Aalbog Univesitet Modeling of pulveized coal and biomass co-fiing in a 15 KW swiling-stabilized bune and expeimental validation Yin, Chungen; Kæ, Søen Knudsen; Rosendahl, Lasse Aistup; Hvis, Søen Lovmand Published in: Poceedings of the Intenational Confeence on Powe Engineeing, ICOPE '9 Publication date: 9 Document Vesion Publishe's PDF, also known as Vesion of ecod Link to publication fom Aalbog Univesity Citation fo published vesion (APA): Yin, C., Kæ, S. K., Rosendahl, L., & Hvis, S. L. (9). Modeling of pulveized coal and biomass co-fiing in a 15 KW swiling-stabilized bune and expeimental validation. In Poceedings of the Intenational Confeence on Powe Engineeing, ICOPE '9 (Vol., pp. 5-1). The Japan Society of Mechanical Enginees. Geneal ights Copyight and moal ights fo the publications made accessible in the public potal ae etained by the authos and/o othe copyight ownes and it is a condition of accessing publications that uses ecognise and abide by the legal equiements associated with these ights.? Uses may download and pint one copy of any publication fom the public potal fo the pupose of pivate study o eseach.? You may not futhe distibute the mateial o use it fo any pofit-making activity o commecial gain? You may feely distibute the URL identifying the publication in the public potal? Take down policy If you believe that this document beaches copyight please contact us at vbn@aub.aau.dk poviding details, and we will emove access to the wok immediately and investigate you claim. Downloaded fom vbn.aau.dk on: juli 14, 18
Poceedings of the Intenational Confeence on Powe Engineeing-9 (ICOPE-9) Novembe 16-, 9, Kobe, Japan MODELING OF PULVERIZED COAL AND BIOMASS CO-FIRING IN A 15 KW SWIRLING-STABILIZED BURNER AND EXPERIMENTAL VALIDATION Chungen YIN*, Søen Knudsen KÆR*, Lasse ROSENDAHL* and Søen Lovmand HVID** *Dept. of Enegy Technology, Aalbog Univesity, Pontoppidanstaede 11, 9 Aalbog East, Denmak **DONG Enegy, Kaftvæksvej 5, 7 Fedeicia, Denmak ABSTRACT A compehensive CFD modeling study has been undetaken to investigate the co-fiing of pulveized coal and biomass in a 15 kw Fuel swil-stabilized bune, which is simila in flow patten to a typical low-no x bune. The objective of the pesent study is to deive a eliable modeling methodology fo design and optimization of low NO x bunes co-fiing pulveized coal and biomass. Fo this pupose, the effects of meshes, global eaction mechanisms fo homogeneous combustion, tubulence models, tubulence-chemisty inteactions, popeties of the solid fuels, and solid-fuel paticle convesion models ae finely examined. The modeling esults ae compaed with detailed mapping of mola factions of main species, obtained fom FT-IR and a Hoiba gas analyze. This pape pesents mainly the compaison of diffeent global homogenous combustion mechanisms and diffeent solid-fuel paticle convesion models, in modeling of pulveized coal and biomass co-combustion. Keywods: Biomass, Coal, Co-combustion, CFD, Swil bune, Global eaction mechanism, Paticle convesion 1. INTRODUCTION Biomass fuels ae one of the most impotant enegy esouces and ae also consideed envionmentally fiendly and enewable, which has geatly spued inteest in using biomass fo heat and powe poduction in ode to mitigate the poblems of global waming and the limited availability oossil fuels. Co-fiing biomass in existing pulveized coal-fied powe plants is one of the main options cuently used in biomass combustion fo enegy poduction. Most pulveized-coal boiles use low-no x bunes, which ae optimized to educe NO emission while maintaining complete coal-bunout. Replacing a fuel o even changing a small potion of the fuel may upset this optimization [1]. This pape pesents a compehensive CFD modeling study of co-fiing pulveized coal and biomass in a 15 kw Fuel swil-stabilized bune, which is simila in flow patten to a typical low NO x bune. The modeling esults ae compaed with the expeimental data, measued by FT-IR and a Hoiba gas analyze. Due to editoial limits, this pape highlights mainly the effects of diffeent global eaction mechanisms fo the volatiles of the solid fuels, and diffeent models fo the solid-fuel paticle convesion.. EXPERIMENTAL WORKS Unde the famewok of the same poject, all the expeimental woks wee done in a bune flow eacto, located in Bigham Young Univesity. A scaled cosssection view of the inne chambe of the eacto indicating locations of the bune qual and access windows is shown in Fig. 1. The eacto has a vetical height of 4 cm, and an inne diamete of 75 cm. The dual-feed swil bune (not shown in Fig. 1) fies the fuels downwad (down-fied) fom the top of the eacto, with a self-sustaining flame. The combustion poducts pass though an ash sepaato and scubbe befoe being vented into the ai. The main measuing effot was to poduce high quality, quantitative maps of gas species (CO, CO, H O, O ; CH 4, C H, C H 4, C 6 H 6 ; NO, NH, HCN) in the flame. The coodinate system (z, ) in Fig. 1 shows the measuing gid. All the details, including the bune and eacto, fiing conditions and measued species maps, can be found in []. Fig. 1 A scaled view of the eacto inteio. The fuels and the opeational conditions of co-fiing pulveized coal and biomass ae given in Table 1.
Table 1 Fuels and opeational conditions Coal Staw Poximate analysis Moistue [wt%, as eceived].1 7.7 Fixed cabon [wt%, dy] 51.5 15.6 Volatiles [wt%, dy] 4.6 79.5 Ash [wt%, dy] 7.89 4.91 HHV [kj/kg, dy] 71 1849 Ultimate analysis C [wt%, dy] 74.8 47. H [wt%, dy] 5.8 5.68 N [wt%, dy] 1.5.54 O [wt%, dy] 1.1 41.6 S [wt%, dy].58 <.1 Paticle Rosin-Rammle paametes Min. diamete [µm] 5 5 Max. diamete [µm] 1 Mean diamete [µm] 11.4 451. Spead paamete [-] 4.4.1 Opeational paametes Cente Annula Cente Annula Seconday Swiling fuel [kg/s] fuel [kg/s] ai [kg/s] ai [kg/s] ai [kg/s] numbe.4194.8.5..44444 1 (staw) (coal). CFD MODELING Compaed with pulveized coal paticles, staw paticles ae elatively big in size, light in density, and non-spheical in shape, which geatly affect the motion and convesion ouel paticles in combustos []. In this wok, the effect of non-spheicity of the fuel paticles on thei motion and convesion has not been accounted fo. Both the coal and staw paticles ae assumed to be spheical. When the fuel paticles tavel though gas and inteact with gas in the eacto, they heat up, devolatilize and undego cha combustion, ceating souces fo eaction in gas phase. Two diffeent ways ae used fo the gas phase combustion of the fuel-volatiles. Two diffeent methods ae also employed fo the convesion of the elatively big staw paticles. All the CFD simulations (66 cells, steady, axisymmetic swil) ae done using FLUENT v.6..6 [4]. The use-defined sub-models ae implemented into the code via UDFs..1 Global eaction mechanisms fo volatile combustion Mechanism-1 (the simplified -step global eaction mechanism). Two atificial species ae used to epesent the coal-volatile and staw-volatile, espectively. The combustion of the two volatiles is modeled using Finite Rate / Eddy Dissipation Model (EDM) and a two-step eaction mechanism with CO as the intemediate species, CH.6O.46N.56S.9 +.997 O CO 14444 44444 coal volatile (1) + 1.H O +.9SO +.8 N CH.14O.985N.146 +.541O CO 1444 4444 staw volatile + 1.67 HO +.7 N CO +.5O CO () The compositions and fomation enthalpies of the two volatiles ae detemined fom the poximate and ultimate analysis data of coal and staw, espectively. Thee ae 8 species and eactions, Eq.(1)~(), in this mechanism. () Mechanism- (the global mechanism of Jones and Lindstedt). The two volatiles ae consideed as a mixtue of eal species, as shown in Eq. (4) and (5), espectively. CH.5998O.56N.566S.9.14CH8 +.41H 144444 44444 coal volatile (4) +.7CO +.51HCN +.56 NH +.9SO CH.15O.986N.145.17CH4 +.575CH8 +.47H 14444 4444 staw volatile +.CO +.15HCN +.1NH +.CO Then the global eaction mechanisms, poposed by Jones and Lindstedt (1988) fo hydocabon combustion [5], ae used fo combustion of the hydocabons (i.e., CH 4 and C H 8 ) in the mixtue. The tubulence-chemisty inteaction is modeled by Eddy-Dissipation Concept (EDC). CH 4.5O CO + H CH 4 HO CO + H H.5O HO CO HO CO + H C H8 1.5O CO + 4H C H8 HO CO + 7H (5) + (6) + (7) + (8) + (9) + (1) + (11) In this mechanism, thee ae 1 species in total (H, H O, O, NO, HCN, NH, CH 4, C H 8, CO, CO, SO, N ) and 6 gas-phase eactions, Eq.(6) Eq.(11). In the above two diffeent mechanisms, the global ate expessions fo the 9 involved eactions ae given as,. 1. [ vol_coal] [ ]. 1. [ vol_staw] [ ]. 5 [ CO] [ ].5 1. 5 [ CH4 ] [ ] [ CH4 ] [ H O].5 [ H] [ ] 1. 5 [ CO] [ H O].5 1. 5 [ CH8] [ ] [ C H ][ H O] 1 = k1 ( T ) O = k ( T ) O = k ( T ) O 6 = k6 ( T ) O 7 = k7 ( T ) 8 = k8 ( T ) O 9 = k9 ( T ) 1 = k1( T ) O 11 = k11( T ) 8 (1) (1) (14) (15) (16) (17) (18) (19) () in which the eaction ate coefficients can be expessed as f b ki ( T ) A T exp( E /( RuT )), (1) with the kinetic paametes given in Table. Table Kinetic paametes fo the gas-phase eactions Reaction Rate coefficient A b E [J/kmol] 1 k f ( ).119 1 11.7 1 8 1 T k f ( ).119 1 11.7 1 8 T k f ( ).9 1 1 1.7 1 8 T 6 k f ( ) 4.4 1 11 1.6 1 8 6 T 7 k f ( ) 1 8 1.6 1 8 7 T 8 k f ( ) 6.8 1 15-1 1.67 1 8 8 T 9 k f ( ).75 1 9 8.4 1 7 9 T 1 k f ( ) 4 1 11 1.6 1 8 1 T 11 k f ( ) 1 8 1.6 1 8 11 T
. Convesion of big staw paticles Convesion of solid fuel paticles is modeled by DPM (Discete Phase Model) laws. Unde the taditional conditions in a pulveized coal/biomass co-fied combusto, the Biot numbe of pulveized coal paticles is well below.1, and coal paticles ae unde isothemal conditions. Howeve, fo big biomass paticles, the Biot numbe is above.1, indicating that isothemal conditions may not hold any moe [6]. Fo compaison, two diffeent ways ae used fo the convesion of big staw paticles in the eacto. Default DPM laws. The paticle is teated as a lumped system: diffeent convesion pocesses (e.g., dying, devolatilization, cha oxidation) occu in seies. Fo example, volatile elease begins only afte all the moistue is diven off and the tempeatue of the paticle eaches the vapoization tempeatue. This is the default way used in FLUENT to update the mass, tempeatue, size and density of a combusting paticle when it tavels in a combusto [4]. Custom DPM laws. The big staw paticles ae discetized into a numbe of shells. Fo each of the shells, tempeatue, convesion ates and compositions will be updated. By using the custom DPM laws, the diffeent convesion pocesses occu in paallel, with the convesion fonts moving fom the suface to the cente of the paticle. Basically, the custom DPM laws ae simila in mechanism with inta-paticle heat, momentum and mass tanspot modeling, e.g., in [7]. Hee only a conceptual compaison between the default and the custom DPM laws is sketched in Fig.. Fig. Conceptual compaison of diffeent DPM laws.. Summay oive diffeent cases Table biefly summaizes the five diffeent cases, whose esults ae pesented in this pape. Table Main mechanisms/models used in the five cases. Case label Volatile Staw paticle Othe models combustion convesion taditional Mechanism-1 default DPM law RKE (Realizable k-ε) fo tubulence; DO fo adiation SKE-default Mechanism- default DPM law SKE (Standad k- ε); DO fo adiation SKE-custom Mechanism- custom DPM law SKE fo tubulence; DO fo adiation RKE-default Mechanism- default DPM law SKE fo tubulence; DO fo adiation RKE-custom Mechanism- custom DPM law RKE fo tubulence; DO fo adiation 4. CALIBRATION OF UDFS AND CFD MODELING Befoe being compaed with the expeimental data, the UDFs (e.g., the UDF fo gas species souces fom paticles which is used fo global eaction mechanism-; the UDF fo the custom DPM laws) and the CFD modeling have been calibated, against the expectation o in tems of oveall heat/mass/element balance. As a demonstation, Table 4 lists pat of the calibation esults fo of the cases. One may conclude fom Table 4 that the UDFs wok popely in tems of species souces and both the cases ae well conveged. Fo the case RKE-default, in which the default DPM laws ae used fo both coal and staw paticles, the eo with the H O souce is elatively big (about -.5%), which was poven to have nothing to do with the species souce UDF and is likely an inheent poblem with wet combustion ouel paticles. The compaison also shows the custom laws poduce a highe DPM enthalpy souce than the default laws. This diffeence is geate than what one could expect fom the diffeence in the cha convesion ates, hinting that the custom DPM laws may need to be locally efined fo the heat geneation and tansfe. Table 4 Pat of the calibation esults. (1) Calibation of the DPM souce UDF RKE-default RKE-custom DPM souce Expected FLUENT Relative FLUENT Relative value [kg/s] epot, kg/s Eo, % epot, kg/s Eo,% H souce 1.986e-4 1.11e-4.187 1.986e-4 -.1 HCN souce 5.89e-5 5.817e-5.14 5.87e-5 -. NH souce.6759e-5.6816e-5.15.6759e-5 -.1 CH 4 souce.86e-4.81e-4.51.86e-4. C H 8 souce 6.5e-4 6.16e-4.11 6.51e-4 -. CO souce 1.616e- 1.69e-.18 1.616e- -.1 SO souce.65e-5.65e-5 -.4.65e-5 -.4 H O souce.667e-4.5747e-4 -.5.671e-4.1 CO souce 7.556e-* 6.517e- 6.746e- () Calibation of the oveall mass balance DPM mass souce [kg/s].5699.5656 Net mass flux ove all the boundaies [kg/s] -.5649 -.5656 Relative eo [%] -.59. () Calibation of the oveall heat balance DPM enthalpy souce[w] -81-4841 Net total heat flux ove all the boundaies [W] 78 4947 Relative eo [%].148 -.46 * it s the maximum value coesponding to 1% cha convesion; O souce and othes ae not given hee. 5. CFD VALIDATION WITH EXPERIMENTAL DATA The detailed gas species wee measued at the gid points (6 6=6 along the axis diection; 9 along the adial diection, as shown in Fig. 1). Figue and Figue 4 show the compaison between the measued main species and the CFD pedicted values on two epesentative measuing lines, y=.75m and y=.5m, espectively. The CFD esults show a geneally acceptable ageement with the measued data, except fo CO mole faction, in paticula on the line y=.5m whee CO concentation is too low to be measued accuately. RKE makes a much bigge diffeence with SKE in the coe egion than in the nea-wall egion (i.e., Fig. vs. Fig. 4), just as expected. Fo combusting swiling flows, the fidelity of CFD modeling is mainly detemined by the accuacy of the tubulence model in the coe egion, athe than in the nea-wall zone. The custom DPM laws make quite significant diffeence with the default DPM laws in the pedicted gas species, especially in the coe zone, which is out of expectation.
Hoiba (=cm) Hoiba (=45cm) FTIR (=cm) FTIR (=45cm) taditional SKE-default Hoiba (=15cm) Hoiba (=6cm) FTIR (=15cm) FTIR (=6cm) taditional SKE-default.18.15 Mole faction of CO [-].15.1.9. Mole faction of CO [-].1.11.9.7.5. Axial position, x [m], along line y=.75m (= o 45cm) Axial position, x [m], along the line y=.5m ( =15 o 6cm) Hoiba (=cm) Hoiba (=45cm) taditional SKE-default Hoiba (=15cm) Hoiba (=6cm) taditional SKE-default Mole faction of O [-].1.18.15.1.9. Axial position, x [m], along line y=.75m ( = o 45cm) Mole faction of O [-].1.11.9.7.5..1 Axial position, x [m], along line y=.5m ( =15 o 6cm).18 FTIR (=cm) FTIR (=45cm) taditional SKE-default.14 FTIR (=15cm) FTIR (=6cm) taditional SKE-default Mole faction of H O [-].15.1.9. Axial position, x [m], along line y=.75m (= o 45cm) Mole faction of H O [-].1.1.8.4 Axial position, x [m], along line y=.5m ( =15 o 6cm) Hoiba (=cm) Hoiba (=45cm) FTIR (=cm) FTIR (=45cm) taditional SKE-default.1 Hoiba (=15cm) Hoiba (=6cm) FTIR (=15cm) FTIR (=6cm) taditional SKE-default.5 Mole faction of CO [-].8.4. Mole faction of CO [-]..15.1.5 Axial position, x [m], along line y=.75m (= o 45cm) Axial position, x [m], along line y=.5m (=15 o 6cm) Fig. CFD vs. measuements on line y=.75m Fig. 4 CFD vs. measuements on line y=.5m
The esults of the custom paticle convesion model show the tempeatue gadient inside the staw paticles is small: only a few degees in diffeence between the paticle suface and the paticle cente. The epoted staw cha convesion ates in the eacto ae also close to each othe when diffeent DPM laws ae used fo staw paticle convesion. Fo instance, staw cha convesion ates ae 67% and 7% fo the cases, RKE-default and RKE-custom, espectively. As a esult, the custom DPM laws may be expected not to have a big influence on the pedicted gas species in the eacto. The elatively big diffeence, exhibited in Fig., might be due to the too high DPM enthalpy souce fom the custom laws, in elative to the default laws. The fou cases, which ae based on Jones & Lindstedt s global eaction mechanism with EDC model used fo tubulence-chemisty inteaction, make significant diffeence with the taditional case which is based on simplified two-step mechanism with EDM. Jones & Lindstedt s fou-step eaction mechanism was deived using analysis olame stuctues and the final kinetic paametes fo the esulting ate equations wee detemined by compaison with expeimental data fo pemixed methane and popane flames, along with diffusion flame data fo a methane-ai flame. Thei global eaction mechanisms have been found to have good ageement fo a ange of paametes such as flame speed, flame thickness, and species pofiles [5]. Jones & Lindstedt s schemes ae believed to pefom much bette than the ove-simplified two-step global mechanism to descibe high tempeatue oxidation of gaseous alkane hydocabons up to butane. EDM has been widely used in combustion simulation, fo example [8], which pesents also a CFD modeling of co-fiing pulveized coal and biomass. EDM is used only in the taditional case in ou study, fo compaison. One of the dawbacks of EDM is the lack of geneality of the model constants, A and B, which ae vey case-dependent. Moeove, EDM will most likely to poduce incoect solutions when multi-step eaction mechanisms ae used [4]. Multi-step eaction mechanisms ae based on chemisty kinetics, which diffe fo each eaction; whilst in EDM evey eaction has the same, tubulent ate and theefoe the same eaction ate. As a esult, Jones & Lindstedt s eaction mechanism (i.e., mechanism-) with EDC model is expected to bette pedict the species distibution than the simplified -step mechanism (i.e., mechanism-1) with EDM. Figues and 4 do show this tend, fom the compaison between the two cases, taditional vs. RKE-default. Thee ae still some discepancies between the measued data and the CFD esults, as shown in Fig. and Fig. 4, which can be attibuted to both the expeimental and modeling sides. (1) Fom the geometies of the bune and eacto and the opeational conditions, one would expect an axisymmetic combustion flow in the eacto, based on which axisymmetic swil flow is used in all the CFD modeling. Howeve, the measued mapping of the species indicates that the eacting flow in the eacto is not axisymmetic, which can also be seen in Fig. and 4. () The measuements wee done on diffeent days. Though effots wee made to maintain the same opeation conditions thoughout the measuing peiod, diffeences in opeations wee inevitable, which would lead to inconsistent measued esults. () Fo the CFD modeling, the uncetainties with the fuels may be one of the majo souces fo the discepancies. The two fuels ae not eally known well enough to cay out a eliable CFD modeling of such a co-fiing flame, although the poximate analysis, the ultimate analysis and the paticle sizes of both the fuels ae given. The knowledge of the kinetic paametes (e.g., appopiate devolatilization and cha oxidation sub-models) and the physical popeties (e.g., paticle density and specific heat) also plays a vital ole in CFD modeling of solid fuel combustion and demands additional fuel chaacteization expeiments. Unfotunately, most of the existing CFD modeling effots fo solid fuel combustion may have been done without such knowledge. In this study, we may ague that the CFD modeling could be based on the cuent best available knowledge of the kinetic paametes and the physical popeties of the staw and coal fied in this eacto. The sensitivity study of those paametes, which is not pesented in this pape, does show that they have vey significant influence on the CFD esults. As a conclusion of this section, RKE-default has been ecommended as the baseline case at this stage, patly fom expectation and patly fom the compaison with the expeimental data. This baseline case can be used fo futhe investigation of the bune and also fo the NO x pediction. The detailed CFD esults of RKE-default ae shown in Fig. 5, along with those of the taditional case (shown in Fig. 6), which ae used fo compaison. Compaed to the taditional case, the main eaction and flame pedicted by RKE-default tend to be confined to a long and naow coe zone aound the centeline of the eacto. (a) Mola faction of CO [-] (b) Mola faction of H O [-] (c) Mola faction of O [-] (d) Mola faction of CO [-] (e) Gas tempeatue [K] (f) Velocity vecto gey-scaled by gas tempeatue [K] Fig. 5 CFD esults of case RKE-default
(a) Mola faction of CO [-] (b) Mola faction of H O [-] (c) Mola faction of O [-] (d) Mola faction of CO [-] (e) Gas tempeatue [K] (f) Velocity vecto gey-scaled by gas tempeatue [K] Fig. 6 CFD esults of case taditional 6. CONCLUSIONS A modeling appoach fo co-fiing pulveized coal and biomass in swiling-stabilized bunes is pesented and demonstated in this pape, in which ealizable k-ε model is used fo tubulence and Jones & Lindstedt s global eaction schemes with EDC model ae used fo homogeneous combustion ouel-volatiles. The CFD esults ae calibated fist and then validated with measued gas species mapping. The following conclusions can be dawn fom this study. Realizable k-ε model pefoms bette than standad k-ε model in pedicting swiling combustion flows, which is not supising. Howeve, the main diffeence is confined to the coe zone. In the zone fa away fom the coe zone, the diffeence is negligible. Although the simplified -step global mechanism with EDM has been vey widely used in CFD modeling of industial coal and/o biomass combustion pocesses, Jones & Lindstedt s global eaction schemes with EDC is expected to be a bette option. Though Jones & Lindstedt s schemes can not be as geneally applicable as detailed elementay mechanism, they do show excellent ageement with measued majo species pofiles fo a wide ange of pemixed and non-pemixed flames. As a esult, Jones & Lindstedt s global schemes ae expected to poduce bette combustion pedictions than the ove-simplified -step global mechanism. EDC model is geneal concept based on a detailed desciption of the dissipation of tubulent eddies, and allows fo multi-step eaction mechanism. This study does demonstate that Jones & Lindstedt s global eaction schemes with EDC model is a bette option in pedicting coal/biomass co-combustion flames. Knowledge about the solid fuels plays a vital ole in CFD modeling of solid-fuel flames. Howeve, such knowledge is vey often incomplete. Besides the poximate analysis and ultimate analysis, additional fuel chaacteization expeiments ae absolutely needed to povide the knowledge of the kinetic paametes, as well as the physical popeties of the solid fuels fied, in ode to cay out a eliable CFD modeling. The sensitivity analysis done in this study does show that the kinetic paametes and the physical popeties of the solid fuels have a geat influence on the pedicted species and tempeatue pofiles. As demonstated in ou ealie wok [], the shape and size of biomass paticles damatically influence thei tajectoy and thus thei convesion in a combusto. In this study, one could conclude that the inta-paticle heat and mass tansfe may be a seconday issue at most in the convesion of the biomass paticles. Fo the staw paticles fied in this eacto, the tempeatue diffeences inside the paticles ae quite small and the custom DPM laws do not eally pedict a significantly impoved cha convesion ate in this eacto, in compaison with the default DPM laws (7% in the fome vs. 67% in the latte). Moe expeiments ae called fo the study of pulveized coal and biomass co-combustion and its modeling, in ode to povide moe complete knowledge about the solid fuels themselves, to collect data fo single biomass paticle convesion, to poduce consistent, quantitative maps of not only gas species but also gas tempeatue fo the pupose of CFD validation. 7. ACKNOWLEDGEMENT This eseach was financially suppoted by Gants PSO 485 and 486. 8. REFERENCES 1. Yin, C., Rosendahl, L, Kæ, S.K., Gate-Fiing of Biomass fo Heat and Powe Poduction, Pogess in Enegy and Combustion Science, 4 (8), pp.75-754.. Damstedt, B, Pedesen, J.M., Hansen, D., Knighton, T., Jones, J., Chistensen, C., Baxte, L. Tee, D. Biomass cofiing impacts on flame stuctue and emissions. Poceedings of the Combustion Institute, 1 (7), pp.81-8.. Yin, C., Rosendahl, L, Kæ, S.K. and Conda, T., Use of Numeical Modeling in Design fo Co-Fiing Biomass in Wall-Fied Bunes, Chemical Engineeing Science, 59 (4), pp.81-9. 4. FLUENT Inc., FLUENT 6..6 Use s Guide, 6. 5. Jones, W.P. and Lindstedt, R.P., Global Reaction Schemes fo Hydocabon Combustion, Combustion and Flames, 7 (1988), pp.-49. 6. Gea, D., Mathu, M.P., Feeman, M.C. and Robinson, A., Effect of Lage Aspect Ratio of Biomass Paticles on Cabon Bunout in a Utility Boile, Enegy & Fuels, 16 (), pp.15-15. 7. Di Blasi, C., Heat, Momentum and Mass Tanspot Though a Shinking Biomass Paticle Exposed to Themal Radiation, Chemical Engineeing Science, 51 (1996), pp.111-11. 8. Backeedy, R.I., Fletche, L.M., Jones, J.M., Ma, L., Poukashanian, M. and Williams, A., Co-Fiing Pulveized Coal and Biomass: A Modeling Appoach, Poceedings of the Combustion Institute, (5), pp.955-964.