Bioresource Technology 100 (2009) 56095615

Contents lists available at ScienceDirect

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

Anaerobic digestion of glycerol derived from biodiesel manufacturing

Jos ngel Siles Lpez *, Mara de los ngeles Martn Santos, Arturo Francisco Chica Prez,
Antonio Martn Martn
Departamento de Qumica Inorgnica e Ingeniera Qumica, Facultad de Ciencias, Universidad de Crdoba, Campus Universitario de Rabanales, Edicio C-3, Ctra. Madrid-Cdiz,
km 396, 14071 Crdoba, Spain

a r t i c l e

i n f o

Article history:
Received 6 March 2009
Received in revised form 1 June 2009
Accepted 6 June 2009
Available online 9 July 2009
Keywords:
Biodiesel manufacturing
Glycerol-containing waste
Anaerobic digestion
Kinetic constants
Biodegradability

a b s t r a c t
The anaerobic digestion of glycerol derived from biodiesel manufacturing, in which COD was found to be
1010 g/kg, was studied in batch laboratory-scale reactors at mesophilic temperature using granular and
non-granular sludge. Due to the high KOH concentration of this by-product, H3PO4 was added to recover
this alkaline catalyst as agricultural fertilizer (potassium phosphates). Although it would not be economically viable, a volume of glycerol was distilled and utilised as reference substrate. The anaerobic
revalorisation of glycerol using granular sludge achieved a biodegradability of around 100%, while the
methane yield coefcient was 0.306 m3 CH4/kg acidied glycerol. Anaerobic digestion could be a good
option for revalorising this available, impure and low priced by-product derived from the surplus of biodiesel companies. The organic loading rate studied was 0.210.38 g COD/g VSS d, although an inhibition
phenomenon was observed at the highest load.
2009 Elsevier Ltd. All rights reserved.

1. Introduction
Concern has recently risen over the use of fossil resources, their
cost, their sustained availability and their impact on global
warming and pollution (Hansen et al., 2005). Since biodiesel fuels
have various advantages such as a complement to petroleumbased fuel, renewable fuel, a favourable energy balance, less harmful emissions and non-toxic fuel, these have drawn much attention
recently (Ito et al., 2005).
During the biodiesel production process, oils/fats (triglycerides)
are mixed with methyl alcohol and alkaline catalysts to produce free
fatty-acid esters, with glycerol as a primary by-product (Chi et al.,
2007). The production of 100 kg of biodiesel yields approximately
10 kg of impure glycerol, with 5590% glycerol (Hazimah et al.,
2003). Glycerol is generated as a by-product not only when biodiesel fuels are produced chemically, but also when they are manufactured enzymatically (Du et al., 2003; Vicente et al., 2004) and during
the production of bioethanol (Yazdani and Gonzalez, 2007). The dramatic growth of the biodiesel industry has created a surplus of glycerol that has resulted in a dramatic 10-fold decrease in crude
glycerol prices over the last few years and generated environmental
concerns associated with contaminated glycerol disposal (Yazdani
and Gonzalez, 2007; Sandun et al., 2007).
At present, due to its properties, pure glycerol has more than
2000 different applications (Elvers et al., 1990). However, the pro* Corresponding author. Tel.: +34 957 218586; fax: +34 957 218625.
E-mail address: a92siloj@uco.es (J.. Siles Lpez).
0960-8524/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2009.06.017

duction of crude glycerol exceeds the present commercial demand

for puried glycerol. Furthermore, the purication of glycerol (with
a view to being sold) that is generated during biodiesel manufacturing is not a viable option for the biodiesel industry (Chi et al., 2007).
Although crude glycerol can be burnt, with the consequent energetic advantages, the setting up of bioreneries that co-produce
products of higher economic value along with biofuels has been proposed as a solution for the economic viability of this product. Several
strategies based on chemical and biological transformations are
being pursued to convert glycerol into more valuable products (Yazdani and Gonzalez, 2007). An example of some of these includes:
The conversion of glycerol into propylene glycol and acetone,
through thermo-chemical processes (Chiu et al., 2006; Dasari
et al., 2005).
The etherication of glycerol with either alcohols (e.g. methanol
or ethanol) or alkenes (e.g. isobutene) and the production of
oxygen-containing components, which could have suitable
properties for use in fuel or solvents (Karinen and Krause, 2006).
The microbial conversion (fermentation) of glycerol to 1,3-propanediol, which can be used as a basic ingredient of polyesters
(Barbirato et al., 1998; Ito et al., 2005).
Other products such as butanol (Biebl, 2001), propionic acid
(Bories et al., 2004), ethanol and formate (Jarvis et al., 1997),
succinic acid (Lee et al., 2001), dihydroxyacetone (Bories et al.,
1991; Claret et al., 1994), polyhydroxyalkanoates (Koller et al.,
2005), or hydrogen and ethanol (Ito et al., 2005) were also
obtained using glycerol as a carbon source.

5610

J.. Siles Lpez et al. / Bioresource Technology 100 (2009) 56095615

Nomenclature
Alk
alkalinity (mg CaCO3/L)
CODremoved removed chemical oxygen demand (mg/L)
CODsoluble soluble chemical oxygen demand (mg/L)
COD STO added chemical oxygen demand (mg/L)
CODtotal total chemical oxygen demand (mg/L)
G
cumulative methane volume (mL)
cumulative methane volume at innite time (mL)
Gm
experimental maximum methane volume (mL)
GT
specic methane production kinetic constant (L/g VSS h)
KG
apparent kinetic constant (1/h)
K 0G
MS
total mineral solids (mg/L)
MSS
mineral suspended solids (mg/L)
OLR
organic loading rate (g COD/g VSS d; kg COD/m3 d)

Anaerobic digestion is another possible way to revalorise abundant and low-priced glycerol streams. This process may be dened
as the biological conversion of organic material to a variety of end
products including biogas whose main constituents are methane
and carbon dioxide (Gujer and Zehnder, 1983; Olthof and
Oleszkiewick, 1982; Speece, 1983; Wheatley, 1990). The advantages of anaerobic digestion include low levels of biological sludge,
low nutrient requirements, high efciency and the production of
methane, which can be used as an energy source. The stoichiometry of the anaerobic digestion of glycerol can be summarised as follows (Christensen and McCarty, 1975; McCarty, 1975):

C3 H8 O3 aNH3 ! bCH4 cCO2 dC5 H7 NO2 eNH4 HCO3

The products of the reaction are methane, carbon dioxide, biomass and ammonic bicarbonate, where a, b, c, d and e are stoichiometric coefcients. In order to obtain the value of these
coefcients and determine the methane yield of the anaerobic
digestion of glycerol, a mass balance was made taking into consideration a biomass yield of anaerobic bacteria of 0.05 (w/w). The
values of a, b, c, d and e were found to be 0.663, 1.648, 0.526,
0.041 and 0.622 mol, respectively. In contrast, when an electron
balance was carried out assuming that every electron was used
in methane generation, b reached a value of 1.750. A simple calculation showed that the theoretical methane yield is 94.2%, which is
a useful compound due to its caloric power (Lower Caloric Power):
35,793 kJ/m3, equivalent to 9.96 kW h/m3. Revalorising glycerol is
of special interest as it gives the highest reduced carbon with the
cost advantage of anaerobic processes. The aim of this work was
to evaluate the performance and the stability of anaerobic digestion process of glycerol-containing waste derived from biodiesel
manufacturing. The study was carried out in six batch laboratory-scale reactors at mesophilic temperature (35 C).
2. Methods
2.1. Experimental set-up
The experimental set-up used for the anaerobic digestion of the
biodiesel-derived glycerol consisted of six 1-L Pyrex reactors with
four connections to allow for the loading of feedstock, the ventilation of the biogas, the injection of inert gas (nitrogen) to maintain
the anaerobic conditions and the removal of efuent. The content
of the reactors was magnetically stirred and temperature was
maintained by means of a thermostatic jacket containing water
at 37 C. The volume of methane produced during the process
was measured by using 1-L BoyleMariotte reservoirs connected
to each reactor. To remove the CO2 produced during the process,
tightly closed bubblers containing a NaOH solution (6 N) were con-

r0
specic rate of methane production (mL CH4/g VSS h)
Sremoved removed substrate (g COD)
added substrate (g COD)
STO
t
time (h)
TS
total solids (mg/L)
TSS
total suspended solids (mg/L)
VA
volatile acidity (mg acetic/L)
VS
total volatile solids (mg/L)
VSS
volatile suspended solids (mg/L)
X
biomass concentration (mg VSS/L)
Y CH4 =Sremoved methane yield coefcient (mL CH4/g COD removed)

nected between the two elements. The methane volume displaced

an equal measurable volume of water from the reservoir.
The reactors were inoculated with methanogenically-active
granular biomass obtained from a full-scale anaerobic reactor used
to treat brewery wastewater from the Heineken S.A. Factory (Jaen,
Spain) and non-granular sludge from a full-scale anaerobic reactor
used to treat urban wastewater in Jerez de la Frontera (Cadiz,
Spain). The granular sludge contained 37,500 mg VSS/L and
31,875 mg MSS/L, while the non-granular sludge contained
28,400 mg VSS/L and 20,330 MSS/L. The inocula were selected on
the basis of their high methanogenic activity (Field et al., 1988)
with values ranging from 0.87 to 0.99 g COD/g VSS d.
2.2. Glycerol
The raw material used as substrate was the glycerol-containing
waste discharged after the biodiesel manufacturing process at the
BIDA S.A. Factory in Fuentes de Andalucia (Seville, Spain). In general,
this waste contained glycerol, water, methanol, salts and fatty acids.
2.3. Raw material pre-treatment
The substrate was previously treated in two different ways: (a)
acidication with phosphoric acid and centrifugation in order to recover the catalyst used in the transesterication reaction (KOH) as
agricultural fertilizer (potassium phosphates). Additionally, methanol and water were removed by vacuum distillation. We call this
substrate acidied glycerol. (b) Acidication followed by distillation (rectication). After the same acidication process, a rectication at laboratory-scale (135140 C; 1.62.0  103 atm) was
carried out. Subsequently, the organic impurities in the distillate
were removed by liquid/liquid extraction with hexane, which was
eliminated by vacuum distillation. We call this substrate distilled
glycerol. Table 1 shows the characteristics and features of acidied,
distilled and commercially available pure glycerol (Elvers et al.,
1990). Due to its high COD, the acidied and distilled glycerol were
diluted using distilled water until reaching 81.6 and 85.7 g COD/L,
respectively, and neutralized by adding sodium hydroxide. Finally,
several nutrients and alkalinity (NaHCO3) were added to the distilled glycerol to provide the necessary nutrients for the appropriate
metabolism of the anaerobic microorganisms (DiStefano and
Ambulkar, 2006). Table 2 shows the composition of the nutrient
and trace element solutions added to the distilled glycerol.
2.4. Anaerobic digesters. Experimental procedure
The anaerobic reactors were initially loaded with 12 g VSS of
granular sludge as inoculum, and the anaerobic digestion of

5611

J.. Siles Lpez et al. / Bioresource Technology 100 (2009) 56095615

2.6. Software

Table 1
Composition and features of the acidied, distilled and pure glycerol.
Parameter

Acidied
glycerol

Distilled
glycerol

Pure
glycerol

Density at 20 C (g/mL)
Refraction index at 20 C
COD (g/kg)
Dynamic viscosity at 50 C (mPa s)
Colour

1.044
1.4440
1010
57
Brown

1.260
1.4728
1155
150
Colourless

1.261
1.4746
1217
152
Colourless

Sigma-Plot software (version 9.0) was used to create graphs,

perform the statistical analysis and t the experimental data presented in this work.
3. Results and discussion
3.1. Stability

Table 2
Composition of the nutrient and trace element solutions added to distilled glycerol.
Nutrient solution

Trace element solution

Compound

g/La

Compound

NH4CI
K2HPO43H2O
KH2PO4
MgCl26H2O
Resazurine
FeCl24H2O
NaS9H2O
NaHCO3

0.200
0.100
0.055
0.200
0.001
0.100
0.500
5.000

MnCl24H2O
0.100
CoCl26H2O
0.170
ZnCl2
0.100
CaCl2
0.200
H3BO4
0.019
NiCl26H2O
0.050
Na2MoO4H2O
0.100
Ad 10 mL of trace element solution per litre of diluted
glycerol

g/L

Concentration of each compound per litre of diluted glycerol.

The stability of the process is evaluated based on the evolution

of the pH, alkalinity, volatile acidity and volatile acidity/alkalinity
ratio (VA/Alk) during the anaerobic digestion process of the different substrates. Table 3 shows the mean value and standard deviation of the pH and volatile acidity/alkalinity ratio in the efuents of
the reactors with the different substrates and sludge types. The pH
was approximately constant across experiments, with a mean
value of 7.72 0.24 using acidied glycerol-granular sludge,
7.74 0.30 using acidied glycerol-non-granular sludge and
7.81 0.17 with distilled glycerol-granular sludge. These values remained within the optimal range for methanogenic bacteria
(Fannin, 1987; Wheatley, 1990). On the other hand, the volatile
acidity/alkalinity ratio was always found to be lower than 0.30
0.40, thus indicating that the process operated favourably without
the risk of acidication (Balaguer et al., 1992).
3.2. Methane yield coefcient and biodegradability of glycerol

acidied and distilled glycerol was studied. Another set of reactors

were then loaded with non-granular sludge and the anaerobic
digestion of acidied glycerol was studied. In all cases, the nutrient
and trace element solutions described by Fannin (1987) and Field
et al. (1988) were added when the sludge was loaded. Both
solutions are very important for activating bacterial growth and
metabolism at the beginning of the process.
In order to activate the biomass prior to the experiments, the
reactors were rst fed with a synthetic solution composed of glucose, sodium acetate and lactic acid (GAL solution) at concentrations of 50 g/L, 25 g/L and 20.8 mL/L, respectively. During this
initial period, the organic load added to the reactors was gradually
increased from 0.25 to 1.00 g COD over a 16-day period. After this
previous stage, biomass acclimatization was carried out. The reactors were fed with 1 g COD, in which the percentage of glycerol
used in the COD was increased from 25% to 100% after four loads.
During this acclimatization period, the volume of methane was
measured as a function of time. The maximum duration of each assay was 48 h; the time interval required for the complete biomethanization of each load. Once this preliminary acclimatization step
was nished, a series of batch experiments were carried out using
both sludge types in addition to the acidied and distilled glycerol
as substrates. During each set of experiments, the organic load
added to the reactors was gradually increased from 1.0 to 1.5
and 2.0 g COD with distilled glycerol and from 1.0 to 1.5, 2.0 and
3.0 g COD with acidied glycerol. In all cases, the volume of methane was measured as a function of time and samples were taken
and analysed before and after feeding. The duration of each experiment was equal to the time interval required for maximum gas
production and COD removal. Each glycerol solution load was carried out at least in duplicate and the results expressed as means.
2.5. Chemical analyses
The following parameters were determined in the efuents of
each load: pH, COD total, COD soluble, TS, MS, VS, TSS, MSS, VSS,
volatile acidity (VA) and alkalinity (Alk). All analyses were carried
out in accordance with the Standard Methods of the APHA (APHA,
1989).

Methane yield coefcient was determined from the experimental maximum methane volume produced (GT) and the nal and initial COD, which were known in all loads. By tting (GT, COD
removed) value pairs to a straight line (Fig. 1), the methane yield
coefcient coincides with the slope of the regression line and
was found to be 292 mL CH4/g COD removed (at 1 atm, 25 C) using
granular sludge-acidied glycerol, 288 mL CH4/g COD removed
using non-granular sludge-acidied glycerol and 356 mL CH4/g
COD removed using granular sludge-distilled glycerol. Similar values have been described in the literature. Gross and Lanting (1988),
for example, studied the anaerobic digestion of wastewater generated during the manufacturing of fuel ethanol from corn, which
contained a highly soluble mixture of weak organic acids such as
acetic, propionic, lactic and butyric acid. The organic strength of
the wastewater typically ranged from 3000 to 5000 mg/L COD soluble. The methane yield coefcient was found to be 330 mL CH4/g
COD removed.
According to Wheatley (1990), and considering the biomass
growth and cell maintenance null, 382 mL of methane are theoretically produced (at 1 atm, 25 C) per gram of COD removed. Experimentally, the effectiveness of the process in each case was: 76%
using granular sludge-acidied glycerol, 75% using non-granular
sludge-acidied glycerol and 93% with granular sludge-distilled
glycerol. When only taking into account these results, granular
sludge-distilled glycerol would be the best option. Studying the removed COD percentage is as important as evaluating the methane
production coefcient in order to determine waste biodegradability. The high biodegradability of glycerol can be demonstrated by

Table 3
pH and volatile acidity/alkalinity ratio values in the efuents of the reactors obtained
for the different substrates and sludge studied.
Substrate and sludge type

pH

VA (eq acetic)/Alk
(eq CaCO3)

Acidied glycerol and granular sludge

Acidied glycerol and non-granular sludge
Distilled glycerol and granular sludge

7.72 0.24
7.74 0.30
7.81 0.17

0.05 0.02
0.03 0.01
0.05 0.01

5612

J.. Siles Lpez et al. / Bioresource Technology 100 (2009) 56095615

plotting the amount of substrate removed against the substrate

added for each experiment. Fig. 2 shows these data in such a manner that the slope of the straight line denotes the percentage biodegradability of the waste, which was found to be around 100%
using granular sludge-acidied glycerol, 75% using non-granular
sludge-acidied glycerol and 85% using granular sludge-distilled
glycerol. These values remained constant throughout the loads.
The biodegradability obtained with glycerol was high and similar
to bioethanol wastewater, which achieved 86% biodegradability
(Gross and Lanting, 1988).
In all the cases, the reactors contained soluble and non-biodegradable COD before adding the substrate. In the same gures,
the initial non-biodegradable COD value was found to be 0.502 g
COD when granular sludge-acidied glycerol was used, 0.093 g

COD when non-granular sludge-acidied glycerol was added and

0.665 g COD for granular sludge-distilled glycerol.
Taking into account both the methane production percentage
and the waste biodegradability, the granular sludge-acidied glycerol would be the best choice. All the substrate concentration
would be removed, thus producing methane and biomass growth
and obtaining 100% biodegradability and 76% effectiveness in generating methane. This fact would simultaneously ensure cell maintenance and non-waste accumulation in the reactors.
3.3. Kinetics of methane production
Fig. 3 shows the variation in the volume of methane accumulated (G) as a function of time for the loads with the different substrates and sludge types. The variable concerned is the amount of

3.5

1000

Granular sludge-Acidified glycerol

Granular sludge-Acidified glycerol

3.0

S removed (g COD)

GT (mL CH4 )

800

600

400

2.5
2.0
1.5
1.0
x = 0.502 g COD

200

m = 1.01 g COD removed / g COD S TO

0.5

r = 0.9946

YCH4/S removed = 292 mL CH4 / g COD removed

3.5

1000

Non-granular sludge-Acidified glycerol

Non-granular sludge-Acidified glycerol

3.0

S removed (g COD)

GT (mL CH4 )

800

600

400

2.5
2.0
1.5
1.0
x = 0.093 g COD

200

0.5

m = 0.74 g COD removed / g COD S TO

2

r = 0.9542

YCH4/S removed = 288 mL CH4 / g COD removed

3.5

0
1000

Granular sludge-Distilled glycerol

Granular sludge-Distilled glycerol

3.0

S removed (g COD)

GT (mL CH4 )

800

600

400

2.5
2.0
1.5
1.0

200

x = 0.665 g COD

0.5

m = 0.85 g COD removed / g COD S TO

2

YCH4/S removed = 356 mL CH4 / g COD removed

r = 0.8925

0.0

0
0

S removed (g COD)
Fig. 1. Variation of the experimental maximum methane volume produced (GT) (at
1 atm, 25 C) with the COD consumed to obtain the methane yield coefcient of the
process using granular sludge-acidied glycerol, non-granular sludge-acidied
glycerol and granular sludge-distilled glycerol.

STO (g COD)
Fig. 2. Plot of the amount of substrate removed against the substrate added for all
the experiments to obtain the biodegradability percentage of the by-product using
granular sludge-acidied glycerol, non-granular sludge-acidied glycerol and
granular sludge-distilled glycerol.

5613

J.. Siles Lpez et al. / Bioresource Technology 100 (2009) 56095615

G Gm 1  expK 0G  t

1400
1200

1.0 g COD
1.5 g COD
2.0 g COD
3.0 g COD

Non-granular sludge-Acidified glycerol

1.0 g COD
1.5 g COD
2.0 g COD

Granular sludge-Distilled glycerol

800
600

200

where Gm is the maximum methane volume accumulated at an innite digestion time; G is zero at t = 0 and the rate of gas production
becomes zero at t equal to innite.
K 0G is an apparent kinetic constant for methane production (h1)
which included the biomass concentration:

0
1400
1200

2
G (mL CH4 )

1000
800
600
400
200
0
1400
1200
1000

G (mL CH4 )

where KG is the specic methane production kinetic constant (L/g

VSS h) and X is the biomass concentration (g VSS/L).
As it was previously mentioned, a larger amount of gas was produced as the load increased. Moreover, the slopes of the curves also
decreased with increasing time. This decrease in the slopes with
time can be explained by the gradual decrease in the concentration
of biodegradable substrate. Eq. (1) shows a good t with the experimental data. Thus, it is correct to apply the proposed kinetic model
to all the loads studied. The values of K 0G and Gm for each load were
calculated numerically from the experimental data obtained by
non-linear regression using SigmaPlot (version 9.0) (Borja et al.,
1995). Table 4 shows the Gm and K 0G values and their standard deviations obtained in the experiments. The standard deviations of Gm
and K 0G values were lower than 5% and 15%, respectively, for all the
loads studied. This suggests that the proposed model adequately
ts the experimental data.
Once the biomass concentration values were determined, the
values of the specic methane production kinetic constant, KG,
were calculated using Eq. (2). Table 4 also shows these KG values.
The specic constant, KG, decreased with substrate concentration
for all the experiments, showing an inhibition process. Specically,
in the case of acidied glycerol, the KG value decreased by 42%
when loading 1.03.0 g COD using granular sludge and by 58%
using non-granular sludge. Surprisingly, a 76% decrease was obtained when loading between 1.0 and 2.0 g COD of distilled glycerol. This marked decrease in the KG value when using distilled
glycerol could be explained by the accumulation in the reactors
of the alkalinity that was added as a nutrient. In contrast, the inhibition observed when using acidied glycerol could be the result of
two causes: the increasing load or the accumulation of phosphorous due to pre-treatment or even both effects simultaneously. In
order to avoid the phosphorous concentration effect, CaO was
added to the reactors during the experiments until a nal concen-

Granular sludge-Acidified glycerol

400

K 0G K G  X

1.0 g COD
2.0 g COD
3.0 g COD

1000

G (mL CH4 )

substrate (g COD) added to the reactors. The results show that a

larger volume of methane was produced when the substrate load
increased and the time to complete the biodegradable fraction removal was 40 h for both sludge types.
In order to characterize each set of experiments kinetically, and
thus facilitate comparisons, the methane production rst order kinetic model described by Borja et al. (1995) was used to t the
experimental data. According to this model, the volume of methane accumulated (G) (mL, at 1 atm, 25 C) at a given time t (h) ts
the following equation:

800
600
400
200
0
0

10

20

30

40

50

60

Time (h)
Fig. 3. Variation of the methane volume accumulated (G) as a function of time for
the loads of 1.0, 1.5, 2.0 and 3.0 g COD using granular sludge-acidied glycerol, nongranular sludge-acidied glycerol and granular sludge-distilled glycerol.

tration of 25 4 mg PO3
4 =L was obtained. The occurrence of a similar inhibitory phenomenon has been described in the literature
(Ito et al., 2005; Barbirato et al., 1998).

Table 4
Gm, K 0G and KG values obtained for all the loads studied using acidied glycerol-granular sludge, acidied glycerol-non-granular sludge and distilled glycerol-granular sludge.
Substrate and sludge type

Load (g COD)

Gm (mL CH4)

K 0G (h1)

R2

KG (Lg1 h1)

Acidied glycerol and granular sludge

1.0
2.0
3.0

323 5
592 5
857 7

0.2075 0.0117
0.1858 0.0053
0.1316 0.0036

0.9676
0.9909
0.9798

0.0384
0.0338
0.0246

Acidied glycerol and non-granular sludge

1.0
1.5
2.0
3.0

318 11
451 8
590 15
789 20

0.1655 0.0177
0.1175 0.0151
0.1216 0.0091
0.0749 0.0059

0.9527
0.9960
0.9698
0.9716

0.0203
0.0178
0.0190
0.0128

Distilled glycerol and granular sludge

1.0
1.5
2.0

375 6
578 5
766 15

0.2006 0.0105
0.1372 0.0048
0.0829 0.0046

0.9548
0.9784
0.9697

0.0131
0.0101
0.0068

5614

J.. Siles Lpez et al. / Bioresource Technology 100 (2009) 56095615

1000
m = 0.9942
Confidence intervals of 95%

800

GT experimental (mL CH4 )

ned as the volume of methane generated per gram of volatile suspended solid and per hour for each set of experiments. As can be
observed in Fig. 5, the r0 values were always higher when using
granular sludge and acidied glycerol at the same time. The lowest
values were reached when using distilled glycerol with a high alkalinity concentration.

600

3.4. Organic loading rate (OLR)

400

200

200

400

600

800

1000

Gm(mL CH4 )
Fig. 4. Comparison between the experimental maximum methane production (GT)
values and the theoretical values (Gm) predicted by Eq. (1).

Moreover, the theoretical values of maximum methane production (Gm) were calculated using Eq. (1) and plotted against their
corresponding experimental values (Fig. 4). The deviations obtained were lower than 5% in practically all cases, again suggesting
that the proposed model can be used to accurately predict the
behaviour of the process and that the kinetic parameters obtained
represented the microorganism behaviour affecting the anaerobic
digestion of this by-product.
In order to compare the different substrates and sludge types,
the specic methane production rate, r0 (mL CH4/g VSS h), was de-

The organic loading rate (OLR) for each substrate and sludge
type was determined using the experimental data. The OLR (kg
COD/m3 d) values were not excessively high for any set of experiments (0.922.00 kg COD/m3 d) when compared to other biodegradable substrates. For food, organic chemical, soft drink and
bakery wastes, which range from 5000 to 11,000 mg/L in raw
COD, removals of 6893% were obtained at organic loadings of
327 kg COD/m3 d (Jeris, 1983).
The OLR values, expressed as g COD/g VSS d, take into account
both the biomass concentration and the organic load, thus making
this an interesting variable for comparing the results. The highest
values were obtained using granular sludge-acidied glycerol:
0.210.38 g COD/g VSS d. On the contrary, when non-granular
sludge-acidied glycerol was used, the ORL reached a value ranging from 0.12 to 0.26 g COD/g VSS d and 0.075 to 0.080 g COD/g
VSSd when granular sludge-distilled glycerol was used. Fig. 6
shows the evolution of the specic methane production kinetic
constant, KG (L/g VSSh), with the organic loading rate (g COD/g
VSSd) for each substrate and sludge type studied. A strong decrease in the KG value was observed when the distilled glycerol
OLR was increased. Although the constant also decreased when
using acidied glycerol, the kinetic constant slowdown was much

14
Granular sludge-Acidified glycerol
Non-granular sludge-Acidified glycerol
Granular sludge-Distilled glycerol

12

r' (mL CH4 /g VSSh)

10
8

1.0 g COD

1.5 g COD

2.0 g COD

3.0 g COD

6
4
2
0
25

r' (mL CH4 /g VSSh)

20

15

10

0
0

10

20

30

Time (h)

40

50

60

70

10

20

30

40

50

60

70

Time (h)

Fig. 5. Variation of the specic methane production rate, r0 , as a function of time for all the loads using granular and non-granular sludge and acidied and distilled glycerol.

J.. Siles Lpez et al. / Bioresource Technology 100 (2009) 56095615

0.06
Granular sludge-Acidified glycerol
Non-granular sludge-Acidified glycerol
Granular sludge-Distilled glycerol

KG (L/g VSSh)

0.05
0.04

0.03

0.02

0.01

0.00

0.0

0.1

0.2

0.3

0.4

0.5

OLR (g COD/g VSSd)

Fig. 6. Variation of the specic kinetic constant KG with the Organic Loading Rate
(OLR), using granular and non-granular sludge and acidied and distilled glycerol.

lower. Additionally, this substrate has allowed the reactors to operate in a stable way with higher OLR. Finally, according to economic
criteria, acidied glycerol is the target substrate. Given that the
granular sludge KG shows values that are twice as high as the
non-granular sludge, the glycerol anaerobic treatment could be
carried out using granular sludge in addition to acidied glycerol.
However, one option to improve the OLR could be to mix glycerol with another organic waste such as wastewater generated
during the biodiesel manufacturing process and study its co-digestion. This research would form part of what is known as the Biorenery concept; a relatively new term referring to the conversion
of feedstock into a host of valuable chemicals and energy. Indeed,
this is considered an important strategy for achieving sustainable
development.
4. Conclusions
The results obtained through this research study reveal that
glycerol-containing waste derived from biodiesel production after
pre-treatment has a high level of anaerobic biodegradability and
that a substantial quantity of methane can be obtained this way.
The highest methane yield coefcient was obtained using granular
sludge-distilled glycerol, but a strong inhibition was observed
when the load was increased. The use of granular sludge and acidied glycerol was found to be the best option for revalorising glycerol anaerobically given that biodegradability was found to be
around 100% and the methane yield coefcient was 0.306 m3
CH4/kg acidied glycerol (at 1 atm, 25 C).
Acknowledgements
The authors are very grateful to the BIDA S.A. Factory (Fuentes
de Andaluca, Seville, Spain) and to the Spanish Ministry of Science
and Innovation for funding Jos ngel Siles Lpez through Project
CTM2005-01293 and Grant BES2006-14074. This study was co-nanced by the European Social Fund. We also wish to express
our gratitude to the laboratory technician Inmaculada Bellido
Padillo for her help.
References
APHA (American Public Health Association), 1989. Standard Methods for the
Examination of Water and Wastewater, 17th ed. Washington, DC, USA.
Balaguer, M.D., Vicent, M.T., Paris, J.M., 1992. Anaerobic uidized bed reactor with
sepiolite as support for anaerobic treatment of vinasses. Biotechnol. Lett. 14,
433438.

5615

Barbirato, F., Himmi, E.H., Conte, T., Bories, A., 1998. 1,3-Propanediol production by
fermentation: an interesting way to valorise glycerine from the ester and
ethanol industries. Ind. Crops Prod. 7, 281289.
Biebl, H., 2001. Fermentation of glycerol by Clostridium pasteurianum batch and
continuous culture studies. J. Ind. Microbiol. Biotechnol. 27, 1826.
Bories, A., Claret, C., Soucaille, P., 1991. Kinetic study and optimisation of the
production of dihydroxyacetone from glycerol using Gluconobacter oxydans.
Process Biochem. 26, 243248.
Bories, A., Himmi, E., Jauregui, J.J.A., Pelayo-Ortiz, C., Gonzales, V.A., 2004. Glycerol
fermentation with Propionibacteria and optimisation of the production of
propionic acid. Sci. Aliments. 24, 121135.
Borja, R., Martn, A., Alonso, V., Garca, C.J., Banks, C.J., 1995. Inuence of different
pretreatments on the kinetics of anaerobic digestion of olive mill wastewater.
Water Res. 29, 489495.
Chi, Z., Pyle, D., Wen, Z., Frear, C., Chen, S., 2007. A laboratory study of producing
docosahexaenoic acid from biodiesel-waste glycerol by microalgal
fermentation. Process Biochem. 42, 15371545.
Chiu, C.W., Dasari, M.A., Sutterlin, W.R., Suppes, G.J., 2006. Removal of residual
catalyst from simulated biodiesels crude glycerol for glycerol hydrogenolysis to
propylene glycol. Ind. Eng. Chem. Res. 45, 791795.
Claret, C., Salmon, J.M., Romieu, C., Bories, A., 1994. Physicology of Gluconobacter
oxydans during dihydroxyacetone production from glycerol. Appl. Microbiol.
Biotechnol. 41, 359365.
Christensen, D.R., McCarty, P.L., 1975. Multi-process biological treatment model. J.
Water Pollut. Contr. Fedn. 47, 26522664.
Dasari, M.A., Kiatsimkul, P.P., Sutterlin, W.R., Suppes, G.J., 2005. Low-pressure
hydrogenolysis of glycerol to propylene glycol. Appl. Catal. A-Gen. 281, 225
231.
DiStefano, T.D., Ambulkar, A., 2006. Methane production and solids destruction in
an anaerobic solid waste reactor due to post-reactor caustic and heat treatment.
Water Sci. Technol. 53 (8), 3341.
Du, W., Xu, Y., Liu, D., 2003. Lipase-catalysed transesterication of soya bean oil for
biodiesel production during continuous batch operation. Biotecnol. Appl.
Biochem. 38, 103106.
Elvers, B., Hawkins, S., Weinheim, G.S., 1990. Ullmanns Encyclopaedia of Industrial
Chemistry, fth ed. VCH, NY, Basel (Switzerland), Cambridge, New York.
Fannin, K.F., 1987. Start-up, operation, stability and control. In: Chynoweth, D.P.,
Isaacson, R. (Eds.), Anaerobic Digestion of Biomass. Elsevier, London, pp. 171
196.
Field, J., Sierra, R., Lettinga, G., 1988. Ensayos anaerobios. In: Fdz-Polanco, F., Garca,
P.A., Hernando, S. (Eds.), 4 Seminario de Depuracin Anaerobia de Aguas
Residuales. Valladolid, Spain, Secretariado de Publicaciones, Universidad de
Valladolid, pp. 5282.
Gross, L.R., Lanting, J., 1988. Anaerobic wastewater treatment of a fuel ethanol
facility. In: Torpy, M.F. (Ed.), Anaerobic Treatment of Industrial Wastewaters.
Pollution Technology Review, vol. 152 . Noyes Data Corporation, Park Ridge,
New Jersey, USA, pp. 2334.
Gujer, W., Zehnder, A.J., 1983. Conversion process in anaerobic digestion. Water Sci.
Technol. 15, 123167.
Hansen, A.C., Zhang, Q., Lyne, P.W.L., 2005. Ethanol-diesel fuel blends a review.
Bioresour. Technol. 96, 277285.
Hazimah, A.H., Ooi, T.L., Salmiah, A., 2003. Recovery of glycerol and diglycerol from
glycerol pitch. J. Oil Palm. Res. 15, 15.
Ito, T., Nakashimada, Y., Senba, K., Matsui, T., Nishio, N., 2005. Hydrogen and Ethanol
production from glycerol-containing wastes discharged after biodiesel
manufacturing process. J. Biosci. Bioeng. 100 (3), 260265.
Jarvis, G.N., Moore, E.R.B., Thiele, J.H., 1997. Formate and ethanol are the major
products of glycerol fermentation produced by a Klebsiella planticola strain
isolated from red deer. J. Appl. Microbiol. 83, 166174.
Jeris, J.S., 1983. Industrial wastewater treatment using anaerobic uidized bed
reactors. Water Sci. Technol. 15, 169176.
Karinen, R.S., Krause, A.O.I., 2006. New biocomponents from glycerol. Appl. Catal. AGen. 306, 128133.
Koller, M., Bona, R., Braunegg, G., Hermann, C., Horvat, P., Kroutil, M., Martinez, J.,
Neto, J., Pereira, L., Varila, P., 2005. Production of polyhydroxyalkanoates from
agricultural waste and surplus materials. Biomacromol. 6 (2), 561565.
Lee, P.C., Lee, W.G., Lee, S.Y., Chang, H.N., 2001. Succinic acid production with
reduced by-product formation in the fermentation of Anaerobiospirillum
succiniciproducens using glycerol as a carbon source. Biotechnol. Bioeng. 72,
4148.
McCarty, P.L., 1975. Stoichiometry of biological reactions. Prog. Water Technol. 7,
157172.
Olthof, M., Oleszkiewick, J., 1982. Anaerobic treatment of industrial wastewater.
Chem. Eng. 15, 13211326.
Sandun, F., Sushil, A., Kiran, K., Ranjitha, B., 2007. Glycerol based automotive fuels
from future bioreneries. Fuel., doi: 10.1016/j.fuel.2007.03.030.
Speece, R.E., 1983. Anaerobic biotechnology for industrial wastewater treatment.
Environ. Sci. Technol. 17, 416427.
Vicente, G., Martinez, M., Aracil, J., 2004. Integrated biodiesel production: a
comparison of different homogeneous catalyst systems. Bioresour. Technol.
92, 297305.
Wheatley, A., 1990. Anaerobic Digestion: A Waste Treatment Technology. Elsevier,
London.
Yazdani, S.S., Gonzalez, R., 2007. Anaerobic fermentation of glycerol: a path to
economic viability for the biofuels industry. Curr. Opin. Biotech. 18, 213219.