Aspen RateSep Absorber Model for CO 2 Capture CASTOR Pilot Plant IFP – Lyon, France by: Ross Dugas January 11, 2008

Scope of the Presentation Objective Objective Introduction Introduction Data Improvements in Aspen Data Improvements in Aspen Density Density Viscosity Viscosity Thermodynamics – Heat of Formation, Heat Capacity Thermodynamics – Heat of Formation, Heat Capacity Kinetics Kinetics Model Parameters Model Parameters k g, k L, liquid holdup, film discretization, etc. k g, k L, liquid holdup, film discretization, etc. Results Results Conclusions Conclusions

Objective Create an Aspen RateSep model to simulate absorber pilot plant data from the CASTOR project Create an Aspen RateSep model to simulate absorber pilot plant data from the CASTOR project CO 2 profiles, Temp profiles CO 2 profiles, Temp profiles The absorber model will aid in the evaluation and optimization of operating conditions The absorber model will aid in the evaluation and optimization of operating conditions liquid rate, lean loading, gas temperature, packing height, packing type, etc. liquid rate, lean loading, gas temperature, packing height, packing type, etc.

Introduction CASTOR Project CASTOR Project 12 experimental runs 12 experimental runs 1.1 meter diameter absorber 1.1 meter diameter absorber Four 4.25 meter beds of IMTP-50 (17m total) Four 4.25 meter beds of IMTP-50 (17m total) MEA Concentration: 30 – 33 wt% (CO 2 -free basis) MEA Concentration: 30 – 33 wt% (CO 2 -free basis) Lean Loading: mol/mol Lean Loading: mol/mol Lean Flow Rate: 13 – 24 m 3 /m 2 h Lean Flow Rate: 13 – 24 m 3 /m 2 h T LEAN = 40C T FG ≈ 48C T LEAN = 40C T FG ≈ 48C y CO2 = 10 – 12% (Saturated basis) y CO2 = 10 – 12% (Saturated basis) Q FG ≈ 5000 Nm 3 /h Q FG ≈ 5000 Nm 3 /h

Data Improvements – Density Aspen defaults incorrectly predicted decreasing density with increasing loading Aspen defaults incorrectly predicted decreasing density with increasing loading Adjust Aspen parameters Adjust Aspen parameters Weiland (1998), wt%, 40-80C correlations Weiland (1998), wt%, 40-80C correlations Parameters for MEA redefined Parameters for MEA redefined MEAH + /MEACOO - and MEAH + /HCO 3 - defined MEAH + /MEACOO - and MEAH + /HCO 3 - defined

Data Improvements – Viscosity Aspen defaults underestimated viscosity Aspen defaults underestimated viscosity Adjust Aspen parameters Adjust Aspen parameters Weiland (1998), wt%, 40-80C correlations Weiland (1998), wt%, 40-80C correlations Parameters for MEA redefined Parameters for MEA redefined MEAH +, MEACOO - and HCO 3 - defined MEAH +, MEACOO - and HCO 3 - defined

Data Improvements – Heat of Formation Heat of absorption was inconsistent within Aspen Heat of absorption was inconsistent within Aspen 5 reactions (Freguia (2002)) 5 reactions (Freguia (2002)) K eq data - Van't Hoff equation K eq data - Van't Hoff equation Heat of formation data in Aspen Heat of formation data in Aspen Heat of formation defined at 25C Heat of formation defined at 25C updated: MEAH +, MEACOO -, HCO 3 -, CO 3 -2 updated: MEAH +, MEACOO -, HCO 3 -, CO 3 -2

Data Improvements – Heat Capacity C p used to match ∆H abs at higher temperatures C p used to match ∆H abs at higher temperatures 40, 60, 80, 100, 120C 40, 60, 80, 100, 120C C p of MEAH + and MEACOO - set to C p of MEA C p of MEAH + and MEACOO - set to C p of MEA Empirically known from heat exchangers Empirically known from heat exchangers

Data Improvements – ∆H abs ∆H abs – K eq equations vs Aspen (∆H form, C p ) parameters ∆H abs – K eq equations vs Aspen (∆H form, C p ) parameters CO 2 Loading - 0.2, 0.3, 0.4, 0.45, 0.5 CO 2 Loading - 0.2, 0.3, 0.4, 0.45, 0.5 Temperature – 25, 40, 60, 80, 100, 120C Temperature – 25, 40, 60, 80, 100, 120C Discrepancy of ±3% Discrepancy of ±3% Improves energy-material balance consistency Improves energy-material balance consistency

Data Improvements – Kinetics Over 25 sources of MEA kinetics for dilute, unloaded solutions Over 25 sources of MEA kinetics for dilute, unloaded solutions Currently only 1 data source for highly concentrated, highly loaded MEA solutions Currently only 1 data source for highly concentrated, highly loaded MEA solutions Aboudheir (2002), laminar jet absorber Aboudheir (2002), laminar jet absorber Rate constants from dilute, unloaded systems don't directly apply to CASTOR conditions Rate constants from dilute, unloaded systems don't directly apply to CASTOR conditions Aboudheir data was verified by matching to dilute, unloaded literature data Aboudheir data was verified by matching to dilute, unloaded literature data Activity coefficient and D CO2 corrections Activity coefficient and D CO2 corrections

Aboudheir data presented on unloaded basis

Data Improvements – Kinetics Ionic strength effect quantified and implemented into Aspen kinetics Ionic strength effect quantified and implemented into Aspen kinetics

Model Parameters Aspen RateSep Aspen RateSep 1.1 m diameter 1.1 m diameter 17m of IMTP-50 packing 17m of IMTP-50 packing Aspects Considered Aspects Considered Solvent Degradation Heat Loss Number of Stages Reaction Film Discretization Pressure Drop Interfacial Area Liquid Holdup Gas Film MT Coefficient (k G ) Liquid Film MT Coefficient (k L )

Model Parameters Reaction Film Discretization Reaction Film Discretization RateSep feature allows the reaction film to be subdivided. RateSep feature allows the reaction film to be subdivided. Reaction rates calculated for each segment Reaction rates calculated for each segment Reaction film broken into 6 non-equal segments Reaction film broken into 6 non-equal segments Larger segments near bulk liquid Larger segments near bulk liquid Smaller segments near gas-liquid interface Smaller segments near gas-liquid interface

Model Parameters Pressure Drop Pressure Drop Billet-Schultes pressure drop model to determine ∆P in packing Billet-Schultes pressure drop model to determine ∆P in packing Matched very well with data Matched very well with data ≈ 70% of measured ∆P attributed to packing ≈ 70% of measured ∆P attributed to packing Implemented into Aspen Implemented into Aspen >80% capacity factor – high vapor rates >80% capacity factor – high vapor rates Interfacial Area Interfacial Area CASTOR tests CASTOR tests a e = f(Q L, V sG, ρ G ) a e = f(Q L, V sG, ρ G ) a e ≈1.5a p a e ≈1.5a p

Model Parameters Liquid Holdup Liquid Holdup Gamma topography with 400mm transparent column Gamma topography with 400mm transparent column h L = f(μ L, V sL, ρ L, a G ) h L = f(μ L, V sL, ρ L, a G ) Gas Film MT Coefficient (k G ) Gas Film MT Coefficient (k G ) Calculated from Onda (1968) Calculated from Onda (1968) Liquid Film MT Coefficient (k L ) Liquid Film MT Coefficient (k L ) A value of 5x10 -4 m/s A value of 5x10 -4 m/s Absorber operated >80% capacity Absorber operated >80% capacity

Case 1A

Conclusions An Aspen RateSep absorber model was created using CASTOR dimensions An Aspen RateSep absorber model was created using CASTOR dimensions Improved thermodynamic, kinetic and physical property data for H 2 O-MEA-CO 2 system were implemented into the Aspen model Improved thermodynamic, kinetic and physical property data for H 2 O-MEA-CO 2 system were implemented into the Aspen model The absorber model was not adjusted to fit experimental performance. The absorber model was not adjusted to fit experimental performance. The absorber model did a very good job of predicting the temperature and CO 2 profiles of the CASTOR data The absorber model did a very good job of predicting the temperature and CO 2 profiles of the CASTOR data