Solar reactors for thermochemical CO 2 and H 2 O splitting via metal oxide redox reactions P. Furler a, J.R. Scheffe a, D. Marxer a, A. Steinfeld a,b a Dempartment of Mechanical and Process Engineering, ETH Zurich, Switzerland b Solar Technology Laboratory, Paul Scherrer Institute, Switzerland SFERA II SUMMER SCHOOL 2014, Odeillo, France 02.07.2014 1
Outline Introduction Thermochemical fuel production 2-step thermochemical cycles Metal oxide redox pairs Solar reactor concepts (DLR, SNL, ETH) Syngas production by simultaneous splitting of H 2 O and CO 2 Reactive metal oxide structure Summary and conclusions 02.07.2014 2
Solar thermochemical fuel production Solar concentration H 2 O/CO 2 H 2 O/CO 2 dissociation CO 2 capture Q solar H 2 O H 2 + 0.5 O 2 O 2 Absorber CO 2 CO + 0.5 O 2 H 2 /CO Heliostat array Catalytic fuel synthesis nco + 2n H 2 C n H 2n + nh 2 O Liquid fuels (Diesel, Jet fuel, Ethanol ) Image: http://www.brightsourceenergy.com/ 02.07.2014 3
H 2 O / CO 2 dissociation pathways Direct solar thermolysis Solar thermochemical redox cycles H 2 O CO 2 T >> 2500 C H 2 O H 2 + 1 2 O 2 CO 2 CO + 1 2 O 2 1 st step : solar reduction MO OX MO red + O 2 2 nd step: oxidation T ~1500 C T ~900 C O 2 H 2 O MO red + H 2 O MO OX + H 2 H 2 CO 2 MO red + CO 2 MO OX + CO CO H 2 / CO / O 2 Very high temperatures High temperature gas separation or rapid quenching No separation needed: H 2 /CO generated separately from O 2 Operating temperature much less than direct gas splitting
Solar H 2 O / CO 2 splitting 2-Step thermochemical cycles MO ox MO red = metal oxide = metal or lower-valence metal oxide MO ox 1 st step : solar reduction MO ox MO red + 1 2 O 2 MO red O 2 H 2 O/CO 2 CO 2 capture recycle 2 nd step: oxidation MO red + H 2 O MO ox + H 2 MO red + CO 2 MO ox + CO CeO 2 H 2 / CO liquid fuels (diesel, jet fuel ) Solar Energy, 2005, 78, 603-615 02.07.2014 5
Common metal oxide redox pairs Volatile metal oxides Undergo gas-solid phase transition: Zinc oxide: ZnO(s) Zn(g) Tin oxide: SnO 2 (g) SnO(g) Greater oxygen exchange capability than non-volatile MO: More O 2 release / uptake => more fuel per mass of oxide Fast reaction kinetics Volatile products must be separated or rapidly quenched: not solved today in an energ. efficient fashion today J. R. Scheffe and A. Steinfeld. Oxygen exchange materials for solar thermochemical splitting of H 2 O and CO 2 : a review. Materials Today 2014, 0. Non-volatile metal oxides Remain in the solid state during reduction Non-volatile (stoichiometric): Iron oxide: Fe 3 O 4 FeO Ferrite: M x Fe 3-x O 4 xmo+(3-x)feo Hercynite Gen. form solid solutions upon reduction: i.e. Fe 3 O 4 (s) FeO-Fe 3 O 4 (s) Non-volatile (non-stoichiometric): Remain crystall. stable while lattice accommodates changes in anion or cation vacancies concentrations Ceria: CeO 2 CeO 2-δ Perovskite: ABO 3 ABO 3-δ 02/07/2014 6
Solar H 2 O / CO 2 splitting 2-Step CeO 2 cycle CeO 2 1 st step : solar reduction CeO 2 CeO 2 δ + δ 2 O 2 CeO 2 δ O 2 H 2 O/CO 2 recycle 2 nd step: oxidation CeO 2 δ + δh 2 O CeO 2 + δh 2 CeO 2 δ + δco 2 CeO 2 + δco CeO 2 H 2 / CO liquid fuels (diesel, jet fuel ) CO 2 capture Journal of Physics and Chemistry of Solids, 1975, 36, 1213-1222. 02.07.2014 7
Solar reactor technology DLR, Germany ZnFe 2 O 4 U. of CO, USA NiFe 2 O 4 Niigata U., Japan NiFe 2 O 4 H 2 N 2 + O 2 H 2 O N 2 Solar Energy, 2011, 85, 634-644 Solar Energy, 2013, 90, 68-83 JSEE, 2008, 130, 014504-1 SNL, USA CoFe 2 O 4, CeO 2 CNRS-PROMES, F CeO 2 ETH / PSI, CH CeO 2 JSEE, 2008, 130, 041001-1 Solar Energy, 2006, 80, 1611-1623 Science 2010, 330 (6012), 1797-1801
Inlet DLR Hydrosol reactor (ferrite) Quartz window T = 800 1200 C C O 2 O 2 H 2 H 2 H 2 Concentrated solar radiation O 2 Honey comb structure (SiC) Insulation Steel housing Reactive coating (Zn x Fe 3-x O 4 ) 1 st step : solar reduction Zn x Fe 3 x O 4 Zn x Fe 3 x O 4 δ + δ 2 O 2 O 2 H 2 O 2 nd step: oxidation Zn x Fe 3 x O 4 δ + δh 2 O Zn x Fe 3 x O 4 + δh 2 H 2 Materials, 2013, 6, 421-436. JSSE, 2006, 128, 125-133. Int.J.Hydrogen Energy, 2009, 34, 4537-4545. 02.07.2014 9
DLR Hydrosol reactor Experimental results Experimental conditions: Reduction Oxidation Power input: > 45 kw < 45 kw Gas flows: N 2 : 15Nm 3 h -1 N 2 : 15 Nm 3 h -1 H 2 O: 3.5 kg h -1 Accurate temperature control by adjusting P solar Dual-cavity concept worked Successfully produced H 2 But decreasing rates and yield over time indicate instable material Solar Energy, 2011, 85, 634-644. 02.07.2014 10
DLR Hydrosol reactor Simple and scalable concept No moving / rotating parts No gas phase reactions Quasi continuous product stream because of dual-cavity approach Good control of gas flows across honey comb structure Solar engineering concept workes Accurate temperature control by P solar Volumetric absorption of concentrated radiation because of honey comb structure Low mass loading of active material < 10% because ferrites are coated on inert honey combs Inert material has to be heated in every cycle => limits efficiency The more active material the more fuel can be produced per cycle Side reactions with supporting structure Active ferrite coating is deactivated over time Sensible heat recuperation from solid not implemented 02.07.2014 11
Sandia NL CR5 reactor T = 1550 C CeO 2 1 st step : solar reduction O 2 CeO 2 CeO 2 δ + δ 2 O 2 CeO 2 δ T = 1100 C H 2 O/CO 2 2 nd step: oxidation CeO 2 δ + δh 2 O CeO 2 + δh 2 CeO 2 δ + δco 2 CeO 2 + δco H 2 CO JSEE, 2008, 130, 041001-1 recycle 02.07.2014 12
Sandia NL CR5 reactor Heat recuperation concept Each Ring rotates in opposite direction to its neighbors Recuperation of sensible heat by thermal radiation At high T radiation is dominating mode of heat transfer (ε σ T 4 ) Heat recuperation is important for achieving high efficiencies R. Hogan et al.. ASME Proceedings of the 6 th International Conference on Energy Sustainability, San Diego, USA, 2012. 02.07.2014 13
CR5 reactor Experimental results Experimental conditions: Reduction Oxidation Power input: 2-9 kw Gas flows: Ar: 20 L min -1 CO 2 : 20 L min -1 Reactor was successfully operated on sun for several hours Good temperature control Problems with mechanical stability of rotating rings Stoichiometric fuel production (CO:O 2 = 2) as expected Sandia Report SAND2012-5658, 2012. 02.07.2014 14
CR5 reactor Sensible heat recuperation inherently included Important aspect for achieving high efficiencies Continuous process Reduction and oxidation proceed simultaneously Hot and cold zones are maintained at constant temperatures Reduces energy requirements for heating of reactor components Accurate temperature control by adjustment of the RPM Low mass loading of active material Only fraction of the rings consists of active redox material Inert material of the rings need to be heated in every cycle => limits efficiency High optical thickness of CeO 2 structure Limits heat transfer => limits fuel production and efficiency Rotating high temperature ceramic parts Mechanical failure of rotating rings was observed and caused major problems Complex gas flow control Cross-flow of O 2 from the reduction zone to the oxidation zone Fraction of CeO 2-δ is oxidized with O 2 High gas flows needed => low product gas concentration 02.07.2014 15
ETH Solar Reactor Porous CeO 2 Al2O3-SiO2 insulation Reactor 1. 2. Thermal Oxidation Reduction: ΔH CeO 2 CeO2 δ + δ 2 δ + δh 2 O H CeO 2 O 2 2 + δh 2 Reactor Front (water-cooled) CeO 2 δ + δco 2 H CeO2 + δco X 1000 100 μm Quartz Window X 1000 Syngas (H 2 / CO) 100 μm O 2 CO 1500 900 C O H 2 2 H2 O 2 Concentrated Solar Radiation 10 mm Inconel Shell Purge H 2 O / Gas CO 2 (Ar) Science 2010, 330 (6012), 1797-1801 Energy Environ. Sci. 2012, 5 (3), 6098-6103 Energy Fuels 2012, 26 (11), 7051-7059 02.07.2014 16
Experimental setup Energy Environ. Sci. 2012, 5 (3), 6098-6103 02.07.2014 17
Syngas production by simultaneous splitting of CO 2 and H 2 O 02.07.2014 18
Experimental results Simultaneous H 2 O / CO 2 splitting (H 2 O:CO 2 = 5.7) Experimental conditions: Reduction Oxidation Power input: 3.6 kw 0.8 kw Gas flows: Ar: 2 l min -1 Ar: 2 l min -1 Ceria structure: Ceria Felt Total Mass: 127 g CO 2 : 0.375 l min -1 H 2 O: 2.15 l min -1 Stoichiometric fuel production as expected (O 2 :fuel = 1:2) (H 2 O H 2 + 0.5 O 2 ) (CO 2 CO + 0.5 O 2 ) No gas phase hydrocarbons detected No C depositions No impurities detected which could damage catalyst of FT unit Energy Environ. Sci. 2012, 5 (3), 6098-6103 02.07.2014 19
Experimental results Syngas composition (H 2 :CO) vs. co-feeding ratio (H 2 O:CO 2 ) Experimental conditions: Reduction Oxidation Power input: 3.6 kw 0.8 kw Gas flows: Ar: 2 l min -1 Ar: 2 l min -1 CO 2 : 0.29-1.39 l min -1 H 2 O: 1.11-2.21 l min -1 Ceria structure: Ceria felt Total mass: 127 g Syngas composition (H 2 :CO) adjustable in a wide range (0.25-2.34) by controlling the co-feeding ration (H 2 O:CO 2 ) Syngas composition increases linearly with increasing H 2 O:CO 2 ratios e.g. co-feeding with H 2 O:CO 2 = 6.5 yields syngas with H 2 :CO = 2 Syngas suitable for FT synthesis Energy Environ. Sci. 2012, 5 (3), 6098-6103 02.07.2014 20
Experimental results Consecutive H 2 O / CO 2 splitting (H 2 O:CO 2 = 6.7) Experimental conditions: Reduction Oxidation Power input: 3.6 kw 0.7 kw Gas flows: Ar: 2 l min -1 Ar: 2 l min -1 Ceria structure: Ceria Felt Total Mass: 127 g CO 2 : 0.33 l min -1 H 2 O: 2.20 l min -1 Energy Environ. Sci. 2012, 5 (3), 6098-6103 Philipp Furler 02.07.2014 21
Solar-to-fuel energy conversion efficiency η solar to fuel = ΔH fuel r fuel dt P solar dt + E inert r inert dt Direct impact on fuel production cost Reactive Structure Heat and mass transfer properties η solar-to-fuel 0.15 20 % Reactive Material Doped CeO 2 Other metal oxides Perovskites Reactor technology Geometry and scale-up Operating conditions Heat recovery 22
Reactive CeO 2 structures 02.07.2014 23
Structural improvements of Ceria x300 300 μm High SSA rapid oxidation kinetics but low specific mass HT limited during reduction step Limitations: HT limited during reduction pores in mm-range vol. absorption fast heating rates high specific mass high mass loading 02.07.2014 24 Science 2010, 330 (6012), 1797-1801 Energy Environ. Sci. 2012, 5 (3), 6098-6103 Energy Fuels 2012, 26 (11), 7051-7059
CeO 2 Structure CeO 2 RPC (Reticulated Porous Ceramic) Macro-porous structure enables penetration and absorption of concentrated solar radiation Faster heating rates Temperature gradients minimized High mass loading within reactor (~1500 g) CeO 2 laminate CeO 2 RPC Energy Fuels 2012, 26 (11), 7051-7059 Al 2 O 3 -SiO 2 insulation Inconel shell 02.07.2014 25
Comparison felt vs. RPC Experimental Conditions Reduction Oxidation Power Input: 3.6 kw 0.7 kw CeO 2 felt Gas Flow: 2 l min -1 Ar 3 l min -1 CO 2(g) CeO 2 RPC Ceria Structure: Ceria RPC: 1416 g, Ceria felt; 90 g Rate (ml min -1-1 g -1-1 CeO Rate (ml min -1 2 ) ) 0,20 200 0,15 150 0,10 100 0,05 50 0,00 0 O 2 evolution CO CO evolution O 22 evolution CeO 22 felt O 22 evolution CeO 22 RPC - - - - Heating Rate CeO2 felt felt - - - - Heating Rate CeO2 RPC 500 ----- ----- O2 x O2 1000 evolution 100 CeO2 μm felt felt ----- ----- O2 O2 evolution CeO2 RPC 2,5 x 1500 50 μm Rate Rate (ml min (ml -1-1 min g -1-1 CeO -1 ) 2 ) 400 2,0 300 1,5 200 1,0 O : -1 : -1 2 Felt : 0.24 l CO CO Felt : 0.49 l O 2 2 : -1 : -1 2 Felt RPC : 2.66 : 4.00 ml l g -1 Felt : 5.39 ml g -1 100 0,5 CO RPC : 8.27 l CO O 2 RPC : 2.83 ml g -1 RPC : 5.86 ml g -1 0,0 0 5 5 10 15 20 25 30 35 40 50 60 70 80 90 3,0 Time (min) Time (min) CO evolution CeO 22 felt CO evolution CeO 22 RPC x 1000 100 μm x 1500 50 μm Overall 17 times higher fuel production per cycle with RPC structure Energy Fuels 2012, 26 (11), 7051-7059 02.07.2014 26
Experimental results Conditions Reduction Oxidation Power Input: 2.8-3.8 kw 0 kw Gas Flow: 2 l min -1 Ar 2 l min -1 CO 2 Mass loading: 1416 g 0,6 reduction oxidation 2000 Rate (ml min -1 g -1 CeO 2 ) 0,5 0,4 0,3 0,2 0,1 T ceria O 2 : 2.76 ml g -1 2.01 ml g -1 1.06 ml g -1 CO: 5.69 ml g -1 4.12 ml g -1 1.47 ml g -1 3.8 kw 3.4 kw 2.8 kw 1500 1000 500 Temperature ( C) 0,0 30 40 50 60 70 80 90 0 Time (min) Energy Fuels 2012, 26 (11), 7051-7059
Experimental results Solar-to-fuel conversion efficiency solar-to-fuel (%) 1.8 1.6 1.4 1.2 1.0 0.8 0.6 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 η RPC = 1.7 % t Reduction T Ceria P solar (kw) solar-to-fuel => 12x η Felt (0.15 %) => 4 x η bricks (0.4%) η average = 24 22 20 18 16 14 Reduction step duration (min) ΔH fuel r fuel dt P solar dt + E inert r inert dt 1700 1600 1500 1400 Nominal reactor temperature ( C) Improved heat-transfer Improved specific mass Energy Fuels 2012, 26 (11), 7051-7059 02.07.2014 28
Structural improvements of Ceria x300 300 μm High SSA rapid oxidation kinetics but low specific mass HT limited during reduction step Limitations: 1) HT limited during reduction pores in mm-range vol. absorption fast heating 2) SSA limited during oxidation pores in the µm-range increase SSA fast oxidation Science 2010, 330 (6012), 1797-1801 Energy Environ. Sci. 2012, 5 (3), 6098-6103 Energy Fuels 2012, 26 (11), 7051-7059 02.07.2014 29 Phys. Chem. Chem. Phys. 2014, 16 10503
SEM micrographs single-scale RPC vs. multi-scale RPC 250 µm 250 µm Phys. Chem. Chem. Phys. 2014, 16 10503 02.07.2014 30
Experimental results Solar reactor experiment Conditions Reduction Oxidation Power Input: 3.8 kw 0 kw Gas Flow: Mass loading: 2 l min -1 Ar (single) 6 l min -1 Ar (dual) single: 1416 g, dual: 998 g 2 l min -1 CO 2 (single) 6 l min -1 CO 2 (dual) 1,4 reduction oxidation 2000 Rate (ml min -1 g -1 CeO 2 ) 1,2 1,0 0,8 0,6 0,4 0,2 T Ceria O 2,single : 2.3 ml g -1 O 2,dual : 2.52 ml g -1 single-scale RPC dual-scale RPC CO single : 4.7 ml g -1 CO dual : 5.02 ml g -1 1800 1600 1400 1200 1000 800 600 Temperature (K) Energy Fuels 2012, 26 (11), 7051-7059 Phys. Chem. Chem. Phys. 2014, 16 10503 0,0 400 10 20 30 40 Time (min) 02.07.2014 31
1 g -1 CeO2 ) C) %) C) Peak rate (ml min -1 g -1 CeO2 ) Temperature ( C) Experimental results 290 redox cycles Conditions Reduction Oxidation Power Input: 3.8 kw..0 kw Gas Flow: 6 l min -1 Ar 6 l min -1 CO 2.. L min -1 H 2 O Mass loading: 998 g (a) 1.4 (I) (II) (III) (IV) 1600 1.2 1550 1.0 0.8 0.6 0.4 CO rate H 2 rate Temperature 1500 1450 1400 1350 1300 0.2 1250 0.0 0 50 100 150 200 250 1200 300 Cycle number (-) (b) (c) 2.0 3.8kW 3.8kW 2.8kW 1600 40 H 2 1500 Submitted to Energy Environ. Sci. June 2014 1.5 1550 1500 30 T 1250 02.07.2014 32
ETH reactor Simple and scalable concept No moving / rotating parts No gas phase reactions Cavity receiver ensures high solar absorption efficiency 90% of incoming solar radiation is absorbed Ceria RPC structure offers advantageous properties Volumetric absorption of concentrated radiation because mm-sized pores -> high heating rates during solar reduction SSA because of µm-sized pores -> fast oxidation reaction kinetics High mass loading of active material and no side reactions Stable syngas production over time (no side reactions) Not a continuous process Sensible heat recuperation from solid not implemented 02.07.2014 33
Summary and conclusions Important aspects of solar reactors Simple and scalable design No moving / rotating high temperature ceramic parts Efficient absorption of concentrated radiation Cavity receiver-type concepts Fast and uniform heating of metal oxide Requires metal oxide structure which can volumetrically absorb radiation Foam type or honey comb structures Particles High mass loading of redox active material High fuel production and η solar-to-fuel Stable metal oxide material Side reactions or strong sintering and loss of activity are not permitted Potential for reaching high η solar-to-fuel Heat recuperation from the solids and gases should be possible 02.07.2014 34
Summary and conclusions Experiments ETH reactor Syngas production by simultaneous splitting of H 2 O and CO 2 Syngas composition can be controlled by the H 2 O/CO 2 co-feeding radio Syngas is suitable for FT-synthesis (composition, purity) Heat transfer limitation during thermal reduction: RPC structure with pores in the mm-range Volumetric absorption => fast / homogenous heating Higher specific mass => higher ceria mass loading Surface limitation during oxidation: Dual-scale RPC with pores in the mm-range and µm-range Increased SSA => fast oxidation kinetics RPC structure allowed to enhance: the fuel production per unit time η solar-to-fuel = 1.7 % 02.07.2014 35
Acknowledgments PREC ETH & STL PSI EMPA: Dr. Ulrich Vogt, Michal Gorbar, Alexander Bonk EU-SOLAR-JET European Commission Contract No. 285098 CCEM EU-ERC Advanced Grant SUNFUELS No. 320541
Thank you! Institute of Energy Technology Philipp Furler Dr. sc. ETH ETH Zurich Sonneggstrasse 3 8092 Zurich Switzerland phone: +41 44 633 9378 furlerp@ethz.ch www.pre.ethz.ch