Pyrolysis Characteristics and Kinetics of Municipal Solid Waste

Transactions of Tianjin UniversityISSN 1006- 4982pp353- 359Vol. 11No. 5Oct.2005Pyrolysis Characteristics and Kinetics of Municipal Solid W asteLI Haiying( 李海英)"2, ZHANG Shuting( 张书廷)'**, ZHAO Xinhua(赵新华)', EIJI Suzuki3( 1. School of Environment Science and Technology, Tianjin University , Tianjin 300072, China;2. School of Metallurgy and Energy ，Hebei Polytechnic University ，Tangshan 063009，China;3. Department of Fine Material Engineering, Shinshu University, Tokida Ueda 386 -8567, Japan )Abstract : Based on a systemic survey, the pyrolysis characteristics and apparent kinetics of the mu-nicipal solid waste ( MSW ) under different conditions were researched using a special pyrolysis re-actor, which could overcome the disadvantage of thermo-gravimetric analyzer. The thermal decom-position behaviour of MSW was investigated using thermo-gravimetric ( TG ) analysis at rates of4.8, 6.6, 8.4, 12.0 and 13.2 K/min. The pyrolysis characteristics of MSW were also studied in dif-ferent function districts. The pyrolysis of MSW is a complex reaction process and three main stagesare found according to the results. The first stage represents the degradation of cellulose and hemi-cellulose, with the maximum degradation rate occuring at 150 C- -200 C; the second stage repre-sents dehydrochlorination and depolymerization of intermediate products and the differential thermo-gravimetric ( DTG ) curves have shoulder peaks at about 300 C; the third stage is the decomposi-tion of the residual big molecular organic substance and lignin at 400 C- 600 C. Within the rangeof given experimental conditions, the results of non-linear ftting algorithm and experiment are inagreement with each other and the correlation coefficients are over 0.99. The kinetic characteristicsare concerned with the material component and heating rate. The activation energy of reaction de-creases with the increase of heating rate.Keywords : municipal solid waste; pyrolysis characteristics; kinetic; non-linear ftting algorithmAt present, the annual production of the munici- ingly， the organism-bearing level is rising year bypal solid waste( MSW ) in China has accounted for year, causing the calorific value increase. For exam-nearly one fourth of that in the world. The stored-up ple, the organism-bearing levels of double-gased areawaste is about 6 million tons, occupying the land area ( coal gas and central heating ) in China are as fol-of 0.5 bllion square meters, and more than 200 cities lows: Beijing, 90. 9%[3]; Tianjin, 77% ; Shanghai,are surrounded by MSW. Moreover, MSW is increas-82%; Shenyang, 86% ; Mudanjiang, 88%- 4J. There-ing at a rate of 8%-10% every year-1.2. In the .fore，alternatives of waste management technologiessolid waste, many kinds of substances with high mois-must be studied. Pyrolysis is a suitable technique,ture content are easy to be biodegraded and fermen-which can reduce , recycle and reuse MSW.ted. They send off bad smell, and affect the environ-Studies on the pyrolysis of typical components ofmental sanitation. The leakage containing poisonousMSW have been performed by many researchers.and harmful heavy metal ions pollutes the soil andThese studies were mostly concerned with the thermal .groundwater.Nowadays, landiling and incineration of MSw decomposition of single composition, such as paper,are practised on a large scale. However, the cost is plastics and so on-5- 8. Garcfa et al studied the py-rising rapidly and the acceptance of these methods is rolysis of MSW by thermo-gravimetric analysisdecreasing for the environmental problem. Correspond- ( TGA )L t which considered two中国煤化工YHCNMHG* Accepted date :2005- 05- 10.LI Haiying, born in 1971 , female，doctorate student，lecturer.* Supported by National Natural Science Foundation of China( No. 50378061 ).* * Correspondence to ZHANG Shuting. E-mail : stzhang@ tju. edu. cn.Transactions of Tianjin UniversityVol. 11 No.5 2005independent reactions was applied to simulate the1 Experimentalprocess9. Lin et al calculated the pyrolysis reactionrate of those key component fractions ( paper, LDPE，1. 1 MaterialsHDPE，PS and PVC ) of RDF using the weight sumBecause MSW comprises complex componentsmethod10]. Jin et al and Li et al made a detailed varying in different places and seasons, so the experi-analysis on the pyrolysis of the single component of ment material is a simulated fraction similar to theMSWl 11,12J. These researches reveal that pyrolysis components found in MSW in China. The mixed com-may be complicated by the fact that the MSW is a ponents based on the make-up of the MSW in differentkind of poor thermal conductor and heterogeneous ma- function districts in Beijing are as follows 131: feedterial.refuses, textiles，plastics, wood dust， rubber, wasteThis paper studies the thermal decomposition of paper, glass, ash and so on. After excluding inorga-MSW with the objective to obtain pyrolysis character- nic materials such as dust, glass, metal etc, the com-istics and kinetics of the solid wastes under different position of the sample MSW is shown in Tab. 1. Theconditions using specially-made pyrolysis equipment，primary analysis and ultimate analysis of MSW areto understand the effects of the heating rate and com- shown in Tab.2. The values ( except Qad ) in Tab. 1ponents on MSW pyrolysis behavior and to obtain ba- and Tab. 2 are weight percentage.sic data which can be applied in practice.Tab.1 Composition of the MSW in different function districts in Bejjing%Waste paperPlastics .TextileVertFeed refusesDouble-gased districts18.2525.305.354.8746. 23Bungalow districts29.1018. 662.2414. 1835. 82Commercial districts53. 3325.122. 790. 6218. 14.Institutions24. 6126.184.4510.4734. 29Hospitals38.7821. 84.11.027. 7520. 61Tab. 2 Results of primary analysis and ultimate analysis of MSWPrimary analysisUltimate analysis/%Index markWa/%A.u/%v1/% FC.2s/% Qun/(kJ.kg-')HsC5. 398.0770. 6915. 8518 8616.830.291.0547.1731.205. 2311. 8068. 624.3518 1476.270.301.1043.04 32.264. 2511.04 .71.5413. 1718 8126.550.230.5444.06 33. 334.948. 7714.0519 6470. 280.9846.75 31. 84Hospitals .4.78 .5.9173. 7615. 5519 3246.740.250.7447.79 33. 791.2 Apparatus and procedurethick )were placed at the bottom of the reactor. WhenThe process diagram of the experiment system for the control units were ready， the experiment started.the pyrolysis of MSW is illustrated in Fig. 1. The At the same time, the variation of the mass of themeasurement error is土0.2%. The samples were first sample was recorded continuously and the reactionbroken into particles with the diameter less than 2 temperature ( probed with a thermocouple ) was recor-mm，then were dried for4 h at 105 C before the ex- ded at the intervals of 1 min by the data processingperiment started. An electrical balance ( 0. 1 mg unit. The pyrolysis process was operated under atmos-readability ) was used. A sample of fixed mass ( 100 pheric pressure. The effluent stream leaving the outlet+0.000 1 g ) was placed in the reactor, which was was combusted. Once the pyrolysis finished, the fur-hung on the suspension wire of the balance and had no nace and the control units were turned off. The carriercontact with the inwall of the furnace. The power of gas( p中国煤化工until the temperaturethe furnace was4 kW. The high-purity N2( 99. 99%，of react:HCNMH0 °C.at the flow of 300 mL/min ) was first introduced intoBecause“ule llllidl cuduction coefficient ofthe reactor for 1 h so as to purge away the residual 0X- municipal solid waste was low, there were temperatureygen. In order to disperse the stream uniformly and gradients in the sample, which might change duringavoid bonding， some small refractory balls( 30 mm the reaction, along the longitudinal and radial direc-- 354-LI Haiying et al: Pyrolysis Characteristics and Kinetics of Municipal Solid Wastetions of the reactor. So the recorded temperature may1 000-not match the actual sample temperature. Experimen-800-tal data obtained under different operating conditions( e. g. different positions， components and heating600rates) are shown in Fig.2， Fig.3 and Fig. 4. From400-4.8K 1 min--- 6.6K 1 minthe experimental results, the temperature difference in---- 12.0K / minradial direction was more than 220 C. In order to... 13.2 K / minavoid the error as much as possible， in this work ,100200300400TC2 was chosen as the thermometric position. .Time 1 minFig. 4 Temperature variation of MSWat various heating rates( TC2 )2 Results and discussionEfluent -T-N2.1 Pyrolysis of MSW in different function dis-trictsA study on the behavior of the MSW mixtureswas performed in order to check possible interactionsdamong components. According to the composition othe MSW in China, the raw materials were formulated1-Electronic balance; 2- -Data recorder; 3- Power supply andwith reference to those in the double-gased districts,temperature controller; 4一K-type thermocouple; 5- F urnacebungalow districts， commercial districts， institutionsFig.1Thermal gravimetric analysis reaction systemand hospitals. Simulated mixtures of different function00-districts were tested in the instrument， the heatingTw- TCrate was 12 K/min. The TG and DTG curves areP 600-shown in Fig. 5.Double-gased districts90医200-80Bungalow districtsCommercial districtsInstitutionsHospitals50100150200250 30060Time 1 min .马5040TW- -Temperature of furmace wall;30TC1- -Temperature of the positionr=0;100 200 300 400 500 600 700 800TC2- -Temperature of the position r=d/4;Temperature/ CTC3- -Temperature of the position r=d/2;TC4一Temperature of gas in the reactor.( a) TG curveFig.2 Temperature distribution of MSWin various positions。5(---- Bungalow districts: Commercial ditricts30-ρ 600-与20-三10Institutions'中国煤化工，:00 600 700 8000MHCNM H Gc020406080100120140Time / min( b) DTG curveFig.5 TG and DTG curves of the MSW inFig.3 Temperature variation of MSWdifferent function districtsin different function districts( TC2 )355Transactions of Tianjin UniversityVol. 11 No.5 2005The TG and DTG curves of the MSW show that rine of PVC converts into hydrogen chloride' 15]. Allthe total thermal degradation ratios are 65%- -75%. these reactions have no apparent bounds and some ofThe initial degradation temperatures are 70- -80 C them overlap. In the first stage, the reaction is endo-lower than the results of Garcia et al9J and Sφrum et thermic. With the temperature rising, the main reac-al14」, which is mainly caused by the difference of tion is the cracking of intermediate products into C2the thermometric positions. In the test, the thermo- C5 hydrocarbon compound. At 400- -600 C, the in-couples measure the material temperatures directly.termediate fragments remaining plastics and lignin fur-The shapes of the MSW pyrolysis curves in the differ- ther decompose. At over 600 C , the inorganic matterent function districts are similar under the same pyrol- begins decomposition and gives off carbon dioxide andysis condition. The major weight losses occur at 70一so onFinally, the solid char is composed of char-600 C. DTG curves are characterized by shoulders coal, macromolecular compound and ash.and double peaks because more than one reaction isThe heterogeneity of MSW can be observed in theinvolved. Three main distinct weight loss stages are TG curves. Because the MSW of commercial districtsfound. The first stage represents the degradation of contains a lot of waste paper and plastics , the degrada-cellulose and hemicellulose with the maximum degra-tion rate is faster than other districts. The curves havedation rate occurring at 150- -200 C. The second crossing points at high temperature , probably due tostage represents dehydrochlorination and depolymer- the sample heterogeneity and the different percentagesization of intermediate products , with the DTG curves of components and ash of the samples. Similar resultsdisplaying shoulder peaks at temperature around 300were found by Garcfa et al9] and Urban et al 16].C. The third stage is the decomposition of the residu- 2.2 Pyrolysis of MSW at diferent heating ratesal big molecule organic substance and lignin at 400一The raw material simulating the MSW of the600 C. Because dfferent components are involved in double-gased disticts of Beijing, was pyrolyzed underthe material，the process is different from the single-different heating rates. The temperature testing pointcomponent pyrolysis and these components have inter- of the thermo-couple was in the position TC2, in thenal reactions in the reactor. For example，the HClmiddle of the material. The testing place was the sameliberated by PVC may interact chemically with the for every working state. The pyrolysis weight losscellulose, catalyzing an acid hydrolysis type of reac- curves at different heating rates are shown in Fig. 6.tion and making it less stable. Catalytic effects caused100by inorganic species such as ash and residues from the90■4.8K 1 min●6.6K 1 min80sulphate production process, may cause the decompo-▲8.4K/ minsition of cellulose at lower temperatures.70▼12.0K1min60● 13.2 K1 minPyrolysis is a complex chemical reaction process三° 50including a series of reactions such as the breaking of三40-C- -C bond and isomerization of organic matter. Dur-30ing the pyrolysis, the intermediate products have two2(0 100 200 300400 500 600700 800 900trends: the big molecules may be cracked into smallTemperature/ Cmolecules or even gas，and the small molecules may( a) TG curvebe polymerized into big molecules. At the low temper-一-4.8K1minature range，the MSW has dehydration reaction， and---. 6.6 K / minthen demethyl and demethyne reactions， leading to.... 8.4 K 1 min三4(---- 12.0 K 1 minthe formation of aromatic hydrocarbon compound.导30-13.2K1minSynthesized water may interact with the methyne andrelease carbon monoxide and hydrogen gas. W hen the中国煤化工temperature gets higher, the big molecule hydrocar-MHCNMHGriGms-bons formed in the first stage go through the following0100200300400500600700800process: cracking，dehydrogenation， condensation,hydrogenation，and the release of hydrogen， meth-( b) DTG curveane, ammonia and so on. At the same time, the chlo-Fig.6 TG and DTG curves at different heating rates一356LI Haiying et al: Pyrolysis Characteristics and Kinetics of Municipal Solid WasteThe experimental results show that, within the In Eq.( 4 ), m is the actual sample mass; mo and m。range of4.8- -12.0 K/min，with the increase of the are the initial sample mass and char yield. Integratingheating rate, the maximum degradation rate becomesEq.( 3 ) becomeslarger，the corresponding temperature of maximum_doweight loss rate ( Tmax ) becomes lower and the peakb (1 -a)" . βRJxox2m βR\area enlarges. The same conclusion is reached byf_. =dx)=AEP(x) (5)using sewage sludge as the raw material. The maxi-βRmum degradation rate as a function of the temperaturewherex=-ETo calculate P( x ), we have chosenof maximum weight loss rate can be written asRT[ -d( m/mo )/dt ]max =ζTma(1 )i 19]where ζ is a coefficient correlated with the raw mate-to use Lyon's approximation'rial, heat transfer and mass transfer.P(x)=-(6 )x( x-2 )As the heating rate exceeds 12. 0 K/min, theTherefore, the overall rate of conversion is givenpeaks of DTG curve move to high temperature sideand the widths of peaks enlarge. Wen et al also foundthis phenomenon in the pyrolysis process of wood”.1-a=[1-(1-n)A F72e-E/RT]β E+2RTBecause the MSW is the thermal insulation material,(7 )at the low heating rates， the heat transfer and themass transfer can proceed more completely, and them_ m∞+(1_m∞degradation may be finished more sufficiently, causingmo momothe smaller remaining solid char. At the high heating[1-(1-n)R_r2"(8)rates, the reaction time is short. The temperature gra-dient results in the shift of peaks toward the high tem-In the kinetic analysis, the nonlinear least squareperature and the decrease of the maximum weight loss algorithm is used, which identifies parameters( A, Erate.and n ). The minimized values of the objective func-2.3 Kinetic analysistions are given below:2.3.1 Kinetic model. m/mo )xP-( m/mo )The kinetic constant is expressed by the Arrhe-(9)σ;nius equation L18 aswhere( m/mo );xP is the experimentally observed TGda/dt=k●f( a)=A●exp( - E/RT)●curve，and( m/m0 );al is the TG curve calculated by( 1 -a)”(2)numerical solution of the kinetic differential equationorda_4e-EIRr(1-a)"(3) with the given set of parameters.dT- β2.3.2Kinetic evaluation of MSWwhere A is frequency factor; E is apparent activationThe results of the calculation performed for allenergy; n is the reaction stage number;T is tempera-the investigated heating rates and MSW are collectedture; t is time; β is heating rate; R is gas factor; x isin Tab.3 and Tab.4. Fig. 7 presents the fitting for the .he conversion ( reacted fraction ) which can be ex-experimentscarriedout under different conditions.pressed byThe constancy of the calculated parameters for all themo-m( 4 ) heating rates and materials shows good ftting.mo-m。Tab.3 Pyrolysis dynamic parameters of MSW at different heatino rates_Heating rates/( K●min -1 )A/s-E/(kJ. mol -1)中国煤化工R4.80. 304 +0.10721.945+1.175TY HCNMHG0. 9976.60.056 +0.01021. 832 +0.7171. 102 +0. 0450.9998.40.051 +0. 01419. 755 +0. 9731.641+0.0780. 99512.00. 160 +0.04518. 260 +0.916.1.823 +0.0730.99713.20.091 +0.02218. 232 +0. 8561.755 +0. 0650. 998Transactions of Tianjin UniversityVol. 11 No.5 2005Tab.4 Pyrolysis dynamic parameters of MSW in different function districtsRaw materialA/s-1E/(kJ. mol-')nRDouble-gased districts1.005 +0.37722. 672+1.1622.265+0.0890. 998Bungalow disticts1. 591 +0.69723. 640+1.3412. 366 +0.1030.998.Commercial districts0. 220 +0.06918.732 +0. 9852. 045 +0. 0830.997Institutions0. 049 +0.01415. 056 +0.9151. 552 +0. 0800. 996.Hospitals0. 163 +0.04518. 262 +0. 879 _1. 999 +0. 076less energy is needed for pyrolysis in the districts with- 4.8K 1 min.0----- 6.6K 1 minmore fiber components ( e. g. commercial district ).0.8-.... 8.4K 1 minThe apparent kinetic parameters were determined-_ 12.0K 1 min--13.2K1minusing the Levenberg- Marquardt nonlinear fitting algo-ε 0.6-rithm. From the values of R2 in Tab.3 and Tab.4, a0.4 tgood agreement between experimental data and simu-lation data can be obtained. The parameters can be0.25200400600800 1 000simulated at given experimental temperature rangingTemperature/ 1from room temperature to 850 C.( a) TG curve at diferent heating rates3 ConclusionsThe analysis of TG and DTG curves of MSW un-.0--der the given experimental conditions, show that the.9---- Bungalow districts.... Commercial districtspyrolysis of MSW is a complex reaction process com-品0.8---- Institutionsprising three main stages. The first stage is the degra-E 0.6-dation of cellulose and hemicellulose, with the maxi-L 0.5-mum degradation rate occurring at 150- 200 C. The0.4-second stage represents dehydrochlorination and depo-0.3七200400 600 800 1 000lymerization of intermediate products, and the DTGTemperature/ Ccurves have shoulder peaks at temperature around 300C. The third stage is the decomposition of the residu-( b) TG curve in diferent function districtsFig.7 Pyrolysis dynamic parametersal big molecule organic substance and lignin at 400-fitting curves of MSW600 C. When the heating rate is within the range of4.8- -12. 0 K/min, with the increase of the heatingIt is possible to simulate the DTG curves accord- rate， the maximum degradation rate becomes largering to .and the corresponding temperature of maximum weightd( m/mom.oloss rate( T,) becomes lower and the peak area en-dt(1-molarges. When the heating rate exceeds 12. 0 K/ min,the highest peak of DTG curve moves towards the high[1-( 1-n,AR_ T2. E/RT一n( 10) temperature side and the width of peak enlarges.β E +2RTAccording to the analysis on the apparent kineticThe dynamic parameters were calculated byparameters determined by a nonlinear ftting algo-Garcia et al9)， Sφrum et all4J and Ferriol et rithm, the material components and heating rates aral 20. Both TG and DTG simulation methods were main factors afecting the kinetic characteristics. Theapplied. Compared with these results, the tendency is results中国煤化工rithm and experimentsimilar but the activation energy of this work is smal- are inr within the range ofYHCNMHGler. The difference is caused by the diferent thermo- given expenmeman Cutuouis.metric positions. The activation energy of reaction dcreases with the increase of heating rate. In the pyrol-Referencesysis process of MSW from different function districts，[ 1 ] Li Guojian, Zhao Aihua, Zhang Yi. Processing Engi-- 358-LI Haiying et al: Pyrolysis Characteristics and Kinetics of Municipal Solid Wasteneering of the Municipal Solid Waste[ M ]. ScienceStudy on pyrolysis characteristics and kinetics of mu-Press, Beijing ,2003.13- - 18( in Chinese ).nicipal solid waste [ J ]. Journal of Engineering Ther-[2] Wang Hua. The Technology of Zero Emission of MSWmophysics, 1999, 20(4 ): 509- 514( in Chinese ).[ M ]. Metallurgical Industry Press, Beijing， 2001.[ 12 ] Li Ji, Zhang Zheng， Yang Xuemin et al. TG-DSC1- 3( in Chinese ).study on pyrolysis characteristics of municipal solid[ 3] Wu Wenwei. The Use of the Municipal Solid Waste as .waste[ J ]. Journal of Chemical Industry and Engi-Resources [ M ]. Science Press, Bejjing, 2003. 12-neering ( China ), 2002, 53 ( 7 ): 759- -764 ( in16( in Chinese ).Chinese ).[ 4] Su Yujiang， Luo Liangmeng. Component analysis of[ 13] Zhang Yi, Tao Hua. Treatment Technology and Engi-urban household garbage of Luoyang[J ]. Urban En-neering of MSW [ M ]. Chemical Industry Press, Bei-vironment & Urban Ecology, 2002, 15 (5 ): 8-10jing, 2002. 141-147 ( in Chinese ).( in Chinese ).[ 14] Sprum L, Gronli M G, Hustad J E. Pyrolysis charac-[ 5] Li Aimin. A study of the pyrolytic semicoke character-teristics and kinetics of municipal solid waste[ J ]. Fuistics of solid waste [ J ]. Journal of Engineering forel, 2002, 80(9): 1 217-1 227.Thermal Energy and Power, 2002, 17 ( 2 ): 132-[ 15] Bockhorn H, Hornung A, Hornung U et al. Dehydro-139 ( in Chinese ).chlorination of plastic mixtures [ J ]. Journal of Ana-[ 6]Williams E A, Williams P T. Analysis of products de-lytical and Applied Pyrolysis, 1999, 49( 1-2 ): 97-rived from the fast pyrolysis of plastic waste[ J ].106.Journal of Analytical and Applied Pyrolysis, 1997, 40- [ 16] Urban D L, Antal M J. Study of the kinetics of sew-41( 5): 347- -363.age sludge pyrolysis using DSC and TGA [ J ]. Fuel,[7 ]Sivalingam G, Karthik R, Giridhar M. Kinetics of1982, 61(9): 799- -806.thermal degradation of poly ( ε-caprolactone ) [ J ].[ 17] Wen Lihua, Wang Shurong, Shi Haiyun et al. KineticJournal of Analytical and Applied Pyrolysis, 2003, 70study on the pyrolysis of wood[ J ]. Fire Science and(2): 631- -647.Technology, 2004, 23(1 ): 2- -5( in Chinese ). .I8]Kim S. Pyrolysis kinetics of waste PVC pipe [ J ].[ 18] Zhao Xuezhuang. Kinetic Principle of Chemical Reac-Waste Management, 2001, 21 ( 7 ): 609- -616.tion [ M ]. Higher Education Press, Beijing， 1984.[ 9] Garcfa A N, Marcilla A, Font R. Thermo-gravimetric29- 33( in Chinese ).kinetic study of the pyrolysis of municipal solid waste[ 19 ] Lyon R E. An integral method of nonisothermal kine-[ J ]. Thermochimica Acta, 1995, 254( 4 ): 277-tic analysis [ J ]. Thermochimica Acta， 1997， 297304.( 1-2): 117- -124.[ 10] LinKS, WangHP, LiuS H el al. Pyrolysis kinetics[ 20] Ferriol M, Gentilhomme A, Cochez M el al. Thermalof refuse-derived fuel [ J ]. Fuel Processing Technolo-degradation of poly ( methyl methacrylate )( PMMA ):gy，1999, 60( 2): 103-110.Modelling of DTG and TG curves[J ]. Polymer Deg-[ 11] Jin Baosheng， Zhong Zhaoping， Zhou Shanming.radation and Stability, 2003, 79( 2 ): 271-281.中国煤化工MHCNMH G