You are on page 1of 8

Bioresource Technology 311 (2020) 123504

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Using an expended granular sludge bed reactor for advanced anaerobic T


digestion of food waste pretreated with enzyme: The feasibility and its
performance

Sitong Zhanga, Lianpei Zoua, Yulan Wana, Min Yea,b, Jiongjiong Yea, Yu-You Lia,b, Jianyong Liua,
a
School of Environmental and Chemical Engineering, Shanghai University, 333 Nanchen Road, Shanghai 200444, China
b
Department of Civil and Environmental Engineering, Graduate School of Engineering, Tohoku University, 6-6-06 Aza, Aramaki, Aoba-ku, Sendai, Miyagi 980-8579,
Japan

G R A P H I C A L A B S T R A C T

A R T I C LE I N FO A B S T R A C T

Keywords: The high content of solid organics in food waste (FW) results in a low and unstable anaerobic digestion (AD)
Food waste efficiency. Improving methane production rate and process stability is attracting much attention towards ad-
Advanced anaerobic digestion vanced AD of FW. The feasibility of advanced AD of FW pretreated with enzyme was investigated by batch
Suspended solid experiments and 164 days running of an expanded granular sludge bed (EGSB) reactor. Simulation study based
Expanded granular sludge bed
on the results of batch experiments indicates it is possible to treat enzymatically pretreated FW using an EGSB
Enzymatic pretreatment
reactor. During the running of an EGSB reactor, the organic loading rate went up to 20 g chemical oxygen
demand (COD)/L.d, and the total COD removal rate reached 88%. The significance of this study is to achieve an
advanced AD of enzymatically pretreated FW with a stable and efficient methane production with biogas residue
being reduced greatly.

1. Introduction 2018). First, the high suspended solids (SS) content in FW leads to
hydrolysis being the rate-limiting step (Ma et al., 2018). Second, AD of
Food waste (FW) contains a large amount of organic matters and has FW is always lagged or inhibited by easy acidification and existing li-
a good potential for methane production via anaerobic digestion (AD) pids as well as its byproducts, i.e., long chain fatty acids (LCFAs) (Qi
(Cheng et al., 2018). However, the poor system stability and the low et al., 2019; Zhang et al., 2017). In order to get an advanced AD (AAD)
methane production efficiency are the two main problems (Ye et al., to improve the efficiency of methane production from FW, many studies


Corresponding author.
E-mail address: liujianyong@shu.edu.cn (J. Liu).

https://doi.org/10.1016/j.biortech.2020.123504
Received 6 April 2020; Received in revised form 4 May 2020; Accepted 5 May 2020
Available online 08 May 2020
0960-8524/ © 2020 Elsevier Ltd. All rights reserved.
S. Zhang, et al. Bioresource Technology 311 (2020) 123504

were performed to improve the hydrolysis rate (i.e., using pretreatment Table 1
methods), to maintain the system stability (always using additives) Characteristics of the control and enzymatically pretreated FW and inoculum.
(Gao et al., 2015; Lin et al., 2013). Parameters Pretreated Control Inoculum sludge for batch
Among the pretreatment methods, which include physical pre- experiments
treatment (e.g., thermal/hydrothermal, ultrasonic, microwave), che-
pH 4.3 ± 0.1 4.6 ± 0.1 7.4 ± 0.1
mical pretreatment (e.g., alkali or acid) and enzymatic pretreatment
TCOD (g/L) 124.1 ± 2.3 122.3 ± 0.4 29.6 ± 0.9
(Stabnikova et al., 2008; Tampio et al., 2014), enzymatic pretreatment SCOD (g/L) 103.2 ± 0.3 71.6 ± 0.5 0.5 ± 0.1
performs well with the best hydrolysis effect (Braguglia et al., 2018). TSS (g/L) 21.2 ± 0.5 40.6 ± 0.2 30.7 ± 0.2
The effect of enzymatic pretreatment on hydrolysis kinetics has been VSS (g/L) 20.8 ± 0.4 40.0 ± 0.2 14.5 ± 0.1
studied and it has been proved that enzymatic pretreatment can ef- TS (g/L) 92.0 ± 1.1 93.3 ± 1.0 37.3 ± 1.4
VS (g/L) 82.2 ± 1.0 89.1 ± 1.5 19.1 ± 1.0
fectively improve the hydrolysis rate (Luo et al., 2012). In addition, it is
NH4+-N (mg/L) 1168.9 ± 49.7 70.1 ± 3.6 560.2 ± 1.0
reported that 64% of SS was converted to soluble chemical oxygen
demand (SCOD) with subsequent methane production being improved
by 2–3 times (Kiran et al., 2015). In another study, more solid organics 2.2. Batch anaerobic digestion experiments of FW
in FW were hydrolyzed (80–90%) by using in situ-produced enzyme
(Pleissner et al., 2014). Enzymatic pretreatment is a promising method The batch AD experiments were set up in serum bottles (effective
to improve methane production from FW. Not only that, the effective volume was 1.4 L). Three experimental bottles were set up: only sludge,
enzymatic pretreatment of FW also results in an obvious reduction of control with unpretreated FW, and enzymatically pretreated FW. The
solid content of FW by improving hydrolysis effect greatly, meaning F/M ratio was set at 0.6. The pH was adjusted to an initial value of 7.5
that the SS concentration of FW being reduced much before AD process. using sodium hydroxide (NaOH, 8 M). Each serum bottle was purged
This could be much beneficial to improve methane production with a with high purity nitrogen for 10 min. On day 1, the pH was adjusted
higher organic loading rate (OLR) of the AD reactor. using NaOH (8 M) to suppress the pH drop. The air bath oscillator were
As well known, because of the high SS concentration of FW as a used to maintain the temperature at 35 ± 1 °C for 34 days, and
feedstock, continuous stirred tank reactor (CSTR) is the only used re- samples were taken and analyzed on day 0, 2, 4, 6, 9, 11, 14, 17, 22, 27,
actor to treat FW in practice, with a low OLR of 2–4.5 g VS/L.d and a and 34, respectively.
long hydraulic retention time (HRT) of 15–30 d (Braguglia et al., 2018;
Qiang et al., 2012). If the solid organic matters in FW can be hydrolyzed
2.3. EGSB reactor operation
enough by enzymatic pretreatment, could it be possible to treat FW
using an efficient anaerobic digestion reactor (such as an expanded
Methane conversion of enzymatically pretreated FW was carried out
granular sludge bed (EGSB) reactor)? If so, AAD of FW could be well
using an EGSB reactor (with an effective volume of 6 L). As shown in
expected. The bioconversion of FW to methane as well as the reactor
Fig. 1, two pumps were used to feed enzymatically pretreated FW and
efficiency could both be improved tremendously. Furthermore, a higher
circulate the effluent. The effluent circulation was used to mix the ef-
biodegradation of FW pretreated with enzyme will result in a higher
fluent with the influent with the purpose of buffering the pH value. The
ammoniacal nitrogen and alkalinity production in the AAD system,
temperature of the reactor was maintained at 35 ± 1 °C using a con-
which can help much to address the acidification problem (Gu et al.,
stant temperature water bath. The experiment was divided into two
2019; Zhao et al., 2019).
phases. During phase I (day 0 to day 27), the OLR increased from 1 to
The key point of using an EGSB reactor for AAD of FW is to make
20 g chemical oxygen demand/L.d (g COD/L.d) by increasing the COD
clear the SS accumulation rule and investigate how it affects the long-
concentration of the influent using the supernatant of the enzymatically
term running of an EGSB reactor. Hence, the objectives of this study
pretreated FW. While during phase II (day 28 to day 164), the en-
were: 1) to investigate the hydrolysis and degradation performance of
zymatically pretreated FW was used directely. In phase I, the HRT was
enzymatically pretreated FW via batch AD experiments; 2) to predict
7 days. In phase II, when the OLR was 12, 14, 16 g COD/L.d, the HRT
the SS accumulation rule in an EGSB reactor by mathematical simula-
was 7 days; and when the OLR was 18 and 20 g COD/L.d, the HRT was
tion, and to study the feasibility of AAD of FW using an EGSB reactor; 3)
6 days. The characteristics of the FW used in the EGSB reactor are
to investigate the methane production and reactor performance of an
shown in Table 2. The concentrations of COD, NH4+-N, volatile fatty
EGSB reactor treating FW pretreated with enzyme.
acids (VFAs), pH of the effluent, and the methane production were
measured at same time.
2. Materials and methods
2.4. Analytical methods
2.1. FW and inoculum sludge
The pH was measured by METTLER TOLEDO (FE20). The total so-
The FW was pulverized to 4 mm with a grinder, and then storaged in lids (TS), volatile solids (VS), TSS, VSS, NH4+-N were all measured
a refrigerator (4 °C). Fungal α-amylase and glucoamylase were added according to the standard methods (Baird et al., 2012; Yang et al.,
with the dosage of 32 U/g FW, acid protease was added with the dosage 2019). The COD was measured by a COD digital reactor block and a
of 50 U/g FW. Then, pretreatment was performed in a water bath at visible spectrophotometer (DR-3900 and DRB-200, HACH, America).
55 °C for 8 h. In fact, the increase of SCOD and the reduction of VSS The biogas composition (CH4, CO2) was measured by a gas chromato-
ended at 4 h. In order to make sure an enough hydrolysis reaction, 8 h graphy (GC-7900, Techcomp, China). The VFA was measured by a gas
of hydrolysis was used. The AD inoculum for the batch experiments was chromatography (GC2010-Plus, Shimadzu, Japan).
sampled from an AD digester in a sewage treatment plant in Shanghai. Origin software (version 9) was used to fit the VSS concentrations of
The inoculation sludge was domesticated in a CSTR reactor to maintain sludge, control FW group (FW and sludge) and pretreated FW group
its microbial activity before using. The inoculum of anaerobic granular (pretreated FW and sludge) in the first 22 days. And then, the VSS
sludge used in an EGSB reactor was obtained from an internal circu- concentration of control FW (FW and sludge) and enzymatic pretreated
lation (IC) reactor in a chewing gum factory. The characteristics of the FW (pretreated FW and sludge) subtracted the VSS concentration of the
control FW, pretreated FW and inoculum sludge for batch experiments sludge group to obtain the VSS concentration of the control FW (only
are shown in Table 1. FW) and pretreated FW (only pretreated FW), respectively. Fitting the
VSS data of two groups to obtain the equations of remaining VSS

2
S. Zhang, et al. Bioresource Technology 311 (2020) 123504

Fig. 1. The flow chart of the EGSB reactor.

Table 2 L); C is the OLR (g COD/L.d); R is COD removal rate (%); SRT is the
Characteristics of the enzymatically pretreated FW used in the EGSB reactor. sludge retention time (d); Y is the sludge yield coefficient (kg VSS/kg
Parameters Pretreated FW supernatant Pretreated FW Granular sludge
COD).

pH 4.4 ± 0.1 4.2 ± 0.1 7.5 ± 0.1


TCOD (g/L) – 133.3 ± 2.3 –
3. Results and discussion
SCOD (g/L) 103.2 ± 0.3 108.9 ± 2.4 –
TSS (g/L) – 14.9 ± 0.1 85.3 ± 0.1 3.1. The feasibility of AAD of enzymatically pretreated FW using an EGSB
VSS (g/L) – 14.4 ± 0.1 75.9 ± 0.1 reactor
TS (g/L) – 101.4 ± 2.0 –
VS (g/L) – 97.0 ± 2.5 –
NH4+-N (mg/L) 725.9 ± 35.2 823.1 ± 4.1 – 3.1.1. The methane production and SS reduction of enzymatically
pretreated FW
Fig. 2a and b show the COD removal and methane production of the
concentration (g/L) ((Eq. (1) and (Eq. (2)). Eq. (1) is for control FW, FW without pretreatment (as the control) and the enzymatically pre-
and Eq. (2) is for pretreated FW. Based on the equations for remaining treated FW. During the first 6 days, the methane production of both
VSS concentration, equations of the VSS accumulation (g/L) are ob- groups were only 793.0 mL and 837.6 mL, respectively, while the solid
tained by integration method, Eq. (3) is for control FW, and Eq. (4) is organic matters were hydrolyzed in a large quantity, resulting in an
for pretreated FW. In order to explore the feasibility of treating en- VFA accumulation and a lag phase of methanogenesis (Kawai et al.,
zymatically pretreated FW in the EGSB reactor, it is necessary to cal- 2014; Zou et al., 2016). From day 6, the VFA started to be consumed
culate and compare the daily discharged sludge (g/L.d) and the daily largely. The NH4+-N was produced during methane production pro-
sludge yield (g/L.d). Eq. (5) is the calculation equation for the daily vided a good buffer system (Wang et al., 2013). On day 34, the methane
solid accumulation, and Eq. (6) is for the daily sludge yield. production of the control and the enzymatically pretreated FW were
x
4065.4 mL and 5212.1 mL, respectively. The methane production of the
m (x ) = P × (0.96832 × e(− 2.40525 ) +0.03168) (1) enzymatically pretreated FW was 28.2% higher than that of the control.
x
More SCOD was utilized from the enzymatically pretreated FW, thus
m (x ) = P × (0.97506 × e(− 2.23035 ) +0.02494) (2) producing more methane (Wu et al., 2015). Therefore, enzymatic pre-
treatment is an efficient way to increase the methane production.
y (x ) = P × (2.63349 × (1 − (0.63349)(x + 1) ) + (x + 1) × 0.03463) (3) Fig. 2c and d show the degradation/reduction of the TSS and VSS
y (x ) = P × (2.71590 × (1 − (0.63870)(x + 1) ) + (x + 1) × 0.02494) during the batch experiments. From day 0 to day 34, the TSS con-
(4)
centration of the control and the enzymatically pretreated FW went
XR from 29.9 g/L, 27.8 g/L to 22.9 g/L, 22.9 g/L, respectively. The VSS
Daily discharged sludge =
SRT (5) went from 16.6 g/L, 14.6 g/L to 9.7 g/L, 9.6 g/L, respectively. In fact,
on day 22, the TSS concentration of the control and the enzymatically
Daily sludge yield = C × R × Y (6)
pretreated FW decreased to 23.3 g/L, 23.6 g/L, and the VSS con-
where, “P” is the daily influent SS concentration (g/L); “x” is time (d); centration decreased to 10.0 g/L, 10.3 g/L. It can be seen that the
“e” values 2.718; XR is the initial sludge concentration in the reactor (g/ amount of SS degradation from day 22 to day 34 was very little, so it

3
S. Zhang, et al. Bioresource Technology 311 (2020) 123504

14000

Cumulative methane production (mL)


(a) Control (b) Control
5000 Pretreated Pretreated
12000

4000 10000

SCOD (mg/L)
8000
3000

6000
2000
4000
1000
2000

0 0
0 3 6 9 12 15 18 21 24 27 30 33 0 2 4 6 9 11 14 17 22 27 34

Time (d) Time (d)

30
(c ) Sludge ( d) Sludge
16
Control Control
28 Pretreated Pretreated

14
26

VSS (g/L)
TSS (g/L)

24 12

22 10

20
8
18
0 2 4 6 9 11 14 17 22 27 34 0 2 4 6 9 11 14 17 22 27 34
Time (d) Time (d)

Fig. 2. The methane production (a) and the concentrations of SCOD (b), TSS (c), and VSS (d) in batch experiments.

can be considered that the SS degradation was over on day 22. In this enzymatically pretreated FW, the COD concentration was 124.1 g/L,
study, the FW was mainly composed of easily biodegradable compo- and the concentrations of VSS and inorganic matters were 20.8 g/L and
nents such as starch and protein. Therefore, on day 22, the SS of the 0.4 g/L, respectively. The value of COD was 6.0 and 310.3 times that of
control could be degraded to a comparable value. However, it is worth VSS and inorganic matters. When the OLR of an EGSB reactor was 20 g
noting that the degradation process of SS was much different between COD/L.d, it means that the daily influent COD concentration in the
two groups. As shown in Fig. 2c and d, in the first two days, the solid reactor was 20 g COD/L. According to the ratio of COD to VSS and
organic matters of enzymatically pretreated group were significantly inorganic matters, it could be known that the daily influent VSS con-
lower than that of the control. In addition, due to the high hydrolysis of centrations of control and pretreated group were 6.7 g/L and 3.3 g/L,
enzymatic pretreatment, the SS of the enzymatically pretreated FW on and daily influent inorganic matters concentrations were both 0.1 g/L,
day 0 was ~49% of the control. It is worth noting that there was few respectively. Therefore, the cumulative results were calculated ac-
microorganism in FW, and the pretreatment time was 8 h, so the high cording to the Eqs. (3) and (4) (as shown in Fig. 3). On day 22, the
hydrolysis performance of pretreatment must not due to biological accumulated concentrations of VSS in control and enzymatically pre-
hydrolysis, but due to the enzyme. In an EGSB reactor, large amounts of treated group were 22.9 g/L and 11.0 g/L, respectively. At the same
SS accumulation can affect the reactor stable running with the effective
reactor volume being reduced greatly (Abbassi-Guendouz et al., 2012). 30
Therefore, compared with the control, the enzymatically pretreated
Control
group might have less SS accumulation, so enzymatically pretreated FW Pretreated
25
might be beneficial to an EGSB reactor.
VSS accumulation (g/L)

3.1.2. The prediction of SS accumulation in an EGSB reactor treating 20


enzymatically pretreated FW
The key point of using an EGSB reactor for AAD of FW is to ensure a 15
stable operation of EGSB with a suitable SS accumulation in the reactor.
Thus, a mathematical simulation based on the SS reduction results of
the batch AD experiments was performed to predict how would the SS 10
accumulate in an EGSB reactor. It is important to assume an SRT to
calculate the accumulation of SS. In Fig. 2c and d, as pointed out above,
5
the degradation of the VSS and the methane production stopped on day
22. Thus, the SRT for prediction of SS accumulation in an EGSB reactor
is assumed to be 22 days. 0
As shown in Table 1, the COD concentration of control FW was 0 2 4 6 8 10 12 14 16 18 20 22
122.3 g/L, and the concentrations of VSS and inorganic matters (TSS- Time (d)
VSS) were 40.0 g/L and 0.6 g/L, respectively. The value of COD was 3.0
and 203.8 times that of VSS and inorganic matters. In addition, in Fig. 3. The prediction of the VSS accumulation in an EGSB reactor.

4
S. Zhang, et al. Bioresource Technology 311 (2020) 123504

time, the undegraded inorganic matters of control and pretreated FW slightly.


were both 1.7 g/L, respectively. As a result, in 22 days, the total ac- When the OLR reached 12 g COD/L.d, the methane yield was
cumulated SS were 25.1 g/L and 13.2 g/L, respectively. It is assumed 0.30 m3 CH4/kg CODre, the COD removal rate was 96.7 ± 1.5%. The
that the initial SS concentration in the reactor was 35 g/L, and the high performance was because the growth of biomass in the EGSB re-
accumulated SS concentration of control and pretreated FW accounted actor, the variety of microorganisms in the reactor was abundant, and
for 71.7% and 37.7% of the initial concentration. The pretreated group the degradation of organic matters by microorganisms was good (Liu
accumulated less SS. Too much SS might limit mass transfer, resulting et al., 2010). When the OLR increased to 14, 16, 18 and 20 g COD/L.d,
in the accumulation of VFA and inhibiting the methanogenesis. Even- the methane yield and the COD removal rate decreased slightly, which
tually, the reactor might collapse (Abbassi-Guendouz et al., 2012). might be because the contact between microorganisms and organic
Compared with the enzymatically pretreated FW, the control (FW matters was slightly affected after the sludge concentration increased,
without pretreatment) might be more difficult to facilitate the stable and the degradation of organic matters slightly decreased (Abbassi-
operation of an EGSB reactor. Guendouz et al., 2012). It's worth noting that the AAD of FW was stable
even without sludge discharge during the experiment, which can be
3.1.3. The feasibility of AAD of enzymatically pretreated FW using an EGSB represented by VFA, pH, and NH4+-N in Fig. 4d, Fig. 4e and Fig. 4f. In
reactor studies of Andalib et al. and Fuess et al., corn thin stillage and su-
In order to ensure the effective operation of the EGSB reactor, garcane vinasse were used to produce methane with OLRs of 29 g COD/
sludge discharge is needed. At the same time, it is necessary to consider L.d and 25 g COD/L.d, respectively (Andalib et al., 2012; Fuess et al.,
the amount of microbial growth in the reactor and compare the amount 2017). The OLRs were higher than that of this study, which might be
of sludge discharge with microbial growth. This is because the long- because the FW used in this study also contained cellulose, lipids and
term stable operation of the reactor is possible only when there is a protein, whose biodegradability were lower than that of starch. In ad-
balance between the discharged and the growing microorganisms. dition, in order to reduce the biogas residue as much as possible, the
When the initial SS concentration in the EGSB reactor is 35 g/L and OLR was eventually maintained at a slightly lower value (20 g COD/
the SRT is 22 days, according to Eq. (5), the daily discharge sludge is L.d) in this study. In fact, in studies of methane production from FW,
1.6 g/L.d. Generally, the sludge yield coefficient is low (Contrera et al., CSTR reactors achieved a stable operation at OLRs of 2–4.5 g VS/L.d,
2018). However, in the simulation calculation, considering the sludge and the methane yield ranged from 0.32 to 0.44 L CH4/g VS (Braguglia
discharge is frequent and the sludge growth is faster when the OLR is et al., 2018). It's worth noting that, in this study, the OLR could reach
20 g COD/L.d, the sludge yield coefficient is assumed to be 0.1 kg VSS/ 20 g COD/L.d (~15.4 g VS/L.d), the EGSB reactor had a higher me-
kg COD. If the COD removal rate of the enzymatically pretreated group thane yield. These results show the great advantages of using an EGSB
is 80%, according to Eq. (6), the daily microorganisms growth can be as reactor to treat enzymatically pretreated FW for AAD.
high as 1.6 g/L.d. Therefore, there is a balance between the daily dis-
charge and the daily growth of microorganisms. However, as for the 3.2.2. The reactor stability
control, the accumulated SS can be much higher than that of the pre- The pH, VFA and NH4+-N concentrations are important parameters
treated FW, the effect on the efficiency of the reactor is greater, the affecting methane production and system stability (Kumari et al.,
COD removal rate might be lower, and the growth of microorganisms is 2018). As shown in Fig. 4d–f, in phase II, the nitrogen-containing
correspondingly lower, which is not conducive to the retention of mi- substrate was well degraded, and the NH4+-N concentration increased
croorganisms in the reactor and cause the collapse of the reactor (Cheng from the initial of 790.0 mg/L to ~1500.0 mg/L finally. With the in-
et al., 2018). As a result, the EGSB reactor has the potential for a long- crease of OLR, the VFA increased slightly, and maintained at 380.6 mg/
term efficient treatment of enzymatically pretreated FW. L with OLR of 20 g COD/L.d. This was due to the increase of SS in the
reactor, which affected mass transfer, resulting in a slight increase of
3.2. The performance of an EGSB reactor treating FW pretreated by enzyme VFA (Abbassi-Guendouz et al., 2012), but this did not affect the per-
formance of the reactor. A previous study showed that when the VFA
3.2.1. The methane production and COD removal efficiency with a high concentration of the reactor was lower than 400 mg/L, the reactor
OLR could be operated stably (Bakraoui et al., 2020). In this study, the
In order to verify the above conclusion of batch experimental and balance between VFA and NH4+-N kept the pH in the range of 7.6–7.8,
theoretical calculation study, AAD of the enzymatically pretreated FW which ensured the stable methane production in the reactor. The EGSB
was performed using an EGSB reactor based on the operation factors reactor can achieve a good buffering performance and a long-term
obtained above. The OLR of the reactor was aimed to be 12 g COD/L.d. stable operation for AD of enzymatically pretreated FW.
The reactor operation was divided into two phases, as shown in Fig. 4a.
In phase I, only the supernatant of the pretreated FW was used to in- 3.2.3. The accumulation of SS in the EGSB reactor and operation
crease the OLR from 1 to 12 g COD/L.d. In phase II, the pretreated FW suggestions
was used for the performance study, with a stable running of 85 days at During the whole EGSB reactor running (164 days including phase I
OLR of 12 g COD/L.d and then being increased to 20 g COD/L.d finally. and phase II), none sludge discharge was performed because the reactor
As illustrated in Fig. 4b and Fig. 4c, the methane production and was ever stable with good methane production. The MLSS and MLVSS
COD removal performances of the EGSB reactor with different OLRs concentrations in the reactor were tested for a few times during the two
were investigated. In phase I, the stable operation of the reactor with phases, as shown in Fig. 5a. It could be seen that, from day 38 to day
rapidly increase of OLR indicated the EGSB reactor had a high impact 164, the concentrations of the MLSS and MLVSS increased from 36.6
resistance (Zhao et al., 2019). In phase II, with the OLR of 12, 14, 16, 18 and 32.5 g/L to 59.6 and 53.6 g/L, respectively. The SS accumulation
and 20 g COD/L.d, the biogas production increased with the increase of can be divided into two stages, stage I (from day 38 to day 44) and stage
OLR, the biogas production rates were 5.2 ± 0.3, 5.7 ± 0.3, II (from day 44 to day 164). As shown in Fig. 5b, the SS accumulation
6.4 ± 0.2, 6.9 ± 0.2 and 7.8 ± 0.1 L/L.d, respectively. Meanwhile, presented in a relatively smooth linear relationship. In stage I and stage
the methane yields were 0.30, 0.28, 0.28, 0.26, 0.27 m3 CH4/kg CODre, II, the daily SS growth rates were 0.3 g/L.d and 0.2 g/L.d, respectively.
respectively. In addition, the proportion of methane was mostly in the The SS growth rate of stage II was much lower than that of stage I, and
range of 60–64%, which indicated that the digestion process was stable. much lower than expected from the above mathematical calculation
The corresponding COD removal rates were 96.7 ± 1.5%, (the daily growth rate was 0.6 g/L.d). Less SS was accumulated in the
92.1 ± 3.4%, 88 ± 2.3%, 84.1 ± 2.3%, 88.4 ± 3.2%. As the OLR EGSB reactor might be due to the better degradation performance and
increased, the methane yield and the COD removal rate decreased better mass transfer in the EGSB reactor. In addition, the

5
S. Zhang, et al. Bioresource Technology 311 (2020) 123504

24 Phase I Phase II
(a)

OLR (g COD/L.d)
20
16
12
8
4
0
50000 (b) Biogas production
Gas production (mL)

Methane production
40000
30000
20000
10000
0
COD concentration (mg/L)

(c) 100
120000

Removal rate (%)


90000 80

Influent 60
60000
Effluent
Removal rate 40
30000
20
0

9.5 (d) Reactor


Effluent
9.0
8.5
pH

8.0
7.5
7.0

2000 (e)
NH4+-N (mg/L)

1500

1000

500

600 (f) VFA


Acetic acid
VFAs (mg/L)

Propionic acid
400 Butyric acid

200

0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170
Time (d)
Fig. 4. The OLR (a), gas production (b), COD (c), pH (d), NH4+-N (e), VFAs (f) in the EGSB reactor.

microorganisms were maintained in the reactor resulting in a large optimal conditions in practical engineering practice could be predicted.
number of methanogens (Luo et al., 2015; Ma et al., 2018). When the initial sludge concentration in the reactor is 30–35 g/L and
In fact, it is not economically feasible to use commercial enzymes in the daily discharged sludge is 0.2 g/L.d. According to Eq. (5), SRT could
practical engineering practice. If use FW which contains a large amount be set at between 150 and 175 days depending on the components of
of organic matters as a substrate for in situ enzyme production can the FW and the effect of enzymatic pretreatment, and the OLR could be
greatly reduce the cost of enzymatic pretreatment, and it is possible to set at between 15 and 20 g COD/L.d. The HRT could be set depending
obtain better hydrolysis performance and higher methane production on the OLR and SRT, and the SS concentration of the feeded FW.
than commercial enzyme (Khan et al., 2015). At the same time, ac- Anyway, all the operation factors, which depend on the FW char-
cording to SS accumulation results from the EGSB reactor running, the acteristics and the performance of enzymatic pretreatment greatly,

6
S. Zhang, et al. Bioresource Technology 311 (2020) 123504

Sludge concentration (g/L)


60
(a) MLSS
50 MLVSS
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170
40
Increased sludge (g/L)

(b) MLSS
30

20

10

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170


40
Increased sludge (g/L)

(c) MLVSS
30

20

10

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170


Time (d)
Fig. 5. The SS concentration (a) and TSS (b) and VSS (c) accumulation in the EGSB reactor.

should be worked out by pre-experiments. CRediT authorship contribution statement

3.2.4. The COD balance Sitong Zhang: Investigation, Writing - original draft. Lianpei Zou:
The COD balance of the EGSB reactor during the 164 days running Formal analysis, Methodology. Yulan Wan: Investigation, Data cura-
was calculated to better evaluate its performance. The total COD con- tion. Min Ye: Methodology. Jiongjiong Ye: Methodology, Writing -
sisted of methane production COD, biomass COD, effluent COD and review & editing. Yu-You Li: Conceptualization. Jianyong Liu: Project
sampling COD. The synthesis of 1 g of biomass would consume 1.42 g administration, Funding acquisition, Writing - review & editing,
COD, and 1 g COD could be converted into 0.35 L CH4 in the standard Supervision.
state (Chen et al., 2019). As illustrated in Fig. 6, 81.1% of the total COD
was converted into methane, meaning that an AAD of FW was obtained.
Declaration of Competing Interest
The biomass COD was 1.8%, indication most of the solid organics were
degraded with a very low SS accumulation, which is very important for
AD of FW using an EGSB reactor. About 6.8% of the total COD cannot The authors declare that they have no known competing financial
be accounted due to sampling and test errors, which is reasonable for a interests or personal relationships that could have appeared to influ-
pilot EGSB reactor treating FW (Khan et al., 2015). ence the work reported in this paper.

4. Conclusions Acknowledgements

The feasibility and performance of AAD of enzymatically pretreated This study was financially supported by the National Natural
FW in an EGSB reactor was investigated using batch experiments and Science Foundation of China (51778352), the Science and Technology
EGSB reactor operation. Simulation study based on results of batch Commission of Shanghai Municipality (18230710900).
experiments indicated enzymatically pretreated FW could be treated
using an EGSB reactor. The performance of the EGSB were investigated
with different OLRs. When the OLR was 20 g COD/L.d, the methane Appendix A. Supplementary data
yield and COD removal rate were 0.27 m3 CH4/kg CODre and 88%,
respectively. The significance of this study is to make it possible to Supplementary data to this article can be found online at https://
realize AAD of enzymatically pretreated FW in an EGSB reactor. doi.org/10.1016/j.biortech.2020.123504.

7
S. Zhang, et al. Bioresource Technology 311 (2020) 123504

COD loss during sampling


Collected methane
Unaccouted COD
Biomass COD
Effluent COD
1.8%

6.8%

81.1%

8.1%

2.2%

Fig. 6. The COD balance.

References 245–253.
Liu, J., Zhong, J., Wang, Y., Liu, Q., Qian, G., Zhong, L., Guo, R., Zhang, P., Xu, Z.P., 2010.
Effective bio-treatment of fresh leachate from pretreated municipal solid waste in an
Abbassi-Guendouz, A., Brockmann, D., Trably, E., Dumas, C., Delgenes, J.P., Steyer, J.P., expanded granular sludge bed bioreactor. Bioresour. Technol. 101 (5), 1447–1452.
Escudie, R., 2012. Total solids content drives high solid anaerobic digestion via mass Luo, J., Qian, G., Liu, J., Xu, Z.P., 2015. Anaerobic methanogenesis of fresh leachate from
transfer limitation. Bioresour. Technol. 111, 55–61. municipal solid waste: A brief review on current progress. Renew. Sustain. Energy
Andalib, M., Hafez, H., Elbeshbishy, E., Nakhla, G., Zhu, J., 2012. Treatment of thin Rev. 49, 21–28.
stillage in a high-rate anaerobic fluidized bed bioreactor (AFBR). Bioresour. Technol. Luo, K., Yang, Q., Li, X.-M., Yang, G.-J., Liu, Y., Wang, D.-B., Zheng, W., Zeng, G.-M.,
121, 411–418. 2012. Hydrolysis kinetics in anaerobic digestion of waste activated sludge enhanced
Baird, R.B., Bridgewater, L., Clesceri, L.S., Eaton, A.D., Rice, E.W., 2012. Standard by α-amylase. Biochem. Eng. J. 62, 17–21.
Methods for the Examination of Water and Wastewater. American Public Health Ma, C., Liu, J., Ye, M., Zou, L., Qian, G., Li, Y.-Y., 2018. Towards utmost bioenergy
Association. conversion efficiency of food waste: Pretreatment, co-digestion, and reactor type.
Bakraoui, M., Karouach, F., Ouhammou, B., Aggour, M., Essamri, A., El Bari, H., 2020. Renew. Sustain. Energy Rev. 90, 700–709.
Biogas production from recycled paper mill wastewater by UASB digester: Optimal Pleissner, D., Kwan, T.H., Lin, C.S., 2014. Fungal hydrolysis in submerged fermentation
and mesophilic conditions. Biotechnolo. Rep. 25, e00402. for food waste treatment and fermentation feedstock preparation. Bioresour. Technol.
Braguglia, C.M., Gallipoli, A., Gianico, A., Pagliaccia, P., 2018. Anaerobic bioconversion 158, 48–54.
of food waste into energy: A critical review. Bioresour. Technol. 248, 37–56. Qi, G., Meng, W., Zha, J., Zhang, S., Yu, S., Liu, J., Ren, L., 2019. A novel insight into the
Chen, R., Wen, W., Jiang, H., Lei, Z., Li, M., Li, Y.Y., 2019. Energy recovery potential of influence of thermal pretreatment temperature on the anaerobic digestion perfor-
thermophilic high-solids co-digestion of coffee processing wastewater and waste ac- mance of floatable oil-recovered food waste: Intrinsic transformation of materials and
tivated sludge by anaerobic membrane bioreactor. Bioresour. Technol. 274, 127–133. microbial response. Bioresour. Technol. 293, 122021.
Cheng, H., Hiro, Y., Hojo, T., Li, Y.Y., 2018. Upgrading methane fermentation of food Qiang, H., Lang, D.L., Li, Y.Y., 2012. High-solid mesophilic methane fermentation of food
waste by using a hollow fiber type anaerobic membrane bioreactor. Bioresour. waste with an emphasis on Iron, Cobalt, and Nickel requirements. Bioresour.
Technol. 267, 386–394. Technol. 103 (1), 21–27.
Contrera, R.C., Lucero Culi, M.J., Morita, D.M., Rodrigues, J.A.D., Zaiat, M., Schalch, V., Stabnikova, O., Liu, X.Y., Wang, J.Y., 2008. Digestion of frozen/thawed food waste in the
2018. Biomass growth and its mobility in an AnSBBR treating landfill leachate. Waste hybrid anaerobic solid-liquid system. Waste Manage. 28 (9), 1654–1659.
Manage. 82, 37–50. Tampio, E., Ervasti, S., Paavola, T., Heaven, S., Banks, C., Rintala, J., 2014. Anaerobic
Fuess, L.T., Kiyuna, L.S.M., Ferraz, A.D.N., Persinoti, G.F., Squina, F.M., Garcia, M.L., digestion of autoclaved and untreated food waste. Waste Manage. 34 (2), 370–377.
Zaiat, M., 2017. Thermophilic two-phase anaerobic digestion using an innovative Wang, Q., Peng, L., Su, H., 2013. The effect of a buffer function on the semi-continuous
fixed-bed reactor for enhanced organic matter removal and bioenergy recovery from anaerobic digestion. Bioresour. Technol. 139, 43–49.
sugarcane vinasse. Appl. Energy 189, 480–491. Wu, C., Jin, L., Zhang, P., Zhang, G., 2015. Effects of potassium ferrate oxidation on
Gao, S., Huang, Y., Yang, L., Wang, H., Zhao, M., Xu, Z., Huang, Z., Ruan, W., 2015. sludge disintegration, dewaterability and anaerobic biodegradation. Int. Biodeter.
Evaluation the anaerobic digestion performance of solid residual kitchen waste by Biodegr. 102, 137–142.
NaHCO3 buffering. Energy Convers. Manage. 93, 166–174. Yang, H., Liu, J., Hu, P., Zou, L., Li, Y.Y., 2019. Carbon source and phosphorus recovery
Gu, N., Liu, J., Ye, J., Chang, N., Li, Y.Y., 2019. Bioenergy, ammonia and humic sub- from iron-enhanced primary sludge via anaerobic fermentation and sulfate reduction:
stances recovery from municipal solid waste leachate: A review and process in- Performance and future application. Bioresour. Technol. 294, 122174.
tegration. Bioresour. Technol. 293, 122159. Ye, M., Liu, J., Ma, C., Li, Y.-Y., Zou, L., Qian, G., Xu, Z.P., 2018. Improving the stability
Kawai, M., Nagao, N., Tajima, N., Niwa, C., Matsuyama, T., Toda, T., 2014. The effect of and efficiency of anaerobic digestion of food waste using additives: A critical review.
the labile organic fraction in food waste and the substrate/inoculum ratio on anae- J. Cleaner Prod. 192, 316–326.
robic digestion for a reliable methane yield. Bioresour. Technol. 157, 174–180. Zhang, W., Lang, Q., Fang, M., Li, X., Bah, H., Dong, H., Dong, R., 2017. Combined effect
Khan, A.A., Mehrotra, I., Kazmi, A.A., 2015. Sludge profiling at varied organic loadings of crude fat content and initial substrate concentration on batch anaerobic digestion
and performance evaluation of UASB reactor treating sewage. Biosyst. Eng. 131, characteristics of food waste. Bioresour. Technol. 232, 304–312.
32–40. Zhao, L., Su, C., Chen, S., Ye, Z., Wei, X., Yao, T., Li, G., Wang, P., 2019. Expanded
Kiran, E.U., Trzcinski, A.P., Liu, Y., 2015. Enhancing the hydrolysis and methane pro- granular sludge blanket reactor treatment of food waste at ambient temperature:
duction potential of mixed food waste by an effective enzymatic pretreatment. Analysis of nitrogen compositions and microbial community structure. Bioresour.
Bioresour. Technol. 183, 47–52. Technol. 294, 122134.
Kumari, K., Suresh, S., Arisutha, S., Sudhakar, K., 2018. Anaerobic co-digestion of dif- Zou, L., Ma, C., Liu, J., Li, M., Ye, M., Qian, G., 2016. Pretreatment of food waste with
ferent wastes in a UASB reactor. Waste Manage. 77, 545–554. high voltage pulse discharge towards methane production enhancement. Bioresour.
Lin, Y., Lu, F., Shao, L., He, P., 2013. Influence of bicarbonate buffer on the methano- Technol. 222, 82–88.
genetic pathway during thermophilic anaerobic digestion. Bioresour. Technol. 137,

You might also like