Bench-scale experiments and CFD simulations for Low Aqueous Solvent with different packings

Zhijie (Jay) Xu, Pacific Northwest National Lab (PNNL)Richard Zheng, Pacific Northwest National Lab (PNNL)

Christine Anderson-Cook, Los Alamos National Lab (LANL)

Challenges & Objectives

22

Challenges:Packing characterization at device scale: Effective mass transfer area ae

Gas-film mass transfer coefficient kg

Liquid-film mass transfer coefficient kl

Counter-current

flow

Flue Gas

Lean Solvent

PurifiedGas

Rich Solvent

to Stripper

film thickness~ O(0.1mm)

Layer height:~20cm

Corrugation height:O(10 mm)

Column:Φ ~10mH ~30m

Source4

Structured

Random

Objectives: Using CFD to

directly model mass transfer area at bench-scale

understand the local hydrodynamics /mass transfer with complex geometry

Using bench-scale column exp. to study the performance of

solvent/packing validate the CFD area model

Outline

Bench-scaled Packed Column (Overview) Experiments (EEMPA & MEA) CFD Simulations

• 1:1 full size column CFD modeling• Representative column CFD modeling

Experiment / CFD Comparison Plan for Sequential Design of Experiment (SDoE) Conclusion

3

Outline

Bench-scaled Packed Column Overview Experiments (EEMPA & MEA) CFD Simulations

• 1:1 full size column CFD modeling• Small size column CFD modeling

Experiment/CFD Comparison Plan for Sequential Design of Experiment (SDoE) Conclusion

4

Bench-scaled Packed Column Overview

Column Design

Gas

Solv

ent

Experiment Column

Solv

ent

Gas

1:1 Full Scale CFD Model

RepresentativeColumn Model

5.0 cm

1.0 cm

Packing Region

Sensitivity study and

SDoE(Sequential Design

of “Experiment”)

Bench-scale Column Design: Glass Jacket Column Diameter 3’’, Height 21’’Packing Type: Raschig rings

Diameter: 6 mm Height: 6 mm

8000-9000 rings Material: 316SS, Nylon 6 Porosity: 68% Specific area: 835 m2/m3

Solvents: MEA EEMPA

MEA EEMPAViscosity 𝜇𝜇

[cP] 1.4 7.1

Surface Tension 𝜎𝜎[N/m] 0.067 0.034

Outline

Bench-scaled Packed Column Overview Experiments (EEMPA & MEA) CFD Simulations

• 1:1 full size column CFD modeling• Representative column CFD modeling

Experiment/CFD Comparison Conclusion

6

Experiments of Bench-scale Packed Column

Packing Material 316 Stainless Steel Nylon 6

H = 6mmOD= 6mm

Wall=0.82 mm

Solvent/packing pairs: MEA and 316SS EEMPA and 316SS MEA and Nylon 6 EEMPA and Nylon 6

Measured Carbon Capture Efficiency (CE) CE increase with 𝑢𝑢𝐿𝐿/𝑢𝑢𝐺𝐺 Effective area back out from CE

0.005 0.01 0.015 0.020.15

0.2

0.25

0.3

0.35

0.4

0.45

Capt

u re

Effic

ienc

y[%

]

EEMPA+SS316EEMPA+Nylon 6MEA+SS316MEA+Nylon 6

Outline

Bench-scaled Packed Column Overview Experiments (EEMPA & MEA) CFD Simulations

• 1:1 full size column CFD modeling• Representative column CFD modeling

Experiment/CFD Comparison Conclusion

8

1:1 Full Size CFD Model

9

Gas

Liq

uid

Gas Outlet

SolventInlet

Green: Gas InletRed: Solvent Outlet

Top Boundary

Bottom Boundary

Arrangement of Raschig Rings

Sectional plane view

Packed Column

Dimension:ID: 3’’

Height=21’’

Full Size CFD Model Setup Numerical Packing ProcessComposite

Particle Model

Experiment Packing

CFD Simulated Packing

Improved ring contact modeling

Stainless Steel Glass Ring

H = 6mm, OD= 6mm, Wall=0.82 mm 200 Small Ball (D=0.82 mm)

1:1 Full-size CFD ModelSimulation Conditions: Total 9434 Rings Total ring surface area: 1.99 m2 Liquid flow rate: 0.1-0.8 SLPM Gas flow rate: 25 SLPM Solvent viscosity: [1.4, 7.1] cP

10

x

Entrance Effect(with single

point injection)

Findings: Significant entrance effect for single

point injection, distributor required. 20% to 30% of column height before

reach fully distribution Stronger entrance effect for more

viscous solvent

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.00 0.10 0.20 0.30 0.40

Nor

mal

ized

Int

erfa

ce A

rea

a e/a

p

Distance from solvent inlet (m)

SLPM=0.115SLPM=0.215SLPM=0.415SLPM=0.760

Liquid fully distribution distance: ~0.15 m

Solvent: EEMPAViscosity: 𝜇𝜇 = 7.1 cP

Inlet

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.00 0.10 0.20 0.30 0.40Nor

mal

ized

Inte

rfac

e A

rea

a e/a

p

Distance from solvent inlet (m)

SLPM=0.1920SLPM=0.2793SLPM=0.3920SLPM=0.4595

Liquid fully distribution distance: ~0.10 m

Solvent: MEAViscosity: 𝜇𝜇 = 1.4 cP

Inlet

11

12

1:1 Full-size CFD ModelDistributor design in experiment

13

CFD Comparison w and w/o Distributor

3D Printed Distributor

12 drip holes for solvent injection

Distributor will be used for all experiments

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.00 0.05 0.10 0.15 0.20 0.25

Nor

mal

ized

Inte

rfac

e Ar

eaa e

/ap

Distance from solvent inlet (m)

Flow Rate: 0.76 SLPM

No Distributor

With Distributor

10 cm shorter in entrance length

Inlet

Outline

Bench-scaled Packed Column Overview Experiments (EEMPA & MEA) CFD Simulations

• 1:1 full size column CFD modeling• Representative column modeling

Experiment/CFD Comparison Conclusion

14

Representative Column Model

Representative Column Setup Column Size: H = 6 cm, OD= 7.62 cm Packing Height Z =[0 5] cm Raschig Ring Size 6 mm Specific Area ap = 857 m2/m3

15

Top Boundary Bottom Boundary

Gas Inlet/OutletLiquid Inlet/OutletRing Surface

2.9 million mesh with mesh size Run to 8s (3 hours Simulation)

for converged solution Good size for sensitivity study

Mesh Scene5.0 cm

1.0 cm

Packing Region

0 2 4 6 8 100

0.05

0.1

0.15

0.2

0.25

0.3

Nor

mal

ized

Inte

rfac

e ar

ea a

i/ap

27.7 mN/m, = 10o

27.7 mN/m, = 40o

27.7 mN/m, = 60o

27.7 mN/m, = 90o

Sensitivity Study For Solvent Parameters Fixed Parameter:𝜌𝜌𝐿𝐿 = 1077 kg/m3,𝑢𝑢𝐿𝐿 = 1.46 × 10−3m/s Varying one parameter at a time (~30 runs);

16

Contact angle 𝜃𝜃 [10o 90o] Surface tension 𝜎𝜎 [25 70] N/m Solvent viscosity 𝜇𝜇𝐿𝐿 [2 10] cP

Range covers MEA and EEMPA:

20 40 60 80 1000

0.05

0.1

0.15

0.2

0.25

0.3

Nor

mal

ized

Inte

rfac

e ar

ea a

i/ap

= 27.7 mN/m, L

=2.46 cP

= 27.7 mN/m, L

=5 cP

= 27.7 mN/m, L

=10 cP

0 20 40 60 800

0.05

0.1

0.15

0.2

0.25

0.3

Nor

mal

ized

Inte

rfac

e ar

ea a

i/ap

= 2.46 cP, = 10o

= 2.46 cP, = 40o

= 2.46 cP, = 60o

= 2.46 cP, = 90o

Representative Column Model

0 2 4 6 8 100

0.05

0.1

0.15

0.2

0.25

0.3

Nor

mal

ized

Inte

rfac

e ar

ea a

i/ap

27.7 mN/m, = 10o

27.7 mN/m, = 40o

27.7 mN/m, = 60o

27.7 mN/m, = 90o

Sensitivity Study For Solvent Parameters Fixed Parameter:𝜌𝜌𝐿𝐿 = 1077 kg/m3,𝑢𝑢𝐿𝐿 = 1.46 × 10−3m/s Varying one parameter at a time (~30 runs);

17

Contact angle 𝜃𝜃 [10o 90o] Surface tension 𝜎𝜎 [25 70] N/m Solvent viscosity 𝜇𝜇𝐿𝐿 [2 10] cP

Range covers MEA and EEMPA:

20 40 60 80 1000

0.05

0.1

0.15

0.2

0.25

0.3

Nor

mal

ized

Inte

rfac

e ar

ea a

i/ap

= 27.7 mN/m, L

=2.46 cP

= 27.7 mN/m, L

=5 cP

= 27.7 mN/m, L

=10 cP

0 20 40 60 800

0.05

0.1

0.15

0.2

0.25

0.3

Nor

mal

ized

Inte

rfac

e ar

ea a

i/ap

= 2.46 cP, = 10o

= 2.46 cP, = 40o

= 2.46 cP, = 60o

= 2.46 cP, = 90o

Contact angel 𝜽𝜽 has the most significant impact on interface area prediction

Contact angle >> Surface Tension > Viscosity

Representative Column Model

18

SDoE for CFD Initial 50 runs to explore 5-dim

parameter (𝑢𝑢𝐿𝐿,𝑢𝑢𝐺𝐺 , 𝜇𝜇𝐿𝐿,𝜎𝜎,𝜃𝜃) space Selected to cover bench experiment

conditionsCFD Parameters RangeViscosity 𝜇𝜇𝐿𝐿 [cP] [5 15]Surface Tension 𝜎𝜎 [N/m] [0.01 0.04]Contact Angle 𝜃𝜃 [°] [5 80]Solvent Flow Rate 𝑢𝑢𝐿𝐿 [L/min] [0.1 0.9]Gas Flow Rate 𝑢𝑢𝐺𝐺 [SLPM] [10 100]

Effect Summary Sensitivity Score

Sensitivity Rank

Contact Angle 27.341 1

Solvent Flow Rate 13.375 2

Surface Tension 9.434 3Solvent Flow Rate*Solvent Flow Rate 6.484

Others …

Statistical Analysis of 50 CFD Runs

Representative Column Model

Outline

Bench-scaled Packed Column Overview Experiments (EEMPA & MEA) CFD Simulations

• 1:1 full size column CFD modeling• Small size column CFD modeling

Experiment/CFD Comparison Conclusion

19

Experiment / CFD Comparison

Physical Properties of Solvent Solvent type: MEA, EEMPA Three key parameters:

Viscosity 𝜇𝜇𝐿𝐿 Surface tension 𝜎𝜎 Contact angle 𝜃𝜃 (much larger uncertainty)

Three data sources: Measured at PNNL Aspen predicted Existing correlation

Identify the range of these parameters in packed column Quantify the uncertainty in CFD interface area prediction

20

Experiment/CFD Results Comparison

21

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3

CFD

Pre

dict

ed

EEMPA+SS316

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3C

FD P

redi

cted

EEMPA+Nylon 6

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3

CFD

Pre

dict

ed

MEA+SS316

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3

CFD

Pre

dict

ed

MEA+Nylon 6Normalized Interface Area 𝒂𝒂𝒊𝒊/𝒂𝒂𝒑𝒑 Comparison

Experiment Pair Contact Angle [o] (Best Guess)

Surface Tension[N/m] (From

Aspen & Paper)

Liquid Viscosity[cP] (From

Aspen & Paper)EEMPA+SS316 [30, 46] [0.026, 0.038] [6.2, 7.6]

EEMPA+Nylon6 [15, 23] [0.026, 0.038] [6.2, 7.6]

MEA+SS316 [19.8, 52.6] [0.054, 0.07] [1.38, 2.54]

MEA+Nylon6 [19.8, 52.6] [0.054, 0.07] [1.38, 2.54]

Contact angle in column:(difficult to precisely determine) Roughness Geometry Loading Temperature, etc.

Experiment/CFD Results Comparison

22

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3

CFD

Pre

dict

ed

EEMPA+SS316

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3C

FD P

redi

cted

EEMPA+Nylon 6

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3

CFD

Pre

dict

ed

MEA+SS316

0 0.1 0.2

Aspen Predicted

0

0.1

0.2

0.3

CFD

Pre

dict

ed

MEA+Nylon 6Normalized Interface Area 𝒂𝒂𝒊𝒊/𝒂𝒂𝒑𝒑 Comparison

Experiment Pair Contact Angle [o] (Best Guess)

Surface Tension[N/m] (From

Aspen & Paper)

Liquid Viscosity[cP] (From

Aspen & Paper)EEMPA+SS316 [30, 46] [0.026, 0.038] [6.2, 7.6]

EEMPA+Nylon6 [15, 23] [0.026, 0.038] [6.2, 7.6]

MEA+SS316 [19.8, 52.6] [0.054, 0.07] [1.38, 2.54]

MEA+Nylon6 [19.8, 52.6] [0.054, 0.07] [1.38, 2.54]

Contact angle in column:(difficult to precisely determine) Roughness Geometry Loading Temperature, etc.

Contact angel 𝜽𝜽 contribute the most uncertainty to the interface area prediction

Contact angle >> Surface Tension > Viscosity

Next Step: Plan for SDoE

23

CFD Parameters Space Filling

Experiment Parameters Space Filling

SDoEDesign

Aspen PlusBench Cart Experiment

CFD Simulation

SDoEOptimization

Interface Area(ai)

Capture Efficiency (CE)

Compare

MeasuredInput

Improve CFD Modeling

CFD Parameters RangeViscosity 𝜇𝜇𝐿𝐿 [cP] [5 15]Surface Tension 𝜎𝜎 [N/m] [0.01 0.04]

Contact Angle 𝜃𝜃 [°] [5 80]Solvent Flow Rate [L/min]

Same as Exp

Gas Flow Rate [SLPM] Same as Exp

Experiment Parameters

Range

Solvent Flow Rate [L/min]

[0.1 0.9]

Gas Flow Rate [SLPM]

[10 100]

Loading/Reboiler Temperature

<135°C

Absorber Temperature [°C]

[30 60]

Measured Results

Nominal Input

Effective Area (ae)

Regression

Aspen calculated viscosity/surface tensionwill put constraints while

exploring CFD parameter space

Conclusion

24

Validation Against Bench-cart Column Experiment Solvents: MEA, EEMPA Packing: SS316, Nylon 6 Leverage Aspen prediction of properties

CFD Approach Optimized Direct calculation of interface area Full-size column for entrance effect:

Viscous solvent has stronger effect Liquid distributor will reduce 2/3 of the effect

Computationally efficient representative column modelSensitivity Study

Contact angle 𝜃𝜃 has the largest impact to the CFD interface area prediction. We need better understanding of its role and influential factors.

Acknowledgements

25

PNNL: Dushyant Barpaga, Charlie Freeman, David Heldebrant, Jie Bao, Rajesh

Singh, Chao Wang, Yucheng Fu

LANL: Christine Anderson-Cook, Sham Bhat, John Baca, Christopher Russell

NETL: Michael Matuszewski, Benjamin Omell, Joshua Morgan, Grigorios Panagakos

UT Austin: Gary Rochelle

Disclaimer: This work is made available as an account of work sponsored by an agency of the United States Government. Neither the UnitedStates Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability orresponsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its usewould not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United StatesGovernment or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the UnitedStates Government or any agency thereof.

For more informationhttps://www.acceleratecarboncapture.org/

Your Contact InfoZhijie.Xu@pnnl.gov

27

Our research Featured in i&ec JOurnal (POsters)