A general model for devolatilization of large coal particles

Tm~cnv-SiuthSy~nposi,l~ll (I~ltcr~iationalj on CulnbustionlThe Colribustion Institute, ISSC,/pp. 31153151

A GENERAL MODEL FOR DEVOLATILIZATION OF LARGE COAL PARTICLES YUXIN ZZHAO, MICHAEL A. SERIO AND PETER R. SOLOMON Atlcc~nccdFi1c.1 Hesr*c~rcl~, Inc 87 Clnlrch Strcct Eo,t Hnrtford, CT 06108, USA

There ha\-? I~eenseveral studies on the tlevolatilization of'l~~rge coal particles under la1)oratoryconditions sim~llatinglarge-scale flrlidizetl or fixctl-hed coii11)ustion conditions. all of which de~nonstratethat the proccss of devolatilization of large coal particles occurs over a 111ucl1 longer tiinescale than that of pul\~erized coals un~lcrthe same conditioiis. This paper reports the develop~nentof a model for large ~(131 particle dcvolatilizatior~under fl~~idized and fixed-bed con~bustionconditions. The modcl combines a heat transfer lnodel with a gencral and co~nprchensivecoal devolatilization model, FG-DVC. It inherits the generillity of FG-D\'C and can be applied to coals of v;~riousGpes \*ithout cxtensi\,e prior kno\\rledge of the coals. This nod el iilcl~~des a detailed treatmcnt of'co;~lpyrolpis reactions includir~gthe yields of individual gas species, tar !.ields, and tar lnolecular n~eiglltdistrih~itions.The cl~angesin coal physical properties are inod(~1cdwith the applic;~tionof hl?rlick's coal physical propel-ty sl~l~models for specific heat and therlnal conductivity. Both the convective and r;idiati\re heat tra~isfcrt~et\veentlic gas pliasc ;~ndcoal particles are considered and Gl~nn'scorrelation for the surface heat transfer coefficient is used. Model predictions are co~nparedwith the data of large coal particle de\-olatilization ineasurrtl in fluidized beds and tubular reactors, for particle thermal response, total \\,eight loss. and individual gas cvolt~tionrates. 111general, the agreement \vith the data is w r y good for the particle center tc~n~perat~ire and weight loss curves and is fair for tlle indi\~idualgas evolution cllives.

Introduction

as they exit. These reactions would include tar crackiiig. gasification reactions with char, and gas-phase There liave been several studies o n tlie de\.olutil- reactions. However, there is no solid evidence supization of large coal particles under laboratory con- porting claims that these reactions are significant ditions sinlulating large-scale fluidized o r fixed-bed enough t o b e modeled [S]. Experiments have shown combustion contlitions, all of wvliicli demonstrate onlv moderate, at most. cllanges in )ields with that the process of devolatilization of large coal par- changing particle size. Tliese c l i n g e s may b e clue ticles occlurs over a ~nticlilonger timescale than that primarily t o tlie indirect effects of heat transfer (i.e., r same conditions. Soine tliat larger particles have slower thermal response). of pulverized coals ~ m d ethe studies [l-31 have addressed tlie tiniescale of t h e An exception is tlie for~natioilof HCN and NH,]. dew.olatilization process for large particles and have Recent exl?erimental results llave indicated tliat estnl)lished tliat t h e tinlescale is proportiorla1 t o cP, large lxtrticle size enhances the formation of NH,? at wit11 11 ranging from 0 . 3 2 t o 2.6 and tl lleing tlie t h e expense of I I C N [9]. Although mass transport cliaineter of t l ~ eparticle. T h e large variation of 1 1 in seeins not to cause significant secondary reactions tlie reported correlations reflects the clifficlilty in dt.for tlie ~ n a j o rvolatile gas species, there is at)undant ternlining the termination point of the tlevolatiliza- e ~ i d e n c ethat it plays an important role on the prition llrocess. More r e c e ~ i tstudies 1iaw.e ~.e~ortc,cl maiy devolatilization reactions determining the Inore detailed m e a s u r t ~ n ~ e n tof' s particle te~l~pc,ra- )ields of tlie tar. S o l o ~ n o ne t al. [8] have given a deture, particle weiglit loss, and gas species coiicentra- tailed review and pointed out that longer residence tions as a functio~iof time [4-71. Tlies? data sets ti~iiesof tar could cause its precursor to be relmlywere collected under relatively well-deiined condi- merizetl illto t h e char str~ictlire,leading t o lower tar tions and can be used I;,r nod el \.crification. yields. This view is incorporated in a coal pyrolysis It is g e ~ ~ e r aaccepted ll~ tliat tr;insport ollieat mid model, FC-DVC [8,10],which accounts for the mass Illass lro111the interior of particles to their s11rfac.e transport limitations associated with tar evolution. controls tlie timescale ol' large coal piirticle devolaModels have been developed to model tlie devoltilization. Besides thct e f k c t on tlit, de\-olatilization atilization of large coal particles. Aganval e t al. rates, there are speculations as to wllether the vol[11,12]and Hajaligol et al. [13]developed large coal a t i l e ~undergo s e c o n d a q reactions within lxu-ticles particle devolatilization models that combine