- Email: [email protected]
Effect of heating power on the scrap tires pyrolysis derived oil
Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Contents lists available at ScienceDirect
Journal of Analytical and Applied Pyrolysis journal homepage: www.elsevier.com/locate/jaap
Effect of heating power on the scrap tires pyrolysis derived oil Radwan Alkhatib a,b , Khaled Loubar a,∗ , Sary Awad a , Eskandar Mounif b , Mohand Tazerout a a b
GEPEA, UMR 6144, Ecole des Mines de Nantes, La chantrerie, 4 rue Alfred Kastler B.P. 20722, 44307 Nantes Cedex 3, France Higher Institute for Applied Science and Technology, B.P. 31983, Damascus, Syria
a r t i c l e
i n f o
Article history: Received 12 June 2014 Received in revised form 26 October 2015 Accepted 26 October 2015 Available online 31 October 2015 Keywords: Pyrolysis Waste tire Diels–Alders reaction Heating rate
a b s t r a c t Experimental investigation is carried out in order to study the effect of power input on the pyrolysis process of waste tires (WT). Three different power input levels i.e., 750 W, 1500 W and 3000 W have been used to heat the reactor to 500 ◦ C. The part of energy that was effectively consumed for thermal cracking of rubber and in evaporating the volatile products was deduced by subtracting reactor heating energy (blank tests) from total power input. Changes in physico-chemical characteristics of products were noticed while changing power inputs, which can be the consequence of changes in energy distribution between heating and cracking processes. In addition to the liquid product yield, its tar content is considered to be one of the most important characteristics according to the European standard EN590. From one side, the yield of liquid product for both 1500 W and 3000 W is higher than 45% while it is around 40% for 750 W. On the other side, tar content in the liquid product is higher than 35% at 1500 W and 3000 W power inputs while it is around 11% at 750 W. When high power is introduced to the reactor, it promotes the cracking of CC and CH bonds leading to higher liquid yield and H2 formation during the pyrolysis. Furthermore, it promotes dehydrogenation/cyclization/aromatization/reactions leading to tar and coke formation. Diels–Alders reaction mechanism explains as well the tar yield in the liquid product via dehydrogenation/aromatization/cyclization/reactions. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Each year 1.3–1.5 billion tires reach the end of their life cycle all over the world. For instance, the USA production accounts 500 million tires alone and European Union produces around 289 million tires/year [1,2]. Approximately 64% of these tires go to landfill or they are illegally dumped or stockpiled. Waste tires (WT) resist degradation due to the vulcanization process during its production. Thus, in landfills, tires do not degrade easily, but tend to float on the top over time, due to trapped gases, breaking landfill covers. Although initially, WT do not represent immediate danger, its inappropriate disposal or production in large quantities can seriously pollute the environment or cause problems when they are not properly treated. Both, environmental concerns and high oil prices have increased the interest in solid waste treatment in recent years. Specially, rubber is among the problematic waste materials
Abbreviations: WT, waste tires; GCV, gross calorific value; BR, butadiene rubber; TGA, thermo-gravimetric analysis; W, watt; HR, heating rate; SBR, styrene–butadiene rubber; NR, natural rubber. ∗ Corresponding author. Fax: +33 251858299. E-mail address: [email protected] (K. Loubar). http://dx.doi.org/10.1016/j.jaap.2015.10.014 0165-2370/© 2015 Elsevier B.V. All rights reserved.
on one hand and a valuable potential as recycled raw materials on the other hand. Tires consist mainly of natural rubber (NR) 10–30%, styrene–butadiene rubber (SBR) 30–50%, butadiene rubber (BR) up to 30%, carbon black ≈30%, sulphur ≈1%, and small quantity of organic and inorganic additives; depending on the manufacturer [3–4]. This means that tires are composed mainly of long-chain hydrocarbons, and reprocessing tires to utilize these hydrocarbons could introduce an economical way to treat this form of waste. Pyrolysis of WT is currently receiving high interest and seems to be more attractive as a method to treat the huge quantities of disposal WT. Pyrolysis is a favourable method because of the high conversion rate to oil that can be obtained, in addition to a high calorific value gas which may be used to fuel the process and a residual char which may also be useful either as a smokeless fuel, carbon black or activated charcoal [5]. The main advantage of pyrolysis is that it could deal with waste, which is otherwise difficult to be recycled. Pyrolysis is, in general, an endothermic process that implies many reactions such as cracking, dehydrogenation, and cyclisation/aromatization which are affected by temperature. Therefore, temperature has a considerable effect on the products and conver-
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
sion rate and, for this reason, it is the governing factor/parameter. In addition, other parameters such as heating rate (HR), particle size, pyrolysis time and volatile residence time involved in the process also have remarkable effects [3]. For this reason, depending on the final objective of the pyrolysis, it is necessary to find the optimal conditions for the feedstock. Many efforts have been paid over the last few decades to produce oil fuel from WT by pyrolysis. As the quality of produced oil depends on the above mentioned conditions, fuel characterization is a very important step on the waste-to-fuel chain. Tar content represents one of the most important aspects that should be taken in account for the optimization of pyrolysis conditions. Tar is known as bitumen or asphalt which is a sticky, black and highly viscous liquid or semi-solid form of petroleum with boiling points range higher than 360 ◦ C [6]. Its chemical structure shows that it is a complex of heavy polyaromatic hydrocarbons (PAHs), so any reaction of dehydrogenation/aromatization/cyclization will cause tar formation. Bibliographic survey [1–3,7–10] shows that all the conditions of pyrolysis contribute – at different scales – to the promotion of these reactions. However, temperature and HR stay the most influencing parameters. In other words, temperature and HR express the quantities of heat (energy) used during the pyrolysis process. Rodrigues et al. [11] found that 75 wt% of the pyrolytic oils had a boiling point below 370 ◦ C, which is lower than the 95% set by EN590 for diesel oil. Chaala and Roy [12] found that heavy oil fraction with boiling point higher than 350 ◦ C is 56 wt%. This is attributed to a stronger effect of thermal cracking on conversion products. The objective of this work is to study the effect of the power inputs (expressed sometimes indirectly as HR) over different heating programs on the liquid product yield and its tar content. Two series of experiments have been realized for three different values of power inputs. In the first series, pyrolysis takes place without limiting the heating rate. The maximum temperature was set to 500 ◦ C. In this case, the resistance is used at its maximum power. While in the second series, the time of heating between 250 ◦ C and 500 ◦ C is limited (250 min). In this case, the heating procedure is fluctuating up and down to control and equilibrate between the time and the temperature till arriving to 500 ◦ C. The temperatures 250 ◦ C and 500 ◦ C have been chosen depending on the thermogravimetric analysis (TGA) results [13]. No inert gas current has been used during the pyrolysis, but the system was only purged for 30 min before the beginning of heating. Thus, volatiles leave reactor under the effects of thermal expansion and evaporation. Three power inputs values were investigated 750, 1500 and 3000 W.
11
100 ml/min nitrogen flow for 30 min before the beginning of each experiment to make sure that the system is filled with inert gas and no oxidation process will take place. This is verified by gas analysis at the end of purging. The produced volatiles are evaporated out of the reactor into a water-cooled (15 ◦ C) condenser, where liquid fraction is condensed in the receiving flask. Samples of noncondensing gases are taken each 10–15 min to be analyzed by a micro GC instrument in order to identify the gas composition. Two thermocouples are attached to the reactor, the first is located in the middle of the reactor (central point), while the second is located on the outlet of reactor before the condenser. The first thermocouple measures the temperature inside the reactor while the second one measures the temperature of produced vapour. The reactor is heated up to 500 ◦ C, then the temperature is stabilized at 500 ◦ C for 30 min (it is chosen depending on the results of TGA, gas analysis and temperature of outlet volatile). The collected liquid in the receiving flask is filtered from carbon soot, solid particles and heavy gum compounds (<1%). Then the liquid is characterized (density, viscosity, flash point, Gross Calorific Values (GCV) and GC–MS). 2.3. Measuring equipments Flash point was measured using NORMALAB NPM 440 instrument; it measures up to 350 ◦ C. Viscosity was measured using AND vibro viscometer SV-10, with measuring range 0.3–10,000 mPa.s. GC–MS analyses were performed using a PerkinElmer gas chromatograph Clarus 680 connected to a mass spectrometer PerkinElmer Clarus 600S operating with electron ionisation at 70 eV. GC separation was conducted with a ALBTM -5 ms Supleco capillary column (30 m × 0.25 mm × 0.25 m film thickness) with the following thermal program: from 60 ◦ C to 180 ◦ C at 7 ◦ C/min, held at 180 (1 min) then ramped to 250 ◦ C at 5 ◦ C/min and to 350 ◦ C at 2 ◦ C/min, held at 350 ◦ C for 5 min. In addition, GCV has been obtained using Parr 6200 calorimeter; upper limit of detection 55 kJ/g. Gaseous products (H2 , CH4 , C2 H2 , C2 H4 , C2 H6 , C3 H6 , C3 H8 , C4 H10 , CO, CO2 , O2 and N2 ) were analyzed using a Micro-GC: Agilent technologies 3000A with a thermal conductivity detector. 2.4. Distillation system Distillation system is composed of a 500 ml plated bottom flask, a plated heater with stirrer, a 60 cm condenser and two thermocouples for measuring the temperatures of both gas and liquid phases.
2. Material and methods 2.1. Raw materials Waste passenger car shredded tires have been brought from a WT Collection Company in France. Shreds dimensions ranged between 1 × 1 × 1 cm3 and 6 × 5 × 1 cm3 . 2.2. Pyrolysis reactor A 19 cm diameter and 24 cm height cylindrical, stainless steel, fixed-bed, external electrical wall heated batch reactor was used. Its total capacity of WT is ≈1500 g. The reactor was purged using a
3. Results and discussion 3.1. First series: different power inputs and different residence times The objective of this series is to investigate the effect of power input on the tar content. Heating up to 500 ◦ C is carried out without limiting in heating rate or residence time. Behaviour of temperature and energy consumption within the pyrolysis process is shown in Figs. 1 and 2.
Table 1 Liquid products’ characteristics with different power, series 1. Experiment No.
Power input (W)
HR (◦ C/min)
Yield (wt%)
Density (kg/l)
Viscosity (mPas @ 40 ◦ C)
GCV (MJ/kg)
Flash point (◦ C)
1 2 3
750 1500 3000
1.81 7.14 16.31
40.51 53.49 50.86
0.85 0.91 0.93
1.17 3.3 2.35
43.43 43.47 43.47
25 nda
12
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Figs. 1 and 2 show that the slopes of both curves, reactor temperature and consumed energy rate, increase with increasing power
hydrogen molecules as shown in the following simple schema (Schema I).
(I) inputs. The consequences of this increment will be clearer in the yield and physico-chemical characterisation of liquid products as shown in Table 1. However, all liquid products are similar in appearance for different experiments i.e.,: dark brown colour, medium viscosity and offensive smell. Fig. 2 and Table 1, as it is predicted, show that higher power inputs provide higher HR. Although the comparable GCVs of three
• As the pyrolysis temperature increases, the PAHs concentration increases as well at the expense of aliphatic compounds concentration [8]. This suggests that the unsaturated aliphatic constituents are transformed to aromatic and polyaromatic structures throughout the temperature increment. The aromatization process is well known and confirmed to be taking place by Diels–Alders reaction (Schema II).
(II) liquid products, significant differences between their other characteristics were noticed. These differences are more obvious between 750 W and 1500 W than between 1500 W and 3000 W. Thus the 750 W experiment was labelled “low power input reaction” (low HR) and the other two experiments were labelled “high power input reactions” (high HR). Liquid product yield trends are coherent with literature survey [1–3], as the rate of degradation increases
• When the introduced energy is high, higher number of both C C and C H bonds will be broken. This will increase the dehydrogenation process, and as a consequence, the cyclisation/aromatization reaction will be increased as well producing higher quantity of undesired heavy molecules like PAHs and tar (Schema III).
(III) with HR increment [8]. In order to investigate the effects of power input on products quality, their characteristics were compared. High power inputs resulted in relatively higher density and viscosity of the liquid products. This could indicate that the presence of higher amounts of heavy compounds. During distillation under atmospheric pressure fractions separation has been realized depending EN590 requirements i.e boiling temperature range between ≈150 ◦ C and ≈360 ◦ C and 95% of diesel like fuel should evaporate under 360 ◦ C [7]. No distillation was detected between 110 ◦ C and 140 ◦ C; the results are given in Table 2. It is clear from Table 2 that the tar content in the liquid product is increased considerably when the power input is increased. While it approaches to the EN590 value (tar = 5% max) with low power input. An explanation to the tar formation could be obtained via the following facts: a) Rubber structure: NR, SBR and BR contain mainly C C, C C and C H bonds either in the aliphatic or aromatic forms. C C dissociation bond energy is higher than the other two bonds C C and C H. Therefore, thermal cracking (pyrolysis) will take place in priority and mainly on CC and C H bonds producing free radicals of hydrogen (H• ) and carbon (R(Ar)-C• ). b) C C cracking is coincided with C H breaking (dehydrogenation) in order to respect the stabilisation rules of carbon atom. This leads to form unsaturated alkenes (or alkynes compounds depending on the dehydrogenation severity) in addition to
Tar formation could be inferred during the pyrolysis process by some significant methods: i.) Volatiles temperature behaviour: regarding the temperature of volatiles, it can be noticed that temperature was raised to 198 ◦ C for lower power input and up to 260 ◦ C for both higher power inputs as shown in Fig. 3. High volatiles temperature within short time indicates that heavy molecular weight compounds (with high boiling points) have escaped the reactor. Taking into account that these temperatures are less than the actual boiling points of some heavy molecules, but they leave the reactor under the force of convection with lighter volatiles. ii.) Gas product analysis: used micro GC instrument determines the followed gases: H2 , CH4 , C2 H2 , C2 H4 , C2 H6 , C3 H6 , C3 H8 , C4 H10 , CO, CO2 , O2 and N2 . CO and CO2 appear at the beginning when the reactor temperature is between 250 ◦ C and 350 ◦ C, this may be due to degradation of some oxygen containing additives (rubber structure does not contain oxygen). Concentrations of other detected gases showed similar behaviours with respect to time. Higher concentrations for H2 and CH4 were detected all over reaction duration. The similarity between trends means that they are produced at the same time and from the same degradation process. The formation of these light linear saturated/unsaturated hydrocarbons may be attributed to cracking of C C and C H bonds in the rubber structure and reforming in aspect, while H2 formation
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
13
Table 2 Distillation fractions of different liquid products. Experiment no.
Power input (W)
Light naphtha <110 ◦ C (wt%)
Middle naphtha 140 ◦ C-360 ◦ C (wt%)
Tar >360 ◦ C (wt%)
1 2 3
750 1500 3000
16.33 15.84 12.28
71.81 42.97 49.81
11.86 41.19 37.91
Fig. 1. Temperature behaviour inside the reactor, series 1.
Fig. 3. Volatile temperatures at the pyrolysis reactor outlet, series 1.
Fig. 2. Consumed energy during pyrolysis experiments and blank tests, series 1.
Fig. 4. Hydrogen concentration in outlet volatile at the end of pyrolysis experiment, series 1.
should be due to dehydrogenation process of C H bond. These observations go along with the degradation mechanism predicted in Schema I. Hydrogen behaviour could be considered as a representative for all other gases. Again, the dehydrogenation process is always coincided with cyclization/aromatization process via Diels–Alder reaction. Hydrogen concentrations in the gas product during the pyrolysis process are showed in Figs. 4 and 5. iii.) High H2 concentration in the gas product within shorter time explains the severe dehydrogenation reaction that the rubber is exposed to at high power inputs. This indicates that some cyclization/aromatization reactions take place forming more tar in the liquid product. Hydrogen concentration is used here as a directional indicator and not for a crucial decision.
iv.) GC–MS analysis of liquid showed complex mixtures of aliphatic (cyclic and linear) and aromatic compounds. It contains more than 100 compounds as showed in Fig. 6. As the dlimonene (RT = 5.21 min) is very important compound in the liquid product, its content change between the three power inputs indicates that the liquid product composition is very sensible to the pyrolysis conditions. Presence of aromatic compounds (for example RT = 3.83 (C8 H10 ), RT = 3.46 (C8 H10 ), RT = 6.71 (C10 H12 ) and RT = 11.09 (C16 H20 )) and cyclic compounds (RT = 5.52 (C10 H14 ) and RT = 7.69 (C11 H14 )), emphasizes the cyclisation/aromatization reactions. Finally it is useful to investigate the partition of the consumed energy between heating and thermal degradation processes. In
14
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Table 3 Division of energy between heating and pyrolysis. Experiment
Consumed energy (kWh)
Blank energy (Heating) (kWh)
Pyrolysis energy (kWh for 1000 g)
1 2 3
2.76 2.11 1.96
2.60 1.71 1.37
0.19 0.52 0.81
Table 4 Liquid products’ characteristics with different power inputs, series 2. Experiment no.
Power input (W)
HR (◦ C/min)
Yield (wt%)
Density (kg/l)
Viscosity (mPas @ 40 ◦ C)
GCV (MJ/kg)
Flash point (◦ C)
4 5 6
750 1500 3000
1.12 3.72 4.86
40.40 48.02 46.36
0.84 0.88 0.88
1.11 1.47 1.10
44.45 43.50 43.50
25 25 29
Table 5 Distillation fractions with different power inputs, series 2. Experiment no.
Power input (W)
Light naphtha <110 ◦ C (wt%)
Middle naphtha 140–360 ◦ C (wt%)
Tar >360 ◦ C (wt%)
4 5 6
750 1500 3000
9.88 32.6 8.6
79.78 43.6 74.05
10.34 23.78 17.35
Table 6 Middle naphtha fraction properties, series 1&2. Power input (W)
Density (g/ml)
Viscosity (mpas @ 40 ◦ C)
GCV (MJ/kg)
Flash point (◦ C)
750 1500 3000
0.86–0.88 0.86–0.88 0.86–0.88
1.30–1.40 1.40 1.10
44.21 43.92 43.89
≈35 ≈35 ≈35
Table 7 Solid product composition, series 1 & 2. Element
N
C
H
S
O
other metals (Zn, Fe, Si, Cu)
Average (%) SD
0.34 0.09
79.79 1.66
1.25 0.10
2.24 0.13
1.4 0.32
14.97 1.12
reach the requested temperature (500 ◦ C), but on other hand, it consumes a lower amount of energy for degradation as compared to the other experiments due to a smaller rate of degradation process (bond cracking). To make sure that part of power input (introduced energy) affects the cracking reaction in addition to the system heating, another series of experiments have been realized but with fixed heating time in order to reduce the HR by heat controlling. In this series, the same previous power inputs are used, but the difference in HR between the experiments is dramatically reduced because of HR control in order to respect the demanded heating time. This reduction in HR is not reflected at the same proportion of tar content in the liquid product. 3.2. Second series: different power rates and constant heating time
Fig. 5. Hydrogen concentration during the pyrolysis process, series 1.
order to estimate this partition, a blank experiment (empty reactor) was performed for each power input. Blank experiment represents approximately the needed heating energy; it is subtracted from the total consumed energy to give the reaction energy part. This procedure gives an approximate value of the energy consumed for degradation (pyrolysis) process. Results are shown in Table 3. On one hand, low input power (experiment 1) consumes high amount of energy to be heated (blank) as it takes longer time to
The objective of this series is to show that power input has more significant effect on tar content than the effect of both heating time and HR on it. Throughout these experiments, temperature was raised from 250 ◦ C to 500 ◦ C at the different power input (750 W, 1500 W and 3000 W) during 250 min which corresponds to a mean HR of 1 ◦ C/min. In order to reach this objective, a PID controller was used. To keep the mean value of HR as close as possible to 1 ◦ C/min, the controller turns the resistance on and off alternatively during the heating phase. Thus, as it can be noticed in Figs. 7 and 8, at 750 W the experiment did not have significant changes as compared to the corresponding one in the first series because it is lower than the required value.
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Fig. 6. GC–MS analysis of liquid for three power inputs. (a) 750 W, (b) 1500 W, (c) 3000 W.
15
16
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Fig. 7. Temperature behaviour inside the reactor, series 2.
Fig. 9. Volatile temperatures at the pyrolysis reactor outlet, series 2.
concentration in the gas product increases considerably with high power input increment. From one side, GC–MS analysis shows that light naphtha fraction is rich in benzene, toluene and xylene (BTX) compounds which are important in many applications in feedstock industry, solvent, etc. On the other hand, middle naphtha fraction contains complex mixtures of aliphatic/aromatic compounds. Middle naphtha characterization is shown in the Table 6. Its characterization reveals that middle naphtha fraction is close to the diesel and could be used as fuel directly. Finally, the analysis of the solid product (Table 7) shows that it is similar in all experiments; it means that power input affects mainly the composition of both liquid and gas. The pyrolysis has been finished as the remained hydrogen in the solid is less than 3% of the total H in the tire. 4. Conclusion
Fig. 8. Consumed energy during pyrolysis experiments, series 2.
However, at high power rates experiments, heating process is fluctuating. In order to take into account the direct heating process, only the “power on” phases were taken into consideration for calculations of mean effective HR. Liquid product properties are shown in Table 4. It is clear that the evolutions of HR, of the second series with respect to power input, are less important than those noticed in the first series. This trend is more obvious at higher power rates. Liquid products properties show that the composition is substantially different between experiments. Volatiles temperature behaviour, as shown in the Fig. 9, emphasizes that the scrap tire is exposed to different thermal treatments. Distillation results of liquid product are shown in Table 5. For low power input, the product characteristics are independent of heating time. While with high power inputs changes are drastic (experiments 2, 3, 5 and 6). Although the HR for experiments 5 and 6 are close to each other, the tar content is different. This emphasizes that the product characteristics are related to the power input. In addition, as the dehydrogenation reaction occurs simultaneously with the cyclization/aromatization process, hydrogen
Although all pyrolysis conditions are interdependent, and affect the characteristics of pyrolytic products; yet, power input remains the most influential factor. In addition to increasing the temperature of the system (reactor and raw material), the power input effect is reflected in the strength of cracking process of CC and C H bonds proportionally, affecting the aromatization/cyclization/dehydrogenation reactions. This simultaneously changes the tar content in the pyrolytic oil. Low power input reduces the severity of cracking. This improves the quality of pyrolytic oil; however, it decreases the product yield. As the power input affects, inversely, the quality of the liquid product and its yield, in addition to process time, more intensive study and adjustment are required to get the optimal value. References [1] P.T. Williams, Pyrolysis of waste tires: a review, Waste Manag. 33 (2013) 1714–1728. [2] A. Quek, R. Balasubramanian, Liquefaction of waste tires by pyrolysis for oil and chemicals—a review, J. Anal. Appl. Pyrolysis 101 (2013) 1–16. [3] J.D. Martınez, N. Puy, R. Murillo, T. Garcia, M.V. Navarro, A.M. Mastral, Waste tire pyrolysis—a review, Renew. Sustain. Energy Rev. 23 (2013) 179–213. [4] F. Chen, J. Qian, Studies on the thermal degradation of cis-1,4-polyisoprene, Fuel 81 (2002) 2071–2077. [5] P.T. Williams, D.T. Taylor, Aromatization of tire pyrolysis oil to yield polycyclic aromatic hydrocarbons, Fuel 72 (1993) 1469–1474. [6] J.G. Speight, Updated by Staff, Asphalt. Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, 2013, pp. 1–38.
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17 [7] J.D. Martínez, M. Lapuerta, R. García-Contreras, R. Murillo, T. García, Fuel properties of tire pyrolysis liquid and its blends with diesel fuel, Energy Fuels 27 (2013) 3296–3305. [8] P.T. Williams, S. Besler, Pyrolysis—thermogravimetric analysis of tires and tire components, Fuel 74 (1995) 1277–1283. [9] O. Senneca, P. Salatino, R. Chirone, A fast heating-rate thermogravimetric study of the pyrolysis of scrap tires, Fuel 78 (1999) 1575–1581. [10] P.T. Williams, S. Besler, D.T. Taylor, Pyrolysis of scrap automotive tires: the influence of temperature and HR on product composition, Fuel 69 (1990) 1474–1482.
17
[11] I.M. Rodriguez, M.F. Laresgoiti, M.A. Cabrero, A. Torres, M.J. Chomon, B. Caballero, Pyrolysis of scrap tires, Fuel Process. Technol. 72 (2001) 9–22. [12] A. Chaala, C. Roy, Production of coke from scrap tire vacuum pyrolysis oil, Fuel Process. Technol. 46 (1996) 227–239. [13] R. Alkhatib, K. Loubar, S. Awad, E. Mounif, M. Tazerout, Waste tires pyrolysis: thermogravimetric study and conditions optimization in a batch reactor, in: International Conference on Applied Energy, ICAE 2013, July 1–4, 2013, Pretoria, South Africa., 2015.
Contents lists available at ScienceDirect
Journal of Analytical and Applied Pyrolysis journal homepage: www.elsevier.com/locate/jaap
Effect of heating power on the scrap tires pyrolysis derived oil Radwan Alkhatib a,b , Khaled Loubar a,∗ , Sary Awad a , Eskandar Mounif b , Mohand Tazerout a a b
GEPEA, UMR 6144, Ecole des Mines de Nantes, La chantrerie, 4 rue Alfred Kastler B.P. 20722, 44307 Nantes Cedex 3, France Higher Institute for Applied Science and Technology, B.P. 31983, Damascus, Syria
a r t i c l e
i n f o
Article history: Received 12 June 2014 Received in revised form 26 October 2015 Accepted 26 October 2015 Available online 31 October 2015 Keywords: Pyrolysis Waste tire Diels–Alders reaction Heating rate
a b s t r a c t Experimental investigation is carried out in order to study the effect of power input on the pyrolysis process of waste tires (WT). Three different power input levels i.e., 750 W, 1500 W and 3000 W have been used to heat the reactor to 500 ◦ C. The part of energy that was effectively consumed for thermal cracking of rubber and in evaporating the volatile products was deduced by subtracting reactor heating energy (blank tests) from total power input. Changes in physico-chemical characteristics of products were noticed while changing power inputs, which can be the consequence of changes in energy distribution between heating and cracking processes. In addition to the liquid product yield, its tar content is considered to be one of the most important characteristics according to the European standard EN590. From one side, the yield of liquid product for both 1500 W and 3000 W is higher than 45% while it is around 40% for 750 W. On the other side, tar content in the liquid product is higher than 35% at 1500 W and 3000 W power inputs while it is around 11% at 750 W. When high power is introduced to the reactor, it promotes the cracking of CC and CH bonds leading to higher liquid yield and H2 formation during the pyrolysis. Furthermore, it promotes dehydrogenation/cyclization/aromatization/reactions leading to tar and coke formation. Diels–Alders reaction mechanism explains as well the tar yield in the liquid product via dehydrogenation/aromatization/cyclization/reactions. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Each year 1.3–1.5 billion tires reach the end of their life cycle all over the world. For instance, the USA production accounts 500 million tires alone and European Union produces around 289 million tires/year [1,2]. Approximately 64% of these tires go to landfill or they are illegally dumped or stockpiled. Waste tires (WT) resist degradation due to the vulcanization process during its production. Thus, in landfills, tires do not degrade easily, but tend to float on the top over time, due to trapped gases, breaking landfill covers. Although initially, WT do not represent immediate danger, its inappropriate disposal or production in large quantities can seriously pollute the environment or cause problems when they are not properly treated. Both, environmental concerns and high oil prices have increased the interest in solid waste treatment in recent years. Specially, rubber is among the problematic waste materials
Abbreviations: WT, waste tires; GCV, gross calorific value; BR, butadiene rubber; TGA, thermo-gravimetric analysis; W, watt; HR, heating rate; SBR, styrene–butadiene rubber; NR, natural rubber. ∗ Corresponding author. Fax: +33 251858299. E-mail address: [email protected] (K. Loubar). http://dx.doi.org/10.1016/j.jaap.2015.10.014 0165-2370/© 2015 Elsevier B.V. All rights reserved.
on one hand and a valuable potential as recycled raw materials on the other hand. Tires consist mainly of natural rubber (NR) 10–30%, styrene–butadiene rubber (SBR) 30–50%, butadiene rubber (BR) up to 30%, carbon black ≈30%, sulphur ≈1%, and small quantity of organic and inorganic additives; depending on the manufacturer [3–4]. This means that tires are composed mainly of long-chain hydrocarbons, and reprocessing tires to utilize these hydrocarbons could introduce an economical way to treat this form of waste. Pyrolysis of WT is currently receiving high interest and seems to be more attractive as a method to treat the huge quantities of disposal WT. Pyrolysis is a favourable method because of the high conversion rate to oil that can be obtained, in addition to a high calorific value gas which may be used to fuel the process and a residual char which may also be useful either as a smokeless fuel, carbon black or activated charcoal [5]. The main advantage of pyrolysis is that it could deal with waste, which is otherwise difficult to be recycled. Pyrolysis is, in general, an endothermic process that implies many reactions such as cracking, dehydrogenation, and cyclisation/aromatization which are affected by temperature. Therefore, temperature has a considerable effect on the products and conver-
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
sion rate and, for this reason, it is the governing factor/parameter. In addition, other parameters such as heating rate (HR), particle size, pyrolysis time and volatile residence time involved in the process also have remarkable effects [3]. For this reason, depending on the final objective of the pyrolysis, it is necessary to find the optimal conditions for the feedstock. Many efforts have been paid over the last few decades to produce oil fuel from WT by pyrolysis. As the quality of produced oil depends on the above mentioned conditions, fuel characterization is a very important step on the waste-to-fuel chain. Tar content represents one of the most important aspects that should be taken in account for the optimization of pyrolysis conditions. Tar is known as bitumen or asphalt which is a sticky, black and highly viscous liquid or semi-solid form of petroleum with boiling points range higher than 360 ◦ C [6]. Its chemical structure shows that it is a complex of heavy polyaromatic hydrocarbons (PAHs), so any reaction of dehydrogenation/aromatization/cyclization will cause tar formation. Bibliographic survey [1–3,7–10] shows that all the conditions of pyrolysis contribute – at different scales – to the promotion of these reactions. However, temperature and HR stay the most influencing parameters. In other words, temperature and HR express the quantities of heat (energy) used during the pyrolysis process. Rodrigues et al. [11] found that 75 wt% of the pyrolytic oils had a boiling point below 370 ◦ C, which is lower than the 95% set by EN590 for diesel oil. Chaala and Roy [12] found that heavy oil fraction with boiling point higher than 350 ◦ C is 56 wt%. This is attributed to a stronger effect of thermal cracking on conversion products. The objective of this work is to study the effect of the power inputs (expressed sometimes indirectly as HR) over different heating programs on the liquid product yield and its tar content. Two series of experiments have been realized for three different values of power inputs. In the first series, pyrolysis takes place without limiting the heating rate. The maximum temperature was set to 500 ◦ C. In this case, the resistance is used at its maximum power. While in the second series, the time of heating between 250 ◦ C and 500 ◦ C is limited (250 min). In this case, the heating procedure is fluctuating up and down to control and equilibrate between the time and the temperature till arriving to 500 ◦ C. The temperatures 250 ◦ C and 500 ◦ C have been chosen depending on the thermogravimetric analysis (TGA) results [13]. No inert gas current has been used during the pyrolysis, but the system was only purged for 30 min before the beginning of heating. Thus, volatiles leave reactor under the effects of thermal expansion and evaporation. Three power inputs values were investigated 750, 1500 and 3000 W.
11
100 ml/min nitrogen flow for 30 min before the beginning of each experiment to make sure that the system is filled with inert gas and no oxidation process will take place. This is verified by gas analysis at the end of purging. The produced volatiles are evaporated out of the reactor into a water-cooled (15 ◦ C) condenser, where liquid fraction is condensed in the receiving flask. Samples of noncondensing gases are taken each 10–15 min to be analyzed by a micro GC instrument in order to identify the gas composition. Two thermocouples are attached to the reactor, the first is located in the middle of the reactor (central point), while the second is located on the outlet of reactor before the condenser. The first thermocouple measures the temperature inside the reactor while the second one measures the temperature of produced vapour. The reactor is heated up to 500 ◦ C, then the temperature is stabilized at 500 ◦ C for 30 min (it is chosen depending on the results of TGA, gas analysis and temperature of outlet volatile). The collected liquid in the receiving flask is filtered from carbon soot, solid particles and heavy gum compounds (<1%). Then the liquid is characterized (density, viscosity, flash point, Gross Calorific Values (GCV) and GC–MS). 2.3. Measuring equipments Flash point was measured using NORMALAB NPM 440 instrument; it measures up to 350 ◦ C. Viscosity was measured using AND vibro viscometer SV-10, with measuring range 0.3–10,000 mPa.s. GC–MS analyses were performed using a PerkinElmer gas chromatograph Clarus 680 connected to a mass spectrometer PerkinElmer Clarus 600S operating with electron ionisation at 70 eV. GC separation was conducted with a ALBTM -5 ms Supleco capillary column (30 m × 0.25 mm × 0.25 m film thickness) with the following thermal program: from 60 ◦ C to 180 ◦ C at 7 ◦ C/min, held at 180 (1 min) then ramped to 250 ◦ C at 5 ◦ C/min and to 350 ◦ C at 2 ◦ C/min, held at 350 ◦ C for 5 min. In addition, GCV has been obtained using Parr 6200 calorimeter; upper limit of detection 55 kJ/g. Gaseous products (H2 , CH4 , C2 H2 , C2 H4 , C2 H6 , C3 H6 , C3 H8 , C4 H10 , CO, CO2 , O2 and N2 ) were analyzed using a Micro-GC: Agilent technologies 3000A with a thermal conductivity detector. 2.4. Distillation system Distillation system is composed of a 500 ml plated bottom flask, a plated heater with stirrer, a 60 cm condenser and two thermocouples for measuring the temperatures of both gas and liquid phases.
2. Material and methods 2.1. Raw materials Waste passenger car shredded tires have been brought from a WT Collection Company in France. Shreds dimensions ranged between 1 × 1 × 1 cm3 and 6 × 5 × 1 cm3 . 2.2. Pyrolysis reactor A 19 cm diameter and 24 cm height cylindrical, stainless steel, fixed-bed, external electrical wall heated batch reactor was used. Its total capacity of WT is ≈1500 g. The reactor was purged using a
3. Results and discussion 3.1. First series: different power inputs and different residence times The objective of this series is to investigate the effect of power input on the tar content. Heating up to 500 ◦ C is carried out without limiting in heating rate or residence time. Behaviour of temperature and energy consumption within the pyrolysis process is shown in Figs. 1 and 2.
Table 1 Liquid products’ characteristics with different power, series 1. Experiment No.
Power input (W)
HR (◦ C/min)
Yield (wt%)
Density (kg/l)
Viscosity (mPas @ 40 ◦ C)
GCV (MJ/kg)
Flash point (◦ C)
1 2 3
750 1500 3000
1.81 7.14 16.31
40.51 53.49 50.86
0.85 0.91 0.93
1.17 3.3 2.35
43.43 43.47 43.47
25 nda
12
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Figs. 1 and 2 show that the slopes of both curves, reactor temperature and consumed energy rate, increase with increasing power
hydrogen molecules as shown in the following simple schema (Schema I).
(I) inputs. The consequences of this increment will be clearer in the yield and physico-chemical characterisation of liquid products as shown in Table 1. However, all liquid products are similar in appearance for different experiments i.e.,: dark brown colour, medium viscosity and offensive smell. Fig. 2 and Table 1, as it is predicted, show that higher power inputs provide higher HR. Although the comparable GCVs of three
• As the pyrolysis temperature increases, the PAHs concentration increases as well at the expense of aliphatic compounds concentration [8]. This suggests that the unsaturated aliphatic constituents are transformed to aromatic and polyaromatic structures throughout the temperature increment. The aromatization process is well known and confirmed to be taking place by Diels–Alders reaction (Schema II).
(II) liquid products, significant differences between their other characteristics were noticed. These differences are more obvious between 750 W and 1500 W than between 1500 W and 3000 W. Thus the 750 W experiment was labelled “low power input reaction” (low HR) and the other two experiments were labelled “high power input reactions” (high HR). Liquid product yield trends are coherent with literature survey [1–3], as the rate of degradation increases
• When the introduced energy is high, higher number of both C C and C H bonds will be broken. This will increase the dehydrogenation process, and as a consequence, the cyclisation/aromatization reaction will be increased as well producing higher quantity of undesired heavy molecules like PAHs and tar (Schema III).
(III) with HR increment [8]. In order to investigate the effects of power input on products quality, their characteristics were compared. High power inputs resulted in relatively higher density and viscosity of the liquid products. This could indicate that the presence of higher amounts of heavy compounds. During distillation under atmospheric pressure fractions separation has been realized depending EN590 requirements i.e boiling temperature range between ≈150 ◦ C and ≈360 ◦ C and 95% of diesel like fuel should evaporate under 360 ◦ C [7]. No distillation was detected between 110 ◦ C and 140 ◦ C; the results are given in Table 2. It is clear from Table 2 that the tar content in the liquid product is increased considerably when the power input is increased. While it approaches to the EN590 value (tar = 5% max) with low power input. An explanation to the tar formation could be obtained via the following facts: a) Rubber structure: NR, SBR and BR contain mainly C C, C C and C H bonds either in the aliphatic or aromatic forms. C C dissociation bond energy is higher than the other two bonds C C and C H. Therefore, thermal cracking (pyrolysis) will take place in priority and mainly on CC and C H bonds producing free radicals of hydrogen (H• ) and carbon (R(Ar)-C• ). b) C C cracking is coincided with C H breaking (dehydrogenation) in order to respect the stabilisation rules of carbon atom. This leads to form unsaturated alkenes (or alkynes compounds depending on the dehydrogenation severity) in addition to
Tar formation could be inferred during the pyrolysis process by some significant methods: i.) Volatiles temperature behaviour: regarding the temperature of volatiles, it can be noticed that temperature was raised to 198 ◦ C for lower power input and up to 260 ◦ C for both higher power inputs as shown in Fig. 3. High volatiles temperature within short time indicates that heavy molecular weight compounds (with high boiling points) have escaped the reactor. Taking into account that these temperatures are less than the actual boiling points of some heavy molecules, but they leave the reactor under the force of convection with lighter volatiles. ii.) Gas product analysis: used micro GC instrument determines the followed gases: H2 , CH4 , C2 H2 , C2 H4 , C2 H6 , C3 H6 , C3 H8 , C4 H10 , CO, CO2 , O2 and N2 . CO and CO2 appear at the beginning when the reactor temperature is between 250 ◦ C and 350 ◦ C, this may be due to degradation of some oxygen containing additives (rubber structure does not contain oxygen). Concentrations of other detected gases showed similar behaviours with respect to time. Higher concentrations for H2 and CH4 were detected all over reaction duration. The similarity between trends means that they are produced at the same time and from the same degradation process. The formation of these light linear saturated/unsaturated hydrocarbons may be attributed to cracking of C C and C H bonds in the rubber structure and reforming in aspect, while H2 formation
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
13
Table 2 Distillation fractions of different liquid products. Experiment no.
Power input (W)
Light naphtha <110 ◦ C (wt%)
Middle naphtha 140 ◦ C-360 ◦ C (wt%)
Tar >360 ◦ C (wt%)
1 2 3
750 1500 3000
16.33 15.84 12.28
71.81 42.97 49.81
11.86 41.19 37.91
Fig. 1. Temperature behaviour inside the reactor, series 1.
Fig. 3. Volatile temperatures at the pyrolysis reactor outlet, series 1.
Fig. 2. Consumed energy during pyrolysis experiments and blank tests, series 1.
Fig. 4. Hydrogen concentration in outlet volatile at the end of pyrolysis experiment, series 1.
should be due to dehydrogenation process of C H bond. These observations go along with the degradation mechanism predicted in Schema I. Hydrogen behaviour could be considered as a representative for all other gases. Again, the dehydrogenation process is always coincided with cyclization/aromatization process via Diels–Alder reaction. Hydrogen concentrations in the gas product during the pyrolysis process are showed in Figs. 4 and 5. iii.) High H2 concentration in the gas product within shorter time explains the severe dehydrogenation reaction that the rubber is exposed to at high power inputs. This indicates that some cyclization/aromatization reactions take place forming more tar in the liquid product. Hydrogen concentration is used here as a directional indicator and not for a crucial decision.
iv.) GC–MS analysis of liquid showed complex mixtures of aliphatic (cyclic and linear) and aromatic compounds. It contains more than 100 compounds as showed in Fig. 6. As the dlimonene (RT = 5.21 min) is very important compound in the liquid product, its content change between the three power inputs indicates that the liquid product composition is very sensible to the pyrolysis conditions. Presence of aromatic compounds (for example RT = 3.83 (C8 H10 ), RT = 3.46 (C8 H10 ), RT = 6.71 (C10 H12 ) and RT = 11.09 (C16 H20 )) and cyclic compounds (RT = 5.52 (C10 H14 ) and RT = 7.69 (C11 H14 )), emphasizes the cyclisation/aromatization reactions. Finally it is useful to investigate the partition of the consumed energy between heating and thermal degradation processes. In
14
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Table 3 Division of energy between heating and pyrolysis. Experiment
Consumed energy (kWh)
Blank energy (Heating) (kWh)
Pyrolysis energy (kWh for 1000 g)
1 2 3
2.76 2.11 1.96
2.60 1.71 1.37
0.19 0.52 0.81
Table 4 Liquid products’ characteristics with different power inputs, series 2. Experiment no.
Power input (W)
HR (◦ C/min)
Yield (wt%)
Density (kg/l)
Viscosity (mPas @ 40 ◦ C)
GCV (MJ/kg)
Flash point (◦ C)
4 5 6
750 1500 3000
1.12 3.72 4.86
40.40 48.02 46.36
0.84 0.88 0.88
1.11 1.47 1.10
44.45 43.50 43.50
25 25 29
Table 5 Distillation fractions with different power inputs, series 2. Experiment no.
Power input (W)
Light naphtha <110 ◦ C (wt%)
Middle naphtha 140–360 ◦ C (wt%)
Tar >360 ◦ C (wt%)
4 5 6
750 1500 3000
9.88 32.6 8.6
79.78 43.6 74.05
10.34 23.78 17.35
Table 6 Middle naphtha fraction properties, series 1&2. Power input (W)
Density (g/ml)
Viscosity (mpas @ 40 ◦ C)
GCV (MJ/kg)
Flash point (◦ C)
750 1500 3000
0.86–0.88 0.86–0.88 0.86–0.88
1.30–1.40 1.40 1.10
44.21 43.92 43.89
≈35 ≈35 ≈35
Table 7 Solid product composition, series 1 & 2. Element
N
C
H
S
O
other metals (Zn, Fe, Si, Cu)
Average (%) SD
0.34 0.09
79.79 1.66
1.25 0.10
2.24 0.13
1.4 0.32
14.97 1.12
reach the requested temperature (500 ◦ C), but on other hand, it consumes a lower amount of energy for degradation as compared to the other experiments due to a smaller rate of degradation process (bond cracking). To make sure that part of power input (introduced energy) affects the cracking reaction in addition to the system heating, another series of experiments have been realized but with fixed heating time in order to reduce the HR by heat controlling. In this series, the same previous power inputs are used, but the difference in HR between the experiments is dramatically reduced because of HR control in order to respect the demanded heating time. This reduction in HR is not reflected at the same proportion of tar content in the liquid product. 3.2. Second series: different power rates and constant heating time
Fig. 5. Hydrogen concentration during the pyrolysis process, series 1.
order to estimate this partition, a blank experiment (empty reactor) was performed for each power input. Blank experiment represents approximately the needed heating energy; it is subtracted from the total consumed energy to give the reaction energy part. This procedure gives an approximate value of the energy consumed for degradation (pyrolysis) process. Results are shown in Table 3. On one hand, low input power (experiment 1) consumes high amount of energy to be heated (blank) as it takes longer time to
The objective of this series is to show that power input has more significant effect on tar content than the effect of both heating time and HR on it. Throughout these experiments, temperature was raised from 250 ◦ C to 500 ◦ C at the different power input (750 W, 1500 W and 3000 W) during 250 min which corresponds to a mean HR of 1 ◦ C/min. In order to reach this objective, a PID controller was used. To keep the mean value of HR as close as possible to 1 ◦ C/min, the controller turns the resistance on and off alternatively during the heating phase. Thus, as it can be noticed in Figs. 7 and 8, at 750 W the experiment did not have significant changes as compared to the corresponding one in the first series because it is lower than the required value.
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Fig. 6. GC–MS analysis of liquid for three power inputs. (a) 750 W, (b) 1500 W, (c) 3000 W.
15
16
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17
Fig. 7. Temperature behaviour inside the reactor, series 2.
Fig. 9. Volatile temperatures at the pyrolysis reactor outlet, series 2.
concentration in the gas product increases considerably with high power input increment. From one side, GC–MS analysis shows that light naphtha fraction is rich in benzene, toluene and xylene (BTX) compounds which are important in many applications in feedstock industry, solvent, etc. On the other hand, middle naphtha fraction contains complex mixtures of aliphatic/aromatic compounds. Middle naphtha characterization is shown in the Table 6. Its characterization reveals that middle naphtha fraction is close to the diesel and could be used as fuel directly. Finally, the analysis of the solid product (Table 7) shows that it is similar in all experiments; it means that power input affects mainly the composition of both liquid and gas. The pyrolysis has been finished as the remained hydrogen in the solid is less than 3% of the total H in the tire. 4. Conclusion
Fig. 8. Consumed energy during pyrolysis experiments, series 2.
However, at high power rates experiments, heating process is fluctuating. In order to take into account the direct heating process, only the “power on” phases were taken into consideration for calculations of mean effective HR. Liquid product properties are shown in Table 4. It is clear that the evolutions of HR, of the second series with respect to power input, are less important than those noticed in the first series. This trend is more obvious at higher power rates. Liquid products properties show that the composition is substantially different between experiments. Volatiles temperature behaviour, as shown in the Fig. 9, emphasizes that the scrap tire is exposed to different thermal treatments. Distillation results of liquid product are shown in Table 5. For low power input, the product characteristics are independent of heating time. While with high power inputs changes are drastic (experiments 2, 3, 5 and 6). Although the HR for experiments 5 and 6 are close to each other, the tar content is different. This emphasizes that the product characteristics are related to the power input. In addition, as the dehydrogenation reaction occurs simultaneously with the cyclization/aromatization process, hydrogen
Although all pyrolysis conditions are interdependent, and affect the characteristics of pyrolytic products; yet, power input remains the most influential factor. In addition to increasing the temperature of the system (reactor and raw material), the power input effect is reflected in the strength of cracking process of CC and C H bonds proportionally, affecting the aromatization/cyclization/dehydrogenation reactions. This simultaneously changes the tar content in the pyrolytic oil. Low power input reduces the severity of cracking. This improves the quality of pyrolytic oil; however, it decreases the product yield. As the power input affects, inversely, the quality of the liquid product and its yield, in addition to process time, more intensive study and adjustment are required to get the optimal value. References [1] P.T. Williams, Pyrolysis of waste tires: a review, Waste Manag. 33 (2013) 1714–1728. [2] A. Quek, R. Balasubramanian, Liquefaction of waste tires by pyrolysis for oil and chemicals—a review, J. Anal. Appl. Pyrolysis 101 (2013) 1–16. [3] J.D. Martınez, N. Puy, R. Murillo, T. Garcia, M.V. Navarro, A.M. Mastral, Waste tire pyrolysis—a review, Renew. Sustain. Energy Rev. 23 (2013) 179–213. [4] F. Chen, J. Qian, Studies on the thermal degradation of cis-1,4-polyisoprene, Fuel 81 (2002) 2071–2077. [5] P.T. Williams, D.T. Taylor, Aromatization of tire pyrolysis oil to yield polycyclic aromatic hydrocarbons, Fuel 72 (1993) 1469–1474. [6] J.G. Speight, Updated by Staff, Asphalt. Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, 2013, pp. 1–38.
R. Alkhatib et al. / Journal of Analytical and Applied Pyrolysis 116 (2015) 10–17 [7] J.D. Martínez, M. Lapuerta, R. García-Contreras, R. Murillo, T. García, Fuel properties of tire pyrolysis liquid and its blends with diesel fuel, Energy Fuels 27 (2013) 3296–3305. [8] P.T. Williams, S. Besler, Pyrolysis—thermogravimetric analysis of tires and tire components, Fuel 74 (1995) 1277–1283. [9] O. Senneca, P. Salatino, R. Chirone, A fast heating-rate thermogravimetric study of the pyrolysis of scrap tires, Fuel 78 (1999) 1575–1581. [10] P.T. Williams, S. Besler, D.T. Taylor, Pyrolysis of scrap automotive tires: the influence of temperature and HR on product composition, Fuel 69 (1990) 1474–1482.
17
[11] I.M. Rodriguez, M.F. Laresgoiti, M.A. Cabrero, A. Torres, M.J. Chomon, B. Caballero, Pyrolysis of scrap tires, Fuel Process. Technol. 72 (2001) 9–22. [12] A. Chaala, C. Roy, Production of coke from scrap tire vacuum pyrolysis oil, Fuel Process. Technol. 46 (1996) 227–239. [13] R. Alkhatib, K. Loubar, S. Awad, E. Mounif, M. Tazerout, Waste tires pyrolysis: thermogravimetric study and conditions optimization in a batch reactor, in: International Conference on Applied Energy, ICAE 2013, July 1–4, 2013, Pretoria, South Africa., 2015.