wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 0

Avai lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ie r . com/ loca te /wat res

Development of granular sludge for textilewastewater treatment

Khalida Muda a,*, Azmi Aris a, Mohd Razman Salim a, Zaharah Ibrahim b, Adibah Yahya b,Mark C.M. van Loosdrecht c, Azlan Ahmad a, Mohd Zaini Nawahwi b

aDepartment of Environmental Engineering, Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310, Skudai, Johor, MalaysiabDepartment of Biological Sciences, Faculty of Biosciences and Bioengineering, Universiti Teknologi Malaysia, 81310, Skudai, Johor, MalaysiacDepartment of Biotechnology, Delft University of Technology, Julianalaan 67, 2628BC Delft, The Netherlands

a r t i c l e i n f o

Article history:

Received 18 January 2010

Received in revised form

22 April 2010

Accepted 2 May 2010

Available online 25 May 2010

Keywords:

Granulation

Granule characterization

Textile wastewater

Sequencing batch reactor

Color removal

* Corresponding author. Tel.: þ60 607 553152E-mail address: khalida@utm.my (K. Mud

0043-1354/$ e see front matter ª 2010 Elsevdoi:10.1016/j.watres.2010.05.023

a b s t r a c t

Microbial granular sludge that is capable to treat textile wastewater in a single reactor

under intermittent anaerobic and aerobic conditions was developed in this study. The

granules were cultivated using mixed sewage and textile mill sludge in combination with

anaerobic granules collected from an anaerobic sludge blanket reactor as seed. The gran-

ules were developed in a single sequential batch reactor (SBR) system under alternating

anaerobic and aerobic condition fed with synthetic textile wastewater. The characteristics

of the microbial granular sludge were monitored throughout the study period. During this

period, the average size of the granules increased from 0.02 � 0.01 mm to 2.3 � 1.0 mm and

the average settling velocity increased from 9.9 � 0.7 m h�1 to 80 � 8 m h�1. This resulted in

an increased biomass concentration (from 2.9 � 0.8 g L�1 to 7.3 � 0.9 g L�1) and mean cell

residence time (from 1.4 days to 8.3 days). The strength of the granules, expressed as the

integrity coefficient also improved. The sequential batch reactor system demonstrated

good removal of COD and ammonia of 94% and 95%, respectively, at the end of the study.

However, only 62% of color removal was observed. The findings of this study show that

granular sludge could be developed in a single reactor with an intermittent anaero-

biceaerobic reaction phase and is capable in treating the textile wastewater.

ª 2010 Elsevier Ltd. All rights reserved.

1. Introduction granules as seeding material for aerobic granules develop-

The development of aerobic granules as a novel treatment

technology for wastewater has been extensively reported

using sequencing batch reactor (SBR) systems. The systemhas

been used to treat different types of wastewater and pollu-

tants such as dairy effluent (Arrojo et al., 2004), soybean-pro-

cessing wastewater (Su and Yu, 2005), nitrogen and

phosphorus-rich effluent (Cassidy and Belia, 2005), phenol

effluent (Carucci et al., 2009) and also municipal wastewater

(de Kreuk and van Loosdrecht, 2006). The use of anaerobic

2/81; fax: þ60 607 556615a).ier Ltd. All rights reserved

ment has recently been reported by Linlin et al. (2005).

In recent years, the ability of biodegradation for textile

dyeing wastewater and dyestuffs involving both anaerobic

and aerobic processes has been widely reported in the litera-

ture (Ong et al., 2005; Isik and Sponza, 2008; Franciscon et al.,

2009). Color removal and complete mineralization of the dyes

have been achieved through the combination of both

processes. For azo dyes, the cleavage of N]N bond, which

results in the removal of color, occurs during anaerobic stage

along with the generation of aromatic amines, a toxic

7.

.

Sampling point

DO probe

pH probe

Effluent

Influent

Influent tank

Air compressor controlled by a

timer

Effluent tank

Peristaltic pumps controlled by timers

Mass-flow controller

Fig. 1 e Schematic layout of the SBR reactor system.

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 04342

compoundwhich is detrimental to human health (van der Zee

and Villaverde, 2005). Under aerobic condition, mineralization

of the amines takes place to complete the treatment process.

Several studies have been carried out using anaerobic

granular sludge but this doesn’t lead to complete removal of

the dyes. Study conducted using granular sludge grown in

anaerobic/aerobic system for complete dye removal is

apparently missing.

Since complete dye degradation requires both anaerobic

and aerobic conditions, several studies have been focused on

treatment systems which utilized two different reactors to

fulfill both conditions (Ong et al., 2005; Moosvi and

Madamwar, 2007; Isik and Sponza, 2008). The operation of

such a system is rather complicated as the anaerobic micro-

organisms in the anaerobic tank need to be separated before

the wastewater can be pumped to the aerobic tank. To

simplify the system, a study was conducted to develop

microbial granular sludge that can survive and function in

both anaerobic and aerobic conditions and hence requires

only one reactor. Potential strict anaerobic microorganisms

can survive easily since oxygen only penetrates partially in

the granules during aerated phase of the process. The study

focused on the development of this type of granular sludge

and the effectiveness of the system in treating synthetic

textile dyeing wastewater.

2. Materials and methods

2.1. Wastewater composition

Synthetic wastewater with the following composition was

used: NH4Cl 0.16 g L�1, KH2PO4 0.23 g L�1, K2HPO4 0.58 g L�1,

CaCl2$2H20 0.07 g L�1, MgSO4$7H2O 0.09 g L�1, EDTA

0.02 g L�1and trace solution 1 mL L�1. The carbon sources used

in this experiment were glucose (0.5 g L�1), ethanol (0.125 g L�1)

and sodium acetate (0.5 g L�1). The trace elements used were

based on the composition recommended by Smolders et al.

(1995). The composition of the trace element was H3BO3

(0.15 g L�1), FeCl3$4H2O (1.5 g L�1), ZnCl2 (0.12 g L�1),MnCl2$4H2O

(0.12 g L�1), CuCl2$2H2O (0.03 g L�1), NaMoO4$2H2O (0.06 g L�1),

CoCl2$6H2O (0.15 g L�1), and KI 0.03 g L�1. Mixed dyes consisted

of Sumifix Black EXA, Sumifix Navy Blue EXF and Synozol Red

K-4B with total concentration of 50 mg L�1 was used in this

study. The mixture gave an initial COD of 1270 mg L�1; 1020

ADMI (American Dye Manufacturing Index) and average

ammonia concentration of 38 mg L�1. The pH of the synthetic

wastewater was adjusted to 7.0 � 0.5 before feeding.

2.2. Reactor set-up

The schematic representation of the reactor set-up is given in

Fig. 1. A column reactor was designed based on Wang et al.

(2004) and Zheng et al. (2005) with several modifications. The

column was designed for a working volume of 4 L with

internal diameter of 8 cm and total height of 100 cm. The

wastewater was fed into the reactor from the bottom of the

column. Air was supplied into the reactor by a fine air bubble

diffuser also from the bottom of the column. The decanting of

the wastewater took place via an outlet port located at 40 cm

height from the bottom of the reactor. The mean cell resi-

dence time (SRT) was set by the discharge of suspended solids

with the effluent.

2.3. Analytical methods

The morphological and structural observations of the

granules were carried out by using a stereo microscope

equipped with digital image processing and analyzer ((PAX-

ITv6, ARC PAX-CAM)). The microbial compositions within the

granules were observed qualitatively with scanning electronic

microscope (FESEM-Zeiss Supra 35 VPFESEM). The granules

were left to dry at room temperature prior to gold sputter

coating (Bio Rad Polaron Division SEM Coating System) with

coating current of 20 mM for 45 s. Other parameters such as

mixed liquor suspended solid (MLSS), mixed liquor volatile

suspended solid (MLVSS), COD, color and NH3 were analyzed

according to the Standard Methods (APHA, 2005).

The granules developed in the SBR column were analyzed

for their physical, chemical and biological characteristics.

Physical characteristics include settling velocity, sludge

volume index (SVI) and granular strength. The settling

velocity was determined by averaging the time taken for an

individual granule to settle at a certain height in a glass

column filled with tap water. The SVI assessment was carried

out according to the procedure described by Beun et al. (1999).

Determination of the granular strength was based on

Ghangrekar et al. (2005). Shear force on the granules was

introduced through agitation using an orbital shaker at

200 rpm for 5 min. At certain amplitude of the shear force,

parts of the granules that are not strongly attached within the

granules were detached. The quantity of the ruptured gra-

nules was separated by allowing the fractions to settle for

1 min in a 150 ml measuring cylinder. The dry weight of the

settled granules and the residual granules in the supernatant

weremeasured. The ratio of the solid in the supernatant to the

Fig. 2 e Change in biomass concentration during the

formation of granular sludge in the SBR. (C) MLSS, (,)

MLVSS, (A) Suspended solid in the effluent.

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 0 4343

total weight of the granular sludge used for granular strength

measurement was expressed in percentage as an integrity

coefficient (IC). This percentage indirectly represents the

strength of the granules. Smaller IC value indicates stronger

granule and vice versa.

The granules were analyzed chemically for their mineral

content which includes Ca2þ, Mg2þ, Naþ, Kþ and Fe2þ. Themineral content was determined using Perkin Elmer Analyst

400 Flame Atomic Absorption Spectrophotometer (FLAA). The

microbial activity of the microbial granular sludge was con-

ducted by measuring the oxygen utilization rate (OUR),

specific oxygen utilization rate (SOUR) and specific methano-

genic activity (SMA). The OUR (mg.O2 L�1 h�1) measurement

was performed according to the Standard Methods (APHA,

2005). The SMA measurements were conducted according to

Erguder and Demirer (2008) with several modifications where

a 500 mL BOD bottle seeded with FSG with final concentration

of 1e2 g VSS L�1 and basal medium (250 mL effective volume).

The bottle was flushed with N2 gas mixture for 5 min to obtain

an anaerobic condition. The bottle was then sealed with

rubber septum. Acetic acid (HAc) was fed into the serumbottle

at a concentration of 3000 mg L�1. The experiments were

conducted in a temperature controlled condition of 30 � 2 �C.The production of methane gas (CH4) was determined

according to Erguder et al. (2001). The production of methane

gas (CH4) was determined daily for 5e7 days using liquid

displacement methods containing concentrated KOH stock

solution (20 g L�1). After each gasmeasurement, the bottlewas

manually shaken. At the end of the SMA assay, the VSS

content in the bottle was measured. The SMA was calculated

as the maximum CH4 produced per gram of VSS per hour

(mL CH4/g�1 VSS h�1) (Zitomer and Shrout, 1998).

2.4. Experimental procedures

A mixture containing an equal volume of sludge from

a municipal sewage treatment plant and a textile mill

wastewater treatment plant that gave a total volume of 2 L

was used in this experiment. The sludge inoculums were

sieved with amesh of 1.0 mm to remove large debris and inert

impurities. The sludge mixture was acclimatized for two

months with 2 L of synthetic wastewater containing dye

degrading microbes. The dye degrading microbes used in this

study was based on the study conducted by Nawahwi (2009)

and Ibrahim et al. (2009). Together with the sludge mixture,

about 100 mL of anaerobic granules with size less than 1 mm

diameter were used as seed for the granulation process. The

anaerobic granules were collected from an anaerobic sludge

blanket reactor system treating paper mill industrial effluent.

The MLSS of the anaerobic granules were 3.3 g L�1.

During the start-up period, 2 L of mixed sludge and 2 L of

synthetic textile wastewater were added into the reactor

system making the final volume of 4 L with total sludge

concentration after inoculation of 5.5 g L�1. The system was

supplied with external carbon sources consist of glucose,

sodium acetate and ethanol which gave a substrate loading

rate of 2.4 kg COD m�3 d�1. The reactor was operated in

successive cycles of 6 h. Each cycle comprised of 5 min filling,

340 min reaction, 5 min settling, 5 min decanting and 5 min

idle. The reaction phase started with an anaerobic phase of

40min, followed by aerobic phase for 130min, another 40 min

of second anaerobic phase and 130 min for second aerobic

phase. The dissolved oxygen (DO) concentration remained

low during the anaerobic condition (0.2 mg L�1) and reached

saturation concentration during the aerobic phase. The

superficial air velocity during the aerobic phase was

1.6 cm s�1. The pH during the reaction process varied in the

range of 6.0e7.8 and the temperature of the reactor system

was set at 30 � 2 �C. The reactor system was operated for

a period of 66 days. Two liters of the wastewater remained in

the reactor after the decanting stage yielding a volumetric

exchange rate (VER) of 50%. 20 mL of sample from the influent

and effluent (wastewater released after decanting stage) of the

reactor system were collected for the measurement of COD,

ammonia and color removal (Fig. 1).

3. Results and discussions

3.1. Biomass profile

The change in biomass concentration (i.e. MLSS) from the

start-up until the end of the study is shown in Fig. 2. During

the first few days of the experiment, almost half of the sludge

was washed-out from the reactor causing a rapid decrease in

the biomass concentration. The MLSS reduced from initial

concentration of 5.5 g L�1 to 2.9 g L�1mainly due to the short

settling time used in the cycle (i.e. 5 min). During this initial

stage, the anaerobic granules were also observed to disinte-

grate into smaller fragmented granules and small debris

resulted from shear force caused by the aeration during the

aerobic stage. These small fragments have poor settling ability

and were washed out from the reactor. This caused an

increase of the suspended solids concentration in the effluent

as shown in Fig. 2. As the experiment continued and granules

with adapted biomass were formed, the concentration of the

biomass in the reactor increased and finally reached

7.3 � 0.9 g MLSS L�1 when the experiment was discontinued

on the 66th day. The MLVSS followed the same trend as MLSS,

Fig. 3 e The changes of dissolved oxygen and oxygen uptake rate in one complete cycle of the SBR reactor system (A) DO,

(,) OUR at day 66th of the experiment. (PI) First aerobic phase, (PII) First anaerobic phase, (PIII) Second aerobic phase, (PIV)

Second anaerobic phase.

Table 1 e The OUR levels during the aerobic reactionphase of one complete cycle.

Aerobic reactionphase

OUR (mg L�1 h�1)

1st stage (PII) 2nd stage (PIV)

Begin react 281 � 39 167 � 51

End react 14 � 2 11 � 2

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 04344

ranging from 1.9 � 0.5 g L�1 to 5.6 � 0.8 g L�1. The SRT also

increased from 1.4 days at the initial stage to 8.3 days on the

66th day, indicating the accumulation of the biomass in the

reactor.

3.2. Bioactivities of the granules

A typical DO concentration profile for one complete cycle and

the OUR profile during both of the aerobic reaction phases are

shown in Fig. 3. Stage PI and PIII show the first and second stage

of anaerobic reaction phase, respectively. At these anaerobic

stages, most of the dye degradation is expected to occur where

amines, as the byproduct,were released (Sponzaand Isik, 2005).

Stage PII and PIV represent the first and second stage of aerobic

reaction phase, respectively. Most of the substrates provided to

the reactor system were anticipated to be consumed within

a fewminutes of the first aerobic reaction phase (PII), known as

the feast period. During the feast period, the DO concentration

in the reactor was low (about 4 mg L�1). The high utilization of

DO during the feast period was also indicated by the high OUR

which was 281 mg L�1 h�1. The amines, which were produced

during anaerobic reaction phase (PI), were mineralized under

this aerobic condition (PII) as they cannot be further degraded

under anaerobic phases (Sponza and Isik, 2005).

When all the carbon sources (substrate and amines) in the

wastewater were utilized, endogenous respiration process

took place, which is referred as the famine period. The tran-

sition from the feast to famine phase was clearly observed

with the drastic increase of the dissolved oxygen and the

extreme drop of the OUR within few minutes of the aerobic

reaction phase (PII). The DO concentration immediately

increased to around 7.0 mg L�1 which was closed to the DO

saturation level. The OUR also reduced to 14 mg L�1 h�1

indicating low utilization of DO.

Since there was no addition of substrate during the second

aerobic reaction phase (PIV), the consumption of DO during

this phase was also low. This is shown by high DO level

reaching saturation value of 7.6 mg L�1. A sharp increase in

the OUR was observed at the beginning of this phase but at

a lower value than the one observed in Stage PII. Apparently,

the residual dyes which were not degraded in Stage PI and PII

were transformed into smaller molecules (e.g. amines) during

the second stage of the anaerobic phase (PIII). These smaller

molecules were further mineralized in Stage PIV which

resulted in a sharp increase in the OUR. As the concentration

of these molecules were reduced, the OUR also became lower

until it reached aminimumof 11mg L�1 h�1 Table 1 shows the

OUR value during both of the aerobic reaction phases.

The SOUR of themicrobial granular sludgewas determined

before the termination of the experiment. The SOUR was

51.1 � 6.8 mg DO g�1 VSS h�1. This value was slightly lower

than those of the aerobic granules reported by Tay et al. (2001)

which ranged from 55.9 to 69.4 mg DO g�1 VSS h�1and higher

than the coupled granules reported by Erguder and Demirer

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 0 4345

(2005) (6e47 mg DO g�1 VSS h�1). The specific methanogenic

activity (SMA) of the microbial granular sludge is lower

(10.3 mL CH4 g�1 VSS h�1) than the one reported by Erguder

and Demirer (2005) (14e42 mL CH4 g�1 VSS h�1). However,

despite the low SMA emission, it provides the evidence of the

existence of methanogens within the microbial granular

sludge. Obviously, the granular sludge offered the metha-

nogens sufficient protection from the toxic oxygen concen-

tration in the bulk liquid.

3.3. Morphology of granular sludge

A week after inoculating the reactor, visual and microscopic

observations of granule formation were made. At this stage,

the developed granules were composed mostly of loosely

clumped sludge which could easily break up into pieces if

placed under vigorous shaking. Within a week, the anaerobic

seed granules had undergone morphological changes from

spherical in shape and black in color with average diameter of

1mm into smaller grey granules. It is likely that the sulfides in

the anaerobic sludge were oxidized due to exposure to the

aerobic condition. On day 30, two different types of granules

were clearly observed in the reactor as shown in Fig. 4.

Fig. 4a shows mainly irregular shaped, yellow colored

granules (Type A) that are solely developed from the activated

sludge. In Fig. 4b, the anaerobic granules that have frag-

mented into smaller pieces have formed different sizes of

granules (Type B) that contained pieces of anaerobic granules.

The outer layer of the latter were yellow in color indicating the

Fig. 4 e Morphological development of granular sludge from an

Pictures were taken using stereo microscope with magnification

from the activated sludge. (b) Granules developed from anaerobi

experiment. (d) Microbial granular sludge at the 66 days of the

domination of aerobic or facultative microorganisms while

the darker spots within the granules indicate the presence of

anaerobic fragments originated from anaerobic granules. The

formation of Type A granules could be elucidated by the

mechanisms explained by Beun et al. (1999). The development

was initiated from the mycelial pellets that were retained in

the reactor due to high settling velocity. Thesemycelial pellets

eventually become the supportmatrix for the bacteria growth.

Bacteria that were able to attach to this matrix were retained

and suppressed the filamentous growth and became the

dominant species in the reactor.

The formation of Type B granules has been discussed by

Linlin et al. (2005). These granuleswere formed through a series

of physical and morphological changes. The anaerobic gra-

nules initially disintegrated into smaller size flocs and debris

when exposed to aeration forces in the SBR column. Some of

the granules and debris that were too small were washed-out

with the effluent while the heavier ones were retained in the

column and acted as nuclei for the formation of the aerobic

granules. This type of granules that consisted of combination of

aerobic and anaerobic portions within the granules could

increase the possibility of degradation process that requires

both aerobic and anaerobic conditions for complete degrada-

tion particularly for textile wastewater that contains azo-dyes.

Fig. 4c shows the sludge particles during the initial stage of the

experimentwith an average size of 0.02� 0.01mmwhile Fig. 4d

shows the granules at the final stage (day 66) of the experiment

with the average particle diameter size of 2.3 � 1.0 mm with

maximum size reaching up to 4 mm.

aerobic granular sludge and aerobic activated sludge.

of 6.33, Scale bar equals to 1 mm. (a) Granules developed

c granules patches. (c) Sludge particles during early stage of

experiment.

Fig. 5 e SEM microstructure observation on mature microbial granular sludge at a magnification of 10,000 K. (a) Coccoid

bacteria tightly linked to one another. (b) Cavities that appear between bacteria clumped inside the granules.

Fig. 6 e The relationship between the settling velocity of

the microbial granular sludge and the biomass

concentration retained in the reactor. (,) Settling velocity

(C) Biomass concentration.

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 04346

The microstructure of the microbial granular sludge was

examined using SEM and shown in Fig. 5. The SEM observation

of the mature granules shows the domination of non-fila-

mentous coccoid bacteria that are tightly linked and embedded

to one another and form a rounded shape on the surface of the

granule, covered with extrapolysaccarides (EPS) (Fig. 5a). The

absence of filamentous bacteria in the developed granulesmay

be due to the experimental conditions that did not favor the

growth of filamentous bacteria at such high concentration of

DO during aerobic phase (i.e. 7.0� 0.5mg L�1) and considerably

high organic loading rate (2.4 kg COD m�3 d�1) (Chudoba, 1985;

Zheng et al., 2006). Fig. 5b shows the presence of cavities

between the clumpedbacteria. These cavities are anticipated to

be responsible in allowing smooth mass transfer of substrates

or metabolite products in and out of the granules.

3.4. Physical characteristics of granular sludge

3.4.1. SizeThe shear force imposed in the development of granules in

this experiment, in terms of superficial upflow air velocity

which was 1.6 cm s�1, resulted in the development of micro-

bial granular sludge with average diameter of 2.3 � 1.0 mm.

According to Peng et al. (1999), the diameter of the developed

aerobic granules is in the range of 0.3e0.5 mm which is much

smaller as compared to anaerobic and anoxic granules that

could develop up to 2 to 3 mm. The strong shearing force

produced by mixing and aeration during the reaction phase

could prevent the development of bigger granules which can

be achieved in an anaerobic system (van Benthum et al., 1996;

Kwok et al., 1998).

3.4.2. Settling velocityThe average settling velocity of the seed sludge and seed

anaerobic granular sludge was 9.9 � 0.7 m h�1 and

42 � 8 m h�1 respectively. The average settling velocity of the

anaerobic granular seed is in accordance with those reported

by Schmidt and Ahring (1996) which is in the range of

18e100 m h�1. The average settling velocity of the granular

sludge developed in this study increased from 17.8� 2.6m h�1

to 80 � 8 m h�1 at the end of experiment. The settling velocity

obtained in this study is almost three times greater than the

settling velocity of the aerobic granules reported by Zheng

et al. (2005) (i.e. 18e31 m h�1).

The increase in settling velocity has given significant

impact on the biomass concentration in the reactor. The

relationship between settling velocity of the granules and the

concentration of the MLSS is shown in Fig. 6. Despite the short

settling time (5 min), the high settling velocity possessed by

the developed microbial granular sludge enabled the granules

to escape from being flushed out during the decanting phase.

Such conditions have caused more microbial granular sludge

to be retained in the system and resulted in the increase of

biomass concentration in the reactor. In this experiment, the

SRT value was 1.4 days during the start-up (partly low due to

wash out of inoculated sludge) and rose up to 8.3 days at the

end of experiment. As less biomass was washed-out during

the decanting period, the increase in SRT is another mani-

festation of good settling characteristics resulting from the

high settling velocity.

The SVI value has improved from 276.6mL g�1 at the initial

stage of the experiment to 69 mL g�1 at the end of the

experiment indicating the good settling properties of the

Fig. 7 e The SVI and mean cell residence time (SRT) profile.

(B) SVI, (-) SRT.

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 0 4347

granules which is favorable in wastewater treatment plant

operation. The change of the SVI value and the SRT as

a function of time are given in Fig. 7. The SVI value achieved in

this experiment is in agreement with the result reported by

McSwain et al. (2004) with SVI values of 115 � 36 mL g�1

(settling time 10 min) and 47 � 6 mL g�1 (settling time 2 min).

The higher settling velocity and lower SVI value of the mature

microbial granular sludge as compared to previous reports by

other researchers indicate that the formation of granules

seeded with anaerobic granules would develop better settling

properties of the granules. It may also be due to the specific

reaction condition of anaerobic/aerobic setting in the experi-

ment that induced the well settling of the granule.

3.4.3. Granular strengthThe granular strength of the granules was measured based on

the integrity coefficient (IC) as described by Ghangrekar et al.

(2005). The smaller the IC value, the higher the strength and

ability of the granules to remain as high structural integrity

granules during aeration phase that caused the shear force.

Fig. 8 e The change in integrity coefficient (representing

the granular strength) during the formation of granular

sludge in the anaerobic/aerobic SBR.

Fig. 8 shows the IC profile of the developed granular sludge as

a function of time. The IC reduces as the granules developed.

With an initial value of 30 � 0.3, the IC was reduced to about

9.4 � 0.5 at the termination of the experiment. A sharp

reduction of IC was observed after 40 days of the experimental

run. According to Ghangrekar et al. (2005) granules with

integrity coefficient of less than 20 were considered high

strength granules. The reduction in IC value indicates the

increase in the strength of the bond that holds the microor-

ganisms together within the developed granules.

During the early stage of the granule development, the

microbes within the granules were loosely bounded to each

other. When the microbes were loosely linked together, the

granules may contain more cavities which make the granules

less dense, as shown by low settling velocity value. As more

microbes were linked together, the granules increased in size.

Under the applied selective pressures (i.e. short settling time,

hydrodynamic shear force, feast-famine regime) within the

reactor, microbes may produce more EPS (Qin et al., 2004). As

reported by Adav et al. (2008), the EPS could contribute greatly

to the strength and the stability of anaerobic granules. When

more EPS are being produced by the microbial cells, they form

a cross-linked network and further strengthen the structural

integrity of the granules. The cavities within the granulesmay

be filled with the EPS as it is a major component of the bio-

granule matrix material in both anaerobic and aerobic gra-

nules. This caused the granules to become denser and

stronger as shown by their high settling velocity and low IC

value at the end of the experiment.

3.4.4. Mineral contentThe concentration of minerals in granular sludge, newly

developed and matured granular sludge were determined in

mg/g of dry sludge and presented in Table 2. The concentra-

tion of Naþ and Kþ are not much different in the sludge, newly

developed and matured granules. However, there is an

obvious increase in the concentration of Ca2þ andMg2þwithin

the matured granules. The concentration of Fe2þ was slightly

reduced in the matured and newly developed granules as

compared to the sludge.

Basically, an unchanged concentration of Naþ and slightly

decreasing Kþ concentration in the sludge and matured

granulesmay indicate that thesemonovalent cations may not

be involved in the granulation process. It has been reported

that high concentration of the Naþ and Kþ may cause adverse

Table 2 e Comparison of mineral content at differentstages during the development of microbial granularsludge.

Mineral contents (mg g�1 of dry sludge)

Mineral Sludge Newly developedmicrobial granularsludge (1 week)

Matured microbialgranular sludge

(10 weeks)

Ca2þ 1.53 � 0.02 2.0 � 0.6 4.65 � 0.04

Mg2þ 0.13 � 0.01 0.322 � 0.003 1.75 � 0.08

Naþ 0.22 � 0.07 0.25 � 0.04 0.24 � 0.05

Kþ 1.31 � 0.07 1.15 � 0.03 0.93 � 0.05

Fe2þ 2.32 � 0.02 1.90 � 0.04 1.98 � 0.08

Fig. 9 e Percentage removal for (a) COD, (b) ammonia and (c)

color of the SBR system. (A) Influent, (B) Percentage

removal, (-) Effluent. (ADMI) American Dye Manufacturing

Units.

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 04348

effect on the granules formation. It could cause reduction in

sludge concentration, settling velocity of the sludge, granular

strength and treatment efficiency (Ghangrekar et al., 2005).

The monovalent cation, Naþ has also been reported causing

detrimental impact on the flocculation system (Sobeck and

Higgins, 2002). Nevertheless, there are contradictory reports

related to the effect of these monovalent cations in granula-

tion development processes. Fernandez et al. (2008) reported

that the concentration of granular biomass has improved

significantly in the reactor system fedwith high concentration

of inorganic salt influent. A quick decrease of solids concen-

tration in the effluent was observed after the addition of NaCl.

The development of the granular sludge in this study

showed higher accumulation of Ca2þ and Mg2þ towards the

end of the experiment. This may indicate the involvement of

the inorganic elements in the granulation process. Based on

the divalent cation bridging theory, the presence of Ca2þ and

Mg2þ promotes equivalent floc properties (Sobeck and

Higgins, 2002). According to Ren et al. (2008), granule-rich

Ca2þ showed more rigid granular structure and higher

strength as compared to granule without Ca2þ accumulation.

3.5. Removal performance

The performance of the reactor system from start-up until the

end of granules development period based on the removal of

COD, color and ammonia is given in Fig. 9. At the initial stage

of the operation, the percentage removal for COD and

ammonia was 71% and 67% respectively (Fig. 9a and b). The

removal efficiency increased to 94% for COD and 95% for

ammonia at the end of the experiment. The increase in the

removal efficiency indicates the occurrence of high biological

activity in the reactor system. During the first month, the

removal efficiency for COD and ammonia fluctuated but the

removal became stable for the remaining period. The removal

efficiency for color was fluctuating almost throughout the

study period (Fig. 9c). The percentage of color removal was

about 25% during the start up and increased to 62% at the end

of the experiment. The average of color removal was 55%. This

low percentage of the color removalmay be due to insufficient

adaptation time. As dye substances are recalcitrant and

difficult to be degraded, more time is required to accumulate

sufficient organisms which degrade the dyes in the reactor.

The inconsistent percentage for color removal may also be

contributed by the unstable condition of the aromatic amines,

the byproduct of dye degradation which easily oxidized and

recolorwhen exposed to oxygen during the aerobic phase. The

increase of color during autoxidation of aromatic amines was

confirmed by several researchers (Cruz and Buitron, 2001;

Libra et al., 2004; Sponza and Isik, 2005).

The inconsistency of color removal may be also influenced

by the absorption of color into sludge biomass throughout the

experiment. The absorption of color into the sludge biomass

has been reported by other researchers (Otero et al., 2003;

Wang et al., 2006; Sirianuntapiboon and Srisornsak, 2007).

Fig. 10 shows the percentage removal of COD, ammonia

and color in a complete 340-min reaction phase of the SBR

system recorded on the 66th days of experiment. The profile

and the percentage removal for COD and ammonia were

almost the same while the removal of color was much lower.

After 340 min of intermittent anaerobic and aerobic modes,

about 93%, 95% and 62% of COD, ammonia, and color

respectively were removed.

During the first anaerobic phase (PI) (0e40 min), approxi-

mately 15% and 4% of COD and ammonia respectively, were

removed. In the first aerobic phase (PII) (40e170 min), about

68% of the COD was removed while 80% of the ammonia was

oxidized. Most of the organic compounds and nitrification of

ammonia were achieved during this stage. The supply of

oxygen at this stage enabled good oxidation of these

compounds (Brauer and Henning, 1986). The nitrate produced

Fig. 10 e The removal for COD, ammonia and color in one

complete cycle of the SBR system. (-) Color, (B) COD, (:)

Ammonia. (PI) First aerobic phase, (PII) First anaerobic

phase, (PIII) Second aerobic phase, (PIV) Second anaerobic

phase.

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 0 4349

will be removed through the denitrification process that will

occur in the second stage of anaerobic phase. The second

anaerobic phase (PIII) (170e210 min) showed only about 5% of

COD and ammonia being removed while the remaining

(about 4%) was removed in the second aerobic phase (PIV)

(210e340 min). As for color, about 45% and 16% were removed

during anaerobic and aerobic phases respectively. The result

shows the ability of anaerobic microbes within the granule to

degrade the dye. The high percentage for color removal indi-

cated active cleavage of dye compound took place during both

of the anaerobic phases (PI and PIII).

The degradation and decolorization of dye during anaer-

obic condition has been widely reported in the literatures (dos

Santos et al., 2007). In anaerobic condition, the electrons from

the electron donor are transferred to the N]N bond of the azo

dye causing the cleavage of the bond forming aromatic

amines. The amines are then degraded under aerobic condi-

tion reducing the COD value of the wastewater. In addition to

the degradation mechanism, dye removal may also occur via

adsorption onto the biomass (Aksu, 2001 and Crini, 2006).

Amines, the colorless byproduct of anaerobic degradation of

dye compound are unstable compound that could easily be

oxidized during the presence of oxygen. These autoxidation of

the amines may form different intensity colored compound

(Cruz and Buitron, 2001; Libra et al., 2004; Sponza and Isik,

2005). This reaction may cause the reduction on the overall

percentage of color removal during aerobic condition (PII and

PIV). Based on the removal performance of the system, it has

been proven that the developed microbial granular sludge is

capable to perform the degradation process during anaerobic

and aerobic phases. This indicates the presence of aerobic,

facultative and anaerobic microorganisms in the microbial

granular sludge. According to Li and Liu (2005), when the

granules grew to a size larger than 0.5 mm, the diffusion of

oxygen into the inner part of the granules became a limitation.

This may give an indication of the presence of anaerobic

microorganisms within centre part of the microbial granular

sludge since the average size of microbial granular sludge

developed in this study was 2.3 � 1.0 mm. Aerobic microor-

ganismsmay be found at the outer layer of the granuleswhich

can easily access the oxygenmolecule andmainly responsible

for the COD removal. The facultative microorganisms may be

found in any part of the microbial granular sludge due to its

capability to live both under anaerobic and aerobic condition.

4. Conclusion

� Stablemicrobial granular sludgecouldbecultivated inasingle

SBRsystemwiththeapplicationof intermittentanaerobicand

aerobic reactionmode during the reaction phase.

� The matured granules showed the domination of non-fila-

mentous bacteria that were tightly linked and embedded to

one another and covered with EPS. The SVI value of the

biomass has decreased from 276.6 mL g�1 to 69 mL g�1at the

end the 66 days, also indicating the excellent settling

properties of the granules.

� The development of the granular sludge is positively

correlated with the accumulation of divalent cationic Ca2þ

and Mg2þ in the granules suggesting the role played by the

cations in the granulation process.

� The results indicate the viability of the single reactor system

for treating textile wastewater under intermittent anaerobic

and aerobic phase strategy.

� The OUR/SOUR and SMA analyses indicate the presence of

anaerobic and aerobic microorganisms activities in the

granular sludge which is capable to perform degradation

process both in anaerobic and aerobic conditions.

Acknowledgements

The authorswish to thank theMinistry of Science, Technology

and Innovation (MOSTI), Ministry of High Education (MOHE)

and Universiti Teknologi Malaysia for the financial supports of

this research (Grants No.: 79137, 78211 and 75221).

r e f e r e n c e s

Adav, S.S., Lee, D.J., Tay, J.H., 2008. Extracellular polymericsubstances and structural stability of aerobic granule. WaterResearch 42, 1644e1650.

Aksu, Z., 2001. Biosorption of reactive dyes by dried activatedsludge: equilibrium and kinetic modeling. BiochemicalEngineering Journal 7, 79e84.

Arrojo, B., Mosquera-Coral, A., Garrido, J.M., Mendez, R., 2004.Aerobic granulation with industrial wastewater in sequencingbatch reactors. Water Research 38, 3389e3399.

Beun, J.J., Hendriks, A., van Loosdrecht, M.C.M., Morgenroth, E.,Wilderer, P.A., Heijnen, J.J., 1999. Aerobic granulation ina sequencing batch reactor. Water Research 33 (10), 2283e2290.

Brauer, H., Henning, R., 1986. Aerobic two-stage biologicaldegradation of organic carbon compounds and ammonium.Bioprocess Engineering 1, 13e22.

Carucci, A., Milia, S., De Gioannis, G., Piredda, M., 2009. Acetate-fed aerobic granular sludge for the degradation of 4-chlorophenol. Journal of Hazardous Materials 166, 483e490.

wat e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 3 4 1e4 3 5 04350

Chudoba, J., 1985. Control of activated sludge filamentous bulkingVI formulation of basic principle. Water Research 19, 471e479.

Crini, G., 2006. Non-conventional low-cost adsorbents for dyeremoval: a review. Bioresource Technology 97, 1061e1085.

de Kreuk, M.K., van Loosdrecht, M.C.M., 2006. Formation ofaerobic granules with domestic sewage. Journal ofEnvironmental Engineering 132, 694e697.

Erguder, T.H., Demirer, G.N., 2005. Investigation of granulation ofa mixture of suspended anaerobic and aerobic cultures underalternating anaerobic/microaerobic/aerobic conditions.Process Biochemistry 40, 3732e3741.

Erguder, T.H., Demirer, G.N., 2008. Low-strength wastewatertreatment with combined granular anaerobic and suspendedaerobic cultures in upflow sludge blanket reactors. Journal ofEnvironmental Engineering 134, 295e303.

Erguder, T.H., Tezel, U., Guven, E., Demirer, G.N., 2001. Anaerobicbiotransformation and methane generation potential ofcheese whey in batch and UASB reactors. Waste Management21, 643e650.

Franciscon, E., Zille, A., Garboggini, F.F., Silva, I.S., Paulo, A.C.,Durrant, L.R., 2009. Microaerophiliceaerobic sequentialdecolourization/biodegradation of textile azo dyes bya facultative Klebsiella sp. strain VN-31. Process Biochemistry44, 446e452.

Ghangrekar, M.M., Asolekar, S.R., Joshi, S.G., 2005. Characteristicsof sludge developed under different loading conditions duringUASB reactor start-up and granulation. Water Research 39,1123e1133.

Ibrahim, Z., Amin, M.F.M., Yahya, A., Aris, A., Umor, N.A.,Muda, K., 2009. Characterisation of microbial flocs formedfrom raw textile wastewater in aerobic biofilm reactor (ABR).Water Science and Technology 60.3, 683e688.

Isik, M., Sponza, D.T., 2008. Anaerobic/aerobic treatment ofa simulated textile wastewater. Separation and PurificationTechnology 60, 64e72.

Linlin, H., Jianlong, W., Xianghua, W., Yi, Q., 2005. The formationand characteristics of aerobic granules in sequencing batchreactor (SBR) by seeding anaerobic granules. ProcessBiochemistry 40, 5e11.

McSwain, B.S., Irvine, R.L., Wilderer, P.A., 2004. The influence ofsettling time on the formation of aerobic granules. WaterScience and Technology 50 (10), 195e202.

Moosvi, S., Madamwar, D., 2007. An integrated process for thetreatment of CETP wastewater using coagulation, anaerobicand aerobic process. Bioresource Technology 98, 3384e3392.

Nawahwi, M.Z., 2009. Degradation of Reactive Red 195 by SelectedBacteria from Textile Wastewater, Master Thesis.

Ong, S.A., Toorisaka, E., Hirata, M., Hano, T., 2005. Treatment ofazo dye orange II in aerobic and anaerobic-SBR systems.Process Biochemistry 40, 2907e2914.

Peng, D., Bernet, N., Delgenes, J.P., Moletta, R., 1999. Aerobicgranular sludge-a case study. Water Research 33, 890e893.

Qin, L., Tay, J.H., Liu, Y., 2004. Selection pressure is a driving forceof aerobic granulation in sequencing batch reactors. ProcessBiochemistry 39, 579e584.

Ren, T.T., Liu, L., Sheng, G.P., Liu, X.W., Yu, H.Q., Zhang, M.C., etal., 2008. Calcium spatial distribution in aerobic granules andits effects on granule structure, strength and bioactivity.Water Research 42, 3343e3352.

Schmidt, E., Ahring, B.R., 1996. Granular sludge formation inupflow anaerobic sludge blanket (UASB) reactors.Biotechnology Bioengineering 49, 229e246.

Sirianuntapiboon, S., Srisornsak, P., 2007. Removal of dispersedyes from textile wastewater using bio-sludge. BioresourceTechnology 98, 1057e1066.

Sobeck, D.C., Higgins, M.J., 2002. Examination of three theories formechanisms of cation-induced bioflocculation. WaterResearch 36, 527e538.

Sponza, D.T., Isik, M., 2005. Reactor performances and fate ofaromatic amines through decolorization of Direct Black 38 dyeunder anaerobic/aerobic sequentials. Process Biochemistry 40,35e44.

Tay, J.H., Liu, Q.S., Liu, Y., 2001. Microscopic observation ofaerobic granulation in sequential aerobic sludge reactor.Journal of Applied Microbiology 91, 168e175.

van Benthum, W.A.J., Garrido, J.M., Tijhuis, L., Van Loosdrecht, M.C.M., Heijnen, J.J., 1996. Formation and detachment of biofilmsand granules in a nitrifying biofilm airlift suspension reactor.Biotechnology Progress 12, 764e772.

Zheng, Y.M., Yu, H.Q., Sheng, G.P., 2005. Physical and chemicalcharacteristics of granular activated sludge from a sequencingbatch airlift reactor. Process Biochemistry 40, 645e650.

Zheng, Y.M., Yu, H.Q., Liu, S.J., Liu, X.Z., 2006. Formation andinstability of aerobic granules under high organic loadingconditions. Chemosphere 63, 1791e1800.

Zitomer, D.H., Shrout, J.D., 1998. Feasibility and benefits ofmethanogenesis under oxygen-limited conditions. WasteManagement 18 (2), 107e116.

APHA, 2005. Standard Methods for the Examination of Water andWastewater. American Public Health Association,Washington, DC.

Cassidy, D.P., Belia, E., 2005. Nitrogen and phosphorus removalfrom an Abattoir wastewater in a SBR with aerobic granularsludge. Water Research 39 (19), 4817e4823.

Cruz, A., Buitron, G., 2001. Biodegration of disperse blue 79 usingsequenced anaerobic/aerobic biofilters. Water Science andTechnology 44, 159e166.

Fernandez, I., Vazquez-Padin, J.R., Mosquera-Corral, A.,Campos, J.L., Mendez, R., 2008. Biofilm and granular systemsto improve Anammox biomass retention. BiochemicalEngineering Journal 42 (3), 308e313.

Kwok, W.K., Picioreanu, C., Ong, S.L., Loosdrecht, M.C.M., Ng, W.J.,Heijinen, J.J., 1998. Influence of biomass production anddetachment forces on biofilm structures in a biofilm airliftsuspension reactor. Biotechnology and Bioengineeing 58 (4),400e407.

Li, Y., Liu, Y., 2005. Diffusion of substrate and oxygen in aerobicgranule. Biochemical Engineering Journal 27, 45e52.

Libra, J.A., Maren Borchert, M., Vigelahn, L., Storm, T., 2004. Twostage biological treatment of a diazo reactive textile dye and thefate of the dye metabolites. Chemosphere 56, 167e180.

Otero, M., Rozada, F., Calvo, L.F., Garcı’a, A.I., Mora’n, A., 2003.Kinetic and equilibrium modelling of the methylene blueremoval from solution by adsorbent materials produced fromsewage sludges. Biochemical Engineering Journal 15, 59e68.

dos Santos, A.B., Cervanes, F.J., van Lier, J.B., 2007. Review paperon current technologies for decolourisation of textilewastewaters: perspective for anaerobic biotechnology.Bioresource Technology 98, 2369e2385.

Smolders, G.J.F., Klop, J., van Loosdrecht, M.C.M., Heijnen, J.J.,1995. A metabolic model of the biological phosphorus removalprocess. Effect of the sludge retention time. Biotechnology andBioengineering 48, 222e233.

Su, K.Z., Yu, H.Q., 2005. Formation and characterization of aerobicgranules in a sequencing batch reactor treating soybeen-processing wastewater. Environmental Science Technology39, 2818e2827.

Wang, Q., Dua, G., Chena, J., 2004. Aerobic granular sludgecultivated under the selective pressure as a driving force.Process Biochemistry 39, 557e563.

Wang, Y., Mu, Y., Zhao, Q.B., Yu, H.Q., 2006. Isotherms, kineticsand thermodynamics of dye biosorption by anaerobic sludge.Separation and Purification Technology 50, 1e7.

van der Zee, F.P., Villaverde, S., 2005. Combinedanaerobiceaerobic treatment of azo dyes-a short review ofbioreactor studies. Water Research 39, 1425e1440.