Enabling 10 mol/kg swing capacity via heat · 2017. 12. 19. · Enabling 10 mol/kg swing capacity...

Enabling 10 mol/kg swing capacity via heat integrated sub-ambient pressure swing adsorption

Ryan P. Lively

Georgia Institute of TechnologySchool of Chemical & Biomolecular Engineering

Atlanta, GA 30332

DOE-NETL Kick-off MeetingThursday, December 3rd, 2015

2

Chemical & Biomolecular Engineering at Georgia Tech

ATLANTA, GEORGIA

• Economy is 6th in United States, 15th in world• Airport is world’s busiest

• Largest college of engineering in the United States

• Only U.S. institution with all graduate engineering programs in top 10 (USNWR 2014, 2015)

• Only U.S. institution in top 5 for R&D spending in both chemistry and chemical engineering (NSF data, 2013)

• 1000+ undergraduate students

• 200+ Ph.D. students• No. 1 in United

States in NTU 2013 Performance Ranking

GEORGIA TECH CHBE ATGEORGIA TECH

24 sponsored research projects with budgets over $1M active in 201439% of research funds in 2014 were industry research

3

Inmondo Tech, Inc.www.inmondotech.com

Focused on scale-up and supply of metal-organic frameworks for adsorption applications

Key Capabilities:• Production of selected MOFs at

10 g – 1 kg scales• Delivery of MOFs as fine

powders and as engineered forms

• Custom synthesis of MOFs• Partnering with customers on

application-oriented prototyping and testing

4

http://www.inmondotech.com

BudgetDOE Contribution1st year: $705,4412nd year: $681,8453rd year: $599,698Total: $1,986,984 (79%)

Cost Share Provided by Georgia Tech: $513,792 (21%)

Total Budget: $2,500,776

5 primary researchers supported (2 post-doctoral researchers, 3 graduate student researchers)

5 PIs supported (Lively, Kawajiri, Realff, Sholl, Walton)

Major equipment purchases/construction: Sub-ambient rapid pressure swing adsorption units

5

Personnel

Principal Investigators:

Georgia TechRyan Lively, Project Director, Inmondo liaison, hollow fibers and RCPSA system

Yoshiaki Kawajiri, Process optimization, cyclic adsorption processes

Matthew Realff, Process systems engineering, technoeconomic analysis

David Sholl, Adsorption and diffusion in nanoporous materials

Krista Walton, Adsorption in MOFs and MOF synthesis

Inmondo TechDr. Amy Beaird, Inmondo PI, Sorption and gas storage

6

Key Idea:Combine:

(i) Sub-ambient gas processing and energy recovery with

(ii) ultra-porous metal-organicframeworks and

(iii) space- and energy-efficient fiber sorbent contactors

to yield a game-changing process strategy

7

Background—Adsorbent materials

Chemisorbents-High ∆Hads-Steep isotherm-Strong binding

--amines--CaO--hydrotalcites

Physisorbents-Low ∆Hads-Shallow isotherm-Weak binding

--zeolites--carbons--metal-organic frameworks

andMOFs

[1] Choi, Drese, Jones, ChemSusChem 2009, 2(9):796-8548

MOFs

Background: Metal-organic frameworks—State-of-the-artMmen-Co2(dobpdc)

Decreasing temperature

[1] TM McDonald, JR Long et al., Nature, 2015, 519, 303-308

UiO-66

Mmen-Mg2(dobpdc)CO2 Capture Performance

[3] JM Simmons, T Yildirim et al., Energ. Env. Sci., 2011, 4(6), 2177-2185

[2] GE Cmarik, KS Walton et al., Langmuir, 2012, 28(44), 15606-15613

Pressure (torr)

25°C

9

75°C

Background: Metal-organic frameworks—State of the artWater stability

Diffraction patterns before and after H2O

exposure (ZIF-8):

[1] KS Park, OM Yaghi et al., Proc. Natl. Acad. Sci. 2006, 103(27), 10186-10191

Pore volume

(gm3/g)Mg-MO F-

740.65 0.62 1400 238 83

UiO -66 0.52 0.37 1160 1130 2

DMO F-1 0.58 0.04 1960 7 100

HKUST-1 0.62 0.49 1270 945 26

UMCM 2.41 0.11 6010 205 97

Surface Area (m2/g)

MaterialWater

loading, (cm3/g)

Before After %Loss

Comparison of MOF material stability before and after water vaporexposure

[2] NC Burtch, KS Walton et al., Chem. Rev.. 2014, 114(20), 10575-10612

Thermodynamic Stability

Kinetic Stability

vs.

10

Background: Metal-organic frameworks—State of the artMOF Isotherm prediction / Gate opening/Breathing/Structural isomerism

[1] F-X Coudert, C Mellot-Draznieks et al., J. Am. Chem. Soc. 2008, 130(43), 14294-14302

[2] C Zhang, DS Sholl, RP Lively et al., J. Phys. Chem. Lett.. 2014, 118(35), 20727-20733

11

Background: Solid adsorbent contactors

• MOF adsorbents can provide good CO2 capacity and rapid adsorption/desorption kinetics if they can be kept isothermal.

• Heat of adsorptions are low to moderate, but large uptakes necessitate efficient heat management (Qreleased = nadsΔHads).

• An effective, scalable contactor for MOF adsorbents that allows for acceptable adsorption/desorption properties, cost, and ability to yield high purity CO2 has not yet been demonstrated.

-- fluidized bed, looping processes give significant sorbent attritionand typically yield depressed working capacities.

-- fixed bed processes make heat management & Δp highly problematic

• An ideal contactor would offer high surface areas at low cost, allow for isothermal adsorption, have low pressure drops, and would yield a concentrated CO2 product.

12

Background: Hollow fiber sorbents, a mass producible structured sorbent inspired by hollow fiber membrane spinning

Ideal temperature swing adsorption1000 µm

[1] RP Lively, WJ Koros et al., Ind. Eng. Chem. Res., 2009, 48(15), 7314-732413

Bundle of 40 fibers in a 1.5’ module at GT

Background: module evolution moved membranes from lab to large scale–and fiber sorbent modules are now on a similar path

Hollow Fiber

Spiral Wound

Plate & Frame

104 m2/m3

Large CO2/CH4 module76 cm OD x 1.8 m

Used on 700 MMSCFD offshore platformCourtesy, E. S. Sanders NAMS 2003 plenary

Parallel connection of 20 cm OD modules

(Desalination plant) N. Alon: Demotix , Israel

Fluid delivery

Spinneret

Water quench bath

Fiber take-up

800-1200 µm OD

103 m2/m3102 m2/m3

Area packing density

High contact area, low pressure drop and mass transfer resistance make fibers the most evolved module

form

Toyobo Water Treatment Plant

14

http://www.demotix.com/photo/1273705/minister-gilad-erdan-prepares-rio-de-janeiro-conference-ashkelon


120°C

Background: Rapid temperature swing adsorption (RTSA)

120°C34°C

0.15 psi/ft Δp

4 min

[1] RP Lively, WJ Koros et al., Int. J. Greenhouse Gas Control 2012, 10, 285

Plug of CO2

15

2.5 minute cycles are possible

15

0 10 20 30 40 500.00.20.40.60.81.01.21.4

q b (m

mol

/g)

Number of Cycle 16

qb,cooled = 1.33 mmol/gqb,uncooled = 1.10 mmol/g

Reduced 8 oC by flowing CW

Background: Cooled Torlon-C803-PEI Fiber Sorbent Generation 3 Fibers

Lab scale heat capture efficiency during adsorption: ~72%

qb remains ~ 1.1 mmol/g over 50 cycles Gen. 2 fibers

Gen. 3 fibers:qb~1.4 mmol/g

qswing~0.85 mmol/g

Target qS ~1 mmol/g

Estimated escalation factor:

1.30-1.35

17

Past case studies for amine RTSA system: swing capacity is critical for capture cost

Particle size (µm) Equilibrium capacity (mmol/g-fiber)

Processoperation

Base 4.0 1.33 RTSA

(a) 4.0 1.33 RVTSA(b) 1.0 1.33 RVTSA(c) 4.0 2.13 (PIM) RVTSA(d) 1.0 2.13 (PIM) RVTSA

(a)

(b)(c)

(d)

Base

Background: Fiber sorbents for PSA applications

2 µm

H2/CO2 separations

18

[1] RP Lively, WJ Koros et al., Int. J. Hydrogen Energ., 2012, 37(20), 15227-15240

PROCESS & PROJECT SCOPE

19

Project scope—Re-thinking general assumptions about post-combustion CO2 capture • Rapid pressure swing adsorption is

more straightforward than rapid temperature swing adsorption (has been commercialized)

• Sub-ambient conditions increase adsorption selectivity and working capacity—even without adsorbent structural changes

• Immense pore volume and surface area of MOFs can be advantageously utilized at sub-ambient conditions and moderate CO2 partial pressures (~1-2 bar)

• Weaknesses of MOFs addressed through contactor (hollow fiber sorbents) and through process strategy

20

RCPSA

Sub-ambient membrane system

Process Scope—Key Topics

Five major activity areas are proposed in this work:

(1) UiO-66 synthesis, sub-ambient adsorption characterization, and stability,

(2) Composite hollow fiber spinning (cellulose acetate/polysulfonefibers containing UiO-66 sorbents),

(3) RCPSA system construction and testing of fiber sorbent modules and hollow fiber sorbent modules with bore-side phase change material,

(4) Modeling and optimization of fiber and hollow fiber module operation as well as flue gas conditioning optimization, and

(5) Overall system techno-economic analysis.

21

Process overview

Advantages: • Low dew point flue gas• improved SOX removal from cold water• CO2 liquefaction and pumping can be used instead of CO2 compression• sub-ambient HEX and CO2 liquefaction are commercially demonstrated

22

Ads

Rapid pressure swing adsorption (Tasks 4, 6, 17, 19, 20)The current project aims to produce polymeric hollow fiber contactors loaded with MOF adsorbent particles for sub-ambient post-combustion CO2 capture.

-- use known and scalable MOFs (e.g., UiO-66)

-- use known polymers for fiber spinning (e.g., cellulose acetate, Matrimid®)

-- adapt lessons learned during NaY RPSA case to MOF case.

-- evaluate base process economics and optimize to minimize costs.

23

MOF Hollow fiber sorbents (Tasks 2, 3, 9) ZIF-8 fiber sorbents

24

Fluid delivery

Spinneret

Water quench bath

Fiber take-up

MOF synthesis and sub-ambient isotherms (Tasks 2, 7, 16)25

[1] PM Schoenecker, KS Walton et al., AIChE J. 2013, 94(4), 1255-1262

V1

V2

P1

Gas dosing

Q (helium cryostat) T• 5 mm diameter pellets• 40-50 mg of MOF per pellet• ~ 5000 pellets per hour

Fiber module modeling (Task 12-13)

z

r Fiber + MOF

Gas flow

PCM

Lumen layer

( )2

2

( , , ) ( , , ) ( , , ) ( , , )11

f f f f if pf f ads

f

T r z t λ T r z t T r z t q r z tρ C ρ Ht r r r tε

∂ ∂ ∂ ∂− + = ∆ ∂ ∂ ∂ ∂−

PCM PCM

PCMPCM PCM,sol PCM PCM

PCMPCM PCM,liq PCM P

latent

CM

: ( ( , , ) )( ) ( ) 1

( )

, 0

( ): ( ( , , ) ( ) 0

(

),

( ): ( ( , , ) ( )),

PCM melt f ID melt

PCM melt f ID

PCM melt f I

I

p I

p I D

T T A h T r z t T

T tT C A h T r z t T tt

T tT T C A h T r z t T tt

tH tt

T t

t

ρ

ρ

ρ

α α

α

α

= = − <

∂< = −

∂∂

> = −∂

∂∆ <

∂

=

) 1

=

Mass balance:

( )

, ,,

,

( , ) ( , ) ( , )( , ) ( , )

( , ) ( , , ) 0

i g i g gg i g

g O f ii g OD

c z t c z t u z tu z t c z t

t z zk A c z t c r z tε

∂ ∂ ∂+ +

∂ ∂ ∂+ − =

2

2

( , ) ( , ) ( , ) ( , )1 (1 ) 0i i i ieff f

c z t c z t c z t q z tDt r r r t

ε ε ρ ∂ ∂ ∂

− + + − = ∂ ∂ ∂

∂∂

( )( , , ) ( , , ) ( , , )eqii i

q r z t k q r z t q r z tt

∂= −

∂( , , )( , , )

1 ( , , )

seq i i ii

i i

q b P r z tq r z tb P r z t

=+

0,0

1 1exp( , , )

ii i

f

Hb bR T r z t T

−∆= −

Heat balance:

26

Model validation (Task 18)

Rezaei, Subramanian, Kalyanaraman, Lively, Kawajiri, Realff Chem. Eng. Sci., 2014, 113, 62-76. Kalyanaraman, Fan, Lively, Koros, Jones, Realff, Kawajiri Chem. Eng. J., 2015, 259, 737-751

Fitting and validation of model against experimental data using fiber module

27

Model integration into steady-state flowsheet model (Tasks 14-15)

Flowsheet model

Minimize energy usage

gPROMS dynamic model

Maximize CO2 recovery & purityMaximize CO2 throughput

Update:Cycle timesFiber dimensions

Time averaged concentration & temperatures

z

r Fiber + MOF

Gas flow

PCM

Lumen layer

Swernath, S.; Searcy, K.; Rezaei, F.; Labreche, Y.; Lively, R. P.; Realff, M. J.; Kawajiri, Y., Computer Aided Chemical Engineering, You, F., Ed. Elsevier: Waltham, MA, 2015; Vol. 36, pp 253-278.

28

High-level Technoeconomic Assessment (Task 21)

• Levelized cost of electricity (mills/kWh)• Cost of CO2 capture ($/ton) • Capital Cost ($) • Escalation factor

Key metrics for economics

1. US Department of Energy, November 2010. Cost and Performance Baseline for Fossil Energy Plants. Rev 2. In: Bituminous Coal and Natural Gas to Electricity, vol. 1. DOE/ NETL-2010/1397.

2. US Department of Energy, April 2011. Quality Guideline for Energy System Studies (QGESS) Cost Estimation Methodology for NETL Assessments of Power Plant Performance. DOE/NETL-2010/1455.

3. US Department of Energy, January 2011. Bench-scale and Slipstream Development and Testing of Post-Combustion Carbon Dioxide Capture and Separation Technology for Application to Existing Coal-fired Power Plants. DOE/NETL DE-FOA-0000403.

4. US Department of Energy, August 2012. Updated Costs (June 2011 Basis) for Selected Bituminous Baseline Cases. DOE/NETL-341/082312.5. US Department of Energy, January 2013. Quality Guideline for Energy System Studies (QGESS), Capital Cost Scaling Methodology.

DOE/NETL-341/013113.

Uncertainty: cost of sub-ambient heat exchanger Sensitivity analysis will be performed

29

Initial high level process analysis using 10 mol/kg swing 30

• Plant parasitic load: 16.7%

• Installed capital costs: $188.2 x 106

• Total annual cost (operating expenses + amortized capital): $72.8 x106/yr

• CO2 captured per year: 4.50 x 106 tons CO2/yr

• CO2 capture cost: $19.0/ton, $21/tonne

• Number of 8” module elements needed: 36,000

• Estimated footprint (assuming modules stacked 10 high): ~200 m2 fora 500 MWe coal-fired power plant



RisksDescription of Risk Probability (Low,

Moderate, High)Impact

(Low, Moderate, High)Risk Management

Mitigation and Response StrategiesTechnical Risks:MOFs do not exhibit ~10 mol/kg swing capacity between 1 and 2 bar CO2 partial pressure

Moderate Moderate (a) Technoeconomic analysis and modeling effort will determine impact of > 10 mol/kg swing vs. 5-10 mol/kg swing.

(b) Other scalable MOFs can be considered if >10 mol/kg is critical

MOF particles do not survive spinning process

Low High If required, MOFs can be grown within porous polymer supports post-spinning. Preliminary data from Lively indicates that MOFs retain their porosity and crystallinity post-spinning.

Failure of sealing for fiber modules

Low Moderate If required, specialty epoxies resistant to temperature changes will be used to seal modules.

Instability of MOF to flue gas contaminants

Low High Functionalized MOFs will be tested, providing various materials for use. Additional flue gas processing can be used if necessary.

Resource Risks:Delays in production of MOFs by Inmondo Tech

Low High MOFs will be produce in excess of minimum requirement in year 1 to ensure availability. Capability to deliver materials in this manner has already been demonstrated.

Management Risks:Difficulty in recruiting postdocs/grad students at GT

Low Moderate Shift personnel between tasks to manage temporary vacancies

Lack of coordination among project partners

Low Moderate Project partners already have a proven record of collaboration; regular project meetings are scheduled with all partners

31

BP 1 Task List & Milestones 32

Budget Period

Task/Subtask No. Milestone Description Planned Completion Actual Completion

Verification Method

1 1 Updated Project Management Plan

08/31/2015 Project Management Plan file

1 1 Kickoff Meeting 10/01/2015 Presentation file1 2.1 Produce 250+ g of UiO-66

@ >900 m2/g surface area and >2.5 mol CO2/kg @ 273K & 1 bar

01/31/16 Report to DOE

1 2.2 Generate sub-ambient isotherms


1 2.3 Syringe fibers using UiO-66


1 3 Spin monolithic fibers 09/30/16 Report to DOE1 4 RCPSA module

construction & seal testing09/30/16 Report to DOE

1 5 Bare fiber module model development


BP2 Task List & Milestones 33

2 6 Test PSA using syringe fiber samples


2 7 Produce 250+ g of UiO-66 @ >900 m2/g surface area and >2.5 mol CO2/kg @ 273K & 1 bar


2 8 MOF moisture and acid gas test (SO2 and steam exposure)


2 9 Lumen layer synthesis and barrier properties


2 10 Demonstrate hollow fiber lumen layer synthesis


2 11 PCM integration into modules


2 12 Model development of fibers with PCM


2 13 Modeling phase change and adsorption using experimental data


2 14 Process flowsheet optimization


BP3 Task List & Milestones 34

3 15 Process flowsheet refinement


3 16 Produce 250+ g of UiO-66 @ >900 m2/g surface area and >2.5 mol CO2/kg @ 273K & 1 bar


3 17 Construct/test RCPSA for dirty gas testing


3 18 Model validation for hollow fiber module—model validation for composite fibers


3 19 Monolithic fiber sorbent stability in dirty gas RCPSA


3 20 Test hollow fibers containing phase change material in PSA


3 21 Complete Technoeconomic assessment


3 22 Test a demonstration module


Summary 35

• Novel polymer/MOF sorbent composite hollow fibers will be used in new sub-ambient RPSA process for post-combustion CO2 capture

• 50% experimental demonstration• 50% prediction, modeling, optimization, and economic feasibility analysis

• Viability of concept will be demonstrated• Potential for game-changing swing capacities by utilizing MOFs in sub-ambient

conditions

• Georgia Tech and Inmondo Tech are partners on this project

• Annual reports, annual review meetings and conferences presentations and quarterly reports will be used to update DOE on team activities

• DOE contribution: ~$2.0M Georgia Tech contribution: ~$0.5M

Enabling 10 mol/kg swing capacity via heat · 2017. 12. 19. · Enabling 10 mol/kg swing capacity...

Documents

Transcript of Enabling 10 mol/kg swing capacity via heat · 2017. 12. 19. · Enabling 10 mol/kg swing capacity...