Comparison of commercial and new developed adsorbent materials for pre-combustion CO 2 capture by pressure swing adsorption

Transcription

1 Comparison of commercial and new developed adsorbent materials for pre-combustion CO 2 capture by pressure swing adsorption Johanna Schell, Nathalie Casas, Lisa Joss, Marco Mazzotti - ETH Zurich, Switzerland Richard Blom, SINTEF Materials and Chemistry, Oslo, Norway TCCS-6 Trondheim, Norway

2 CO 2 capture in an IGCC power plant Coal Gasifier Shift CO 2, H 2 CO 2 / H 2 separation CO 2 CO 2 compression CO 2 O 2 H 2 Air ASU air Gas turbine Steam cycle electricity TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 2

3 CO 2 capture in an IGCC power plant Pressure 35 bar CO 2 fraction 40% H 2 fraction 60% CO 2 / H 2 separation CO 2 to storage: CO 2 purity > 95% Capture > 90% Pressure 110 bar H 2 to gas turbine: H 2 purity 94% Pressure 25 bar CO 2 / H 2 separation by PSA promising TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 3

4 PSA cycle possibilities Classical Skarstorm cycle used to produce high purity high p product (e.g. H 2 ) consists of 4 basic steps: Pressurization with Feed Adsorption Low pressure product Countercurrent blowdown Purge with high pressure product In more advaced cycles more steps are applied, e.g.: Pressure equalization HP LP LP LP HP HP Purge with other than product Cocurrent blowdown Steps can be combined in multiple ways High pressure product Complex process design TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 4

5 Approach Materials Commercial and new materials TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 5

6 Approach Materials Commercial and new materials Static experiments Rubotherm MSB TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 6

7 Approach Materials Commercial and new materials Static experiments Rubotherm MSB Dynamic experiments Two column PSA setup PI 110 cm 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 7

8 Approach Materials Commercial and new materials Process modeling Mass, energy and momentum balances, Isotherms & EOS u T w voids adsorbent ε (1 ε ) c i n i P T T s Static experiments Rubotherm MSB Dynamic experiments Two column PSA setup PI 110 cm 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 8

9 Approach Materials Commercial and new materials Process modeling Mass, energy and momentum balances, Isotherms & EOS u T w voids adsorbent ε (1 ε ) c i n i P T T s Static experiments Rubotherm MSB Model-based Process Design Dynamic experiments Two column PSA setup 110 cm PI 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 9

10 Approach Materials Commercial and new materials Process modeling Mass, energy and momentum balances, Isotherms & EOS u T w voids adsorbent ε (1 ε ) c i n i P T T s Static experiments Rubotherm MSB Model-based Process Design Dynamic experiments Two column PSA setup 110 cm PI 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 10

11 Adsorbent materials Commercial material Activated carbon AC AP3-60, Chemviron, Germany New adsorbent material (SINTEF) USO-2-Ni MOF (Ni 2 (1.4-bdc) 2 (dabco) 4DMF 0.5H 2 O) Mesoporous silica MCM-41 TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 11

12 Adsorbent materials Commercial material Activated carbon AC AP3-60, Chemviron, Germany New adsorbent material (SINTEF) USO-2-Ni MOF (Ni 2 (1.4-bdc) 2 (dabco) 4DMF 0.5H 2 O) Mesoporous silica MCM-41 New materials synthesized as powder For process application: formulation as pellets crucial No well established method 4 Different formulation methods investigated Particle size: mesh TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 12

13 Approach Materials Commercial and new materials Process modeling Mass, energy and momentum balances, Isotherms & EOS u T w voids adsorbent ε (1 ε ) c i n i P T T s Static experiments Rubotherm MSB Model-based Process Design Dynamic experiments Two column PSA setup 110 cm PI 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 13

14 Equilibrium measurements Isotherm measurements using Rubotherm MSB CO 2 H 2 N 2 Mix AC T C C C 25 C p bar bar bar bar MOF T C C - - p bar bar - - MCM-41 T C In progress - - p bar - - Exp. excess Isotherms TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 14

15 Equilibrium measurements Isotherm measurements using Rubotherm MSB CO 2 H 2 N 2 Mix AC T C C C 25 C p bar bar bar bar MOF T C C - - p bar bar - - MCM-41 T C In progress - - p bar - - Exp. excess Isotherms Excess Absolut Mathematical description of adsorption equilibrium Selection Isotherm eq. Parameter Fitting TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 15

16 AC: pure component isotherms H 2 N 2 CO 2 AC AC AC TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 16

17 AC: pure component isotherms H 2 N 2 CO 2 AC AC AC CO 2 25 C 45 C N 2 25 C 45 C 45 C 25 C H 2 Isotherm parameters used for: Absolute adsorption Heat of adsorption (Clausius Clapeyron) Prediction of binary adsorption, e.g.: o Empirical binary isotherm (e.g. Langmuir) o IAST Cyclic capacity evaluation of suitability TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 17

18 Comparison: pure component isotherms CO 2 isotherms at 25 C MOF MCM-41 AC TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 18

19 Comparison: pure component isotherms CO 2 isotherms at 25 C Cyclic capacity p Ads = 1.5 MPa, p Des varying MOF MCM-41 MOF AC AC MCM-41 Zeolite [1] [1] Xiao et al., Adsorption 14 (2008) TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 19

20 Approach Materials Commercial and new materials Process modeling Mass, energy and momentum balances, Isotherms & EOS u T w voids adsorbent ε (1 ε ) c i n i P T T s Static experiments Rubotherm MSB Model-based Process Design Dynamic experiments Two column PSA setup 110 cm PI 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 20

21 Model equations of an adsorption column 1. Mass balances species i ( uc ) ε c i ni 1 i ci + ν + D = Li ε ε 0 t t z z z Accumulation Convection n i t 2. Energy balances ( ut ) n T T ν s 1 n j 2hL ε KL T + νγ + ( Hj ) + ( T Tw ) = ε = ε ε 0 t t z C j 1 t r C C z z Accumulation ( ) = ka n n i p i i Convection g Dispersion T s 1 n nj ha = + t C t C Acc s s p ( Hj ) ( T Ts ) j = 1 Heat of adsorption Linear driving force s Exchange gas - solid Heat of adsorption Exchange wall i 3. Constitutive equations g u T w voids adsorbent 1. Non linear adsorption isotherm: * n i = n pt yi i (,, ) 2. Equation of State: Ideal gas law 3. Pressure: Ergun equation g Dispersion ε (1 ε ) c i n i P T T s TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 21

22 Model equations of an adsorption column 1. Mass balances species i ( uc ) ε c i ni 1 i ci + ν + D = Li ε ε 0 t t z z z u T w voids adsorbent ε (1 ε ) c i n i P T T s n i t 2. Energy balances Linear driving force T s 1 n nj ha = + t C t C s ( ) = ka n n i p i i s p ( Hj ) ( T Ts ) j = 1 s 3. Constitutive equations 1. Non linear adsorption isotherm: * n i = n pt yi i (,, ) 2. Equation of State: Ideal gas law 3. Pressure: Ergun equation ( ut ) n T T ν s 1 n j 2hL ε KL T + νγ + ( Hj ) + ( T Tw ) = ε = ε ε 0 t t z C j 1 t r C C z z g i g g TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 22

23 Approach Materials Commercial and new materials Process modeling Mass, energy and momentum balances, Isotherms & EOS u T w voids adsorbent ε (1 ε ) c i n i P T T s Static experiments Rubotherm MSB Model-based Process Design Dynamic experiments Two column PSA setup 110 cm PI 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 23

24 Experimental setup: 2-column Lab PSA Breakthrough experiments PSA cycles including p equalization Fully automated Premixed gases Column insulation and heating p and T measurements Product streams online analyzed by MS TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 24

25 Breakthrough and PSA experiments 10 cm Breakthrough experiments Fit the missing model parameters T = 25 C p = 15 bar H 2 CO 2 40 cm 60 cm 85 cm 110 cm TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 25

26 Breakthrough and PSA experiments 10 cm Breakthrough experiments Fit the missing model parameters T = 25 C p = 15 bar H 2 CO 2 40 cm 60 cm 85 cm 110 cm PSA experiments Validate the PSA simulation tool Exp. testing of full PSA cycles H2 rich product H 2 CO 2 CO 2 rich product CO 2 H 2 TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 26

27 Approach Materials Commercial and new materials Process modeling Mass, energy and momentum balances, Isotherms & EOS u T w voids adsorbent ε (1 ε ) c i n i P T T s Static experiments Rubotherm MSB Model-based Process Design Dynamic experiments Two column PSA setup 110 cm PI 85 cm 60 cm Column C 40 cm 10 cm PI TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 27

28 Process Design criteria Feed H 2 to gas turbine CO 2 / H 2 separation CO 2 to storage: CO 2 purity > 95% Capture > 90% Pressure 110 bar Specifications & boundary conditions: CO 2 purity and capture rate purge with the feed co-current blowdown Minimize energy penalty (compression costs): no repressurization increase CO 2 desorption pressure Investments cost: max. 3 p-equalization steps TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 28

29 Process Design criteria Feed H 2 to gas turbine CO 2 / H 2 separation CO 2 to storage: CO 2 purity > 95% Capture > 90% Pressure 110 bar Specifications & boundary conditions: CO 2 purity and capture rate purge with the feed co-current blowdown Minimize energy penalty (compression costs): no repressurization increase CO 2 desorption pressure Investments cost: max. 3 p-equalization steps For fixed material, T and p Des : process performance dependent on t ads, t blow & t purge TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 29

30 Influence of the adsorption step time (AC) Change of time of the adsorption step Time of blowdown and purge step already optimized CO 2 purity CO 2 capture rate T = 308 K p Des = 1 bar TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 30

31 Influence of the adsorption step time (AC) Change of time of the adsorption step Time of blowdown and purge step already optimized CO 2 purity Trade-off Better shown as pareto front CO 2 capture rate T = 308 K p Des = 1 bar T = 308 K p Des = 1 bar TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 31

32 AC comparison: process conditions Different desorption pressures T = 308 K p Des = 1 bar p Des = 2 bar TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 32

33 AC comparison: process conditions Different desorption pressures T = 308 K Different process temperatures p Des = 1 bar p Des = 1 bar T = 308 K T = 338 K p Des = 2 bar TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 33

34 Comparison: AC MOF MOF with real physical properties compared to theoretical MOF T = 308 K, p Des = 1 bar theoretical MOF real MOF TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 34

35 Comparison: AC MOF MOF with real physical properties compared to theoretical MOF T = 308 K, p Des = 1 bar Theoretical MOF compared to AC T = 308 K, p Des = 1 bar theoretical MOF theoretical MOF real MOF AC TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 35

36 Conclusions PSA for pre-combustion very promising because of boundary conditions and process specifications Model-based process design beneficial due to various process configurations Model parameters have to be determined in a reliable way Pareto front to compare different process conditions and materials MOF shows promising behavior however material formulation very important for process performance TCCS-6 Trondheim Schell, Casas, Joss, Blom, Mazzotti 36