Bioprocess Biosyst Eng (2010) 33:267–276

DOI 10.1007/s00449-009-0321-1

ORIGINAL PAPER

Treatment of domestic wastewater in an up-flow anaerobic sludge

blanket reactor followed by moving bed biofilm reactor
A. Tawfik Æ F. El-Gohary Æ H. Temmink

Received: 30 December 2008 / Accepted: 14 April 2009 / Published online: 30 April 2009
Ó Springer-Verlag 2009

Abstract The performance of a laboratory-scale sewage 5.0 h, corresponding to overall log10 reduction of 2.3, 1.4
treatment system composed of an up-flow anaerobic sludge and 0.7, respectively. The discharged sludge from UASB–
blanket (UASB) reactor and a moving bed biofilm reactor MBBR exerts an excellent settling property. Moreover, the
(MBBR) at a temperature of (22–35 °C) was evaluated. mean value of the net sludge yield was only 6% in UASB
The entire treatment system was operated at different reactor and 7% in the MBBR of the total influent COD at a
hydraulic retention times (HRT’s) of 13.3, 10 and 5.0 h. An total HRT of 13.3 h. Accordingly, the use of the combined
overall reduction of 80–86% for CODtotal; 51–73% for UASB–MBBR system for sewage treatment is recom-
CODcolloidal and 20–55% for CODsoluble was found at a mended at a total HRT of 13.3 h.
total HRT of 5–10 h, respectively. By prolonging the HRT
to 13.3 h, the removal efficiencies of CODtotal, CODcolloidal Keywords Sewage
UASB
MBBR
COD

and CODsoluble increased up to 92, 89 and 80%, respec- Nitrification
Faecal coliform
Sludge
tively. However, the removal efficiency of CODsuspended in
the combined system remained unaffected when increasing
the total HRT from 5 to 10 h and from 10 to 13.3 h. This Introduction
indicates that, the removal of CODsuspended was indepen-
dent on the imposed HRT. Ammonia-nitrogen removal in Within the spectrum of anaerobic sewage treatment
MBBR treating UASB reactor effluent was significantly technologies, the up-flow anaerobic sludge blanket
influenced by organic loading rate (OLR). 62% of ammo- (UASB) reactor offers great promise, especially in devel-
nia was eliminated at OLR of 4.6 g COD m-2 day-1. The oping countries that are usually located in hot and moderate
removal efficiency was decreased by a value of 34 and 43% climatic zones [1, 2]. These reactors remain robust high-
at a higher OLR’s of 7.4 and 17.8 g COD m-2 day-1, rate treatment systems, generally without moving
respectively. The mean overall residual counts of faecal mechanical parts, limiting both capital and operating costs
coliform in the final effluent were 8.9 9 104 MPN per [3]. Like many high-rate systems, the UASB retains a high
100 ml at a HRT of 13.3 h, 4.9 9 105 MPN per 100 ml at amount of biomass in the form of flocculant sludge, gran-
a HRT of 10 h and 9.4 9 105 MPN per 100 ml at a HRT of ules or aggregates of microorganisms. Furthermore, good
contact between biomass and wastewater is ensured due to
mixing as a result of biogas production. The configuration
A. Tawfik (&)
F. El-Gohary of these reactors has proven to be efficient in removing
Water Pollution Research Department, National Research
organic matter and total suspended solids (TSS), as well as
Center, P. Box 12622, El-Behouth St., Dokki, Cairo, Egypt
e-mail: [email protected] in producing smaller amounts of excess sludge compared
to aerobic reactors [4, 5]. However, the performance of the
H. Temmink UASB reactors is affected by operational conditions, one of
Sub-department of Environmental Technology,
which is the hydraulic retention time (HRT). Castillo et al.
Agrotechnology and Food Sciences Department,
Wageningen University and Research Centre, [6] investigated the effect of different HRT’s on a pilot-
P.O. Box 8129, 6700EV Wageningen, The Netherlands scale UASB reactor (750 l), fed with domestic wastewater

123
268 Bioprocess Biosyst Eng (2010) 33:267–276

Fig. 1 Integrated up-flow

anaerobic Sludge blanket
(UASB)–moving bed biofilm
reactor (MBBR) treating
domestic wastewater

(COD inf. = 600 mg l-1), at temperatures ranging from 13 the nitrification rate. Also the characteristics of the sludge
to 20 °C. Their results showed that the removal values of produced in the combined system will be considered.
the different COD fractions increased by the increase of the
HRT. However, there has been a tendency for this to
become constant at a HRT more than 6 h. The reactor Materials and methods
achieved 66% for COD removal at an HRT of 8 h. In
another study, A’lvarez et al. [7] investigated the perfor- Up-flow anaerobic sludge blanket (UASB) reactor
mance of an UASB reactor treating domestic wastewater at
an HRT of 11 h and a temperature of 14 °C. After a start- A schematic diagram of the UASB reactor is shown in
up period of 75 days, the UASB removal values were 58% Fig. 1. The reactor had a working volume of 10 l and a
for TSS, 41% for CODtotal and 54% for total biochemical height of 1.35 m. Eight ports for obtaining sludge samples
oxygen demand (BOD5 total). It is worth-mentioning how- are arranged along the reactor height, the first one at 0.1 m
ever, that in spite of the advantages of the UASB reactors above the base of the column. The reactor is provided by a
as an advanced primary treatment, post-treatment step is conical gas solids separator (GSS) at the top of the tank
required to achieve the emission standards set by regula- with a height of 0.2 m. The gas production was measured
tory authorities. Yet it is not a priori clear which of the by a wet gas meter (Schlumberger P. Max: 100 m bar).
different post-treatment units can be the best alternative. Initially, the UASB reactor was inoculated with 6 l
The choice depends on: the required effluent quality; the digested sludge. The initial concentration of the sludge in
available land area, the treatment cost, the simplicity and the reactor was 18 g VSSl-1. The system was fed with raw
operational stability of the treatment system, the indepen- sewage from a nearby sewer network (Dokki, Cairo) using
dence on imported equipment and material and operational a peristaltic pump. During the study period, temperature
flexibility. So far the results obtained from lab sale and full varied from 22 to 35 °C. The main characteristics of the
sale units suggest useful application of moving bed biofilm domestic wastewater are given in Table 1.
reactors (MBBR) systems for aerobic post-treatment [8, 9].
MBBR provides a long biomass retention time and Moving bed biofilm reactor (MBBR)
accommodate high loading rates without any problems of
clogging [10]. In a MBBR, the bacteria are fixed in a The MBBR consists of a reactor vessel with a volume of
biofilm on a carrier. The carrier is suspended and moves 8.0 l and a depth of 0.5 m (Fig. 1). The reactor was filled
freely in the reactor. The MBBR has been applied for with 1158 carrier media. The carrier elements represent
organic matter removal [11], for nitrification [12], and for 70% of the total reactor volume [14]. The carriers are made
nutrient (N and P) removal [13]. of polyethylene, with a specific gravity of 0.95 and an
The objective of this study is to assess the performance effective specific surface area of 363 m-2 m-3. The media
of the combined UASB–MBBR system for domestic is shaped in a cylindrical form and has a length of 1.8 cm
wastewater treatment at different HRT’s, consequently and a diameter of 1.85 cm. Complete mixing of the media
different OLR’s. This will be carried out by monitoring the is ensured by means of a central stirrer with blades placed
removal of the COD fractions (CODsuspended, CODcolloidal, at 10 and 40 cm below top-water level; the stirrer is driven
CODsoluble), and faecal coliform (FC) removal as well as by a 0.37-kW geared electric motor at a rotational speed of

123
Bioprocess Biosyst Eng (2010) 33:267–276 269

Table 1 Mean characteristics of domestic wastewater

Parameters COD fractions (mg O2 l-1) Nitrogen (mg N l-1) Faecal coliform (MPN per 100 ml)
pH
Total Suspended Colloidal Soluble NH4-N TKN

6.9 (0.3) 740 (238) 488 (268) 63 (33) 189 (87) 27 (5) 40 (4) 1.1 9 107 (7.9 9 106)
Standard deviations are presented between brackets

Table 2 Operational conditions of the combined system (UASB–MBBR)

Operational conditions HRT (h) OLR Flow rate (m3 day-1)
UASB MBBR UASB (kg COD m-3 day-1) MBBR (g COD m-2 day-1) UASB

Run 1 8 5.3 1.5 4.6 0.036

Run 2 6 4 2.4 7.4 0.048
Run 3 3 2 5.8 17.8 0.096

about 90 rpm. A screen is provided at the outfall end of the residence time (SRT). Volatile suspended solids (VSS) of
reactor to keep the media from clogging the effluent spout the attached biofilm on the carrier amounted to 7.6
or passing out of the reactor. The MBBR was continuously (HRT = 5.3 h), 9.0 (HRT = 4 h), and 11 gVSS/l media
operated and fed with UASB reactor effluent (Fig. 1). Pure (HRT = 2 h). The calculated biofilm thickness was ranged
oxygen is supplied from the bottom of the reactor through a from 420 to 750 lm and from 720 to 934 lm depending on
diffuser. Dissolved oxygen (DO) is measured in the reactor the applied loading rate.
by a portable DO-meter and the flow rate of oxygen was Microscopic examination test show that there were a
controlled by visual inspection of a flowmeter. In this way, large numbers of ciliates and rotifers in the biofilm adhered
the oxygen supply rate was adjusted in order to keep the to the carrier elements.
concentration of DO fairly constant at a level of not less
than 2.0 mg O2 l-1 [14] during the whole experimental Excess sludge in the combined UASB–MBBR
period.
The sludge bed of the UASB reactor was kept below tap 5,
Operational conditions ca. 80 cm from the UASB bottom, by opening this tap once
a week for discharging the sludge accumulated above.
The operational conditions of the combined UASB–MBBR Additionally, the sludge from the MBBR was daily dis-
are shown in Table 2. The UASB–MBBR was operated for charged and collected in a storage tank of 10 l for mea-
290 days, 39–98; 130–183; and 210–290 days at HRT’s of, surement of total and volatile solids. The sludge residence
respectively 8 ? 5.3; 6 ? 4 and 3 ? 2 h. The first 38 days time (SRT) of the UASB and MBBR was calculated
of operation were considered as a start-up period, while the according to the following equation;
periods from day 99 to 129 and from 184 to 209 were  
VX
considered as acclimatization periods to the new HRT. SRT ¼
Qw Xw þ QXe
Statistical analysis at different HRT’s has been done
according to Snedecor and Cochran [15]. where: V, reactor volume; X, average biomass concentra-
tion of the reactor (mg VSSl-1); Qw, excess sludge
Characteristics of biofilm carriers (l day-1); Xw, concentration of the excess sludge
(mg VSSl-1); Q wastewater flow rate (l day-1); Xe effluent
Representative samples of colonized carrier media were concentration (mg VSSl-1) and according to Zeeuw [17]
taken from the reactor three times in each run. The har- Xe = CODsuspended/1.4.
vested carriers with biomass was washed in a sodium
hypochlorite solution (6% active chlorine) and then Sampling and analytical methods
exposed to ultrasound for 3.0 h with a rinse step every hour
with the chlorinated solution and a final rinse with deion- Grab samples of the influent and the effluents of the UASB
ized water [16]. The concentration of biomass is expressed and MBBR were collected and immediately analysed for
as g VSS l-1 media to be able to calculate the sludge pH, temperature and dissolved oxygen (DO). The COD

123
270 Bioprocess Biosyst Eng (2010) 33:267–276

was analysed using the micro-method as described by Table 3 COD fractions (CODsuspended CODcolloidal and CODsoluble) in
APHA [18]. Raw samples were used for CODtotal, 4.4-lm an UASB–MBBR treating domestic wastewater at different HRT’s
folded paper filtered (Schleicher and Schuell 595 1/2) Parameters COD fractions (mg l-1)
samples for CODfiltrate and 0.45-lm membrane filtered Samples
Total Suspended Colloidal Soluble
(Schleicher and Schuell ME 25) samples for dissolved
COD (CODsoluble). The CODsuspended and CODcolloidal Run 1
were calculated by the difference between CODtotal Sewage 699 (190) 409 (239) 71 (33) 219 (85)
and CODfiltered, CODfiltered and CODsoluble, respectively. UASB effluent 203 (50) 49 (26) 53 (33) 101 (34)
Ammonia, total Kjeldahl nitrogen (TKN), nitrite, nitrate, %R* 69 (10) 85 (10) 17 (41) 46 (29)
sludge analysis and faecal coliform (FC) was determined MBBR effluent 54 (8) 12 (2) 6 (3) 37 (7)
according to APHA [18]. %R* 71 (10) 69 (17) 85 (11) 58 (20)
Overall removal 92 (1.3) 96 (2) 89 (9) 80 (10)
efficiency
Results and discussion Run 2
Sewage 733 (236) 485 (199) 64 (34) 185 (103)
Effect of HRT on the performance of the combined UASB effluent 244 (73) 80 (51) 42 (16) 122 (37)
UASB–MBBR system %R* 64 (13) 80 (15) 9 (69) 24 (24)
MBBR effluent 95 (21) 17 (4) 12 (3) 65 (16)
COD fractions removal %R* 57 (17) 63 (43) 64 (23) 42 (21)
Overall removal 86 (6) 96 (2) 73 (22) 55 (28)
The results presented in Table 3 and Fig. 2 indicate that efficiency
decreasing the total HRT from 13.3 to 10 h and from 10 to Run 3
5 h exerted a negative impact on the efficiency of the total Sewage 803 (301) 603 (349) 49 (27) 151 (53)
system (UASB–MBBR) as reflected in the residual COD UASB effluent 293 (71) 122 (66) 44 (18) 127 (36)
fractions values (CODtotal, CODsuspended, CODcolloidal and %R* 58 (22) 74 (23) 10 (41) 15 (12)
CODsoluble). At a total HRT of 13.3 h, the total process MBBR effluent 142 (23) 25 (7) 19 (5) 97 (17)
provided a final effluent quality with 54 mg l-1 CODtotal, %R* 49 (14) 69 (25) 51 (20) 20 (16)
6 mg l-1 CODcolloidal and 37 mg l-1 CODsoluble. Approx- Overall removal 80 (9) 94 (5) 51 (23) 20 (54)
imately, the same result, at a total HRT of 13.3 h, was efficiency
achieved in the MBBR system treating chemically pre- Standard deviations between brackets
treated sewage [10] at shorter HRT of 8.0 h. Residual COD %R*: percentage removal
values at a total HRT of 10 and 5.0 h were 95 and
142 mg l-1 for CODtotal, 65 and 97 mg l-1 for CODsoluble,
respectively. As expected, the UASB reactor achieved a 1400
Run 1 Run 2 Run 3
COD total (mgl )
-1

poor removal efficiency of CODcolloidal as shown in 1200

1000
Table 3. This low removal efficiency mainly can be due to 800
a poor physical removal in the system [19]. On the other 600
400
hand, an almost complete removal of CODcolloidal was 200
achieved in the MBBR, i.e. only 6, 12 and 17 mg l-1 0
remained in the final effluent when operated at HRT’s of 0 50 100 150 200 250 300 350

5.3, 4 and 2 h, respectively. The removal of CODcolloidal in Time (days)

the MBBR occurred mainly due to adsorption followed by Influent UASB eff. MBBR eff.
hydrolysis and biodegradation.
The results in Table 3 revealed that the removal of Fig. 2 Variation of CODtotal in the combined system at different
CODsuspended in the combined system was not significantly HRT’s
affected by decreasing the total HRT from 13.3 to 10 h and
from 10 to 5 h. The major part of CODsuspended was
removed in the UASB reactor, and little additional removal The fate of the COD in the domestic wastewater fed to
occurred in the MBBR system (Table 3). At a total HRT’s the combined UASB–MBBR units during experimental
of 13.3, 10 and 5.0 h, percentage removal values for the runs 1, 2 and 3 is presented in Figs. 3 and 4. Approximately
combined system were 96, 96 and 94%, respectively. This 12.7, 4.4 and 16.2% of the influent COD could not be
indicates that the removal of CODsuspended independent on accounted for COD balance in the test runs 1, 2 and 3,
the imposed HRT. respectively. These results are in agreement with Singh and

123
Bioprocess Biosyst Eng (2010) 33:267–276 271

80 Nitrification efficiency
60
The nitrification efficiency in the MBBR treating UASB
reactor effluent at different organic loading rates (OLR,s) is
%

40
shown in Table 4 and Fig. 5. The results show that
20 increasing the organic loading rate (OLR) from 4.6 to 7.4
0
and from 7.4 to 17.8 g COD m-2 day-1, results in an
R1 R2 R3 increase of the ammonia concentration in the final effluent
Run no. from 13 to 18 and from 18 to 21 mg l-1, respectively. At
OLR of 4.6, 7.4, and 17.8 g COD m-2 day-1, ammonia
E(%) R(%) M(%) S(%) G1 (%) G2(%) L(%)
was removed by a value of 62, 28 and 19%, while at the
Fig. 3 COD balance for UASB reactor during experimental runs 1, 2 same time 11, 4.4 and 0.3 mg l-1 of nitrate were, respec-
and 3. Slices represent the terms of the balance as percentage (%) of tively produced. Based on these results, it can be concluded
influent COD: E (%): effluent COD; R(%): removed COD; M that the OLR imposed to the MBBR reactor should remain
(%):COD converted to biomass; S(%):COD assimilated by sulphate below 7.4 g COD m-2 day-1 to enhance the nitrification
reducing bacteria (approximately 0.67 g COD per g SO4 reduced)
[22]; G1(%): COD converted to CH4; G2 (%): dissolved CH4 in the process as also found by Rusten et al. [14] for MBBR
treated effluent (calculated according to Henry’s law) and L (%): treating pre-settled sewage.
unaccounted fraction of COD The results presented in Fig. 6 revealed that the nitrifi-
cation rate in MBBR was strongly dependant on CODsus-
pended/N ratio. A low nitrification rate was achieved in the
MBBR at the high influent CODsuspended/N ratio of 2.1 and
80 3.1, the nitrification rate was 0.1 and 0.03 g NO3 ?
70 NO2 m-2 day-1 as compared to CODsuspended/N ratio of
60
1.36, the nitrification rate amounted to 0.26 g NO3 ?
50
NO2 m-2day-1. This can be attributed to attachment of the
%

40
30 suspended solids on the surface of the nitrifying biofilm
20 where they take away oxygen which otherwise would have
10
been available for nitrifiers [23].
0
R1 R2 R3
Run no. Nitrogen loss
E(%) R(%) M(%) C (%)

The nitrogen removal in the MBBR treating UASB reactor

Fig. 4 COD balances for MBBR treating UASB reactor effluent
during experimental runs 1, 2 and 3. Slices represent the terms of the
effluent was 26% at an OLR of 4.6 g COD m-2 day-1 as
balance as percentage (%) of influent COD: E (%): effluent COD; R compared to 16% at higher OLR’s of 7.4 and 17.8 g COD
(%): COD removal; M (%): sludge production and C (%): COD m-2 day-1 (see Table 4; Fig. 7). The nitrogen loss can be
conversion due to (1) assimilation of biomass (2) denitrification
occurring in the anoxic zone of the biofilm [24].

Viraraghavan [20] who found a COD gap of about 10–15% Faecal Coliform (FC) removal
of the total input COD of the UASB reactor treating
sewage at 20 °C. This is partially attributed to COD con- The results in Fig. 8 show that a significantly improved FC
sumption for cell synthesis. A higher value of unaccounted reduction at increasing the HRT from 5.0 to 10 h and from
COD of 40.8–41.5% was recorded for anaerobic filter (AF) 10 to 13.3 h. The mean overall residual counts of FC at an
treating municipal wastewater [21]. HRT’s of 13.3, 10 and 5 h were 8.9 9 104, 4.9 9 105 and
Figure 4 shows the COD balance in the MBBR system 9.4 9 105 MPN per 100 ml, corresponding to overall log10
treating UASB reactor effluent. The average COD removal reduction of 2.3, 1.4 and 0.7, respectively. The results
efficiency R (%) was 73.3% (HRT = 5.3 h), 61% obtained revealed that FC removal mainly proceeds in the
(HRT = 4 h), and 51.5% (HRT = 2.0 h), of which a MBBR system as shown in Fig. 8.
fraction of 7, 20 and 30% is discharged as surplus sludge M The results presented in Figs.9 and 10 show that the
(%). The remaining portion of the removed COD (66.4, 41 removal of FC only significantly improved once the con-
and 21.5%) can be due to (1) biological conversion C (%) centration of the dispersed CODsuspended and CODcolloidal
of biodegradable organic matter, and (2) assimilation of has become very low and the HRT has increased from 2
heterotrophic and autotrophic bacteria. to 5.3 h. Apparently, dispersed COD removal is very

123
272 Bioprocess Biosyst Eng (2010) 33:267–276

Table 4 Nitrogen species removal in an MBBR treating UASB reactor effluent at different OLR’s
Parameters Nitrogen species (mg l-1) Nitrification rate
Samples g NO3 ? NO2-N (m-2 day-1)
TKN NH4-N NO2-N NO3-N Nitrogen loss

Run 1
Sewage 38 (5) 31 (5)
UASB effluent 36 (4) 34 (5)
%R* 5 (4) -9.7 (4)
MBBR effluent 15 (3) 13 (3) 0.4 (0.2) 11 (2) 10 (5) 0.3 (0.05)
%R* 58 (9) 62 (9) 26 (12)
Overall removal efficiency 60 (9) 59 (10) 30 (12)
Run 2
Sewage 41 (4) 24 (3)
UASB effluent 37 (5) 24 (3)
%R* 9 (9) 0.0
MBBR effluent 26 (3) 18 (2) 0.4 (0.2) 4.4 (1.2) 7 (5) 0.14 (0.04)
%R* 30 (10) 28 (10) 16 (12)
Overall removal efficiency 37 (7) 27 (10) 25 (9)
Run 3
Sewage 42 (5) 25 (3)
UASB effluent 35 (6) 27 (3)
%R* 17 (16) -8 (5)
MBBR effluent 29 (7) 21 (2) 0.17 (0.1) 0.33 (0.1) 5 (2) 0.03 (0.01)
%R* 17 (7) 19 (10) 16 (7)
Overall removal efficiency 31 (15) 14 (11) 29 (15)
Standard deviations between brackets
%R*: percentage removal
NH4-N and nitrification rate

Nitrification rate

30 0.3
(gNOxm-2d-1)

Run 1 Run 2 Run 3

0.25
25
0.2
20 0.15
y = -0.1242x + 0.4158
0.1
15 R2 = 0.9966
0.05
10 0
5 0 0.5 1 1.5 2 2.5 3 3.5

0 COD suspended/N ratio

0 50 100 150 200 250 300 350
Nitrification rate (gNOx-Nm-2d-1)
Time (days) Linear (Nitrification rate (gNOx-Nm-2d-1))

OLR (gCODm-2d-1) NH4-N (mg/l) Nitrification(NO3+NO2)

Fig. 6 Relationship between the CODsuspended/N ratio and nitrifica-
Fig. 5 Variation of the OLR (g COD m day ); NH4-N and -2 -1 tion rate in the MBBR system treating anaerobically pretreated
nitrification rate in the MBBR system treating anaerobically sewage
pretreated sewage
(E. coli) in the colloidal form is limiting step in the biofilm
system.
important to achieve a satisfactory FC removal. The frac-
tion of FC attached on the suspended solids will be Excess sludge production
removed as a result of sedimentation; while the free dis-
persed FC (attached to colloidal particles will be adsorbed The characteristics of the excess sludge of the combined
on the carrier material [25]. However, longer HRT is UASB–MBBR are presented in Table 5. The sludge vol-
required for removal of FC in the colloidal form. Tawfik ume index (SVI) of the wasted sludge from the UASB and
et al. [26] found that the removal of Escherichia coli MBBR system is below 74 ml g TS-1, which indicates

123
Bioprocess Biosyst Eng (2010) 33:267–276 273

Run 1 Run 2 Run 3

excellent settleability. The VSS/TSS ratio of 0.5–0.6
TN inf., TN eff. and TN

OLR(gCODm-2d-1)
50 30

NOx-N(mg/l) and
45
40 25 indicates that the wasted sludge from the UASB reactor is
loss (mg/l)

35
30
20 well stabilized, while the wasted sludge from MBBR still
25 15
20 needs post-stabilization [27] as the VSS/TSS ratio of this
15 10
10 5
sludge was higher than 0.5. The sludge yield coefficient is
5
0 0 strongly affected by the imposed SRT, because the results
0 50 100 150 200 250 300 350 in Table 5 reveal that the sludge yield coefficient in a
Time(days)
combined system (UASB–MBBR) operated at a SRT
Total N loss TN (effluent) NO3+NO2
TN (influent) OLR (gCODm-2d-1)
(21 days for UASB ? 5.0 day for MBBR) is almost three
times higher than at a SRT (118.3 days for UASB ? 22
Fig. 7 Nitrogen loss, nitrification rate and organic loading rate days for MBBR). However, the combined UASB–MBBR
(g COD m-2 day-1) of the MBBR system treating anaerobically still produced a relatively low amount of wasted sludge
pretreated sewage compared to conventional activated sludge processes [28].

8.5
Run 1 Run 2 Run 3
Discussion
FC (log10/100ml)

7.5

6.5

5.5
The results obtained in this study indicated that the com-
bined system consisting of UASB–MBBR system treating
4.5
domestic wastewater at a total HRT of 13.3 h is very
3.5
0 50 100 150 200 250 300 350
effective for removal of COD fractions, ammonia and FC.
Time (days) The total system removed over 92% of CODtotal; 96% of
CODsuspended 89% of CODcolloidal and 80% of CODsoluble.
Sewage UASB MBBR
These results are similar to those reported for UASB–septic
Fig. 8 Variation of FC (log10) at different HRT’s tank in combination with MBBR treating black wastewater
[13]. The combined system removed 92% of CODtotal and
99% of BOD7. A lower removal efficiency of COD (71.3–
Run 1 Run 2 Run 3 77.1%) was achieved in an MBBR treating domestic
COD suspended (mgl-1)

45 6.5
wastewater at an HRT of 6 h [10]. This indicates that the
(log10/100ml)

35 6
F.coliform

5.5 introduction of an UASB reactor (as a pretreatment) prior

25
5 to MBBR (as a post-treatment) increased the removal
15 4.5 efficiency of COD. Comparison of the results obtained
5 4 from the present study with that published by other
0 50 100 150 200 250 300 350
investigators indicates a considerable variation depending
Time (days)
on the treatment system and the operating conditions. Kim
COD suspended effluent F.coliform effluent(Log10) et al. [29] reported a similar COD removal value (92%) for
aerobic filter in combination with UASB reactor at a total
Fig. 9 CODsuspended and faecal coliform (FC) in the treated effluent HRT of 13.5 h (8 h for UASB ? 5.5 h for aerobic filter).
of MBBR
COD removal ranging from 90 to 94% has been found by
Tawfik et al. [30] at lower HRT (10.7 h) using a combined
Run 1 Run 2 Run 3 system consisting of UASB–down flow hanging sponge
F.coliform (log10/100ml)

30 6.5
COD colloidal (mgl-1)

25
(DHS) system. Sousa and Foresti [2] investigated UASB-
6
20 sequencing batch reactor (SBR) for sewage treatment. The
5.5
15 total system achieved an overall removal efficiency of 95%
5
10 of COD. An UASB-activated sludge (AS) system treating
5 4.5
domestic wastewater was investigated by Sperling et al.
0 4
0 50 100 150 200 250 300 350
[31]. The integrated system achieved a removal efficiency
Time (days) of COD (85–93%) at a total HRT of 7.9 h (4.0 h
UASB ? 3.9 h aerobic reactor). Coletti et al. [32] achieved
COD colloidal effluent F.coliform effluent(Log10)
similar removal efficiencies of 95% (BOD5) and 88%
Fig. 10 CODcolloidal and faecal coliform (FC) in the treated effluent (COD) in a compartmentalized UASB reactor followed by
of MBBR activated sludge system. Bodı́k et al. [3] studied sewage

123
274 Bioprocess Biosyst Eng (2010) 33:267–276

Table 5 Characteristics of wasted sludge in a combined UASB–MBBR treating domestic wastewater at different HRT’s
Parameters SV* SW* (105 °C) SW (550 °C) SVI* VSS/ SYC* (g sludge g COD SRT*
Samples (ml l-1) (gl-1) (gl-1) (ml g TS-1) TSS removed-1day-1) (day-1)

Run 1
UASB reactor 220 6 3 39 0.5 0.06 118.3
MBBR 100 2 1.4 50 0.7 0.07 22
Run 2
UASB reactor 593 10 6 59 0.6 0.18 51.0
MBBR 140 4 2.4 35z 0.6 0.2 11.4
Run 3
UASB reactor 890 12 6 74 0.51 0.2 21
MBBR 160 7 5.2 23 0.74 0.3 5.0
SV* sludge volume, SW* sludge weight, SVI* sludge volume index, SYC* sludge yield coefficient, SRT* sludge residence time

treatment system consisting of an anaerobic baffled filter Intermittently aerated MBBR treating anaerobically pre-
reactor followed by aerobic post-treatment (hanging poly- treated wastewaters at an OLR of 0.023–0.093 kg COD
propylene cords). The HRT in anaerobic and aerobic unit m-3 day-1 was investigated by Luostarinen et al. [13]. The
were 15 and 4 h, respectively. The total process achieved reactor removed 57% of total nitrogen (TN). Improvement
the following removal efficiencies; COD (78.6–83.0%); of nitrogen removal in an integrated system consisting of
BOD5 (92.5–94.0%) and TSS (80.9–92.7%). UASB and MBBR may occur by using UASB reactor for
CODsuspended removal values for the UASB–MBBR denitrification in combination with methanogens as
were 96, 96 and 94% at a total HRT’s of 13.3, 10 and 5.0 h, described by several researchers [29, 36]. In the UASB
respectively. Similar removal efficiency of 92% for TSS reactor any external electron donor is not required, and
was achieved using UASB–AS system at a HRT of 9.9 h denitrification at the inlet of the reactor would improve the
[31]. In another study, UASB reactor in combination with a COD removal. A combined system consisting of an UASB
submerged aerated biofilter provided 94% for TSS removal reactor and an aerobic membrane bioreactor (MBR) was
[4]. operated at 28–30 °C for the treatment of low-strength
The results of the average concentration of ammonia, synthetic wastewater. The nitrified effluent from MBR was
nitrate and nitrite and their removal efficiencies (Table 4) recirculated into the UASB with a ratio of 50–800%. Under
showed that when DO = 2 mg l-1 and OLR = 4.6 g COD these conditions, the denitrification became the preferred
m-2 day-1, the nitrification efficiency reached above 62%, pathway rather than methanogenesis in the UASB reactor.
which was in consistent with Painter [33] who reported that The combined system achieved total nitrogen (TN)
a DO value of at least 2.0 mg l-1 is essential to maintain removal efficiency of up to 82.8% [37]. Jun et al. [38]
complete nitrification in biological wastewater-treatment reported a maximum TN removal efficiency of 70% with a
systems. Wang et al. [10] investigated the nitrification recirculation ratio of 300% at a total HRT of 24 h for the
efficiency in MBBR system treating domestic wastewater treatment of raw sewage in a combined up-flow anaerobic
at different DO levels (6, 4, 2 and 1 mg l-1). The results sludge reactor and aerobic bio-filtration system.
showed that when DO [ 2 mg l-1, the efficiency of nitri- Our results obtained with UASB–MBBR system, oper-
fication reached 94.3%. Increasing DO concentration up to ated at a total HRT of 13.3 h show a high percentage
6 mg l-1 slightly improved the nitrification efficiency by a removal of FC (98.8%), corresponding to 2.3 log10
value of only 9.0%. When DO was lowered to 1 mg l-1, reduction. These results are comparable to those obtained
the ammonia removal amounted to 56%. in other biofilm systems, i.e. RBC system achieved a
The nitrogen removal in the MBBR treating UASB removal efficiency of 99–99.8% for E. coli at longer
reactor effluent amounted to 26% at an OLR of 4.6 g COD retention time Tawfik et al. [36]. The removal efficiency of
m-2 day-1. These results are in agreement with the studies FC by a combined process (UASB–DHS) system was
by Tawfik et al. [34] who found that 22% of the nitrogen investigated by Tawfik et al. [30]. The total process
remained unaccountable in a rotating biological contactor achieved 99.8% for FC removal.
(RBC) system treating UASB reactor effluent. Total The major part of FC was removed in the MBBR system
nitrogen losses of up to 30% have been recorded in the treating UASB reactor effluent indicating that, the biofilm
aeration tank of full scale, biological nitrogen removal play a role for removal of FC. Two possible mechanisms
(BNR) processes treating domestic wastewater [35]. have been reported for FC removal by biofilm processes

123
Bioprocess Biosyst Eng (2010) 33:267–276 275

[25]. The first mechanism involves adsorption of FC onto 2. Ammonia removal in the MBBR was significantly
the biofilm, while the second mechanism involves the influenced by the OLR. The overall ammonia removal
removal of FC through predation by other microbes such as ranged from 47 to 75% at 4.6 g COD m-2 day-1 and
protozoa, and metazoans, i.e. Se0 las et al. [39] observed in a from 26 to 47% at 7.4 g COD m-2 day-1. At OLR of
sand column that pathogenic bacteria removal was better in 17.8 g COD m-2 day-1, the ammonia removal was
the presence of a biofilm than without it. Bellamy et al. largely deteriorated (9–51%).
[40] explained the removal of bacteria in slow sand filters 3. As the overall HRT increased, faecal coliform (FC)
by adsorption of the bacteria on the biofilm attached to removal increased. The mean overall log10 reduction
sand grains. However, adsorption of pathogenic bacteria to was 2.3 and 1.4 at total HRT of 13.3 and 10 h,
the media is influenced by several factors such as the respectively. There was a very low FC reduction of
content of organic matter, the degree of biofilm develop- 0.7 log10 at a total HRT of 5.0 h.
ment, temperature; ionic strength and pH value [41]. The 4. The sludge yield coefficient in a combined system
other causes of pathogenic bacteria removal in the MBBR (UASB–MBBR) operated at a total SRT (21 days for
system could be predation (filter feeding) [26]. Sylvaine UASB ? 5.0 days for MBBR) is almost three times
et al. [42] studied the efficiency of pathogenic bacteria higher than at a total SRT (118.3 days for UASB ?
removal (1) with a biofilm surface and active protozoa, (2) 22 days for MBBR).
with a biofilm surface and inactivated protozoa, (3) with a 5. In view of these results, we recommend to use a
clean surface. Protozoa in the presence of a biofilm were combined UASB–MBBR system for sewage treatment
responsible for 60% of bacteria removal. Biofilm without at an HRT of 8 and 5.3 h, respectively.
protozoa and a clean surface each removed similar quan-
tities of bacteria. Further investigation for mechanism
removal of FC in MBBR system is required.
The discharged sludge from the UASB reactor treating References
domestic wastewater is rather well stabilized, i.e. VSS/TSS
ratio = 0.5–0.6 at imposed operational conditions. There- 1. Halalsheh M (2002) Anaerobic pre-treatment of strong sewage. A
fore, the UASB reactor can simultaneously treat domestic proper solution for Jordan. PhD thesis, Environmental Technol-
wastewater and stabilize sludge produced. In this case, a ogy Department, Wageningen University and Research Center,
The Netherlands
conventional digestion tank can be eliminated from the 2. Sousa JT, de Foresti E (1996) Domestic sewage treatment in an
process, especially in tropical and subtropical countries up-flow anaerobic sludge blanket-sequencing batch reactor
where the temperature exceeding 20 °C. It is known the system. Water Sci Technol 33(3):73–84
investment cost of the digestion system generally amounts 3. Bodı́k I, Kratochvı́l K, Gapariková E, Hutňan M (2003) Nitrogen
removal in an anaerobic baffled filter reactor with aerobic post-
to 30–40% of the total cost of the whole sewage treatment treatment. Bioresour Technol 86(1):79–84
plant. Accordingly, a large portion of the investment can be 4. Goncalves R, Araujo V, Bof VS (1999) Combining up-flow
saved by using UASB reactor for sewage treatment. anaerobic sludge blanket (UASB) reactors and submerged aerated
Moreover, the mean value of the net sludge yield coeffi- biofilters for secondary domestic wastewater treatment. Water Sci
Tech 40(8):71–79
cient found amounted to only 0.06, 0.18 and 0.2 g sludge g 5. Aiyuk S, Forrez I, Lieven DK, van Haandel A, Verstraete W
COD removed-1 day-1 for the UASB reactor when oper- (2006) Anaerobic and complementary treatment of domestic
ated at SRT’s of 118.3, 51 and 21 days, respectively. The sewage in regions with hot climates: a review. Bioresour Technol
excess sludge fraction corresponding to only approxi- 97:2225–2241
6. Castillo A, Cecchi F, Mata-Alvarez J (1997) A combined
mately, 6% of the total influent COD at an HRT of 8.0 h anaerobic–aerobic system to treat domestic sewage in coastal
and to 20% at the shorter one (HRT = 3 h). Similar results areas. Water Res 31(12):3057–3063
has been achieved by Seghezzo et al. [43] who found that 7. A0 lvarez JA, Armstrong E, Go0 mez M, Soto M (2008) Anaerobic
the excess sludge from the UASB reactor treating pre- treatment of low-strength municipal wastewater by a two-stage
pilot plant under psychrophilic conditions. Bioresour Technol
settled sewage at an HRT of 6.3 h was 0.18 kg sludge kg 99:7051–7062
COD removed-1 day-1 . 8. Luostarinen S, Rintala J (2006) Anaerobic on-site black water
and kitchen waste treatment using UASB-septic tanks at low
temperatures. Water Sci Technol 54(2):143–149
Conclusions 9. Ji M, Yu J, Chen H, Yue PL (2001) Removal of slowly bio-
degradable COD in combined thermophilic UASB and MBBR
1. CODtotal removal in the combined UASB–MBBR is Systems. Environ Technol 22(9):1069–1079
10. Wang XJ, Xia SQ, Chen L, Zhao JF, Renault NJ, Chovelon JM
significantly influenced by the HRT and OLR. The (2006) Nutrients removal from municipal wastewater by chemi-
overall removal efficiencies in this study were 92, 86 cal precipitation in a moving bed biofilm reactor. Process
and 80% at HRT’s of 13.3, 10 and 5.0 h, respectively. Biochem 41:824–828

123
276 Bioprocess Biosyst Eng (2010) 33:267–276

11. Rusten B, Hem L, Ødegaard H (1995) Nitrification of municipal 28. Carlsson B, Hasselblad S, Plaza G, Martensson S, Lindberg GF
wastewater in moving bed biofilm reactors. Water Environ Res (1997) Design and operation of a pilot scale activated sludge
67(1):75–86 plant. Vatten 53(1):27–32
12. Hem L, Rusten B, Ødegaard H (1994) Nitrification in a moving 29. Kim Y, Mikawa K, Saito T, Tanaka K, Emori H (1997) Devel-
bed biofilm reactor. Water Res 28:1425–1433 opment of novel anaerobic/aerobic filter process for nitrogen
13. Luostarinen S, Luste S, Valentı0 n L, Rintala J (2006) Nitrogen removal using immobilized nitrifiers pellets. Water Sci Tech
removal from on-site treated anaerobic effluents using intermit- 36(12):151–158
tently aerated moving bed biofilm reactors at low temperatures. 30. Tawfik A, Ohashi A, Harada H (2006) Sewage treatment in a
Wat Res 40:1607–1615 combined up-flow anaerobic sludge blanket (UASB)—down-flow
14. Rusten B, Mike PR, Siljudalen JG (1998) The innovative moving hanging sponge (DHS) system. Biochem Eng J 29(3):210–219
bed biofilm reactor/solids contact re-aeration process for sec- 31. Sperling MV, Freire VH, Chernicharo CAD (2001) Performance
ondary treatment of municipal wastewater. Water Environ Res evaluation of an UASB-activated sludge system treating muni-
70(5):1083–1093 cipal wastewater. Water Sci Tech 43(11):323–328
15. Snedecor GW, Cochran WG (1980) Statistical methods. The 32. Coletti FJ, Povinelli J, Daniel LA (1997) Pos-tratamento por
Iowa state university press, Ames lodos ativados de efluentes provenientes de processos anaerobios
16. Dupla M, Yves C, Serge P, Richard V, Mario J (2006) Design de tratamento de esgoto sanitario; determinacao de constants
optimization of a self-cleaning moving-bed bioreactor for sea- cineticas. In: anais:19 congresso brasileiro de engenharia sani-
water denitrification. Water Res 40:249–258 taria e ambiental, for fo iguacu, set/97 (in Portuguese)
17. Zeeuw WJ de (1984) Acclimatization of anaerobic sludge for 33. Painter HA (1997) Microbial transformations of inorganic nitro-
UASB reactor start-up. Ph-D thesis, Department of Environ- gen. Prog Water Technol 8(4/5):3–11
mental Technology, Wageningen University, The Netherlands 34. Tawfik A, Klapwijk A, El-Gohary F, Lettinga G (2002) Treat-
18. APHA (2005) Standard methods for the examination of water and ment of anaerobically pre-treated domestic sewage by a rotating
wastewater, 21st edn. American Public Health Association/ biological contactor. Water Res 36(1):47–155
American Water Works Association/Water Environment Feder- 35. Randall CW, Barnard JL, Stensel HD (1992) Design and retrofit
ation, Washington DC of wastewater treatment plants for biological nutrient removal.
19. Elmitwalli TA, Soellner J, de Keizer A, Zeeman G, Bruning H, Water quality management library, vol 5. Technomic publishing
Lettinga G (2001) Biodegradability and change of physical company Inc, Lancaster, PA
characteristics of particles during anaerobic digestion of domestic 36. Barber WP, Stuckey C (2000) Nitrogen removal in a modified
sewage. Water Res 35(5):1311–1317 anaerobic baffled reactor (ABR): 1, denitrification. Water Res
20. Singh KS, Viraraghavan T (1998) Start up and operation of 34(9):2413–2422
UASB reactors at 20 C for municipal wastewater treatment. 37. Yingyu A, Fenglin Y, Hwee CC, Fook SW, Wu Bing (2008) The
J Ferment Bioeng 85(6):609–614 integration of methanogenesis with shortcut nitrification and
21. Ioannis DM, Sotirios GG (2008) Restart of anaerobic filters denitrification in a combined UASB with MBR. Bioresour
treating low-strength wastewater. Bioresour Technol 99:3579– Technol 99:3714–3720
3589 38. Jun HB, Park SM, Park JK, Choi CO, Lee JS (2004) Nitrogen
22. Metcalf and Eddy Inc (2003) In: Tcobanoglous G, Burton FL, removal in an up-flow sludge blanket (USB) reactor combined by
Stensel HD (eds) Wastewater engineering: treatment and reuse, aerobic biofiltration systems. Water Sci Technol 49(5):191–197
4th edn. McGraw-Hill, New York 39. Se0 las B, Lakel A, Andres Y, Le Cloirec P (2002) Wastewater
23. Temmink H, Klapwijk A, de Korte KF (2001) Feasibility of the reuse in on-site wastewater treatment: bacteria and virus move-
biofix—process for treatment of municipal wastewater. Water Sci ment in unsaturated flow through sand filter. Water Sci Technol
Tech 43(1):241–249 47(1):59–64
24. Holman JB, Wareham DG (2005) COD, ammonia and dissolved 40. Bellamy WD, Hendricks DW, Logsdon GS (1985) Slow sand
oxygen time profiles in the simultaneous nitrification/deni- filtration: influences of selected process variables. J Am Water
trification process. Biochem Eng J 22(2):125–133 Well Assoc 12:62–66
25. Tawfik A, Klapwijk A, Van Buuren J, El-Gohary F, Lettinga G 41. Stevik TK, Aa K, Ausland G, Hanssen JF (2004) Retention and
(2004) Physico-chemical factors affecting the E. coli removal in a removal of pathogenic bacteria in wastewater percolating through
rotating biological contactor (RBC) treating UASB effluent. porous media: a review. Water Res 38:1355–1367
Water Res 38:1081–1088 42. Sylvaine C, Yves A, Abdel L, Pierre Le C (2006) Bacteria
26. Tawfik A, Ohashi A, Harada H (2006) The influence of physical- removal in septic effluent: influence of biofilm and protozoa.
chemical and biological factors on the removal of faecal coliform Water Res 40:3109–3114
through down flow hanging sponge (DHS) system treating UASB 43. Seghezzo L, Gutierrez MA, Trupiano AP, Figueroa ME, Cuevas
reactor effluent. Water Res 40(9):1877–1883 CM, Zeeman G, Lettinga G (2002) The effect of sludge dis-
27. Mahmoud N, Zeeman G, Gijzen H, Lettinga G (2004) Anaerobic charges and up-flow velocity on the removal of suspended solids
stabilization and conversion of biopolymers in primary sludge. in a UASB reactor treating settled sewage at moderate tempera-
Effect of temperature and sludge residence time. Water Res tures. In: Proceedings of the VII Latin-American work-shop and
38(4):983–991 seminar on anaerobic digestion, Merida, Mexico, 22–25 October

123