Waste Management 31 (2011) 2167–2173

Contents lists available at ScienceDirect

Waste Management
journal homepage: www.elsevier.com/locate/wasman

Anaerobic digestion of the liquid fraction of dairy manure in pilot plant

for biogas production: Residual methane yield of digestate
Carlos Rico a,⇑, José Luis Rico b, Iñaki Tejero a, Noelia Muñoz b, Beatriz Gómez b
a
Department of Sciences and Techniques of Water and Environment, University of Cantabria, Los Castros s/n, 39005 Santander, Spain
b
Department of Chemical Engineering and Inorganic Chemistry, University of Cantabria, Los Castros s/n, 39005 Santander, Spain

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

Article history: The performance of the only dairy manure biogas plant in Cantabria (Northern coast of Spain) was eval-
Received 27 January 2011 uated in terms of liquid–solid separation and anaerobic digestion of the liquid fraction. Screened liquid
Accepted 22 April 2011 fraction was satisfactorily treated in a CSTR digester at HRTs from 20 to 10 days with organic loading
Available online 25 May 2011
rates ranging from 2.0 to 4.5 kg VS/(m3 d). Stable biogas productions from 0.66 to 1.47 m3/(m3 d) were
achieved. Four anaerobic effluents collected from the digester at different HRTs were analyzed to mea-
Keywords: sure their residual methane potentials, which ranged from 12.7 to 102.4 L/g VS. These methane potentials
Biogas
were highly influenced by the feed quality and HRT of the previous CSTR anaerobic digestion process. Bio-
Pilot plant
Anaerobic digestion
methanization of the screened liquid fraction of dairy manure from intensive farming has the potential to
Dairy manure provide up to 2% of total electrical power in the region of Cantabria.
Screw press Ó 2011 Elsevier Ltd. All rights reserved.
Digestate

1. Introduction solids are recalcitrant or barely biodegradable (Rico et al., 2007).

Separation of liquid and solid fractions of manure is a desirable
One of the new trends of dairy farming in Spain is to intensify upstream operation in the treatment process since dewatering
the animal production, which is always accompanied by the the solid fraction lowers the cost of shipping, which facilitates
production of large amounts of manure, not properly managed the export of nutrients from the areas with excess of manure and
by farmers, presenting a considerable environmental threat in the redistribution of these nutrients to other areas in need of them
these zones. In humid areas, particularly where the ground shows (Holm-Nielsen et al., 2009). Before its transport, the solid fraction
pronounced slopes, the liquid fraction of dairy manure is the most can be subjected to dry anaerobic digestion and/or the composting
problematic due to run-off and lack of sufficient land for disposal, process which would represent an additional reserve of more
resulting in a surplus of nitrogen in the areas where intensive stable organic carbon and nitrogen for the cultivated soil (Atallah
farming is concentrated. One such region is that of Cantabria, et al., 1995).
located in Northern Spain, which has a bovine population of Moreover, the separated liquid fraction, which has much less
around 280,000 livestock units (mainly milk). suspended solids content, would be more easily subjected to an
Anaerobic digestion of dairy manure has been demonstrated to anaerobic process requiring a simpler reactor as well as lower tem-
be an attractive treatment that provides several benefits including perature and hydraulic retention time (HRT) than those conditions
the improvement of manure fertilizer quality; reduction of odors, for unscreened dairy manure (Liao et al., 1984). The digested liquid
pathogens and greenhouse gas emissions; and production of a fraction can be used as a fertilizer for agriculture (Morris and
renewable fuel, the biogas (Albertson et al., 2006; Chae et al., Lathwell, 2004; Mantovi et al., 2010). In addition, it would be
2008; Hartmann and Ahring, 2005; Hawkes et al., 1984; Kaparaju favorable to further refine the liquid fraction into concentrated
and Rintala, 2003). fertilizers or to receive post-treatments to obtain clean water, suit-
Dairy manure composition depends on the cows’ diet and the able for recycling (Holm-Nielsen et al., 2009).
conditions under which animals (dairy cows, breeding cows, and Separation of dairy manure can be performed before or after
calves) are kept in the farm. In addition to solid and liquid dejec- anaerobic digestion. The main advantage of separation before
tions, dairy waste extracted from cow houses contains the remains digestion is the removal of the fibrous part, which can cause
of food and bedding (straw, sand, sawdust, etc.). Many of these clogging problems within the reactor and pipelines (Wen et al.,
2007). In addition, fibrous materials hinder pumping and mixing.
⇑ Corresponding author. Tel.: +34 942201848; fax: +34 942201703. However, about half of VS remains in the solid fraction (Møller
E-mail address: [email protected] (C. Rico). et al., 2002) which implies a lower amount of recoverable energy.

0956-053X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.wasman.2011.04.018
2168 C. Rico et al. / Waste Management 31 (2011) 2167–2173

One way to harness this potential energy contained in the solid maintain the temperature of the process since the recirculation
fraction is through the dry anaerobic digestion process, which is pipe passes through a heat exchanger. Water was used as heating
usually performed discontinuously without any mixing of the solid fluid and was heated up to 55 °C in a 500 L electric heat
substrate (Weiland, 2006). accumulator tank. The heat exchanger (HRS Spiratube) consisted
Anaerobic digestion in CSTR reactors has been widely studied of a corrugated tube within a tube.
by several researchers with different manures at both the lab scale Biogas left the reactor by its own pressure and was measured
(Boe and Angelidaki, 2009; Hawkes et al., 1984; Liao et al., 1984; Lo using a mass flow meter (Bronkhorst, F111-BI). Biogas passed
et al., 1983; Karim et al., 2005; Zeeman et al., 1988) and the pilot through a catch pot and a coalescing filter to protect the biogas
scale (Kaparaju et al., 2008). However the majority of these works meter from condensates and particulate matter. A sampling port
utilize the lab scale in which manure is conserved at low temper- was installed between the coalescence filter and the mass flow me-
atures or even frozen prior to its use to prevent degradation of the ter. At the end of the line biogas was burned in a flare (Emison, A-5
feed. Also there is a lack of information about pilot CSTR installa- ESP).
tions processing the liquid fraction of dairy manure by employing The control panel was located in a closed, protective box. This
industrial or semi-industrial equipment. panel contained all the electric system controls required for the
In this work, the performances of the screw press separator and functioning of the pilot plant. The heating system was commanded
the 1.5 m3 CSTR anaerobic digester of the only dairy manure biogas by the temperature probe to maintain the temperature of the
plant in Cantabria, a pilot plant located in the Agrarian Secondary digester at 37 ± 1 °C.
School ‘‘La Granja’’ (Heras, Cantabria), were tested. The CSTR reac- In addition to the equipment described above, the pilot plant
tor was fed with the screened liquid fraction of dairy manure in also had a decanter centrifuge to refine the digested screened
mesophilic conditions. Residual methane production from the pilot liquid fraction in a nutrient-rich solid fraction and a liquid fraction
plant screened/digested liquid fraction was also measured at the free of suspended solids in addition to a UASB reactor for imple-
lab scale. menting the process performed by García et al. (2008) at lab scale.
In addition, the pilot plant had facilities for recovering nutrients
from anaerobic liquid effluents by struvite precipitation and other
2. Materials and methods post-treatments to obtain a final liquid effluent suitable for reutili-
zation. The performance of these processes was not considered in
2.1. Pilot plant the present work.

The pilot plant was projected for 2008 and inaugurated in June
2009. During the realization of the present work, a screw press 2.2. Mode of operation
separator (Doda MS5CE, 0.8 mm mesh) and a CSTR digester were
employed as the main components of the installation. Fig. 1 shows The CSTR reactor was filled with screened manure in September
the scheme of the installation and technical parameters of the 2009 and the heating system was programmed to slowly obtain
components used in the pilot plant tests. the operation temperature, achieved after 3 days. The recirculation
A chopper centrifugal pump (Cri-man, ETO series) was used to pump then was programmed to work eight times per day for
recirculate the raw manure slurry, avoiding stratifications, into 30 min cycles at a flow rate of 1000 L/h to keep the reactor content
the storage tanks and to pump the slurry to the separator. Two well mixed. In addition, the recirculation pump was programmed
mono pumps (Seepex, BN-025-12 and BN-2-6L) were employed together with the heat exchanging system to automatically start
for feeding the reactor and recirculation of the reactor content. when the temperature controller detected the need for heating to
The CSTR digester consisted of a cylindrical fixed dome vessel maintain a stable reactor temperature of 37 ± 1 °C.
made of 304 stainless steel and was equipped with temperature The reactor was not fed again to allow for the growth and
and pH probes. Inside, the digester is covered with epoxy painting acclimatization of methanogenic biomass and organic matter
and the exterior wall is thermally insulated to prevent heat loss. stabilization of the reactor content. After 35 days, analysis of the
The reactor has two sampling ports, one in the effluent tube and reactor content showed that volatile fatty acids (VFA) had been
the other in the recirculation pipe, as well as a purge valve in the totally removed. The reactor then was started up at an HRT of
bottom to remove sinking materials. Mixing is accomplished by 20 days and fed in a semi-continuous mode in 30 min cycles. Feed-
recirculation of reactor content. Recirculation was also used to ing cycles were equally distributed throughout the day and the

Fig. 1. Scheme of the pilot installation.

 C. Rico et al. / Waste Management 31 (2011) 2167–2173 2169

Table 1 the solid-support Chromosorb W-AW 80/100 mesh. Nitrogen was

Reactor operation data. used as the carrier gas at a flow rate of 14 mL/min, and a FID
HRT (days) OLR Pump flow Feeding Feed rate detector was installed. VFA concentrations are expressed in COD
(kg VS/(m3 d)) (L/h) cycles (L/d) units. The volume of biogas generated in the pilot plant CSTR
a day reactor was measured by a mass flow meter and registered by an
(30 min.)
electronic unit. Biogas composition was assayed on a 2 m Poropak
20 2.0 30 5 75 T column in a HP 6890 GC system with helium as the carrier gas at
16.7 2.1 30 6 90
12.5 3.5 30 8 120
a flow rate of 15 mL/min with a TCD detector. Biogas and methane
10 4.5 30 10 150 volumes are expressed at 0 °C and 1 atm. All other analyses, includ-
ing TS, VS, pH, COD, total Kjeldahl nitrogen (TKN-N), NH4+-N and
total phosphorus (PT), were performed according to standard
methods (APHA, 1998). All the analyses were performed in dupli-
recirculation pump was programmed to work at the same time as
cate; the mean values are reported herein.
the feeding pump to enhance mixing. Feeding cycle data are
presented in Table 1. 75 L of reactor content were purged weekly
from the bottom of the reactor to remove inorganic and sinking
materials that enter the reactor with the feed (e.g. sand) or are 3. Results and discussion
formed during the process.
3.1. Performance of the screw press separator

2.3. Manure characteristics

The screw press produced a screened solid fraction (SSF) and a
screened liquid fraction (SLF). The amounts of these fractions were
Dairy manure was collected from the dung pit of a near 500-free
measured twice because large quantities are difficult to handle. A
stall dairy cow farm equipped with scrape systems. The diet of the
mass balance of the separation process is presented in Table 2.
dairy cows was a constant during the experimental period. Manure
The screw press separator produced an SSF upon an increase in
was extracted from the dung pit by a tractor equipped with a
the concentrations of TS, VS, TKN and PT of 2.76, 3.11, 1.14 and 1.23
vacuum tank system and then transported and discharged into the
times, respectively. About 18% of the manure mass was found in the
raw manure tanks of the pilot plant. Because manure remained
solid fraction, implying that 49.8% and 55.9% of the TS and VS were
stored in the raw manure tanks during the pilot scale experiments,
transferred to the solid fraction. Regarding the removal of nutrients,
the effect of manure storage time on biogas production could be ob-
20.6% and 22.2% of TKN and PT, respectively, were transferred to the
served. The characteristics of raw manure are reported in Table 2.
solid fraction. Wu (2007) reported similar recovery percentages in
the solid fraction (48.2% for dry matter (DM), 12.5% for PT, and
2.3.1. CSTR feed preparation 22.7% for TKN). However, these values were higher than those
Dairy manure was screened by the screw press separator. The found by Møller et al. (2002), who reported values of 29.9–13.1%
feeding tank received the screened liquid fraction for feeding the for TS, 12.7–4.0% for TKN and 15.4–8.0% for PT in the screw press
CSTR reactor. The feeding tank was filled up and when it became separation of cattle manure. The lower DM content of manure in
empty, it was cleaned with a vacuum cleaner to remove sediments that work (63.7–44.9 g/l) could be the reason for this difference.
from the bottom and filled again. The solid fraction was collected Whereas the SSF-manure ratio in the present work obtained per-
in a 400 L container and composted. However, composting of the centages of 18%, Møller et al. (2002) reported only 5.2% and 2.4%.
solid fraction was not under consideration in this study. However, Zhang and Westerman (1997) reported that the perfor-
mance data of mechanical separators vary widely not only because
2.4. Residual methane tests of the different testing and reporting procedures, but also because
the characteristics of the manure used are sometimes different. La-
Digestate samples were collected at the exit of the CSTR in 5 L ter tests performed with the screw press separator showed that
containers for transport to the lab. Batch tests were started within with a smaller DM content in the manure slurry, a smaller SSF-man-
2 h of their collection. Tests were performed in 2.5 L PVC reactors ure ratio, in terms of the mass and lower removal efficiencies of DM
and incubated at 35 °C in a thermostatic bath for 60 days without and nutrients, was achieved. Specifically, an SSF-manure ratio of 8%
stirring. Methane production was measured by means of a dis- and TS, TKN and PT removal efficiencies of 33.5%, 12.5% and 14.9%
placement system using an alkaline solution to absorb the CO2 pro- respectively, were obtained when a manure slurry with 53.7 g TS/
duced. Each trial was performed in duplicate. L was subjected to screw press separation. In the light of these re-
sults, nutrient removal was influenced by the solids input concen-
2.5. Analytical techniques tration: as the solids concentration in the influent increased, the
removal of nutrients increased, as reported by Converse et al.
The VFA concentrations of the CSTR influents and effluents were (1999). Furthermore, the screw press was especially efficient in
determined using a HP6890 gas chromatograph (GC) fitted with a removing DM but not nutrients, as reported by Møller et al.
2 m 1/8-in glass column, liquid phase 10% AT 1000, packed with (2000, 2002) and Pain et al. (1978) reported. This phenomenon is
due to a large proportion of small particles containing nutrients
that remain in the liquid fraction after screw press separation.
Table 2
Composition of dairy manure, solid and liquid fractions.

Raw manure Solid fraction Liquid fraction 3.2. Performance of the CSTR pilot reactor
Mass distribution (%) 100 18 82
TS (%) 9.2 25.3 5.8 3.2.1. Feed characteristics and removal percentages of COD and VS
VS (%) 7.3 22.7 4.2 Table 3 shows the characteristics of the screened dairy manure
VS/TS 0.79 0.90 0.72 employed as feed and the effluent for the four HRT tested. Because
TKN-N (g/L) 3.8 4.3 3.6
the manure came from the uncovered dung pit of a near farm, cli-
PT (g/L) 1.1 1.4 1.0
matological conditions (especially rain for the HRT 16.7 day feed)
2170 C. Rico et al. / Waste Management 31 (2011) 2167–2173

Table 3 3.2.2. Biogas production rates and methane content of biogas

Steady-state data for affluent and effluent of CSTR reactor (Mean ± SD). Fig. 3 shows daily biogas production rates throughout the
HRT: HRT: HRT: HRT: experimental period. Whenever raw manure tanks were near
20 days 16.7 days 12.50 days 10 days empty or a shorter HRT was set, fresh manure was taken from
Influent the near farm, which occurred on days 43, 80, 110 and 140.
pH 6.7 ± 0.1 6.9 ± 0.1 6.7 ± 0.1 6.6 ± 0.1 Depending on the ambient temperature, stored manure in the
CODT (g/L) 51.6 ± 2.5 46.8 ± 1.5 57.4 ± 2.0 59.6 ± 3.1 tanks was more or less rapidly degraded. For this reason, during
CODVFA (g/L) 9.2 ± 1.1 6.4 ± 0.7 12.4 ± 1.1 14.3 ± 0.6
TS (g/L) 56.4 ± 1.6 51.7 ± 2.1 60.3 ± 2.4 62.7 ± 3.3
the last days of operation for each HRT, biogas production suffered
VS (g/L) 40.5 ± 1.7 35.8 ± 1.2 43.4 ± 1.5 44.9 ± 2.6 pronounced decreases (days 36–43, 70–80 and 130–140) due to
TKN-N (g/L) 3.4 ± 0.3 3.0 ± 0.3 3.7 ± 0.3 4.0 ± 0.4 the aging of manure. To calculate the average steady state values
NH4+-N (g/L) 2.3 ± 0.2 1.6 ± 0.3 1.9 ± 0.2 2.1 ± 0.3 of biogas production for each HRT, the days of the experiment
PT (g/L) 0.98 ± 0.05 0.90 ± 0.08 1.1 ± 0.1 1.1 ± 0.1
when biogas production diminished due to feed degradation were
Effluent not considered.
pH 7.7 ± 0.1 7.8 ± 0.1 7.9 ± 0.1 7.8 ± 0.1
In Table 4, the mean specific biogas and methane production as
CODT (g/L) 30.2 ± 2.1 29.3 ± 1.8 33.2 ± 1.9 36.9 ± 2.6
CODVFA (g/L) 0 0 0.27 ± 0.09 0.69 ± 0.10 well as volumetric biogas and methane production rates are
TS (g/L) 39.2 ± 2.1 37.7 ± 1.9 41.1 ± 1.8 45.3 ± 1.2 shown. The reactor started to be fed in October at an HRT of
VS (g/L) 23.7 ± 1.6 22.4 ± 2.5 25.1 ± 2.3 27.8 ± 1.6 20 days (day 0) and a mean OLR of 2.0 kg VS/(m3 d). After 8 days,
TKN-N (g/L) 3.2 ± 0.3 2.9 ± 0.3 3.5 ± 0.3 3.8 ± 0.3 the biogas production reached stable values. The mean biogas pro-
NH4+-N (g/L) 2.3 ± 0.1 2.1 ± 0.2 2.1 ± 0.2 2.2 ± 0.2
duction rate for this period was 992 L biogas/d, which represents
PT (g/L) 0.95 ± 0.03 0.88 ± 0.05 1.0 ± 0.1 1.0 ± 0.1
327 L biogas/kg VS and a volumetric biogas production rate of
0.66 m3/(m3 d). A peak in biogas production at day 25 (1150 L bio-
gas/d) was observed, which was due to a failure in the Tª probe that
50 increased the temperature of the reactor to 41 °C. At day 43 biogas
production was reduced to 596 L/d due to degradation of the feed.
40 From 44 to 80 days, the HRT was set at 16.7 days. Due to heavy
rains in the days prior, the SLF presented a lower organic load than
that of the SLF processed before at 20 days HRT. This implied that
30
% removed

OLR was almost the same (2.1 kg VS/(m3 d)) and volumetric biogas
production rate was only a bit higher (0.72 m3/(m3 d)) than those
20 with the previous HRT. Specific biogas production for this period
was 336 L/kg VS, quite similar to that of the previous period. The
peak in biogas production at day 54 was caused by another failure
10
in the Tª probe. The digester performance remained stable until
degradation of the feed caused a slow decrease in biogas
0 production.
1 2 3 4 5
OLR (Kg VS/(m3·d))
VS COD

Fig. 2. Removal percentage of VS and COD during CSTR operation. White rhombus 2.5 22
represent mean VS removal percentage and black squares represent mean removal
percentage of COD for the four HRT tested. 20
2
18
m3 biogas/day

affected the characteristics of manure because of dilution with rain 1.5

HRT (days)
16
water.
The CSTR reactor operation started at an HRT of 20 days with an 14
1
organic loading rate (OLR) of 2.0 kg VS/(m3 d). At the last HRT
performed (10 days), the OLR applied was 4.5 kg VS/(m3 d). At the 12
16.7 days HRT, the CSTR feed (SLF) presented a lower organic load, 0.5
10
resulting in a similar OLR for the 20 and 16.7 days HRT. The average
steady-state removal percentages of CODT and VS are shown in
0 8
Fig. 2. For all the HRTs, CODT and VS reduction were about 0 20 40 60 80 100 120 140 160 180
37–41% and 30–36%, respectively, showing a similar performance biogas HRT Day
of the reactor when the OLR increased. Generally, when the HRT de-
creases and the OLR increases, the percentage of COD removed Fig. 3. CSTR daily biogas production. In this figure, daily biogas production (black
circles) is represented during all the experimental period. The black line represents
should decrease if the methanogenic activity of the biomass in-
the HRT during the test.
volved in the process was maintained. However, as observed in this
work, biomass growth and acclimatizing throughout the experi-
mental period should help in compensating for the increase in the
Table 4
OLR, maintaining the COD removal percentage of the reactor. Specific biogas and methane production and volumetric biogas and methane
Compared with previous works, Lo et al. (1983) reported production rates during CSTR operation.
maximum COD and VS removal rates of 22–24% and 23–27% at
HRT (days) 20 16.7 12.5 10
HRTs between 8 and 10 days when treating diluted screened dairy
manure (4.4% TS, 3.4% VS) at a lab CSTR at 30 °C. Hawkes et al. L biogas/kg VS 327 336 349 328
L CH4/kg VS 206 219 220 223
(1984) reported VS removal percentages of 17%, 33% and 40% at
m3 biogas (m3 d) 0.66 0.72 1.21 1.47
5, 15 and 18 days HRT on the anaerobic digestion of screened cattle m3 CH4/(m3 d) 0.42 0.47 0.76 1.00
slurry, with a mean VS content of 2.84%, at 35 °C.
 C. Rico et al. / Waste Management 31 (2011) 2167–2173 2171

With fresh manure, the HRT was set at 12.5 days on day 80. As Typical methane yields ranging from 75 to 223 L CH4/kg VS were
observed from the data in Table 3, the SLF presented a higher reported for dairy manure with up to 7% TS in various digester con-
organic load than the SLF processed previously. Due to decreasing figurations (Ogejo and Li, 2010). Lo et al. (1983) reported methane
HRT and increasing organic load of the SLF, OLR increased by up to production of 0.50 L CH4/(L d) and 0.106 L CH4/g VS for screened
3.5 kg VS/(m3 d). Biogas production strongly increased, reaching a dairy manure and 0.25 L CH4/(L d) and 0.025 L CH4/g VS for
mean value of 1816 L/d and a volumetric biogas production rate unscreened manure. Hawkes et al. (1984) reported methane yields
value of 1.21 m3/(m3 d). The better characteristics of the feed also of 0.119, 0.166 and 0.204 L CH4/g VS at 5, 10 and 15 days HRT,
resulted in a higher specific biogas production, 349 L/kg VS. The fall respectively, on the anaerobic digestion of screened cattle slurry.
in biogas production on day 120 was caused by an electrical failure Dugba and Zhang (1999) obtained a maximum of 0.82 L CH4/(L d)
that stopped the recirculation pump, causing a drop in tempera- when treating screened dairy manure with two-stage (thermo-
ture reactor to 29.5 °C. The problem was solved and the system philic–mesophilic) anaerobic sequencing reactor systems. Wen
rapidly recovered temperature and performance. In this sense, et al. (2007) reported 0.88 L CH4/(L d) for the anaerobic digestion
Chae et al. (2008) reported that although temperature shocks led of liquid dairy manure using a sequential CSTR system.
to a reduction in biogas production, methanogenic bacteria have The biogas production rate of the CSTR digester was affected not
a considerable ability to adapt to moderate temperature changes. only by HRT but also especially by the quality of the feed. This
On day 140, the HRT was set at 10 days. For this experimental quality was related to the VFA content of the SLF. In this work,
period, the combination of the shortest HRT and the best character- the feed characteristics were not the same for the different HRTs
istics of the feed in terms of organic load and VFA content resulted tested. Mean CODVFA concentrations were 9.2, 6.4, 12.4 and 14.3
in an OLR of 4.5 kg VS/(m3 d) and the highest biogas yields. Specific for the SLF used as feed at 20, 16.7, 12.5 and 10 days HRTs, respec-
biogas production slightly decreased due to the lower HRT (328 L/ tively. For the shorter HRTs, 12.5 and 10 days HRT, the VFA concen-
kg VS), but mean biogas production increased to 2210 L/d and tration of the feed was also higher. For this reason, biogas
volumetric biogas production rate reached the maximum value production increased significantly, which explains the higher
of the test, 1.47 m3 biogas/(m3 d). The reactor showed a stable specific methane production and biogas production found in this
behavior, but the presence of VFA in the effluent was observed. work compared with those reported by other authors. In terms of
The methane content of biogas throughout the experimental methane yield per liter of feed, from the data in Table 4 and meth-
period ranged from 57% to 72% (Fig. 4). Under steady state condi- ane content of biogas, it can be deducted that the CSTR system
tions, the methane content of biogas was consistent: 63%, 65%, yielded 8.3, 7.8, 9.5 and 10.0 L CH4/L feed (SLF) at 20, 16.7, 12.5
63% and 68% for 20, 16.7, 12.5 and 10 days HRT, respectively. and 10 days HRT, respectively, under steady state conditions.
Decreases in methane content were detected when HRT was
reduced until steady state conditions were reached. On the other
hand, the highest methane content in biogas (72%) was reached 3.2.3. Volatile fatty acids in anaerobic effluent
between 121 and 123 days when the reactor performance was Fig. 5 shows the CODVFA concentration values in influents and
recovering from a drop in temperature. That was probably due to effluents during the experimental period. With regard to the VFA
the higher solubility of CO2 at the lower temperature. content in influents, differences were observed amongst feeds.
The results obtained in this test showed that the anaerobic For the shortest HRT, feed showed higher organic loads resulting
digestion of the screened liquid fraction of dairy manure at 37 °C in VFA contents of the SLF between 12,000 and 14,000 mg/L. Cold
permits stable operation with HRTs as short as 10 days (OLR winter temperatures allowed for good conservation of manure in
4.5 kg VS/(m3 d)) and the obtainment biogas and methane volu- the storage tanks during operational periods of 12.5 and 10 days
metric production rates of 1.47 m3 biogas/(m3/d) and 1.00 m3 HRT. There was no VFA accumulation in effluents for 20 and
CH4/(m3/d). The system was able to rapidly adapt to changes in 16.7 days HRT. At 12.5 days HRT concentrations between 186 and
operating conditions. A few days were enough to reach stable 423 mg CODVFA/L were observed. On day 120, the drop in the
conditions when passing from 2.1 to 3.5 kg VS/(m3 d) and from temperature reactor induced an increase in the VFA concentration
3.5 to 4.5 kg VS/(m3 d) with significant increases in biogas produc- up to 1764 mg CODVFA/L. Subsequently, when reactor recovered
tion. Likewise, the system rapidly recovered to steady state condi- temperature, the VFA concentration decreased to 238 mg CODVFA/
tions after moderate temperature shocks. L on day 130. At 10 days HRT the concentration of VFA in the

16000 22
80 22

14000 20
70 20
12000
60 18
18
CODVFA (mg/L)

10000
HRT (days)

50 16
HRT (days)

16
% CH4

8000
40
14
14 6000
30
12
12 4000
20
2000 10
10 10
0 8
0 8 0 20 40 60 80 100 120 140 160 180
0 20 40 60 80 100 120 140 160 180 Day
Day VFA inf VFA ef HRT
%CH4 biogas HRT
Fig. 5. VFA in affluent and effluent during CSTR operation. This figure represents
Fig. 4. Biogas methane content during CSTR operation. The methane content of punctual CODVFA concentration during the experimental period for influents (white
biogas (black circles) is represented during all the experimental period. The black circles) and effluents (black circles). The black line represents the HRT during the
line represents the HRT during the test. test.
2172 C. Rico et al. / Waste Management 31 (2011) 2167–2173

120 16
(1) (2)
14
100
12

CH4 (mL/(g VS·d))

80
CH4 (mL/g VS)

10

60 8

6
40
4
20
2

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Day Day
S-20 S-16.7 S-12.5 S-10 S-20 S-16.7 S-12.5 S-10

Fig. 6. Methane cumulative (1) and daily specific methane yields (2) of digestate samples. Graphic (1) represents methane accumulated at batch tests for digestate samples
and graphic (2) the daily specific methane yields for the same tests (White circles-digestate from CSTR at 20 days HRT; black rhombus-digestate from CSTR at 16.7 days HRT;
white triangle digestate from CSTR at 12.5 days HRT; black squares-digestate from CSTR at 10 days HRT).

effluent rose to values between 621 and 835 mg CODVFA/L. The this case, manure from the same farm resulted in quite different
only VFAs detected in effluents were acetic and propionic acids; methane yields.
acetic was predominant. The effluent pH was maintained close to If we use the results obtained in this work to estimate the
8, the VFA accumulation was not critical, and the biogas production amount of methane potential in Cantabria from the SLF of manure
and reactor performance were stable. However, the presence of produced by intensive dairy farming (1,844,391 tons manure per
VFA in the effluent indicated that the system was not able to pro- year according to PSE-PROBIOGAS), 29.8 and 19.4 million Nm3 bio-
cess all the incoming substrate and monitoring of the reactor was gas and methane could be produced per year respectively. These
stopped. volumes of biogas would represent an electricity production of
67.9 GWh/year in combined heat and power production systems,
2% of total electricity consumed in 2008 in Cantabria.
3.3. Residual methane from digestate

Digestate samples were collected at the exit of the CSTR on days 4. Conclusions
28, 69, 117 and 174 (Samples: S-20, S-16.7, S-12.5, S-10), corre-
sponding to HRTs of 20, 16.7, 12.5 and 10 days respectively. The performance of a 1.5 m3 volume CSTR digester processing
Fig. 6 shows the cumulative and daily specific methane yields for the screened liquid fraction of dairy manure was analyzed herein.
the different digestates. Sample S-10 still had a considerable meth- At HRTs as short as 10 days, the digester showed a stable operation
ane potential (103 mL CH4/g VS), whereas sample S-12.5 yielded reaching a volumetric biogas production rate close to 1.5 m3/
64 mL CH4/g VS. That was predictable due to the short HRT in the (m3 d). Under these conditions, the digestate yielded an attractive
CSTR reactor and the presence of VFA in the digestate. For samples amount of gas, 28.4% of that produced in the CSTR, which implies
S-20 and S-16.7, the residual methane was practically negligible that the digestate tank should be covered to capture its residual
(less than 15 mL CH4/g VS). After 60 days in batch residual meth- methane yield. At 20 days HRT, a larger reactor size would be re-
ane assays, samples S-10 and S-12.5 yielded 2.85 and 1.60 L CH4/ quired, but since the residual methane yield of digestate was neg-
L digestate, respectively, whereas the yields recorded from samples ligible, the storage tank of digested material could remain
S-20 and S-16.7 were significantly lower, 0.35 and 0.28 L CH4/L uncovered without significant gaseous emissions to the atmo-
digestate. That is in accordance with Menardo et al. (2011) who re- sphere and methane losses. The potential of biogas production
ported that high OLR and short HRT promote that digestates still from dairy cattle manure in the region of Cantabria represents a
contain considerable amounts of undigested organic matter. good opportunity. Only the screened liquid fraction could produce
Residual methane recovered from digestates represent 4.2%, the 2% of total electricity consumed in the region.
3.6%, 16.8% and 28.4% of the methane yields obtained in the CSTR
for HRT of 20, 16.7, 12.5 and 10 days, respectively. Weiland (2003) Acknowledgements
reported additional methane yields, as high as 15%, for digestate
post-methanization. These differences are explained by the differ- This project was made possible by a contract between the Envi-
ent characteristics of the feed in terms of methane potential during ronment and Education Councils of the Government of Cantabria
the experimental period. and the University of Cantabria and was supported by the Environ-
Adding methane yields during CSTR operation and residual ment Council, Project 05-640.02-2174 entitled ‘‘Environmental
methane yields in batch tests would result in the following global Management of Dairy Cattle Manure in Cantabria’’. The study pre-
methane potentials: 8.7, 8.1, 11.1 and 12.9 L CH4/L SLF, which cor- sented herein was performed within the framework of the NOVE-
responds to materials employed during CSTR operation at 20, 16.7, DAR Consolider CSD2007-00055.
12.5 and 10 days HRT, respectively. The SLF employed during the
last period (10 days THR) yielded 59% more methane than the References
SLF used at 16.7 HRT, which indicates the importance of good man-
agement practices with manure from its generation to the entrance Albertson, M.L., Pruden, A., Oliver, R.T., 2006. Enhanced anaerobic digestion of
into the anaerobic digester. Uncovered pits combined with heavy biomass waste for optimized production of renewable energy and solids for
compost. Int. Congress Series 1293, 221–229.
rains and manure storage under temperate temperatures of 20 °C APHA, 1998. Standard methods for the examination of water and wastewater, 18th
leads to lower methane production per mass or volume of SLF. In ed. American Public Health Association, Washington, DC, USA.
 C. Rico et al. / Waste Management 31 (2011) 2167–2173 2173

Atallah, T., Andreux, F., Choné, T., Gras, F., 1995. Effect of storage and composting on Lo, K.V., Bulley, N.R., Liao, P.H., Whitehead, A.J., 1983. The effect of solids-separation
the properties and degradability of cattle manure. Agr. Ecosyst. Environ. 54, pre-treatment on biogas production from dairy manure. Agr. Wastes 8, 155–
203–213. 165.
Chae, K.J., Jang, A., Yim, S.K., Kim, I.S., 2008. The effects of digestion temperature and Mantovi, P., Fabbri C., Soldano M., Piccinini S., 2010. Effect of solid/liquid separation
temperature shock on the biogas yields from the mesophilic anaerobic on raw and digested slurries. Proceedings of the 14th International Ramiran
digestion of swine manure. Bioresource Technol. 99, 1–6. Conference.
Boe, K., Angelidaki, I., 2009. Serial CSTR digester configuration for improving biogas Menardo, S., Gioelli, F., Balsari, P., 2011. The methane yield of digestate: effect of
production from manure. Water Res. 43, 166–172. organic loading rate, hydraulic retention time, and plant feeding. Bioresource
Converse, J.C., Koegel, R.G., Straub, R.J., 1999. Nutrient and solids separation of dairy Technol. 102, 2348–2351.
and swine manure using a screw press separator. ASAE, St. Joseph, MI. ASAE Møller, H.B., Lund, I., Sommer, S.G., 2000. Solid–liquid separation of livestock slurry:
paper no 99–4050. efficiency and cost. Bioresource Technol. 74, 223–229.
Dugba, P.N., Zhang, R., 1999. Treatment of dairy wastewater with two-stage Møller, H.B., Sommer, S.G., Ahring, B.K., 2002. Separation efficiency and particle size
anaerobic sequencing batch reactor systems-thermophilic versus mesophilic distribution in relation to manure type and storage conditions. Bioresource
operations. Bioresource Technol. 68, 225–233. Technol. 85, 189–196.
García, H., Rico, C., García, P.A., Rico, J.L., 2008. Flocculants effect in biomass Morris, D.R., Lathwell, D.J., 2004. Anaerobically digested dairy manure as fertilizer
retention in a UASB reactor treating dairy manure. Bioresource Technol. 99, for maize in acid and alkaline soils. Commun. Soil Sci. Plan. 35, 1757–1771.
6028–6036. Ogejo, J.A., Li, L., 2010. Enhancing biomethane production from flush dairy manure
Hartmann, H., Ahring, B.K., 2005. Anaerobic digestion of the organic fraction of with turkey processing wastewater. Appl. Energ. 87, 3171–3177.
municipal solid waste: influence of co-digestion with manure. Water Res. 39, Pain, B.F., Hepherd, R.Q., Pittman, R.J., 1978. Factors affecting the performance of
1543–1552. four slurry separating machines. J. Agr. Eng. Res. 23, 231–242.
Hawkes, F., Rosser, B., Hawkes, D., Statham, M., 1984. Mesophilic anaerobic Rico, J.L., García, H., Rico, C., Tejero, I., 2007. Characterisation of solid and liquid
digestion of cattle slurry after passage through a mechanical separator: fractions of dairy manure with regard to their component distribution and
factors affecting gas yield. Agric. Wastes 10, 241–256. methane production. Bioresource. Technol. 98, 971–979.
Holm-Nielsen, J.B., Al Seadi, T., Oleskowicz-Popiel, P., 2009. The future of Weiland, P., 2003. Production and energetic use of biogas from energy crops and
anaerobic digestion and biogas utilization. Bioresource Technol. 100, 5478– wastes in Germany. Appl. Biochem. Biotech. 109, 263–274.
5484. Weiland, P., 2006. Biomass digestion in agriculture: a successful pathway for the
Kaparaju, P.L.N., Rintala, J.A., 2003. Effects of temperature on post-methanization of energy production and waste treatment in Germany. Eng. Life Sci. 6, 302–309.
digested dairy cow manure in a farm-scale biogas production system. Environ. Wen, Z., Frear, C., Chen, S., 2007. Anaerobic digestion of liquid dairy manure using a
Technol. 24, 1315–1321. sequential continuous-stirred tank reactor system. J. Chem. Technol. Biot. 82,
Kaparaju, P., Buendia, I., Ellegaard, L., Angelidaki, I., 2008. Effects of mixing on 758–766.
methane production during thermophilic anaerobic digestion of manure: lab- Wu, Z., 2007. Phosphorus and Nitrogen distribution of screw press separated dairy
scale and pilot-scale studies. Bioresource Technol. 99, 4919–4928. manure with recovery of bedding material. Appl. Eng. Agric. 23, 757–762.
Karim, K., Klasson, K., Hoffman, R., Drescher, S., De Paoli, D., Al-Dahhan, M., 2005. Zhang, R.H., Westerman, P.W., 1997. Solid-liquid separation of animal manure for
Anaerobic digestion of animal waste: effect of mixing. Bioresource Technol. 96, odor control and nutrient management. Appl. Eng. Agric. 13, 385–393.
1607–1612. Zeeman, G., Sutter, K., Vens, T., Koster, M., Wellinger, A., 1988. Psychrophilic
Liao, P.H., Lo, K.V., Chieng, S.T., 1984. Effect of liquid-solids separation on biogas digestion of dairy cattle and pig manure: start-up procedures of batch, fed-
production from dairy manure. Energ. Agr. 3 (1), 61–69. batch and CSTR-type digesters. Biol. Wastes 26, 15–31.