采用TG-FTIR联用研究烟煤热解及热解动力学参数的确定

第10卷第27期2010年9月科学技术与工程Vol 10 No 27 Sep. 2010l671-1815(2010)27-6642-07Science Technology and Engineering⊙2010 Sci Tech. Engng动力技术采用TG-rTIR联用研究烟煤热解及热解动力学参数的确定刘栗邱朋华吴少华张纪锋秦裕琨(哈尔滨工业大学燃烧工程研究所哈尔滨150001)摘要研究煤热解时的组分析出规律对进一步研究低NO燃烧或煤粉再燃时的均相NO,还原反应来说都是非常重要的。釆用 TG-FTIR实验装置对两种中国烟煤在不同升温速率(10,20,50和80℃/min)下的失重及气体释放规律进行了研究,并将实验数据与 FG-DVC软件的模拟结果进行了对比。通过对比发现,其中一种烟煤的模拟结果与实验数据比较相符,但另一种烟煤的模拟结果与实验数据偏差较大。偏差主要是由于FG-DvC模型中提供的有关动力学参数不准确所导致。基于 FG-DVC模型的假设,官能团热解形成的轻气体产物的释放过程可以用一系列平行独立的单方程模型描述,应用FTR的实验结果对热解气体组分的动力学参数进行了修正。采用修正后的动力学参数, FG-DVC能更准确的模拟该煤的热解过程。关鍵词煤热解 TG-FTIR动力学参数 FG-DVC中图法分类号TK6;文献标志码AThe evolution of volatile species during pyrand char. The DVC model is employed to determinehas a significant effect on coal combustionthe amount and molecular weight of macromolecularSince fragments. The lightest of these fragments evolv1970, some kinetic models such as single equation tar. Nine well-characterized coals were selected to formmodel, two-equation model, and Solomon's general an FG-DvC data base. The minimum input to the FG-model etc have been presented in order to simulate the dVC model provided by the user is the ultimate analyPyisis more reasotonably. Solomon@3)researched the sis of coal on dry ash free basis. If the H/C and O/Cslow pyrolysis process for several American coals, ob- atomic ratios for the interested coal fall into the grid oftained some kinetic parameters for gas evolution param- FG-DVC coal data base a pre-processor subroutine willeters for gas evolution and developed a general model generate the FG-DVC needed input files for the interest-called FG-DVC to simulate the pyrolysis process. This ed coal. Although the parameters used in FG-DVC codemodel takes account of the evolution of gases, tar, char were determined by slow pyrolysis process, Solomon etand adsorbed molecular in detail. The FG-DVC model al have compared the measured data with the FG-DVCcombines two models, one is the Functional Group model predictions and found the model can accurately(FG)model and the other one is the Depolymeriza- simulate the process of rapid pyrolysis under the condition, Vaporization, Cross-linking( DVC)model. The tion of American coal 4 5). Because of the difference beFG model is used to simulate the gas evolution and the tween the American coals and the Chinese coals, the kielemental and functional group compositions of the tar netic parameters for many Chinese coals usually cantbe中国煤化工ation scheme based010年6月21日收到,7月1日修改国家基金研究计划onthCNMHparameters of the(2006c8200303)和国家自然科学基金(50706011)资助期刘栗,等:采用TFmR联用研究烟煤热解及热解动力学参数的确定6643diagram( a plot of HC versus O/C atomic ratios)were CO, /N, with known concentration was mixed with n2sed to simulate and error will be observed. However, and then entered into the gas cell. The species con-there is seldom study on the quantitatively analysis of centration in the gas cell was changed by changing thethe gas evolution for domestic coals using TG-FTIRmixing ratio( gas mixture to n2). A method was thenThe aim of this work is to quantitatively analyze the py- established through polynomial fitting the known con-rolysis process for two Chinese coals and to determine centrationthe kinetic parameters related to the main species for thecoal which cant be simulate reasonably by FG-Dvc 2 Kinetie ModelAccording to the FG-DVC model, the evolution1 Experimentalof each species is assumed to be independent fromthe other species and the evolution rate can be repre1.1 Coal samplessented by a first-order rate with a Gaussian distribuTwo bituminous coals, Zhunger coal and Shenhua tion of activation energiespal, were dryinged at 50t for (4-5)h and then der reaction rate for release of the ith functionalng with caelian mortar before experiment. Thegroup(Xi, also called precursor pool in the FGcoal analysis of samples was shown in table 1DVC input files) can be expressed as followingTab. 1 The proximate and ultimate analysis of coal samples equate(1)adhunger coal2.7521.8228.2847.1577.415.0415.501.500.55And the rate constant k, in eq. (1)is given by anShenhua coal5.7110.5327,7556.0180.644.8512.501.430.58Arrhenius expression with a Gaussian distribution ofactivation energes:1. 2 TG-FTIR experimentk1=A,exp((-E±)/RT)(2)The pyrolysis of coal was performed at a thermo-Where A; is the pre-exponential factor, E, is thegravimetric analyzer(TGA/SDTA851")coupling with average active energy, o, is the width of the GaussianFTIR (Nicolet 5 700). The pyrolysis conditions were distribution and R is the gas constant. A non-isotheras follows: coal sample weight, 50 mg: gas atmos- mal method is used to obtain the kinetic paramephere, N2; pressure, 0.1 MPa; total gas flow ters[9-13), As the coal sample is heated at a constantthrough the furnace, 150 mL/min. After purging, theheating rate Hample was heated from room temperature to 105 C(at 10 C/min)for 20 min to dry it and then to 900(3)℃for20min(at10℃/min,20℃/min,50℃/ then the eq,(1) can be transformed to:min and 80C / min respectively). At the same timedx kXthe volatile species were introduced to FTIR for quaHtative or quantitative analysis. In order to make thewhich the rate of中国煤化工quantitative analysis, the gas cell must be calibratefor interested gas species. A gas mixture of CH/Co/ derivalNMHGual to zero, l.e.6644科学技术与工程10卷dt0. at T=T.-DTGEq (5)can be rewritten to the following form bysubstituting eq (1)and eq. (2)into eq. (5):dt000EE0 at T=T(6)Fig. 1 TG/DTG and simulated weight loss curvessquare brackets equals to zero, i. e.(7)DTG010.02It can be found from eq. (7)that thetion with I After the evolution rates fe106each species at different heating rates are measuredthrough experiment, the kinetic parameters then can beTemperature/Cdetermined from the slope and the intercept in eqFig. 2 TG/ DTG und simulated weight loss curves of(7). After A, and Ei are determined, o, and Y, can behenhua coal at50℃/minfitted to experimental data using a trial-and-error ap-Tab. 2 The characteristic parameters of coal samplesSampleT/(℃)W。/(w%)W。/(w%)(10-2w%s1)3 Results and discussionShenhua7.6531.13. 1 Thermogravimetrie characteristicThe TG/DtG curves and simulated weight loss byShenhua coal had higher R and lower Tr.ThisG-DVC code of Zhunger coal and Shenhua coal dur- meant that the active energy for Shenhua coal was smalling pyrolysis were presented in figure I and figure 2.The weight loss was increased as temperature goes The final weight loss calculated by FG-DVC (WL)up. After pyrolysis finished, Zhunger coal had higher also contained in table 2. From figurel, figure 2 andWL, than Shenhua coal, as shown in table 2. This may table 2, it was found that the calculation error forbe caused by higher Va in Zhunger coal. Meanwhile, Zhunger coal was very small, but the difference be-from the DtG curves, the R and T could be ob- tween TG curve and FG-DVC curve for Shenhua was中国煤化工10%. The coordi-CNMHGvelen diagram werein figure 3. It can be seen from the figure 3 that27期刘栗,等:釆用TG-FTR联用研究烟煤热解及热解动力学参数的确定6645the Zhunger coal coordinates in the van Krevelen dia- with higher heating rates. This was caused by the moregram fall into the grid of library-coal data, while the difference between coal sample and thermocouple withShenhua coal coordinates are far from the grid. So the higher heating rate The width of temperature related toFG-DVC input files for the Zhunger coal can be gener- CO, evolution also became bigger with the heating rateated by means of an interpolation scheme which is increased. This was caused by deeper overlapping de-based on the three surrounding coals database. And gree with higher heating ratehe input files for Shenhua coal are generated based onThe evolution curves of CH, at different heatingthe library coal most closely located in the van Krev- rate were shown in figure 5. Formation of CH, startedelen diagram. As the parameters in the input files are at about 340C and reached the maximum evolution atnot accuracy, the calculation error for Shenhua coal about525℃,556℃and577℃for20℃/min,50was bigger. The following section will focus on theC/min and 80 C/min respectively. The formation ofkinetic parameters in the input files and improve the CH finished at the end of the liner heating step and nolution of gases for Shenhua coal in order to modifyobvious formation of CHa was found during temperatureults eventually20℃min3983850℃min80C/mind High-Rank Coals0.050.100.150.20025/cFig. 4 Evolution rate curves of CO, during pyrolysis forFig 3 The van Krevelen diagram forhenhua coal sample at diferent heating ratesals and the library coals of the FG-DVC code00153.2 Evolution characteristic of gasesThe variation of gas evolution with temperature at001050℃/mindifferent heating rates could be seen from figure 4 to80℃/minfigure 6. For 10 C/min, evolution of CO, started at a-o2890005bout 180C. and reached the first maximum at about457C and then reached the second maximum around716C. The first peak appeared due to the decomposi2004006008001000tionTemperature/Chigher temperature was caused by more stable functiongroup such as ether. It was also found that the secondFig. 5 Evolution rate curves of CH, during pyrolysis forpeak was more important than the first peak exceptsample at different heating rates中国煤化工80/min curve. The Tm for both peaks at differentCNMHGheating rate were shown in table 3. The T_ was higher科学技术与工程10卷formula (7). The data in table 4 were in the input for-0035mat of the FG-DVC model. Y is the initial fraction of a10℃20℃/minparticular function group with the modified kinetic pa-￥00254-50℃/minrameters in the input files, the weight loss000-80cmiyield of gases during pyrolysis were recalculated℃/min.Tab. 4 The kinetic parameters for gasesrsor pool A/s" x10(Eo/R)K(o/R)/K004001-CO2-looe0.381323020000.0777903-C02- tight0.12229520000.530000Fig. 6 Evolution rate curves of Co during pyrolysis for321687818000.025215Shenhua coal sample at different heating ratesi8117holding step. As shown in table 3, the T also shifted10-C0- tight0.101938830000.031924to higher value for higher heating rate. The CHa wasmainly formed by the reaction in which the methylOriginal FG-DVCodified-FG-DVCchain and aliphatic bridges of a larger molecule werebrokeThe evolution curves of Co during pyrolysis wereshown in figure 6. Evolution of CO started at about280%c. For 20C/min, evolution curves reached thefirst peak at about 626.C and then decreased a littleCO also had two shoulder peaks. The second and max-imum peak was around 727C. The Co was releasedform the ether 0 group in the original coal. As sheFig. 7 Comparison of modified FG-DVC simulation onin table 3, the T_ also shifted to higher valueeight los with experimental result and original FG-DVCsimulation on weight loss at 80 C/minhigher heating rateTab. 3 The T for gas at different heating rates2 Experimental results Original FG-DVC80℃/minModified FG-DVC45668℃488.00℃510.94℃716.05℃757.90℃773.76℃CH4525.49℃556.06℃576.64℃663.71℃686.15℃726.64℃771.36℃79566℃l目3. 3 Modified FG-DVC ModelFi中国煤化工: simulation withThe kinetic parameters for gas, as shown in tableYHCNMH Guation at 80C/minwere determined by employed the value in table 3 to27期刘粟,等:采用 TG-FTIR联用研究烟煤热解及热解动力学参数的确定6647As shown in figure 7 and figure 8, the simulationReferencesement. Especially, the simulationyield of CO2 and CO fitted the experimental data very I Qiu P H, WuS H, Sun S Z,et al. Industrial test on coal re-buming atwell. It also can be found the simulated weight loss600 MW utility boiler and NO, reduction. Korean Joumal of Chemicalwas still a little more was still a little more TG resultsEngineering,2007;24(4):683-6872 LZQ,Jing J P, Chen Z C, et al. Combustion characteristics and NOxIt was because that the kinetic parameters were deteremissions of two kinds of swir burners in a 300-MWe wall-fired pul-mined only for CO2, CH, and CO and the kinetic pautility boiler. Combustion Science anrameters for other species such as Tar, H, and H, 0 2008: 180(7): 1370-1394were not available. This may be improved in later3 Solomon P R, Hamblen D G, Carangelo R M, e al. General model ofcoal devolatilization. Energy Fuels, 1988; 2(4): 405-4224 Dutton K. Functional-group, depolymerization, vaporization, cross linkingmodeadVaneedFuelReseachIne.http://www.afrine.com4 Conclusionsproducts/fgdve/default. htm, 20075 Li Xiaoli, Sun Rui, Zhang Xiao-hui, e al. Simulation study on NOThe pyrolysis experiment and numerical simulationreduction by volatiles from coal devolatilization. Proceedings of theCSEE2008;28(11):3035( in Chineeof two types of Chinese bituminous coal at different6 Yang Jingbiao, Cai Ningheng. A TG-FTiR study on catalytic pyroly-heating rates were performed using TG-FTIR analysissis of coal. Joumal of Fuel Chemistry and Technology, 2006; 34 (6):and FG-DVC model separately. The weight loss andthe evolution rate of CH4, Co and Co, during pyrolysis 1 Zhou Junhu, Ping Chuanjun, Yang Weijuan, et al. Experimentalwere measured. The thermogravimetric characteristicstudy on the pyrolysis characteristic o coal blends using TGA-FTIR.and the evolution characteristic of gases were obtainedJoumal of Fuel Chemistryand Technology, 2004, 32(6): 658-662(infrom the experiment data.8 Solomon P R. Hamblen D G, Serio M Aet aL. A characterization(1)Shenhua coal had smaller active energy andmethod and model for predicting coal conversion behaviour. Fuelmore weak aliphatic chains than Zhunger coal(2)CO2 and CO had two shoulder peaks during9 de Jong W, Pirone A, Wojtowicz A. Pyrolysis of Miscanthus Giganevolution process while CH, had only one peak aroundteus and wood pellets: TG-FTR analysis and reaction kinetics. Fuel2003;82(9):1139-114550C. The T shifted to higher value for higher heat- 10 de Jong W, Di N C, Vemneker B C H,etal. TG-FTIR pyrolysising rate. The width of temperature related to gas evolucoal and secondary biomass fuels: Determination of pyrolysis kinetiction also became bigger with higher heating rate.parameten for main species and NOr precurson, Fuel, 2007: 86(3)FG-DVC model can simulate the pyrolysis(15):2367-237611 Braun R L, Bumham A K. Analysis of chemical reaction kinetics u-very well forfor Zhunger coal, but the difference beween experimental data and FG-DVC curve for Shen& Fuels,1987;1(2):153-161hua was obvious12 Solomon P R, Seno M A, Carangelo R M, et al. Analysis of the Ar-(4)The kinetic parameters for CH,, CO and COgonne premium coal samples by thermogravimetric Fourier transformwere obtained from experiment data. The FG-DVCinfrared spectroscopy. Energy Fuels, 1990; 4(3): 319--333model was modified with the calculated kinetic parame-13 Wang Hui, Jiang Xiumin, Yuan Dequan, e al. Pyrolysis o coalwater slurry volatile matter by using FG-DVC model. Journal ofters of Shenhua coal. Using the modified model, the中国煤化工(10):2428-2432(inumerical simulation fitted the experimental resultsCNMHGmore reasonably for Shenhua coal下转第6652页)6652科学技术与工程10卷3 Karimierczk M K. Transfer function of current modulator in PWM4阮新波严仰光直流开关电源的软开关技术北京:科学出版mental Theory and Applications, IEEE Transactions∞m, Volume:475谭阳红蒋文科,何怡刚,基于OCAD10.5的电子电路分析与设lsue:9,Sept,20001407-1412计北京国防工业出版社,20The Simulation Research of Buck-Boost ConverterLI Xue-JiWeifang Oil Transportation Station, Pipeline Storage and Transportation Corporation, SINOPEC, Binzhou 256600, P. R. ChinaAbstract] PSpice is a powerful simulation software, simulation results are very close to the true state of the cir-cuit. The overall working stages of Buck-Boost converter is simulated and analyzed by PSpice. The working processBuck-Boost circuit includes the transient process of start-up circuit and the steady working process. All thestages of stored energy elements of Buck-Boost converter are also introduced. The large number of visual simulationwaveforms are given. Thus the understanding of Buck-Boost converter is deepenedKey words] Buck-Boost converter Pspice transient analysis steady-statey(上接第6647页)Experiment Research on Bituminous Coal Pyrolysis byTG-FTIR and Determination of Pyrolysis Kinetic ParametersLIU Li, QIU Peng-hua, WU Shao-hua, ZHANG Ji-feng, QIN Yu-kunCombustion Engineering Research Institute, Harbin Institute of Technology, Harbin 150001, P. R. China)Abstract] It is significant to study the components and the relevant concentration of volatile matters released dur-ing pulverized coal pyrolysis, which is fundamental for the further study of low NO, combustion and NO, reductionduring coal rebuming process. The devolatilisation experiments of two types of Chinese bituminous coal were per-formed using TG-FTIR(Thermogravimetry combined with Fourier Transform Infrared Spectroscopy analysis. Fourheating rates(10, 20, 50 and 80C/min)were adopted to research the weight loss and gases evolution. The numeri-cal simulations were performed by using FG-DVC( Functional Group and Depolymerization, Vaporization, Cross-linking)model on the experimental coals. It was indicated that the simulation results were well fitted for one of the two typesof coal but not very well for another. The emor was caused by the inaccuracy of the kinetic parameters of the main spe-cies provided by FG-DVC model. The kinetic parameters are then corrected by introducing FTIR results to a series offirst-order formulation by assuming that the light gases evolution are parallel and independent in FG-DVC model. Byadopting the corrected kinetic parameters the simulation results ar中国煤化工h betterI Key words] coal pyrolysis TG-FTIR kinetic parameteCNMHG