cfd-assistedprocess intensification for biomass fast ... · cfd-assistedprocess intensification for...

14
CFD-assisted Process Intensification for Biomass Fast Pyrolysis in Gas-solid Vortex Reactor Technology Shekhar R. Kulkarni, Arturo Gonzalez Quiroga, Patrice Perreault, Geraldine Heynderickx, Kevin M. Van Geem, Guy B. Marin 3 rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Upload: others

Post on 17-Mar-2020

12 views

Category:

Documents


0 download

TRANSCRIPT

CFD-assisted Process Intensification for Biomass FastPyrolysis in Gas-solid Vortex Reactor Technology

Shekhar R. Kulkarni, Arturo Gonzalez Quiroga, Patrice Perreault, Geraldine Heynderickx, Kevin M. Van Geem, Guy B. Marin

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Biomass Fast Pyrolysis

Short gas residence time Effective heat transferFast removal of bio-oil vapours & rapid condensation

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Multiphase Chemical Reactors

Process intensification in terms of heat & mass transfer

Gas-Solid slip velocities

Packed beds

Short Gas Residence time

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

G

Gas-SolidVortex Reactors

Centrifugal

Drag

- Gas flow restrictions- Dilute beds

S

G

Drag

Gravity

Fluidized Bed Reactor

GSVR Research @ LCT

Cold-Flow GSVR Hot-Flow GSVR Reactive GSVRCFD

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Reactive GSVR | Tailored Design

Reduced backflow due to profiled bottom plate

Pressure drops ~ 9 kPa (slots) | ~ 20 kPa (total)

Pa

Uniform velocity across slotsm s-1

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Top View

Front View

CFD Simulations

Parameter Full Geometry Pie Geometry

Air inlet temp (K) 289

Air inlet flow (kg s-1) 0.0143 0.00179

Aluminium loading (kg) 0.0107

Aluminium density (kg m-3) 2700

Aluminium dp (m) 0.0005

Aluminium feeding Via UDF (0.0385 < r < 0.0395 m)

Turbulence model Re-Normalization Group k-ε

Shear Stress Transport k-w

Time step (s) 2 X 10-5 10-4

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Pie Geometry

~ 0.25 x 106 cells

Full Geometry

~ 2.5 x 106 cells

ANSYS FLUENT v18.0

Eulerian – Eulerian

Full vs Pie Geometry Comparison

Pie-geometry can be chosen for computational ease.

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Biomass Fast-Pyrolysis Reactive Simulations

Fast Pyrolysis Lumped Reaction Mechanism

Biomass (dry) v.Hemicellulose

v.Lignin

v.Cellulose

(36 %)

(47 %)

(17 %)

a.Hemicellulose

a.Lignin

a.Cellulose

Gas

Gas

Gas

Bio-oil

Char_c

Bio-oil

Char_h

Bio-oil

Char_l + Gas

+ Gas

+ Gas

Biomass Phasedp = 0.5 mm

Char Phasedp = 0.2 (0.3 mm)

Xue, Qingluan, T. J. Heindel, and R. O. Fox. "A CFD model for biomass fast pyrolysis in fluidized-bed reactors." Chemical Engineering Science 66, no. 11 (2011): 2440-2452.

Parameter Value

N2 inlet temp (K) 842

N2 inlet flow (kg hr-1) 18

Biomass loading (kg) 0.001 (batch feed)

Biomass feed temp (K)

842

Time step (s) 10-4

Turbulence model k-ε RNG

Primary phase Gas Mixture

Secondary phase – I Biomass Phase

Secondary phase – II Char Phase

Interphaseinteractions

Drag: GidaspowHeat Transfer: Gunn

Simulation Settings

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Key Results (Results scaled for the full reactor configuration)

Product

Distribution

Previous (2D)

Simulations1

Current (3D)

Work

Char 14 – 17 % 20.92 %

Bio-oil 73 – 76 % 68.19 %

Pyrolysis Gas 8.5 - 9.5 % 10.89 %

1 Ashcraft, Robert W., Geraldine J. Heynderickx, and Guy B. Marin. "Modeling fast biomass

pyrolysis in a gas–solid vortex reactor." Chemical engineering journal 207 (2012): 195-208.

Time required for

complete conversion

~ 8 sec

Slot ~ 4-5 kPa

Bed ~ 1-2 kPa

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Process Intensification : Diameter based Segregation

Biomassdp = 0.5 mmρ = 450 kg m-3

Chardp = 0.2 mmρ = 500 kg m-3

Solids v.f. profiles displayed at axial plane: z = 0.01 m

- Density ratio of 0.9 & dp ratio of

2.5 show positive radial

segregation

- Streamlines near the outlet

indicate likeliness of char exiting

the reactor as rather than

biomass

- Segregation is transient and char

bed moves radially outwards as

biomass reacts.

- To sustain segregation and

reduce char residence time in

reactor, continuous biomass

feeding could be implemented.

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Fast Pyrolysis Advanced Reaction Mechanism

Levoglucosan, Glyoxal, Acetaldehyde, Hydroxymethylfurfural, Permanent Gases

Formaldehyde, Xylan, Permanent Gases

Phenol, Acrylic Acid, pCoumaryl, Permanent Gases

Gas-PhaseBiomass Phase

v.Hemicellulose

v.Lignin

v.Cellulose

a.Hemicellulose - 1

Lignin - C

a.Cellulose

Lignin - HLignin - O

a.Hemicellulose - 2 Char

Char-Phase

Permanent Gases : N2, H2, CO, CO2, CH4, etc

1. Ranzi, Eliseo, et al. "Chemical kinetics of biomass pyrolysis." Energy & Fuels 22.6 (2008): 4292-4300.

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Biomass (dry)

Advanced Kinetics Scheme : Preliminary product fields

Levoglucosan v.f.Biomass v.f. Char v.f.

Gas Temp Biomass Temp

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Solids v.f. profiles displayed at axial plane: z = 0.01 m

Simulation settings same as

mentioned previously

Concluding Remarks Lumped and advanced kinetic mechanisms implemented successfully in the CFD of GSVR.

Pie-geometry suitable for running qualitative reactive simulations.

3D Simulations indicate transient radial char and biomass segregations within a range of biomass tochar particle diameter ratios � process intensification favorable for fast pyrolysis.

Char, bio-oil yields lower than those in previous (2D) reactive simulations, indicating strong influence ofend-wall effects.

13/73

3rd Annual SMARTCATS Meeting, October 25 - 27, Prague

Future Work Reactive simulations with realistically shaped biomass particles

Validation of reaction UDF with simulations in a fluidized bed.

Acknowledgements3rd Annual SMARTCATS Meeting, October 25 - 27, Prague