD curves were standardised experiments, the working flow was 2.75 L h−1 , which implied
for the experimental RTD. To simulate the reactors, the DTS, a mean residence time of 10 h for each reactor, without water
PRO v.4.2 was used [16]. It made possible the construction of recirculation. To ensure proper mixing, air is dispatched into the
properly interconnected complex network of elementary reac- biological reactors through many diffusers.
tors (such as plug-flow reactors, perfect mixing cells in series As shown Fig. 2, little difference in mixing or dead space
and perfect mixing cells in series exchanging with a dead zone) was observed in this test, because the obtained curves look like
and the optimisation of the parameters of the experimental theoretical RTD, without significant loss of shape. A little short
curves. cut noticed in the first compartment of the ASBR, and the aver-
age dead zones measured within the reactors, were between 1.5
2.4. Analytical procedures and 10.1%. Flow patterns within the compartments of the ASBR
showed an intermediate behaviour between plug-flow and ide-
The TSS, VSS and SVI were analysed in accordance with ally mixed. The RTD experimental results represent the main
Standard Methods for the Examination of Water and Wastewa- flow in the compartments and the obtained theoretical volume
ter [17] (after filtration with a 1.2 ␮m membrane). The COD in each compartment were nearly equal (≤±10.1).
and, NH4 + -N, were analysed according to Dr. Lange’s test, A model, including ideal perfect mixing cells in series, has
(LCK 414, LCK 114), (LCK303), respectively. Temperature, two parameters: the mean residence time (τ) and the number of
pH, dissolved oxygen (DO) were also determined using a WTW mixing cells (J) were used. Fig. 2 shows the comparison between
Multiline P4. experimental (Fig. 2a and c) and theoretical RTD results (Fig. 2b

Fig. 2. Comparison between experimental and theoretical RTD in each compartment of the HASBR for J = 4 (a and b) and in the CAS reactor for J = 1 (c and d).
554 M. Tizghadam et al. / Journal of Hazardous Materials 154 (2008) 550–557

and d), using J = 1 for the CAS reactor (Fig. 2d) and J = 4 for Fig. 3 also shows the character of NH4 + -N removal in
the ASBR (Fig. 2b). The agreement between the experimental the reactors. The positive effect of fixed biomass on NH4 + -N
and theoretical results, and the consistency of the parameters, removal efficiency can be clearly demonstrated in the reactors.
demonstrate the validity of the proposed model. Fig. 2b shows The removal of NH4 + -N in steady state, was about 97 and 98%
the result of simulation in each compartment carried out with on average, for the HAS reactor and the HASBR, respectively.
J = 4, that is to say four reactors in series. The value of J = 4 is The removal of NH4 + -N in steady state in the CAS reactor and
sufficient for global simulation of flow regime into the ASBR the ASBR are less than 10% in average.
(curve E(t) 4 in Fig. 2a and b). It is possible to consider, in a The important factor that may have contributed to these
simple approach, that the CAS reactor is really a completely results is the presence of biofilm support media in which a nitri-
mixed reactor and that the hydrodynamics in the ASBR tend fying population was established. A previous study showed that
towards a cascade reactor with an axial dispersion, modelled all the consumption of NH4 + -N was due to the biofilm synthesis
with four perfect mixing cells in series. The same experiments in a hybrid system [21].
were done including plastic brushes (but without biomass) and The average SVI-values in steady state are shown in Fig. 3.
the difference between experimental results, with and without As the SVI decreases, the settling efficiency increases [8]. SVI-
plastic brushes, was negligible for two reactors, because the vol- values in steady state were about 277 and 126 mL g−1 on average
ume of the plastic brushes was negligible in the reactors (3–6% in the CAS and ASBR reactors, respectively. The sludge had
of total volume). This result could be used in further studies in very poor settling properties in the CAS reactor, but they were
order to combine biological reactions and the hydrodynamics much better in the ASBR. In the previous study, it was showed
model [6]. that SVI was lower than 250 mL g−1 in plug-flow reactor but it
did exceed to 800 mL g−1 in the CAS reactor, under the same
3.2. Reactor performance operation conditions [8]. The other study carried out by Roche
[22] indicated that, changing the hydraulic regime of a reactor
Fig. 3 shows the total COD and NH4 + -N removal efficiency from complete-mix to plug-flow regime has a significant effect
and sludge settling properties (in term of SVI-values). The oper- on the sludge settling properties.
ation conditions were similar in respect to organic loading rate No problems have been noticed with the excessive growth of
(OLR) of 1.4 kg COD m−3 day−1 , SRT of 13 days and hydraulic filamentous microorganisms in the HAS and HASBR. The SVI-
retention time (HRT) of 10 h. The organic (volumetric) load- values in the hybrid systems (HAS, HASBR) did not exceed
ing rate defined as the amount of COD applied to the aeration 100 mL g−1 . Moreover, the sludge settling properties are main-
tank volume per day. The average dissolved oxygen (DO) was tained in the processes (e.g. SVI, settling rate, . . .) [23]. The
maintained above 2 mg L−1 in the reactors, allowing oxygen to results presented in Fig. 3 agree well with the data from lit-
penetrate into the biofilm. The average of 87% and 97% total erature [21]. In the other study, the SVI decreased from 350 to
COD removal efficiencies were obtained in steady state, for the 38 mL g−1 , after the addition of the biofilm into the CAS system
CAS reactor and the HASBR, respectively, as the minimum and [13].
maximum of the total COD removal efficiency among four reac-
tors. The reduction of the total COD in the hybrid systems (HAS 3.3. Effects of increasing the organic loading rate on COD
and HASBR) was better than the suspended growth reactors removal efficiency and sludge settling in the CAS and
(CAS and ASBR); and in the ASBR, it was better than the CAS HASBR
reactor in the same operation mode (Fig. 3). The difference is
due to the existence of the biofilm in the hybrid reactors and the The effect of the organic loading rate on the COD removal
plug-flow hydraulic regime in the ASBR. These results confirm efficiency of CAS and HASBR reactors was evaluated (Fig. 4a).
the competence of our hypotheses and agree well with data from The OLR was increased from 1.4 to 5.6 kg COD m−3 day−1 , at
the literature [8,18–20]. four different HRT of 10, 7, 5 and 3 h.
The HASBR clearly demonstrates greater total COD removal
efficiency than the CAS reactor.
The average total COD removal efficiency at 1.4, 2.1, 3.24
and 5.6 kg COD m−3 day−1 corresponds to HRT of 10, 7, 5 and
3 h were 93, 91, 86 and 60% in CAS reactor and 98, 97, 96
and 90% in HASBR. The total COD removal efficiency of the
CAS reactor decreased by 33%, but it decreased by 8% in the
HASBR reactor, while increasing the organic loading rate to
four-fold. Previous studies showed that the reduction of COD
increased from 51 to 90% in an upgraded activated sludge system
with floating carriers [20]. These results indicate that the hybrid
system is resilient to organic shock loads, thereby offering a
viable upgrading option for CAS treatment plants [11].
Fig. 3. The average total COD and NH4 + -N removal efficiencies and the average The SVI-values of the CAS and HASBR reactors is shown in
SVI values in four reactors. Fig. 4b. When the HRT was decreased, filamentous bulking took
 M. Tizghadam et al. / Journal of Hazardous Materials 154 (2008) 550–557 555

Fig. 6. Course of total COD (a) and NH4 + -N (b) concentrations in influent and
effluent of HASBR reactor.
Fig. 4. The effect of organic loading rate (OLR) on average COD removal
efficiency (a) and sludge settling (b) in the CAS and HASBR.
3 h were 8, 9, 10 and 9% in average corresponding to 98, 55,
place in the CAS reactor. On OLR of 5.6 kg COD m−3 day−1 37 and 27% in an average of the HASBR (Fig. 5). During the
(HRT of 3 h), the SVI value in average increased rapidly and operation of the CAS, no significant consumption of NH4 + -N
exceeded 400 mL g−1 in the CAS reactor, but in the HASBR, the was observed in this system. All NH4 + -N consumption in the
SVI value in average was about 110 mL g−1 in the same oper- HASBR was due to the attached biomass synthesis, because the
ation mode. In a previous study, an improvement of the sludge HRT value was small to admit the growth of the nitrifying bacte-
settling rate after adding fixed-film support media was reported ria. The NH4 + -N removal efficiency in the HASBR significantly
[21]. No problems associated with excessive growth of filamen- decreased, with decreasing the HRT, but it was about 27% at a
tous microorganisms were reported for hybrid reactors [9]. low HRT (Fig. 5). These results agree well with the data from
literature [14,24].
3.4. Effects of decreasing the HRT on the nitrification
performance of the CAS and HASBR reactor 3.5. Long period experiments of the CAS and HASBR
reactors
Fig. 5 shows the NH4 + -N removal efficiency values in aver-
age in the CAS and HASBR reactors. The positive effect of fixed Fig. 6a shows the characters of COD removal under five
biomass on NH4 + -N removal efficiency can demonstrate in the different conditions, which included influent and effluent total
HASBR reactor. The HASBR and the CAS reactors were oper- COD. In the first run, the HASBR was compared to the HAS
ated at a SRT of 13 days in different HRT of 10, 7, 5 and 3 h. In reactor and in runs 2–5, the HASBR, was compared to the CAS
the CAS, average NH4 + -N removal efficiencies at 10, 7, 5 and reactor, under the same operation conditions (Table 2). The total
HRT was 10, 7, 5 and 3 h, respectively, excluding the time in the
clarifier. The results of the HASBR for the influent and effluent
total COD concentrations are shown in Fig. 6a.
It was observed that average total COD in the effluent was
8 mg L−1 at an HRT of 10 h, but it was 48 mg L−1 at an HRT
of 3 h, in the reactor in steady state. As seen in Fig. 6a, the
effluent total COD concentration decreased slightly with a four-
fold increase of OLR.
The average effluent concentration of NH4 + -N in the HASBR
is presented in Fig. 6b. The effluent concentrations of NH4 + -
N in steady state were about 0.8, 0.6, 18, 32 and 32 mg L−1
on average, corresponding to 36, 46, 43, 49 and 42 mg L−1 of
Fig. 5. Effect of HRT on NH4 + -N removal efficiency of the CAS and HASBR influent concentration of NH4 + -N. In the second run, as the
reactors. biofilm area per unit wastewater volume was increased from
556 M. Tizghadam et al. / Journal of Hazardous Materials 154 (2008) 550–557

Acknowledgments

The authors wish to thank the Limoges wastewater treatment

plant staff for its help and the reviewers for their suggestions.

References

[1] Metcalf and Eddy Inc, Wastewater Engineering: Treatment, Disposal,

Reuse, McGraw-Hill, New York, 2003.
[2] M. Casellas, C. Dagot, M. Baudu, Set up and assessment of a control
strategy in a SBR to enhance nitrogen and phosphorus removal, Process
Fig. 7. Course of NH4 + -N concentrations in influent and effluent in the CAS Biochem. 41 (9) (2006) 1994–2001.
reactor at HRT = 10 h. [3] S. Gonzalez, M. Petrovic, D. Barcelo, Removal of a broad range of
surfactants from municipal wastewater-comparison between membrane
14.6 to 29.8 m2 m−3 , the effluent concentrations of NH4 + -N bioreactor and conventional activated sludge treatment, Chemosphere 67
then decreased very slightly; but it increased significantly when (2007) 335–343.
[4] C. Nicolella, M.C.M. van Loosdrecht, J.J. Heijnen, Wastewater treatment
the HRT decreased. The best NH4 + -N removal efficiency was with particulate biofilm reactors, J. Biotechnol. 80 (2000) 1–33.
occurred at HRT of 10 h. Yang and co-workers [25] reported that, [5] V. Chaignon, B. Lartiges, A. Samrani, C. Mustin, Evolution of size distri-
the removal of ammonia in a submerged membrane bioreactor at bution and transfer of mineral particles between flocs in activated sludge:
a HRT ≥10 h was over 98%, whereas this rate was significantly an insight into floc exchange dynamics, Water Res. 36 (2002) 676–684.
lower in CAS reactor in the same operation conditions. [6] O. Potier, J.P. Leclerc, M.N. Pons, Influence of geometrical and operational
parameters on the axial dispersion in an aerated channel reactor, Water Res.
Fig. 7 shows the NH4 + -N concentration values in the influent 39 (2005) 4454–4462.
and effluent of the CAS reactor in the first and second runs. About [7] J.I. Borroto, J. Dominguez, J. Griffith, M. Fick, J.P. Leclerc, Technetium-
3 weeks after adding biofilm support media, the effluent NH4 + - 99m as a tracer for the liquid RTD measurement in opaque anaerobic
N started to decrease significantly and any days after removing digester: application in a sugar wastewater plant, Chem. Eng. Process. 42
the biofilm support media, the effluent NH4 + -N concentration (2003) 857–865.
[8] A.A. Azimi, M. Zamanzadeh, The effect of selectors and reactor configu-
started to increase significantly. ration on filamentous sludge bulking control in activated sludge, Pakistan
J. Biol. Sci. 9 (3) (2006) 345–349.
4. Conclusion [9] T. Sriwiriyarat, C.W. Randall, Evaluation of integrated fixed film activated
sludge wastewater treatment processes at high mean cell residence time
and low temperatures, J. Environ. Eng. 131 (11) (2005) 1550–1556.
The following conclusions were drawn from the results
[10] T. Sriwiriyarat, C.W. Randall, Performance of IFAS wastewater treatment
obtained during this investigation. processes for biological phosphorus removal, Water Res. 39 (16) (2005)
3873–3884.
(1) Approximately 98 ± 2% of total COD and 98 ± 2% of [11] H.A. Al-Sharekh, M.F. Hamoda, removal of organics from wastewater
ammonia of the influent was removed in the HASBR, when using a novel biological hybrid system, Water Sci. Technol. 43 (1) (2001)
321–326.
the influent wastewater concentration was 593 ± 11 and
[12] J.L. Su, C.F. Ouyang, Nutrients removal using a combined process with
43 ± 5 mg L−1 , respectively, at HRT of 10 h. These results activated sludge and fixed biofilm, Water Sci. Technol. 34 (1) (1996)
were 93 ± 3 and 6 ± 3%, respectively, for the CAS reactor. 477–486.
(2) Approximately 90 ± 7% of total COD was removed in [13] 9F. Gebara, Activated sludge biofilm wastewater treatment system, Water
the HASBR, when the influent wastewater concentration Res. 33 (1) (1999) 230–238.
was 654 ± 17 mg L−1 at HRT of 3 h and an OLR of [14] W. Jianlong, S. Hanchang, Q. Yi, Wastewater treatment in hybrid biological
reactor (HBR): effect of organic loading rates, Process Biochem. 36 (2000)
5.36 kg COD m−3 day−1 . The result for the CAS reactor was 297–303.
60 ± 3%. [15] E.V. Münch, K. Barr, S. Watts, J. Keller, Suspended carrier technology
(3) The SVI-values were 400 and 112 mL g−1 at HRT of 3 h in allows upgrading high-rate activated sludge plants for nitrogen removal
the CAS and the HASBR reactors, respectively. via process intensification, Water Sci. Technol. 41 (2000) 5–12.
[16] J.P. Leclerc, C. Detrez, A. Bernaed, D. Schweich, DTS: Un logiciel d’aide
à l’élaboration de modèles d’écoulement dans les réacteurs, Revue de
Some of the implications derived from this study are: l’Institut Français du Pétrole 50 (5) (1995) 641–656.
[17] APHA., Standard Methods for the Examination of Water and Wastew-
- Existing activated sludge systems can be upgraded easily, by ater, 19th ed., American Public Health Association, Baltimore, MD,
1995.
simply fitting baffles and biofilm support media inside the
[18] J.X. Liu, J.W. van Groenestijn, H.J. Doddema, B.Z. Wang, Removal of
aeration tank, thus eliminating the need for constructing new nitrogen and phosphorus using a new biofilm-activated sludge system,
extensions. Water Sci. Technol. 34 (1–2) (1996) 315–322.
- COD and nitrogen removal efficiency and system stability [19] V. Aravinthan, S. Takizawa, K. Fujita, K. Komatsu, Factors affecting nitro-
can be increased in existing activated sludge systems. The gen removal from domestic wastewater using immobilized bacteria, Water
Sci. Technol. 38 (1998) 193–202.
HASBR gives more flexibility in the operation of an over-
[20] E. Hansen, L. Zadura, S. Frankowski, M. Wachowicz, Upgrading of an
loaded wastewater treatment plant. Increasing the organic activated sludge plant with floating biofilm carriers at Frantschach Swiecie
loading rate by four-fold had a minor effect on the total COD S.A. to meet the new demands of year 2000, Water Sci. Technol. 40 (11–12)
removal efficiency in the HASBR. (1999) 207–214.
 M. Tizghadam et al. / Journal of Hazardous Materials 154 (2008) 550–557 557

[21] J. Wanner, K. Kucman, P. Grau, Activated sludge process combined with [24] G. Andreottola, P. Foladori, G. Gatti, P. Nardelli, M. Pettena, M. Ragazzi,
biofilm cultivation, Water Res. 22 (2) (1988) 207–215. Upgrading of a small overloaded activated sludge plant using a MBBR
[22] N. Roche, Influence de l’hydrodynamique du bassin d’aération sur la system, J. Environ. Sci. Health 10 (2003) 2317–2328.
décantabilité des bous activées, Ph.D. Thesis, INPL, 1989. [25] H. Li, M. Yang, Y. Zhang, M. Gao, Y. Kamagata, Comparison of nitrifica-
[23] A.A.M. Langenhoff, N. Intrachandra, D.C. Stucky, Treatment of dilute tion performance and microbial community between submerged membrane
soluble and colloidal wastewater using anaerobic baffled reactor: influence bioreactor and conventional activated sludge system, Water Sci. Technol.
of hydraulic retention time, Water Res. 34 (4) (2000) 1307–1317. 5 (6–7) (2005) 193–200.