Overview of the BISON Multidimensional Fuel Performance Code

Transcription

1 Overview of the BISON Multidimensional Fuel Performance Code Rich Williamson BISON Team Jason Hales, Steve Novascone, Ben Spencer, Danielle Perez, Giovanni Pastore IAEA Technical Meeting: Modeling of Water-Cooled Fuel Including Design-Basis and Severe Accidents October 28 November 1, Chengdu, China

2 Outline Background Code Verification and Validation Applications in 3D Missing Pellet Surface Halden IFA-431 Rod 4 (eccentric vs concentric pellets) Priorities for the Future Summary

3 MOOSE-BISON-MARMOT The MOOSE-BISON-MARMOT codes provide an advanced, multiscale fuel performance capability Atomistic/Mesoscale Material Model Development Predicts microstructure evolution in fuel Used with atomistic methods to develop multiscale materials models Multiphysics Object-Oriented Simulation Environment Simulation framework allowing rapid development of FEM-based applications Advanced 3D Fuel Performance Code Models LWR, TRISO and metal fuels in 1D, 2D and 3D Steady and transient reactor operations

4

5 Fuel Performance Code Solution method: Implicit finite element solution of the coupled thermomechanics and species diffusion equations using the MOOSE framework Multiphysics constitutive models: large deformation mechanics (plasticity and creep), cracking, thermal expansion, densification, radiation effects (swelling, thermal conductivity, etc.). Massively parallel, has been run on 1-12,000 cpus. Substantial experimental validation is underway R. L. Williamson, J. D. Hales, S. R. Novascone, M. R. Tonks, D. R. Gaston, C. J. Permann, D. Andrs and R. C. Martineau, Multidimensional Multiphysics Simulation of Nuclear Fuel Behavior, Journal of Nuclear Materials, 423, 149 (2012)

6 BISON Governing Equations Energy conservation (transient heat conduction with fission source) c p T t ( k T ) E f F Fission density rate = f(x, t) Species conservation (transient oxygen or fission product diffusion with radioactive decay) C t D C CQ* 2 FRT Fickian diffusion Soret diffusion T C S Radioactive decay Momentum conservation (Cauchy s equation of equilibrium)

7 BISON LWR Capabilities General Capabilities Finite element based 1D spherical, 2D-RZ and 3D fully-coupled thermo-mechanics with species diffusion Linear or quadratic elements with large deformation mechanics Steady and transient operation Massively parallel computation Gap/Plenum Behavior Gap heat transfer with k g = f (T, n) Mechanical contact (master/slave) Plenum pressure as a function of: evolving gas volume (from mechanics) gas mixture (from FGR model) gas temperature approximation Meso-scale informed material models Oxide Fuel Behavior Temperature/burnup dependent conductivity Heat generation with radial and axial profiles Thermal expansion Solid and gaseous fission product swelling Densification Thermal and irradiation creep Fracture via relocation or smeared cracking Fission gas release (two stage physics) transient (ramp) release grain growth and grain boundary sweeping Temperature Cladding Behavior Thermal expansion Thermal and irradiation creep Irradiation growth Gamma heating Combined creep and plasticity Coolant Channel Closed channel thermal hydraulics with heat transfer coefficients

8 Tangential Stress (MPa) Temperature (K) Thermal Conductivity (W/m/K) What makes BISON different? Results Missing Pellet Surface Parallel Computing Some tasks are embarrassingly parallel meaning that they do not require any communication. For example, taking the square root of a list of numbers. However, most parallel computing requires communication. For example, summing a list of numbers. displacements magnified 25x MPS defect results in higher pellet temperatures and much higher clad stress; need for 3D analysis is clear 3D/Arbitrary Geometry CPU 0 CPU 1 CPU 2 Parallel Computing Simulation of Aspherical TRISO Particle Aspherical particles are fairly common Single facet aspherical particle problem has been solved in BISON assuming 2D axisymmetry 200 mm Irradiation 2.5 yrs 12% FIMA Step K Storage 100 days Step 2 Furnace Heating 200 hrs Step 3 Thermal Conductivity from MARMOT AEH_x AEH_y Direct_x Direct_y The direct method is sensitive to bubbles on the boundary Boundary bubbles result in low local temperature This results in a low average temperature and thus a low thermal conductivity The AEH technique is not sensitive to local effects 500 Time 300 K Aspherical Spherical Time (sec) During accident testing, asphericity raises peak tensile stress in SiC containment layer by almost 4x Typical run times of a few minutes on 8 processors Fuel Types Time (Ms) Temperature from the direct approach with flux applied in the x (left) and y (right) directions. Note areas of low temperature in the x-direction plot. Coupling

9 Outline Background Code Verification and Validation Applications in 3D Missing Pellet Surface Halden IFA-431 Rod 4 (eccentric vs concentric pellets) Priorities for the Future Summary

10 Verification, Validation, and Uncertainty Analysis BISON and associated software is supported by extensive regression testing (>800 tests for BISON thru MOOSE) Validation to a wide variety of integral rod tests in progress BISON has been coupled with SNL s DAKOTA code for systematic sensitivity analysis and uncertainty quantification parameters DAKOTA BISON response metrics Beginning of life temperature validation Calibration of relocation model using DAKOTA

11 BISON Code Assessment Comparisons to integral fuel rod data (21 rods, ~ 35 measurements) Experiment Rod FCT - BOL FCT TL IFA-431 1, 2, 3 X FCT - Ramps FGR Clad - Elong IFA-432 1, 2, 3 X X IFA-513* 1, 6 X X IFA A1 X X IFA X X IFA-597.3* 8 X RISO-3* AN3 X X RISO-3* AN4 X X FUMEX-II 27(1) X FUMEX-II 27(2a) X FUMEX-II 27(2b) X FUMEX-II 27(2c) X Clad Dia (PCMI) RISO-3 GE7 X X OSIRIS J12 X REGATE X X IFA-431 (3D) 4 X *Early User Assessment problems D. M. Perez, R. L. Williamson, S. R. Novascone, T. K. Larson, J. D. Hales, B. W. Spencer and G. Pastore, An Evaluation of the Nuclear Fuel Performance Code BISON, Int. Conf. on Mathematics and Computation Applied to Nuclear Science & Engineering (M&C 2013) Sun Valley, Idaho, May 5-9, 2013.

12 Beginning of Life Fuel Centerline Temperature Very good comparisons for the eleven measurements considered to date BOL comparisons validate important physics such as power input, fuel and clad thermal conductivity, gap gas conductivity, fuel and clad thermal expansion, gap closure and fuel relocation

13 RISØ-AN3 Power Ramp Base irradiated in the Biblis A PWR over four reactor cycles. Re-fabricated rod was shortened and instrumented with a fuel centerline thermocouple and pressure transducer. Ramp tested at the RISØ DR3 water-cooled HP1 rig Assumed short rod length through base irradiation and ramp Fuel Centerline Temperature Fission Gas Release

14 RISØ-3 GE7 PCMI During Power Ramp Base irradiated in the Quad Cities-1 BWR for ~5 years; Ramp tested in the RISØ DR3 water-cooled HP1 rig under BWR conditions 72-pellet stack modeled in axisymmetry with discrete pellets Strong axial power profile during ramp test

15 Outline Background Code Verification and Validation Applications in 3D Missing Pellet Surface Halden IFA-431 Rod 4 (eccentric vs concentric pellets) Priorities for the Future Summary

16 PCMI - Missing Pellet Surface Analysis Missing pellet surface He fill gas Cross-section of rod that failed due to MPS defect (from Aleshin et al, 2010) Zr-4 clad UO 2 fuel High resolution 3D calculation (250,000 elements, 1.1x10 6 dof) run on 120 processors Simulation from fresh fuel state with a typical power history, followed by a late-life power ramp

17 Results Temperature 0.25 mm deep defect 0.5 mm deep defect 0.25 mm deep defect 0.5 mm deep defect (a) pellets and cladding (b) only cladding shown (deformations magnified 15x)

18 Results Clad Creep Strain 0.25 mm deep defect 0.5 mm deep defect Zoomed-in view of 0.5 mm deep defect Creep during base irradiation plays significant role in relaxing stresses

19 Results Frictionless vs. Glued Frictionless Glued Contact 0.5 mm defect model run with glued contact to determine the effect of contact friction coefficient on stresses Friction coefficient has strong influence on stresses in defect region Frictionless and glued contact are bounding cases frictional contact is clearly needed, and is a current development effort

20 IFA-431 Rod 4 Concentric vs. Eccentric Pellets Several pellets locked in position (radially) by oversize pellets, thermocouples and Mo rods Top section held concentrically; Bottom section held eccentrically Rod Fill gas 100% Xe

21 IFA-431 Model Geometry Modeled as two separate shortened rods, with a single larger-diameter pellet at the top and bottom. The four-pellet test section is modeled as a single smeared column. In both cases, the thermocouple sheath and moly rods were centered in the largerdiameter pellets. In the eccentric case, the test pellets are shifted radially as shown.

22 Temperature (C) IFA-431 Relocation Activation Study 2000 BISON-Xe Concentric (19.7 kw/m activation) x x x x x x x x BISON-Xe Concentric (9.85 kw/m activation) BISON-Xe Concentric (4.93 kw/m activation) BISON-Xe Eccentric (19.7 kw/m activation) BISON-Xe Eccentric (9.85 kw/m activation) BISON-Xe Eccentric (4.93 kw/m activation) Halden-Xe Concentric Halden-Xe Eccentric x x x x x x x x ALHR (W/m) x xx Error bars are +/- 5% power uncertainty x xx x Original predicted temperatures (green) clearly follow trends in the experimental data (concentric hotter than eccentric), but the differences increase with increasing power. The simple empirical relocation model in BISON activates at 19.7 kw/m. As a parametric study, the relocation activation energy was simply reduced by a factor of two (9.85 kw/m) and four (4.93 kw/m), with the results plotted here. Other more rigorous studies also point to the need for a lower activation power. Better comparisons are obvious for lower activation power.

23 IFA-431 BOL Fuel Centerline Temperature Shown is a comparison of predicted temperatures for the concentric and eccentric rods at a LHGR of 20 kw/m. The peak temperature for eccentrically positioned pellets is clearly lower than for concentric pellets Clad inner wall temperature: Additionally, in the eccentric case, the thermocouple is not sensing the highest temperature region in the fuel.

24 AP-1000 Preliminary Simulation Loosely coupled: BISON RattleS n ake R7 Marmot Temperature and magnified displacement results for an AP1000 core. Each rod shown is a unique BISON simulation.

25 Outline Background Code Verification and Validation Applications in 3D Missing Pellet Surface Halden IFA-431 Rod 4 (eccentric vs concentric pellets) Priorities for the Future Summary

26 Path Forward Validation and UQ Completed: ~ 20 LWR cases, 13 TRISO cases Many, many more are needed; major emphasis for FY-14 - FUMEX-II and -III priority cases - Halden collaboration - NNL collaboration on ENIGMA cases Participation in FUMAC Develop systematic approach to frequently run and compare all cases and update documentation Sensitivity analyses and UQ studies DAKOTA and RAVEN

27 Modeling Accident Behavior (RIA and LOCA) Much of the required framework is in place: Transient operators with adaptive time stepping Arbitrary geometry with large deformation mechanics Implicit numerics with fully coupled physics Fission gas coupled to swelling Development areas: Multiphysics coupling Neutronics (RIA) Nakamura et al., 2000 Thermal-fluids (LOCA) Fuel material behavior Nakamura et al., 2000 Fission gas burst model Novel fracture models for extensive fuel cracking

28 Modeling Accident Behavior - continued Clad material behavior Temperature and strain rate dependent plasticity models Rapid steam oxidation, with a moving material interface Hydrogen diffusion and embrittlement High temperature creep and plasticity Ballooning (large plasticity) and burst (failure models and discrete fracture via XFEM)

29 1. J. D. Hales, R. L. Williamson, S. R. Novascone, D. M. Perez, B. W. Spencer, and G. Pastore, Multidimensional Multiphysics Simulation of TRISO Particle Fuel, J Nuc Mat, in press (2013). 2. M. C. Teague, M. R. Tonks, S. R. Novascone, and S. R. Hayes, Microstructural Modeling of Thermal Conductivity of High Burnup Mixed oxide Fuel, J Nuc Mat, in press (2013). 3. M. R. Tonks, P. C. Millett, P. Nerikar, S. Du. D. Andersson, C. R. Stanek, D. Gaston, D. Andrs, and R. Williamson, Multiscale Development of a Fission Gas Thermal Conductivity Model: Coupling Atomic, Meso and Continuum Level Simulations, J Nuc Mat, 440, (2013). 4. J. D. Hales, M. R. Tonks, M. R. Chockalingam, D. M. Perez, S. R. Novascone, and B. W. Spencer, Multiscale Nuclear Fuel Analysis via Asymptotic Expansion Homogenization, Transactions of SMiRT-22, San Francisco, California, August S. R. Novascone, B. W. Spencer, R. L. Williamson, D. Andrs, J. D. Hales, and D. M. Perez, The Effects of Thermomechanics Coupling Strategies in Nuclear Fuel Performance Simulations, Transactions of SMiRT-22, San Francisco, California, August F. N. Gleicher, S. R. Novascone, B. W. Spencer, R. L. Williamson, R. C. Martineau, M. Rose, and T. Downar, Coupling the Core Analysis Program DeCART to the Fuel Performance Program BISON, Proceedings of the International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering, Sun Valley, Idaho, 5-9 May J. D. Hales, D. Andrs, and D. R. Gaston, Algorithms for Thermal and Mechanical Contact in Nuclear Fuel Performance Analysis, Proceedings of the International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering, Sun Valley, Idaho, 5-9 May D. M. Perez, R. L. Williamson, S. R. Novascone, T. K. Larson, J. D. Hales, B. W. Spencer, and G. Pastore, An Evaluation of the Nuclear Fuel Performance Code BISON, Proceedings of the International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering, Sun Valley, Idaho, 5-9 May S. R. Novascone, B. W. Spencer, D. Andrs, R. L. Williamson, J. D. Hales, and D. M. Perez, Results from Tight and Loose Coupled Multiphysics in Nuclear Fuels Performance Simulations using BISON, Proceedings of the International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering, Sun Valley, Idaho, 5-9 May L. P. Swiler, R. L. Williamson, and D. M. Perez, Calibration of a Fuel Relocation Model in BISON, Proceedings of the International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering, Sun Valley, Idaho, 5-9 May J. D. Hales, D. M. Perez, R. L. Williamson, S. R. Novascone, B. W. Spencer, and R. C. Martineau, Validation of the BISON 3D Fuel Performance Code: Temperature Comparisons for Concentrically and Eccentrically Located Fuel Pellets, Enlarged Halden Programme Group Meeting: Proceedings of the Fuels and Materials Sessions, Storefjell Resort Hotel, Norway, March BISON Release and Documentation Code Release (Ver 1.0) on Sept. 30, 2013 Theory, User and Assessment Manuals Journal and Conference Papers (2013)

30 Summary BISON is being leveraged across multiple US-DOE programs and is in use at multiple national and international laboratories, many universities and in industry User base is expanding user support efforts are significant Roughly 20 integral rod LWR and 13 TRISO validation cases have been completed many more are needed BISON used to do first-ever simulation of 3D pellet eccentricity experiment (invited paper at Halden Reactor Program Meeting) Major new capability completed for TRISO-coated particle fuel Numerous new applications (metal plate fuel, fast reactor oxide fuel, accident tolerant fuel designs, ATR experiment design) demonstrate BISON s versatility First official code release was September 30, 2013 FCRD 30

31 The National Nuclear Laboratory 31

32 Animal Hierarchy Individual Applications BISON Peregrine Common fuels related physics Common non-fuels related physics Common component interface and solution scheme FOX ELK MOOSE

33 RISO-3 AN3 FUMEX-II Priority Case Fuel pin CB8 was base irradiated in the Biblis A PWR over four reactor cycles Re-fabricated fuel pin (CB8-2R) was shortened from the CB8 rod and instrumented with a fuel centerline thermocouple and pressure transducer. CB8-2R was ramp tested at the Riso DR3 water-cooled HP1 rig Assumed short rod length through base irradiation for simulation Modeled as 2D-RZ axisymmetric with smeared pellets; Quadratic elements Aspect ratio scaled 10x

34 RISO-3 GE7 PCMI He fill gas (0.29 MPa) UO 2 fuel FUMEX-III Priority Case Fuel segment ZX115 base irradiated in the Quad Cities-1 BWR for ~5 years Ramp tested in the Riso DR3 water-cooled HP1 rig under BWR conditions 72-pellet stack was modeled as 2D-RZ axisymmetric with individual (discrete) pellets; quadratic elements Strong axial power profile during ramp test Zr-2 clad 110 mm gap radius expanded 6x

35 RISO-3 GE7 Continued 100 s ramp

36 Results Missing Pellet Surface displacements magnified 25x MPS defect results in higher pellet temperatures and much higher clad stress; need for 3D analysis is clear

37 Path Forward Enhanced Oxide Fuel Capabilities Fuel fracture (smeared and discrete) Thermomechanical contact Material model development - Fuel creep - Solid swelling - Clad behavior coupling primary/secondary creep to instantaneous plasticity - High burnup fuel structure (with coupling to Marmot and MAMMOTH) General restart Smeared cracking Discrete cracking HBS Micrographs, Photos courtesy of Glyn Rossiter, National Nuclear Laboratory

38 IFA-431 Eccentricity Study The precise radial location of the eccentric pellets is not clear in the above description, which was extracted from the report NUREG/CR-0560 (PNL-2673). Note that for the concentric portion of the rod, the radial pellet-clad gap for the standard diameter pellets is 114 mm. However the drawing indicates the pellets were displaced radially 127 mm in the eccentric case, which would place them beyond the inner wall of the clad.

39 Temperature (C) IFA-431 Eccentricity Study, Continued Error bars are +/- 5% power uncertainty ALHR (W/m) BISON - Flush BISON - 10 mm off BISON - 20 mm off HALDEN Since the precise radial location of the eccentric pellets was unknown, an additional parametric study involved varying the location of the eccentric pellets from flush with the larger pellets to 10 and 20 mm less than flush. For the non-flush cases, the fuel pellets were moved inward thus increasing the local pelletclad gap. The results show the expected effect of varying eccentricity and seem to indicate that the level of eccentricity is irrelevant once the test pellets come into contact with the clad. Relocation activation level was 4.9 kw/m per prior parametric study.

40 User Training Workshop Materials ~ 200 slides Overview, Getting Started, Theory, Example Problem, Mesh Generation, Post Processing, Adding a New Material BISON 2-day Workshops INL January 2012 MIT January 2012 Anatech Feb 2012 INL May 2012 (23 participants) NNL (UK) September 2012 INL December 2012 (15 participants) INL June 2013 (11 participants) WEC Aug 2013 (10 participants) INL Dec 2013 (planned)