Modelling of liquid-vapor-solid equilibria in the -CO 2 -H 2 O system Maria G. Lioliou, Statoil ASA
Introduction CO 2 capture by amines is state-of-the-art technology Characteristics of a desired technology Temperature stability over a temperature range High selectivity for CO 2 Non-corrosive High cycle life Low energy regeneration (stripping) Rapid kinetics for scrubbing and stripping Can ammonia processes challenge amines with respect to energy requirements reduced environmental impact Photo: Marit Hommedal, Statoil 2 -
Ammonia for CO 2 capture Used in De-NOx processes & to remove SO 2 and HCl Low environmental impact Several approaches on the thermodynamics Edwards (1978), Pawlikowski (1982), Bieling (1989), Kurz (1995), Krop (1999) Tool to predict phase behavior, mass & energy balances 3 -
Components & equilibrium reactions Vapour CO 2 CH 4, N2, O2, Ar HO 2 Soave-Redlich-Kwong EOS for gas fugacities Aqueous CO 2 HO 2 CO + H O H + HCO + 2 2 3 HCO H + CO + 2 3 3 NH H + NH + + 4 3 NH + HCO NH COO + H O 3 3 2 2 H O H + OH 2 + NH HCO NH HCO s + 4 + () 3 4 3 Henry s law Pitzer model & equilibrium constants Solubility product Solids NH () 4HCO3 s Pure solids 4 -
Thermal model Heat capacities, standard state enthalpies & entropies were used T T 0 f = f + 0 pi ()( ) H H C T dt For water: C 10 C C C T C T T 5 pc, 6 2 p( HO 2 ) = pa, + pb, + + 2 pd, 10 Aqueous & dissolved species: C pc, = + + 200 Cpi () CpA, CpB, T T Solid(s): treated as particles/species in a water stream Gas components: C = C + C id res pi () p() i p () i C = x C id id p ( m) i p () i i C = C + C T + C T + C T id 2 3 p() i pa, pb, pc, pd, residual term: calculated through the EOS 5 -
Data Model is based entirely on open literature data Approximately 3500 experimental data on: VLE SLE heat capacities, enthalpies, entropies Scattered SLE data NH 4 HCO 3(s) of primary importance Other possible salts: (NH 4 ) 2 CO 3 NH 4 CO 2 NH 4 NH 4 HCO 3 -NH 4 CO 2 NH 2 (NH 4 ) 2 CO 3 -NH 4 HCO 3 -H 2 O 6 -
Model testing -CO 2 -H 2 O system 0.20 80 Vapor pressure (bar) p ( ) 0.16 0.12 0.08 0.04 Temp=20 o C =0.49752 =0.5078 =1.02846 =0.1288 =1.56242 =2.11016 0.00 0.0 0.3 0.6 0.9 1.2 1.5 1.8 100 75 50 25 T=20 o C 0 0 2 4 6 8 10 12 =0.591 =1.087 =2.006 =4.14 =5.93 =9.03 =12.17 60 40 20 T=60 Temp=60 o C o C : 0.5-13 mole/kg H 2 O =0.721 =0.728 =0.966 =1.0285 =1.129 =2.1102 =2.186 =2.658 =3.837 =3.977 =8.14 =11.69 =11.79 0 0.0 2.5 5.0 7.5 10.0 12.5 15.0 Temp=80 o C Temp=100 o C T=80 o C T=100 o C p ( ) 80 60 40 20 NH3=1.079 NH3=3.12 NH3=9.57 NH3=11.2 NH3=14.19 0 0 2 4 6 8 10 12 CO 2 concentration (mole/kgh 2 O) CO 2 : up to 15 mole/kg H 2 O Reliable results at least up to 100 o C Data from: Pexton and Badger (1938), Van Krevelen et al. (1949), Otsuka et al. (1960), Kurz et al. (1995), Verbrugge (1979), Göppert and Maurer (1988), Müller et al. (1988), Pawlikowski et al. (1982), Mezger and Payer (1925) 7 -
Model testing NH 4 HCO 3 solubility 3.0 + NH 4 HCO 3 SR = K NH HCO sp ( ) 4 3 Supersaturation, SR 2.5 2.0 1.5 1.0 SR=1 perfect match SR>1 model predicts too low solubility SR<1 model predicts too high solubility 0.5 0 20 40 60 80 100 120 Temperature ( o C) Data from: Jänecke (1927), Jänecke (1929) 8 -
Simulation tool Excel based process simulator, thermodynamic calculations in dll 9 -
Application in a generic ammonia process Chilled Ammonia Process patented by Eli Gal (2006) Cleaned gas Air cooler Pure CO 2 Powerspan s ECO 2 technology CSIRO s ammonia PCC CAER pilot studies (Univ. of Kentucky) Absorber Heat exchanger Desorber Flue gas CO 2 rich CO 2 lean 10 -
Simulation results Cleaned gas Absorber Solven t in cleaned gas 3.5% CO 2, G: 10 kg/s, L: 200 kg/s, (NH 4 ) 2 CO 3 : 3,4,6 mol/l Flue gas CO 2 rich slip (mmole/s) 30 25 20 15 10 5 Solvent: (NH4) 2 CO 3 3 mol/l 4 mol/l 6 mol/l 0 5 10 15 20 25 30 35 40 45 Temperature in absorber ( o C) Cooling duty of flue gas 3.5% CO 2, G: 10 kg/s, L: 200 kg/s, (NH 4 ) 2 CO 3 : 4 mol/l Cooling duty of flue gas (kj/s) 2400 2000 1600 1200 800 400 Flue gas temperature 80 o C 90 o C 120 o C 0 5 10 15 20 25 30 35 40 45 Temperature in absorber ( o C) 11 -
Simulation results Energy needed to heat CO 2 rich stream Pure CO 2 T(regen): 60, 70, 80, 90 o C, T(abs): 12, 20, 35 o C Desorber Heating duty (kj/s) 250 200 150 100 T in absorber 15 o C 20 o C 35 o C CO 2 lean 50 0 60 65 70 75 80 85 90 Temperature in desorber ( o C) 12 -
Thank you Modelling of liquid- vapor- solid equilibria in the - CO 2 - H 2 O system Maria G. Lioliou TNE RD NEH CCA mlio@statoil.com, tel: +47 94424249 www.statoil.com 13 -
14 - Backup slides
Calculations Calculate thermodynamic equilibrium and other P-T constants EOS solved for gas fugacity coefficients Pitzer solved for aqueous components activity coefficients Equilibria, mass balances for CO2, NH3, H2O & alkalinity equation Update EOS and SRK with new values Typically 3-6 iterations 15 -
Parameters & simplifications in Pitzer model Similar to Bieling et al. (1995) Theoretically: 36 binary and 120 ternary parameters Interactions neglected: H + /[OH - ] << /NH 4+ and CO 2 /HCO 3- /CO 2-3 CO 2 and cannot coexist ions of the same sign ion and a neutral Ternary parameters: important only at very high concentrations Same approach for triple interactions 16 -
solubility Comparison with HYSYS x% 20 kg/s o C Gas out, T o C 3.0x10 7 Heat Flow (J/s) 2.5x10 7 2.0x10 7 1.5x10 7 1.0x10 7 5.0x10 6 0.0-5.0x10 6 HYSYS 40% Model 40% HYSYS 60% Model 60% HYSYS 80% Model 80% Water: 10 kg/s Gas: 20 kg/s Water 10 kg/s o C 50 40 Q Liquid out, T o C experimental points developed model Peng-Robinson NRTL -1.0x10 7-1.5x10 7 20 30 40 50 60 70 Temperature ( o C) solubility 30 20-9.00x10 5-1.05x10 6 10 Heat Flow (J/s) -1.20x10 6-1.35x10 6-1.50x10 6-1.65x10 6-1.80x10 6-1.95x10 6-2.10x10 6 Peng Robinson EOS model developed model Redlich Kwong 10 20 30 40 50 60 70 80 Temperature ( o C) 0 0 10 20 30 40 50 60 70 80 90 100 Temperature ( o C) Different fluid packages in HYSYS non ideal behaviour of in water Solubility of is calculated and compared to various models 17 -