Chemical Engineering Science 55 (2000) 2931}2944

Dynamic behaviour of strati"ed downdraft gasi"ers

Colomba Di Blasi*Dipartimento di Ingegneria Chimica, Universita& degli Studi di Napoli **Federico II++, P.le V. Tecchio, 80125 Napoli, Italy

Received 20 November 1998; received in revised form 13 September 1999; accepted 8 November 1999

Abstract

A one-dimensional unsteady model is formulated for biomass gasi"cation in a strati"ed concurrent (downdraft) reactor. Heat andmass transfer across the bed are coupled with moisture evaporation, biomass pyrolysis, char combustion and gasi"cation, gas-phasecombustion and thermal cracking of tars. Numerical simulation has allowed to predict the in#uence of model parameters, kineticconstants and operational variables on process dynamics, structure of the reaction front and quality of the producer gas. In particular,two di!erent stabilization modes of the reaction front have been determined. For high values of the air-to-fuel ratio and of the primarypyrolysis rate, the process is top-stabilized, resulting in a high conversion e$ciency and good gas quality. As the air #ow is decreasedbelow a critical limit value, the reaction front becomes grate-stabilized. The two di!erent con"gurations are largely determined by thegas-phase combustion of volatile pyrolysis products. Finally, the predictions of the gas composition and the axial temperature pro"lesare in agreement with experimental data. ( 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Mathematical modelling; Simulation; Packed bed; Gasi"cation; Biomass

1. Introduction

Biomass and waste are widely recognized to be a majorpotential for energy production. Gasi"cation enablesconversion of this material into combustible gas, mech-anical and electrical power and synthetic fuels and chem-icals. In principle, the gasi"cation units employed for coalcan also be applied for biomass and waste, but signi"cantdi!erences exist between the two fuel categories. Coalpyrolysis yields 60}80% char the balance coming fromgases and tars. Updraft, countercurrent gasi"ers are wellsuited for the conversion of low-reactive char into gas(around 90% of the coal gasi"ed in the world makes useof this con"guration (Hobbs, Radulovic & Smoot, 1993)).When biomass is pyrolyzed, gases and tars represent70}90% of the total mass fed, whereas only 30}10% isa highly reactive char. In updraft air gasi"cation, theoxygen is consumed at the grate, essentially throughpartial combustion of char. The resulting hot gases causechar gasi"cation and biomass pyrolysis. The relativelylow temperatures and the absence of oxygen result inlarge amounts of tars in the producer gas.

*Tel.: 39-081-7682232; fax: 39-081-2391800.E-mail address: diblasi@unina.it (C. D. Blasi)

Downdraft gasi"ers, characterized by the concurrent#ows of solid fuel and gas, show little #exibility withrespect to the fuel moisture content and size. However,these units are usually preferred for small-scale processes(below 500 kWe) because they produce a cleaner gas,resulting in a less complicated cleaning process. In theclassical design, the packed bed is supported acrossa constriction (throat), where most of the gasi"cationtakes place. Air is fed above this zone and, due to therestriction, the #ow is highly turbulent, thus favouringmixing. The more recent version is strati"ed (or open-core), where there is no restriction and the bed is sup-ported on a grate. This design has been shown to copewell with problems deriving from the gasi"cation of loosematerials, that is, poor oxygen distribution with conse-quent instabilities and ash melting (Buekens, Bridgwater,Ferrero & Maniatis, 1990).

Di!erences between the countercurrent and concur-rent gasi"cation are also due to the processes whichdetermine the composition of the producer gas. In the"rst case, the quality of the gas is largely dependent onthe presence of devolatilization products. On the con-trary, the heating value of the producer gas of concurrentgasi"ers is determined by the amount of carbon monox-ide and hydrogen present. Both species are formed by thecracking of tars and the slow heterogeneous gasi"cation

0009-2509/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved.PII: S 0 0 0 9 - 2 5 0 9 ( 9 9 ) 0 0 5 6 2 - X

Nomenclature

Ai

pre-exponential factor (Eqs. (18)}(22), (24), (25),(27))

Bg

permeability to gas #ow, m2

c speci"c heat, J/kg KC molar concentration, kmol/m3

d particle diameter, mD reactor diameter, mD

idi!usion coe$cient, m2

Ei

activation energy, kJ/molh heat transfer coe$cient, W/m2 KH

ispeci"c species enthalpy, kJ/kg

km

mass transfer coe$cient, m/skHm

maximum value of the mass transfer coe$cient,m/s

mM

moisture evaporation rate, kg/m3 sM molecular weightp gas pressure, kPapvs

vapor pressure expressed by theClausius}Clapeyron equation

Pr particle Prandtl numberrc

current particle radius, mR initial particle radiusR universal gas constantRe particle Reynolds numberR

jreaction rate, kg/m3 s

Sc particle Schmidt numbert time, s¹ temperature, K; total moisture content, d.b., kg/kg;

ggas velocity, m/s

;s

solid velocity, m/s< particle volume, m3

=a

air feed rate, kg/h=

bbiomass feed rate, kg/h

z space, m

Greek letters

a stoichiometric coe$cientb stoichiometric coe$cientc stoichiometric coe$cient*h reaction enthalpy, kJ/kge porosityjH thermal conductivity, W/mKK moisture (evaporation) enthalpy, kJ/kgk viscosity, kg/mslp

particle density number, 1/m

oc0

constant bed density in the combustion/gasi"-cation zone, kg/m3

oi

gas phase (i"O2, CO,2) mass concentration

(mass/gas volume), kg/m3

ok

apparent solid (k"B, M) density (mass/totalvolume), kg/m3

p Stephan}Boltzmann constants ash content of the biomass, % of initial dry

massu

irate of species production (devolatilization),kg/m3 s

f correction factor for the solid/gas heat transfercoe$cient

Subscripts

B biomass (wood)c1 tar combustionc2 methane combustionc3 carbon monoxide combustionc4 hydrogen combustionc5 char combustionC CHARCH

4methane

CO carbon monoxideCO

2carbon dioxide

E equilibriumg total volatiles (vapour#gas)gw gas/wallg1 carbon dioxide gasi"cationg2 hydrogen gasi"cationg3 steam gasi"cationH

2hydrogen

H2O steam

i gas-phase chemical speciesj chemical reactionk solid-phase chemical speciesM moistureO

2oxygen

p1 primary pyrolysisp2 secondary pyrolysiss solidsg solid/gassw solid/wall¹ TARv vapourw wallwg water gas shift0 ambient value

of char, so that the size of the reduction zone and theresidence time of the gasifying agents play a fundamentalrole on the performances of the gasi"er. These are, on theother hand, dependent not only on the size of the gasi"er

but also on the position where the reaction zone stabil-izes (Reed & Markson, 1985).

Kinetics-free, equilibrium models can predict the exitgas composition, given the solid composition and the

2932 C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944

Table 1Model equations

* biomass

L.B

Lt#

L;s.

BLz

"!Rp1

, (1)

* moisture

L.M

Lt#

L;s.

MLz

"!mM

, (2)

* gas-phase species

eL.

iLt

#

L(.i;

g)

Lz"

LLzAeDi

.g

L>i

Lz B#Mi+j

lijR

j#u

pi,

i"O2, H

2, CO,CO

2, CH

4, j"c1}c4, g1}g3,wg, (3}7)

* steam

eL.

H2OLt

#

L(.H2O;

g)

Lz"M

H2O+j

lH2Oj

Rj

#

LLzAeDv

.g

L>H2O

Lz B#mM#u

pH2O, j"c1}c4, g1}g3, wg, (8)

* vapor-phase tar

eL.

TLt

#

L(.T;

g)

Lz"

LLzAeDT

.g

L>T

Lz B#lTR

p1!R

p2, (9)

* nitrogen

.N2

".g! +

iEN2

.i, (10)

* total gas continuity

eL.

gLt

#

L(.g;

g)

Lz"+

i

+j

lijM

iR

j#m

M#(1!l

C)R

p1, (11)

i"N2,O

2, H

2, CO,CO

2, CH

4,H

2O, j"c1}c4, g1}g3,wg,

* solid-phase energy

L(+i.

iH

i)

Lt"

LLz AjHs

L¹s

Lz B#L(;

s+

i.

iH

i)

Lz!+

j

Rj*H

j

!Qsg#Q

sw!m

MK, (12)

Hi"c

si(¹

s!¹

0), i"B,C,M, j"c5, g1}g3,p1,

* gas-phase energy

eL(+

i.

iH

i)

Lt"

LLzAjHg

L¹g

Lz B!L(;

g+

i.

iH

i)

Lz#Q

sg#Q

gw

!+j

Rj*H

j, (13)

hi"c

gi(¹

g!¹

0), i"N

2, O

2, H

2, CO, CO

2, CH

4, H

2O,T,

j"c1}c4,wg, p2,

Qsg"h

sglp(¹

s!¹

g), Q

sw"

4hsw

D(¹

w!¹

s),

Qgw

"

4hgw

D(¹

w!¹

g),

* ideal gas law, modi"ed Darcy law

P"

.gR¹

gM

g

, (14)

Bg

kLP

Lz";

s!;

g. (15)

equilibrium temperature, but they cannot be used forreactor design. Only in a very few cases chemical reactionkinetics and transport phenomena have been properlycoupled to model conventional (Groeneveld & vanSwaaij, 1980) and strati"ed (Manurung & Beenackers,1994) downdraft gasi"ers. However, the description ofthe `#aming pyrolysisa (Reed & Markson, 1985), that isbiomass pyrolysis and combustion of volatile pyrolysisproducts, is still based on equilibrium models or onhighly simpli"ed treatments, which assume the existenceof a stable combustion zone with in"nitely oxygen con-version rate near the air inlet. Furthermore, these modelsdescribe steady-state conditions. Thus, they do not allowthe prediction of the dynamic behaviour of strati"edgasi"ers and of the di!erent modes of stabilization of thereaction front and thus the size of the reduction zone.This study proposes a more advanced, dynamic model,which includes "nite rate kinetics for biomass pyrolysisand combustion of char, gaseous species and tars. Themodel was used to simulate the e!ects of changes intransport coe$cients, chemical kinetics and operatingconditions on the performances of the gasi"cation pro-cess, in view of reactor design and optimization, espe-cially in relation to the dynamic behaviour of downdraftreactors, i.e. to top- or grate-stabilized operation.

2. Mathematical model

The strati"ed gasi"er model is based on mass andenergy balances for the solid phase and mass and energybalances for the gas phase, written for a one-dimensional,unsteady state system. Species considered are: oxygen,nitrogen, hydrogen, steam, carbon dioxide, carbon mon-oxide, methane and hydrocarbons (which also includetars). The pressure drop in the reactor is modelled usingthe generalized Darcy law but, given the large bed per-meabilities, simulations have been carried out with theassumption of constant pressure (the gas velocity is deter-mined from the continuity equation and the density ofthe mixture from the ideal gas law). Model equations arelisted in Table 1. As the concurrent #ows of solid and gasdescend across the reactor, several processes take place,namely, moisture evaporation, biomass pyrolysis, charcombustion and gasi"cation, combustion of the gasesand thermal cracking of the tars, in accordance with theschematic representation reported in Fig. 1. The mainfeatures of submodels applied for these processes can bederived from Table 2 and the relevant approximationsmade are listed below.

Recently, moisture evaporation has been described insome detail for a single biomass particle (Di Blasi, 1998).For conditions similar to those of "xed-bed gasi"cation,it has been shown that both gas-phase and liquid-phasetransport phenomena can play a controlling role. How-ever, the single-particle e!ects are usually neglected in the

C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944 2933

Fig. 1. Schematic of the strati"ed concurrent (downdraft) gasi"er.

Table 2Moisture evaporation, chemical reactions and transport coe$cients

Evaporation/condensation

mM"l

pkm(.

v!.

H2O), (16)

pv/p

vs"exp[(17.884!0.1423¹

s#23.63]10~5¹2

s)1.0327!67.41

]10~5¹s)92U]. (17)

Pyrolysis reactions

B kp1P l

cCHAR#l

GGAS

1#l

T¹AR, (p1)

¹AR kp2P GAS

2(p2)

Rp1

"Ap1

expA!Ep1

R¹sB. B

, (18)

Rp2

"eAp2

expA!E

p2R¹

gB. T

. (19)

Gas-phase combustion

(tar) CH1.522

O0.0228

#0.867O2

kc1P CO#0.761H

2O, (c1)

CH4#1.5O

2kc2P CO#2H

2O (c2)

2CO#O2

kc3P 2CO

2, (c3)

H2#O

2kc4P 2H

2O, (c4)

Rj"A

jexpA!

Ej

R¹gB¹g

C0.5i

CO2

, j"c1, c2, i"¹, CH4, (20)

Rc3"A

c3expA!

Ec3

R¹gBCCO

C0.25O2

CH2O

, (21)

Rc4"A

c4expA!

Ec4

R¹gBCH2

CO2

. (22)

Gas-phase water gas shift

CO#H2O

kwg

H CO2#H

2, (wg)

Rwg

"ekwgACCO

CH2O

!

CCO2

CH2

KEB, (23)

kwg

"Awg

expA!E

wgR¹

gB, K

E"A

EexpA

EE

R¹gB. (24)

KE"A

EexpA

EE

R¹gB. (25)

Heterogeneous reactions of char

CHaOb#cO2

kc5P A2!2c!b#

a2BCO#A2c#b!

a2!1BCO

2

#a2H2O, (c5)

CHaOb#CO2

kg1P 2CO#bH

2O#A

a2!bBH2

, (g1)

CHaOb#A2!a2#bBH2

kg2P CH

4#bH

2O, (g2)

CHaOb#(1!b)H2O kg3P CO#A1!b#

a2BH2

, (g3)

formulation of reactor models, because the characteristictimes of moisture evaporation are orders of magnitudeshorter than those of char gasi"cation. Therefore, in mostcases the process is assumed to take place instanta-neously (for instance Yoon, Wei & Denn, 1978) or to bea di!usion limited process (Hobbs et al., 1993). The lattertreatment is retained in this study with the use of anempirical expression for the vapour pressure, to take intoaccount the di!erences in evaporation rates for capillaryand bound water (Di Blasi, 1998) (p

vs(¹

s) is the

Clausius}Clapeyron expression).It is well known that pyrolysis of large biomass par-

ticles is a process controlled by heat and mass transfer(for instance, Di Blasi, 1996a). More speci"cally, theconversion times are largely determined by the rate ofinward heat conduction, whereas product distribution ishighly dependent on intra-particle residence times ofvolatiles and thus on the extent of secondary cracking.Given the very steep temperature pro"les along the gasi-"er axis in the concurrent con"guration, it is often (Bliek,1984) assumed that intra-particle heat transfer resist-ances are much lower than extra-particle axial resist-ances. Thus, the particle is assumed to be thermally thin.In this study, this treatment is still retained but apparent

2934 C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944

Table 2 (Continued)

Rj"l

p

Ci

(1/km)#(1/k

j), k

j"A

jexpA!

Ej

R¹sB, j"c5, g1!g3,

i"O2, CO

2,H

2, H

2O (26}27)

Arc

RB3"(1!s)X#s, (28)

X"

;s

;s0

, (29)

lp(rc)"

3(1!e)rc

, (30)

.c0

L;s

Lz"!M

C(R

c5#R

g1#R

g2#R

g3). (31)

Solid/gas heat transfer coe$cient, mass transfer coe$cient

hsg"f

2.06cpg

.g;

ge

Re~0.575Pr~2@3, (32)

km"

2.06cpg;

ge

Re~0.575Sc~2@3. (33)

Properties

jHg"ej

g, jH

s"ej

rg#e

js

(js/(dj

rs)#1.43(1!1.2e))

, (34)

jrg"4p0.05¹3

g, (35)

jrs"4p0.85¹3

s, (36)

js"0.0013#0.05(¹

s/1000)#0.63(¹

s/1000)2

J

ms K(37)

jg"4.8]10~4¹6.717

g

J

ms K(38)

k"1.98]10~5(¹g/300)2@3

kg

sm(39)

Table 3Devolatilization data: gas composition is expressed as percent of theinitial dry biomass

I II III

lC

0.33 0.41 0.33lG

0.48 0.4 0.5lT

0.19 0.19 0.2

Gas 1 CO 7.5% 5.5% 11%CO

213% 10.5% 11%

H2

1% 0.02% 1%CH

41.5% 1% 2%

H2O 25% 23% 25%

Gas 2 CO 9.5% 9.5% 13%CO

25.7% 5.7% 3%

CH4

3.8% 3.8% 4%

kinetics are used for the pyrolysis process. A one-stepglobal reaction is considered, where the fractions ofgases, tars and chars (Di Blasi, 1993) produced and thegas composition should be speci"ed. Tars undergo sec-ondary cracking in the void spaces of the bed (one-stepglobal reaction), to produce secondary gases, whose com-position should again be speci"ed. The size (volume) andthe solid velocity of the spherical particles remain con-stant as the bed density decreases during moisture evap-oration and biomass devolatilization. Consequently, theporosity of this portion of the bed varies. However, thise!ect is not taken into account as single-particle simula-tions (Di Blasi, 1997) show that porosity variation exertsa negligible in#uence on the devolatilization character-istics.

For primary wood pyrolysis, apparent activation ener-gies are reported to be in the range 63}125 kJ/mol de-pending on the wood species, sample size and heatingconditions (Roberts, 1970). Reference values of the ac-tivation energy and pre-exponential factor are chosen as:

Ap1

"1.516]103 s~1, Ep1

"105 kJ/mol, determinedfor beech wood cylinders 1}2 cm thick. Kinetic constantsfor tar cracking from Liden, Berruti and Scott (1988) areused. While the composition of the secondary gas hasbeen estimated on the basis of literature data obtainedfor wood (Boroson, Howard, Longwell & Peters, 1989),an experimental study (Di Blasi, Signorelli, Di Russo& Rea, 1999) has been carried out to determineprimary pyrolysis product yields and gas composition forconditions corresponding to downdraft gasi"cation.Packed-beds (diameter 4 cm) of biomass particles havebeen exposed to di!erent radiation intensities, whichcorrespond to maximum (surface) temperatures in thebed in the range 650}1000 K. A continuous nitrogen #owacross the bed reduces the residence times of volatileproducts and makes the activity of extra-particle second-ary reactions negligible. As expected, devolatilizationcharacteristics, which are representative of intra-particleprimary and secondary reactions, have been found todepend both on the heating conditions and the biomasstype. Assuming a surface temperature of 850 K, three setsof devolatilization data ha ve been considered for thesimulation of downdraft gasi"cation (Table 3). The "rstdata set (I) refers to 0.5}1 cm thick wood chips, thesecond (II) to rice husks and the third (III) has beenconsidered for a parametric investigation of the e!ects ofvariations in pyrolysis products on the gasi"cation char-acteristics.

Combustion of volatile products is an important pro-cess in downdraft gasi"cation. Here, the treatment pro-posed by Bryden and Ragland (1996) for the "xed-bedcombustion of biomass is adopted. Hydrocarbons, whichinclude tars and methane, react with oxygen to formwater vapour and carbon monoxide, according toa "nite-rate, global reaction. The formation of the reac-tion intermediate, carbon monoxide, is important for thecorrect prediction of the ignition delay time. Global,

C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944 2935

"nite rate reactions are also considered for the combus-tion of carbon monoxide and hydrogen. However, forcomputational simplicity, instead of an in"nitely fast rateof hydrogen consumption, a high pre-exponential factorand a low activation energy have been introduced fora second-order reaction rate (E"83 kJ/mol, A"

105 m3/s/mol). Tars are modelled as hydrocarbonsCH

1.522O

0.0228(Bryden & Ragland, 1996), with molecu-

lar weight equal to 95 (Corella, Aznar, Delgado & Aldea,1991).

As for homogeneous reactions, the water gas shiftreaction is also considered and is described by "nite ratekinetics, with the equilibrium constant reported by Yoonet al. (1978) and the kinetic constants derived by Biba,Macak, Klose and Malecha (1978).

Combustion and gasi"cation reactions of char are het-erogeneous and are described by the unreacted core,shrinking particle model, where a steep reaction zonepropagates through the isothermal char. Two mecha-nisms responsible for the global reaction rate are con-sidered: di!usion through the gas "lm, surrounding theparticle, and intrinsic chemical kinetics. To account forthe simultaneous e!ects of the two resistances, an e!ec-tive volumetric reaction rate is introduced, based on theassumption of a linear dependence of the reaction rate onthe oxidizing/gasifying species concentration. As a conse-quence of the heterogeneous reactions, the particle dia-meter shrinks and the density of the bed (and porosity)remains constant, causing a gradual decrease in the solidvelocity. The minimum size of the particle and conse-quently the maximum particle density number (l

p(Eq.

(23))) depend on the ash content of the solid (Hobbs et al.,1993).

For simplicity, chars are assumed to consist of purecarbon (a"b"0) and the products of heterogeneouscombustion to be carbon dioxide only (c"1), thoughthis assumption is not valid in general (Hobbs et al.,1993). The intrinsic kinetics of wood char combustion aredescribed according to the data determined for cellulosicfuels (Kashiwagi & Nambu, 1992). Kinetics of char gasi"-cation by carbon dioxide and steam are those reportedby Groeneveld and van Swaaij (1980). Hydrogasi"cationis usually negligible under atmospheric conditions and itsrate is assumed to be slower by three orders of magnitudecompared with carbon dioxide and steam gasi"cation(Hobbs et al., 1993).

Literature correlations are used for the e!ective ther-mal conductivity of the bed (Goldman, Xieu, Oko, Milne& Essenhigh, 1984), the e!ective bed-to-wall heat transfercoe$cients (Hobbs et al., 1993) (not reported in Table 2),the solid/gas heat transfer and the mass transfer coe$-cients (Gupta & Thodos, 1963). However, the solid/gasheat transfer coe$cient estimated from non-reactingsystem data can exceed the experimental values inreacting gasi"ers (Hobbs et al., 1993), as a consequenceof unsteady heat transfer. Therefore, the experimental

correlation is multiplied by empirical factors (f) withvalues in the range 0.02}1 (Cho & Joseph, 1981;Radulovic, Ghani & Smoot, 1995). In this study thereference value for f is taken equal to 1. Some di$cultiesalso exist in the application of the Gupta and Todoscorrelation for the mass transfer coe$cient, mainly be-cause of the di!erent scales of the conversion units (inparticular, there is large uncertainty for low Reynoldsnumbers) and again the changes introduced by chemicalreactions. Apart from the correlation suggested by Bhat-tacharya, Salam, Dudukovic and Joseph (1986), in allcases the mass transfer coe$cient decreases as the par-ticle size is increased, but di!erences become very high asthe particle diameter reduces to very low values. This iscritical, because in the unreacted-core, shrinking particlemodel, the maximum temperature predicted in the oxida-tion/reduction zone becomes highly dependent on thecorrelation used. On the other hand, at high temper-atures, "lm transfer becomes the controlling resistance.To avoid unrealistic temperature values, corrective fac-tors, which limit the maximum value of k

m(Bhattacharya

et al., 1986) or reduce its value for all conditions (forinstance, 0.5 in Goldmann et al., 1984) have been sugges-ted. Both procedures have been applied to the Gupta andTodos correlation, used in this model. Given the propervalues of the correction factor or its maximum allowed inthe course of the reaction, the two procedures do notresult in any change from the qualitative point of view inthe predictions of process characteristics. Therefore, onlythe e!ects associated with the changes in the maximumvalue of k

m(indicated as kH

m) on the prediction of the

gasi"cation process will be discussed. The reference valuechosen for kH

mis 0.045 m/s.

The variation of the gas thermal conductivity andviscosity with temperature is described as in Purnomo,Aerts and Ragland (1990). A constant value is assumedfor di!usivities (0.2]10~4 m2/s) and speci"c heats ofgaseous species (evaluated for a temperature of 1000 K)and speci"c heats of biomass and char (1.34 kJ/kg) (DiBlasi, 1996b). All the thermochemical data referred to orlisted in this section, with the inclusion of the devolatiliz-ation data set I, will be indicated in the following asreference data (a summary of kinetic constants is re-ported in Table 4).

The numerical solution of the model equations isbased on operator splitting procedures and "nite-di!er-ences approximations. The reactor is divided into a set ofelementary cylindrical cells, whose cross sections co-incide with the reactor cross section, while the height canbe variable. The solution procedure is divided into threestages, corresponding to chemical reaction processes,heat exchange (between phases and with the reactor wall)and transport phenomena. For each time step, in the "rsttwo stages, the solution is calculated, for each controlvolume, of linearized ordinary di!erential equations, bymeans of a "rst-order implicit Euler method. In the third

2936 C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944

Table 4Reference value for kinetic constants

PyrolysisA

pl(s~1) 1.516]103 Roberts (1970)

Epl

(kJ/mol) 105 Roberts (1970)A

p2(s~1) 4.28]106 Liden et al. (1988)

Ep2

(kJ/mol) 107 Liden et al. (1988)*h

p1(kJ/kg) !420 Di Blasi (1996b)

*hp2

(kJ/kg) 42 Di Blasi (1996b)

Char combustionA

c5(s~1) 5.67]109 Kashiwagi and Nambu (1992)

Ec5

(kJ/mol) 160 Kashiwagi and Nambu (1992)*h

c5(kJ/kg) 2.5]107 Kashiwagi and Nambu (1992)

Char gasi"cationA

g1"A

g3(m3/mol s) 7.92]104 Groeneveld and van Swaaij (1980)

Ag2

(m3/mol s)! 79.2 Groeneveld and van Swaaij (1980)Eg1"E

g2"E

g3(kJ/mol) 218 Groeneveld and van Swaaij (1980)

*hg1

(kJ/kg) !9.3]106 Groeneveld and van Swaaij (1980)*h

g2(kJ/kg) 7.2]106 Groeneveld and van Swaaij (1980)

*hg3

(kJ/kg) !6.4]106 Groeneveld and van Swaaij (1980)

Water-gas shiftA

E(!) 0.0265 Yoon et al. (1978)

EE

(kJ/mol) 65.8 Yoon et al. (1978)A

wg(m3/mol s) 2.78 Biba et al. (1978)

Ewg

(kJ/mol) 12.6 Biba et al. (1978)*h

wg(kJ/mol) 41 Biba et al. (1978)

Gas-phase combustionA

c1"A

c2((m3/mol)0.5/sK) 9.2]106 Bryden and Ragland (1996)

Ec1"E

c2(kJ/mol) 80 Bryden and Ragland (1996)

*hc1

(kJ/kg) 17 473 Bryden and Ragland (1996)*h

c2(kJ/kg) 50 190 Bryden and Ragland (1996)

Ac3

((m3/mol)0.75/sK) 1017.6 Bryden and Ragland (1996)Ec3

(kJ/mol) 166 Bryden and Ragland (1996)*h

c3(kJ/kg) 10 107 Bryden and Ragland (1996)

Ac4

((m3/s mol)! 1]1011 Bryden and Ragland (1996)Ec4

(kJ/mol)! 42 Bryden and Ragland (1996)*h

c4(kJ/kg) 142919 Bryden and Ragland (1996)

!Estimated.

step the transport equations, after discretization with thehybrid scheme, are solved through a semi-implicit pro-cedure, that is, each conservation equation is implicit inthe corresponding variable being conserved, whereas theother variables are taken as the last available values. Forall stages, the equations for the condensed-phase vari-ables are solved "rst, followed by those of gaseous com-ponents and enthalpies. All the simulations presentedhere were carried out with a time step of 1]10~4 s andspace steps of 0.001 m, giving grid-independent solutionsfor the reference values of the process parameters.

3. Results

Simulations have been carried out for biomass par-ticles 1 cm thick, with an ash content of 10.5% and an

initial moisture content of 10% d.b. (dry basis). A pilotscale unit (ID 0.45 and 0.50 m high (Manurung &Beenackers, 1994) is considered, where the bed density is200 kg/m3 and the void fraction is 0.5. Ignition of the bedis caused by hot air. A steady-state scenario (time equalzero is referred to this initial condition), corresponding toa reaction zone located near the top of the reactor, hasbeen chosen as initial condition for the simulations byvarying model parameters (f, kH

m, devolatilization data),

kinetic constants (pyrolysis, gasi"cation, combustion)and the two operational variables: the biomass feed rateand the air-to-biomass ratio. A "rst set of simulations hasbeen carried out by varying the biomass feed rate (=

b) in

the range 7}18 kg/h for an air-to-fuel ratio of 1.5. Thee!ects of the air-to-fuel ratio (=

a/=

b) have also been

investigated for variations in the range 0.6}2.6 for a bi-omass feed rate of 18 kg/h. In this section, the e!ect of

C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944 2937

Fig. 2. A. Axial pro"les of solid and gas temperature and gas velocity(reference data, =

b"18 kg/h,=

a/=

b"1.5). B. Axial pro"les of spe-

cies molar fractions (reference data, =b"18 kg/h,=

a/=

b"1.5).

Fig. 3. A. Structure of the reaction front: oxygen molar fraction, charmass fraction and temperature pro"les (reference data,=

b"18 kg/h,

=a/=

b"1.5). B. Structure of the reaction front: axial pro"les of

reaction rates (reference data,=b"18 kg/h, =

a/=

b"1.5).

these parameters on the process characteristics are dis-cussed. Finally, a comparison between model predictionsand experimental measurements is given.

The main characteristics of the gasi"cation process canbe observed from Figs. 2 and 3, which report the axialpro"les of the main variables of interest at steady condi-tions for =

b"18 kg/h and =

a/=

b"1.5. The leading

edge of the reaction zone is stabilized slightly below thefeeding section (about 2 cm), so that the gasi"er is topstabilized. The air and biomass co-fed at the top of thereactor constitute a non-reactive fuel/oxidant mixture,which makes the #ows homogeneous, providing a uni-form air/fuel distribution to the zones below. This regionis followed by the evaporation of moisture and the partialcharring of the biomass particles, essentially because ofradiative/conductive heat transfer through the solid-phase. Indeed, oxygen soon meets the steep char front,causing its exothermic combustion and a large increasein the gas velocity. As a consequence of the increase in thesolid-phase temperature, the gasi"cation rates also attaina local maximum. However, the endothermicity of thisprocess and the convective cooling of the inlet gasquench the heterogeneous consumption of char. Then airmixes with volatile products, downstream of the charring

front, resulting in a pre-mixed #ame. Here, when thegaseous mixture attains temperatures su$ciently high,combustion of volatile pyrolysis products becomes animportant mechanism to provide heat for further chargasi"cation, thus increasing the hydrogen and carbonmonoxide content of the gas. Consequently, the particlesize and the solid velocity continuously decrease,whereas the particle density number increases. Homo-geneous, gas-phase combustion partially destroys thetars and, through high temperatures, also favours theircracking.

It should be noted that the moisture evaporation frontis very steep, probably as a consequence of the highlysimpli"ed description (di!usional resistance only), whichdoes not include intra-particle temperature and moisturegradients. The endothermicity of the process has beenfound to contribute signi"cantly to limit the maximumtemperature. Biomass devolatilization is also character-ized by spatial temperature and solid-phase speciesgradients much larger than those observed for the up-draft con"guration. Indeed, the concurrent solid and gas#ow towards the combustion zone makes the heat trans-fer di$cult towards the virgin solid region. Di!erencesbetween the solid and gas temperatures are signi"cant at

2938 C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944

Fig. 4. Conversion characteristics (particle volume, solid velocity andparticle density number pro"les) for several biomass feed rates (refer-ence data, =

a/=

b"1.5).

Fig. 5. Composition of the (dry) gas and yield of unreacted char (as% of the amount formed from pyrolysis) for several biomass feed rates(reference data, =

a/=

b"1.5).

Fig. 6. Axial temperature pro"les for several times (reference data,=

b"11 kg/h, =

a/=

b"1).

the leading edge of the reaction zone, but they tend todisappear in the reduction zone. Among the gasi"cationproducts, the most abundant is CO, followed by H

2and

CH4. The tar content of the gas is very small (0.2% of

the mass of the gaseous e%uents), while H2O and CO

2are still present in signi"cant amounts. On the whole, theprocesses of primary and secondary pyrolysis and thehomogeneous and heterogeneous combustion reactionstake place simultaneously, with high rates mainly alonga distance of about 0.2 m.

The main characteristics of the process remain un-altered as the biomass feed rate is varied in the range7}18 kg/h. However, as =

bis decreased, the charring

front becomes successively steeper and the amount ofunreacted carbon discharged at the grate lower. Thedi!erences in the temperature pro"les are small, but inthe "rst part of the reacting bed the gas temperature issuccessively higher (as=

bis decreased). The particle and

conversion characteristics (axial, steady pro"les of par-ticle volume, solid velocity and particle density number)and the gas composition are shown in Figs. 4 and5 (=

a/=

b"1.5) for di!erent biomass feed rates. It can

be seen that complete particle burn-out is attained for=

b)7 kg/h. Given the better conversion, the quality of

the gas slightly improves as=b

is decreased, with a tarcontent that goes to zero for =

b)11 kg/h.

Signi"cant changes are introduced in process dynam-ics, structure of the reaction front and gas quality if theair-to-fuel ratio is varied. For the reference values of thethermochemical properties, kinetic constants and trans-port coe$cients, a top-stabilized process is possible onlyfor=

a/=

b'1.2 (for =

b)19 kg/h). For lower values,

both homogeneous and heterogeneous combustion ratesdo not attain su$ciently high values for the stabilizationof the reaction front. The two processes appear to beclosely connected and give rise to a front propagatingwith a constant speed towards the bottom of the gasi"er,

where eventually extinction takes place. An example ofthis situation is shown in Fig. 6, through the axialtemperature pro"les, simulated for several times,(=

a/=

b"1 and =

b"11 kg/h). The lower the air-to-

fuel ratio, the faster the propagation rate of the reactionfront and the shorter the extinction time. As an increasein the biomass feed rate is associated with an increase inthe rate of solid (char and ash) discharge at the bottom ofthe gasi"er, high values of=

balso result in an extinction

process with dynamics qualitatively similar to those pre-sented in Fig. 6 (for instance, for =

b"20 kg/h and

=a/=

b"1.4).

For a top-stabilized reactor, as the air-to-fuel ratio isincreased, no signi"cant changes are observed at theleading edge of the reaction zone. This is understandableas the characteristics of this zone are determined essen-tially by the rate of moisture evaporation and primarypyrolysis kinetics, which are not varied. However, thesize of the reaction zone becomes successively narrowerand the homogeneous combustion more favoured, givingrise to signi"cant temperature increases. It should be

C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944 2939

Fig. 7. Conversion characteristics (particle volume, solid velocity andparticle density number pro"les) for several air-to-fuel ratios (referencedata,=

b"18 kg/h).

Fig. 8. Composition of the (dry) gas and yield of unreacted char (as% of the amount formed from pyrolysis) as functions of the air-to-biomass ratio (reference data,=

b"18 kg/h: solid lines for devolatiliz-

ation data I and dashed lines for devolatilization data III).

Fig. 9. Structure of the reaction front: axial pro"les of reaction rates andtemperature for the reference data,=

b"18 kg/h,=

a/=

b"2.4 (dashed

lines: devolatilization data I, solid lines: devolatilization data III).

Fig. 10. Structure of the reaction front: axial pro"les of combustionrates and temperature for the reference data, =

b"18 kg/h,

=a/=

b"2 (dashed lines: E

p1"105 kJ/mol, solid lines: E

p1"

84 kJ/mol).

noted that the maximum temperature remains almostconstant as long as there is carbon available for en-dothermic gasi"cation. As shown by Fig. 7, completeburn-out of the solid particles is achieved for values of=

a/=

bslightly above 1.75, corresponding to a max-

imum in the rate of char consumption and a slow com-bustion rate of volatile species. The tar content of the gasdecreases as the air-to-fuel ratio is increased, from about0.4% (=

a/=

b"1.3) to 0.1% (=

a/=

b"1.75) of the

total mass of the gaseous e%uents and goes to zero for=

a/=

b*2.

The gas composition and the char yield discharged atthe grate are reported in Fig. 8 as functions of the air-to-fuel ratio for the reference data and the pyrolysis charac-teristics I (solid lines) and III (dashed lines) (=

b"

18 kg/h). As expected, the contents of hydrogen andcarbon monoxide initially increase and then decrease, asa consequence of the improved gasi"cation e$ciency andthe enhanced combustion rate of the combustible gas,

respectively. It can be seen that, though from the quali-tative point of view the same trend in the producer gasquality is predicted for both the devolatilization dataI and III, the quantitative di!erences are signi"cant. Ofcourse the quality of the producer gas is better whenhigher amounts of carbon monoxide are assumed to beformed from both primary and secondary reactions. Thehigher concentration of combustible gas also resultin larger combustion rates and in higher temperatures(Fig. 9).

The e!ects of the apparent pyrolysis rate on the predic-tions of the gasi"cation process are particularly impor-tant, because the rate of charring largely determines thestructure of the leading edge of the reaction zone. A com-parison between the structures of the reaction zone fortwo di!erent activation energies can be made throughFig. 10 (kH

m"0.06 m/s,=

b"11 kg/h,=

a/=

b"2). The

increase in the pyrolysis rate is associated with an in-crease in the rate of heterogeneous combustion and the

2940 C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944

Fig. 11. Axial temperature pro"les for several times (reference data,devolatilization data III, E

p1"63 kJ/mol, kH

m"0.08 m/s, no methane

combustion, heterogeneous reaction rates increased by a factor of 2,=

b"9.5 kg/h, =

a/=

b"1.1).

Fig. 12. Conversion characteristics (particle volume, solid velocity andparticle density number pro"les) for several times (reference data, de-volatilization data III, E

p1"63 kJ/mol, kH

m"0.08 m/s, no methane

combustion, heterogeneous reaction rates increased by a factor of 2,=

b"9.5 kg/h, =

a/=

b"1.1).

presence of a successively steeper charring front. Themaximum temperature at the leading edge of the reactionzone slightly decreases, as a consequence of variation inthe relative position of the exothermic homogeneouscombustion zone and the exothermic/endothermicchar consumption zone. Indeed, two distinct maximaappear in the solid-phase temperature pro"le, corre-sponding to heterogeneous and homogeneous combus-tion. Between the two, there is a minimum caused bythe convective cooling of the air fed at the top ofthe reactor and the endothermicity of the gasi"cationreactions. The large di!erences between the heterogen-eous and the homogeneous combustion rates, establishedin the presence of low activation energies of the primarypyrolysis, appear to be the key feature for a grate-stabil-ized gasi"er.

Besides extinction (low air-to-fuel ratio) and a top-stabilized reactor (high air-to-fuel ratio), another interest-ing behaviour is simulated for intermediate values of theair-to-fuel ratio, as shown through Figs. 11 and 12. Thesesimulations have been obtained for E

p"63 kJ/mol,

kHm"0.08 m/s, rates of heterogeneous reaction increased

by a factor of 2, no methane combustion and devolatiliz-ation data III. For values of=

a/=

bin the range 1}1.2,

associated with the prompt formation of char, heterogen-eous combustion also occurs at some extent near the topof the reactor. However, because of the reduced temper-atures, the subsequent gas-phase combustion and chargasi"cation zones move downward. After unsteadypropagation, this front stabilizes at a certain distance(0.22}0.20 m) from the grate, where both homogeneousand heterogeneous reactions take place to some extent.This mode of stabilization, observed also in the experi-ments by Reed and Markson (1985) and indicated asgrate-stabilization, is not very e!ective. Indeed, becauseof the reduced size of the homogeneous combus-tion/gasi"cation zone, the tar content of the gas is high

and char conversion is low. It is important to point outthat a grate-stabilized process is simulated only if a sig-ni"cant axial temperature gradient is established, so thata temperature su$ciently high for the stabilization of thereaction front is attained at a certain distance from thefeed section. In the absence of such a gradient and forlow air-to-fuel ratios extinction takes place (for instance,Fig. 6).

The limit value of the air-to-fuel ratio for a top stabil-ized reactor is a!ected by the intrinsic kinetics of theheterogeneous reactions. Thus, it moves to lower valuesas the activation energy of the combustion reaction isdecreased and/or the activation energy of the gasi"cationreactions is increased. Activation energies of the combus-tion reaction in the range 117}159 kJ/mol do not give riseto variations in the characteristics of the gasi"cationprocess, given a su$ciently high value of the air-to-fuelratio (*1.2). However, this parameter in#uences thedynamic behaviour of the reaction front (stabilization)for low air-to-fuel ratios. Activation energies of the gasi"-cation reactions below the reference value again do notresult in signi"cant changes in the gasi"cation process.However, higher values are associated with higher tem-peratures, given the reduced activity of endothermic chargasi"cation.

The maximum value of the e!ective mass transfercoe$cient, kH

m, and the solid/gas heat transfer coe$cient

(factor f) exert a signi"cant in#uence on the structure ofthe reaction zone. Fig. 13 shows the conversion charac-teristics for three values of kH

m(=

a/=

b"2, =

b"

11 kg/h, reference data). For very high kHm

values, theheterogeneous reactions predominate: particle burnoutand oxygen consumption take place along a very thinregion, the solid temperature presents a very sharp peakat the leading edge of the reaction zone (char combus-tion) followed by a rapid decay (char gasi"cation). As

C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944 2941

Fig. 13. Conversion characteristics (particle volume, solid velocity andparticle density number pro"les) for three values of kH

m(reference data,

=b"11 kg/h, =

a/=

b"2).

Table 5Comparison between the predicted and measured composition of the producer gas

% vol Groeneveld and van Swaaij (1980) Walawender, Chern and Fan (1985) Wang and Kinoshita (1994) This study

CO 17 18}20 18}20 20.3}18.5H

214 12}16 10}14 16.8}9.8

CO2

13.6 14}16 9.6}11 15.3}9.4CH

40.9 2.5 2.5}4.8 4.5}2.4

N2

46.5 53}45 52}54 43}60=

a/=

b* 1}2 * 1.4}2.2

carbon dioxide is the only product of char combustion(the predominating reaction) and the rather thin zone ofhigh temperature does not allow the tars to be destroyed,the quality of the gas is poor (low contents of carbonmonoxide and hydrogen and high amounts of tars).As kH

mis reduced to very low values, heterogeneous com-

bustion/gasi"cation processes become controlled bychemical kinetics, so that the rates of heterogeneousreactions become comparable or slower than those ofthe gas-phase combustion. Indeed, for low values of themaximum k

m, oxygen is burned mainly by volatile prod-

ucts, the reaction zone enlarges, tars are completelycracked (or burned) and the quality of the gas is good.The partial combustion of the volatile products alsoresults in a wide region of high temperature, though thepeak at the leading edge of the reaction is slightly re-duced. Variations in the solid/gas heat transfer coe$-cient, through the factor f, result in changes in thereaction front and gasi"cation characteristics qualitat-ively similar to those discussed in relation to kH

m.

Another important point is the comparison betweenthe model predictions and the experimental measure-ments. It is well known (Britten, 1988) that the predic-tions of maximum temperature and other parameters,such as extension of the combustion and gasi"cationzone, are signi"cantly a!ected by parameter values,

which are not generally well known, in particular heatand mass transfer coe$cients. However, it is also wellknown (Amundson & Arri, 1978, Yoon et al., 1978;Hobbs et al., 1993) that overall conversion rates ande%uent gas composition are remarkably insensitive tomodel details. Indeed, in agreement with previous "nd-ings, it has been found that the gas composition is mainlydetermined by the composition of the pyrolysis gas andthe equilibrium conditions of the water gas shift reaction.As can be seen from Table 5, the use of reference data andthe set of devolatilization data I (wood chips) give rise topredictions of producer gas composition very close to thevalues reported for woody biomass by several references.It is believed that the agreement can be further improvedby taking into account the variation in the productdistribution and gas composition with the reaction tem-perature.

Contrary to the case of gas composition, the literaturedoes not report extensive measurements of the temper-ature pro"le in downdraft gasi"ers. Few exceptions arethe two-dimensional temperature measurements, madeavailable by Wang and Kinoshita (1994) for a con-ventional (throated) gasi"er and the measurements byManurung and Beenackers (1994) for a strati"ed, ricehusk downdraft gasi"er. This second case is consideredfor comparison. With respect to the reference data, theinitial particle diameter has been chosen as 0.5 cm, theinitial ash content as 16% and the devolatilization char-acteristics, indicated as data set II, have been used. Fur-thermore, the activation energy of the apparent primarypyrolysis degradation rate has been taken as 63 kJ/molbecause a reduction in the particle size should be asso-ciated with a faster degradation rate (no indication iscurrently available on the apparent degradation kineticsof rice husks at high temperatures). Also, the value ofkHm

has been taken equal to 0.02 m/s. In agreement withexperimental observation (Manurung & Beenackers,1994), optimal gasi"cation conditions, in terms of ther-mal e$ciency, are simulated for intermediate values(about 1.7) of the air-to-fuel ratio, when the char conver-sion is high and the calori"c value of the producer gas isat the maximum (molar composition consisting of about20% carbon monoxide, 15% hydrogen, 13% carbondioxide and minor fractions of methane). Measured and

2942 C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944

Fig. 14. Comparison between axial temperature pro"les as predictedby the model (solid and dashed lines) and measured by Manurung andBeenackers (1994) (symbols) for rice husk gasi"cation (=

b"10.5 kg/h,

=a/=

b"1.3).

predicted temperature pro"les (Fig. 14) show some dis-agreement mainly at the bottom of the gasi"er, probablycaused by the absence of grate heat losses in the model.Also, it should be noted that the predicted temperaturepro"le is signi"cantly dependent on the fraction of wallheat losses, f, retained from the correlation proposed forindustrial gasi"er (Hobbs et al., 1993), a parameter whichcannot be assigned a priori. However, on the whole, theagreement is acceptable.

4. Conclusions and further developments

A mathematical model has been formulated, whichincludes the description of all the main chemical andphysical processes taking place during the "xed-beddowndraft gasi"cation of lignocellulosic fuels. Probably,the most innovative aspects of the model, with respect tothe state of the art, are represented by the description ofthe #aming pyrolysis process through "nite rate kineticsof primary (apparent) pyrolysis, secondary cracking oftars and combustion of carbon monoxide, hydrogen, tarsand methane. These features, coupled with the formula-tion of transient equations for heat and mass transportacross the bed, have allowed the key characteristics ofdowndraft gasi"cation to be analysed. In particular, nu-merical simulation has been applied to investigate thee!ects of the two most important operational variables,i.e. biomass feed rate and air-to-fuel ratio, on processdynamics, composition and quality of the producer gasand overall e$ciency of the process.

The model predictions reproduce well, from the quali-tative point of view, the dynamic behaviour and thesteady-state con"gurations, on dependence on the air/fuel feed rate, of downdraft wood gasi"ers. Gas-phasecombustion and primary pyrolysis appear to play a con-trolling role for the mode of stabilization of the reaction

front, that is, for a top- or grate-stabilized reactor. Fromthe quantitative point of view, the description of thedevolatilization stage (product yields and gas composi-tion) and the correlations/values of the transport coe$-cients are crucial points. A `reasonablea selection forthese variables leads to good quantitative agreements interms of gas compositions, but marginal agreement is stillshown for the temperature pro"les. Model validation hasalways been carried out only to a limited extent in "xed-bed gasi"er modelling (Hobbs et al., 1993), because of thevery few experimental results available in the literature(often, not enough information is reported for compari-son with model predictions). This is particularly true forthe downdraft gasi"cation of wood. Therefore, modelvalidation should be further addressed, possibly in con-junction with an adequate experimental program.

There are numerous other aspects which need furtherinvestigation in the concurrent wood gasi"cation. Thecomprehensive model presented in this study has beenapplied only for a pilot-scale gasi"er fed with a "xed-property lignocellulosic biomass, by varying the operat-ing conditions (air and/or biomass feed rate). Also, thein#uences of few key parameters (primary gas composi-tion, adjustable factors in the heat and mass transfercoe$cients and pyrolysis kinetic constants) have beenexamined only for a narrow range of values, in order toassess their role in the process dynamics. A successivestudy should provide extensive sensitivity analyses forthe model parameters, the physico-chemical properties offeedstocks and the plant size. Further development canalso be important in relation to the accuracy of thedi!erent process submodels. For instance, though thesubmodel for the primary pyrolysis process, proposed inthis study, is more accurate than those currently appliedin biomass gasi"er modelling, it is believed that single-particle e!ects in the description of the devolatilizationstage during gasi"cation deserve a more accurate treat-ment. Indeed, while the assumption of a thermally thinparticle was found to work well for coal gasi"cation,where the amount of volatiles released is relatively low,for the present case it may result in a too rapid biomassconversion and wrong predictions of the position of thereaction front in strati"ed downdraft reactors or in incor-rect evaluation of the size of the pyrolysis zone in updraftreactors.

Single-particle e!ects in the description of primarypyrolysis and moisture evaporation are also importantfor a correct prediction of the tar evolution and thus forthe quality of the producer gas. Indeed, the simulationshave shown that the maximum temperature at the lead-ing edge of the reaction zone is highly a!ected by the rateof moisture evaporation. Too low temperatures maycause that incompletely devolatilizated char enters thereduction zone, tars are produced there and remain un-converted in the gas. Also, low temperatures may causeincomplete conversion of tars in the oxidation zone.

C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944 2943

It should be noted that a correct prediction of thetar content of the gaseous e%uents is important for ane!ective design of the gasi"er, the selection of the optimaloperation condition and the choice of the most adequategas cleaning procedure.

Another important factor for the conversion process isthe char reactivity (gasi"cation and combustion), whichin the model is described by the unreacted core model, onthe basis of the e!ective mass transfer coe$cient and theintrinsic kinetics. Critical points in this approach, whichneeds further investigation, are represented by the in"-nitely thin reaction zone and by the absence of the charconcentration in the rates of the gasi"cation reactions.

Finally, more reliable input data are needed for thesimulation of the gasi"cation process, in relation to bothtransport coe$cients and intrinsic reaction kinetics. In-deed, these have been, for large part, investigated underthermogravimetric conditions, which do not reproducethe true gasi"cation conditions.

Acknowledgements

The research was funded in part by the EuropeanCommission in the framework of the Non Nuclear En-ergy Programme (JOULE III), Contract JOR3-CT95-0021.

References

Amundson, N. R., & Arri, L. E. (1978). Char gasi"cation in a counter-current reactor. A.I.Ch.E. Journal, 24, 87.

Bhattacharya, A., Salam, L., Dudukovic, M. P., & Joseph, B. (1986).Experimental and modelling studies in "xed-bed char gasi"cation.Industrial & Engineering Chemical Process Design and Development,25, 988.

Biba, V., Macak, J., Klose, E., & Malecha, J. (1978). Mathematicalmodel for the gasi"cation of coal under pressure. Industrial & Engin-eering Chemical Process Design and Development, 17, 92.

Bliek, A. (1984). Ph.D. thesis, Mathematical modelling of a cocurrent xxedbed coal gasixer. The Twente University of Technology, The Neth-erlands.

Boroson, M. L., Howard, J. B., Longwell, J. P., & Peters, A. W. (1989).Products yields and kinetics from the vapor phase cracking of woodpyrolysis tars. A.I.Ch.E. Journal, 35, 120.

Britten, J. A. (1988). Extinction phenomena in countercurrent packed-bed coal gasi"ers: A simple model for gas production and charconversion rates. Industrial & Engineering Chemistry Research, 27,197.

Bryden, K. M., & Ragland, K. (1996). Numerical modeling of a deep,"xed-bed combustor. Energy & Fuels, 10, 269.

Buekens, A. G., Bridgwater, A. V., Ferrero, G. L., & Maniatis, K. (1990).Commercial and marketing aspects of gasi"ers. Commission of theEuropean Communities, EUR 12736.

Cho, Y. S., & Joseph, B. (1981). Heterogeneous model for moving-bedcoal gasi"cation reactors. Industrial & Engineering Chemical ProcessDesign and Development, 20, 314.

Corella, J., Aznar, M. P., Delgado, J., & Aldea, E. (1991). Steamgasi"cation of cellulosic wastes in a #uidized bed with downstreamvessels. Industrial & Engineering Chemistry Research, 30, 2252.

Di Blasi, C. (1993). Modeling and simulation of combustion processesof charring and non-charring solid fuels. Progress in Energy andCombustion Science, 19, 71.

Di Blasi, C. (1996a). Kinetic and heat transfer control in the slow and#ash pyrolysis of solids. Industrial & Engineering Chemistry Re-search, 35, 37.

Di Blasi, C. (1996b). Heat, momentum and mass transfer througha shrinking biomass particle exposed to thermal radiation. ChemicalEngineering Science, 51, 1121.

Di Blasi, C. (1997). In#uences of physical properties on biomass de-volatilization characteristics. Fuel, 76, 957.

Di Blasi, C. (1998). Multi-phase moisture transfer in the high-temper-ature drying of wood particles. Chemical Engineering Science, 53,353.

Di Blasi, C., Signorelli, G., Di Russo, C., & Rea, G. (1999). Productdistribution from pyrolysis of wood and agricultural residues. Indus-trial & Engineering Chemistry Research, 38, 2216.

Goldman, J., Xieu, D., Oko, A., Milne, R., & Essenhigh, R. H. (1984).A comparison of predictions and experiment in the gasi"cation ofanthracite in air and oxygen-enriched/steam mixtures. 20th Interna-tional Symposium on Combustion. The Combustion Institute, Pit-tsburgh, p. 1365.

Groeneveld, M. J., & van Swaaij, W. P. M. (1980). Gasi"cation ofchar particles with CO

2and H

2O. Chemical Engineering Science,

35, 307.

Gupta, A. S., & Thodos, G. (1963). Direct analogy between mass andheat transfer to beds of spheres. A.I.Ch.E. Journal, 9, 751.

Hobbs, M. L., Radulovic, P. T., & Smoot, L. D. (1993). Combustion andgasi"cation of coal in "xed-beds. Progress in Energy and CombustionScience, 19, 505.

Kashiwagi, T., & Nambu, H. (1992). Global kinetic constants forthermal oxidative degradation of a cellulosic paper. Combustion andFlame, 88, 345.

Liden, A. G., Berruti, F., & Scott, D. S. (1988). A kinetic model for theproduction of liquids from the #ash pyrolysis of biomass. ChemicalEngineering Communications, 65, 207.

Manurung, R. K., & Beenackers, A. A. C. M. (1994). Modeling andsimulation of an open-core downdraft moving bed rice husk gasi"er.Advances in Thermochemical Biomass Conversion. London: BlackieA & P, p. 288.

Purnomo, Aerts, D. J., & Ragland, K. W. (1990). Pressurized downdraftcombustion of wood chips, Twenty-third International Symposium onCombustion. The Combustion Institute (pp. 1025}1032).

Reed, T. B., & Markson, M. (1985). Biomass gasi"cation reactionvelocities. In R. P. Overend, T. A. Milne, L. K. Mudge, Funda-mentals of Thermochemical Biomass Conversion (p. 951). Amsterdam:Elsevier.

Radulovic, P. T., Ghani, M. U., & Smoot, L. D. (1995). An improvedmodel for "xed-bed coal combustion and gasi"cation. Fuel, 74, 582.

Roberts, A. F. (1970). A review of the kinetic data of the pyrolysis ofwood and related substances. Combustion and Flame, 14, 261.

Yoon, H., Wei, J., & Denn, M. M. (1978). A model for moving bed coalgasi"cation reactors. A.I.Ch.E. Journal, 24, 885.

Walawender, W. P., Chern, S. M., & Fan, L. T. (1985). Wood chipgasi"cation in a commercial downdraft gasi"er. In R. P. Overend, T.A. Milne & L. K. Mudge, Fundamentals of Thermochemical BiomassConversion (p. 911). Amsterdam: Elsevier.

Wang, Y., & Kinoshita, C. M. (1994). Temperature "elds in downdraftbiomass gasi"cation, Proceedings of the International Conference onAdvances in Thermochemical Biomass Conversion (p. 280). London:Blackie A & P.

2944 C. Di Blasi / Chemical Engineering Science 55 (2000) 2931}2944