Biotechnol Lett
DOI 10.1007/s10529-017-2323-4
ORIGINAL RESEARCH PAPER
Improving EGSB reactor performance for simultaneous
bioenergy and organic acid production from cheese whey
via continuous biological H2 production
Lucas Rodrigues Ramos . Edson Luiz Silva
Received: 6 February 2017 / Accepted: 9 March 2017
Ó Springer Science+Business Media Dordrecht 2017
Abstract
Objectives To evaluate the influence of hydraulic
retention time (HRT) and cheese whey (CW) substrate
concentration (15 and 25 g lactose l-1) on the
performance of EGSB reactors (R15 and R25, respec-
tively) for H2 production.
Results A decrease in the HRT from 8 to 4 h favored
the H2 yield and H2 production rate (HPR) in R15, with
maximum values of 0.86 ± 0.11 mmol H2 g COD-1
and 0.23 ± 0.024 l H2 h-1 l-1, respectively. H2 pro-
duction in R25 was also favored at a HRT of 4 h, with
maximum yield and HPR values of 0.64 ± 0.023 mmol
H2 g COD-1 and 0.31 ± 0.032 l H2 h-1 l-1, respec-
tively. The main metabolites produced were butyric,
acetic and lactic acids.
Conclusions The EGSB reactor was evaluated as a
viable acidogenic step in the two-stage anaerobic
treatment of CW for the increase of COD removal
efficiency and biomethane production.
Keywords Cheese whey wastewater  Hydraulic
retention time  Organic loading rate  Substrate
concentration  Whey
L. R. Ramos  E. L. Silva (&)
Department of Chemical Engineering, Federal University
of São Carlos, Rod Washington Luis, Km 235,
São Carlos, SP CEP 13565-905, Brazil
e-mail: edsilva@ufscar.br
Introduction
The dairy industry generates considerable quantities
of residues, one of which is cheese whey (CW). CW is
produced during the precipitation and removal of
casein in the cheese-making process. For every 1 kg of
cheese, 9 kg CW are produced (Siso 1996). The
chemical oxygen demand (COD) of CW varies from
50 to 102 kg m-3, with most of the organic load
comprising lactose from the milk (39–60 kg m-3)
(Carvalho et al. 2013). Because of this high carbohy-
drate content, CW is an ideal resource for use in
anaerobic treatment and bioenergy production tech-
nologies (Karadag et al. 2014). However, anaerobic
digestion in a single methanogenic stage faces chal-
lenges in the low alkalinity of this wastewater and the
rapid pH decrease via the accumulation of organic
acids. Biological H2 production in the first stage is
therefore an alternative allowing for the synergy of
CW treatment and bioenergy production (Karadag
et al. 2014).
However, the economic feasibility of fermentative
H2 production from CW depends on the development
and design of reactors that can improve the H2 yield
(HY) and production rate (HPR) (Show et al. 2012).
Regarding the choice of reactor configuration, the
reactor should be able to operate under low hydraulic
retention time (HRT), high biomass retention, and
optimal good mass transfer for the increase of
substrate conversion and prevention of high H2 partial
pressures. For lower-cost substrates, HPR becomes
123

more important than HY because higher substrate
concentrations can increase HPR. Thus, the feasibility
of H2 production from CW requires that reactors can
operate in high substrate concentrations with low HRT
and high biomass retention. Costs are reduced through
the design of smaller reactors (Pawar and van Niel
2013).
H2 production from CW has been evaluated using
upflow anaerobic sludge bed (UASB) reactors (Ki-
rankumar et al. 2016), anaerobic packed-bed reactors
(APBRs) (Perna et al. 2013), anaerobic fluidized-bed
reactors (AFBRs) (Rosa et al. 2014), anaerobic
sequencing biofilm batch reactors (AnSBBRs) (Lima
et al. 2016), and continuous stirred-tank reactors
(CSTRs) (Venetsaneas et al. 2009). The most used
system is the CSTR because it facilitates suspended
cellular growth with high mass transfer between the
substrate and the microbial community. However,
with a low HRT, a high biomass concentration cannot
be maintained because the H2-producing microbial
community in suspension is flushed out of the CSTR
(Barca et al. 2015).
This drawback can be overcome by biofilm reac-
tors, such as the expanded granular sludge bed reactor
(EGSB), via decoupling of the HRT and the biomass
retention time. In biofilm reactors, a carrier material is
used for bacterial adhesion, increasing the biomass
concentration in the reactor (Guo et al. 2008). Thus,
the EGSB stands out among reactor configurations for
H2 production, as it favors mass transfer between the
substrate and the microbial community, allows for
high substrate conversion even at high organic loading
rates (OLR), and reduces HRT (Wang et al. 2009).
Although the EGSB has many advantages, few
studies have evaluated its efficiency for H2 production
using simple substrates (Abreu et al. 2012) and real
wastewaters (Guo et al. 2008; Wang et al. 2009). In
addition, identifying the optimal range of operation is of
vital importance to ensuring that substrate overload does
not occur with high H2 production. Thus, for the future
design of high-rate anaerobic reactors with immobilized
biomass in two-stage anaerobic digestion, identifying
the effects of the operational parameters related to OLR,
which are HRT and substrate concentration, is neces-
sary. Therefore, the objective of this study was to
evaluate the effect of HRT (8–0.5 h) on continuous
biological H2 production in EGSB reactors with the
substrate concentrations of 15,000 and 25,000 mg
lactose l-1 using CW as a carbon source.
123
Biotechnol Lett
Materials and methods
Substrate feed
The dairy company industry Elegê Laticı́nios S.A.,
located in Teotônia, Brazil, supplied the CW powder
used in the CW wastewater fed to the EGSB reactors.
Different solutions with 10 g CW powder l-1 were
characterized as containing 18 ± 1 mg Kjeldalh
nitrogen g CW-1, the equivalent of 116 ± 10 mg
protein g CW-1, 7822 ± 697 mg total sugars (lac-
tose) l-1, and 11,009 ± 1030 mg total COD l-1 and
having a pH of 6.3 ± 0.1. Nutrients required for cell
growth were added to the affluent according to the
methods of Rosa et al. (2014).
Description of EGSB reactor
The reactors were constructed in transparent acrylic
using an external water jacket in which water from an
ultrathermostatic bath controlled the reactor temper-
ature at 30 °C. The reactors had a 4.3 cm internal
diam., 150 cm ht, and total volume of 2177 cm3. The
superficial velocity of the reactors was maintained at
5.5 m h-1 via recycling of the reactor effluent, which
ensured a granular bed expansion of 30%.
The granules were obtained from the granular
sludge of a UASB reactor used for poultry slaughter-
house wastewater treatment (Avı́cola Dakar S/A,
Tietê/SP). The granules were heat-treated using a
method adapted from Kim et al. (2006). The granules
were boiled at 100 °C for 1 h and then cooled to
ambient temperature in an ice bath. The goal of this
treatment was to inhibit methanogenic Archaea and
select for acidogenic bacteria that could form
endospores. A diagram of the EGSB reactors used in
this study is presented in Fig. 1.
Seed sludge
The seed sludge used for the operational start-up of the
mesophilic EGSB reactors was obtained from the
granulated sludge of a UASB reactor used for poultry
slaughterhouse wastewater treatment (Avı́cola Dakar
S/A, Tietê/SP). For the selection of H2-producing
communities, the sludge was subjected to a thermal
pretreatment at 90 °C for 10 min and cooled to
ambient temperature in an ice bath according to the
methodology of Kim et al. (2006).
Biotechnol Lett
Fig. 1 Schematic diagram
of reactor setup
Start-up and operation of EGSB reactors
The reactors were fed with the CW powder solution
with concentrations of 15,000 (R15) and 25,000
mg l-1 (R25) of total carbohydrates (lactose). The
reactors were operated in batch mode via recycling of
the inoculum with CW in their respective concentra-
tions in order for microorganisms to adapt to the CW
and attach to the surface of the granules. Following the
inoculation step, the reactors were switched to contin-
uous mode for 8 h. After verification of steady-state H2
production through the attainment of stable HPR and
HY values, the operational conditions were adjusted
via decreasing the HRT to 6, 4, 2, 1, and 0.5 h, for a
total of 260 days of operation.
Analytical methods
Total carbohydrate concentrations were measured
using the method of Dubois et al. (1956) with lactose
as a standard. COD, pH, and Kjeldahl nitrogen
analyses were performed according to APHA (2012).
The concentrations of organic acids and alcohols were
determined by HPLC according to the methodology of
Penteado et al. (2013). H2 and CO2 gas compositions
were measured by GC according to Rosa et al. (2014).
123

The reactor volumetric biogas production was mea-
sured using the fluid displacement method adapted
from Walker et al. (2009).
Results and discussion
The carbohydrate conversion values observed in the
R15 and R25 reactors increased with the reduction of
the HRT values from 8 to 4 h, with maximum values
of 60 ± 3.3% (R15) and 38.5 ± 3.94% (R25). By
decreasing the HRT to values below 2 h, the carbo-
hydrate conversion values were reduced to the min-
imum values of 32.9 ± 4.5% in R15 and 8.7 ± 3.2%
in R25. Reduced conversion of carbohydrates with the
reduction of the HRT has also been observed in other
studies. In evaluating the H2 produced from CW of
different Thermotoga strains in batch tests, Cappelletti
et al. (2012) observed substrate conversions ranging
from 39% to 82%. Using an AFBR in mesophilic
conditions, Rosa et al. (2014) evaluated the effect of
different HRT conditions (1, 2, and 4 h) on the
production of H2 from CW with an organic matter
concentration of 5000 mg COD l-1. With the reduc-
tion of the HRT, the conversion of lactose was reduced
from 95 to 85%.
During the operation of the mesophilic EGSB
reactors, H2 content (H2 %) (Fig. 2) was analyzed for
the evaluation of H2 production.
The reduction of HRT values from 8 to 4 h favored
the H2 % of the R15 reactor, which reached a
maximum value of 23.6 ± 4.9%. However, with
HRT values lower than 4 h, H2 production was
hampered, with no production at HRT values of 1
and 0.5 h. On the other hand, the decrease in HRT
Fig. 2 Variation in H2 % as a function of HRT for R15 and R25
123
Biotechnol Lett
from 8 to 0.5 h in the R25 reactor raised the % H2
values from 0.7 ± 0.1% to 37.5 ± 3.1%.
The maximum H2 % value was 37.5 ± 3.1% in
R25. This value is higher than those of other studies
that used continuous reactors. For example, Perna
et al. (2013) used a mesophilic APBR (30 °C) for
continuous H2 production from CW and observed a
maximum H2 % value of 10% with an increase in the
substrate concentration from 25 to 37 g COD l-1 at an
HRT of 24 h. Similarly, Fernández et al. (2015)
achieved a maximum H2 % of 18.2% by decreasing
the HRT from 3 to 2 days using a CSTR for H2
production from CW (38 g COD l-1).
pH values between 4 and 4.3 in the R15 and R25
reactors favored H2 production and inhibited the
methane producers. The COD removal values were
between 4.6 ± 1.5 and 9.9 ± 1.6%. The maximum
values of 9.9 ± 1.6% (R15) and 7.1 ± 0.9% (R25)
were obtained at the HRT of 4 h.
HPR is a useful design parameter, as it allows for
the evaluation of the production capacity of a reactor
from its volume or estimation of these same values for
the achievement of a desired production level. The
HPR values of reactor R15 increased with the
reduction of the HRT from 8 to 4 h, with a maximum
value of 0.228 ± 0.024 l H2 h-1 l-1 (Fig. 3). How-
ever, the further decrease of the HRT restricted H2
production, which did not occur with HRT values
lower than 2 h. In the R25 reactor, the decrease in
HRT from 8 to 2 h favored HPR, which reached a
maximum value of 0.312 ± 0.032 l H2 h-1 l-1 at the
HRT of 4 h. At the HRT of 1 h, the HPR was reduced
to 0.173 ± 0.051 l H2 h-1 l-1, and at the HRT of
0.5 h, the HPR was 0.302 ± 0.033 l H2 h-1 l-1.
Similarly, Fernández et al. (2015), using a mesophilic
Fig. 3 Variation in HPR as a function of HRT for R15 and R25
Biotechnol Lett
sequencing batch reactor for H2 production from CW,
observed that with the reduction of HRT from 3 to
1.5 days, the HPR increased from 0.15 l H2 l-1 days-1
to 0.19 l H2 l-1 days-1.
The substrate concentration of 25 g carbohydrate
l-1 produced higher HPR values (with a maximum of
0.312 ± 0.032 l H2 h-1 l-1 in R25) than did that of
15 g carbohydrate l-1. Similar results were obtained
by Yu and Fang (2001), who found that the increase in
dairy wastewater concentration from 2 to 30 g COD
l-1 favored H2 production by the inhibition of H2-
consuming microorganisms, such as methanogens. In
evaluating the H2 production from CW, Kargi et al.
(2012) achieved a maximum cumulative H2 produc-
tion of 257 ml by increasing the substrate concentra-
tion from 5 to 20 g total sugars l-1. Similar to this
study’s findings for R25, Davila-Vazquez et al. (2008)
observed a maximum HPR of 0.2 l H2 h-1 l-1 with
CW at 25 g l-1 when evaluating the kinetics of H2
production from CW in concentrations between 0.86
and 29.1 g l-1. Additionally, Amorim et al. (2012)
evaluated the effect of HRT values from 1 to 8 h in
four mesophilic AFBRs (30 °C) at different concen-
trations of glucose (2–25 g l-1). The authors observed
a maximum HPR of 1.46 l H2 h-1 l-1 with 10 g l-1 of
glucose. In the reactor fed with the concentration of
25 g l-1, the HPR obtained a maximum value of
0.71 l H2 h-1 l-1 at the HRT of 2 h, decreasing to
0.61 l H2 h-1 l-1 at the HRT of 1 h.
Another important parameter, HY relates the pro-
duction of H2 per unit time to the organic load applied
to the reactor, allowing for the observation of the
inhibitive effects of substrate overload. In the R15 and
R25 reactors, the reduction of HRT from 8 to 4 h
favored HY, which reached maximum values of
0.856 ± 0.106 (R15) and 0.638 ± 0.023 mmol H2 g
COD-1 (R25) (Fig. 4). Similarly, Kirankumar et al.
(2016) evaluated the effect of HRT on the continuous
production of H2 from dairy wastewater in a
mesophilic UASB reactor and obtained a maximum
HY of 5.0 mmol H2 g COD-1 by reducing HRT from
36 to 6 h.
The strategy used for reactor R15 obtained the
highest HY value of 0.856 ± 0.106 mmol H2 g
COD-1. Similarly, in studying the kinetics of H2
production from CW, Davila-Vazquez et al. (2008)
observed the maximum HY value of 3.1 mol H2 mol
lactose-1 in the concentration of 15 g l-1.
Fig. 4 Variation in the HY values as a function of HRT in R15
and R25
However, in a mesophilic AFBR (30 °C), Amorim
et al. (2012) observed a maximum HY of 2.5 mol
H2.mol glucose-1 at 2 g glucose l-1 with a decrease in
the HRT from 8 to 2 h. On the other hand, the HY
decreased from 2.5 to 0.6 mol H2 mol glucose-1 with
an increase in glucose from 2 to 25 g l-1. The authors
concluded that the progressive increase in H2 produc-
tion with the increase in glucose concentration was
attributable to the restriction of effective metabolism
for H2 production by substrate in low concentrations.
However, the reduction of HY above the optimum
concentration was due to the limited use of glucose
during both transport and metabolism.
In this study, the reduction of HRT favored both HY and
HPR in reactors R15 and R25 via the increase of the OLR.
With a decrease in the HRT from 8 to 4 h, the OLR of the
R15 reactor increased from 67.5 to 135 kg COD m-3
days-1, producing the maximum values of HPR
(0.228 ± 0.024 l H2 h-1 l-1) and HY (0.856 ±
0.106 mmol H2 g COD-1). In the same way, the R25
reactor achieved maximum values of HPR (0.312 ±
0.032 l H2 h-1 l-1) and HY (0.638 ± 0.023 mmol H2 g
COD-1) at the HRT of 4 h, equivalent to an OLR of 210 kg
COD m-3 days-1, and achieved an HPR of 0.302 ±
0.033 l H2 h-1 l-1 at an OLR of 1680 kg COD m-3 days-1
(HRT = 0.5 h). The increase of the OLR guaranteed the
greater availability of substrate, accelerating bacterial
metabolism. Similarly, Kargi et al. (2012) observed that
the increase of the substrate concentration from 5.2 to
20 g l-1 allowed for increases in HY and HPR because low
concentrations limited the availability of the substrate.
However, when OLR values were greater than 135 kg
COD m-3 days-1 in R15 and 210 kg COD m-3 days-1 in
R25 in this study, in HY values decreased.
123

Table 1 shows the OLR values used in and the HY
and HPR results obtained from this and past studies.
The findings here are in agreement with those in the
literature, including that the decrease of the HY is
associated with the increase of the OLR. In evaluating
mesophilic H2 production from CW in a CSTR,
Venetsaneas et al. (2009) observed maximum HY
(1.56 mol H2 mol lactose-1) and HPR (0.12 l H2 h-1
l-1) values at the HRT of 24 h, equivalent to the OLR
of 30 kg COD m-3 days-1 (30 g COD l-1). Further-
more, using a fixed-bed mesophilic reactor at an HRT
of 24 h, Perna et al. (2013) obtained a maximum HY
of 1.1 mol H2 mol lactose-1 by raising the OLR to
37 kg COD m-3 days-1 and increasing the substrate
concentration from 22 to 37 g COD l-1. Using a
sequential batch reactor (AnSBBR), Lima et al. (2016)
attained increases in HPR (0.045 l H2 h-1 l-1) and in
HY (1.12 mol H2 mol lactose-1) by increasing the
applied OLR from 14.6 to 18.5 kg COD m-3 days-1
and the affluent concentration from 4.25 to 5.4 g COD
l-1, equivalent to 4.8 g carbohydrate l-1.
It should be noted that an excessive increase in
OLR might overload the reactor. Rosa et al. (2014)
found a decrease in HY from 1.33 to 0.44 mol H2 mol
lactose-1 with an increase in the OLR from 30 kg to
120 kg COD m-3 days-1 (5 g COD l-1) via decreas-
ing the HRT from 4 to 1 h. According to the authors,
the increase in HY is related to the selection of OLR
values within an appropriate range. The challenge in
the selection of the OLR is associated with the
complexity of the microbial community and the
dependence of the different metabolic levels on the
applied environmental conditions. In this study, H2
production was not inhibited by the high OLR of
1680 kg COD m-3 days-1 (HRT = 0.5 h) in reactor
R25 with an HPR of 0.302 ± 0.033 l H2 h-1 l-1.
The quantification of the liquid metabolites pro-
duced in the reactor, such as volatile organic acids and
Table 1 Optimum values of HY, HPR, HRT, and OLR observed in studies using CW
Reference Reactor HRT (h) OLR (kg COD m-3 days-1)
Venetsaneas et al. (2009) CSTR 24 30
Perna et al. (2013) APBR 24 37
Rosa et al. (2014) AFBR 4 30
Lima et al. (2016) AnSBBR 3 18.5
R15 EGSB 4 135
R25 EGSB 4 210
123
Biotechnol Lett
alcohols, is necessary because of the functions of these
metabolites as regulators of possible routes for the
metabolism of the microbial community, including the
H2 producers and consumers and the competitors for
substrate. In addition, some of these metabolites may
play a toxic or inhibitory role for populations of H2-
producing microorganisms.
Lactic, butyric and acetic acids were identified in
higher contents in R15 (Fig. 5). Other metabolites
observed were propionic, formic, and succinic acids.
The decrease in HRT from 8 to 0.5 h suppressed
lactate production, decreasing the molar fraction from
73.6% at an HRT of 8 h to 1.8% at an HRT of 2 h. This
decrease in HRT was accompanied by the predomi-
nance of acetate (10.4–43.1%), butyrate (1.5–46.3%),
and propionate (1.1–23.2%). The molar fractions of
HFo and HSu remained constant between 2.1–12.7
and 2.5–5.3%, respectively.
In reactor R25, the main metabolites detected were
lactate, butyrate and acetate. Other metabolites
observed were propionate, formate and succinate.
With a decrease in the HRT from 8 to 0.5 h, HLa
production was suppressed (74.3–1.8%) in favor of
acetate (4.7–33.1%), butyrate (1.1–53.3%), and pro-
pionate (0.9–11.9%). The productions of formate and
succinate remained constant between 1.9–8.4 and
1.7–5.3%, respectively.
In the R15 and R25 reactors, higher concentrations
of lactate (up to 74.3%) were observed with the HRT
values of 8 and 6 h, while HY and HPR were the
lowest. Noike et al. (2002) observed that the lactic acid
pathway is substrate-competitive with H2 production.
When investigating the coexistence of H2-producing
bacteria (Clostridium acetobutylicum or Clostridium
butylicum) and lactic acid bacteria (Lactobacillus
paracasei and Enterococcus durans) using tofu-pro-
cessing wastewater, the authors also found that H2
production was inhibited by lactic acid bacteria.
HPR (l H2 h-1 l-1) HY (mol H2 mol lactose-1)
0.12 1.56
0.042 1.1
0.023 1.33
0.045 1.12
0.228 0.96 ± 0.07
0.312 0.96 ± 0.10
Biotechnol Lett
Fig. 5 Distribution of the
main metabolites produced
as a function of HRT in
a R15 and b R25
The distribution of metabolites in all reactors
indicates the greater production of acetate and butyrate
with the decrease in the HRT from 8 to 0.5 h. These
are the major metabolites related to high HPR. The
production of H2 through the acetic and butyric routes
can result in maximum HY values of 8 and 4 mol H2
mol lactose-1, respectively, as shown in Eqs. 1 and 2
(Davila-Vazquez et al. 2008), as follows:
C12 H22 O11 þ 5H2 O ! 8H2 þ 4CO2 þ 4CH3 COOH
ð1Þ
C12 H22 O11 þ H2 O ! 4H2 þ 4CO2
þ 2CH3 CH2 CH2 COOH ð2Þ
At the HRT of 4 h, reactors R15 and R25 produced
high concentrations of acetate (37.3%) and butyrate
(46.3%) and also resulted in the maximum HY
(0.86 mmol H2 g COD-1) in R15 and maximum
HPR (0.312 l H2 h-1 l-1) in R25. Similarly, Perna
et al. (2013) observed a maximum HY of 1.1 mol H2
mol lactose-1 associated with high production of
butyrate (10,000 mg l-1) and a reduction in lactate
production (2000 mg l-1). Rosa et al. (2014) observed
increased production of acetate and butyrate with the
decrease in HRT from 4 to 1 h.
The presence of acetate at 30.1 to 43.1% in the HRT
of 4 and 0.5 h, conditions with greater H2 production,
indicates that the increases in OLR from 67.5 to
135 kg COD m-3 days-1 (R15) and 105 to 210 kg
COD m-3 days-1 (R25) have directed H2 production
via the acetic route. A similar profile of acetate and H2
production was observed by Guo et al. (2008). In their
study, with an increase in the OLR to 120 kg COD
m-3 days-1 via a decrease in the HRT from 6 to 2 h,
123

the acetate and ethanol concentrations reached their
maximum values, representing 89% of the metabolites
produced. In this condition, the maximum HPR value
of 0.71 l H2 h-1 l-1 was observed.
High production of HPr (18.1–23.2%) in reactor
R15 at HRT values lower than 2 h was responsible for
the consumption of the H2 produced and related to
substrate overload due to inhibition by products at
OLR values higher than 135 kg COD m-3 days-1.
Guo et al. (2008) observed an increase in the propionic
acid concentration to 280 mg l-1 associated with the
overload of the EGSB reactor with the increase of the
OLR to 192 kg COD m-3 days-1.
Although reactor R25 was operated at OLR values
up to 1680 kg COD m-3 days-1, which were higher
than those in R15, decreasing the HRT from 8 to 0.5 h
helped maintain high H2 production (0.302 ± 0.033 l
H2 h-1 l-1), and the main metabolite was HBu (36.0%).
However, at the HRT of 1 h and with the increase in the
HPr concentration to 614 mg l-1, the HY was reduced
to 0.101 mmol H2 g COD-1. According to Hallenbeck
(2009), the NADH produced by glycolysis must be
reoxidized so that carbohydrate consumption can
continue. If NADH cannot be oxidized by H2 produc-
tion in the acetic pathway, other metabolic routes with
more reduced products, such as the production of
butyric acid, are favored. Likewise, Amorim et al.
(2012) observed the predominance of acetate
(38.1–55.1%) and butyrate (39.6–47.4%) in mesophilic
AFBRs (30 °C) with 2 and 4 g glucose l-1. With
glucose at 10 and 25 g l-1, the metabolic routes of
methanol (0–42.3%) and ethanol (48.5–62.7%) pro-
duction were predominant in the acetic and butyric
routes, and HY decreased to 0.6 mol H2 mol glucose-1.
Conclusions
The increase in OLR favored mesophilic H2 produc-
tion from CW in the EGSB reactor. Efficient immo-
bilization of the H2-producing microbial community
facilitated increases in HY and HPR with the reduction
of HRT. A lower substrate concentration (R15) led to a
maximum HY value, while HPR was favored by
higher concentrations (R25). The acetic and butyric H2
production routes were favored over the production of
lactic and propionic acids when HRT decreased and
substrate concentration increased.
123
Biotechnol Lett
Acknowledgements This work was supported by the CNPq
(National Council for Scientific and Technological Development),
CAPES (Coordination for the Improvement of Higher Education
Personnel), and FAPESP (São Paulo Research Foundation).
References
Abreu AA, Karakashev D, Angelidaki I, Sousa DZ, Alves MM
(2012) Biohydrogen production from arabinose and glu-
cose using extreme thermophilic anaerobic mixed cultures.
Biotech Biofuels 5:1–12
Amorim ELC, Sader LT, Silva EL (2012) Effect of substrate
concentration on dark fermentation hydrogen production
using an anaerobic fluidized bed reactor. Appl Biochem
Biotech 166:1248–1263
APHA (American Public Health Association), American Water
Works Association, Water Environment Federation (1998)
Standard methods for the examination of water and
wastewater, 20th edn. APHA, Washington, DC
Barca C, Soric A, Ranava D, Giudici-Orticoni M, Ferrasse L
(2015) Anaerobic biofilm reactors for dark fermentative
hydrogen production from wastewater: a review. Bioresour
Technol 185:386–398
Carvalho F, Prazeres AR, Rivas J (2013) Cheese whey
wastewater: characterization and treatment. Sci Total
Environ 445–446:385–396
Davila-Vazquez G, Alastriste-Mondragón F, León-Rodrı́guez
A, Razo-Flores E (2008) Fermentative hydrogen produc-
tion in batch experiments using lactose, cheese whey and
glucose: influence of initial substrate concentration and pH.
Int J Hydrogen Energy 33:4989–4997
Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F (1956)
Colorimetric method for determination of sugars and
related substances. Anal Chem 28:350–356
Fernández C, Cuetos MJ, Martı́nez EJ, Gómez X (2015) Ther-
mophilic anaerobic digestion of cheese whey: coupling H2
and CH4 production. Biomass Bioenerg 81:55–62
Guo WQ, Ren NQ, Wang XJ, Xiang XJ, Meng ZH, Ding J, Qu
YY, Zhang LS (2008) Biohydrogen production from
ethanol-type fermentation of molasses in an expanded
granular sludge bed (EGSB) reactor. Int J Hydrogen
Energy 33:4981–4988
Hallenbeck PC (2009) Fermentative hydrogen production:
principles, progress and prognosis. Int J Hydrogen Energy
34:7379–7389
Karadag D, Köroğlu OE, Ozkaya B, Cakmakci M, Heaven S,
Banks C (2014) A review on fermentative hydrogen pro-
duction from dairy industry wastewater. J Chem Technol
Biotechnol 89:1627–1636
Kargi F, Eren NS, Ozmihci S (2012) Hydrogen gas production
from cheese whey powder (CWP) solution by thermophilic
dark fermentation. Int J Hydrogen Energy 37:2260–2266
Kim DH, Han SK, Kim SH, Shin H (2006) Effect of gas sparging
on continuous fermentative hydrogen production. Int J
Hydrogen Energy 31:58–69
Kirankumar P, Krishna SV, Chaitanya N, Bhagawan D, Hima-
bindu V, Narasu ML (2016) Effect of operational param-
eters on biohydrogen production from dairy wastewater in
batch and continuous reactors. Biofuels 1–7
Biotechnol Lett
Lima DMF, Lazaro CZ, Rodrigues JAD, Ratusznei SM, Zaiat M
(2016) Optimization performance of an AnSBBR applied
to biohydrogen production treating whey. J. Environ
Manag 169:191–201
Noike T, Takabatake H, Mizuno O, Ohba M (2002) Inhibition of
hydrogen fermentation of organic wastes by lactic acid
bacteria. Int J Hydrogen Energy 27:1367–1371
Pawar SS, van Niel WJ (2013) Thermophilic hydrogen pro-
duction: how far are we? Appl Microbiol Biotechnol 97:
7999–8009
Penteado ED, Lazaro CZ, Sakamoto IK, Zaiat M (2013) Influ-
ence of seed sludge and pretreatment method on hydrogen
production in packed-bed anaerobic reactors. Int J
Hydrogen Energy 38:6137–6145
Perna V, Castelló E, Wenzel J, Zampol C, Fontes Lima DM,
Borzacconi L, Varesche MBA, Zaiat M, Etchebehere C
(2013) Hydrogen production in an upflow anaerobic
packed bed reactor used to treat cheese whey. Int J
Hydrogen Energy 38:54–62
Rosa PRF, Santos SC, Sakamoto IK, Varesche MBA, Silva EL
(2014) Hydrogen production from cheese whey with
ethanol-type fermentation: effect of hydraulic retention
time on the microbial community composition. Int J
Hydrogen Energy 161:10–19
Show KY, Lee DJ, Tay JH, Lin CY, Chang JS (2012) Biohy-
drogen production: current perspectives and the way for-
ward. Int J Hydrogen Energy 37:15616–15631
Siso MIG (1996) The biotechnological utilization of cheese
whey: a review. Biores Technol 57:1–11
Venetsaneas N, Antonopoulou G, Stamatelatou K, Kornaros M,
Lyberatos G (2009) Using cheese whey for hydrogen and
methane generation in a two-stage continuous process with
alternative pH controlling approaches. Biores Technol
100:3713–3717
Walker M, Zhang Y, Heaven S, Banks C (2009) Potential errors in
the quantitative evaluation of biogas production in anaerobic
digestion processes. Bioresour Technol 100:6339–6346
Wang X, Ding J, Ren NQ, Liu BF, Guo WQ (2009) CFD sim-
ulation of an expanded granular sludge bed (EGSB) reactor
for biohydrogen production. Int J Hydrogen Energy
34:9686–9695
Yu HQ, Fang HH (2001) Acidification of mid- and high-strength
dairy wastewaters. Water Res 35:3697–3705
123