Bioresource Technology 102 (2011) 10922–10928

Contents lists available at SciVerse ScienceDirect

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

Batch and continuous biogas production from grass silage liquor

Jehad Abu-Dahrieh a, Angela Orozco b, Elaine Groom b, David Rooney a,⇑
a
CenTACat, Queen’s University Belfast, Belfast BT9 5AG, Northern Ireland, UK
b
QUESTOR, Queen’s University Belfast, Belfast BT9 5AG, Northern Ireland, UK

a r t i c l e i n f o a b s t r a c t

Article history: Herein batch and continuous mesophilic anaerobic digestion of grass silage liquor was studied. The
Received 1 August 2011 continuous process was carried out in Armfield digesters with an OLR ranging from 0.851 to
Received in revised form 15 September 1.77 kg COD m 3 day 1. The effect of recirculation of effluent from the digester was investigated using
2011
different OLRs of grass silage liquor feed. These results showed that as the OLR increased, the methane
Accepted 16 September 2011
Available online 22 September 2011
yield decreased for the reactor with no recycle and increased for the reactor with recycle. However,
the COD removal for both digesters was nearly the same at the same OLR. Overall these studies show that
grass silage liquor can produce a high quality methane steam between 70% and 80% and achieve methane
Keywords:
Anaerobic digestion
yields of 0.385 m3 kg 1 COD.
Biochemical methane potential Ó 2011 Elsevier Ltd. All rights reserved.
Grass silage liquor
Biogas

1. Introduction or grass silage is not a mature feed-stock for AD purposes and

Due to increasing political social and environmental pressures
research into substantial energy has grown rapidly over recent
years within both the academic and industrial community. While
a number of potential options exist, it is increasingly recognised
that future energy production will utilise a significantly more
diverse mix of generating facilities combined with effective energy
storage. Biomass is nature’s preferred method of solar energy stor-
age and if used for renewable energy production a wide range of
materials can be converted, including wood, food waste, energy
crops and grass. Globally biomass offers an abundant and rela-
tively cheap source of energy however the quality of the energy
obtained, e.g. through combustion, can be significantly lower than
non-renewable petroleum resources.
One of the most important factors when considering agricul-
tural feed-stocks for biogas production is the net gross energy
yield per hectare. Energy crops, crop residues and grasses are
among the options available for methane production using anaer-
obic digestion (AD). Of these grasslands play an important role in
the global agriculture covering 69% of the world’s agricultural
area or 26% of the total land area (Prochnow et al., 2009) and over
90% of Irish agricultural land (Murphy and Power, 2009). Recently
many studies have investigated using grass/grass silage as a
source for renewable energy through biogas production. However
if grass is to be used as feedstock for AD for energy production, it
should be converted to silage (Smyth et al., 2009). However grass
thus needs further investigation to optimize the operating condi-
tions for maximum methane yield with low operating costs.
Lehtomaki et al. (2008) studied AD of grass silage in batch leach
bed processes for methane production. In another paper, they also
classified that grasses have the highest methane potential among
potential crops for biogas production in northern countries as
Finland (Lehtomaki et al., 2008). Smyth et al. (2009) compared
the net energy of the grass to biomethane systems with other
energy crops and they found that grass has higher gross energy
than rapeseed biodiesel and wheat ethanol systems. Recent strat-
egies to improve grass biomethane include reactor design (Singh
et al., 2011), pre-conditioning (Richter et al., 2011) and the use of
adsorbents (Weiß et al., 2011).
There are of course many factors that affect the performance of
the AD process: pH, temperature, hydraulic retention time (HRT)
and organic loading rate (OLR) for example. The processes them-
selves can be classified according to operating temperature and
here there are two main types: mesophilic, where the digester is
heated to 35 °C and thermophilic where the digester operates at
55 °C. Traditionally, AD plants have operated under mesophilic
conditions as thermophilic digesters require greater input energy
for heat and are very sensitive to any changes in the conditions.
This also increases the risk of failure, however at the same time
they do produce more biogas.
Anaerobic digesters themselves can be classified into three
main groups:

⇑ Corresponding author.  Single stage continuously fed systems, where all the biological
E-mail address: d.rooney@qub.ac.uk (D. Rooney). reactions take place in one sealed reactor.

0960-8524/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2011.09.072
 J. Abu-Dahrieh et al. / Bioresource Technology 102 (2011) 10922–10928
 Two stage continuously fed reactors where these digesters are
optimized to bring maximum control of each biological stage
by separating the hydrolysis and acidification stages from
acetogenesis and methanogenesis stages.
 Batch reactors, where the reactors are filled once with the feed-
stock and left for a period of time that allows biodegradation of
the substrate. The simplest example for the batch reactor is a
laboratory system commonly used to measure the biochemical
methane potential (BMP).
The BMP assay process was first established in (1979) by Owen
et al. as a simple and inexpensive procedure to measure biodegrad-
ability and toxicity of substrates subjected to anaerobic digestion.
Here the estimate of the ultimate methane production is considered
a key parameter for assessing design and implementation for the
full scale design of anaerobic digestion processes (Angelidaki
et al., 2009). To ensure CH4 production is not limited by substrate
or inoculum availability multiple substrate to inoculum ratios
should be tested. The substrate to inoculum ratio should not affect
the final volume of methane produced but should control the rate at
which methane is produced (Moody et al., 2009).
The above discussion has highlighted that biogas is considered
to be an important component of the future renewable energy
mix. Given its nature it has great flexibility to be converted to elec-
tricity, stored as a pressurised gas or cleaned and used in a gas grid
or as a transport fuel. Many options exist for its production and of
these grasslands show significant promise. To support this area the
main objectives of this paper are to determine the BMP for grass
silage at different inoculum to substrate ratio, the effect of contin-
uous recycling of biomass on the performance of the digester and
the effect of the OLR on biogas production, methane yield and
COD removal %.
2. Methods
2.1. Materials
2.1.1. Inoculum
The inoculum used to inoculate the BMP reactors and Armfield
digesters was from a mesophilic anaerobic digester obtained from
the Agri-Food and Biosciences Institute (AFBI), N. Ireland which
was treating cow manure. Prior to inoculation into the equipment,
the inoculum was first sieved through a 1 mm mesh to remove any
entrained solids or particulates. The sieved inoculum was analysed
using standard methods to determine the VS, total suspended sol-
ids (TSS), soluble COD.
2.1.2. Silage Liquor
The silage liquor was provided by Irish Agricultural Institute
‘Teagasc’, Grange Research Centre, Dunsany, Co., Meath Ireland.
Upon receiving the effluent the liquid was sieved through a
1 mm mesh to remove any entrained solids or detritus. The total
volume of filtered effluent was collected into a drum container
and vigorously mixed to be homogenised. At this stage a sample
was taken to determine the COD, pH, volatile fatty acid (VFA)
and alkalinity. The silage liquor was immediately frozen and kept
at 20 °C in the freezer to prevent further degradation. Before
the anaerobic digestion process silage liquor was defrosted and
silage liquor solutions were prepared and degassed, the pH was ad-
justed to 7 with 1 N NaOH and stored in the fridge until use.
2.2. Experimental setup
2.2.1. BMP
In the BMP reactor setup four units were used, one containing
inoculum alone (sludge) (R1), one with cellulose as a reference
10923
(R2), one containing silage liquor with substrate to inoculum (S/I)
ratio of 1 g COD/gVS and one containing silage with a (S/I) ratio of
1.5 g COD/gVS (R4). In the test runs it was not practical to run trip-
licate sludge blanks due to equipment restrictions. The BMP tests
were carried out once due to equipment availability, although each
sample assay was carried out in triplicate. Each assay was per-
formed in a 1 l reactor with 0.75 l working volume, all reactors
were continuously mixed by a magnetic stirrer. A blank assay (only
inoculum) was used to determine methane production resulting
from inoculum itself and a control assay (cellulose) used as a refer-
ence to test the quality of the inoculum by comparing the results
with other studies. The same amount of inoculum was added to
each reactor. The inoculum was kept homogenous by continuous
mixing. An appropriate amount of cellulose was added to R2 with
an inoculum to cellulose VS-ratio of 2:1, silage to inoculum ratio of
1:1 (COD silage to inoculum) in R3 and silage to inoculum ratio of
1.5:1 (COD silage to inoculum) in R4. All reactors were filled with
distilled water up to 0.75 kg weight and 3 g/l of NaHCO3 was added
as a buffer. 0.75 ml of trace minerals were also added to each reac-
tor to improve performance (Gonzalez-Gil et al., 2001).
Finally, the reactors were sealed and the headspace flushed
with nitrogen to remove oxygen. Then these assays were continu-
ously stirred while immersed in a temperature controlled water
bath at 38 °C. The gas production volume and composition were
recorded on a daily basis. At the same time the temperature and
pressure were noted so the gas volumes were adjusted to standard
temperature and pressure (STP: 0 °C and 1 bar) for comparison
with other studies. The batch reactors were incubated until there
was no further gas produced. Methane potentials were calculated
as m3-CH4 kg 1-COD added in the case of silage liquor substrate
and m3-CH4kg 1-VS added in the case of cellulose substrate with
CH4 of the inoculum subtracted. Once a week, a 5 ml sample of
each reactor was taken in order to analyse for, TS, VS, COD, pH
and alkalinity.
The substrate (grass silage liquor) and inoculum used in this
study in both the BMP and Armfield digesters had the following
properties: suspended solids 23,400 mg/l and 12,170 mg/l, volatile
solids 17,890 mg/l and 10,370, COD 350,000 mg/l and 227,500 mg/l
and pH 8.01 and 3.88, respectively.
2.2.2. Armfield digesters unit
2.2.2.1. Armfield operating conditions. Fig. 1 gives a schematic repre-
sentation for the laboratory-scale Armfield digester. The digester
unit has 5 l upward-flow packed bed reactor with feed rate and
temperature control facilities to allow continuous operation at up
to 7 l/day. A packed section promoted good mixing of the feed with
the biomass charge. The flow rates to the vessels were set and con-
trolled by calibrated peristaltic pumps. The temperature of each
reactor was controlled by an electric heating mat wrapped around
the external wall. The temperature distribution within the reactor
was maintained at ±0.5 °C. Reactor temperatures could be sepa-
rately set at any desired value within a range ambient to 55 °C
(Engineering Teaching Equipment - Discover with Armfield,
2005). Given that these reactors were operated continuously these
experiments were only performed once.
The biogas produced from the reactor was collected in the gas
collection tank using a water displacement system which was de-
signed to impart a small and constant back pressure to the reactor
vessel. Water displaced by the gas overflowed from the base of the
vessel to drain (Armfield Limited, 1996).
The Armfield reactors were operated in parallel with the influent
silage being fed to first reactor (A1) with recirculation of the efflu-
ent. The second reactor (A2) was then fed with silage influent only.
In this part and for each reactor three organic loading rates (OLR)
were studied: 0.851 kg COD m 3 day 1, 1.42 kg COD m 3 day 1
and 1.77 kg COD m 3 day 1 respectively with 0.5 l/day flow rate.
10924 J. Abu-Dahrieh et al. / Bioresource Technology 102 (2011) 10922–10928

Fig. 1. Schematic diagram for Armfield digester. (1) Feed pump: peristaltic pump (flow rates from 0.2 to 5.8 l/day), (2) temperature control: 200 W heating jacket with PID
control from a temperature sensor positioned inside the reactor, (3) 5 l volume reactor: 150 mm diameter, (4) 20 cm liquid height, (5) 10 cm packing height, (6) gas collection
vessel: 0–5 l capacity and (7) Liquid effluent collector.

2.2.2.2. Armfield initial setup. Here the equipment was assembled
and a synthetic feed was added to promote the growth of the inoc-
ulum at the temperature of the experiment. This synthetic feed
was primarily composed of glucose and other nutrients. It was ini-
tially fed to the reactors at a low feed rate (0.3 l/day) which then
increased by 0.2 l/day. When the flow rate reached to 1.5 l/day,
the feed was fed for another two without further increase. This
was in order to acclimatise the bacteria to the new conditions
within the digesters and to increase the biomass in the reactor.
2.3. Analysis techniques
Soluble VFA and COD were measured by Hach DR2700 Spectro-
photometer according to method 8196 and method 8000 in the
Hach procedures manual, respectively. The pH was measured with
a Hanna Instruments HI 8424 microcomputer pH meter. The alka-
linity of the samples was measured with a Mettler DL21 Titrator.
TSS and VS were analysed according to standard methods for the
examination of water and wastewater APHA (1998).
3. Results and discussion
only a slight increase in pH from 7.3 to 8.01 over the course of
the experiment for all reactors (R1, R2, R3 and R4). This indicates
that the reaction conditions were stable. This is also an indication
that the anaerobic processes within these reactors were not under
inhibitory effects.
Fig. 2 shows methane concentration in biogas samples as ob-
served over the experiment period. It can be seen that for all reac-
tors there are a sharp increase for methane concentration over the
first 5 days for all reactors and then a slight increase was observed
for remaining period. The methane concentration for R1, R2, R3
and R4 at the end of period is 69%, 62%, 83% and 72%, respectively.
This can be attributed to the fact that at the beginning of the exper-
iment there was a rapid growth in the different bacteria required
to convert the high concentration of substrate available and these
bacteria would have hydrolysed those compounds which break
down quickly. This would have provided substrate for respiration
for the methanogens and thus allowed for increased growth in
the methanogenic cultures. After this substrate was consumed,
the availability of the feed for the methanogens would have been
greatly reduced. Therefore the remaining biomass would be com-
posed of tougher plant components which are not quickly hydroly-

3.1. Inoculum and silage liquor properties

90
It is clear that grass silage liquor has a low pH which can be
80
attributed to the high concentration of VFA’s in the silage of above
Methane Concentration, (%)

800 mg/l. The increased VFA content could be as a result of the fer- 70
mentation of the grass in order to form silage. The VFA’s in the si-
lage are then used directly in the acetogenesis stage to produce 60
acetic acid, which is then used by the methanogens to produce
50
methane.
40
3.2. BMP
30
Blank- R1
As per the experimental section the BMP tests were carried out 20
Control - R2
in four units, one containing inoculum alone (sludge) (R1), one Silage grass/Inoculum =1:1 - R3
10
with cellulose as a reference (R2), one containing silage liquor with Silage grass/Inoculum =1.5:1 - R4
(S/I) ratio of 1 and one containing silage liquor with (S/I) ratio of 0
1.5 (R4). Biogas production for all reactors was run over a period 0 5 10 15 20 25 30 35 40 45
of 40 days. Biochemical methane potential (BMP) experiments for Time (days)
silage were conducted for different initial CODs of grass silage to
VS of inoculum ratio. The weekly samples show that there was Fig. 2. Methane concentration in the biogas.
 J. Abu-Dahrieh et al. / Bioresource Technology 102 (2011) 10922–10928 10925

sed and thus decreased methane production. Under the conditions is approximately 1 (Chynoweth et al., 1993; Gunaseelan, 1997),
used herein it can be stated that the appropriate duration for bio- which is in agreement with our results. It can also be seen also that
gas production, as determined from BMP tests, is less than 20 days. the methane yield of the control at STPis 0.35 m3 kg 1 VS cellulose
Fig. 3 shows the methane yield of R2, R3 and R4. Methane yield similar to that found by Nallathambi Gunaseelan (2009) (0.404 m3
was calculated by subtracting the amount of methane produced by kg 1 VS cellulose) and Hansen et al. (2004) (0.379 m3 kg 1 VS cel-
the blank reactor (sludge only) from methane production of each lulose), which means that the BMP system is under the anaerobic
reactor (R2, R3 and R4) and dividing the difference by the COD in conditions and works properly.
the Silage fed to the reactor (R3 and R4) or by the VS in the cellu-
lose fed to the reactor (R2). The purpose of the blank was to deter- 3.3. Armfield digesters
mine the amount of methane that could be generated as a result of
AD of the biomass contained in the sludge itself. Gunaseelan 3.3.1. Armfield daily operation
(1997) reported in his review that BMP of grasses (Napier grass, On a daily basis, the biogas volume and composition produced
energy cane, Alemangrass-6A, turf grass, wheat straw, paddy were recorded. The volume of the effluent, volume of water dis-
straw, millet straw, oats crop, maize crop, corn stover and placed, and the volume of the feed remaining were also recorded
sorghum) exhibited CH4 yields as high as 0.3 m3 kg 1 VS added in order to determine the actual amount of the feed fed to the reac-
without pre-treatment. From Fig. 3, it is clear that the silage liquor tor. A 50 ml sample of the effluent was taken every day from each
produced a maximum yield of 0.385 m3 kg 1 COD at (S/I) ratio of 1, reactor to determine COD, VFA, pH, alkalinity, TSS and VS. From
whereas it was 0.197 m3 kg 1 COD at (S/I) ratio of 1.5. These results this information the internal conditions of the reactors could be
indicate that as the (S/I) increased the methane yield decreased, tracked and any change in alkalinity or pH monitored and adjusted
this is attributed to the fact that as the amount of inoculum accordingly with sodium carbonate solution.
increases in the reactor it also increases the availability of the The feed was prepared daily from the defrosted concentrated
nutrients needed to complete digestion of the substrate (grass solutions stored in the freezer and after an initial acclimatisation
silage liquor) or that the inoculum bacteria are inhibited by the period, both reactors were fed with silage feed. For each reactor
high COD loading of the substrate (grass silage liquor). Several three organic loading rates (OLRs) were tested: 0.851
researchers have studied the effect of (S/I) ratio on the methane kg COD m 3 day 1, 1.42 kg COD m 3 day 1 and 1.77 kg COD m 3
yield (Gunaseelan, 1995, 1997; Liu et al., 2009). Gunaseelan day 1 to investigate the performance of the process and the
(1995) for example studied the effect of inoculum to substrate ra- changes in methane production. Both reactors worked at an OLR
tio on methane yield from Parthenium as a substrate. As shown of 0.851 kg COD m 3 day 1 between days 1 and 16, at an OLR of
here they found that as the S/I ratio increased the methane yield 1.42 kg COD m 3 day 1 between days 17 and 23 and at 1.77 kg
decreased and with large inoculum size the batch digestion can COD m 3 day 1 between days 24 and 33.
be completed. It is recognised that this may not be relevant to
3.3.2. pH of the Armfield digesters
the specific grass silage liquor used here, however these results
In these experiments both reactors were running without
do agree with those found here and thus the same conclusions
adjustment of the pH inside the reactor. The pH of the effluent
are drawn.
from A1 and A2 ranged from 7.4 to 8.4 and from 8 to 8.5, respec-
For example, González-Fernández and García-Encina (2009)
tively. These values could be an indication that the digesters were
studied the impact of three experimental ratios of S/I (1, 2 and 3)
working within the optimum interval for methane formation,
using swine slurry as substrate. Their results showed that methane
which means that the reactors are not under inhibitory conditions.
production rates decreased with increasing ratios. They explained
this behaviour by the accumulation of VFAs (predominantly acetic 3.3.3. Alkalinity and VFA
(HAc) and propionic acid (HPr)) which was reflected in the soluble Fig. 4 shows the relationship between VFA and alkalinity. It is
COD profile. In the standard BMP procedure the optimum S/I ratio clear that both reactors were operating with a good buffering

0.40
Silage grass/Inoculum =1:1 0.40
Silage grass/Inoculum =1.5:1
Methane Yield (m3/kg COD of grass silage) @ STP

0.35
Methane Yield (m3/kg VS of cellulose) @ STP

Control(Cellulose) 0.35
0.30
0.30
0.25
0.25
0.20
0.20

0.15
0.15

0.10 0.10

0.05 0.05

0.00 0.00

-0.05 -0.05
-5 0 5 10 15 20 25 30 35 40
Time (days)

Fig. 3. Cumulative methane yield at STP.

10926 J. Abu-Dahrieh et al. / Bioresource Technology 102 (2011) 10922–10928

500
4400
450 VFA-A1 4200
VFA-A2 4000
Alkalinity-A1
400 Alkalinity-A2 3800
3600
350 3400
3200

Alkalinity, mg/l
300
3000
VFA (mg/l)

250 2800
2600
200 2400
2200
150 2000
1800
100
1600
50 1400
1200
0 1000
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34

Time (days)
Fig. 4. Relationship of volatile acids to alkalinity of A1 and A2.

capacity (indicated by the low VFA less than 200 mg/l compared to
alkalinity of 2000–3600 mg/l). The fluctuation of VFA and alkalinity
during the time, did not affect the methanogenic activity as the
methane concentration in the gas did not drop which demon-
strates that the acetogens and methanogens were able to cope with
the fluctuations of the VFA and alkalinity within the reactor. This
indicates that the conditions within A1 and A2 were stable and
that the possibility of methanogen inhibition was low. The VFA
for the effluent for both reactor ranges from 53 to 250 mg l 1,
the low VFA in the effluent means that almost all the VFA formed
is converted to methane and carbon dioxide.
3.3.4. Results of the Armfield digesters
Table 1 shows the Armfield digester results where it can be seen
that as the OLR increased from 0.851 to 1.77 kg COD m 3 day 1, the
methane yield decreased for A2 from 0.751 to 0.52 m3 kg 1 COD
and increased for A1 from 0.463 to 0.66 m3 kg 1 COD. The decline
of methane yield in A2 can be explained by the results of the
BMP, which identified that as the S/I ratio increased the potential
for methane production decreased. Recirculation of the effluent
from A1 improved the methane production yield; this may be
attributed to the fact that during the recycling the amount of the
inoculum is kept constant inside the reactor which thus stimulated
the overall degradation and contact between the bacteria and their
food. This enables the bacteria to more readily access the sub-
stances they are feeding off and increases the speed of gas yields
(Wikimedia Foundation Inc., 2011).
For A2, there is no significant difference in methane yield, biogas
composition and COD removal when the OLR increased from 1.42 to
1.77 kg COD m 3 day 1, which means that increasing the OLR did
not decrease the COD removal efficiency so the digester was still
working under good conditions with a high COD. For an OLR of
1.42 kg COD m 3 day 1 both reactors produced the same amount
of methane yield but with different CO2 concentrations, meaning
that hydrogen was combining with CO2 to produce methane, indi-
cated by methanogenesis step, which is the last of the typical reac-
tions that take place in the anaerobic digestion process (Madsen
et al., 2011). The COD removal for both reactors was nearly the
same at the same OLR. For A1, the COD removal was 88.1%, 97.8%
and 96.05% and for A2 86.1%, 98.3% and 97% at OLR of 0.851, 1.42
and 1.77 kg COD m 3 day 1, respectively. This could be attributed
to good distribution or spread of the organic loading through the
bed that all of the available biomass in the substrate was being con-
sumed. It can be also observed from the table that the biogas in A1
contains a low concentration of H2S (4.5 ppm), which could be
attributed to recirculation of the effluent or possibly due to the

Table 1
Results of Armfield digesters.

Property Unit COD for silage feed (mg/l)

8510 8510 14220 14220 17720 17720
A1 A2 A1 A2 A1 A2
OLR kg COD m 3 day 1
0.851 0.851 1.422 1.422 1.772 1.772
HRT days 0.65 10 0.65 10 0.65 10
COD removal % 88.1 86.1 97.82 98.3 96.05 97
Methane yield m3/kg COD 0.463 0.751 0.53 0.54 0.66 0.52
CH4 % 70 77 71 73 76 70
CO2 % 7 22 14 22 19 22
O2 % 3.5 0.8 2 0.6 0.4 0.3
H2S ppm 4.5 >>a 99 >>a 320 279
H2 ppm 5.8 >>a 128 >>a 30 >>a
a
>> means the value cannot be read in the measuring device (out of range).
 J. Abu-Dahrieh et al. / Bioresource Technology 102 (2011) 10922–10928 10927

presence of O2 in the digester (3.5%). From the literature, the sim-
plest method for the desulphurization of biogas is by injection of
small amount of oxygen or air (2–6% in the biogas).
Figs. 5 and 6 show the daily biogas production of the Armfield
reactor without and with recycle, respectively. Both reactors
worked at an OLR of 0.851 kg COD m 3 day 1 between days 1 and
16, at an OLR of 1.42 kg COD m 3 day 1 between days 17 and 23
and at 1.77 kg COD m 3 day 1 between days 24 and 33. These fig-
ures show also that the main component in the biogas is methane
with concentration range from 70 to 76 after the adaptation period
for both reactors, this demonstrates that the grass silage liquor sub-
strate can produce a high methane content in the biogas (CH4 rich
biogas). This can be attributed to the presence of the VFA’s initially
in the silage liquor which negates the need for the hydrolysis and
acidogenesis stages. This would indicate that the feed can be more
readily converted into methane and increase the speed at which
production occurs.
The other components of the biogas are carbon dioxide, hydro-
gen, oxygen, hydrogen sulphide and other gases including nitrogen
and ammonia. For the low OLR of A1 there is a significant quantity
of other gases in the biogas produced, which is here attributed to
N2 or NH3 gases. It can be seen that as the OLR of A1 increased
the concentration of CO2 increased and the other gases concentra-
tion decreased, whereas for A2 there was no significant difference
in the gas composition at different OLRs. Low concentrations of CO2
in the biogas for A1 at low OLR may be attributed to the ammonia
formed by the acidogenic (fermentative) bacteria which was not
converted by acetogens to produce acetic acid as well as carbon
dioxide and hydrogen.
Detailed economics concerning the overall production of meth-
ane using this route have not been investigated here. However
given that this study utilises grass silage liquor rather than solid
grass, it is likely that this method would be economic for larger
installations, particularly if used to supplement anaerobic manure

Fig. 5. Daily biogas concentration of Armfield reactor without recycle (A2).

Fig. 6. Daily biogas concentration of Armfield reactor with recycle (A1).

10928 J. Abu-Dahrieh et al. / Bioresource Technology 102 (2011) 10922–10928
treatments. Furthermore it is known that the results of such tests
are scalable to industrial size units and therefore the work dis-
cussed here is of relevance to industrial practice.
4. Conclusions
Using a combination of previously unreported techniques, this
work has demonstrated that grass silage liquor can be used to
produce a high quality methane stream between 70% and 80% in
both batch and continuous reactors. Investigation of inoculum to
substrate ratio on methane yield from grass silage liquor using
BMP showed that the optimum S/I ratio is approximately 1 with a
maximum methane yield of 0.385 m3 kg 1 COD. These results indi-
cate that grass silage liquor can be successfully converted using
anaerobic digestion and while further investigation into the tech-
no-economics is required it is expected that this process is econom-
ical and scalable.
Acknowledgements
The authors would like to thank Agri-Food and Biosciences
Institute (AFBI), N. Ireland AFBI for supplying the inoculums used
in these experiments as well to the Irish Agricultural Institute
‘Teagasc’, Grange Research Centre, Dunsany, Co. Meath Ireland for
supplying the grass silage. We would also like to thank the
Department of Agriculture and Food, DAFF, Research Stimulus Fund
Project: ‘‘Green Grass’’) for the research funding the EPSRC and the
ATU Unit at the QUESTOR Centre.
References
Angelidaki, I., Alves, M., Bolzonella, D., Borzacconi, L., Campos, J.L., Guwy, A.J.,
Kalyuzhnyi, S., Jenicek, P., van Lier, J.B., 2009. Defining the biomethane potential
(BMP) of solid organic wastes and energy crops: a proposed protocol for batch
assays. Water Science and Technology 59, 927–934.
Armfield Limited, 1996. Instruction manual of Anaerobic Digester W8. Armfield
Limited.
Chynoweth, D.P., Turick, C.E., Owens, J.M., Jerger, D.E., Peck, M.W., 1993.
Biochemical methane potential of biomass and waste feedstocks. Biomass and
Bioenergy 5, 95–111.
Engineering Teaching Equipment – Discover with Armfield, 2005. W8 Anaerobic
Digester Engineering Teaching Equipment – Discover with Armfield.
?tul=0?>González-Fernández, C., García-Encina, P.A., 2009. Impact of substrate to
inoculum ratio in anaerobic digestion of swine slurry. Biomass and Bioenergy
33, 1065–1069.
Gonzalez-Gil, G., Seghezzo, L., Lettinga, G., Kleerebezem, R., 2001. Kinetics and
mass-transfer phenomena in anaerobic granular sludge. Biotechnology and
Bioengineering 73, 125–134.
Gunaseelan, V.N., 1995. Effect of inoculum/substrate ratio and pretreatments on
methane yield from Parthenium. Biomass and Bioenergy 8, 39–44.
Gunaseelan, V.N., 1997. Anaerobic digestion of biomass for methane production: a
review. Biomass and Bioenergy 13, 83–114.
Hansen, T.L., Schmidt, J.E., Angelidaki, I., Marca, E., la Cour Jansen, J., Mosbaek, H.,
Christensen, T.H., 2004. Method for determination of methane potentials of
solid organic waste. Waste Management (Amsterdam, Netherlands) 24, 393–
400.
Lehtomaki, A., Huttunen, S., Lehtinen, T.M., Rintala, J.A., 2008a. Anaerobic digestion
of grass silage in batch leach bed processes for methane production. Bioresource
Technology 99, 3267–3278.
Lehtomaki, A., Viinikainen, T.A., Rintala, J.A., 2008b. Screening boreal energy crops
and crop residues for methane biofuel production. Biomass and Bioenergy 32,
541–550.
Liu, G., Zhang, R., El-Mashad, H.M., Dong, R., 2009. Effect of feed to inoculum ratios
on biogas yields of food and green wastes. Bioresource Technology 100, 5103–
5108.
Madsen, M., Holm-Nielsen, J.B., Esbensen, K.H., 2011. Monitoring of anaerobic
digestion processes: a review perspective. Renewable and Sustainable Energy
Reviews 15, 3141–3155.
Moody L., Burns R., Wu-Haan W., R., S., 2009. Use of Biochemical Methane Potential
(BMP) Assays for Predicting and Enhancing Anaerobic Digester Performance.
Agricultural Engineering, 44th Croatian & 4th International Symposium on
Agriculture.
Murphy, J.D., Power, N.M., 2009. An argument for using biomethane generated from
grass as a biofuel in Ireland. Biomass and Bioenergy 33, 504–512.
Nallathambi Gunaseelan, V., 2009. Biomass estimates, characteristics, biochemical
methane potential, kinetics and energy flow from Jatropha curcus on dry lands.
Biomass and Bioenergy 33, 589–596.
Owen, W.F., Stuckey, D.C., Healy Jr., J.B., Young, L.Y., McCarty, P.L., 1979. Bioassay for
monitoring biochemical methane potential and anaerobic toxicity. Water
Research 13, 485–492.
Prochnow, A., Heiermann, M., Ploechl, M., Linke, B., Idler, C., Amon, T., Hobbs, P.J.,
2009. Bioenergy from permanent grassland – a review: 1. Biogas. Bioresource
Technology 100, 4931–4944.
Richter, F., Fricke, T., Wachendorf, M., 2011. Influence of sward maturity and pre-
conditioning temperature on the energy production from grass silage through
the integrated generation of solid fuel and biogas from biomass (IFBB): 1. The
fate of mineral compounds. Bioresource Technology 102, 4855–4865.
Singh, A., Nizami, A.-S., Korres, N.E., Murphy, J.D., 2011. The effect of reactor design
on the sustainability of grass biomethane. Renewable and Sustainable Energy
Reviews 15, 1567–1574.
Smyth, B.M., Murphy, J.D., O’Brien, C.M., 2009. What is the energy balance of grass
biomethane in Ireland and other temperate northern European climates?
Renewable & Sustainable Energy Reviews 13, 2349–2360.
Weiß, S., Zankel, A., Lebuhn, M., Petrak, S., Somitsch, W., Guebitz, G.M., 2011.
Investigation of mircro organisms colonising activated zeolites during
anaerobic biogas production from grass silage. Bioresource Technology 102,
4353–4359.
Wikimedia Foundation Inc., 2011. Anaerobic digestion. Wikimedia Foundation
Inc.