Journal of Hazardous Materials 299 (2015) 181–191

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Recycling of MSWI fly ash by means of cementitious double step cold

bonding pelletization: Technological assessment for the production of
lightweight artificial aggregates
Francesco Colangelo a,b,∗ , Francesco Messina a,b , Raffaele Cioffi a,b
a
Department of Engineering, University of Naples ‘Parthenope’ Centro Direzionale, Isola C4, 80143 Naples, Italy
b
INSTM, Consorzio Interuniversitario per la Scienza e Tecnologia dei Materiali, Via G. Giusti, 9. 50121 Firenze, Italy

h i g h l i g h t s

• MSWI fly ash is recycled up to 70% in the production of cold bonded aggregates.

• Double-step pelletization allows to improve mechanical and stabilization properties.
• Cement and coal fly ash/lime systems give higher engineering performance.
• Concrete containing cold bonded aggregates has satisfactory engineering performance.
• Pilot equipment upscale is technologically simple and economically feasible.

a r t i c l e i n f o a b s t r a c t

Article history: In this work, an extensive study on the recycling of municipal solid waste incinerator fly ash by means of
Received 22 April 2015 cold bonding pelletization is presented. The ash comes from an incineration plant equipped with rotary
Received in revised form 8 June 2015 and stoker furnaces, in which municipal, hospital and industrial wastes are treated. Fly ash from waste
Accepted 9 June 2015
incineration is classified as hazardous and cannot be utilized or even landfilled without prior treatment.
Available online 18 June 2015
The pelletization process uses cement, lime and coal fly ash as components of the binding systems. This
process has been applied to several mixes in which the ash content has been varied from 50% (wt.%)
Keywords:
up to a maximum of 70%. An innovative additional pelletization step with only cementitious binder has
Municipal solid waste incinerator fly ash
Double step granulation
been performed in order to achieve satisfactory immobilization levels. The obtained lightweight porous
Cold bonding pelletization aggregates are mostly suitable for recovery in the field of building materials with enhanced sustainability
Stabilization/solidification properties. Density, water absorption and crushing strength ranged from 1000 to 1600 kg/m3 , 7 to 16%
Artificial aggregate concrete and 1.3 to 6.2 MPa, respectively, and the second pelletization step increased stabilization efficiency. The
Leaching feasibility of the process has been analyzed by testing also concrete specimens containing the artificial
aggregates, resulting in lightweight concrete of average performance.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction undoubtedly worthy of consideration [2–14]. Such process has eco-

Cold bonding pelletization of waste is often proposed a sta-
bilization/solidification technique for waste reuse in low cost
building materials production. Particularly, artificial aggregates are
of course one of the most interesting technological solutions for
waste recovery [1]. Cement-based cold bonding pelletization pro-
cess has recently received a quite relevant attention in literature
and its suitability for the manufacture of artificial aggregates is
∗ Corresponding author. Fax: +39 081 547 6777.
E-mail address: colangelo@uniparthenope.it (F. Colangelo).
nomic and environmental advantages due to the reduced energy
requirement (process carried out at room temperature) respect
to industrial alternatives such as sintering [e.g. 15,16], which
is an energy intensive process. More recently, also alternative
cement-free binding matrices with reduced embedded CO2 have
been proposed for stabilization/solidification [e.g. 17–20], such as
geopolymer and alkali activated ones. These systems have gained
an increasing interest from researchers thanks to promising results
in terms of physical, mechanical and durability properties and
possibility of synthesis starting from natural/industrial wastes
[21–24][e.g. 21–24].

http://dx.doi.org/10.1016/j.jhazmat.2015.06.018
0304-3894/© 2015 Elsevier B.V. All rights reserved.
182 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191
Fig. 1. Image of an artificial aggregate obtained by means of double-step pelletization after splitting in two halves.
In this paper a cement-based pelletization process has been
studied for the manufacture of artificial aggregate starting from
two municipal solid waste incinerator (MSWI) fly ash samples
coming from different furnaces of the same plant. The stabiliza-
tion/solidification treatment investigated is based on cementitious
cold bonding pelletization and makes use of a rotary plate pelletiza-
tion pilot-scale apparatus with binding mixes composed of cement,
lime and coal fly ash. Cold bonding pelletization is performed by
means of a rotating and tilting plate. The waste is added in pow-
der form into the apparatus together with binders. As soon as the
plate starts to rotate, a spray doser immediately supplies water
homogeneously into the plate avoiding waste dispersion into work-
ing environment, allowing both the safe handling of hazardous of
MSWIFA and the synthesis of the stabilizing matrices. MSWI fly ash
may contain chlorides and sulfates which could be harmful for the
cementitious binding matrix and, so, a pre-treatment is needed.
In order to reduce the content of chlorides and sulfates a washing
pre-treatment has been carried out. This washing pre-treatment is
described in a previous work by the authors [25]. Wastewater from
washing process can be easily processed inside incineration plants
by means of evaporation recovering heat from wastes combustion.
The evaporation treatment removes salts from water, minimizing
the total waste production and producing water that is suitable for
the reuse in the industrial cycle. In our study, in addition to tra-
ditional single step pelletization, a second pelletization step has
been carried out to have final products with improved properties
in view of reuse for artificial aggregate production. In the one-step
pelletization the waste is incorporated within the binding matrix
in a measure ranging from 50 to 70%. In the two-step pelletization
a second step is carried out with pure binder to get the aggregates
from the one-step process encapsulated within an outer shell able
to improve the technological and leaching properties. A graphical
scheme of the double-step cold bonding pelletization is reported
in Fig. 1.
The obtained granules have been tested in order to assess
their physico-mechanical and leaching properties. Specifically,
measurements have been carried out regarding the following
properties: density, water adsorption capacity, Los Angeles coef-
ficient, crushing strength, leaching behavior, mesostructural and
microstructural characterization by means of optical microscopy
and scanning electron microscopy, respectively. Moreover, con-
crete mixes have been prepared with some of the artificial
aggregates produced by means of the above mentioned pelleti-
zation process. These mixes have been tested and classified from
the technological point of view by evaluating density, compressive
strength and dynamic modulus of elasticity. The obtained values
have revealed the feasibility of this application for MSWI fly ash
reuse in a sustainability-oriented framework.
2. Materials and experimental methods
2.1. Residues characterization and washing pre-treatment
The MSWI fly ash employed in this work comes from an incin-
eration plant located in Melfi (Potenza, Italy) in which municipal,
hospital and industrial wastes are treated. The plant is equipped
with rotary and stoker furnaces and fly ash samples from both fur-
naces have been individually employed. According to the European
Waste Catalogue, this waste is given the code 19.01.13* and classi-
fied as hazardous. The content of the main components in both the
ash samples has been determined through X-ray fluorescence. The
ash samples have been submitted to total acid digestion (ASTM
5258-92) and subsequent chemical analysis through inductively
coupled plasma atomic emission spectrometry (ICP-AES) technique
for the determination of metals contents. Chlorides and sulfates
contents have been determined by means of the Mohr method
and ionic liquid chromatography, respectively. The residues have
been submitted to UNI 10802:2004 leaching test, whose procedure
derives from EN 12457-2:2002. This is a static test that makes use
of deionized water as leachant with a 10 l/kg liquid to solid ratio.
Its duration, in the case of granular solids (size <3 mm), must be
24 h. To prevent metals precipitation after leaching, the leachate
pH must be brought to 2 by means of 1 N HNO3 solution. Par-
ticle size distribution of ashes has been determined by means
of a Malvern Instrument Mastersizer 2000 laser scattering ana-
lyzer (particles observation range: 0.02 ÷ 2000 ␮m). The binding
materials employed for the manufacture of the granules are CEM
II/A-L 42.5R (UNI EN 197-1:2011), commercial hydrated lime and
 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191 183

Table 1 used in amount equal to 40% of granules weight. In the follow-

Compositions (wt.%) of stabilizing matrices.
ing sections, these systems are distinguished by placing the sign
Ash Origin Binder “+” at the end of each label. The granules size distribution has
compositios been determined by sieving according to UNI EN 933-1 standard.
Rotary Stoker CEM Lime CFA w/S Each specimen was soaked into water at 20 ± 2 ◦ C and then super-
furnace furnace ficially wiped with a moist cloth. This was repeated every 24 h until
R70C 70 – 30 – – 0.25 constant weight was reached. The water saturated granules have
S70C – 70 30 – – 0.41
been used for the determination of the specimen volume by means
R70L 70 – – 30 – 0.25
S70L – 70 – 30 – 0.36 of water displacement. The mass of the same specimen has been
R60LA 60 – – 15 25 0.32 determined after drying at 105 ± 5 ◦ C. Following this procedure,
S60LA – 60 – 15 25 0.39 the granules apparent density was obtained. The water absorption
R50LA 50 – – 30 20 0.35
capacity (WAC) was determined on the granules which have been
S50LA – 50 – 30 20 0.40
dried at 40 ± 2 ◦ C until constant weight was reached. Then, WAC
has been determined by means of the following relationship:
coal fly ash supplied by the ENEL (Italian Electricity Board) power mw − md
plant located in Brindisi (Italy). It is well known that cementi- WAC = 100 (1)
md
tious stabilization processes are not very effective towards MSWI
fly ashes. The reason for this is the scarce immobilizing capabil- where mw is the water saturated granules mass and md is the
ity of a cementitious matrix toward chlorides and other soluble 40 ± 2 ◦ C dried granules mass. The Los Angeles coefficient has been
salts. As a consequence, based on previous research [25], a prelim- determined as a measure of aggregate degradation resulting from a
inary washing step was applied to develop a cementitious process combination of actions including abrasion, impact and grinding in
with a reasonable low binder/waste ratio. Specifically, the two a rotating steel drum containing steel spheres according to UNI EN
MSWI fly ash samples have been submitted to a two-step wash- 1097-2 standard (ASTM C131). The specimen has been washed and
ing pre-treatment with liquid/solid ratio equal to 2:1. The washing oven dried at 105 ± 5 ◦ C until constant weight was reached. Then,
procedure is extensively described in the mentioned work [25]. it was loaded into the rotating drum together with the specified
number of spheres. After 500 drum revolutions at 30-33 rpm, the
2.2. Aggregates and concrete manufacturing and characterization specimen has been discharged and sieved through a 1.7 mm sieve
(No. 8 of UNI 2331/2332 sieve series or 12 mesh of ASTM E 11-70
The pre-washed MSWI fly ash have been employed in the prepa- series). The Los Angeles coefficient indicates exactly the difference
ration of eight pelletization mixtures, whose compositions are between the original charge weight and the recovered one after
reported in Table 1. The granules manufacture has been carried sieving expressed in terms of percentage of the original weight.
out by means of a pilot scale granulator apparatus. This apparatus The result is rounded to 1%. The measurement of the granules com-
is equipped with a rotating and tilting plate (d = 80 cm) for which pressive (crushing) strength has been carried out according to UNI
rotating speed and tilting angle can be adjusted between wide lim- EN 13055-1 standard by means of a 3000 kN Controls® MC60 press.
its. In the experiment, the settings have been fixed at 45 rpm for the After physical and mechanical characterization, leaching tests have
rotating speed and 45◦ for the tilting angle, according to optimiza- been carried out on artificial aggregates. The analysis of heavy met-
tion performed in a previous study [3]. The pilot scale cold bonding als release behaviour is of course the most important experimental
pelletization apparatus is shown in Fig. 2. test in order to evaluate the feasibility of the recovery process. The
The granules were cured for 12 h in climatic chamber where aggregates have been submitted to UNI 10802 leaching test for
temperature and relative humidity were kept at 50 ◦ C and 95%, monolithic material, carried out with leachant renewals at 2 and
respectively. This curing phase favors granules hardening, neces- 18 h and total duration of 48 h ensuring a liquid volume/solid sur-
sary for successive handling. Finally, the granules were submitted face ratio of 10 cm. As in the granular material case, the pH must
to a 14-day curing at room temperature and humidity. It was be brought to 2 by means of 1 M HNO3 before the metals released
checked that further curing in these conditions did not improve are analyzed. The cumulative leaching data after 72, 336 and 672 h
significantly the technological properties of the granules. have been also determined.
All the systems listed in Table 1 were submitted to a second Concrete mixtures have been reported in Table 2. Acrylic admix-
granulation step in which a 1:1 cement/coal fly ash binder was ture content has been modified in order to get fresh concrete of the

Fig. 2. Pilot scale cold bonding pelletization apparatus.

184 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191

Table 2
Compositions of concretes.

Components [kg/m3 ] R70C R70C+ S70C S70C+ R60LA R60LA+ S60LA S60LA+ R50LA R50LA+ S50LA S50LA+

CEM I 42.5R 350 350 350 350 350 350 350 350 350 350 350 350
Water 175 175 175 175 175 175 175 175 175 175 175 175
S:AA* 3:4 3:4 3:4 3:4 3:4 3:4 3:4 3:4 3:4 3:4 3:4 3:4
Acrylic 7.1 6.2 6.3 5.4 6.8 5.9 6.0 5.1 6.3 6.1 5.7 5.3
admixture
[l/m3 ]

same workability (slump range 170–200 mm) considering the dif- Table 4
Particle size distribution of the two MSWI fly ash samples.
ferent water absorption of artificial aggregates. The whole coarse
fraction (4–18 mm) of designed concrete mixtures is made with Ash origin
the produced cold bonded aggregates. The fine fraction (0–4 mm)
Size Rotary furnace Stoker furnace
has been realized with natural sand. In a sustainable-oriented [␮m]
practice, the amount of sand could be substituted by properly
<1 0.21 1.81
pre-treated industrial by-products, wastes, marble powder from 1–10 8.46 14.02
quarrying activities, etc., as demonstrated by many studies avail- 10–20 19.98 12.22
able in literature [26–33]. In order to evaluate the technological 20–45 44.04 21.49
performance of produced aggregates, the following mechanical 45–80 20.28 19.76
80–120 5.31 14.13
properties have been determined. Concrete cubic specimens (15 cm >120 1.72 16.57
in size) have been employed for the measurement of compressive
strength, according to UNI EN 12390-3:2009 standard. Concrete
dynamic modulus of elasticity has been determined using cylindri- microstructure, dramatically reducing the stabilization potential of
cal specimens (15 cm in diameter and 30 cm in height) by means the system. The results of physical characterization of MSWI fly ash
of ultrasonic pulse measurement, according to UNI EN 13412:2007 are reported in Table 4 and show that significant differences exist
standard. between the two ash samples. Specifically, the resulting volume
weighted mean diameter is 36.6 ␮m for the rotary furnace incin-
3. Experimental results and discussion erator ash sample and 64.4 ␮m for the stoker furnace incinerator
ash sample. The different feed of the rotary and stoker furnaces
3.1. MSWI fly ash chemical and physical characterization and determines, so, differences in both chemical and physical charac-
release behavior teristics. In order to evaluate the environmental hazard related to
MSWI fly ash recycling, the heavy metals release behavior has been
The results of chemical characterization of MSWI fly ash are evaluated on the granular raw waste. Heavy metals concentrations
reported in Table 3. Differences can be observed depending on in the leachate are reported in Table 5, together with Italian regu-
different feed to the two furnaces. Actually, the rotary furnace incin- lation limits for material recovery. Despite the differences in heavy
erator is fed with wastes from industrial processes, while the stoker metals content, the release is similar for the two ash samples, even
furnace incinerator is fed with municipal and hospital wastes. In if rotary furnace fly ash still represents a more relevant issue.
terms of chlorides, sulfates and heavy metals contents rotary fur- Apart from the evaluation of harmful substances such as chlo-
nace fly ash shows higher environmental risks in respect to stoker rides and sulfates and pollutants as heavy metals, it is important
furnace one. As already pointed out in previous sections, stabiliz- to consider also the potential contribution of the MSWI fly ash
ing matrices containing soluble salts such as chlorides and sulfates wastes to the reaction kinetics of the binding matrix. For example,
may exhibit expansive behavior due to alteration of hydration in previous study [4], three industrial wastes, namely weathered
products. These expansive products, in turn, may determine the fly ash, wastewater treatment sludge and desulfurization device
development of both a higher porosity and cracking patterns in the sludge have been employed in the manufacture of cold bonded arti-
ficial aggregates. In that case, authors showed that the waste is able
to contribute to the stabilization/solidification process by giving
Table 3 reaction products such as ettringite and C–S–H. In this study, our
Chemical composition of MSWI fly ash samples and main heavy metals detected.
attention focused on the indirect evaluation of this issue, by means
Component Ash origin of the technological testing of cold bonded aggregates and con-
Rotary furnace Stoker furnace
crete containing the stabilized MSWI fly ash. Considering the data
presented in Table 5, the limits of heavy metals release is danger-
Al2 O3 (wt.%) 7.61 3.89
ously overcome by MSWI fly ash raw waste. This means that, before
Na2 O (wt.%) 3.36 2.62
K2 O (wt.%) 3.85 2.32
CaO (wt.%) 32.11 41.23 Table 5
Fe2 O3 (wt.%) 2.03 1.02 Heavy metals release from raw MSWI fly ash compared to limits provided by in force
MgO (wt.%) 3.24 1.69 legislation.
MnO2 (wt.%) 0.56 0.11
P2 O5 (wt.%) 1.12 0.53 Ash origin
TiO2 (wt.%) 1.85 0.42
Metal Rotary furnace Stoker furnace Limits*(␮g/l)
SiO2 (wt.%) 7.93 11.06
(␮g/l) (␮g/l) * M.D. 5/4/2006
SO3 (wt.%) 9.12 7.12
n.186
Cl− (wt.%) 3.92 1.41
Cu (mg/kg) 4927 2261 Cd 630 540 5
Cd (mg/kg) 62 23 Zn 8270 7320 3000
Pb (mg/kg) 3621 1720 Pb 1860 1530 50
Zn (mg/kg) 4643 2645 Cu 1090 1080 50
Cr (mg/kg) 88 57 Cr 1320 1180 50
 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191
Fig. 3. Artificial aggregates obtained by means of cold bonding pelletization and divided in two main fractions.
Fig. 4. Artificial aggregates obtained starting from unwashed MSWI fly ash showing
excessive expansion.
going into the details of the chemical processes involved in the sta-
bilization/solidification process in terms of chemical nature of the
binding matrices, the overall process, starting from the two-step
pre-washing treatment, needs to be assessed in terms of technical
feasibility.
3.2. Aggregates physical and mechanical characterization
After two-step pre-washing treatment, MSWI fly ash have been
submitted to a pilot scale cementitious cold bonding pelletization
process considering the binding matrices presented in Table 1.
For each mixtures and pelletization session, the obtained artifi-
cial aggregates have been divided in two main fractions: the first
one is finer, with aggregate size ranging from 4 to 12 mm; the sec-
ond one is coarser, with aggregate size ranging from 12 to 18 mm.
Fig. 3 shows the morphology of the obtained fractions of aggregates,
which will be tested separately in terms of physical and mechanical
properties. Cold bonding pelletization gives mostly rounded aggre-
gates, which are highly desirable for a better concrete technology,
since the water demand respect to crushed ones (most of the natu-
ral aggregates) is decreased because a reduction of internal friction
is obtained. Fig. 4 shows the aggregates resulting from unwashed
MSWI fly ash. The raw waste without pre-treatment has been pel-
letized in a preliminary experimental phase to get insight about
the issue related to chlorides and sulfates content. Almost all the
aggregates exhibit a widely cracked surface which is dramatically
different from the surfaces shown in Fig. 3.
Figs. 5 and 6 show the results of density and WAC measurements
separately carried out on the two different size fractions of all the
granules tested. As far as the density is concerned, it is clearly seen
185
that this property increases as the ash content increases. Further-
more, the density is greater for the smaller size fraction. This is
due to the fact that, during the final compacting phase of granula-
tion, water is more easily expelled in the case of smaller granules.
The ash origin has no significant effect on granules density, despite
the greater water requirement in the granules manufacture start-
ing from stoker furnace incinerator ash. By analyzing the results
for the systems containing 70% ash, it is possible to draw the con-
clusion that higher density is obtained when cement is used in the
granules manufacture instead of lime.
The values of density found for the granules manufactured in
this study are such that they can be classified as lightweight aggre-
gate. In fact, commercially available lightweight aggregates (mostly
sintered ones) for civil applications exhibit density in the range
1100–1500 kg/m3 , while typical natural aggregate density ranges
from 2400 to 2800 kg/m3 . The obtained density data are lower than
the limit of 1500 kg/m3 and, particularly, 4–12 mm size aggregates
show an average density of about 1050 kg/m3 . Artificial lightweight
aggregates are so obtained by means of the designed pelletization
process, revealing that at least the reduction of dead load in struc-
tures could be achieved by employing these granules. The values of
WAC are clearly affected by the size of the fraction under consider-
ation. In fact, higher values are found for the 12–18 mm fraction.
This result is in agreement with those relative to density, inas-
much as less dense (more porous) granules are obviously able to
absorb water to a greater extent. It is seen that lime-based systems
have greater WAC than those cement-based, again in agreement
with density data. When compared to the commercial lightweight
aggregate referred to above, the values of WAC found for the gran-
ules tested in this study lay within a range whose limits are slightly
lower. In fact, WAC generally ranges between 10 and 18% for com-
mercial artificial aggregate, while, for the granules manufactured
in this study, the range is 8.9–15.5% for one-step granulation and
7.3–13.7% in the case of two-step granulation, which, of course,
reduce the porosity of the external part of the matrix which absorbs
water. Since the second step of pelletization is carried out with only
binder, this results in a lower value of WAC. These values of WAC are
interesting because aggregates with such absorption capacity could
be employed for internal curing, which in the last two decades has
been proposed as a mitigation technique for shrinkage phenomena
in concrete [e.g. 34,35].
Crushing strength and Los Angeles coefficients of the stabilizing
aggregates are reported in Figs. 7–9 , respectively. As in the case
of commercial alternatives, crushing strength of our lightweight
aggregates relies in the range of 2–6 MPa. Mechanical results
are fundamental to understand two main factors: (i) the maxi-
mum amount of waste that can be recycled in the cold bonded
artificial aggregates; (ii) the service conditions (load bearing vs non-
structural applications) in which the material can be employed and,
as a consequence, the achievable market. The combination of these
two factors determines the final volume of MSWI fly ash that can
186 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191

Fig. 5. Density of cold bonded artificial aggregates (sign “+” indicates aggregates realized with a double-step pelletization).

be recycled and total volume of land saved by avoiding landfilling.
It is generally observed in Fig. 7 that the smaller fraction of cold
bonded aggregates (4–12 mm) which is also the more dense/less
porous, exhibits much higher strength than the coarser one: for
smaller aggregates crushing strength tends to increase by a factor
of 2. Looking at different systems presented in Table 1, it is gener-
ally observed that, for each system, the second step of pelletization
(see systems indicated with “+” sign in Fig. 7) always determines
a relevant increase in crushing strength, which is more significant
for cold bonded aggregates with lower strength. The cement-based
systems are able to maximize simultaneously the amount of MSWI
fly ash recycled (up to 70%) and the mechanical performance, reach-
ing a maximum value for S70C+ equal to 6.35 MPa. Lime-based
systems (R70L and S70L) are the ones exhibiting minimum values
of strength. In that case, anyway, the beneficial effect of the second
pelletization step is more significant.
Mechanical degradation resistance, represented by Los Ange-
les coefficient values (plotted in Fig. 8), shows a similar trend to
that of crushing strength, as expected. According to Table 9 of UNI
EN 13242:2008 (Italian-European standard “Aggregates for unbound
and hydraulically bound materials for use in civil engineering work and
road construction”), all the produced aggregates can be classified
LA50 category, except for R70L, R70L+, S70L and S70L+ which can be
instead classified as LA60 . As far as the relationship between crush-
ing strength and Los Angeles coefficient is of concern, we can find
that, as an example, the class of aggregates with highest crushing
resistance (S70C+) show also the best LA class. Furthermore it can be
observed that, for the same mixture, the second pelletization step

Fig. 6. WAC of cold bonded artificial aggregates (sign “+” indicates aggregates realized with a double-step pelletization).
 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191 187

Fig. 7. Crushing strength of cold bonded artificial aggregates (sign “+” indicates aggregates realized with a double-step pelletization).

improves the performance of the specimens. The parameter repre-
sented by Los Angeles coefficient is fundamental when aggregates
are proposed for the construction of roads, where the resistance to
abrasion and mechanical degradation are key parameters.
3.3. Heavy metals release from cold bonded aggregates and
microstructural characterization
Table 6 shows the results of the leaching test carried out on
the size fraction 12–18 mm of the granules manufactured starting
from the rotary furnace incinerator ash. The size fraction 4–12 mm
has been excluded from this test because the granules have shown
better physical properties (higher density, lower porosity). More-
over, all the samples obtained from stoker furnace incinerator ash
have not been tested for leaching behavior. The reason for this
is that granules based on this ash proved to have better physical
and mechanical properties (density, WAC, crushing strength, Los
Angeles coefficient) in relation to civil engineering applications. In
addition, as can be seen in Table 5, stoker furnace incinerator ash
is poorer of heavy metals and, so, has a reduced environmental
concern respect to MSWI fly ash coming from rotary furnace. The
metal release values of Table 6 are cumulative and relative to three
times: 3, 14 and 28 days. According to Italian regulation, only the
cumulative release value after 3 days is to be taken into account
for comparison with the limit fixed for each metal. The compliance
leaching test employed for the granules was extended to 14 and

Fig. 8. Los Angeles coefficient of cold bonded artificial aggregates (sign “+” indicates aggregates realized with a double- step pelletization).
188 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191

Table 6
Cumulative release of heavy metals from cold bonded aggregates.

Cd, ␮g/l Limit = 5 Zn, ␮g/l Limit = 3000 Pb, ␮g/l Limit = 50 Cu, ␮g/l Limit = 50 Cr, ␮g/l Limit = 50

Time [days] Time [days] Time [days] Time [days] Time [days]

Aggregates 3 14 28 3 14 28 3 14 28 3 14 28 3 14 28 PH

R70C 3.07 3.89 4.13 2.13 2.49 2.71 12.8 15.3 21.2 27.1 32.2 38.4 32.3 38.6 43.1 11.08
R70C+ 1.24 1.81 1.93 0.67 0.79 0.91 3.25 4.47 4.92 10.4 12.3 15.7 15.8 19.1 22.8 10.47
R70L 6.34 7.76 8.11 3.62 4.31 4.87 47.1 54.5 68.8 41.3 59.5 68.7 44.5 65.3 75.2 10.97
R70L+ 1.02 1.68 1.85 0.92 1.02 1.09 3.15 3.74 4.13 13.4 16.2 19.5 23.5 25.2 28.4 10.24
R60LA 3.19 3.92 4.23 2.31 2.38 2.93 12.6 17.3 19.1 24.3 26.9 29.8 28.3 32.1 34.7 11.03
R60LA+ 1.05 1.78 1.89 0.58 0.73 0.78 3.31 4.32 4.67 9.12 9.87 11.4 14.2 18.5 19.1 10.15
R50LA 3.56 4.62 4.91 2.21 2.48 2.83 13.3 14.9 20.3 23.2 28.1 31.7 29.1 36.4 38.2 11.14
R50LA+ 1.08 1.77 1.89 0.61 0.70 0.86 3.18 3.76 4.15 10.5 13.6 15.2 17.1 22.9 24.3 10.92

of binder employed, referring back to physical and mechanical

Fig. 9. Optical microscope image showing the external layer made of only binder.
28 days to have a better understanding of the long term leaching
behavior of the granules.
As the leaching test is of compliance type, the first observation
on the data of Table 6 must be related to the cases in which the
limits are exceeded. This only happens for 3 days leaching results
for Cd and Zn in system R70L. In all the other cases the cumulative
amounts of metals released after 3 days are below the relative lim-
its. As expected, the release is much lower in the case of granules
manufactured via the two-step granulation. Metal release is also
affected by the type of binder. In fact, cement-based granules con-
taining 70% ash behave better than the corresponding lime-based
ones. This is quite an expected result, as physico-mechanical prop-
erties of the granules based on lime alone as binder revealed to be
of the worst rank. The effect of ash content on leaching behavior
does not seem significant at least in the composition range inves-
tigated. A final consideration can be made with regard to release
at times longer than 3 days (14 and 28 days). It is seen that in all
the cases investigated the release at these longer times increases
over that at 3 days but to a limited extent, and this is a very positive
result in view of long-term behavior in civil applications. Regarding
relationship between leaching behavior and physico-mechanical
properties, it can be argued that the system with highest heavy
metal release, namely R70L, is also the one with worst physico-
mechanical performance.
A more detailed picture respect to Fig. 1 has been obtained by
means of optical microscopy. Fig. 9 shows the interface between
the inner core of the aggregate (binder + waste) and the exter-
nal layer (only binder). The maximum size of the external layer
ranges from 1.0 to 2.0 mm. It is important to observe the signifi-
cantly different volumes of waste containing phase and only binder
phase. The second one is quite limited. Despite the low amount
performance of aggregates, the influence of second pelletization
step is much relevant. Considering that maximum size of cold
bonded aggregates is 18 mm, it is demonstrated here that even
if the external layer width is one order of magnitude smaller
than inner phase, it still determines a significant improvement
of engineering and stabilizing performance, as shown also in
Table 6.
Fig. 10 shows a SEM image which is able to give more detailed
information about the interface and different porosities between
internal and external layer resulting from second step of pelleti-
zation process. Fig. 10b shows the interface between the inner
stabilizing matrix (Fig. 10a) and the external layer (Fig. 10c). The
higher porosity of the waste containing matrix is clearly shown.
Referring back to all images provided (Figs. 1, 9 and 10), going
from macrostructure to microstructure through mesostructure, all
information inferred from physical and mechanical testing have
been confirmed, revealing the efficiency of the designed process in
giving artificial aggregates of satisfactory engineering performance
and good stabilization of heavy metals.
3.4. Concrete physical and mechanical characterization
Technological properties of concrete designed with artificial
aggregates can be found in Fig. 11 (density), Fig. 12 (compres-
sive strength) and Fig. 13 (dynamic modulus of elasticity). The
specimens have been cured for 28 days at room temperature and
100% R.H. According to UNI EN 206-1:2006 (Italian–European stan-
dard “Concrete – Part 1: Specification, performance, production and
conformity”) concretes containing R70C, R70C+, S70C and S70C+
aggregates belong to density class D2.0 while the remaining one
belong to D1.8 class (Fig. 12).
All the concretes designed, as the artificial aggregates, can be
classified as lightweight concrete, which could be suitable for pur-
poses such as dead load reduction, thermal/acoustic insulation, etc.
Unconfined compressive strength of cubic specimens ranged from
15 to 35 MPa, revealing that structural elements with average per-
formance could be obtained. It can be observed that, concerning
the influence of aggregates mechanical performance on concrete
behavior, the highest value of compressive strength is obtained for
S70C+ which is also the class of aggregates with best mechani-
cal properties. As far as the modulus of elasticity is concerned, it
should be observed that typical values for concrete blocks contain-
ing natural aggregates range from 25 to 50 GPa. We can observe
that the three concrete systems with lowest compressive strength
values (namely R70C, R60LA and R50LA) are at the same time the
ones with lowest dynamic modulus of elasticity, even if the rela-
tion is of course nonlinear. This trend loses in significance for higher
compressive strength values (Fig. 13).
 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191 189

Fig. 10. SEM image of an artificial aggregate made by means of double-step pelletization showing the internal waste-containing phase (10.a), the interface (10.b), and the
external layer (10.c) obtained by means of double-step cold bonding pelletization.

Fig. 11. Density of concrete containing cold bonded artificial aggregates.

Fig. 12. Compressive strength of concrete containing cold bonded aggregates.

190 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191
Fig. 13. Dynamic modulus of elasticity of concrete containing cold bonded aggregates.
4. Conclusions
By observing the set of experimental data that have been previ-
ously discussed, the following conclusions can be drawn:
• MSWI fly ash from municipal, hospital and industrial solid wastes
incineration can be successfully recycled up to 70% content in the
production of cold bonded artificial aggregate;
• The innovative approach of double-step pelletization allows to
improve both physico-mechanical and stabilization properties of
binding matrices for any mixture studied; this technique pro-
duces an encapsulating outer layer which is made of waste-free
binder and, so, it is able to reduce the porosity, the heavy metals
release and improve the physical and mechanical properties;
• Different binders can be used for the production of the granules,
but only cement and coal fly ash/lime systems give satisfac-
tory engineering performance; lime-based systems proved to be
unsuitable as granules binder;
• Comparing the two MSWI fly ash samples employed in this study,
coming from stoker and rotary furnaces, respectively, the former
was able to give granules of better technological properties;
• The influence of aggregates size on physical and mechanical prop-
erties is in line with other literature studies [e.g.,1];
• In the majority of the cases tested, the release of heavy metals
was below the limits fixed by Italian law for civil engineering
applications. Exceptions to this have only been observed when
lime alone is used as a binder for the production of the granules;
• The cold bonded aggregates designed and tested in this study
have proved to be suitable for the manufacture of concrete-
based products. Depending on process variables (type of binder,
granules composition, type of pelletization etc.), it is possible to
produce artificial aggregates of satisfactory physico-mechanical
properties for concrete-based facilities of average performance;
• The upscale of the here used pilot scale pelletization equipment
does not pose specific technological issues and, furthermore,
it is economically feasible, allowing to use this technology in
developing countries to improve resource efficiency and reduce
landfilling.
Further studies will focus on the durability aspects and interplay
between long term development of mineralogical and mechanical
properties of stabilizing matrices and long term leaching behavior.
References
[1] G. Baykal, A.G. Döven, Utilization of fly ash by pelletization process; theory:
application areas and research results, Resour. Conserv. Recycl. 30 (1) (2000)
59–77.
[2] R. Cioffi, F. Colangelo, F. Montagnaro, L. Santoro, Manufacture of artificial
aggregate using MSWI bottom ash, Waste Manage. 31 (2) (2011) 281–288.
[3] F. Colangelo, R. Cioffi, Use of cement kiln dust, blast furnace slag and marble
sludge in the manufacture of sustainable artificial aggregates by means of
cold bonding pelletization, Materials 6 (2013) 3139–3159.
[4] V. Ducman, B. Mirtič, The applicability of different waste materials for the
production of lightweight aggregates, Waste Manage. 29 (8) (2009)
2361–2368.
[5] V. Ducman, B. Mirtič, Water vapour permeability of lightweight concrete
prepared with different types of lightweight aggregates, Constr. Build. Mater.
68 (2014) 314–319.
[6] C. Ferone, F. Colangelo, F. Messina, F. Iucolano, B. Liguori, R. Cioffi, Coal
combustion wastes reuse in low energy artificial aggregates manufacturing,
Materials 6 (11) (2013) 5000–5015.
[7] S. Geetha, K. Ramamurthy, Environmental friendly technology of cold-bonded
bottom ash aggregate manufacture through chemical activation, J. Clean.
Prod. 18 (15) (2010) 1563–1569.
[8] M. Gesoglu, E. Güneyisi, H.Ö. Öz, Properties of lightweight aggregates
produced with cold-bonding pelletization of fly ash and ground granulated
blast furnace slag, Mater. Struct. 45 (2012) 1535–1546.
[9] P. Gomathi, Sivakumar, A, Accelerated curing effects on the mechanical
performance of cold bonded and sintered fly ash aggregate concrete, Constr.
Build. Mater. 77 (2015) 276–287.
[10] E. Güneyisi, M. Gesoğlu, İ. Altan, H.Ö. Öz, Utilization of cold bonded fly ash
lightweight fine aggregates as a partial substitution of natural fine aggregate
in self-compacting mortars, Constr. Build. Mater. 74 (2015) 9–16.
[11] E. Güneyisi, M. Gesoğlu, E. Booya, K. Mermerdaş, Strength and permeability
properties of self-compacting concrete with cold bonded fly ash lightweight
aggregate, Constr. Build. Mater. 74 (2015) 17–24.
[12] P. Scanferla, G. Ferrari, R. Pellay, A.V. Ghirardini, G. Zanetto, Libralato, G, An
innovative stabilization/solidification treatment for contaminated soil
remediation: demonstration project results, J. Soils Sediments 9 (3) (2009)
229–236.
[13] V. Vasugi, K. Ramamurthy, Identification of design parameters influencing
manufacture and properties of cold- bonded pond ash aggregate, Mater. Des.
54 (2014) 264–278.
[14] V. Vasugi, K. Ramamurthy, Identification of admixture for pelletization and
strength enhancement of sintered coal pond ash aggregate through
statistically designed experiments, Mater. Des. 60 (2014) 563–575.
[15] C.R. Cheeseman, A. Makinde, S. Bethanis, Properties of lightweight aggregate
produced by rapid sintering of incinerator bottom ash, Resour. Conserv.
Recycl. 43 (2) (2005) 147–162.
 F. Colangelo et al. / Journal of Hazardous Materials 299 (2015) 181–191
[16] J. Chen, D. Yang, W. Tang, S.Y. Wang, Producing synthetic lightweight
aggregates from reservoir sediments, Constr. Build. Mater. 28 (2012) 387–394.
[17] C. Ferone, F. Colangelo, F. Messina, L. Santoro, R. Cioffi, Recycling of
pre-washed municipal solid waste incinerator fly ash in the manufacturing of
low temperature setting geopolymer materials, Materials 6 (8) (2013)
3420–3437.
[18] Y.L. Galiano, C.F. Pereira, J. Vale, Stabilization/solidification of a municipal
solid waste incineration residue using fly ash-based geopolymers, J. Hazard.
Mater. 185 (1) (2011) 373–381.
[19] L. Zheng, W. Wang, Y. Shi, The effects of alkaline dosage and Si/Al ratio on the
immobilization of heavy metals in municipal solid waste incineration fly
ash-based geopolymer, Chemosphere 79 (6) (2010) 665–671.
[20] L. Zheng, C. Wang, W. Wang, Y. Shi, X. Gao, Immobilization of MSWI fly ash
through geopolymerization: effects of water-wash, Waste Manage. 31 (2)
(2011) 311–317.
[21] I. Kourti, D.A. Rani, D. Deegan, A.R. Boccaccini, C.R. Cheeseman, Production of
geopolymers using glass produced from DC plasma treatment of air pollution
control (APC) residues, J. Hazard. Mater. 176 (1) (2010) 704–709.
[22] C. Lampris, R. Lupo, C.R. Cheeseman, Geopolymerisation of silt generated from
construction and demolition waste washing plants, Waste Manage. 29 (1)
(2009) 368–373.
[23] F. Messina, C. Ferone, F. Colangelo, R. Cioffi, Low temperature alkaline
activation of weathered fly ash: influence of mineral admixtures on early age
performance, Constr. Build. Mater. 86 (1) (2015) 169–177.
[24] B. Molino, A. De Vincenzo, C. Ferone, F. Messina, F. Colangelo, R. Cioffi,
Recycling of clay sediments for geopolymer binder production. A new
perspective for reservoir management in the framework of Italian legislation:
the Occhito reservoir case study, Materials 7 (8) (2014) 5603–5616.
191
[25] F. Colangelo, R. Cioffi, F. Montagnaro, L. Santoro, Soluble salt removal from
MSWI fly ash and its stabilization for safer disposal and recovery as road
basement material, Waste Manage. 32 (6) (2012) 1179–1185.
[26] B. Taha, G. Nounu, Properties of concrete contains mixed colour waste
recycled glass as sand and cement replacement, Constr. Build. Mater. 22 (5)
(2008) 713–720.
[27] H.Y. Wang, A study of the effects of LCD glass sand on the properties of
concrete, Waste Manage. 29 (1) (2009) 335–341.
[28] A. Sales, Lima, A, Use of Brazilian sugarcane bagasse ash in concrete as sand
replacement, Waste Manage. 30 (6) (2010) 1114–1122.
[29] Z.Z. Ismail, E.A. Al-Hashmi, Use of waste plastic in concrete mixture as
aggregate replacement, Waste Manag. 28 (11) (2008) 2041–2047.
[30] K.S. Al-Jabri, M. Hisada, S.K. Al-Oraimi, Al-Saidy, AH, Copper slag as sand
replacement for high performance concrete, Cem. Concr. Compos. 31 (7)
(2009) 483–488.
[31] A. Shayan, A. Xu, Value-added utilisation of waste glass in concrete, Cem.
Concr. Res. 34 (1) (2004) 81–89.
[32] B. Taha, G. Nounu, Utilizing waste recycled glass as sand/cement replacement
in concrete, J. Mater. Civ. Eng. 21 (12) (2009) 709–721.
[33] F. Iucolano, B. Liguori, D. Caputo, F. Colangelo, Cioffi, R, Recycled plastic
aggregate in mortars composition: Effect on physical and mechanical
properties, Mater. Des. 52 (2013) 916–922.
[34] D.P. Bentz, Snyder, KA, Protected paste volume in concrete extension to
internal curing using saturated lightweight fine aggregate, Cem. Concr. Res.
29 (11) (1999) 1863–1867.
[35] D. Cusson, T. Hoogeveen, Internal curing of high-performance concrete with
pre-soaked fine lightweight aggregate for prevention of autogenous
shrinkage cracking, Cem. Concr. Res. 38 (6) (2008) 757–765.