Production of lightweight aggregates from mining and industrial wastes

B. Gonza lez-Corrochano
a
, J. Alonso-Azca rate
a,
*
, M. Rodas
b
a
Department of Chemistry-Physics, School of Environmental Sciences, University of Castilla-La Mancha, Avenida Carlos III, s/n, 45071 Toledo, Spain
b
Department of Crystallography and Mineralogy, School of Geological Sciences, Complutense University of Madrid, Ciudad Universitaria, 28040 Madrid, Spain
a r t i c l e i n f o
Article history:
Received 23 July 2008
Received in revised form
24 February 2009
Accepted 19 March 2009
Available online 21 April 2009
Keywords:
Lightweight aggregate
Washing aggregate sludge
Sewage sludge
Bloating
a b s t r a c t
Washing aggregate sludge from a gravel pit, sewage sludge from a wastewater treatment plant (WWTP)
and a clay-rich sediment have been physically, chemically and mineralogically characterized. They were
mixed, milled and formed into pellets, pre-heated for 5 min and sintered in a rotary kiln at 1150

C,
1175

C, 1200

C and 1225

C for 10 and 15 min at each temperature. The effects of the raw material
characteristics, heating temperatures and dwell times on the loss on ignition (LOI), bloating index (BI),
bulk density (r
b
), apparent and dry particle densities (r
a
, r
d
), voids (H), water absorption (WA
24h
) and
compressive strength (S) were determined. All the mixtures presented a bloating potential taking into
consideration the gases released at high temperatures. The products obtained were lightweight
aggregates (LWAs) in accordance with Standard UNE-EN-13055-1 (r
b
_1.20 g/cm
3
or particle
density _2.00 g/cm
3
). LWAs manufactured with 50% washing aggregate sludge and 50% clay-rich sedi-
ment were expanded LWAs (BI >0) and showed the lowest apparent particle density, the lowest water
absorption and the highest compressive strength. It was possible to establish three groups of LWAs on
the basis of their properties in comparison to Arlita G3, F3 and F5, commercially available lightweight
aggregates manufactured in Spain. Our LWAs may have the same or similar applications as these
commercial products, such as horticulture, prefabricated lightweight structures and building structures.
2009 Elsevier Ltd. All rights reserved.
1. Introduction
In industrial and mining activities, the elimination of wastes is
one of the main concerns (Blanco, 2004). Washing aggregate sludge
is a siltclay waste that is generated in the classication cycle of
sand and gravel. The accumulation of washing aggregate sludge in
rafts is the most widely used method because it is the most
economical, although this process has numerous limitations and
disadvantages (e.g. large areas are required for the storage of
wastes) (Blanco, 2004).
Sewage sludge is the nal waste from the water purication
process in treatment plants. It is decomposed and stabilized in
order to reduce the odor, and often undergoes treatment for
disposal and destruction. Solidication/stabilization (S/S) has
been successfully applied to several industrial wastes. However,
this technology does not always produce satisfactory
results, since each waste responds differently to the S/S (Sali-
hoglu et al., 2007).
Therefore, the study and recovery of these wastes is a challenge
and an alternative that represents a future course of action.
The reutilization of industrial wastes and the use of recycled
materials in construction applications have been a common prac-
tice and have grown worldwide over the last decades (Gao et al.,
2008; Sanchez et al., 2009). Heat treatment at high temperatures
can convert some types of wastes into ceramic products with
a wide range of microstructural features and properties (Bethanis
et al., 2004).
A lightweight aggregate (LWA) is a granular material with
a bulk density (r
b
) not exceeding 1200 kg/m
3
(1.20 g/cm
3
) or
with a particle density not exceeding 2000 kg/m
3
(2.00 g/cm
3
)
(UNE-EN-13055-1, 2003). LWAs can be manufactured from
materials extracted from different types of geological materials,
such as igneous rocks like pumice and tuffs (De Gennaro et al.,
2004; Riley, 1951), sedimentary and metamorphic rocks
(claystone, slate and shale; Conley et al., 1948; Dahab, 1980;
Decleer and Viaene, 1993). They can also be manufactured from
wastes and industrial by-products such as glass and y ash
(Baykal and Doven, 2000; Kim et al., 2005; Videla and Martinez,
2002).
Articial lightweight aggregates are formed by the rapid heating
at high temperature of materials that have the ability to expand.
* Corresponding author. Tel.: 34 625168760; fax: 34 925268840.
E-mail address: jacinto.alonso@uclm.es (J. Alonso-Azca rate).
Contents lists available at ScienceDirect
Journal of Environmental Management
j ournal homepage: www. el sevi er. com/ l ocat e/ j envman
0301-4797/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jenvman.2009.03.009
Journal of Environmental Management 90 (2009) 28012812
Two conditions are necessary to achieve an appropriate expanded
material (Riley, 1951):
(1) it must contain substances that release gases at high tempe-
rature (T
1
),
(2) a plastic phase with the appropriate viscosity must be produced
so that it can trap the gases released at temperature T
1
.
Most natural aggregates have a particle density between 2.4 and
2.8 g/cm
3
, typically 2.6 g/cm
3
, while lightweight aggregates have
a particle density between 0.8 and 2.0 g/cm
3
(Cheeseman and Virdi,
2005). They have become a focus of interest because the low
particle density and the low bulk density entail a decrease in the
load transmitted to the ground, and less work and effort are
required to transport them (De Gennaro et al., 2004). The benets
associated with these lowdensities, which are due to the formation
of voids and pores, are very good thermal and acoustic insulation
and materials with a good resistance to re (Benbow, 1987; Fakh-
fakh et al., 2007).
The objective of this research is to recycle different types of
wastes, including sewage sludge and washing aggregate sludge, in
order to obtain a usable material, such as lightweight aggregates,
and to ensure that they are of good quality for different
applications.
2. Raw materials
Washing aggregate sludge was taken from a representative
gravel pit located along the middle sector of the river Tagus. The
aggregates were extracted from the ood plain and the lower
terrace deposits using a dry method. The sediments are made up of
conglomerates and sandstones with small amounts of silt and clay
in the matrix. The <5-mm sieve undersize was washed in a gravi-
tational hydraulic classier; the undersize particles (silt and clay
wastes) were removed with the overow and transported through
channels to ponds, where they settled by gravity. Samples collected
from the sediment rafts were transported to the laboratory in
plastic bags and dried at 60

C for several days.
The sewage sludge came from the wastewater treatment plant
(WWTP) located in the Technological Campus of Toledo, University
of Castilla-La Mancha (Spain). The samples were dried at 60

C for
approximately one week, as they had a very high humidity.
Due to the high carbonate content of the washing aggregate
sludge, this material was also mixed with a clay-rich sediment in
order to control the effect of the carbonate content. This natural
sediment is a miocene lutite from the Red Clays of the La Sagra
formation (Garca, 2004). Approximately 50 kg of material was
taken to the laboratory in plastic containers. The initial treatment in
the laboratory was manual disaggregation of the sediment, which
was left to dry at roomtemperature for several days onplastic trays,
and subsequently in an oven at 60

C.
3. Methods
3.1. Characterization of raw materials
The particle size distribution was analyzed by laser diffraction
(Tay et al., 2000) in the range between 0.04 mm and 2000 mm with
a Coulter

LS 230 equipment.
After fusion with lithium metaborate ux at 1000

C and acid
dissolution (Ingamells, 1970), the chemical composition of the
samples was analyzed in triplicate using inductively coupled
plasma-atomic emission spectroscopy (ICP-AES, Thermo Electron
6500 ICAP) (De Gennaro et al., 2004; Lin et al., 2008). The loss on
ignition (LOI
rm
) during fusion was evaluated in triplicate as the
percentage weight difference in samples heated at 1000

C, for
45 min.
The composition of the raw materials was plotted in the ternary
Riley diagram (Riley, 1951) in order to evaluate the bloating
behavior of the samples. The SiO
2
/Flux ratio (Si/
P
F) (uxing
elements =CaOMgOK
2
ONa
2
OFeOFe
2
O
3
) was deter-
mined (Fakhfakh et al., 2007).
The carbon content (total organic TOC and inorganic IC) of
the raw materials was estimated using a Shimadzu

TOC-V
CSH
analyzer.
The qualitative and semi-quantitative mineralogical analysis of
the washing aggregate sludge was performed by X-ray diffraction
(XRD, Bethanis et al., 2002; Cheeseman and Virdi, 2005). The bulk
mineralogy was determined using the polycrystalline disoriented
powder method. The ne and medium silt fraction mineralogy and
the clay fraction mineralogy were studied through the preparation
of oriented aggregates treated with ethylene glycol and heated to
550

C for 2 h. Carbonates were eliminated by an acid treatment
(Jackson, 1969) and, subsequently, grain-size fractions were
obtained by sedimentation.
X-ray diffractograms were performed with a Phillips XPert MPD
diffractometer, using CuKa radiation, programmable divergence
slit, a graphite monochromator and a sealed xenon gas proportional
detector. The conditions used during the measurement were an
accelerating voltage of 40 kV and a current of 40 mA.
3.2. LWA manufacturing
The raw materials were mixed at four levels of washing aggre-
gate sludgesewage sludge mixing proportions, named W100,
W75S25, W50S50 and W25S75, corresponding to 100%, 75%, 50%
and 25% of washing aggregate sludge by weight, respectively. A
mixture of washing aggregate sludgeclay-rich sediment, named
W50C50, with 50% of clay-rich sediment by weight, was also
prepared.
The mixtures were milled to a grain size of less than 200 mm
(Yasuda, 1991) and homogenized using a Restch

SK 100/C Spe-
zialstahl arm mill.
The plasticity of the mixtures was studied by Atterberg limits
calculation, using the procedure specied in Standards UNE-103-
103 (1994) and UNE-103-104 (1993).
A controlled amount of water (23% above the plastic limit) was
added to the ground mixtures and they were extruded using
a Nannetti

pneumatic extruder. Cylinders 1.5 cm in length were

cut and rolled by hand until spherical pellets with an approximate
diameter of 810 mm were obtained. The green pellets were dried
for 48 h at room temperature and, subsequently, placed in an oven
at 105

C for 48 h.
Prior to sintering the dry pellets in the rotary kiln, the following
tests were carried out.
- Determination of the melting behavior of the base powder
using a Misura

heating microscope. The heating rate was

20

C/min, in an air atmosphere, until melting was achieved.
- Simultaneous thermogravimetrydifferential thermal analysis
(TGDTA). Changes in weight and energy of approximately
20 mg of the mixtures were checked using a Setaram

ther-
moanalyzer equipment. The operating conditions were an air
atmosphere and a heating rate of 10

C/min (Ducman et al.,
2002) in a temperature range from 25

C to 1300

C (Man-
gialardi, 2001).
The results obtained from these tests contributed to deciding
the temperature and dwell time conditions during the subsequent
sintering of the lightweight aggregates in the rotary kiln.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2802
Sintering of the pellets was carried out in a tubular rotary kiln
(TO-R 120-14, Nannetti

), which has an Al
2
O
3
refractory tube with
an internal diameter of 6 cm, an external diameter of 7 cm and
a length of 120 cm. Three zones can be differentiated in the tube on
the basis of the temperature reached: a central area (2, heating area),
where the temperature is highest (temperature selected by the
user), and two lateral areas, one located at the entrance of the tube
(1, pre-heating area) and one located at the exit (3, cooling area). The
temperature of zones 1 and 3 is lower than that of the central zone.
There were two distinct steps in the heating of about 300
granules of each mixture (in groups of 25 pellets).
- Burnout. In order to prevent the deterioration of the dried balls
subjected to rapid ash heat, and avoid their explosion and
black core formation during the nal ring process, pre-treat-
ment in zone 1 of the rotary kiln, about 20 cm from the entry,
was maintained for 5 min, with temperatures of 1150

C,
1175

C, 1200

C and 1225

C in zone 2. As may be observed in
Fig. 1, the temperature about 20 cm from the entrance ranges
between 300

C and 450

C, depending on the selected
temperature for zone 2. The loss onignitionof the pellets inzone
1 (LOI
z1
) was calculated as the percentage weight difference.
- Sintering. Subsequently, the pellets passed on to zone 2 of the
rotary kiln at 1150

C, 1175

C, 1200

C and 1225

C (T
heating
) for
10 and 15 min at each temperature, in order to achieve
a viscosity state in the mass of the lightweight aggregate that
can trap the internally generated gases.
The speed of rotation was xed at 2.2 rpm. During the burnout
and sintering steps, the tilt was kept constant at 0

. The granules
were moved from the entrance through zones 1, 2 and 3 due to
slight increases in the tilt angle of the rotary tube.
3.3. Physical property tests
The loss on ignition during heating in the rotary kiln (LOI
10/15
) of
all the compositiontemperaturetime lightweight aggregates was
calculated using the following equation:
LOI
XYZ
= 100 (Msc Mc)=Msc
where,
- LOI
XYZ
(%) is the loss on ignition of pellets with composition X,
heated at temperature Y for a dwell time of Z minutes.
- Mc (g) is the average mass of 25 pellets with composition X,
heated at temperature Y for a dwell time of Z minutes.
- Msc (g) is the average mass of 25 pellets with composition X,
prior to sintering.
The total loss on ignition (LOI
t
) of the pellets was evaluated as
the percentage weight difference in samples heated in a mufe at
1150

C, 1175

C, 1200

C and 1225

C for 24 h.
The Bloating Index (BI) was expressed as the volume change
after ring, in accordance with the equation BI =100 (d
2
d
1
)/d
1
,
where d
1
and d
2
are the diameters of the pellets before and after
ring, respectively (Fakhfakh et al., 2007).
The bulk density (r
b
, expressed in g/cm
3
) was calculated as the
W/V ratio specied by Standard UNE-EN-1097-3 (1999), where Wis
the weight of the aggregates in a container with volume V.
The particle density (apparent and dry, expressed in g/cm
3
) and
the water absorption of 25 sintered pellets after 24 h were deter-
mined using an established procedure described in UNE-EN-1097-6
(2000), by means of a pycnometer. According to this standard:
- the apparent particle density (r
a
) is the ratio between the mass
of a sample of aggregates dried in an oven and the volume
these aggregates occupy in water, including the watertight
internal pores and excluding the pores available to water;
- the dry particle density (r
d
) is the ratio between the mass of
a sample of aggregates dried in an oven and the volume these
aggregates occupy in water, including the watertight internal
pores and the pores available to water.
The equations used were:
Apparent particle density : r
a
= M
4
=[M
4
(M
2
M
3
)[
Dry particle density : r
d
= M
4
=[M
1
(M
2
M
3
)[
Water absorption : WA
24h
= 100 (M
1
M
4
)=M
4
where:
- M
1
is the mass (g) of saturated pellets on the dry surface.
- M
2
is the mass (g) in the pycnometer containing saturated
pellets and water.
- M
3
is the mass (g) in the pycnometer containing water.
- M
4
is the mass (g) of the dry sample, prior to immersion in
water.
0
200
400
600
800
1000
1200
Distance to the entrance (cm) and zone
T
e
m
p
e
r
a
t
u
r
e

(

C
)
0 (entry) 20(z1) 40(z1) 60(z2) 80(z3) 100(z3) 120(z3)
Tsl=1150C Tsl=1175C
Tsl=1200C Tsl=1225C
Fig. 1. Temperature gradient in the tube of the rotary kiln. z1, z2, z3: zone 1 (pre-
heating area), zone 2 (heating area) and zone 3 (cooling area), respectively. Tsl: selected
temperature by the user.
Fig. 2. Grain size distribution curves of the raw materials.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2803
The void percentage (H) (air space between aggregates in
a container) was calculated using the method specied in Standard
UNE-EN-1097-3 (1999):
H = 100 (r
d
r
b
)=r
d
where,
- H is the void percentage (%), r
b
is the bulk density and r
d
is the
dry particle density of the sample.
The compressive strength (S) of a sphere that is tested on two
diametrically opposed points is given by (Li et al., 2000; Yashima
et al., 1987):
S = (2:8 Pc)=

pX
2

where,
- X is the distance between the loading points (sphere diameter)
and Pc is the rupture load.
The fracture force of a single aggregate (expressed in N/mm
2
)
was measured by pressing the aggregate in a Nannetti

FM 96
press until it cracked. The mean compressive strength values
were calculated fromtests carried out on 25 pellets prepared at
each compositiontemperaturetime.
These physical properties were compared to a Spanish
commercial LWA, G3 variety of Arlita

.
3.4. Analyses and processing of results
A contrast independence test was carried out in order to
determine whether signicant relationships exist between the
variables.
4. Results and discussion
4.1. Characterization of raw materials
Grain size analyses indicate that the washing aggregate sludge
and the clay-rich sediment are relatively ne. The sewage sludge
has a coarser grain size and a smaller clay fraction (Fig. 2).
The chemical analyses and the mineralogy of the bulk samples
are shown in Tables 1 and 2, respectively. Silica and alumina oxides
are the main constituents of the washing aggregate sludge and the
clay-rich sediment; calcium oxide is also signicant in the rst
sample. The SiO
2
, Al
2
O
3
and FeOFe
2
O
3
contents mostly corre-
spond to quartz and phyllosilicates present in the samples; the CaO
Table 1
Chemical composition of the raw materials and mixtures (M) studied.
Samples SiO
2
Al
2
O
3
FeO Fe
2
O
3
CaO MgO K
2
O Na
2
O SO
3
TiO
2
LOI
rm
a
TOC
b
IC
c
P
F Si/
P
F
d
W100 53.59 10.48 1.34 1.49 13.03 1.74 2.51 0.81 0.02 0.44 15.72 0.17 2.82 20.93 2.56
W75S25 (M) 43.90 17.49 1.19 1.32 10.67 1.54 2.05 0.75 0.06 0.36 24.39 1.82 2.46 17.52 2.51
W50S50 (M) 34.22 24.50 1.04 1.15 8.30 1.34 1.58 0.70 0.10 0.28 33.73 2.87 1.98 14.10 2.43
W25S75 (M) 24.53 31.51 0.88 0.98 5.94 1.14 1.11 0.64 0.14 0.21 47.31 4.97 1.18 10.69 2.29
Sewage sludge 14.84 38.52 0.73 0.81 3.57 0.93 0.64 0.58 0.18 0.13 51.19 7.39 0.36 7.28 2.04
W50C50 (M) 53.35 14.79 2.25 2.50 6.80 3.12 3.29 0.80 0.01 0.65 12.68 0.00 1.50 18.78 2.84
Clay-rich sediment 53.11 19.09 3.17 3.52 0.57 4.50 4.07 0.79 0.00 0.86 7.58 0.00 0.00 16.62 3.20
a
Loss on ignition of raw materials.
b
Organic carbon content.
c
Inorganic carbon content.
d
SiO
2
/
P
F (CaOMgOK
2
ONa
2
OFeOFe
2
O
3
; Riley, 1951).
Table 2
Mineralogical composition of the washing aggregate sludge and clay-rich sediment studied. TR =minor trace component; =120%; =2140%; =4160%;
=6180%; =81100%.
Sample Bulk mineralogy Fine and medium silt fraction Clay fraction
Qtz
a
Fl
b
Cal
c
Dol
d
Phi
e
Ill
f
Kao
g
Sm
h
Ill Kao Sm
W100 TR
Clay-rich sediment TR TR TR
a
Quartz.
b
Feldspar.
c
Calcite.
d
Dolomite.
e
Philosilicates.
f
Illite.
g
Kaolinite.
h
Smectite.
100% SiO
2
50% Al
2
O
3
50% Fluxing
elements
W100
W75S25
W50C50 60%
70%
80%
90%
Fig. 3. Representation of raw materials composition in Rileys (1951) ternary diagram
showing the area of bloating.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2804
comes from the carbonate minerals. Na
2
O and K
2
O contents are
mainly attributed to the feldspars and the interlayer of clay
minerals (Fakhfakh et al., 2007).
In the sewage sludge, the main compound is Al
2
O
3
. It comes
from the coagulant added in the water purication process
(aluminum polychloride, 40%). The sludge is formed by amorphous
compounds and/or organic matter. The high loss on ignition (LOI
rm
)
of the sample is mainly attributed to the oxidation of this organic
matter.
The chemical data of the ground mixtures (recalculated
to 100%) were plotted on the SiO
2
Al
2
O
3
-uxing elements
(CaOMgOK
2
ONa
2
OFeOFe
2
O
3
) diagram (Riley, 1951).
The raw materials with a chemical composition located within the
bloating area are mixtures W50C50 and W75S25 (Fig. 3), hence it
is expected that they form a sufcient viscous phase to be able to
trap a signicant amount of gas and, therefore, be good source
materials for LWA manufacturing (Hung and Hwang, 2007).
The SiO
2
/
P
F ratios and the carbon content of the raw materials
and mixtures are shown in Table 1. All the samples have ratios
above 2, which is the minimum required to obtain a suitable
viscosity during the ring process (Fakhfakh et al., 2007).
The organic components in the washing aggregate sludge are in
trace amounts, with a TOC value of 0.17%. On the contrary, there is
a high content of inorganic carbon, coming from calcite and
dolomite.
The clay-rich sediment shows no carbon, because it does not
contain carbonate minerals and/or organic matter. The sewage
sludge has more than 7% organic carbon and only 0.36% inorganic
carbon, because it is mainly composed of organic matter (faeces,
waste paper, etc.).
4.2. LWA manufacturing
Fig. 4 shows the plastic behavior of the wastes and mixtures in
Casagrandes diagram. W100 exhibits an appropriate behavior,
which is almost optimal for extrusion (Casagrande, 1932, 1948).
W50C50 is located outside but very close to area A (appropriate
behavior for extrusion), while the rest of the mixtures are outside
and more distant from the two areas. The good extrusion behavior
of W100 and W50C50 is related to the high clay fraction content.
The presence of plate-like morphology particles, such as clay
minerals, makes it possible for them to slide over each other due to
the water retained in the interstitial spaces, which acts as a lubri-
cant (Ribeiro et al., 2005). Despite the fact that the mixtures
containing sewage sludge (W75S25, W50S50 and W25S75) are
more distant fromareas A and B, they can be perfectly extruded. It is
difcult to exactly determine which factors most inuence the
location of the mixtures with respect to the areas in Casagrandes
diagram, as the plastic behavior of an extrudable ceramic paste is
controlled by different factors (Benbow et al., 1987): (i) volumetric
fraction, shape and size distribution of the particles, (ii) type of
dominant interparticle forces, (iii) surface chemistry, (iv) packing
density and (v) rheological behavior of the liquid phase.
The results of the heating microscope test are summarized in
Table 3. No expansion was observed. This does not necessarily
mean that the dry pellets will not bloat, because the continuous
temperature increase in the heating microscope (20

C/min)
partially hides the real potential for expansion, as the samples do
not have enough time to complete the bloating before melting
begins (De Gennaro et al., 2007).
Thermogravimetrydifferential thermal curves of the milled
washing aggregate sludge and the milled mixtures are shown in
Fig. 5. The initial weight loss (TG lines) from ambient temperature
to 120

C is due to the evaporation of the remaining moisture
(Bethanis et al., 2002). Studies on other organic-containing wastes
have reported weight losses up to 550

C resulting from the release
of occluded gases, primarily CO
2
(Zheng and Kozinski, 2000). The
DTA lines of all the samples show a low-temperature endothermal
peak at around 110

C140

C (P1) due to the expulsion of water
absorbed by the clay. Awide range exothermic effect, with one peak
at around 350

C (P2) (Fig. 5b and c) or two peaks at around 350

C
(P2) and 430

C (P2
/
) (Fig. 5d), results from the oxidation of organic
matter; this exothermic effect is not present in the washing
aggregate sludge (Fig. 5a) or in mixture W50C50 (Fig. 5e), because
their organic matter content is very low or zero (Table 1). The
endothermic peak at around 520

C (P3) in W100 (Fig. 5a) and in
the mixture with the clay-rich sediment (Fig. 5e) is caused by the
dehydroxylation of the clay minerals, such as kaolinite and illite
(Chopra et al., 1964) (Table 2). The endothermic peak at 575

C (P4)
could be due to the transformation from a- to b-quartz (Chopra
et al., 1964), and the sharp endothermic peak at 750

C (P5)
(associated with a signicant weight loss) in sample W100, which
decreases as the sewage sludge in the mixture increases, is due to
the decarbonation of calcium carbonate. Pure calcium carbonate
shows an endothermic peak at around 850

C900

C, but, in the
presence of impurities, the temperature of the peak is known to
decrease by 100

C (Chopra et al., 1964). The last endothermic peak,
at approximately 1140

C (P6), is believed to be due to the gas
evolution produced in the reaction of Fe
2
O
3
. The reaction involved
is represented by 6Fe
2
O
3
/4Fe
3
O
4
O
2
(Huang et al., 2007).
The above results suggest that the potential sources of gas
evolution in articial aggregates could be one or more of the
following: (a) the clay mineral itself (absorbed as well as chemically
held water), (b) organic matter, (c) calcium carbonate, (d) ferric
oxide and (e) entrapped air.
W100 could be cooked at 1150

C and at 1175

C, but not up to
1200

C, because at this temperature softening starts (Table 3).
Fig. 4. Representation of wastes and mixtures in Casagrades diagram. Solid line (area
A) and dashed line (area B) mark theoretically appropriate and optimal elds,
respectively, for extrusion moulding (Casagrande, 1932, 1948).
Table 3
Result of fusibility tests. Characteristic temperatures (in

C) of: Ts =sintering,
Tr =starting to soften, Tme =maximumbloating, Taf =advanced fusion, Tf =fusion,
TafTs =thermal wide suitable for ring. (M) =mixture, n.d =not determined,
N.B =no bloating.
Sample Ts Tr Tme Taf Tf TafTs
W100 1195 1200 N.B 1235 1255 40
W75S25 (M) 1195 n.d N.B 1285 1295 90
W50S50 (M) 1150 n.d N.B 1395 1400 245
W25S75 (M) 905 1265 N.B 1345 1355 440
W50C50 (M) 1175 1245 N.B 1295 1310 120
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2805
W50C50 pellets could not be cooked to 1225

C despite the fact
that its starting softening temperature (Tr) is 1245

C; this could be
due to the fact that dwell times in the rotary kiln are much
higher (1015 min) than dwell times in the heating microscope
(heat rate =20

C/min). The aggregates manufactured with
sewage sludge could be subjected to all temperaturedwell
time conditions, as they have Taf values higher than the ring
temperatures.
4.3. Characterization of the articial lightweight aggregates
Fig. 6 shows the results of the loss on ignition in pre-heating
zone 1 (LOI
z1
), at 10 and 15 min of heating in the rotary kiln (LOI
10
,
LOI
15
), and the total loss on ignition in the mufe (LOI
t
).
LOI
z1
is in all cases much lower than LOI
10
, LOI
15
and LOI
t
,
because the temperature and the dwell time (5 min) in zone 1 are
lower than those in the heating area of the rotary kiln (at least
0 400 800
HeatFlow/V
Exo
-16
-12
-8
-4
0
TG-C/%
-8
-4
0
T/C
1200
P1
(100C)
P3
(520C)
P4
(575C)
P6
(1150C)
P5
(750C)
0 400 800
HeatFlow/V
Exo
-10
0
10
TG-C/%
-60
-50
-40
-30
-20
-10
0
1200
T/C
P1
(150C)
P2 (350C)
P2 (430C)
P5
(750C)
0 400 800
HeatFlow/V
Exo
-10
-5
0
5
10
TG-C/%
-30
-20
-10
0
1200
T/C
P1
(150C)
P2 (350C)
P5
(750C)
0 400 800
HeatFlow/V
Exo
-15
-10
-5
0
5
10
-20.
-15.
-10.
-5.
0.0
TG-C/%
1200
T/C
P1 (140C)
P2 (350C)
P5
(750C)
P6
(1175C)
0 400 800
HeatFlow/V
Exo
-12
-6
TG-C/%
-20
-10
0.0
1200
P1
(110C)
P3
(520C)
P5
(750C)
P6
(1140C)
W75S25
W25S75 W50S50
W50C50
W100
0
a b
c d
e
Fig. 5. Results of simultaneous thermogravimetrydifferential thermal analysis (TGDTA). P1: expulsion of adsorbed water on clay; P2 and P2
/
: oxidation of organic matter; P3:
dehydroxylation of the clay minerals; P4: transformation of a- to b-quartz; P5: decarbonation of calcium carbonate; P6: reaction of Fe
2
O
3
. Dashed line: TG lines; solid line: DTA line.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2806
10 min and 1150

C) and those in the mufe (at least 24 h and
1150

C).
In order to obtain a thermally expandable material, a suitable
chemical composition that inuences the softening at high
temperature, as well as the presence of substances able to develop
gas (such as water, carbon dioxide, oxygen, etc.), is required (Riley,
1951). In addition, these substances must release the gas above the
pre-treatment temperatures and/or must form intermediate pha-
ses which may retain gas and release it around the swelling
temperature (Fakhfakh et al., 2007).
LOI results show that this release of gases is possible after the
pre-heating treatment, when all the aggregates are introduced in
zone 2 of the rotary kiln, but, among the samples with good
chemical characteristics (inside the swollen area of Rileys diagram,
Fig. 3), expanded aggregates (BI >0) were only obtained from
mixture W50C50 heated at 1200

C (Fig. 7).
Previous studies (Fakhfakh et al., 2007) also show that some
samples which meet the two conditions necessary to achieve an
appropriate expanded material (according to Riley, 1951) do not
expand, like sample W75S25. Rileys composition diagram for good
bloating clays may not be fully applicable to other types of raw
materials or to different heating conditions (Conley et al., 1948;
Kung and Liu, 1996; Utley et al., 1965). The bloating area was drawn
around the points which represented the clays that bloated at
a temperature of 1320

C or less when red for 15 min in an electric
Globar furnace (Utley et al., 1965) and, in this study, the raw
materials (a ne sediment mixed with sewage sludge), the kiln and
the heating conditions are slightly different.
Some studies (Huang et al., 2007) show that feeding tempera-
tures affect the BI. If the feeding temperature is sufciently high for
the surfaces of the raw aggregates to become quickly vitried, then
the gas produced in the process is enveloped. On the contrary,
when the feeding temperature is too low, the articial aggregates
will experience a temperature increase, both internal and external,
during sintering. When gas is produced, the surface is not
completely vitried, making it possible for the gas to escape.
Instead of expanding, W75S25 undergoes a progressive decrease in
the pellet sizes (BI <0) as the heating temperature increases. This
could be due to the absence of the formation of a vitried layer and
the loss of mass by gas release (Fig. 6).
The bulk density is less than 1.20 g/cm
3
in all cases (Fig. 8);
therefore, all the articial aggregates are LWAs. The mixtures
containing sewage sludge produce LWAs with a bulk density
comparable to that of the so-called expanded clays, which is in
the 0.50.7 g/cm
3
range (De Gennaro et al., 2004). It has been
reported that when the grain size of the aggregate increases, the
bulk density gradually decreases (De Gennaro et al., 2005). It
can be seen that when the BI (Fig. 7) of a given sample
W100
0
10
20
30
40
1150 1175
Temperature (C)
L
O
I

(
%
)
preheating
heating 10
min
heating 15
min
total
(muffle)
W75S25
0
10
20
30
40
1150 1175 1200 1225
Temperature (C)
L
O
I

(
%
)
W50S50
0
10
20
30
40
1150 1175 1200 1225
Temperature (C)
L
O
I

(
%
)
W25S75
0
10
20
30
40
1150 1175 1200 1225
Temperature (C)
L
O
I

(
%
)
W50C50
0
10
20
30
40
1150 1175 1200
Temperature (C)
L
O
I

(
%
)
Fig. 6. Loss on ignition of granules in pre-heating zone, at 10 and 15 min of heating in the rotary kiln and total loss on ignition in the mufe.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2807
decreases (granules are smaller), the bulk density of this sample
increases.
The dry particle density values obtained for all the articial
aggregates are less than 2 g/cm
3
; therefore, all the articial aggre-
gates manufactured in this study are classied as lightweight
aggregates according to UNE-EN-13055-1 (2003). Fig. 9 shows the
effect of the composition, heating temperature and dwell time on
the dry particle density of the LWAs. The dry particle density
decreases as the sewage sludge content increases and, on the
contrary, it increases as the clay-rich sediment increases. Mun
(2007) attributed the decrease in density to the internal sintering
due to the calorication and combustion of organic materials and
the accelerated expansion caused by gas generation. In this study,
the decrease in density cannot be attributed to expansion, because
BI is negative in the aggregates containing sewage sludge (Fig. 7).
The dry particle density shows some of the highest signicant
correlations in absolute value with LOI
rm
(0.95) (Table 4). This
result is in accordance with the data shown in Table 1: aggregates
with the lowest LOI
rm
have the highest dry particle density;
therefore, the mass of the individual aggregate components is one
of the most important factors in the dry particle density of the
LWAs.
Among all the samples studied, only W50C50 shows a reduction
in density with the increase in heating temperature for dwell times
of 10 and 15 min. This is due to the bloating that only this sample
experiences. The volume of new internal pores is proportionally
higher than the mass loss due to the increase in temperature (Fig. 6
and Table 4; non-signicant correlation between T
heating
and
LOI
10/15
).
The apparent particle density for the LWAs prepared from the
clay-rich sediment decreases from 2.23 g/cm
3
to 1.84 g/cm
3
(Fig. 10), with the lowest value at a sintering temperature of
1200

C. It has been reported that this property of LWAs is related to
the expansion percentage (Fakhfakh et al., 2007). This sample
shows an increase in BI from 2.5% to 4.3% by increasing the pore
size of the LWAs, resulting in a decrease in the apparent particle
density if these pores are impermeable to water. As shown in Fig. 11,
the water absorption after 24 h of W50C50 decreases with
temperature, so that the pores formed are impermeable and they
-20
-15
-10
-5
0
5
10
1150 1175 1200 1225
Temperature (C)
B
l
a
o
t
i
n
g

I
n
d
e
x

(
%
)
W100, 10 min
W75S25, 10 min
W50S50, 10 min
W25S75, 10 min
W50C50,10 min
W100, 15 min
W75S25, 15 min
W50S50, 15 min
W25S75, 15 min
W50C50, 15 min
Bloating
Fig. 7. Relationship between heating temperature, dwell time, composition and
bloating index.
Dry pellets 1150 1175 1200 1225
Temperature (C)
B
u
l
k

d
e
n
s
i
t
y

(
g
/
c
m
3
)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
W100, 10 min
W75S25, 10 min
W50S50, 10 min
W25S75, 10 min
W50C50,10 min
Arlita G3
W100, 15 min
W75S25, 15 min
W50S50, 15 min
W25S75, 15 min
W50C50, 15 min
LWA<=1.20 g/cm
Fig. 8. Relationship between heating duration, composition and bulk density at
different sintering temperatures.
Dry pellets 1150 1175 1200 1225
Temperature (C)
D
r
y

p
a
r
t
i
c
l
e

d
e
n
s
i
t
y

(
g
/
c
m
3
)
0
0.5
1
1.5
2
2.5
W100, 10 min
W75S25, 10 min
W50S50, 10 min
W25S75, 10 min
W50C50,10 min
Arlita G3
W100, 15 min
W75S25, 15 min
W50S50, 15 min
W25S75, 15 min
W50C50, 15 min
LWA<=2.00 g/cm
Fig. 9. Relationship between heating duration, composition and dry particle density at
different sintering temperatures.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2808
are the reason for the decrease in the apparent particle density of
W50C50. On the contrary, the mixtures containing sewage sludge
do not showthe previously mentioned relationship; their apparent
particle density does not increase even though their BI decreases.
LOI
10/15
is maintained as the temperature increases and BI
decreases; therefore, if r
a
is also maintained or does not increase
proportionally, this could be due to the volume of new imperme-
able internal pores having increased and would compensate for the
above-mentioned decrease in external volume.
A particle with isolated pores or a vitried surface tends to
absorb little water, whereas one with connected or open pores will
absorb water like a sponge (Hung and Hwang, 2007). The water
absorption of the aggregates increases as the proportion of sewage
sludge in the raw pellets increases (Fig. 11). The surface of the
pellets with different proportions of sewage sludge is not vitried
during heating and, therefore, gases will escape, causing an
increase in the open porosity of the granules. As Fig. 6 shows, LOI
(i.e. generation of gas) increases as the sewage sludge increases;
therefore, the open porosity will increase, with a consequent
increase in the WA
24h
. Moreover, Table 4 shows that a high
signicant correlation exists between WA
24h
LOI
z1
(0.83) and
WA
24h
LOI
10/15
(0.91).
The compressive strength (S) is affected by interrelated factors,
such as the density and shape of the aggregates (high density
values and more spherical shapes produce high S values), surface
Table 4
Spearman Rank Order Correlations (rs) (All samples). Marked correlations are signicant at p < 0.05. LOI
rm
: loss on ignition of rawmaterials; IC: inorganic carbon content; TOC:
organic carbon content; Si/
P
F: SiO
2
/
P
Fluxing ratio; T
heating
: heating temperaute in zone 2 of the rotary kiln; LOI
z1
: loss on ignition of articial aggregates in zone 1 of rotary kiln;
LOI
10/15
: loss on ignition of articial aggregates in zone 1 + zone 2 + zone 3 of rotary kiln (for 10 or 15 min in zone 2); LOI
t
: total loss on ignition of articial aggregates in mufe; BI:
bloating Index; r
b
: Bulk density; r
a
: apparent particle density; r
d
: dry particle density; WA
24h
: water absorption after 24 h; H: void percentage; S: compressive strength.
LOIrm IC TOC Si/
P
F T
heating
Dwell time LOI
z1
LOI
10/15
LOI
t
BI r
b
r
a
r
d
WA
24h
H S
LOIrm 1.00
IC L0.49 1.00
TOC 1.00 L0.49 1.00
Si/
P
F 1.00 0.49 1.00 1.00
T
heating
0.21 0.12 0.21 0.21 1.00
Dwell time 0.00 0.00 0.00 0.00 0.00 1.00
LOI
z1
0.94 L0.58 0.94 L0.94 0.49 0.00 1.00
LOI
10/15
0.98 L0.49 0.98 L0.98 0.31 0.06 0.95 1.00
LOI
t
0.98 L0.49 0.98 L0.98 0.31 0.00 0.95 0.97 1.00
BI L0.79 0.44 L0.79 0.79 L0.53 0.03 L0.86 L0.84 L0.77 1.00
r
b
L0.94 0.47 L0.94 0.94 0.05 0.02 L0.80 L0.88 L0.90 0.56 1.00
r
a
0.72 0.22 0.72 L0.72 0.08 0.07 0.73 0.83 0.85 L0.58 L0.70 1.00
r
d
L0.95 0.51 L0.95 0.95 0.00 0.03 L0.83 L0.90 L0.91 0.61 0.99 L0.70 1.00
WA
24h
0.96 L0.51 0.96 L0.96 0.01 0.06 0.83 0.91 0.92 L0.65 L0.98 0.73 L0.99 1.00
H L0.37 0.13 L0.37 0.37 0.08 0.05 0.32 0.31 0.34 0.41 0.32 0.20 0.40 L0.42 1.00
S L0.80 0.73 L0.80 0.80 0.02 0.03 L0.81 L0.87 L0.89 0.48 0.80 L0.46 0.93 L0.92 0.22 1.00
0.7
1.2
1.7
2.2
2.7
Dry pellets 1150 1175 1200 1225
Temperature (C)
A
p
p
a
r
e
n
t

p
a
r
t
i
c
l
e

d
e
n
s
i
t
y

(
g
/
c
m
3
)
W100, 10 min
W75S25, 10 min
W50S50, 10 min
W25S75, 10 min
W50C50,10 min
Arlita G3
W100, 15 min
W75S25, 15 min
W50S50, 15 min
W25S75, 15 min
W50C50, 15 min
Fig. 10. Relationship between heating duration, composition and apparent particle
density at different sintering temperatures.
0
10
20
30
40
50
60
70
80
90
100
Dry pellets 1150 1175 1200 1225
Temperature (C)
W
A

2
4
h
(
%
)
W100, 10 min
W75S25, 10 min
W50S50, 10 min
W25S75, 10 min
W50C50,10 min
Arlita G3
W100, 15 min
W75S25, 15 min
W50S50, 15 min
W25S75, 15 min
W50C50, 15 min
Fig. 11. Relationship between heating duration, composition and water absorption at
different sintering temperatures.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2809
features (cracking or fractures due to thermo-stress produce low S
values), water absorption (low WA
24h
values produce high S
values), pore size and distribution, densication effects due to
sintering, bloating index, mineral species present in the LWAs and
the nature of the newly formed phases (Chang et al., 2007; Fakh-
fakh et al., 2007; Lin, 2006). In this study, the compressive strength
ranges from 0.5 MPa to at least 13.31 MPa (most LWAs with clay-
rich sediment cannot be tested because the load necessary to break
them is greater than 160 kg and the press is not prepared to endure
such a high load). The highest values, 13.31 MPa, 12.9 MPa and
12.55 MPa, are obtained for W75S25 heated at 1225

C for 15 min,
W75S25 heated at 1225

C for 10 min and W50C50 heated at
1150

C for 10 min, respectively (Fig. 12). These values are clearly
related to the dry particle density and the water absorption of the
pellets (rs =0.93 and 0.92, respectively, Table 4). The correlations
between SLOI
z1
(0.81) and SLOI
10/15
(0.87) are also signicant,
as the quantity of gases released can affect the microstructure of
the sintered articial aggregates, thereby decreasing the
compressive strength (Lin, 2006).
The fact that the best compressive strength values are achieved
with the highest dry particle density values leads to the creation of
a parameter that represents a measure of the compromise between
the need for high resistance and low density for certain applica-
tions (De Gennaro et al., 2007). This parameter is the S/r
d
ratio: the
best LWAs would be those that show the highest values. The S/r
d
ratios of the aggregates and of Arlita G3 are shown inTable 5. All the
articial aggregates with clay-rich sediment and washing aggregate
sludge, and those with 25% sewage sludge and 75% washing
aggregate sludge heated at 1225

C, are the best LWAs when taking
into consideration the compromise S/r
d
, whose values are higher
than those obtained for Arlita G3.
4.4. Comparison of the laboratory-produced and commercial LWAs
Table 6 shows the varieties of Arlita

, the most widely marketed

LWAs in Spain, their main features and applications.
The Arlita G3 values shown in Table 6 are slightly different from
those obtained in this study due to the grain size selected (<10 mm)
for this variety and changes in the method used to obtain the values.
The LWAs obtained may be used for different applications,
depending on the desired property (i.e. bulk density, WA
24h
or S).
Specically, if the compressive strength is taken into account, but
the values of the other two properties are considered, it is possible
to establish three groups of LWAs by comparison to the different
types of Arlita varieties.
(a) LWAs with low resistance (lower than Arlita G3) and high
water absorption (higher than Arlita A5). These LWAs are all
manufactured with 50% and 75% sewage sludge, W50S50 and
W25S75, respectively, regardless of the heating temperature
and dwell time. They may be used for the same applications as
Arlita G3: insulation, geotechnical applications, gardening and/
or horticulture.
(b) LWAs with medium resistance and water absorption. These are
all manufactured with 100% washing aggregate sludge (W100),
regardless of the heating temperature and dwell time, and
manufactured with 25% sewage sludge and heating at 1150

C,
1175

C and 1200

C. These LWAs may have similar uses as
Arlita G3, previously mentioned, and/or Arlita F3 (i.e. pre-
fabricated lightweight structures and lightweight insulating
concretes).
(c) LWAs with high resistance and low water absorption (lower
than Arlita F7). These LWAs are manufactured with 25% sewage
sludge (W75S25) and heating at 1225

C, and are all manu-
factured with clay-rich sediment (W50C50), regardless of the
heating temperature and dwell time. These may have similar
uses as Arlita F5: concrete slabs and building structures.
The above-mentioned applications and uses are preliminary;
the nal investigationwill be the specic objectives for future work.
0
2
4
6
8
10
12
14
Dry pellets 1150 1175 1200 1225
Temperature (C)
C
o
m
p
r
e
s
s
i
v
e

s
t
r
e
n
g
h
t

(
M
P
a
)
W100, 10 min
W75S25, 10 min
W50S50, 10 min
W25S75, 10 min
W50C50,10 min
ArlitaG3
W100, 15 min
W75S25, 15 min
W50S50, 15 min
W25S75, 15 min
W50C50, 15 min
Fig. 12. Relationship between heating duration, composition and compressive
strength at different sintering temperatures.
Table 5
Compressive strength/dry particle density ratios.
T
heating
(

C)
1150 1175 1200 1225
Time
(min)
10 15 10 15 10 15 10 15
W100 2.76 3.70 3.26 4.05
W75S25 2.29 2.24 2.65 2.76 3.70 3.54 7.21 7.35
W50S50 1.41 1.50 1.56 1.48 1.37 1.47 1.44 1.31
W25S75 0.56 0.85 0.97 0.93 0.96 0.87 1.05 1.16
W50C50 Min.
a
6.68
Min.
6.68
Min.
6.54
Min.
6.37
Min.
6.68
Min.
7.13
Arlita G3 4.41
a
Minimum value.
Table 6
Varieties of Arlita

, features and applications (Optiroc A

ridos Ligeros, S.A., 2000)

Variety Bulk
density
(g/cm
3
)
Water
absorption
(%)
Compressive
strength
(MPa)
Applications
G3 0.32550 20 0.981 Insulation, geotechnical applications,
gardening and horticulture
F3 0.35050 2025 1.962 Prefabricated lightweight structures
and insulation lightweight concretes
F5 0.55050 1520 4.905 Concrete slabs, building structures
A5 0.57550 3035 Refractory mortars, super
lightweight concretes
F7 0.75050 1015 6.867 Prestressed concrete, civil works
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2810
5. Conclusions
We have evaluated washing aggregate sludge, sewage sludge
from a WWTP and clay-rich sediment for use in the production of
lightweight aggregates. The physical, chemical and mineralogical
analyses of the raw materials, the laboratory-scale manufacturing
of articial aggregates and the technological characterization of the
products show that obtaining LWAs from mining and industrial
wastes is possible.
The following specic conclusions have been deduced from this
study.
v All the mixtures show bloating potential taking into consider-
ation the gases released at high temperatures. Only two
mixtures are located within the bloating area with a proper
chemical composition to trap the gases.
v The supply of raw materials such as those used in this study is
guaranteed due to the continuous production of these types of
wastes associated with the growth of industrial and mining
activities in the region studied.
v All the products obtained in this study are lightweight aggre-
gates, in accordance with Standard UNE-EN-13055-1, 2003.
v Only LWAs manufactured with 50% washing aggregate sludge
and 50% clay-rich sediment are expanded LWAs (BI >0).
v The expansion largely depends on the formation of a proper
viscous phase that traps gases and releases these gases at the
same time. If the gases are not released when a sufcient
viscous phase has been formed, they will escape and bloating
will not occur.
v LWAs that show the lowest bulk density and dry particle
density are manufactured with 75% sewage sludge and 25%
washing aggregate sludge. Those with the lowest apparent
particle density and lowest water absorption are LWAs with
50% clay-rich sediment and 50% washing aggregate sludge. This
type of aggregates has the highest compressive strength.
v S/r
d
ratios higher than those of Arlita G3 are shown by all the
LWAs with 50% clay-rich sediment and those with 25% sewage
sludge, heated at 1225

C.
v The possible applications of the synthesized LWAs, taking into
consideration the compressive strength and the water
absorption, are the following: LWAs suitable for concrete slabs
and building structures that require high compressive strength.
LWAs appropriate for applications that do not require very high
compressive strength, such as prefabricated lightweight
structures and lightweight insulating concretes. LWAs appro-
priate for insulation, geotechnical applications, gardening and
horticulture due to their lowcompressive strength and/or their
high water absorption.
The manufacturing of articial aggregates through waste recy-
cling is not only a useful alternative to the extraction of natural
aggregates, but also offers a feasible solution to waste management
and contributes to the recycling of wastes.
Acknowledgements
Financial support from the Junta de Comunidades de Castilla-La
Mancha (Project PBI-05-044 and Project PCI08-0027) of Spain is
gratefully acknowledged.
References
Baykal, G., Doven, A.G., 2000. Utilization of y ash by pelletization process: theory,
application areas and research results. Resources Conservation and Recycling 30
(1), 5977.
Benbow, J., September 1987. Mineral in re protection construction support market.
Industrial Minerals, 6173.
Benbow, J.J., Oxley, E.W., Bridgwater, J., 1987. The extrusion mechanics of pastes
the inuence of pastes formulation on extrusion parameters. Chemical Engi-
neering Science 42, 21512162.
Bethanis, S., Cheeseman, C.R., Sollars, C., 2002. Properties and microstructure of
sintered incinerator bottom ash. Ceramics International 28 (8), 881886.
Bethanis, S., Cheeseman, C.R., Sollars, C.J., 2004. Effect of sintering temperature on
the properties and leaching of incinerator bottom ash. Waste Management and
Research 22 (4), 255264.
Blanco, I., 2004. Caracterizacio n de los residuos y este riles generados en las graveras
del ro Jarama y sus aplicaciones en la industria cera mica. PhD thesis, Uni-
versidad Complutense, Madrid, 563 pp.
Casagrande, A., 1932. Research in Atterberg limits of soil. Public Roads 13.
Casagrande, A., 1948. Classication and identication of soils. Transactions of the
American Society of Civil Engineers 113 (2351), 901903.
Chang, F.C., Lo, S.L., Lee, M.Y., Ko, C.H., Lin, J.D., Huang, S.C., Wang, C.F., 2007.
Leachability of metals from sludge-based articial lightweight aggregate.
Journal of Hazardous Materials 146 (12), 98105.
Cheeseman, C.R., Virdi, G.S., 2005. Properties and microstructure of lightweight
aggregate produced from sintered sewage sludge ash. Resources, Conservation
and Recycling 45 (1), 1830.
Chopra, S.K., Lal, K., Ramaahandran, V.S., 1964. Gas-producing agents in production
of lightweight aggregates. Journal of the Applied Chemistry of the Ussr 14 (5),
181.
Conley, J.E., Wilson, H., Kleinfelter, T.A., 1948. Production of Lightweight Concrete
Aggregates from Clays, Shales, Slates and Other Materials. Report of Investi-
gation, 4401. US Bureau of Mines, 121 pp.
Dahab, A.S., 1980. Les argiles expansibles: caracte risation mine ralogique et chimi-
que. Expansion statique: application a` la culture sans sol. PhD thesis, Labo-
ratoire Pale oge ographie-Substances utiles, ENSG, Nancy, 150 pp.
Decleer, J., Viaene, W., 1993. Rupelian boom clay as raw material for expanded clay
manufacturing. Applied Clay Science 8, 111128.
De Gennaro, R., Cappelletti, P., Cerri, G., De Gennaro, M., Dondi, M., Langella, A.,
2004. Zeolitic tuffs as raw materials for lightweight aggregates. Applied Clay
Science 25 (12), 7181.
De Gennaro, R., Cappelletti, P., Cerri, G., De Gennaro, M., Dondi, M., Langella, A., 2005.
Neapolitan YellowTuff as rawmaterial for lightweight aggregates in lightweight
structural concrete production. Applied Clay Science 28 (14), 309319.
De Gennaro, R., Cappelletti, P., Cerri, G., De Gennaro, M., Dondi, M., Graziano, S.F.,
Langella, A., 2007. Campanian Ignimbrite as raw material for lightweight
aggregates. Applied Clay Science 37 (12), 115126.
Ducman, V., Mladenovic, A., Suput, J.S., 2002. Lightweight aggregate based on waste
glass andits alkali-silicareactivity. Cement andConcrete Research32(2), 223226.
Fakhfakh, E., Hajjaji, W., Medhioub, M., Rocha, F., Lo pez-Galindo, A., Setti, M.,
Kooli, F., Zargouni, F., Jamoussi, F., 2007. Effects of sand addition on production
of lightweight aggregates from Tunisian smectite-rich clayey rocks. Applied
Clay Science 35, 228237.
Gao, X., Wang, W., Ye, T., Wang, F., Lan, Y., 2008. Utilization of washed MSWI y ash
as partial cement substitute with the addition of dithiocarbamic chelate. Jour-
nal of Environmental Management 88, 293299.
Garca, G., 2004. Aragonitos de Pantoja (Toledo). Bocamina 13, 1027.
Huang, S.C., Chang, F.C., Lo, S.L., Lee, M.Y., Wang, C.F., Lin, J.D., 2007. Production of
lightweight aggregates from mining residues, heavy metal sludge, and incin-
erator y ash. Journal of Hazardous Materials 144 (12), 5258.
Hung, M.F., Hwang, C.L., 2007. Study of ne sediments for making lightweight
aggregate. Waste Management and Research 25, 449456.
Ingamells, C.O., 1970. Lithium metaborate ux in silicate analysis. Analytica Chimica
Acta 52, 323334.
Jackson, M.L., 1969. Soil Chemical Analysis Advanced Course, second ed. Univer-
sity of Wisconsin, Madison, WI. Published by the author.
Kim, Y., Kim, J.H., Lee, K.G., Kang, S.G., 2005. Recycling of dust wastes as lightweight
aggregates. Journal of Ceramic Processing Research 6 (2), 9194.
Kung, L.S., Liu, C.P., 1996. Lightweight Aggregate Concrete. China Railroad Publisher,
Beijing, China.
Li, Y., Wu, D., Zhang, J., Chang, L., Fang, Z., Shi, Y., 2000. Measurement and statistics
of single pellet mechanical strength of differently shaped catalysts. Powder
Technology 113, 176184.
Lin, K.L., 2006. Mineralogy and microstructures of sintered sewage sludge ash as
lightweight aggregates. Journal of Industrial and Engineering Chemistry 12 (3),
425429.
Lin, K.L., Chang, W.C., Lin, D.F., Luo, H.L., Tsai, M.C., 2008. Effects of nano-SiO2
and
different ash particle sizes on sludge ashcement mortar. Journal of Environ-
mental Management 88, 708714.
Mangialardi, T., 2001. Sintering of MSW y ash for reuse as a concrete aggregate.
Journal of Hazardous Materials 83 (13), 225239.
Mun, K.J., 2007. Development and tests of lightweight aggregate using sewage
sludge for nonstructural concrete. Construction and Building Materials 21 (7),
15831588.
Optiroc A

ridos Ligeros, S.A., 2000. La Arcilla Expandida. In: Arlita Arcilla Expandida,
Manual General, Cap, 1, pp. 529.
Ribeiro, M.J., Ferreira, J.M., Labrincha, J.A., 2005. Plastic behaviour of different
ceramic pastes processed by extrusion. Ceramics International 31, 515519.
Riley, C.M., 1951. Relation of chemical properties to the bloating of clays. Journal of
the American Ceramic Society 34, 121128.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2811
Salihoglu, G., Pinarli, V., Salihoglu, N.K., Karaca, G., 2007. Properties of steel foundry
electric arc furnace dust solidied/stabilized with Portland cement. Journal of
Environmental Management 85, 190197.
Sanchez, F., White, M.K.L., Hoang, A., 2009. Leaching from granular cement-based
materials during inltration/wetting coupled with freezing and thawing. Jour-
nal of Environmental Management 90, 983993.
Tay, J.H., Hong, S.Y., Show, K.Y., 2000. Reuse of industrial sludgeas pelletizedaggregate
for concrete. Journal of Environmental Engineering-ASCE 126 (3), 279287.
UNE-103-104, 1993. Test for plastic limit of a soil.
UNE-103-103, 1994. Determination of the liquid limit of a soil by the Casagrande
apparatus method.
UNE-EN-1097-3, 1999. Test for mechanical and physical properties of aggregates
determination of loose bulk density and voids.
UNE-EN-1097-6, 2000. Test for mechanical and physical properties of aggregates
determination of particle density and water absorption.
UNE-EN-13055-1, 2003. Lightweight aggregates lightweight aggregates for
concrete, mortar and grout.
Utley, R.W., Lovell, H.L., Spicer, T.S., 1965. The preparation of coal refuse for the
manufacture of lightweight aggregate. Transactions of the Society of Mining
Engineers of AIME 234 (4), 346352.
Videla, C., Martinez, P., 2002. Physical, mechanical and microscopic characterization
of cold bonded y ash lightweight aggregates. Materiales de Construccio n 52
(268), 518.
Yashima, S., Kanda, Y., Sano, S., 1987. Relationship between particle size and fracture
energy or impact velocity required to fracture as estimated from single particle
crushing. Powder Technology 51, 277282.
Yasuda, Y., 1991. Sewage-sludge utilization in Tokyo. Water Science and Technology
23 (1012), 17431752.
Zheng, G., Kozinski, J.A., 2000. Thermal events during the combustion of biomass
residue. Fuel 79, 181192.
B. Gonzalez-Corrochano et al. / Journal of Environmental Management 90 (2009) 28012812 2812