Kinetics of thermophilic batch anaerobic digestion of thermal hydrolysed waste activated sludge

Abstract

Enhancing waste activated sludge (WAS) anaerobic digestion by thermal pretreatment has become of high interest. However, thermal treatment has been mainly combined to mesophilic anaerobic digestion. This paper presents the combination of sludge thermal pretreatment (110, 165 and 220 °C) and batch thermophilic anaerobic digestion (55 °C). Optimal conditions of thermal pretreatment were shown to be 165 °C, involving a chemical oxygen demand (COD) and volatile solids (VS) solubilisation of 18 and 15% and a biodegradability increase from 47 to 61%. Treatments at 165 °C were carried out in electric and steam modes and no significant difference on the impact of heating mode on sludge anaerobic biodegradability was observed. Moreover, it may be recommended not to carry out successive batch experiments to assess thermophilic BMP of sludge as accumulation of volatile fatty acids (VFA), particularly propionate, and a decrease of VFA uptake rates may occur. However, thermal pretreatment at 165 °C allowed the decrease of propionate accumulation and an higher methane production.

Full text

Page 1

Biochemical Engineering Journal 46 (2009) 169–175
Contents lists available at ScienceDirect
Biochemical Engineering Journal
journal homepage: www.elsevier.com/locate/bej
Kinetics of thermophilic batch anaerobic digestion of thermal hydrolysed waste
activated sludge
A. Motteta,b, J.P. Steyera, S. Délérisb, F. Vedrenneb, J. Chauzyc, H. Carrèrea,∗
aINRA, UR50, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne, F-11100, France
bVeolia Environnement R&D, Centre de Recherche sur l’Eau, F-78603 Maisons-Laffitte, France
cVeolia Water Direction Technique, 1 rue Battista Pirelli, F-94410 Saint Maurice, France
a r t i c l ei n f o
Article history:
Received 20 October 2008
Received in revised form 30 April 2009
Accepted 5 May 2009
Keywords:
Anaerobic digestion
Sludge management
Thermal pretreatment
Methane potential
Volatile fatty acids
a b s t r a c t
Enhancing waste activated sludge (WAS) anaerobic digestion by thermal pretreatment has become of
high interest. However, thermal treatment has been mainly combined to mesophilic anaerobic digestion.
This paper presents the combination of sludge thermal pretreatment (110, 165 and 220◦C) and batch
thermophilic anaerobic digestion (55◦C). Optimal conditions of thermal pretreatment were shown to be
165◦C, involving a chemical oxygen demand (COD) and volatile solids (VS) solubilisation of 18 and 15%
andabiodegradabilityincreasefrom47to61%.Treatmentsat165◦Cwerecarriedoutinelectricandsteam
modes and no significant difference on the impact of heating mode on sludge anaerobic biodegradability
was observed. Moreover, it may be recommended not to carry out successive batch experiments to assess
thermophilic BMP of sludge as accumulation of volatile fatty acids (VFA), particularly propionate, and a
decrease of VFA uptake rates may occur. However, thermal pretreatment at 165◦C allowed the decrease
of propionate accumulation and an higher methane production.
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
Biological processes are developed as the main process to
improve efficiently the quality of the effluent in municipal
wastewater treatment plants. Waste activated sludge (WAS), as a
by-product, are generated in large and increasing quantities. Since
the development of the biogas sector in France, following the July
2006 publication of new sufficiently attractive prices, WAS is more
and more considered as renewable energy source and becomes an
interesting substrate to anaerobic digestion. However, WAS anaer-
obic digestion is more difficult than for primary sludge and, with
usualtechnologiesnowadaysavailable,onlyapproximately20–30%
of the sludge organic matter are mineralized [1].
Anaerobic digestion process follows four major steps: hydroly-
sis, acidogenesis, acetogenesis and methanogenesis. Hydrolysis is
a slow step and is considered as the rate-limiting step of the over-
all process in the case of sludge degradation [2]. During hydrolysis
step, organic compounds, such as polysaccharides, proteins and
fats, are hydrolysed by extracellular enzymes to monomer or dimer
compounds.
∗Corresponding author. Tel.: +33 468 425 168; fax: +33 468 425 160.
E-mail addresses: mottet@supagro.inra.fr (A. Mottet), steyer@supagro.inra.fr
(J.P. Steyer), Stephane.DELERIS@veolia.com (S. Déléris),
Fabien.VEDRENNE@veolia.com (F. Vedrenne), Julien.CHAUZY@veoliaeau.fr
(J. Chauzy), carrere@supagro.inra.fr (H. Carrère).
Over the last years, pretreatment steps, such as physical treat-
ment with bead mill, sonication or high-pressure homogenizer,
biological treatment with enzymatic hydrolysis, chemical treat-
ment with alkaline addition were applied to improve hydrolysis
of particulate organic matter and substantially biodegradability of
sludge. Thermal treatment has been widely combined to anaero-
bic digestion performed in the mesophilic range and this resulted
in an increase of biogas production and of the kinetic rates [3] and
energy costs can be covered by biogas production [4,5]. Climent
et al., Bougrier et al. and Jeong et al. [6–8] underlined the impact
of solubilisation of particulate organic matter on the biogas pro-
duction enhancement during anaerobic digestion. Another way to
increase anaerobic digestion performances consists in operating
digestors in thermophilic mode. Indeed, thermophilic anaerobic
digestionallowstoimprovedegradationrates,henceshorterreten-
tion times and provides an enhancement of the biogas production
[9]. However, very few studies deal with the combination of ther-
mal pretreatment and thermophilic anaerobic digestion of sludge
[10–12].
The objective of this work was to carefully analyse the impact
of thermal pretreatment on thermophilic anaerobic digestion of
WAS. Efficiency of thermal pretreatment was evaluated by solu-
bilisation of COD and volatile solids (VS). Methane production and
kinetics of volatile fatty acids (VFAs) and soluble chemical oxygen
demand (COD) were measured during batch anaerobic biodegra-
dation in order to investigate in details the hydrolysis of pretreated
and untreated organic matter. Moreover, electric and steam modes
1369-703X/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.bej.2009.05.003

Page 2

170
A. Mottet et al. / Biochemical Engineering Journal 46 (2009) 169–175
for thermal pretreatment at 165◦C were compared. Indeed, electric
mode is generally used at laboratory scale whereas steam mode is
used in industrial plants [13].
2. Materials and methods
2.1. Waste activated sludge samples
Sludge samples were taken from a wastewater treatment
plant (France), treating urban wastewaters and working with
high load aerobic process at a sludge retention time of 0.4 days.
Characteristics of WAS or raw sludge were: Total solids (TS):
46.6±2.3gL−1,volatilesolids(VS):36.6±1.8gL−1,totalsuspended
solids (TSS): 40.5±2.0gL−1(87%TS), volatile suspended solids
(VSS):31.5±1.6gL−1(86%VS),COD:64.0±9.6gCODL−1andsoluble
COD (<0.45?m): 5.6±0.3gCODL−1.
2.2. Thermal hydrolysis
Thermal pretreatments were performed in a 10L agitated auto-
clave (Autoclave, class IV), allowing a temperature increase by
electricmodeorbysteammode.Sludgesamplevolumewasaround
6L. Temperatures of treatment were 110◦C, 165◦C and 220◦C in
electric mode and 165◦C in steam mode. Once temperature was
reached, treatments lasted for 30min. In general, at laboratory
scale, autoclave has only a temperature increase by electric mode.
Thus the process, used for thermal pretreatment, allowed us to
compare performances obtained in electric mode against perfor-
mances obtained in steam mode that is generally used at industrial
scale.
The solubilisation of COD and VS were chosen to evaluate
the degree of disintegration of pretreated sludge. This parameter
defines the transfer of particulate organic matter to soluble organic
matter. The solubilisation was expressed as a percentage, following
the equation:
SX(%) =
SS− SS0
Xp0
× 100
where SSand SS0are the concentrations in soluble fraction, mea-
sured in treated and untreated sludge, respectively; Xp0is the
concentrationinparticulatefraction,measuredinuntreatedsludge.
2.3. Determination of anaerobic degradation tests
Biochemical methane potential (BMP) tests were used to estab-
lish anaerobic biodegradability and determination of the methane
potential of WAS samples. The method was based from Buffiere et
al.[14].Anaerobicbatchreactorswereoperatedunderthermophilic
conditions. The temperature was kept at 55◦C by water circulation
in a water jacket. Six reactors, with a volume of 3.5L each, were
used in parallel. The inoculum was taken from a full scale sludge
anaerobic digester. One reactor was used with no feed to quan-
tify the endogenous activity of the inoculum. Others reactors were
fed with untreated sludge and with sludge treated at 110◦C, 165◦C
(electricandsteammodes)and220◦C.Organicloadingwas0.5gCOD
of sludge per gVSof inoculum. Large inoculation ratio ensured high
microbialactivity,lowriskforoverloadingandlowriskofinhibition
[15]. For each condition, four successive 22 days batch experiments
were carried out in a same reactor to minimise the effect of the
inoculum and two successive batch experiments were separated
by 2 days. At the beginning of each BMP test, the reactors were
purged with N2/CO2(75/25) gas mixture [16]. Biogas production
and pH were measured continuously. An electronic volumetric gas
counter was used to monitor biogas production. BMP is expressed
as the produced methane relatively to the amount of introduced
COD (mLCH4gCODin−1), in STP conditions.
It has to noted that endogenous activity decreased from batch
1 to 2. It represented about 12% of cumulated methane production
from raw sludge in batch 1, 6% in batch 2, 3% in batch 3 and 5% in
batch 4.
The biodegradability of the sludge was estimated from the BMP
value and the theoretical maximal methane yield of 350mLgCOD−1
in STP conditions, following the equation:
Biodegradability(%) = 100 ×BMP(mLCH4gCODintroduced−1)
350(mLCH4gCOD−1)
During each batch anaerobic digestion, total and soluble COD, VFA,
biogas composition were regularly monitored in order to follow
the formation of by-products, involved in the biological reactions
(Fig. 2).
Newman–Keuls tests, with a confidence interval at 90%, were
realised to compare biodegradability values of each treated sludge
and raw sludge and to estimate the significant differences between
each biodegradability values.
The maximum accumulation rates of propionate and acetate
and the maximum production rate of methane can be used to
evaluate the impact of thermal pretreatment on the steps of aceto-
genesis and methanogenesis. Propionate and acetate accumulation
rates (Kpropionateand Kacetate) were calculated from the variation
of propionate and acetate concentrations versus time during the
accumulation phases. Kpropionateand Kacetatewere determined by
the slope of the linear regression line between day 0 and the
day corresponding to the maximal accumulation. These param-
eters were expressed in gCODL−1day−1. Accumulation of each
compound is the result of production minus consumption. For
example, acetate accumulation is the result of acetate production
minus acetate conversion into methane and carbon dioxide. How-
ever, acetate production rate could not be calculated by taking
into account methane production because we could not determine
the parts of methane originating from acetoclastic methanogene-
sis and from hydrogenotrophic methanogenesis. For the methane
production rate, the methane volume was converted to gCODand
KCH4, expressed in gCODgCODintro−1day−1, was determined from
the maximum value of the methane specific production rate,
which was obtained by the derivative of the methane specific
production.
2.4. Analysis
The particulate fraction was separated by centrifugation at
50000×g, 15min and 5◦C. Supernatant was then filtered through
a cellulose acetate membrane with 0.45?m pore size to obtain the
soluble fraction.
COD measurements were realised on total and soluble fractions
and measurements of TS, VS, TSS and VSS were realised on sludge
andonsolidsofcentrifugation,accordingtostandardmethods[17].
The oven-dried step, at 105◦C for 24h, involved a volatilisation of
VFA. TS and VS concentrations were thus corrected by taking into
account VFA concentrations.
Proteins were measured according to the Lowry method [18].
The technique quantified the peptidic bounds. By using different
knownsolutionsofbovineserumalbumin(BSA),acalibrationcurve
was obtained and protein concentrations were determined in BSA
equivalent gram per litre. Carbohydrates were measured with the
anthrone reduction method [19]. It dosed carbohydrate concentra-
tions by quantifying the carbonyl functions (C O). A calibration
curve was obtained from glucose (Gluc) and carbohydrate concen-
trations were determined in glucose equivalent gram per litre.
VFAconcentrationsweremeasuredbyusinggaschromatograph
(GC-8000 Fisons instrument), equipped with a flame ionisation
detector with an automatic sampler AS 800. The internal stan-

Page 3

A. Mottet et al. / Biochemical Engineering Journal 46 (2009) 169–175
171
Table 1
COD and VS solubilisation, methane potential and biodegradability of untreated and pretreated WAS at different conditions.
Conditions Solubilisation of COD (%)Solubilisation of VS (%) BMP (mLCH4gCODin−1) Biodegradability (%)
Raw sludge
110◦C
165◦C, electric
165◦C, steam
220◦C
165 ± 17
186 ± 10
195 ± 9
215 ± 7
142 ± 22
47
53
56
61
41
3.8 ± 0.2
18.0 ± 1.0
17.8 ± 0.4
27.0 ± 1.0
1.9 ± 0,2
16.0 ± 2.0
14.0 ± 1.0
24.0 ± 2.0
dard method allowed to measure acetate, propionate, butyrate and
iso-butyrate, valerate and iso-valerate concentrations.
Thecompositionofbiogaswasdeterminedwithagaschromato-
graph (Shimadzu GC-8A), with a CTRI Alltech column, with argon
as the carrier gas, equipped with a thermal conductivity detector
and connected to an integrator (Shimadzu C-R8A).
3. Results and discussion
3.1. Impact of thermal pretreatment on sludge solubilisation and
biodegradability
Solubilisation of COD and VS are often used to evaluate the
impact of pretreatment on the sludge maximal biodegradability.
Results are summarised in Table 1. As already shown [6,7], the val-
ues of COD and VS solubilisation increased with temperature until
220◦C, from 3.8 to 27% and from 1.9 to 24%, respectively. The ther-
malpretreatmentledtoatransferofparticulateorganicmatterinto
the soluble phase, lower than 0.45?m and could be assimilated
to a thermal hydrolysis. Thus, the application of thermal pretreat-
mentonalargelyparticulaterawsludge(86%VS)shouldmakemore
availableorganiccomponentstotheanaerobicmicroorganismsand
couldimplyanincreaseofdegradationratesandofproducedbiogas
volume.
Nevertheless, the values of biodegradability (Table 1) showed a
threshold value in the increase of methane production. Indeed, it
increased with thermal pretreatment until a temperature of 165◦C,
from165mLCH4gCODin−1foruntreatedsludgeto215mLCH4gCODin−1
for sludge pretreated at 165◦C in steam mode. Thus, among the
tested temperatures, the optimum was 165◦C. It is worth noting
that an identical optimal temperature pretreatment was found for
anaerobic digestion under mesophilic condition [7].
Moreoverat220◦C,althoughalargesolubilisationofparticulate
organic matter occurred, sludge biodegradability was lowered to
raw sludge biodegradability with 142mLCH4gCODin−1. This may be
explained by the composition of solubilised organic matter (Fig. 1).
At 220◦C, protein solubilisation was similar to the one obtained
at 165◦C in steam mode, i.e. 40.1% and carbohydrates solubilisa-
tion strongly decreased from 15% at 165◦C to 1.2%. However COD
solubilisation increased from 18 to 27%. Thus, at 220◦C, carbo-
Fig. 1. Impact of treatment temperature on solubilisation of WAS.
hydrates in the soluble phase reacted with other components to
form products slowly or hardly biodegradable. These results are
in agreement with those of Bougrier et al., Müller and Stuckey
and McCarty [5,20,21]. They suggested that the presence of “burnt
sugar” reactions and Maillard reactions for high pretreatment tem-
peratures. The brown colour of the soluble phase of sludge treated
at 220◦C confirmed the presence of new compounds, like Amadori
compounds and melanoidins which are recalcitrant to anaerobic
degradation.
An objective of this study was to assess the impact of heating
modes on solubilisation and biodegradation results. Solubilisation
of COD, VS and carbohydrates, obtained at 165◦C in both modes,
did not show significant differences: it reached around 18, 15 and
15%, respectively with both modes. On the other hand, protein sol-
ubilisation was slightly higher for sludge treated with steam (40.2%
against 34.5% in electric mode). Moreover, the Newman–Keuls test
showed that the difference of BMP between treatments at 165◦C
(in electric and steam modes) were not significant with a 90% con-
fidence interval (Table 2). Thus, it can be concluded that laboratory
thermalhydrolysiscarriedoutinelectricmodeproperlyrepresents
industrial thermal hydrolysis carried out with steam injection.
Newman–Keuls test also showed that the improvement of
sludge thermophilic anaerobic biodegradability was not significant
afterthe110◦Cpretreatmentandconfirmedthereductionofsludge
biodegradability after 220◦C treatment.
Solubilisation values and methane production values, repre-
senting initial and final conditions of the anaerobic digestion, are
not sufficient to obtain a full understanding of anaerobic digestion
mechanisms. Kinetics of intermediate products (VFAs and soluble
COD) and final products of anaerobic digestion (CH4and CO2) are
indeedinterestingtoexplainbiodegradabilitydifferences,involved
by thermal pretreatment and to observe the potential degradation
limits.
3.2. Kinetics of batch anaerobic digestion
The batch anaerobic digestion experiments of the five sludge
samples were realised in five reactors, which were used in parallel.
As in Buffiere et al. [14], four successive batch experiments were
Table 2
Comparison of each methane production obtained with different thermal pretreat-
ment conditions (mean values of 4 successive batch experiments).

Page 4

172
A. Mottet et al. / Biochemical Engineering Journal 46 (2009) 169–175
Fig. 2. Soluble compound concentrations during batch anaerobic digestion of each
sludge (batch 4).
carried out in a same reactor for each substrate in order to evaluate
the inoculum adaptation. The experiments were monitored during
22 days.
Fig. 2 represents the measured CODsolubleminus VFAs in batch
4. It can represent the transfer of compounds from the particu-
late phase to the soluble phase, i.e., the behaviour of hydrolysis,
thus the production of soluble compounds (<0.45?m), before their
degradation by acetogenesis and acidogenesis.
Fig. 3 represents methane production and VFA concentrations
for batch 2, 3 and 4 for raw sludge and sludge treated at 110, 165◦C
and 220◦C. VFA concentrations were not monitored for batch 1.
It is important to define the VFA concentration variations moni-
tored during experiments; they result from the VFA production by
acidogenesis and from the VFA uptake by acetogenesis and aceti-
clastic methanogenesis. For example, an accumulation of acetate
can be due to the step of acetate production faster than the step
Fig. 3. Methane production and VFA concentrations during batch anaerobic digestions of (A) untreated sludge, (B) pretreated at 110◦C, (C) pretreated at 165◦C in electric
mode, (D) pretreated at 165◦C in steam mode and (E) pretreated at 220◦C.

Page 5

A. Mottet et al. / Biochemical Engineering Journal 46 (2009) 169–175
173
of acetate uptake. These values can bring interesting information
into the understanding of mechanisms involved in the methane
production.
The soluble COD concentration variations (Fig. 2) of sludge pre-
treated in optimal conditions (165◦C) showed a direct degradation
ofsolubleorganicmatterandnosignificantaccumulationofsoluble
COD. Thus, pretreatment allowed a better accessibility of organic
matter, which was directly available for the biological steps, and
could minimise the limiting effect of biological hydrolysis.
For the sludge pretreated at 220◦C, despite a large solubilisa-
tion of COD (27%), soluble compound concentration at the end of
experimentwashighanddecreasedveryslowly.Itseemstoindicate
that compounds, formed at 220◦C (e.g. memanoidins and Amadori
compounds) were hardly biodegradable.
In the case of raw sludge and pretreated at 110◦C, a first phase of
degradation of particulate organic matter from day 0 to day 3 and
a second phase of degradation from day 3 to day 6 were observed.
The significant accumulation of soluble components seems to lead
totheconclusionthattwotypesofparticulatematterconstitutethe
sewage sludge: one fraction is quickly hydrolysed whereas the sec-
ond fraction is more slowly hydrolysed. These two fractions are less
significant on the soluble COD concentration variations of sludge
pretreated at 165◦C. But they can be clearly observed for all sludge
samples in Fig. 2.
Indeed methane production curves showed two phases (Fig. 3).
A first one, with a fast production corresponding to the degra-
dation of readily accessible organic matter, like monomers and
dimers compounds or exopolymers, lasted from day 2 to day 6. It is
also characterised by the production and the accumulation of VFAs
(from day 0 to day 4). The second phase, with a slower produc-
tion, corresponding to the degradation of hardly accessible organic
matter, like particulate macromolecules strongly linked in sludge
structure or compounds located inside the cells lasted from day 6
to days 12–14. Finally, methane production was null or very low
from day 14 to day 22.
Thevariationsofacetateandpropionateconcentrationscouldbe
associated to the first and the second phases of methane produc-
tion, respectively. Batch experiment 2 with sludge treated at 165◦C
in electric mode (Fig. 3C) could confirm this hypothesis. It was
observed that the first phase of methane production was mainly
associatedtotheacetateuptake,fromtheday0today6.Duringthis
phase, the highest methane volume was produced and the acetate
concentration decreased strongly. The second phase of methane
production was associated to the propionate uptake. Indeed the
start-up of the second methane production phase with the uptake
of propionate at day 7 could be observed and a plateau was reached
at day 12 when the VFA uptake was complete. According to Vavilin
et al. [22], the acetogenesis and methanogenesis can be the rate-
limiting steps in anaerobic digestion of a complex substrate at a
high organic loading. These observations seem to show that the
acetate and propionate degradation steps can also be considered as
the limiting step for the anaerobic digestion of a complex substrate
(waste activated sludge) with an organic loading of 0.5gCODper gVS
of inoculum.
As discussed above, total methane production increased with
thermalpretreatmenttemperature.However,asignificantdecrease
of methane production for raw sludge was observed through suc-
cessive batch experiments (Fig. 3A). This seems to be linked to the
propionate accumulation. Thus successive thermophilic digestion
batches of raw sludge may lead to an accumulation of propionate
as already observed by Speece et al. [23].
Inthecaseofsludgetreatedat220◦C,thefirstphaseofmethane
production, which was from day 0 to day 6 for the batch 2, had
a tendency to increase in time through consecutive batch experi-
ments, while a small methane production and a slow degradation
rate of acetate were observed. This seems to indicate that the
Fig. 4. Acetate (A), propionate (B) accumulation maximum rates and methane
(C) production specific maximum rates with different conditions of pretreatment
through batch experiments.
hardly biodegradable components produced at 220◦C are slowly
biodegradable. Thus the results can confirm that the melanoidins
and Amadori compounds are hardly and slowly biodegradable.
The enhancement of methane production observed during batch 4
could arise from an overestimation of produced volume (resulting
from very slowly biodegradable compounds from previous experi-
ments) and not from an adaptation of inoculum.
The values of propionate and acetate accumulation maximum
rates, K expressed in gCODL−1day−1, can be used to accurately eval-
uate the impact of thermal pretreatment on the mechanisms of
acetogenesisandmethanogenesissteps(Fig.4AandB).Thisparam-
eter was determined from the slope of the linear regression line
associated to the accumulation phase of propionate and acetate
from day 0 to the day corresponding to the maximal accumulation
of each batch experiment.
The values of Kpropionatedid not show statistically significant dif-
ferences (Fig. 4A). However, important differences were observed
betweenthevariationsofpropionateconcentrationsofeachsludge
with an accumulation of propionate in the case of the raw sludge