Bio H Producere

7232019 Bio H Producere

httpslidepdfcomreaderfullbio-h-producere 111

Biohydrogen production from food waste by

coupling semi-continuous dark-photofermentation

and residue post-treatment to anaerobic digestion

A synergy for energy recovery

Anish Ghimire a Serena Valentino b Luigi Frunzo c Eric Trably dRenaud Escudie d Francesco Pirozzi b Piet NL Lens e Giovanni Esposito a

a Department of Civil and Mechanical Engineering University of Cassino and Southern Lazio via Di Biasio 43

03043 Cassino FR Italyb Department of Civil Architectural and Environmental Engineering University of Naples Federico II via Claudio 21

80125 Naples Italyc Department of Mathematics and Applications Renato Caccioppoli University of Naples Federico II via Cintia

Monte S Angelo I-80126 Naples Italyd INRA UR0050 Laboratoire de Biotechnologie de l Environnement F-11100 Narbonne Francee UNESCO-IHE Institute for Water Education Westvest 7 2611 AX Delft The Netherlands

a r t i c l e i n f o

Article history

Received 3 July 2015

Received in revised form

10 September 2015

Accepted 27 September 2015

Available online xxx

Keywords

Biohydrogen

Food waste

Dark fermentation

Photofermentation

Anaerobic digestion

a b s t r a c t

This study aimed at maximizing the energy yields from food waste in a three-step con-

version scheme coupling dark fermentation (DF) photofermentation (PF) and anaerobic

digestion (AD) Continuous H2 production was investigated over a period of nearly 200 days

in a thermophilic semi-continuous DF process with no pH control The highest H 2 yield of

12145 plusmn 4455 N L H2 kg VS was obtained at an organic loading rate of 25 kg VSm3 d and a

hydraulic retention time of 4 days The DF effluents mainly contained volatile fatty acids

(VFAs) and alcohols as metabolites and un-hydrolyzed solid residues The supernatant

after separation was used to recover H 2 in a PF using Rhodobacter sphaeroides The solid

residual fraction along with PF effluent was converted into methane by anaerobic diges-

tion By combining DF and PF the H2 yield from the food waste increased 175 fold

Moreover by adding AD as a post treatment of the DF residue the total energy yield was

substantially increased to reach 555 MJkg VSfood waste added versus 355 MJkg VSfood waste

when applying solely AD

Copyright copy 2015 Hydrogen Energy Publications LLC Published by Elsevier Ltd All rights

reserved

Corresponding author Tel thorn39 081 7683 436 fax thorn39 081 5938 344E-mail address luigifrunzouninait (L Frunzo)

Available online at wwwsciencedirectcom

ScienceDirect

j o u r n a l h o m e p a g e w w w e l s e v i er c o m l o c a t e h e

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1e1 1

httpdxdoiorg101016jijhydene201509117

0360-3199Copyright copy 2015 Hydrogen Energy Publications LLC Published by Elsevier Ltd All rights reserved

Please cite this article in press as Ghimire A et al Biohydrogen production from food waste by coupling semi-continuous dark-photofermentation and residue post-treatment to anaerobic digestion A synergy for energy recovery International Journal of Hydrogen Energy (2015) httpdxdoiorg101016jijhydene201509117



Introduction

The inherent characteristics of hydrogen (H2) such as higher

energy content (142 MJ per kg) energy and water as the only

by-products generated from its combustion application in

fuel cells for electricity generation and the ability to be pro-

duced biologically makes H2 a very interesting alternativefuture sustainable energy carrier [1] Among several biological

technologies proposed for H2 production dark fermentation

(DF) is emerging as one of the prominent options shown by

the increasing research interests in this technology [2] The

advantages such as the flexibility to operate under different

conditions of temperature and pressure higher production

rates possibility to use renewable waste biomass as feedstock

and the treatment capability make the DF process attractive

Waste biomass such as agricultural residues the organic

fraction of municipal solid waste (OFMSW) and agro-

industrial wastes are economically competitive when

considering a supply of sustainable feedstock aiming at the

industrial development of DF systems for biological treatmentof waste [3e5]

OFMSW which is mainly composed of food waste (FW) has

been receiving a lot of attention because of its potential to be

used for the production of biofuels and other value added

products [6] Especially about 13 billion tonnes of food per

year get wasted which is approximately one-third of the food

produced for human consumption [7] FW is generated from

agricultural production industrial manufacturing processes

and final consumption in households In the European Union

the total annual generation of FW is estimated around 893

million tonnes comprising 377 million tonnes generated

from household consumption alone [8] The volatile solids

content in FW ranges from 21 to 27 which shows its highorganic carbon contentthat can be furthervalorized [9]andin

particular for H2 production by DF as demonstrated in the

literature [10e15] Some studies have reported the operational

feasibility of continuous H2 production using food or kitchen

wastes as a feed in DF processes [101416]

With the advantage of a stable operation continuous DF

processes are usually preferred and scaling-up is more viable

in comparison to batch processes which involve regular

downtime periods of maintenance [17] However stable

operation of continuous DF of FW is highly influenced by

bioreactor operating parameters such as pH temperature

organic loading rates (OLRs) and hydraulic retention times

(HRTs) [4518] These factors also influence the microbialcommunities and thus the biochemical pathways that can

affect the total H2 yields in mixed cultures [19] In addition

there is growing interest in coupling DF either with photo-

fermentation (PF) [2021] or bioelectrochemical systems (BES)

[22] to obtain higher overall H2 yields or with anaerobic

digestion (AD) for methane production [23e25] due to the

post-treatment requirement of DF effluents (DFEs) and net

positive energy gain from coupling these bioprocesses [26]

H2 production rates and total H2 yields are mainly a func-

tion of substrate types and OLRs applied [2] A varying rangeof

optimal OLR values has been reported for dark fermentative

H2 conversion from FW carried out in thermophilic DF pro-

cesses [2] Shin et al [27] found an optimal H2 yield of 12625 L

H2 kg VS at an OLR of 8 kg VSm3 d while the H2 production

decreased when the OLR was increased to 10 kg VSm 3 d The

authors reported 8 kg VSm3 d 5 days and a pH of 55

respectively as optimal OLR HRT and culture pH In a study

coupling DF and AD Cavinato et al [10] reported 667 L H2 kg

VS added at an optimum OLR of 163 VSm3 d a HRT of 33

days and for a pH maintained in the range of 5e6 through the

recirculation of AD effluent Generally HRTs in a range of 2e

6days have been reported as optimum for DF of organic FW in a

CSTR process [2] This range ofHRTs is similar to the first stage

of a two-stage AD process [28] Moreover the HRT is also a

function of the substrate type and bioreactor operational

parameters

It has been well documented that dark fermentative H2

production is generally due to the conversion of the initial

soluble fraction of carbohydrates present in the complex

organic biomass that will lead to accumulation of volatile

fatty acids (VFAs) and alcohols in DFEs [2930] Some recent

studies have shown the potential of these DFEs to be utilized

in PF processes for H2 production [2021] Combining DF with

PF Su et al [31] achieved an increase in H2 yield from 767 to5961 L H2 kg VS from water hyacinth Meanwhile Rai et al

[20] achieved 43 higher volumetric H2 yields from acid hy-

drolyzed sugarcane bagasse in two step DF-PF systems

However during the conversion of complex organic biomass

like FW a part of the unhydrolyzed solid residues will remain

that can be further valorized in AD systems producing

methane (CH4) in a three steps conversion scheme (Fig 1) Xia

et al [3233] reported that a three-step conversion of algal

biomass combining DF-PF-AD can achieve 17 and 13 times

higher energy yields in comparison to a two-stage DF-AD and

an one stage AD process respectively

High OLRs are often responsible for a decrease in culture

pH due to the accumulation of VFAs present in DFE Thusmost of the continuous DF systems utilizing acidic substrates

such as food waste requires constant addition of external

alkalinity sources such as alkaline chemicals (NaOH or KOH)

or buffering agents (bicarbonate or phosphate buffers)

[142734] A long-term study of continuous H2 production at

varying operating conditions of OLR and HRT to establish a

long-term operability for continuous H2 production in relation

with the production of metabolites could provide further in-

sights for the development of scaled-upDF systems Similarly

a three-step conversion process (DF PF and AD) might

contribute to an increase in overall energy yield and could

provide the biological treatment to the by-products generated

from DF systemsThis study aims to demonstrate the long-term operational

feasibility of continuous H2 production from FW using a semi-

continuous thermophilic DF reactor at various low OLRs and

HRTs without pH control The experiment also aimed at

reducing the dependency on chemical buffering agents that

are used to maintain the culture pH at working conditions H2

production through different possible biochemical pathways

was discussed in relation to major metabolites present in

DFEs obtained during the varying experimental conditions

The potential of coupling DF with photofermentative H 2 pro-

duction was investigated in batch PF experiments by using the

liquid fraction of the DFE after physical separation Further

the waste streams generated from the coupling of DF-PF were





utilized in AD to maximize the energy yields and provide in-

tegrated waste treatment solutions

Materials and methods

Preparation of feedstock

An average composition of waste as found in European

countries was prepared as cited elsewhere [9] The waste

mixture was prepared at the laboratory and was composed of

(in by wet-weight) fruit and vegetables 72 cooked pasta

and rice 10 bread and bakery 5 dairy products (cheese) 2meat and fish 8 and snacks (biscuits) 3 TheFW ingredients

were freshly bought at municipal markets in Naples (Italy)

shredded with a blender and immediately stored at 20 C to

avoid acidification The FW characteristics were (in gkg FW)

chemical oxygen demand (COD) 3476 plusmn 474 carbohydrate

content 10580 plusmn 07 total Kjeldahl nitrogen (TKN) 64 plusmn 018

lipids 1750 plusmn 119 total solids (TS) 2379 plusmn 044 volatile

solids (VS) 228 plusmn 042 and the pH was 44 plusmn 01

DFE was collected from the outlet of the fermenter and had

a pH of 45 plusmn 01 After undergoing settling for 30 min and

centrifugation at 4500 rpm for 20 min the supernatant was

collected The DFE characteristics are presented in Table 1

The DFE was supplemented with KH2PO4 3 gLNaHCO307gL ferric citrate 245 mgL and 10 mL of a trace metals solution

(for composition see below) pH was adjusted to 65 and then

the DFE medium was autoclaved at 121 C for 20 min

The solid residues left after settling and centrifugation of

DFE along with the PF effluents mainly containing photo-

fermentative biomass were used as feed for AD The charac-

teristics of the solid residues generated from solideliquid

separation were comprised of undigested FW which had a pH

of 45 plusmn 01 and solid DF residue with a content of COD

264 plusmn 04 gkg residue TS 242 plusmn 002 and VS 231 plusmn 002

The PF effluent had a pH of 726plusmn 001 and contained a soluble

COD of 14077 plusmn 109mgLwith071plusmn 001TS and 028plusmn 001

VS contents

Experimental setup and operational conditions

Dark fermentation bioreactor

Anaerobic digested sludge was collected from an anaerobic

digestion plant of the farm ldquoLa Perla del Mediterraneordquo

(Campania Italy) The sludge was used as start-up seed

inoculum after thermal pretreatment at 105 C for 4 h to

enrich the microbial consortia of H2 producers like spore

forming Clostridia and to inhibit the methanogens [35] The

inoculum had (in gL) TS 2954 plusmn 022 VS 1836 plusmn 014

ammonium (NHthorn4 ) 028 plusmn 0011 total alkalinity (as CaCO3)

144 plusmn 0014 and had a pH of 83 plusmn 01

A continuously stirred serum bottle of 1500 ml working volume was used as DF bioreactor which was maintained at a

constant thermophilic temperature (55 plusmn 2 C) The reactor

was started with an initial SX ratio (substrate to inoculum

ratio as gVS substrategVS inoculum) of 05 and operated in

semi-continuous mode with three different HRTs and four

OLRs in six different operational conditions (Table 2) The pH

of the initial feed (45plusmn 01)was adjusted manually to an initial

pHof 70 with1 M NaOH The culture pHin the reactor was not

adjusted allowing the digesting mixture to reach indigenous

chemical equilibrium

Fig 1 e Schematic of the three-stage conversion of FW to hydrogen and methane

Table 1 e Characteristics of the DFE used in PFexperiments

Parameters Values (mgL)

Chemical Oxygen Demand (COD) 35618 plusmn 1311

TKN 2080 plusmn 7

NHthorn4 114 plusmn 03

Phosphate (PO34 ) 1305 plusmn 1

Total iron (Total-Fe) 07

Lactic Acid 330

Acetic Acid 4660

Propionic Acid 4496

Butyric Acid 10754

Ethanol 3230

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1e1 1 3




Effluent and gas samples from the reactor were analyzed

daily for determining the major metabolic intermediates ie

acetate propionate butyrate lactate ethanol and the gas

composition (H2 and CO2) The total gas volumewas measured

by volumetric water displacement The gas was passed

through acidic water (15 HCl) and the volume of water dis-

placed corresponded to the volume of total gas produced The

volume of hydrogen produced was calculated by considering

this volume and the gas composition and was then normal-

ized for standard conditions

Photofermentation bioreactorRhodobacter sphaeroides AV1b (kindly provided by professor

Roberto De Philippis University of Florence Italy) was pre-

viously isolated from the Averno lake in Naples (Italy) as

described elsewhere in Bianchi et al [36] and was used as

inoculum for PF R sphaeroides AV1b was first grown in a

medium as previously described by Bianchi et al [36]

which was composed of (in gL) DL-malic acid 2 sodium

glutamate 17 K2HPO4 05 KH2PO4 03 MgSO47H2O 04

NaCl 04 CaCl22H2O 0075 ferric citrate 0005 yeast

extract 04 and 10 mL of trace metals solution containing (in

mgL) ZnSO47H2O 10 MnCl24H2O 3 H3BO3 30

CoCl26H2O 20 CuCl22H2O 1 NiCl26H2O 2 and

Na2MoO42H2O 30The R sphaeroides AV1b pre-culture was grown again in a

DFE supplemented with appropriate chemicals and auto-

claved as explained in preparation of feedstock It was mainly

composed of (in mgL) acetic acid 848 propionic acid 457

butyric acid 1184 NHthorn4 6 phosphate (as PO3

4 ) 358 and total

Fe 0045 Ten mL of the culture (152 g TSSL) that represents

25 VVof the reactor working volume was used as inoculum

in the PF experiments with DFE (Table 1)

Transparent 500 mL borosilicate serum glass bottles

(Simax Czech Republic) with 400 mL working volume were

used as photofermentative batch reactor The batch reactors

were maintained at room temperature (24 plusmn 2 C AprileMay)

under a luminance of about 4000 Lux and positioned on thetop of the stirrers Caps of the reactors presented two separate

ports for biogas and culture medium sampling The bottles

were sealed with silica and flushed with argon to ensure

anaerobic conditions and eliminate the nitrogen gas (N2) from

the headspace since N2 can inhibit the activity of the nitro-

genase enzyme responsible for photofermentative H2 pro-

duction [37] The H2 production was quantified as described in

DF bioreactor setup

AD of residues from DF-PF process

A batch test was carried out in 1 L transparent borosilicate

serum glass bottles (Simax Czech Republic) and was main-

tained at 34 plusmn 1 C in a water bath The working volume of the

reactor was 600 mL with an initial SX ratio of 05 with a

substrate concentration of 45 g VSL A low SX ratio 05 was

selected to assess the biomethane potential of the feed used

Based on the substrate type a range of SX ratio 05e23 gVS

substrategVS inoculum is suggested to prevent the acidifi-

cation of the AD reactor [38] The source of inoculum used in

the tests was the same as the start up inoculum used in the

semi-continuous DF reactor The characteristics of the inoc-ulum were (in gL) TS 2371 plusmn 017 VS 1455 plusmn 011 ammo-

nium (NHthorn4 ) 046 plusmn 002 and had a pH 82 plusmn 01 The tests were

carried out in duplicates

Analytical methods

Hydrogen was quantified with a Varian Star 3400 gas chro-

matograph equipped with a ShinCarbon ST 80100 column

and a thermal conductivity detector Argon was used as the

carrier gas with a front and rear end pressure of 20 psi The

duration of analysis was 14 min The fermentation products

(lactic acetic propionic and butyric acids) were quantifiedby High Pressure Liquid Chromatography (HPLC) (Dionex LC

25 Chromatography Oven) equipped with a Synergi 4u Hydro

RP 80A (size 250 460 mm) column and UV detector (Dionex

AD25 Absorbance Detector) The gradient elution consisted

of 20 methanol and 10 acetonitrile in 5 mM H 2SO4 pum-

ped at a rate of 09 mLmin using a Dionex GP 50 Gradient

pump The elution time was 185 min Ethanol was quanti-

fied by HPLC Aminex HPX-87H column (300 mm on 78 mm

Bio-rad) using 5 mM H2SO4 as an eluent The COD of the FW

was measured as described elsewhere [39] The

carbohydrate content was determined according to the

Dubois method [40] Total lipids were measured following a

Bligh and Dyer chloroformmethanol total lipid extractionmethod [41] The light intensity was measured with a light

meter (Lutron-LX-107) The TS and VS of the seed sludge and

TKN were determined according to the Standard Methods

[42]

Data analysis

Hydrogen production rates (HPR) were expressed in L H2 m3

d while the H2 yields (HY) were determined considering the

total daily organic load fed to the reactor and expressed as L

H2 kg VS added Average and deviations for daily production

were determined during the steady state reached after 3e

4days of operation The H2 Production Stability Index (HPSI)

was evaluated by considering the ratio of standard deviation

and average HPR as reported by Tenca et al [16]

HPSI frac14 1 SDethHPRTHORN

AvgHPR (1)

A HPSI index closer to 1 represents a stable hydrogen

production

FactoMineR an extension on R software was used for

multivariate analysis of the metabolite distribution from the

different experimental periods in relation to the hydrogen

yields and co-relation circles of the major metabolites were

generated

Table 2 e Experimental design used for the operation of semi-continuous reactor

Experimental periods I II III IV V VI

OLR (kg VSm3 d) 1 1 15 2 2 25

HRT (d) 12 6 6 6 4 4

Concentration (kg VSm3) 12 6 9 12 8 10





Results and discussion

Continuous dark fermentative biohydrogen production

Effect of operational parameters on H2 production rate and yield

The results in terms of H2 yields (HY) hydrogen production

rates (HPR) and H2 Production Stability Index (HPSI) during

the different OLRs and HRTs investigated in the six operationperiods (Table 2) are summarized in Table 3 Fig 2 shows the

HPR (a) and pH trends (b) over the operation period of 193

days The results show an increase in HPR when OLRs were

increased During the operating periods II III and IV at a

constant HRT of 6 days the HPR increased from 541 plusmn 41 to

1095 plusmn 33 and 2102 plusmn 30 N Lm3 d when the OLR was

increased from 1 to 15 and 2 kg VSm 3 d respectively (Tables

2 and 3) Meanwhile the overall HY increased from

541 plusmn 413 N Lkg VSadded to 1051 plusmn 149 N L H2 kg VSadded

During the experimental period IV the H2 production had a

comparatively better stability as shown by a HPSI of 086

However no significant effect was observed on the total HY

and HPR when the HRT changed to 4 days during operationalperiod V (Table 3) When the OLR was changed from 2 to

25 kg VSm3 d during period VI both HY and HPR increased

However the H2 production was not stable supported by a

lower value of HPSI of 063 This instability could be

explained by the accumulation of acids and a subsequent

decrease in pH to 44 plusmn 01 which might have affected the

microbial community

During a short operation period (at the end of period IV)

the culture pH inside the reactor was regulated manually to an

initial culture pH 55 with 1 M NaOH during feeding with the

objective to assess the influence of pH on the H2 productionperformance (Fig 2b) However pH regulation did not show

any effect on the HPR (Fig 2a) Nevertheless the increased

HPSI (Table 3) showed that H2 production was stable during

that period in comparison to the experimental period when

the culture pH was uncontrolled The percentage of H2 and

CO2 in the gas averaged 59 plusmn 6 and 39 plusmn 6 respectively

when the H2 production stabilized However the H2 produc-

tion performances in experimental period IV (HPR

2102 plusmn 298 N Lm3 d and HY 1051 plusmn 149 N Lkg VSadded at a

HRT of 6 days and OLR 2 g VSLm3 d) were comparable to

experimental period V (HPR 2080 plusmn 348 N Lm3 d and HY

1040 plusmn 174 N Lkg VSadded at a HRT of 4 days and OLR of 2 kg

VSm3 d) Thus the operational conditions of period V wereconsidered as ideal for the DF of FW in thermophilic semi-

continuous reactors as a lower HRTs are generally more

economically efficient in terms of bioreactor design and

operation

A comparison of previous studies on dark fermentative H2

production from FW with the results from this study (Table 4)

suggests that comparable results in terms of H2 production

can be achieved even at low OLRs and without pH control

Nonetheless the characteristics of FW can also affect the

overall HY as H2 production is mainly function of the soluble

fraction of carbohydrates present in the substrate [30] The

OLRs reported in the past studies were higher than in this

study and thus a source of alkalinity to balance the pH con-ditions at optimum was required Valdez-Vazquez et al [14]

used NaHCO3 and K2HPO4 to maintain the optimum pH at

64 while Lee et al [43] used NaOH and H3PO4 to maintain the

Table 3 e H2 production rate yields and productionstability from FW by mixed anaerobic cultures

Exp Period HPR(N Lm3 d)

HY(N Lkg VSadded)

H2 inbiogas ()

HPSI

I 1169 plusmn 401 1169 plusmn 401 528 plusmn 1 066

II 541 plusmn 413 541 plusmn 413 312 plusmn 1 024

III 1095 plusmn 328 730 plusmn 219 438 plusmn 20 070

IV 2102 plusmn 298 1051 plusmn 149 594 plusmn 6 086

V 2080 plusmn 348 1040 plusmn 174 572 plusmn 6 083

VI 3036 plusmn 1114 1214 plusmn 445 558 plusmn 10 063

Fig 2 e HPR (L H2 m3 d) (a) and pH trends in semi-continuous thermophilic reactor (b) shaded region represents the

experimental period when the culture pH inside the reactor was adjusted daily to pH 55 during the feeding operation





culture pH at 6 Thus this pH decrease resulting from the

production of acids can be minimized by the use of lower

OLRs Higher OLRs can exert detrimental effects on the mi-crobial community and thus H2 production by decreasing the

pH due to the accumulation of metabolites [44]

Metabolic intermediates

Lactate acetate propionate butyrate and ethanol were the

main metabolic intermediates observed during the different

experimental periods Such a mixture of intermediates is

characteristic of mixed fermentation pathways occurring

with complex substrates [30] Average concentrations of the

main metabolites during the six different experimental pe-

riods are summarized in Table 5 There can be a number of

possible H2 production pathways during mixed type fermen-

tation as represented by equations (2)e

(5) (Table 6) whereasH2 consuming or unfavorable pathways presented in equa-

tions (6)e(9) might exist at the same time [1719] The presence

of ethanol acetate and butyrate are evidences for the pres-

ence of an ethanol-acetate or butyrateeacetate pathway for

H2 production in the DF of the FW investigated On the other

hand the presence of lactate or propionate can be attributed

to fluctuations in H2 production resulting in low H2 yields

Fig 3 shows the plot of correlation circles of the five major

metabolites and the HY Fig 3(a) shows that the butyrate and

acetate concentration is well correlated with the HY values

Not surprisingly propionate lactate and ethanol are in the

Dim 2 and are not correlated with the HY which is supported

by equations (6)e

(9) (Table 6) in a DF with glucose as modelsubstrate However the pathways leading to ethanol-acetate

also yield H2 as shown in Equation (4) [5051] Nonetheless

Fig 3 shows that the ethanol is not correlated with acetate

Therefore most of the H2 yields can be attributed to the

butyrateeacetate pathways which showed a good correlation

and is explained in Dim 1 The variable Dim 3 is mostlyexplained by lactate concentrations (Fig3 b) which correlated

oppositely with HY and is an orthogonal and independent

variable The proximity of butyrate ethanol and propionate

suggests that these metabolites can be expected from DF by

mixed microbial consortia This is also supported in a study by

Hwang et al [50] who obtained butyrate ethanol and propio-

nate as the major metabolites during the DF at a pH range of

4e45 45e50 50e6 respectively

Photofermentative H2 production from the liquid fraction of

DF

The DFE from the semi-continuous DF reactor obtained during

experimental period VI was further converted to H2 by R

sphaeroides AV1b in a PF process Cumulative H2 production

and VFA consumption trends during the PF experiments are

shown in Fig 4(a) and (b) respectively VFA and ammonium

concentrations in the DFE medium (shown in Table 2) were

both at non-inhibiting levels for photofermentative H2 pro-

duction Han et al [52] reported that concentrations equal to

98 mM 109 mM and 42 mM respectively for acetate buty-

rate and propionate gave the optimum H2 yield using R

sphaeroides However concentrations up to 30 mM of acetate

have been reported by Hustede et al [53] Similarly the

ammonium concentration was at non-inhibitory levels asonly a concentration higher than 2e5 mM of NHthorn

4 has been

reported to inhibit the photofermentative H2 production

[5455]

Table 4 e Comparison of dark fermentative H2 production using FW by anaerobic mixed cultures

Substrate type Reactor ype T (C) pH OLR(kg VSm3

$d)Maximum assessed H2

yield (N L H2 kg VSadded)H2 in

biogas ()Reference

FW Batch 55 45 (initial) 6 463 23 [45]

Vegetable kitchen

waste

Intermittent-CSTR 55 60 28a 381b 40 [43]

FW and sewage sludge Batch 35 50e60 e 1229ce [46]

OFMSW (FW thorn paper) Semi-continuous CSTR 55 64 11d 360 58 [14]

OFMSW Packed bed reactor 38 plusmn 2 56 plusmn 02 16e 99 47 [47]

FW Semi-continuous CSTR 55 plusmn 2 47 plusmn 02 2 1040 plusmn 174 572 (plusmn6) This study

FW frac14 food waste OFMSW frac14 organic fraction of municipal solid wastea gCODLdb mL H 2 g CODc mL H 2 g carbohydrate CODd g VSkg wet mass reactorde g VSkgd

Table 5 e Characteristics of influent and effluents from DF of FW during different experimental periods

Exp Period pH_IN pH_OUT Lactate (mM) Ethanol (mM) Acetate (mM) Propionate (mM) Butyrate (mM)

I 700 47 plusmn 03 01 plusmn 02 48 plusmn 02 131 plusmn 36 385 plusmn 221 104 plusmn 28

II 700 45 plusmn 01 06 plusmn 14 54 plusmn 35 32 plusmn 20 344 plusmn 233 62 plusmn 42

III 700 45 plusmn 02 40 plusmn 91 87 plusmn 27 49 plusmn 06 597 plusmn 216 110 plusmn 16

IV 700 49 plusmn 04 00 plusmn 00 172 plusmn 86 85 plusmn 18 965 plusmn 291 120 plusmn 29

V 700 47 plusmn 02 00 plusmn 00 171 plusmn 66 67 plusmn 19 570 plusmn 215 99 plusmn 32

VI 700 44 plusmn 01 05 plusmn 09 94 plusmn 53 57 plusmn 28 589 plusmn 270 111 plusmn 75





The PF of spent DFE yielded a cumulative production of

3656 plusmn 32 NmL H2 corresponding to a volumetric yield of

914 plusmn 8 N L H2 m3 and a substrate yield of 427 plusmn 6 N L H2 kg

COD consumed The batch experiments were carried out for

40 days until the H2 production completely ceased (Fig 4(a))

This is longer than any H2 production time reported else-

where [2033] The long lag phase (9 days) can partly explain

this result The final effluents were analyzed for COD VFAs

and biomass concentration which showed a COD reduction

of 601 while more than 98 plusmn 1 of VFAs were removed to

reach a final biomass concentration of 16 g TSSL Theoret-

ical COD removal calculated from the VFA concentration in

final effluents showed a COD removal efficiency of 992

However the production of biomass and other bacterial ca-

rotenoids increased the final total COD of the PF effluent and

thus reduced the total COD removal efficiency This was

evident by the reddish brown color of the effluent The

maximum percentage of H2 in the biogas was 89 with 89

of CO2

The volumetric H2 production obtained in this study

(914 plusmn 8 N L H2 m3) is higher than the study of Rai et al [20]

using Rhodopseudomonas BHU 01 with a volumetric H 2 yield

of 755 L H2 m3 In another study by Uyar et al [56] using

Rhodobacter capsulatus (DSM 155) as biomass and DFE of Mis-

canthus hydrolysate as substrates a volumetric yieldof 1000 L

H2 m3 was obtained which is slightly higher than in this

study The present study showed the potential of an inte-

grated DF-PF system to achieve higher H 2 yields Thus the

combined DF-PF processes can help in the industrial devel-

opment of DF processes using FW The residues generated

from the downstream of these processes can nevertheless

still be treated with anaerobic digestion in order to provide

additional conversion of organic matter to further recover

energy

AD of DF-PF waste stream

The solid residues generated by the coupled DF-PF process can

be ideal for AD as the undigested FW residues from the DF

process and the PF effluent containing biomass generated

from the PF can be converted to methane in a biorefinery

model (Fig 1) The result of the average cumulative methane

Table 6 e Reaction stoichiometry in DF of glucose

Possible H2 producing pathways Metabolic pathway DG00

a (kJmol) Eqn

C6H12O6 thorn 2H2O 2CH3COOH thorn 2CO2 thorn 4H2 Acetate 2063 (2)

C6H12O6 CH3CH2CH2COOH thorn 2CO2 thorn 2H2 Butyrate 2548 (3)

C6H12O6 thorn 2H2O CH3CH2OH thorn CH3COOH thorn 2CO2 thorn 2H2 Ethanol amp acetate 2157 (4)

4C6H12O6 thorn 2H2O 3CH3CH2CH2COOH thorn 2CH3COOH thorn 8CO2 thorn 10H2 Butyrate amp acetate 2540 (5)

Unfavorable and H2 consuming pathways

C6H12O6 thorn 2H2 2CH3CH2COOH thorn 2H2O Propionate 3596 (6)

15C6H1206 2C2H5COOH thorn CH3COOH thorn CO2 thorn H2O Propionate amp acetate 3100 (7)

C6H12O6 2CH3CH2OH thorn 2CO2 Ethanol 2350 (8)

C6H12O6 2CH3CHOHCOOH Lactate 1981 (9)

aDG0

0 values are adapted from Refs [4849]

Fig 3 e Correlation circle of five metabolites and HY formed by the first three principle components Dim1 Dim 2 and Dim 3

representing 3500 1803 and 1654 of the total variance respectively Projections according to the first two (Dim 1 and Dim

2) (a) and first and third factors (Dim 1 and Dim 3) (b)





production trends during the biomethane potential test using

the waste stream generated from the DF-PF process is pre-

sented in Fig 5 The cumulative CH4 production stabilized

after 50 days and the average cumulative CH4 production was871 plusmn 16 mL corresponding to a total average yield of

324 plusmn 6 N L CH4 g VS added (feed) and 09 kg CODkg VS

removed (calculated from CH4 produced) evaluated after

subtracting the endogenous methane produced in the con-

trols The initial and final average pH in the BMP tests was 70

and 77 respectively while the pH of the dark fermentation

and photofermentation residues were respectively 433 and

726 The pH was not adjusted with a buffering agent because

the alkalinity of the inoculum was sufficient to maintain the

pH this further adds practicability to AD as a post-treatment

option

Energy yields from gas biofuels produced from food waste

When considering the conversion of the initial VS added at the

beginning of the DF process the overall average H2 yield from

coupling ofthe DF-PF process was increasedfrom 1051 N L H2

kg VSinitial to 1843 N L H2 kg VSinitial with an additional

792NLH2 kg VSinitial fromPFand993NLCH4 kg VSinitial from

AD The increase in energy yields obtained in this study wascompared with energy yields from the coupled process pre-

viously reported in the literature (Table 7) The energy yields of

hydrogen and methane from the stand alone DF as well as the

two stage DF-PF and DF-AD was calculated based on the

heating values of H2 (242 kJmol) and methane (801 kJmol)

These calculated energy yields represent the energy gain from

the conversion of substrates by biological processes However

the net energy gain can be estimated by considering the en-

ergy input in the processes which is not representative in lab

scale reactors and thus not calculated in this study

By coupling DF with PF and AD processes an additional

44 MJkg VS of energy yield can be achieved from food waste

which is higher than the coupled DF - AD process or standalone DF processes (Table 7) Out of the overall energy recov-

ered fromthe three-stage conversion (DF-PF-AD) of food waste

H2 contributes only 358 out of 555 MJkg VS However this

may be a positive add-on to the overall economic return

compared to CH4 productivity only Therefore the three-step

process can definitely increase the recovered energy yield

Moreover it is a very good solution for waste treatment as a

higher FW conversion was accomplished Table 7 shows that

the energyyieldof DFand PFfrom the study ofZong et al [57] is

higher than the energy yield reported in this study This is

likely because of the difference in H2 yield achieved in these

studies In other studies by Xia et al[3233] and Wang etal [58]

although the overall energy yields obtained from the respec-tive three and two step conversion were high the pre-

treatment of the substrate required an energy input There-

fore the overall energy yields obtained from the coupling of

various processes depends on the H2 and CH4 yields and pro-

duction rates in individual processes which are mainly a

function of process operational conditions such as pH tem-

perature HRT and OLR as well as carbohydrate content and

nature of the feedstock Moreover the coupling of the PF and

AD processes in the downstream process is not only advan-

tageous from the energy point of view but it also provides

biological treatment of the waste stream generated by the DF

processes (COD and pathogen removal) [59]

Conclusion

This study has shown the long-term feasibility of continuous

H2 production as well as the possibility to further recover

energy through integration of PF and AD using FW as the

substrate In addition the viability of H2 production at low

OLRs without the culture pH control can minimize the

excessive use of chemical buffering agents for pH control The

integration of DF with PF can increase the overall H2 yield 175

fold On the other hand applying AD for the post treatment of

waste streams generated by the coupling of the DF-PF

Fig 4 e Cumulative hydrogen production (a) and depletion

of major VFAs (acetate propionate and butyrate) (b) in PF

tests using DFE and R sphaeroides AV1b

Fig 5 e Methane yields from mesophilic AD of waste

stream generated in the coupled DF-PF processes





processes can further increase the overall energy yield by

555 MJkg VS of food waste adding a synergistic effect to the

overall energy recovery during the conversion of food waste

Acknowledgments

The authors would like to thank Prof Roberto De Philippis of

University of Florence (Italy) for providing the purple non

sulfur bacteria strains The authors would also like to

acknowledge the Erasmus Mundus Joint Doctorate Pro-

gramme ETeCoS3 (Environmental Technologies for Contami-

nated Solids Soils and Sediments) under the EU grant

agreement FPA No 2010-0009 This research was further sup-

ported by the project ldquoModular photo-biologic reactor for bio-

hydrogen application to dairy waste e RE-MIDArdquo from the

Agriculture Department of the Campania Region in the

context of the Programme of Rural Development 2007e2013

Measure 124

r e f e r e n c e s

[1] Kotay SM Das D Biohydrogen as a renewable energyresource -Prospects and potentials Int J Hydrogen Energy200833258e63 httpdxdoiorg101016 jijhydene200707031

[2] Ghimire A Frunzo L Pirozzi F Trably E Escudie R Lens PNLet al A review on dark fermentative biohydrogen productionfrom organic biomass process parameters and use of by-products Appl Energy 201514473e95 httpdxdoiorg101016japenergy201501045

[3] Chong M Sabaratnam V Shirai Y Ali M Hassan MABiohydrogen production from biomass and industrial wastesby dark fermentation Int J Hydrogen Energy2009343277e87 httpdxdoiorg101016 jijhydene200902010

[4] Ntaikou I Antonopoulou G Lyberatos G Biohydrogen

production from biomass and wastes via dark fermentation

a review Waste Biomass Valorization 2010121e39 httpdxdoiorg101007s12649-009-9001-2

[5] De Gioannis G Muntoni A Polettini A Pomi R A review of dark fermentative hydrogen production from biodegradablemunicipal waste fractions Waste Manag 2013331345e61httpdxdoiorg101016jwasman201302019

[6] Uckun Kiran E Trzcinski AP Ng WJ Liu Y Bioconversion of food waste to energy a review Fuel 2014134389e99 httpdxdoiorg101016jfuel201405074

[7] Gustavsson J Cederbery C Sonesson U van Otterdijk RMeybeck A Global food losses and food waste-Extent causesand prevention Rome Food and Agriculture Organization of the United Nations 2011 Available from httpwwwfao

orgdocrep014mb060emb060epdf [accessed 100714][8] European Commission DG ENV Prepatory study on food

waste across EU 27-final report European Commision 2010Available from httpeceuropaeuenvironmenteussdpdfbio_foodwaste_reportpdf [accessed 100714]

[9] VALORGAS Compositional analysis of food waste from studysites in geographically distinct regions of Europe-valorisationof food waste to biogas 2010 Available from httpwwwvalorgassotonacukDeliverablesVALORGAS_241334_D2-1_rev[1]_130106pdf [accessed 12012013]

[10] Cavinato C Giuliano A Bolzonella D Pavan P Cecchi F Bio-hythane production from food waste by dark fermentationcoupled with anaerobic digestion process a long-term pilotscale experience Int J Hydrogen Energy 20123711549e55httpdxdoiorg101016jijhydene201203065

[11] Xiao L Deng Z Fung KY Ng KM Biohydrogen generationfrom anaerobic digestion of food waste Int J HydrogenEnergy 20133813907e13 httpdxdoiorg101016 jijhydene201308072

[12] Han SK Shin H Biohydrogen production by anaerobicfermentation of food waste Int J Hydrogen Energy200429569e77 httpdxdoiorg101016 jijhydene200309001

[13] Elbeshbishy E Hafez H Nakhla G Viability of ultrasonicationof food waste for hydrogen production Int J Hydrogen Energy2012372960e4 httpdxdoiorg101016 jijhydene201101008

[14] Valdez-vazquez I Riosleal E Esparzagarcia F Cecchi FPoggivaraldo H Semi-continuous solid substrate anaerobicreactors for H2 production from organic waste mesophilic

versus thermophilic regime Int J Hydrogen Energy

Table 7 e Comparison of energy yields from gaseous biofuels produced out of FW as feedstock using stand alone orcoupling of different technologies

Feedstock Processtype H2 yield fromDFDF thorn PF

(N L H2 kg VS)

a Energy yieldfrom H2

(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference

FW thorn paper Semi-continuous DF 360 389 e 389 [45]

FW DF thorn PF (batch) 671b 725 e 725 [57]

Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]

FW DF thorn AD (batch) 55 060 94 396 [25]

N oceanica c DF thorn PF thorn AD

(batch)

1839 198 1613 774 [33]

C pyrenoidosa d DF thorn PF thorn AD

(batch)

1983 214 1862 666 [32]

FW Semi-continuous DF thorn PF

(batch) thornAD (batch)

184 199 993 555 This study

a The energy yield was calculated from the yield of biogas based on the heating values of hydrogen (242 kJmol) and methane (801 kJmol)b L H 2 kg food wastec Algal biomass pre-treatment by microwave heating with dilute H2SO4d Algal biomass pre-treatment by steam heating with dilute H 2SO4





2005301383e91 httpdxdoiorg101016 jijhydene200409016

[15] Sreela-or C Imai T Plangklang P Reungsang A Optimizationof key factors affecting hydrogen production from foodwaste by anaerobic mixed cultures Int J Hydrogen Energy20113614120e33 httpdxdoiorg101016 jijhydene201104136

[16] Tenca A Schievano A Perazzolo F Adani F Oberti R

Biohydrogen from thermophilic co-fermentation of swinemanure with fruit and vegetable waste maximizing stableproduction without pH control Bioresour Technol20111028582e8 httpdxdoiorg101016 jbiortech201103102

[17] Hawkes F Hussy I Kyazze G Dinsdale R Hawkes DContinuous dark fermentative hydrogen production bymesophilic microflora principles and progress Int JHydrogen Energy 200732172e84 httpdxdoiorg101016 jijhydene200608014

[18] Guo XM Trably E Latrille E Carrere H Steyer J-P Hydrogenproduction from agricultural waste by dark fermentation areview Int J Hydrogen Energy 20103510660e73 httpdxdoiorg101016jijhydene201003008

[19] Li C Fang HHP Fermentative hydrogen production from

wastewater and solid wastes by mixed cultures Crit RevEnviron Sci Technol 2007371e39 httpdxdoiorg10108010643380600729071

[20] Rai PK Singh SP Asthana RK Biohydrogen production fromsugarcane bagasse by integrating dark- and photo-fermentation Bioresour Technol 2014152140e6 httpdxdoiorg101016jbiortech201310117

[21] Chookaew T O-thong S Prasertsan P Biohydrogenproduction from crude glycerol by two stage of dark andphoto fermentation Int J Hydrogen Energy 20152e7 httpdxdoiorg101016jijhydene201502133

[22] Chookaew T Prasertsan P Ren ZJ Two-stage conversion of crude glycerol to energy using dark fermentation linkedwith microbial fuel cell or microbial electrolysis cell NBiotechnol 201431179e84 httpdxdoiorg101016

jnbt201312004[23] Wieczorek N Kucuker MA Kuchta K Fermentative hydrogen

and methane production from microalgal biomass (Chlorellavulgaris) in a two-stage combined process Appl Energy2014132108e17 httpdxdoiorg101016 japenergy201407003

[24] Pisutpaisal N Nathao C Sirisukpoka U Biological hydrogenand methane production in from food waste in two-stageCSTR Energy Procedia 201450719e22 httpdxdoiorg101016jegypro201406088

[25] Nathao C Sirisukpoka U Pisutpaisal N Production of hydrogen and methane by one and two stage fermentationof food waste Int J Hydrogen Energy 20133815764e9 httpdxdoiorg101016jijhydene201305047

[26] Ruggeri B Tommasi T Sassi G Energy balance of dark

anaerobic fermentation as a tool for sustainability analysisInt J Hydrogen Energy 20103510202e11 httpdxdoiorg101016jijhydene201008014

[27] Shin H-S Youn J-H Conversion of food waste into hydrogenby thermophilic acidogenesis Biodegradation 20051633e44

[28] Aslanzadeh S Rajendran K Taherzadeh MJ A comparativestudy between single- and two-stage anaerobic digestionprocesses effects of organic loading rate and hydraulicretention time Int Biodeterior Biodegrad 2014951e8 httpdxdoiorg101016jibiod201406008

[29] Monlau F Sambusiti C Barakat A Guo XM Latrille E Trably Eet al Predictive models of biohydrogen and biomethaneproduction based on the compositional and structuralfeatures of lignocellulosic materials Environ Sci Technol20124612217e25 httpdxdoiorg101021es303132t

[30] Guo XM Trably E Latrille E Carrere H Steyer J-P Predictiveand explicative models of fermentative hydrogen productionfrom solid organic waste role of butyrate and lactatepathways Int J Hydrogen Energy 2013391e10 httpdxdoiorg101016jijhydene201308079

[31] Su H Cheng J Zhou J Song W Cen K Hydrogen productionfrom water hyacinth through dark- and photo- fermentationInt J Hydrogen Energy 2010358929e37 httpdxdoiorg

101016jijhydene201006035[32] Xia A Cheng J Ding L Lin R Huang R Zhou J et al

Improvement of the energy conversion efficiency of Chlorella

pyrenoidosa biomass by a three-stage process comprising dark fermentation photofermentation andmethanogenesis Bioresour Technol 2013146436e43 httpdxdoiorg101016jbiortech201307077

[33] Xia A Cheng J Lin R Lu H Zhou J Cen K Comparison in darkhydrogen fermentation followed by photo hydrogenfermentation and methanogenesis between protein andcarbohydrate compositions in Nannochloropsis oceanica

biomass Bioresour Technol 2013138204e13 httpdxdoiorg101016jbiortech201303171

[34] Elsamadony M Tawfik A Potential of biohydrogenproduction from organic fraction of municipal solid waste

(OFMSW) using pilot-scale dry anaerobic reactor BioresourTechnol 20151969e16 httpdxdoiorg101016 jbiortech201507048

[35] Ghimire A Frunzo L Salzano E Panico A Lens PNL Pirozzi FBiomass enrichment and scale-up implications for darkfermentation hydrogen production with mixed culturesChem Eng Trans 201543391e6 httpdxdoiorg103303CET1543066

[36] Bianchi L Mannelli F Viti C Adessi A De Philippis RHydrogen-producing purple non-sulfur bacteria isolatedfrom the trophic lake Averno (Naples Italy) Int J HydrogenEnergy 20103512216e23 httpdxdoiorg101016 jijhydene201008038

[37] Koku H Eroglu I Gunduz U Yucel M Turker L Aspects of themetabolism of hydrogen production by Rhodobacter

sphaeroides Int J Hydrogen Energy 2002271315e

29 httpdxdoiorg101016S0360-3199(02)00127-1

[38] Esposito G Frunzo L Liotta F Panico A Pirozzi F Enhancedbio-methane production from co-digestion of differentorganic wastes Open Environ Eng J 201251e8

[39] Noguerol-Arias J Rodrıguez-Abalde A Romero-Merino EFlotats X Determination of chemical oxygen demand inheterogeneous solid or semi-solid samples using a novelmethod combining solid dilutions as a preparation stepfollowed by optimized closed reflux and colorimetricmeasurement Anal Chem 2012845548e55 httpdxdoiorg101021ac3003566

[40] DuBois M Gilles K Hamilton J Rebers P Smith FColorimetric method for determination of sugars and relatedsubstances Anal Chem 195628350e6

[41] Bligh EG Dyer WJ A rapid method of total lipid extractionand purification Can J Biochem Physiol 195937911e7

[42] American Public Health Association (APHA) Standardmethods for the examination of water and wastewater 21sted 2005 Washington DC

[43] Lee Z-K Li S-L Kuo P-C Chen I-C Tien Y-M Huang Y-J et alThermophilic bio-energy process study on hydrogenfermentation with vegetable kitchen waste Int J HydrogenEnergy 20103513458e66 httpdxdoiorg101016 jijhydene200911126

[44] Van Ginkel S Logan BE Inhibition of biohydrogen productionby undissociated acetic and butyric acids Environ SciTechnol 2005399351e6

[45] Shin H-S Youn abd J-H Kim S-H Hydrogen production fromfood waste in anaerobic mesophilic and thermophilic





acidogenesis Int J Hydrogen Energy 2004291355e63 httpdxdoiorg101016jijhydene200309011

[46] Kim S-H Sun-Kee H Hang-Sik S Feasibility of biohydrogenproduction by anaerobic co-digestion of food waste andsewage sludge Int J Hydrogen Energy 2004291607e16httpdxdoiorg101016jijhydene200402018

[47] Alzate-Gaviria LM Sebastian PJ Perez-Hernandez AEapen D Comparison of two anaerobic systems for hydrogen

production from the organic fraction of municipal solidwaste and synthetic wastewater Int J Hydrogen Energy2007323141e6 httpdxdoiorg101016 jijhydene200602034

[48] Thauer RK Jungermann K Decker K Energy conservation inchemotrophic anaerobic bacteria Bacteriol Rev 197741100e80

[49] Kim S-H Han S-K Shin H-S Effect of substrate concentrationon hydrogen production and 16S rDNA-based analysis of themicrobial community in a continuous fermenter ProcessBiochem 200641199e207 httpdxdoiorg101016 jprocbio200506013

[50] Hwang MH Jang NJ Hyun SH Kim IS Anaerobic bio-hydrogen production from ethanol fermentation the role of pH J Biotechnol 2004111297e309 httpdxdoiorg101016

jjbiotec200404024[51] Lin C Hung W Enhancement of fermentative hydrogen

ethanol production from cellulose using mixed anaerobiccultures Int J Hydrogen Energy 2008333660e7 httpdxdoiorg101016jijhydene200804036

[52] Han H Liu B Yang H Shen J Effect of carbon sources on thephotobiological production of hydrogen using Rhodobacter

sphaeroides RV Int J Hydrogen Energy 20123712167e74httpdxdoiorg101016jijhydene201203134

[53] Hustede E Steinbiichel A Schlegel HG Relationship betweenthe photoproduction of hydrogen and the accumulation of PHB in non-sulphur purple bacteria Appl MicrobiolBiotechnol 19933987e93

[54] Lee C-M Hung G-J Yang C-F Hydrogen production byRhodopseudomonas palustris WP 3-5 in a serial photobioreactor

fed with hydrogen fermentation effluent Bioresour Technol20111028350e6 httpdxdoiorg101016 jbiortech201104072

[55] Argun H Kargi F Kapdan I Light fermentation of darkfermentation effluent for bio-hydrogen production bydifferent Rhodobacter species at different initial volatile fattyacid (VFA) concentrations Int J Hydrogen Energy2008337405e12 httpdxdoiorg101016 jijhydene200809059

[56] Uyar B Schumacher M Gebicki J Modigell MPhotoproduction of hydrogen by Rhodobacter capsulatus from

thermophilic fermentation effluent Bioprocess Biosyst Eng 200932603e6 httpdxdoiorg101007s00449-008-0282-9

[57] Zong W Yu R Zhang P Fan M Zhou Z Efficient hydrogen gasproduction from cassava and food waste by a two-stepprocess of dark fermentation and photo-fermentationBiomass Bioenergy 2009331458e63 httpdxdoiorg101016jbiombioe200906008

[58] Wang Z Shao S Zhang C Lu D Ma H Ren X Pretreatment of vinegar residue and anaerobic sludge for enhanced hydrogenand methane production in the two-stage anaerobic systemInt J Hydrogen Energy 2015404494e501 httpdxdoiorg101016jijhydene201502029

[59] Ward AJ Hobbs PJ Holliman PJ Jones DL Optimisation of theanaerobic digestion of agricultural resources BioresourTechnol 2008997928e40 httpdxdoiorg101016

jbiortech200802044

Glossary

AD anaerobic digestionCOD chemical oxygen demandCSTR continuously stirred tank reactorDF dark fermentationFW food wasteHRT hydraulic retention timeOFMSW organic fraction of municipal solid wasteOLR organic loading ratePF photofermentationTS total solids

VFA volatile fatty acidsVS Volatile solids


Please cite this article in press as Ghimire A et al Biohydrogen production from food waste by coupling semi-continuous dark-



Introduction

The inherent characteristics of hydrogen (H2) such as higher

energy content (142 MJ per kg) energy and water as the only

by-products generated from its combustion application in

fuel cells for electricity generation and the ability to be pro-

duced biologically makes H2 a very interesting alternativefuture sustainable energy carrier [1] Among several biological

technologies proposed for H2 production dark fermentation

(DF) is emerging as one of the prominent options shown by

the increasing research interests in this technology [2] The

advantages such as the flexibility to operate under different

conditions of temperature and pressure higher production

rates possibility to use renewable waste biomass as feedstock

and the treatment capability make the DF process attractive

Waste biomass such as agricultural residues the organic

fraction of municipal solid waste (OFMSW) and agro-

industrial wastes are economically competitive when

considering a supply of sustainable feedstock aiming at the

industrial development of DF systems for biological treatmentof waste [3e5]

OFMSW which is mainly composed of food waste (FW) has

been receiving a lot of attention because of its potential to be

used for the production of biofuels and other value added

products [6] Especially about 13 billion tonnes of food per

year get wasted which is approximately one-third of the food

produced for human consumption [7] FW is generated from

agricultural production industrial manufacturing processes

and final consumption in households In the European Union

the total annual generation of FW is estimated around 893

million tonnes comprising 377 million tonnes generated

from household consumption alone [8] The volatile solids

content in FW ranges from 21 to 27 which shows its highorganic carbon contentthat can be furthervalorized [9]andin

particular for H2 production by DF as demonstrated in the

literature [10e15] Some studies have reported the operational

feasibility of continuous H2 production using food or kitchen

wastes as a feed in DF processes [101416]

With the advantage of a stable operation continuous DF

processes are usually preferred and scaling-up is more viable

in comparison to batch processes which involve regular

downtime periods of maintenance [17] However stable

operation of continuous DF of FW is highly influenced by

bioreactor operating parameters such as pH temperature

organic loading rates (OLRs) and hydraulic retention times

(HRTs) [4518] These factors also influence the microbialcommunities and thus the biochemical pathways that can

affect the total H2 yields in mixed cultures [19] In addition

there is growing interest in coupling DF either with photo-

fermentation (PF) [2021] or bioelectrochemical systems (BES)

[22] to obtain higher overall H2 yields or with anaerobic

digestion (AD) for methane production [23e25] due to the

post-treatment requirement of DF effluents (DFEs) and net

positive energy gain from coupling these bioprocesses [26]

H2 production rates and total H2 yields are mainly a func-

tion of substrate types and OLRs applied [2] A varying rangeof

optimal OLR values has been reported for dark fermentative

H2 conversion from FW carried out in thermophilic DF pro-

cesses [2] Shin et al [27] found an optimal H2 yield of 12625 L

H2 kg VS at an OLR of 8 kg VSm3 d while the H2 production

decreased when the OLR was increased to 10 kg VSm 3 d The

authors reported 8 kg VSm3 d 5 days and a pH of 55

respectively as optimal OLR HRT and culture pH In a study

coupling DF and AD Cavinato et al [10] reported 667 L H2 kg

VS added at an optimum OLR of 163 VSm3 d a HRT of 33

days and for a pH maintained in the range of 5e6 through the

recirculation of AD effluent Generally HRTs in a range of 2e

6days have been reported as optimum for DF of organic FW in a

CSTR process [2] This range ofHRTs is similar to the first stage

of a two-stage AD process [28] Moreover the HRT is also a

function of the substrate type and bioreactor operational

parameters

It has been well documented that dark fermentative H2

production is generally due to the conversion of the initial

soluble fraction of carbohydrates present in the complex

organic biomass that will lead to accumulation of volatile

fatty acids (VFAs) and alcohols in DFEs [2930] Some recent

studies have shown the potential of these DFEs to be utilized

in PF processes for H2 production [2021] Combining DF with

PF Su et al [31] achieved an increase in H2 yield from 767 to5961 L H2 kg VS from water hyacinth Meanwhile Rai et al

[20] achieved 43 higher volumetric H2 yields from acid hy-

drolyzed sugarcane bagasse in two step DF-PF systems

However during the conversion of complex organic biomass

like FW a part of the unhydrolyzed solid residues will remain

that can be further valorized in AD systems producing

methane (CH4) in a three steps conversion scheme (Fig 1) Xia

et al [3233] reported that a three-step conversion of algal

biomass combining DF-PF-AD can achieve 17 and 13 times

higher energy yields in comparison to a two-stage DF-AD and

an one stage AD process respectively

High OLRs are often responsible for a decrease in culture

pH due to the accumulation of VFAs present in DFE Thusmost of the continuous DF systems utilizing acidic substrates

such as food waste requires constant addition of external

alkalinity sources such as alkaline chemicals (NaOH or KOH)

or buffering agents (bicarbonate or phosphate buffers)

[142734] A long-term study of continuous H2 production at

varying operating conditions of OLR and HRT to establish a

long-term operability for continuous H2 production in relation

with the production of metabolites could provide further in-

sights for the development of scaled-upDF systems Similarly

a three-step conversion process (DF PF and AD) might

contribute to an increase in overall energy yield and could

provide the biological treatment to the by-products generated

from DF systemsThis study aims to demonstrate the long-term operational

feasibility of continuous H2 production from FW using a semi-

continuous thermophilic DF reactor at various low OLRs and

HRTs without pH control The experiment also aimed at

reducing the dependency on chemical buffering agents that

are used to maintain the culture pH at working conditions H2

production through different possible biochemical pathways

was discussed in relation to major metabolites present in

DFEs obtained during the varying experimental conditions

The potential of coupling DF with photofermentative H 2 pro-

duction was investigated in batch PF experiments by using the

liquid fraction of the DFE after physical separation Further

the waste streams generated from the coupling of DF-PF were





































VS contents


























TKN 2080 plusmn 7




Lactic Acid 330

Acetic Acid 4660

Propionic Acid 4496

Butyric Acid 10754

Ethanol 3230
































4 ) 358 and total















DF bioreactor setup















Analytical methods





















[42]

Data analysis










AvgHPR (1)


production





generated



OLR (kg VSm3 d) 1 1 15 2 2 25

HRT (d) 12 6 6 6 4 4





























microbial community



















operation














HY(N Lkg VSadded)

H2 inbiogas ()

HPSI







































by equations (6)e















DF














4 has been


[5455]





biogas ()Reference

FW Batch 55 45 (initial) 6 463 23 [45]

Vegetable kitchen

waste





































of CO2















energy









a (kJmol) Eqn










aDG0






















option











































Conclusion






















Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary







































VS contents


























TKN 2080 plusmn 7




Lactic Acid 330

Acetic Acid 4660

Propionic Acid 4496

Butyric Acid 10754

Ethanol 3230
































4 ) 358 and total















DF bioreactor setup















Analytical methods





















[42]

Data analysis










AvgHPR (1)


production





generated



OLR (kg VSm3 d) 1 1 15 2 2 25

HRT (d) 12 6 6 6 4 4





























microbial community



















operation














HY(N Lkg VSadded)

H2 inbiogas ()

HPSI







































by equations (6)e















DF














4 has been


[5455]





biogas ()Reference

FW Batch 55 45 (initial) 6 463 23 [45]

Vegetable kitchen

waste





































of CO2















energy









a (kJmol) Eqn










aDG0






















option











































Conclusion






















Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary


































4 ) 358 and total















DF bioreactor setup















Analytical methods





















[42]

Data analysis










AvgHPR (1)


production





generated



OLR (kg VSm3 d) 1 1 15 2 2 25

HRT (d) 12 6 6 6 4 4





























microbial community



















operation














HY(N Lkg VSadded)

H2 inbiogas ()

HPSI







































by equations (6)e















DF














4 has been


[5455]





biogas ()Reference

FW Batch 55 45 (initial) 6 463 23 [45]

Vegetable kitchen

waste





































of CO2















energy









a (kJmol) Eqn










aDG0






















option











































Conclusion






















Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary






























microbial community



















operation














HY(N Lkg VSadded)

H2 inbiogas ()

HPSI







































by equations (6)e















DF














4 has been


[5455]





biogas ()Reference

FW Batch 55 45 (initial) 6 463 23 [45]

Vegetable kitchen

waste





































of CO2















energy









a (kJmol) Eqn










aDG0






















option











































Conclusion






















Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary

































by equations (6)e















DF














4 has been


[5455]





biogas ()Reference

FW Batch 55 45 (initial) 6 463 23 [45]

Vegetable kitchen

waste





































of CO2















energy









a (kJmol) Eqn










aDG0






















option











































Conclusion






















Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary

























of CO2















energy









a (kJmol) Eqn










aDG0






















option











































Conclusion






















Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary




















option











































Conclusion






















Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary










Acknowledgments












Measure 124

r e f e r e n c e s





















(N L H2 kg VS)


(MJkg VS)

CH4 yieldfrom AD

(L CH4 kg VS)

a Totalenergy yield

(MJkg VS)

Reference



Vinegar residue

treated by HCl

DF thorn AD (batch) 532 057 192 74 [58]



(batch)

1839 198 1613 774 [33]


(batch)

1983 214 1862 666 [32]



184 199 993 555 This study







































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary








































29 httpdxdoiorg101016S0360-3199(02)00127-1

































jbiortech200802044

Glossary



























jbiortech200802044

Glossary





Documents

Bio H Producere