Mathematical Modelling of the Anaerobic Digestion Including the Syntrophic

Acetate Oxidation
Ivan Simeonov*, Dimitar Karakashev#

*Corresponding author
Department of Mathematical Modelling and Computer Sciences
Institute of Microbiology, Bulgarian Academy of Sciences,
Acad.G. Bonchev St., Block 26, Sofia 1113, Bulgaria
E-mail: issim@microbio.bas.bg
#New address: Department of Environmental Engineering, DTU Environment,
Technical University of Denmark, 2800 Lyngby, Denmark

Abstract: Anaerobic digestion is an effective biotechnological process for treatment of different

agricultural, municipal and industrial wastes. However, it is a very unstable process in regard to the
biogas reactors operation. This is due to the complicated interactions between different microbial species
as well as of the complex transformations of the organic matter affected by a variety of environmental
factors. Anaerobic digestion process can be divided in four main stages: hydrolysis, acidogenesis,
acetogenesis and methanogenesis. The syntrophic acetate oxidation is a two-stage process mediating by
two different phylogenetical microbial groups, living in syntrophic consortia: acetate oxidizing
Eubacteria which convert acetate to carbon dioxide and hydrogen, and hydrogenotrophic methanogenic
Archae which use carbon dioxide and hydrogen for production of methane. Use of mathematical models is
a powerful tool for investigations and optimisation of the AD. In this paper a new mathematical model of the
AD of activated sludge from municipal wastewater treatment plants including hydrolysis and the syntrophic
acetate oxidation process has been developed and verified experimentally and by computer simulations using
Simulink. Analytical computer studies of the model have been performed using Symbolic Toolbox of Matlab.
Keywords: anaerobic digestion, CSTR, mathematical model, parameter estimation, simulation, steady-
states analysis.

List of symbols main component is methane (energetical aspect). AD process

-1 can be divided in four main stages (Gerardi, 2003):
D – dilution rate, (day )
- hydrolysis of undissolved high-molecular weight
Xi (i = 1, 2, . . . ,6) - concentration of microorganisms for
compounds (proteins, sugars, fats) to soluble low-molecular
population “i”, (g/dm3)
weight compounds (monosugars, aminoacids, long - chain
Si (i = 0, 1, . . . ,6) – concentration of substrate “i”, (g/dm3)
fatty acids, glycerol);
Qsp – specific biogas yield, (dm3 biogas.dm-3 medium-.day-1)
- acidogenesis – fermentation of low-molecular weight
t – time, (day)
compounds from previous stage to VFA (propionate, butirate,
d/dt – first time derivative
acetate), hydrogen and carbon dioxide;
Abbreviations - acetogenesis from VFA to acetate, hydrogen and
carbon dioxide;
AD - anaerobic digestion - methanogenesis mediating by aceticlastic
VFA - volatile fatty acids
methanogens (convert acetate to methane and carbon dioxide)
CSTR – continuously stirred tank reactor and hydrogenotrophic methanogens (produce methane from
COD - chemical oxygen demand hydrogen and carbon dioxide).
TS – total solids
VS - volatile solids The key metabolites for methanogenesis – hydrogen, carbon
dioxide and acetate could be utilized as carbon sources not
1. INTRODUCTION only from methanogenic Archae but also from other
microorganisms. For example, hydrogen and carbon dioxide
Anaerobic digestion (AD) is an effective biotechnological are used from homoacetogenic bacteria for acetate synthesis;
process for treatment of different agricultural, municipal and alternatively the acetate can be transformed to methane via
industrial wastes (Ahring, 2003; Deublein and Steinhauser, syntrophic acetate oxidation (Batstone et al., 2002). The
2008). It combines environmental depollution (ecological syntrophic acetate oxidation is a two-stage process mediating
aspect) with production of renewable energy – biogas, which by two different phylogenetical microbial groups, living in

1
syntrophic consortia: acetate oxidizing Eubacteria which
convert acetate to carbon dioxide and hydrogen, and
hydrogenotrophic methanogenic Archae which use carbon
dioxide and hydrogen for production of methane.
AD is a very unstable process in regard to the biogas reactors
operation. This is due to the complicated interactions between
different microbial species as well as of the complex
transformations of the organic matter affected by a variety of
environmental factors (Ahring, 2003). In this context use of
mathematical models is a powerful tool for investigations and
optimisation of the AD (Batstone et al., 2002; Angelidaki et
al., 1993; Dochain and Vanrolleghem, 2001; Lyberatos and
Skiadas, 1999; Simeonov et al., 1996; Simeonov, 2010).
The aim of this paper is to develop and verify experimentally a
new mathematical model of the anaerobic digestion of activated
sludge from the municipal wastewater treatment plants including
hydrolysis and the syntrophic acetate oxidation process.
2. MATERIALS AND METHODS
Laboratory equipment. A laboratory scale anaerobic digester
with working volume of 2 dm3 operating in CSTR mode and
equipped with system for automatic maintenance of constant
stirring and mesophilic temperature (34 ± 0.50C) has been
used. The reactor has been fed once daily with 50 mL of
substrate. During the experiments samples for analysis of
glucose, VFA and ammonia have been taken out.
0
Materials. Sterilized by autoclaving (115 C, 20 min) waste
activated sludge from Sofia municipal wastewater treatment
plant has been used for daily feeding of the reactor. The
chemical composition of this waste is given in Table 1. For
start-up the reactor has been initially inoculated from another
bioreactor in operation.
Table 1. Chemical composition of the waste activated
sludge.
Total Glucose Soluble Total Acetate Ammonia
Glucose. The concentration of glucose has been determined by
GOD-PAP method (Stefanie et al,. 1998) in the supernatant after
centrifugation of the samples at conditions mentioned above.
Total solids (TS) and volatile solids (VS). TS and VS have
been measured according to (Gerardi, 2003).
Ammonia. Concentration of ammonia nitrogen has been
determined according to the method of Nesler (Gerardi, 2003).
Biogas yield. Biogas yield has been measured in gasholder by
water replacement method.
Volatile fatty acids (acetate, propionate). VFA
concentrations have been measured by HPLC using a
chromatograph Schimadzu LC-4A equipped with ZORBAX
OD SC18 (25x4.6 cm) column at following conditions:
eluent 0.1 % H3PO4, eluation rate 0.6 mL/min, detection by
flame ionization detector at λ = 210 nm .
Computer studies. Computer investigations (parameters
optimisation, simulations and determinations of static
characteristics) of the model have been performed using
Simulink, Optimisation and Symbolic Toolboxes of Matlab.
Chemicals. All chemicals used have been analytical/HPLC
grade and have been obtained by commercial sources.
3. RESULTS AND DISCUSSION
3.1. Model structure.
The model of the AD in CSTR has the following structure:
dS0 So.X1 6
= −β. + D.Ye .Soi +λ(∑bi Xi ) −D.S0 (1)
dt S2 +S3 +S4 + Ki0 i=1
dX1
= (μ1 − b1) X1 − D.X1 (2)
dt

solids 3
(g/dm ) protein protein (g/dm )3
(gNH4+/dm3) dS1 S o .X 1
= −Yglu / X 1 .μ1 . X 1 + β .
(%) (g/dm3) (g/dm3) dt S 2 + S 3 + S 4 + K io (3)

2.3 0.03 0.2 1.2 0.09 0.5 − D.S1 + S1i + S1control

dX2
Analytical methods. = (μ2 − b2 ) X 2 − D.X 2 (4)
All analyses have been carried out in accordance with the
dt
Standard Methods of the American Public Health Association
(APHA, 1985). dS 2
= Y prop / X 1 .μ1 . X 1 − Y prop / X 2 .μ 2 . X 2
Protein. The concentration of the protein (soluble and total) has dt (5)
been performed according to the method from (Bradford, 1976). − D.S 2 + S 2i
Soluble protein has been measured in the supernatant after
centrifugation of the probes for 5 min at 5000 rpm. Total protein
dX3
has been measured in the supernatant, collected after hydrolysis = (μ3 − b3 ) X 3 − D.X 3 (6)
of the sample (0.5 g sample was dissolved in 1 mL 0.2 M NaOH dt
in glass tube placed in 100 0C water bath) for 10 min. and
centrifugation at 5000 rpm for 5 min.

2
dS 3 μ max5 .S 4
= Ybut / X 1 .μ1 . X 1 − Ybut / X 3 .μ 3 . X 3 μ5 = − specific growth rate of acetate oxidizers X5
dt (7) KS5 + S4
− D.S 3 + S 3i (day-1);
μ max 6 .S 5 .S 6
dX4 μ6 = − specific growth rate of
= (μ4 − b4 ) X 4 − D.X 4 (8) ( K S 6 + S 5 ).(K S 6 + S 6 )
dt hydrogenotrophic methanogens X6 (day-1); Yglu/X1(g/g biomass)
= 12.9, Yacet/X1 (g/g biomass) = 3.54, Yprop/X1(g/g biomass) =
dS4 2.94, Yprop/X2(g/g biomass) = 10.2, Ybut/X1(g/g biomass) =
= Yacet/ X1.μ1.X1 + Yacet/ X 2.μ2.X2 + Yacet/ X 3.μ3.X3
dt (9) 3.08, Ybut/X3(g/g biomass) = 11.9, Yacet/X2(g/g biomass) = 8,
Yacet/X3(g/g biomass) =1.54, Yacet/X4(g/g biomass) = 24.14,
− Yacet/ X 4.μ4.X4 − Yacet/ X 5.μ5.X5 − D.S4 + S4i + S4control Yacet/X5 (g/g biomass) = 1.2, YH2/X1(g/g biomass) = 0.125,
YH2/X2(g/g biomass) = 1.1, YH2/X3(g/g biomass) = 0.62,
dX5 YH2/X5(g H2/g biomass Х5) = 0.11, YH2/X6(g H2/g biomass Х6)
= (μ5 − b5 ) X 5 − D.X 5 (10) = 4, YСО2/X6(g /g biomass) = 2.3, YСО2/X1(g/g biomass) = 2.4,
dt YСО2/X2(g/g biomass) = 7.6; YСО2/X4g/g biomass) = 16.7,
YСО2/X5(g/g biomass) = 1.2, YСН4/X6(dm3 CH4/g biomass Х6) =
dS5 0.14, YСН4/X4 (dm3 CH4/g biomass Х4) =8.51 – yield
= YH 2 / X 1 .μ1 . X 1 + YH 2 / X 2 .μ 2 . X 2 coefficients; β (day-1) = 0.31 – hydrolytic rate; Ki,o(g/ dm3) =
dt 0.23 – inhibition constant, reflecting the decrease of hydrolytic
+ YH 2 / X 3 .μ 3 . X 3 + YH 2 / X 5 .μ 5 . X 5 (11) rate due to VFA (propionate, butyrate and acetate)
accumulation; Ki,NH4+(g/dm3)=0.26 – inhibition constant
− YH 2 / X 6 .μ 6 . X 6 − K H 2 .S 5 − D.S 5 reflecting the decrease of aceticlastic methanogenesis rate due to
ammonia accumulation; Ki,acet/prop (g/dm3)=0.96 – product
inhibition constant, reflecting the decrease of propionate
dX 6
= (μ6 − b6 ) X 6 − D.X 6 (12)
degradation rate due to acetate accumulation; Ki,acet/but
dt (g/dm3)=0.72 – product inhibition constant, reflecting the decrease
of butyrate degradation rate due to acetate accumulation; D (day-1)
– dilution rate; Ye=0.55 – coefficient of decomposition, counting
dS 6 what part of insoluble organic compounds are transformed to
= YCO 2 / X 1 .μ 1 . X 1 + YCO 2 / X 2 .μ 2 . X 2 soluble compounds; Soi=30.6 g/dm3 – concentration of insoluble
dt organic compounds, measured as TS; So(g/dm3) – concentration
+ YCO 2 / X 4 .μ 4 . X 4 + YCO 2 / X 5 .μ 5 . X 5 (13) of soluble organic compounds, measured as VS; SNH4+(g/dm3) –
concentration of ammonia; S1(g/dm3) – concentration of glucose;
− YCO 2 / X 6 .μ 6 . X 6 − D.S 6 − K CO 2 .S 6 S2(g/dm3) – concentration of propionate; S3(g/dm3) –
concentration of butyrate; S4(g/dm3) – concentration of acetate;
Q = YCH 4 / X 4 .μ 4 . X 4 + YCH 4 / X 6 .μ 6 . X 6 S5(g/dm3) – concentration of hydrogen in liquid phase; S6(g/dm3)
(14), – concentration of carbon dioxide in liquid phase; Si1, Si2, Si3 and
+ k CO 2 .S 6 Si4, are the concentrations of the corresponding substrates in the
influent; S1control and S4control - concentrations of the corresponding
stimulating substances (glucose or acetate containing wastes)
μ max 1 .S1 (Simeonov et al., 2010); X1(g/dm3) – concentration of glucose-
where: μ1 = − specific growth rate of glucose-
K S 1 + S1 fermenting acidogens; X2(g/dm3) – concentration of propionate-
fermenting acidogens X1 (day-1); degrading acetogens; X3(g/dm3) – concentration of butyrate-
degrading acetogens; X4(g/dm3) – concentration of acetoclastic
μ max2 methanogens; X5(g/dm3) – concentration of acetate oxidizers;
μ2 = − specific growth X6(g/dm3) – concentration of hydrogenotrophic methanogens;
(1 + K s 2 / S 2 ).(1 + S 4 / K i,.acet / prop )
Q(dm3.day-1) – biogas yield; Km = 1.12 – coefficient in the
rate of propionate - degrading acetogens X2 (day-1); Contois growth rate model for μ4, reflecting the decrease of
aceticlastic methanogenesis rate due to biomass accumulation;
μ max3 KS1(g/dm3) = 0.5 – saturation constant for glucose-fermenting
μ3 = − specific growth rate
(1 + K s3 / S 3 ).(1 + S 4 / K i,.acet / but ) acidogens; KS2(g/dm3) = 0.259 – saturation constant for
of butyrate - degrading acetogens X3 (day-1); propionate-degrading acetogens; KS3(g/dm3) = 0.176 – saturation
constant for butyrate-degrading acetogens; KS5(g/dm3) = 0.21–
μmax4 .Ki, NH4+ .S 4 saturation constant for acetate oxidizers; KS6(g/dm3) = 0.01 –
μ4 = − specific growth rate saturation constant for hydrogenotrophic methanogens; μmax1(day-
(Km .X 4 + S 4 ).(Ki, NH4+ + S NH4+ ) 1
) = 5 – maximum specific growth rate of glucose-fermenting
of aceticlastic methanogens X4 (day-1); acidogens at 340C; μmax2(day-1) = 0.54 – maximum specific growth
rate of propionate-degrading acetogens at 340C; μmax3(day-1) = 0.68

3
– maximum specific growth rate of butyrate-degrading acetogens
at 340C; μmax4(day-1) = 0.6 – maximum specific growth rate of
aceticlastic methanogens at 340C; μmax5(day-1) = 0.35 – maximum
specific growth rate of acetate oxidizers at 40C; μmax6(day-1) = 1.4
– maximum specific growth rate of hydrogenotrophic
methanogens at 340C; bi (i=1,…6) – mortality rates for each of the
sixth bacterial populations (it was supposed that bi = 0.05μmaxi);
KCO2(day-1) = 0.096 – mass-transfer coefficient of СО2 from liquid
to gas phase; KH2(day-1) = 9.6 – mass-transfer coefficient of H2
from liquid to gas phase; kCO2 (dm2.g/day) = 0.034 – mass-transfer
coefficient of СО2 from liquid to gas phase.
We assume that a part of the dead cells are transformed into soluble
organics with recycling conversion factor λ (λ >0 and λ<bi ).
The model from (Angelidaki et al, 1993) has been used as a
basis. Four new equations: (10), (11), (12) and (13), describing
acetate oxidation and hydrogenotrophic methanogenesis, were
developed. Additionally, the part - Yacet/X5.μ5.X5, describing acetate
concentration decreasing due to the acetate consumption from the
acetate oxidizers, has been incorporated in the equation (9). The
analytical expression of the growth rate μ4 has been modified
taking into account the decrease of the growth rate due to cell
biomass accumulation.
3.2. Model calibration and verification.
Parameters identification (estimation of the coefficients),
including theoretical and practical identifiability, of such
complex non-linear model is a very hard problem (Dochain,
and Vanrolleghem, 2001; Simeonov, 2010). That is why in
this case a pragmatical engineering approach has been
adopted - the initial values of all coefficients have been taken
from (Angelidaki et al, 1993; Dochain, and Vanrolleghem,
2001) excepting the coefficient for the acetate oxidation
stage, which have been taken from (Strous et al, 1999). The
last part of parameters has been determined by experiments
with pure cultures of psychrophilic acetate oxidizers isolated
from granular sludge. In our reactor experiments however,
we have used mesophilic microbial consortium mediating the
anaerobic digestion of non-granular sludge and a new model
calibration using our experimental data have been obligatory.
This calibration consists of finding out a new set of
coefficients giving good coincidence between computer
model simulations and the laboratory experimental results.
For this purpose the changes in the concentrations of
ammonia, glucose, propionate, acetate and biogas yield after
pulls addition of ammonia (in the form of NH4OH with
amplitude corresponding to 0.5 g/dm3) to the feeding
substrate have been studied. Background ammonia
concentrations in the reactor before the pulls experiment were
around 2 g/dm3. Ammonia ions have been used because the
aceticlastic methanogens, producing up to 2/3 of the methane
in the nature (Schmidt et al., 2002) are most sensitive to their
action compared to other trophic microbial groups mediating
the anaerobic digestion. Consequently this inhibition of the
aceticlastic methanogenesis will lead to very strong response
on the system in regard to the biogas yield and other
connected parameters - glucose, VFA. Experimental data
obtained have been compared with the computer simulations.
The initial analyses have shown considerable differences
between the laboratory and simulation results in a
4
quantitative aspect but not in qualitative. A great number of
optimisation procedures and computer simulations using
expert knowledge and the methodology for parameters
estimation from (Simeonov et al., 1996) have been performed
in order to receive good coincidence between experimental
and simulation data. The following new values of the
coefficients have been obtained: μmax1=0.7; μmax4=0.45;
μmax5=0.037; KS1=4.8; KS2=0.93; KS5=0.4; Km=1.3; kCO2=0.9;
Ki,NH4+=0.5; Yacet/X1=20; Yacet/X4=16; YСН4/X4=4; YСН4/X6 =0.8.
The simulation and the experimental results are shown on
Fig.1 (Karakashev, 2004). After the calibration, the model
verification has been performed using pulls addition of
ammonia (Fig.2) with bigger amplitude equal of 0.75 g/dm3.
Fig 1 and Fig. 2 show that the calibrated model reflects very
good the process kinetics as some differences in quantitative
attitude between model predicted values and the experimental
data are observed only for VFA – acetate and propionate. It
could be explained with the fact that the kinetics of VFA (key
metabolites for many microbial populations) production and
uptake is more complicated that described in the present
model. The acetate increase and the biogas yield decrease
after the ammonia pulls are due to the NH4+ inhibition of the
aceticlastic methanogens. In that case the methanogenesis
most probably became via syntrophic cooperation between
acetate oxidizers and hydrogenotrophic methanogenss. On
the other hand the decrease in the glucose concentration (Fig
1b, Fig 2b) observed after the pulls could be explained with
inhibition of the hydrolysis of polysaccharides to glucose in
result of VFA accumulation. That suggestion is also
supported by analysis of the expression
S o .X 1
β.
S 2 + S 3 + S 4 + K io
in equation (3) of the model. Analogically, the propionate
concentration increase (Fig. 1c, Fig 2c) is a result of product
inhibition (from acetate) of propionate degrading acetogens,
which is also supported by analysis of the analytical
expression of the specific growth rate μ2. It has been also
observed that after ammonia pulses the specific biogas yield
did not return to the initial values of Qsp=0.175 (dm3 gas.dm-3
medium-.day-1) – Fig.1d (respectively 0.16 (dm3 gas.dm-3
medium.day-1) – Fig 2d), and is returned to lower values –
0.15 (dm3 gas.dm-3 medium-.day-1) (resp. 0.12 (dm3 gas.dm-3
medium-.day-1)). A possible explanation is that the acetate
accumulation leads to increase production of CO2 from the
acetate oxidizers. The large amounts of CO2 produced results
to its bigger solubility in the liquid phase leading to little CO2
quantities in the gas phase and total decreasing in the biogas
yields. Finally, big increase in the ammonia concentrations
(up to 5.93 (g/dm3) at pulls with amplitude 0.5 (g/dm3), Fig
1a, and up to 13.39 (g/dm3) at pulls with amplitude 0.75
(g/dm3), Fig 2a) after the start of the experiment has been
registered. This could be explained with influence of the
ammonia ions (substrate inhibition) on their own metabolism
(uptake) at anaerobic conditions. The main process
responsible for the ammonia uptake in anaerobic conditions
is the ammonia oxidation mediating from two phylogenetic
groups hemolitotrophic microorganisms – Annamox bacteria,
conducting nitrite – dependent oxidation of ammonia to N2
and nitrate, and Nitrosomonas bacteria mediating NO2 (gas) – increase of ammonia concentration from 2 g/dm3 to 2.5 g/dm3
dependent ammonia oxidation to N2 and NO (Van de Graff et (resp. 2.75 g/dm3) could be result to substrate inhibition of
al., 1995; Shmidth et al., 2002). It was proved that the the anaerobic ammonia oxidizers manifested by big increase
ammonia oxidation is inhibited at ammonia concentration of ammonia content after the pulls additions.
above 1 (g/dm3) (Stefanie et al., 1998). For this reason the

(a) (a)

experimen tal

Ammonia (g/dm3)
14
14 experime ntal
12 data 12 data
Ammonia

10
(g/dm3)

10
8
8
6
4
sim ula tion 6

2
data 4
sim ulation
2
0
0
data
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 2 2 4 26 28 30 32

Days
Days

(b)
(b)
Glucose (g/dm3)

0 ,25
0,25

Glucose (g/dm3)
0,2

0 ,15 0,2
0,1
0,15
0 ,05

0 0,1
0 2 4 6 8 10 12 14 16 1 8 20 2 2 24 2 6 28 3 0 32
0,05

Days 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

(c) Days

0, 16
(c)
0, 14
Propionate

0, 12
(g/dm3)

0 ,1 0,15
Propionate

0, 08
(g/dm3)

0, 06 0,1
0, 04
0, 02 0,05
0
0 2 4 6 8 10 12 14 16 18 2 0 2 2 2 4 2 6 2 8 3 0 3 2 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
Days
Days
(d)

1 ,6
(d)
Acetate (g/dm3)

1 ,4
1 ,2
Acetate (g/dm3)

1 2
0 ,8
1,5
0 ,6
0 ,4 1
0 ,2
0 0,5

0 2 4 6 8 1 0 12 1 4 1 6 18 2 0 22 24 2 6 28 30 3 2 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
Days
Days

(e) (e)
0 ,18
0,2
Biogas yield (dm3

0 ,16
gas.L medium .
Biogas yield (L
-1
medium.day-1)

0 ,14
0,15
gas.dm-3

0 ,12
day )
-1

0,1
0,1
0 ,08
0 ,06
0,05
0 ,04
0 ,02
0
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
0 2 4 6 8 10 12 1 4 16 18 20 2 2 2 4 26 28 30 3 2
Days Days

Fig. 1. Simulation and experimental data with pulls addition Fig. 2. Simulation and experimental data with pulls addition
of ammonia with amplitude of 0.5 g/dm3 of ammonia with amplitude of 0.75 g/dm3

5
3.3. Static characteristics of the model.
Study of the input-output static characteristics of the model is
important in regards to the determination of optimal working
points at different optimality criteria. These characteristics
can be found by nullifying the right parts of the model
differential equations.
Analytical study of the model have been performed using
Symbolic Toolbox of Matlab and the input-output static
characteristics Q=Q(D) and COD = COD(D) have been
obtained. It is assumed COD to be equal to the sum S0 + S1 +
S2 + S3 + S4+ S5 + S6 (Dochain and Vanrolleghem, 2001).
They are shown on Fig. 3. It is evident that maximum of the
function Q=Q(D) exists, which is in accordance with
previously known results for simplest models (Simeonov,
1999). The following important parameters have been
calculated:
- Dsup=0.147 (day-1), above which the microorganisms will be
washed out. Consequently the bounds of changes of D are 0 < D <
0.147 (day-1);
- Dmax=0.1072 (day-1) at which Q obtain maximal value
Qmax=0.195 (dm3 gas.dm-3 medium.day-1).
It could be noticed that the increase in the biogas yield with
increase of D was combined with increase of COD,
respectively low extent of biodegradation.
4. CONCLUSION
A new mathematical model of the anaerobic digestion of activated
sludge from municipal wastewater treatment plants, including
hydrolysis and the syntrophic acetate oxidation process, has
been presented in this work. This model could be used for study,
monitoring and optimisation of the AD, however it is a little bit
complex for control algorithms design.
For the model calibration optimisation procedures, computer
simulations and expert knowledge have been used and finally a
new set of values for some model coefficients has been obtained.
Q (dm3 biogas.dm-3 medium.day-1), COD (2.10-2 gOdm-3)
0.15
Q Simeonov I., Momchev, V. Grancharov, D. (1996). Water
0.1
0.05 COD
0
0.02 0.04 0.06 0.08 0.1 0.12 0.14
D (day-1)
Fig. 3. Static input-output characteristics Q=Q(D) and
COD = COD(D)of the model
Static input-output characteristics Q=Q(D) and COD =
COD(D) have been obtained using the new model developed.
Two main conclusions can be drawn from the new model
analytical study:
1. Maximal biogas production exists for a value of D in
the admissible range of this control action.
2. The increase of the biogas yield with increase of D
was combined with increase of COD, respectively low extent
of biodegradation.
Further efforts will be concentrated on clarification of the
role of the different microorganisms populations in the
methane production via the model developed and for stability
analysis of this model
Acknowledgements. This work was supported by contract
DO 02-190/08 of the Bulgarian National Science Fund”.
REFERENCES
Ahring, B.K. (Editor)(2003). Biomethanation. Springer-
Verlag, Berlin Heidelberg,.
Angelidaki, L., Ellegaard, L. E., Ahring, B. K. (1993).
Biotech.Bioeng., 42, 159-166.
APHA-AWWA-WPCF. (1985). American Public Health
Association Standard methods for the examination of
waste and wastewater, D.C., Washington.
Batstone, D. J, Keller, J., Angelidaki, I., Kalyuzhnyi, S. V.,
Pavlostathis, S. G., Rozzi, A., Sanders, W. T., Siegrist,
M. H., Vavilin, V. A. (2002). Anaerobic digestion model
no. 1 (ADM1), IWA Task Group for Mathematical
Modelling of Anaerobic Digestion Processes, IWA,
London.
Bradford, M. M. (1976). Analytical Biochemistry, 72, 248-
254.
Deublein D., Steinhauser A. (2008). Biogas from waste and
renewable resources. Wiley-VCH Verlag, Weinheim.
Dochain, D., Vanrolleghem P.(2001). Dynamical Modelling
and Estimation in Wastewater treatment Processes. IWA
Publishing, UK.
Gerardi M.H (2003). The microbiology of anaerobic
digesters. John Wiley&Sons, Inc. New Jersey.
Karakashev D. (2004) PhD thesis, Institute of Microbiology
of BAS.
Lyberatos, G., Skiadas, I. V. (1999). Global Nest: Int. J. 1,
63-76.
Schmidt, I., Sliekers, O., Schmid, M., Ciprus, I., Strous, M.,
Bock, E., Kuenen, J. G., Jetten, M. S. M. (2002). FEMS
Microbiology Ecology, 39, 175-181.
Schnurer A., Zellner, G., Svensson, B. H. (1999). FEMS
Microbiology Ecology, 29, 249-261.
Research, 30, 1087-1094.
Simeonov, I. (1999). Bioprocess Eng., 21(4), 377-38.
Simeonov I. (2010). Chapter 2. Modelling and control of the
anaerobic digestion of organic wastes in continuously
stirred bioreactors. In Tzonkov S. (ed.), Contemporary
approaches to modeling, optimization and control of
biotechnological processes, 41-76. Prof. Marin Drinov
Acad. Publ. House, Sofia.
Stefanie J. W. H., Elferink, O., Luppens, S. B. I., Marcelis, C.
L. M., Stams, A. J. M. (1998). Appl. Env. Microb., 64
(6), 2301-2303.
Strous, M., Kuenen, J. G., Jetten, M. S. M. (1999). Appl. Env.
Microbiol., 65 (7), 3248-3250.
Van de Graff, A. A., Mulder, A., Bruijn, P. Jetten, M. S. M.,
Robertson, L. A. Kuenen, J. G., (1995). Appl. Env.
Microbiol., 61 (4), 1246-1251.