Prediction of gas composition in biomass gasifiersgrupo.us.es/bioenergia/templates/jubilee/pdf/presentacion8-1.pdfPrediction of gas composition in biomass gasifiers A. Gómez-Barea,

Prediction of gas composition in biomass gasifiers

A. Gómez-Barea, M. Campoy, P. OlleroETSI of Seville

ETSI, University of Seville (Spain)

B. Leckner, H. ThunmanChalmers University of Technology (Sweden)

2nd International Congress of Energy

and Environment Engineering and Management

Badajoz, 6-8 June 2007

Content

1. Motivation and objective

2. Background: Existing evidence in gasification

3. Modelling

4. Experiments and application

5. Conclusions

Content



3. Modelling


5. Conclusions

Motivation for a new method

• In the preliminary design of a FBG, the knowledge of the main components of the gas produced in the gasifier is a key factor

• Advanced models for FBG exist but require physical and kinetic inputs difficult to estimate and sometimes are not available to industrial applications

• Simple and reliable semi-empirical methods to predict gas composition and reactor performance are not common in literature, and there is a need for such modelling tools


Difficulty in modelling(complex Chemestry and transport phenomena at particle level)


CharSecondary

Char (carbonizado)

Gases condensables(Tar: CCiiHHjjOOzz)

Primary light gasDEVOLATILISACION COCO22, H, H22OOCO, HCO, H22,CH,CH4 4

(Gas)(Gas)

COCO22HH22OO HH22

GASIFICATION

COCO HH22 CHCH44

OO22

COMBUSTION

COCO22

Light gas(secondary)

TarSecondary(Tar + soot)

R-C-OPOO22

COCO22

QQ

HH22OO

Char (S)

CCiiHHjjOOzz (L)(L)

Oxidant

Biomass

air

Producer GASCO, H2, CH4, CO2, N2, H2O,

CxHyOz

Fuel

Bed

Freeboard

air

Producer GASCO, H2, CH4, CO2, N2, H2O,

CxHyOz

Fuel

Bed

Freeboard

Difficulty in modelling (complex flow pattern and transport phenomena at reactor level)

emulsion

bubble

emulsion

bubble

N2

CO2

H2O

H2CO

N2

CO2

H2O

H2CO

Char particle

Boundary layer

N2

CO2

H2O

H2CO

N2

CO2

H2O

H2CO

Char particle

Boundary layer

N2

CO2

H2O

H2CO

N2

CO2

H2O

H2CO

Char particle

Boundary layer

N2

CO2

H2O

H2CO

N2

CO2

H2O

H2CO

Char particle

Boundary layer

H2O + CO ⇔ H2 + CO2

N2

C C C C

H2O + CO ⇔ H2 + CO2

N2

C C C C

H2O + CO ⇔ H2 + CO2

N2

C C C C

H2O + CO ⇔ H2 + CO2

N2

C C C C

H2O + CO ⇔ H2 + CO2

N2

C C C C

H2O + CO ⇔ H2 + CO2

N2

C C C C

Volatiles(H2, CO, CxHy)

Char

Volatiles(H2, CO, CxHy)

Char


Past trials for simple modelling of FBG

• Equilibrium models (EM)• Quasi-Equilibrium models (QEM)• Empirical models


Equilibrium models (EM)

Advantages– Simple to apply – Independent of gasifier design– Widely used

Failures– Overestimates yields of H2 and CO – Underestimates the yield of CO2

– Prediction of gas nearly free of CH4 and tar – No char in the gas phase over 1000 K


Advantages– Improvement of EM – Simple to apply

Failures– Need correlations– Dependent of gasifier design– Most cases do not predict tar and/or char– Sometimes recommendations can avoid

correlation but this make QEM non-predictive

Quasi-Equilibrium models (QEM)(Gumz, 1950)


Empirical models

Advantages– Simple to apply – The best predictions

Failures– Needs a lot of experimental data – Only valid for a given facility and biomass

Example– Maniatis et al (1994) Correlations based on

one parameter (ER)


To develop a model (method):

– Based on QEM (simple)– With predictive capability– Free from correlations– Able to estimate tar and char– Based on established evidences

Objective


Content


2. Background: existing evidence in gasification

3. Modelling


5. Conclusions

Existing evidence for corrections

• Heterogeneous or homogeneous equilibrium?– In EM no solid carbon in the gas phase over 1000 K

• Steam Reforming of Methane (SRM) in equilibrium?– Steam reforming of methane is kinetically limited

below 1300 K– methane in the exit stream of the gasifier ~ that

formed in devolatilisation

2. Evidences for correction

Existing evidence for corrections

• Water Gas Shift Reaction (WGSR) in equilibrium?– Equilibrium for the WGSR reached at 1273 K and

residence time about 1 s– Between 1073 K and 1273 K the attainment of

equilibrium has to be confirmed– This confirmation depends on the use of catalysts

and steam presence:• Synthetic (Ni) vs. minerals (dolomite, olivine, etc)

catalyst• Steam vs. air gasification


Conclusions from the existing evidence for the model

– Homogeneous equilibrium is enough for practical applications

– Modified equilibrium based on WGSR and SRM is convenient

– Kinetic rates of SMR should be included in the model

– CH4 in the exit is nearly that formed during devolatilisation (air gasification without catalyst)

– Equilibrium of WGSR is nearly attained: an approach to equilibrium method based on T, tres, type of catalyst and the presence of steam seems to be convenient


Content



3. Modelling


5. Conclusions

Aim: Overall model

ash (Solid) (Gas)

(Gas condensable )

→α β 2 2 2 2

1

2 2 3 4 2 5 2 6 4 7 2

8 j k l

CH O + aO + b H O + cCO + d N

x C(s) ++x H + x CO + x H O + x CO + x CH + x Nx C H O

α βCH O

2

2

2

4

2

j k l

HCOH OCOCHNC H OAshC(s) +

2

2

2

2

OH OCON

OxidantOxidant

BiomassBiomass SolidSolid

CondensableCondensable

Light GasLight GasHeatHeat

GASIFIERGASIFIER

TTbb

3. Modelling

Simplified model of a FB gasifier (Nseg>>1)

FLUIDISATIONAGENT

(a)

FUEL

FUELPARTICLE

(b)

Flaming-pyrolysis zone

(FPZ) FUELDEVOLATILISATION

(FDZ)

CHARGASIFICATION

(CRZ)

C(s)+R(g) P(g)

OXIDATIONOF CHAR AND

VOLATILESTAR CONVERSION

(OZ)

FUEL

Initial conditionsfor Char

Reduction Zone(CRZ) RFPZ=(CO2+H2O)FPZ

FLUIDISATIONAGENTFLUIDISATION

AGENT(a)

FUEL

FUELPARTICLE

(b)


(FPZ)


(FPZ) FUELDEVOLATILISATION

(FDZ)

CHARGASIFICATION

(CRZ)

C(s)+R(g) P(g)

OXIDATIONOF CHAR AND

VOLATILESTAR CONVERSION

(OZ)

FUEL


Reduction Zone(CRZ)


Reduction Zone(CRZ) RFPZ=(CO2+H2O)FPZ

FLUIDISATIONAGENT

3. Modelling

NNsegseg = = segregationsegregation time / devolatilisation timetime / devolatilisation time

Steps in modelling

1. Estimation yields of light gases, char and tar from FPZ (CH4, tar and char are estimated as function of T)

2. Estimation tar, methane and char conversion in CRZ by application simple kinetic models

3. QE model:- Unconverted CH4, tar and char are removed from this analysis formulation of C-H2-O2-N2 mass balances - Mass balances with corrected C-H-O inputs - two equilibrium (or approach to equilibrium) relationships (WGSR and SRMR)

4. Restoration of unconverted CH4, tar and char Application of heat balance over the corrected exit streams

3. Modelling

Model concept adopted

3. Modelling

OXYGEN STEAMFUEL

CHAR GASIFICATION

(with H2O and CO2)

TAR CONVERSION

(Reforming+Cracking)

Produced gasDischarged ashes

FLAMING PYROLYSIS ZONE (FPZ)Devolatilisation + char and volatile oxidation

NITROGEN

METHANE CONVERSION


4 ,FDZCHx ,FDZcharx ,FDZtarx

,gptarx,gpcharx4 ,gpCHx

CH

AR R

EDU

CTI

ON

ZO

NE

(CR

Z)N

ON

-EQ

UIL

IBR

IUM

FA

CTO

RS

(RA

TE M

OD

LES

)

EQ

UIL

IBR

IUM

CA

LCU

LATI

ON

S

4( )CHX ( )charX ( )tarX

OXYGEN STEAMFUEL

CHAR GASIFICATION

(with H2O and CO2)

TAR CONVERSION


Produced gasDischarged ashes

FLAMING PYROLYSIS ZONE (FPZ)Devolatilisation + char and volatile oxidation

NITROGEN

METHANE CONVERSION


4 ,FDZCHx ,FDZcharx ,FDZtarx

,gptarx,gpcharx4 ,gpCHx

CH

AR R

EDU

CTI

ON

ZO

NE

(CR

Z)N

ON

-EQ

UIL

IBR

IUM

FA

CTO

RS

(RA

TE M

OD

LES

)

EQ

UIL

IBR

IUM

CA

LCU

LATI

ON

S

4( )CHX ( )charX ( )tarX

INPUTSINPUTS

STEPSTEP 11

STEPSTEP 22

STEPSTEP 33STEPSTEP 44OUTPUTSOUTPUTS

Char conversion sub-model

3. Modelling

• Based on a recent simple method for non-catalytic gas-solid reactions for one reaction

• Reaction: Char + R P, being R = H2O + CO2 and P = H2 + CO

• Population balance and any kinetic models with any structural behavior and nth order kinetics respect to R is solved in one-envelope calculation

Tar and CH4 conversion sub-model

3. Modelling

• Calculated by single-flow kinetic models (CSTR, PFR)

– Initial conditions established by solution of FPZ

• Adequate selection of tar and methane model could be challenge:

– Tar: mainly depends on the biomass nature, operating conditions (T, t), presence of catalyst

– Methane: The use of catalyst and the presence of steam

Content



3. Modelling


5. Conclusions

10 kWth Lab and pilot scale experiments


Devolatilisation studies

Crucible

Furnace heater

Reactive gas

Horizontal arm

Alumina

TGATGA

Lab FBLab FB


AnalysisCO, H2,CO2, CH4, (O2, C2+, N2,Tar)

CO2-H2O-O2

Tbedbed

B

N2 (He)

¿ CHAR ?

¿ CHAR ?

FUR

NA

CE

CO, H2,CO2, CH4, (O2, C2+, N2,Tar)

CO2-H2O-O2

T

BIOMASS

N2 (He)

¿ CHAR ?

¿ CHAR ?

Char reactivity studies


AnalysisComb: CO2 COGas: (CO, H2)

Comb: O2 Gas: (CO2-H2O)

Tbed

BIOMASS(or CHAR)

FUR

NA

CE

CombustiónC + ½ O2 CO2

C + O2 CO2

GasificaciónC + CO2 ? 2CO C + H2O ? CO + H2

Comb: CO2 COGas: (CO, H2)

Comb: O2 Gas: (CO2-H2O)

T

BIOMAS(

CombustiónC + ½ O2 CO2

C + O2 CO2

CombustionC + ½ O2 CO2

C + O2 CO2

GasificaciónC + CO2 ? 2CO C + H2O ? CO + H2

GasificationC + CO2 2CO C + H2O CO + H2

150 kWth pilot scale experiments


Scaling-up

3 MWth BFB Gasifier4. Experiments and application

Applications

• Optimisation for gasification with wood and orujillo at 150 kWth pilot scale

• Test programme for the 3 MWth BFB gasifier• Preliminary design of BFB gasifier for

processing MBM• The tool developed improves significantly the

capability of equilibrium


Validation: Gasification of wood pellets at 150 kWthpilot gasifier (ER=0.28)


92.510091.5%CC (tar included)

0.28−0.32kgC/kgdaxC,ash

2.12.42.1mole gas/kg fuel dafFgp,d (Gas yield)

12.50.012g/Nm3CkHlOm (tar)

524051% vvN2

5.10.04.5% vvCH4

115.512.5% vvH2O

11.925.512.5% vvH2

18.57.117.2% vvCO2

14.324.515.0% vvCO

Pilot resultsEquilibriumThis methodUnits

Summary and Conclusions

1. The development of a model based on QEA with predictive capability and easy to apply

2. Used as tool for design and optimisation: improves significantly equilibrium predictions

3. Valid for preliminary design4. Yields of char, methane and tar during devolatilisation

steps need to be estimated5. Proper selection of kinetic parameters for tar and CH4 may

be critical

5. Conclusions

Conclusions

Thank you for your kind attention

5. Conclusions

Documents

Prediction of gas composition in biomass gasifiersgrupo.us.es/bioenergia/templates/jubilee/pdf/presentacion8-1.pdfPrediction of gas composition in biomass gasifiers A. Gómez-Barea,