Bioresource Technology 102 (2011) 3451–3456

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Pyrolysis of ramie residue: Kinetic study and fuel gas produced in a cyclone furnace

Gong Cheng a, Leguan Zhang a,b, Piwen He a, Feng Yan a, Bo Xiao a,⇑, Tao Xu a, Chengcheng Jiang a,Yanli Zhang a, Dabin Guo a

a School of Environmental Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, PR Chinab Department of Environmental and Municipal Engineering, Henan University of Urban Construction, Pingdingshan 467044, PR China

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

Article history:Received 4 August 2010Received in revised form 23 October 2010Accepted 25 October 2010Available online 30 October 2010

Keywords:Pyrolysis kineticsCyclone furnaceRamie residueEquivalence ratio

0960-8524/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.biortech.2010.10.120

⇑ Corresponding author. Tel./fax: +86 027 8755746E-mail address: xiaobo1958@126.com (B. Xiao).

The thermal decomposition behavior of ramie residue (RR) and the characteristics of fuel gas produced in acyclone furnace were studied. The pyrolysis kinetics was investigated using thermogravimetric analysis(TGA) at heating rates of 5–20 �C/min. The results showed that RR mainly decomposed between 250 and390 �C, and the apparent activation energy ranged from 200 to 258 kJ/mol. In the cyclone furnace, fast pyro-lysis, partial combustion and gasification occurred almost simultaneously, and the thermal energy was sup-plied by partial combustion of RR powder at the hypo stoichiometric amount of air. Higher effect ofequivalence ratio (ER) led to higher reaction temperature and fewer contents of tar and char, but too highER lowered fuel gas content and degraded fuel gas quality. Over the ranges of the experimental conditions,the gas yield varied between 1.07 and 2.08 N m3/kg and the LHV was between 3350 and 4798 kJ/N m3.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

The bioenergy from biomass was recognized as a potentialrenewable energy provided for the shortage of the oil (Park et al.,2009). A significant way to obtain the bioenergy was using biomasswastes such as wheat straw, sewage sludge and oil-palm emptyfruit bunch (Budarin et al., 2009; Li et al., 2009; Thipkhunthodet al., 2007), because of low cost, extensive source and pollutionreduction (Tu et al., 2009). Ramie (Boehmeria nivea), namely‘‘China grass,’’ was widely cultivated in China as an important tex-tile raw material. In China, the production of ramie was more than500,000 tons per year. Generally, ramie fiber stripped from stembast of the plant was used for textile process. However, residualdecorticated stems accounted for more than 90% of the total ramieproduction and they were often treated by burning as agriculturalwastes. It led to serious environmental problems and enormouswaste of resources. So, the reutilization of this waste and theexploitation for potential bioenergy would be indispensable.

Biomass thermochemical processing including pyrolysis, com-bustion and gasification was an advanced technology for bioener-gy conversion (Cao et al., 2006; Dare et al., 2001; Xiao et al., 2010).An innovative cyclone furnace, similar to a cyclone dust catcher,was widely applied in combustion and gasification of biomass(Guo et al., 2009b; Luo et al., 2009). In the cyclone furnace, fastpyrolysis, partial combustion and gasification took place almostsimultaneously, and the thermal energy was provided by partial

ll rights reserved.

4.

combustion of fuel particle. Owing to the combined reactor withpyrolysis, combustion and gasification, heat transfer and masstransfer were intense in the cyclone furnace. It would result in ahigh temperature for pyrolysis and a low content of tar in fuelgas (Sun et al., 2009). Furthermore, the separation of char fromthe flow was achieved in the furnace without additional cleaningequipment, which reduced initial investment significantly. Com-pared with the fluidized bed, the cyclone furnace was simple,low-cost and easy to operate (Guo et al., 2009b; Luo et al., 2010).

Moreover, it was crucial that an understanding in the decompo-sition kinetics of biomass material prior to practical design forpyrolysis process. Thermogravimetric analysis (TGA) was one ofthe most common techniques used to investigate thermal behaviorand kinetics during pyrolysis of various materials (Font et al., 2009;Zhou et al., 2009).

The aim of this study was to examine the pyrolysis of ramie res-idue for fuel gas in the cyclone furnace with the energy self-suffi-ciency. The pyrolysis kinetics was determined by calculatingdifferential thermogravimetric (DTG) curves based on the Arrhe-nius equation. And the influence of equivalence ratio (ER) onreaction temperature, production distribution, and fuel gas charac-terization was also studied.

2. Methods

2.1. Materials

The RR was residual decorticated stem of ramie which was fromfarm in Xianning City, Hubei Province, China. The samples were

3452 G. Cheng et al. / Bioresource Technology 102 (2011) 3451–3456

dried under the sun for a period of 7 days to reduce the moisturecontent and then were shredded into particles of sizes of approxi-mately 0.125 mm. The proximate and ultimate analyses and theheating value of the sample were listed in Table 1.

2.2. Analysis methods

Thermogravimetric analysis of the RR sample was carried outby TA Instruments system (Diamond TG/DTA, PerkinElmerInstruments). A sample mass of 3.5 ± 1.0 mg was used for thethermogravimetric analysis in each experiment. Nitrogen wasused as a carrier gas. The heating rate was controlled at 5, 10,15 and 20 �C/min from 20 to 800 �C. Ultimate analysis of the sam-ple was obtained with a CHNS/O analyzer (Vario Micro cube, Ele-mentar). Such analysis gave the weight percent of carbon,hydrogen, nitrogen, and sulfur in the sample simultaneously,and the weight percent of oxygen was determined by difference.

2.3. Apparatus and procedures

The equipment included cyclone furnace, screw feeder, air com-pressor, flowmeter, gas sampling, and thermocouple (K-type). Theschematic configuration was illustrated in Fig. 1. The cyclone fur-nace (high 600 mm, outer diameter 300 mm) was made of heat-resistant stainless steel tubing with a 5 mm of wall thickness andwrapped with insulation materials.

At the start-up of each experimental run, RR powder wasadded to the hopper; the air compressor was turned on to pro-vide an excessive amount (60 m3/h) of air at first. Meanwhile,the screw feeder was turned on at the desired rotate speed tokeep a 15 kg/h of feedstock flow rate, and then oxy-acetyleneflame was used to ignite RR powder in the cyclone furnace. After5 min of complete combustion, the temperature in the reactor

Fig. 1. Experimental apparatus 1, air compressor; 2, screw feeder; 3, cyclonefurnace; 4, thermocouples; 5, flow meter; 6, gas sampling port; 7, temperaturemeasurement system; 8, gas flow pipe.

Table 1Ultimate analysis and proximate analysis of the sample.

Ultimate analysis (wt.%) Proximate analysis (wt.%)

C 44.54 ± 0.56 Higher heating value (MJ/kg) 16.62 ± 0.32H 5.36 ± 0.13 Moisture content 4.52 ± 0.18Oa 47.78 ± 0.40 Volatile matter 78.46 ± 0.48N 0.41 ± 0.09 Fixed carbon 15.17 ± 0.54S 0.06 ± 0.008 Ash 1.85 ± 0.39

a By difference.

was about 800 �C, and then the air flow was adjusted by adjustingvalve of air compressor to obtain the ER of 0.21, 0.25, 0.28, 0.32,and 0.35. When the temperature in the cyclone furnace kept astable state for 3 min, the tests began and experimental data wererecorded. At the same time, the temperature and outlet gas flowrate were recorded. Normally each test was repeated three times.Because of the energy supplied by partial combustion, the tem-perature in the reactor might have the difference of 10–30 �Camong parallel tests.

The fuel gas was sampled with a sampler fixed a cotton filter fordrying and cleaning. The gaseous products of H2, CH4, N2, CO, CO2,C2H4 and C2H6 were analyzed by GC 9800T with a thermal conduc-tivity detector (TCD). The columns used were TDX-01 for the anal-ysis of H2, CH4, CO, CO2 and 5A, porapak Q for the analysis of C2H4,C2H6. The temperatures of injector, oven and detector were at 200,85 and 90 �C, respectively. The carrier gas was argon in all analyses.The gas standards were mixtures of H2, CH4, CO, CO2, C2H4 andC2H6. With further dilutions by different volumes of pure N2 stan-dard, different concentrations of each gas compound were ob-tained to generate a calibration curve.

2.4. Methods of data processing

The lower heating value (LHV) of fuel gas was defined as(Lv et al., 2004; Mun et al., 2010; He et al., 2009),

LHVðkJ=N m3Þ ¼ ðCO� 126:36þH2 � 107:98þ CH4

� 358:18þ C2H4 � 59:036þ C2H6 � 63:772Þ ð1Þ

where, CO, H2, CH4, C2H4 and C2H6 were the molar percentages ofcomponents of the product gas.Equivalence ratio was calculated by,

ER ¼ actually fed air weight=stoichiometrical air weight ð2Þ

The carbon conversion efficiency was calculated by,

Xcð%Þ ¼ ½12YðCO%þ CO2%þ CH4%þ 2� C2H4%

þ 2� C2H6%Þ=22:4� C%� � 100% ð3Þ

where, Y was the product gas yield (N m3/kg), C% was the mass per-centage of carbon in ultimate analysis of the sample, and CO, CO2,CH4, C2H4 and C2H6 were the molar percentages of components ofthe product gas.

0

20

40

60

80

100

0 200 400 600 800

Temperature ( )

TG

(%

)

5 /min

10 /min

15 /min

20 /min

Fig. 2. TG curves of RR at different heating rates.

0

5

10

15

20

0 200 400 600 800

Temperature ( )

DT

G (

%•m

in-1

)

20 /min

15 /min

10 /min

5 /min

Fig. 3. DTG curves of RR at different heating rates.

-4

-3.5

-3

20%

30%

40%

50%

G. Cheng et al. / Bioresource Technology 102 (2011) 3451–3456 3453

3. Results and discussion

3.1. Pyrolysis kinetics

3.1.1. Pyrolysis at different heating ratesFigs. 2 and 3 showed the TG and DTG curves for RR at heating

rates of 5, 10, 15 and 20 �C/min. From Fig. 2, a small change of sam-ple mass at the temperature lower than 200 �C was attributed tovaporization of moisture, and one major mass loss step occurredbetween 250 and 390 �C. Meanwhile, the DTG curve had one over-lapping peak and a flat tailing section, and the maximum mass lossrate at different heating rates occurred at about 350 �C. The resultwas similar to other Chinese biomasses such as rice straw, cottonstraw and maize straw reported by Hu et al. (2007). It had beenwell known that lignocellulosic biomass mainly consisted ofcellulose, hemicellulose and lignin, and these compositions signif-icantly influenced the biomass pyrolysis process (Sonobe andWorasuwannarak, 2008). Under a slow heating regime, decompo-sition of cellulose was complete under around 360 �C, hemicellu-lose decomposes between 200 and 260 �C and lignin between280 and 500 �C (Aguiar et al., 2008). So the shoulder peak at lowertemperature was attributed to the thermally labile hemicellulose,and the maximum peak represented the decomposition of the cel-lulose (Fisher et al., 2003; Müller-Hagedorn et al., 2003). Moreover,the DTG curves presented that increasing the heating rate tendedto slightly delay thermal degradation processes towards highertemperatures. With raising the heating rate, the maximum rateof mass loss also increased from 4.65%/min to 18.97%/min.

-2.5

-2

-1.5

-1

-0.5

0.0015 0.0016 0.0017 0.0018 0.0019

1/T (K-1)

ln(d

X/d

t)

60%

70%

75%

Fig. 4. Relationship between ln (dX/dt) and 1/T at various conversions.

3.1.2. Kinetics parametersKinetic model was studied for RR pyrolysis to obtain the kinetic

parameters including activation energy, reaction order, and pre-exponential factor. In this work, the isoconversional method wasemployed as a function of the extent of conversion and the differ-ential method (Kim and Kim, 2004) was used to obtain the pyroly-sis kinetic parameters from the thermogravimetric data.

The results of thermogravimetric experiments were expressedas a function of conversion X, which was defined as

X ¼ ðW0 �W tÞ=ðW0 �W f Þ ð4Þ

where W0 was the initial mass of sample; Wt was the mass ofpyrolyzed sample and Wf was the final residual mass.

The rate of conversion, dX/dt, in thermal decomposition wasexpressed as

dX=dt ¼ kf ðXÞ ð5Þ

The reaction rate constant k was expressed by the Arrheniusequation

k ¼ A expð�E=RTÞ ð6Þ

According to the nth-order reaction mechanism, f(X) wasexpressed as (Li et al., 2010; Hu et al., 2007):

f ðXÞ ¼ ð1� XÞn ð7Þ

Substituting Eqs. 6 and 7 into Eq. (5) and taking a natural loga-rithm, the above equation yielded

ln ðdX=dtÞ ¼ lnAþ nln ð1� XÞ � E=RT ð8Þ

In Eqs. (5)–(8), A and E were the Arrhenius parameters, pre-exponential factor and activation energy, respectively; f(X) wasthe reaction model, that was a function of X; n was reactionorder.

Based on Eq. 8, activation energy E can be determined from theplot ln (dX/dt) versus 1/T. For example, when the conversion was40%, the corresponding temperatures at heating rates of 5, 10, 15and 20 �C/min were 602, 593, 587 and 579 K, respectively. The cor-responding ln (dX/dt) values were �1.5262, �2.2797, �2.7992 and�3.2333 at 1/T values of 1.726, 1.704, 1.687 and 1.622 � 10�3 K�1,respectively. Therefore, the activation energy E at conversion of40% can be obtained from the slope of the straight line, which plot-ted based on ln (dX/dt) versus 1/T. When the apparent order wasassumed to be 0, 1 or 2, the pre-exponential factor can be also ob-tained from the intercept, ln A + n ln (1 � X).

The relationship between ln (dX/dt) and 1/T at various conver-sions from 0.2 to 0.75 was shown in Fig. 4. The activation energiesat various conversions were presented in Table 2. It can be seenthat the activation energy for RR pyrolysis was from 200 to258 kJ/mol at conversion from 0.2 to 0.75. A relatively small varia-tion was in the first step for conversion from 0.2 to 0.6. Then, theactivation energy increased sharply for the higher conversion va-lue. Similar results were studied on several lignocellulosic biomassresidues. Sonobe and Worasuwannarak (2008) investigated thepyrolysis behaviors of several agricultural residues using the dis-tributed activation energy model. The activation energies of rice

Table 2Calculated kinetic parameters for RR pyrolysis.

Conversion (%) 20 30 40 50 60 70 75

R2 0.969 ± 0.014 0.965 ± 0.019 0.974 ± 0.011 0.987 ± 0.025 0.983 ± 0.027 0.964 ± 0.028 0.928 ± 0.030E (kJ/mol) 200 ± 3 218 ± 3 223 ± 3 224 ± 3 224 ± 3 242 ± 4 258 ± 4A n = 0 3.19 � 1019 6.22 � 1020 6.14 � 1020 3.88 � 1020 2.68 � 1020 3.95 � 1021 5.22 � 1023

n = 1 3.99 � 1019 8.89 � 1020 1.02 � 1021 7.76 � 1020 6.71 � 1020 1.32 � 1022 2.09 � 1023

n = 2 4.98 � 1019 1.27 � 1021 1.71 � 1021 1.55 � 1021 1.68 � 1021 4.39 � 1022 8.35 � 1023

Table 3

3454 G. Cheng et al. / Bioresource Technology 102 (2011) 3451–3456

straw, rice husk and corncob were about 120–205, 150–210 and140–210 kJ/mol, respectively. Biagini et al. (2008) studied kineticsof three biomass residues (rice husks, olive cake and cacao shells)by applying traditional isoconversional methods. In contrast to arapid increase of activation energy for conversion below and above0.2 and 06, respectively, a plateau can be observed for conversionbetween 0.25 and 0.55 in the case of the biomass residues.

Table 2 also showed the pre-exponential factors at the conver-sion from 0.2 to 0.75. The value ranged from 1019 to1023 min�1 andit was lower at the conversions of 0.5 and 0.6. However, pre-expo-nential factor increased sharply when the conversion was above0.7. This result was consistent with study on pyrolysis of orientalwhite oak in Korea (Park et al., 2009). According to Park et al.(2009), the highest value at 0.8 of conversion probably attributedto further devolatilization of residual char.

3.2. Fuel gas production in the cyclone furnace

As soon as RR powder was fed along the tangent direction of thefurnace wall with air, the pyrolysis promptly occurred and thematerial was decomposed to char, tar and light gases such as H2,CO, CH4, CO2, H2O, light hydrocarbons. Partial fuel gas and charunderwent combustion reactions (Eqs. (9)–(11)), which effectivelyprovided the thermal energy for biomass pyrolysis.

Cþ O2 ¼ CO2 þ 408:8kJ=mol ð9Þ2Cþ O2 ¼ 2COþ 246:6kJ=mol ð10Þ2H2 þ O2 ¼ 2H2OðgÞ þ 484kJ=mol ð11Þ

Then gas products and solid residues were revolved anddropped along furnace wall. During this process, gasification reac-tions between gaseous reactant and solid reactant took place asfollows (Eqs. (12)–(15)). Finally, residual ash separated from thegas flow and fell into the taper container, and the fuel gas flew intothe pipe with high temperature. In this case, the tar cracking andgas phase reactions such as steam reformation for H2 production(Eqs. (16)–(18)) were major reactions.

Cþ CO2 ¼ 2CO� 172kJ=mol ð12ÞCþH2OðgÞ ¼ COþH2 � 131kJ=mol ð13ÞCþ 2H2OðgÞ ¼ CO2 þ 2H2 � 76kJ=mol ð14Þ2Cþ 2H2OðgÞ ¼ CH4 þ CO2 þ 11kJ=mol ð15ÞCOþH2OðgÞ ¼ H2 þ CO2 þ 42:2kJ=mol ð16ÞCH4 þH2OðgÞ ¼ COþ 3H2 � 206kJ=mol ð17ÞCH4 þ 2H2OðgÞ ¼ CO2 þ 4H2 � 165kJ=mol ð18Þ

Overall material balance.

In (kg/h) Out (kg/h) Overall closure (%)

Feedstock Air Total Gas Tar Char Total

15 15.52 30.52 21.60 2.98 5.46 30.04 98.4315 18.10 33.10 25.31 2.49 4.81 32.61 98.5215 20.69 35.69 30.89 1.63 2.69 35.21 98.6615 23.27 38.27 34.99 1.15 1.52 37.66 98.4115 25.86 40.86 39.41 0.56 0.74 40.71 99.63

3.2.1. Gas residence time and material balance calculationThis type of pyrolysis reactor was similar to a cyclone dust

catcher. Residual ash was collected in the taper slag container. Ingeneral, the cyclone furnace could operate continuously 2 h. In thistest, gas residence time and material balance calculation werestudied.

In the cyclone furnace, gas residence time can be determined by

T ¼ 3600V=u ð19Þ

where T was the gas residence time (s); V was the volume of thereactor (m3); and u was the outlet gas flow rate (m3/h).

When the inlet air flow rate were 12, 14, 16, 18 and 20 m3/h, theoutlet gas flow rate measured were 16.1, 19.7, 24.0, 28.2 and31.2 m3/h, and the gas residence time were 9.5, 7.8, 6.4, 5.4 and4.9 s, respectively.

Table 3 showed the material balance calculation at different airflow rates. The overall material balance was above 98%. When theair flow rate was 20 Nm3/h, the gas produced reached 39.41 kg/h,and the production of char and tar were 0.74 and 0.56 kg/h,respectively.

3.2.2. Effect of ERIn the cyclone furnace, equivalence ratio (ER) was the main con-

trolling parameter because it affected reaction temperature. Thetemperatures were measured at different points in the reactorand an average value was recorded in this test. Effect of ER onthe reaction temperature was presented in Table 4. It can be seenthat the average reaction temperature increased from 657 to948 �C with the increasing ER from 0.21 to 0.35. The reason maybe that more oxygen at higher ER was available for combustionof volatiles and char from pyrolysis (Sun et al., 2009). With thetemperature increasing from 657 to 948 �C, the char decreasedgradually from 36.43% to 4.90%, and the tar decreased graduallyform 19.88% to 3.73%, while the gas increased drastically from40.53% to 90.33%. It could be explained that a part of char wasburnt and the tar was further thermally cracked with increasingtemperature.

As shown in Table 4, the concentrations of H2, CO, CO2, CH4,C2H4 and C2H6 were listed at different ERs. When ER increasedfrom 0.21 to 0.28, the concentrations of H2 and CO increased, whilethe ER was higher than 0.28, the yields of H2 and CO decreased syn-chronously. Increased ER tended to an increasing content of CO2

due to the reaction between char and oxygen. However, the con-tents of CH4, C2H4 and C2H6 were contrary because of decomposi-tion and combustion of these gases at high temperature. Themaximum yield of fuel gas was presented at the ER of 0.25. Toosmall or too large ER was unfavorable for fuel gas production inthe cyclone furnace. Generally, ER was a crucial factor affectinggas quality because it not only affected the reaction temperaturebut also represented the oxygen quantity introduced into the fur-nace (Guo et al., 2009a). When the ER was lower, pyrolysis and

Table 4Experimental results of different ERs.

ER 0.21 0.25 0.28 0.32 0.35

Temperature (�C) 657 ± 14 749 ± 12 823 ± 11 889 ± 8 928 ± 8

Product distribution (wt.%)Gas 40.53 ± 1.24 48.07 ± 1.24 68.01 ± 1.10 78.12 ± 0.61 90.33 ± 0.89Tar 19.88 ± 1.41 16.61 ± 0.79 10.84 ± 0.86 7.66 ± 0.63 3.73 ± 0.45Char 36.43 ± 3.77 32.06 ± 2.62 17.91 ± 1.54 10.12 ± 0.95 4.90 ± 0.48

Gas composition (mol%, dry basis)H2 6.21 ± 0.07 7.16 ± 0.09 8.51 ± 0.10 8.38 ± 0.08 8.20 ± 0.08CO 16.07 ± 0.21 17.99 ± 0.20 18.22 ± 0.15 18.02 ± 0.16 15.95 ± 0.14CO2 9.68 ± 0.10 10.66 ± 0.13 14.21 ± 0.18 18.66 ± 0.24 20.72 ± 0.36CH4 5.54 ± 0.07 4.75 ± 0.06 2.80 ± 0.03 1.85 ± 0.03 1.24 ± 0.03C2H4 1.19 ± 0.032 0.73 ± 0.037 0.57 ± 0.059 0.54 ± 0.066 0.06 ± 0.085C2H6 0.31 ± 0.043 0.11 ± 0.018 0.09 ± 0.011 0.05 ± 0.009 0.03 ± 0.007

Gas characterizationGas yield (N m3/kg) 1.07 ± 0.02 1.31 ± 0.02 1.60 ± 0.02 1.88 ± 0.02 2.08 ± 0.02Fuel gas (mol%, dry basis) 29.32 ± 0.49 30.94 ± 0.46 30.19 ± 0.38 28.84 ± 0.24 25.48 ± 0.37LHV (kJ/N m3) 4776 ± 49 4798 ± 56 4263 ± 55 3880 ± 48 3350 ± 42H2/CO (mol/mol) 0.39 ± 0.01 0.40 ± 0.01 0.47 ± 0.01 0.47 ± 0.01 0.51 ± 0.01Carbon conversion efficiency (wt.%) 44.13 ± 2.33 55.27 ± 2.06 70.34 ± 2.77 89.79 ± 2.97 95.29 ± 2.94

G. Cheng et al. / Bioresource Technology 102 (2011) 3451–3456 3455

incomplete combustion of RR played a more prevailing role. Whenthe ER was higher, oxidization reactions became stronger and morecombustible gas was consumed (Lv et al., 2004). However, the LHVof the produced gas gradually decreased with the increasing ER. Itwas due to the shortened resident time of gas products and dilu-tion by the addition of nitrogen. The increasing ER and reactiontemperature promoted the pyrolysis of biomass, the secondarypyrolysis of tar and carbon combustion, which resulted in theincreasing carbon conversion efficiency.

4. Conclusions and future work

The kinetic parameters for pyrolysis of ramie residue wereobtained by thermogravimetric analysis. The apparent activationenergy was in the range of 200–258 kJ/mol when the conversionvaried from 0.2 to 0.75. The fuel gas produced from RR pyrolysisin the cyclone furnace with the energy self-sufficiency was alsostudied. ER was a crucial factor affecting reaction temperatureand fuel gas quality. The higher ER resulted in a higher reactiontemperature. However, the LHV of the gas gradually decreasedwith the increasing ER. The optimal value of ER for the content offuel gas was 0.25. Under the operating conditions, the yieldsof tar, char and gas were 3.73–19.88%, 4.90–36.43% and 40.53–90.33%, respectively; the LHV of fuel gas was between 3350 and4776 kJ/N m3; carbon conversion efficiency ranged from 44.13%to 95.29%.

It was feasible that the pyrolysis of RR in the cyclone furnacewith the energy self-sufficiency. This type of pyrolysis reactorhad a great potential in the commercial scale application. Our fur-ther studies included: an economic benefit analysis and an energyconsumption calculation of the reactor system; a design of theautomatic slag treatment system for improving the sustainabilityof the reactions; using oxygen-enriched air combustion to promotethe LHV of the fuel gas.

Acknowledgements

This research was supported by the National Natural ScienceFoundation of China (No. 20876066) and Hubei Province Scientificand Technological Project (No. 2007AA204B01). The authors aregrateful to the Analytical and Testing Center of Huazhong Universityof Science and Technology for carrying out the ultimate analysis andthe thermogravimetric analysis of the samples.

References

Aguiar, L., Márquez-Montesinos, F., Gonzalo, A., Sánchez, J.L., Arauzo, J., 2008.Influence of temperature and particle size on the fixed bed pyrolysis of orangepeel residues. J. Anal. Appl. Pyrol. 83, 124–130.

Biagini, E., Fantei, A., Tognotti, L., 2008. Effect of the heating rate on thedevolatilization of biomass residues. Thermochim. Acta. 472, 55–63.

Budarin, V.L., Clark, J.H., Lanigan, B.A., Shuttleworth, P., Breeden, S.W., Wilson, A.J.,Macquarrie, D.J., Milkowski, K., Jones, J., Bridgeman, T., Ross, A., 2009. Thepreparation of high-grade bio-oils through the controlled, low temperaturemicrowave activation of wheat straw. Bioresour. Technol. 100, 6064–6068.

Cao, Y., Wang, Y., Riley, J.T., Pan, W.P., 2006. A novel biomass air gasification processfor producing tar-free higher heating value fuel gas. Fuel process. Technol. 87,343–353.

Dare, P., Gifford, J., Hooper, R.J., Clemens, A.H., Damiano, L.F., Gong, D., Matheson,T.W., 2001. Combustion performance of biomass residue and purpose grownspecies. Biomass Bioenergy 21, 277–287.

Fisher, T., Hajaligol, M., Waymack, B., Kellogg, D., 2003. Pyrolysis behavior andkinetics of biomass derived materials. J. Anal. Appl. Pyrol. 62, 331–349.

Font, R., Moltó, J., Gálvez, A., Rey, M.D., 2009. Kinetic study of the pyrolysis andcombustion of tomato plant. J. Anal. Appl. Pyrol. 85, 268–275.

Guo, X.J., Xiao, B., Liu, S.M., Hu, Z.Q., Luo, S.Y., He, M.Y., 2009a. An experimentalstudy on air gasification of biomass micron fuel (BMF) in a cyclone gasifier. Int.J. Hydrogen Energy 34, 1265–1269.

Guo, X.J., Xiao, B., Zhang, X.L., Luo, S.Y., He, M.Y., 2009b. Experimental study on air–stream gasification of biomass micron fuel (BMF) in a cyclone gasifier.Bioresour. Technol. 100, 1003–1006.

He, M.Y., Xiao, B., Liu, S.M., Guo, X.J., Luo, S.Y., Xu, Z.L., Feng, Y., Hu, Z.Q., 2009.Hydrogen-rich gas from catalytic steam gasification of municipal solid waste(MSW): influence of steam to MSW ratios and weight hourly space velocity ongas production and composition. Int. J. Hydrogen Energy 34, 2174–2183.

Hu, S., Jess, A., Xu, M.H., 2007. Kinetic study of Chinese biomass slow pyrolysis:comparison of different kinetic models. Fuel 86, 2778–2788.

Kim, S.S., Kim, S., 2004. Pyrolysis characteristics of polystyrene and polypropylenein a stirred batch reactor. Chem. Eng. J. 98, 53–60.

Li, D.M., Chen, L.M., Zhao, J.S., Zhang, X.W., Wang, Q.Y., Wang, H.X., Ye, N.H., 2010.Evaluation of the pyrolytic and kinetic characteristics of Enteromorpha proliferaas a source of renewable bio-fuel from the Yellow Sea of China. Chem. Eng. Res.Des. 88, 647–652.

Li, J.F., Yin, Y.F., Zhang, X.M., Liu, J.J., Yan, R., 2009. Hydrogen-rich gas production bysteam gasification of palm oil wastes over supported tri-metallic catalyst. Int. J.Hydrogen Energy 34, 9108–9115.

Luo, S.Y., Xiao, B., Hu, Z.Q., Liu, S.M., Guan, Y.W., 2009. Experimental study onoxygen-enriched combustion of biomass micro fuel. Energy 34, 1880–1884.

Luo, S.Y., Xiao, B., Hu, Z.Q., Liu, S.M., He, M.Y., 2010. Experimental study oncombustion of biomass micron fuel (BMF) in cyclone furnace. Energy Convers.Manage. 51, 2098–2102.

Lv, P.M., Xiong, Z.H., Chang, J., Wu, C.Z., Chen, Y., Zhu, J.X., 2004. An experimentalstudy on biomass air–steam gasification in a fluidized bed. Bioresour. Technol.95, 95–101.

Müller-Hagedorn, M., Bockhorn, H., Krebs, L., Müller, U., 2003. A comparative kineticstudy on the pyrolysis of three different wood species. J. Anal. Appl. Pyrol. 68–69, 231–249.

Mun, T.Y., Seon, P.G., Kim, J.S., 2010. Production of a producer gas from woody wastevia air gasification using activated carbon and a two-stage gasifier andcharacterization of tar. Fuel 89, 3226–3234.

3456 G. Cheng et al. / Bioresource Technology 102 (2011) 3451–3456

Park, Y.H., Kim, J.S., Kim, S.S., Park, Y.K., 2009. Pyrolysis characteristics and kineticsof oak trees using thermogravimetric analyzer and micro-tubing reactor.Bioresour. Technol. 100, 400–405.

Sonobe, T., Worasuwannarak, N., 2008. Kinetic analyses of biomass pyrolysis usingthe distributed activation energy model. Fuel 87, 414–421.

Sun, S.Z., Zhao, Y.J., Tian, H.M., Ling, F., Su, F.M., 2009. Experimental study oncyclone air gasification of wood powder. Bioresour. Technol. 100, 4047–4049.

Thipkhunthod, P., Meeyoo, V., Rangsunvigit, P., Rirksomboon, Thirasak., 2007.Describing sewage sludge pyrolysis kinetics by a combination of biomassfractions decomposition. J. Anal. Appl. Pyrol. 79, 78–85.

Tu, W.K., Shie, J.L., Chang, C.Y., Chang, C.F., Lin, C.F., Yang, S.Y., Kuo, J.T., Shaw, D.G.,You, Y.D., Lee, D.J., 2009. Products and bioenergy from the pyrolysis of rice strawvia radio frequency plasma and its kinetics. Bioresour. Technol. 100, 2052–2061.

Xiao, R.R., Chen, X.L., Wang, F.C., Yu, G.S., 2010. Pyrolysis pretreatment of biomassfor entrained-flow gasification. Appl. Energy 87, 149–155.

Zhou, L.M., Luo, T.A., Huang, Q.W., 2009. Co-pyrolysis characteristics and kinetics ofcoal and plastic blends. Energy Convers. Manage. 50, 705–710.