Chemical Engineering Journal 139 (2008) 236–244

Effects of thermal treatments on five different waste activated sludge

samples solubilisation, physical properties and anaerobic digestion
Claire Bougrier, Jean Philippe Delgenès, Hélène Carrère ∗
INRA, UR050, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne F-11100, France
Received 18 July 2006; received in revised form 26 July 2007; accepted 27 July 2007

Abstract
In order to face excess waste activated sludge management problems, sludge anaerobic digestion with thermal pre-treatment is of great interest.
If most of works agree on the optimal treatment temperature (160–180 ◦ C), results of thermal pre-treatments in terms of biogas production are very
dispersed. With the aim of analysing the impact of sludge samples, thermal pre-treatments were carried on five different waste activated sludge
samples.
For temperatures lower than 200 ◦ C, COD solubilisation was found to increase linearly with treatment temperature and all the different sludge
samples behaved in the same way. For temperatures lower than 150 ◦ C, carbohydrates solubilisation was more important than proteins solubilisation.
Analyses of sludge apparent viscosity, settleabilty and dewaterability (CST) of pre-treated sludge pointed out a threshold temperature of 150 ◦ C.
Thermal treatments up to 190 ◦ C allowed the biogas production to increase during batch anaerobic digestion of sludge. Biogas volume enhancement
was linked to sludge COD solubilisation and to untreated sludge initial biodegradability; the lower the initial biodegradability, the higher is the
impact of thermal treatment.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Methanisation; Biogas; Viscosity; Filterability; Settleability; Proteins; Carbohydrates; Lipids

1. Introduction thermal pre-treatments and anaerobic digestion is widely inves-

tigated in literature. Thermal treatments were first applied to
Sludge represents the major solid waste from biological sludge to improve their dewaterability [4]. Sludge dewaterability
wastewater treatment processes. In response to the strengthening was reported to be improved after a treatment at a temperature
of the European legislation regarding the Urban Wastewa- higher than 150 ◦ C [5] or higher than 180 ◦ C [6]. But a large
ter Treatments Directive (91/271/EEC), sludge production is part of works deals with the effect of sludge thermal treatments
increasing whereas disposal routes are more reduced. It is thus on biogas production enhancement during anaerobic digestion.
essential to develop processes to reduce sludge quantity. One of Table 1 presents some examples of such studies. At laboratory
the most interesting processes is anaerobic digestion or methani- scale, anaerobic digestion is carried out either in batch reac-
sation. It leads to sludge stabilisation by converting a part of its tors or in CSTR. It was operated in a fixed bed reactor in one
organic matter into biogas which is valuable as a renewable industrial reference [16]. Most of the studies reported an opti-
energy source. The rate-limiting step of this biological process mal temperature in the range from 160 to 180 ◦ C and treatment
is organic matter hydrolysis. Methanisation process can thus times from 30 to 60 min. However, treatment time is often shown
be improved by a thermal pre-treatment in order to lyse sludge to have little effect in this temperature range [4] and Dohanyos
cells [1]. Moreover, thermal pre-treatments imply sludge san- et al. [14] proposed a very fast thermal treatment, lasting only
itation and energy costs can be covered by biogas production 60 s. On the other hand, thermal treatments at moderate tem-
[2]. Therefore, some industrial plants such as Cambi process [2] perature (70 ◦ C) lasted several days [11]. If we consider their
and BioTHELYS® [3] are commercialised and combination of impact on biogas production, thermal treatments can be clas-
sified into two groups: treatments at temperatures of 70 ◦ C or
121 ◦ C which led to a 20–30% biogas production increase and
∗ Corresponding author. Tel.: +33 4 68 42 51 68; fax: +33 4 68 42 51 60. treatments at 160–180 ◦ C which led to a 40–100% biogas pro-
E-mail address: carrere@ensam.inra.fr (H. Carrère). duction increase. The 160–180 ◦ C pre-treatments are thus most

1385-8947/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.cej.2007.07.099
 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244 237

bilisation. The impact of thermal treatments on sludge viscosity,

Nomenclature settleability and dewaterability (CST) was investigated in a sec-
ond part. Finally, batch anaerobic biodegradation of pre-treated
Bo untreated sludge biodegradability (%) and untreated sludge was measured.
BSA bovine serum albumin (–)
COD chemical oxygen demand (g O2 /l) 2. Experimental
CODp COD in particles (g O2 /l)
CODs soluble COD (g O2 /l) 2.1. Waste activated sludge samples characteristics
CST capillary suction time (s)
CSTR continuous stirred tank reactor (–) The experiments were carried out using five different WAS
Gluc glucose (–) samples. Samples were collected from various plants located
HRT hydraulic retention time (d) in the South of France. Table 2 presents sludge samples main
N-NH4 + ammonium nitrogen (g N/l) characteristics. Samples used could be ranged according to two
SCOD COD solubilisation (–) parameters: the process from which they originated and their
SVI sludge volume index (ml/g) organic content.
TS total solids (g/l)
TSS total suspended solids (g/l) 2.2. Thermal treatment conditions
VFA volatile fatty acids (g/l)
VS volatile or organic solids (g/l) The reactor used for thermal treatment was a Zipperclave
VSS volatile or organic suspended solids (g/l) (Autoclave France) controlled in temperature via a regulation
WAS waste activated sludge (–) using a proportional integral derivative (PID). Sample volume
was around 0.7 l. Temperature of treatment varied from 90 to
Greek letters
210 ◦ C according to the sludge samples. The rise in temperature
θ temperature of treatment (◦ C)
lasted from 25 to 60 min, depending on the desired temperature.
μap apparent viscosity (Pa s)
Once temperature was reached, treatments lasted 30 min.
In table or figure presentations, untreated sludge was arbi-
trarily associated to 20 ◦ C temperature.
efficient to enhance sludge anaerobic digestion but they lead to
very dispersed results in terms of increase in biogas production 2.3. Samples analyses
(from 40 to 100%). Such a wide range of results was obtained
with different anaerobic digestion process conditions (batch pro- 2.3.1. Sludge composition
cesses of 7 to 24 days, CSTR with hydraulic retention time of In order to determine sludge composition, several measure-
5 to 25 days). Moreover, we may wonder what is the impact of ments were made on samples according to Standard methods
the sludge samples on the increase of biogas production due to [19]. First, soluble and particulate fractions were obtained
thermal treatment. Or is it interesting to treat all types of waste after centrifugation (Beckman J2 MC – 25,000 rpm, 15 min,
activated sludge? 5 ◦ C).
The objective of this study was to investigate the impact of COD measurement can permit to achieve pollution concen-
thermal pre-treatment on five different sludge samples. In the tration, expressed in oxygen (O2 ) consumed to oxidise matter.
first part, sludge solubilisation was analysed in terms of COD, A known amount of oxidant is added to samples and, after
total and volatile solids, proteins, carbohydrates and lipids solu- reaction, the excess oxidant is analysed. By this difference in

Table 1
Literature review on impacts of thermal pre-treatments on waste activated sludge mesophilic anaerobic digestion
Reference Thermal treatment Anaerobic digestion Results

Haug et al. [7] 175 ◦ C, 30 min CSTR, HRT = 15 d Increase of CH4 production from 115 to 186 ml/g CODin (+62%)
Stuckley and McCarty [8] 175 ◦ C, 60 min Batch, 25 d Increase of convertibility of COD to CH4 from 48 to 68% (+42%)
Li and Noike [1] 175 ◦ C, 60 min CSTR HRT = 5 d Increase of gas production from 108 to 216 ml/g CODin (+100%)
Tanaka et al. [9] 180 ◦ C, 60 min Batch, 8 d Increase of methane production (+90%)
Fjordside [10] 160 ◦ C CSTR, 15 d Increase of biogas production (+60%)
Gavala et al. [11] 70 ◦ C, 7 d Batch Increase of CH4 production from 8.30 to 10.45 mmol/g VSin (+26%)
Barjenbruch and Kopplow [12] 121 ◦ C, 60 min CSTR, 20 d Increase of biogas production from 350 to 420 ml/g VSSin (+20%)
Kim et al. [13] 121 ◦ C, 30 min Batch, 7 d Increase of biogas production from 3657 to 4843 l/m3 WASin (+32%)
Dohanyos et al. [14] 170 ◦ C, 60 s Batch, 20 d Thermophilic Increase of biogas production (+49%)
Valo et al. [15] 170 ◦ C, 60 min Batch, 24 d Increase of biogas production (+45%)
Valo et al. [15] 170 ◦ C, 60 min CSTR, 20 d Increase of CH4 production from 88 to 142 ml/g CODin (+61%)
Graja et al. [16] 175 ◦ C, 40 min Fixed film reactor, HRT = 2.9 d 65% reduction of TSS
Bougrier et al. [17] 170 ◦ C, 30 min Batch, 24 d Increase of CH4 production from 221 to 333 ml/g CODin (+76%)
Bougrier et al. [18] 170 ◦ C, 30 min CSTR, 20 d Increase of CH4 production from 145 to 256 ml/g VSin (+51%)
238 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244

Table 2
Untreated WAS samples characteristics
Sludge samples A B C D E

Size of the plant (PE) 500,000 90,000 10,000 33,000 60,000

Process High load High load Medium load Medium load Extended aeration
Wastewater Urban and industrial Urban and industrial Urban Urban Urban and industrial
TS concentration (g/l) 33.7 ± 0.2 15.3 ± 0.2 17.1 ± 0.2 15.0 ± 0.1 14.8 ± 0.2
VS content (%TS) 81 ± 2 76 ± 2 70 ± 2 82 ± 1 76 ± 2
TSS content (%TS) 93 ± 1 93 ± 2 80 ± 1 97 ± 1 95 ± 2
Total COD (g/l) 35.1 ± 0.1 14 ± 1 17.4 ± 0.5 10.5 ± 0.7 11.4 ± 0.4
Soluble COD (g/l) 1.52 ± 0.04 1.3 ± 0.2 0.47 ± 0.02 0.04 ± 0.01 0.04 ± 0.01
Total proteins (g eqBSA/l) 12.2 ± 0.3 nd nd 4.3 ± 0.2 5.1 ± 0.4
Soluble proteins [g eqBSA/l] 0.31 ± 0.03 nd nd 0.30 ± 0.05 0.20 ± 0.03
Total carbohydrates (g eqGluc/l) 2.66 ± 0.07 nd nd 1.34 ± 0.02 1.45 ± 0.05
Soluble carbohydrates (g eqGluc/l) 0.13 ± 0.01 nd nd 0.020 ± 0.001 0.02 ± 0.01
Total lipids (g/l) 1.5 ± 0.3 nd nd nd nd
VFA (g/l) 0.23 ± 0.01 nd nd 0.065 ± 0.001 0.016 ± 0.001

TS: total solids; VS: volatile solids (or organic solids); TSS: total suspended solids (solids concentration in the particular fraction of sludge); nd: not determinated.

the oxidant concentration, it is possible to calculate the oxy- using different known solutions of glucose (Gluc), a calibra-
gen quantity necessary to degrade pollution, that is to say the tion curve was obtained and carbohydrate concentrations were
COD concentration. COD was measured on the total sludge determined in glucose equivalent gram per litre. The error due
and on the supernatant. For this paper, COD measured on to this measure was around 10%.
supernatant will be called “soluble COD” (CODs) and the dif- Lipid concentration was determined using two techniques.
ference between total COD and soluble COD will be called First, fatty solids concentration was measured by extraction by
“particulate COD” (CODp). The error due to this measure was hexane according to Bridoux et al. [22]. Samples were acid-
around 10%. ified in order to maintain fatty acids in the non-dissociated
Measures of total and organic solids (TS and VS) were form. Hexane was added to samples which were agitated;
realised on sludge and on solids of centrifugation (Total and fatty acids transferred from sludge to hexane fraction. Then
Volatile Suspended Solids: TSS and VSS). Samples were heated the hexane phase was collected and evaporated (Rotavapor R,
at 105 ◦ C for 24 h; water was evaporated. That led to total matter Büchi). By weighting the extracted compounds, and by know-
concentration. Then, samples were heated at 550 ◦ C for 2 h. That ing the initial volume of sludge, it was possible to determine
led to mineral matter concentration. Organic matter concentra- fatty solids concentration. The error due to this measure was
tion was then deduced. Solids concentration of the supernatant, around 15%.
that is to say the soluble phase, was deduced from the difference Beside, volatile fatty acids (VFA) concentrations were
between total solids and suspended solids concentrations. All determined in the soluble fraction by gas chromatography
these concentrations led to the composition in the different parts (GC800, Fisons Instruments). The internal standard method
of the sludge. The error due to this measure was around 3–5%. allowed to measure total VFA concentration (acetic, propi-
Measures of ammonium nitrogen (N-NH4 + ) was realised in onic, butyric and iso-butyric, valeric and iso-valeric acids)
the soluble fraction by colorimetric dosage [19]. The error due in the range 0.25–1 g/l. The error due to this measure was
to this measure was around 10%. around 3%.
In order to better know solids and soluble fractions, proteins,
carbohydrates and lipids concentrations were measured. 2.3.2. Sludge physical properties
Protein concentration was determined on total sludge and on Viscosity measurements were carried out using a RT 10 Roto-
supernatant using the Lowry method [20]. The technique quan- visco (Haake Fisons) connected to a computer. The system was
tified the peptidic bounds. After reactions with salts and Folin a coaxial cylinder system with a gap of 0.925 mm. Measures
reagent, absorbance of samples was determined at 750 nm, using were realised by increasing the shear stress from 3 to 10 Pa and
a spectrophotometer (DV-640, Beckman). By using different were duplicated.
known solutions of bovine serum albumin (BSA), a calibration Settleability was estimated using the sludge volume index
curve was obtained and protein concentrations were determined (SVI). Sludge was diluted several times (twice, four times).
in BSA equivalent gram per litre. The error due to this measure One litre was introduced in test tubes. After 30 min, the settled
was around 15%. volume was measured.
As well as proteins, carbohydrate concentration was deter- The filterability was measured using the capillary suction
mined on total sludge and soluble fraction. The anthrone method time (CST). The apparatus was a Triton type 319 Multi-CST
[21] was used. It dosed carbohydrates concentration by quantify- (Triton Electronics Ltd.). The CST permits to estimate the sludge
ing the carbonyl functions (C O). After reaction with anthrone ability to dewater; water is absorbed by CST paper by capillary.
and sulphuric acid, absorbance of samples was determined at The CST measure corresponds to the time needed for water to
625 nm using a spectrophotometer (DV-640, Beckman). By cross a fixed distance in the filter paper.
 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244 239

2.4. Batch anaerobic digestion

Batch anaerobic digestion tests were carried out to assess

sludge biodegradability. For these experiments the inoculum
was a sludge treating a mixture of wine effluents (80%)
and sludge (20%). The inoculum was diluted to 4 g/l of
volatile suspended solids (VSS – equivalent to organic sus-
pended solids). For each pre-treatment, samples of treated
or untreated sludge were added to 400 ml of inoculum. The
pollution to degrade was equivalent to 0.5 g COD/g VSS of
inoculum. Plasma bottles were agitated (200 rpm). Several con-
trol samples were realised: a blank (water), an ethanol sample
(completely biodegradable compound) and an untreated sam-
ple (untreated WAS added). Produced biogas volume was
measured by movement of liquid (water, pH = 2, NaCl 10%).
Experiments lasted from 17 to 24 days according to sludge Fig. 1. Effect of thermal treatment on COD solubilisation for the various sludge
samples. samples and comparison with literature. The points at 20 ◦ C correspond to
Enhancement of biodegradability was evaluated by com- untreated sludge.
parison of biogas volumes produced by treated and untreated
samples. Biogas ratio is biogas volume produced with treated expressed by:
sludge divided by biogas volume produced with untreated
CODs − CODs0 CODs − CODs0
sludge. Moreover, the biodegradability percentage was SCOD = = (1)
estimated by comparing the biogas volume produced with COD0 − CODs0 CODp0
sludge (treated or not) to the biogas volume produced with
3.1.1. COD solubilisation
ethanol. For all substrates, volumes of biogas were reported
Fig. 1 presents results obtained for COD solubilisation of the
to the COD introduced in the bottle (quantity of pollution to
five sludge samples. All the points, obtained from this study
degrade).
or in the literature, seemed to be put in almost a line. But,
for temperature higher than 200 ◦ C, results seemed to be more
2.5. Partial least square analyses dispersed.
Therefore, COD solubilisation was linked to temperature of
Partial least squares (PLS) regression technique is based in pre-treatment whatever be the sludge samples characteristics.
constructing PLS factors (also called principal components) by
minimising the covariance between the dependent variables (Y 3.1.2. Solids solubilisation
block) and the explicative variables (X block). Then, the predic- At the same time, solids solubilisation was observed. Table 3
tion of the Y block was calculated with a multivariable linear presents results obtained for the five sludge samples tested.
regression on X block through PLS1 models using the software For all sludge samples, TSS/TS ratio decreased with treatment
R version R 1.2.2 for Windows and by using PLS functions temperature, as well as the ratio VSS/TSS. Therefore, solid
developed by Durand [23]. The algorithm constructs orthogonal concentration in particles decreased and particles became more
PLS factors in each block by minimising the covariance between mineral. For all sludge tested, the solubilisation level increased
the X and Y blocks. The first PLS factor contains the highest
percentage of variance, and the following factors account for Table 3
decreasing amounts of variance. The number of PLS factors Solubilisation of solids due to thermal treatments
(also called dimensions, dim) of the model was determined by Temperature (◦ C) 20 95 110 130 150 170 190 210
minimising the mean squared prediction error (PRESS) through
Sludge sample
a cross-validation procedure.
TSS/TS (%)
A 93 82 74 64 57
B 93 77 70 54 33 27
3. Results and discussion
C 80 64 49 29
D 97 84 79 72 63 59 46 45
3.1. Effects of thermal treatments on sludge solubilisation E 95 83 87 77 68 58 45 47
VSS/TSS (%)
In this study, the term “solubilisation” was used in order to A 82 81 79 74 72
describe the transfer from the particular fraction to the super- B 79 76 70 65 66 68
natant of centrifugation. Solubilisation of COD, proteins and C 82 77 79 64
carbohydrates has been defined as the ratio of the soluble D 83 80 82 78 74 71 65 61
E 78 71 73 72 68 66 68 56
fraction minus the initial soluble fraction divided by the ini-
tial particulate fraction. For example, COD solubilisation was 20 ◦ C data corresponds to untreated sludge.
240 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244

regularly with the rise in temperature. A treatment of 170–190 ◦ C

led to a solubilisation level of almost 40–60% for sludge A, D
and E and to a solubilisation level of 75–80% for sludge B and C.
No global parameter seemed to explain this difference between
sludge samples.
Due to thermal treatments, solids were solubilised, especially
organic solids. Organic solids were very affected by treatment;
for a temperature of treatment higher than 150 ◦ C, organic solids
solubilisation was more than 43% for all sludge samples tested.
But, according to literature [26], it seemed that all organic com-
pounds did not react in the same way. For instance, Barlindhaug
and Odegaard [27] emitted the hypothesis that carbohydrates
were more easily degraded than proteins, but that proteins were
better solubilised. In order to better understand behaviour of
each kind of compounds, carbohydrate, protein and lipid con- Fig. 3. Proteins and carbohydrates solubilisation for three sludge samples (A,
centrations were determined in the sludge samples before and D and C). The points at 20 ◦ C correspond to untreated sludge.
after thermal treatment.
This decrease in carbohydrates concentrations could be
3.1.3. Carbohydrates and proteins solubilisation explained by the measurement method. Indeed, carbohydrates
Further to thermal treatment, proteins and carbohydrates were were measured by spectrophotometry by quantifying the car-
solubilised. Fig. 2 shows proteins concentration (curves in black) bonyl function (C O). If carbohydrates had reacted with
and carbohydrates concentration (curves in grey) in the whole other carbohydrates (“burnt sugar” reactions) or with proteins
sludge sample A and in the soluble fraction. Experiments have (Maillard reactions), these carbonyl functions disappeared and
also been realised on two other sludge samples (D and E). Results compounds were not measured.
were in agreement with conclusions exposed here. Nevertheless, it seemed that carbohydrates were in a
Total proteins concentration seemed to increase. This could first time more easily solubilised than proteins. Fig. 3
be explained by a better quantification of proteins after heat- presents solubilisation rate calculated for proteins and carbo-
ing. However, as the error of measure was around 15%, the hydrates for the three sludge samples tested. For all samples,
total proteins concentration could also be considered as con- solubilisation of carbohydrates was initially higher than pro-
stant and equal to almost 12.5 g eqSAB/l (sd: 1.2 g/l). Soluble teins one. Then, for temperature higher than 130 ◦ C or
proteins concentration increased strongly: from 0.31 g eqSAB/l 170 ◦ C (depending on samples), proteins solubilisation became
(sd: 0.03 g/l) for untreated sludge to 5.9 g eqSAB/l (sd: 0.8 g/l) higher.
for sludge treated at 170 ◦ C. On the contrary, total carbo- Thus, this could suggest a hypothesis on the location of these
hydrates concentration seemed to decrease; whereas soluble compounds. It seemed that carbohydrates were mainly located
carbohydrates concentration first increased strongly: from 0.13 g in the exopolymers of sludge structure whereas proteins were
eqGluc/l (sd: 0.01 g/l) for untreated sludge to 1.04 g eqGluc/l (sd: mainly located inside the cells. So, for “low” temperatures, only
0.12 g/l) for sludge treated at 130 ◦ C. Then the soluble carbohy- exopolymers were affected by thermal treatment: carbohydrates
drates concentration decreased in order to reach 0.78 g eqGluc/l were solubilised and also few proteins. For higher temperatures,
(sd: 0.13 g/l) for sludge treated at 170 ◦ C. cell walls were lysed; proteins were no more protected and were
strongly solubilised. At the same time, soluble carbohydrates
reacted (with themselves or with soluble proteins) and formed
organic compounds which were not quantified. We could suggest
that these new compounds were like Amadori compounds or
melanoidins and that they coloured the supernatant of sludge in
brown.
At the same time, in order to know if proteins were degraded
by heat, N-NH4 + measures were realised. N-NH4+ concen-
tration first increased with treatment temperature and then
remained constant: from 0.35 g N/l (sd: 0.05 g/l) for untreated
sludge to almost 0.7 g N/l (sd: 0.1 g/l) for temperatures higher
than 90 ◦ C. Therefore, it seemed that a small part of proteins
were completely degraded. So, proteins were solubilised and
only few degraded due to thermal treatment.

3.1.4. Lipids solubilisation

Fig. 2. Proteins and carbohydrates concentrations in the sludge sample A after Lipids concentration was determined by solvent extraction
thermal treatment. The points at 20 ◦ C correspond to untreated sludge. (hexane) for long chains fatty acids (what were called “lipids”)
 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244 241

Fig. 5. Apparent viscosity of sludge sample A at different shear rate after thermal
treatment. The points at 20 ◦ C correspond to untreated sludge.

Fig. 4. Lipids and VFA concentrations in sample A after thermal treatment. The
points at 20 ◦ C correspond to untreated sludge. could be linked to degradation of macromolecules into acidic
compounds.
and volatile fatty acids concentration (VFA) was determined by Thermal treatment had effects on sludge viscosity. Untreated
gas chromatography. Fig. 4 shows lipid (total and soluble) and sludge was pseudo-plastic fluid. Fig. 5 presents apparent viscos-
VFA concentrations in the sludge sample A. The total lipids ity of sludge sample A after thermal treatment, for different shear
concentration increased with thermal treatment. This increase rates. Apparent viscosity of sludge decreased with treatment
could be due to a better extractability of lipids and a better affin- temperature. Moreover, for a temperature higher than 150 ◦ C,
ity between hexane and heated lipids. Flocs were destructurated apparent viscosity remained constant, for a given shear rate.
by thermal treatments. Then, the mixing between sludge and This comes along with modifications on sludge settleability.
solvent was more efficient and the transfer of lipids towards the Indeed, sludge volume index decreased with the rise in tem-
hexane phase was facilitated. Therefore, the total lipids con- perature. For sludge A (Fig. 6), SVI was almost 140 ml/g for
centration was better determined. At the same time, the soluble untreated sludge. It was equal to 47 ml/g for a temperature of
lipids concentration remained low: less than 0.5 g/l. That sug- 150 ◦ C, then it seemed stabilised around 36 ml/g for higher tem-
gests that lipids, due to their high hydrophobicity, were not peratures. Thus, thermal treatment enhanced sludge settleability.
solubilised in the aqueous phase. Nevertheless, VFA concentra- This is due to modification in sludge structure. Neyens et al. [26]
tion strongly increased with treatment: from 0.2 g/l for untreated explained this by EPS solubilisation. EPS are hydrated com-
sludge to almost 2.4 g/l for sludge treated at 170 ◦ C. We may sup- pounds which can absorb huge quantity of water. Thanks to EPS
pose that this increased was linked to lipids degradation [28]. solubilisation, a part of linked water is also released. Moreover,
Indeed, due to thermal treatment, long chains fatty acids may be it seemed that 150 ◦ C was a temperature threshold.
reduced to form fatty acids of lower molecular weights, which Beside, filterability of sludge was evaluated by CST (Fig. 7).
themselves may be degraded in low chain fatty acids (i.e. VFA) It varied with temperature of treatment. First, CST values
and acetic acid. But, due to the difficulty in determining the total increased: for instance, for sample A, it increased from 1300 s
lipids concentration, it was impossible to determine a solubil- to 2030 s for a temperature of 130 ◦ C. Then for temperature
isation rate of lipids. VFA production may also originate from higher than 150 ◦ C, CST strongly decreased and reached very
proteins degradation. low values: 31 s for a temperature of treatment of 190 ◦ C for
sample A and 13 s for samples D and E. Thus, for temperatures
3.2. Sludge physical properties lower than 130 ◦ C, filterability of sludge was deteriorated. This
could be linked to sludge solubilisation and the amount of small
Thermal treatment led to modifications on sludge character- particles. Nevertheless, this hypothesis has not been verified.
istics.
For instance, pH has also been measured during experiments
with sludge from B and C (Table 4). For both sludge samples,
pH values first increased (from 6.9 to 7.1 or 7.4 at 150 ◦ C) and
then decreased with temperature of treatment (to 6.4 or 7.1 at
170 ◦ C). The pH increase could be due to proteins desorption or
acidic compounds volatilisation [24]. And then, pH decreased

Table 4
pH variation for sludge samples B and C
Temperature (◦ C) 20 130 150 160 170

Sample B 6.9 7.15 7.4 7.8 7.1

Sample C 6.8 7.25 7.1 / 6.4 Fig. 6. Sludge volume index of sludge sample A after thermal treatment. The
point at 20 ◦ C corresponds to untreated sludge.
242 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244

Fig. 9. Biogas production from sludge particulate and soluble fractions, sludge
sample A.
Fig. 7. CST Measurement for sludge samples A, D and E after thermal treatment.
The points at 20 ◦ C correspond to untreated sludge samples.
cant effect on biogas composition. Indeed, methane content was
73 ± 3% for the reactor fed with untreated WAS, 74 ± 3% and
On the contrary, for temperatures higher than 150 ◦ C, filterabil-
73 ± 4% for reactors fed with sludge samples treated at 135 ◦ C
ity was improved. This could be explained by the modification
and 190 ◦ C, respectively.
of sludge structure and the released of linked water [26]. The
Some batch anaerobic digestion tests were carried out in order
observed threshold temperature of 150 ◦ C is in agreement with
to evaluate biogas volumes produced by the soluble and particu-
the one reported by Fisher and Swanwick [5].
late fractions of sludge samples. Three test bottles were realised
for each sample; a given volume of total sludge was introduced
3.2.1. Batch anaerobic digestion in the first one. The same volume of the sample was centrifuged.
For all sludge samples, all pre-treatments from 90 ◦ C to Sludge particulate fraction was introduced in the second bottle
210 ◦ C led to higher sludge biodegradabilities than the untreated and the soluble fraction was introduced in the last one. The sum
sludge ones, as shown in Fig. 8. Sludge biodegradability of biogas volumes produced by the particular soluble fractions
increased with pre-treatment temperature for temperatures up to was equal to the biogas volume produced by the total sludge, the
190 ◦ C and it slightly decreased after a pre-treatment of 210 ◦ C. error being less than 10%. Fig. 9 shows the repartition of biogas
This decrease in sludge biodegradability has already been volume between soluble and particular fractions. For untreated
observed for temperatures higher than 175 ◦ C which enhanced sludge, biogas was produced at 91% from particulate matter,
sludge solubilisation but not biogas production [8,29]. This was whereas for sludge sample treated at 190 ◦ C, only 43% of biogas
explained by the formation of inhibitory or toxic compounds. came from particulate fraction. Thus biogas volume produced
Moreover, Pinnekamp [30] observed a sharp decrease of biogas from the sludge soluble fraction increased with temperature pre-
production and ascribed it to the products of Maillard reac- treatment but this increase was more important for temperatures
tions. Indeed, reaction of carbohydrates with amino acids forms higher than 130 ◦ C.
melanoidins which are difficult or impossible to degrade. The biogas volume produced by each fraction is plotted ver-
Biogas composition was not measured in this study. However, sus the amount of introduced COD in Fig. 10 that shows that
another work [31] carried out with semi-continuous anaerobic biogas volume was proportional to the amount of introduced
reactors fed with untreated or treated WAS samples showed COD, except for the last point for soluble fraction. This point
that thermal pre-treatment (at 135 ◦ C or 190 ◦ C) had no signifi-

Fig. 8. Biodegradability enhancement (at 17 to 24 days of batch experiments) Fig. 10. Relationship between produced biogas volume and the amount of intro-
of pre-treated sludge. Sludge samples A, B, C, D and E. duced COD for particulate and soluble fractions, sludge sample A.
 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244 243

corresponded to 190 ◦ C treatment and showed the presence of

non biodegradable compounds in the soluble phase (notably,
products of Maillard reactions). Thus, for temperatures lower
than 170 ◦ C, 1 g of COD introduced in the test bottle led to the
production of 398 ml of biogas. The conversion was lower for
the particulate fraction as 1 g of COD led to the production of
279 ml of biogas. This shows that organic matter present in the
soluble fraction was more biodegradable than particulate mat-
ter. It has also to be noticed that data corresponding to untreated
sludge (lowest amount of soluble COD and highest amount of
particulate COD) could be represented by the same lines as the
data for pre-treated samples at temperatures lower than 170 ◦ C.
This means that biodegradability matter present in both fractions Fig. 11. Biodegradability enhancement (at 17 to 24 days of batch experiments)
was not significantly changed by thermal treatments (at temper- versus COD solubilisation and untreated sludge biodegaradability (Bo). Sludge
samples A, B, C, D and E.
atures lower than 170 ◦ C). The increase in biogas production
was due to the transfer of organic matter from the particulate
a high initial biodegradability and sludge from extended aera-
fraction to the soluble fraction.
tion (E) process had a poor initial biodegradability (29%). On
Fig. 8 gathers experimental biogas volume enhancements
the other hand, sludge samples from medium load processes had
obtained for the five studied sludge samples. We tried to analyse
either a high (C) or a low (D) initial biodegradability.
the impact of the sludge sample on the biogas volume enhance-
ment. In a first time, a PLS analysis was carried out to see the
correlation between different measured explicative variables. 4. Conclusion
Three sludge samples were taken into account (samples A, C and
D for which all the measures were available) and temperature, Thermal treatments are efficient to solubilise sludge. For tem-
COD solubilisation, TS solubilisation, VS solubilisation, pro- peratures lower than 200 ◦ C, COD solubilisation was found
teins solubilisation, carbohydrates solubilisation and untreated to increase linearly with treatment temperature and all the
sludge biodegradability (Bo) were considered. As expected, all different sludge samples behaved in the same way. For temper-
these parameters, except untreated sludge biodegradability were atures above 200 ◦ C, solubilisation results seemed to be more
correlated. Thus, COD solubilisation was kept as an overall dispersed. For temperatures lower than 150 ◦ C, carbohydrates
parameter to represent organic matter solubilisation. In a sec- solubilisation was more important than proteins solubilisation
ond time, a PLS regression was carried out considering the five as carbohydrates are located in exopolymers whereas proteins
sludge samples. The Y block was constituted of biogas vol- are mainly inside the cells. Moreover, carbohydrates concentra-
ume enhancements (27 points). The X block was constituted tion decreased at high temperature as they reacted with other
of sludge initial biodegradability (Bo) and COD solubilisation carbohydrates or solubilised proteins.
(SCOD ). The result was a dimension 2 model with an average Analyses of physical characteristic of sludge pointed out a
prediction error of 25%. threshold temperature of 150 ◦ C. Indeed, sludge dewaterability
was improved above this temperature but deteriorated for lower
Biogas volume enhancement temperatures. Moreover, sludge apparent viscosity and sludge
volume indexes were reduced first with temperature and then
= 2.156 + 1.155 SCOD − 2.348 Bo (2) remained almost constant for temperatures higher than 150 ◦ C.
The results of this model are plotted in Fig. 11, which shows the Thermal treatments up to 190 ◦ C allowed the biogas pro-
good agreement between experiment and calculated data. Thus, duction to increase during batch anaerobic digestion of sludge.
for temperatures lower or equal to 170 ◦ C, the impact of thermal Biogas volume enhancement was linked to sludge COD solubil-
treatments on batch anaerobic biogas production can be sim- isation and to untreated sludge initial biodegradability, the lower
ply represented by considering sludge COD solubilisation and the initial biodegradability, the higher is the impact of thermal
untreated sludge initial biodegradability. The higher the COD treatment.
solubilisation, the higher is the biogas production. Besides, the
impact of pre-treatment was more important when it was applied Acknowledgement
to sludge samples with a low initial biodegradability. However,
untreated sludge biodegradability did not seem to be linked to Authors want to thank ADEME for financial contributions.
the sludge samples compositions. Indeed, sludge samples A and
D had the same organic solids concentration (81–82% of TS) and References
their initial biodegradabilities were 53% and 33%, respectively.
In the same way, sludge samples B and E had an organic solids [1] Y.Y. Li, T. Noike, Upgrading of anaerobic digestion of waste activated
sludge by thermal pretreatment, Water Sci. Technol. 26 (1992) 857–866.
concentration of 76% and initial biodegradabilities of 48% and [2] U. Kepp, I. Machenbach, N. Weisz, O.E. Solheim, Enhanced stabilization
29%, respectively. If wastewater treatment plants processes are of sewage sludge through thermo hydrolysis – three years of experience
considered, sludge from both high load processes (A and B) had with full scale plant, Water Sci. Technol. 42 (9) (2000) 89–96.
244 C. Bougrier et al. / Chemical Engineering Journal 139 (2008) 236–244

[3] J. Chauzy, D. Cretenot, L. Patria, P. Fernandez, P. Sauvegrain, J.P. Lev- [17] C. Bougrier, C. Albasi, J.P. Delgenès, H. Carrère, Effect of ultrasonic,
asseur, Bio Thelys: a new sludge reduction process. Biosolids 2003, thermal and ozone pre-treatments on waste activated sludge solubilisation
wastewater sludge as a resource, Trondheim, 23–25 June 2003, pp. and anaerobic biodegradability, Chem. Eng. Proc. 45 (8) (2006) 711–718.
473–479. [18] C. Bougrier, J.P. Delgenès, H. Carrère, Combination of thermal treatments
[4] E. Neyens, J. Baeyens, A review of thermal sludge pre-treatment processes and anaerobic digestion to reduce sewage sludge quantity and improve
to improve dewaterability, J. Hazard. Mater. 98 (1–3) (2003) 51–67. biogaz yield, Process Saf. Environ. Protect. 84 (B4) (2006) 280–284.
[5] R.A. Fisher, S.J. Swanwick, High temperature treatment of sewage sludge, [19] American Public Health Association, American Water Works Association,
Water Pollut. Control 71 (3) (1971) 255–370. Water Pollution Control Federation, Standard methods for the examination
[6] N.J. Anderson, D.R. Dixon, P.J. Harbour, P.J. Scales, Complete character- of water and wastewater, in: L.S. Clesceri, A.E. Greenberg, R.R. Trussel
isation of thermally treated sludges, Water Sci. Technol. 46 (10) (2002) (Eds.) 18th ed., APHA, 1992.
51–54. [20] O.H. Lowry, N.J. Rosebrough, A.L. Fau, R.J. Randall, Protein measurement
[7] R.T. Haug, D.C. Stuckey, J.M. Gossett, P.L. Mac Carty, Effect of thermal with the Folin reagent, J. Biol. Chem. 193 (1951) 265–275.
pretreatment on digestibility and dewaterability of organic sludges, J. Water [21] R. Dreywood, Qualitative test for carbohydrate material, Ind. Eng. Chem.
Pollut. Control Fed. (1978) 73–85. Res. 18 (1946) 199.
[8] D.C. Stuckey, P.L. Mc Carty, Thermochemical pretreatment of nitrogenous [22] G. Bridoux, P. Dhulster, J. Manem, Greases analysis in waste water treate-
materials to increase methane yield, Biotechnol. Bioeng. Symp. 8 (1978) ment plant, Techniques-Sciences-Méthodes 5 (mai 1994) (1994) 257–262
219–233. (in French).
[9] S. Tanaka, T. Kobayashi, K.I. Kamiyama, L.N. Bildan, Effects of thermo- [23] J.F. Durand, Partial least-squares regression: from linearity to nonlinear
chemical pretreatment on the anaerobic digestion of waste activated sludge, additive spline modeling Web site: http://jf-durand-pls.com/.
Water Sci. Technol. 35 (8) (1997) 209–215. [24] M. Murillo Murillo, Characterisation of the effect of a hydrogen peroxide
[10] C. Fjordside, An operating tale from Naeested Sewage Treatment plant, treatment of a sludge. Application to the reduction of sludge production,
Municipal Wastewater Treatment Nordic Conference, Copenhagen, 17–19 Ph.D. thesis, INSA Toulouse, 2004 (in French).
January 2001. [25] P. Camacho, Study of sludge reduction processes by combibingchemi-
[11] H.N. Gavala, U. Yenal, I.V. Skiadas, P. Westermann, B.K. Ahring, cal or physical treatments and biological treatments, Ph.D. thesis, INSA
Mesophilic, thermophilic anaerobic digestion of primary and secondary Toulouse, 2001 (in French), cited by Murillo-Murillio [24].
sludge. Effect of pre-treatment at elevated temperature, Water Res. 37 [26] E. Neyens, J. Baeyens, R. Dewill, B. De Heyder, Advanced sludge treat-
(2003) 4561–4572. ment affects extracellular polymeric substances to improve activated sludge
[12] M. Barjenbruch, O. Kopplow, Enzymatic, mechanical and thermal pre- dewatering, J. Hazard. Mater. 106B (2004) 83–92.
treatment of surplus sludge, Adv. Environ. Res. 7 (3) (2003) 715–720. [27] J. Barlindhaug, H. Odegaard, Thermal hydrolysis for the production of
[13] J. Kim, C. Park, T.H. Kim, M. Lee, S. Kim, S.W. Kim, J. Lee, Effects of var- carbon source for denitrification, Water Sci. Technol. 34 (1–2) (1996)
ious pretreatments for enhanced anaerobic digestion with waste activated 371–378.
sludge, J. Biosci. Bioeng. 95 (3) (2003) 271–275. [28] A. Shanableh, S. Jomaa, Production and transformation of volatile fatty
[14] M. Dohanyos, J. Zabranska, J. Kutil, P. Jenicek, Improvement of anaerobic acids from sludge subjected to hydrothermal treatment, Water Sci. Technol.
digestion of sludge, Water Sci. Technol. 49 (10) (2004) 89–96. 44 (10) (2001) 129–135.
[15] A. Valo, H. Carrère, J.P. Delgenès, Thermal, chemical and thermo-chemical [29] R.T. Haug, T.J. Le Brun, D. Tortorici, Thermal pretreatment of sludges: a
pre-treatment of waste activated sludge for anaerobic digestion, J. Chem. field demonstration, J. Water Pollut. Control Fed. 55 (1) (1983) 23–34.
Technol. Biotechnol. 79 (2004) (2004) 1197–1203. [30] J. Pinnekamp, Effect of thermal pretreatment of sewage sludge on anaerobic
[16] S. Graja, J. Chauzy, P. Fernandes, L. Patria, D. Cretenot, Reduction of digestion, Water Sci. Technol. 21 (1989) 97–108.
sludge production from WWTP using thermal pretreatment and enhanced [31] C. Bougrier, J.P. Delgenès, H. Carrère, Impact of thermal pre-treatments
anaerobic methanisation, Water Sci. Technol. 52 (1–2) (2005) 267– on the semi-continuous anaerobic digestion of waste activated sludge,
273. Biochem. Eng. J. 34 (2007) 20–27.