ENVIRONMENTAL MICROBIOLOGY

Kinetic Modelling and Characterization of MicrobialCommunity Present in a Full-Scale UASB Reactor TreatingBrewery Effluent

Abimbola M. Enitan & Sheena Kumari &Feroz M. Swalaha & J. Adeyemo & Nishani Ramdhani &Faizal Bux

Received: 9 August 2013 /Accepted: 15 November 2013 /Published online: 12 December 2013# Springer Science+Business Media New York 2013

Abstract The performance of a full-scale upflow anaerobicsludge blanket (UASB) reactor treating brewery wastewaterwas investigated by microbial analysis and kinetic modelling.The microbial community present in the granular sludge wasdetected using fluorescent in situ hybridization (FISH) andfurther confirmed using polymerase chain reaction. A groupof 16S rRNA based fluorescent probes and primers targetingArchaea and Eubacteria were selected for microbial analysis.FISH results indicated the presence and dominance of a sig-nificant amount of Eubacteria and diverse group of methano-genic Archaea belonging to the order Methanococcales ,Methanobacteriales , and Methanomicrobiales within in theUASB reactor. The influent brewery wastewater had a rela-tively high amount of volatile fatty acids chemical oxygendemand (COD), 2005 mg/l and the final COD concentrationof the reactor was 457 mg/l. The biogas analysis showed 60–69 % of methane, confirming the presence and activities ofmethanogens within the reactor. Biokinetics of the degradableorganic substrate present in the brewery wastewater was fur-ther explored using Stover and Kincannon kinetic model, with

the aim of predicting the final effluent quality. The maximumutilization rate constantUmax and the saturation constant (KB)in the model were estimated as 18.51 and 13.64 g/l/day,respectively. The model showed an excellent fit between thepredicted and the observed effluent COD concentrations. Ap-plicability of this model to predict the effluent quality of theUASB reactor treating brewery wastewater was evident fromthe regression analysis (R2=0.957) which could be used foroptimizing the reactor performance.

Introduction

Brewery industries produce millions of litres of beer eachyear which results in the release of large amounts ofwastewater with high organic content. The reduction ofthis high-strength wastewater is mandatory to protect theenvironment as well as to reduce the cost of penalties thatmight be incurred due to unlawful effluent discharge. Recent-ly, the use of anaerobic treatment technology such as upflowanaerobic sludge blanket (UASB) reactors has become apopular biological treatment method for both industrial anddomestic waste treatment [1].

The anaerobic breakdown of the complex organic com-pounds involve the action of several groups of microorgan-isms which results in a variety of intermediates includingbiogas such as hydrogen, methane, and carbon dioxide[2–4]. The microbial species involved in the conversion oforganic material in anaerobic digesters are grouped based ontheir biochemical activities. The group includes hydrolytic,acidogenic, acetogenic, and methanogenic organisms [5].These organisms grow in a syntrophic manner when the

A. M. Enitan (*) : S. Kumari (*) :N. Ramdhani : F. BuxInstitute forWater andWastewater Technology, Durban University ofTechnology, P.O. Box 1334, Durban 4000, South Africae-mail: enitanabimbola@gmail.come-mail: sheenas@dut.ac.za

A. M. Enitan : F. M. SwalahaDepartment of Biotechnology and Food Technology, DurbanUniversity of Technology, P.O. Box 1334, Durban 4000, SouthAfrica

J. AdeyemoDepartment of Civil Engineering and Surveying, Durban Universityof Technology, P.O. Box 1334, Durban 4000, South Africa

Microb Ecol (2014) 67:358–368DOI 10.1007/s00248-013-0333-x

digester is operated under optimum reaction conditions [6],however, acetogenesis and methanogenesis are the rate limit-ing steps during the anaerobic digestion process [7, 8]. Con-sequently, to regulate the UASB reactor performance effi-ciently, it is vital to understand the microbial community andtheir activity at different reactor operational parameters [9]. Inaddition, a thorough understanding of the methanogenic ac-tivity within the UASB granules would enhance biogas(methane) production and yield [10].

Studies have shown that the microbial community in theUASB reactor responds to any sudden change in the environ-mental conditions, thus leading to a shift in the type ofmicrobial species found in the reactor, their population sizeand activities. Therefore, an in-depth understanding of themicrobial consortium and the associated activities are essentialfor effective reactor operation. However, using conventionalmethods, it is difficult to assess the diversity, colonization andtopological distribution of these microorganisms due to thestructural complexity of the granular sludge [11]. Recently,molecular techniques such as denaturing gradient gel electro-phoresis (DGGE), fluorescence in situ hybridization (FISH)and pyrosequencing have been successfully adopted to studythese complex microbial populations [12–16]. Furthermore,development of a suitable mathematical model, which ade-quately describes the overall process performance in the bio-reactor is an important tool for process control strategiesresulting in better effluent quality and biogas production[17]. Mass balances, kinetic, and stoichiometric models aresome of the methods that are being employed in describing theprinciple of different anaerobic digesters [18, 19]. Simple andmore sophisticated models such as the Monod, Chen andHashimoto, Contois, Michaelis-Menten, Haldane, Grausecond-order and anaerobic digestion model 1 (ADM1) havealso been developed to improve the reactor performance [20,21].

Kinetic analysis is an acceptable method to describe andpredict the performance of any biological treatment unit [22,23]. It can be applied in the optimization and control ofanaerobic wastewater treatment processes, to determine therelationship between fundamental parameters needed for an-aerobic reactions [19, 24]. Among several kinetic modelsdeveloped for organic substance removal in the UASB reac-tor, the Stover–Kincannon model has been well documented[19, 24]. However, the modified form of this model is one ofthe most widely adopted methods for the determination ofkinetic constants and has been successfully applied for anaer-obic treatment of distillery wastewater [24], poultry manurewastewater [19], poultry slaughterhouse waste [23], and mu-nicipality wastewater [25].

Therefore, monitoring the environmental conditions andidentification of the functional microbial population as wellas analysing the kinetic process of UASB reactors is crucialfor reactor design, maintenance and its efficient operation to

increase methane production as a source of renewable energyand for better effluent quality. Hence, this study focused onunderstanding the microbial composition of a UASB reactortreating brewery wastewater using advanced molecular tech-niques and to understand the biokinetics of the degradableorganic substrates present in brewery wastewater using akinetic model to predict the effluent quality and biogasproduction.

Materials and Methods

Reactor Description and Sample Collection

A full-scale industrial UASB reactor treating brewery waste-water in Prospecton Brewery, South Africa was selected forthe current study. This reactor was commissioned in 1985 andpatented as biothane process. The biomass for the seeding wasimported from Holland due to the good quality which wasimperative for the initial start up of the reactor. The reactorwas constructed of concrete with a series of settlers and baffleplates arranged at the bottom for even distribution with a pre-conditioning tank. Without the pre-conditioning tank, theoperating capacity was 1,480 m3 [25], and in addition to thepre-conditioning tanks and screens the total capacity increasedto 1,700 m3 [26]. The wastewater from the conditioning tankenters from the bottom of the UASB reactor through spargepiping systems and a pulsed system was used to regulate theflow rate through each set of pipes for even distribution ofwastewater across the biomass blanket. Intermittent feedingwith the brewery wastewater into the reactor was maintainedbetween 37±2 °C at a pH range of 6.5–7.2. Retention timevaried with influent flow rate between 8 and 12 h for bacteriato make use of the organic content of the wastewater. Thebiogas produced in the reactor was separated from the effluentand the biomass in three-phase separators at the top of thereactor.

Prior to sample collection, the sampling valves wereopened for 5 min to flush out the sampling tubes and valves.Thereafter, granular sludge samples were collected for micro-bial analysis in sterile glass bottles and flushed with nitrogengas and sealed immediately to maintain anaerobic conditionsduring transportation to the laboratory. A series of pre-screened and acidified wastewater samples, before enteringthe reactor (digester in) and effluent leaving the reactor (di-gester out), were also collected in one litre sterile glass bottles.Both granular sludge samples and wastewater collected weretransported to the laboratory at 4 °C for analysis. Physico-chemical analyses were done within 48 h of collection withnecessary preservation techniques adapted from StandardMethods for Examination of Water and Wastewater [34].The biogas was collected in a Tedlar bag (Sigma-Aldrich)for analysis.

Kinetics and Characterization of Microbes of an UASB Reactor 359

Analytical Procedure

Wastewater Characterization

Brewery wastewater samples were analysed for parameterssuch as alkalinity, biological oxygen demand (BOD5),chemical oxygen demand (COD), volatile suspended solids(VSS), total suspended solids (TSS), sulphates, nitrites,nitrates, ammonia and ortho-phosphate using standardmethods [26]. Total dissolved solids (TDS), conductivity(microsiemens per centimeter) and oxidation–reduction po-tential (O/R potential) were measured using a YSI meter(YSI 556MPS, Yellow Springs, USA). The temperatureand pH were measured using a pH meter (Beckman pH211 Microprocessor, USA). Protein concentration was an-alyzed using a UV–VIS spectrophotometer (Merck,Spectroquants Pharo 300, Germany) [27]. Volatile fattyacids (VFA) (acetic, propionic, isobutyric, butyric, valericand isovaleric acids) were quantified using HPLC (ModelLC-20AT, Shimadzu, Japan) equipped with a UV detector(SPD-20A) and analysed using a Metrosep organic acidcolumn (250×7.8 mm) at a flow rate of 0.6 ml/min and aninjection volume of 20 μl at 210 nm. The mobile phaseconsisted of a 0.5 mM H2SO4 solution. The composition ofmethane in the biogas produced was analyzed using GasChromatography (Shimadzu GC-2014, Japan) equipped witha thermal conductivity detector (TCD). A Porapak Q (1.8 m×2.10 mm) column was used at 40 °C with injector and detec-tors were set at 100 °C. Helium served as the carrier gas at aflow of 20 ml/min.

Fluorescent In Situ Hybridization

Fluorescent in situ hybridizations were carried out accord-ing to the protocol described by Amann et al. [28] withminor modifications. Sludge granules were fixed in 4 %paraformaldehyde (Gram negative) and in PBS–ethanol(Gram positive) [28]. Fixed samples were then sonicated

at 2 W for 5 min using an Ultrasonic Liquid Processor(Misonix XL-2000 Series). Thereafter, granules were treat-ed with 10 μl of lysozyme (10 mg/ml) at 37 °C for30 min; then with proteinase K (1 mg/ml) at 50 °C for45 min. Samples were diluted further by the addition of500 μl of sterile water for even dispersion and quantifica-tion with the group specific and Archaea domain probes(Table 1). For the hybridisation, the treated samples (5–10 μl) were fixed on poly-L -lysine coated slides andallowed to air dry at room temperature for 10 min, thendehydrated by a series of ethanol washes (50, 80 and100 %). The oligonucleotide probes were labelled withrhodamine (FAM) and tetramethylrhodomine-5-isothiocyanate (TAMRA) dye at the 5′-end, respectively(Table 1). The hybridisation and wash buffers were pre-pared according to the formamide stringency (Table 1).Samples were hybridised by the addition of 9 μL ofhybridisation solution (10 % SDS, 1 M Tris/HCl (pH 8),5 M NaCl, and formamide concentrations (Table 1); togeth-er with 1 μl of oligonucleotide probe, incubated in thehybridisation oven at 46 °C, overnight. After hybridisation,the slides were washed with pre-warmed washing buffer(1 M Tris/HCl, 10 % SDS, 0.5 M EDTA and 5 M NaCl;Table 1) for 1 h at 48 °C; subsequently rinsed with distilledwater and then air-dried. The slides were counter-stainedwith 4′-6-diamino-2-phenylindole (DAPI) for 10 min atroom temperature. Slides were rinsed in pre-warmed dis-tilled water and air-dried in the dark. The samples werethen mounted with an anti-fading solution (Vectashield,Vector Laboratories, Inc. Burlingame). The hybridisedslides were viewed using a Zeiss Axio-Lab HB050/ACmicroscope (Carl Zeiss, Jena, Germany) equipped with anHBO 50 W Hg vapour lamp, with appropriate filter sets,specific for TAMRA and FAM using ×100 PlanApochromat Objective. Images were captured using ZeissAxioCam MRC camera (Carl Zeiss, Gottingen, Germany)and analysed using a Zeiss Axio vision Release 4.8 imag-ing system.

Table 1 16S rRNAoligonucleotide probes with the corresponding formamide stringency and NaCl concentrations used in this study

Target group Oligonucleotides Fluorescentlabel

Formamide concentration (%)/NaCl (μl) References

Probe name Probe sequence (5′ – 3′)

Archaea ARC915 GTG CTC CCC CGC CAATTC CT FAMa 30/1020 [55]

Methanosarcina MS821 CGC CAT GCC TGA CAC CTA GCG AGC FAMa 40/460 [37]

Methanosaeta MX825 TCGCACCGTGGCCGACACCTAGC FAMa 50/180 [37]

Eubacteria EUB338 GCTGCCTCCCGTAGGAGT TAMRAb 30/1020 [56]

EUB338 II GCAGCCACCCGTAGGTGT TAMRAb 30/1020 [57]

EUB338 III GCTGCCACCCGTAGGTGT TAMRAb 30/1020 [57]

a Rhodamineb Tetramethylrhodamine-5-isothiocyanate

360 A.M. Enitan et al.

Genomic DNA Extraction and Polymerase Chain Reaction

The direct isolation of total genomic DNA from granularsludge samples was carried out using phenol extraction meth-od [29]. PCR reactions were performed to detect the methylcoenzyme-M reductase (mcrA) gene using the methanogenmcrA forward primer 5′-GGTGGTGTMGGATTCACACARTAYGCWACAGC-3′ and reverse primer 5′-TTCATTGCRTAGTTWGGRTAGTT-3′, respectively [30]. The PCR mix-ture contained 25 μl reaction volume of 0.3 μl of Taq DNApolymerase (5 U/ml), 2.5 μl of PCR reaction buffer, 1 μl ofeach of the primers (10 mM), 0.5 μl of dNTPs (10 mM), and

2 μl of the extracted DNA. The modified PCR amplificationconditions of Luton et al. [30] was used as follows: initialdenaturation was performed at 94 °C for 5 min; followed by40 cycles of denaturation at 92 °C for 1 min; primer annealingat 52 °C for 1 min; and elongation at 72 °C for 1 min and afinal extension was performed at 72 °C for 5 min. The PCRamplification was carried out by an automatic thermal cyclerVeriti (Applied Biosystems). The successful PCR ampliconswere further ligated into the pTZ57R/T cloning vector of theTA cloning kit (Invitrogen) and transformed into DH5α E.Coli cells using an insTAclone PCR cloning kit (Fermentas)following the manufacturer’s instructions. Random positive

Table 2 Characterization of granular sludge used for molecular analysis

Parameter TCOD SCOD PCOD TSS VSS TS VS PO4 NO2 NH4 pH Temperature (°C)

Concentration (mg/l) 1,700 1,220.58 479.15 70.54 62.27 83.42 70.38 70.59 0.12 1.50 6.78 28

Fig. 1 a Images of granuleshybridised by highly rhodaminelabelled archaeal-domainoligonucleotide probes(ARC915) showing diversespecies of methanogens (green).b Images of granules hybridisedby archaeal-domainoligonucleotide probes(ARC915) showing diversespecies of methanogens in DAPIunder ×100 Plan ApochromatObjectives (blue). c Granularsludge FISH labelled withtetramethylrhodomine-5-isothiocyanate using the universalprobes for Eubacteria (Eub338).d The MX825 probe-labelledsample to confirmed theacetoclasticMethanosaeta group,and e the corresponding DAPIstained cells

Kinetics and Characterization of Microbes of an UASB Reactor 361

clones were selected for further sequence analysis (InqabaLaboratories, South Africa). The obtained mcrA gene se-quences were manually edited and similarity search for theDNA sequences were carried out using the Basic Local Align-ment Search Tool (BLAST) program to search in the NationalCentre for Biotechnology Information (NCBI) sequence data-base (http://www.ncbi.nlm.nih.gov/BLAST). The nucleotidessequences obtained from the GenBank were converted toamino acid sequences and then aligned in CLUSTAL X.The aligned mcrA amino acid gene sequences were editedusing BioEdit and exported to MEGA5.10. Evolutionary

analyses were conducted in MEGA5.10 [31]. Thephylogenetic tree was constructed from the alignments andbootstrap analysis was performed using 1000 replicates byneighbour-joining method. The nucleotide sequences weresubmitted to National Centre for Biotechnology Information(Gen bank accession number: KF715644–KF715648).

Kinetic Analysis Using the Stover–Kincannon Model

According to Stover–Kincannon model [32], the organic sub-strate utilization rate in a UASB reactor process can be

Fig. 2 Phylogenetic tree for methanogenic archaea obtained from granular sludge of UASB reactor treating brewery wastewater, using methylcoenzyme-M reductase (mcrA) gene primer set. The evolutionary history was inferred using the neighbour-joining method [59]

362 A.M. Enitan et al.

expressed as a function of the organic loading rate. Thesubstrate consumption rate can be expressed as [18, 19, 25]:

dS

dt¼ Q

V rSi−Seð Þ ð1Þ

The original Stover–Kincannon model is described inEq. (1) as;

dS

dt¼ Q Si−Seð Þ

V r¼

UmaxQSiV r

� �

KB þ QSiV r

� � ð2Þ

Where dS /dt is the substrate removal rate (grams COD perliter per day) in the UASB reactor, S is the reactor substrateconcentration (grams per liter), Umax is the maximum utiliza-tion rate constant (grams per liter per day), Vr is the workingvolume of reactor (liter), KB is the saturation constant (gramsper liter per day), Q is the flow rate (liters per day), S i and Se

are the influent and effluent substrate concentrations (gramsper liter), respectively.

Combining Eqs. (1) and (2) gives the modified Stover–Kincannon model for a UASB reactor at steady state.

dS

dt

� �−1

¼ Vr

Q Si−Seð Þ ¼KBVr

Umax QSið Þ þ1

Umax

Y ¼ λX þ λ0; YVr

Q Si−Seð Þ λ ¼ KB

Umax;X ¼ Vr

Q Sið Þ;λ0 ¼ 1

Umax

ð3ÞConsidering the mass balance of substrate present in waste-

water that flows into the reactor and out of the reactor plus thetotal amount of substrate degraded, at a specific flow rate, controlvolume and time, then the mass balance can be written as;

QSi ¼ QSi þ VrdS

dt

� �ð4Þ

Substituting dS /dt from the Eq. (3) into the Eq. (4) and byrearranging the expression, it will give Eqs. (5) and (6).

Se ¼ Si−UmaxSi

KB þ QSi=V rð Þ ð5Þ

V r ¼ QSiUmaxSiSi−Se

� �−KB

ð6Þ

At a given influent concentration, organic loading rate (QSi/Vr) and known volume of anaerobic reactor, Eq. (5) can be usedto estimate the concentration of substrate present in the reactor

effluent when KB and Umax values are obtained. Equation (6)can be used to determine the required volume of anaerobicreactor needed to reduce effluent substrate concentration in orderto meet the discharge standard. Equation (3) can be used todetermine theKB andUmax of the reactor. The inverse of loading

rate VrQ Si−Seð Þ

� �is plotted against the total loading rate of the

reactor VrQ Sið Þ . A straight line result is obtained; the slope and

intercept of the line are KBUmax

and 1Umax

, respectively.

Statistical Analysis

The full sets of experimental data were used for the statisticalanalysis. Evaluation using ANOVA was used to test the sig-nificant effects of the measured and predicted results at analpha level of 0.05. GraphPad Prism v.5, software packagewas used for the statistical analysis and graphs.

Results and Discussion

Detection of Methanogenic Bacteria from Granular SludgeUsing FISH and PCR

The physico-chemical characterization of granular sludge col-lected from UASB reactor treating brewery wastewater usedfor microbial analysis in this study is shown in Table 2.Identification of microbial consortium present in the granularsludge of brewery wastewater was carried out by FISH using

Table 3 Biochemical properties of pre-conditioned brewery wastewaterentering the UASB before treatment

Parameters Average concentration valuesa

Temperature (°C) 29.21

pH 6.87

COD 2,005.73

TSS 2,449.46

TS 4,520.00

TDS 1,792.80

Protein content 134.40

Phosphates 21.25

TON 0.52

NH4 21.64

NO2 2.30

NO3 0.07

Sulphate 178.25

ORP (mv) −144.78Conductivity (mS/cm) 2.18

Alkalinity (mg CaCO3/l) 2,880.52

a All parameters are in milligrams per liter except otherwise stated

Kinetics and Characterization of Microbes of an UASB Reactor 363

Eubacteria and Archaea specific probes (Table 1). The resultsindicated that the most common acetoclastic methanogenspresent in the UASB reactor belongs to the orderMethanobacteriales , Methanococcales and Methano-microbiales . They are short, straight to slightly curved rodsor irregular cocci and packet-like in nature. These groups ofmethanogens are previously being reported from the anaero-bic reactors which showed more than 70 % methane produc-tion [33]. In situ hybridization analysis of the samples stainedwith ARC 915 and EUB 388 mix probes revealed the domi-nance of both rod and coccoid-shaped methanogens in thereactor (Fig. 1a, b and c). The thick cell wall with long andshort curved rods, cocci and irregular cocci packet shapesindicated the presence of diverse group of methanogenicArchaea belonging to the order Methanobacteriales ,Methanococcales and Methanomicrobiales . Detection of rodand cocci packet shapes by ARC 915 probe shows thatMethanosaeta and Methanosarcinal -like species are alsopresent in the UASB reactor. The presence of cocci with thickcell wall and packet-like shape, typical to the genusMethanosarcina was further confirmed by MS821 probe.

These results are in agreement with the previous findingswhere the genus Methanosarcina were detected in granularsludge samples [34–36]. Further, the positive hybridization ofMX825 probe confirmed the presence and dominance ofacetoclastic Methanosaeta group in the samples (Fig. 1d ande), which is distinguished by their typical rod-shape and is inline with the previous reports [37–40].

The FISH results were further confirmed using PCR andphylogenetic analysis using methanogenic specific primers(mcrA) [41]. The successful PCR products after sequencingand phylogenetic analysis showed 96 to 100 % similarity tomethanogen ic Archaea be long ing to the o rde rMethanobacteriales and Methanomicrobiales (Fig. 2). Simi-lar results were previously reported from the UASB reactorgranules treating brewery wastewater [14] and also from otheranaerobic reactors producing biogas [35, 41, 42]. The mcrAsequences clustered around the Methanobacteriales such asMethanobacterium beijingense strain, Methanobacteriumarrhuesense and Methanothermobacter crinale showing96 % similarity [43–45]. This further confirms the dominanceof hydrogenotrophic Methanomicrobiales within the UASB

0 1 2 3 4 5 60

20

40

60

0

1000

2000

3000

4000

5000

% TSS removal efficiencyTSS digester in TSS digester out

Duration (Months)

TS

S r

emo

val e

ffic

ien

cy (

%) T

SS

con

centratio

n (m

g/l)

Fig. 4 Relationship betweenconcentrations of total suspendedsolid in influent, effluent andremoval efficiency of theanaerobic treatment of brewerywastewater

40 50 60 70 80 90 1000

1000

2000

3000

4000COD in COD out

% COD removal efficiency

CO

D c

on

cen

trat

ion

(m

g/l)

Fig. 3 Average values of influent(S i) and effluent CODconcentration (Se) with thecorresponding COD removalefficiency

364 A.M. Enitan et al.

reactor granules. However, in this study, the amplification ofthe mcrA primer sets using PCR did not detect theMethanosarcina andMethanosaeta spp. in the granular sam-ples as reported by Luton et al. [30]. Most clones belonged tothe order Methanomicrobiales and few clones wereMethanosarcina spp. while none was reported forMethanosaeta spp. [30]. Similar observations were also madeby Castro et al. [41] and Smith et al. [46]. Therefore, acombination of both PCR and fluorescence in situ hybridiza-tion (RNA-based methods) techniques could demonstrate abetter understanding of the microbial consortium present inthe reactor.

UASB Reactor Performance and Biogas Production

The performance of the UASB reactor to treat the brewerywastewater displays the ability to reduce the COD concentra-tion to a considerable level. The characteristics of the influentbrewery wastewater are shown in Table 3. The average influ-ent COD concentration was 2,005.73 mg/l at 28 °C (Table 3)with a COD removal efficiency of 78.97 %. The average datafor influent (S i) and effluent substrate concentration (Se) withthe corresponding COD removal rate during the samplingperiod is represented in Fig. 3. The average effluent substrateconcentration (S e) from the UASB reactor was lower(457.25 mg/l) than the influent substrate concentration (S i).This might be due to the low level of total solids introducedinto the reactor which helped the reactor performance. Thetotal TSS removal in the UASB reactor with an inlet and outletTSS concentration of the brewery wastewater is shown in

Fig. 4. The correlation analysis between inlet and outlet TSSconcentration showed a significant correlation between inletTSS and outlet TSS with high R 2 value of 0.9929 (P =0.0003). The ANOVA analysis results showed a significantdifference between COD and TSS removal efficiency duringtreatment (P =0.024). The efficiency of COD removal and themethanogenic activity were further shown in the compositionof biogas generated from the UASB reactor with methanecontent of 60–69 % (Table 4). There was a significant differ-ence between the percentage COD removal and biogas yieldusing ANOVA with high R2 value (0.9702). The results ofcorrelation analysis of the major content of biogas (carbondioxide and methane) shows that there was a significantcorrelation between the percentage of carbon dioxide andmethane produced during treatment of brewery wastewater(P <0.05, R2=0.954). The ANOVA results showed that meth-ane yield depended on the substrate present in the wastewaterin terms of COD removal efficiency. In addition, the microbialcharacterization and biogas production results further con-firmed the presence of methanogens in the UASB sludge.

The influent characterization confirmed the presence ofsignificant amount of volatile fatty acids in the brewery waste-water which could serve as substrate for the methanogens toproduce biogas [47]. The volatile fatty acids such as acetate(538.3 mg/l), propionic acid (237.5 mg/l), butyric acid(50.06 mg/l) and valeric acid of 16.54 mg/l were detected inthe influent wastewater with no detection of these acids in theeffluent. This shows that the methanogens metabolized theVFA present in the brewery wastewater to produce methane.Among all the methanogens detected using FISH,Methanosaeta sp. and Methanosarcina sp. were reported topossess the ability to metabolize acetate [48]. The abundanceofMethanosarcina sp. at high acetate level was in agreementwith previous studies [40, 47, 49]. Delbès et al. [50] reportedthat species closely related to the family Methanobacterialesand Methanobacterium formicicum were found dominant inan anaerobic bioreactor during acetate accumulation. However,the current study has shown a reduction in methanogenicactivities when there was a high nitrogen and ammonium

Fig. 5 Effect of organic loadingrate on COD removal rate usingthe modified Stover–Kincannonmodel to determinate the kineticconstants

Table 4 Average com-positions of biogas pro-duced in this study

Biogas composition Values (%)

Methane, CH4 65.9

Carbon dioxide, CO2 30.7

Nitrogen, N2 3.4

Hydrogen sulphide, H2S Not detected

Hydrogen, H2 Not detected

Kinetics and Characterization of Microbes of an UASB Reactor 365

nitrogen content in the effluent (data not shown). This could bethe result of unfavorable conditions in the reactor leading to areduction or inhibition of methanogenic growth in the reactor.In addition, the FISH analysis of these samples have also showna poor fluorescent signal during hybridization which could beattributed to a high protein content of the granules due to thelow methanogenic activities [51].

Kinetic Modelling and Model Validation

Kinetic studies are critical for the design and operation of anyfull-scale reactors to determine the substrate removal rates. Var-ious kinetic models viz., Monod, Contois, modified Stover–Kincannon and Grau second order have been tested [32, 52].Among these, the modified Stover–Kincannon kinetic modelwas selected for the present study which has been widelyemployed for high-strength wastewater samples [19, 25, 53].

From Eq. 3, the saturation constant KB and the maximumutilization rate constant Umax in the model was estimated as13.64 and 18.51 g/l/day, respectively. The applicability ofEq. (3) was shown by the regression analysis which showed thatthe utilization rates were directly proportional to the reactorefficiency (R2=0.978; Fig. 5). The comparison studies exploringthe modified Stover–Kincannon model for anaerobic treatmentof different types of wastewater under different experimentalconditions is shown in Table 5. From the table, the maximumutilization constant (Umax) values (11.83 and 1.996 g/l/day)reported by Yetilmezsoy [19] is lower than the value obtainedin this study, however, lower than the estimated value obtainedfor synthetic-based wastewater [54]. The high Umax in thesynthetic wastewater could be attributed to the presence ofreadily biodegradable substrates that are easily accessible tomicroorganisms [54]. On the contrary, industrial scale waste-waters such as brewery effluent might contain different

R² = 0.957

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.15 0.25 0.35 0.45 0.55

Obs

erve

d S

e (g

CO

D/l)

Predicted Se (g COD/l)

Fig. 6 Relationship between theobserved effluent CODconcentration and predictedeffluent COD concentration bymodified Stover–Kincannonmodel

Table 5 Comparison of different types of anaerobic wastewater treatment processes using modified Stover–Kincannon model

Digester type Type of substrate Operating temperature (°C) Modified Stover–Kincannonmodel kinetic and estimatedcoefficients

References

KB (g/L/day) Umax(g/L/day) R2

UASB Brewery wastewater 28–32 13.64 18.51 0.978 Present study

UASB Poultry manure wastewater 30–34.5 13.02 11.83 0.991 [19]

Anaerobic biphasicfixed film reactor

Distillery wastewater 37 1.69(kg/m3/d) 2 (kg/m3/d) 0.992 [24]

UASB Municipal wastewater 17.1–21 1.536 1.996 0.972 [25]

UASB Synthetic wastewater(2,4-dichlorophenol)

– 0.0098 (mg/l/day) 0.01 (mg/l/day) 0.992 [58]

Anaerobic filter Synthetic wastewater (saline) 37 5.3 7.05 0.910 [53]

Mesophilic anaerobic filter Synthetic wastewater (starch) 35 50.6 49.8 0.998 [54]

Mesophilic anaerobic filter Paper pulp liquor 35 6.14 6.71 0.998 [54]

366 A.M. Enitan et al.

recalcitrant and more complex compounds that are less de-gradable [19]. Furthermore, the operating conditions of anaer-obic reactors could also influence the activity of microorgan-isms which can affect the kinetic rates. The biochemical andthe kinetic data obtained in this study confirms the efficiencyof the microbial community present within the UASB reactorin degrading the organic matter present in brewery wastewaterto produce optimum biogas that can serve as source of energy.Further, to test the validity of the model, the observed effluentCOD values and predicted values obtained from the modelwere compared (Fig. 6). The results indicated the significanceof the model with an excellent fit between the predictedeffluent COD concentrations and the observed concentrations(Fig. 6). The high R2 value of 0.957 between the observed andpredicted values suggested that the predicted results are inaccordance with the observed results (Fig. 6). This furthervalidates the applicability of modified Stover–Kincannonmodel to predict effluent concentrations from any anaerobictreatment system treating different types of wastewater.

Conclusions

The current study confirms the presence of a diverse group ofbiogas producing methanogens within the UASB reactor gran-ules treating brewery wastewater. The methanogens assist inreducing the intermediate metabolites that are produced duringthe breaking down of the organic matter to large volume ofbiogas (especially methane and carbon dioxide) for energy gen-eration. The UASB reactor showed an average COD removalefficiency of 78.97 % and biogas production of 66–69 % duringthe investigation period. The significance of the modified Sto-ver–Kincannonmodelwith an excellent fit between the predictedand observed effluent COD concentrations confirmed the appli-cability of this model to be used for predicting and improving thereactor performance in treating brewery wastewater.

Acknowledgments The authors gratefully acknowledge South AfricanBreweries (SAB) for their continuous support for this study and theDurban University of Technology for the financial and laboratorysupport.

References

1. Parawira W, Kudita I, Nyandoroh MG, Zvauya R (2005) A study ofindustrial anaerobic treatment of opaque beer brewery wastewater ina tropical climate using a full-scale UASB reactor seeded withactivated sludge. Process Biochem 40:593–599

2. Appels L, Baeyens J, Degréve J, Dewil R (2008) Principles andpotential of the anaerobic digestion of waste-activated sludge. ProgEnerg Combust 34:755–781

3. Mirzoyan N, Parnes S, Singer A, Tal Y, Sowers K, Gross A (2008)Quality of brackish aquaculture sludge and its suitability for anaero-bic digestion and methane production in an upflow anaerobic sludge

blanket (UASB) reactor. Aquaculture 279:35–41. doi:10.1016/j.aquaculture.2008.04.008

4. Amani T, Nosrati M,Mousavi SM, Kermanshahi RK (2011) Study ofsyntrophic anaerobic digestion of volatile fatty acids using enrichedcultures at mesophilic conditions. Int J Environ Sci Tech 8:83–96

5. Hulshoff L, Lens P, Castro S, Lettinga G (2004) Anaerobic sludgegranulation. Water Res 38:1376–1389

6. Crocetti G, Murto M, Björnsson L (2006) An update and optimisa-tion of oligonucleotide probes targeting methanogenic Archaea foruse in fluorescence in situ hybridisation (FISH). J Microbiol Methods65:194–201

7. Chulhwan P, Chunyeon L, Sangyong K, Yu C, Howard CH (2005)Upgrading of anaerobic digestion by incorporating two differenthydrolysis processes. J Biosci Bioeng 100:164–167

8. Mumme J, Linke B, Tolle R (2010) Novel upflow anaerobic solid-state (UASS) reactor. Bioresour Technol 101:592–599

9. Batstone DJ, Keller J, Blackall LL (2004) The influence of substratekinetics on the microbial community structure in granular anaerobicbiomass. Water Res 38:1390–1404

10. Yu Y, Lee C, Hwang S (2004) Analysis of community structure inanaerobic processes using a quantitative real-time PCR method.Proceedings of the 10th Anaerobic Digestion Conference,Montréal, Canada., pp. 459-465

11. Liu WT, Chan OC, Fang HHP (2002) Microbial community dynam-ics during start-up of acidogenic reactors. Water Res 36:3203–3210

12. Ziganshin A, Schmidt T, Scholwin F, Il’inskaya O, Harms H,Kleinsteuber S (2011) Bacteria and archaea involved in anaerobicdigestion of distillers grains with solubles. Appl MicrobiolBiotechnol 89:2039–2052. doi:10.1007/s00253-010-2981-9

13. Keyser M, Witthuhn RC, Lamprecht C, Coetzee MPA, Britz TJ(2006) PCR-based DGGE fingerprinting and identification ofmethanogens detected in three different types of UASB granules.Syst Appl Microbiol 29:77–84. doi:10.1016/j.syapm.2005.06.003

14. Diaz EE, Stams AJM, Amils R, Sanz JL (2006) Phenotyic propertiesand microbial diversity of methanogenic granules from a full scaleupflow anaerobic sludge bed reactor treating brewery wastewater.Appl Env Microbiol 72:4942–4949

15. McHugh S, O’Reilly C, Mahony T, Colleran E, O’Flaherty V (2003)Anaerobic granular sludge bioreactor technology. Rev Environ SciBiotechnol 2:225–245

16. Zhang L, SunY, GuoD,Wu Z, JiangD (2012)Molecular diversity ofbacterial community of dye wastewater in an anaerobic sequencingbatch reactor. Afr J Microbiol Res 6444–6453

17. Pontes RFF, Pinto JM (2006) Analysis of integrated kinetic and flowmodels for anaerobic digesters. Chem Eng J 122:65–80. doi:10.1016/j.cej.2006.02.018

18. Acharya BK, Mohana S, Madamwar D (2008) Anaerobic treatmentof distillery spent wash-A study on upflow anaerobic fixed filmbioreactor. Bioresour Technol 99:4621–4626

19. Yetilmezsoy K (2012) Integration of kinetic modeling and desirabil-ity function approach for multi-objective optimization of UASBreactor treating poultry manure wastewater. Bioresour Technol118(118):189–101

20. Batstone D, Keller J, Angelidaki I, Kalyhuzhnyi S, Pavlostathis S,Rozzi A, Sanders W, Siegrist H, Vavilin V (2002) The IWAAnaerobic Digestion Model No.1 (ADM1). Water Sci Technol 45:65–73

21. Parsamehr M (2012) Modeling and analysis of a UASB reactor.Luleå University of Technology

22. Yetilmezsoy K, Sakar S (2008) Development of empirical models forperformance evaluation of UASB reactors treating poultry manurewastewater under different operational conditions. J Hazard Mater153:532–543. doi:10.1016/j.jhazmat.2007.08.087

23. Debik E, Coskun T (2009) Use of the static granular bed reactor(SGBR) with anaerobic sludge to treat poultry slaughterhouse waste-water and kinetic modelling. Bioresour Technol 100:2777–2782

Kinetics and Characterization of Microbes of an UASB Reactor 367

24. Acharya BK, PathakH,Mohan S, ShoucheY, Singh V,Madamwar D(2011) Kinetic modelling and microbial community assessment ofanaerobic biphasic fixed film bioreactor treating distillery spent wash.Water Res 45:4248–4259

25. Turkdogan-Aydinol FI, Yetilmezsoy K (2010) A fuzzy-logic-basedmodel to predict biogas and methane production rates in a pilot-scalemesophilic UASB reactor treating molasses wastewater. J HazardMater 182:460–471. doi:10.1016/j.jhazmat.2010.06.054

26. APHA–AWWA–WPCF (1998) Standard methods for the examina-tion of water and wastewater. 20th ed. Washington, DC, USA.American Public Health Association/American Water WorksAssociation/Water Environment Federation

27. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Proteinmeasurement with the folin phenol reagent. J Biol Chem 193:265–275

28. Amann RI, Ludwig W, Schleifer KH (1995) Phylogenetic identifica-tion and in situ detection of individual microbial cells without culti-vation. Microbiol Rev 59:143–169

29. Sekiguchi Y, Kamagata Y, Syutsubo K, Ohashi A, Harada H,Nakamura K (1998) Phylogenetic diversity of mesophilic and ther-mophilic granular sludges determined by 16 s rRNA gene analysis.Microbiology 144:2655–2665

30. Luton PE, Wayne JM, Sharp RJ, Riley PW (2002) The mcrA gene asan alternative to 16S rRNA in the phylogenetic analysis ofmethanogen populations in landfill. Microbiology 148:3521–3530

31. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S(2011) MEGA5: molecular evolutionary genetics analysis usingmaximum likelihood, evolutionary distance, and maximum parsimo-ny methods. Mol Biol Evol 28:2731–2739

32. Kincannon DF, Stover EL (1982) Design methodology for fixed filmreaction—RBCs and biological towers. Pergamon, New York

33. Krzysztof Z, Frac M (2012) Methane fermentation process as anaer-obic digestion of biomass: transformations, stages and microorgan-isms. Afr J Biotechnol 11:4127

34. Sekiguchi Y, Kamagata Y, Nakamura K, Ohashi A, Harada H (1999)Fluorescence in situ hybridization using 16S rRNA-targeted oligo-nucleotides reveals localization of methanogens and selected uncul-tured bacteria in mesophilic and thermophilic sludge granules. ApplEnviron Microbiol 65:1280–1288

35. Kovacik WP, Scholten JCM, Culley D, Hickey R, Zhang W,Brockman FJ (2010)Microbial dynamics in upflow anaerobic sludgeblanket (UASB) bioreactor granules in response to short-term chang-es in substrate feed. Microbiology 156:2418–2427. doi:10.1099/mic.0.036715-0

36. Jupraputtasri W, Boonapatcharoen N, Cheevadhanarak S,Chaiprasert P, Tanticharoen M, Techkarnjanaruk S (2005) Use ofan alternative Archaea-specific probe for methanogen detection. JMicrobiol Methods 61:95–104. doi:10.1016/j.mimet.2004.11.017

37. Raskin L, Stromley JM, Rittmann BE, Stah DA (1994) Groupspecific 16S rRNA hybridization probes to describe natural commu-nities of methanogens. Appl Environ Microbiol 60:1232–1240

38. Sekiguchi Y, Takahashi H, Kamagata Y, Ohashi A, Harada H (2001)In situ detection, isolation, and physiological properties of a thinfilamentous microorganism abundant in methanogenic granularsludges: a novel isolate affiliated with a clone cluster, the greennon-sulfur bacteria, subdivision I. Appl Environ Microbiol 67:5740–5749. doi:10.1128/aem.67.12.5740-5749.2001

39. Gomec CY, Letsiou I, Ozturk I, Eroglu V, Wilderer PA (2008)Identification of Archaeal population in the granular sludge of anUASB reactor treating sewage at low temperatures. J Environ SciHealth Part A 43:1504–1510

40. Vavilin VA, Qu X, Mazeas L, Lemunier M, Duquennoi C, He P,Bouchez T (2008) Methanosarcina as the dominant aceticlasticmethanogens during mesophilic anaerobic digestion of putresciblewaste. Antonie Van Leeuwenhoek 94:593–605. doi:10.1007/s10482-008-9279-2

41. Castro H, Ogram A, Reddy KR (2004) Phylogenetic characterizationof methanogenic assemblages in eutrophic and oligotrophic areas ofthe Florida everglades. Appl Environ Microbiol 70:6559–6568

42. Cardinali-Rezende J, Debarry R, Colturato LDB, Carneiro E,Chartone-Souza E, Nascimento AA (2009) Molecular identificationand dynamics of microbial communities in reactor treating organichousehold waste. Appl Microbiol Biotechnol 84:777–789. doi:10.1007/s00253-009-2071-z

43. Cheng L, Dai L, Li X, Zhang H, Lu Y (2011) Isolation and charac-terization of Methanothermobacter crinale sp. nov, a novelhydrogenotrophic methanogen from Shengli Oilfields. ApplEnviron Microbiol 77:5212–5219

44. Shlimon AG, Friedrich MW, Niemann H, Ramsing NB, Finster K(2004)Methanobacterium aarhusense sp. nov., a novel methanogenisolated from a marine sediment (Aarhus Bay, Denmark). Int J SystEvol Microbiol 54:759–763

45. Kampmann K, Ratering S, Baumann R, Schmidt M, Zerr W, SchnellS (2012) Hydrogenotrophic methanogens dominate in biogas reac-tors fed with defined substrates. Syst Appl Microbiol 35:404–413

46. Smith JM, Castro H, Ogram A (2007) Structure and function ofmethanogens along a short-term restoration chronosequence in theFlorida Everglades [down-pointing small open triangle. ApplEnviron Microbiol 73:4135–4141

47. Karakashev D, Batstone DJ, Trably E, Angelidaki I (2006) Acetateoxidation is the dominant pathway from acetate in the absence ofMethanosaetaceae. Appl Environ Microbiol 72:5138–5141. doi:10.1128/AEM.00489-06

48. Ferry J (1993) Methanogenesis: ecology, physiology, biochemistryand genetics. Chapman & Hill, New York

49. McMahon KD, Stroot PG, Mackie RI, Raskin L (2001) Anaerobiccodigestion of municipal solid waste and biosolids under variousmixing conditions—II. Microbial population dynamics. Water Res35:1817–1827. doi:10.1016/S0043-1354(00)00438-3

50. Delbès C, Moletta R, Godon J-J (2001) Bacterial and archaeal 16SrDNA and 16S rRNA dynamics during an acetate crisis in an anaer-obic digester ecosystem. FEMS Microbiol Ecol 35:19–26

51. Wikström T, Nordmark D, Pelkonen M, Lagerkvist A (2012)Fluorescent in situ hybridization technique in anaerobic processstudies. In: Lagerkvist A (ed) Abstract proceedings of 7thIntercontinental Landfill Research Symposium. Luleå TekniskaUniversitet, Sunderbyn, Luleå, Sweden

52. Pandian M, H-H NGO, Pazhaniappan S (2011) Substrate removalkinetics of an anaerobic hybrid reactor treating pharmaceutical waste-water. J Water Sustain 1:301–312

53. Kapdan IK, Erten B (2007) Anaerobic treatment of saline wastewaterby Halanaerobium lacurosei . Process Biochem 42:449–453

54. Ahn JH, Forster CF (2000) Kinetic analyses of the operation ofmesophilic and thermophilic anaerobic filters treating a simulatedstarch wastewater. Process Biochem 36:19–23

55. Stahl DA, Amann R (1991) Development and application of nucleicacid probes. In: Stackebrandt E, Goodfellow M (eds) Nucleic acidtechniques in bacterial systematics, vol 8. Wiley, London, England,pp 207–248

56. Amann RI, Binder BJ, Olson RJ, Chisholm SW, Devereux R, StahlDA (1990) Combination of 16S rRNA-targeted oligonucleotideprobes with flow cytometry for analyzing mixed microbial popula-tions. Appl Environ Microbiol 56:1919–1925

57. Daims H, Bruhl A, Amann R, Schleifer K-H, Wagner M (1999) Thedomain-specific probe EUB338 is insufficient for the detection of allbacteria: development and evaluation of a more comprehensive probeset. Syst Appl Microbiol 22:434–444

58. Sponza D, Uluköy A (2008) Kinetic of carbonaceous substrate in anupflow anaerobic sludge blanket (UASB) reactor treating 2,4 dichlo-rophenol (2,4 DCP). J Environ Manag 86:121–131

59. Saitou N, Nei M (1987) The neighbor-joining method: a new methodfor reconstructing phylogenetic trees. Mol Biol Evol 4:406–425

368 A.M. Enitan et al.