1. FRAPCON/FRAPTRAN Code Application NRC Office of Research
Patrick Raynaud, Ph.D.

2. FRAPCON/FRAPTRAN User Group - Manchester, UK
Outline
Background on NRC fuel behavior codes
Predicting LOCA and RIA limits
FRAPCON hydrogen models
SCIP-2 Modeling Workshop power ramps
FRAPCON/FRAPTRAN comparison
Recommendations for power ramp modeling
Fuel dispersal effort: core-wide realistic rod burst inventory
FRAPCON/FRAPTRAN and TRACE integration
Input generator improvements for detailed coolant boundary conditions
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

3. Background NRC Fuel Behavior Codes
FRAPCON-3.4
FRAPTRAN-1.4
Steady-state and slow transients
Minutes to many days
Equilibrium solution
Thermal, mechanical, fission gas, rod internal pressure response, corrosion, hydriding, cladding creep
No failure models
Warnings and/or stops when certain limits are reached (1% hoop strain, fuel melt…)
Rapid transients
Milliseconds to a few minutes
Transient solution
Thermal, mechanical, fission gas, rod internal pressure response, high temperature corrosion, fuel cladding interaction, cladding failure (PCMI, ballooning)
Failure models (ballooning and burst, PCMI)
RIA, LOCA, BWR oscillations
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

4. Predicting RIA and LOCA Limits Objectives of the Study
Obtain hydrogen content [H] as a function of rod burnup BU to investigate burnup dependence of LOCA and RIA criteria
Use FRAPCON-3.4 to predict [H] vs. BU for U.S. cladding alloys
Zircaloy-2: BU dependent hydrogen pickup model: direct [H] vs. BU relationship
Zircaloy-4, ZIRLO™, M5™: heat flux, neutron flux, temperature, and time dependent model: complex indirect [H] vs. BU relationship
Constant alloy-dependent hydrogen pickup fractions
Generate BU dependent allowable ECR and ∆h for U.S. cladding alloys for different core axial elevations and power histories
Compare BU dependent LOCA and RIA limits as a function of cladding alloy
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

5. FRAPCON Fuel Performance Modeling
BWR
Typical GE BWR/4
10x10 fuel assembly, Zircaloy-2 cladding
Typical Westinghouse 4-loop PWR
17x17 fuel assembly, ZIRLO™ cladding
Core-load patterns and rod average power from plant safety analysis reports
Best guess at representative power histories for 2-cycle and 3-cycle lifetimes
7 PWR and 18 BWR power histories
PWR histories based on ZIRLO™ cladding used for Zircaloy-4 and M5™
BWR
PWR
PWR
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

6. FRAPCON Hydrogen Models
Zircaloy-2 [H] only dependent on BU: no axial-node dependence
PWR alloys: strong temperature and heat flux dependence results in large axial variations
Zircaloy-4: high oxidation and H pickup fraction
Early transition in oxidation kinetics (δoxide>2μm)
Oxidation beyond allowable limits
ZIRLO™: intermediate behavior
M5™: low oxidation and H pickup fraction
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

7. LOCA and RIA Criteria Alloy Model Comparison
Zircaloy-2 in BWR
Better predicted performance than Zircaloy-4 and ZIRLO™ in PWR
Rapid degradation at high burnup
PWR alloys:
Highest H pickup for Zircaloy-4: lowest margin
M5™ plants less challenged by new criteria
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

8. SCIP-2 Modeling Workshop Ramps and Fission Gas Release
Studsvik slow power ramps performed in R-2 reactor
16 rods modeled
12 rods modeled under FRAPCON only
KKL-1, M5-H2, O-2, Z-4
OL1-1, OL1-2, OL1-3, OL1-4
OA1-1, OA1-2, OA1-3, OA1-4
4 rods with FRAPCON and then with FRAPTRAN initialized by FRAPCON
GE-1 (feasibility scoping study)
xM1, xM2, xM3
FRAPCON and FRAPTRAN predictions compared for 4 cases
Trends based on ramp characteristics were investigated
No specific PCI models in FRAPCON and FRAPTRAN
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

9. FRAPCON Modeling Methodology Slow Power Ramps
Base irradiation and ramp simulated in same run
Used automatic input generator
Input given power histories and shapes and then made small adjustments to power level to match discharge burnups
Initial rod internal pressure adjusted to match refabricated rodlet pressure after base irradiation and before ramp test
Initial and final step at cold zero power
Allows for free volume calibration and residual hoop strain and gap predictions
Fission gas release (FGR)
1st run for base FGR: FGR turned ON for base irradiation and ON for ramp test
2nd run for ramp FGR: FGR turned OFF for base irradiation and ON for ramp test
Reminder: stepwise ramp approximation in FRAPCON
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

10. FRAPTRAN for Slow Power Ramps Evaluation with the GE-1 Case
FRAPTRAN is less suited to model phenomena over a long time scale
No creep
No steady-state fission gas release models
No pellet radial relocation and relaxation models
Different hypotheses were investigated to determine if FRAPTRAN can be used to model slow power ramps:
Default FRAPTRAN models
Transient FGR model
User-input rod internal pressure (RIP) to match FRAPCON predictions (no FGR modeled)
User-input fission gas release (FGR) to match FRAPCON predictions (FRAPTRAN default pressure calculation)
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

11. FRAPTRAN for Slow Power Ramps GE-1 Plenum Pressure
Poor agreement during ramp unless RIP or FGR is imposed
Imposing RIP does not match FGR, imposing FGR matches RIP and FGR
FRAPTRAN-1.4 defaults
FRAPTRAN-1.4 transient FGR
FRAPCON shown in red
FRAPTRAN-1.4 imposed RIP
FRAPTRAN-1.4 imposed FGR
FRAPTRAN shown in blue
Good Agreement
Good Agreement
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

12. FRAPTRAN for Slow Power Ramps GE-1 Gap Conductance or HTC
Poor agreement during preconditioning (gap open/closed)
Good agreement once gap closes in FRAPTRAN for imposed FGR
FRAPTRAN-1.4 defaults
FRAPTRAN-1.4 transient FGR
FRAPCON shown in red
FRAPTRAN-1.4 imposed RIP
FRAPTRAN-1.4 imposed FGR
Good agreement during most of ramp
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

13. FRAPTRAN for Slow Power Ramps GE-1 Cladding Perm. Hoop Strain
Similar trends predicted, but FRAPTRAN under-prediction
Improved agreement if FGR is imposed
FRAPTRAN-1.4 defaults
FRAPTRAN-1.4 transient FGR
FRAPCON shown in red
FRAPTRAN-1.4 imposed RIP
FRAPTRAN-1.4 imposed FGR
FRAPTRAN shown in blue
Best Agreement
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

14. FRAPTRAN for Slow Power Ramps Summary Based on GE-1 Case
Large differences between the two codes
Gap closure and heat transfer coefficient, rod internal pressures, cladding stresses, etc…
Agreement between the codes can be improved
Turning FGR ‘on’ in FRAPTRAN
Small improvement on residual gap, gap conductance, and RIP
Imposing RIP
Same RIP but degraded free volume and permanent cladding hoop strain
Imposing FGR
Large improvements in gap conductance, fuel temperature, RIP, and cladding permanent hoop strain
Relative Agreement Between FRAPCON and FRAPTRAN
Gap Size
Gap HTC
RIP
FGR
Free Volume
Cladding Hoop Strain
Fuel Temperature
FRAPTRAN Model
Transient FGR
Better
No Change
Imposed RIP
Lesser
Best
Worst
Imposed FGR
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

15. FRAPTRAN for Slow Power Ramps Comparison with FRAPCON
Large differences between the two codes related to mechanical models
Fuel relocation recovery in FRAPCON but not FRAPTRAN
Absence of creep and differences in rod internal pressure in FRAPTRAN
Differences in predicted permanent hoop strain (creep + plastic in FRAPCON versus just plastic in FRAPTRAN)
Impact on gap closure and stresses
Fission Gas Release absent in FRAPTRAN
Must be imposed manually based on FRAPCON calculation
Impact of FGR and mechanical predictions on thermal predictions are significant
Very different gap heat transfer coefficients due to very different gap sizes and different RIP and gas composition
Different fuel temperature predictions >100 K
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

16. FRAPTRAN for Slow Power Ramps Conclusions & Recommendations
FRAPCON overall better captures the phenomena at play during a relatively slow transient such as a power ramp
Creep, fission gas release, fuel relocation recovery, gap heat transfer
FRAPTRAN can be used but with caution
Fission gas release should be manually added to each time step
Gap size and heat transfer coefficient should be looked at closely and matched with FRAPCON when possible to improve predictions
For events longer than 3-8 seconds, such as the SCIP power ramps, thermal equilibrium will be reached in the fuel rod
FRAPCON is preferred over FRAPTRAN for the SCIP power ramps
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

17. FRAPCON/FRAPTRAN and TRACE for LOCA Rod Burst Inventory
Objective: calculate a best-estimate number of fuel rods that rupture
Supports efforts to assess impact of fuel dispersal during a LOCA
1st case chosen: large-break LOCA in a 4-loop PWR with large dry containment
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

18. FRAPCON Power History Modeling
Generate power bins and histories to maximize core average discharge burnup (final core average BU ~ 51 GWd/MTU)
Keep track of every assembly throughout life in the core
Peak assembly (center of core)
Cycle 1 power: 1G
Cycle 2 power: 2L History: 1G2L3A
Cycle 3 power: 3A
Discharge BU: GWd/MTU
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

19. TRACE Output Processing for FRAPTRAN
8 coolant zones, with corresponding fuel rod heat structures
1-to-1 axial zone correspondence (14 axial nodes)
Extracted variables for coolant boundary conditions:
‘Coolant’ option
Attempt to accurately model coolant conditions
Inlet pressure, enthalpy, and mass flux
‘Heat’ option
Trick to impose cladding OD temperature
Coolant pressure, temperature, and HTC
Impose coolant temperature = cladding OD temperature
Impose very high HTC (to force cladding OD temperature equal to coolant temperature)
Reflood option
Determine reflood rate based on core level vs. time
‘coolant’ or ‘heat’ options ignored once reflood begins
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

20. FRAPCON/FRAPTRAN User Group - Manchester, UK
FRAPTRAN Modeling
Initialize FRAPTRAN with FRAPCON base irradiation runs at MOC
22 different possible power histories
For a given power history, choice restart time determines the assembly burnup (1st cycle, 2nd cycle, or 3rd cycle)
Use coolant boundary conditions from TRACE
Run FRAPTRAN until after quench and determine whether rod has ruptured
43 groups of rods * 8 azimuthal coolant sectors = 344 FRAPTRAN runs
10 first cycle power bins
22 second cycle power bins
11 third cycle bins
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

21. Input Generator Enhancements for Advanced Coolant Modeling
‘coolant’ option: up to 50 time/parameter pairs
‘heat’ option: up to 100 time/parameter pairs and 20 axial zones
‘reflood’ option: up to 20 time/parameter pairs for inlet temperature and pressure, and up to 100 time/parameter pairs for reflood rate
Except for increased number of coolant zones in ‘heat’ option, capabilities already existed in the code, but were added to input generator
Coolant zones were increased from 10 to 20 for ‘heat’ option
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK

22. FRAPCON/FRAPTRAN User Group - Manchester, UK
Summary
NRC is continuing in-house use of the code
PNNL continues to be very supportive of knowledge transfer activities to NRC, and as a result, NRC is actively participating in the code development effort that it sponsors at PNNL
FRAPCON/FRAPTRAN analyses support regulatory decision-making as well as safety scoping studies, and benchmarking exercises
NRC and PNNL are seeking additional opportunities to collaborate and exchange with other code users
Debugging, novel code applications, code interfacing, etc…
09/07/2012
FRAPCON/FRAPTRAN User Group - Manchester, UK