Professional Documents
Culture Documents
Received 20 August 2005; received in revised form 14 December 2005; accepted 17 December 2005
Available online 9 February 2006
Abstract
Anaerobic digestion of organic fraction of municipal solid waste was conducted in pilot-scale reactor based on high-solid combined
anaerobic digestion process. This study was performed in two runs. In Run 1 and Run 2, pre-stage Xushing and micro-aeration were con-
ducted to determine their eVect in terms of enhancing hydrolysis and acidiWcation in ambient condition. In Run 2, after pre-stage, the
methane phase (methanogenesis) was started-up after pH adjustment and inoculum addition in mesophilic condition. AcidiWed leachate
produced in pre-stage was used for percolation during active methane phase. At the end of methane phase, air Xushing was conducted
before unloading the digesters. Hydrolysis and acidiWcation yield of 140 g C/kg TS and 180 g VFA/kg TS were achieved, respectively in
pre-stage. Micro-aeration exhibited an equivocal result in terms of enhancing hydrolysis/acidiWcation; however it showed a positive eVect
in methane phase performance and this needed further investigation. Leachate percolation during methane phase showed an enhanced
methanization when compared to the reactors without leachate percolation. After 60 days, 260 l CH4/kg VS was obtained. Based on the
waste methane potential, 75% biogas conversion and 61% VS degradation were achieved.
© 2006 Elsevier Ltd. All rights reserved.
0960-8524/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2005.12.018
P.H.L. Nguyen et al. / Bioresource Technology 98 (2007) 380–387 381
2.2. Process features the waste was unloaded from the reactor and the digested
waste sample was taken for characterization at the end of
Fresh market waste were loaded into the reactors and the process.
Xushed with tap water to produce eZuent in the form of
leachate. The purpose of Xushing was to leach out pollu- 2.4. Experimental set-up
tant in order to reduce the organic load from the waste
bed and at the same time to reduce the constraint of high Three digestion systems ran in parallel to optimize the
organic loading in high solid batch system. Fig. 2 illus- pre-stage and methane phase. Pre-stage Xushing and micro-
trates the three stages involved in combined anaerobic aeration conditions were examined in Run 1 and Run 2
digestion process. (1) Pre-stage: volatile fatty acids whereas methane phase was conducted only in Run 2. The
(VFA) and other dissolve organic compounds produced pre-stage operational sequence is schematized in Fig. 3.
by the fresh waste were Xushed into leachate. Flushing In Run 1, the reactors were loaded with fresh waste
and micro-aeration at diVerent strategies were provided to a compaction density of 500 kg/m3 with bulking agent.
to optimize the hydrolysis/acidiWcation process at ambi- Pre-stage was conducted for seven days. AcidiWed leachate
ent temperature. (2) Methane phase: biogas production was removed daily and was replaced by another 200 l of
at mesophilic temperature. Start-up was conducted with tap water. Daily water replacement during pre-stage under
pH adjustment and inoculum addition. An active meth- ambient condition was found to enhance leaching (Dayan-
ane phase was indicated by the presence of about 50% thi et al., 2004). Initially Xushing at a rate of 5 l/min for 4 h
methane in biogas and the pH of the system around 7. (3) run/4 h stop was performed followed by micro-aeration at a
Final stage: the waste bed was Xushed with fresh air to rate of 1 l/min (0.4 l/kg h) for 2 h run/4 h stop. It should be
remove the remaining biogas in digester before unload- noted that reactor 1 was non-micro-aerated while reactors
ing the waste. 2 and 3 was only provided with micro-aeration for the Wrst
three and seven days of operation, respectively. The results
2.3. Feedstock characteristics and preparation of pre-stage performance in Run 1 are illustrated in Fig. 4.
In general, it could be deduced that after Wve days of Xush-
The substrate used was market waste collected from ing in three reactors, the additional removal of DOC load
Bangkok, Thailand. The waste was characterized for of only 64% was very low. In this regard, Wve days of Xush-
moisture content (MC D 90%), volatile solids (VS D 79%), ing could be enough.
and total solids (TS D 10%). Fresh waste was manually In Run 2, pre-stage was conducted for only Wve days.
segregated to remove bulky and inert materials. The Daily water replacement was applied in reactor 1, whereas
sorted waste was subjected to size reduction to <60 mm by in reactors 2 and 3, water was replaced after day 1 and on
using mechanical pulverizer. The waste was loaded into day 3 of operation. As a result, a total of 1000 l water (48.8 l/
the reactor together with bamboo cutlets (10% volume of kg TS) was used for reactor 1, and only 600 l water (29.3 l/
the loaded waste) as bulking agent. The purpose of kg TS) was used for reactors 2 and 3. In reactor 3, Wve days
employing bulking material was to create void space in of micro-aeration was provided at a rate of 1 l/min (0.4 l/
order to facilitate the distribution of Xushing water and kg h) for 2 h run/4 h stop; while in reactors 1 and 2, micro-
aeration. Bamboo cutlets were separated manually when aeration was not applied. The reactors were loaded initially
P.H.L. Nguyen et al. / Bioresource Technology 98 (2007) 380–387 383
Run 1 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Run 2 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5
(g/kgTS)
5.5
1 conducted on fresh waste based on the method established
5.0 by Hansen et al. (2004). During pre-stage, the representa-
0 4.5 tive leachate sample from three reactors were collected
1 2 3 4 5 6 7
daily from the leachate tank and analyzed for dissolved
Run time (day)
organic carbon (DOC); total kjeldahl nitrogen (TKN);
Fig. 4. Variation of DOC, pH, and alkalinity in pre-stage leachate of Run 1 VFA including acetic acid (HAc), propionic acid (HPro),
(broken lines corresponding to primary Y-axis, continuous line correspond- butyric acid (HBu), and valeric acid (HVa); pH; and
ing to secondary Y-axis); (–䊏–) reactor 1 (non-micro-aerated); (–䊉–) reactor
alkalinity. All parameters were determined based on the
2 (three-day micro-aerated); (–䉱–) reactor 3 (seven-day micro-aerated).
analytical procedures in standard methods (APHA
et al., 1998). Daily biogas production was determined
of 150 kg of waste with bulking agent. Around 30 kg of from wet gas meter and biogas was analyzed in volumet-
fresh waste was added after two and four days of Xushing. ric composition (CO2, CH4, O2 and N2) by using Gas chro-
This additional amount of waste corresponded to the avail- matography.
able reactor’s headspace resulting from waste settlement
during pre-stage, also an optimum density of 500 kg/m3 was 3. Results and discussion
maintained in the new layer of waste.
Following pre-stage, new condition was provided in 3.1. Pre-stage
order to enhance the start-up of methane phase. The pH of
the system was adjusted to 6.5 and was followed by inocu- Fig. 4 exhibits the variation of DOC, pH, and alkalinity
lum addition. Mixture of cow dung, stabilized/digested of leachate during pre-stage. In three digesters, highest
waste and anaerobic sludge was used as seeding material DOC concentration (3.5 g/l) was noted in the Wrst day of
totally accounting for 16% VS of the loaded waste. Percola- Xushing. The result showed the DOC concentration in daily
tion was performed for two days to distribute inoculums leachate reduced sharply with run time to around 0.5 g/l in
throughout the waste bed. Reactors were incubated to a Wve days. Similar trends were observed for TKN and VFA.
temperature of 37 °C. DiVerent strategies were applied for The VFA concentration during Wrst few days of Xushing
three digesters. In reactors 1 and 3, leachate percolation was about 3–4 g/l and decreased to around 0.7 g/l on day 5.
was only practiced by the time the reactor shifted to active Although, Xushing could enhance acid production, it
methane phase (CH4 D 50%) and that was on day 40 and diluted the produced acids from the digester too early. The
30, respectively. However, in reactor 2, leachate percolation low pollutant concentration in leachate from day 5 was
was not provided. Pre-stage leachate was percolated at a due to the early extraction of hydrolyzed materials and the
384 P.H.L. Nguyen et al. / Bioresource Technology 98 (2007) 380–387
dilution of waste bed by Xushing. There was no further in Run 2, pH adjustment and inoculum addition was per-
signiWcant organic removal was noticed after Wve days. As formed after pre-stage. Since it was also observed from Run
mentioned, the purpose of pre-stage operation was to 1 that VFA concentration in leachate was low on day 3 (less
partly remove the organic fraction of the waste bed in the than 2 g/l), it was possible to reduce the amount of water
form of leachate in preparation for methane phase; this used without causing VFA inhibition. Thus, in Run 2, the
suggested that Xushing was insigniWcant after day 5. The amount of Xushing water was varied; 1000 l used in reactor
observation demonstrated that the short duration was 1, and 600 l was supplied in reactors 2 and 3. Also, addi-
needed for Xushing. tional 30 kg of waste added into the digesters in Run 2
Moreover, pH value in three reactors was low in the showed an insigniWcant diVerence in terms of pollutant
range of 5–6 (Fig. 4). This might be due to high VFA pro- load that could be extracted from the waste bed into leach-
duction and low buVering capacity (alkalinity) of the waste. ate. Optimum compaction density of 500 kg/m3 in the new
The pH value in reactor 3 (with seven days micro-aeration) waste layer was hypothesized to be the reason for a com-
was quite high due to bicarbonate buVering capacity caused parable yield.
by the formation of carbon dioxide in aerobic metabolism.
AcidiWcation occurs strongly at the early Xushing period 3.2. Methane phase
because acidogens are known to be fast-growing bacteria
with a minimum doubling time of around 30 min and are Fig. 5a and b show biogas composition and cumulative
capable of fermenting part of the soluble fraction of production in methane stage, respectively. During start-up,
organic refuse to produce VFA in a short time interval biogas production was low and methane content increased
(Mosey, 1983). Since hydrolysis is not rate limiting, acids slowly to 50% in reactors. The system was successfully
are produced quickly. At the end of Xushing, a maximum started-up after 25 days. Gas production rate in reactor 3
VFA yield of 180 g VFA/kg TS was achieved in reactor 1. was higher than reactors 1 and 2. The possible explanation
Lower acid yield of 160 g VFA/kg TS was observed in reac- was due to early micro-aeration in pre-stage which might
tors 2 and 3. have resulted in better hydrolysis/acidiWcation during start-
The cumulative load of single VFA distribution during up of methanization period providing substrate for metha-
pre-stage in three digesters did not vary much. Generally, nogens.
acetic acid dominates over propionic, butyric, and valeric DiVerent behaviors could be observed after start-up
acids. Since acetic acid is the direct substrate for methano- (Fig. 5b). In reactor 3, it was observed that cumulative gas
gens, its higher concentration shows that methanogenesis production increased immediately after lag phase, when
do not occur during pre-stage. A comparison of DOC and methane composition was stable and leachate percolation
DOC equivalent of TVFA (Table 1) exhibited that over was practiced. The curve implied that the waste bed was
half of the soluble organic carbon in leachate was acidiWed suYciently inoculated and buVered so that the methano-
into VFA. This showed that acidiWcation was strong over gens were activated. During batch 1 of leachate percolation
hydrolysis. The high fraction of VFA in leachate favored biogas production kept increasing but there was insigniW-
the proposal of feeding it back into the digester during cant change in VFA in reactor 3 (Fig. 6a and b). The reason
methane phase by leachate percolation. Early extraction of could be due to the continuous production of VFA from
VFA in pre-stage leachate would prevent imbalance the waste bed. Thus, based on this observation, the phase
between acidogenesis and methanogenesis in methane change from acidic to methanogenic stage had not yet com-
phase which was normally considered to cause instability in pleted. The cumulative methane yield rapidly increased and
high-solid digestion system (Mata-Alvarez, 2003). the high methane content (>60%) indicated a balanced
Low concentration of VFA at the end of pre-stage sug- methane fermentation.
gested that strict separation of acidogenesis and methano- Regarding the cumulative gas production, reactor 3
genesis could not be maintained at this point and it would gained the highest biogas production of about 5000 l (256 l/
be better to shift the reactor to a new stage. In this regard, kg TS) after 60 days of operation. This implied that the
Table 1
Load of DOC and DOC equivalent of VFA in pre-stage leachate (Run 1 and Run 2)
R1 (without micro-aeration) R2 (3 days micro-aeration) R3 (7 days micro-aeration)
Run 1 (7 days pre-stage)
DOC 140 127 119
DOC equivalent of TVFA 86 72 70
R1 (1000 l water) R2 (600 l water) R3 (600 l water)
Run 2 (5 days pre-stage)
DOC 140 128 129
DOC equivalent of TVFA 83 80 88
Note: Unit expressed in g C/kg TS.
P.H.L. Nguyen et al. / Bioresource Technology 98 (2007) 380–387 385
% CH4 in Biogas
50
40
30
Reactor 1
20 Reactor 2
10 Reactor 3
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
(a) Run time (day)
5000
Reactor 1 250
Reactor 2
4000 Reactor 3
200
3500
3000
150
2500
2000
100
1500
1000 50
500
0 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
(b) Run time (day)
Fig. 5. (a) Biogas composition and (b) cumulative biogas production (reactors 1 and 3: leachate percolation on day 40, 30; reactor 2: no leachate percola-
tion).
reactor reached a mature phase in which the hydrolyzed In reactor 3, at the commencement of leachate percola-
and acidiWed products were almost consumed so that a tion, the daily gas production was increasing (Fig. 6a). This
long term reduction of daily gas production was observed. showed that leachate percolation had a beneWcial eVect in
In reactor 2, only 2700 l of biogas (146 l/kg TS) was pro- enhancing biogas generation. However, after nine days of
duced after 60 days because leachate percolation was not leachate percolation (batch 1), the daily gas production
practiced that caused the gas production rate to increase reduced suddenly without reducing VFA concentration. It
very slowly. It reached a highest daily gas production of was observed that propionic acid in leachate kept on increas-
100 l/day on day 60. This rate was obtained earlier in reac- ing and reached nearly 4 g/l (Fig. 6b) and might positively
tors 1 and 3 at day 47 and 30, respectively. This suggested cause the sudden decrease of biogas from 160 l to 110 l. It
that leachate percolation during methane phase could pos- could be likely that there was an inhibition caused by high
itively enhance biogas production. This was also proven to concentration of this acid, as reported by other authors
be an eVective means of mixing, provided moisture (Gourdon and Vermande, 1987; Mawson et al., 1991; Inanc
(Ghosh, 1985; Chanakya et al., 1992; O’Keefe et al., 1993) et al., 1996; Pullammanappallil et al., 2001). New batch of
and appropriate pressure to help release biogas (Mohee pre-stage leachate with low propionic acid concentration
and Ramjeawon, 2003). On the other hand, Chan et al. was percolated causing daily gas production to increase back
(2002) demonstrated that leachate recirculation was not to the level of 150 l/day. The daily gas production was stable
only eVective in enhancing the degradation rate of waste for a week before it started to decrease again. This time, the
and gas production but also in the reduction of the overall drop in biogas production was not accompanied by high
leachate loading. concentration of propionic acid. The reduction of acetic acid,
386 P.H.L. Nguyen et al. / Bioresource Technology 98 (2007) 380–387
6
100
4
50
2
0 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
(a) Run time (day)
0
30 35 40 45 50 55 60
(b) Run time (day)
Fig. 6. (a) Daily gas production and (b) single VFA concentration (reactor 3).
a direct substrate for methanogens, was the reason for bio- system had stabilized and the leachate was mature. The
gas production to decrease. Third batch of leachate was pH value increased from 5 (pre-stage) to a range of 7.3–7.8
replaced, in order to supply VFA for digester. The system (methane phase).
kept on producing biogas but not as high as observed before
and the rate was slowly dropping. Long-term decreasing 3.3. Air Xushing stage
trend of biogas production was observed showing that
hardly biodegradable waste remained. In reactor 1, similar In Run 2, the reactor 3 reached a mature phase after 60
behavior of biogas production was observed, daily biogas days and air Xushing was practiced to Xush out the remain-
generation started to increase at the commencement of ing biogas before unloading the digester. After one day of
leachate percolation. aeration, methane and carbon dioxide content dropped
In reactor 3, propionic acid was the highest accumu- to <5%. Thus, one day of aeration was suYcient to remove
lated acid in the system during Wrst batch of leachate per- the remaining biogas and the reactor could be safely
colation. It reached the level of nearly 4 g/l as compared to unloaded.
2 g/l of acetic acid. Butyric and valeric acids concentration
were low. Due to a favorable environmental condition for 3.4. BMP and actual methane yield
acid formation, mainly high pH, other acids started to
appear. The stable concentration of acetic acid indicated In reactor 3, approximately 5 m3 of biogas with meth-
the balance between the rate of acid consumption and ane content of 55% was obtained, which was equivalent
production. Propionic acid and other acids were produced to 230 l CH4/kg VS. The actual methane yield was 230 l
more while acetic acid was converted to methane and car- CH4/kg VS as compared to 300 l CH4/kg VS in lab-scale
bon dioxide. According to Mrz.-Viturtia et al. (1994), the BMP test. This highlighted that almost 75% methane
low acetic acid concentration compared with propionic conversion was obtained in the pilot scale system under
acid indicated a stable methane performance. In mature the inXuence of pre-stage (Wve days micro-aeration,
phase, most acids started to decrease revealing that the Xushing with 600 l of water) in ambient condition and
P.H.L. Nguyen et al. / Bioresource Technology 98 (2007) 380–387 387
methane phase (with leachate percolation) in mesophilic national Symposium on Anaerobic Digestion of Solid Waste, Barce-
condition. lona, June 15–17, 1999; 113–120.
Chan, G.Y., Chu, L.M., Wong, M.H., 2002. EVects of leachate recirculation
on biogas production from landWll co-disposal of municipal solid
4. Conclusions waste, sewage sludge and marine sediment. Environmental Pollution
118 (3), 393–399.
The application of micro-aeration during pre-stage Chanakya, H.N., Sushama, B., Rajan, M.G.C., Wahi, M., 1992. Two-phase
showed an equivocal result on hydrolysis/acidiWcation anaerobic digestion of water hyacinth for urban garbage. Bioresource
Technology 42, 123–131.
enhancement. Flushing for short duration of Wve days with Dayanthi, W.K.C.N., Visvanathan, C., Tränkler, J., Kuruparan, P. 2004.
lesser volume of water (29.3 l/kg TS) was found beneWcial to Pretreatment of domestic solid waste by enhanced leaching. In: Pro-
partly remove the organic matter from waste bed in pre- ceedings of 3rd Asian PaciWc LandWll Symposium (APLAS), Japan,
paration for methane phase. Hydrolysis yield of 140 g/kg October 27–29, 2004, 232–239.
TS which is equivalent to 30% of C from the waste bed Ghosh, S., 1985. Solid phase methane fermentation of solid waste. Journal
of Energy Resources and Technology 107, 402–406.
was removed in leachate form and acidiWcation yield of Gourdon, R., Vermande, P., 1987. EVects of propionic acid concentration
180 gVFA/kg TS were obtained. The importance of leach- on anaerobic digestion of pig manure. Biomass 13 (1), 1–12.
ate percolation during methane phase was recognized to Hansen, T.L., Schmidt, J.E., Angelidaki, I., Marca, E., Jansen, J.C., Mos-
enhance biogas production. Nevertheless, micro-aeration baek, H., Christensen, T.H., 2004. Method for determination of meth-
during pre-stage may have a positive eVect in methanogene- ane potential of solid waste. Waste Management 24, 393–400.
Inanc, B., Matsui, S., Ide, S., 1996. Propionic acid accumulation and con-
sis since an active methane phase was reached early com- trolling factors in anaerobic treatment of carbohydrate: eVects of
pared to other reactors without micro-aeration. This might hydrogen and pH. Water Science and Technology 34 (5–6), 317–325.
have resulted in better hydrolysis/acidiWcation during the Mata-Alvarez, J., 2003. Biomethanization of the organic fraction of
start-up of methanization period providing substrate for municipal solid wastes. IWA Publishing, 1 900222 14 0.
methanogens. Thus, micro-aeration exhibits an equivocal Mawson, A.J., Earle, R.L., Larsen, V.F., 1991. Degradation of acetic and
propionic acids in the methane fermentation. Water Research 25 (12),
eVect in terms of enhancing anaerobic digestion process 1549–1554.
and this needs further investigation. The actual methane Mohee, R., Ramjeawon, T., 2003. Appropriate of the sequential batch
yield of waste (260 l CH4/kg VS) was compared with the anaerobic composting (SEBAC) technology as a treatment method
result of BMP test (300 l CH4/kg VS), it revealed that 75% for organic waste in Mauritius. In: Proceedings of ORBIT 2003, 341–
process eYciency can be obtained in this process. 350.
Mosey, F.E., 1983. Mathematical modeling of the anaerobic digestion pro-
cess: regulatory mechanisms for the formation of short-chain volatile
Acknowledgements acids from glucose. Water Science and Technology 15, 209–232.
Mrz.-Viturtia, A., Mata-Alvarez, J., Cecchi, F., 1994. Two phase continu-
The authors are grateful to the Swedish International ous anaerobic digestion of fruit and vegetable wastes. Resources, Con-
versation and Recycling 13, 257–261.
Development Co-operation Agency (SIDA) for Wnancing O’Keefe, D.M., Chynoweth, D.P., Barkdoll, A.W., Nordstedt, R.A., Owens,
this research, which is a part of the Sustainable Solid Waste J.M., Sifontes, J., 1993. Sequential batch anaerobic composing of
LandWll Management in Asia under the Asian Regional municipal solid waste (MSW) and yard waste. Water Science and
Research Program on Environmental Technology. Technology 27 (2), 77–86.
Pullammanappallil, P.C., Chynoweth, D.P., Lyberatos, G., Svoronos, S.A.,
2001. Stable performance of anaerobic digestion in the presence of a
References high concentration of propionic acid. Bioresource Technology 78 (2),
165–169.
APHA, AWWA, WEF, 1998. Standard methods for the examination of Visvanathan, C., Tränkler, J., Chiemchaisri, C., Basnayake, B.F.A., Gong-
water and wastewater, 20th ed. Washington, DC, USA ISBN: 0-87553- ming, Z., 2004. Municipal solid waste management in Asia: Asian
235-7. regional research program on environmental technology. AIT Publica-
Biey, E.M., Musibono, E.D., Verstraete, W., 2003. Start-up of a multi-stage tion, Thailand. ISBN: 9-744-17258-1. Available from: <http://www.
system for biogas production and solid waste treatment in low-tech swlf.ait.ac.th/>.
countries. Water Science and Technology 48 (4), 1–8. Wellinger, A., Widmer, C., Schalk, P. 1999. Percolation-a new process to
Capela, I.F., Azeiteiro, C., Arroja, L., Duarte, A.C. 1999. EVects of pre- treat MSW. In: Proceedings of Second International Symposium on
treatment (composting) on the anaerobic digestion of primary sludge Anaerobic digestion of Solid Waste, Barcelona, June 15–17, 1999, 315–
from a bleached kraft pulp mill. In: Proceedings of Second Inter- 322.