Predicting the energy efficiency of a recuperative bayonet decomposition reactor for thermochemical sulfur cycles Dr. Max Gorensek, PE; Dr. Tommy Edwards Computational Sciences Directorate April 15, 2009 4th NEA Info Exchange – Nuclear H2 Production Apr 13-16, 2009 Oakbrook, IL

Outline Background Bayonet Reactor Pinch Analysis Recuperation Helium Heating Statistical Design of Model Experiments and Analyses Results Conclusions Gorensek, MB; Edwards, TB. Energy efficiency limits for a recuperative bayonet sulfuric acid decomposition reactor. submitted to I&EC Research 23 Feb 2009

SNL Bayonet Decomposition Reactor Concept All-silicon carbide construction Bayonet reactor consists of one closed ended tube co-axially aligned with an open ended tube to form two concentric flow paths Heat applied externally Liquid fed to annulus, vaporized, passed through catalyst bed Product returns through center, heats feed through recuperation Advantages include internal heat recuperation, only one connection at cool end, corrosion resistance, and low fabrication cost Silicon carbide bayonets are an off-the-shelf item (thermocouple tubes) H2SO4 in SO2, O2, H2O out catalyst High- Temperature Heat Cool base

Bayonet Decomposition Reactor Scale-up Concept Multiple parallel bayonet reactors manifolded at base Hot He applied externally with baffling to ensure counter-current flow (w.r.t. fluid being heated in annulus) Insulation layer as needed to keep base cool, prevent cross-pinch heat transfer Hot (~850°C) He from heat source Feed, product manifold Insulation Cool (~500°C) He to heat source He flow baffling Multiple parallel bayonets Catalyst beds Graphic courtesy P.S. Pickard, SNL

HTGR Heat Transfer System 50°C min ΔT between primary He and secondary He at IHX, 25°C min ΔT between secondary He and process fluid at PCHX.

Pinch Analysis of Bayonet Reactor Assume bayonet design achieves good heat transfer Reaction accomplished with practical bayonet length Reasonable temperature differences attained 25°C approach between helium heat source and annulus 10°C approach between annulus and inner tube Other assumptions Steady-state operation Plug (single-phase) or homogeneous (two-phase) fluid flow Fluid flow path cross-sections well-mixed Local thermodynamic and phase equilibrium Minimum heating target for decomposition reaction can be calculated from a pinch analysis. Heating target provides a lower limit on the energy requirement – at least this much will be needed.

Pinch Analysis Methodology Track sulfuric acid fluid element passing through bayonet with Aspen Plus* Ramp temperature from inlet to Tmax and back in increments Maintain phase and dissociation reaction equilibria Liquid-phase electrolyte dissociation Vapor-phase (H2SO4  SO3 + H2O) dissociation Add SO3 decomposition equilibrium when T ≥ Tcat, remove when T reaches Tmax Determine temperature – enthalpy relationship Prepare heating (annulus) and cooling (center-tube) curves Import into Aspen Energy Analyzer * Gorensek, M.B.; Summers, W.A. Hybrid Sulfur flowsheets using PEM electrolysis and a bayonet decomposition reactor. Int. J. Hydrogen Energy 2008, doi:10.1016/j.ijhydene.2008.06.049.

High-temperature heating target Bayonet Reactor Pinch Analysis – Recuperation High-temperature heating target 320.9 kJ/mol 80.1% H2SO4 feed, 90-bar pressure, 900°C peak process temperature, equilibrium decomposition, 10°C min ΔT Heat rejection target 94.5 kJ/mol

Bayonet Reactor Pinch Analysis – He Heating 80.1% H2SO4 feed, 90-bar pressure, 900°C peak process temperature, equilibrium decomposition, 10°C min ΔT recuperation, 25°C min ΔT He heating

Statistical Analysis – Control Variables These variables “control” the process and are of interest when the output of a computer experiment is a measure of process performance Variable Minimum Maximum ΔTmin,recup, °C 10 100 ΔTeq, °C -25 Pi, bar 90 Tcat, °C 600 740 Ti, °C 48 150 Tmax, °C 750 900 xi, mole fraction SO3 (wt% H2SO4) 0.068 (30) 0.384 (90) Min ΔT for recuperation Temp approach to equil. for SO3 decomposition Inlet pressure Catalyst bed inlet temp Inlet temperature Catalyst bed outlet temp Acid feed concentration

Statistical Analysis – Response Variables These variables are a measure of process performance and respond predictably to changes in the input or control variables in a computer experiment QH,min – minimum heating target, kJ/mol SO2 Normalized to 1 mol SO2 product Primary measure of performance Should be <450 kJ/mol SO2 (including any pre-concentration) X – fractional conversion of SO3 to SO2 Secondary performance measure Tp,He – helium pinch temperature, °C Impacts utilization of nuclear heat for decomposition QC,min – minimum cooling target, kJ/mol SO2 Measure of waste heat available for acid concentration

Statistical Analysis – Experimental Design JMP used to generate 80-point Latin Hypercube Design (LHD) Space-filling design 80 levels for each factor Minimum distance between design points maximized subject to even spacing of levels constraint for each factor

Statistical Analysis – Results Most interesting result at left QH,min strongly dependent on xi QH,min<400 kJ/mol SO2 iff xi>0.228 Lowest value of QH,min calculated was 357.6 kJ/mol SO2 at xi = 0.308 (81.4 wt% H2SO4) Attempts made to apply response surface modeling Random data subsets excluded from model, used for validation No success Neural network modeling also failed Heuristic approach adopted instead

Heuristic Analysis LHD supplemented with additional runs seeking to minimize QH,min Over 80 more combinations of control variables simulated Definitive operating envelope was established Effect of individual control variables on QH,min (within the given control variable domain) Pi and Tmax should be as high as possible (90 bar, 900°C) ΔTmin,recup and –ΔTeq should be as small as possible (10°C and 0°C) Ti and Tcat have no effect in this domain Optimal xi between 0.28 and 0.3, depending on other variables Lowest value of QH,min found within given control variable domain is 320.9 kJ/mol SO2 xi = 0.298 (80.1 wt% H2SO4)

Effect of Peak Temperature on QH,min at Various Pressures xi = 0.298, Ti = 25°C, Tcat = 675°C, –ΔTeq = 0, and ΔTmin,recup = 10°C

Effect of Pressure on QH,min at Various Peak Temperatures xi = 0.298, Ti = 25°C, Tcat = 675°C, –ΔTeq = 0, and ΔTmin,recup = 10°C

Lowest QH,min at High and Low Temperatures Ti = 25°C, Tcat = 675°C, –ΔTeq = 0, and ΔTmin,recup = 10°C

Helium Pinch at High and Low Temperatures Ti = 25°C, Tcat = 675°C, –ΔTeq = 0, and ΔTmin,recup = 10°C

He Heating Pinch Analysis – Low Temperature 77.6% H2SO4 feed, 15-bar pressure, 700°C peak process temperature, equilibrium decomposition, 10°C min ΔT recuperation, 25°C min ΔT He heating

Acid Concentration Trade-off 90-bar pressure, 900°C peak process temperature, equilibrium decomposition, 10°C min ΔT recuperation, 25°C feed temperature, 675°C catalyst bed inlet temperature Heating Target Cooling Target

Conclusions (1 of 2) Pinch analysis used in conjunction with statistical methods to establish limiting operating envelope of bayonet reactor without detailed heat transfer design To achieve lowest heat target, operate bayonet at highest possible temperature and pressure and around 80% H2SO4 feed concentration For Tmax = 900°C (ROT = 975°C), Pi = 90 bar, and 80.1% H2SO4 feed QH,min = 320.9 kJ/mol, lowest value observed Tp.He = 539.5°C, allowing 385.5°C coolant temperature change There is disincentive for feeding > 80% H2SO4, but effect of feeding < 80% H2SO4 partially offset by increase in cooling target, displacing some of heat needed for acid concentration – potential trade-off Feed and catalyst bed inlet temperatures have no effect on QH,min, but catalyst bed inlet temperature can determine helium pinch at lower peak process temperatures

Conclusions (2 of 2) For peak process temperatures above 800°C, increasing pressure lowers the heat target; below 800°C it raises the target Reactor operating temperature may be lowered 150°C (to 825°C, or Tmax = 750°C) to alleviate materials concerns, but potential efficiency advantage of sulfur cycle over direct electrolysis impacted Dropping ROT below 825°C (Tmax below 750°C) does not lead to an efficient process For Tmax = 700°C (ROT = 775°C), Pi = 15 bar, and 77.6% H2SO4 feed QH,min = 409 kJ/mol, 25% more than at 900°C QH,min increases to 461 kJ/mol at 60 bar, 469 kJ/mol at 90 bar, 46% more than at 900°C Tp.He = 631.4°C, allowing <100°C coolant temperature change