Bioresource Technology 129 (2013) 620–623

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Short Communication

Cultivation of aerobic granular sludge for rubber wastewater treatment

0960-8524/$ - see front matter � 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.biortech.2012.12.113

⇑ Corresponding author. Tel.: +60 607 5531125/738; fax: +60 607 5531575.E-mail addresses: aznah@utm.my, nhasyimah9@gmail.com (A. Nor Anuar).

Noor Hasyimah Rosman a, Aznah Nor Anuar b,⇑, Inawati Othman a, Hasnida Harun a,Muhammad Zuhdi Sulong (@ Abdul Razak) a, Siti Hanna Elias a, Mohd Arif Hakimi Mat Hassan a,Shreesivadass Chelliapan b, Zaini Ujang b

a Department of Environmental Engineering, Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysiab Institute of Environment and Water Resource Management, WATER Research Alliance, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia

h i g h l i g h t s

" Aerobic granular sludge showed excellent settling ability and sludge volume index." A COD removal rate of 96.5% was observed at an OLR of 3.7 kg COD m�3 d�1." Ammonia and total nitrogen removal efficiencies were 94.7% and 89.4%, respectively.

a r t i c l e i n f o

Article history:Received 12 October 2012Received in revised form 13 December 2012Accepted 14 December 2012Available online 22 December 2012

Keywords:Aerobic granular sludgeGranule characterizationNitrificationRubber wastewaterSequential batch reactor

a b s t r a c t

Aerobic granular sludge (AGS) was successfully cultivated at 27 ± 1 �C and pH 7.0 ± 1 during the treat-ment of rubber wastewater using a sequential batch reactor system mode with complete cycle time of3 h. Results showed aerobic granular sludge had an excellent settling ability and exhibited exceptionalperformance in the organics and nutrients removal from rubber wastewater. Regular, dense and fast set-tling granule (average diameter, 1.5 mm; settling velocity, 33 m h�1; and sludge volume index,22.3 mL g�1) were developed in a single reactor. In addition, 96.5% COD removal efficiency was observedin the system at the end of the granulation period, while its ammonia and total nitrogen removal efficien-cies were up to 94.7% and 89.4%, respectively. The study demonstrated the capabilities of AGS develop-ment in a single, high and slender column type-bioreactor for the treatment of rubber wastewater.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Rubber is one of the main agro-based industrial sectors thatplay an important role in Malaysia’s economy. Presently, Malaysiais the fourth largest rubber producer in the world after Thailand,Indonesia and India (Vijayaraghavan et al., 2008). There are twotypes of processes in raw natural rubber processing; the produc-tion of latex concentrate and the Standard Malaysian Rubber(SMR) (Sulaiman et al., 2010). SMR is the current bulk of Malaysianrubber which is produced in the form of technically specifiedcrumb rubber. Large quantities of effluent were produced fromthe processing of raw natural rubber since it required huge amountof water for its operation. The effluent typically contains a smallamount of uncoagulated latex, serum with substantial quantitiesof proteins, carbohydrates, sugars, lipids, carotenoids, as well asinorganic and organic salts and also includes washings water from

the various processing stages (Mohammadi et al., 2010). Bothbiological and chemical methods are used to treat the rubberwastewater.

Research into aerobic granular sludge technology applicationsin the real-industrial wastewater has been initiated by previousresearchers. The applicability of aerobic granulation in treating awide variety of industrial wastewater under laboratory scale con-ditions like abattoir (Yilmaz et al., 2008), dairy (Wichern et al.,2008), livestock (Kishida et al., 2009), winery (López-Palau et al.,2009), chemical industrial wastewater (Liu et al., 2011), palm oilmill effluent (Abdullah et al., 2011), landfill leachate (Wei et al.,2012) as well as industrial effluents (Val del Río et al., 2012) werealso been investigated. To date, there is no reported study on theuse of aerobic granular sludge system for the treatment of waste-water from rubber industry.

Hence, aerobic granular sludge technology seems to be a rea-sonable option for studying rubber wastewater treatment. In thisstudy, the process of aerobic granulation in a Standard MalaysianRubber (SMR) wastewater treatment system were examined byfocusing on the feasibility of aerobic granular sludge in treating

N.H. Rosman et al. / Bioresource Technology 129 (2013) 620–623 621

rubber wastewater, as well as the investigation of organics andnitrogen removal in the granular system. This work could be usefulfor further understanding of the aerobic granulation mechanism aswell as for the development of granule-based systems in treatingindustrial wastewaters.

2. Methods

2.1. Experimental procedures and bioreactor set-up

Experiments were carried out in a cylindrical column bioreactor(internal diameter of 5 cm and total height of 35 cm) with aworking volume of 600 mL. During the start-up period, 300 mL ofsludge from a municipal sewage treatment plant was added intothe bioreactor system as inoculums, resulting in an initial mixed li-quor suspended solid (MLSS) of 5300 mg L�1 in the bioreactor. Aprogrammable logic controller (PLC) was operated to control theactuations of the pumps; influent pump, effluent pump, and aera-tor pump with the setting time for each phase in the sequentialbatch reactor (SBR) mode. The bioreactor was operated in SBRmode at a cycle of 3 h: 5 min of feeding without stirring, 155 minof aerobic reaction, 15 min of settling, 3 min of effluent with-drawal, and 2 min of idling. During the filling phase, wastewaterwas introduced through ports located at the bottom of the bioreac-tor. While fine air bubbles for aeration were supplied by means ofair bubble diffusers placed at the bottom at a volumetric flow rateof 0.12 m3 h�1 (1.70 cm s�1 superficial air flow velocity). The efflu-ent was withdrawn through the outlet ports positioned at mediumheight in the column bioreactor, resulting in volumetric exchangeratio (VER) of 50%. The sludge retention time was set by the dis-charge of suspended solids with the effluent. The bioreactor wasoperated at room temperature (27 ± 1 �C) and without oxygenand pH control.

2.2. SMR wastewater characteristics and seed sludge

The SMR wastewater was obtained from the Chip Hong RubberSdn. Bhd., Johor, Malaysia. It was collected from the rubber factoryplant once a week and stored in cold storage room at a tempera-ture of 4 �C. Prior feeding into the reactor, the pH was adjustedto a level within neutral condition in the range of 7.0 ± 0.5. Thecharacteristics of SMR wastewater used throughout the experi-ment were presented in Table 1, in comparison with SMR waste-water as described in the literature (Vijayaraghavan et al., 2008).The seed sludge was taken from an aeration tank of a local, muni-cipal wastewater treatment plant. It had a MLSS concentration of9.6 g L�1 and a sludge volume index (SVI) of 84.7 mL g�1.

2.3. Analytical methods

Measurements of the parameters such as mixed liquor sus-pended solid (MLSS), mixed liquor volatile suspended solid

Table 1Characteristics of raw Standard Malaysian Rubber process wastewater.

Parametersa SMR values (this study) SMR concentrationb

pH 7.35 7.5COD 1850 2960BOD5 890 1380Suspended solids (SS) 270 310Total nitrogen (TN) 248 n.s.c

Ammoniacal nitrogen (AN) 49 57

a All other parameters are in mg L�1 except pH.b Source: Vijayaraghavan et al. (2008).c n.s. = Not specified.

(MLVSS), COD, BOD5, NH3, total nitrogen (TN) and sludge volumeindex (SVI) were carried out according to Standard Methods forthe Examination of Water and Wastewater (APHA, 2005). The pHand DO were continuously monitored with the sensors insertedin the bioreactor and being recorded by a pH/DO meter (Orion4-Star Benchtop pH/DO Meter). The SVI procedure was carriedout according to the procedure described by de Kreuk et al.(2005). The morphological and structural observations of granularsludge were conducted periodically by using a stereo microscopeequipped with digital image analyzer (PAX-ITv6, ARC PAX-CAM).The microstructure compositions within the granule were ob-served with scanning electron microscope (FESEM-Zeiss Supra 35VPFESEM). For the pre-treatment procedure for SEM image, thegranules were left dried at room temperature prior to gold sputtercoating (Biorad Polaron Division SEM Coating System).

3. Results and discussion

3.1. Aerobic granular sludge formation and morphology observation

Under microscopic examination, the initial seed sludge showeda fluffy, irregular, loose-structure morphology and had relativeabundance of filamentous microorganisms. The sludge color grad-ually changed from dark brown to yellowish brown at the end ofthe experimental period. In the initial stage of granulation (firstweek), the loose flocs were easily broke up into small pieces ifplaced under vigorous shaking.

Over the next seven weeks, the flocs-like sludge gradually dis-appeared and was replaced by small granules with 0.7 mm in aver-age diameter. The flocs became denser under high shear forcewhich induces the biomass aggregates to secrete more exopolysac-charides (Dulekgurgen et al., 2008). The evolution of seed sludgefrom flocs to granules was achieved due to the interactions be-tween inter-particle bridging process among EPS, microbial cellsand ion (Sheng et al., 2010). EPS has the ability to increase the cellhydrophobicity and alter the surface charges on the microorgan-isms (Zhu et al., 2012), which in turn promote microbial cell adhe-sion and granulation.

Subsequently, the small granules changed to more regular inshape and gradually increased in size in the following weeks, whilemore flocculent sludge washed out from the bioreactor, resultingin the accumulation of the aerobic granules with high settlingvelocity. Finally, mature granules formed after ten weeks of inocu-lation, leading to a stable operation of the bioreactor. The compactmature granules were smooth with a solid surface, and the averagediameter is 1.5 mm. Towards the end of the experiment, granuleskept growing in a much lower speed up.

The outer morphology and inner structures of the aerobic gran-ules was further examined using SEM. The SEM observationshowed that the granules had a round shape with a clear boundaryoutline and compact structure. A closer examination revealed thatthe granules consisted of a wide variety of non-filamentous coccal-shaped bacteria which were tightly linked one another to form thecompact structure. Meanwhile, the presence of multiple cavitiescould also be observed between the clumped bacteria. These cavi-ties were believed to enhance the transport of substrate and oxy-gen into the inner cores of the granules and metabolic productstransfer in and out of the granules (Gao et al., 2011).

3.2. Biomass profile and settling properties of granules

Fig. 1 shows the MLSS, MLVSS and SVI variation in the SBR sys-tem from the start-up until the end of the study. At the beginningof the experiment, most of the sludge was washed-out from thebioreactor causing a rapid decreased in the biomass concentration

Fig. 1. Variation of biomass concentrations and sludge volume index in SBR for90 days. (d) MLSS concentration; (s) MLVSS concentration; (�) SVI.

Fig. 2. Profile of removal performances in the SBR system within 90 days for (a)COD, (b) ammonia and (c) TN. (N) Influent concentration; (j) effluent concentra-tion; (s) percentage removal.

622 N.H. Rosman et al. / Bioresource Technology 129 (2013) 620–623

and increased the effluent solids concentration. It was found thatthe MLSS reduced from 5300 mg L�1 to 2200 mg L�1 in the first15 days, presumably due to the short settling time applied in thecycle system. During the days between 15 and 30, both MLSSand MLVSS concentration kept increasing but with occasionalsporadic decrease. A possible explanation could be that somemicroorganisms in the bioreactor were adapting themselves to de-grade non-active microorganisms and the seed microorganismswere to adapt to the rubber wastewater. As the small granules be-gan to appear in the bioreactor on the day-39, the concentration ofthe biomass is also improving. Thereafter, both MLSS and MLVSSconcentrations increased steadily and achieved steady state atabout 8200 mg L�1 and 6500 mg L�1, respectively during days 75and 90. A similar trend was also observed for the MLVSS contentwhich ranging from initial concentration of 2800 mg L�1 to6500 mg L�1 at the end of the experiment. The MLVSS to MLSS ra-tio is about 0.79 and a stable condition of biomass concentrationindicates a good accumulation of biomass in the bioreactor.

With the increase of biomass concentration, the settle ability ofthe developed aerobic granules with rubber wastewater had im-proved, resulted in granule with good settling properties in termsof SVI. The SVI value has improved from 84.7 mL g�1 at the startingof the experiment to 22.3 mL g�1 at the end of the experiment. Inthe first two weeks, the SVI slightly increased due to severe bio-mass washout in the effluent which related to the decreasing ofMLSS content in the bioreactor. Then, the SVI gradually decreasedwith the formation of fast settling and denser biomass towards theend of the experiment. Meanwhile, the average settling velocity ofthe mature granular sludge developed in this study was 33 m h�1

which is in agreement with the result reported by Su andYu (2005) with settling velocity values in the range of36.6 ± 8.8 m h�1 in cultivating aerobic granules fed with soybean-processing wastewater. The higher settling velocity and lowerSVI value of the mature granular sludge as compared to the con-ventional activated sludge is an indication of good settling charac-teristics of the granules. The well settling of the granules, favorablegood separation of biomass from the treated effluent, thus, leaves aclear supernatant in the bioreactor system.

3.3. COD and nitrogen removal efficiencies of granules

The removal efficiencies of organic and nutrients in the SBR sys-tem with respect to COD, ammonia and total nitrogen from begin-ning until the end of granules development period is illustrated inFig. 2. At the initial stage of the operation, the concentration ofeffluent COD decreased from 568 mg L�1 to 493 mg L�1, presum-ably due to the adapting process of the sludge with rubber waste-water. The effluent quality in the bioreactor was unsatisfactory

since the percentage removal of COD was around 70%. During thefirst month, the COD removal efficiency was fluctuating, but the re-moval became stable for the remaining period. As the evolution offlocculent sludge into granular sludge gradually takes place in thebioreactor system, the degradation ability for COD removal effi-ciency had improved in between 85.6% and 89.9%, which is compa-rable to previous result obtained by Abdullah et al. (2011) intreating palm oil mill effluent using aerobic granular sludge. Even-tually, the percentage of COD removal efficiency increased up to96.5% at the end of the experiment indicates the high biologicalactivity occurred during microbial aerobic degradation process ofrubber wastewater in the bioreactor.

The percentage removal of ammonia was 68.8% at the beginningof the bioreactor start-up and then increased when granules wereformed. The effluent ammonia concentrations show a significantdecrease from 18.5 mg L� to 8.4 mg L�1, afterwards. The removalefficiencies for ammonia and total nitrogen fluctuated for aboutone month and remained stable for the subsequent period. Upon

N.H. Rosman et al. / Bioresource Technology 129 (2013) 620–623 623

aerobic granular sludge formation, the ammonia concentrations inthe effluent shows a significant better quality and maintained be-low 10 mg L�1 indicates a good ammonia removal efficiency. After90 days of operation, the removal efficiency increased to 94.7% forammonia and 89.4% for total nitrogen. Nitrifying bacteria popula-tion within the aerobic granules became predominant after thebiodegradation of organics which contribute to the occurrence ofnitrification. As reported by Belmonte et al. (2009), an improve-ment of nitrification process could be achieved by promoting theformation of granules that enhance the retention of large amountsof nitrifying bacteria in the bioreactor system which allows a high-er ammonia removal of above than 94%. The sufficient amount ofoxygen supplied in the bioreactor system enabled a good oxidationfor ammonia and about more than 90% of ammonia being removedin the aerobic reaction phase which demonstrates a stable and aneffective nitrification process.

4. Conclusions

Compact and stable aerobic granular sludge with an excellentsettling ability and an average diameter size of 1.5 mm were suc-cessfully cultivated in an SBR system fed with rubber wastewater.After 90 days of operation, a good COD removal rate of 96.5% wasobserved at a load of 3.7 kg COD m�3 d�1. In addition, ammonia re-moval efficiency of 94.7% and a total nitrogen removal efficiency of89.4% were achieved in the single bioreactor system for treatingrubber wastewater. The study presented herein suggested feasibil-ity of the developed aerobic granules for the treatment of waste-water from rubber factory industry.

Acknowledgements

The authors wish to thank the Ministry of Higher Education(MOHE) and Universiti Teknologi Malaysia for the financial sup-ports of this research (UTM Research University Grant).

References

Abdullah, N., Ujang, Z., Yahya, A., 2011. Aerobic granular sludge formation for highstrength agro-based wastewater treatment. Bioresour. Technol. 102, 6778–6781.

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

Belmonte, M., Vázquez-Padín, J.R., Figueroa, M., Franco, A., Mosquera-Corral, A.,Campos, J.L., Méndez, R., 2009. Characteristics of nitrifying granules developedin an air pulsing SBR. Process Biochem. 44, 602–606.

de Kreuk, M.K., Pronk, M., van Loosdrecht, M.C.M., 2005. Formation of aerobicgranules and conversion processes in an aerobic granular sludge reactor atmoderate and low temperatures. Water Res. 39, 4476–4484.

Dulekgurgen, E., Artan, N., Orhon, D., Wilderer, P.A., 2008. How does shear affectaggregation in granular sludge sequencing batch reactors? Relations betweenshear, hydrophobicity, and extracellular polymeric substances. Water Sci.Technol. 58, 267–276.

Gao, D., Liu, L., Liang, H., Wu, W.M., 2011. Aerobic granular sludge: characterization,mechanism of granulation and application to wastewater treatment. Crit. Rev.Biotechnol. 31 (2), 137–152.

Kishida, N., Tsuneda, S., Kim, J.H., Sudo, R., 2009. Simultaneous nitrogen andphosphorus removal from high-strength industrial wastewater using aerobicgranular sludge. J. Environ. Eng. ASCE 135 (3), 153–158.

Liu, L., Sheng, G.P., Li, W.W., Tong, Z.H., Zeng, R.J., Liu, J.X., Xie, J., Peng, S.C., Yu, H.Q.,2011. Cultivation of aerobic granular sludge with a mixed wastewater rich intoxic organics. Biochem. Eng. J. 57, 7–12.

López-Palau, S., Dosta, J., Mata-Álvarez, J., 2009. Start-up of an aerobic granularsequencing batch reactor for the treatment of winery wastewater. Water Sci.Technol. 60, 1049–1054.

Mohammadi, M., Man, H.C., Hassan, M.A., Yee, P.L., 2010. Treatment of wastewaterfrom rubber industry in Malaysia-review. Afr. J. Biotechnol. 9 (38), 6233–6243.

Sheng, G.P., Yu, H.Q., Li, X.Y., 2010. Extracellular polymeric substances (EPS) ofmicrobial aggregates in biological wastewater treatment systems: a review.Biotechnol. Adv. 28, 882–894.

Su, K.Z., Yu, H.Q., 2005. Formation and characterization of aerobic granules in asequencing batch reactor treating soybean-processing wastewater. Environ. Sci.Technol. 39 (8), 2818–2827.

Sulaiman, N.M.N., Ibrahim, S., Abdullah, S.L., 2010. Membrane bioreactor for thetreatment of natural rubber wastewater. Int. J. Environ. Eng. 2 (1–3), 92–109.

Val del Río, A., Figueroa, M., Arrojo, B., Mosquera-Corral, A., Campos, J.L., García-Torriello, G., Méndez, R., 2012. Aerobic granular SBR systems applied to thetreatment of industrial effluents. J. Environ. Manage. 95, S88–S92.

Vijayaraghavan, K., Ahmad, D., Ahmad Yazid, A.Y., 2008. Electrolytic treatment ofStandard Malaysian Rubber process wastewater. J. Hazard. Mater. 150, 351–356.

Wei, Y.J., Ji, M., Li, R.Y., Qin, F.F., 2012. Organic and nitrogen removal from landfillleachate in aerobic granular sludge sequencing batch reactors. Waste Manage.(Oxford) 32, 448–455.

Wichern, M., Lübken, M., Horn, H., 2008. Optimizing sequencing batch reactor (SBR)reactor operation for treatment of dairy wastewater with aerobic granularsludge. Water Sci. Technol. 58, 1199–1206.

Yilmaz, G., Lemaire, R., Keller, J., Yuan, Z.G., 2008. Simultaneous nitrification,denitrification, and phosphorus removal from nutrient-rich industrialwastewater using granular sludge. Biotechnol. Bioeng. 100 (3), 529–541.

Zhu, L., Lv, M.L., Dai, X., Yu, Y.W., Qi, H.Y., Xu, X.Y., 2012. Role and significance ofextracellular polymeric substances on the property of aerobic granule.Bioresour. Technol. 107, 46–54.