torrefied wood and coal blends

Applied Energy 105 (2013) 57–65

Contents lists available at SciVerse ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Thermogravimetric analysis and kinetics of co-pyrolysis of raw/torrefied wood and coal blends Ke-Miao Lu a, Wen-Jhy Lee a, Wei-Hsin Chen b,⇑, Ta-Chang Lin a a b

Department of Environmental Engineering, National Cheng Kung University, Tainan 701, Taiwan, ROC Department of Greenergy, National University of Tainan, Tainan 700, Taiwan, ROC

h i g h l i g h t s " Raw Cryptomeria japonica is torrefied at 250 and 300 °C for 1 h. " Raw/torrefied Cryptomeria japonica are blended with an anthracite coal. " A thermogravimetric analyzer is used to examine the co-pyrolysis characteristics. " Five different biomass blending ratios of 100, 75, 50, 25, and 0 wt.% are considered. " Interactions or synergistic effects between raw/torrefied biomass and coal are slight.

a r t i c l e

i n f o

Article history: Received 20 August 2012 Received in revised form 10 December 2012 Accepted 19 December 2012 Available online 20 January 2013 Keywords: Torrefaction Blends Thermogravimetric analysis (TGA) Co-pyrolysis Kinetics Interaction or synergistic effect

a b s t r a c t The properties of biomass can be improved via torrefaction, and torrefied wood is a fuel with the potential to partially replace coal. In this study, raw Cryptomeria japonica (WRaw) is torrefied at 250 (TW250) and 300 °C (TW300) for 1 h, and then mixed with an anthracite coal to undergo co-pyrolysis. A thermogravimetric analyzer is used to examine the co-pyrolysis characteristics of fuel blends and five different biomass blending ratios (BBRs) of 100, 75, 50, 25, and 0 wt.% are taken into consideration. When WRaw, TW250, and the coal are tested, the pyrolysis is characterized by a three-stage reaction, whereas four-stage thermal degradation is found for TW300 and fuel blends. The predictions from the linear superposition of the thermal decomposition of individual fuels fit the experimental data of the fuel blends, suggesting that the interaction or synergistic effect of co-pyrolysis between the raw/torrefied C. japonica and the coal is slight. The co-pyrolysis kinetics of the fuel blends is also analyzed. The variation of chemical kinetics with decreasing BBR in the second stage is different from that in the third stage. That is, an increase in BBR leads to an increase in the activation energy in the second stage, whereas it causes a decrease in the third stage. This is attributed to that the reactivities of cellulose and lignin in biomass are different from that of coal in the two stages. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Economic growth increases the demand for energy. Fossil fuels are currently the most important energy sources [1] and they are mainly consumed via combustion. However, the burning of fossil fuels releases a large amount of carbon dioxide into the atmosphere so as to raising environmental concerns. Over the past several years, a number of countries have considered making greater use of nuclear energy. But enthusiasm for this has fallen in the wake of the accident at Fukushima, Japan, in 2011. Biomass has a much shorter life cycle compared to fossil fuels, and it has the ability to mitigate the greenhouse effect, due to its carbon neutral nature. Therefore, if biomass is used as an alternative fuel it cannot ⇑ Corresponding author. Tel.: +886 6 2605031; fax: +886 6 2602205. E-mail address: [email protected] (W.-H. Chen). 0306-2619/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.apenergy.2012.12.050

only reduce the consumption of fossil fuels, but also reduce the environmental pollution and greenhouse effect [2,3]. Among the various ways of using bioenergy, one that has attracted particular attention is the use of biomass blended with coal, as such blends can be applied for co-firing [3–5], co-gasification [6–8], and co-pyrolysis [9–21]. Co-firing biomass and coal blends is able to reduce NOx and SOx levels in flue gases [4], and can also reduce emissions of greenhouse gases at lower capital and operating costs [5]. With regard to co-gasification, it has been pointed out [6] that increasing the biomass ratio in the mixture of residual biomass and low-grade coal in a fluidized-bed reactor had the potential to intensify the yield of product gas and its heating value. Seo et al. [8] indicated that an increase in the biomass ratio led to increases in H2 and CH4 in the product gas, as well as greater total carbon conversion and cold gas efficiency. McLendon et al. reported [7] that the transport (rheological) properties of coal and

58

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65

Nomenclature A B Ea k n R T

pre-exponential factor (min1) heating rate (K min1) activation energy (kJ mol1) rate constant (min1) order of reaction (–) universal gas constant (=8.314 J mol1 K1) temperature (K)

t X W

heating time (min) conversion of sample (–) weight of sample (mg)

Subscript i initial state (at 105 °C) f final state (at 800 °C)

biomass blends were significantly better than those of coal alone consumed in a fluidized bed gasifier, while the plugging behavior was also greatly reduced and fuel handling was easier. In regard to co-pyrolysis, Moghtaderi [10] showed that blends of coal and biomass had lower pyrolysis temperatures than coal, while Meesri and Moghtaderi [13,15] and Zhang et al. [18] noted that the gas and liquid yields from co-pyrolysis increased along with the biomass ratio. Seo et al. [8] found that after undergoing pyrolysis the micropore ratio in a biomass and coal blend increased, while the average pore size of the blend decreased. Because the reaction took place on the active carbon sites in the micropores, the enhanced micropore structure was conducive to gasification reactivity. Despite the numerous advantages from the reactions of biomass and coal blends, as mentioned above, the energy density of raw biomass is relatively low compared to that of coal, due to its larger volume (or lower density), higher moisture content, and lower heating value. Raw biomass also has other poor characteristics such as hygroscopic behavior, low grindability, and non-uniform properties due to the diversity of biomass sources [22,23]. These issues make present problems with regard to the storage, transport, and utilization of raw biomass, and thus a number of studies have attempted to overcome these issues via torrefaction. Torrefaction is a thermal pretreatment process which is carried out at temperatures of 200–300 °C under an inert or nitrogen atmosphere for several minutes to several hours [24–26]. A number of review papers [27–29] have summarized the advantages of using torrefaction to improve the characteristics of biomass with regard to its use as a fuel. In brief, torrefaction is able to decrease the atomic O/C and H/C ratios, reduce the moisture level, and intensify the energy density of biomass. It transforms biomass from a hygroscopic material to a hydrophobic one, as well as improving the grindability and homogeneity. These changes mean that biomass can be used in wider range of applications. Considering the co-pyrolysis of biomass and coal, while some studies [9,12–16,19,21] report that no interactions occur in the fuel blend reactions, others [8,18,20] present the opposite results. In reviewing recent studies concerning the reactions of biomass and coal blends, it is noted that no research has been carried out that examine the co-pyrolysis of torrefied biomass and coal blends, even though numerous studies have been performed on biomass torrefaction. For this reason, the present study aims to investigate the thermal degradation characteristics and pyrolysis kinetics of raw/torrefied wood (Cryptomeria japonica) and coal blends. The raw or torrefied wood is mixed with a coal at various weight ratios. The mixtures are pyrolyzed in a non-isothermal environment, and thermogravimetric analysis is carried out. The co-pyrolysis and kinetics of the fuel blends are then explained in detail.

in the present study. The carbon content in biomass is lower than that in coal, and anthracite coal has high carbon content among the various coals that are available. When biomass and anthracite coal are blended, the carbon content of the blend is closer to that of bituminous or subbituminous coal. Because of this, an anthracite coal from Australia was adopted for this work’s fuel blend tests. Prior to performing torrefaction, the reaction system was leak tested using nitrogen, in order to ensure that no oxygen was entrained in the reaction system when torrefaction was carried out. The torrefied wood was made in several steps. To begin with, raw C. japonica wood was cut into small chips with sizes of about 30 mm  30 mm  40 mm. The wood chips were then dried in an oven at 105 °C for 24 h to provide a basis for experiments. Finally, 9.5 g (±5%) of wood chips was placed in a reactor in each experiment followed by being treated via either mild torrefaction at 250 °C or severe torrefaction at 300 °C [31] for 1 h. While the samples were torrefied, nitrogen at the flow rate of 100 mL min1 (25 °C) was continuously blown into the reactor to keep the reaction tube in an inert atmosphere.

2. Methodology

2.3. Pyrolysis kinetics

2.1. Preparation of samples

The non-isothermal kinetics for solid decomposition is usually written as follows [32]:

C. japonica is abundant in Taiwan, accounting for around 58.5% of the forest therein, and is characterized by rapid growth and good wood quality [30]. It was thus chosen as the experiment material

2.2. Thermogravimetric analysis (TGA) The pyrolysis characteristics of the samples were examined using a thermogravimetric analyzer (TG, PerkinElmer Diamond TG/DTA). A crucible loaded with sample was placed inside the TG where the weight of the sample was constantly measured. The functions of the TG were to measure and record the dynamics of sample weight loss with increasing temperature or time. For each experimental run, around 5 mg of sample at particle sizes of 100– 200 mesh was used. When the pyrolysis of the samples was carried out, the heating temperature in the TG was in the range of 25– 800 °C, and the heating rate was controlled at 20 °C min1. Once the heating temperature reached 105 °C, it was held for 10 min to completely remove moisture and provide a basis for analysis. The temperature in the TG was detected and recorded at a frequency of 2 Hz. Nitrogen was used as the carrier gas in the TG, so that the sample was pyrolyzed in an inert environment without oxygen. The flow rate of the carrier gas was fixed at 100 cc min1 (STP). Based on the distributions of weight loss, it was possible to carry out the thermogravimetric analyses (TGA) and derivative thermogravimetric (DTG) analyses of the samples. To ensure the measured quality of the TGA, the TG was periodically calibrated using indium (In), tin (Sn) and calcium oxalate (CaC2O4), and the experiment under any given condition was usually carried out more than twice. The relative error among the measurements of TGA was controlled to less than 5%.

dX ¼ kf ðXÞ dt

ð1Þ

59

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65 Table 1 Properties of raw/torrefied wood and coal. Material Photograph

WRaw

TW250

TW300

Coal

HHV (MJ/kg)

20.29

22.55

26.60

27.34

Elemental analysis (wt.%, dry-ash-free) C 51.73 H 5.30 O (by diff.) 42.80 N 0.17 S 0.00 H/C (atomic ratio) 1.23 O/C (atomic ratio) 0.62

58.16 5.19 36.65 0.00 0.00 1.07 0.47

73.30 4.35 22.35 0.00 0.00 0.71 0.23

86.56 4.93 6.20 1.70 0.61 0.68 0.05

Proximate analysis (wt.%, dry basis) Volatile matter (VM) Fixed carbon (FC) Ash

80.82 19.18 0.00

72.59 27.12 0.29

42.33 56.89 0.78

31.71 54.58 13.71

Fiber analysis (wt.%) Hemicellulose Cellulose Lignin Other

16.01 43.60 32.20 8.19

7.65 46.86 42.11 3.38

5.03 24.80 69.36 0.81

where X is fuel conversion, and is given by

X¼

or

Wi  W Wi  Wf

ð2Þ

In the above equation, W, Wi, and Wf represent the instantaneous, initial (at 105 °C), and final (at 800 °C) weights of the sample. The reaction rate constant k is expressed in terms of the Arrhenius equation as

k ¼ A exp



Ea RT

 ð3Þ

n

ð4Þ

Substituting Eqs. (3) and (4) into Eq. (1) gives

  dX Ea ¼ A exp  ð1  XÞn dt RT

ð5Þ

For a constant heating rate b = dT/dt, Eq. (5) can be rearranged to the following equation

    dX 1 Ea A exp  ð1  XÞn ¼ dT b RT

ð6Þ

The integral method based on Coats and Redfern (CR) equation [33–35] is used in this work, and the approximate integration of Eq. (6) gives

ln

   lnð1  XÞ T

2

¼ ln

  AR 2RT Ea  1 bEa E RT

if n ¼ 1

ð7Þ

or

ln

" #  lnð1  XÞ1n ð1  nÞ  T 2

¼ ln

  AR 2RT Ea  1 bEa E RT

if n–1

ð8Þ

For most reactions, the value of 2RT/E is very small (i.e. 2RT/ Ea<<1), and thus the preceding two equations can be approximated by

ln

   lnð1  XÞ T

2

¼ ln

AR Ea  bEa RT

if n ¼ 1

ln

1  ð1  XÞ1n ð1  nÞ  T 2

# ¼ ln

AR Ea  bEa RT

if n–1

ð10Þ

The plot of ln[ln(1  X)/T2] versus 1/T becomes a linear line for n = 1; the plot of ln[1(1  X)1n/T2] versus 1/T is also a linear line for n – 1. Accordingly, the apparent activation energy (Ea) and the apparent frequency factor (A) can be determined from the slope and intercept of the regression line, respectively. 3. Results and discussion

and the function f(X) can be written as

f ðXÞ ¼ ð1  XÞ

"

ð9Þ

Five different biomass blending ratios (BBRs) of raw/torrefied biomass and the anthracite coal are taken into consideration; they are 100, 75, 50, 25, and 0 wt.%. When BBR is equal to 100%, it means that biomass alone is tested, while only coal is tested when BBR is 0%. In the following discussion, the untreated (raw) C. japonica wood and torrefied samples at 250 and 300 °C are denoted as WRaw, TW250, and TW300, respectively. 3.1. Properties of C. japonica and coal The properties of WRaw, TW250, TW300, and coal, as drawn from the elemental and proximate analyses, are presented in Table 1. The results of fiber analysis of WRaw, TW250, and TW300 are also given in the table. The hemicellulose, cellulose, and lignin contents in WRaw are 16.01, 43.60, and 32.20 wt.%, respectively. When the wood is torrefied at 250 °C, significant amounts of hemicellulose are consumed, and thus the of cellulose and lignin contents in TW250 increase. For the wood torrefied at 300 °C, cellulose is also thermally degraded. As a consequence, the lignin content in TW300 increases markedly. The coal used in this work has a significantly higher carbon content and lower oxygen content compared to both the raw and torrefied biomass. After undergoing torrefaction, the carbon content and higher heating value (HHV) of the wood are higher than those of raw biomass. For example, the calorific values of the biomass torrefied at 250 and 300 °C (i.e. TW250 and TW300) are increased by 11% and 31%, respectively. It should be noted that the HHV of TW300 (=26.60 MJ kg1) is close to that of

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65

2

nd

1.2

rd

stage

3 stage 1

100

TGA (wt%)

TGA DTG

80

0.8

o

368 C o 1.0181 wt% / C

60

0.6 0.4

40

0.2

20 0

0 100

200

300

400

500

600

700

800

o

Temperature ( C)

(b) 120

2 nd

3 rd

stage

stage

st

1 stage

1.2

th

4 stage

100

1 o

80

364 C o 0.6163 wt% / C

0.8 0.6

60

442 oC 0.1468 wt% / oC

20 0

o

0.4

40

DTG (wt% / C)

The TGA and DTG curves of WRaw and coal are shown in Fig. 1. As a whole, the pyrolysis processes of the two fuels are characterized by a three-stage thermal degradation, and the decomposition of the fuels mainly takes place in the second stage. The secondstage reactions of WRaw and coal approximately develop at 230– 410 and 420–520 °C, respectively. The peak temperature of coal (=465 °C) is much higher than that of WRaw (=367 °C), but the decomposition intensity of the former (=0.17 wt.% °C1) is substantially lower than that of the latter (=0.93 wt.% °C1). This is due to differences in the fuel structures of coal and biomass. Specifically, the peak exhibited in Fig. 1a results from the thermal degradation of cellulose, and it is due to the release of volatile matter from the coal, as shown in Fig. 1b. In addition, the coal has a high carbon content (=87 wt.%), and thus does not decompose quickly in the course of the pyrolysis reaction. With regard to TW250, Fig. 2a shows that its peak temperature (=368 °C) in the DTG curve is very close to that of WRaw (=367 °C). Due to the depletion of a large portion of hemicellulose (Table 1), which causes relatively more cellulose to be contain in TW250, the peak intensity of TW250 (=1.02 wt.% °C1) is even higher

st

1 stage

o

3.2. Pyrolysis characteristics of materials

(a) 120

DTG (wt% / C)

coal (=27.34 MJ kg1). The results of the proximate analysis show that the volatile matter in raw C. japonica is 80.82 wt.%, which is more than two times higher than that of coal (=31.71 wt.%). However, the ash contents in WRaw, TW250, and TW300 are 0, 0.29, and 0.78 wt.%, respectively, while 13.71 wt.% in the coal. This implies that using raw or torrefied C. japonica as an alternative to coal can significantly reduce the costs associated with ash disposal.

TGA (wt%)

60

0.2 0

100

200

300

400

500

600

700

800

Temperature (oC)

(a)

120 st

1 stage

2

nd

1.2

rd

stage

100

1 0.8

o

367 C o 0.9345 wt% / C

60

0.6 0.4

40

o

TGA (wt%)

80

DTG (wt% / C)

TGA DTG

0.2

20 0

Fig. 2. Pyrolysis distributions of the TGA and DTG curves of (a) TW250 and (b) TW300.

3 stage

0 100

200

300

400

500

600

700

800

o

Temperature ( C)

(b)

120

2 nd

st

1 stage

1.2

rd

3 stage

stage

1

80

0.8

60

0.6 465 C o 0.1692 wt% / C

0.2

20 0

0.4

o

40

o

DTG (wt% / C)

TGA (wt%)

100

0 100

200

300

400

500

600

700

800

Temperature (oC) Fig. 1. Pyrolysis distributions of the TGA and DTG curves of (a) WRaw and (b) coal.

than that of WRaw (=0.93 wt.% °C1). However, when the wood is torrefied at 300 °C for 1 h, the peak is reduced to 0.62 wt.% °C1. This is consistent with the results of the fiber analysis (Table 1), where cellulose is also consumed to a certain extent. Chen and Kuo [36] reported that hemicellulose and cellulose in biomass were significantly decomposed through torrefaction at 290 °C. For TW300, in addition to the thermal degradation of cellulose at 364 °C, a mild decomposition process also develops at 442 °C (Fig. 2b). This is attributed to the reaction of lignin, in that relatively more lignin is retained in the wood from torrefaction at 300 °C. Instead of the three-stage thermal degradation seen with WRaw, TW250, and coal, four-stage thermal decomposition is found with TW300. According to the distributions of TGA curves shown in Figs. 1 and 2 , the solid residues of WRaw, TW250, and TW300 at 800 °C are 15, 21, and 40 wt.%, respectively, while they are 70 wt.% for coal. The higher the torrefaction temperature, the greater the amount of solid residue in the biomass. 3.3. Pyrolysis characteristics of blends The TGA and DTG curves of raw and torrefied C. japonica blended with coal at the BBRs of 75%, 50%, and 25% are shown in Figs. 3–5. When WRaw is mixed with coal, the pyrolysis is characterized by a four-stage reaction rather than the three-stage one observed in Figs. 1 and 2b. The first peaks, which occur in the secondstage reaction between 368 and 372 °C (Fig. 3), are due to the thermal decomposition of WRaw in the mixtures. The second peaks in the third-stage reaction between 464 and 468 °C stem from coal decomposition. In other words, in the DTG curves, the peaks exhibited come from a combination of the thermal degradation of

61

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65

(a) 120

st

2

1 stage

nd

1

3 rd stage stage

th

4 stage

80

TGA DTG

o

367 C 0.7803 wt% / oC

0.6

60

0.4

40 454 oC 0.0918 wt% / oC

20 0

200

300

400

500 o

600

700

Temperature ( C)

(b) 120

nd

stage

1

3 rd

th

4 stage

stage

st

1 stage

2

nd

1

3 rd stage stage

th

4 stage

80 370 oC

0.8 0.6

0.7168 wt% / oC

60

0.4 467 oC 0.0760 wt% / oC

20 0

0.2

200

300

400

500

600

700

(c) 120

800

nd

stage

1

3 rd

th

4 stage

stage

368 C o 0.4846 wt% / C

0.6

60

0.4

o

40

468 oC 0.1116 wt% / oC

0.2

200

300

400

500

600

700

200

300

400

500

600

700

800

2

1 st stage

nd

stage

3

1

rd

stage

4 th stage

0.8

80 60

0.6

367 oC o 0.2761 wt% / C

0.4

40

o

464 C 0.1491 wt% / oC

0.2

20 0

0 100

200

300

400

500

600

700

800

o

0 100

0 100

Temperature ( C)

20 0

0.2

o

80

0.8 o

DTG (wt% / C)

TGA (wt%)

100

470 oC 0.1156 wt% / oC

100

TGA (wt%)

2

0.4

369 oC o 0.5331wt% / C

DTG (wt% / C)

st

40

o

o

1 stage

60

0

Temperature ( C)

(b) 120

0.6

Temperature ( C)

0 100

80

20

o

40

0.8

o

TGA DTG

DTG (wt% / C)

TGA (wt%)

100

TGA (wt%)

2

800

DTG (wt% / C)

st

1 stage

0.2 0

100

100

(a) 120

0.8

o

TGA (wt%)

100

DTG (wt% / C)

biomass and coal in the mixtures. The peak intensity in the thirdstage reaction is lower than that of the second-stage one. Meanwhile, decreasing BBR increases the peak intensity of the thirdstage reaction. The TGA curves shown in Fig. 3 also suggest that when the BBR is higher, the weight of solid residue at the end of pyrolysis reaction (i.e. 800 °C) is lower. For instance, the solid residues at BBRs of 25, 50, and 75 wt.% are 56.3, 42.7, and 29.2 wt.%, respectively. The pyrolysis characteristics of the mixtures of TW250 and coal (Fig. 4) are similar to those of the mixtures of WRaw and coal. TW300 experiences severe torrefaction and more hemicellulose and cellulose are depleted (Figs. 1 and 2). Accordingly, the first and the second peaks shown in the DTG curves wither in a significant way (Fig. 5). The pyrolysis peaks of coal and lignin are obtained at approximately 442 and 465 °C (Figs. 1b and 2b), and these temperatures are close to each other. Therefore, the peaks exhibited in the third stage are the consequence of the overlap of the decomposition of coal and lignin.

Fig. 4. Co-pyrolysis distributions of the TGA and DTG curves of TW250 and coal at BBRs of (a) 75, (b) 50, and (c) 25 wt.%.

800

o

Temperature ( C) 3.4. Interaction of biomass and coal

(c) 120

1 st stage

1

rd

3 2 nd stage stage

4 th stage

0.8

80 60

0.6 372 oC 0.2731 wt% / oC

0.4

o

40

DTG (wt% / C)

TGA (wt%)

100

464 oC 0.1276 wt% / oC

0.2

20 0

0 100

200

300

400

500

600

700

800

o

Temperature ( C) Fig. 3. Co-pyrolysis distributions of the TGA and DTG curves of WRaw and coal at BBRs of (a) 75, (b) 50, and (c) 25 wt.%.

If there are no interactions in the thermal decomposition of the biomass and coal, the pyrolysis characteristics of the blends will follow the behaviors of their parent materials in an additive manner. To evaluate the interaction between the raw/torrefied wood and the coal, the experimental and calculated TGA and DTG curves are plotted in Figs. 6 and 7, respectively. It can be seen that all the calculated curves almost overlap with the experimental ones. The maximum relative errors between the experimental and calculated TGA curves of various blends are further examined in Table 2, and a comparison of weight percentages of solid residue at 800 °C in the TGA curves is given in Table 3. The predicted results are obtained from the calculations in terms of the weight percentage of every single material. A comparison of the experimental and calculated TGA curves indicates that the relative error is within 5%, except for the case of TW250 at BBR of 50% (Table 2). With regard to the solid residue, its relative error is within 6.8% (Table 3). It is thus

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65

(a) 120

2 nd

st

1 stage

stage

1

3 rd

TGA DTG

365 oC 0.4656 wt% / oC

0.8 0.6

60

0.4 436 oC 0.1438 wt% / oC

o

40

80

DTG (wt% / C)

TGA (wt%)

100 80

(a) 100

th

4 stage

stage

0.2

TGA (wt%)

62

0 100

200

300

400

500

600

700

40 Exp. (75%WRaw+25%Coal) Exp. (50%WRaw+50%Coal) Exp. (25%WRaw+75%Coal) Cal. (75%WRaw+25%Coal) Cal. (50%WRaw+50%Coal) Cal. (25%WRaw+75%Coal)

20

20 0

60

0

800

o

100

200

Temperature ( C)

(b) 120

2 nd

st

1 stage

stage

1

3 rd

0.8 0.6

363 oC 0.1170 wt% / oC

0.4 461 oC 0.1427 wt% / oC

o

40

0.2

600

700

800

600

700

800

0 100

(c) 120

200

300

400

500

600

700

0

800

2

1 st stage

nd

stage

3

1

rd

0.8

300

400

500 o

367 oC 0.1694 wt% / oC

0.4

o

468 oC 0.1643 wt% / oC

0.2

20

TGA (wt%)

0.6

80

DTG (wt% / C)

80

40

200

(c) 100

4 th stage

stage

100

Temperature ( C)

100

TGA (wt%)

800

40

o

0

700

60

Temperature ( C)

60

600

20

20 0

500

80

TGA (wt%)

80

DTG (wt% / C)

TGA (wt%)

100

60

400

(b) 100

th

4 stage

stage

300

Temperature (oC)

60 40 20

0 100

200

300

400

500 o

600

700

800

Temperature ( C) Fig. 5. Co-pyrolysis distributions of the TGA and DTG curves of TW300 and coal at BBRs of (a) 75, (b) 50, and (c) 25 wt.%.

concluded that the synergistic effects or interactions between the raw/torrefied biomass and coal are very slight when they are copyrolyzed. The results also reveal that the pyrolysis behavior of the mixture can be predicted through linear superposition in terms of the weight percentage of every single material, regardless of whether raw or torrefied biomass is employed. 3.5. Pyrolysis kinetics Reaction temperature and time are two crucial factors influencing the pyrolysis reaction of fuels. The pre-exponential factors and activation energies of all the fuels, including the raw/torrefied biomass, coal, and their blends, in the second and the third stages are tabulated in Table 4, where the Arrhenius equation and the first order reaction are adopted. Meanwhile, the linear regression for the extraction of the kinetic parameters of the fuel blend of 50 wt.%

0

100

200

300

400

500 o

Temperature ( C) Fig. 6. Comparisons of experimental and calculated TGA curves of (a) WRaw and coal blends, (b) TW250 and coal blends, and (c) TW300 and coal blends.

TW300 +50 wt.% coal at the second and third stages are presented in Fig.8a and b, respectively. Whether in the second or third stage, the value of R2 is not less than 0.94, reflecting that the pyrolysis processes in these two stages are well correlated, as shown in Fig. 8. When the weight percentage of raw or torrefied C. japonica (or BBR) in a mixture decreases, the values of the pre-exponential factor and activation energy in the second stage have a decreasing trend. Because of the sharper peak exhibited in biomass pyrolysis compared to coal pyrolysis, the activation energy decreases as BBR goes down. The reactivity of biomass is more than that of coal, and this is the reason why the pre-exponential factor rises when BBR increases. On the other hand, the thermal decomposition in the third stage is dominated by lignin or coal. Unlike the second stage, the activation energy in the third stage is sensitive to BBR, and an increase in this decreases the activation energy, reflecting the fact that the activation energy of lignin is much less than that

63

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65

(a) 1.2 1

DTG (wt% / oC)

Table 2 A comparison of sample conversion between the experimental and calculated TGA curves.

Exp. (75%WRaw+25%Coal) Exp. (50%WRaw+50%Coal) Exp. (25%WRaw+75%Coal) Cal. (75%WRaw+25%Coal) Cal. (50%WRaw+50%Coal) Cal. (25%WRaw+75%Coal)

0.8 0.6 0.4 0.2 0

100

200

300

400

500

600

700

800

Temperature (oC)

(b)

1.2

DTG (wt% / oC)

1

a

BBR (wt.%)

75

50

25

WRaw Temperaturea (°C) Experiment Calculation Relative error (%)

372 56.0 54.0 3.6

370 70.4 67.6 3.9

366 83.6 82.1 1.8

TW250 Temperature (°C) Experiment Calculation Relative error (%)

250 63.7 61.1 4.1

354 83.7 69.1 17.4

345 91.8 91.5 0.3

TW300 Temperature (°C) Experiment Calculation Relative error (%)

361 85.3 79.9 6.3

800 56.6 53.8 5.0

778 63.4 63.3 0.1

Temperature location at the maximum relative error.

0.8 0.6

Table 3 A comparison of solid residue between the experimental and calculated TGA curves (at 800 °C).

0.4

BBR (wt.%)

0.2 0

100

200

300

400

500 o

600

700

800

Temperature ( C)

(c)

TW250 Experiment Calculation Relative error (%)

1.2 1

DTG (wt% / oC)

WRaw Experiment Calculation Relative error (%)

TW300 Experiment Calculation Relative error (%)

0.8

100

75

50

25

0

14.5

29.2 28.5 2.4

42.7 42.1 1.4

56.3 55.9 0.7

70.2

21.0

33.9 33.1 2.4

45.6 42.5 6.8

57.2 58.1 1.6

70.2

39.9

47.8 46.3 3.1

56.6 53.8 4.9

62.5 62.5 0.0

70.2

0.6 0.4 0.2 0

100

200

300

400

500 o

600

700

800

Temperature ( C) Fig. 7. Comparisons of experimental and calculated DTG curves of (a) WRaw and coal blends, (b) TW250 and coal blends, and (c) TW300 and coal blends.

of coal. In view of the marked increase in activation energy when BBR decreases, the pre-exponential factor is also raised to compensate for the significant reduction in chemical kinetics due to the growth of the exponential term. When the activation energies of WRaw, TW250, and TW300 in the second stage are compared with each other, TW250 at a given BBR generally has the highest value in that relatively more cellulose is contained the sample (Table 1). In the third stage, lignin plays an important role in determining the activation energy and the relative amount of lignin in TW300 is the highest. As a consequence, the activation energy is characterized by TW300 > TW250 > WRaw, as observed in Table 4.

4. Conclusions The pyrolysis characteristics of raw/torrefied C. japonica and an anthracite coal, as well as the co-pyrolysis of their blends, have

been examined in this study through thermogravimetric analyses. The results indicate that the thermal degradation processes of the single materials are characterized by a three-stage reaction, except for TW300, which demonstrates a four-stage one, resulting from relatively more lignin being retained in the torrefied biomass. When raw (WRaw), mildly torrefied (TW250), or severely torrefied (TW300) C. japonica is mixed with coal under various weight percentages, the pyrolysis process changes from a three-stage reaction to a four-stage one. The results of the analyses suggest that the pyrolysis characteristics of the mixtures of biomass and coal are very close to the combination of those of the individual materials. It is thus concluded that the interactions or synergistic effects between raw/torrefied biomass and coal are slight. In other words, the pyrolysis behavior of fuel blends can be determined in terms of the weight percentages of biomass and coal, regardless of whether biomass is torrefied or not. An examination of the copyrolysis kinetics of fuel blends indicates that an increase in BBR leads to an increase in the activation energy in the second stage, whereas it causes a decrease in the third stage. This arises from that fact that the co-pyrolysis kinetics in the second stage is mainly affected by cellulose in biomass, and it is influenced by lignin in the third one. Similarly, the effects of cellulose and lignin on co-pyrolysis kinetics lead to that the activation energy of TW250 at a given BBR is generally higher than those of WRaw and TW300 in the second stage, whereas the activation energy of TW300 in the third stage is higher than the other two samples.

64

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65

Table 4 Chemical kinetics of biomass and coal blends in the second and the third stages. BBR (wt.%)

TW250

WRaw Temperature (°C)

The second-stage reaction 100 230–410 75 240–410 50 240–410 25 270–410 The third-stage reaction 100 – 75 410–510 50 410–520 25 410–520 Coal 0

(a)

420–520

1

Ea (kJ mol1)

A (min

76.81 77.89 71.89 63.88

7.05  105 7.49  105 1.74  105 2.28  104

– 2.94 9.83 18.71 73.59

104.59 96.49 92.40 76.63

7.19  107 1.28  107 5.00  106 1.48  105

0.98 0.98 0.97 0.98

– 1.01  101 5.49  101 4.30  101

– 0.99 0.99 0.98

410–500 410–500 400–520 400–520

13.89 19.31 23.75 37.11

7.32  101 2.20  101 4.66  101 4.89  101

0.98 1.00 0.99 0.99

1.97  104

0.98

420–520

73.59

1.97  104

0.98

– 3.73  102 2.99  101 1.72  101

– 0.94 0.97 0.98

– 410–500 420–510 410–530

– 5.62 12.51 23.72

1.97  104

0.98

420–520

73.59

2

1/T

0.0016

-13

0.0017

Regression line Experimental data

-13.2

2

ln[G(X)/T ]

300–410 300–410 330–400 330–400

2.94  108 1.05  108 4.36  106 4.84  105

y = -11114.32 x + 3.11 R 2 = 0.97

-13.4 y = -2856.83 x - 9.41 2 R = 0.99

-13.6

0.0013

0.99 0.99 0.98 0.98

108.68 104.34 89.64 80.50

-14.5

-13.8 0.0012

R2

270–410 280–410 280–420 300–410

-15.5

(b)

A (min1)

1.00 1.00 1.00 0.99

-14

0.0015

Ea (kJ mol1)

A (min

R

Regression line Experimental data

-16

Temperature (°C)

Ea (kJ mol1)

)

-13

-15

TW300 1

Temperature (°C)

-13.5

ln[G(X)/T ]

2

0.0014

0.0015

0.0016

1/T Fig. 8. The linear regression for the extraction of the kinetic parameters of the fuel blend of 50 wt.% TW300 + 50 wt.% coal at (a) the second and (b) the third reaction stages.

Acknowledgments The authors gratefully acknowledge the financial support of the National Science Council, Taiwan, ROC, for this study. References [1] International Energy Agency, World Energy, Outlook; 2011. [2] Hamelinck CN, Faaij APC, Den Uil H, Boerrigter H. Production of FT transportation fuels from biomass; technical options, process analysis and optimisation, and development potential. Energy 2004;29:1743–71.

)

R

2

[3] Sami M, Annamalai K, Wooldridge M. Co-firing of coal and biomass fuel blends. Prog Energy Combust Sci 2001;27:171–214. [4] Liu DC, Zhang CL, Mi B, Shen BK, Feng B. Reduction of N2O and NO emissions by co-combustion of coal and biomass. J Inst Energy 2002;75:81–4. [5] Dai J, Sokhansanj S, Grace JR, Bi X, Lim CJ, Melin S. Overview and some issues related to co-Firing biomass and coal. The Canadian J Chem Eng 2008;86:367–86. [6] Pan YG, Velo E, Roca X, Manya JJ, Puigjaner L. Fluidized-bed co-gasification of residual biomass/poor coal blends for fuel gas production. Fuel 2000;79: 1317–26. [7] McLendon TR, Lui AP, Pineault RL, Beer SK, Richardson SW. High-pressure cogasification of coal and biomass in a fluidized bed. Biomass Bioenergy 2004;6:377–88. [8] Seo MW, Goo JH, Kim SD, Lee SH, Choi YC. Gasification characteristics of coal/ biomass blend in a dual circulating fluidized bed reactor. Energy Fuels 2010;24:3108–18. [9] Pan YG, Velo E, Puigjaner L. Pyrolysis of a blend of biomass with poor coals. Fuel 1996;75:412–8. [10] Moghtaderi B. The safety implication of low heating rate pyrolysis of coal/ biomass blends in pulverised fuel boilers. J Loss Prev Process Ind 2001;14: 161–5. [11] Biagnini E, Lippi F, Petarca L, Tiognotti L. Devolatilization rate of biomasses and coal-biomass blends: an experimental investigation. Fuel 2002;81:1041–50. [12] Kastanaki E, Vamvuka D, Grammelis P, Kakaras E. Thermogravimetric studies of the behavior of lignite–biomass blends during devolatilization. Fuel Process Technol 2002;77–78:159–66. [13] Meesri C, Moghtaderi B. Lack of synergistic effects in the pyrolytic characteristics of woody biomass/coal blends under low and high heating rate regimes. Biomass Bioenergy 2002;23:55–66. [14] Vuthaluru HB. Thermal behaviour of coal/biomass blends during co-pyrolysis. Fuel Process Technol 2003;85:141–55. [15] Moghtaderi B, Meesri C, Wall TF. Pyrolytic characteristics of blended coal and woody biomass. Fuel 2004;83:745–50. [16] Vuthaluru HB. Investigations into the pyrolytic behaviour of coal/biomass blends using thermogravimetric analysis. Bioresour Technol 2004;92:187–95. [17] Lester E, Gong M, Thompson A. A method for source apportionment in biomass/coal blends using thermogravimetric analysis. J Anal Appl Pyrolysis 2007;80:111–7. [18] Zhang L, Xu S, Zhao W, Liu S. Co-pyrolysis of biomass and coal in a free fall reactor. Fuel 2007;86:353–9. [19] Chen WH, Wu JS. An evaluation on rice husks and pulverized coal blends using a drop tube furnace and a thermogravimetric analyzer for application to a blast furnace. Energy 2009;34:1458–66. [20] Haykiri-Acma H, Yaman S. Interaction between biomass and different rank coals during co-pyrolysis. Renew Energy 2010;35:288–92. [21] Idris SS, Rahman NA, Ismail K, Alias AB, Rashid ZA, Aris MJ. Investigation on thermochemical behaviour of low rank Malaysian coal, oil palm biomass and their blends during pyrolysis via thermogravimetric analysis (TGA). Bioresour Technol 2010;101:4584–92. [22] Chen WH, Cheng WY, Lu KM, Huang YP. An evaluation on improvement of pulverized biomass property for solid fuel through torrefaction. Appl Energy 2011;88:3636–44. [23] Chen WH, Lu KM, Tsai CM. An experimental analysis on property and structure variations of agricultural wastes undergoing torrefaction. Appl Energy 2012;100:318–25. [24] Rousset P, Aguiar C, Labbe N, Jean-Michel. Commandre enhancing the combustible properties of bamboo by torrefaction. Bioresour Technol 2011; 102:8225–31. [25] Li H, Liu X, Legros R, Bi XT, Lim CJ, Sokhansanj S. Torrefaction of sawdust in a fluidized bed reactor. Bioresour Technol 2012;103:453–8. [26] Wu KT, Tsai CJ, Chen CS, Chen HW. The characteristics of torrefied microalgae. Appl Energy 2012;100:52–7.

K.-M. Lu et al. / Applied Energy 105 (2013) 57–65 [27] Tumuluru JS, Sokhansanj S, Hess JR, Wright CT, Boardman RD. A review on biomass torrefaction process and product properties for energy applications. Ind Biotechnol 2011;7:384–401. [28] Stelt MJC, Gerhauser H, Kiel JHA, Ptasinski KJ. Biomass upgrading by torrefaction for the production of biofuels: a review. Biomass Bioenergy 2011;35:3748–62. [29] Chew JJ, Doshi V. Recent advances in biomass pretreatment-torrefaction fundamentals and technology. Renew Sust Energy Rev 2011;15:4212–22. [30] Lin YJ, Hwang GS. Charcoal from biomass residues of a Cryptomeria plantation and analysis of its carbon fixation benefit in Taiwan. Biomass Bioenergy 2009;33:1289–94. [31] Chen WH, Kuo PC. Isothermal torrefaction kinetics of hemicellulose, cellulose, lignin and xylan using thermogravimetric analysis. Energy 2011;36:6451–60.

65

[32] Ma Y, Li S. The pyrolysis, extraction and kinetics of Buton oil sand bitumen. Fuel Process Technol 2012;100:11–5. [33] Coats AV, Redfern JP. Kinetic parameters from thermogravimetric data. Nature 1964;201:68–9. [34] Guo J, Lua AC. Kinetic study on pyrolytic process of oil palm solid waste using two step consecutive reaction model. Biomass Bioenergy 2001;20: 223–33. [35] Sait HH, Hussain A, Salema AA, Ani FN. Pyrolysis and combustion kinetics of date palm biomass using thermogravimetric analysis. Bioresour Technol 2012;118:382–9. [36] Chen WH, Kuo PC. Torrefaction and co-torrefaction characterization of hemicellulose, cellulose and lignin as well as torrefaction of some basic constituents in biomass. Energy 2011;36:803–11.