Modeling and simulation of chemically reacting flows in gas-solid catalytic and non-catalytic proces

Particuology 8 (2010) 525- -530Contents lists available at ScienceDirectPARTICUOLOEYParticuologyEL .SEVIERjournal homepage: www. elsevier. com/locate/particModeling and simulation of chemically reacting flows in gas- solid catalytic andnon-catalytic processesChangning Wu, Binhang Yan, Yong Jin, Yi Cheng *Department of Chemical Engineering, Tsinghua University, Beiing 10084, ChinaART1CLE1NFoA BSTRACTArticle history:This paper gives an overview of the recent development of modeling and simulation of chemically react-Received 5 May 2010ing flows in gas-solid catalytic and non-catalytic processes. General methodology has been focused onAccepted 12 August 2010the Eulerian-Lagrangian description of particulate flows, where the particles behave as the catalysts orthe reactant materials. For the strong interaction between the transport phenomena (i.e.. momentum,Keywords:heat and mass transfer) and the chemical reactions at the particle scale, a cross-scale modeling approach,Gas-solidchemically reacting f1ow_i.e., CFD-DEM or CFD-DPM, is established for describing a wide variety of complex reacting flows in mul-Eulerian-Lagrangian scheme”ulationtiphase reactors. Representative processes, including fluid catalytic cracking (FCC), catalytic conversion ofsyngas to methane, and coal pyrolysis to acetylene in thermal plasma, are chosen as case studies to demon-Computational fluid dynamics (CFD)Discrete element method (DEM)strate the unique advantages of the theoretical scheme based on the integrated particle-scale informationDiscrete phase model (DPM)with clear physical meanings. This type of modeling approach provides a solid basis for understandingthe multiphase reacting flow problems in general.◎2010 Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy ofSciences. Published by EIsevier B.V. AlI rights reserved.1. IntroductionFor well-known complexity such as the multi-scale phenom-ena (Li & Kwauk, 2003; Sundaresan, 2000), understanding theGas-solid fluidized bed reactors (e.g. bubbling fluidized beds,multiphase chemically reacting flows in a theoretical way isturbulent fluidized beds, risers, and downers) are widely appliedextremely challenging. So far, reactor modeling mostly lies inin diverse catalytic and non-catalytic processes, such as in the man-the category of simplifed fluid dynamics plus simplified chemi-ufacture of petroleum-based fuels and products, production ofcal kinetics. For example, the plug flow model, CSTR model, andcommodity and specialty chemicals, refining of ores, coal combus-the convection-diffusion-reaction model in most of the reactiontion and gasification, production of polymers and other materialsengineering textbooks have been widely applied to characterizeand so on. For a gas-solid catalytic process, the solid (particle) phasea reactor performance, where only reactor-scale information isgenerally exists in the form of a heterogeneous catalyst for catalyz-involved. In the past decade, CFD technique has received muching the gas phase reactions. The dispersed motion of the particlemore attention when modeling a multiphase flow reactor. In thisphase may lead to different movement pathways and therefore res-category, the discrete particle phase can be treated as a pseudo-idence times, along which the catalyst activity would change duecontinuous fluid (i.e., Eulerian- Eulerian scheme) or a discreteto the influence of the complex reaction micro-environment (e.g..element (i.e., Eulerian-Lagrangian scheme ). The advantage of thetemperature and concentration fields). In a gas- solid non-catalyticEulerian- Eulerian scheme is the consistent form of governing equa-process, the particles take their role as a reactant material or ations, which allows for efficient computation to handle practicalheat carrier. For example, coal particles react with the surroundingengineering applications. However, since the particle-scale infor-gases at certain temperatures to implement different coal conver-mation is not included in such models, a chemically reacting flowsion processes. For such cases, tracking each particle's history inathat is sensitive to the instantaneous particle properties such asmultiphase flow reactor is crucial to master the macro-scale reac-the catalyst activity in time cannot be reasonably described by thetor performance based upon the understanding of particle-scaleEulerian-Eulerian models from a theoretical point of view. As aphysics.comparison, particle-scale information can be tracked using theEulerian- Lagan-Lagrangian model,del, though the computational cost mightbe high. The pur中国煤the recet develpment of using ths to simulate severalgas-solid chemidE-mail address: yicheng@tsinghua.edu.cn (Y. Cheng).:fHc N M H Gifferent from most1674. 2001/1$ - see front matter 0 2010 Chinese Society of Particuology and Institute of Process Engineering. Chinese Academy of Sciences. Published by Elsevier B.V All rights reserved.doi:10.1016/j.partic.2010.08.003526C. Wu et al. / Particuology 8 (2010) 525 -530Control volume人少Discrete phase●momentum●heat/mass●position●reactions, particle-particle collisionContinuous phaseGovemning equations:Interphase exchangesmass, momentum, energy●massturbulence, species) momentumreactions●energyFig. 1. Schematic diagram of the Eulerian-Lagrangian model for reacting flow simulation.available references, heat and mass transfer as well as the chemicalfor the nature of dilute conditions. Compared to the above DEMreactions are incorporated into the hydrodynamic descriptions ofapproach, DPM is relatively simple, where the gas- phase simu-the multiphase flows.lation becomes dominant as the particles show clear tendencyto follow gas phase flows. Particle-scale physics and chemistry2. Eulerian-Lagrangian models for gas-solid chemicallycan be modeled, and versatile models have been included in thereacting flow problemscommercial software FLUENT. Upon that, new physics for specificprocesses can be modeled using the user defined functions (UDFs)A general modeling scheme is ilustrated in Fig. 1 for describingin the FLUENT environment. In this case, we will show a successfulthe chemically reacting flow problems with a continuous gas phaseapplication of this modeling scheme in simulating the coal pyroly-and a discrete particle phase. Two types of Eulerian-Lagrangiansis behavior under ultra-high temperatures generated by thermalmodels have been developed for characterizing gas- solid twoplasma.phase flows with the consideration of whether or not theparticle-particle collision is taken into account.3. Case studies2.1. CFD- DEM for dense gas-solid flows3.1. FCC process in risers and downersIn most of the multiphase flows in chemical engineering appli-cations, e.g. in bubbling/turbulent fluidized bed, riser and denseproducts such as gasoline and light olefins using cracking cata-downer reactors, the volume fraction of particles is relatively highlyst in short gas- solid contact time. The riser has served as theso that the interaction between particles must be considered tomajor commercial reactor for conventional FCC processes for overcapture the major features of the underlying multiphase trans-50 years, while the downer reactor has recently been widely inves-port phenomena. In this category, the discrete element methodtigated and promoted as a new-generation FCC reactor (Aitani,(DEM), also called soft sphere model, has been widely appliedYoshikawa, & Ino, 2000; Cheng, Wu, Zhu, Wei, & Jin, 2008; Zhu, Yu,to simulate the particulate flows (e.g, Deen, van Sint Annaland,jin, Grace, & Issangya, 1995). Although the only major differencevan der Hoef, & Kuipers, 2007; Tsuji, Kawaguchi, & Tanaka, 1993;between the riser and downer reactors is that the gas and solidsZhu, Zhou, Yang, & Yu, 2007). In the literature, the aforementionedmove co-currently along or against the gravity force in macro-scaleCFD-DEM approach was mostly used to simulate the multiphaseoperation, these two kinds of reactors show great difference in flowflows without chemical reactions, where only few of the referencespattern and reactor performances for FCC processes (Wu, Cheng,introduced the heat transfer model (Kaneko, Shiojima, & Horio,& jin, 2009). In an FCC process, the catalyst particles inherently1999; Zhou, Yu, & Zulli, 2009). Here, we focus on the extension ofundergo complex trajectories and reaction micro- environment. Asthe DEM approach to reacting flow problems, that is, the particle-a result, the catalyst activity changes in time and in space (e.g..scale heat and/mass transfer, catalyst activity in time, etc., arecoke deposition), which accordingly imposes complex influencesincorporated in the general CFD- DEM governing equations (Wu,on reactor performance (i.e, conversion and selectivity). A physi-Cheng, Ding, & Jjin, 2010). At the current stage, DEM code has notcal description of the catalytic process with deactivation of catalystbeen well commercialized so that in-house codes were developedmust take into account the instantaneous particle-scale informa-in different groups. We have succesfully embedded our own DEMtion along with their movement. In particular, the activity of thecode in FLUENT and CFX commercial software for multiphase flowFCC catalyst drops very rapidly, e.g.. decreased by 1/2 within aboutsimulations.' Two case studies, i.e., CFD-DEM simulations of FCC and1 s. This is also the reason that downer would be superior to risersyngas to methane processes, will be presented in the followingin terms ot1of reactor performance, given its significantly shortenedsection.residence time.Under the fram1sr pproacha, ie. the2.2. CFD-DPM for dilute gas-solid flowsparticle phase is trerk, a large numberof reactor models \1HcNMHG.ractingAowsinThe discrete phase model (DPM) is also referred to as the particleFCC process in the.in, & Heynderickx,trajectory model, in which particles interact with the continu-2003; Gao, Xu, Lin, Yang, & Guo, 1999; Liu et al.. 2008; Theologosous gas phase, but the particle-particle collision is not consideredand Markatos, 1993). However, the inherent assumption that theC. Wu et al. / Particuology 8 (2010) 525. -530527particle phase is a continuous medium brings evident constraints tooidageresidence time, 1 (s)the accurate understanding of the reacting flows in an FCC reactor.One of the key constraints is that the history of the catalyst particlescannot be appropriately modeled. Hence the practical informationof the residence time and the activity of the catalyst particles aremissed. Even if the deactivation function in terms of the particleresidence time (tc) is applied, there is actually still a compromise9to assume that the catalyst activity is uniform across the sametioncross section (Wu et al, 2010a; Wu et al., 2009). If the deactivatifunction in terms of the coke weight percentage on catalyst (Cc)1.43mcomposition of the gas-phase species based on the default assump-tion of no spatial slip between the gas and particle phases (Gao etal, 1999). Therefore, these kinds of reactor models cannot physi-cally describe the effect of the transient hydrodynamics integratedwith the catalytic reactions on the overall reactor performance. Inthe past years, we have developed a DEM module, which has beensuccessfully embedded in the commercial software FLUENT (andCFX). Hydrodynamics in cold-model fluidized beds has been wellpredicted by the CFD-DEM method (e.g.. Zhao, Cheng, Ding, & Jin,2007; Zhao, Cheng. & jin, 2007; Zhao, Cheng, Wu, Ding, & jin, 2010;Zhao, Ding, Wu, & Cheng, 2010) Particularly in line with the exper-imental observations were the simulation of the gas- solid flow ina two- dimensional (2D) downer of 10 m in height and 0.10m in .width, clearly revealing the micro- to macro-scale flow structures.The distinct clustering phenomena and the residence time distri-butions (RTDs) in the fully developed region of a riser and a downerwere observed using the CFD-DEM model (Zhao et al, 2010a).Further, we have extended the CFD-DEM approach to pre-dict the gas-solid chemically reacting flows by incorporating thedescriptions for heat transfer behaviors between particles andbetween gas and particles, the instantaneous catalyst deactivation(Gross, Jacob, Nace, & Voltz, 1976), and a 4-lump chemical kinetics4operating pressur 250 kPain the gas phase for FCC reactions (Wu et al, 2010a). As shown ingas inlet temperature, 823 K0.11 mgasflow rate, 6 kg(m2 s)Fig. 2, the model predictions can reasonably track the history of thecatalst-t--oil ratio, 10particle movement with the time -dependent heat transfer, catalystcatalyst activity,deactivation and the chemical reactions.t=4.0sφ= 1/(1+162.1519.60As a comparison, the reacting flow in a downer reactorFig. 2. llustration of the transient hydrodynamics characterized with residenceapproaches plug flow pattern so that the catalyst activity is uni-formly decayed along the flow development. Accordingly, downerexhibits much better control of selectivity to intermediate prod-ucts. The simulation results captured the major features of FCCprocess very well either in the riser or in the downer, reasonably inline with experimental data from the literature (Corma, Martinez,Recently, we applied the CFD-DEM approach to simulate theMelo, Sauvanaud, & Carriat, 2002; Wu et al, 2010a).gas- solid reacting flows in a lab-scale 2D fluidized-bed reactorfor syngas-to-methane (STM) process (Wu, Tian, & Cheng, 2010).The distinct advantage of this modeling approach is that the par-3.2. Catalytic conversion of syngas to methaneticle temperature (i.e.. reaction temperature) can be calculated intime accurately by tracking the history of the particle movement,The syngas- to-methane (STM) process converts the syngasmaking it possible to judge whether a catalyst particle is over-from the coal or biomass gasifers to substitute natural gas (SNG)heated and sintered or not. A kinetic model was proposed andthrough the well-known highly exothermic methanation reactionsintegrated into the CFD-DEM model which included the metha-(Wender, 1996). The production of SNG from coal provides annation reaction of CO, steam reforming reactions of methane,alternative fuel to supplement the limited petroleum and natu-and forward/reverse water-gas-shift (WGS) reactions. Deactiva-ral gas (Hirsch, Gallagher, Lessard, & Wesselhoft, 1982). Generally, .ion function of catalyst is however not considered in the currentthe methanation reactions take place at around 300- 500°C overmodel. .supported noble metal catalysts in a methanation reactor. The tem-The simulation results captured the major features of the reac-perature inthe reactor should be contolled to preventoverheatingperatorr performnanceincluding unwanted defluidization, as shown inabeof the catalyst particles due to the exothermic reaction. In addition,Fig. 3. At the gas mass flux of 0.33 kg/(m2 s), thecould bea high temperature is also undesirable for methane formation fromwell fluidized“中国性化宁ass fux Was reducedthe viewpoint of thermodynamic equilibrium. Most of the metha-to 0.11 kg(m2s)perficial gas velocitynation reactors reported have been fixed bed reactors. However,(Ug) of 5.1 timesHCNMHG-elocity (Umr), almostthe fluidized-bed reactor is actually better suited for the syngas-to-all of the particJottom region of themethane process because of its excellent heat removal capabilitycolumn (see Fig. 3()). This phenomenon is mainly caused by theand easier scale-up.fast methanation reaction involving a large decrease in the num-528C. Wu et al. / Particuology 8 (2010) 525 -5300.050.04 tparticletemperature(K)1073973923873823773出6736230.02.01 t↑↑↑↑↑1 ↑↑↑↑↑↑↑↑↑↑↑gas: 0.33 kg/(m2s) 0.11 kg/(m2 s)t=9.5st= 12.5s(b)(coperating pressure, 5 atmfree steam temperature, 623 Kgas inlet temprature, 623 K wall heal transfer cofficient, 2500 W/(m2 K)components ofeed gas (w%): 78.0 co, 19.5 H2, 0.5 CO2, 1.0 CH4 and 1.0 H2OFig. 3. Transient spatial dstributions of particles characterized with their respective temperatures.bers of molecules, that is, 4-2 with a corresponding value of Ug/Umfsion at the inlet, which were validated by the direct observation inreduced to 2.6.the cold model experiments. Tian, Xie, Zhu, and Fletcher (2001)Under the same operating conditions and reactor geometry, thesimulated the coal devolatilization using the kinetic model offluidized bed reactor behaves differently from the fixed-bed reac-chemical percolation devolatilization (CPD) under the conditionstor. The results indicated the excellent heat removal capability ofof their lab-scale plasma reactor and showed a sound agree-the fluidized-bed reactor in preventing the catalyst particles fromment with their experimental data at relatively large coal feedoverheating and sintering.rates.Although a full understanding on multiphase flows has not beenFor a thorough understanding of the detailed processes that aachieved yet, the CFD -DEM modeling approach including basiccoal particle undergoes together with its interaction with high-mechanisms of the gas-solid catalytic reacting flows will undergotemperature gas flow, Chen and Cheng (2009) established agteatevarlouschemical processes.It an becomprehensive model, including the following mechanisms:cludedt this type of modeling approach forms a solid basis forthe cross- scale modeling of general multi-phase catalytic reacting●Particle -scale physics: particle movement, heating up,flows.devolatilization, and further thermal conversion of char;●Particle-gas interaction: gas-particle drag, heat and mass trans-3.3. Coal pyrolysis to acetylene in thermal plasma●Gas phase physics and chemistry: turbulent flow, heat transfer,Coal pyrolysis in thermal plasma would open up a direct andthermal conversion of volatile gases (ie, equilibrium chemistry),clean means to convert coal to chemicals as a typical gas-solidand turbulent-chemistry interaction (by β-PDF model).non-caLalyecproesstvever,he extreme operating conditions(i.e., ultrahigh temperature and milliseconds contact time) largelyAmong the sub-models, the chemical percolation devolatiliza-tion (CPD) model. 由南建花tein, Pugmire, anddemics. In order to qualitatively understand the process design,Grant (1990) was uf coal devolatiliza-CFD simulations of the gas- solid two-phase flows were conductedtion. The CPD modeYHCNMHGEcture of coal and itunder the Eulerian -Eulerian scheme by Cheng, Chen, Ding, Xiong,has demonstratedlatilization behav-& jin (2008). These experiments investigated the effect of theior of coal. Especially, the influence of the physical properties ofdesign of the nozzle shape and arrangement on the coal disper-coal on the pyrolysis results can be addressed.C. Wu et al. / Particuology 8 (2010) 525- -530529, HydrogeninletV-shaped夫一Coal inletplasma torch' ←letz=0.021 m2=0.021mMixing zoneTemperature (网Reaction zoneCooling wateraround the wallVelocity Magnitude (m/s)950850=白Quenchz=-0.4mand separator星。CFig. 4. Schematic of the 2-MW plasma reactor (a), and the spatial distributions of gas temperature (b) and velocity magnitude (c) in the reactor calculated by the CFD-DPMsimulations (Chen & Cheng. 2009).The three-dimensional (3D) simulations were carried out tolization processes. Heat conduction resistance in coal particles andanalyze a 2-MW pilot-scale plasma reactor in industry (see thethe mass transfer resistance of the generated volatiles would begeometry shown in Fig. 4). The simulation results revealed the non-the two main resistances impeding the inward heat flow. The pro-uniform flow field and temperature field under the unique designposed model was successfully validated using the available data inof the V-shaped plasma. The high-temperature, high-speed gas jetthe literature. AIll the results demonstrated that both of the resis-in the center hinders the coal particles from penetrating into thetance factors would impede thermal energy transportation into theehot zone of the reactor. In other words, teactor performanceparticle. This effect would become more obvious when the particleof the coal pyrolysis is sensitive to the contact between the coalsize was larger.particles and the hot gas. Under a good phase-contact operation,Yan, Wu, jin, and Cheng, 2010 therefore further improved thethe concentration of acetylene might reach its maximum when thework of Chen and Cheng (2009) by considering the above mecha-power input of plasma well matched the feeding rate of coal.nisms in the CFD model, together with (1 ) the reasonable evaluationOn the basis of the above work, a mechanism model incorpo-of the devolatilization species and the heat of pyrolysis; and (2) arating the heat conduction in solid materials, diffusion of releasedkinetic model for the production of light gas from the tar-crackingvolatile gases and reactions was proposed for a thorough investi-reactions (Ma, 1996). The improved CFD-DPM model was vali-gation of the heat transport inside a coal particle under the extremedated with the experimental data reported by Baumann, Bittner,environmental conditions (Shuang, Wu, Yan, & Cheng, 2010). As is .Beiers, Klein, and Juntgen (1988). As shown in Fig. 5(a), the mod-generally known, volatile matter, i.e., the gaseous product of coalifed CFD-DPM model could give better predictions on the carbondevolatilization, is always yielded in the first chemical conversionconversion to acetylene from feeding the coal at different inletstep for any coal utilization process. Thermal energy is the driv-tempertures of the plasma, in comparison to the reported experi-ing force for devolatilization, which means that the behavior ofmental data. The predicted variations of carbon conversion to lightinward flow of thermal energy to coal is the key issue for coal uti-gas with the axial location were also compared with the experimen-0t鼓Experimentals (Baumann et al.1988)Tg(); Expt.ModelAlI volatiles converted to light gas-- - Tar cracking considered--29005一30啊一a to C2H2) to liaht aso中国煤化工2002250250027500.05 0.1Inlet temperature of plasma (K)Dist:MYHCNMHGFig. 5. Comparison of model predictions with experimental data. (a) Carbon conversion to C2 H2 at different inlet temperature of plasma, and (b) carbon conversion to lightgas at different axial location (Wu et al, 2010).530C. Wu et al. / Particuology 8 (2010) 525- -530tal data, showing reasonably good agreement (see Fig. 5(b); Wu etFletcher, T. H. Kerstein, A. R, Pugmire, R. J.. & Grant, D. M. (1990)]. Chemical per-al., 2010). The model has also been used to predict the 2-MW andcolation model for devolatilization 2. Temperature and heating rate effects on5-MW pilot-plant reactors (Yan et al., 2010).Gao,J.S.. Xu.C M. Lin,S. x. Yang, G. H. & Guo, Y. c. (999. Advanced model forturbulent gas- solid flow and reaction in FCC riser reactors. AIChE Journal, 45,4. Concluding remarks and future perspectives1095- 11Gross, B.Jacob, s. M., Nace, D.M.. & Voltz, s. E.(1976). Simulation ofcatalytic crackingThis paper provides an overview of the recent development ofmodeling and simulation of chemically reacting flows in severalCatalytic coal gasification: An emerging technology. Science, 215， 121-typical gas- -solid catalytic and non- catalytic processes, includingKaneko, . Shiojima, T. & Horio, M. (1999). DEM simulation of fluidized bedsFCC, syngas to methane conversion and coal pyrolysis to acetylene.for gas-phase olefin polymerization. Chemical Engineering Science, 54, 5809-The key issue is to address the signifcant effect of particle-scaleinformation on the meso- and macro-scale behaviors in multiphaseji,」H，&Kwauk,M.(2003).Exploring . complex systems in chemicalengineering- the multi-scale methodology. Chemical Engineering Science, 58,flows especially when chemical reactions occur in the system. In521-535.addition, the reactor performance with chemically reacting flowsLiu, F. Wei, F. Li, G. L, Cheng, Y., Wang. L, Luo, G. H.. et al. (2008). Study onis shown to be distinctly different from a reactor without reactions.the FCC process of a novel riser-downer coupling reactor (II): Industrial trialThe integration of momentum, heat and mass transfer with chem-ical reactions poses the challenges on the solutions to modelingMa.J.(1996). Soot formation and soot secondary reactions during coal pyrolysis. Unpub-and simulation. This leads to a close connection between funda-lished doctoral dissertation. Utah, USA: Brigham Young University.mental research and practical applications of new understandingShuang, Y, Wu, C. N, Yan, B. H, & Cheng. Y. (2010). Heat transfer inside particlesand devolatilization for coal pyrolysis to acetylene at ultrahigh temperatures.and new insights. Since the modeling and simwlationof multiphaseI new insights. Since the modelSundaresan, s. (2000). Perspective: Modeling the hydrodynamics of multiphase flowreacting flows is still at its infant stage, tremendous efforts mustbe made to establish precise descriptions for diverse reacting flow .reactors, current status and challenges. AIChE Journal, 46, 1102-1105.features, which would also lead to diverse contributions to currentTheologos, K. N, & Markatos, N. C (1993). Advanced modeling of fluid catalyticcracking riser type reactors. AIChE Journal, 39, 1007-1017.industrial applications. On the other hand, experimental studies,Tian,' Y.Xie, K., Zhu, S., & Fletcher,T. H.(2001). Simulation ofcoal pyrolysis in plasmainvolving well-designed (hot-model) experiments and advancedTsuji, Y. Kawaguchi, T.. & Tanaka, T. (1993). Discrete particle simulation of twomeasurement techniques for probing the temperature and/or thedimensional fuidized bed. Powder Technology, 77 79-87.progress of chemical reactions, are demanded to establish reliableWender, I. (1996). Reactions of synthesis gas. Fuel Processing Technology, 48,database for straightforward process understanding and modelwu.C.N. Cheng.y.. Ding.Y.L &lin.y.(2010). CFD-DEM simulation of gas-solidvalidation.reacting flows in fluid Catalytic cracking (FCC) process. Chemical EngieeringScience, 65, 542- 549.AcknowledgementsWu, C. N, Cheng, Y.. & jin, Y. (2009). Understanding riser and downer based FCCprocesses by a comprehensive two-dimensional reactor model. Industrial andThe authors would like to acknowledge the financial sup-wu, C.N.Tian.D.Y.. &Cheng. Y.(2010) CFD-DEM simulation of syngas-to-methaneprocess inafluidized-bed reactor. InS. D. Kim.yY. Kang.J.K.Lee.&Y.C.Seo(Eds.).port of the National Natural Science Foundation of China (NSFC)under grants Nos. 20976091 and 20806045, the Key Project ofFluidization XII Engineering Conferences International(p.733). Korea: Gyeong-ju.Wu, C. N., Yan, B. H., Zhang, L. Shuang, Y.. Jjin, Y.. & Cheng, Y. (2010). Analysis ofNational High-tech R&D Program under grant No.2009AA044701,key techniques and economic feasibility for one-step production of acetyleneand the Program for New Century Excellent Talents in universitiesby coal pyrolysis in thermal plasma reactor. Jourmnal of Chemical(NCET).Yan, B. H.. Wu, C. N.. jin, Y. & Cheng. Y. (2010). CFD simulation of the reacting flowprocess of coal pyrolysis to acetylene in hydrogen plasma downer reactors. InS.ReferencesD. Kim, Y. Kang,J. K. Lee, &Y. C. Seo(Eds.).Fluidization XII Engineering ConferencesInternational (p: 725). Korea: Gyeong-ju.Aitani, A. Yoshikawa, T, & Ino, T. (2000). Maximization of FCC light olefinsina2D downerby CFD-DEM simulation. InF. Berruti.X.T. Bi &D.PugIsey(Eds.).by high severity operation and ZSM-5 addition. Catalysis Today, 60, 11Fluidization XII (pp.855-862). Vancouver: Engineering Foundation.Zhao,' Y. Z., Cheng, Y, & Jin, Y. (2007) Kinetic simulation of two-dimensional spoutedBaumann, H. Bittner, D.. Beiers, H. G. Klein.J.& Juntgen H.(1988)bed with draft plates by discrete element method. Chemical Engineering (China),Chen, J. Q.. & Cheng, Y. (2009). Process development and reactor analysis of coalZhao,Y.Z, Cheng,Y., wi.c. N. Ding, y. L. & jin, y. (2010). Eulerian-Lagrangianpyrolysis to acetylene in hydrogen plasma reactor.Journal ofChemical Engineer-simulation of distinct clustering phenomena and RTDs in riser and downer.Particuology, 8, 44- -50.Cheng. Y.. Chen,J. Q, Ding. Y. L. Xiong. X. Y. & jin, Y. (2008). Inlet effect on the coalZhao, Y. Z. Ding, Y. L, Wu, C. N., & Cheng, Y. (2010). Numerical simulation of hydro-PyCnles tretreteri.t. hY5r9Ero, plasma downer reactr Canadian. loumaldynamics in downers using a CFD-DEM coupled approach. Powder Technology,Cheng. Y, Wu, C. N.. Zhu, J. X., Wei, F. & jin, Y. (2008), Downer reactor: FromZhou, z. Y. Yu, A B.. & Zuli, P. (2009). Particle scale study of heat transfer in packedfundamental study to industrial application. Powder Technology, 183, 364-and bubbling fluidized beds. AIChE Journal, 55, 868- -884.Zhu, H. P.. Zhou, Z. Y., Yang, R. Y, & Yu, A. B. (2007). Discrete particle simulationCorma, A.. Martinez C. Melo, F. V.. Sauvanaud, L. & Carriat,J. Y. (2002). A new con-ofparticulate systems: Theoretical developments. Chemical Engineering Science,A:Ceneral,232, 247-263.Das, A. K, Baudrez, E.. Marin, G. B. & Heynderickx, G. J. (2003). Three-dimensionalcirculating fuidized bed(Downer) reactrs-A state of the art review. Canadiansimulation of a fluid catalytic cracking riser reactor. Industrial and EngineeringJournal of Chemical Engineering. 73, 662- 677.Review of discrete particle modeling of fluidized beds. Chemical EngineeringScience, 62, 28- 44.中国煤化工MHCNMHG