Euler-Euler and Euler-Lagrange Modeling of Wood ... · Euler-Euler and Euler-Lagrange Modeling of...

Post on 14-May-2018

273 views 6 download

Transcript of Euler-Euler and Euler-Lagrange Modeling of Wood ... · Euler-Euler and Euler-Lagrange Modeling of...

Euler-Euler and Euler-Lagrange Modeling of

Wood Gasification in Fluidized Beds

Michael Oevermann Stephan Gerber Frank Behrendt

Berlin Institute of Technology

Faculty III: School of Process Sciences and Engineering

Department of Energy Engineering

Chair for Energy Process Engineering and Conversion Technologies for

Renewable Energies

Fasanenstr. 89, 10623 Berlin, Germany

CFB-9, May 13-16, 2008, Hamburg

Overview

Motivation

Modeling approaches

Euler-Lagrange / Discrete Element Method (DEM)

Euler-Euler

Results

Summary and outlook

Motivation - simulation of wood gasification

Renewed interest in producing gas from biomass

Gasification in fluidized bed reactors

Lack of detailed knowledge about the complex interactions

between heterogeneous reactions and the fluid mechanics

in fluidized beds

Develop closure laws for large scale models

CFD modeling for particulate flow

Continuum models Euler-Lagrange DNS

tractable problem size

computational cost

Euler-Lagrange model

Basis:

OpenFOAM (www.opencfd.co.uk/openfoam)

Modules:

1 Solver for the gasphase (OpenFoam)

2 Lagrangian particle tracking / DEM (OpenFOAM)

3 Particle combustion / gasification model

(here: simple zero dimensional particle model)

Euler-Lagrange model – gasphase

Mass / momentum balance:

(ρε)t +∇ · ερu = ερs

(ερu)t +∇ · {ερuu}+∇p −∇ · τ + ερg = Fs

Energy / species balance:

(ρεE)t +∇ · [(ρE + p)εu] +∇ · q = Qs

(ρεuYα)t +∇ · [ρεYαu]− εwα = ερα,s

ρs, Qs, ρα,s

Euler-Lagrange model – particle tracking

Equation of motion for every single particle

mid~vi

dt+ ~vi

dmi

dt=

∑j

~Fij

Jid~ω

dt+ ~ω

dJi

dt=

∑j

~Mij

Fluid dynamic forces ( drag, buoyant, Magnus, etc.) withempirical correlations

Contact forces via spring-damper modell

Multiple particle contacts possible

Euler-Lagrange – contact forces by DEM

(Cundall & Strack, 1979)

i jkn

ηn µt

ktηt

~F nij = knδ~n + ηn~vn

ij

~F tij = min

(µt |~F n

ij |~t , ηt |~v tij |

)

Euler-Lagrange model – coupling

Detailed single particle

(Eulerian)

Solid phase

Qp,m

p s

Rp, m

ρg,~ v

g,~ω

g,T

g

~ Fs ,

Qsi,ε

(e. g. T p, Y ps )

φ

(Lagrangian)

iner

tre

actin

g

~r si(t), ~vsi(t)

~ωpi(t),T si(t),Qsi

~r sr(t), ~vsr(t)

~ωpr(t)

model (1d)Rp(t), mp(t), T p(t)

mps(t), Qp(t)

Particle library

mps(t ; φ), Qp(t ; φ)

Rp(t ; φ), mp(t), T p(t ; φ)

ρg(~x , t), ~vg(~x , t)

T g(~x , t), Y gs (~x , t)

Gas phase

Euler-Euler model

Public Domain Code MFIX (www.mfix.org)

1 gas phase

3 solid phases: wood (4mm), 2 x charcoal (1mm, 2mm)

5 homogeneous gas phase reactions (rev.),

4 heterogeneous reactions carbon

Models for primary and secondary pyrolysis

Gas phase chemistry

Simple reaction mechanism

Species: H2, O2, N2, CO, CO2, CH4, H2O, tar

1. 2 CO + O2 2 CO2

2. 2 H2 + O2 2 H2O

3. 2 CH4 + 3 O2 2 CO + 4 H2O

4. CO + H2O H2 + CO2

5. CH4 + H2O CO + 3 H2

Pyrolysis / heterogeneous reactions

Primary pyrolysis (Grönli, Melan (2000))

wood → H2O(g), CO(g), CO2(g), H2(g), CH4(g), tar(g), char(s).

Secondary pyrolysis (Boroson et al. (1989))

tar → CO(g), CO2(g), H2(g), CH4(g), tarinert.

Heterogeneous reactionsC + O2 → CO2 (Ross and Davidson(1982))

C + CO2 → 2 CO (Biggs and Agarwal(1997))

C + H2O → CO + H2 (Hobbs et al. (1992))

C + 2 H2 → CH4 (Hobbs et al. (1992))

Reactor at the department

Holzdosierung

Holzzufuhr

Reaktor

Gasfilter

Gasbrenner

Ø 95

42

60

0

110

0

Ø 5

0

50

Ø134

fuel inlet

air

productgasoutlet

Reactor at the department

Reactor – inlet boundaries

Ø 95

42

60

0

110

0

Ø 5

0

50

Ø134

fuel inlet

air

productgasoutlet wood inflow:

wood T = 150 ◦C, v = 0.08 cm/s

bulk density 0.385 g/cm3

air inflow:

gas species 21% O2, 79% N2

T = 400 ◦C, v = 25 cm/s

Results - Euler-Euler model

play play

Results – exhaust gas composition

Tar

H2O

CH4

CO

CO2

Time / s

Results - Euler-Euler model

0

0,05

0,1

0,15

0,2

0,25

E1 E2 E3 E4 E5 E6 E7 E8 E9 E10E11

E12E13

E14E15

E16E17

S1 S2

Vol % hydrogen carbon monoxidecarbon dioxide hydrocarbons

Experiments Simulations

Euler-Lagrange / DEM – results

Model assumptions:

proportional particle massloss and shrinking

transient calculation of particle mass-loss and energy

no heterogeneous reaction (in progress)

no particle-particle heat transfer

no radiation

Euler-Lagrange / DEM – results

play play

Summary and outlook

Euler-Lagrange and Euler-Euler model for woodgasification in fluidized beds

Euler-Euler: reasonable agreement with experimental data

Euler-Lagrange: allows detailed investigation ofparticle/chemistry/gas phase interactions

Refined chemistry

Parallelization of the DEM modell

Quantitative comparison between Euler-Euler,Euler-Lagrange/DEM und experiments

Coupling Euler-Lagrange with Euler-Euler methods

The financial support of the

Deutsche Bundesstiftung Umwelt (DBU)

is gratefully acknowledged!

Governing equations – Euler-Euler model

Species mass balance gas phase

∂t(εgρgYgs) +∇ · (εgρg~vgYgs) = ∇ · (Dgs∇Ygs) + Rgs

Mass balance solid phase

∂t(εmρmYms) +∇ · (εmρm~vsmYms) = ∇ · (Dms∇Yms) + Rms

Governing equations – Euler-Euler model

Momentum balance solid phase

∂t(εmρm~vm)+∇·(εmρm~vm~vm) = −εm∇pm+∇·τm−

nm∑l=1l 6=m

~Iml +~Igm+εmρm~g

Governing equations – Euler-Euler model

Energy balance gas phase

εgρgcpg

(∂Tg

∂t+ ~vg · ∇Tg

)= −∇ · ~qg +

Nm∑m=1

γgm(Tm − Tg

)−∆Hg

Energy balance solid phases

εmρmcpm

(∂Tm

∂t+ ~vm · ∇Tm

)= −∇ · ~qm − γgm

(Tm − Tg

)−∆Hm