Resources, Conservation & Recycling 129 (2018) 102–111

Contents lists available at ScienceDirect

Resources, Conservation & Recycling

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

Full length article

Waste tire recycling process for production of steam activated carbon in a MARK
pilot plant
Antonio Molinoa, , Antonio Donatellib, Tiziana Marinoc, Alfredo Aloised, Juri Rimauroa,
⁎

Pierpaolo Iovanea
a
ENEA – Italian National Agency for New Technologies, Energy and Sustainable Economic Development, Territorial and Production System Sustainability Department,
Piazzale Enrico Fermi, 1, 80055 Portici, NA, Italy
b
ENEA – Italian National Agency for New Technologies, Energy and Sustainable Economic Development, Territorial and Production System Sustainability Department, SS.
7 Appia, km 706+00, 72100 Brindisi, Italy
c
CNR, Institute on Membrane Technology, National Research Council, Via Pietro Bucci, Cubo 17C, c/o Campus of University of Calabria, 87036 Rende, CS, Italy
d
University of Calabria, Department of Environmental and Chemical Engineering, Via P. Bucci, Cubo 45A, 87036 Arcavacata di Rende, CS, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: This work is focused on the production of activated carbon from waste tires by means of a two-steps process:
Gasification steam gasification of starting material followed by steam activation of char. In particular the evaluation of the
Tires most appropriate hold time value for the activation step, in terms of quality and amount of activated carbons
Activated carbon produced, and the energy balance of the overall process were performed. The hold time was choose as the best
Pilot plant
compromise between a high surface area of final product and a not excessively low burn-off value. Investigations
Rotary kiln
were carried out through seven experimental tests on a pilot plant based on a rotary kiln reactor. Studies per-
formed pointed out that, when other process parameters were kept constant (in steam gasification: temperature
is 850 °C, steam to feed ratio is 1, mean residence time of material is 6 min, carrier flowrate of N2 is 0.9 Nm3h−1;
in steam activation: temperature is 920 °C, steam to char ratio is 2, carrier flowrate of N2 is 1 Nm3h−1), a hold
time in activation step set to 3 h is the best choice, resulting in a surface area of 786 m2·g−1 and a burn-off value
equal to 78.4%. Moreover, the energy balance of the overall process which goes from waste tires to activated
carbon was performed: if 1 kg·h−1 of waste tires are gasified and activated for three hours, an input power of
3.5 kW is requested. After the experimental tests, a theoretical model was developed, based on the Arrhenius
equation whose parameters were determined from a sigmoidal regression of the experimental data. The theo-
retical model showed results in good agreement with the experimental data, especially for hold times greater
than 2 h, and can represent a useful tool to make decision choices.

1. Introduction treatment technologies such as pyrolysis and gasification, are some of

Europe produces about 355 million of tires per year, corresponding
to 24% of the total world production (ETRMA, 2011), of which the end
of life tires (ELTs) became about 2.6 million tons subtracting the re-
used, exported and reconstructed amounts (about 5%, 4% and 9% re-
spectively). ELTs represent a relevant problem within the waste man-
agement strategy of the European Community and, despite the attempts
of repurposing waste tires in many different ways, a relevant fraction
(nearly 13%), is still landfilled.
Apart from this, Europe is also one of the most active areas in the
world for the recovery of ELTs (about 95%, even if Italy and the Czech
Republic are still between 70% and 90%) (ETRMA, 2014). Thermal
the proven solutions for high-efficiency material and energy recovery
from ELTs (Antoniou et al., 2014; Choi et al., 2014; Karatas et al., 2012;
Donatelli et al., 2010; Kandasamy and Gokalp, 2015; Portofino et al.,
2011; Portofino et al., 2013; Sainz-Diaz and Griffiths, 2000). In parti-
cular, the sustainability of the whole process, on both pilot and in-
dustrial scales, is guaranteed if a valuable exploitation of the solid by-
products is implemented.
More in detail, gasification is a partial oxidation of feedstock, nor-
mally performed at higher temperatures than pyrolysis which instead
occurs in inert conditions. These technologies convert the intrinsic
chemical energy of a carbon-rich material into intermediate products:
syngas (a fuel gas), oil and solid residue (char). Types and yields of

⁎
Corresponding author.
E-mail addresses: antonio.molino@enea.it (A. Molino), antonio.donatelli@enea.it (A. Donatelli), tiziana.marino@yahoo.it (T. Marino), alfredo.aloise@unical.it (A. Aloise),
juri.rimauro@enea.it (J. Rimauro), pierpaolo.iovane@enea.it (P. Iovane).

http://dx.doi.org/10.1016/j.resconrec.2017.10.023

0921-3449/ © 2017 Elsevier B.V. All rights reserved.

Received 20 June 2017; Received in revised form 17 October 2017; Accepted 17 October 2017
A. Molino et al.
Nomenclature
cp Specific heat (kJ·kg−1·K−1)
D Diameter of the kiln (m)
E Activation energy (kJ·mol−1)
G Mass-flow rate (kg·h−1)
kI Surface reaction rate constant (s−1)
L Length of the kiln (m)
m Reaction order
M Mass (kg)
MRT Mean residence time (min)
n Rotational speed (min−1)
HHV Higher heating value (kJ·kg−1)
LHV Latent heating vaporization (kJ·kg−1)
Q̇ Thermal power (kW)
R Molar constant gas (J·mol−1·K−1)
S(Xc) Transient specific surface area (m2∙g−1)
SCR Steam to char ratio (h−1)
SFR Steam to feed ratio
these products can be adjusted by changing the operating parameters
(Portofino et al., 2013). The char, containing 85 wt% carbon, can be
valuable as raw material both as a filler for reuse in the rubber industry
or in the asphalt. As filler, the results showed that the char has similar
properties to a semi-reinforcing filler with prospects for manufacturing
process of the new tire (Shah et al., 2006; Martınez et al., 2013).
Other major application can be its use as activated carbon. Many
researchers studied activated carbon deriving from char tire pyrolysis
with BET surface areas up to 1000 m2·g−1 under well-established pro-
cess conditions (Ariyadejwanich et al., 2003; Mui et al., 2004; Saleh and
Danmaliki, 2016). Usually, the whole process consists of two steps:
thermal pyrolysis at a relatively low temperature (400–700 °C) under
nitrogen or helium gas and followed by activation with an oxidizing
agent at 800–1000 °C. Gonzalez et al. (Gonzalez et al., 2006) prepared
activated carbons with surface areas of 1317 and 496 m2·g−1 by acti-
vation of pyrolytic tire char at 900 and 850 °C in a steam/N2 mixture
and CO2, respectively. The activation time and the nature of the oxi-
dizing agent affected the porosity and the surface area of the activated
carbons while the raising of the temperature had a lower effect. Many
procedures are described in the literature for activating waste tires (Mui
et al., 2004) which suggest that the steps consist of a first micropores
formation, followed by pores enlargement until to crash of the carbo-
naceous structure. Moreover, using steam as activating agent, the ac-
tivated carbons show higher total micropore volumes and BET surface
areas than those produced using CO2; so the steam activation shows
best results in terms of the external surface areas but the process is
energy consuming growing when the activation time increases (Mui
et al., 2004; Gonzalez et al., 2006). On the other hand, the porosity of
the produced activated carbons usually enhances when burn off in-
creases and the carbonaceous adsorbent yield reduces. Higher activa-
tion temperatures and longer hold times should give a lower activated
carbon yield, but leads to higher BET surface area, higher volume and
larger pore sizes (Hadi et al., 2016).
So, it is mainly important to understand the char activation devel-
opment and in particular the reaction kinetics and pore structure evo-
lution. Few researchers studied the kinetics of tires char activation and
some suggested models were considered useful to understand the
gas–solid reaction kinetics between activating gas and char (Mui et al.,
2004; Guo and Lua, 2001).
Physical activation using CO2 or steam as oxidizing agents are the
most commonly used processes in the production of tyre carbons but
the aim of our research is to get good results in terms of the external
surface areas of the carbons assessing also energy consuming of the
process. Moreover, even if the production of activated carbons from
103
Resources, Conservation & Recycling 129 (2018) 102–111
T Temperature (°C)
Xc Moles of carbon (moles)
Greek symbols
β Kiln slope (°)
Δt Hold time (min)
θ Angle of repose (°)
Subscript
a Activation
ac Activated carbon
c Condensable
g Gasification
in Input
out Output
sg Syngas
st Scrap tires
waste tires by means of thermal treatments (gasification) has been
widely studied and experimented, few experimental activities on pilot
or preindustrial scale are available in Literature. The novelty of our
work is the optimization of the activated carbon production from waste
tires on a pilot plant, in view of industrial exploitation. In particular,
the main target of our experimental activities is to study how hold time
affects the BET surface area and the burn-off of activated carbon pro-
duced from waste tires, when other process parameters (mainly tem-
peratures and steam-on-char ratio) were kept constant, in order to find
the best compromise between the surface activity and the amount of
activated carbon obtained. The experimental data help us to propose a
kinetic model for the theoretical study of carbon activation process.
Moreover, with regard to the best compromise found, an energy bal-
ance of the process as a whole (gasification and activation steps) is
performed.
2. Materials and apparatus
2.1. Feedstock matrix
The chosen feedstock for experiments have been rubber grain de-
rived from waste tires, already employed in previous experimental
works (Donatelli et al., 2010; Portofino et al., 2011; Portofino et al.,
2013). The reinforced fibers and the steel belts have been extracted and
removed from waste tires, and then samples have been shredded to
1–2 cm particle size diameter by a cold mechanical grinding. Then, the
rubber grains have been stored under dry conditions before the ex-
perimental tests.
Samples of rubber grain were characterized by an elementary ana-
lyzer EA 1110 CHNS-O (ThermoQuest) which was used to determine
the mass fraction of the carbon, hydrogen, nitrogen, sulfur and oxygen
(in weight%) in agreement with ASTM D5373-08 and ASTM D5016-08.
The material was exposed to a temperature of 1000 °C under an oxygen
atmosphere. The produced gases (CO2, SO2, N2 and H2O) were sepa-
rated in chromatography columns and then measured by selective in-
frared (IR) and thermal conductivity detectors (TCD). Proximate ana-
lysis was investigated with a thermo-balance 2950 (TA Instruments)
which was used to determine the fixed carbon and volatile organic
matter (in dry weight%) in agreement with ASTM D3175-77, while the
moisture content and ash were determined in agreement with ASTM
D3173-11 and ASTM D2866-11. The higher heating value of the waste
tires was investigated by a calorimetric bomb IKA C5000: it determines
the heat produced by the combustion of the sample under examination,
and absorbed by a known mass of water where temperature rise is
A. Molino et al. Resources, Conservation & Recycling 129 (2018) 102–111

checked in agreement with ASTM D5865-13.
The measurement accuracy is guaranteed by the quality of instru-
ments and a deep knowledge of the methods, acquired in particular on
characterization of different types of feedstock in our laboratories
(Molino et al., 2014).
2.2. Pilot plant
The experimental activities were performed on a pilot plant based
on a rotary kiln reactor having an internal diameter of 400 mm, a he-
ated length of 1000 mm and already employed in previous works
(Donatelli et al., 2010; Iovane et al., 2013; Molino et al., 2013). Table 1
summarizes the main geometrical and physical parameters of the
equipments composing the pilot plant and its schematic depiction is
shown in Fig. 1.
It consists of a feed hopper in which the feeding material, waste tires
in this case, is collected and after moved to the rotary kiln by a screw
conveyor. In the electrically heated rotary kiln, HT11-S by PLEQ, the
process takes place with the production of a solid residue and a gaseous
product (syngas). The rotary kiln is slightly tilted and was keeping in
rotation, to allow the solid feedstock to move from the input to the exit
sides.
The mean residence time (MRT) of the solid matrix in the rotary kiln
can be adjusted by changing appropriately the slope and the speed of
the kiln, according to the formula (Liu and Specht, 2006):
L
MRT =
π·D ·n·tanβ
A previous work (Donatelli et al., 2015) performed with a similar
rotary kiln (same ratio L/D) has shown a good accordance between
experimental and theoretical MRTs calculated with the Eq. (1).
At the outer of the kiln the solid residue, composed by large parti-
cles char, is collected into a tank, while the syngas follows the gas
treatment sections composed by a spray tower for removing tar frac-
tion, condensable and fine particles, and a scrubber with an alkaline
solution for removing acid components. Then the gas passes through a
hydraulic guard, to ensure a slight overpressure in order to avoid un-
desired infiltrations of air from outside, and then is sent to an oxidizer
in which is burned. The oxidizer comprises a heat exchanger in which
the heat from the combustion of the syngas is usefully employed for the
production of the steam sent to the rotary kiln reactor. The water steam
employed in the process was obtained by extracting water from the
water supply, directly. The water at room temperature (about 15 °C) is
sent directly into the heat exchanger, where the heat from the com-
bustion of the syngas is useful employed to produce saturated water
steam (about 100 °C and p = 1 atm). A valve and a volumetric counter
installed on the water adduction circuit for monitoring the set Steam-
Feed Ratio (SFR) value, defined as the ratio between the mass flow-rates
of steam and tires, controlled the steam mass flow-rate. The saturated
water steam is then sent into the heated rotary kiln, in co-current flow
with N2 and scrap tires flows.
Moreover, the torch can be also fed with Liquefied Petroleum Gas
(LPG) as support fuel in starting phase. Nitrogen is also employed
during the experiments, as carrier gas or sent to mechanical seals for
preventing gas leakage from the rotary kiln reactor to the external. The
pilot plant is equipped with some volumetric counters for the mea-
surement of flow-rates of syngas, steam and nitrogen. In addition,
samples of syngas before the burning step are collected for the eva-
luation of the volumetric composition in a micro gas chromatograph
Agilent A3000. For safety reason, an emergency line allows to send
directly the syngas to the torch in case of overpressure (upper limit of
pressure is set with a hydraulic guard).
3. Methods
3.1. Experimental
3.1.1. Gasification procedure
The rotary kiln was electrically heated while N2 was continuously
(1)
sent inside (volumetric flow rate of N2 = 0.9 Nm3·h−1) in order to
realize a complete inert atmosphere and when the chosen inner tem-
perature was reached (850 °C), the waste tires previously loaded into
the feed hopper was moved to the rotary kiln by a screw conveyor. The
water steam was produced as described in Section 2.2, and with the
desired stem to feed ratio (SFR) value defined as SFR = G H2 O/ Gst .
The rotary kiln was kept in clockwise rotation in order to discharge
the feeding material and with a slope and a revolution speed that assure
the desired residence time, according with Eq. (1). The solid residue
was collected in the tank placed just after the rotary kiln reactor, while
the syngas was treated as described in paragraph 2.2.
3.1.2. Activation procedure
The activation procedure was similar to the gasification, with some
basic differences hereinafter described.

Table 1
Geometric and physical parameters of the pilot plant.

Process operation Equipment Streams Geometric shape Size T [°C] P [mbar]

enter exit enter exit

Steam Gasification/ Steam Feed hopper Waste tires parallelepiped L=1m 15 15 1000 1000
Activation H = 0.6 m
Screw conveyor Waste tires helicoid 15 450 1000 1010
Rotary kiln (model HT11-S) Waste tires, H2Og, N2, syngas cylindrical D = 0.4 m 850/920 850/920 1010 1010
L=1m
Solid collection Tank Char cylindrical D = 0.6 m 500 1010
L = 0.8 m
Gas Treatment Water spray tower Large particles, syngas, H2Ol cylindrical D = 0.5 m 800 20 1010 1008
L=2m
Alkaline solution syngas, NaOH/H2Ol cylindrical D = 0.5 m 20 20 1008 1006
L=2m
Oxidizer Torch Syngas, LPG, air cylindrical D = 0.2 m 20 800 1006 1000
L=2m
Auxiliaries Hydraulic guard (on process line) H2Ol cylindrical D = 0.2 m 20 20 1006 1000
L = 0.4 m
Hydraulic guard (on emergency H2Ol cylindrical D = 0.2 m 20 20 1050 1000
line) L = 0.4 m
Heat exchanger H2O(from liquid to gas helicoid D = 0.02 m 15 120 1000 1000
phase) L = 10 m
p = 0.2 m

104
A. Molino et al. Resources, Conservation & Recycling 129 (2018) 102–111

Fig. 1. Schematic diagram of rotary kiln pilot plant.

First of all the char was introduced into the rotary kiln only when 1/(m − 1)
Xc = 1 − ⎛ ⎞
1
⎜ ⎟

⎝ 1 + (m − 1) k t ⎠
the temperature rose to the selected value. During this step, and until I
(5)
the end of the experimental test, the rotary kiln was kept in counter-
clockwise rotation, in order to avoid the discharging of the feeding With regard to the surface area, the proposed relation between the
material. BET value and the hold time is based on a random pore model:
The water steam was produced as described in Section 2.2, and with a
the desired Steam to Char ratio (SCR) value. Defined as BET =
(Xc − X0)
SCR = G H2 O/ Mchar . 1 + exp −( b ) (6)
The test was performed as long as the desired hold time was
reached, and then the solid residue (activated carbon) was discharged where the parameters a, b and X0 can be obtained from a sigmoidal
into a tank by switching the rotation of the kiln (from counter-clockwise regression of the experimental data.
to clockwise rotation). The produced gas followed the path already
described for gasification. For each test, the burn-off of solid was cal- 4. Results and discussion
culated, in according with the following equation:
The waste tires ultimate and proximate analysis and the corre-
burn − off (%) = (Mchar − Mactivatedcarbon )/Mchar ·100 (2)
sponding higher heating value measurements are shown in Table 2.
and the activated surface area of the solid residue was determined with The data indicate a very high carbon content (up to 80% in weight),
Micromeritics ASAP 2020 (Accelerated Surface Area and Porosimetry a low ash quantity of 5% in weight, a very low moisture content (down
System). to 1% in weight), a high calorific value (over 35 MJ∙kg−1) and a sulfur
content of about 2% in weight. These features make this material a
3.2. Theoretical good precursor for the production of the activated carbon, or a valid
solid fuel even if the high sulfur content limits its application in
Beside the experimental activities, a theoretical model for the pre- countries with stringent environmental laws.
diction of the expected surface area and burn-off values in steam acti- Experimental trials consisted of a four-hour-long steam gasification
vation of waste tires was developed based on the random pore model of tires, followed by an activation step. More in detail, six tests were
(RPM) applicable to porous systems in the presence of product layer carried out by changing the hold times in activation step from 60 to
diffusion resistance (Bhatia and Perlmutter, 1980). The basic assump- 360 min, with a step of 60 min, and keeping constant the other process
tions for the RPM model are as follows: The reaction proceeds iso- parameters to values listed in Table 3. In particular, the mass flow rate
thermally; diffusion in pores and external mass transfer are not rate- of waste tires in steam gasification was set to 3 kg·h−1 and resulted in
limiting compared to a surface reaction and solid-state diffusion in 1.13 kg·h−1 (4.52 kg in four hours) of char for the next step of activa-
product layers. In particular, the char activation process with steam was tion. Moreover, one test of gasification without subsequent activation
modeled in term of carbon conversion (Xc) with respect to the time. Xc was carried out, to establish the starting value of the surface area for the
is the weight loss of solid during the activation stage. Assuming that the char. The main process parameters are summarized in Table 3, together
partial pressure of the reaction species is constant, the overall reactivity with results.
for the gas-solid phase reaction on the internal surface of a particle can A temperature in gasification step of 850 °C was taken into account
be written as: from a previous study on the influence of temperatures in steam gasi-
fication of waste tires (Portofino et al., 2013) where the temperature of
dXc E
= k I exp ⎛− ⎞ S (Xc )
dt ⎝ RT ⎠ (3) Table 2
I Ultimate and proximate analysis of scrap tires.
where Xc is the solid conversion fraction, t is the time elapsed, k is the
surface reaction rate constant, E is the activation energy and S(Xc) is the Ultimate analysisa[%w/w]
specific surface area, function of the solid conversion fraction. It de-
pends of the active surface area of the activated carbon. N C H S Ob
0.3 84.1 7.3 2.3 0.8
Under isothermal conditions, it is possible to express the activation
rate by the Arrhenius equation as reported below: Proximate analysis [%w/w]
Volatilea Fixed carbona Asha Moisture
d (1 − Xc ) E
= k I Xcm = A·exp ⎛− ⎞ (Xcm )
61.3 33.5 5.2 0.8
dt ⎝ RT ⎠ (4) Higher heating value [MJ·kg−1]
38.6
where A is the frequency factor for the activation process and m is the
order reaction. Resolution of the differential Eq. (4) gives the following a
Measured on dry based.
b
dependence of Xc from time: Calculated by difference.

105
A. Molino et al. Resources, Conservation & Recycling 129 (2018) 102–111

Table 3
Process parameters for the experimental tests.

Steam process parameters T [°C] 850 850 850 850 850 850 850
Gasification gasifying agent water steam water steam water steam water steam water steam water steam water steam
SFR 1 1 1 1 1 1 1
MRT [min] 6 6 6 6 6 6 6
time lenght [h] 4 4 4 4 4 4 4
Input G st[kg·h−1] 3 3 3 3 3 3 3
G N2 [kg·h−1] 1.05 1.05 1.05 1.05 1.05 1.05 1.05
G H2O [kg·h−1] 3 3 3 3 3 3 3
Output G char [kg·h-1] 1.13 1.13 1.13 1.13 1.13 1.13 1.13
G sg[kg·h−1] 3.86 3.86 3.86 3.86 3.86 3.86 3.86
G H2O [kg·h-1] 2.06 2.06 2.06 2.06 2.06 2.06 2.06
syngas composition H2 38.9 38.9 38.9 38.9 38.9 38.9 38.9
(%v/v) CO 10.0 10.0 10.0 10.0 10.0 10.0 10.0
CO2 6.4 6.4 6.4 6.4 6.4 6.4 6.4
CH4 17.9 17.9 17.9 17.9 17.9 17.9 17.9
C2H4 1.4 1.4 1.4 1.4 1.4 1.4 1.4
C2H6 0.1 0.1 0.1 0.1 0.1 0.1 0.1
N2 25.3 25.3 25.3 25.3 25.3 25.3 25.3
Steam Activation process parameters T [°C] 920 920 920 920 920 920 920
activation agent water steam water steam water steam water steam water steam water steam water steam
SCR 2 2 2 2 2 2 2
Solid residence time 0 60 120 180 240 300 360
[min]
Input M char [kg] 0 4.52 4.52 4.52 4.52 4.52 4.52
M N2 [kg] 0 1.16 2.32 3.48 4.64 5.80 6.96
M H2O [kg] 0 9.04 18.08 27.12 36.16 45.20 54.24
Output M ac[kg] 0 4.06 2.35 0.98 0.48 0.36 0.29
M sg [kg] 0 5.18 6.34 7.50 8.66 9.82 10.98
M H2O [kg] 0 5 16 27 36 45 54
output parameters burn off (%) 0 10.2 48.1 78.4 89.3 92 93.5
BET [m2·g−1] 37.0 ± 0.2 154.4 ± 1.0 553.1 ± 3.4 775.8 ± 4.7 693.5 ± 4.3 557.5 ± 3.41 487.5 ± 2.8
syngas composition H2 – 35.15 36.88 35.89 37.31 34.75 34.41
(%v/v) CO – 18.21 20.35 14.69 17.53 15.07 14.27
CO2 – 9.46 10.28 11.99 10.31 9.26 11.45
CH4 – 1.84 1.32 1.29 1.24 1.35 1.31
C2H4 – 0.00 0.00 0.00 0.00 0.00 0.00
C2H6 – 0.00 0.00 0.00 0.00 0.00 0.00
N2 – 35.34 31.17 36.14 33.61 39.57 38.55

steam gasification was changed from 850 °C to 1000 °C and tests per-
formed with the lowest temperature showed highest char yield and
highest fixed carbon in char: these two findings are important if the aim
of the process is the upgrading of the material in order to produce ac-
tivated carbons. A residence time of 6 min in gasification step was
chosen based on previous experimental tests on the same pilot plant:
this residence time assures steady state condition in the gas composi-
tion, and this is a sign of the complete release of the volatiles that are
present in solid phase, a fundamental condition for the next activation
step. Solid and syngas yields, together with the volumetric gas

Fig. 2. Volumetric syngas composition vs activation hold

times.

106
A. Molino et al.
composition of the syngas produced in gasification are shown in
Table 3.
A temperature of 920 °C was chosen in activation steps, based on the
consideration that the physical activation is highly endothermic,
therefore high temperatures and greater than those employed in gasi-
fication steps are required. In particular, we chose an intermediate
value in the range adopted from several previous works (Mui et al.,
2004). In addition, the values of hold times in activation steps were
chosen taken into account the average hold times adopted by other
researchers, many of them are summarized in the overview of Mui et al.
(Mui et al., 2004).
The burn-off of solid was calculated in accordance with the Eq. (2)
and the surface area of the activated solid residue was determined by
the BET method; the corresponding values are summarized in Table 3.
It is worth to notice that the volumetric gas composition obtained
from the activation is almost constant by varying the hold time from 1
to 6 h (Fig. 2).
It is important to state that both the gasification and the activation
steps had a duration of 3 days: first day for the preparation and setting
of the pilot plant, second one for the performing of the test and the last
one for discharging and collect the solid residue. Together with the
limited amount of available waste tires, this is why no repetition was
made for each test, therefore experimental data are provided without
deviations.
Moreover, the BET analysis were carried out on one sample for each
test: the values reported in Table 3 are obtained from analysis per-
formed by Micromeritics ASAP 2020 instrument.
As expected, from Table 3 you can notice that the greater the hold
time, the greater the burn-off: this is due to action of steam, which
literally erodes the char. For the longest hold time (6 h) the burn-off
increased up to 93.5%, and 4 h are enough to achieve about 90%. It is
clear that, in terms of yield in industrial production of activated carbon,
such high values of hold times have to be avoided in order to ensure
acceptable throughput. Anyway, considerations based exclusively on
reasons related to the amount and not the quality of the activated
carbon are considered misleading. Particular attention must be ad-
dressed to the surface area too.
With regard to this aspect, it is well known that in a first stage steam
activation contributes to increase the surface area up to a maximum
value, through the formation of cavities and micropores into the bulk
material, but in a second stage the pores increase their volume and
collapse, forming first mesopores and then macropores and decreasing
the overall surface area. Table 3 points out this behavior also for the
performed experimental tests: surface area increased in the first three
hours, up to maximum value (775.8 m2·g−1), but for higher hold times
it decreased due to collapse of micropores and mesopores. Results ob-
tained suggest working with hold times not higher than 3 h, in corre-
spondence of which a good compromise of high surface area and burn-
107
Resources, Conservation & Recycling 129 (2018) 102–111
off (78.4%) is obtained.
With regard to the theoretical approach, the kinetic parameters m
and kI at a temperature of 920 °C were obtained by a linear regression of
the residual weight fraction versus time, and they were 0.61 and
5.31 × 10−3 min respectively. Trend of the solid conversion fraction
obtained with the predicted random pore model expressed by Eq. (3)
was compared to experimental results (Fig. 3): it could be seen that the
results predicted from the random pore model agreed with the experi-
mental data, especially for hold time greater than 2 h.
Based on the experimental data of the BET surfaces versus burn-off,
parameters a, b and X0 were obtained from the sigmoidal regression of
the experimental data. The values obtained were: a = 772, b = 0.111
and X0 = 0.323 obtained with r2 equal to 0.985, which indicates the
variance about the mean. Fig. 3 shows a comparison between the ex-
perimental and the theoretical surface areas. In particular, the Fig. 4
shows the BET values calculated in accordance with the Eq. (6) (con-
tinuous line) and the experimental values (crosses) obtained from the
analysis performed by Micromeritics ASAP 2020 instrument, together
with the estimated errors given by the instrument (dashed lines). From
the Figure it can be noticed that the theoretical trend overestimates the
surface area of the activated carbon during the growing trend (of about
100–150 m2·g−1), up to the maximum values: this can be explained
with the consideration that the theoretical model considers the perfect
intimate interaction between the char and the gasifying agent, therefore
maximizing the activation results. On the other side, in the experi-
mental it can be inferred that a fraction of the char does not get in
contact for enough time with the water steam, resulting in a lower
surface area of the carbon. Moreover, the model does not take in ac-
count the collapse of the carbon’s pores, which are responsible for the
diminishing of the surface area for activation times higher than that in
which the maximum is obtained. This explains the constant value of
surface area after the maximum, in contrast with the reduction re-
sulting from experimental tests.
Anyway, the theoretical model can represent a useful tool to make
decision choices, because it gives us information about the maximum
surface area we can obtained from the activation process of a carbon
material and can represent a tool for measure the efficacy of activation
process (the lower the difference between the experimental and theo-
retical data, the higher the efficacy of activation process): of course,
some constants employed in the model must be adjusted if process
parameters different from hold time are modified (steam on tires ratio,
temperatures, residence time in gasification). Future works are foreseen
on this scope.
It is worth mentioning that the proper choices in process parameters
must be made based on economic considerations (sale price of activated
carbon, energy consumption). This work is not intended to make purely
economic considerations, but hereinafter an assessment of the energy
consumption required to support the process is made.
Fig. 3. Comparison of carbon conversion versus time between the ex-
perimental data (squares) and the prediction by random pore model
(continuous line) for waste tires at temperature of 920 °C.
A. Molino et al. Resources, Conservation & Recycling 129 (2018) 102–111

Fig. 4. Comparison of BET surface versus activation time between the

experimental data (cross with dashed lines indicating the deviation) and
the prediction by random pore model (continuous line) for waste tires at
temperature of 920 °C.

High hold times impact also on the heat balance of the process; it is demand and the heat losses, respectively:
well known that steam gasification and steam activation processes are
endothermic, resulting in an increment of the power demand when the Q˙ in + Q˙ st + Q˙ H2 O + Q˙ N2 + Q˙ ' N2 + Q˙ ' H2 O + Q˙ char = Q˙ sg + Q˙ ' sg + Q˙ '' H2 O
time length of the process increases. With regard to this aspect, the heat + Q˙ ''' H2 O + Q˙ char + Q˙ ac + Q˙ ashes + Q˙ loss
balance of the whole process, which goes from tires to activated carbon,
was performed for trial carried out with a hold time equal to 3 h, and in (7)
accordance with the scheme of heat flows highlighted in Fig. 5.

Q˙ st = Gst ⎡HHVst + cpst dT⎤ ≅ Gst [HHVst + cpst (Tgin − 20)]

Tgin

∫
The equations considered for the heat balance are the following (a

⎢ ⎥
reference temperature equal to 20 °C has been considered for evalua-
tion of enthalpy values), in which Qin and Qloss represent the energy ⎣ 20 ⎦ (8)

Fig. 5. Mass (above) and heat (below) balance of activated

carbon production process.

108
A. Molino et al. Resources, Conservation & Recycling 129 (2018) 102–111

Q˙ H2 O = G H2 O ⎡ cpH2 Ol dT⎤ ≅ G H2 O cpH2 O (Tgin − 20)

Tgin

∫
determined following the procedure suggested by Eisermann et al.

⎢ ⎥
(Eisermann et al., 1980) and hereafter reported.
⎣ 20 ⎦ (9) Calling:

Q˙ N2 = G N2 ⎡ cpN2 dT⎤ ≅ G N2 cpN2 (Tgin − 20)

Tgin

∫ i primary volatile matter the amount in excess of 10% on the dry, ash
⎢
⎣ 20 ⎥
⎦ (10)
free basis of solid,
ii secondary volatile matter the total volatile matter content if the
amount is less than or equal to 10%. Or 10% if the total volatile
Q˙ ' N2 = G' N2 ⎡ cpN2 dT⎤ ≅ G N2 cpN2 (Tgin − 20)
Tgin

∫ matter content is greater than 10%

⎢
⎣ 20 ⎥
⎦ (11) iii cF, cV1 and cV2 the specific heats of fixed carbon, primary and sec-
ondary volatile matter, respectively (kJ·kg−1·K−1),

Q˙ ' H2 O = G' H2 O ⎡ cpH2 Ol dT⎤ ≅ G H2 O cpH2 O (Tgin − 20)

Tgin

∫
⎢
⎣ 20 ⎥
⎦ (12)
it will be:
cF = −0.218 + 3.807 × 10−3 × T − 1.758 × 10−6 × T 2 (20)

Q˙ sg = Gsg ⎡HHVsg + cpsg dT⎤ ≅ Gsg [HHVsg + cpsg (Tgout − 20)]

Tgout

∫ c v1 = 0.728 + 3.391 × 10−3 × T

⎢ ⎥
(21)
⎣ 20 ⎦
(13) c v2 = 2.273 + 2.554 × 10−3 ×T (22)

in which T is given in K.
Q˙ sg′ = G' sg ⎡HHV ' sg + ∫ cpsg dT⎤ ≅ Gsg [HHVsg + cpsg (Taout − 20)]
Taout
The specific heats of scrap tyres (cpst), char (cpchar) and activated
⎣ 20 ⎦ carbon (cpac) will be the weighted average of these contributions, on the
(14) basis of the fixed carbon and volatile fractions resulting from the im-
mediate analysis:

Q'' H2 O = G'' H2 O ⎡ cpH2 Og dT⎤

Tgout

∫ ∫
⋅ 100
cpH2 Ol dT + LHV + cp = F × cF + V1 × CV 1 + V2 × CV 2 (23)
⎢
⎣ 20 100 ⎥
⎦ (15)
Moreover, for the ash it will be:
cash = 0.594 + 5.86 × 10−4 × T
Q''' H2 O = G''' H2 O ⎡ ∫ ∫ cpH2 Og dT⎤
⋅ 100 Taout
(24)
cpH2 Ol dT + LHV +
⎣ 20 100 ⎦ (16) in which cash and T are given in kJ·kg−1·K−1 and K, respectively.
The HHV for solids (char and activated carbon) were determined by

Q˙ char = Gchar ⎡HHVchar + cpchar dT⎤ ≅ Gchar [HHVchar

Tgout

∫
Dulong’s equation:
⎢
⎣ ⎥
⎦ O
HHV = 32.79 × C + 150.4 × ⎛H − ⎞ + 9.26 × S + 4.97 × O
20

+ cpchar (Tgout − 20)] (17) ⎝ 8⎠

+ 2.42 × N (20)

Q˙ ac = Gac ⎡HHVac + ∫ cpac dT⎤ ≅ Gac [HHVac + cpac (Taout − 20)]

Taout
in which HHV is given in MJ·kg and C, H, O, S and N are the mass
−1

⎣ 20 ⎦ fraction of solids obtained from the ultimate analysis. Table 4 sum-

(18) marizes the so determined specific heat values and the HHVs employed
in the calculations.

∫
Taout
Q˙ ashes = Gashes cpashes dT ≅ Gashes cpashes (Taout − 20)
20
in which Tgin and Tgout are the temperatures before and after the ga-
sification step, respectively (20 °C and 850 °C) and Taout is the tem-
perature after the activation step (920 °C).
In Eqs. from (8) to (9), the specific heat is estimated as the average
between the starting and final temperatures.
The specific heat of the syngases produced from the gasification and
activation processes were determined from Mcbride B. J. et al. (Mcbride
et al., 1993) and their HHV were determined from volumetric compo-
sition and as weighted average of HHV of compounds. The specific
heats for solids (scrap tires, char, activated carbon and ashes) were
Table 5 summarizes the mass and the heat balances for gasification
and activation processes when a mass flow rate of tires set to 1 kg·h−1 is
(19)
considered.
From data in Table 5 and from Eq. (7), it follows that the energy
demand for the overall process after deducting the energy losses
(Q˙ in − Q˙ loss ) is about 12.6 MJ·h−1 (equivalent to about 3.5 kW). Un-
likely it was not possible to estimate Qin and Qloss separately, but the
difference represents the net energy, useful to ensure that the process
takes place completely.
It is worth to note that the mass balance closes to the 100%. This
happens because all the mass flow rates are measured with accuracy,
with the exception of the water contained in the synthesis gas and
which is condensed and removed before the combustion. This fraction

Table 4
Specific heat and higher heating values of streams.

material cp [kJ·kg−1·K−1] HHV [MJ·kg−1] Note

scrap tyres 2.10 38.58 cp at 20 °C

H2Ol 4.19 – cp at 20 °C
H2Og 2.21 – average cp between 100 °C and 850 °C
H2Og 2.23 – average cp between 100 °C and 920 °C
char 1.32 28.42 average cp between 20 °C and 850 °C
activated carbon 1.31 11.86 average cp between 20 °C and 920 °C
ash 0.26 – average cp between 20 °C and 920 °C
syngas from steam gasification 2.52 19.04 average cp between 20 °C and 850 °C
syngas from steam activation 1.74 7.59 average cp between 20 °C and 920 °C

109
A. Molino et al. Resources, Conservation & Recycling 129 (2018) 102–111

Table 5
Mass and heat balance for the overall process (gasification and 3 h activation).

Mass flow-rate [kg·h−1] Thermal power [kJ·h−1]

Steam Gasification Input Gst 1 measured Q̇st 38.58

G H2 O 1 measured Q̇ H2 O –
G N2 1.05 measured Q̇ N2 –
tot 3.05 tot 38.58
⋅
Output G'' H2 O 0.69 calculated 2940.95
Q'' H2 O
Gchar 0.38 measured Q̇char 422.67
Gsg 1.99 measured Q̇sg 4184.34
tot 3.05 tot 7547.97

Steam Activation Input Gchar 0.38 measured Q̇char 422.67

G' N2 1.16 measured Q̇' N2 –
G' H2 O 0.75 measured Q̇' H2 O –
tot 2.29 tot 422.67
Output Gac 0.03 measured Q̇ac 35.65
⋅
G''' H2 O 0.71 calculated 3160.19
Q''' H2 O
⋅
G' sg 1.50 measured 2358.88
Q' sg
Gashes 0.05 measured Q̇ashes 12.15
tot 2.29 tot 5566.87
Overall Output–Input ∆G 0 ∆Q˙ = Q˙ in − Q˙ loss 12653.59

is calculated as the remaining fraction that allows us to close the mass
balance to 100%. Table 5 expressly shows if the mass flow rates are
measured or calculated. Worldwide, there are few works targeted on
the energy balance of activated carbon production, none of which refers
to waste tires and a technology based on rotary kiln reactors. As ex-
ample, Sharifan in his thesis (Sharifan, 2014) has performed an energy
balance of production of activated carbon from 200 g of hazelnut shells,
performed in an electrical furnace box (discontinuous process) at
T = 862 °C and for 3 h. The energy consumption was estimated about
6 MJ, equivalent to 30 MJ·h−1 for the processing of 1 kg·h−1 of ha-
zelnut shells. This value is about three times the one resulting from our
experimental tests.
The energy demand takes in account the power lost in different
ways (conduction, convection, radiation). It is important to underline
some aspects related to the energy balance: first, the experimental tests
were carried out on a pilot plant designed and realized for scientific
research purposes, therefore it is not optimized in terms of reduction of
heat losses. Of course, great attention on this aspect should be reserved
on industrial plants. Moreover, if the main product of the process is the
activated carbon, the syngas produced could be used to provide energy
for the gasification and activation steps, contributing to reduce the
demand of input energy. We want to point out that a partial precaution
in this direction has also been adopted in experimental tests: the oxi-
dizer burned the syngas before release in environment, in order to
produce the water steam employed in the process. Of course, other and
different solutions could and must be adopted in an industrial plant.
5. Conclusions
Gasification of waste tires for the production of activated carbon is
well known and widespread worldwide. Amount and quality of final
product depend from a large number of parameters: temperatures, type
of activation agent, activation agent on waste tires ratio, hold time of
the process. In particular, it is well known that the surface area of ac-
tivated carbons increases with the increasing of hold times, at least until
it reaches a maximum value after which the collapse of the structure
occurs. On the other hand, the burn-off values increase when the hold
times increase, therefore an optimum hold time value, in terms of high
surface area and low burn-off, has to be found for the process.
The present work aimed to investigate how hold times in the water
steam activation of waste tires could influence the surface area of
activated carbon, as well as the burn-off value, when other process
parameters were kept constant (mainly temperatures and steam on
waste tires ratio). On this basis, first several experimental tests on a
pilot plant based on a rotary kiln reactor were performed and then a
theoretical model for the prediction of the burn-off and the surface area
for water steam gasification and activation of tires was developed.
As expected, the greater the hold time, the greater the burn-off
value, while the surface area showed a maximum value equal to
786 m2·g−1 for a hold time set to 3 h. At the same hold time, the burn-
off reached 78.4%. Investigations suggested us that, when other para-
meters were set as reported in the text, a hold time not higher than 3 h
are recommended. Indeed, for hold times greater than 3 h both the
amount of activated carbon and the surface area decrease because of
pores enlargement which converts micropores first in mesopores and
then in macropores, causing a collapse of the activated carbon struc-
ture.
Moreover, the overall process that goes from waste tires to activated
carbon is endothermic, therefore the greater is the time for performing
the process, and the greater is the energy required. In particular an
energy demand of 3.5 kW is requested for the overall process, if
1 kg·h−1 of waste tires are gasified and activated for three hours. Since
the main product of the process is the activated carbon, it is worth to
note that the syngas could be used to provide energy for the gasification
and activation steps, contributing to reduce the demand of input en-
ergy. This has been partially done in the experimental tests: the heat
derived from the combustion of the syngas has been usefully employed
in a heat exchanger for the production of the steam sent to the rotary
kiln reactor.
Finally, a theoretical model based on the Arrhenius equation and a
sigmoidal regression of the experimental data for the estimation of both
the surface area and the burn off of the carbon, was proposed. It showed
results in good agreement with the experimental data, especially for
hold time greater than 2 h. It can represent a useful tool to make de-
cision choices: of course, some constants employed in the model must
be adjusted if process parameters different from hold time are modified
(steam on tires ratio, temperatures, and hold time in gasification).
References
ETRMA –European Tyre Rubber Manufactures Association, 2011. http://www.etrma.org/
uploads/Modules/Documentsmanager/brochure-elt-2011-final.pdf (Accessed
January, 2012).

110
A. Molino et al.
ETRMA –European Tyre Rubber Manufactures Association, 2014. European tyre and
rubber industry –Statistics. http://www.etrma.org/statistics-2 (Accessed February,
2014).
Antoniou, N., Stravropoulos, G., Zabaniotou, A., 2014. Activation of end of life tyre
pyrolytic char for enhancing viability of pyrolysis- critical review, analysis and re-
commendations for a hybrid dual system. Renew. Sustain. Energy 39, 1053–1073.
http://dx.doi.org/10.1016/j.rser.2014.07.143.
Choi, G.G., Jung, S.H., Oh, S.J., Kim, J.S., 2014. Total utilization of waste tire rubber
through pyrolysis to obtain oils and CO2 activation of pyrolysis char. Fuel Process.
Technol. 213, 57–64. http://dx.doi.org/10.1016/j.fuproc.2014.02.007.
Karatas, H., Olgun, H., Akgun, F., 2012. Experimental results of gasification of waste tire
with air & CO2, air & steam and steam in a bubbling fluidized bed gasifier. Fuel
Process. Technol. 102, 166–174. http://dx.doi.org/10.1016/j.fuproc.2012.04.013.
Donatelli, A., Iovane, P., Molino, A., 2010. High energy syngas production by waste tyres
steam gasification in a rotary kiln pilot plant. Experimental and numerical in-
vestigations. Fuel 89, 2721–2728. http://dx.doi.org/10.1016/j.fuel.2010.03.040.
Kandasamy, J., Gokalp, I., 2015. Pyrolysis, combustion, and steam gasification of various
types of scrap tires for energy recovery. Energy Fuels 29, 346–354. http://dx.doi.org/
10.1021/ef502283s.
Portofino, S., Casu, S., Iovane, P., Russo, A., Martino, M., Donatelli, A., Galvagno, S.,
2011. Optimizing H2 production from waste tires via combined steam gasification
and catalytic reforming. Energy Fuels 25, 2232–2241. http://dx.doi.org/10.1021/
ef200072c@proofing.
Portofino, S., Donatelli, A., Iovane, P., Innella, C., Civita, R., Martino, M., Matera, D.A.,
Russo, A., Cornacchia, G., Galvagno, S., 2013. Steam gasification of waste tyre: in-
fluence of process temperature on yield and product composition. Waste Manage. 33,
672–678. http://dx.doi.org/10.1016/j.wasman.2012.05.041.
Sainz-Diaz, C.I., Griffiths, A.J., 2000. Activated carbon from solid wastes using a pilot-
scale batch flaming pyrolyser. Fuel 79, 1863–1871. http://dx.doi.org/10.1016/
S0016-2361(00)00052-1.
Shah, J., Jan, M.R., Mabood, F., Shahid, M., 2006. Conversion of waste tyres into carbon
black and their utilization as adsorbent. J. Chin. Chem. Soc. 53, 1085–1089. http://
dx.doi.org/10.1002/jccs.200600144.
Martınez, J.D., Puy, N., Murillo, R., Garcıa, T., Navarro, M.V., Mastral, A.M., 2013. Waste
tyre pyrolysis –A review. Renew. Sustain. Energy Rev. 23, 179–213. http://dx.doi.
org/10.1016/j.rser.2013.02.038.
Ariyadejwanich, P., Tanthapanichakoon, W., Nakagawa, K., Mukai, S.R., Tamon, H.,
2003. Preparation and characterization of mesoporous activated carbon form waste
tires. Carbon 41, 157–164. http://dx.doi.org/10.1016/S0008-6223(02)00267-1.
Mui, E.L.K., Ko, D.C.K., McKay, G., 2004. Production of active carbons from waste tyres–a
review. Carbon 42, 2789–2805. http://dx.doi.org/10.1016/j.carbon.2004.06.023.
Saleh, T.A., Danmaliki, G.I., 2016. Adsorptive desulfurization of dibenzothiophene from
111
Resources, Conservation & Recycling 129 (2018) 102–111
fuels by rubber tyres-derived carbons: kinetics and isotherms evaluation. Process Saf.
Environ. Prot. 102, 9–19. http://dx.doi.org/10.1016/j.psep.2016.02.005.
Gonzalez, J.F., Encinar, J.M., Gonzalez-Garcia, C.M., Sabio, E., Ramiro, A., Canito, J.L.,
Ganan, J., 2006. Preparation of activated carbons from used tyres by gasification with
steam and carbon dioxide. Appl. Surf. Sci. 252, 5999–6004. http://dx.doi.org/10.
1016/j.apsusc.2005.11.029.
Hadi, P., Yeung, K.Y., Guo, J., Wang, H., McKay, G., 2016. Sustainable development of
tyre char-based activated carbons with different textural properties for value-added
applications. J. Environ. Manage. 170, 1–7. http://dx.doi.org/10.1016/j.jenvman.
2016.01.005.
Guo, J., Lua, A.C., 2001. Experimental and kinetic studies on pore development during
CO2 activation of oil-palm-shell char. J. Porous Mater. 8, 149–157. http://dx.doi.org/
10.1023/A:1009603110828.
Molino, A., Nanna, F., Villone, A., 2014. Characterization of biomasses in the southern
Italy regions for their use in thermal processes. Appl. Energ. 131, 180–188. http://dx.
doi.org/10.1016/j.apenergy.2014.06.013.
Iovane, P., Donatelli, A., Molino, A., 2013. Influence of feeding ratio on steam gasification
of palm shells in a rotary kiln pilot plant. Experimental and numerical investigation.
Biomass Bioenergy 56, 423–431. http://dx.doi.org/10.1016/j.biombioe.2013.05.
025.
Molino, A., Iovane, P., Donatelli, A., Braccio, G., Chianese, S., Musmarra, D., 2013. Steam
gasification of refuse-derived fuel in a rotary kiln pilot plant: experimental tests.
Chem. Eng. Trans. 32, 337–342. http://dx.doi.org/10.3303/CET1332057.
Liu, X.Y., Specht, E., 2006. Mean residence and hold-up of solids in rotary kilns. Chem.
Eng. Sci. 61, 5176–5181. http://dx.doi.org/10.1016/j.ces.2006.03.054.
Donatelli, A., Garzone, P., Iovane, P., 2015. Discharging granular material from a rotary
kiln in a slumping regime: theoretical and experimental studies. Particuology 23,
56–61. http://dx.doi.org/10.1016/j.partic.2015.02.003.
Mcbride, B.J., Gordon, S., Reno, M.A., 1993. Coefficients for Calculating Thermodynamic
and Transport Properties of Individual Species, NASA Report TM-4513. (Accessed
January, 2014, http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19940013151.
pdf).
Eisermann, W., Johnson, P., Conger, W.L., 1980. Estimating thermodynamic properties of
coal, char, tar and ash. Fuel Process. Technol. 3, 39–53. http://dx.doi.org/10.1016/
0378-3820(80)90022-3.
Sharifan, S., 2014. A Comparative Optimisation Study of Activated Carbon Production
from Hazelnut Shells by Thermal and Microwave Heating Methods (Doctoral
Dissertation). (Imperial College London. Retrieved from http://hdl.handle.net/
10044/1/12845).
Bhatia, S.K., Perlmutter, D.D., 1980. A random pore model for fluid-solid reactions .1.
isothermal, kinetic control. AIChE J. 26, 379–386.