A downdraft high temperature steamonly solar gasifier of biomass char: A modelling study

Model development for biomass gasification in an entrained flow gasifierusing intrinsic reaction rate submodel Xiaoyan Gaoa, Yaning Zhanga,b,⇑, Bingxi Lia,⇑, Xiangyu Yua a School of Ener

Model development for biomass gasification in an entrained flow gasifier

using intrinsic reaction rate submodel

Xiaoyan Gaoa, Yaning Zhanga,b,⇑, Bingxi Lia,⇑, Xiangyu Yua

a

School of Energy Science and Engineering, Harbin Institute of Technology, Harbin, China

b

School of Chemical Engineering and Technology, Harbin Institue of Technology, Harbin, China

a r t i c l e i n f o

Article history:

Received 30 June 2015

Accepted 28 October 2015

Keywords:

Model development

Intrinsic reaction rate model

Biomass gasification

Entrained flow gasifier

a b s t r a c t Intrinsic reaction rate submodel is established in this study to consider the effects of diffusion rate and kinetic rate for simulating the char reactions due to their slow reaction rates and important controlling steps The biomass gasification model for an entrained flow gasifier is developed with the Euler–Lagrange method using ANSYS FLUENT software Gas phase is treated as continuous phase in standard k–emodel to close governing equations whereas biomass particles are treated as discrete phase in discrete phase model (DPM) to track the movement of particles For homogeneous phase reactions, finite rate/eddy dis-sipation model is applied to calculate the reaction rates For heterogeneous phase reactions, intrinsic reaction rate model is realized by coding the user-defined functions (UDFs) to calculate char reaction rates The results obtained from this study show that the relative errors of volumetric concentrations are mainly within the range of 1–18% and the relative errors of lower heating value, gas production, cold gas efficiency and carbon conversion efficiency are within the ranges of 1–13%, 1–8%, 1–12% and 1–11%, respectively The CFD model developed in this study can be used to simulate biomass gasification pro-cesses for entrained flow gasifiers

Ó 2015 Elsevier Ltd All rights reserved

1 Introduction

Energy shortages and pollution emissions are still the

problem-atic issues all over the world, and the threats of resource

exhaus-tion and environment polluexhaus-tion stress the need for exploring new

energy resources Biomass resource, an abundant resource on the

earth, is therefore becoming an important alternative for the world

[1–3] According to the Annual Energy Outlook 2014 (AEO2014)

released by the U.S Energy Information Administration (EIA),

bio-mass power generation would grow with the increased use of

co-firing technology in the near term and it would grow with the

increased capacities of power plants in the long run As a result,

the electricity generation from biomass would increase

signifi-cantly with an estimated average annual growth rate of 4.4% from

2012 to 2040[4]

Entrained flow gasification is an important thermochemical

conversion method to convert solid fuels into high value

gaseous fuels and chemical products [1,5,6], even on a small

scale because it is capable of gasifying any fuels to produce a

clean and almost tar-free syngas in a short residence time [7,8] Entrained flow gasification is therefore widely studied and used [5,9–11] Fig 1 illustrates the working principle of a typical downdraft entrained flow gasifier Generally, biomass fuels and gases are introduced from the top of the reactor, the biomass particles then mix with gases thoroughly through the reactor From the outlet at the bottom of the gasifier, the produced gas and ash exit

Computational fluid dynamics (CFD) is a useful tool for design-ing gasifiers, predictdesign-ing performances and optimizdesign-ing structures [1,12,13] A CFD model can offer both spatial and temporal field solutions of temperature, velocity, pressure, etc., and it is therefore able to provide the predictions of operating performances as well

as gasification mechanisms inside a gasifier [14,15] Several researchers developed mathematical models for simulating bio-mass gasification in entrained flow gasifiers Kobayashi et al.[16] built a simple thermodynamic equilibrium model for biomass gasi-fication in an entrained gasifier Coda et al.[17]took the slagging/ melting tendencies into account and then built a thermodynamic equilibrium model Valero and Usón[18]divided gasification pro-cess into two isothermal zones and developed a model for a pres-surized entrained flow gasifier Fletcher et al.[19]developed a CFD model to study the flow and reactions inside an entrained flow

http://dx.doi.org/10.1016/j.enconman.2015.10.070

0196-8904/Ó 2015 Elsevier Ltd All rights reserved.

⇑Corresponding authors at: School of Energy Science and Engineering, Harbin

Institute of Technology, Harbin 150001, China Tel./fax: +86 451 86412078.

E-mail addresses: ynzhang@hit.edu.cn (Y Zhang), libx@hit.edu.cn (B Li).

Contents lists available atScienceDirect Energy Conversion and Management

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / e n c o n m a n

gasifier using CFX package Ku et al.[20]considered the

interac-tions between gas phase and particle devolatilizationphase in an

Euler–Lagrangian CFD model by using Open FOAM software It is

known that char reaction rate is much slower than devolatilization

(pyrolysis) rate, and it is therefore the limiting step in gasification

[21–24] However, intrinsic char reactivity of biomass fuel has not

been considered in model development for biomass gasification in

entrained flow gasifiers

The objective of this study is twofold: (a) to develop a

compre-hensive CFD model for simulating biomass gasification in an

entrained flow gasifier by considering the intrinsic char reactivity

of biomass fuels, (b) to determine the relative errors between

sim-ulated and experimental results based on the entrained flow

gasi-fier built in our institute

2 Model development 2.1 Main assumptions The following assumptions are made for the model:

1 The gravitational force of gas phase is neglected[19,25]

2 The gasifier is operated under steady state conditions

3 The gas phase is regarded as uncompressible ideal gas, and air is composed of 21% oxygen and 79% nitrogen

4 All biomass granules are spherical in shape and uniform size, and slags during gasification are neglected[26]

5 The gas phase species include CO, CO2, CH4, C2H4, H2, H2O, O2,

N and tar (described as CHO)[12,27]

Nomenclature

A, B Magnussen constants for reactants and products;

pre-exponential factor

a absorption coefficient

ap absorption coefficient of particle

C linear-anisotropic phase function coefficient

C1e, C2e k–emodel constants

Cj,r molar concentration of species j in r reaction

(kmol m
3)

Cs, C1 vapor concentration at particle surface and in the bulk

gas (kmol m
3)

Csw swelling coefficient

cp heat capacity (J kg
1K
1)

De effective diffusion coefficient (m2s
1)

Di molecular diffusion coefficient of gas component i

(m2s
1)

Di,m mass diffusion coefficient of chemical species i in the

gas mixture (m2s
1)

Dk,i Knudsen diffusion coefficient of gas species i (m2s
1)

D0 molecular diffusion coefficient at reference temperature

(m2s
1)

DT,i thermal diffusion coefficient (m2s
1)

dp particle diameter (m)

dp,0 initial particle diameter (m)

E activation energy (kJ mol
1)

fv,0, fw,0 initial volatile fraction and initial moisture fraction of

fuel

G incident radiation

Gk generation term for turbulence kinetic energy

h specific enthalpy of gas phase (J kg
1) and convective

heat transfer coefficient (W m
2K
1)

hfg latent heat of moisture/volatile matters (J kg
1)

Hrec enthalpy of reaction (J kg
1)

k turbulence kinetic energy (m2s
2) and reaction kinetic

rate (s
1)

kc thermal conductivity of bulk gas (W m
1K
1)

kg mass transfer coefficient between vapor and bulk gas

(m s
1)

kgp mass transfer coefficient between gas phase and particle

phase (m s
1)

kint intrinsic reaction rate (1 s
1atm
m)

kf,r, kb,r forward rate constant and backward rate constant for r

reaction

Mc, Mw molecular weight of carbon and water vapor (kg

kmol
1)

Mi, Mj molecular weight of chemical species i and j (kg kmol
1)

mp mass of the tracked particle (kg)

mp,0 initial mass of the tracked particle (kg)

pg bulk gas pressure (Pa)

pg,j partial pressure of species j (atm)

ps,j partial pressure of species j at particle surface (atm)

R universal gas constant (J kmol
1K
1) and reaction rate

(kmol m
3s
1)

Ri,r net production rate of chemical species i in r reaction

bRi;r Arrhenius molar rate of production/consumption of

chemical species i in r reaction

Rint intrinsic char reactivity (s
1)

Rp,j particle reaction rate with gas species j (kg m
2s
1)

Rp ;j particle reaction rate with gas species j (kg s
1)

rpore average pore radius (m)

Sm, SF, Shsource term for mass, momentum and energy

SpY i, RfYi mass fraction source terms for chemical species i

Tg, Tp temperature of gas phase and tracked particle (K)

Tm mean temperature (K)

T0 reference temperature (K)

ui, uj gas phase velocity components (m s
1)

u0

i; u0

j fluctuating velocity of gas phase (m s
1)

v velocity of particle phase (m s
1)

vg stoichiometric ratio of gas moles to carbon moles

X carbon conversion degree

xi, xj global coordinates (m) and mole fraction

Yi mass fraction of chemical species i

YP, YR mass fraction of product species and reactant species

Sct turbulent Schmidt number

a, b rate exponent

d temperature exponent

e dissipation rate of turbulence kinetic energy (m2s
3)

g effectiveness factor

g0

j ;r; g00

j ;r rate exponents for reactant j and product j in r reaction

hR radiation temperature (K)

h0 initial porosity

k thermal conductivity of gas phase (W m
1K
1)

l gas phase viscosity (kg m
1s
1)

lt turbulent viscosity (kg m
1s
1)

t0

i ;r; t00

i ;r stoichiometric coefficients for reactant i and product i in

r reaction

qg,qp density of gas phase and particle phase (kg m
3)

rk,re turbulent Prandtl numbers for k ande

rs scattering coefficient

ep particle emissivity

s tortuosity of pores

6 During drying process, moisture evaporation is described as a

diffusion limited process[28]

7 The contents of sulfur and nitrogen and associated reactions are

neglected

2.2 Continuous phase model

The gas phase is modeled in Eulerian coordinates, and all

erning equations are given in Reynolds-averaged manner The

gov-erning mass equation for gas phase is:

@ðqguiÞ

@xi

where the mass source Smis the mass added to the gas phase from

the particle phase

The governing momentum equation for gas phase is:

@ðqguiujÞ

@xj ¼
@pg

@xiþ@x@

j

l@ui

@xj
qgu0iu0j

where the term
qgu0iu0j is the Reynolds stress (turbulent stress)

which is expressed according to the hypothesis of Boussinesq

[29], and the mass source SFis the external body force from the

interaction with the dispersed phase

The governing energy equation is:

@ðqguihÞ

@xi ¼ @

@xj

k@Tg

@xj

where the energy source Shis the source term due to the heat

trans-fer of convection and radiation between gas phase and particle

phase, latent heats of drying and pyrolysis, as well as the homoge-neous and heterogehomoge-neous reactions heat

In this study, standard k–eturbulent model is utilized to solve the turbulent stress The transport equations for turbulence kinetic energy k and its dissipation rateeare as follows[29]:

@

@xiðqgkuiÞ ¼ @

@xj

lþlt

rk

@k

@xj

@

@xi

ðqgeuiÞ ¼ @

@xj

lþlt

re

@e

@xj

þ C1ee

kGk
C2eqg

e2

The governing equation for chemical species i is given by:

@

@xjðqgujYiÞ ¼ @

@xj qgDi;mþlt

Sct

@Yi

@xjþDT ;i

Tg

@Tg

@xj

þ SpY iþ RfY i ð6Þ

where the source term SpY i is caused by the presence of particle phase, and the source term RfYiis due to the production/consump-tion in chemical reacproduction/consump-tions The turbulent Schmidt number Sct is set to be 0.7[5]

2.3 Particle transport model The particle phase is modeled in Lagrangian coordinates using discrete phase model (DPM) The impact of turbulence in gas phase

on the particle is predicted by the stochastic tracking model The governing equations for a tracked particle are:

dmp

dt ¼ dmp

dt

drying

þ dmp

dt

pyrolysis

þ dmp

dt

reaction

ð7Þ

mpdv

mpcp

dTp

dt ¼ hpd2p Tg
Tp

þeppd2pr h4
T4

p

þ dmp

dt

drying

hfgþ dmp

dt

pyrolysis

hfg

þ dmp

dt

reaction

where the termP

Fiis the sum of forces between particle phase and gas phase

Biomass particles in an entrained flow gasifier undergo the pro-cesses of drying, devolatilization, oxidation and gasification The detailed expressions for source terms are introduced in the follow-ing chemical reaction models

2.4 Chemical reaction models The chemical reactions inside a gasifier include the moisture release, pyrolysis, homogeneous reactions (oxidation and gasifica-tion of volatile matters) and heterogeneous reacgasifica-tions (oxidagasifica-tion and gasification of biomass char)

(1) Drying Moisture release is simulated through using wet combustion model When the particle temperature reaches the evaporization temperature, moisture is released The evaporation rate is given by:

dmp

dt

drying

If the temperature is higher than water boiling temperature, the evaporation rate is:

biomass particles

& oxidant

Particle

Gas path

Particle path

Fig 1 Schematic sketch of gas–solid flow in a downdraft entrained flow gasifier [1]

dt

drying

¼pdpkc

cp

2þ 0:46Re0:5

ln 1þcpðTg
TpÞ

hfg

ð11Þ

(2) Pyrolysis

The pyrolysis process is modeled by single kinetic rate model

where the biomass pyrolysis is represented by a one-stage global

reaction:

Biomass! Char þ Volatile

Volatiles¼ x1COþ x2CO2þ x3H2þ x4CH4þ x5C2H4þ tar ðR1Þ

In Fluent, the biomass char contains only solid carbon and ash,

and the composition of volatile can be obtained through

Thun-man’s method [30] based on the mass balance with proximate

and ultimate analyses The pyrolysis rate depends on the amount

of volatiles remaining in the biomass particle, so the

decomposi-tion rate is given by:

dmp

dt

pyrolysis

¼
k mp
ð1
fv ;0Þð1
fw ;0Þmp ;0

ð12Þ

The parameter needed is only kinetic constant obtained in

Arrhenius expression Single kinetic rate model has been widely

used in pyrolysis simulations due to its simplicity making it

com-putationally tractable[12] In this study, the pre-exponential factor

is 4.88 1012

s
1and the activation energy is 177 kJ mol
1[31]

Since the fraction of volatile mater in biomass is significant, the

effect of shrinkage/swelling during the pyrolysis process should be

taken into account[32] In this study, swelling coefficient Cswof 1.8

[32] is used to describe the change of particle diameter during

devolatilization, and the particle diameter can be expressed as:

dp

dp;0¼ 1 þ ðCsw
1Þð1
fw;0Þmp;0
mp

(3) Homogeneous phase reactions

After the pyrolysis, the combustible gases (CO, H2, CH4, etc.)

among volatile will react with oxidant fed into the reactor With

insufficient oxidant, gasification reactions will also happen among

the volatile gases The homogeneous phase reactions taken into

account in this study are as follows:

CxHyOzþ x þy

2
z

O2! xCO þy

Reaction rates of homogeneous phase reactions are calculated through finite-rate/eddy-dissipation model Both the Arrhenius and eddy-dissipation reaction rates are calculated in finite-rate/ eddy-dissipation model, and then the minimum of these two rates

is chosen as the homogeneous reaction rate[33] Arrhenius expres-sion is:

bRi;r¼ t00

i ;r
t0

i ;r

kf;rYN j¼1

Cj;rg 0 j;r
kb;r

YN j¼1

Cj;rg 00 j;r

!

ð14Þ

The kinetic rates for gas phase reactions used in this study are listed inTable 1

Eddy-dissipation rate is determined by the smaller of the expressions below:

Ri ;r¼t0

i ;rMiAqge

kminR

YR

t0 R;rMR

!

ð15Þ

Ri ;r¼t0 i;rMiABqg

e k

P

PYP

PN

jt00

j ;rMj

ð16Þ

In this study, A is equal to 4.0, and B is equal to 0.5[33,37] (4) Heterogeneous phase reactions

In order to improve the char reaction model, the intrinsic reac-tion rates of char reacreac-tions (intrinsic reacreac-tion rate model) are applied to take into account both diffusion effect and chemical reaction effect when establishing the heterogeneous phase reac-tion model[38,39] The heterogeneous phase reactions considered

in this study are as follows:

Table 1

Kinetic parameters for homogeneous phase reactions.

CO C b

O 2

[27]

H 2 C b

O 2

CH 4 C b

O 2

[12]

CO C b

H 2 O

R ib ¼ A expð
E=RT g ÞC a

CO 2 C b

H 2

2 H 4 C b

O 2

[36]

g expð
E=RT g ÞC a

x H y O z C b

O 2

[28]

R i in kmol m
3 s
1 , E in kJ mol
1 , C in kmol m
3 , A in (koml m
3 ) m K
d s
1 with m = 1
a
b.

Chsi þ 0:5O2! CO ðR8Þ

In the intrinsic reaction rate model, both effectiveness factorg

and intrinsic reaction rate Rintare used to express the char reaction

rates, and the reaction rate of heterogeneous reaction can be

expressed as:

Rp;j¼qpdp

The order of reaction can be represented with m, so the intrinsic

rate of reaction (Rint) is expressed as[12,40]:

Rint¼ kintpm

where the term F(X) is the surface function which depicts the

vari-ation of active site concentrvari-ation depending on the carbon

conver-sion degree

The effectiveness factorg, the ratio of the actual reaction rate to

the intrinsic rate, is defined as[38]:

g¼3

/

1

tanh /
1

/

ð19Þ

/¼dp

6

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ðm þ 1ÞkintFðXÞqpvgRTgpm
1

s ;j

2McDe

s

ð20Þ

The universal gas constant R in Eq.(20)is in the unit of atm m3

kmol
1K
1 The effective diffusivity Deis the diffusion coefficient

of gas reactant through the particle pores Here, both molecular

diffusion and Knudsen diffusion are taken into account, and the

definition of Deis[41]:

De¼h

s

1

Diþ 1

Dk;i

ð21Þ

The molecular diffusion coefficient is a function of temperature

at a certain pressure[42] The coefficients for molecular diffusion

and Knudsen diffusion are given as follows:

Di¼ D0 Tm

T0

 n

ð22Þ

Tm¼Tgþ Tp

Dk ;i¼ 97:0rpore

ffiffiffiffiffiffi

Tp

Mi

s

ð24Þ

The porosity h of char particles and the mean pore radiusrpore

are obtained by[38,41]:

rpore¼ 2h

Through the bulk diffusion, the partial pressure of species j at

particle surface can be obtained by[38]:

kgp

RTg

ðpg ;j
ps ;jÞ ¼Rp ;jvg

Mc

ð27Þ

The mass transfer coefficient kgpis determined by the Frössling

equation[38]:

kgpdp

In an entrained flow gasifier, the relative velocity between gas phase and particle phase is small[38], so Eq.(28)can be simplified to:

kgpdp

Di

The pressure ps,jcan be expressed as:

ps ;j¼ pg ;j
Rp ;j

Kd

ð30Þ

Kd¼ 2McDi

Finally, the reaction rate of biomass char is given by:

Rp ;j¼qpdp

6 gFðXÞkint pg ;j
Rp;j

Kd

ð32Þ

In order to solve Rp,j, Brent’s iteration method is applied in this study In Fluent, the unit for particle reaction rate is kg s
1, so mul-tiply Rp,jby the external surface area to give:

The sub-model for determining the reaction rates of heteroge-neous reactions is developed as user-defined functions to be com-piled in the Fluent

2.5 Radiation model The radiation flux during the entrained flow gasification is cal-culated through P-1 model The radiation heat flux is:

3ða þrsÞ
Crs

@G

@xi

 

ð34Þ

Input parameter:

f( Rp,j)=0

Root?

No Yes Brent method

Return

Char reaction?

Yes

No

Exit Access

For simplification, the parameterCis introduced:

Thus, for particles, the transport equation for the incident radi-ation G is:

@

@xj

C@G

@xj

ða þ apÞG þ 4p arT4

p þep

!

2.6 Calculation procedure for char reactions Fig 2introduces the UDF flow chart for char reaction rates When the solver starts to calculate the char reaction, the DEFI-NE_PR_RATE macro is called According to the previously defined constants, the intermediate parameters in char reaction submodel including surface function, average pore radius, etc are computed

in order to fill the char reaction rate function (Eq.(32)) Brent iter-ation is continued to solve the Eq.(32)

3 Model validation 3.1 Experimental apparatus and material

An entrained flow gasification system was built in the School of Energy Science and Engineering, Harbin Institute of Technology (HIT), China The gasification system consists of a downdraft entrained flow reactor, a biomass fuel feeder, a gas supplying sys-tem, a heating and temperature measuring system and a sampling system A detailed description of the experimental apparatus and experimental procedures can be found in Ref.[36] The schematic diagram of the entrained flow reactor is shown inFig 3 For simpli-fication, a 2D geometric model of this entrained flow gasifier was built and the geometric dimensions were detailed in our previous work[43] The grid independence of the geometric model was ver-ified based on the previously developed model [43] where five grids (0.04, 0.08, 0.17, 0.23 and 0.30 million cells) were examined The relative differences of gas volumetric concentrations between 0.17 million and 0.30 million were less than 2%, and the grid of 0.17 million cells is therefore adopted in this study The standard wall function is adopted for near-wall treatment, and second order upwind scheme is used as the discretization scheme The conver-gence criteria for energy and P1 are set to be 10
6and the conver-gence criteria for the other variables are set to be 10
3

In this study, air gasification of sawdust in the entrained flow gasifier is simulated at various equivalence ratios and gasification temperatures Equivalence ratio is defined as the ratio of the actual air supplied to the stoichiometric air required for complete com-bustion The variation of equivalence ratio is controlled by altering the air supplying rate while keeping the fuel feeding rate and other operating parameters fixed The variation of gasification tempera-ture is controlled by the electrical heating element installed in the entrained flow reactor while other operating parameters are fixed The main characteristics and pyrolysis coefficients for the sawdust used in this study are listed inTables 2and3, respectively, and the

Biomass & oxidant

inlet

Carrier gas inlet

Outlet

Fig 3 Schematic diagram of the entrained flow gasifier at HIT.

Table 2

Main characteristics of sawdust.

Proximate analysis a

[36]

Ultimate analysis a

[36]

a

Table 3

Pyrolysis coefficients for sawdust.

Table 4 Reaction rate constants for char reactions [40]

tar produced during the pyrolysis process is described by

FðXÞ ¼ 94:95X5
190:37X4
47:08X2þ 6:14X þ 0:29 ð37Þ

In addition, the intrinsic reaction rate constants for sawdust

char are listed inTable 4

3.2 Boundary conditions

The boundary conditions for simulating the entrained flow

gasifier are given as follows:

(a) The inlet conditions for oxidant gas (air) are: flow rates (given inTable 5), air inlet temperature Tin,air= 300 K, and turbulence specification adopts turbulent intensity Iin, air= 5% and hydraulic diameter Din,air= 8 mm

(b) The inlet conditions for carrier gas (N2) are: flow rates (given

inTable 5), N2inlet temperature Tin,nitro= 300 K, and turbu-lence specification adopts turbulent intensity Iin,nitro= 5% and hydraulic diameter Din,nitro= 8 mm

(c) The outlet conditions are: gauge pressure Po= 0 Pa, and tur-bulence specification adopts turbulent intensity Io= 5% and hydraulic diameter Do= 100 mm

(d) Wall condition: no slip shear condition together with con-stant wall temperature and the wall temperature is equal

to gasification temperature

The flow rates of sawdust particles are given inTable 5, and the initial conditions for particles are given inTable 2 For the particle phase, the maximum number of tracking step is set as 105, and the tracking length scale is specified as 0.001 m

3.3 Results and discussion The CFD model is validated with experimental data taken from the published work[36] The relative errors between the simulated and experimental data are detailed in this study The relative error

is defined as the absolute difference between the simulated and experimental values divided by the experimental value

3.3.1 Gas composition The simulated and experimental volumetric concentrations of the produced gas compositions at different equivalence ratios and gasification temperatures are shown inFig 4 InFig 4(a), when equivalence ratio varies in the range of 0.22–0.34, the simulated volumetric concentrations of CO, CO2, H2, CH4 and C2H4 are in the ranges of 20.72–30.20%, 9.64–11.89%, 6.50–9.45%, 2.24–3.52% and 0.73–1.13% whereas the corresponding experimental values are in the ranges of 21.87–30.64%, 9.10–10.90%, 6.23–8.05%, 1.76–3.33% and 0.94–1.40%, respectively It is observed that increasing the equivalence ratio from 0.22 to 0.34 increases the yield of CO2 and decreases the yields of CO and H2, however, it

Table 5

Parameters for experiments on sawdust gasification [36]

(L min
1)

N 2 flow rate (L min
1)

Sawdust feeding (g min
1)

0

10

20

30

40

50

Experiment CO CO 2 H 2 CH 4 C 2 H 4

Equivalence ratio

Simulation CO CO 2 H 2 CH 4 C 2 H 4

(a) at different equivalence ratios (gasification temperature = 800 oC)

0

10

20

30

40

50

Experiment CO CO 2 H 2 CH 4 C 2 H 4

Simulation CO CO 2 H 2 CH 4 C 2 H 4

(b) at different gasification temperatures

(equivalence ratio = 0.28)

Fig 4 Simulated and experimental volumetric concentrations of produced gas

0 10 20 30

40

CO2

Experimental volumetric concentration (%)

CH4

CO

20%

H2

20%

C2H4

shows slight reducing effects on the yields of CH4and C2H4 The

changes are caused by the facts that higher equivalence ratio can

promote the oxidation exothermic reactions (CO, H2, CH4, etc.)

and cause higher temperature inside the reactor[44], which would

support the endothermic gasification reactions[8,45]

In Fig 4(b), when gasification temperature increases from

800°C to 1000 °C, the simulated volumetric concentrations of CO,

CO2, H2, CH4 and C2H4 vary within the ranges of 20.57–24.09%,

11.58–13.28%, 8.33–13.61%, 1.31–2.81% and 0.13–0.90% whereas

the corresponding experimental values are in the ranges of

23.38–25.99%, 10.05–12.11%, 7.62–13.49%, 1.55–3.33% and 0.19–

1.31%, respectively The simulated and experimental results show

that when gasification temperature increases from 800°C to

1000°C, the yields of CO2and H2increase whereas the yields of

CO, CH4 and C2H4 decrease Endothermic char reaction and

methane oxidation reaction are favorable at higher temperature,

which improves the production of H2and weakens the production

of CH4 [20,46] However, in the range of 700–900°C, water gas

shift reaction accelerates with the increasing temperature,

result-ing in a decrease in CO yield whereas an increase in CO2yield[8]

The relative errors between the simulated and experimental gas

compositions are presented inFig 5 The relative errors are 1–13%,

5–15%, 1–18%, 4–30%, 16–35% for CO, CO2, H2, CH4, C2H4,

respec-tively Several researchers reported similar results for relative

errors between the simulated and experimental gas compositions The relative errors reported by Ku et al.[20]were 25%, 25%, 19% and 19% for CO, H2, CO2and CH4, respectively The relative errors

of H2, CO, CO2and CH4reported by He et al.[47]were in the ranges

of 10–40%, 20–35%, 15–20% and 32–65%, respectively The relative errors reported by Liu et al.[48]were 12%, 1%, 50% and 50% for CO,

H2, CH4and C2H4, respectively The relative errors reported by Zhao [36] were 4–12%, 1–20%, 0–17%, 9–40%, and 1–55% for CO, CO2,

CH4, C2H4, and H2, respectively The results obtained in this study show that the relative errors between the simulated and experi-mental gas compositions are mainly within 18% except for a few points related to CH4and C2H4 Ku et al.[20]stated that the rela-tive error of CH4 may be somewhat large (due to its small amounts), however, the small amounts can usually be neglected

In this study, although the maximum relative errors are up to 30% and 35% for CH4 and C2H4, the corresponding differences between simulated and experimental volumetric concentrations are 0.58% and 0.29%, respectively, being very small amounts and therefore can be neglected As the developed CFD model can pre-dict well for the compositions of most of the small-amount gases (CH and CH ) and the other gases, the developed CFD model

0

2

4

6

8

10

-3 )

Equivalence ratio

Experiment Simulation

(a) at different equivalence ratio

o

s (gasification temperature = 800 C)

0

2

4

6

8

10

Experiment Simulation

-3 )

Temperature ( o C)

(b) at different gasification temperatures

(equivalence ratio = 0.28) Fig 6 Simulated and experimental lower heating values of produced gas.

0 1 2

3

Experiment Simulation

3 kg

-1 biomass)

Equivalence ratio

(a) at different equivalence ratios (gasification temperature = 800 oC)

0 1 2

3

Experiment Simulation

3 kg

-1 biomass)

Temperature ( o C)

(b) at different gasification temperatures (equivalence ratio = 0.28) Fig 7 Simulated and experimental gas productions.

therefore can be used to predict the produced gas compositions of

biomass gasification in the entrained flow gasifier

3.3.2 Lower heating value

The simulated and experimental lower heating values of

pro-duced gas at different equivalence ratios and gasification

tempera-tures are given in Fig 6 In Fig 6(a), when equivalence ratio

increases from 0.22 to 0.34, the simulated lower heating values

of produced gas are in the range of 4.57–6.80 MJ N m
3whereas

the relevant experimental values are in the range of 4.65–

6.67 MJ N m
3 It is observed that the lower heating value of

pro-duced gas monotonically decreases when the equivalence ratio

increases, this is due to the decreases in the main combustible

spe-cies (CO and H2)[49]

In Fig 6(b), when gasification temperature increases from

800°C to 1000 °C, the simulated lower heating values of produced

gas vary in the range of 4.65–5.51 MJ N m
3whereas the relevant

experimental values are in the range of 5.08–6.00 MJ N m
3 Both

the simulated and experimental data show that increasing the

gasification temperature decreases the lower heating value of

pro-duced gas Although the H2production increases with the

gasifica-tion temperature, the other combustible species (CO, CH4 and

C2H4) decrease, making the lower heating value of produced gas decrease with the rise of gasification temperature[50]

The relative errors between the simulated and experimental lower heating values of produced gas are within 1–13%, being lower than the relative errors of about 20% reported by Miao

et al.[51]for the lower heating values of produced gas from a cir-culating fluidized bed reactor and the maximum relative error of 28% reported by Ngo et al.[50]for a three-stage gasification model

3.3.3 Gas production Gas production is determined as the total volume of the pro-duced gas per kilogram biomass (N m3kg
1biomass) The simu-lated and experimental gas productions at different equivalence ratios and gasification temperatures are shown inFig 7 InFig 7 (a), when equivalence ratio varies from 0.22 to 0.34, the simulated gas productions are in the range of 1.47–1.83 N m3kg
1biomass whereas the corresponding experimental values are in the range

of 1.42–1.82 N m3kg
1 biomass Both the predicted and experi-mental results show that the gas production increases monotoni-cally with the rise of equivalence ratio, this is due to the fact that higher temperature (caused by higher equivalence ratio) favors the cracking of tar and more gas could be produced[44]

0

20

40

60

80

100

Experiment Simulation

Equivalence ratio

(a) at different equivalence ratios (gasification temperature = 800 oC)

0

20

40

60

80

100

Experiment Simulation

Temperature (oC)

(b) at different gasification temperatures

(equivalence ratio = 0.28)

0 20 40 60 80 100

Experiment Simulation

Equivalence ratio

(a) at different equivalence ratio (gasification temperature = 800 oC)

0 20 40 60 80 100

Experiment Simulation

Temperature ( o C)

(b) at different gasification temperatures (equivalence ratio = 0.28) Fig 9 Simulated and experimental carbon conversion efficiencies.

InFig 7(b), when gasification temperature varies from 800°C to

1000°C, the predicted gas production varies in the range of 1.66–

1.82 N m3kg
1biomass whereas the corresponding experimental

gas production varies in the range of 1.68–1.76 N m3kg
1biomass

Lapuerta et al.[52]also reported that the gas production varied

slightly when the gasification temperature increased from 750°C

to 1000°C

The relative errors between simulated and experimental gas

productions are in the range of 1–8% These values are lower than

the maximum relative errors of 128% and 20% reported by Ngo

et al.[50]and Miao et al.[51]for the predicted gas productions,

respectively

3.3.4 Cold gas efficiency

Cold gas efficiency is defined as the ratio of the lower heating

value of the fuel gas to the lower heating value of the raw biomass

feedstock The simulated and experimental gasification efficiencies

at different equivalence ratios and gasification temperatures are

given in Fig 8 In Fig 8(a), when equivalence ratio rises from

0.22 to 0.34, the simulated cold gas efficiency varies within the

range of 55.43–66.53% whereas the relevant experimental value

varies within the range of 56.05–62.81% Since cold gas efficiency

is the product of gas production and lower heating value, it is

therefore determined by the gas production and lower heating

value collectively

InFig 8(b), when gasification temperature rises from 800°C to

1000°C, the simulated cold gas efficiency varies in the range of

52.77–60.78% whereas the relevant experimental value varies in

the range of 59.36–66.71% van der Meijden et al.[53]stated that

high temperature can decrease the cold gas efficiency, both the

simulated and experimental results in this study also show that

increasing gasification temperature generally decreases the cold

gas efficiency

The relative errors between simulated and experimental

gasifi-cation efficiencies are in the range of 1–12% These values are lower

than the maximum relative error of around 20% reported by Miao

et al.[51]for the predicted gasification efficiencies

3.3.5 Carbon conversion efficiency Carbon conversion efficiency is defined as the ratio the amount

of carbon in the final produced gas to the amount of carbon in the biomass feedstock The simulated and experimental carbon con-version efficiencies at different equivalence ratios and gasification temperatures are shown in Fig 9 InFig 9(a), when equivalence ratio increases from 0.22 to 0.34, the simulated carbon conversion efficiency varies between 88.26% and 89.54% whereas the experi-mental value varies between 85.93% and 92.81% InFig 9(b), when gasification temperature increases from 800°C to 1000 °C, the sim-ulated carbon conversion efficiency varies between 81.39% and 89.11% whereas the experimental value varies between 87.38% and 92.81%

The relative errors between simulated and experimental carbon conversion efficiencies are in the range of 1–11% These values are lower than the values of 1–25% reported by Nikoo and Mahinpey [54]for the simulated carbon conversion efficiencies

3.3.6 Gasification phenomena The temperature and species mass fraction contours of a basic case inside the entrained flow gasifier when the gasification tem-perature and equivalence ratio are respectively 800°C and 0.28 are shown inFig 10 There is a highest-temperature zone located

in the upper section of the gasifier, which is due to the exothermic oxidation reactions The temperature decreases along the reactor because of the occurrence of the endothermic reduction reactions Biomass particles injected from the top go through drying and pyrolysis rapidly, meanwhile the gas species of CO, H2, CO2, etc are released And then the oxidation reactions take place in the combustion zone, the mass fraction of CO2reaches the maximum around the highest-temperature zone As the mixture of gas and particles moves further up into the gasification zone (where

(a) Temperature (K) (b) CO mass fraction (c) H2mass fraction (d) CO2 mass fraction