You are on page 1of 10

Environmental Technology

ISSN: 0959-3330 (Print) 1479-487X (Online) Journal homepage: http://www.tandfonline.com/loi/tent20

Performance assessment of two-stage anaerobic
digestion of kitchen wastes

Zhang Bo & He Pin-jing

To cite this article: Zhang Bo & He Pin-jing (2014) Performance assessment of two-stage
anaerobic digestion of kitchen wastes, Environmental Technology, 35:10, 1277-1285, DOI:
10.1080/09593330.2013.866169

To link to this article: http://dx.doi.org/10.1080/09593330.2013.866169

Accepted author version posted online: 15
Nov 2013.
Published online: 17 Dec 2013.

Submit your article to this journal

Article views: 340

View related articles

View Crossmark data

Citing articles: 14 View citing articles

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tent20

Download by: [University of Canberra] Date: 23 October 2017, At: 07:00
Environmental Technology, 2014
Vol. 35, No. 10, 1277–1285, http://dx.doi.org/10.1080/09593330.2013.866169

Performance assessment of two-stage anaerobic digestion of kitchen wastes
Zhang Boa,b∗ and He Pin-jingb
a School
of Environmental Science & Engineering, Shanghai Jiaotong University, 800 Dongchuan Road, Shanghai 200240,
People’s Republic of China; b State Key Laboratory of Pollution Control & Resources Reuse, Tongji University, Shanghai 200092,
People’s Republic of China
(Received 10 April 2013; accepted 9 November 2013 )

This study is aimed at investigating the performance of the two-phase anaerobic digestion of kitchen wastes in a lab-scale
setup. The semi-continuous experiment showed that the two-phase anaerobic digestion of kitchen wastes had a bioconversion
rate of 83%, biogas yield of 338 mL·(g chemical oxygen demand (COD))−1 and total solid conversion of 63% when the
entire two-phase anaerobic digestion process was subjected to an organic loading rate (OLR) of 10.7 g·(L d)−1 . In the
hydrolysis–acidogenesis process, the efficiency of solubilization decreased from 72.6% to 41.1%, and the acidogenesis
Downloaded by [University of Canberra] at 07:00 23 October 2017

efficiency decreased from 31.8% to 17.8% with an increase in the COD loading rate. On the other hand, the performance of
the subsequent methanogenic process was not susceptible to the increase in the feeding COD loading rate in the hydrolysis–
acidogenesis stage. Lactic acid was one of the main fermentation products, accounting for over 40% of the total soluble COD
in the fermentation liquid. The batch experiments indicated that the lactic acid was the earliest predominant fermentation
product, and distributions of fermentation products were pH dependent. Results showed that increasing the feeding OLR of
kitchen wastes made the two-stage anaerobic digestion process more effective. Moreover, there was a potential improvement
in the performance of anaerobic digestion of kitchen wastes with a corresponding improvement in the hydrolysis process.
Keywords: two-phase; kitchen wastes; anaerobic digestion; hydrolysis–acidogenesis; methanogenesis

1. Introduction wastes, two-stage wet anaerobic digestion has usually been
Kitchen wastes are one of the major components of the employed in methane recovery from kitchen wastes. Cho
municipal organic wastes. Considering the city of Shang- and Park [13] applied the two-stage anaerobic digestion
hai for instance, about 3 million tons of kitchen wastes are process to experiments on Korean food wastes, Bouallagui
discharged each year, which accounts for about 58% of the et al. [14] on fruit and vegetable wastes and Wang et al.
total municipal solid wastes.[1,2] Kitchen wastes are the [15,16] developed a hybrid bioreactor for food wastes. Most
main sources of decay, odour and leachate due to their high of these studies focused on the evaluation of the ultimate
volatile solids (VSs) and moisture content.[3] Regulations bioenergy production potential. However, there is lack of
and laws on the discharge of kitchen wastes have been grad- information regarding the limiting factors and mass balance
ually established in China to promote the reduction and of the two-stage anaerobic digestion process.
recovery of kitchen wastes. Anaerobic digestion is consid- It is generally considered that the slowest step in a
ered to be an appropriate technology for kitchen wastes process is the limiting factor. Anaerobic degradation of
considering its biodegradability, recovery of energy and cellulose-poor wastes such as putrefactive organic wastes,
structural strength such as moisture content, bulk density e.g. kitchen wastes, is limited by methanogenesis rather
and compressibility, compared with landfill, incineration than by hydrolysis in a single-stage anaerobic digestion
and composting. It has been reported that the anaerobic process.[14] However, it is still not clear which step, hydrol-
digestion technology is effective for the reduction of kitchen ysis or methanogenesis, is the limiting step for the two-stage
wastes and recovery of methane biogas.[4–9] anaerobic digestion process of kitchen wastes, because
Two-stage anaerobic digestion, compared with its kitchen wastes were considered to be easily biodegradable.
single-stage counterpart, could separately optimize the con- In the two-stage anaerobic digestion process, the solubi-
ditions for the hydrolytic and acidogenic microorganisms as lization efficiency and distribution of organic acids in the
well as for the acetogenic–methanogenic microorganisms, hydrolysis–acidogenesis process determine the subsequent
and thus improve the stability of the system.[10–12] Due to methanogenesis efficiency. The performance of two-stage
the high moisture and high organic matter content of kitchen anaerobic digestion can be enhanced by regulating and

∗ Corresponding author. Email: zhangbo214@sjtu.edu.cn

© 2013 Taylor & Francis
1278 Z. Bo and H. Pin-jing

controlling the solubilization process, such as controlling 2.2. Two-stage anaerobic digestion operation
pH, bulking agent and inoculum to substrate ratio.[17,18] Anaerobic digestion of kitchen wastes was conducted in
Organic acids produced in biowastes digestion are mainly a two-phase anaerobic system. The working volumes of
volatile fatty acids (VFAs), and their production rate and the hydrolysis–acidogenesis reactor and the methanogenic
species distribution are dependent on pH and oxidative reactor were 2 and 1.6 L, respectively. The two reactors were
and reductive potential in the surrounding media for the placed in a chemostat room under a temperature of 35◦ C.
microorganisms.[19–21] However, for kitchen wastes, lac- The reactors were fed intermittently and operated under
tic acid is also one of the main organic acids, constitut- a cycle of 12 h including 30 min for feeding, 11 h for reac-
ing approximately 50% of the chemical oxygen demand tion and 30 min for sedimentation and sludge removal. The
(COD) in the hydrolysis–acidogenesis liquid. In addition, acidified kitchen wastes in the hydrolysis reactor were first
its production is considered to have an adverse impact on centrifuged with the supernatant later being pumped into the
the following methanogenesis.[11,19,22,23] However, the subsequent methanogenic reactor, while the solid fractions
VFAs were paid more attention in the past work; it is rarely are being recycled into the hydrolysis–acidogenesis reac-
reported for the role of lactic acid in the two-stage anaerobic tor. Both the hydrolysis–acidogenesis and methanogenic
digestion of kitchen wastes.[24,25] reactors were mixed by recycling the effluent.
To have a full understanding of the two-phase anaerobic The kitchen wastes were diluted with tap water until
digestion process of kitchen wastes, this study investigated reaching a TS of 100 g L−1 . The feeding volume of kitchen
Downloaded by [University of Canberra] at 07:00 23 October 2017

the hydrolysis–acidogenesis process of kitchen wastes and wastes was increased from 10 mL to 20, 40, 60, 80 and
methanogenesis of the hydrolysis–acidogenesis liquid in 100 mL in every cycle. The methanogenic reactor effluent
batch and semi-continuous mode. The pH, VFAs, lactic was recycled into the feeding mixture for the hydrolysis–
acid, biogas constituents at different organic loading rates acidogenesis stage to control the feeding flow volume at
(OLRs) in the hydrolysis–acidogenesis process, the COD 400 mL and hydraulic retention time (HRT) on 5 d. The
conversion efficiency, biogas production rate and the con- HRT in the methanogenic stage was maintained on 4 d.
stituents of organic acids in the effluent in the methanogenic No extra inoculum was introduced into the hydrolysis–
process were analysed. In addition, mass balance analysis acidogenesis reactor.
was also carried out to further assess the performance of the The methanogenic reactor was inoculated with
two-stage anaerobic digestion of kitchen wastes. mesophilic-digested sludge obtained from a wastewater
plant’s sludge digester located in Shanghai, China. The
sludge mixed liquor suspended solid was maintained at
10 g L−1 . The sludge in the methanogenic reactor was ini-
2. Materials and method
tially cultivated with a synthetic medium with glucose and
2.1. Origin and characteristics of kitchen wastes sodium acetate as the carbon source, ammonium chloride
The kitchen wastes collected from a student restaurant as the nitrogen source and dipotassium hydrogen phos-
mainly contained cooked rice, vegetables, meat, eggs, pota- phate as the phosphorus source. The COD, N and P ratio
toes and common salt; the major characteristics of which was 200:5:1. The fresh synthetic medium with a COD of
are shown in Table 1. The total solid (TS) content and 25 g L−1 was fed once a day and the HRT controlled on
VS were 20% and 19%, respectively. The ratio of total 5 d. When about 87% of the feeding COD was converted to
organic carbon (TOC), total nitrogen and total phosphate CH4 at a loading rate of 5 g·(L d)−1 , the start-up period was
was 100:4.16:0.25. The mean pH of feeding kitchen wastes considered to be complete.
was as low as 3.9 due to natural acidogenesis. The kitchen The whole experiment was carried out over six
wastes were composed of 13% proteins, 10% lipids and runs (Run1: OLR = 2.3 g · (L d)−1 ; Run2: OLR = 4.1 g ·
70% carbohydrate. The collected wastes were ground and (L d)−1 ; Run3: OLR = 6.3 g · (L d)−1 ; Run4: OLR =
homogenized in a blender and were stored at 4◦ C before 8.4 g · (L d)−1 ; Run5: OLR = 10.7 g · (L d)−1 and Run6:
feeding. OLR = 13.1 g · (L d)−1 ). For each OLR to the hydrolysis–
acidogenesis reactor, the kitchen wastes were prepared for
the entire experiment, hence maintaining the same feeding
Table 1. Characteristics of kitchen wastes. COD concentration.
Parameters Value

TS (%) 20 ± 3.5 2.3. Batch experiments of the hydrolysis–acidogenesis
VS (%) 19 ± 2.3
Carbohydrates (%TS) 70 ± 2.1 of kitchen wastes
Protein (%TS) 13 ± 2.1 The batch experiments were conducted to evaluate the vari-
Lipids (%TS) 10 ± 1.5 ations in pH, VFA and lactic acid as the functions of time in
pH 3.9 ± 0.1 the hydrolysis–acidogenesis stage. The batch experiments
C/N/P 100:4.16:0.25
were performed for kitchen wastes with 12.5% TS in 500 ml
Environmental Technology 1279

glass-flask reactors at a constant temperature of 35◦ C. Three Acidogenesis efficiency
identical reactors were employed with two reactors being Output VFA(calculated as COD, g)
batch operated for 15 days at controlled-pH values of 5 and = %
7, with the third reactor without pH control serving as the Total input COD (g)
control. The flasks were sealed with neoprene bags and sil-
icone rubber sealant and agitated in a water bath shaker at 3. Results and discussion
100 rpm. The flasks were connected to inverted glass tubes 3.1. Hydrolysis–acidogenesis process
filled with water acidified to pH 2. Biogas was collected
by water displacement. About 300 ml of kitchen wastes 3.1.1. Solubilization efficiency and acidogenesis
were placed into every flask, and no extra inoculum was efficiency
introduced into the reactors. The pH was monitored with The variations in pH, VFA and COD concentration in
a pH meter, and NaOH (5 mol L−1 ) and HCl (5 mol L−1 ) the hydrolysis–acidogenesis liquid with OLR increases are
were automatically added to control the pH at a desired shown in Figure 1(a). In addition, the average operational
value. result in every run of the hydrolysis–acidogenesis stage is
summarized in Table 2.
Although the pH initially declined gradually with
2.4. Analytical methods the increasing COD loading rate from Run 1 to 3, the
Downloaded by [University of Canberra] at 07:00 23 October 2017

pH, TS, VS, COD, NH+ 4 -N and TP were analysed in
steady state of pH at about 5 could be maintained, which
accordance with the Standard Methods.[26] TOC was mea- was attributed to the recirculation of the methanogenic
sured by catalytic oxidation on a multi-N/C 3000 analyser effluent in Run 4–6. The methanogenic effluent pro-
(Vario TOC, German). The protein content was deter- vided the bicarbonate alkalinity for pH buffering in the
mined based on Kjeldahl nitrogen, which was measured hydrolysis–acidogenesis process and could introduce anaer-
by acid hydrolysis of the insoluble organic nitrogen and obic activated sludge into the hydrolysis–acidogenesis
the Kjeldahl nitrogen titration procedure.[26] The analy- reactor.[29–31]
sis of the digestion and distillation process was carried As shown in Figure 1(a and b), and Table 2, despite
out using the nitrogen analyser (BÜCHI, Digestion Unit the VFA and COD concentrations increasing to 12.1 and
K-424; BÜCHI, distillation Unit B-324). The measured 33.8 g L−1 in Run 5, respectively, the solubilization and
Kjeldahl nitrogen was multiplied by 6.25 to give the crude acidogenesis efficiencies showed an obvious decrease with
protein content.[27] Lipids concentration was determined increase in the COD loading rate. The solubilization effi-
gravimetrically after extraction of lipids by petroleum ether ciency was kept at about 70% in Run 1–3, about 60%
according to the Soxlett extraction method.[27] Carbohy- in Run 4 and 5 and about 40% in Run 6. The acido-
drate in the wastes was determined in accordance with the genesis efficiency also changed gradually from 31.8% to
phenol–sulphuric acid method.[28] The VFA and lactic acid 17.8%. This showed that the performance of the hydrolysis–
were determined by a high-performance liquid chromatog- acidogenesis stage in two-phase anaerobic digestion was
raphy (Agilent 2010A, USA) equipped with a C18 column. susceptible to increases in the COD loading rate.
Gas chromatography (GC; Shimadzu GC-14B, Japan) was
used to determine the biogas composition. The GC was 3.1.2. Distribution of hydrolysis–acidogenesis products
equipped with a 3 mm × 2 m TDX-102 filled column and
a thermal conductivity detector. For the determination of The two-phase anaerobic digestion process is able to detect
Na, K and Ca in the liquid phase, the sample was diluted the distribution of fermentation products, thus making
and then filtered through a 4.4-μm paper and a 0.45- it possible to clarify the fermentation products of the
μm membrane filter. The filtrate was analysed by means hydrolysis–acidogenesis stage and the feeding substrate
of an inductively coupled plasma atomic emission spec- constituents of the methanogenic stage. The distribution of
trophotometer (Jarrell-Ash ICAP 9000, Fisher Scientific, hydrolysis–acidogenesis products at different COD load-
USA). ing rates is shown in Figure 1(b). The VFA and lactic
acid were the main constituents of organic acids. Despite
the decrease in acidogenesis efficiency, there was no obvi-
2.5. Calculation methods of solubilization efficiency ous decrease in VFA to the soluble COD ratio. However,
and acidogenesis efficiency the lactic acid concentration in the soluble COD increased
gradually with increases in OLR from Run 1 to 4, and
The efficiencies of solubilization and acidogenesis were
constituted about 40% of the output soluble COD in Run
calculated as follows:
4–6. In addition, as seen in Figure 1(b), the ratio of total
hydrolysis–acidogenesis products to soluble output COD
Solubilization efficiency showed an increase with the increase in the COD loading
Output soluble COD (g) rate. A ratio greater than 80% was attained for Run 4–6.
= %
Total input COD (g) This implied that the accumulation of monomeric products
1280 Z. Bo and H. Pin-jing

(a) 20 40
Run1 Run2 Run3 Run4 Run5 Run6
35
16
30

pH and OLR(g.(L.d)–1

COD and VFA(g.L–1)
12 25

20
8 15

10
4
5

0 0
0 20 40 60 80 100 120
Time (d)
Downloaded by [University of Canberra] at 07:00 23 October 2017

OLR VFA
pH COD

(b)
13.4 VFA(43%)

10.7 VFA(35%)
OLR(g.(L.d)–1)

8.4 VFA(40%)
7
6.3 VFA(37%) Ethanol
Butyric acid
4.1 VFA(35%) Propionic acid
Acetic acid
2.3 VFA(44%) Lactic acid

0 20 40 60 80 100
Fraction of fermentation products in output soluble COD(%)

(c) 100 1.0
H2
N2
80 0.8
CH4
H2,N2,CH4and CO2(%)

Biogas productivity

CO2
60 0.6
(L.(L.d)–1)

Biogas productivity
40 0.4

20 0.2

0 0.0
0 2 4 6 8 10 12 14

Feeding COD loading rate
(g.(L.d) –1)
Figure 1. Hydrolysis and acidogenesis of kitchen wastes in semi-continuous experiments: (a) pH, VFA and COD; (b) constituents of
soluble output COD and (c) biogas constituents, at different feeding COD loading rates.

rarely occurred. Whereas the fermentation of monomers Figure 1(b) shows that the butyric acid was the predomi-
proceeded quickly, the hydrolysis of particulate polymers nant VFA species for all COD loading rates. The acetic acid
became the rate-limiting step. concentration decreased with the increase in COD loading
Environmental Technology 1281

Table 2. Average operational results of the hydrolysis–acidogenesis process at different COD loading rates.

Run 1 Run 2 Run 3 Run 4 Run 5 Run 6

OLR (g (L d)−1 ) 2.3 4.1 6.3 8.4 10.7 13.1
pH 6.98 ± 0.23 6.69 ± 0.11 6.22 ± 0.41 4.98 ± 0.16 5.05 ± 0.34 5.01 ± 0.26
VFA (g L−1 ) 3.7 ± 0.2 5.2 ± 0.9 8.4 ± 0.7 10.3 ± 0.7 12.1 ± 0.3 11.6 ± 0.4
COD (g L−1 ) 8.4 ± 2.4 14.8 ± 0.9 22.5 ± 1.9 25.6 ± 4.4 33.8 ± 7.5 26.9 ± 8.5
Solubilization efficiency (%) 72.6 72.6 70.9 61.3 63.1 41.1
Acidification efficiency (%) 31.8 25.4 26.6 24.4 22.6 17.8
TS conversion efficiency (%) 75 75 68 65 63 60

40 3. It has been reported that lactic acid fermentation could
VFA(g.L–1) Lactic acid(g.L–1)

pH=5
32 no pH adjustment inhibit the growth of putrefactive and food poisoning bacte-
24 pH=7 ria responsible for the production of VFA if the pH was not
16 controlled.[22,23] When micro-organisms are concerned,
8
lactic acid is usually the precursor of propionic acid during
40
the anaerobic digestion,[33,34] whereas propionic acid was
Downloaded by [University of Canberra] at 07:00 23 October 2017

6
32
24 not favourable to the following methanogenesis.[35]
16 The single-stage anaerobic digestion of kitchen wastes
8 easily failed due to the fast decrease in pH at the ini-
0
0 tial period. It was generally considered that VFA caused
8
the decrease in pH. It can be speculated from this study
6 that lactic acid probably contributed to pH decrease at
pH

4 the initial period for the single-stage anaerobic digestion
2 process because the pH in the single-phase anaerobic pro-
0 2 4 6 8 10 12 14 16 cess was maintained at about 7, whereas neutral pH was
Time(d) favourable to the production of lactic acid at the initial
period. It is favourable to reduce the production of lactic
Figure 2. Hydrolysis–acidogenesis of kitchen wastes in batch
experiments: variations of VFA and lactic acid at different pHs. acid or encourage the conversion of lactic acid to VFA in
the hydrolysis–acidogenesis process for the improvement
of two-phase anaerobic digestion of kitchen wastes.
rate, while that of propionic acid indicated an increase in
Run 1–3 and stayed constant for Run 4–6.
3.1.4. Biogas productivity and constituents
3.1.3. VFA and lactic acid It was observed from the variations of the biogas production
To further determine the metabolic route of the hydrolysis– rate and the constituents at different feeding COD loading
acidogenesis products, the VFA and lactic acid variations rates (Figure 1(c)) that the production of methane (CH4 )
under pH values of 5 and 7 and the control were examined declined from 56% to 3% with an increase in the COD
in batch experiments. The results are shown in Figure 2, loading rate. On the other hand, the production of CO2
in which it was found that the lactic acid and VFA were increased from 30% to 77%, and there were no obvious
generated rapidly at the initial period. The lactic acid con- variations in the production of H2 and N2 . The generation
centration for neutral pH declined following a maximum of CH4 was inversely proportional to the decrease in pH
concentration, but for pH 5 and the control (without pH (Figure 1(a)) in the hydrolysis–acidogenesis process due to
adjustment), it gradually increased to the maximum value the inhibition of low pH on the activity of methanogens.
and entered a plateau. However, VFA continued to increase Due to the recirculation of methanogenic effluent into
to the maximum concentration and kept constant at the the hydrolysis–acidogenesis reactor, it is unavoidable to
plateau under all three pH conditions. bring methanogenic microorganisms to the hydrolysis–
The appearance of the lactic acid would have an acidogenesis reactor. Good separation of acid and methane
important influence on both hydrolysis–acidogenesis and phases was achieved by low and high methane yields
methanogenic process because the ionization constant of in the hydrolysis–acidogenesis reactor and methanogenic
lactic acid is lower than that of the acetic acid, propionic acid reactor, respectively.[36] The increase in CO2 content
and butyric acid.[32] The production of lactic acid could and the decrease in CH4 content in this study sug-
decrease the pH to a relatively low value unless the pH gest that methanogenesis was inhibited completely in the
was controlled. During the batch test, it was proved that the hydrolysis–acidogenesis process in spite of the recycling
pH in the reactor without pH control would decrease below methanogenic effluent, and phase separation was achieved.
1282 Z. Bo and H. Pin-jing

3.2. Methanogenic process deviations for all the measurements for all runs. The results
The soluble output COD at different OLRs in the showed that the COD conversion efficiency increased to
hydrolysis–acidogenesis reactor was fed into the methanogenic 70–80% with increases in the feeding COD concentrations.
reactor. The COD conversion (Figure 3(a)), variations in The biogas productivity continued to rise from Run 1 to 5,
biogas productivity (Figure 3(b)), pH, VFA and alkalinity and in Run 6 it decreased a little. In addition, the biogas
(Figure 3(c)) in the methanogenic reactor were monitored yield maintained at about 500 mL·(g COD)−1 in every run.
continuously. Table 3 summarizes the means and standard The effluent VFA concentration decreased gradually from
Run 1 to 5, and in Run 6, it slightly increased. The pH in
the methanogenic reactor remained close to 7.5, whereas
(a) 40 100 the alkalinity kept a decreasing tendency.
Influent COD Figure 4 shows the distribution of organic acids in the

COD bioconversion efficiency (%)
35
Effluent COD
80 methanogenic effluent corresponding to the OLRs used
30 COD removal rate(%)
in the hydrolysis–acidogenesis stage. In spite of the high
25 feeding concentration, lactic acid concentration was below
COD(g.L–1)

60
20 1 g·L−1 in the methanogenic effluent. In Run 1 and 2, the
40
propionic acid concentration almost accounted for 78%
15
of the total VFA due to the initial domestication of the
Downloaded by [University of Canberra] at 07:00 23 October 2017

10 methanogenic microorganism for the fermentation liquid
20
5
of kitchen wastes. In the subsequent runs, the propionic
acid accounting for over 60% of the VFA was still the
0 0 dominant constituent. The accumulation of propionic acid
0 20 40 60 80 100 120
in the anaerobic process will result in low efficiency of
Time (d)
the methanogenic phase due to the low acetogenic rate of
(b) 6
propionic acid.[37,38]
The COD conversion efficiency rose rapidly from
Biogas production rate (L.(L.d)–1)

Run 2 to 4. In addition, the biogas productivity almost
4
ascended linearly with increasing OLR in the hydrolysis–
acidogenesis stage in the methanogenic stage, as Table 3
indicated. This further suggested that the feeding COD
loading rate increase in the hydrolysis–acidogenesis stage
2 had no obvious influence on the performance of the sub-
sequent methanogenesis despite using the inoculum of
anaerobic activated sludge. In relation to the previous results
observed in the hydrolysis–acidogenesis reactor, over 80%
0 of the total soluble COD was from the fermentation prod-
0 20 40 60 80 100 120 ucts, whereas monomeric products rarely accumulated. This
Time (d) study revealed that hydrolysis was still the rate-limiting step
(c) 8 14 for the two-phase anaerobic digestion of kitchen wastes
Alkalinity though kitchen wastes in this study contained high VS
VFA 12
content, and considered to be easily biodegradable. The
Alkalinity and VFA (g.L–1)

6 pH slow hydrolysis rate might be due to the high organic
10
acid concentration in the fermentation liquid.[39] For the
8 overall enhancement of the two-phase anaerobic digestion
4
6 process of kitchen wastes, it is also necessary to regulate
pH

and control the dependent factors such as pH, tempera-
4 ture and particle size in order to accelerate the hydrolysis
2
process.[40]
2

0 0
0 20 40 60 80 100 120
3.3. Mass balance of the whole two-phase anaerobic
Time (d)
digestion process
Figure 3. Methanogenesis of hydrolysis–acidogenesis liquid of
kitchen wastes in semi-continuous experiments: (a) variations in The performance of the whole two-phase anaero-
influent COD, effluent COD and COD bioconversion efficiency; bic digestion process under a COD loading rate of
(b) variations in biogas productivity rate and (c) variations in pH, 10.7 g·(L d)−1 was assessed by the mass balance analy-
VFA and alkalinity. sis, as shown in Figure 5. About 83.1% of bioconversion
Environmental Technology 1283

Table 3. Average operational results in the methanogenic stage.

Run 1 Run 2 Run 3 Run 4 Run 5 Run 6

OLR (g (L d)−1 ) in the hydrolysis– 2.3 4.1 6.3 8.4 10.7 13.1
acidogenenesis reactor
Input COD (g L−1 ) 8.4 ± 2.4 14.8 ± 0.9 22.5 ± 1.9 25.6 ± 4.4 33.8 ± 7.5 26.9 ± 8.5
OLR (g (L d)−1 ) in the methanogenic reactor 2.1 ± 0.6 3.7 ± 0.23 5.6 ± 0.48 6.4 ± 1.1 8.5 ± 1.9 6.7 ± 2.1
Output COD (g L−1 ) 5.9 ± 0.6 9.3 ± 1.1 8.6 ± 1.3 7.7 ± 1.6 7.2 ± 1.1 6.9 ± 2.7
Biogas productivity (L (L d)−1 ) 1.0 ± 0.3 1.8 ± 0.2 2.8 ± 0.7 3.3 ± 0.4 4.2 ± 0.4 3.6 ± 0.9
Biogas yield (mL (g COD)−1 ) 476.2 ± 32.3 486.5 ± 21.3 500.0 ± 25.8 515.6 ± 15.6 494.1 ± 20.2 537.7 ± 32.3
CH4 (%) 70 ± 1.5 72 ± 1 72 ± 2 75 ± 1.5 75 ± 2.3 72 ± 3
pH 7.5 ± 0.2 7.6 ± 0.1 7.6 ± 0.03 7.45 ± 0.1 7.7 ± 0.1 7.6 ± 0.01
VFA (g L−1 ) 3.3 ± 1.6 3.0 ± 0.5 2.3 ± 0.2 1.9 ± 0.2 1.4 ± 0.2 1.5 ± 0.4
Alkalinity (g L−1 ) 4.7 ± 1.01 4.3 ± 0.3 3.6 ± 0.2 3.8 ± 0.3 3.1 ± 0.1 2.2 ± 0.1

4 efficiency (calculated as [19.84(input)-2.29(discharge)-
Downloaded by [University of Canberra] at 07:00 23 October 2017

Lactic acid 1.06(output)]/19.84(input)) and the biogas yield of
Distribution of organic acids (g.L–1)

Propionic acid
Acetic acid 338.7 mL·(g COD)−1 (calculated as [6.72/19.84]) were
3 Butyric acid achieved. The TS conversion was 63% in the whole
two-phase anaerobic digestion process, which was mainly
reached in the hydrolysis–acidogenesis stage. Therefore,
2
the hydrolysis–acidogenesis stage needs to be improved if
the TS conversion is increased. The calculated methane
1
yield was 195 mL·(g input TS)−1 and 310.5 mL·(g con-
verted TS)−1 .
The operational performance of the two-phase anaero-
0 bic digestion in this study was comparable to those in the
0 2 4 6 8 10 12 14 16 reported literatures.[14,41] Bouallagui et al. [14] obtained
OLR(g.(L.d)–1) the maximum feeding OLR of 7.5 g COD·(L d)−1 when
Figure 4. Methanogenesis of hydrolysis–acidogenesis liquid of two-stage anaerobic sequence batch reactor (ASBR) was
kitchen wastes in semi-continuous experiments: distribution of used to treat kitchen wastes. The stable feeding OLR is
organic acids. lower than 8 gCOD·(L d)−1 [13,42,43] in most two-stage

Figure 5. Mass balance of two-phase anaerobic digestion for kitchen wastes at the OLR of 10.7 gCOD·(L d)−1 .
1284 Z. Bo and H. Pin-jing

anaerobic digestion of kitchen wastes. In this study, it was [8] Foresti E. Perspective on anaerobic treatment in developing
favourable to increase the feeding OLR by recirculating countries. Water Sci Technol. 2001;44:141–148.
methanogenic effluent into the hydrolysis reactor. At high [9] De Baere L. Anaerobic digestion of solid waste: state-of-the-
art. Water Sci Technol. 2000;41:283–290.
OLR of 10.7 g·(L d)−1 , the methane yield reached 315 mL· [10] Zhu H, Parker W, Conidi D, Basnar R, Seto P. Eliminat-
(g converted TS)−1 . However, the solubilization efficiency ing methanogenic activity in hydrogen reactor to improve
and acidogenesis efficiency were 63% and 22.6%, respec- biogas production in a two-stage anaerobic digestion pro-
tively, at a high OLR (10.7 g·(L d)−1 ). Therefore, it can cess co-digesting municipal food waste and sewage sludge.
be observed from the mass balance analysis that there is Bioresour Technol. 2011;102:7086–7092.
[11] Zhang B, Cai WM, HE PJ. Influence of lactic acid on the
a potential to improve the digestion performance if the two-phase anaerobic digestion of kitchen wastes. J Environ
hydrolysis process is improved. Sci. 2007;19:244–249.
[12] Yilmazer G, Yenigün O. Two-phase anaerobic treatment of
cheese whey. Water Sci Technol. 2002;40:289–295.
4. Conclusions [13] Cho JK, Park SC. Biochemical methane potential and solid
This study assessed the performance of two-stage anaerobic state anaerobic digestion of Korea food wastes. Bioresour
digestion of kitchen wastes by mass balance. It can be con- Technol. 1995;52:245–253.
[14] Bouallagui H, Touhami Y, Ben Cheikh R, Hamdi M. Bioreac-
cluded that it is important to increase the feeding OLR for tor performance in anaerobic digestion of fruit and vegetable
two-stage anaerobic digestion process of kitchen wastes. At wastes. Process Biochem. 2005;40:989–995.
Downloaded by [University of Canberra] at 07:00 23 October 2017

a COD loading rate of 10.7 g·(L d)−1 , a bioconversion effi- [15] Wang JY, Xu HL, Tay JH. Semi-continuous anaerobic
ciency of about 83.1% and a biogas yield of 338.7 mL·(g digestion of waste using a hybrid anaerobic solid–liquid
COD)−1 were achieved. The TS conversion was 63% in bioreactor. Water Sci Technol. 2003;48:169–174.
[16] Wang JY, Xu HL, Tay JH. A hybrid two-phase system
the whole two-phase anaerobic digestion process. Based for anaerobic digestion of food waste. Water Sci Technol.
on the results observed in the study, there is a potential to 2002;45:159–165.
improve the digestion performance if the hydrolysis process [17] Xu SY, Karthikeyan OP, Selvam A, Wong JWC. Effect of
is improved. inoculum to substrate ratio on the hydrolysis and acidifica-
tion of food waste in leach bed reactor. Bioresour Technol.
2012;126:425–430.
Acknowledgements [18] Xu SY, Lam HP, Karthikeyan OP, Wong JWC. Optimiza-
tion of food waste hydrolysis in leach bed coupled with
George Kirumba was instrumental in proof-reading the
methanogenic reactor: effect of pH and bulking agent.
manuscript. The authors of this manuscript express their heartfelt
Bioresour Technol. 2011;102:3702–3708.
gratitude to all that made the study a success.
[19] Zhang B, Zhang LL, Zhang SC, Shi HZ, Cai WM. The
influence of pH on hydrolysis and acidogenesis of kitchen
wastes in two-phase anaerobic digestion. Environ Technol.
Funding
2005;26:329–339.
The work was financially supported by Social Development Sci- [20] Zhang B, He PJ, Shao LM, Ye NF. Enhanced isomer purity
ence Foundation of Shanghai, China [12231202003]; National of lactic acid from the non-sterilized fermentation of kitchen
Natural Science Foundation of China [No. 21177084]; Shanghai wastes. Bioresour Technol. 2008;99:855–862.
Solid Waste Disposal Center, China. [21] Ten Brummeler E. Dry anaerobic digestion of the organic
fraction of municipal solid waste [Ph.D. thesis]. The Nether-
lands: Wageningen Agricultural University.
References [22] Wang QH, Narita JY, Xie WM, Ohsumi Y, Kusano K,
[1] Fang L. The present situation and solution of municipal solid Shirai Y, Ogawa HI. Effects of anaerobic/aerobic incuba-
wastes in China. Sci Technol Daily. 1999; December 19. tion and storage temperature on preservation and deodor-
[2] Wang HT, Nie YF. Municipal solid waste characteristics ization of kitchen garbage. Bioresour Technol. 2002;84:
and management in China. J Air Waste Manage Assoc. 213–220.
2001;51:251–272. [23] Wang QH, Yamabe K. Suppression of growth of putrefactive
[3] Diggelman C, Ham RK. Household food waste to wastewater and food poisoning bacteria by lactic acid fermentation of
or to solid waste? That is the question. Waste Manage Res. kitchen waste. Process Biochem. 2001;37:351–357.
2003;21:501–514. [24] Hill DT, Cobb SA, Bolte JP. Using volatile fatty acid rela-
[4] Banks CJ, Chesshire M, Heaven S, Arnold R. Anaerobic tionships to predict anaerobic digester failure. Trans ASAE.
digestion of source-segregated domestic food waste: perfor- 1987;30: 496–501.
mance assessment by mass and energy balance. Bioresour [25] Siegert I, Banks C. The effect of volatile fatty acid additions
Technol. 2011;102:612–620. on the anaerobic digestion of cellulose and glucose in batch
[5] Li Y, Park SY, Zhu J. Solid-state anaerobic digestion for reactors. Process Biochem. 2005;40:3412–3418.
methane production from organic waste. Renew Sust Energ [26] APHA. Standard method for the examination of water and
Rev. 2011;15:821–826. wastewater. 20th ed. Washington, DC, USA: American
[6] Li C, Champagne P, Anderson BC. Evaluating and model- Public Health Association; 1995.
ing biogas production from municipal fat, oil, and grease [27] Nielsen SS. Food analysis. Beijing: China Light Industry
and synthetic kitchen waste in anaerobic co-digestions. Press; 2002; p. 219–223.
Bioresour Technol. 2011;102:9471–9480. [28] Nielsen SS. Food analysis. Beijing: China Light Industry
[7] Molino A, Nanna F, Ding Y, Bikson B, Braccio G. Press; 2002; p. 182–183.
Biomethane production by anaerobic digestion of organic [29] Shin HS, Han SK, Song YC, Lee CY. Performance of UASB
waste. Fuel. 2013;103:1003–1009. reactor treating leachate from acidogenic fermenter in the
Environmental Technology 1285

two-phase anaerobic digestion of food waste. Water Res. [37] Wang L, Zhou Q, Li FT. Avoiding propionic acid accumu-
2001;35:3441–3447. lation in the anaerobic process for biohydrogen production.
[30] Xu SY, Karthikeyan OP, Selvam A, Wong JWC. Effect of Biomass Bioenergy. 2006;30:177–182.
inoculum to substrate ratio on the hydrolysis and acidifica- [38] Fischer JR, Ianotti EL, Sievers DM. Anaerobic digestion of
tion of food waste in leach bed reactor. Bioresour Technol. swine manures at various influent solids concentration. Agric
2012;126:425–430. Wastes. 1981;6:157–166.
[31] Reinhart DR, Al-Yousfi AB. The impact of leachate recircu- [39] Vavilin VA, Lokshina LY, Jokela JPY, Rintala JA. Modelling
lation on municipal solid waste landfill operational charac- solid waste decomposition. Bioresour Technol. 2004;94:
teristics. Waste Manage Res. 1996;14:337–346. 69–81.
[32] Lu F, He PJ, Shao LM, Lee DJ. Lactate inhibits hydrolysis [40] Veeken BA, Kalyuzhnyi S, Scharff H, Hamelers B. Effect of
of polysaccharide-rich particulate organic waste. Bioresour pH and VFA on hydrolysis of organic solid waste. J Environ
Technol. 2008;99:2476–2482. Eng. 2000;126:1076–1081.
[33] Ren NQ, Wang BZ, Huang JC. Ethanol-type fermenta- [41] Bouallagui H, Torrijos M, Godon JJ, Moletta R,
tion from carbohydrate in high rate acidogenic reactor. Ben Cheikh R, Touhami Y, Delgenes JP, Hamdi M.
Biotechnol Bioeng. 1997;54:128–433. Two-phases anaerobic digestion of fruit and vegetable
[34] Min H. Anaerobic microorganism. Hangzhou: Zhejiang wastes: bioreactors performance. Biochem Eng J. 2004;21:
University Press; 1993. 193–197.
[35] Marchaim U, Krause C. Propionic to acetic acid ratios [42] Han SK, Shin HS, Song YC, Lee CY, Kim SH. Novel anaer-
in overloaded anaerobic digestion. Bioresour Technol. obic process for the recovery of methane and compost from
1993;43:195–203. food waste. Water Sci Technol. 2002;45:313–319.
Downloaded by [University of Canberra] at 07:00 23 October 2017

[36] Ince O. Performance of a two-phase anaerobic diges- [43] Muroyama K, Mochizuki T, Wakamura T. Methane fer-
tion system when treating diary wastewater. Water Res. mentation of bean curd refuse. J Biosci Bioeng. 2001;91:
1998;32:2707–2713. 208–212.