Availableonlineatwww.sciencedirect.comJOURNAL OFDcienceireciCHEMISTRYELSEVIERJournal of Natural Gas Chemistry 21(2012)571-580www.elsevier.com/locate/jngoOxidative reforming of methane for hydrogen and synthesis gasproduction: Thermodynamic equilibrium analysisAntonio C D. Freitas, Reginaldo guirardelloinstein 500, 13083-852, Campinas-SP BrazilManuscript received November 30, 2011; revised February 6, 2012 1abstractA thermodynamic analysis of methane oxidative reforming was carried out by Gibbs energy minimization(at constant pressure and temperature)and entropy maximization (at constant pressure and enthalpy) methods, to determine the equilibrium compositions and equilibriumtemperatures,respectively. Both cases were treated as optimization problems(non-linear programming formulation ) The GAMS@ 23. 1 software and the CONOPT2 solver were used in the resolution of the proposed problems. The hydrogen and syngas production were favored athigh temperatures and low pressures, and thus the oxygen to methane molar ratio(O2/CHa)was the dominant factor to control the compositionof the product formed. For O2/CH4 molar ratios higher than 0.5, the oxidative reforming of methane presented autothermal behavior in thecase of either utilizing O2 or air as oxidant agent, but oxidation reaction with air possessed the advantage of avoiding peak temperatures inthe system, due to change in the heat capacity of the system caused by the addition of nitrogen, The calculated results wered withpreviously published experimental and simulated data with a good agreement between therKey wordsthermodynamic analysis; methane oxidative reforming; Gibbs energy minimization; entropy maximization; hydrogen and syngas production1. Introductionproduces a H2/CO molar ratio around two, which is more adequate for Fischer-Tropsch synthesis [11]. The main reactionsB Res In recent years, hydrogen has been attracting great in- in the oxidative reforming of methane are illustrated in fol-terest as a clean fuel for combustion engines and fuel cells lowing equationsI. Among all the potential sources of hydrogen, natural gas,hich has methane as its main component, has been considCH4+202→CO2+2H2O△H98=-8023 k/mol(1)ed as a good option because it is clean and abundant, andcan be easily converted to hydrogen [2].CH4+0.502→CO+2H2△H298=-35.7kmol(2)Synthesis gas(Syngas, a mixture of hydrogen and carbonmonoxide)is of great importance as a major chemical interCH4+O2→CO2+2H2△H298=-3189kJ/mol(3)mediate in chemical processes for the synthesis of several fuels and chemicals [3-7]. The syngas can be used in Fischer-CO+H2O“CO2+H2△H29g=-41.0kmol(4)Tropsch synthesis(FTS), which can produce a large variety ofhydrocarbons ranging from light gases to heeavy waxCH4+H2O←CO+3H2△H298=206.0k/mol(5)Steam reforming of methane is the main industrial routeto produce hydrogen and syngas [9, 10]. This reaction proCH4+CO2←2C0+2H2△H298=2470 kJ/mol(6)duces the syngas with a H2/Co molar ratio equal to three,which is very high when compared to other reforming proCO+H2 C(s+H20 AH298=-1310 kJ/mol(7)cesses for the application in FTSOxidative reforming of methane has been investigated asCH4“C(s)+2H2△H298=74.85 kj/mol(8)n alternative to the process of steam reforming. The par-tial oxidation of methane is a slightly exothermic reaction that2C0CO2+Crel AH2og--172.4 kJ/mol (9)Corresponding author. Te: +55-19-35213955; Fax: +55-19-35213910: E-mail: guira @feq中國(guó)煤化工This work was supported by CAPES-Coordenacao de aperfeicoamento de Pessoal de eHCN Glo Nacional de Desen-volvimento Cientifico e Tecnologico-Brazilrighto2012, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. All rights reserved0.1o6S1003-9953(11)604064572Antonio C. D. Freitas et al./ Journal of Natural Gas Chemistry Vol. 21 No 5 2012where, All reactions involving O2 are, for practical purposes,The oxidative reforming of methane typically occurs inthermodynamically irreversible. Hydrogen and syngas pro- low or moderate pressures(1-15 atm) and high temperaturesduction varies significantly according to the operating condi-(above 1000 K)[7]. The hypothesis of ideal gas(pi= 1),thermodynamic analysis provides important knowledge about the pure solid compound in vapor phase was admitted tion etions such as pressure, temperature and reactant ratio. The absence of liquid phase and rejection of the molar fractionthe effects of those variables on the reforming process andEquation(10), therefore, can be simplified, and the Gibbsmakes it possible to predict the technical and economic feasi- energy can be expressed as follows:bility of the process [ 12].In this paper we report the thermodynamic analysis ofmethane oxidative reforming reaction, where Gibbs energyG=∑雞(°+m(mP+(mn-n∑minimization(in conditions of constant pressure(P)and tem-perature(T) was employed to calculate equilibrium compo-ions and entropy maximization(at constant P and enthalpy∑nA°(H)) was employed to determine the equilibrium temperature(15)of the reaction. The effect of processes variables such as presIn the gibbs energy minimization, the calculations weresure, temperature and reactant ratio was studied. The catalytic performed considering two different situations: one is whereeffect of inhibition of coke formation was evaluated under the coke may or may not be formed, depending on the reactionequilibrium compositions, utilizing the gibbs energy mini- conditions( thermodynamic controlled ) and the other is wheremization. The thermal effect of use of air as oxidant agent coke is not allowed to be formed in any situation, in order towas evaluated by entropy maximizationinclude the catalytic effect of inhibition on the coke forma-tion. Comparing these two situations is interesting, because2. Thermodynamic modelavoiding coke formation extends the catalyst life in reformingprocesses [14-19] when the inhibition effect is not enough2.1. Equilibrium at constant P and T: formulation as a prob. and the use of adequate operating conditions may naturallylem of minimization of Gibbs energyavoid coke formation( thermodynamic controlled)The thermodynamic equilibrium condition for reactive2. 2. Equilibrium at constant P and H: formulation as a probmulticomponent closed system, at constant P and T, with lem of entropy maximizationgiven initial composition, can be obtained by the minimization of Gibbs energy (G)of the system, with respect of theThe thermodynamic equilibrium condition for reactivenumber of moles of each component in each phase, given by: multicomponent closed systems, at constant P and H, withgiven initial composition, can be obtained by maximization ofmG-∑嗎4+∑n+∑nA(0) entropy(S)of the system, with respect to ngi=1While satisfying the restrictions of non-negative numbermaxs=∑ns+∑nS+∑nS;(16of moles of each component in each phasen2,n1,n2≥0(11)The same above mentioned restrictions of non-negativityof number of moles(Equation 11)and mole balances(EquaAnd the restriction of mole balances, given by atom bal- tion 12) should be satisfied here too. Usually, physical properance for reactive systems:ies are given as functions of composition, pressure and temperature, not enthalpy. Therefore an additional restriction, ref-∑am(n}+n+2)=∑amn?m=1,…,NE(12) erent to enthalpy balance, must be satisfiedSmith and Missen [13] demonstrated that the stoichiomet-∑(m2H+n班+nF)=∑(n9B9=H(17ric formulation is equivalent to the non-stoichiometric onei=1provided that all independent reactions are consideredThe entropy of each component in the mixture and theThe values of pg can be calculated from the formation enthalpy balance can be determined using the following ther-values given at some reference conditions, using the followmodynamic relationsing thermodynamic conditionsaH中國(guó)煤化工1,…,NCCp;=1,…,NC(13)CNMHG(篇)。=m1=1NC040i=1.....NCP咚hemistry VoL. 21 No 5 2012The hypothesis of ideal gas (oi=1), the absence of liq- namic analysis of methane oxidative reforming [1, 29-311uid phase and rejection of the molar fraction of the pure solid none of them realized a complete thermodynamic study ofcompound in vapor phase were admitted, so the entropy can these reaction, evaluating the effects of pressure temperabe expressed as followsture and molar composition of the reactants under the productivity and thermal behavior of the reaction. Works usingthe methodology of entropy maximization for thermodynamics=∑?2s"-R(mnP+(mn2-hn∑nanalysis of methane oxidative reforming were not found in theliterature∑nSAlthough the Gibbs energy minimization and entropymaximization methods are equivalent, the formulation of the(20)optimization problem in form of entropy maximization is veryIn the entropy maximization method the thermal effect ofinteresting for evaluating mixtures that could result in hotuse of air as oxidant agent will be analyzedspots in reactors undergoing exothermic reactions [27]During the process of optimization, utilizing the GibbsSince the system was analyzed at low pressure and highenergy minimization method the number of moles of the temperature, generally, it can be considered as ideal, and thegaseous(n), liquid(nd )and solid (n) phase are considered solid phase formed is considered as pure component. Anddecision variables, while T, P and the chemical potential of such consideration showed good results in previous worksthe pure component in the reference state (w!)are considered (12, 26, 27,32, 33]parametersIn this work, the thermodynamic analysis was carried outIn the maximization of entropy, the variables are n, ni, over the following variable ranges: pressure 1, 5 and 10 atm,n, Tand all quantities that depend on them, such as physical initial temperature 600 at 1600 K, an oxygen/methane molarproperties of pure components(which depend on temperature) ratio (OCR)between, O 1/1-1/1 and an air/methane molar ra-and molar fractions. The parameters are physical properties of tio(ACR)between 0.5/1-5/1(with an air molar compositionpure components at some reterence temperature, and the ini- of 80%N2 and 20%O2)tial molar amount(n!)3. Results and discussion2.3. Numerical procedureA thermodynamic analysis based on Gibbs energy minAlthough the formulated problem is non-linear, the usedimization and entropy maximization was carried out fomethodology guarantees the global optimum, since in this methane oxidative reforming. The thermodynamic equilibcase the problem is convex [20, 21]. GAMS@ 23. 1 software rium calculations showed a low computational time(in allGeneral Algebraic Modeling System) with CONOPT2(Con- cases less than 1 s)tinus optimizer version 2)solver executed in a Pentium IllThe main species in the methane reforming processes are(512 MB, 900 MHz), was used in the resolution of the calcu- CH4, COz, co, O2, H2, H20 and solid carbon(Cs)[34,35lation problem of chemical and phase equilibrium.in this work, N2 was inserted to represent air in the simula-The method of thermodynamic analysis by minimization tions when it was considered as oxidant agent. The thermo-of Gibbs energy is commonly employed and others works use dynamic data used in the calculations(Cpa equation paramthis technique with excellent results [22-27]. The entropy eters Cpai, Cpbi, Cpci and Cpdi, the standard enthalpy ofmaximization methodology is less used, but recent works formation(AHM)and the standard gibbs energy of formationshow good results in applications of this technique [27, 28AG)were taken from Refs. [36-39] and are presented inAlthough, some works have performed the thermodTable 1Table 1. Thermodynamic data for the components considered in the simulationsComponentCpb×10Cpc1×105Cpd1×105△ Hf (/mo)△Gr(mo)9081745200.2275457045-3943590.5570031-110525l37169H203470.12yH24181822857204220.083中國(guó)煤化工0.593CNMHGTm: temperature limit of Cpa expression. P=Cpa, +Cpb, xT+ Cpc; XT2+Cpd, xT-2574Antonio C. D. Freitas et al/ Journal of Natural Gas Chemistry VoL. 21 No. 5 20123.1. Results for Gibbs energy minimization: equilibrium cein the system had no significant influence on H2 productiotionIf OCR is above 0.5, the mole of H2 produced began to decrease as O2 was added to the system. Figure 1(b)shows thatFigure I depicts the thermodynamic equilibrium results CO production increased with O2 addition until an OCR offor mole of H2, CO, solid carbon produced and CHa conver- 0.5, which reached its maximum. If OCR was 0.75 or 1, thesion(%), respectively, at different temperatures and OCRs. production of Co decreasedAll calculations were performed by the method of minimizaFrom the analysis of Figure 1(c), we can see that the adtion of Gibbs energy, allowing the formation of solid car- dition of O2 and high temperatures caused a decrease in cokebon,at a constant pressure of l atm. The calculation of CH formation, and total inhibition of Cs formation began in theconversion was carried out utilizing the following definition OCR of 0.5 and temperature of 1200 K. CHa conversion in-(Equation 21):creased with the increases in the temperature and OCRs(seeCH4 conversion(9%)=1CH -CHAFigure ld). Similar behaviors were obtained in simulationsx 100%(21) realized by Enger and coworkers [40 by finding the iof the conversion of methane with an increase in ocR in theFigure 1(a)shows that if OCR was below 0.5, O2 addition feed, and with the rise of reaction temperature2.51.006-4-0.250.50一0.250.501400Temperature(K)Temperature(K●0.100.250.250.500.500.751110001400Figure l. (a) Moles of H2, (b)moles of CO, (c) moles of coke and(d)CH4 conversion. Results obtained for methane partial oxidation reaction allowing theformation of solid carbon at 1 atmThe effect of pressure increase on the number of moles of tures above 1000 K, as we can see in Figure 2(c)hydrogen produced can be seen in Figure 2(a). From the anal-A comparison between the results obtained for the simuysis of this figure we can see that H2 production decreased lations realized avoiding and allowing the coke formation iswhen the pressure increased. Similar effects were observed show in Figure 3. Figures 3(a),(b)and(c) show the comparfor moles of Co produced and CH4 conversion(see Figureison for the中國(guó)煤化工 and ch4 conversion2(b)and(d), respectively). Similar results for the effect of respectivelylpressure elevation were observed by Enger et al. [40]. The 0.5 in the feeCAMH Ged using an OCr ofwo oulu because it showed thecoke formation showed a reverse behavior, i. e, the moles of most significant productions of H2 and CO, and enabled thecoke formed increased with increasing pressure, at tempera- inhibition of coke formation at temperatures above 1200 K, asJournal of Natural Gas Chemistry Vol 21 No, 5 2012575shown above. Through the analysis of Figure 3(a)we see that figure the number of moles of co produced by the processthe number of moles of H2 produced was greater for the sys- simulated avoiding Cs formation is greater than that producedtem simulated allowing Cs formation, and this behavior could by the simulated allowing the Cs formationbe explained by the occurrence of the methane decompositionFor temperatures above 1200 K, the behavior of the tworeaction(Equation 8)simulated systems tended to be equal. This behavior was extended to be higher for the system simulated allowing Cs for- formation was achieved by the thermodynamic effec? h of CsEvaluating this reaction we can see that H2 production pected, since for these reaction conditions, the inhibitionmation. The occurrence of this reaction also explained theAs we can see from Figures 1, 2 and 3, lower pressures,behavior observed for CHa conversion too(see Figure 3c)lesser intermediates or lower OCRs combined with higherIn Figure 3(b)we can see the comparison of moles of Co temperatures increased the hydrogen yield in methane oxida-produced by the two processes simulated. By evaluating this tive reformingI atm:10 atm0.0L1600Temperature(K)0.25(c)800205I at00510 atm0.0012001400Temperature(K)Figure 2. Effect of increased pressure on(a)moles of H2, (b)moles of CO, (c)moles of coke and (d) CHa conversionwith cWith CWith CWithout Cwithout中國(guó)煤化工--- without C0080010001200140016006008001000120014CNMHG1200140160Temperature(K)erature(K)Figure 3. Comparison of the simulations realized allowing and preventing the coke formation with O2/CH=0.5/1.(a)Moles of H2 produced, (b)moles of COproduced, (c) CHa conversion576Antonio C. D. Freitas et al./ Joumal of Natural Gas Chemistry VoL. 21 No. 5 20123.2. Results for entropy maximization: equilibrium tempera- ing O2 or air as oxidants, can be seen in Figures 4(a)and(b),turesrespectively. Systems simulated considering O2 or air had asimilar behavior, but final temperatures for the system simuThe results obtained for equilibrium temperatures, deter- lated with O2 were higher than the system simulated usingmined using the method of entropy maximization, consider- as oxidant, with the same concentration of O2 in the system4000OCRACR0.10-0.750.2525001.252.5023000+3.75蘭&E5500L114001600nitial temperature(K)Initial temperature(K)igure 4, Final temperature for the system simulated (a)utilizing O2 as oxidant agent and(b)utilizing air as oxidant agentEffect of the reduction of final temperatures when air is air, and this effect was observed in experiments carried out byutilized as an oxidant agent is show in Figure 5(a). for an OCR Jiet al. [41]. Effect of the use of air proved to be interesting toof 0.5 and an ACR of 2.5 (O2/CH4=0.5/1). The reduction of decrease the final temperature of reaction and to prevent thefinal temperatures was associated with the presence of N2 in occurrence of peak temperatures in reactorsO/CH, 0.5/13000l4001200 F With O,1500T=1400K1000With air100012000.20.60.8Initial temperature(K)o, CH molar ratioFigure 5. Comparison of the simulations realized considering O2 and air as oxidant agent for an OCR of 0.5 and an ACR of 2.5(a)for different initialtemperatures and(b)at constant initial temperature of 1400 K for different O2/CH4 molar ratios in the feedAs can be seen from Figure 5(b), when raising the concen. molar ratios higher than 0.5 were used, in all simulated casestration of O2 in the feed, and keeping the initial temperature (using 2 or air as oxidant agent ). This behavior can be seenconstant(1400 K), there was an increase of the final tempera- in the Figure 5(b). The exothermic behavior of the oxidativeture. This behavior could probably be explained by Equations reforming of methane was one of the main advantages of this(I),(2)and (3). These reactions were exothermic and their process and this characteristic was emphasized by others studoccurrence was favored by increasing the concentration of O2 ies in literati中國(guó)煤化工concentration higherin the feed, thus increasing the energy released during the re- than 0.5, thecontrol in practice,actionsince theCN MH Gasy to be safely andIt was also interesting to verify that the oxidative reform. carefully controlleding of methane presented autothermal behavior when O2/CH4The effect of pressure increase in the final temperaturesJournal of Natural Gas Chemistry VoL 21 No 5 2012577is shown in Figure 6, as a function of initial temperature for were observed for lower initial temperatures. With increas-methane oxidative reforming with O2(a)and air(b)for an ing initial temperature of the reaction, the pressure increaseconstantO2/CHA molar ratio of 0.5/1(OCR of 0.5 and ACR of showed a less significant influence on the final temperatures2.5)in the feed. Systems simulated with Oz and air presented determined.similar behavior. In both cases, most significant changesOCR=0.50ACR2.501800140021200110 atr0010 ab1600Initial temperature(K)Figure 6. Effect of pressure on the final reaction temperature for methane oxidative reforming with O2(a)and air(b)3.3. Comparison with experimental and simulated data forThe simulations were performed in the following con-isothermal reactorsitions: temperature between 700 and 1200 K, atmosphericpressure and an OCR of 0.5. The experimental runs of Dal3.3.1. Comparison with experimental dataSanto et al. [43 Hong and Wang[44], Lanza et al. [45]andOzdemir et al. [46] were performed in the same conditions ofFigure 7(a) and(b)shows a comparison between sim- pressure and Oz concentrationulated and experimental data for methane conversion andThe comparison with experimental data for methane conH2/CO molar ratio, respectively. The experimental data for version and H2/Co molar ratio in the product, allowed usmethane conversion was obtained in Refs. [43-46]. The to verify that Gibbs energy minimization methodology pos-experimental data for H2/CO molar ratio was obtained in sessed a good predictive ability for determine the compositionRef.[46]of this type of system.Simulated dataet al. [46]合aa Hong and Wang(44 nickel nanowire catalyst△ Dal Santo et al.[43Hong and Wang [44] metallic nickel catalystYH中國(guó)煤化工⊥923723CNMHGFigure 7.(a) Comparison between simulated and experimental data for partial oxidation of methane, (b)comparison between simulated data and experimentalta for H2/Co ratio578Antonio C. D. Freitas et al/ Joumal of Natural Gas Chemistry VoL. 21 No. 5 20123.3.2. Comparison with simulated dataculated at a constant pressure of l atm and at constant tem-perature of 873 K, and OCR was evaluated in a range fromIn Table 2, we can view a comparison of the data pre- 0.5 to 1.5sented in the works of Avila-Neto et al. [35] and Zhu etWe can see that, by comparing the data presented in Taal.[47], both calculated by Gibbs energy minimization, and ble 2, the data calculated by the present work were in reason-the data calculated by this work using the same methodol- able agreement with the data found in the literature that usedogy, allowing and avoiding Cs formation. The data were cal- the same methodology.Table 2. Comparison of the simulated data of Avila-Neto et al. [35], Zhu et al. [47] and the present work allowing and avoiding Cs formationO2/CHa in the feed150Lit.1 Lit. 2Calc 22 Lit. 1 Lit. 2 Calc. 1 Calc. 2 Lit. 1 Lit. 2 Calc. 1 Calc. 2CH4021440.1560.1400.2120.0600.0230.05900590005000000050005070.0780.1740.1090.2370.1980.24202420.4090401041204120.1550.2000.1060.1590.1180.1480.120.1230062006300660066CO20.10700780.0940.1030.1950.1750.1900.19002700.2680.2660266041704880.48404160.3900.4560.3860.254Lit.1:simulated data from Avila-Neto et al. [35]: Lit. 2: simulated data from Zhu et al. [47]: Calc. 1 data calculated by the present work allowing Csformation; Calc. 2: data calculated by the present work avoiding Cs formation3.4. Comparison with experimental and simulated data forIn Table 3 we can see that the method of entropy maxthermal profile of reactorsimization for systems with constant P and H showed goodpredictive ability to determine the final temperatures of the3.4. 1. Comparison with experimental datamethane oxidative reforming reaction, in both cases, either usO2agent.In this section the data calculated by the present work us-ing the methodology of entropy maximization were compa3.4.2. Comparison with simulated datawith experimental data in order to verify the predictability ofthe methodology applied against real data from temperatureFigure 8 shows a comparison between the data obtainedprofiles for the reaction of partial oxidation of methane.in the work of Reyes and coworkers [48] calculated for an adiTemperature profile data for the reaction of partial oxida- abatic oxidative reforming reactor and the data calculated bytion of methane with air or O2 were obtained in Ref. [41].Ji the present work utilizing the entropy maximization methodand coworkers [41] analyzed the reaction of oxidative reform-The simulations were performed with a constant O2/CH4ing of methane using O2 or air as oxidants. The experiments molar ratio of 0.5/1 and at a constant pressure of 1 atm forwere performed in a fixed bed reactor with a nickel based cat- a temperature range of 573.15 to 87315K, for the oxidationalyst. A CHa/O2 molar ratio of 2/1(equivalent of an OCR of reactions performed with O2 or air0.5 and an aCr of 2.5 in this work) and atmospheric pressureAs can be seen in Figure 8, the data calculated by Reyeswere used by the authorsand coworkers [48] were in good agreement with the data calThe initial temperature used for the oxidation reactionulated by the present work in both cases(utilizing O or airwith O2 was 580C(853. 15 K), and as for the oxidation with as oxidant agent)air was 646C(919.15 K). The equilibrium temperatures de-termined by the experiments of Tanaka et al. [42], the finaltemperatures determined by this work utilizing the entropymaximization and the calculated percentage deviations(cal1150culated using Equation 22)can be seen in Table 3.1100Deviation100%Table 3. Comparison of equilibrium temperatures calculated bythe methodology of entropy maximization for the methaneoxidative reactions with Oz or air and the experimentaldata obtained in Ref. 41]Oxidant agent Ti (K)(K)T“(K) Deviation(853.151167.151172.170.56中國(guó)煤化工(919.151093.15111097Figure 8. CCNMHG Reyes et al. [48] for finalTi: initial reaction temperature; Tf P final experimental tempera-temperature in the methane oxidative reforming with O2()and air(o), andture determined by Jiet al. [41]; Tcalc final temperature determinedthe data calculated for the present work utilizing the entropy maximizationby entropy maximization method in present workethod for methane oxidative reforming with O2(..)and air(-)Journal of Natural Gas Chemistry VoL. 21 No. 5 20124. Conclusionsne Number of elements in the systemR Universal gas constanta thermodynamic analysis of methane oxidative reform- s Entropying for hydrogen or syngas production has been performed. Sk Entropy of component i in phase kThe more favorable conditions, determined by G minimiza- PRessuretion for H2 production are as follows: temperature above t Temperature1200 K, atmospheric pressure and low OCRs, in which conditions the number of moles of H2 produced is high; for syn- Greek lettersgas production the better conditions determined are: temper- Wi Chemical potential of component i in phase kature above 1200 K, at atmospheric pressure and intermediateOCRs, in which conditions the production of carbon monoxSuperscripts and subscriptside( co)is increased and the H2/Co ratio is close to 2o Reference stateThe results of G minimization verified that the use of cat0 Inicialalysts that inhibited or decreased coke formation was interest-f Firing in promoting the production of syngas, but, the amount of g Gas phaseproduced hydrogen tended to decrease when the effect of thisk Phase in the elementtype of catalyst was considered by inhibiting the occurrence 1 Liquid phaseof methane decomposition reactionThe S maximization showed that if oCr was higher than i Component in the mixture0.5, in the whole range of examined temperature, the oxida- m Elements in component itive reforming reaction of methane was exothermic, the sametrend was observed in the oxidation with air but oxidationreaction with air had the advantage of avoiding peak temperaReferencestures in the system due to the energy released in reaction, andthis effect may be associated with the presence of nitrogen in [1 Ayabe S, Omoto H, Utaka T. Kikuchi R Sasaki. K. Teraoka.guchi K Appl Catal A, 2003, 241The comparison with experimental and simulated data al- [2] Dias J A C, Assaf JM. J Power Sources, 2004, 130:106lowed us to verify that the methodologies employed by the [3] Edwards J H, Maitra A M. Fuel Process Technol, 1995,42:269present work showed a good predictive ability and were in 14 Knifton JF. JAm Chem Soc, 1981, 103: 3959ell agreement with similar methodologies found in litera- [5] Xu B Q, Sun k Q, Zhu Q M, Sachtler W M H. Catal Today2000,63:453The methodologies used and applied in the software [6] Yang Y C, Liu x Q, Luo S Z, wu Y T, Jia CX, Li S E. FuelGAMSG23. 1, and solved with the solver CONOPT2 provedEnergy Abstr, 1999, 40(3): 197[7]Rostrup-Nielsen J R. Catal Today, 2002, 71: 243to be quick and effective in the resolution of the proposed [8] Sie S T, Krishna R Appl Catal A, 1999, 186: 55problems, with computational time inferior to I s in all cases [91 al-Oahtani H. Chem Eng / 1997, 66: 51analyzed[10] Pedemera MN, Pina J, Borio DO. Chem Eng J, 2007, 134: 138[11] Corbo P, Migliardini F Int J Hydrogen Energy, 2007, 32: 55Acknowledgements[12] Rossi C CR S, Alonso C G, Antuneso A C, guirardello RThis work was supported by CAPES-Coordenacao deCardozo-Filho L Int J Hydrogen Energy, 2009, 34: 323Aperfeicoamento de Pessoal de Ensino Superior-Brazil and CNPq (133 Smith WR, Missen R w Chemical Reaction Equilibrium AnalConselho Nacional de desenvolvimento Cientifico e tecnologicosis: Theory and algorithms, original edition by John wileyBrazilSons, reprinted edition by Krieger Publishing Company, 1982[14] Dong WS, Roh HS, Jun KW, Park S E, Oh Y S Appl Catal ANomenclature2002226:63Cpai Cp equation parameter for component i[15] Dong WS, Jun K W, Roh H S, Liu Z W, Park S E Catal Lettami Number of atoms of element m in component i2002,78:215Cpbi Cp equation parameter for component i[16] Laosiripojana N, Assabumrungrat S. Appl Catal A, 2005, 290Cpc: Cp equation parameter for component iCpi Heat capacity for component i[17] Souza MMVM, Schmal M. Appl Catal A, 2005, 281: 19[18]Chen ZX, Grace JR, Lim C J, Li A W Int J Hydrogen EnergyCpdi Cp equation parameter for component i2007,32:2359GGibbs energy[19) Koh ACw, Chen L W, Leong WK, Johnson B FG, KhimyakHLinjiEnorm an 32: 725Hk Enthalpy of component i in phase k[20] White W中國(guó)煤化工 Chem Phys,195.,28nk Number of moles for component i in the phase kCNMHG[21] McDonaldastA. na cng inem Res, 1995, 34: 1674n, Initial number of moles for component i[22] Castillo J, Grossman I E Comput Chem Eng, 1981, 5: 99Nc Number of components in the system[23 Mather A E Fluid Phase equilib, 1986, 30: 83Antonio C. D. Freitas et al Journal of Natural Gas Chemistry Vol. 21 No. 5 2012[24]Castier M. Rasmussen P, Fredesnslund A Chem Eng Sci, 1989, [36] Poling B E, Prausnitz J M, O'Connell P J. The Properties of44:237Gases and Liquids. 5th ed. New York: MC Graw Hill, 2000[25] Souza A T, Corazza M L, Cardozo L, Guirardello R, Meireles [37] Reid R C, Prausnitz J M, Poling B E. The Properties of GasesMAA. J Chem Eng Data, 2004, 49: 352and Liquids. 4th ed. New York: MC Graw Hill, 1987[26]Adhikari S, Fernando S, Haryanto A. Energy Fuels, 2007, 21: [38]Reid R C, Prausnitz J M, Poling B E. The Properties of Gasesand Liquids, 3rd ed. New York: MC Graw Hill, 1977[27] RossiCCR S, Berezuk M E, Cardozo-Filho L, Guirardello R. [39] DIADEM Public v 1.2CDIPPRCDesign Institute for PhysicalComput Chem Eng, 2011, 35: 1226Property Data Information and Data Evaluation Manager, 2000[28] Castier M. Fluid Phase Equilib, 2009, 276: 7[40] Enger B C, Lodeng R, Holmen A. Appl Catal A, 2008, 346: 1[291 Papp H, Schuler P, Zhuang Q. Top Catal, 1996, 3: 299[41] Ji YY, Li Z, Chen Y X, Yu C Y, Yu FY, Xu H YJ Nat GasChem,200,9:291[30 Ostrowski T, Giroir-Fendler A, Mirodatos C, Mleczko L CatalToday,1998,40:181(42 Tanaka H, Kaino R, Okumura K, Kizuka T, Tomishige K Jcatal.2009.268[31] Newson E, Truong T B Int J Hydrogen Energy, 2003, 28: 1379 [431 Dal Santo V, Mondelli C, De Grandi V, Gallo A, Recchia S,Sordelli L, Psaro R. Appl Catal A, 2008, 346: 126Energy,200,25:47[44] Hong X B, Wang Y Q J Nat Gas Chem, 2009, 18: 98[33] Vasudeva K, Mitra N, Umasankar P, Dhingra S C. Int J Hydro- [45] Lanza R, Canu P, Jaras SG. Appl Catal A, 2008, 348: 221gen Energy, 1996, 21(1):13[46 Ozdemir H, Oksuzomer M F, Gurkaynak M A Int HYdrogen[34] Rakass S, Oudghiri-Hassani H, Rowntree P, Abatzoglou N JEnergy,2010,35:12147Power Sources, 2006, 158(1): 485[471 Zhu J, Zhang D. King K D. Fuel, 2001, 80: 899[35] Avila-Neto CN, Dantas S C, Silva F A, Franco TV, Romanielo [48] Reyes S C, Sinfelt J H, Feeley JS. Ind Eng Chem Res, 2008, 42L L, Hori C E, Assis A JJ Nat Gas Sci Eng, 2009; 1: 2051588中國(guó)煤化工CNMHG