Gas Reactor TRISO-Coated Particle Fuel Modeling Activities at the - - PowerPoint PPT Presentation

Idaho National Engineering and Environmental Laboratory

Gas Reactor TRISO-Coated Particle Fuel Modeling Activities at the Idaho National Engineering and Environmental Laboratory

David Petti, Gregory Miller, John Maki, Dominic Varacalle, and Jacopo Buongiorno Paris, France October 12, 2001

Idaho National Engineering and Environmental Laboratory

Outline

• Fuel Performance Model Objectives
• Structural Modeling

– Normal particle – Property Database – Debonded particle – Cracked particle

• Chemistry module

– Fission Gas and CO Release Model

• Comparison of PARFUME predictions to recent US

irradiations

• Preliminary predictions from PARFUME for German

LEU particle at high burnup

• Future Work

Idaho National Engineering and Environmental Laboratory

PARticle FUel ModEl (PARFUME)

• Objective: To develop a mechanistic fuel

performance model that – Describes the relevant behavior of TRISO-coated fuel during irradiation – Explains past poor performance for US gas reactor fuel – Can be used as a design tool to develop improved coated particle fuel – Can be used to establish a linkage between acceptable in-reactor performance and a fuel product/process fabrication specification

Idaho National Engineering and Environmental Laboratory

Integrated Mechanistic Fuel Performance Model Multidimensional Finite Element Structural Analysis Physio-chemical Behavior Under Irradiation

• Fission Gas Release and Swelling
• Fission Product Chemistry Behavior

(Diffusion, Migration, and Segregation)

• Gas production (CO and CO2)
• Irradiation-induced dimensional changes of

the coating layers

• Kernel Migration
• Restructuring and grain growth

Structural Properties

• fracture strength
• elastic modulus
• poisson's ratio
• creep constants
• Interfacial bond

strength Multidimensional Statistical Fit Of Stress Results

Idaho National Engineering and Environmental Laboratory

Radial crack in IPyC Debonding of IPyC from SiC SiC crack near IPyC crack Examination of NPR Fuel indicates that asymmetric radiation shrinkage is important

Photomicrograph of failed NPR fuel particle

Need to consider cracking and debonding in model development

Idaho National Engineering and Environmental Laboratory

Three Conceptual Models for Particle Failure

Standard 3 Layer Model 3 Layer Cracked Model 3 Layer Debonded Model

OPyC SiC IPyC

Idaho National Engineering and Environmental Laboratory

ABAQUS Results from Standard and Cracked Models

Note: Model contains ~ 800 nodes and takes about 2- 8 hours to run

Stress Concentration at Crack Tip

SiC Layer

Standard Particle Cracked Particle

Standard/Nominal Particle is in compression; Particle with Cracked IPyC has SiC layer in tension

2 4 6 8 10 12 Time (x 106 seconds) 200

• 200
• 400
• 600

Principal stress (MPa) Cracked Particle Normal Particle

Idaho National Engineering and Environmental Laboratory

ABAQUS results for debonded particle

Stress Concentrations at Edge of Debonding

SiC Stress

Idaho National Engineering and Environmental Laboratory

PyC shrinkage is a function of temperature, Bacon Anisotropy Factor (BAF), density and fluence

Radial Change at 1032°C

• 2
• 1

1 2 3 1 2 3 4

Fluence (*10^25 n/m*2)

BAF=1.28 BAF=1.17 BAF=1.08 BAF=1.02

Tangential Change at 1032°C

• 7
• 6
• 5
• 4
• 3
• 2
• 1

1 2 3 4

Fluence (*10^25 n/m*2)

BAF=1.02 BAF=1.08 BAF=1.17 BAF=1.28

Radial (BAF=1.08)

• 5

5 10 15 20 2 4 6 8 Fluence (x 10^25 n/m*2)

Data - 910°C Fit 910°C Data - 700°C Fit 700°C Data - 1215°C Fit 1215°C

Tangential (BAF=1.08)

• 15
• 12
• 9
• 6
• 3

2 4 6 8

Fluence (x 10^25 n/m*2)

Data 910°C Fit 910°C Data 700°C Fit 700°C Data 1215°C Fit 1215°C

Idaho National Engineering and Environmental Laboratory

Pyrocarbon irradiation induced creep rate has large influence on stress in IPyC and stress concentration in SiC

SiC Stress in Cracked Model at 1200°C using different PyC creep data

Using historical creep value of 3.29 *1027 (psi-nvt)-1 from GA Using new creep value of 1.4*1027 (psi-nvt)-1 based

• n broad

assessment of data from GA in 1993 Note: STRESS3 code uses 3.4 *1027 (psi-nvt)-1

Idaho National Engineering and Environmental Laboratory

Development of Key Parameters and Levels

• KEY PARAMETERS

– Temperature (K) – IPyC Thickness (um) – IPyC Anisotropy (BAF) – IPyC Density (g/cc) – SiC Thickness (um) – OPyC Thickness (um) Low Medium High 873 1073,1273 1473 30 40 50 1.0 1.16 1.33 1.8 1.9 2.0 30 40 50 33 43 53

972 calculations! Statistical fit good to 0.5%

Idaho National Engineering and Environmental Laboratory

Effect of Temperature: Differences in PyC creep have large differences in calculated stress in IPyC and SiC

IPyC Layer In SiC near crack tip

• --- 600 °C
• --- 1200°C

Idaho National Engineering and Environmental Laboratory

Fission Gas Release Model

• Release fraction considerations

– Fission recoil contribution based upon range distributions tabulated by Littmark and Ziegler – Diffusion coefficient used in the equivalent sphere model based upon the correlation developed by Turnbull et al. This correlation is the sum of three contributions:

• At high temperatures, intrinsic diffusion dominates
• At intermediate temperatures, radiation enhanced

vacancy diffusion dominates

• At low temperatures, an athermal radiation induced

contribution dominates (possibly ascribed to a knock-out mechanism)

Idaho National Engineering and Environmental Laboratory

CO Generation and Release Model

• CO yield comparison for UO2 fuel:

– NPR review: CO yield may be as high as 0.38 based upon available oxygen – ORNL compilation: CO yield is approximately 0.13 but may be as high as 0.40 depending on thermodynamic calculations – GA model for LEU fuel: (yield)CO = 1.64 exp(-3311/T) where T = temperature (K) at 1273 K, (yield)CO = 0.12

• CO gas release is temporarily using the GA model:

– For UCO fuel, (yield)CO = 0 – For LEU UO2 fuel, (yield)CO = 1.64 exp(-3311/T) where T is temperature in degrees K

• For UO2 fuel, (release fraction)CO = 1

Idaho National Engineering and Environmental Laboratory

Model results for German particle

1 2 3 4 5 6 500 1000 1500 Irradiation Time (days)

Total internal pressure CO partial pressure Fission gas partial pressure

Idaho National Engineering and Environmental Laboratory

Chemistry Model

• Physics results

– Power density and burnup for 4%, 8%, 12%, 16%, and 20% enrichment – Used HTR module as design point (no inner reflector or mixing zone)

• ORIGEN2 calculations are complete

– Elemental fission product concentrations as a function of initial uranium loading in the pebble and burnup – Use either lookup tables or statistical fits to develop input for HSC chemistry calculations

• Chemistry calculations expected this fall
• Goal is to get fission gas release and CO pressure as

a function of O/C ratio in kernel and state of important fission products in the kernel (oxide versus carbide)

Idaho National Engineering and Environmental Laboratory

Comparison of Ceramographic Observations from Recent US Irradiations to PARFUME Calculations for TRISO Coated Fissile Fuel Particles

• Fuel Compact

% Cracked % Cracked Fast Fluence Irradiation Burnup

• IPyC layer

SiC layer 1025 n/m2 temperature % FIMA °C

• PIE PARFUME

PIE PARFUME

• NPR-2 A4

65 100 3 8.2 3.8 746 79

• NPR-1 A5

31 100 0.6 1.6 3.8 987 79

• NPR-1 A8

6 100 4.9 2.4 845 72

• NPR-1A A9

18 100 1 0.9 1.9 1052 64

• HRB-21 1C

1 100 7.7 1.5 800 14

• HRB-21 2B

3 100 1.9 2.3 980 18

• HRB-21 4A

33 100 5 1.6 3.5 1000 22.5

• NPR fuel is UCO, 93% U-235
• HRB-21 fuel is UCO, 20% U-235

Idaho National Engineering and Environmental Laboratory

Increase of creep coefficient by a factor of 2 to 3.5 will give results that are in much better agreement with the data and is more consistent with creep values used previous fuel performance codes

NPR-2 A4 20 40 60 80 100 120 1 2 3 4 5 Factor Applied to Creep Coefficient

Calculated Data

NPR-1 A8 20 40 60 80 100 120 1 2 3 4 5 Factor Applied to Creep Coefficient

Calculated Data

NPR-1A A9 20 40 60 80 100 120 1 2 3 4 5 Factor Applied to Creep Coefficient

Calculated Data

NPR-1 A5 20 40 60 80 100 120 1 2 3 4 5 Factor Applied to Creep Coefficient

Calculated Data

Idaho National Engineering and Environmental Laboratory

Preliminary PARFUME Predictions of Structural Behavior of German LEU Particle at High Burnup

• 600
• 500
• 400
• 300
• 200
• 100

5 10 15 20 25 End-of-life burnup (%FIMA) SiC Stress (MPa)

CEGA creep Amplified creep

• 300
• 250
• 200
• 150
• 100
• 50

5 10 15 20 25 End-of-life burnup (%FIMA) SiC Stress (MPa) 700 C 900 C 1100 C

SiC layer in compression under all conditions

Idaho National Engineering and Environmental Laboratory

Future Work (next 1-3 years)

• PARFUME calculations of fuel failures from previous

US irradiations if there are enough data

• Complete fission product chemical modeling
• Start fission product diffusion modeling
• Integrate all pieces into PARFUME, the overall fuel

performance model

• Exercising the code to understand the thermo-

mechanical and physico-chemical limits of TRISO- coated fuel at very high burnup and high fluence

• Consider other advanced coating systems, other

kernel types (Th, Pu) and other reactor concepts (e.g., gas cooled fast reactor)