Neutronics of Prismatic Fluoride Salt Cooled High Temperature Reactors

Transcription

1 Neutronics of Prismatic Fluoride Salt Cooled High Temperature Reactors Workshop on Advanced Reactors Concepts PHYSOR 2012 Knoxville, TN April 15, 2012 Dan Ilas

2 Main Neutronic Design Characteristics of AHTR AHTR : Advanced High Temperature Reactor Plate type of fuel in a hexagonal prismatic assembly design Power (thermal): 3,400 MW Coolant: FLiBe (99.995% 7 Li enrichment) Coolant inlet temperature: 650 C Coolant outlet temperature: 700 C Moderator: Graphite Fuel: U-C-O in TRISO (AGR-2, AGR-5/6) Fuel elements: Slabs Fuel assembly: Hexagonal, in channel box Volumetric power density: ~13 MW/m 3 Cycle length: >2 years (19.75% enrichment) 2 Managed by UT-Battelle

3 Constraints and Design Decisions Can regard the AHTR neutronic design as a constrained optimization problem. Constraint Consequence Power (3,400 MWth) Size of reactor High temperature operation Choice of structure materials, coolant Low fuel temperature Power density (<13MW/m 3 ), striped fuel configuration Use of realistic, qualified fuel Fuel kernel size (AGR-2 type of fuel), packing fraction upper limit (<40%) Use of low fuel enrichment Upper limit of fuel enrichment < 20% Use of realistic graphite materials Density of C-C composite below 2g/cm 3 Achieve high burnup, long core life Online refueling, multi-batch Coolant flow velocity < 3m/s Size and arrangement of fuel plates (spacing between fuel plates) Large Reynolds number Size and arrangement of fuel plates (spacing between fuel plates) Low coolant pressure drop (<1.5 atm) Size and arrangement of fuel plates (spacing between fuel plates) Mechanical refueling necessities Spacing between fuel assemblies 3 Managed by UT-Battelle

4 Constraints and Design Decisions TRISO fuel design Grain design : AGR-2 (being qualified) fuel design used as prototype Packing fraction : 40% for AHTR base design (AGR-2: <40%) Fuel enrichment : 19.75% for AHTR base design (AGR-2: 14%) Outer pyrolitic graphite, 40 μm thick UCO kernel, d=427 μm Graphite buffer, 100 μm thick Silicon Carbide coating, 35 μm thick Inner pyrolitic graphite, 35 μm thick 4 Managed by UT-Battelle

5 Constraints and Design Decisions Fuel plates Dimensions of fuel plates <= input from mechanics, materials Annular (stripes) fuel distribution in the plates: Low fuel temperature (<1000 C maximum fuel temperature) Higher burnup achievable (~80MWd/MTHM for once-through) Fuel stripes Central graphite meat 5 Managed by UT-Battelle

6 Constraints and Design Decisions Fuel assembly Hexagonal, parallel arrangement of 6 plates in each third of the hexagon Distance between plates <= input from thermal hydraulics Y-shape structure (C-C) at center Channel box holder (C-C) 4-cm wide C-C Y-shape 1-cm thick channel box (C-C) Control blade Fuel plates Intra-assembly coolant Inter-assembly coolant 6 Managed by UT-Battelle

7 Constraints and Design Decisions Arrangement of fuel assemblies in the core 253 assemblies separated by interstitial coolant Inner reflector? : No. Use uniform distribution of assemblies Thickness of radial reflector : ~0.7m Fuelled region Removable reflector blocks Permanent reflector Core barrel Coolant downcomer Reactor vessel 7 Managed by UT-Battelle

8 Constraints and Design Decisions Design of control rods Materials : MHC Shape : Blade Placement : In Y-shaped C-C structure Number : 253 (each fuel assembly) Burnable poison Grains of Europium Oxide in fuel plates => reduces reactivity swing 8 Managed by UT-Battelle

9 Computational Tools Use the in-house TRITON module of SCALE 6.1 for depletion calculations and lifetime assessments Use the KENO-VI (continuous energy) module of SCALE for confirmatory calculations Advantage: Easy to use, easy to ask developers Issue: SCALE (multigroup) cannot handle double heterogeneity for plates and annular geometry (either cylinders or plates) => approximations needed 9 Managed by UT-Battelle

10 Computational Tools Large errors in lifetime with straightforward methods! explicit grains homogenized explicit grains with Dancoff correction keff Time (years) No experimental depletion results to benchmark against! =>Dancoff correction thought to produce good results. Tested on NGNP prismatic design with very good agreement vs. reference (also computational) results Very long computational time (weeks of runtime) 10 Managed by UT-Battelle

11 Computational Tools Dancoff factor depletion procedure (SCALE) compared to VESTA continuous energy (R. Kelly, D. Ilas, Verification of a Depletion Method in SCALE for the Advanced High Temperature Reactor, PHYSOR 2012) VESTA 1.2 TRITON k-eff Managed by UT-Battelle Time (days)

12 End of FY2011 Reference Model D. Holcomb, D. Ilas, V. Varma, A. Cisneros, R. Kelly, J. Gehin, Core and Refueling Design Studies for the Advanced High Temperature Reactor, ORNL report (ORNL/TM-2011/365) Parameter Value Units Power (thermal) 3,400 MW Average power density per grain 41 mw/particle Volumetric power density 12.9 MW/m 3 Mass of heavy metal 32.9 MT Fuel enrichment % Mass of fissile 6.5 MT Inlet coolant temperature 650 C Outlet coolant temperature 700 C Maximum fuel temperature (average assembly) 854 C Coolant flow velocity (intra-assembly flow only) 1.93 Reynolds number (intra-assembly flow only) 7,200 - m/s 12 Managed by UT-Battelle

13 End of FY2011 Reference Model Core lifetime (one batch) Cycle length (years) Cycle length Burnup CHM Burnup (GWd/MTHM) 13 Managed by UT-Battelle

14 End of FY2011 Reference Model Core lifetime (one batch) enrichment dependent % 5% 10% 15% % 5% 10% 15% keff keff Time (years) Burnup (MWd/kg) 14 Managed by UT-Battelle

15 End of FY2011 Reference Model Other calculations (D. Ilas, D. Holcomb, V. Varma Advanced High-temperature Reactor Neutronic Core Design, PHYSOR 2012) Isothermal reactivity coefficient Negative for all configurations Becomes more positive for higher Carbon-to-Heavy Metal (CHM) ratios Becomes more negative with burnup Void reactivity coefficient Becomes positive above CHM~450 for 19.75% enrichment Fast fluence distribution Within accepted graphite irradiation limits 15 Managed by UT-Battelle

16 End of FY2011 Reference Model Other calculations (D. Ilas, D. Holcomb, V. Varma Advanced Hightemperature Reactor Neutronic Core Design, PHYSOR 2012) Power (thermal flux) distribution Flattens with burnup Control blades Provide both reactor control and primary shut-down mechanism Reactivity worth ~ 10%Δk/k for the chosen design Secondary shut-down system Injection of EuF 3 into FLiBe coolant Burnable poison Eu 2 O 3 grains Preliminary analysis 16 Managed by UT-Battelle

17 FY2011 Refueling Analyses Procedures described in: A. T. Cisneros, D. Ilas Neutronics and Depletion Methods for Parametric Studies of Fluoride-Salt-Cooled High-Temperature Reactors with Slab Fuel Geometry and Multi- Batch Fuel Management Schemes, PHYSOR 2012) Two possibilities analyzed: Offline refueling (2-3 day outage, every 6 months) Pseudo-online (frequent refueling 2-3 weeks, short outages) Pseudo-online impacted by Xe accumulation: ρ (pcm) % 10% 15% 19.75% 17 Managed by UT-Battelle Time after shutdown (hours)

18 FY2012 Reference Model AGR-5/6 fuel, 9% enriched, 40% packing fraction, CHM=400 Lifetime for this design: ~9 months (once-through) => 6 months refueling interval (2-batch - checkerboard) => 12 months residence time for 2-batch (Linear Reactivity Model prediction) => ~73 GWd/MTIHM discharge burnup reference core, AGR-2 (CHM~200) 2011 reference core, AGR-5 (CHM~205) 2011 reference core, but 9% enrichment (CHM~205) 2012 reference, AGR-5: enrichment 9%, CHM= reference LRM-anticipated 2-batch keff Managed by UT-Battelle Time (Years)

19 FY2012 Reference Model Isothermal Temperature Coefficient Evaluations for fresh initial core More negative with lower enrichment More positive (but still negative) with higher CHM More negative with burnable poison (Eu) enrich. 9%, CHM~205 enrich. 9%, CHM=400 enrich. 9%, CHM=400, with Eu 2011 reference: enrich %, CHM~200-4 ρ/ T (pcm/k) Managed by UT-Battelle T (⁰C)

20 FY2012 Reference Model Void Coefficient Evaluations for fresh initial core (α=void fraction) Positive for CHM=400, 9% enrichment Stays positive for End-Of-Cycle (EOC) and Beginning-Of- Equilibrium-Cycle (BOEC) without Eu as BP Preliminary estimations show it can become negative for BOEC with Eu as BP Conclusion: BP should be part of the design Average: pcm/% ρ/ α (pcm/%) Managed by UT-Battelle 10 0 Average: +7.9 pcm/% Average: pcm/% FY 2011 reference average: pcm/% 9% enrichment, CHM=200, BOC 9% enrichment, CHM=400, BOC 9% enrichment, CHM=400, BOC, with Eu Void Fraction (%)

21 FY2012 Reference Model Burnable Poison BP as Eu 2 O 3 particles at the center of the fuel plate Eu 2 O 3 chosen on chemical compatibility criteria Parameters to monitor: Maximum reactivity Initial reactivity With BP With No BP Reactivity swing keff Cycle length penalty Managed by UT-Battelle Reactivity penalty Time (Years)

22 FY2012 Reference Model Burnable Poison The Dancoff factor method used with multigroup TRITON No benchmarking against experiments or comparison to other codes yet Self shielding an important factor for accuracy (can lead to more than 1% differences in reactivity penalty) A compromise must be reached: Minimize Initial and Maximum Reactivities, Reactivity Swing, Cycle Penalties Preserve a reasonable reserve of reactivity (especially after Xe buildup) Different BP particle sizes and BP loadings considered 22 Managed by UT-Battelle

23 FY2012 Reference Model Burnable Poison Different BP particle sizes (R=150, 250, 350 μm) and BP loadings ( kg of Eu per core) considered %, AGR-5, rings, 8x250μm 9%, AGR-5, rings, 10x250μm 9%, AGR-5, rings, 12x250μm 9%, AGR-5, rings, 47x150μm 9% enric, CHM=400, 40% PF, AGR-5, rings, DF keff Time (Years) 23 Managed by UT-Battelle

24 FY Power Distribution More detailed model (9 rings, 5 axial nodes) developed to estimate power distribution and peaking factors (fresh core) Conclusions: Higher enrichment Lower CHM Flatten power distribution Use of BP Relative Radial Power Distribution M/A =Maximum-to-Average power ratios M/A=2.09 M/A=1.98 M/A=1.75 M/A= R (cm) Enrich=9% CHM=400 Eu Enrich=9% CHM=400 no Eu Enrich=9% CHM=200 noeu Enrich=19.75% CHM=200 noeu 24 Managed by UT-Battelle

25 Conclusions Prismatic FHR (AHTR) neutronic design studies are possible with current computational tools Reasonable neutronic design options exist Computational tools must be further benchmarked, refined and improved (desirable features: parallelism, continuous energy) for more detail and more accuracy 25 Managed by UT-Battelle