Shing Lee 2006 EMT

8/3/2019 Shing Lee 2006 EMT

1/8

Enzyme and Microbial Technology 35 (2004) 605612

Operation strategies for biohydrogen production with a high-rateanaerobic granular sludge bed bioreactor

Kuo-Shing Leea, Yung-Sheng Lob, Yung-Chung Loa, Ping-Jei Lina, Jo-Shu Changb,

a Department of Chemical Engineering, Feng Chia University, Taichung, Taiwanb Department of Chemical Engineering, National Cheng Kung University, Tainan 701, Taiwan

Abstract

Long-term operation for biohydrogen production with an efficient carrier-induced granular sludge bed (CIGSB) bioreactor had encoun-

tered problems with poor biomass retention at a low hydraulic retention (HRT) as well as poor mass-transfer efficiency at a high HRT or

under a prolonged operation period. This work was undertaken to develop strategies enabling better biomass retention and mass-transfer

efficiency of the CIGSB reactors. Supplementation of calcium ion was found to enhance mechanical strength of the granular sludge. Ad-

dition of 5.427.2 mg/l of Ca2+ also led to an over three-fold increase in biomass concentration and a nearly five-fold increase in the H 2production rate (up to 5.1 l H2 /h/l). Two reflux strategies were utilized to enhance the mass-transfer efficiency of the CIGSB system. The

liquid reflux (LR) strategy enhanced the H2 production rate by 2.2-fold at an optimal liquid upflow velocity of 1.09 m/h, which also gave a

maximal biomass concentration of ca. 22 g VSS/l. Similar optimal H2 production rate was also obtained with the gas reflux (GR) strategy

at a rate of 1.01.49 m/h, whereas the biomass concentration decreased to 27 g VSS/l and thereby the specific H2 production rate was

higher than that with LR. The operation strategies applied in this work were effective to allow stable and efficient H 2 production for nearly

100 days.

2004 Elsevier Inc. All rights reserved.

Keywords: Biohydrogen production; Granular sludge; Calcium ion; Liquid/gas reflux strategy

1. Introduction

Being clean, recyclable, and efficient, H2 is considered a

promising energy carrier of the future [1,2]. Thus, the need

of a sufficient supply of H2 is in great demand. In contrast to

conventional H2 production methods that convert fossil fuels

into H2 by thermal/chemical means, H2 production via bio-

logical pathways may be a cost-effective and pollution-free

alternative [1]. Early researches on biohydrogen productionwere mainly focused on bio-photolysis of water by algae and

cyanobacteria [3] as well as photo-fermentation of organic

substrates by photosynthetic bacteria [4]. However, after the

mid 1990s, much attention has been paid to dark fermen-

tation, in which H2 is produced from organic compounds

(especially carbohydrates) by anaerobic bacteria (e.g., aci-

Corresponding author. Fax: +886 6 235 7146/234 4496.

E-mail address: [email protected] (J.-S. Chang).

dogenic bacteria). This transition in biohydrogen research

is primarily because dark fermentation normally achieves a

much higher H2 production rates than water photolysis and

photo-fermentation. In addition, dark fermentation also owns

the advantage of simultaneous waste reduction and H2 gen-

eration.

Recentstudies on H2 fermentation with anaerobic bacteria

utilized pure bacterial strains to produce H2 [1,58]. Accli-

mated sewage sludge or microflora has also been used for H2production in some reports [813]. CSTR (continuous stirred

tank reactor) operations have been the most frequently used

configuration for H2-producing bioreactor [6,12,14,15]. Due

to the slow-growing feature of the H2-producing bacteria, the

CSTR reactor becomes unstable when it is conductedat a high

dilution rate because of the washout of cells [13]. This causes

operational instability of the bioreactor and also limits the H2production rate. Consequently, immobilized cells created by

natural or synthetic matrices [4,1619] were often used to al-

0141-0229/$ see front matter 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.enzmictec.2004.08.013


2/8

606 K.-S. Lee et al. / Enzyme and Microbial Technology 35 (2004) 605612

low better retention of H2-producing bacterial cells for stable

operationsat higher dilution rates. Although beingcommonly

used in a variety of bioprocesses, the gel-entrapped-type im-

mobilized cells have the drawback of being cost intensive

and less feasible when a large-scale mixed culture (such as

an activated sludge system) is involved. In contrast, cell im-

mobilization via surface attachment (e.g., biofilm) [16,17,20]or via self-flocculation [11,17,2126] (e.g., granular sludge

bed) may have higher feasibility in practical environmental

applications.

We recently developed a carrier-induced granular sludge

bed (CIGSB) bioreactor able to significantly increase the

retention of H2-producing sludge and thereby being very

efficient in biohydrogen production [21,33]. Although the

CIGSB system is an excellent candidate of a practical

biohydrogen-producing process, operation at a very low hy-

draulic retention time (HRT) (e.g, HRT < 1 h) still weakened

the stability of the granular sludge bed, leading to washout

of the H2-producing sludge. Meanwhile, mass-transfer ef-

ficiency of the bioreactor markedly decreased after a long-term operation due to clot formation resulting from massive

production of bacterial cells and lack of efficient mixing.

Therefore, this study was attempted to resolve the forego-

ing problems by supplementation of calcium ions and by

employment of two reflux strategies (liquid and gas reflux)

for better mixing. This study aimed to develop optimal op-

eration strategies for an efficient and stable performance of

H2 fermentation in the novel CIGSB system. The informa-

tion obtained from this study is expected to provide basic

knowledge for the bioreactor design of practical bioprocesses

capable of long-term production of H2 from wastewater or

wastes.

2. Materials and methods

2.1. Preparation of H2-producing sludge

The seed sludge obtained from a wastewater treatment

plant located in central Taiwan was subjected to acidic

pretreatment [27] to inactivate methanogenic bacteria and

acid-sensitive non-H2-producing bacterial population in the

sludge. The acidic pretreatment involved a decrease in sludge

pH to ca. 3.0 by 0.1 N HCl for 24 h and a readjustment of pH

back to 7.0 by 0.1 N NaOH. Prior to start-up of bioreactor,

the acid-treated sludge was acclimated in a continuous-flow

reactor at a hydraulic retention time of 612 h and a temper-

ature of 35 C.

2.2. Fermentation medium

Composition of the medium used for H2 fermentation was

described in our recent work [21]. In some experiments,

20100 mg/l of CaCl22H2O (i.e., 5.427.2 mg/l of Ca2+)

was supplemented into the feed to improve the mechanical

strength of the granular sludge.

2.3. Carrier support

The carrier used for the CIGSB reactor was cylindrical

activated carbon(CAC) obtained from China CarbonCo. Ltd.

(Taipei, Taiwan). The CAC carrier was 34 mm in diameter

and 515 mm in height. The gravity of the carrier was 1.34,

and the surface area was greater than 1200 m2

/g. The parentmaterial of the cylindrical activated carbon was bituminous

coal with a bulk density of 0.400.48 g/cm3 and an iodine

number of 1150 mg/g.

2.4. Hydrogen-producing granular sludge bed

bioreactor

The CIGSB reactor consisted of an acrylic resin-made

column with 8 cm in diameter, 40 cm in height, and a work-

ing volume of 1 l. The column was initially packed with the

carrier support (activated carbon) at a bed height of 4 cm

and bed porosities (void fractions of the column) of 90%.

Biofilms rich in H2-producing bacteria were grown on thecarrier supports at batch mode according to our recent work

[20] before the reactor was switched to a continuous mode.

The effluent of the reactor went to a gasliquid separator,

where the gaseous and soluble products were collected sep-

arately. The CIGSB reactor was started up by feeding fresh

medium at a hydraulic retention time (HRT) of 4 h. After

reaching steady-state operation, the HRT was decreased to

2 h, at which self-flocculated sludge granulation started to

occur [21]. The HRT was subsequently decreased from 2

to 0.5 h, during which addition of Ca2+, at a concentration

of 5.427.2 mg/l, as well as liquid and gas refluxes were

employed at different stages using peristaltic pumps. Therange of liquid reflux rate was 010.5 l/h/l, corresponding

to liquid upflow velocity of 0.22.29 m/h. The gas reflux

rates applied were 2.07.5 l/h/l, corresponding to gas up-

flow velocity of 0.41.49 m/h. The compositions of gas prod-

ucts (predominantly, H2 and CO2) and soluble metabolites

(volatile fatty acids, alcohols, etc.) produced during H2 fer-

mentation were monitored as a function of time. The pH and

biomass concentration (represented by volatile suspended

solid, VSS) in the effluent were also recorded. The reac-

tor was operated at a constant temperature of 35 C and a

pH of 6.7 [6,14]. A gas meter (Type TG1; Ritter Inc., Ger-

many) was used to measure the amount of gas products gen-

erated, and the gas volumes were calibrated to 25 C and 760

mmHg.

2.5. Analytical methods

The gas products (mainly H2 and CO2) and soluble

metabolites (volatile fatty acids and ethanol) were analyzed

by gas chromatography (GC). Detailed operation conditions

of the GC measurements were described elsewhere [21].

Standard methods [28] were used to determine biomass con-

centration (in terms of volatile suspended solid) of samples

taken separately from granular sludge portion and suspended-


3/8

K.-S. Lee et al. / Enzyme and Microbial Technology 35 (2004) 605612 607

cell portion in the reactor. The carbohydrate concentration in

theeffluent was also measured accordingto standard methods

[28]. The size of granular sludge was determined by measur-

ing the size of 1015 granules with a ruler, and the mean

values are reported.

3. Results and discussion

Long-term operation (over 90 days) of the CIGSB reac-

tor was conducted with a variety of operation conditions and

strategies. As indicated in Fig. 1, there are four stages dur-

ing the course of experiments. From days 0 to 24, it was a

HRT-varying stage (HVS stage), during which the reactor

was carried out at a progressively decreased HRT from 4 to

0.5 h. The next stage (from days 24 to 35) was the calcium

addition stage (CAS stage); during which 5.427.2 mg/l of

Ca2+ was added into the reactor operated at HRT = 0.5 h. Af-

ter the CAS stage, the HRT was increased from 0.5 to 1 h, and

the operation entered the liquid-reflux stage (LRS stage; days

3567), during which liquid reflux was employed at a liquid

upflow velocity of 0.42.29 m/h. The last stage (days 6792)

was gas-reflux stage (GRS stage), during which gas reflux

was employed at a gas upflow velocity of 0.41.49m/h. The

resulting time-course profiles of H2 production rate, H2 con-

tent in biogas, and concentration of soluble metabolites were

illustrated in Fig. 1. Detailed description and interpretation of

the experimental data are presented in the following sections.

Fig. 1. Time-course profileof H2 production rate, H2 content in biogas, and soluble metabolite concentration with respect to differentHRT and operationstages

in a carrier-induced granular sludge bed bioreactor. HVS: HRT-varying stage; CAS: calcium addition stage; LRS: liquid reflux stage; GRS: gas reflux stage.

3.1. Effect of HRT on H2 production (HVS stage)

The granular sludge bed reactor was started up at a HRT of

4 h with a steady-state H2 production rate (vH2 ) of 0.41 l/h/l

and a H2 content (CH2 ) of 26% (Table 1; Fig. 2). The vH2and CH2 were enhanced by three-fold and 32%, respectively,

when theHRTwas cutdownto 2 h, dueprimarily to formationof granular sludge, which was ca. 1 mm in diameter. Sludge

granulation at HRT = 2 h also gave rise to a dramatic increase

in biomass concentration (Cb) from1.2g VSS/l (at HRT=4 h)

to7.8gVSS/l(Fig.2). Decrease of HRT to 1 h also resulted in

a bettervH2 and Cb of 2.15 l/h/l and 8.3 g VSS/l, respectively,

while the CH2 value (33.6%) was similar to that obtained

at HRT = 2 h. However, when HRT was further decreased

to 0.5 h, performance of biohydrogen production diminished

sharply, as thevH2 andCH2 decreased to 1.14 l/h/l and 21.4%,

respectively(Table 1; Fig. 2). Thebiomass concentration also

dropped to 4.2 g VSS/l (only 50% of that observed at HRT

= 1 h), indicating that severe wash-out of biomass occurred

when the reactor was conducted at a HRT of 0.5 h. Since vH2and Cb both dropped nearly 50% for a decrease of HRT from

1to0.5h(Fig. 2), the poor performance of the reactor at HRT

= 0.5 h appeared to be attributed to the loss of H2-producing

bacterial population due to the wash-out effect. The H2 yield

(YH2 ) remained fairly stable at about 2 mol H2 /mol sucrose

for HRT = 14 h, but decreased to only 1 mol H2/mol sucrose

at HRT = 0.5 h (Fig. 2). The sucrose conversion maintained

higher than 70% at HRT of 14 h with a maximal value of


4/8


5/8


Fig. 2. The effect of hydraulic retention time (HRT) on H2 production rate,

H2 yield, total biomass, ratio of granular sludge to total biomass, and the

size of granular sludge during the HRT-varying stage.

3.3. Effect of liquid reflux on H2 production (LRS stage)

After the CAS stage, the HRT was switched from 0.5 to

1 h with a continual additionof 5.4 mg/l of Ca2+ in an attempt

for long-term operation. Unfortunately, during the period of

days 3540,the H2 production and H2 content both decreased

abruptly to 1.26 l/h/l and 29%, respectively (Fig. 1; Table 1).

Inspection of thebioreactor shows that there areseveral dead

zones in the granular sludge bed due to poor mixing effi-

ciency of the reactor. Thus, the downhill performance may

be caused by mass-transfer problems. It seems not feasible

to apply a mechanical stirrer for mixing, because the stirrer

may easily destroy the structure of granular sludge, acceler-

ating washout of cells. Therefore, we attempted to recycle

the liquid effluent to improve the mixing efficiency. The se-

quence of adjustment in liquid reflux rate was indicated in

Fig. 1. For the liquid reflux rate (represented by liquid up-

flowvelocity;Vup,liq)of0.21.09m/h,theH2 productionrate,

H2 yield, and H2 content all considerably increased with an

increase in Vup,liq (Fig. 4; Table 1). However, when Vup,liqwas increased to 1.53 and 2.29 m/h, the H2 production per-

Fig. 3. The effect of Ca2+ supplement on H2 production rate, H2 yield, total

biomass, ratio of granular sludge to total biomass, and the size of granular

sludge during the calcium addition stage.

formance decreased sharply. It is likely that the high upflow

velocity exceeded the limit of hydraulic tolerance of the gran-

ular sludge, evidenced by a dramatic decrease of biomass

concentration from 1522 g VSS/l to 1.8 g VSS/l in response

to an increase of Vup,liq from 1.09 to 2.29 m/h (Fig. 4). In

addition, a high liquid reflux may also dilute the sucrose con-

centration in the fermentation culture, causing negative effect

on H2 production. Therefore, a liquid reflux rate of 1.09 m/h

was appropriate to enhance and stabilize the performance of

the bioreactor. A repeated operation at Vup,liq = 1.09 m/h dur-

ing the final period (days 6167 in Fig. 1) of LRS stage con-

firmed that operation at liquid reflux of 1.09 m/h resulted in

optimal H2 production performance (see 1.09R in Fig. 3).

During the LRS stage, the H2 yield and sucrose conversion

were maintained at a level of 2.02.4 mol H2 /mol sucrose

and 8290%, respectively, except for much lower values at

Vup,liq = 2.29 m/h (Fig. 4; Table 1). The H2 content was quite

stable and accounted for 3036% in the biogas (Table 1),

suggesting a stable population structure of H2 producers in

the sludge despite the employment of different liquid reflux

rates.


6/8


Fig. 4. The effect of liquid upflow velocity on H2 production rate, H2 yield,

total biomass, ratio of granular sludge to total biomass, and the size of gran-ular sludge during the liquid reflux stage. The 1.09R in x-axis denotes

repeated operation at Vup,liq = 1.09m/h.

3.4. Effect of gas reflux (GR) on H2 production (GRS

stage)

The second approach to improve mass-transfer efficiency

was via gas reflux strategies, in which gas effluent was recy-

cled through the bottom of reactor at a gas upflow velocity of

0.41.49 m/h. The liquid upflow rate was kept constant at ca.

0.2 m/h as a result of a constant feeding of medium at a HRT

of 1 h. The adjustment of gas reflux rate (represented by gas

upflow velocity; Vup,gas) followedthe sequence of 1.49 1.0

0.4 1.49m/h (Fig. 1). The results show that the best H2production occurred at a Vup,gas of 1.0 and 1.49 m/h, which

gave similar H2 production rate (vH2 ) of ca. 2.9l/h/l and a H2yield (YH2 )of2.4molH2/mol sucrose (Fig. 5). Operation at a

Vup,gas of 0.4 m/h resulted in much lower vH2 and YH2 . When

the Vup,gas was once again switched from 0.4 to 1.49 m/h (as

indicated as 1.49R in Fig. 5), the H2 production rate and

yield both rose but did not reach the levels as those obtained

when Vup,gas of 1.49 m/h was first applied (Fig. 5). The key

factor that influenced the H2 production performance during

the GRS stage seems to be the biomass concentration. Fig. 5

Fig. 5. Theeffectof gasupflowvelocityon H2 production rate,H2 yield, total

biomass, ratio of granular sludge to total biomass, and the size of granularsludge during the gas reflux stage. The 1.49R in x-axis denotes repeated

operation at Vup,gas = 1.49m/h.

shows that the trends ofvH2 andYH2 were consistent with that

of biomass concentration. The effect of gas reflux on biomass

concentration was two-folds; gas reflux may decrease the set-

tling ability of granular sludge but, on the other hand, may

also facilitate the contact of medium and cells to allow a bet-

ter cell growth. The latter is confirmed by the substantially

high substrate conversions (9192%) when Vup,gas of 1.0 and

1.49 m/h was employed (Table 1). The results show that a

Vup,gas of 1.0 m/h was appropriate for maintaining a solid

performance in terms of H2 production rate and yield. Wealso observed that a violent increase in gas reflux rate (e.g.,

Vup,gas of 0.41.49 m/h) should be avoided because it may

cause significant loss of biomass due to a sudden increase in

shear stress and upflow driving force.

3.5. Liquid reflux versus gas reflux

Recycle of liquid and biogas to improve mixing efficiency

of anaerobic bioreactors has been carried out in other stud-

ies [14,32]; there is still a lack of information concerning

comparison of the performance between the two mixing


7/8


strategies. The results obtained from LRS and GRS stages

show that although the optimal H2 production rate and yield

were quite similar, the biomass concentration in the GRS

stage was, in general, much lower than that obtained in LRS

(Figs.4and5). Asa result, the specific H2 productionrate was

significantly higher in GRS stage than in LRS stage (Table 1).

This information seems to imply that gas reflux had bettermixing efficiency than that of liquid reflux on the basis of

similar upflow velocity and also achieved higher specific H 2production activity. Employment of gas reflux would face the

consequence of severe reduction in biomass concentration in

the bioreactor. However, as long as the overall H2 produc-

tion activity is sustained, lower biomass production would

be favorable due to a smaller amount of waste sludge being

produced, alleviating the loading of downstream sludge treat-

ment. Therefore, it is proposed that the gas reflux strategy be

used if desired settling ability and mechanical strength of the

granular sludge can be achieved.

3.6. Characteristics of granular sludge

Approximately, 40 h after the switch of HRT from 4 to 2 h,

sludge granulation was observed in the bioreactor, accompa-

nied by a marked increase in biomass concentration (Fig. 2).

The size of granular sludge was initially ca. 1 mm in diam-

eter, and the size increased to nearly 2 mm as the HRT was

increased to 0.5 h. The size of granular sludge was enlarged

to 33.5 mm with the addition of Ca2+ (Fig. 3). During the

LRS stage, the size of granular sludge decreased as the liquid

reflux rate increased, andthe diameter becameless than 1 mm

as Vup,liq > 1.09 m/h (Fig. 4). Employment of gas reflux also

led to a small size of granular sludge (ca. 1.5 mm) (Fig. 5).The results show that the size of granular sludge (GS) was

larger when a GS-strengthening agent (Ca2+) was supplied

or when the bioreactor was operated under a lower hydraulic

loading and lower shear stress.

The ratio of granular sludge to total biomass retained in

the bioreactor (Rsludge) is considered a good indicator to the

efficiency of sludge granulation. The data in Figs. 25 show

that at CAS and LRS stages, the amount of granular sludge

accounted for the majority of total biomass, with an average

Rsludge value of 9598%. In comparison, the Rsludge dropped

to 8590 and 7184% at HVS and GRS stages, respectively.

These observation are consistent with the results discussed

earlier that Ca2+ supplementation facilitated the formation

and mechanical strength of granular sludge, and that gas re-

flux tended to decrease settling ability of the granular sludge,

probably due to disintegration of granular sludge structure.

3.7. Composition of soluble metabolites during H2fermentation

The data for production of soluble metabolites are indi-

cated in Fig. 1. Except for the early HVS stage, where ethanol

production was abnormally high, the predominant soluble

metabolite was acetic acid (HAc), followed by butyric acid

(HBu), ethanol (EtOH), and propionic acid (HPr). In addi-

tion, a trace amount of valeric acid (HVa) was also detected.

It can also be observed that production of the major soluble

products (HAc and HBu) had a similar trend to that of H2production rate, while production of ethanol rose when H2production was inefficient (e.g., days 812 and days 7983

in Fig. 1). These results indicate that H2 production in thesludge was via acidogenesis pathways and that solvent for-

mation (e.g., ethanol) inhibited production of H2 [6]. A re-

cent phylogenic study on the H2-producing sludge used in

the work showed that the dominant H2 producers in our sys-

tem belonged to the strains ofClostridium butyricum [33,34],

which is a spore-forming acidogenic bacterium that produces

H2 during degradation of organic substrates [1].

3.8. Highlights of special features of the granular sludge

bed bioreactor

The optimal H2

production rate obtained from this work

was nearly 5 l/h/l (i.e., 204 mmol/h/l) (Table 1), which is

much higher than most of the results reported by compara-

ble studies [11,16,17,2426]. The CIGSB reactor presented

here can also be stably operated at an extremely low HRT

of 0.5 to 1 h and a high organic loading of 2040 kg COD/h

for a prolonged time span with a high substrate conversion of

8093%. In contrast, most of the H2-producing bioreactors

in comparable studies were unable to be stably maintained at

HRT 2 h [11,14,24,26]. Moreover, the bioreactor presented

here was able to retain up to 22 g VSS/l of biomass under a

high dilution rate of 1 h1 (i.e., a HRT of 1 h) during liquid-

reflux stage (Fig. 4). The ability to retain a large amount

of biomass for high-cell-density, H2 fermentation makes itanother advantageous feature of our system. The operation

strategies (e.g., calcium supplement, liquid, and gas reflux)

were effective in maintaining efficient and stable H2 produc-

tion in the granular sludge bed bioreactor for nearly 100 days.

Thus, the bioreactor may be applied in practicalproduction of

H2 from pure substrates or wastes. The bioreactor strategies

may also be a good reference for performance improvement

in bioreactors that have similar configurations.

Acknowledgements

The authors gratefully acknowledge the financial supportof National Science Council (Grant no. NSC91-2815-C-035-

013-E) and of Energy Commission, Ministry of Economic

Affairs of Taiwan (Grant no. NSC93-ET-7-006-001-ET). The

authors also thank Miss Ji-Fung Wu for her assistance on

bioreactor operations and Professor C.-Y. Lin of Feng Chia

University for providing the H2-producing seed sludge.

References

[1] Das D, Veziroglu TN. Hydrogen production by biological processes:

a survey of literature. Int J Hydrogen Energy 2001;26:1328.


8/8


[2] Levin DB, Pitt L, Love M. Biohydrogen production: prospects

and limitations to practical application. Int J Hydrogen Energy

2004;29:17385.

[3] Asada Y, Miyake J. Photobiological hydrogen production. J Biosci

Bioeng 1999;88:16.

[4] Zhu H, Suzuki T, Tsygankov AA, Asada Y, Miyake J. Hydro-

gen production from tofu wastewater by Rhodobacter sphaeroides

immobilized in agar gels. Int J Hydrogen Energy 1999;24:30510.

[5] Nandi R, Sengupta S. Microbial production of hydrogen: an

overview. Crit Rev Microbiol 1998;24:6184.

[6] Kataoka N, Miya A, Kiriyama K. Studies on hydrogen production by

continuous culture system of hydrogen producing anaerobic bacteria.

Water Sci Technol 1997;36:417.

[7] Sparling R, Risbey D, Poggi-Varalldo HM. Hydrogen produc-

tion from inhibited anaerobic composters. Int J Hydrogen Energy

1997;22:5636.

[8] Ueno Y, Haruta S, Ishii M, Igarashi Y. Characterization of a mi-

croorganism isolated from the effluent of hydrogen fermentation by

microflora. J Biosci Bioeng 2001;92:397400.

[9] Nakamura M, Kanbe H, Matsumoto J. Fundamental studies on hy-

drogen production in the acid-forming phase and its bacteria in anaer-

obic treatment processes: the effects of solids retention time. WaterSci Technol 1993;28:818.

[10] Yokoi H, Tokushige T, Hirose J, Hayashi S, Takasaki Y. H2 produc-

tion from starch by a mixed culture of Clostridium butyricum and

Enterobacter aerogenes. Biotechnol Lett 1998;20:1437.

[11] Fang HHP, Liu H, Zhang T. Characterization of a hydrogen-

producing granular sludge. Biotechnol Bioeng 2002;78:4452.

[12] Lay JJ. Modeling and optimization of anaerobic digested sludge

converting starch to hydrogen. Biotechnol Bioeng 2000;68:269

78.

[13] Chen CC, Lin CY, Chang JS. Kinetics of hydrogen production with

continuous anaerobic cultures utilizing sucrose as the limiting sub-

strate. Appl Microbiol Biotechnol 2001;57:5664.

[14] Lin CY, Chang RC. Hydrogen production during the anaerobic

acidogenic conversion of glucose. J Chem Technol Biotechnol

1999;74:498500.[15] Majizat A, Mitsunori Y, Mitsunori W, Michimasa N, Junichiro M.

Hydrogen gas production from glucose and its microbial kinetics in

anaerobic systems. Water Sci Technol 1997;36:27986.

[16] Kumar N, Das D. Continuous hydrogen production by immobilized

Enterobacter cloacae IIT-BT 08 using lignocellulosic materials as

solid matrices. Enzyme Microb Technol 2001;29:2807.

[17] Yokoi H, Tokushige T, Hirose J, Hayashi S, Takasaki Y. Hydrogen

production by immobilized cells of aciduric Enterobacter aerogenes

strain HO-39. J Ferment Bioeng 1997;83:4814.

[18] Wu SY, Lin CN, Chang JS, Lee KS, Lin PJ. Microbial Hydro-

gen Production with Immobilized Sewage Sludge. Biotechnol Prog

2002;18:9216.

[19] Wu SY, Lin CN, Chang JS. Hydrogen production with immobilized

sewage sludge in three-phase fluidized-bed bioreactors. Biotechnol

Prog 2003;19:82832.

[20] Chang JS, Lee KS, Lin PJ. Biohydrogen production with fixed-bed

bioreactors. Int J Hydrogen Energy 2002;27:116774.[21] Lee KS, Lo YS, Lo YC, Lin PJ, Chang JS. H2 production with anaer-

obic sludge using activated-carbon supported packed-bed bioreactors.

Biotechnol Lett 2003;25:1338.

[22] Liu H, Fang HHP. Hydrogen production from wastewater by acido-

genic granular sludge. Water Sci Technol 2002;47:1538.

[23] Palazzi E, Fabino B, Perego P. Process development of continuous

hydrogen production by Enterobacter aerogenes in a packed column

reactor. Bioprocess Eng 2000;22:20513.

[24] Tanisho S, Ishiwata Y. Continuous hydrogen production from mo-

lasses by fermentation using urethane foam as a support of flocks.

Int J Hydrogen Energy 1995;20:5415.

[25] Rachman MA, Nakashimada Y, Kakizono T, Nishio N. Hydrogen

production with high yield and high evolution rate by self-flocculated

cells of Enterobacter aerogenes in a packed-bed reactor. Appl Mi-

crobiol Biotechnol 1998;49:4504.[26] Chang FY, Lin CY. Biohydrogen production using an up-flow anaer-

obic sludge blanket reactor. Int J Hydrogen Energy 2004;29:339.

[27] Chen CC, Lin CY, Lin MC. Acid-base enrichment enhances

anaerobic hydrogen production process. Appl Microbiol Biotechnol

2002;58:2248.

[28] APHA. Standard Methods for the examination of water and wastew-

ater. New York American Public Health Association. 1995.

[29] Yu HQ, Tay JH, Fang HHP. The roles of calcium in sludge granu-

lation during UASB reactor start-up. Water Res 2001;35:105260.

[30] Kosaric N, Blaszczyk R. Microbial aggregates in anaerobic wastew-

ater treatment. Adv Biochem Eng Biotechnol 1990;42:2762.

[31] Morgan JW, Evison LM, Forster CF. Changes to the microbial ecol-

ogy in anaerobic digesters treating ice cream wastewater during start-

up. Water Res 1991;25:63953.

[32] Jeison D, Chamy R. Comparison of the behaviour of expandedgranular sludge bed (EGSB) and upflow anaerobic sludge blanket

(UASB) reactors in dilute and concentrated wastewater treatment.

Water Sci Technol 1999;40:917.

[33] Lee KS, Wu JF, Lo YS, Lo YC, Lin PJ, Chang JS. Anaerobic

hydrogen production with an efficient carrier-induced granular sludge

bed bioreactor. Biotechnol Bioeng 2004;87(5):64857.

[34] Chen WM, Tseng ZJ, Lee KS, Chang JS. Fermentative hydrogen

production with Clostridicem butyricum CGS5 isolated from anaer-

obic sewage sludge. Inter J Hydrog Energy 2004, in press.

Documents

Shing Lee 2006 EMT