- Email: [email protected]
Simulation of large coal particles pyrolysis by circulating ash heat carrier toward the axial dimension of the moving bed
Fuel Processing Technology 154 (2016) 227–234
Contents lists available at ScienceDirect
Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc
Research article
Simulation of large coal particles pyrolysis by circulating ash heat carrier toward the axial dimension of the moving bed Ya-Qing Zhang, Jia-Long Zhu, Xiao-Hang Wang, Xi-Wang Zhang, Shi-Xue Zhou, Peng Liang ⁎ College of Chemical and Environmental Engineering, Shandong University of Science and Technology, Qingdao 266590, Shandong, China
a r t i c l e
i n f o
Article history: Received 19 January 2016 Received in revised form 24 August 2016 Accepted 25 August 2016 Available online 2 September 2016 Keywords: Coal pyrolysis Solid heat carrier Large coal particle Heat transfer Volatile evolution
a b s t r a c t A heat transfer, reaction, pyrolysis mathematical model for the non-isothermal coal particles by using circulating ash as heat carrier toward the moving bed has been established. Combined with the Thermogravimetry-Mass spectrometry technology and Coats–Redfern integral method, the model has the ability to predict the temperature distribution of pyrolysis gas-coal-ash as well as the evolution characteristics of the main volatile products (such as CH4, CO2, H2, CO, C2H4, C2H6, C6H6, C7H8, C8H10, C10H8). The results show that, the maximum temperature difference between the core and surface of coal (10 mm) has reached 406 K at the bed height of 0.05 m. The layer closer to the coal core has a higher but later peak value of the devolatilization rate. The evolution of the main volatile products is concentrated at the bed height of 0.08–0.24 m. The velocity of the moving bed, blending ratio of ash to coal, coal particle size, preheating temperature of coal and initial temperature of ash have obvious influence on the devolatilization process. Radiation is the most significant factor affecting the devolatilization behavior. The model can be applied to different coal species. This study can provide a theoretic foundation for the amplification design of the moving-bed reactor in the poly-generation system. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The majority of the raw coal in China is used for direct combustion [1] which is inefficient and serious pollution. Fortunately, the circulating fluidized bed (CFB) combustion technology has been developed greatly for its better fuel flexibility, low pollution and load adjustability in recent years [2,3]. Similarly, the moving-bed pyrolyzer is increasingly valued because of its high adjustment capacity, stable operation and uniform heating [4]. So this provides a base for the poly-generation technology of coal pyrolysis and CFB combustion to achieve the grading conversion of coal. High-temperature ash, which comes from the CFB boiler is used as solid heat carrier, mixed with coal is fed into the pyrolyzer simultaneously. Then, the coal is heated to pyrolysis temperature and produce gas and tar. The char produced in the pyrolyzer is returned to CFB boiler to provide electricity and heat [5]. Thus the utilization of coal with high efficiency and low emissions is realized. Furthermore, since the pyrolysis reactor is independent of the CFB boiler, the CFB boiler can be operated separately when the pyrolysis system is out of work. Therefore, the moving-bed pyrolyzer as the critical equipment in the poly-generation technology [6], its operating state is directly relates to the stability and economy of the whole system. The coal pyrolysis heated by solid or gas carriers has a higher efficiency of heat transfer than the traditional heating technologies. And ⁎ Corresponding author. E-mail address: [email protected] (P. Liang).
http://dx.doi.org/10.1016/j.fuproc.2016.08.037 0378-3820/© 2016 Elsevier B.V. All rights reserved.
compared with coal pyrolysis by gas heat carrier, the solid heat carrier not only provide higher heating rate but also avoid the volatiles dilution by the inert gas [7]. So the experimental works on coal pyrolysis by solid heat carrier have been widely valued in recent years [8–11]. However, the simulation studies are relatively few reported. To enable a better understanding on the poly-generation process, it is necessary to establish a mathematical model for the analysis of coal pyrolysis behavior, especially for the prediction and amplification of industrial processes. Some researchers [12–14] carried out the radial stratification of large particles to investigate the temperature-rise period. The studies showed that the coal particle size has significant influence on the pyrolysis behavior. However, the models are inappropriate for the process of coal pyrolysis by solid heat carrier because the heat transfer mechanism outside the coal surface considered only the heat radiation and heat convection. Higuera [15] reported the devolatilization of a moving coal particle by using a competing reaction pyrolysis reaction model. He pointed out that the relative motion of the particles would improve the heat transfer process between the surrounding gas and the particle. Liang et al. [6] reported a steady state, axial-dimensional, numerical model for coal pyrolysis by solid heat carrier toward the moving-bed. However, the coal particle was assumed as an isothermal body in the above two models, the existence of temperature gradient as a result of internal thermal resistance of coal particles should not be ignored. Therefore, a reasonable pyrolysis model of the non-isothermal coal particles by circulating ash carrier toward the axial dimension of the moving bed should be established.
228
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
In this study, the Thermogravimetry-Mass spectrometry (TG-MS) technology and kinetic study were applied to modeling the nonisothermal coal pyrolysis by circulating ash heat carrier. Based on the heat transfer theory, not only the internal pyrolysis behavior of coal particles but also the temperature distribution of pyrolysis gas-coal-ash toward the pyrolyzer can be predicted. Combined with the multiplereaction model, the evolution characteristics of the main volatile products is revealed. This work will provide a better understanding on the poly-generation technology of coal pyrolysis combined with CFB combustion. 2. Experimental As shown in Table 1, Shenmu coal from Shaanxi Province is used in this study. The kinetic parameters for model calculation is obtained in the TG-MS system which is composed of a thermogravimetric analyzer (TG/TGA, Setaram Setsys-Evolution, Caluire, France) and a mass spectrometer (MS, Balzers Omnistar, Brügg, Switzerland). Before the TGMS experiment is carried out, the coal sample was crushed and sieved to ≤ 74 μm, then dried at 105 °C for 4 h. The high-purity nitrogen (99.999 vol.%) was introduced into the TG-MS system with a flow rate of 80 mL/min. In the TG analyzer, the Shenmu coal sample (15 mg) was heated from the room temperature to 700 °C at a heating rate of 50 °C/min. The evolution characteristics of the gaseous products were analyzed in the MS with the ionization voltage of 40 eV and detection charge-mass ratio ranging from 0 to 300 amu. The physical characteristics of CFB ash, coal, and pyrolysis gas used in the experiment are shown in Table 2. The composition analysis of the circulating ash is shown in Table 3. 3. Mathematical model
As Fig. 1 shows that, the coal coupled with high-temperature circulating ash which comes from the mixing section is added into the reactor. As the mixed material moving downwards, the coal is heated by the surrounding ash and pyrolysis gas released from the under layer. Then the pyrolysis reaction occurs with the increase of coal particle temperature. The pyrolysis gas escapes upward, while the ash and char move downwards and return to the CFB boiler. In the moving-bed reactor, the height is axially divided into m layers. At any jth layer in the reactor, the temperature of the pyrolysis gas and ash surrounding the coal particles is respectively Tg,j and Ta,j. In the coal particle, the diameter is radially divided into n layers. From the core to surface, the internal temperature distribution of the coal particle is respectively T0,j, T1,j, … Tn,j. Assumptions of the model simplification are as follows: (1) The particle sizes of coal and ash remain unchanged. (2) Ash is assumed as an isothermal, inert, spherical particle. (3) Spherical coal particle is tightly surrounded by the ash, and the characteristics of a single coal particle can predict the coal cluster in the same bed layer. (4) The diffusion process of pyrolysis products evolved from the internal to external of coal particles is ignored. (5) The moving-bed reactor is operated at a steady state. (6) The calculation time step is small enough to ensure little
Table 1 The proximate and ultimate analyses of Shenmu coal sample (wt%).
M 7.21
A 3.89
V 31.54
Ultimate analysis (daf) FC 57.36
3.2. Mathematical model description of coal pyrolysis by solid heat carrier 3.2.1. Pyrolysis kinetic equation In this model, the multiple independent parallel first-order reactions are assumed to describe the evolution characteristics of the volatile products. Therefore, the kinetic equation of the main volatile products (such as CH4, CO2, H2, CO, C2H4, C2H6, C6H6, C7H8, C8H10, C10H8) and the total volatile (including but not all of the above main volatile products) can be expressed as dw j Ej ¼ k0 j exp − w j −w j dt RT
ð1Þ
where the final volatile product yields wj* are obtained from the test of low temperature distillation of coal by aluminum retort (GB/T4802010), the pre-exponential factor k0j and activation energy Ej are calculated by the TG-MS result. In this study, Coat–Redfern method [16–18] is applied to calculate the k0j and Ej because of its wider applicability and better fitting effect. The equations of Coats-Redfern method can be expressed as. ln 1−w j k0 j R E j ln − − ¼ ln αE j RT T2 " ln
1−n # 1− 1−w j T 2 ð1−nÞ
¼ ln
k0 j R E j − αE j RT
ðn ¼ 1Þ
ð2Þ
ðn≠1Þ
ð3Þ
In Eq. (2) or (3), the left side versus 1/T shows a linear relationship. Therefore, the pre-exponential factor k0j and activation energy Ej can be calculated from the intercept and slope of the straight line.
3.1. Analysis and assumptions
Proximate analysis (ad)
temperature change of coal-ash-gas three phases in any adjacent iterative computation.
C 80.88
H 5.09
N 1.08
S 0.29
Oa 12.66
a By difference; d, dry; daf, dry ash free; M, moisture; A, ash; V, volatile matter; FC, fixed carbon.
3.2.2. Mass conservation equation ρc d dw ρc0 ¼− dt dt
ð4Þ
3.2.3. Heat transfer mechanisms in the moving-bed reactor Five heat transfer mechanisms in the moving-bed reactor are considered as follows: (a) Thermal conduction between the circulating ash and coal particles. In the moving-bed reactor, the coal particles are uniformly distributed in the circulating ash which has a relatively small size. In the moving downwards process of the mixed material, a circular contact surface between the coal and ash particle will be produced due to the mutual extrusion. Then the heat can be passed through the contact surface from the circulating ash to the coal surface. A formula has been proposed by Luikov [19] for calculating the size of the contact surface, dac = 0.2dpa. Therefore, taken the volume thickness of dz. in the pyrolyzer as the calculational unit, the surface area of coal and ash can be respectively writ1−ε ten as sc ¼ ac 1−ε 1þθ Adz and sa ¼ aa 1þθ Aθdz. In the above two equations, θ is the volume ratio of ash to coal in the pyrolyzer, ac and aa are respectively the specific surface area of coal and ash, ac = 6(1 − ε)/dpc, aa = 6(1 − ε)/dpa. The contact area of coal and ash can be given by the equation sac = scRac, where Rac [6] is the ratio of the contact area to coal surface area. Based on the above analysis, the heat flow of thermal conduction between ash and coal particles is expressed as dqcod ¼
2λa λc T a −T c sac λa þ λc dpa
ð5Þ
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
229
Table 2 Parameters of ash, coal, and pyrolysis gas for numerical calculation. ΔH (kJ/kg)
ε
us (m/min)
ρa0 (kg/m3)
ρc0 (kg/m3)
Ta0 (K)
Tc0 (K)
cpa (J/(kg·K))
cpg (J/(kg·K))
λa (W/(m·K))
λg (W/(m·K))
dpa (m)
μg (Pa·s)
β
χ
σ (J/(s·m2·K4))
−300
0.415
0.125
1550
1250
1100
298
840
1214
0.29
0.0742
6.70 × 10−4
3.79 × 10−5
0.8
0.5
5.67 × 10−8
where λc [20] is the thermal conductivity of coal, and it can be given by the equations.
where cpc [23] is the heat capacity of coal, and it can be given by the equations:
λc ¼ 0:23 λc ¼ 0:23 þ 2:24 10−5 ðT−673Þ1:8
for T ≤673 K for T N673K
ð6Þ
(b) Radiation heat transfer between the circulating ash and coal particles dqrad ¼ βσχ ðsc −sac Þ T 4a −T 4c
ð7Þ
(c) Convection heat transfer between the circulating ash and pyrolysis gas dqag ¼ hag ðsa −sac Þ T a −T g
ð8Þ
cpc ¼ 1254 cpc ¼ 1254–1:75ðT−623Þ
for T ≤623K for T N623K
ð14Þ
The boundary conditions of Eq. (13) is given as. Surface of coal particle: λc sc
∂T jr¼rpc ¼ dqcod þ dqrad þ dqcg ∂r
ð15Þ
Center of coal particle: λc sc
∂T jr¼0 ¼ 0 ∂r
ð16Þ
(d) Convection heat transfer between the coal and pyrolysis gas 3.3. Calculation method
dqcg ¼ hcg sc T g −T c
ð9Þ
In Eqs. (4) and (5), the heat transfer coefficient of particle and gas h can be calculated from the Nu number proposed by Gunn et al. [21,22]: Nu ¼
dh ¼ 7−10ε þ 5ε2 1 þ 0:7Re0:2 Pr1=3 λ þ 1:33−2:4ε þ 1:2ε2 Re0:7 Pr1=3
ð10Þ
(e) Endothermic effect of pyrolysis process. As an endothermic process, the coal pyrolysis reaction will decrease the temperature of coal particles. To simplify the calculation in this model, the reaction heat is assumed to be proportional to the reaction conversion, regardless of the reaction conditions and coal species. As shown in Table 2, the value of the reaction heat in this paper is 300 kJ/kg. According to the above heat transfer mechanisms, the heat balance equations of ash and pyrolysis gas can be expressed as.
Four stage Runge–Kutta scheme is used to solve the ordinary differential equations (ODEs) which are composed of Eqs. (11) and (12). Thus, the temperature of Ta and Tg can be obtained. The finitedifference method (FDM) is used to discretize the partial differential equations, such as Eqs. (13), (15) and (16). The discrete algebraic equations are solved by the three diagonal matrix algorithm (TDMA). The continuous solution region is divided into grids of n × m (radial of the coal particle × axial of the moving bed) consisting of finite nodes. The internal radial distribution of coal particles is divided into 10 layers, so the radial step size of the coal particle is Δ r = rpc / 10. The axial step size of the moving bed is Δz = 0.002 m. Therefore, the internal temperature distribution of coal (T0,j, T1,j… Tn,j) versus bed height can be determined. Combining the internal temperature distribution of coal with Eq. (1), the volatile product yields and devolatilization rates of each coal layer versus bed height are calculated. The parameters for numerical calculation are shown in Table 2. 4. Results and discussion
1−ε us Aθ ρ cpa dT a ¼ −dqcod −dqrad −dqag 1þθ a
ð11Þ
1−ε ρ ðw −wÞcpg dT g ¼ dqag −dqcg 1 þ θ a0
ð12Þ
us Aθ
3.2.4. Internal heat conduction equations of coal particle
ρc cpc
! 2 ∂T ∂ T 2 ∂T ∂w −ρc0 ΔH ¼ λc 2 þ ∂t ∂t ∂ r r ∂r
ð13Þ
Table 3 Ash composition (wt%) of the circulating ash. SiO2
Al2O3
Fe2O3
CaO
MgO
TiO2
SO3
K2O
Na2O
P2O5
55.76
33.23
5.03
2.46
0.53
0.86
0.73
1.11
0.16
0.13
4.1. Temperature distribution of pyrolysis gas-ash-coal in the moving bed As shown in Fig. 2, it can be seen that, the temperature of pyrolysis gas firstly soars and then decreases smoothly with the height of the moving-bed. This is mainly because the boundary condition of pyrolysis gas temperature at z = 0 is calculated by the overall heat balance of the pyrolysis system. In other words, the temperature of the gas phase space above the bed material is assumed equal to the average temperature of the moving bed system. After entering the reactor bed, the temperature of pyrolysis gas rapidly increases and is close to the ash temperature, which proves that the convection heat transfer between gas and ash is very intense. Due to the strong radiation and convection heat transfers, the external temperature of coal increase rapidly with the height of moving-bed. However, the internal temperatures rise relatively slow because of the thermal resistance. As a result, the maximum temperature difference between the core and surface of coal has reached 406 K at the bed height of 0.05 m. The result indicates that, the rate-determining step of the large coal particle pyrolysis is the heat transfer, which is consistent with the study of Yan et al. [24].
230
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
Coal and ash particles from the mixing section
Coal
T 0,m
Pyrolysis gas Upper space z=0
T g0
T c0
T 1,m th 1 layer
T g,1 T a,1 T 0,1 , T 1,1 , ...T n,1
j th layer
T g,j T a,j T 0,j, T 1,j, ...T n,j
m
th
layer
z=L
T i,m T n-1,m Ash
T n,m
T g,m T a,m T 0,m , T 1,m ,...T n,m
Pyrolysis gas
Tn,m
Char and ash to the CFB Temperature profile of the moving bed
Temperature profile of coal particle
Fig. 1. Diagram of the simplifying model in the moving-bed reactor.
4.2. Released rate of the total volatile and the main volatile products versus axial bed height Fig. 3 shows the total volatile yields and devolatilization rates with the same calculation conditions as Fig. 2. It can be seen that the local devolatilization rates first increase and then decrease. The layer closer to the coal core has a higher but later peak value of the devolatilization rate. It is suspected that the internal temperature distribution may account for the later peak. Different from the temperature rising process, the devolatilization process of the coal surface do not start until a transient heating process (~0.016 m). As shown in Table 4, it can be seen that the activation energy of different volatile products is largest in the first temperature range, which is mainly due to the formation of the corresponding fragmented ions [13, 25]. As a result of the cleavage or polymerization of various functional groups [26,27], the activation energy of the total devolatilization reaction displays the order as: second range N first range N third range.
Fig. 4 shows that the evolution of the main volatile products is concentrated at the bed height of 0.08–0.24 m. Compared with other volatile products, CO2 is the first to reach the maximum released rate at ~ 0.1 m while C10H8 is the last at ~ 0.16 m. This is mainly because the higher the activation energy, the higher the barrier of product release. 4.3. Effect of operation parameters on the total devolatilization rate As Fig. 5 shows that, with the increase of the moving-bed velocity in Fig. 5(a), the maximum devolatilization rate slightly decreases. The major reason is that the circulating ash ambient temperature decreases along with the axial bed height (as shown in Fig. 2). The faster movingbed velocity will shorten the residence time in the hot circulating ash ambient, then resulting in a lower devolatilization rate and extended reaction time. Fig. 5(b) shows that increase the blending ratio of ash to coal from 3 to 6 will enhance the maximum devolatilization rate of coal particle and shorten the theoretical bed height. This can be
1120
960
Temperature (K)
Circulating ash Pyrolysis gas Coal center th
800
1 layer of coal th
2 layer of coal th
3 layer of coal th
640
4 layer of coal th
5 layer of coal th
6 layer of coal th
480
7 layer of coal th
8 layer of coal th
9 layer of coal th
320
10 layer of coal
0.00
0.06
0.12
0.18
0.24
0.30
0.36
z (m) Fig. 2. Temperature distribution of pyrolysis gas-ash-coal versus bed height. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
231
Rate of total volatile released th 10 layer 2.0
1.6
60
th
5 layer
Total volatile yields
Coal center
Whole particle
Whole particle
Coal center
50
40
th
5 layer th
10 layer
1.2
30
0.8
20
0.4
10
0.0
0
-0.04
0.00
0.04
0.08
0.12
0.16
0.20
0.24
0.28
Total volatile yields (wt%)
-1
Rate of total volatile released (wt%·s )
2.4
0.32
z (m) Fig. 3. Total volatile yields and devolatilization rate of the Shenmu coal particle versus axial bed height. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
explained by the higher blending ratio of ash to coal has increased the pyrolysis temperature. The effect of coal particle size on the total devolatilization rate is shown in Fig. 5(c). When the coal particle size increases from 5 mm to 12 mm, the maximum devolatilization rate of coal particle is decreased and the theoretical bed height is obviously extended. This is mainly because the internal thermal resistance has delayed the temperature increase of the large coal particle. Moreover, the decrease of specific surface area and convective heat transfer coefficient has weakened the radiation and convection heat transfer. It should be note that, by the weighted average of the particle size distribution, the non-uniform coal particle can also be predicted in the model. Fig. 5(d) shows the effect of ash particle diameter on the total devolatilization rate. It can be seen that, the pyrolysis of the larger ash
particle tends to be slower as a result of the smaller heat transfer area of the radiation and convection. In Fig. 5(e) and Fig. 5(f), with the increase of the initial temperature of ash and preheating temperature of coal, the theoretical heights of the moving bed are both decreased. Calculation results show that, the theoretical bed height of pyrolysis drops from 0.375 m to 0.125 m when the preheating temperature of coal increases from 900 K to 1200 K. Similarly, the theoretical bed height drops from 0.17 m to 0.13 m when the initial temperature of ash increases from 298 K to 448 K. In the present model, the calculational result shows that the heat conduction inside the large coal particle is the rate-determining step of the pyrolysis reaction. But for the heat transfer process from the gas/ash phase to the coal surface, which mechanism plays the most
Table 4 Kinetic parameters of the main volatile products of Shenmu coal (n = 1).
CH4 CO2 H2 CO C2H4 C2H6
C6H6 C7H8
C8H10
C10H8
Total volatile
Temperature range (°C)
ln(k0jR/αEj)
−Ej/R
R2
k0j (min−1)
Ej (kJ/mol)
wj* (wt%)
429–622 622–674 330–610 610–660 493–700 465–615 615–673 465–615 615–673 470–577 577–615 615–648 511–620 620–674 493–610 610–640 640–670 494–605 605–631 631–674 511–537 537–560 560–590 270 ~ 375 375 ~ 511 511 ~ 684
3.13 −10.71 −11.81 −13.92 −5.67 −2.19 −12.33 −3.84 −14.00 1.76 −9.69 −15.00 −8.31 −17.04 −3.36 −14.07 −18.33 1.29 −11.23 −16.36 66.88 21.50 −4.40 −3.53 3.11 −8.21
−18,843 −6718 −5598 −3602 −14,292 −15,261 −6408 −15,257 −6397 −17,222 −7642 −2940 −13,295 −5545 −16,126 −6792 −2915 −19,184 −8369 −3728 −71,123 −34,457 −12,915 −8321 −12,710 −4058
0.9885 0.9824 0.9801 0.9801 0.9883 0.9840 0.9878 0.9839 0.9877 0.9883 0.9906 0.9852 0.9959 0.9805 0.9880 0.9911 0.9863 0.9826 0.9945 0.9785 0.9931 0.9936 0.9807 0.9943 0.9856 0.9922
2.16E + 07 7.49 2.08 0.16 491.07 8.58E + 04 1.41 1.64E + 04 0.27 5.01E + 06 23.55 0.05 164.02 0.01 2.81E + 04 0.26 1.60E-03 3.48E + 06 5.58 0.01 3.94E + 35 3.74E + 15 7996.35 1.22E + 04 1.43E + 07 55.35
156.66 55.85 46.54 29.95 118.82 126.88 53.28 126.84 53.19 143.19 63.54 24.44 110.53 46.10 134.07 56.47 24.24 159.49 69.58 31.00 591.32 286.47 107.38 69.18 105.67 33.74
1.82 1.89 0.11 0.49 0.09 1.18
0.49 0.05
0.14
0.28
28.70
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
-1
Rate of volatile products released (wt%·s )
232
CH4 C2 H 6 CO2 CO
-3
3.0x10
-3
2.0x10
-3
1.0x10
0.0 -6
H2 C2H 4 C6H 6
1.8x10
-6
1.2x10
-7
6.0x10
0.0 C7H8 C8H10 C10H8
-5
3.0x10
-5
2.0x10
-5
1.0x10
0.0 0.00
0.08
0.16
z (m) 0.24
0.32
0.40
Fig. 4. Released rate of the main volatile products versus axial bed height. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
important role is still unknown. In order to get a deeper understanding on the heating and pyrolysis behavior of the large coal particle, different mechanisms were calculated and compared separately by ignoring the corresponding single heat transfer mechanism. Then, the important order of the heat transfer mechanisms can be given as: radiation heat transfer N convection heat transfer between coal and pyrolysis gas N endothermic effect. The previous literatures [6,10] have reported the experimental study on coal pyrolysis by the circulating ash heat carrier. The results indicate that the maximum volatile yield can be achieved when the pyrolysis time is ~4 min. However, the modeling data show that residence time of ~2 min is adequate as the velocity of the moving bed is determined to be 0.125 m/min. Considering the ideal assumption of the adiabatic system, perfect mixing of the ash and coal, spherical coal and ash
-1
Rate of total volatile released (wt%·s )
0.9 0.6 0.3 0.0 2.8 2.1 1.4 0.7
particle, etc. in the present model, it can be inferred that the 3–4 min design residence time of the moving bed reactor in the poly-generation technology is reasonable and bounteous. 4.4. Model calculational results of different coal species The model calculational results of different coal species such as the Indian brown coal [28], the Chinese brown coal [29], Datong bituminous coal [30] and Shenmu bituminous coal are compared in Fig. 6. The result indicates that, the evolution characteristics of different coal samples vary obviously as a result of their respective coal properties. Compared with the Chinese and Indian brown coal samples, the theoretical bed heights of the Datong and Shenmu bituminous coal tend to be higher, while the maximum total devolatilization rates are relatively small.
(a) Velocity of moving bed us=0.10m/min us=0.13m/min us=0.15m/min us=0.20m/min
(b) Blending ratio of ash to coal ash/coal=6.0 ash/coal=5.0 ash/coal=4.0 ash/coal=3.0
(c) Coal particle diameter 5mm 8mm 10mm 12mm
(d) Ash particle diameter 0.57mm 0.67mm 0.87mm 0.97mm
(e) Initial temperature of ash 1200K 1100K 1000K 900K
(f) Preheating temperature of coal 298K 348K 398K 448K
0.0 0.9 0.6 0.3 0.0 0.0
0.1
0.2
z (m)
0.3
0.4
0.0
0.1
0.2
0.3
0.4
z (m)
Fig. 5. Effect of operation parameters on the total devolatilization rate. Calculation parameters: β = 0.8; χ = 0.5.
-1
Rate of total volatile released (wt%·s )
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
Rac rpc sac
Coal samples 4
Indian brown coal Chinese brown coal Datong bituminous coal Shenmu bituminous coal
3
2
1
sa sc Ta Tc Tg us w w⁎ wj w⁎j z
0 0.00
0.04
0.08
0.12
0.16
0.20
233
ratio of contact area to coal surface area radius of the particle, m contact surface area of ash and coal in dz. volume thickness pyrolyzer, m2 ash surface area in dz. volume thickness pyrolyzer, m2 coal surface area in dz. volume thickness pyrolyzer, m2 temperature of the ash, K temperature of the external layer of coal, K temperature of the pyrolysis gas, K velocity of the moving bed, m/s mass conversion fraction of coal evolved up to time t, % the value of w at t = ∞, % mass fractional of gas j product evolved up to time t, % the value of wj at t = ∞, % bed height, m
0.24
z (m) Fig. 6. Model calculational results of different coal species. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
This is mainly because the coal rank of bituminous coal is higher than the brown coal. The larger activation energy of the bituminous coal has limited the devolatilization reaction. It can be pointed out that, combined with the kinetic study of coal sample, the model can achieve the adaptability of different coal species. 5. Conclusion In this study, a mathematical pyrolysis model of non-isothermal coal particles by circulating ash carrier toward the axial dimension of the moving bed is developed. Based on the results of Coats–Redfern integral method and TG-MS experiment, the model can predict not only the temperature distribution of pyrolysis gas-coal-ash toward the pyrolyzer height, but also the evolution of total volatile and the main volatile products. For the large coal particle pyrolysis, heat transfer is the ratedetermining step. The devolatilization process has a transient delay compared with the heating process. The evolution of the main volatile products is concentrated at the bed height of 0.08–0.24 m. The calculating result indicates that the 3–4 min design residence time of the moving bed reactor in the poly-generation technology is reasonable and bounteous. Increase the blending ratio of ash to coal, preheating temperature of coal and initial temperature of ash can accelerate the devolatilization process. Combined with the kinetic study of the coal rank and coal species, the model can be applied to different coal. The present model may provide theoretical prediction data for the scaleup of the coal pyrolysis/CFB combustion poly-generation technology as well as a deeper understanding on the evolution mechanism of the coal volatile products. Nomenclature A cross area of the moving-bed pyrolyzer, m2 aa specific surface area of ash particle, m2/m3 ac specific surface area of coal particle, m2/m3 cpa heat capacity of ash, J/(kg K) cpc heat capacity of the coal, J/(kg K) cpg heat capacity of pyrolysis gas, J/(kg K) dac contact area diameter of ash and coal, m dpa diameter of the ash particle, m dpc diameter of the coal particle, m Ej activation energy of reaction forming product j, kJ/mol hag heat transfer coefficient of ash and gas, W/(m2 K) hcg heat transfer coefficient of coal and gas, W/(m2 K) k0j pre-exponential factor of reaction forming product j, min−1 n reaction order R ideal gas constant, J/(mol K)
Greek symbols α β ε ΔH λa λc μg θ ρc ρc0 σ
heating rate, K/min emissivity of coal particle the void ratio reaction heat, kJ/kg thermal conductivity of ash, W/(m K) thermal conductivity of coal, W/(m K) viscosity of the gas, Pa·s volume ratio of ash to coal density of the coal, kg/m3 density of the raw coal, kg/m3 Stefan–Boltzman constant, (W/m2 K4)
Acknowledgement The work was supported by the National Natural Science Foundation of China (Grant No. 21376142). References [1] Z. Li, Z. Guo, X. Gong, H. Tang, Kinetic characteristics of pulverized coal combustion in the two-phase flow, Energy 55 (2013) 585–593. [2] M. Eriksson, M.R. Golriz, Radiation heat transfer in circulating fluidized bed combustors, Int. J. Therm. Sci. 44 (2005) 399–409. [3] X. Qu, P. Liang, R. Zhang, Z.X. Gan, J.C. Bi, Sulfur transformation in the process of circulating fluidized bed combustion combined with coal pyrolysis, Energy Fuel 24 (2010) 5023–5027. [4] Y.J. He, F.Z. Xiao, Z.H. Luo, Numerical modeling of the cavity phenomenon and its elimination way in rectangular radial moving bed reactor, Powder Technol. 274 (2015) 28–36. [5] X. Qu, P. Liang, Z.F. Wang, R. Zhang, D.K. Sun, X.K. Gong, Z.X. Gan, J.C. Bi, Pilot development of a polygeneration process of circulating fluidized bed combustion combined with coal pyrolysis, Chem. Eng. Technol. 34 (2011) 61–68. [6] P. Liang, Z.F. Wang, J.C. Bi, Simulation of coal pyrolysis by solid heat carrier in a moving-bed pyrolyzer, Fuel 87 (2008) 435–442. [7] D.G. Lai, Z.H. Chen, Y. Shi, L.X. Lin, J.H. Zhan, S.Q. Gao, G.W. Xu, Pyrolysis of oil shale by solid heat carrier in an innovative moving bed with internals, Fuel 159 (2015) 943–951. [8] G.X. Hu, X.W. Gong, H. Hao, H.J. Fan, Z. Wang, Experimental studies on flow and pyrolysis of coal with solid heat carrier in a modified rotating cone reactor, Chem. Eng. Process. Process Intensif. 47 (2008) 1777–1785. [9] G.X. Hu, H.J. Fan, Y.Q. Liu, Experimental studies on pyrolysis of Datong coal with solid heat carrier in a fixed bed, Fuel Process. Technol. 69 (2001) 221–228. [10] P. Liang, Z.F. Wang, J.C. Bi, Process characteristics investigation of simulated circulating fluidized bed combustion combined with coal pyrolysis, Fuel Process. Technol. 88 (2007) 23–28. [11] R. Xiong, L. Dong, J. Yu, X.F. Zhang, L. Jin, G.W. Xu, Fundamentals of coal topping gasification: characterization of pyrolysis topping in a fluidized bed reactor, Fuel Process. Technol. 91 (2010) 810–817. [12] X.L. Liu, G. Wang, G. Pan, Z. Wen, Numerical analysis of heat transfer and volatile evolution of coal particle, Fuel 106 (2013) 667–673. [13] C.A. Heidenreich, H.M. Yan, D.K. Zhang, Mathematical modelling of pyrolysis of large coal particles—estimation of kinetic parameters for methane evolution, Fuel 78 (1999) 557–566. [14] P. Liang, Y.Q. Zhang, W.M. Jiang, A.F. Wei, T. Liu, J.F. Wu, Simulation study of Shenmu coal pyrolysis by gas heat carrier based on a moving bed, Energy Fuel 29 (2015) 7727–7733.
234
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
[15] F.J. Higuera, Numerical simulation of the devolatilization of a moving coal particle, Combust. Flame 156 (2009) 1023–1034. [16] S.F. Zhang, F. Zhu, C.G. Bai, L.Y. Wen, C. Zou, Thermal behavior and kinetics of the pyrolysis of the coal used in the COREX process, J. Anal. Appl. Pyrolysis 104 (2013) 660–666. [17] Q. Dong, Y. Xiong, Kinetics study on conventional and microwave pyrolysis of moso bamboo, Bioresour. Technol. 171 (2014) 127–131. [18] S. Ceylan, Y. Topcu, Z. Ceylan, Thermal behaviour and kinetics of alga Polysiphonia elongata biomass during pyrolysis, Bioresour. Technol. 171 (2014) 193–198. [19] A.V. Luikov, A.G. Shashkov, L.L. Vasilev, Y.E. Fraiman, Thermal conductivity of porous systems, Int. J. Heat Mass Transf. 11 (1968) 117–140. [20] B.A. Adesanya, H.N. Pham, Mathematical modelling of devolatilization of large coal particles in a convective environment, Fuel 74 (1995) 896–902. [21] D.J. Gunn, Transfer of heat or mass to particle in fixed and fluidized beds, Int. J. Heat Mass Transf. 21 (1978) 467–476. [22] S.Y. Wang, G.D. Liu, Y.B. Wu, J.H. Chen, Y.J. Liu, L.X. Wei, Numerical investigation of gas-to-particle cluster convective heat transfer in circulating fluidized beds, Int. J. Heat Mass Transf. 53 (2010) 3102–3110. [23] S. Badzioch, P.G.W. Hawksley, Kinetics of thermal decomposition of pulverized coal particles, Ind. Eng. Chem. Process. Des. Dev. 9 (1970) 521–530.
[24] B.H. Yan, Y. Cheng, Y. Jin, C.Y. Guo, Analysis of particle heating and devolatilization during rapid pyrolysis in a thermal plasma reactor, Fuel Process. Technol. 100 (2012) 1–10. [25] C. Lievens, D.H. Ci, Y. Bai, L.G. Ma, R. Zhang, J.Y. Chen, Q.Q. Gai, Y.H. Long, X.F. Guo, A study of slow pyrolysis of one low rank coal via pyrolysis–GC/MS, Fuel Process. Technol. 116 (2013) 85–93. [26] X.C. Lin, C.H. Wang, K. Ideta, J. Miyawaki, Y. Nishiyama, Y.G. Wang, S. Yoon, I. Mochida, Insights into the functional group transformation of a Chinese brown coal during slow pyrolysis by combining various experiments, Fuel 118 (2014) 257–264. [27] M. Wang, Z.S. Li, W.B. Huang, J.X. Yang, H.T. Xue, Coal pyrolysis characteristics by TG–MS and its late gas generation potential, Fuel 156 (2015) 243–253. [28] V.R. Patel, R.N. Patel, V.J. Rao, Kinetic parameter estimation of lignite by thermogravimetric analysis, Proc. Eng. 51 (2013) 727–734. [29] Y. Xu, Y.F. Zhang, Y. Wang, G.J. Zhang, L. Chen, Thermogravimetric study of the kinetics and characteristics of the pyrolysis of lignite, React. Kinet. Mech. Catal. 110 (2013) 225–235. [30] S.F. Zhang, F. Zhu, C.G. Bai, L.Y. Wen, C. Zou, Thermal behavior and kinetics of the pyrolysis of the coal used in the COREX process, J. Anal. Appl. Pyrolysis 104 (2013) 660–666.
Contents lists available at ScienceDirect
Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc
Research article
Simulation of large coal particles pyrolysis by circulating ash heat carrier toward the axial dimension of the moving bed Ya-Qing Zhang, Jia-Long Zhu, Xiao-Hang Wang, Xi-Wang Zhang, Shi-Xue Zhou, Peng Liang ⁎ College of Chemical and Environmental Engineering, Shandong University of Science and Technology, Qingdao 266590, Shandong, China
a r t i c l e
i n f o
Article history: Received 19 January 2016 Received in revised form 24 August 2016 Accepted 25 August 2016 Available online 2 September 2016 Keywords: Coal pyrolysis Solid heat carrier Large coal particle Heat transfer Volatile evolution
a b s t r a c t A heat transfer, reaction, pyrolysis mathematical model for the non-isothermal coal particles by using circulating ash as heat carrier toward the moving bed has been established. Combined with the Thermogravimetry-Mass spectrometry technology and Coats–Redfern integral method, the model has the ability to predict the temperature distribution of pyrolysis gas-coal-ash as well as the evolution characteristics of the main volatile products (such as CH4, CO2, H2, CO, C2H4, C2H6, C6H6, C7H8, C8H10, C10H8). The results show that, the maximum temperature difference between the core and surface of coal (10 mm) has reached 406 K at the bed height of 0.05 m. The layer closer to the coal core has a higher but later peak value of the devolatilization rate. The evolution of the main volatile products is concentrated at the bed height of 0.08–0.24 m. The velocity of the moving bed, blending ratio of ash to coal, coal particle size, preheating temperature of coal and initial temperature of ash have obvious influence on the devolatilization process. Radiation is the most significant factor affecting the devolatilization behavior. The model can be applied to different coal species. This study can provide a theoretic foundation for the amplification design of the moving-bed reactor in the poly-generation system. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The majority of the raw coal in China is used for direct combustion [1] which is inefficient and serious pollution. Fortunately, the circulating fluidized bed (CFB) combustion technology has been developed greatly for its better fuel flexibility, low pollution and load adjustability in recent years [2,3]. Similarly, the moving-bed pyrolyzer is increasingly valued because of its high adjustment capacity, stable operation and uniform heating [4]. So this provides a base for the poly-generation technology of coal pyrolysis and CFB combustion to achieve the grading conversion of coal. High-temperature ash, which comes from the CFB boiler is used as solid heat carrier, mixed with coal is fed into the pyrolyzer simultaneously. Then, the coal is heated to pyrolysis temperature and produce gas and tar. The char produced in the pyrolyzer is returned to CFB boiler to provide electricity and heat [5]. Thus the utilization of coal with high efficiency and low emissions is realized. Furthermore, since the pyrolysis reactor is independent of the CFB boiler, the CFB boiler can be operated separately when the pyrolysis system is out of work. Therefore, the moving-bed pyrolyzer as the critical equipment in the poly-generation technology [6], its operating state is directly relates to the stability and economy of the whole system. The coal pyrolysis heated by solid or gas carriers has a higher efficiency of heat transfer than the traditional heating technologies. And ⁎ Corresponding author. E-mail address: [email protected] (P. Liang).
http://dx.doi.org/10.1016/j.fuproc.2016.08.037 0378-3820/© 2016 Elsevier B.V. All rights reserved.
compared with coal pyrolysis by gas heat carrier, the solid heat carrier not only provide higher heating rate but also avoid the volatiles dilution by the inert gas [7]. So the experimental works on coal pyrolysis by solid heat carrier have been widely valued in recent years [8–11]. However, the simulation studies are relatively few reported. To enable a better understanding on the poly-generation process, it is necessary to establish a mathematical model for the analysis of coal pyrolysis behavior, especially for the prediction and amplification of industrial processes. Some researchers [12–14] carried out the radial stratification of large particles to investigate the temperature-rise period. The studies showed that the coal particle size has significant influence on the pyrolysis behavior. However, the models are inappropriate for the process of coal pyrolysis by solid heat carrier because the heat transfer mechanism outside the coal surface considered only the heat radiation and heat convection. Higuera [15] reported the devolatilization of a moving coal particle by using a competing reaction pyrolysis reaction model. He pointed out that the relative motion of the particles would improve the heat transfer process between the surrounding gas and the particle. Liang et al. [6] reported a steady state, axial-dimensional, numerical model for coal pyrolysis by solid heat carrier toward the moving-bed. However, the coal particle was assumed as an isothermal body in the above two models, the existence of temperature gradient as a result of internal thermal resistance of coal particles should not be ignored. Therefore, a reasonable pyrolysis model of the non-isothermal coal particles by circulating ash carrier toward the axial dimension of the moving bed should be established.
228
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
In this study, the Thermogravimetry-Mass spectrometry (TG-MS) technology and kinetic study were applied to modeling the nonisothermal coal pyrolysis by circulating ash heat carrier. Based on the heat transfer theory, not only the internal pyrolysis behavior of coal particles but also the temperature distribution of pyrolysis gas-coal-ash toward the pyrolyzer can be predicted. Combined with the multiplereaction model, the evolution characteristics of the main volatile products is revealed. This work will provide a better understanding on the poly-generation technology of coal pyrolysis combined with CFB combustion. 2. Experimental As shown in Table 1, Shenmu coal from Shaanxi Province is used in this study. The kinetic parameters for model calculation is obtained in the TG-MS system which is composed of a thermogravimetric analyzer (TG/TGA, Setaram Setsys-Evolution, Caluire, France) and a mass spectrometer (MS, Balzers Omnistar, Brügg, Switzerland). Before the TGMS experiment is carried out, the coal sample was crushed and sieved to ≤ 74 μm, then dried at 105 °C for 4 h. The high-purity nitrogen (99.999 vol.%) was introduced into the TG-MS system with a flow rate of 80 mL/min. In the TG analyzer, the Shenmu coal sample (15 mg) was heated from the room temperature to 700 °C at a heating rate of 50 °C/min. The evolution characteristics of the gaseous products were analyzed in the MS with the ionization voltage of 40 eV and detection charge-mass ratio ranging from 0 to 300 amu. The physical characteristics of CFB ash, coal, and pyrolysis gas used in the experiment are shown in Table 2. The composition analysis of the circulating ash is shown in Table 3. 3. Mathematical model
As Fig. 1 shows that, the coal coupled with high-temperature circulating ash which comes from the mixing section is added into the reactor. As the mixed material moving downwards, the coal is heated by the surrounding ash and pyrolysis gas released from the under layer. Then the pyrolysis reaction occurs with the increase of coal particle temperature. The pyrolysis gas escapes upward, while the ash and char move downwards and return to the CFB boiler. In the moving-bed reactor, the height is axially divided into m layers. At any jth layer in the reactor, the temperature of the pyrolysis gas and ash surrounding the coal particles is respectively Tg,j and Ta,j. In the coal particle, the diameter is radially divided into n layers. From the core to surface, the internal temperature distribution of the coal particle is respectively T0,j, T1,j, … Tn,j. Assumptions of the model simplification are as follows: (1) The particle sizes of coal and ash remain unchanged. (2) Ash is assumed as an isothermal, inert, spherical particle. (3) Spherical coal particle is tightly surrounded by the ash, and the characteristics of a single coal particle can predict the coal cluster in the same bed layer. (4) The diffusion process of pyrolysis products evolved from the internal to external of coal particles is ignored. (5) The moving-bed reactor is operated at a steady state. (6) The calculation time step is small enough to ensure little
Table 1 The proximate and ultimate analyses of Shenmu coal sample (wt%).
M 7.21
A 3.89
V 31.54
Ultimate analysis (daf) FC 57.36
3.2. Mathematical model description of coal pyrolysis by solid heat carrier 3.2.1. Pyrolysis kinetic equation In this model, the multiple independent parallel first-order reactions are assumed to describe the evolution characteristics of the volatile products. Therefore, the kinetic equation of the main volatile products (such as CH4, CO2, H2, CO, C2H4, C2H6, C6H6, C7H8, C8H10, C10H8) and the total volatile (including but not all of the above main volatile products) can be expressed as dw j Ej ¼ k0 j exp − w j −w j dt RT
ð1Þ
where the final volatile product yields wj* are obtained from the test of low temperature distillation of coal by aluminum retort (GB/T4802010), the pre-exponential factor k0j and activation energy Ej are calculated by the TG-MS result. In this study, Coat–Redfern method [16–18] is applied to calculate the k0j and Ej because of its wider applicability and better fitting effect. The equations of Coats-Redfern method can be expressed as. ln 1−w j k0 j R E j ln − − ¼ ln αE j RT T2 " ln
1−n # 1− 1−w j T 2 ð1−nÞ
¼ ln
k0 j R E j − αE j RT
ðn ¼ 1Þ
ð2Þ
ðn≠1Þ
ð3Þ
In Eq. (2) or (3), the left side versus 1/T shows a linear relationship. Therefore, the pre-exponential factor k0j and activation energy Ej can be calculated from the intercept and slope of the straight line.
3.1. Analysis and assumptions
Proximate analysis (ad)
temperature change of coal-ash-gas three phases in any adjacent iterative computation.
C 80.88
H 5.09
N 1.08
S 0.29
Oa 12.66
a By difference; d, dry; daf, dry ash free; M, moisture; A, ash; V, volatile matter; FC, fixed carbon.
3.2.2. Mass conservation equation ρc d dw ρc0 ¼− dt dt
ð4Þ
3.2.3. Heat transfer mechanisms in the moving-bed reactor Five heat transfer mechanisms in the moving-bed reactor are considered as follows: (a) Thermal conduction between the circulating ash and coal particles. In the moving-bed reactor, the coal particles are uniformly distributed in the circulating ash which has a relatively small size. In the moving downwards process of the mixed material, a circular contact surface between the coal and ash particle will be produced due to the mutual extrusion. Then the heat can be passed through the contact surface from the circulating ash to the coal surface. A formula has been proposed by Luikov [19] for calculating the size of the contact surface, dac = 0.2dpa. Therefore, taken the volume thickness of dz. in the pyrolyzer as the calculational unit, the surface area of coal and ash can be respectively writ1−ε ten as sc ¼ ac 1−ε 1þθ Adz and sa ¼ aa 1þθ Aθdz. In the above two equations, θ is the volume ratio of ash to coal in the pyrolyzer, ac and aa are respectively the specific surface area of coal and ash, ac = 6(1 − ε)/dpc, aa = 6(1 − ε)/dpa. The contact area of coal and ash can be given by the equation sac = scRac, where Rac [6] is the ratio of the contact area to coal surface area. Based on the above analysis, the heat flow of thermal conduction between ash and coal particles is expressed as dqcod ¼
2λa λc T a −T c sac λa þ λc dpa
ð5Þ
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
229
Table 2 Parameters of ash, coal, and pyrolysis gas for numerical calculation. ΔH (kJ/kg)
ε
us (m/min)
ρa0 (kg/m3)
ρc0 (kg/m3)
Ta0 (K)
Tc0 (K)
cpa (J/(kg·K))
cpg (J/(kg·K))
λa (W/(m·K))
λg (W/(m·K))
dpa (m)
μg (Pa·s)
β
χ
σ (J/(s·m2·K4))
−300
0.415
0.125
1550
1250
1100
298
840
1214
0.29
0.0742
6.70 × 10−4
3.79 × 10−5
0.8
0.5
5.67 × 10−8
where λc [20] is the thermal conductivity of coal, and it can be given by the equations.
where cpc [23] is the heat capacity of coal, and it can be given by the equations:
λc ¼ 0:23 λc ¼ 0:23 þ 2:24 10−5 ðT−673Þ1:8
for T ≤673 K for T N673K
ð6Þ
(b) Radiation heat transfer between the circulating ash and coal particles dqrad ¼ βσχ ðsc −sac Þ T 4a −T 4c
ð7Þ
(c) Convection heat transfer between the circulating ash and pyrolysis gas dqag ¼ hag ðsa −sac Þ T a −T g
ð8Þ
cpc ¼ 1254 cpc ¼ 1254–1:75ðT−623Þ
for T ≤623K for T N623K
ð14Þ
The boundary conditions of Eq. (13) is given as. Surface of coal particle: λc sc
∂T jr¼rpc ¼ dqcod þ dqrad þ dqcg ∂r
ð15Þ
Center of coal particle: λc sc
∂T jr¼0 ¼ 0 ∂r
ð16Þ
(d) Convection heat transfer between the coal and pyrolysis gas 3.3. Calculation method
dqcg ¼ hcg sc T g −T c
ð9Þ
In Eqs. (4) and (5), the heat transfer coefficient of particle and gas h can be calculated from the Nu number proposed by Gunn et al. [21,22]: Nu ¼
dh ¼ 7−10ε þ 5ε2 1 þ 0:7Re0:2 Pr1=3 λ þ 1:33−2:4ε þ 1:2ε2 Re0:7 Pr1=3
ð10Þ
(e) Endothermic effect of pyrolysis process. As an endothermic process, the coal pyrolysis reaction will decrease the temperature of coal particles. To simplify the calculation in this model, the reaction heat is assumed to be proportional to the reaction conversion, regardless of the reaction conditions and coal species. As shown in Table 2, the value of the reaction heat in this paper is 300 kJ/kg. According to the above heat transfer mechanisms, the heat balance equations of ash and pyrolysis gas can be expressed as.
Four stage Runge–Kutta scheme is used to solve the ordinary differential equations (ODEs) which are composed of Eqs. (11) and (12). Thus, the temperature of Ta and Tg can be obtained. The finitedifference method (FDM) is used to discretize the partial differential equations, such as Eqs. (13), (15) and (16). The discrete algebraic equations are solved by the three diagonal matrix algorithm (TDMA). The continuous solution region is divided into grids of n × m (radial of the coal particle × axial of the moving bed) consisting of finite nodes. The internal radial distribution of coal particles is divided into 10 layers, so the radial step size of the coal particle is Δ r = rpc / 10. The axial step size of the moving bed is Δz = 0.002 m. Therefore, the internal temperature distribution of coal (T0,j, T1,j… Tn,j) versus bed height can be determined. Combining the internal temperature distribution of coal with Eq. (1), the volatile product yields and devolatilization rates of each coal layer versus bed height are calculated. The parameters for numerical calculation are shown in Table 2. 4. Results and discussion
1−ε us Aθ ρ cpa dT a ¼ −dqcod −dqrad −dqag 1þθ a
ð11Þ
1−ε ρ ðw −wÞcpg dT g ¼ dqag −dqcg 1 þ θ a0
ð12Þ
us Aθ
3.2.4. Internal heat conduction equations of coal particle
ρc cpc
! 2 ∂T ∂ T 2 ∂T ∂w −ρc0 ΔH ¼ λc 2 þ ∂t ∂t ∂ r r ∂r
ð13Þ
Table 3 Ash composition (wt%) of the circulating ash. SiO2
Al2O3
Fe2O3
CaO
MgO
TiO2
SO3
K2O
Na2O
P2O5
55.76
33.23
5.03
2.46
0.53
0.86
0.73
1.11
0.16
0.13
4.1. Temperature distribution of pyrolysis gas-ash-coal in the moving bed As shown in Fig. 2, it can be seen that, the temperature of pyrolysis gas firstly soars and then decreases smoothly with the height of the moving-bed. This is mainly because the boundary condition of pyrolysis gas temperature at z = 0 is calculated by the overall heat balance of the pyrolysis system. In other words, the temperature of the gas phase space above the bed material is assumed equal to the average temperature of the moving bed system. After entering the reactor bed, the temperature of pyrolysis gas rapidly increases and is close to the ash temperature, which proves that the convection heat transfer between gas and ash is very intense. Due to the strong radiation and convection heat transfers, the external temperature of coal increase rapidly with the height of moving-bed. However, the internal temperatures rise relatively slow because of the thermal resistance. As a result, the maximum temperature difference between the core and surface of coal has reached 406 K at the bed height of 0.05 m. The result indicates that, the rate-determining step of the large coal particle pyrolysis is the heat transfer, which is consistent with the study of Yan et al. [24].
230
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
Coal and ash particles from the mixing section
Coal
T 0,m
Pyrolysis gas Upper space z=0
T g0
T c0
T 1,m th 1 layer
T g,1 T a,1 T 0,1 , T 1,1 , ...T n,1
j th layer
T g,j T a,j T 0,j, T 1,j, ...T n,j
m
th
layer
z=L
T i,m T n-1,m Ash
T n,m
T g,m T a,m T 0,m , T 1,m ,...T n,m
Pyrolysis gas
Tn,m
Char and ash to the CFB Temperature profile of the moving bed
Temperature profile of coal particle
Fig. 1. Diagram of the simplifying model in the moving-bed reactor.
4.2. Released rate of the total volatile and the main volatile products versus axial bed height Fig. 3 shows the total volatile yields and devolatilization rates with the same calculation conditions as Fig. 2. It can be seen that the local devolatilization rates first increase and then decrease. The layer closer to the coal core has a higher but later peak value of the devolatilization rate. It is suspected that the internal temperature distribution may account for the later peak. Different from the temperature rising process, the devolatilization process of the coal surface do not start until a transient heating process (~0.016 m). As shown in Table 4, it can be seen that the activation energy of different volatile products is largest in the first temperature range, which is mainly due to the formation of the corresponding fragmented ions [13, 25]. As a result of the cleavage or polymerization of various functional groups [26,27], the activation energy of the total devolatilization reaction displays the order as: second range N first range N third range.
Fig. 4 shows that the evolution of the main volatile products is concentrated at the bed height of 0.08–0.24 m. Compared with other volatile products, CO2 is the first to reach the maximum released rate at ~ 0.1 m while C10H8 is the last at ~ 0.16 m. This is mainly because the higher the activation energy, the higher the barrier of product release. 4.3. Effect of operation parameters on the total devolatilization rate As Fig. 5 shows that, with the increase of the moving-bed velocity in Fig. 5(a), the maximum devolatilization rate slightly decreases. The major reason is that the circulating ash ambient temperature decreases along with the axial bed height (as shown in Fig. 2). The faster movingbed velocity will shorten the residence time in the hot circulating ash ambient, then resulting in a lower devolatilization rate and extended reaction time. Fig. 5(b) shows that increase the blending ratio of ash to coal from 3 to 6 will enhance the maximum devolatilization rate of coal particle and shorten the theoretical bed height. This can be
1120
960
Temperature (K)
Circulating ash Pyrolysis gas Coal center th
800
1 layer of coal th
2 layer of coal th
3 layer of coal th
640
4 layer of coal th
5 layer of coal th
6 layer of coal th
480
7 layer of coal th
8 layer of coal th
9 layer of coal th
320
10 layer of coal
0.00
0.06
0.12
0.18
0.24
0.30
0.36
z (m) Fig. 2. Temperature distribution of pyrolysis gas-ash-coal versus bed height. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
231
Rate of total volatile released th 10 layer 2.0
1.6
60
th
5 layer
Total volatile yields
Coal center
Whole particle
Whole particle
Coal center
50
40
th
5 layer th
10 layer
1.2
30
0.8
20
0.4
10
0.0
0
-0.04
0.00
0.04
0.08
0.12
0.16
0.20
0.24
0.28
Total volatile yields (wt%)
-1
Rate of total volatile released (wt%·s )
2.4
0.32
z (m) Fig. 3. Total volatile yields and devolatilization rate of the Shenmu coal particle versus axial bed height. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
explained by the higher blending ratio of ash to coal has increased the pyrolysis temperature. The effect of coal particle size on the total devolatilization rate is shown in Fig. 5(c). When the coal particle size increases from 5 mm to 12 mm, the maximum devolatilization rate of coal particle is decreased and the theoretical bed height is obviously extended. This is mainly because the internal thermal resistance has delayed the temperature increase of the large coal particle. Moreover, the decrease of specific surface area and convective heat transfer coefficient has weakened the radiation and convection heat transfer. It should be note that, by the weighted average of the particle size distribution, the non-uniform coal particle can also be predicted in the model. Fig. 5(d) shows the effect of ash particle diameter on the total devolatilization rate. It can be seen that, the pyrolysis of the larger ash
particle tends to be slower as a result of the smaller heat transfer area of the radiation and convection. In Fig. 5(e) and Fig. 5(f), with the increase of the initial temperature of ash and preheating temperature of coal, the theoretical heights of the moving bed are both decreased. Calculation results show that, the theoretical bed height of pyrolysis drops from 0.375 m to 0.125 m when the preheating temperature of coal increases from 900 K to 1200 K. Similarly, the theoretical bed height drops from 0.17 m to 0.13 m when the initial temperature of ash increases from 298 K to 448 K. In the present model, the calculational result shows that the heat conduction inside the large coal particle is the rate-determining step of the pyrolysis reaction. But for the heat transfer process from the gas/ash phase to the coal surface, which mechanism plays the most
Table 4 Kinetic parameters of the main volatile products of Shenmu coal (n = 1).
CH4 CO2 H2 CO C2H4 C2H6
C6H6 C7H8
C8H10
C10H8
Total volatile
Temperature range (°C)
ln(k0jR/αEj)
−Ej/R
R2
k0j (min−1)
Ej (kJ/mol)
wj* (wt%)
429–622 622–674 330–610 610–660 493–700 465–615 615–673 465–615 615–673 470–577 577–615 615–648 511–620 620–674 493–610 610–640 640–670 494–605 605–631 631–674 511–537 537–560 560–590 270 ~ 375 375 ~ 511 511 ~ 684
3.13 −10.71 −11.81 −13.92 −5.67 −2.19 −12.33 −3.84 −14.00 1.76 −9.69 −15.00 −8.31 −17.04 −3.36 −14.07 −18.33 1.29 −11.23 −16.36 66.88 21.50 −4.40 −3.53 3.11 −8.21
−18,843 −6718 −5598 −3602 −14,292 −15,261 −6408 −15,257 −6397 −17,222 −7642 −2940 −13,295 −5545 −16,126 −6792 −2915 −19,184 −8369 −3728 −71,123 −34,457 −12,915 −8321 −12,710 −4058
0.9885 0.9824 0.9801 0.9801 0.9883 0.9840 0.9878 0.9839 0.9877 0.9883 0.9906 0.9852 0.9959 0.9805 0.9880 0.9911 0.9863 0.9826 0.9945 0.9785 0.9931 0.9936 0.9807 0.9943 0.9856 0.9922
2.16E + 07 7.49 2.08 0.16 491.07 8.58E + 04 1.41 1.64E + 04 0.27 5.01E + 06 23.55 0.05 164.02 0.01 2.81E + 04 0.26 1.60E-03 3.48E + 06 5.58 0.01 3.94E + 35 3.74E + 15 7996.35 1.22E + 04 1.43E + 07 55.35
156.66 55.85 46.54 29.95 118.82 126.88 53.28 126.84 53.19 143.19 63.54 24.44 110.53 46.10 134.07 56.47 24.24 159.49 69.58 31.00 591.32 286.47 107.38 69.18 105.67 33.74
1.82 1.89 0.11 0.49 0.09 1.18
0.49 0.05
0.14
0.28
28.70
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
-1
Rate of volatile products released (wt%·s )
232
CH4 C2 H 6 CO2 CO
-3
3.0x10
-3
2.0x10
-3
1.0x10
0.0 -6
H2 C2H 4 C6H 6
1.8x10
-6
1.2x10
-7
6.0x10
0.0 C7H8 C8H10 C10H8
-5
3.0x10
-5
2.0x10
-5
1.0x10
0.0 0.00
0.08
0.16
z (m) 0.24
0.32
0.40
Fig. 4. Released rate of the main volatile products versus axial bed height. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
important role is still unknown. In order to get a deeper understanding on the heating and pyrolysis behavior of the large coal particle, different mechanisms were calculated and compared separately by ignoring the corresponding single heat transfer mechanism. Then, the important order of the heat transfer mechanisms can be given as: radiation heat transfer N convection heat transfer between coal and pyrolysis gas N endothermic effect. The previous literatures [6,10] have reported the experimental study on coal pyrolysis by the circulating ash heat carrier. The results indicate that the maximum volatile yield can be achieved when the pyrolysis time is ~4 min. However, the modeling data show that residence time of ~2 min is adequate as the velocity of the moving bed is determined to be 0.125 m/min. Considering the ideal assumption of the adiabatic system, perfect mixing of the ash and coal, spherical coal and ash
-1
Rate of total volatile released (wt%·s )
0.9 0.6 0.3 0.0 2.8 2.1 1.4 0.7
particle, etc. in the present model, it can be inferred that the 3–4 min design residence time of the moving bed reactor in the poly-generation technology is reasonable and bounteous. 4.4. Model calculational results of different coal species The model calculational results of different coal species such as the Indian brown coal [28], the Chinese brown coal [29], Datong bituminous coal [30] and Shenmu bituminous coal are compared in Fig. 6. The result indicates that, the evolution characteristics of different coal samples vary obviously as a result of their respective coal properties. Compared with the Chinese and Indian brown coal samples, the theoretical bed heights of the Datong and Shenmu bituminous coal tend to be higher, while the maximum total devolatilization rates are relatively small.
(a) Velocity of moving bed us=0.10m/min us=0.13m/min us=0.15m/min us=0.20m/min
(b) Blending ratio of ash to coal ash/coal=6.0 ash/coal=5.0 ash/coal=4.0 ash/coal=3.0
(c) Coal particle diameter 5mm 8mm 10mm 12mm
(d) Ash particle diameter 0.57mm 0.67mm 0.87mm 0.97mm
(e) Initial temperature of ash 1200K 1100K 1000K 900K
(f) Preheating temperature of coal 298K 348K 398K 448K
0.0 0.9 0.6 0.3 0.0 0.0
0.1
0.2
z (m)
0.3
0.4
0.0
0.1
0.2
0.3
0.4
z (m)
Fig. 5. Effect of operation parameters on the total devolatilization rate. Calculation parameters: β = 0.8; χ = 0.5.
-1
Rate of total volatile released (wt%·s )
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
Rac rpc sac
Coal samples 4
Indian brown coal Chinese brown coal Datong bituminous coal Shenmu bituminous coal
3
2
1
sa sc Ta Tc Tg us w w⁎ wj w⁎j z
0 0.00
0.04
0.08
0.12
0.16
0.20
233
ratio of contact area to coal surface area radius of the particle, m contact surface area of ash and coal in dz. volume thickness pyrolyzer, m2 ash surface area in dz. volume thickness pyrolyzer, m2 coal surface area in dz. volume thickness pyrolyzer, m2 temperature of the ash, K temperature of the external layer of coal, K temperature of the pyrolysis gas, K velocity of the moving bed, m/s mass conversion fraction of coal evolved up to time t, % the value of w at t = ∞, % mass fractional of gas j product evolved up to time t, % the value of wj at t = ∞, % bed height, m
0.24
z (m) Fig. 6. Model calculational results of different coal species. Calculation parameters: dpc = 10 mm; Ta0 = 1100 K; us = 0.125 m/min; θ = 5.0; β = 0.8; χ = 0.5.
This is mainly because the coal rank of bituminous coal is higher than the brown coal. The larger activation energy of the bituminous coal has limited the devolatilization reaction. It can be pointed out that, combined with the kinetic study of coal sample, the model can achieve the adaptability of different coal species. 5. Conclusion In this study, a mathematical pyrolysis model of non-isothermal coal particles by circulating ash carrier toward the axial dimension of the moving bed is developed. Based on the results of Coats–Redfern integral method and TG-MS experiment, the model can predict not only the temperature distribution of pyrolysis gas-coal-ash toward the pyrolyzer height, but also the evolution of total volatile and the main volatile products. For the large coal particle pyrolysis, heat transfer is the ratedetermining step. The devolatilization process has a transient delay compared with the heating process. The evolution of the main volatile products is concentrated at the bed height of 0.08–0.24 m. The calculating result indicates that the 3–4 min design residence time of the moving bed reactor in the poly-generation technology is reasonable and bounteous. Increase the blending ratio of ash to coal, preheating temperature of coal and initial temperature of ash can accelerate the devolatilization process. Combined with the kinetic study of the coal rank and coal species, the model can be applied to different coal. The present model may provide theoretical prediction data for the scaleup of the coal pyrolysis/CFB combustion poly-generation technology as well as a deeper understanding on the evolution mechanism of the coal volatile products. Nomenclature A cross area of the moving-bed pyrolyzer, m2 aa specific surface area of ash particle, m2/m3 ac specific surface area of coal particle, m2/m3 cpa heat capacity of ash, J/(kg K) cpc heat capacity of the coal, J/(kg K) cpg heat capacity of pyrolysis gas, J/(kg K) dac contact area diameter of ash and coal, m dpa diameter of the ash particle, m dpc diameter of the coal particle, m Ej activation energy of reaction forming product j, kJ/mol hag heat transfer coefficient of ash and gas, W/(m2 K) hcg heat transfer coefficient of coal and gas, W/(m2 K) k0j pre-exponential factor of reaction forming product j, min−1 n reaction order R ideal gas constant, J/(mol K)
Greek symbols α β ε ΔH λa λc μg θ ρc ρc0 σ
heating rate, K/min emissivity of coal particle the void ratio reaction heat, kJ/kg thermal conductivity of ash, W/(m K) thermal conductivity of coal, W/(m K) viscosity of the gas, Pa·s volume ratio of ash to coal density of the coal, kg/m3 density of the raw coal, kg/m3 Stefan–Boltzman constant, (W/m2 K4)
Acknowledgement The work was supported by the National Natural Science Foundation of China (Grant No. 21376142). References [1] Z. Li, Z. Guo, X. Gong, H. Tang, Kinetic characteristics of pulverized coal combustion in the two-phase flow, Energy 55 (2013) 585–593. [2] M. Eriksson, M.R. Golriz, Radiation heat transfer in circulating fluidized bed combustors, Int. J. Therm. Sci. 44 (2005) 399–409. [3] X. Qu, P. Liang, R. Zhang, Z.X. Gan, J.C. Bi, Sulfur transformation in the process of circulating fluidized bed combustion combined with coal pyrolysis, Energy Fuel 24 (2010) 5023–5027. [4] Y.J. He, F.Z. Xiao, Z.H. Luo, Numerical modeling of the cavity phenomenon and its elimination way in rectangular radial moving bed reactor, Powder Technol. 274 (2015) 28–36. [5] X. Qu, P. Liang, Z.F. Wang, R. Zhang, D.K. Sun, X.K. Gong, Z.X. Gan, J.C. Bi, Pilot development of a polygeneration process of circulating fluidized bed combustion combined with coal pyrolysis, Chem. Eng. Technol. 34 (2011) 61–68. [6] P. Liang, Z.F. Wang, J.C. Bi, Simulation of coal pyrolysis by solid heat carrier in a moving-bed pyrolyzer, Fuel 87 (2008) 435–442. [7] D.G. Lai, Z.H. Chen, Y. Shi, L.X. Lin, J.H. Zhan, S.Q. Gao, G.W. Xu, Pyrolysis of oil shale by solid heat carrier in an innovative moving bed with internals, Fuel 159 (2015) 943–951. [8] G.X. Hu, X.W. Gong, H. Hao, H.J. Fan, Z. Wang, Experimental studies on flow and pyrolysis of coal with solid heat carrier in a modified rotating cone reactor, Chem. Eng. Process. Process Intensif. 47 (2008) 1777–1785. [9] G.X. Hu, H.J. Fan, Y.Q. Liu, Experimental studies on pyrolysis of Datong coal with solid heat carrier in a fixed bed, Fuel Process. Technol. 69 (2001) 221–228. [10] P. Liang, Z.F. Wang, J.C. Bi, Process characteristics investigation of simulated circulating fluidized bed combustion combined with coal pyrolysis, Fuel Process. Technol. 88 (2007) 23–28. [11] R. Xiong, L. Dong, J. Yu, X.F. Zhang, L. Jin, G.W. Xu, Fundamentals of coal topping gasification: characterization of pyrolysis topping in a fluidized bed reactor, Fuel Process. Technol. 91 (2010) 810–817. [12] X.L. Liu, G. Wang, G. Pan, Z. Wen, Numerical analysis of heat transfer and volatile evolution of coal particle, Fuel 106 (2013) 667–673. [13] C.A. Heidenreich, H.M. Yan, D.K. Zhang, Mathematical modelling of pyrolysis of large coal particles—estimation of kinetic parameters for methane evolution, Fuel 78 (1999) 557–566. [14] P. Liang, Y.Q. Zhang, W.M. Jiang, A.F. Wei, T. Liu, J.F. Wu, Simulation study of Shenmu coal pyrolysis by gas heat carrier based on a moving bed, Energy Fuel 29 (2015) 7727–7733.
234
Y.-Q. Zhang et al. / Fuel Processing Technology 154 (2016) 227–234
[15] F.J. Higuera, Numerical simulation of the devolatilization of a moving coal particle, Combust. Flame 156 (2009) 1023–1034. [16] S.F. Zhang, F. Zhu, C.G. Bai, L.Y. Wen, C. Zou, Thermal behavior and kinetics of the pyrolysis of the coal used in the COREX process, J. Anal. Appl. Pyrolysis 104 (2013) 660–666. [17] Q. Dong, Y. Xiong, Kinetics study on conventional and microwave pyrolysis of moso bamboo, Bioresour. Technol. 171 (2014) 127–131. [18] S. Ceylan, Y. Topcu, Z. Ceylan, Thermal behaviour and kinetics of alga Polysiphonia elongata biomass during pyrolysis, Bioresour. Technol. 171 (2014) 193–198. [19] A.V. Luikov, A.G. Shashkov, L.L. Vasilev, Y.E. Fraiman, Thermal conductivity of porous systems, Int. J. Heat Mass Transf. 11 (1968) 117–140. [20] B.A. Adesanya, H.N. Pham, Mathematical modelling of devolatilization of large coal particles in a convective environment, Fuel 74 (1995) 896–902. [21] D.J. Gunn, Transfer of heat or mass to particle in fixed and fluidized beds, Int. J. Heat Mass Transf. 21 (1978) 467–476. [22] S.Y. Wang, G.D. Liu, Y.B. Wu, J.H. Chen, Y.J. Liu, L.X. Wei, Numerical investigation of gas-to-particle cluster convective heat transfer in circulating fluidized beds, Int. J. Heat Mass Transf. 53 (2010) 3102–3110. [23] S. Badzioch, P.G.W. Hawksley, Kinetics of thermal decomposition of pulverized coal particles, Ind. Eng. Chem. Process. Des. Dev. 9 (1970) 521–530.
[24] B.H. Yan, Y. Cheng, Y. Jin, C.Y. Guo, Analysis of particle heating and devolatilization during rapid pyrolysis in a thermal plasma reactor, Fuel Process. Technol. 100 (2012) 1–10. [25] C. Lievens, D.H. Ci, Y. Bai, L.G. Ma, R. Zhang, J.Y. Chen, Q.Q. Gai, Y.H. Long, X.F. Guo, A study of slow pyrolysis of one low rank coal via pyrolysis–GC/MS, Fuel Process. Technol. 116 (2013) 85–93. [26] X.C. Lin, C.H. Wang, K. Ideta, J. Miyawaki, Y. Nishiyama, Y.G. Wang, S. Yoon, I. Mochida, Insights into the functional group transformation of a Chinese brown coal during slow pyrolysis by combining various experiments, Fuel 118 (2014) 257–264. [27] M. Wang, Z.S. Li, W.B. Huang, J.X. Yang, H.T. Xue, Coal pyrolysis characteristics by TG–MS and its late gas generation potential, Fuel 156 (2015) 243–253. [28] V.R. Patel, R.N. Patel, V.J. Rao, Kinetic parameter estimation of lignite by thermogravimetric analysis, Proc. Eng. 51 (2013) 727–734. [29] Y. Xu, Y.F. Zhang, Y. Wang, G.J. Zhang, L. Chen, Thermogravimetric study of the kinetics and characteristics of the pyrolysis of lignite, React. Kinet. Mech. Catal. 110 (2013) 225–235. [30] S.F. Zhang, F. Zhu, C.G. Bai, L.Y. Wen, C. Zou, Thermal behavior and kinetics of the pyrolysis of the coal used in the COREX process, J. Anal. Appl. Pyrolysis 104 (2013) 660–666.