Bioresource Technology 181 (2015) 238–246

Contents lists available at ScienceDirect

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

Cow, sheep and llama manure at psychrophilic anaerobic co-digestion

with low cost tubular digesters in cold climate and high altitude
J. Martí-Herrero a,b,⇑, R. Alvarez c, R. Cespedes d, M.R. Rojas c, V. Conde c, L. Aliaga e, M. Balboa e, S. Danov a
a
Centre Internacional de Mètodes Numèrics en Enginyeria (CIMNE), Building Energy and Environment Group, Edifici GAIA (TR14), C/Rambla Sant Nebridi 22, 08222
Terrassa, Barcelona, Spain
b
Instituto Nacional de Eficiencia Energética y Energías Renovables (INER), 6 de Diciembre N33-32, Quito, Ecuador
c
IIDEPROQ, UMSA, Plaza del Obelisco 1175, La Paz, Bolivia
d
Estación Experimental de Choquenaira, UMSA, Viacha, Bolivia
e
Energising Development Bolivia (EnDev-Bolivia/GIZ), Bolivia

h i g h l i g h t s

 Low cost tubular digesters adapted to cold climate are monitored.

 Four digesters fed with llama, cow and codigestion of both with sheep manure.
 Cow–sheep co-digestion results in 100% more biogas production respect cow manure.
 The retention of solids is high and a bottom sludge outlet is strongly recommended.
 The increase of OLR leads to an increase in BPR but reduces bioslurry quality.

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

Article history: The aim of this research is to evaluate the co-digestion of cow and llama manure combined with sheep
Received 21 October 2014 manure, in psychrophilic conditions and real field low cost tubular digesters adapted to cold climate. Four
Received in revised form 10 January 2015
digesters were monitored in cold climate conditions; one fed with cow manure, a second one with llama
Accepted 14 January 2015
Available online 22 January 2015 manure, the third one with co-digestion of cow–sheep manure and the fourth one was fed with
llama–sheep manure. The slurry had a mean temperature of 16.6 °C, the organic load rate was
Keywords: 0.44 kgvs m3 d1 and the hydraulic retention time was 80 days. After one hundred days biogas produc-
Anaerobic digestion tion was stable, as was the methane content and the pH of the effluent. The co-digestion of cow–sheep
Low temperature manure results in a biogas production increase of 100% compared to the mono-digestion of cow manure,
Psychrophilic while co-digestion of llama–sheep manure results in a decrease of 50% in biogas production with respect
Co-digestion
to mono-digestion of llama manure.
Biogas
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction tubular digesters that do not require active mixing or heating

Anaerobic digestion has a long history in developing countries
in order to take advantage of the biogas produced as fuel for cook-
ing, the bioslurry as fertilizer and the decontamination process as a
waste treatment system. At the household level, digesters are
mostly fed with pig or cow manure (Bond and Templeton, 2011;
Chen et al., 2012; Martí-Herrero et al., 2014a), which is related to
the household technology implemented so far, mainly low cost
⇑ Corresponding author at: Centre Internacional de Mètodes Numèrics en
Enginyeria (CIMNE), Building Energy and Environment Group, Edifici GAIA (TR14),
C/Rambla Sant Nebridi 22, 08222 Terrassa, Barcelona, Spain.
devices. Since no active heating is employed, these digesters were
installed mainly in warm and hot climate regions, where the ambi-
ent temperature is high enough to permit the anaerobic digestion
inside the digester in a range of mesophilic temperatures. In these
hot climate regions the low cost digesters are buried in the ground
and have a slurry temperature similar to that of the soil. This can
vary from 26 °C to 30 °C (Lansing et al., 2008; Terradas-Ill et al.,
2014). Small and medium farmers from these hot climatic regions
have enough pigs or cows to collect fresh manure for their digest-
ers. Mesophilic anaerobic digestion is well known and most of the
50 million household digesters installed worldwide function in this
range (Bond and Templeton, 2011). However, less information
and dissemination of low cost digesters has been done in hilly

http://dx.doi.org/10.1016/j.biortech.2015.01.063
0960-8524/Ó 2015 Elsevier Ltd. All rights reserved.
 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246
highlands or cold climate conditions. Kanwar and Guleri (1994)
compared the slurry temperature in hilly conditions of a fixed
dome digester and a plastic tubular digester. The slurry tempera-
ture of the fixed dome digester varied from 16 °C to 24 °C for eight
months around the summer period (mean ambient temperature
around 23 °C), and 13–15 °C for the other four months that corre-
spond to the winter period (mean ambient temperature around
9 °C). While the plastic tubular digester reached 2–3 °C higher tem-
peratures during summer periods, and 2–3 °C less during the win-
ter period, in comparison to the fixed dome model. This particular
plastic tubular digester installed at 1300 m above sea level
(m.a.s.l.) did not carry any conditioning for cold climate, thus
Kawar and Guleri concluded that this model ‘‘is not generally
suitable for hilly areas’’. Almost ten years later, in 2003, a low
cost plastic tubular digester adapted to cold climate was
installed at 4005 m.a.s.l. in the surroundings of La Paz – Bolivia
(Martí-Herrero, 2007). This particular digester led to the wide-
spread application of this technology to hilly and cooler climatic
regions, with more than 300 digesters installed in the highlands
of Bolivia (Martí-Herrero et al., 2014a) and Peru (Garfí et al.,
2012). Psychrophilic anaerobic digestion that corresponds to
colder regions with no active heating devices is less reported in
literature, even though there is a huge potential of implementation
of this AD technology (Martí-Herrero et al., 2014a) in a great
number of households.
The cold regions in the so called developing countries, usually
converge at high altitudes as in the highlands of Bolivia, Peru,
Ecuador, Chile, Colombia, Mexico, China, Pakistan, Nepal, etc.,
and are characterized by the lack of biomass to be used as cooking
fuel, the nutrient poverty of the soil and the corresponding low
crop productivity. Usually some cattle are owned by part of the
rural population and dry manure is a common alternative used
as fuel for cooking.
The AD technology provides energy (in the form of biogas), dis-
placing the use of local biomass (deforestation), and the fertilizer
(bioslurry) increases the crops productivity. Both products (biogas
and bioslurry) are the outcome of the digestion of a local resource
easily obtained (fresh manure). Though available manure to feed
the digesters in cold regions comes from animals, about which lit-
erature is scarce. In the Andean highlands there are no pigs, and
cattle are only found up to 4000 m.a.s.l. At higher altitudes and/
or drier regions sheep and llamas take over the role of cattle as
livestock, but no information has been reported so far about
digesters working with these manure types and conditions, except
at laboratory scale data (Alvarez et al., 2006; Alvarez and Lidén,
2008a,b, 2009).
Alvarez et al. (2006) report the biogas production of llama and
cow manure measuring the effects of altitude (pressure), low psy-
chrophilic and mesophilic temperatures, Hydraulic Retention Time
(HRT) and Organic Load Rate (OLR). Alvarez concludes that the
main parameter for biogas production is temperature, followed
by the HRT and OLR, however no significance was found for atmo-
spheric pressure (altitude). The methane yield for cow manure at
11 °C AD had a value of 6.4–33.6 lCH4 kg1 VS and at 35 °C of 49.6–
131.3 lCH4 kg1
VS , while for llama manure the values at 11 °C were
3.3–19.31 lCH4kg1 1
VS and at 35 °C of 35.6–84.1 lCH4kgVS . He also con-
cluded that llama manure has a great potential for biogas production
because of the high content of volatile solids (VS), which is greater
than in other types of manures (44.31% of VS at wet weight).
The effects of daily temperature variation on anaerobic diges-
tion is reported at the laboratory scale by Alvarez and Lidén
(2008a) using a mixture of llama–cattle–sheep manure. In this
study the daily temperature varies between 11–25 °C, 15–28 °C
and 19–32 °C respectively, while other reactors have a fixed slurry
temperature of 18 °C, 25 °C and 35 °C. Alvarez and Lidén found that
during the periods of low temperature the digester almost stopped
239
the biogas production, but during high temperature periods the
biogas production turned ‘‘unexpectedly’’ high. Furthermore,
biogas production of a constant slurry temperature digester is
similar to that of biogas produced in a digester with a similar mean
slurry temperature but with a daily temperature variation of
±6–7 °C around the mean value. Finally, in a daily temperature var-
iation digester, biogas production during low temperature periods
is much less than production in a digester with similar low, but
constant, temperatures pointing out the adaptation of bacteria to
cold thermal conditions.
The slurry temperature of low cost field digesters does not have
this daily temperature variation (3 °C of daily slurry temperature
amplitude, reported by Martí-Herrero et al., 2014b), but presents
annual variations if the digester is not adapted to cold climate con-
ditions. Kalia and Kanwar (1998) report in a long term thermal
monitoring that fixed dome digesters in hilly conditions have a
slurry temperature of 22–23 °C during summer (3 °C less than
the mean ambient one), and 13–14 °C during the winter period
(4 °C above the mean ambient temperature). Tubular digesters
adapted to cold climate seem to fulfill a better thermal perfor-
mance with less annual slurry temperature variation. Perrigault
et al. (2012) report data of a tubular digester adapted to cold cli-
mate working at 3400 m.a.s.l, with a slurry temperature of
24.5 °C, (which is 8.4 °C higher than the mean ambient tempera-
ture). Also, Martí-Herrero et al. (2014b) report a slurry tempera-
ture of 16.4 °C for winter (10.3 °C above the mean ambient
temperature) and 16.7 °C for summer (6.7 °C above the mean
ambient temperature) for a tubular digester adapted to cold
climate working at 3884 m.a.s.l.
To increase biogas production of low cost digesters in cold cli-
mate regions, some strategies were detected to optimize thermal
performance (Perrigault et al., 2012); better design methodology
for tubular digesters (Martí-Herrero, 2011; Martí-Herrero and
Cipriano, 2012), higher rates of load (Ferrer et al., 2011) and the
introduction of soda bottles inside the digesters as a biofilm carrier
(Martí-Herrero et al., 2014b). Thus one of the actual issues in
anaerobic digestion to get better biogas production is the
co-digestion, although 50% of overall papers on this topic were
published between 2012 and 2013 (Mata-Alvarez et al., 2014),
and present no information for digesters in cold climate and high
altitude regions.
Anaerobic co-digestion with animal manures, as with mono-
digestion, is focused on pig and cow manure (Mata-alvarez et al.,
2014), which is the typical livestock of hot and warm regions,
but not for cold climate and high altitude regions. Garfí et al.
(2011) report the results of a psychrophilic co-digestion of cow
and pig manure in low cost tubular digesters at 2800 m.a.s.l, find-
ing that the co-digestion of guinea pig and cow manure (0.08 m3
m3 d1) produces more biogas than mono digestion of guinea
pig manure (0.03 m3 m3 d1), but less than the mono digestion
of cow manure (0.12 m3 m3 d1). At higher altitudes, sheep, llama
and cow are the local livestock of the Andean region. The
co-digestion of these three substrates were studied in laboratory
scale by Alvarez and Lidén (2009), at 25 °C, HRT of 50 days and
1
1.2 kgVS m3 d1, obtaining 0:14 m3CH4 kgVS for 50–50% (VS based)
1
mix of llama and sheep manure, 0:10 m3CH4 kgVS for 50–50% mix
1
of cow and sheep manure, and 0:09 m3CH4 kgVS for 50–50% mix of
llama and cow manure. The biogas production at same conditions
1
for mono-digestion from llama manure were 0:09 m3CH4 kgVS , for
1 1
cow manure 0:10 m3CH4 kgVS and 0:12 m3CH4 kgVS for sheep manure.
In this study, co-digestion of llama-sheep and cow-sheep man-
ure are evaluated in real cold climate and highland conditions, and
are compared to mono-digestion of llama and cow manure in
experimental low cost tubular digesters adapted to cold climate
(Fig. 1). These experimental systems are similar to those hundreds
of household systems already working in the Andean highlands
240 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246
from Bolivia and Peru (Martí-Herrero et al., 2014a,b). Furthermore
the strategy of increasing the OLR in order to produce more biogas
is evaluated as well.
2. Methods
The research has been realized at 3884 m.a.s.l in the highland of
Bolivia, at the Research Center of Digesters, Biogas and Bioslurry
(CIB3: Centro de Investigación en Biodigestores, Biogás y Biól),
sited in Viacha (Bolivia), 22 km from La Paz city (Martí-Herrero
et al., 2014a). At the CIB3, 10 pilot low cost tubular digesters are
operated and monitored in real cold weather conditions. Four of
them are the focal point of this research. The results can also be
compared with other two digesters that are fed with cow manure
in the same laboratory and same weather conditions, though with
lower OLR, higher HRT, and one of them containing a biofilm sup-
port carrier, reported previously in Martí-Herrero et al. (2014b).
2.1. Experimental low cost tubular digesters
The experimental low cost tubular digesters represent the small
scale of the household low cost tubular digesters adapted to cold
climates, designed following the methodology reported by
Martí-Herrero and Cipriano (2012). These household tubular
digesters are fed with 20 kg of manure which is mixed with 60 l
of water every day; have 81 days of HRT; are semi-buried in the
ground; are insulated with straw in the trench; and the biogas bell
can gain solar radiation, thus heat. These digesters adapted to cold
climates are inside a compact greenhouse with massive adobe
walls, oriented east–west, with the shed greenhouse oriented to
Fig. 1. Up-left: Main research infrastructure of the CIB3, with 10 experimental tubular digesters operated and monitored in real weather conditions at 3850 m.a.s.l (Viacha, La
Paz) (Martí-Herrero et al., 2014a). Up-right: Low cost tubular digester adapted for cold climate inside a greenhouse with adobe walls from (Martí-Herrero et al., 2014b).
Below: Design of the typical household digester (Martí-Herrero et al., 2014a).
the winter sun of the north (Perrigault et al., 2012; Martí-Herrero
et al., 2014a).
The experimental low cost tubular digesters are the same in
materials, design and installation methodology to those 300
household systems installed in farms of the Bolivian Altiplano
(Martí-Herrero et al., 2014a), only smaller, being about 1 m3 liquid
volume and 4.7 m long. A comparison between the experimental
digester at CIB3 and the household one can be found in
Martí-Herrero et al. (2014b).
2.2. Analytical analysis
Biogas production has been recorded almost every day in situ,
using a commercial low pressure diaphragm gas meter (G2.5
Metrix). Data from the initial load of the digesters, which started
in May 2011, has been recorded for 300 days, ending on April
2012. Biogas samples were collected in Tedlar bags, once the
digesters were stabilized, to examine the methane and carbon
dioxide concentration using a gas chromatograph (Shimadzu
Model GC14B, Japan) equipped with a thermal conductivity
detector (TCD), and a Carboxen-1010 plot Capillary column
30 m  0.53 mm ID (Supelco, USA). The injector, detector and oven
temperatures were 130 °C, 200 °C, and 100 °C, respectively. Helium
served as the carrier gas at a pressure of 300 kPa.
For the characterization of the influent, samples were collected
in winter and summer season. The bioslurry was collected and ana-
lyzed every 15 days once the digesters were stabilized. Total solids
(TS), volatile solids (VS) and pH were determined according to
standard methods (Clesceri et al., 2000). The total solids (TS) con-
tent was determined after heating (105 °C for 1 h), cooling, desic-
cating, and weighing procedures that were repeated until the
 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246
weight change was less than 4%. Volatile solids were determined
by ignition of the residue produced in TS analysis to constant
weight, in a muffle furnace at a temperature of 550 °C.
2.3. Feedstock
Llama, sheep and cow manure was collected from the CIB3 at
16°410 3900 S and 68°170 1400 W in the Bolivian Highlands. The man-
ure was collected every day from the floor of the cattle shed in
the morning and loaded to the digesters later the same day.
Manure samples from the winter and the summer period were
collected and analyzed finding no significant differences between
seasons. Table 1 shows the main characteristics of the manures
used in the present study as mean values for winter and summer
seasons. Fresh manure from llama and sheep is much drier than
cow manure, and reveals a high content of volatile solids. This
means that the organic matter content of the Andean cow manure
is a quarter of the sheep manure, and about a third of the organic
matter contained in llama manure.
2.4. Temperature monitoring
The temperature of one of the ten digesters installed at the CIB3
has been monitored previously and the results were reported in
Martí-Herrero et al. (2014b). This temperature data will be
employed in the present work, since all 10 digesters are the same
in design, material and orientation; the assumption being that the
same thermal performance as that already reported is to be
expected.
2.5. Operation of digesters
The experimental low cost tubular digesters started being filled
in May 2012; the digesters D50ll and D50lls were loaded with fresh
llama manure in a relation 1:5 (manure:water), while D50c and
D50cs were loaded with cow manure in a relation 1:3 (manure:-
water). These initial ratios water:manure were chosen according
to the standard operation of household tubular digesters, that
correspond to 5–8%TS (3–6%VS w.w.) in the influent, since less
dilution produces clogging inside the digester in the long term
(Martí-Herrero et al., 2014a).
To compare the results of the present study the load for each
digester was calculated for an OLR of 0.7 kgSV m3 d1 and HRT
of 50 days, as noted in Table 2. The OLR and HRT selected were
65% higher and 45% lower, respectively, than reported at previous
studies about anaerobic digestion of cow manure in tubular
digesters at similar conditions (Martí-Herrero et al., 2014b). Also,
the HRT of 50 days is selected as a reference value since previous
laboratory scale experiments already used this value (Alvarez
et al., 2006; Alvarez and Lidén, 2009).
The digesters began to produce the first amount of biogas after
20 days, and 100 days later the production was stabilized. The
schedule for loading the digesters has been consciously charted,
allowing the local operator some freedom to load the digesters
according to his time availability. The operator, as well as the
Table 1
Characteristic of the fresh manure.
Cow Llama Sheep
Total solids (%wet) 16.2 58.3 67.6
Volatile solids (%wet) 12.5 34.4 50.3
Volatile solids (%TS) 68.2 70.9 66.4
Total Kjeldahl N (%TS) 1.6 1.5 1.3
Phosphorus (%TS) 0.3 0.3 0.8
Potassium (%TS) 1.5 0.9 1.1
241
farmer, is aware that a daily loading failure can be compensated
by a double load at the following day. The amount of water and
manure is measured every day with a container that has a mark
at a specific level showing the reference load.
Four experimental low cost tubular digesters were considered
for this study. Two of them have a mono digestion process while
the other two have a co-digestion process of two different man-
ures. Since the loading of the digesters was not accomplished every
day, because of the operation schedule as commented above, these
HRT (50d) and OLR (0.7 kgSV m3 d1) reference values were mod-
ified by real experimental procedures and will be presented in the
results section of the present study. D50ll was loaded with 1.7 kg
of fresh llama manure and 15.1 l of water while D50c was loaded
with 4.8 kg of fresh cow manure and 12.8 l of water. The co-diges-
tion ratios were fixed to 50–50% of both manures based in VS, for
both cases D50lls and D50cs, as realized in Alvarez and Lidén
(2009). D50lls was fed with 0.88 kg of llama manure and 0.6 kg
of sheep manure, mixed with 15.6 l of water, while D50cs was
fed with 2.38 kg of cow manure, 0.6 kg of sheep manure, and
14.3 l of water. These were the amounts of daily load for all four
experimental digesters, and if the operator doubles the amount
of these values, it was because there is an absence of load the
day before. Therefore real mean OLR and HRT can be determined
for this study, differing from the reference design parameters from
Table 2.
3. Results and discussion
3.1. Climate and thermal behavior of digesters
As already mentioned before, the ambient, slurry and 1 m depth
temperatures considered in this study are the same to those
reported in Martí-Herrero et al. (2014b), and are shown in Figs. 2
and 3. The slurry temperature of the digesters shows a daily range
of 3 °C around a mean value of 16.6 °C, according to psychrophilic
conditions. The mean daily ambient temperature is 9.9 °C, with a
mean daily minimum temperature of 4.7 °C and a maximum of
13.4 °C. The soil temperature is almost constant at 14 °C. Therefore,
as the slurry temperature is higher than the maximum ambient
and the soil one, the low cost tubular digesters adapted to cold
climate acts as a solar heat collector with thermal mass, as
has already been reported in Perrigault et al. (2012) and
Martí-Herrero et al. (2014b).
3.2. Real OLR and HRT due operation and biogas production
The biogas production was recorded almost every day, accord-
ing to the days that the local operator fed the digesters. The biogas
production recorded has been divided into stages due to the main-
tenance of the gas meters. These were filled with condensed water
from biogas, preventing normal operation, so it was necessary to
empty them every 2 months. Three biogas production profiles
were recorded for D50ll, D50cs and D50c, while only two produc-
tion profiles for D50lls, as shown in Figs. 2 and 3 (local conditions
of pressure and temperature). In order to estimate the biogas
production per day, the slope of the accumulative biogas profile
is represented in the Figures together with the R2 value of the fit
line, which in all cases is up to 0.99. Figs. 2 and 3 also show the
maximum and minimum daily temperature, with the slurry
temperature if available, and the feeding schedule of the digesters.
3.3. Real OLR and HRT
The digesters were monitored for more than 277 days, in which
173 loads were conducted. This means that the digesters were
242 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246

Table 2
Experimental design and daily load of the digesters.

Digester Liquid volume [l] Total load [l] Manure Water [l] Manure:water ratio References
Llama [kg] Cow [kg] Sheep [kg] OLR [kgSV m3 d1] HRT [days]
D50ll 840 16.799 1.72 15.08 8.7 0.70 50
D50lls 856 17.12 0.88 0.59 15.63 10.6 0.70 50
D50cs 865 17.3 2.38 0.60 14.32 4.8 0.69 50
D50c 880 17.6 4.84 12.76 2.6 0.69 50

Fig. 2. Up, accumulative biogas production (local conditions) in different periods for D50ll (mono-digestion of llama manure). Down, accumulative biogas production (local
conditions) in different periods for D50lls (co-digestion of llama and sheep manure). In both cases with the daily mean slurry temperature (when possible) and the minimum
and maximum ambient temperatures. Also load schedule is shown.

loaded 62.5% of the days, therefore the reference OLR f 0.7 kgSV
m3 d1 decreased to 0.44 kgSV m3 d1 and the HRT increased
from the reference value of 50–80 days (Table 3). This load
schedule is similar to the farmers’ schedule using digesters,
consisting in a load between 23% and 100% of the days
(Martí-Herrero et al., Unpublished results).
3.4. Biogas production
The biogas production of llama manure in psychrophilic diges-
tion varies from 0.11 m3 d1 to 0.14 m3 d1 (at local conditions),
while when llama manure is mixed with sheep manure 50–50%
(VS based) the biogas production declines to a range of 0.065–
0.067 m3 d1. In the case of pure cow manure in psychrophilic
digestion, biogas production varies from 0.09 m3 d1 to
0.12 m3 d1 and when it is mixed with sheep manure (50–50%
VS based) the production increases to 0.17–0.28 m3 d1.
In order to normalize this values to 0 °C and 1 atm, local condi-
tions for all the profiles were considered. Local biogas pressure is
496 mmHg (0.653 atm), and the biogas temperature considered is
equal to the daily mean ambient temperature. Biogas inside the
digester can reach higher temperatures than ambient temperature,
 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246 243

Fig. 3. Up, accumulative biogas production (local conditions) in different periods for D50cs (co-digestion of cow and sheep manure). Down, accumulative biogas production
(local conditions) in different periods for D50c (mono-digestion of cow manure). In both cases with daily mean slurry temperature (when possible) and the minimum and
maximum ambient temperatures. Also Load schedule is shown.

as reported in Martí-Herrero et al. (2014b), where the maximum
biogas temperature varies from 20 °C to 40 °C. But temperature
measurements collected from the gas meter show that the biogas,
once it moves through the instrument, has a temperature that can
be correlated to the daily mean ambient for a normalization
purpose.
Results are shown in Table 3, along with results from other
studies with these Andean manures at laboratory scale, and field
studies.
3.5. Biogas composition
The biogas composition has been measured every month for
D50ll and D50c, but only 7 and 4 samples were taken for D50lls
and D50cs, respectively. There was no annual variation in biogas
composition, while the values for standard deviation that were
found were around 5–9% for %CH4 and %CO2 throughout the mon-
itoring period. The values obtained for methane and carbon dioxide
in this study can be seen in Table 4.
The methane composition of both mono manure digesters,
D50ll and D50C, were in the range of previous results reported at
laboratory scale, with a variety of temperatures, OLR and HRT
conditions, by Alvarez et al. (2006) and Alvarez and Lidén
(2008b, 2009), with values of 42–53%CH4 for llama manure, and
39–61%CH4 for cow manure. In the case of cow manure mono-
digestion in this study, the 47.8%CH4 of methane obtained is close
to the range 47.2–47.5%CH4 previously reported by Martí-Herrero
et al. (2014b), with similar temperature conditions, but higher HRT
and lower OLR in relation to the present study (see Table 3).
For the co-digestion digesters the %CH4 is lower than the values
reported from laboratory scale which have higher temperatures
(25 °C) and OLR (1.2 kgSV m3 d1) by Alvarez and Lidén (2009),
being 45.6%CH4 for D50lls (50.6%CH4 at laboratory scale), and
44.8%CH4 for D50cs (51.5%CH4 at laboratory scale).
3.6. Bio-slurry quality
The digesters were fed with a mixture of water and manure in a
range of 4.47–5.98% TS and 3.52–3.45%SV (w.w.) (Table 5). TS
values were higher in llama manure mono-digestion, followed by
both sheep manure co-digestions and cow manure mono-
digestion. A field survey in the highlands of Bolivia to household
244 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246

Table 3
Characterization of the digesters and comparison with other laboratory and field results.

Manure Type of CH4SBPa %CH4 T slurry HRT OLR SBPa BPR*

study 1
[m3CH4 kgSV =d
1
] [°C] [d] [kgSV m3 d
1
]
1
[m3 kgSV d
1
] [m3 m3 d
1
]

Alvarez et al. (2006) Llama Laboratory 0.019 57 11 50 0.89 0.03 0.03

D50ll Llama Field 0.101 46.7 16.6 80 0.44 0.22 (0.02) 0.10 (0.01)
Alvarez and Lidén (2009) Llama Laboratory 0.090 53 25 50 1.20 0.17 0.20
Alvarez et al. (2006) Llama Laboratory 0.084 42 35 50 0.89 0.20 0.18
D50lls Llama + sheep Field 0.051 45.6 16.6 80 0.44 0.11 (0.00) 0.05 (0.00)
Alvarez and Lidén (2009) Llama + sheep Laboratory 0.140 50.6 25 50 1.2 0.28 0.33
D50cs Cow + sheep Field 0.152 44.8 16.6 80 0.43 0.34 (0.11) 0.15 (0.05)
Alvarez and Lidén (2009) Cow + sheep Laboratory 0.100 51.5 25 50 1.2 0.19 0.23
Alvarez et al. (2006) Cow Laboratory 0.024 39 11 50 0.52 0.06 0.03
Martí-Herrero et al. (2014b) (D90) Cow Field 0.109 47.2 16.6 124 0.26 0.23 0.06
Martí-Herrero et al. (2014b) (D90b) Cow + biofilm carrier Field 0.157 47.5 16.6 124 0.26 0.33 0.09
D50c Cow Field 0.081 47.8 16.6 80 0.43 0.17 (0.03) 0.07 (0.01)
Alvarez and Lidén (2009) Cow Laboratory 0.100 54.8 25 50 1.2 0.18 0.22
Alvarez et al. (2006) Cow Laboratory 0.131 61 35 50 0.52 0.22 0.11

(n) is the standard deviation.

a
Normal conditions (1 atm and 0 °C).

Table 4 3.7. Final considerations

Biogas composition.

%CH4 %CO2
D50ll [14] 46.7 (4.0) 41.9 (3.5)
D50lls [7] 45.6 (2.6) 44.8 (3.7)
D50cs [4] 44.8 (2.2) 47.4 (4.2)
D50c [15] 47.8 (2.8) 40.2 (3.3)
[n], number of samples.
(n), standard deviation.
digesters fed with cow manure report 3.93%TS and 3.20%VS (w.w.)
mean values for the mixing load (Martí-Herrero et al., Unpublished
results). In the present study, the dry matter content of manures
had a great mineral charge, from 1.02% (for cow) to 2.44% (for
llama), in relation to the survey mentioned before, which reports
0.73% of mineral matter for cow manure. Consequently, the high
content of inorganic matter in llama and sheep manure, although
in less degree at the second one, could be related to the fact that
the stables of these animals had a sand or stone floor, so the
manure probably carried some dust.
The bioslurry produced has a pH in a range of 7.5–7.7, close to
the values reported by Alvarez and Lidén (2009). Digester D50lls
produces the highest TS (3.16%) and VS (2.46%) content in the efflu-
ent, also being the digester that performsless reduction in those
parameters (40.91% reduction for TS and 30.15% for SV), while
the codigestion digester D50cs produces a bioslurry with 1.78%TS
and 1.53%SV reaching a reduction from influent of around 61%
for TS and SV.
Solid reduction percentage from inlet to outlet of the digesters
is much higher than that reported in literature. In the present
study, VS reduction varies from 40.73% for D50c to 61.33% for
D50cs, while at laboratory scale VS reduction is in the range of
12.52% for mono-digestion of cow manure, and up to 19.66% for
co-digestion of llama-sheep manure (Alvarez and Lidén, 2009).
The mineral matter of the influent of the four digesters moves in
a range of 2.44–1.02%, while the range for the bioslurries is slim-
mer, being 0.44–0.70%. D50ll is the digester with the highest inor-
ganic matter retention (81.11%), while D50c is the one with the
lowest inorganic matter retention, with 50.06%. In these extreme
cases the final content of mineral matter in the bioslurry is very
similar (0.46% for D50ll to 0.51% for D50c), which can be related
to the hydraulic performance of the tubular digesters and its
capacity to evacuate this kind of matter.
% Others The results obtained from the present study show that anaero-
11.4 (4.9) bic digestion in real cold climate conditions, using low cost tubular
9.6 (5.3) digesters adapted to cold climate, with llama, cow and co digestion
7.8 (4.7) of each with sheep manure is feasible. The pH of bioslurry is almost
11.9 (4.4)
constant with values around 7.5–7.7. Biogas production (Figs. 2
and 3) presents a steady rate of production with a stable methane
content of 44.8–47.8% and a standard deviation in the range of 2.2–
4.0. The anaerobic digestion is produced in psychrophilic condi-
tions, with a mean slurry temperature of 16.6 °C, a daily amplitude
of 3 °C, HRT of 80 days and OLR of 0.44 kgSV m3 d1.
The mono-digestion of cow manure (D50c) has a SBP of
0.17 m3 kg1
SV , while laboratory results (Alvarez and Lidén, 2009)
report a SBP of 0.18 m3 kg1SV for higher constant slurry temperature
(25 °C), HRT of 50 days and almost three times more OLR (1.2 kgSV -
m3 d1). These similar results can be explained by the higher HRT
of D50c, as compensation of the low slurry temperature in compar-
ison to the one from the laboratory. Alvarez and Lidén (2009) con-
sider this biogas yield lower than others reported with cow
manure and 25 °C (Hawkes et al., 1984; Kalia and Sing, 1998), and
point out that this could be explained as a consequence of ‘‘differ-
ences in manure composition, which affects the degradation
process’’.
Nitrogen content can be affected by manure composition. Nitro-
gen is needed to build the methanogenic cell structure, and 0.6%N
(w.w.) is the value recommended as a minimum to sustain the
anaerobic digestion process (NCA, 2001). Insufficient Nitrogen will
limit biogas production, and at the other hand ammonia will be
formed if there is an excess of nitrogen. The excess of ammonia
can inhibit the biogas production (Hansen et al., 1998), and as
more nitrogen goes into the digester, more ammonium nitrogen
will be found in the sludge.
Cow manure in the present study has 1.6% of TS of N, less than
41% of the values reported by ASAE Standard (2003) and only
0.26%N (w.w.), which can explain the low biogas yield obtained
in this study and other experiments in Andean conditions
(Alvarez and Lidén, 2009).
Co-digestion of cow-sheep manure (D50cs) is the mixture that
shows more SBP (0.34 m3 kg1 3
SV ) and BPR (0.15 m m
3 1
d ) produc-
tion, which correlates to the major reduction of SV (61.33%) from the
influent to the bioslurry.
When cow and sheep manure is combined, the nitrogen content
in the feeding substrate is increased by 51%, reaching 0.39%N
 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246 245

Table 5
TS, VS and mineral matter content in influent and bioslurry.

Digester Influent Bio slurry % Reduction ofa

TS VS Mineral matter TS VS Mineral matter PH N P K ST SV Mineral
(%w.w.) (%w.w.) (%w.w.) (%w.w.) (%w.w.) (%w.w.) (%TS) (%TS) (%TS) matter
D50ll 5.98 3.53 2.44 2.50 2.04 0.46 7.5 0.039 0.003 0.016 58.15 42.26 81.11
D50lls 5.34 3.52 1.82 3.16 2.46 0.70 7.6 0.046 0.005 0.023 40.91 30.15 61.67
D50cs 4.59 3.48 1.11 1.78 1.35 0.44 7.7 0.048 0.005 0.026 61.24 61.33 60.96
D50c 4.47 3.45 1.02 2.55 2.04 0.51 7.6 0.042 0.003 0.017 42.86 40.73 50.06
a
‘‘% Reduction of’’ is calculated as ‘‘100(%Xinfluent  %Xefluent)/%Xinfluent’’ for X being TS, VS and mineral matter (w.w.).

(w.w.), which could be the reason for the increase of 100% on bio-
gas production in comparison to cow monodigestion.
For llama manure mono-digestion (D50ll) the SBP is 0.22 m3
kg1
SV , which is close but somewhat higher than the value reported
at laboratory scale of 0.17 m3 kg1 SV from ALvarez and Lidén
(2009).The first value is measured in psychrophilic conditions
though, and the second one in mesophilic conditions, also, the first
value presents higher HRT than the second one. When llama manure
is mixed with sheep manure (D50lls), the SBP and BPR is reduced in
more than 50%, while at the laboratory scale this co-digestion
produces an increase of around 65%.
Llama mono-digestion seems to work positively since it
achieves better results than at the laboratory scale, despite the
0.85%N (w.w.). On the other hand, when sheep and llama manure
are mixed, the nitrogen content rises to 0.87% N (w.w.). At this
point, field experiments break off from laboratory ones at similar
Andean conditions. These field results of 50% less biogas produc-
tion in co-digestion of llama-sheep manure, in comparison to
mono-digestion of llama, should be considered with precaution,
since it contradicts all the literature about co-digestion benefits
(Mata-Alvarez et al., 2014) and specifically literature about
sheep-llama co-digestion (Alvarez and Lidén, 2009). The biogas
production of digester D50lls was collected in two different winter
periods, obtaining very similar SBP and BPR values at both of times,
with a 0.00 standard deviation value. On the contrary the other
three digesters (D50ll, D50cs and D50c) present differences in
biogas production between periods. This could indicate a problem
with this particular system (D50lls), which only records a maxi-
mum amount of biogas production, while the real value might be
higher. But other issues like ammonium nitrogen content, or the
particular thermal performance of this digester D50lls, could be
relevant to explain this behavior. Therefore, deeper experimenta-
tion and research should be undertaken in co-digestion of sheep
and llama manure in order to clear these doubts.
Differences in BPR between laboratory scale, from 0.20 m3
m d for llama mono-digestion to 0.33 m3 m3 d1 for llama-
3 1
sheep co-digestion, and in real cold climate conditions, from
0.07 m3 m3 d1 for cow mono-digestion to 0.15 m3 m3 d1 for
cow–sheep co-digestion, are related to the higher OLR of the
laboratory digesters respect field digesters.
D50c represents a digester where OLR increased (HRT
decreases) in comparison to typical values of household low cost
tubular digesters adapted to cold climate in the Andean region,
that are characterized by digester D90 reported by Martí-Herrero
et al. (2014b). Therefore increasing the OLR 65%, from 0.26 (D90)
to 0.43 kgSV m3 d1 (D50c), and reducing the HRT from 124 to
80 days, turns into an increase of 22% of BPR in comparison to
the typical D90 digester, though it reduces the SBP in 26%. Since
SBP is related to the anaerobic digestion efficiency and BPR to
the reactor efficiency, D50c develops superior reactor efficiency
but worse anaerobic digestion efficiency. So 65% more fresh man-
ure is used to generate 22% more biogas, producing a less digested
effluent, and therefore also lower quality bioslurry. Bioslurry
though, is the most valued product of anaerobic digestion from
farmers in the Andean region (Martí-Herrero et al., Unpublished
results), so to increase the BPR at the expense of SBP and bioslurry
quality does not seem to be socio-culturally acceptable, at least to
most of the Andean farmers. The proposal to add a low cost biofilm
carrier (soda bottle rings) reported in Martí-Herrero et al. (2014b),
increasing BPR by 50% and SBP by 44% with the same amount of
fresh manure that is used in D90, results in a more convenient
strategy to improve the biogas production and hence the bioslurry
quality.
High reduction of total and volatile solids together with mineral
matter were found in all four digesters (40.91–61.24% reduction for
TS, 30.15–61.33% for VS and 50.06–81.11% for mineral matter),
compared with data of similar digesters (D90), with values of
33–35% reduction for TS, VS and mineral matter (Martí-Herrero
et al., 2014b). The higher values presented in this study could be
related to the higher solid content in llama and sheep manure,
the higher OLR of these experiments, and to the fact that more sol-
ids are retained inside the digester, as demonstrated by mineral
matter accumulation. Consequently, in order to avoid possible
slurry clogging, a specific outlet to evacuate sludge accumulation
in the bottom of the digester is recommended when working with
high OLRs of cow, sheep and llama manure.
4. Conclusions
Low cost tubular digesters adapted to cold climate work in
psychrophilic conditions, demonstrating the feasibility of biogas
production by cow, llama and co-digestion of each with sheep
manure, with stabilized pH, biogas production and methane
content. Co-digestion of sheep and cow manure significantly
increases the biogas production in comparison to the mono-
digestion of cow manure. Llama manure digestion gives better field
results than the corresponding data from the laboratory, while
co-digestion of llama and sheep needs further research. The reten-
tion of solids is high and a bottom sludge outlet is strongly
recommended.
Acknowledgements
Financial support to carry out this research was provided by
Hivos and the Endev-Bolivia Program of the GIZ, as well as the col-
laboration of the Caribbean and Latin American Biodigesters Net
(RedBioLAC). Jaime Martí-Herrero is thankful to the Prometeo Pro-
ject of the Secretariat for Higher Education, Science, Technology
and Innovation of the Republic of Ecuador that funded part of his
work. The authors wish to acknowledge the revision of the docu-
ment by Mike Baker and Samay Schütt.
References
Alvarez, R., Lidén, G., 2008a. The effect of temperature variation on biomethanation
at high altitude. Bioresour. Technol. 99, 7278–7284.
Alvarez, R., Lidén, G., 2008b. Anaerobic co-digestion of aquatic flora and quinoa with
manures from Bolivian Altiplano. Waste Manage. 28, 1933–1940.
246 J. Martí-Herrero et al. / Bioresource Technology 181 (2015) 238–246
Alvarez, R., Lidén, G., 2009. Low temperature anaerobic digestion of mixtures of
llama, cow and sheep manure improved methane production. Biomass
Bioenergy 33, 527–533.
Alvarez, R., Villca, S., Lidén, G., 2006. Biogas production from llama and cow manure
at high altitude. Biomass Bioenergy 30, 66–75.
ASAE Standard, 2003. Manure Production and Characteristics, ASAE D384.1 FEB03,
USA, pp. 683–685.
Bond, T., Templeton, M.R., 2011. History and future of domestic biogas plants in the
developing world. Energy Sustainable Dev. 15, 347–354.
Chen, L., Zhao, L., Ren, C., Wang, F., 2012. The progress and prospects of rural biogas
production in China. Energy Policy 51, 58–63.
Clesceri, L.S., Greeberg, A.E., Eaton, A.D., 2000. Standard Methods for the
Examination of Water and Wastewater, 20th ed. American Public Health
Association (APHA), Washington, DC, USA.
Ferrer, I., Garfí, M., Uggetti, E., Ferrer-Martí, L., Calderon, A., Velo, E., 2011. Biogas
production in low-cost household digesters at the Peruvian Andes. Biomass
Bioenergy 35, 1166–1674.
Garfí, M., Ferrer-Martí, L., Irene Perez, I., Flotats, X., Ferrer, I., 2011. Co-digestion of
cow and guinea pig manure in low-cost tubular digesters at high altitude. Ecol.
Eng. 37, 2066–2070.
Garfí, M., Ferrer-Marti, L., Velo, E., Ferrer, I., 2012. Evaluating benefits of low-cost
household digesters for rural Andean communities. Renewable Sustainable
Energy Rev. 16, 575–581.
Hansen, K.H., Angelidaki, I., Ahring, B.K., 1998. Anaerobic digestion of swine manure
inhibition by ammonia. Water Res. 32 (1), 5–12.
Hawes, F.R., Roser, B.L., Statham, M., 1984. Mesophilic anaerobic digestion of cattle
slurry after passage through a mechanical separator: factor affecting gas yield.
Agric. Wastes 10, 241–256.
Kalia, A.K., kanwar, S.S., 1998. Long-term evaluation of a fixed dome Janata biogas
plant in hilly conditions. Bioresour. Technol. 65, 61–63.
Kalia, A.K., Sing, S.P., 1998. Horse dung as a partial substitute for cattle dung for
operating family-size biogas plants in a hilly region. Bioresour. Technol. 64, 63–
66.
Kanwar, S.S., Guleri, R.L., 1994. Performance evaluation of a family-size, rubber-
balloon biogas plant under hilly conditions. Bioresour. Technol. 50, 119–121.
Lansing, S., Botero, R., Martin, J.F., 2008. Waste treatment and biogas quality in
small-scale agricultural digesters. Bioresour. Technol. 99 (13), 5881–5890.
Martí-Herrero, J., 2007. Transfer of low-cost plastic biodigestores technology at
household level in Bolivia. Livest. Res. Rural Dev. 19 (12) (Article #192. [cited
August 2014]. Available from <http://www.lrrd.org/lrrd19/12/mart19192.htm>).
Martí-Herrero, J., 2011. Reduced hydraulic retention times in low-cost tubular
digesters: two issues. Biomass Bioenergy 35 (10), 4481–4484.
Martí-Herrero, J., Cipriano, J., 2012. Design methodology for low cost tubular
digesters. Bioresour. Technol. 108, 21–27.
Martí-Herrero, J., Chipana, M., Cuevas, C., Paco, G., Serrano, V., Zymla, B., Heising, K.,
Sologuren, J., Gamarra, A., 2014a. Low cost tubular digesters as appropriate
technology for widespread application: results and lessons learned from
Bolivia. Renewable Energy 71, 156–165.
Martí-Herrero, J., Alvarez, R., Rojas, M.R., Aliaga, L., Céspedes, R., Carbonell, J., 2014b.
Improvement through low cost biofilm carrier in anaerobic tubular digestion in
cold climate regions. Bioresour. Technol. 167, 87–93.
Martí-Herrero, J., Ceron, M., Garcia, R., Pracejus, L., Alvarez, R., Cipriano, X.,
Unpublished results. The influence of users behaviour on biogas production
from low cost household digesters: technical and socio-cultural field analysis.
Energy Sustainable Dev.
Mata-Alvarez, J., Dosta, J., Romero-Güiza, M.S., Fonoll, X., Peces, M., Astals, S., 2014.
A critical review on anaerobic co-digestion achievements between 2010 and
2013. Renewable Sustainable Energy Rev. 36, 412–427.
NCA, National Academia of Sciences, 2001. Methane Generation from Human,
Animal and Agricultural Wastes. Books for Business, New York, Hong Kong.
Perrigault, V Weatherford., Martí-Herrero, J., Poggio, D., 2012. Towards thermal
design optimization of tubular digesters in cold climates: a heat transfer model.
Bioresour. Technol. 124, 259–268.
Terradas-Ill, Cuong, H., Pham, C.H., Triolo, J.M., Martí-Herrero, J., Sommer, S.G., 2014.
Thermic model to predict biogas production in unheated fixed-dome digesters
buried in the ground. Environ. Sci. Technol. 48, 3253–3262.