ML20236U160
| ML20236U160 | |
| Person / Time | |
|---|---|
| Site: | Mcguire, Catawba, McGuire  |
| Issue date: | 07/31/1998 |
| From: | DUKE POWER CO. |
| To: | |
| Shared Package | |
| ML20138K462 | List: |
| References | |
| DPC-NE-2009, NUDOCS 9807290343 | |
| Download: ML20236U160 (82) | |
Text
_ ._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
f NON-PROPRIETARY DPC-NE-2009 I
DUKE POWER COMPANY WESTINGHOUSE FUEL TRANSITION REPORT July 1998 Nuclear Engineering Division Nuclear Generation Department Duke Power Company pgg72g0gggggggj69 p PDR
Duke Power Company Westinghouse Fuel Transition Report TABLE OF CONTENTS Pace
1.0 INTRODUCTION
1-1 2.0 FUEL DESIGN 2-1 2.1 References 2-3 3.0 CORE DESIGN 3-1 3.1 o Introduction 3-1 3.2 Reload Desien Methodolocv 3-1 3.3 References 3-3 4.0 FUEL ROD ANALYSIS 4-1 4.1 Comnuter Code 4-3 4.2 Fuel Rod Desien Bases and Analyses 4-4 4.2.1 Fuel Rod Internal Pressure 4-4 4.2.1.1 Analysis 4-4 4.2.2 Fuel Rod Cladding Stress 4-5 4.2.2.1 Analysis 4-6 4.2.3 Fuel Rod Cladding Strain 4-6 4.2.3.1 Analysis 4-7 4.2.4 Fuel Rod Cladding Fatigue 4-7 4.2.4.1 Analysis 4-7 4.2.5 Fuel Clad Oxidation and Hydriding 4-8 4.2.5.1 Analysis 4-8 4.2.6 Fuel Temperature 4-9 4.2.6.1 Analysis 4-9 4.2.7 Fuel Clad Flattening 4-10 4.2.7.1 Analysis 4-10 4.2.8 Fuel Rod Axial Growth 4-10 4.2.8.1 Analysis 4-11 4.3 References 4-12 5.0 THERMAL-HYDRAULIC ANALYSIS 5-1 5.1 Plant Specific Data 5-1 ii
Paee 5.2 Thennal-Hydraulic Code and Model 5-2 5.3 Critical Heat Flux Correlation 5-3 5.4 Statenoints 5-3 5.5 Kev Parameters 5-4 5.6 DNB Statistical Desien Limit 5-4 5.7 Transition Cores 5-4 5.8 References 5-5 6.0 UFSAR ACCIDENT ANALYSES 6-1 6.1 Thermal-Hydraulic Transient Analysis Methodology (DPC-NE-3000) 6-1 6.1.1 Plant Description 6-2 6.1.2 McGuire/ Catawba RETRAN Model 6-2 6.1.3 McGuire/ Catawba VIPRE Model 6-2 6.2 Multidimensional Reactor Transients and Safety Analysis 6-3 Physics Parameters Methodology (DPC-NE-3001) 6.2.1 Rod Ejection 6-3 6.2.2 Steam Line Break 6-3 6.2.3 Dropped Rod 6-4 6.3 UFSAR Chapter 15 System Transient Analysis Methodology (DPC-NE-3002) 6-4 6.4 Mass and Enercy Release and Containment Response Methodolocv 6-4 (DPC-NE-3004) 6.5 LOCA Analyses 6-5 6.5.1 Small Break LOCA 6-5 6.5.2 Large Break LOCA 6-6 6.6 Rod Ejection Analysis Usine SIMUL ATE-3K 6-8 6.6.1 Simulation Codes and Models 6-9 6.6.1.1 FASMO-3 & SIMULATE-3P 6-9 6.6.1.2 ARROTTA 6-10 6.6.1.3 SIMULATE-3K 6-10 iii
Pace 6.6.2 Rod Ejection Nuclear Analysis 6-15 6.6.2.1 REA Analytical Approach 6-15 6.6.2.2 SIMULATE-3K Nuclear Analysis 6-16 6.6.2.3 Core Thermal-Hydraulic Analysis 6-20 6.6.2.4 Cvele Snecific Evaluation 6-22 6.6.2.5 Mixed Cores 6-22 6.7 References 6-23 7.0 FUEL ASSEMBLY REPAIR AND RECONSTITUTION 7-1 7.1 References 7-2 8.0 IMPROVED TECHNICAL SPECIFICATION CHANGES 8-1 8.1 Technical Justification for Surveillance Requirement SR 3.2.1.2 8-2 and SR 3.2.1.3 8.2 Technical Justification for Surveillance Requirement SR 3.2.2.2 8-3 8.3 References 8-4 iv
Duke Power Company Westinghouse Fuel Transition Report List of Tables Table Bge 2-1 Comparison of Robust Fuel Assembly and Mark-BW Fuel Assembly 2-4 Design Parameters 3-1 Nuclear Uncertainty Factors 3-4 5-1 Westinghouse Robust Fuel Assembly Data 5-6 5-2 McGuire/ Catawba SCD Statepoints 5-7 5-3 McGuire/ Catawba Statistically Treated Uncertainties 5-8 5-4 McGuire/ Catawba Statepoint Statistical hesults 5-11 5-5 McGuire/ Catawba Key Parameter Ranges 5-13 6-1 Rod Ejection ARROTTA Results 6-28 6-2 Rod Ejection SIMULATE-3K Results 6-28 6-3 Rod Ejection Transient Kinetics Input Parameters 6-29 8-1 Ffx,y,z) Margin Decrease Over 31 EFPD Surveillance Interval 8-5 8-2 Fm(x,y) Margin Decrease Over 31 EFPD Surveillance Interval 8-6 I
v
(
Duke Power Company Westinghouse Fuel Transition Repon List of Fieurel Figure Pace 4-1 Typical Bounding Pin Power History 4-14 6-1 Reference Core Loading Information 6-30 6-2 FSAR Section 15.4.8 - Control Rod Ejection 6-31 BOC HFP Core Power vs. Time 6-3 FSAR Section 15.4.8 - Control Rod Ejection 6-31 BOC HZP Core Power vs. Time 6-4 FSAR Section 15.4.8 - Control Rod Ejection 6-32 EOC HFP Core Power vs. Time 6-5 FSAR Section 15 4.8 - Control Rod Ejection 6-32 EOC HZP Core Power vs. Time vi
1.0 INTRODUCTION
Duke Power Company is currently using Framatome Cogema Fuels (FCF) Mark-BW fuel assemblies in the McGuire and Catawba reactors. Duke Power will transition to the 17x17 Westinghouse 0.374 Robust Fuel Assembly (RFA) design described in Chapter 2 of this report.
i This topical report presents the information required to support the licensing basis for the use of the RFA design in McGuire and Catawba reload cores.
This report describes the core design, fuel rod design, and thermal-hydraulic analyses that are performed to show that all licensing criteria are met for each reload core. This report also discusses the UFSAR Chapter 15 transient and accident analyses methodology that is applicable to each reload design. Previously approved methodologies used by Duke Power Company to perform core design, thermal-hydraulic design, and UFSAR Chapter 15 Non-LOCA analyses for the Mark-BW fuel will be used to analyze the RFA design with the revisions described in Chapters 3,5, and 6, respectively. l Chapter 4 describes the fuel rod design analysis methodology that will be used to analyze the j RFA design. Although the fuel rod analysis methodology is new for Duke Power, the methods are essentially identical to the NRC-approved Westinghouse methods. The Westinghouse LOCA analysis methodology is described in Section 6.5. Section 6.6 presents an improved methodology that will be used to perform the nuclear analysis portion of the rod ejection accident (REA) analysis for McGuire and Catawba. The new methodology is based on the SIMULATE-3K computer code.
Chapter 7 discusses the licensing and analysis approach Duke Power will use for reconstitution of the RFA design. Chapter 8 describes the Technical Specification changes that will be made due to the transition to the RFA design and the analysis methodology described in this report.
1-1 I
I t -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
2.0 FUEL DESIGN Duke Power is transitioning to the Westinghouse 17x17 0.374 robust fuel assembly design for the McGuire and Catawba reactors. For the remainder of this report the fuel design will be referred to as simply the RFA design. The RFA design is based on the VANTAGE + fuel assembly design, licensed by the NRC in Reference 2-1. The RFA design used at .McGuire and Catawba will include the following features initially licensed with the VANTAGE + fuel design:
- ZIRLO clad fuel rods,
- ZIRLO guide thimbles, instrumentation tubes and mid-grids (both structural and Intermediate Flow Mixing (IFM) grids),
- 0.374 inch fuel rod OD,
- Zirconium diboride Integral Fuel Burnable Absorbers (IFBAs), I
- Mid-enriched annular axial blanket pellets,
- High burnup fuel skeleton, and l
Debris Filter Bottom Nozzle (DFBN). I 1
In addition to the VANTAGE + fuel design features listed above, the RFA design used at !
McGuire and Catawba will incorporate the following features that were licensed using the Fuel Criteria Evaluation Process (Reference 2-2) via Reference 2-3:
- Increased guide thimble and instmmentation tube OD (0.482 inch),
- Modified Low Pressure Drop (MLPD) structural mid-grids, and
- Modified Intermediate Flow Mixing (MIFM) grids.
1 2-1
The RFA design used at McGuire and Catawba will include the following additional features to 1
help mitigate debris failures:
i l
- Pre-oxide coating on the bottom of the fuel rods and I
- Protective bottom grid with longer fuel rod end-plugs.
{
The RFA design used at McGuire and Catawba will include the following features to help mitigate Incomplete Rod Insertion (IRI):
i fuel rods positioned on the botton nozzle and
- shorter fuel assembly.
l l
The four features listed above will be evaluated using the 10CFR50.59 process. '
One new feature that will be added to the McGuire and Catawba RFA design is a Quick Release Top Nozzle (QRTN). This top nozzle design is similar to the Reconstitutable Top Nozzle (RTN) design, but has been modified for easier removal. This design change will be licensed by Westinghouse using the Fuel Criteria Evaluation Process (Reference 2-2) and notification will be made to the NRC. i The Westinghouse RFA is designed to be mechanically and hydraulically compatible with the FCF Mark-BW fuel (Reference 2-4) that is currently used at McGuire and Catawba. The basic i I
design parameters of the RFA are compared to those of the Mark-BW fuel assembly in Table 2-1. j i
1 The IFM grids are non-structural members whose primary function is to promote mid-span flow mixing. Therefore, the design bases for the IFM grids are to avoid cladding wear and interactive damage with grids of the neighboring fuel assemblies during fuel handling. Westinghouse fuel with IFM grids has been flow tested both adjacent to another assc mbly with IFM grids and )
l adjacent to an assembly without IFM grids. There was no indication of adverse fretting wear of the fuel rods by the standard structural or IFM grids (Reference 2-5). No adverse fretting wear is j l
l 2-2 l l
I 1
)
expected in transition cores with the Westinghouse RFA design and Mark-BW fuel since the Mark-BW fuel is very similar to Westinghouse fuel assembly designs without IFM grids.
2.1 References 2-1 S. L. Davidson & T. L. Ryan, " Vantage + Fuel Assembly Reference Core Report",
WCAP-12610-P-A, April 1995.
2-2 S. L. Davidson (Ed.), " Westinghouse Fuel Criteria Evaluation Process", WCAP-12488-P-A, October 1994.
2-3 NSD-NRC-97-5189, Letter from N. J. Liparulo (Westinghouse) to J. E. Lyons (USNRC), .
" Transmittal of Response to NRC Request for Information on Wolf Creek Fuel Design Modifications", June 30,1997.
2-4 " Mark-BW Mechanical Design Report", BAW-10172P-A, December 1989.
2-5 S. L. Davidson (Ed.), " Reference Core Report Vantage 5 Fuel Assembly", WCAP-10444-P-A, September 1985.
i i
l 2-3 L1
Table 2-1 Comparison of Robust Fuel Assembly and Mark-BW Fuel Assembly Design Parameters 17x17 Robust Fuel 17x17 Mark-BW Fuel l Assembly Desien Assemb)v Desien Fuel Assembly Length,in.
Assembly Envelope,in.
Fuel Rod Pitch, in.
Fuel Rod Material l
Fuel Rod Clad OD,in.
Fuel Rod Clad nickness,in.
Fuel / Clad Gap, mils Fuel Pellet Diameter,in. !
Fuel Stack Height, in. l I
Guide Thimble Material Outer Diameter of Guide Thimbles, in.
(upper part)
Inner Diameter of Guide Thimbles, in.
(upper part)
Outer Diameter of Guide Himbles,in.
(lower part)
Inner Diameter of Guide Thimbles, in.
(lower part)
Outer Diameter ofInstrument Guide nimbles, in.
Inner Diameter ofInstrument Guide nimbles, in.
End Grid Material Intermediate Grid Material Intermediate Flow Mixing Grid Material l
2-4
3.0 CORE DESIGN 3.1 Introduction The nuclear characteristics of the Westinghouse RFA design and the Mark-BW fuel design are almost identical due to similar dimensional characteristics of the fuel pellet, fuel rod and cladding. As a result, the methods and core models used to perform transition and full core
)
analyses of the Westinghouse RFA design are the same as those currently licensed and employed !
in reload design analyses for McGuire and Catawba.
3.2 Reload Desien Methodology The development of core models, core operational imbalance limits and the evaluation of key physics parameters used to confirm the acceptability of UFSAR Chapter 15 accidents will be performed in compliance with the approved methodology defined in References 3-1 through 3-4.
Conceptual transition core designs using the Westinghouse RFA design have been evaluated and show that current reload limits remain bounding with respect to key physics parameters. In the event that one of the key parameters is exceeded, the evaluation process described in Reference 3-3 would be performed.
The introduction of the Westinghouse RFA design is not expected to change the magnitude of the nuclear uncertainty factors described in Reference 3-1. However, the use of zirconium di-boride Integral Fuel Burnable Absorbers (IFBA) is a fuel design change which is different from the bumable absorber types modeled in Duke's current benchmarking database. The NRC SER for Reference 3-1 requires Duke to re-benchmark the nuclear code package and assure that the nuclear uncertainties remain appropriate for significant changes in fuel design. While the introduction of the IFB A bumable absorber is not considered significant, the nuclear j uncertainties in Reference 3-1 were re-evaluated and confirmed to be bounding.
Duke explicitly modeled Segouyah Unit 2 Cycles 5,6, and 7 and performed statistical analysis of the nuclear uncertainty factors as described in Reference 3-1. These cores were chosen because 3-1
they are very similar to McGuire and Catawba and contained both IFBA and Wet Annular l
Burnable Absorber (WABA) fuel. The results of the statistical analysis are shown in Table 3-1 and show that the current licensed nuclear uncertainty factors bound those for the Westinghouse feel with a combination of IFBA and/or WAB A burnable absorbers. Boron concentrations, rod l worths, and isothermal temperature coefficients were also predicted and found to agree well with the measured data. A 10CFR50.59 USQ evaluation has been performed to demonstrate that the currently approved CASMO-3/ SIMULATE-3P methods and nuclear uncertainties are applicable to the Westinghouse RFA design described in this report.
In all nuclear design analyses, both the Westinghouse RFA and the Mark-BW fuel are explicitly modeled in the transition cores. When establishing Operating and RPS limits (i.e. LOCA kw/ft, l DNB, CFM, transient strain), the fuel specific limits or a conservative overlay of the limits are used.
The nuclear design related Technical Specification limits were reviewed for transition and full j
\
core reloads comprised of the Westinghouse RFA design. The only change required to the Technical Specifications is to replace the factor used to account for possible increases in FAH and Fq between flux maps with a burnup dependent factor (see Chapter 8 for additional details). ,
I I
In surrmary, the steady-statt "iyyics co.M methodology and nuclear uncertainty factors remain
)
unchanged for the transition to the Westinghouse RFA design. The evaluation of conceptual core designs with the RFA design indicate that key physics parameters assumed in the UFSAR I Chapter 15 accident analyses remain bounding. The introduction of the IFB A bumable poison j 1
design will require that the factor used to account for the possible increase in peaking over a 31 '
EFPD surveillance period be replaced by a burnup dependent factor (see Chapter 8).
I 3-2
3.3 References 3-1 " Design Methodology Using CASMO-3/ SIMULATE-3P," DI C-NE-1004A (Revision 1),
SER dated April 26,1996.
3-2. " Duke Power Company Nuclear Design Methodology for Core Operating Limits of Westinghouse Reactors," DPC-NE-201 IPA, March 1990.
3-3 " Multidimensional Reactor Transients and Safety Analysis Physics Parameters Methodology," DPC-NE-3001 PA, November 1991.
3-4 " Duke Power Company McGuire Nuclear Station, Catawba Nuclear Station Nuclear Physics Methodology," DPC-NF-2010A, June 1985.
I l
3-3
I Table 3-1 Nuclear Ur..ertainty Factors (Statistically combined factors 'without Engineering Hot Channel Factor)
Westinghouse Fuel with
! Parameter IFBA/WABA DPC-NE-1004A Fah 1.027 1.028 l
F, 1,049 1.053
[
F, 1,049 1.061 f
i l .
I 1
3-4
I l
l l i
i I 4.0 FUEL ROD ANALYSIS l
1 l
This chapter describes Duke Power's fuel rod mechanical reload analysis methodology for l Westinghouse fuel. The fuel rod analysis methodology discussed in this Chapter is essentially i identical to Westinghouse's approved methodology. The analyses will be performed using the NRC approved Westinghouse fuel perfortnance code, PAD, described in Section 4.1. Fuel rod mechanical analyses for Mark-BW fuel at McGuire and Catawba will continue to be performed i using the NRC-approved methodology given in Reference 4-12. '
( The fuel rods are designed to meet the requirements of 10CFR50, Appendix A, " General Design l Criteria" (Reference 4-1), specifically Criterion 10 " Reactor Design", which states: "The reactor core and associated coolant, control and protection systems shall be designed with appropriate margin to assure that specified acceptable fuel design limits are not exceeded during any -
condition of normal operation including the effects of anticipated operational occurrences."
t l To meet this requirement and the requirements of Section 4.2 of the Standard Review Plan (SRP)
(Reference 4-2), Westinghouse has established specific fuel design criteria associated with Condition 1 and II operation (Reference 4-3). Section 4.2 of this report describes each of the fuel rod design criteria which are evaluated as required by SRP 4.2 for Condition I and II operation.
l A description of the fuel rod analysis methodology which is used to show that the design criteria are met each cycle is also provided.
Detailed fuel rod design analyses consider parameters such as the pellet / clad diametral gap, the size and density of the pellet, the gas plenum volume, and the helium prepressurization. Using l the approved fuel performance models in PAD (Reference 4-4), the analyses also consider effects such as fuel densification and swelling, cladding creep, cladding corrosion, fission gas release and other physical properties which vary with burnup. The integrity of the fuel rods is ensured by designing the rods and operating the core to prevent excessive fuel temperatures, excestive fuel rod internal gas pressures, and excessive cladding stresses and strains. This is achieved by verifying that the conservative design criteria described in Section 4.2 are satisfied during Condition I and II events over the life of the fuel.
4-1
l l
l
! I The fuel rod analyses must consider the uncertainties associated with design models and variations in as-built dimensions. Due to the empirical basis of the perfonnance models used in j
the design codes (e.g., fission gas release, clad creep, etc.), there is variability in the data used for model validation. To have confidence that the extremes of the performance spectrum are l
covered, deviations from best estimate model projections must be accounted for. Each model i which has a significant effect on fuel rod performance includes uncertainty bands defined to l j
bound 95 % of the data. These uncertainty bands are used to define conservative upper bound uncertainty levels in the model predictions. These uncertainty levels are considered in the fuel I rod analyses, assuring that all fuel rods in a core will satisfy the design criteria.
The fuel rod analyses also consider the variations in rod dimensions and fuel fabrication characteristics. Typically drawing tolerances which are assumed to represent at least a 2 sigma i l
bound are used in fuel rod analyses. Actual as-built measurements and bounding values based on j
measured standard deviations may be used for critical fuel parameters. The typical method for including model, rod dimension, and fuel characteristic uncertainties is by statistical convolution. j j
The fuel rod for the RFA design is identical to the fuel rod for the VANTAGE + design, thus the licensed pin burnup for the Westinghouse RFA design is 60.000 mwd /mtU (Reference 4-3).
Using the Westinghouse Fuel Criteria Evaluation Process (FCEP) (Reference 4-13), the burnup limit can be increased to 62,000 mwd /mtU for specific reload cores.
Fuel rod analyses or evaluations to verify that a generic analysis is applicable must be performed for each reload cycle. Typically, generic analyses are completed the.t are expected to envelope the operation of future fuel cycles. The generic fuel rod analyses are then shown to be valid for each reload cycle design. This chapter describes the generic fuel rod analysis methods. In most
- cases, the generic analyses are bounding for each fuel cycle design and no new analyses are required. Cycle specific fuel rod analyses may be performed to obtain additional margin.
l l 4-2 l
4.1 Computer Code The PAD fuel performance code (Reference 4-4) is the main code used for evaluating fuel rod performance. PAD iteratively calculates the interrelated effects of temperature, pressure, cladding clastic and plastic behavior, cladding corrosion, fission gas release, and fuel densification and swelling as a function of time and power. PAD evaluates the power history of a rod as a series of steady-state power levels with instantaneous changes from one power level to another.
PAD divides the fuel rod into several axial segments and each segment is assumed to operate at a constant set of conditions over its length. Fuel densification and swelling, cladding stresses and ,
strains, temperatures, burnup and fission gas release are calculated separately for each axial segment and the effects are integrated to obtain the overall fission gas release and rod internal pressure. The coolant temperature rise along the rod is calculated based on the flow rate and axial power distribution and the cladding surface temperature is calculated considering the effects of corrosion and the possibility oflocal boiling. j PAD considers the fuel pellet as a solid cylinder with allowances for dishing, chamfering, and pellet chipping. To calculate thermal expansion, fuel densification and swelling, and fission gas release, the pellet is divided into equal volume concentric rings and each ring is assumed to be at its average temperature during a given time step. Axial and radial thermal expansion, swelling and densification are determined for each ring and these effects are integrated over the entire fuel rod to calculate the length of the fuel column and the void volume to calculate the rod internal pressure.
l The current version of the PAD code is PAD 3.4 (Reference 4-4). This version of the code j includes an updated fission gas release model, fuel densification and swelling models, and l cladding creep model. The PAD code has been certified for use in safety-related analyses according to Duke Power's Quality Assurance program. When any new versions of the PAD !
l code are submitted to the NRC by Westinghouse, Duke Power plans to use the new version after it is approved for licensing analyses.
l 4-3 L
f i
4.2 Fuel Rod Design Bases and Analyses The design bases for the RFA design that will be used in McGuire and Catawba are identical to those given in Reference 4-3 for Vantage + fuel. 'Ihe fuel rod design bases and analysis methodologies are described below.
4.2.1 Fuel Rod Intemal Pressure The fuel rod internal pressure design basis is that the fuel system will not be damaged due to excessive fuel rod internal pressure (Reference 4-3 and 4-6). The internal pressure of the lead rod in the reactor will be limited to a value below that which could cause (1) the diametral gap to increase due to outward clad creep during steady-state operation and (2) extensive DNB propagation '.o occur.
4.2.1.1 Analysis Part I of this design basis precludes the cladding outward creep rate from exceeding the fuel solid swelling rate, and, thus, ensures that during steady-state operation the fuel-cladding gap will not re-open following contact, or increase in size. The PAD code is used to predict fuel rod internal pressures that are used to verify that the fuel rod internal pressure <*esign basis is met.
The rod average burnup at which the diametral gap begins to increase due to the outward cladding creep rate is calculated. This allowable rod bumup is compared to predicted rod bumups for each reload design to confirm that the rod internal pressure criterion is met for all of the fuel.
A bounding pin power history, similar to that shown in Fig. 4-1, is used to perform a generic rod internal pressure analysis. A cycle-specific rod internal pressure analysis may be performed using predicted limiting pin power histories if the bounding power history does not envelope the pin powers for a future core design. The transient gas release contribution to the rod internal pressure must be included in the rod internal pressure analyses. Both Condition I axial xenon oscillations and Condition II overpower transients are considered in calculating the iod internal pressure.
4-4
Sensitivity studies have been performed to determine the design parameters and PAD models which are the most significant contributors to the uncertainty in the rod internal pressure. An upper bound rod internal pressure is calculated to account for the impact of possible variations in design parameters or models. The bounding pressure is compared to a lower bound steady-state pressure limit.
Part 2 of the rod internal pressure design basis deals with DNB propagation, which is discussed in Reference 4-6. The current methodology for calculating the frequency and expected location of fuel rods experiencing both DNB and internal pressure greater than the reactor coolant system pressure is consistent with that used for the evaluations documented in Reference 4-6. For each rod that is both in DNB and above system pressure, the number of additional rods in DNB due to propagation effects are calculated based on whether the neighboring rods are in DNB or above system pressure. A fuel rod which is both in DNB and above system pressure is assumed to balloon at the location of DNB. When the ballooned clad contacts its neighboring rods, it is assumed that these rods will also experience DNB as a result of the flow blockage. If one of these rods is also above system pressure, it would also balloon to contact its neighboring rods.
This process is assumed to continue if any of the neighbor rods are above system pressure. The total number of rods in DNB initially, rods above system pressure, rods both in DNB and above system pressure, and rods in DNB due to propagation are calculated.
4.2.2 Cladding Stress The cladding stress design basis is the fuel system will not be damaged due to excessive fuel cladding stress (Refemnce 4-3 and 4-9). The volume average effective stress calculated with the Von Mises equation considering interference due to uniform cylindrical pellet cladding contact, caused by thermal expansion, pellet swelling and uniform cladding creep, and pressure differences, is less than the ZIRLO* 0.2 c/c offset yield stress, with due consideration of temperature and irradiation effects under Condition I and II modes of operation. While the cladding has some capability for accommodating plastic strain, the yield stress has been established as a conservative design limit.
4-5
4.2.2.1 Analysis Excessive clad stress can arise due to rapid local power increases such that clad creep cannot accommodate the pellet thermal expansion. The clad stress criterion is applied to the volume average effective stress which occurs as a resp'. of a Condition II transient local power increase.
The primary mechanism which increases the .: lad stresses during a Condition II transient, relative to the steady-state stresses, is the differential thermal expansion between the pellet and the cladding.
l_ ~
For each reload design, the allowable changes in local linear heat rate (A kw/ft) as a function of burnup are compared to predicted peaking changes that result from either Condition I or 11 events.
4.2.3 Cladding Strain The cladding strain design basis is that the fuel system will not be damaged due to excessive fuel cladding strain (Reference 4-3 and 4-9). The design limit is that during steady-state operation, the total plastic tensile creep strain due to uniform cladding creep and uniform fuel pellet expansion associated with fuel swelling and thermal expansion is less than 1% from the unirradiated condMon. The acceptance limit for fuel rod cladding strain during Condition II events is that the total tensile strain due to uniform cylindrical pellet thermal expansion is less than 1% from the pre-transient value (Reference 4-2).
4-6
L 4.2.3.1 Analysis The intent of this criterion is to minimize the potential for clad failure due to excessive clad straining. This criterion addresses slow strain rate mechanisms where the effective clad stress I
never reaches the yield strength due to stress relaxation. Clad strain allowable local power limits (A kw/ft) are calculated using PAD and the methodology discussed above for calculating clad stress local power limits. Analyses have generally shown that the transient clad stress analyses are more limiting than the transient clad strain analyses (i.e., the clad stress A kw/ft limits are typically more restrictive than the clad strain A kw/ft limits).
4.2.4 Cladding Fatigue The cladding fatigue design basis is that the fuel system will not be damaged due to excessive clad fatigue (Reference 4-3 and 4-9). The fatigue life usage factor is limited to less than 1.0 to prevent reaching the material fatigue limit.
4.2.4.1 Analysis A cladding fatigue analyns is performed to consider the accumulated effects of short term, cyclic, cladding stress and strain resulting primarily from daily load follow operation. The accumulated effects of cyclic strains associated with normal plant shutdowns and retums to full power are also considered.
The fatigue model in PAD calculates the low cyclic fatigue and the fatigue life fraction of a fuel rod during load follow operation, as a function of time and irradiation history. The Langer-O'Donnell low cyclic fatigue model (Reference 4-7) constitutes the basic approach used in the fatigue analysis. The empirical factors used in the Langer-O'Donnell fatigue model have been modified to conservatively bound the results of Westinghouse test programs presented in Reference 4-8. The design equations follow the concepts of the fatigue design criterion given in th- ASME Code, Section III:
4-7
The calculated pseudo-stress amplitude (S.) is multiplied by 2 to obtain the allowable number of cycles (Nr)
The allowable cycles for a given S is five percent of Nror a safety factor of 20 on the number of cycles.
The lower of the two allowable number of cycles is selected and the cumulative fatigue life fraction is then calculated as:
nu /Na < l.0 where:
Ng = number of cycles of mode k Na = number of allowable cycles PAD is used to analyze a spectrum of pin power histories to determine the fatigue life.
4.2.5 Fuel Clad Oxidation and Hydriding The fuel clad oxidation and hydriding design basis is that fuel damage will not occur due to excessive clad oxidation or hydriding (Reference 4-3). To limit metal-oxide formation to acceptable values, the ZIRLO* metal-oxide interface temperature is limited
] (Reference 4-3). The clad and structural component hydrogen pickup is limited to [ ] (Reference 4-3) at end oflife to preclude loss of ductility due to hydrogen embrittlement by the formation of zirconium hydride platelets.
4.2.5.1 Analysis A spectrum of pin power histories, including a bounding power history similar to that shown in Fig. 4-1, are analyzed to verify that the cladding metal-oxide interface temperature limits are met 4-8
during steady-state operation and during Condition 11 local power increases. For each steady-i state power history, the temperature of the metal-oxide interface is calculated. The oxide layer on the fuel is calculated using the ZIRLO corrosion model described in Reference 4-3. At various times during the steady-state depletion, Condition II local power increases are simulated.
The local power is increased until the cladding metal-oxide interface temperature is equal to the transient cladding temperature limit. An analysis is performed foreach reload which verifies that the local power limit associated with the transient cladding temperature limit is not exceeded during Condition II events (Reference 4-11).
The methodology for calculating the hydrogen pickup of the cladding is the same as that described above for calculating the metal-oxide interface temperature. In addition to the zire-oxide buildup on the cladding, the hydrogen pickup resulting from the corrosion process is calculated. Corrosion and percent metal wastage for the grids and thimbles is also calculated.
4.2.6 Fuel Temperature The fuel temperature design basis is that fuel rod damage will not occur due to excessive fuel temperatures (Reference 4-3). The fuel system and protection syst m are designed to assure that for Condition I and Il events, the calculated centerline fuel temperature does not exceed the fuel melting temperature. The melting temperature of unirradiated UO2 is taken as 5080 F, decreasing by 58 F per 10,000 mwd /mtU of fuel burnup (Reference 4-3). A centerline fuel temperature of 4700 F has been selected by Westinghouse as the design limit for fuel temperature analyses, References 4-9 and 4-10.
4.2.6.1 Analysis The PAD 3.4 code (Reference 4-4) is used to verify that the fuel temperature design limit is met.
Using a fuel centerline temperature limit of 4700 *F covers both the reduction in melt temperature with burnup and manufacturing and modeling uncertainties. PAD is used to calculate the fuel centerline temperature and the local linear heat rate to prevent fuel melting or linear heat rate to melt (LHRTM). As explained in Reference 4-11 an analysis is performed for l 4-9
, each reload which verifies that this local power limit is not exceeded for Condition I and II i
t events, j 4.2.7 Fuel Clad Flattening From Reference 4-3, the design basis for fuel clad flattening is that fuel rod failures will not occur due to clad flattening.
4.2.7.1 Analysis Westinghouse demonstrated in Reference 4-5 that clad flattening will not occur for current Westinghouse fuel designs. Based on post irradiation examination and in-core flux data Westinghouse confirmed that significant axial gaps in the fuel column due to densification will not occur for current Westinghouse fuel. Therefore, it was concluded that clad flattening will not occur.
A new clad flattening evaluation is required only if any of the following fuel rod design parameters change: cladding creep properties, cladding thickness, fuel densification, rod prepressure, and as-fabricated pellet-clad gap. All of these parameters are related to the fuel design itself; they are not affected by a particular reload core design. For each new region of fuel; the cladding thickness, fuel rod prepressure, and as-fabricated pellet-clad gap will be verified to be within the range of parameters considered in Reference 4-5.
4.2.8 Fuel Rod Axial Growth From Reference 4-3, the fuel rod growth design basis is that the fuel rods will be designed with adequate clearance between the fuel rod end plugs and the top and bottom nozzles to accommodate the difference in the growth of the fuel rods and the growth of the fuel assembly.
The Westinghouse RFA was designed to assure that there is no interference between the fuel l rods and the fuel assembly top and bottom nozzles during the design life of the fuel.
4-10
(
) 4.2.8.1 Analysis l The fuel rod growth model described in Reference 4-4 is used to show that the fuel rod growth criterion is met. The rod growth analysis assumes upper bound fuel rod growth, lower bound fuel assembly growth, minimum initial fuel rod to nozzle gap, upper bound rod fast fluence, and nominal differential thermal expansion between the fuel rod cladding and the fuel assembly structure. A generic analysis is performed to calculate the maximum allowable rod average burnup for which the rod to nozzle gap is zero. For the current RFA design, the allowable rod burnup with respect to the rod growti criterion is greater than the licensed burnup limit of 60,000 mwd /mtU. Using the Westinghouse Fuel Criteria Evaluation Process (FCEP) (Reference 4-13),
the burnup limit can be increased to 62,000 mwd /mtU for specific reload cores.
I i
4 l
t 4-11
(
4.3 References 4-1 Title 10, Chapter 1, Code of Federal Regulations - Energy, Part 50, " Domestic Licensirs l of Production and Utilization Facilities", Appendix A," General Design Criteria for Nuclear Power Plants".
4-2 "Section 4.2, Fuel System Design", Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants - LWR Edition, NUREG-0800, Rev. 2, US Nuclear Regulatory Commission, July 1981.
4-3 S. L. Davidson & T. L. Ryan," Vantage + Fuel Assembly Reference Core Report",
WCAP-12610-P-A, April 1995.
4-4 Weiner, R. A., et al., " Improved Fuel Performance Models for Westinghouse Fuel Rod Design and Safety Evaluations", WCAP-10851-P-A, August 1988.
4-5 P. J. Kersting, et al., " Assessment of Clad Flattening and Densification Power Spike Factor Elimination in Westinghouse Nuclear Fuel", WCAP-13589-A, March 1995.
4-6 Risher, D., et al., " Safety Analysis for the Revised Fuel Rod Internal Pressure Design Basis", WCAP-8963-P-A, August 1978.
4-7 W. J. O'Donnell and B. F. Langer, " Fatigue Design Basis of Zircaloy Components",
Nuclear Science and Engineering. 20,1-12,1964.
4-8 S. L. Davidson and J. A. Iorii. " Reference Core Report 17x17 Optimized Fuel Assembly", WCAP-9500-P-A, May 1982.
4-9 S. L. Davidson (ed.), et al., " Extended Buraup Evaluation of Westinghouse Fuel",
WCAP-10125-P-A, December 1985.
4-12
4-10 S. L. Ellenberger, et al., " Design Bases for the Thermal Overpower Delta-T and Thermal Overtemperature Delta-T Trip Functions", WCAP-8745-P-A, September 1986.
, 4-11 " Duke Power Company Nuclear Design Methodology for Core Operating Limits of 1
l Westinghouse Reactors". DPC-NE-2011P-A, March 1990.
4-12 " Duke Power Company Fuel Rod Mechanical Reload Analysis Methodology Using TACO 3", DPC-NE-2008P-A, SER dated April 3,1995.
4-13 S. L. Davidson (Editor), " Westinghouse Fuel Criteria Evaluation Process", WCAP-12488P-A, October 1994.
4-13
- - 0 y -
.. 5 r -
o t - - -
s i - - - - - -
H r -
e -
.. 4 0
w -
o -
P -
i n -
P
- a -
g -
.. 0 G a u - -
n
- - - 3 i
d u- -
n -
~
u
- - - - - ~
o B
l - - - - - -
a -
.. 0 2
i c -
p -
y -
T -
1 ~
~
4 -
. 0 1
i g -
F c.. -
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 5 4 3 2 1 0 9 1 1 1 1 1 1 1 0 y30c. c a. .
5.0 THERMAL-HYDRAULIC ANALYSIS Steady-state thermal-hydraulic analyses for the Westinghouse RFA design will be performed using the NRC approved methodology given in References 5-1 and 5-4. Reference 5-1 describes the VIPRE-01 core thermal-hydraulic models used for steady state analyses at McGuire and Catawba. The only changes necessary to perform core thermal-hydraulic analyses for the Westinghouse RFA design are to specifically model the fuel (dimensions, form loss coefficients, etc.) and to use the WRB-2M critical heat flux (CHF) correlation (Reference 5-2). The RFA design, VIPRE-01 models, and the WRB-2M CHF correlation are discussed in Sections 5.1,5.2, and 5.3, respectively.
DPC-NE-2005P-A (Reference 5-4) describes Duke Power's NRC-approved methodology for calculating a Statisti:al Core Design (SCD) DNBR limit for application to pressurized water reactors. Individual appendices to the report list information necessary to complete the calculations for specific plants and fuel types. This includes the fuel data for the VIPRE-01 model, parameter uncertainties, the CHF correlation, and the range of conditions analyzed. The remainder of Chapter 5 is written in the same format as an appendix to Reference 5-4. Sections 5.1 through 5.3 list the plant specific data, models, and CHF correlation. Section 5.4 lists the j range of statepoint conditions analyzed and Section 5.5 describes the key parameters and associated uncertainties. The statistical design limit, or SDL, which will be used for licensing analyses for Westinghouse Robust fuel at McGuire and Catawba is discussed in Section 5.6.
Section 5.7 discusses how the impact of the geometric and hydraulic differences between the
! resident Mark-BW fuel and the Westinghouse RFA design is addressed and determines the SDL
! for RFA/ Mark-BW transition cores.
l Unless otherwise noted, all VIPRE-01 modeling inputs listed in Reference 5-1 for the 17x17 fuel at McGuire and Catawba are unchanged. The thermal-hydraulic SCD analysis discussed in this chapter was performed using the approved methodology given in the main body of Reference 5-4.
I I
5-1
5.1 Plant Snecific Data This analysis is for the McGuire and Catawba plants (four-loop Westinghouse PWR's) with the i
RFA design. The Robust fuel design includes 0.374 OD fuel rods and non-stmetural Intermediate Flow Mixing (IFM) grids in the upper three spans to improve DNB performance.
This design also includes the fuel reliability features of a debris filtering bottom and a protective grid between this aozzle and the first strr:tural grid. See Chapter 2 of this report for a complete description of the fuel design.
\
The parameter uncertainties and statepoint ranges were selected to bound the McGuire and Catawba unit and cycle-specific values (see Sections 5.4 and 5.5).
5.2 Thermal-Hydraulic Code and Model The VIPRE-01 thermal-hydraulic computer code described in Reference 5-3 and the McGuire/ Catawba eight channel model approved in Reference 5-1 are used in this analysis. The reference pin power distribution based on a 1.60 peak pin from Reference 5-1 was used. The VIPRE-01 models approved in Reference 5-1 for the Mark-BW fuel are used to analyze th , RFA design with the following changes:
- 1) The RFA design geometry information is listed in Table 5-1. Applicable form loss coefficients as per the vendor were used in the models. Alm, the axial noding was adjusted to be compatible with the Westinghouse WRB-2M CHF correlation.
- 2) The bulk void fraction model was changed from the Zuber-Findlay model to the EPRI model. Correspondingly, the subcooled void model was changed from the Levy to EPRI model.
The Zuber-Findlay bulk void model is applicable only to qualities below approximately 0.7 (void fractions of 0.85) and is discontinuous at higher values (Reference 5 3). The EPRI bulk void 5-2
model is essentially the same as the Zuber-Findlay bulk void model except for the equation used m calculate the drift velocity (Reference 5-3). This eliminates the discontinuity at high qualities and void fractions. Therefore, the EPRI model covers the full range (i.e., void fraction range,0 -
1.0) of void fractions required for performing DNB calculations. Also, for overall void model I compatibility, the subcooled void model was changed from the Levy model, as specified in Reference 5-1, to the EPRI correlation.
To evaluate the impact of changing bulk void models on DNB predictions, fifty-one RFA critical heat fir.x test data points (Reference 5-2) were compared using both the Levy /Zuber-Findlay and EPRl/EPRI subcooled void / bulk void model combinations in VIPRE-01. These data points cover a pressure range of 1519 to 2426 psia and an inlet temperature range 397.4 to 617.6 F.
2 The mass flux at the MDNBR location varied from 1.48 to 3.02 Mlbm/hr-ft . The void fraction at the MDNBR location varied from 0.309 to 0.697. The equilibrium quality at the MDNBR location varied from 0.07 to 0.254. The results of this comparison are as follows:
L evv/Zuber-Findlav EPRI/EPRI Minimum DNBR (Avg.) 1.029 1.028 The minimum DNBR results show a minimal difference of 0.1% (0.001 in DNB). Therefore, the EPRI bulk void model and EPRI subcooled void correlation will be used in RFA analyses.
5.3 Critical Heat Flux Correlation The WRB-2M critical heat flux correlation described in Reference 5-2 is used for all statepoint analyses. This .:orrelation was developed by Westinghouse for application to the RFA design.
As discussed in Reference 5-2 the WRB-2M correlation was developed with the VIPRE-01 thermal-hydraulic computer code. This correlation was programmed into the Duke Power version of VIPRE-01 and will be used in all DNBR calculations for the RFA design, except for the steam line break transient (see Section 6.2.2).
5-3
s
/
5.4 Statenoints The statepoint conditions evaluated in this analysis are listed in Table 5-2. These statepoints i
cover the range of conditions to which the statistical DNBR limit will be applied. The range of key parameter values evaluated in this analysis are listed on Table 5-5.
5.5 Kev Parameters and Uncertainties The key parameters and their uncertainty magnitude and associated distribution used in this analysis are listed on Table 5-3. The uncertainties were selected to bound the values calculated for each parameter at McGuire and Catawba.
5.6 DNB Statistical Desien Limit The statistical DNBR value for each statepoint evaluated is listed on Table 5-4. Section 1 of Table 5-4 contains the 500 case mns and Section 2 contains the 5000 case runs. The number of cases was increased from 3000 to 5000 as described in Attachment 1 of the main body of Reference 5-4. The DNBRs calculated for all of the statepoints are normally distributed. As shown in Section 2 of Table 5-4 the maximum statepoint statistical DNBR value is [ ].
Therefore, the statistical design limit (SDL) using the WRB-2M CHF correlation for the RFA design at McGuire/ Catawba is conservatively determined to be [ ].
5.7 Transition Cores A transition core mod:1 is used to determine the impact of the geometric and hydraulic differences between the resident FCF Mark-BW fuel and the Westinghouse RFA design. The 8 channel model described in Reference 5-1 is used to evaluate the impact of transition cores containing the RFA design. In Figure 5 of Reference 5-1, the RFA design is used instead of Mark BW fuel. Therefore, the limiting assembly (Channels 1 through 7)is modeled as the RFA design and the remainder of the core (Channel 8)is modeled as Mark-BW fuel. The transition core analysis models each fuel type in their respective locations with the correct geometry. The 5-4
form loss coefficients for each fuel design are input so the effect of crossflow out of the IFM grid spans in the limiting channel is calculated.
A transition core DNBR penalty is determined for the RFA design using the 8 channel RFA/ Mark-BW transition core model. A conservative DNBR penalty is applied for all DNBR analyses for RFA/ Mark-BW transition cores.
To evaluate the impact of the transition core on the statistical DNBR limit, the most limiting full core statepoint (Statepoint 12 on Table 5-4) was evaluated using the 8 channel transition core model. This case is designated as statepoint 12TR in Sections 1 and 2 of Table 5-4. The statistical DNBR calculated using the transition core model (statepoint 12TR) is slightly greater than the Statistical DNBR value for the full RFA core (statepoint 12) at both the 500 and 5000 cases levels. As shown in Section 2 of Table 5-4, this value is still less than [ ]. Therefore, the statistical design limit of[ ] is bounding for RFA/ Mark-BW transition cores as well as full RFA cores.
5.8 References 5-1 DPC-NE-20(MP-A, McGuire and Catawba Nuclear Stations Core Thermal-Hydraulic Methodology Using VIPRE-01 Rev 1, February 1997.
5-2 WCAP-15025-P, Modified WRB-2 Correlation, WRB-2M, for Predicting Critical Heat Flux in 17x17 Rod Bundles with Modified LPD Mixing Vane Grids, Westinghouse Energy Systems, Febrrary 1998.
5-3 VIPRE-01: A Thermal-Hydraulic Code For Reactor Cores, EPRI NP-2511-CCM-A, Vol.
1-4, Battelle Pacific Northwest Laboratories, August 1989.
5-4 DPC-NE-2005P-A, Duke Power Company Thermal-Hydraulic Statistical Core Design Methodology, Rev 1, November 1996.
5-5
Table 5-1 RFA Design Data (TYPICAL)
GENERAL FUEL SPECIFICATIONS Fuel rod diameter, inches (Nominal) 0.374 Guide tube diameter, inches (Nominal) 0.482 Fuel rod pitch, inches (Nominal) 0.496 Fuel Assembly pitch, inches (Nominal) 8.466 Fuel Assembly length, inches (Nominal) 159.8 GENERAL FUEL CHARACTERISTICS Component Material Number Location /rvne Grids Inconel 1 Lower Protective Inconel 2 Upper and Lower Non-Mixing Vane ZIRLO 6 Intermediate Mixing Vane ZIRLO 3 Intermediate Flow Mixing (Non-structural)
Nozzles 3(MSS 1 Debris Filtering Bottom 304SS 1 Removable Top 5-6
Table 5-2 McGuire/ Catawba SCD Statepoints, WRB-2M Correlation Core Inlet Stpt Power
- RCS Flow ** Pressure Temperature Axial Peak Radial Peak
& (ch RTP) (K enm) (osia) (*F) (F. @ Z) (FAHj 3
2 3
4 5
6 7
8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 12TR* *
- 100c/c RTP = 3411 Megawatts Thermal Mass Dow rate should be calculated using the given core inlet temp.
TR - transition core model f
5-7
Table 5-3 McGuire/ Catawba Statistically Treated Uncertainties Parameter Uncenainty / Standard Deviation Type Of Distribution
' Core Power * +/- 2% /1.22% Normal Core Flow Measurement +/- 2.2% /1.34% Normal Bypass Flow +/- 1.5% Uniform Pressure +/- 30 psi Uniform Temperature +/- 4 deg F . Uniform i- N AH Measurement +/- 4.0% / 2.43% Normal E
Fg- +/- 3.0% /1.82% Normal l
Spacing +/- 2.0% /1.22% Normal Fz +/- 4.41 % / 2.68% Normal Z +/- 6 inches Uniform DNBR Correlation +/- 10.73% / 6.52% Normal Code /Model [ ] Normal
- Percentage of 100% RTP (3411 MWth) 5-8 t.
Table 5-3 (Continued)
McGuire/ Catawba Statistically Treated Uncertainties Parameter Justification Core Power The core power uncenainty was calculated by statistically combining the uncenainties of the process indication and control channels. The uncenainty is calculated from normally distributed random error terms such as sensor calibration accuracy, rack drift, sensor drift, etc. combined by the square root sum of squares method (SRSS). Since the uncertainty is calculated from normally distributed values, the parameter distribution is also normal.
Core Flow Measurement Same approach as core power.
Bypass Flow The core bypass flow is the parallel core flow paths in the reactor vessel (guide thimble cooling flow, head cooling flow, fuel assembly / baffle gap leakage, and hot leg outlet nozzle gap leakage) and is dependent on the driving pressure drop. Parameterization of the key factors that control
.4P, dimensions, loss coefficient correlations, and the effect of the uncertainty in the driving AP on the flow rate in each flow path, was performed. The dimensional tolerance changes were combined with the SRSS method and the loss coefficient and driving AP uncertainties were conservatively added to obtain the combined uncertainty. This uncertainty was conservatively applied with a uniform distribution.
Pressure The pressure uncenainty was calculated by statistically combining the uncertainties of the process indication and control channels. The uncenainty is calculated from random error terms such as sensor calibration accuracy, rack drift, sensor drift, etc. combined by the square root sum of squares method. The uncertainty distribution was .
conservatively applied as uniform. I Temperature Same approach as pressure.
Fg N l Measurement This uncertainty is the measurement uncenainty for the movable incore instruments. A measurement uncertainty can arise from instrumentation
! drift or reproducibility error, integration and location error, error associated with the burnup history of the core, and the error associated with the conversion ofinstrument readings to rod power. The uncertainty distribution is normal.
5-9 l 1
Table 5-3 (Continued)
McGuire/ Catawba Statistically Treated Uncertainties Pararneter Justification FgE This uncenainty accounts for the manufacturing variations in the variables affecting the heat generation rate along the flow channel. This conservatively accounts for possible variations in the pellet diameter, density, and U235 enrichment. This uncenainty distribution is normal and was conservatively applied as one-sided in the analysis to ensure the MDNBR channellocation was consistent for all cases.
Spacing This uncenainty accounts for the effect on peaking of reduced hot channel flow area and spacing between assemblies. The power peaking gradient becomes steeper across the assembly due to reduced flow area and spacing. This uncertainty distribution is normal and was conservatively applied as one-sided to ensure consistent MDNBR channel location.
FZ This uncertainty accounts for the axial peak prediction uncenainty of the physics codes. The uncenainty distribution is applied as normal.
Z This uncertainty accounts for the possible error in interpolating on axial peak location in the maneuvering analysis. The uncenainty is one of the physics code's axial nodes. The uncenainty distribution is conservatively applied as uniform.
DNBR Correlation This uncertainty accounts for the CHF correhtion's ability to predict DNB.
The uncertainty distribution is applied as nonnal.
Code /Model This uncertainty accounts for the thermal-hydraulic code uncertainties and offsetting conservatism. This uncertainty also accounts for the small DNB prediction differences between the various model sizes. The uncertainty distribution is applied as normal.
5-10
Table 5-4 l McGuire/ Catawba Statepoint Statistical Results SECTION 1 l
WRB-2M Critical Heat Flux Correlation 500 Case Runs Coefficient Statistical Statenoint # Mean g of Variation DNBR 1
2 3
4 5
6 7
8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 12TR*
TR - transition core model 5-11 l
I i Table 5-4 (Continued) l l
McGuire/ Catawba Statepoint Statistical Results 1
SECTION 2 l WRB-2M Critical Heat Flux Correlation l
t 5000 Case Runs Coefficient Statistical Statenoint # Mean g ofVariation DNBR 7
11
! 12 12TR*
l TR - transition core model i
l l
I l
i i
I I
l l
l l
5-12 l
Table 5-5 McGuire/ Catawba Key Parameter Ranges WRB-2M CHF Correlation Parameter Maximum Minimum Core Power * (% RTP)
Pressure (psia)
T inlet (deg. F)
RCS Flow (Thousand GPM)
FAH, Fz, Z i
l 100% RTP = 3411 Megawatts Thermal All values listed in this table are based on the currently analyzed statepoints (Table 5-2). Ranges are subject to change based on future statepoint conditions.
5-13 i
l l
1
l 6.0 UFSAR ACCIDENT ANALYSES l
DPC-NE-3000-PA, " Thermal-Hydraulic Transient Analysis Methodology"(Reference 6-1),
DPC-NE-3001-PA," Multidimensional Reactor Transients and Safe:y Analysis Physics Parameters Methodology"(Reference 6-2), and DPC-NE-3002-A,"UFSAR Chapter 15 System Transient Analysis Methodology" (Reference 6-3) describe the Duke Power NRC-approved models and methodology for analyzing UFSAR Chapter 15 Non-LOCA transients and accidents.
l DPC-NE-3004-PA," Mass and Energy Release and Containment Response Methodology"
! (Reference 6-4), describes the Duke Power NRC-approved models and methodology for analyzing UFSAR Chapter 6.2 mass and energy release accidents and containment response.
UFSAR Chapter 15 non-LOCA analyses will continue to be performed according to the methodologies described previously in Reference 6-1, Reference 6-2, and Reference 6-3, except as noted in Sections 6.1 6.3, respectively. LOCA mass and energy release analyses (UFSAR Chapter 6.2) will continue to be performed according to the methodology described in Reference i l
6-4, except as noted in Section 6.4. LOCA analyses (UFSAR Chapter 15.6.5) will be performed {
by Westinghouse as described in Section 6.5.
6.1 Thermal-Hydraulic Transient Analysis Methodology (DPC-NE-3000) l DPC-NE-3000-PA, " Thermal-Hydraulic Transient Analysis Methodology"(Reference 6-1),
serves as the Duke Power Company response to Generic Letter 83-11," Licensee Qualification for Performing Safety Analyses in Support of Licensing Action," which requires that licensees !
performing their own safety analyses demonstrate their analytical capabilities. Reference 6-1 describes the RETRAN-02 (Reference 6-21) system transient thermal-hydraulic models, and the VIPRE-01 (Reference 6-20) core thermal-hydraulic models developed for Oconee, McGuire, and Catawba Nuclear Stations. The previous comparisons of computer code results to experimental data, plant operational data, and other benchmarked analyses, continue to demonstrate the analytical capability to perform non-LOCA transient thermal-hydraulic analyses. Changing from Mark-BW to the RFA design does not affect this conclusion.
6-1
A review of Reference 6-1 indicates that only portions of Chapter 3 (McGuire/ Catawba Transient Analyses) currently do not support the RFA design from a technical standpoint. Chapters 2 and 4 pertain to Oconee Nuclear Station only, and therefore remain unaffected. Chapter 5 pertains to McGuire/ Catawba RETRAN benchmark analyses, which continue to demonstrate analytical capability to perfotm non-LOCA transient thermal-hydraulic analyses regardless of fuel type.
Chapters 1 (Introduction) and 6 (Summary) are affected from an editorial standpoint only.
6.1.1 Plant Description (Section 3.1 in DPC-NE-3000) 1 i
- The only difference with respect to the plant description will be the change from Mark-BW fuel l
to the RFA design. Chapter 2 ot this report gives a complete description of the RFA design.
l 6.1.2 McGuire/ Catawba RETRAN Model (Section 3.2 in DPC-NE-3000)
Volumes [ ] in the primary system nodalizatior scheme represent the reactor core region from the [ .] Dimensional changes due to the change to j the RFA design will require minor changes to these volume calculations, as well as associated junction and heat conductor calculations.
6.1.3 McGuire/ Catawba VIPRE Model (Section 3.3 in DPC-NE-3000)
The McGuire/ Catawba simplified [ ] channel model in Reference 6-1 is used for analyzing the RFA design. As described in Chapter 5, the reference radial pin power distribution remains unchanged, but the peak pin is increased from 1.50 to 1.60 and the WRB-2M CHF correlation (Reference 6-5) and the SCD limit developed in Chapter 5 are used. The axial node size is adjusted to be compatible with the WRB-2M CHF correlation. The RFA design geometry is listed in Table 5-1 and applicable form loss coefficients are used. The remaining code inputs and options remain identical to that originally approved in Reference 6-1.
No transitit ore transient analyses are performed as the results detennined in Chapter 5 also apply for transient analyses. As discussed in Refert. 1, the [ ] channel model used for I
6-2
transient analyses was originally developed with additional conservatism over the 8 channel model used for steady-state analyses to specifically minimize the impact of changes in core reload design methods or fuel assembly design. Should it be determined in the future that transition core transient analyses are warranted, they will be performed accordingly.
}
6.2 Multidimensional Reactor Transients and Safety Analysis Physics Parameters Methodology (DPC-NE-3001)
DPC-NE-3001-PA, " Multidimensional Reactor Transients and Safety Analysis Physics Parameters Methodology"(Rc,Terence 6-2), describes the Duke Power Company methodologies for simulating the UFSAR Chapter 15 events characterized by multidimensional reactor transients (rod ejection, steam line break, and dropped rod), and for systematically confirming that reload physics parameters important to Chapter 15 transients and accidents are bounded by values assumed in the licensing analyses (the Safety Analysis Physics Parameters (SAPP) methodology). The SAPP methodology remains unchanged when analyzing the RFA design.
Thermal-hydraulic changes for analyzing the RFA design in rod ejection, steam line break, and dropped rod accidents are discussed in the sections that follow.
6.2.1 Rod Ejection The changes presented in Section 6.1 also apply to the rod ejection accident. The nuclear analysis of the rod ejection accident using SIMULATE-3K is presented in Section 6.6. The remainder of the rod ejection thermal-hydraulic methodology presented in Reference 6-2 remains unchanged.
6.2.2 Steam Line Break The changes presented in Section 6.1 also apply to steam line break, with the exception of the CHF correlation. Since the WRB-2M CHF correlation pressure range of applicability is not acceptable for steam line break analyses (see Chapter 5 of this repon for ranges of applicability),
the W3-S CHF correlation will continue to be used as originally documented in Reference 6-2, 6-3
(
l The remainder of the steam line break thermal-hydraulic methodology presented in Reference 6-2 remains unchanged, except for the selection of subcooled and bulk void models for offsite power lost (OSPL) cases for reasons described in Chapter 5. The [
] for steam line break cases for which offsite power is lost. This is acceptable since the [ ]
gives more conservative DNBR results for steady-state cases (according to Reference 6-1), and preliminary studies of steam line break cases show no difference in results.
6.2.3 Dropped Rod The changes presented in Section 6.1 also apply to the dropped rod transient. The remainder of the dropped rod thermal-hydraulic methodology presented in Reference 6-2 remains unchanged.
6.3 UFSAR Chapter 15 System Transient Analysis Methodology (DPC-NE-3002J _
DPC-NE-3002-A, "UFSAR Chapter 15 System Transient Analysis Methodology"(Reference 6-
- 3) documents the conservative modeling assumptions used by Duke Power Company in performing the NSSS primary and secondary system analyses of UFSAR Chapter 15 accidents.
It covers all applicable non-LOCA accidents in UFSAR Sections 15.1-15.6, except those already discussed in Reference 6-2. There are no changes to Reference 6-3 with respect to analyzing the RFA design.
6.4 Mass and Enerev Release and Containment Response Methodology (DPC-NE-3004)
DPC-NE-3004-PA, " Mass and Energy Release and Containment Response Methodology" (Reference 6-4), describes the Duke Power Company methodology for simulating the mass and energy release from high energy line breaks (LOCA and steam line break) and the resulting containment response to demonstrate that the containment peak pressure and temperature limits are not exceeded. Since the fuel stored energy for the RFA design is similar to that for the Mark-BW fuel, there are no changes anticipated for Reference 6-4 with respect to the RFA design except the RETRAN related changes described in Section 6.1 of this report. Similar changes to 6-4
I the RELAPS model, which is used to model the mass and energy release from LOCAs, are also anticipated. The RETRAN and RELAPS model changes for the RFA design are not significant
} enough to require reanalyses. Future reanalyses will incorporate the RFA design model revisions.
6.5 LOCA Analyses Large and small break LOCA analyses will be performed by Westinghouse using approved versions of the Westinghouse Appendix K LOCA evaluation models. All features employed have been approved by the NRC as required and annual model reports for the evaluation models have been supplied to the NRC, the most recent of which is found in Reference 6-22. Therefore, no NRC review of the evaluation model features is necessary, and only methodology with respect to analyzing McGuire/ Catawba will be presented in this section. New LOCA analyses will be performed to support the licensing of McGuire/ Catawba during the transition and full core operation of the RFA design.
6.5.1 Small Break LOCA 2
For small break LOCAs (SBLOCAs) due to breaks less than I ft , Westinghouse developed the NOTRUMP computer code (Reference 6-23) to calculate the transient depressurization of the reactor coolant system (RCS) as well as to describe the mass and enthalpy of flow through the break. The NOTRUMP Small Break LOCA Emergency Core Cooling System (ECCS)
Evaluation Model (References 6-24,6-25,6-26, and 6-27) was developed and licensed by Westinghouse to determine the RCS response to design basis SBLOCAs, and to address NRC concerns expressed in NUREG-0737, Item II.K.3.30.
The NRC approved noding scheme for the NOTRUMP Evaluation Model is shown in Reference 6-24, although minor noding changes to facilitate the modeling of broken loop ECCS were instituted and reported to the NRC in Reference 6-28. Peak cladding temperature (PCT) calculations are performed with the LOCTA-IV code (Reference 6-29) using the NOTRUMP calculated core pressure, fuel rod power history, uncovered core steam flow and mixture heights as boundary conditions. Additional modifications to the LOCTA-IV code to allow the modeling 6-5
of annular fuel pellets in the axial blankets have been reviewed and approved by the NRC in Reference 6-27. The axial shape chosen for McGuire/ Catawba SBLOCA will be based on the desired core operating limits and axial offset control strategy so as to bound all burnups and operating cycles.
\
Due to the nature of SBLOCA transients, the rod heatup and resulting calculated PCT is insensitive to transition core effects, and an evaluation is performed to demonstrate that this is a valid assumption. Therefore, SBLOCA will generally have no additional penalty for transition core effects.
6.5.2 Large Break LOCA For the Westinghouse large break LOCA (LBLOCA) methodology, a major pipe break (large break) is defined as a rupture with a total cross-sectional area equal to or greater than 1.0 ft2 . The most recent version of the 1981 Westinghouse Large Break LOCA ECCS Evaluation Model with B ASH (Reference 6-30) will be used to perform the LBLOCA analysis for the transition of McGuire/ Catawba to the RFA design. A description of the various aspects of the Westinghouse LOCA analysis methodology can be found in WCAP-8339 (Reference 6-31). This document describes the major phenomena modeled, the interfaces among the computer codes, and the features of the codes which ensure compliance with the acceptance criteria. The SATAN-VI(Reference 6-32), WREFLOOD (Reference 6-33), BASH and LOCBART codes, which are used in the LOCA analysis, are described in detail in References 6-30 and 6-34. These codes assess the core heat transfer geometry and determine if the core remains amenable to cooling through and subsequent to the blowdown, refill, and reflood phases of the LOCA. The LOTIC computer code (Reference 6-
- 35) calculates the minimum containment backpressure transient required for LBLOCA analyses in Appendix K to 10 CFR Part 50. Although there have been several updates to the original SATAN-VI code, the most notable upgrade is delineated in Reference 6-34.
The WREFLOOD code has been replaced by the REFILL code as reported in Reference 6-36. The REFILL code is identical to the section of the WREFLOOD code that modeled the refill phase of the transient. There has also been a recent change (the incorporation of the REFILL and LOCTA codes directly into the BASH code as subroutine modules) in the methodology for execution of the l 6-6
BASH Evaluation Model as reported in Reference 6-37. In addition, the LOTIC code has been coupled with the BASH code so that the codes run interactively. The BASH Evaluation Model now utilizes the SATAN code for the blowdown calculations, the BASH code for the refill and reflood phases with interactive LOTIC calculations for containment backpressure, and the LOCBART code for the fuel rod heatup calculations. The most recent version of the LOCBART code employs an improved grid heat transfer model which has been approved by the by NRC in Reference 6-38.
An input parameter that affects LOCA analysis results is the assumed axial power shape at the beginning of the accident. The methodology employed by Westinghouse is termed ESHAPE (Explicit Shape Analysis for Pct Effects). The ESHAPE methodology is based upon explicit analysis of the LBLOCA transient with a set of bounding skewed axial power shapes to supplement the base analysis performed with the chopped cosine power shape. The limiting case break, as demonstrated with a chopped cosine, will be reanalyzed using skewed power shapes and typically demonstrate that the chopped cosine power shape is limiting.
. required in Appendix K to 10 CFR 50, a minimum of a three break spectrum will be analyzed, typically with Moody break discharge coefficients, CD, of 0.4,0.6, and 0.8. In addition, as required in the NRC Safety Evaluation Report (SER) for the BASH Evaluation Model, a maximum Safety Injection flow case will be analyzed.
When assessing the effect of transition cores on the LBLOCA analysis, it must be determined whether the transition core can have a greater calculated peak cladding temperature (PCT) than a complete core of the RFA design. For a given peaking factor, the only mechanism available to cause a transition core to have a greater calculated PCT than a full core of either fuel is the possibility of flow redistribution due to fuel assembly hydraulic resistance mismatch. Hydraulic resictance mismatch will exist only for a transition core and is the only unique difference between a con,plete core of either fuel type and the transition core. Explicit analyses will be performed simulating the cross-flow effects due to any hydraulic mismatch between the current fuel and the Westinghouse fuel. If it is determined that a transition core penalty is required during the cycks that both fuels reside in the core, it will be applied as an adder to the LOCA results for a full core of the RFA design.
6-7
6.6 Rod Eiection Analysis Usine SIMULATE-3K This section presents an improved methodology to be used by Duke Power to perform the l nuclear analysir portion of the rod ejection accident (REA) analysis for the McGuire and Catawba Nuclear Stations. The current approved REA analysis methodology is described in the topical report titled, " Multidimensional Reactor Transients and Safety Analysis Physics Parameters" (Reference 6-2) and uses the computer code ARROTTA to perform the nuclear analysis portion of the REA calculation. A Safety Evaluation Report (SER) for this topical was received on November 15,1991 (Reference 6-6). The new methodology is based on the SIMULATE-3K (Reference 6-9) computer code which employs a three-dimensional neutron kinetics model based on the QPANDA two-group nodal model to calculate three-dimensional power distributions, core reactivity, or a core power level for both static and transient applications.
The SIMULATE-3K methodology affords compatibility with the current SIMULATE-3P nuclear design methodology (Reference 6-8) and will enhance the generation of forcing functions (transient core power distribution and hot assembly peak pin power distribution) at bounding physics parameter conditions for input into fuel enthalpy, peak RCS pressure, and DNB calculations. The SIMULATE-3K cross section model is also more robust than that used by ARROTTA. The transition from ARROTTA to SIMULATE-3K will reduce the engineering resources required to perform future REA analyses and enhance the transition from Mark-BW fuel to Westinghouse RFA or other fuel types in the future.
The basic methodology described in Reference 6-2 for the nuclear analysis ponion of the REA remains intact with only minor differences which are outlined in this report. All other methods described in Reference 6-2 remain unchanged, i.e. core thermal-hydraulic and system thermal-hydraulic analysis. To demonstrate the transient capability of SIMULATE-3K, comparisons between SIMULATE-3'K and ARROTTA reference REA analyses at beginning-of-cycle (BOC) and end-of-cycle (EOC), hot full power (HFP) and hot zero power (H7P) conditions were performed. These comparisons demonstrate the acceptability of the physical and numerical models within the SIMULATE-3K code as compared to the current licensed methodology.
t 6-8
A description of the models employed and the benchmark calculations performed in the verification of the SIMULATE-3K computer code are presented in Section 6.6.1. This section l
also includes a comparison of ARROTTA and SIMULATE-3K REA results applicable to the McGuire and Catawba Nuclear Stations at BOC and EOC, HFP and HZP conditions.
Section 6.6.2 describes the nuclear analysis methodology to be used in the evaluation of the UFSAR Chapter 15 REA using SIMULATE-3K.
f I
l )
6.6.1 SD4JLATION CODES AND MODELS i
6.6.1.1 CASMO-3 & SIMULATE-3P
{
\
CASMO-3 is used to produce two energy group edits of homogenized cross sections, assembly discontinuity factors, fission product data, and pin power data for input to ARROTTA, SIMULATE-3P, and SIMULATE-3K core models. CASMO-3 is a multigroup, two dimensional transport theory code for burnup calculations on PWR or BWR fuel assemblies. The code models a geometry consisting of cylindrical fuel rods of varying composition in a square pitch array with allowance for fuel rods loaded with integral burnable absorber, lumped burnable absorber roi, clustered discrete control rods, incore instrument channels, assembly guide tubes, and intra-assembly water gaps. The program utilizes a cross section library based on ENDF/B-IV with some data taken from ENDF/B-V. Reference 6-11 provides a detailed description of the theory and equations solved by CASMO-3. The use of CASMO-3 in this report is consistent l 1
with the previously approved methodologies of References 6-8 and 6-2. I SIMULATE-3P is used to set up the cycle-specific model and conditions for the REA. It may also be used to generate pin-to-assembly factors for the conversion of nodal powers to pin powers for the REA analyses. SIMULATE-3P is a three-dimensional, two energy group, diffusion theory core simulator program which explicitly models the baffle and reflector regions of the reactor. Homogenized cross sections and discontinuity factors developed with CASMO-3 are used on a coarse mesh nodal basis to solve the two group diffusion equations using the f QPANDA neutronics model. A nodal thermal hydraulics model is incorporated to provide both fuel and moderator temperature feedback effects. Inter- and intra-assembly information from the f
6-9
coarse mesh solution is then utilized along with the pinwise assembly lattice data from CASMO-3 to reconstitute pin-by-pin power distributions in two and three dimensions. The program performs a macroscopic depletion of fuel with microscopic depletion ofiodine, xenon, promethium, and samarium fission products. Reference 6-10 provides a detailed description of l
l the theory and equations solved by SIMULATE-3P. The use of SIMULATE-3P in this repon is consistent with the previously approved methodologies of References 6-8 and 6-2.
6.6.1.2 ARROTTA ARROTTA is a three-dimensional, two energy group diffusion theory core simulator applicable for both static and transient kinetics simulations. Homogenized cross sections, discontinuity factors, and six groups of delayed neutron precursor data are generated with CASMO-3 and used on a coarse mesh nodal basis to solve the two energy group diffusion equations using the QPANDA neutronics model. The thermal-hydraulic model is comprised of both fluid dynamics and heat transfer models. Reference 6-12 provides a detailed description of the theory and equations solved by ARROTTA. The use of ARROTTA for the benchmark calculations performed in this report is consistent with the previously approved methodology documented in Reference 6-2.
6.6.1.3 SIMULATE-3K 6.6.1.3.1 Code Description The SIMULATE-3K code (Reference 6-9)is a three-dimensional transient neutronic version of the SIMULATE-3P code (Reference 6-10). SIMULATE-3K uses the QPANDA full two-group nodal spatial model deseloped in SIMULATE-3P, with the addition of six delayed neutron groups. The program employs a fully-implicit time integration of the neutron flux, delayed neutron precursor, and heat conduction models. Beta is fully functionalized similar to other cross sections to provide an accurate value of beta for the time-varying neutron flux. The control of time step size may be determined either as an automated feature of the program or by user input. Use of the automated feature allows the program to utilize larger time steps (which 6-10
may be restricted to a maximum size based on user input) at times when the neutronics am changing slowly and smaller time steps when the neutronics are changing rapidly.
Additional capability is provided in the form of modeling a reactor trip. The trip may be initiated I
at a specific time in the transient or following a specified excore detector response. Use of the excore detector response model to initiate the trip allows the user to specify the response of individual detectors as required to initiate the trip, as well as the time delay prior to release of the control rods. The velocity of the control rod movement is also controlled by user input.
The SIMULATE-3K thermal-hydraulic model includes a spatial heat conduction and a hydraulic channel model. The heat conduction model solves the conduction equation on a multi-region mesh in cylindrical coordinates. Temperature-dependent values may be empl-oyed for the heat capacity, thermal conductivity, and gap conductances. A single characteristic pin conduction calculation is performed consistent with the radial neutronic node geometry, with an optional calculation of the peak pin behavior available to monitor local maxima. A single characteristic hydraulic channel calculation is perfonned based on the radial neutronic node geometry. The model allows for direct moderator heating at the option of the user. This thermal-hydraulic model is used to determine fuel and moderator temperatures for updating the cross-sections, and may additionally be used to provide edits of fuel temperature throughout the transient.
The SIMULATE-3K program utilizes the same cross-section library and reads the same restart file (exposure and burnup-related information) as SIMULATE-3P. Executed in the static mode, SIMULATE-3K performs the same solution techniques, pin power reconstruction, and cross-section development as SIMULATE-3P. Additional features of SIMULATE-3K include the application of conservatism to key physics parameters through simple user input. Also, the inlet thermal-hydraulic conditions can be provided on a time dependent basis through user input.
6.6.1.3.2 SIMULATE-3K Code Verification The SIMULATE-3K code has been benchmarked against many numerical steady state and transient benchmark problems by the code vendor, Studsvik of America, Inc. The results of these benchmarks are described in Reference 6-9 and show excellent agreement between 6-11
SIMULATE-3K and the reference solutions. Some of the SIMULATE-3K benchmarks which have been performed are: The fuel conduction and thermal-hydraulics model has b:en benchmarked against the TRAC code (Reference 6-13). The transient neutronics model has been benchmarked, using standard LWR problems, to reference solutions generated by QUANDRY l (Reference 6-14), SPANDEX (Reference 6-15), NEM (Reference 6-16), and CUBBOX (Reference 6-17). Finally, a benchmark of the coupled performance of the transient neutronics and thermal-hydraulic models was provided by comparison of results from a standard NEACRP rod ejection problem to the PANTHER code (Reference 6-18). Steady-state components of the SIMULATE-3K model are implemented consistent with the CASMO-3/ SIMULATE-3P methodology and performance benchmarks which were approved for use on all Duke Power reactors in Reference 6-8. In addition, a benchmark to ARROTTA for the Oconee REA analyses was performed in topical report DPC-NE-3005-P, "Oconee UFSAR Chapter 15 Transient Analysis Methodology (Reference 6-19).
6.6.1.3.3 SIMULATE-3K / ARROTTA REA Benchmark The three dimensional neutron kinetics capability of the SIMULATE-3K code is demonstrated by comparing SIMULATE-3K and ARRO'ITA calculations for the reference rod ejection accident analyses performed at BOC and EOC, HFP and HZP conditions for McGuire and Catawba. For the REA benchmark, ARROTTA and SIMULATE-3K are used to calculate the core power level and nodal power distribution versus time during the rod ejection transient for the BOC and EOC, HFP and HZP REA cases. These comparisons demonstrate the acceptability of the physical and numerical models within SIMULATE-3K for application in the REA analyses for McGuire and Catawba Nuclear Station.
The reference core used in the benchmark calculations is a hypothetical Catawba 1 Cycle 15 core. This core represents typical fuel management strategies (i.e. core loadings and cycle lengths) currently being developed for reload core designs at McGuire and Catawba Nuclear Stations. The ARRO'ITA and SIMULATE-3K models for this core were then adjusted to produce a conservative initial condition Doppler and moderator temperature coefficient, ejected rod worth, Beta. and power distribution as described in the " Multidimensional Reactor Transients and Safety Analysis Physics Parameter" topical report DPC-NE-3001 (Reference 6-2).
6-12
The combination of these conservative input parameters produces conservative transient results.
The assembly enrichmems, burnable poison loading, and assembly exposures for the reference core are shown in Figure 6-1. The core consists of all Framatome Mark-BW fuel.
6.6.1.3.3.1 ARROTTA Analysis The ARROTTA REA analysis is based on the methodology described in the " Multidimensional Reactor Transients and Safety Analysis Physics Parameters" topical report DPC-NE-3001 (Reference 6-2) with the exceptions that the initial power conditions have been increased to reflect a design pin FAH of 1.6, and the ARROTTA model was updated to reflect the CIC15 reference core design.
The REA analyses of Reference 6-2 were made limiting by setting key physics parameters to conservative or bounding values. Utilizing this approach produces limiting results which are expected to bound future reload cycles. The ARROTTA model was adjusted to produce conservative MTC, DTC, Beta, and ejected rod worths as identified in Tables 6-3.
6.6.1.3.3.2 SIMULATE-3K Analysis The SIMULATE-3K analysis is performed as described in DPC-NE-3001, Reference 6-2. The SIMULATE-3K model employed in this analysis was adjusted to be functionally equivalent to the ARROTTA model to account for differences in the two codes cross section model. Since ARROTTA is restricted to one node per fuel assembly in the radial direction, the SIMULATE-3K model was set up to be consistent with this assumption. The axial nodalization depends on 6-13
the fuel assembly design, such as whether or not axial blanket fuel is being modeled. For the analysis presented, an axial nodalization of 18 equal length fuel nodes is used.
[
] Additional model adjustments were performed to produce limiting values for the Doppler temperature coefficient, moderator temperature coefficient, ejected rod worth, and Beta. Table 6-3 provides a summary of initial condition values for each of these parameters for the SIMULATE-3K analyses. Trip times were input to be consistent with the ARROTTA analyses.
6.6.1.3.3.3 Results ARROTTA results from each of the four cases evaluated are summarized in Table 6-1. Results from the SIMULATE-3K cases are provided in Table 6-2. Table 6-3 lists the REA initial condition kinetics parameters for both the ARROTTA and SIMULATE-3K benchmarks. Core power versus time for each case is shown in Figures 6-2 through 6-4 For the HFP cases which begin at 102% power, core power increases rapidly as the control rod is ejected. The ejected rod worth in these transients is not sufficient to achieve a prompt critical state. Power increases until Doppler feedback from increasing fuel temperature begins to turn the excursion around. Core power level continues to decrease as the fuel temperature approaches an equilibrium value. A reactor trip signal on high flux occurs very early in these transients but the conservative trip delay time prevents rod motion until after the peak core power occurs.
Additional conservatism applied to the rate of rod insertion and scram worth minimizes the effect of the reactor trip until the rods approach the bottom of the reactor core.
The transients initiated from HZP differ from the at-power initial conditions in that the ejected rod worth is large enough to achieve a prompt critical core. The power increase continues after the control rod is fully ejected until the fuel heats up enough for Doppler feedback to tum the excursion around. Consarvatisms on trip delay time, rate of rod insertion, and scram worth minimize the impact of the reactor trip.
6-14
These results showed good agreement between SB1ULATE-3K and ARROTTA for the reference analyses. The transient power response and time of peak power statepoint agreed well.
The nodal peak powers agreed well with the exception of the EOC HZP case. This was due to the unique combination of adjustments which had to be made for this case to duplicate
, ARROTI'A's initial conditions as specified in Table 6-3. In conclusion, these comparisons demonstrate the acceptability of the physical and numerical models within the SIMULATE-3K code for application in analyses of the REA for McGuire and Catawba Nuclear Station.
6.6.2 Rod Ejection Nuclear Analysis The current approved methodology for the REA utilizes the computer code ARRO'ITA (Reference 6-12) to perform nuclear analysis calculations. This section describes the use of SB1ULATE-3K for the nuclear analysis calculations for the REA analyses as described in topical report, " Multidimensional Reactor Transients and Safety Analysis Physics Parameters" DPC-NE-3001 (Reference 6-2).
6.6.2.1 REA Analytical Approach The complexity of the core and system response to a rod ejection event requires the application of a sequence of computer codes. The rapid core power excursion is simulated with a three-dimensional transient neutronic and therrnal-hydraulic model using the SIMULATE-3K code (Reference 6-9). [ ] The resulting transient core power distribution results are then input to VIPRE-01 (Reference 6-20) core thermal-hydraulic models. The VIPRE models calculate the fuel temperatures, the allowable power peaking to avoid exceeding the DNBR limit, and the core coolant expansion rate. The allowable power peaking is then used along with a post-ejected condition fuel pin census to determine the percentage of pins exceeding the DNB limit. The coolant expansion rate is input to a RETRAN-02 (Reference 6-21) model of the Reactor Coolant System to determine the peak pressure resulting from the core power excursion.
The remainder of this section will address how the nuclear analyses of the REA will be performed with SIMULATE-3K. The basic methodology, as described in Reference 6-2, 6-15
remains unchanged with the exception of minor differences between SIMULATE-3K and ARROTTA which are discussed in the following section.
6.6.2.2 SIMULATE-3K Nuclear Analysis The response of the reactor core to the rapid reactivity insertion from the control rod ejection is simulated with SIMULATE-3K code (Reference 6-9). SIMULATE-3K computes a three-dimensional power distribution (in rectangular coordinates) and reactivity or power level for both static and transient applications. SIMULATE-3K includes a prediction ofindividual pin powers.
Modifications are made to the core model to ensure conservative results. These changes produce a rod ejection model which produces limiting results tisat are expected to bound future reload cycles. A complete description of the SIMULATE-3K code is discussed in Section 6.6.1.3 and Reference 6-9.
The SIMULATE-3K model geometry will typically be [ ] per fuel assembly in the radial direction. The axial nodalization depends on the fuel assembly design, such as whether or not axial blanket fuel is being modeled. The number of axial levels is chosen to accurately describe the axial characteristics of the fuel. For current fuel designs, a typical axial nodalization of 24 equal length fuel nodes in the axial direction is used. The SIMULATE-3K model explicitly calculates neutron leakage from the core by use of reflector nodes in the radial direction beyond the fuel region and in the axial direction above and below the fuel column stack. Required fuel and reflector cross sections are developed consistent with the methodology approved for SIMULATE-3P in topical report DPC-NE-1004A (Reference 6-8).
SIMULATE-3K is used to calculate the core power level and nodal power distribution versus time during the rod ejection transient. [
] This information is used by VIPRE to determine the fuel enthalpy, the percentage of the fuel pins exceeding the DNB limit, and the coolant expansion rate.
l 6-16
6.6.2.2.1 Initial Conditions The SIMULATE-3K rod ejection analysis is analyzed at four statepoints; beginning-of-cycle (BOC) at hot zero power (HZP) and hot full power (HFP) and end-of-cycle (EOC) at HZP and HFP. The conservatism applied to the rod ejection analysis as described in Reference 6 2 are implemented based on the methodology described in Reference 6-9 and are expected to bound future reload cycles. Initial conditions for SIMULATE-3K different than those discussed in Reference 6-2 are described below.
- ~"
The moderator temperature coefficient (MTC) is also adjusted to conservative values at BOC or EOC which bounds the magnitude of the MTC expected in a reload core. The MTC is adjusted in SIMULATE-3K by [
] This adjustment is made via the equation from SIMULATE-3K (Reference 6-9);
6-17
Similar adjustments are made to yield conservative rod worth for control rod withdrawal and rod worth for control rod insertion. .
The Doppler (or fuel) temperature coefficient (DTC) is imponant to this transient because the negative reactivity from the increased fuel temperature is the only effect that limits the power excursion and starts to shut down the reactor. The DTC is adjusted to a conservative value which bounds the magnitude of the DTC expected in a reload core. The DTC is adjusted in SIMULATE-3K by [
]
The effective delayed neutron fraction ( ) and the ejected rod worth both determine the transient power response of the reactor. The peak power level obtained during the transient will increase for small values of and larger values of the ejected od worth. The ejected rod worth and are adjusted to conservative values which bound values expected for a reload core. The ejected rod wonh is adjusted in SIMULATE-3K by [
] can be adjusted in SIMULATE-3K by [
6-18
]
s
[ ]
or can be adjusted by inputting a single set of delayed neutron parameters to be used for all fueled nodes.
The combined effect of all these changes to the SIMULATE-3K model is to produce a model that is expected to bound future reload cycles for both McGuire and Catawba Nuclear Stations.
6.6.2.2.2 Boundary Conditions The fuel and core thermal-hydraulic boundary conditions are established using conservative assumptions. Bounday conditions for initial power, core flow, inlet temperature, reactor pressure, and fission power fraction in the coolant are selected to yield conservative results.
The reactor trip signal is generated when the third highest excore channel reaches either
[ for the HZP cases or [ ] for the HFP cases. This modeling is based on a single failure of the highest channel and a two-out-of-the-remaining-three trip coincidence logic. [
] in SIMULATE-3K (Reference 6-9) can be used. The remaining control rods fall into the reactor assuming a conservative trip delay after the trip signal is generated, Durir.g the reactor trip, the ejected rod and a second rod with the highest worth are assumed not to fall into the reactor. To conservatively model the reactor trip, not all of the control rod banks are allowed to drop, and some of the banks that are dropped have their worth reduced by a cross section adjustment. The rod worth adjustment is made in SIMULATE-3K by [
6-19
] based on Eq. 6.1. Also, negative reactivity inserted due to the reactor trip is not allowed to exceed the conservative trip reactivity curve. The integral worth of the falling control rods is computed for several different axial positions of the rods at the initial conditions. [
]
6.6.2.3 Core Thermal-Hydraulic Analysis g
The core thermal-hydraulic analyses use the VIPRE-01 code for the calculation of peak fuel enthalpy, DNBR, and the coolant expansion rates for various initial and boundary conditions g postulated for the REA transient. All input to the core thermal-hydraulic analyses once supplied by ARROTTA can now be supplied by SIMULATE-3K. The nuclear analysis input boundary conditions supplied by SIMULATE-3K for the thennal-hydraulic analyses are [
]
6.6.2.3.1 Fuel Temperature and Peak Fuel Enthalpy The calculation of the transient maximum hot spot average fuel temperature and the maximum radial average fuel enthalpy requires the following input boundary conditions to be supplied by SIMULATE-3K: [
] This information is consistent with that provided by ARROTTA in Reference 6-2.
6.6.2.3.2 DNBR Evaluation The percentage of the core experiencing DNBR is calculated as explained in Reference 6-2 except SIMULATE-3K results are used instead of ARROTTA results. For the HFP REA cases, .
6-20
[
] For a given axial power profile, the maximum pin radial peak can be determined such that DNB would not occur during the transient.
These DNB limits are referred to as maximum allowable radial peaks (MARP) !imits. A fuel pin census is then performed to determine the number of fuel pins in the core that exceed the power peaking limit.
6.6.2.3.3 Coolant Expansion Rate The calculation of the coolant expansion rate requires th : following input boundary conditions to be supplied by SIMULATE-3K: [
] This SIMULATE-3K information is input to VIPRE to calculate the flow rate in each channel during the transient. Using the VIPRE channel flow rates, the total coolant expansion rate can be calculated. This total coolant expansion rate is input to the RETRAN plant transient model for simulating the resulting pressure response. This SIMULATE-3K information is consistent with that provided by ARROTTA in Reference 6-2.
6-21
6.6.2.4 Cycle-Specific Evaluation Due to the conservative assumptions and modeling used in the SIMULATE-3K model, it is anticipated that for reload cores, no new SIMULATE-3K cases will be necessary. The determination as to whether the existing SIMULATE-3K cases remain bounding will be made by performing a cycle-specific reload check of the key physics input parameters as described in Reference 6-2. These parameters will be calculated using steady-state neutronics codes approved by the NRC for reload design. If the key physics parameters remain bounded then no new SIMULATE-3K analyses are necessary; otherwise, an evaluation, reanalysis, or re-design of the reload core will be performed.
For the HFP REA cases, a DNB pin census will be performed for the reload cycle, as described in Section 4.7 of Reference 6-2, with the radial power information being calculated with an NRC approved steady-state neutronics code. The HZP REA cases are bounded by the HFP cases in the offsite dose analyses, and therefore, a pin census is not required. The ejected rod worth shall be calculated with the fuel and moderator temperatures frozen in the pre-ejected condition or uniform throughout the core (either method will generate conservative results). [
] The power distribution with the ejected rod out will be used for the DNB pin census. The calculated percent fuel failure due to DNB will be compared for each cycle to the fuel failure limit assumed in the dose calculation. If the cycle specific value is less than the limit, then the existing safety analysis is still valid. Otherwise, an evaluation, a new dose calculation, reanalysis, or new reload design will be perfonned as appropriate.
6.6.2.5 Mixed Cores The Westinghouse fuel is expected to beha've neutronically similar to that of the Framatome Cogema Fuels Mark-BW fuel. The steady-state cycle-specific checks will verify that all key physics parameters remain valid and the DNB census will use the appropriate CHF correlations for the various fuel types present in the core.
6-22
6.7 References 6-1 DPC-NE-3000-PA, Thermal-Hydraulic Transient Analysis Methodology, Rev.1, Decerrher 1997.
6-2 DPC-NE-3001-PA, Multidimensional Reactor Transients and Safety Analysis Physics Parameters Methodology, Rev. O, November 1991.
6-3 DPC-NE-3002-A, UFSAR Chapter 15 System Transient Analysis Methodology, Rev. 2, December 1997.
6-4 DPC-NE-3004-PA, Mass and Energy Release and Containment Response Methodology, Rev. O, November 1997.
6-5 WCAP-15025-P, Modified WRB-2 Correlation, WRB-2M, for Predicting Critical Heat Flux in 17x17 Rod Bundles with Modified LPD Mixing Vane Grids, Westinghouse Energy Systems, February 1998.
6-6 Letter, T. A. Reed (NRC) to H. B. Tucker (Duke), November 15,1991 (SER for DPC-NE-3001).
6-7 Letter from Jeffrey A. Borkowski (Studsvik of America) to Robert St. Clair (Duke Power Company), January 12,1996.
6-8 Nuclear Design Methodology using CASMO-3/ SIMULATE-3P, DPC-NE-1004A, Revision 1, Duke Power Company, December 1997.
6-9 SIMULATE-3 Kinetics Theory And Model Description, SOA-96/26, Studsvik of America, April 1996 6-23
6-10 SIMULATE-3: Advanced Three-Dimensional Two-Group Reactor Analysis Code, Studsvik/SOA-92/01, Studsvik of America, April 1992 6-11 CASMO-3: A Fuel Assembly Burnup Program User's Manual, STUDSVIK/NFA-88/48, Studsvik of America, September 1988 6-12 ARROTTA: Advanced Rapid Reactor Operational Transient Analysis, EPRI, August 1993 6-13 TRAC-BFl/ MOD 1, Vol.1, NUREG/CR-4356, J. Borkowski and N. Wade (Editors),
May 1992 6-14 QUANDRY: An Analytical Nodal Method for Solving the Two-Group, Multidimensional, Static and Transient Nodal Diffusion Equations, K. S. Smith, Massachusetts Institute of Technology,1979 6-15 Development of a Variable Time-Step Transient NEM Code: SPANDEX, Trans. Am.
Nucl. Soc. 68- 425, B. N. Aviles,1993 6-16 NEM: A Three Dimensional Transient Neutronics Routine for the TRAC-PFI Reactor Thermal Hydraulic Computer Code, B. R. Fandini, Pennsylvania State University,1990 6-17 CUBBOX: Coarse-Mesh Nodal Diffusioa Method for the Analysis of Space-Time Effects In Large Light Water Reactors, S. Langenbuch, W. Maurer, and W. Werner, Nucl. Sci. Eng.63-437,1977 6-18 NEACRP 3-D LWR Core Transient Benchmark, NEACRP-L-335, H. Finneman and A.
Galati, January 1992 6-19 UFSAR Chapter 15 Transient Analysis Methodology, DPC-NE-3005-P, Duke Power Company, July 1997.
6-24
6 20 VIPRE-01: A Thermal-Hydraulic Code for Reactor Cores, EPRI NP-2511-CCM, Revision 2, EPRI, July 1985 I
6-21 RETRAN-02: A Program for Transient Thermal-Hydraulic Analysis of Complex Fluid Flow Systems, EPRI NP-1850-CCM, Revision 4, EPRI, November 1988.
6-22 NSD-NRC-98-5575, "1997 Annual Notification of Changes to Westinghouse Small Break LOCA ECCS Evaluation Model and Large Break LOCA ECCS Evaluation Model Pursuant to 10CFR50.46(a)(3)(ii)", April 9,1998.
6-23 WCAP-10079-P-A (Proprietary), WCAP-10080-A (Non-Proprietary), Meyer, P.E.,
"NOTRUMP- A Nodal Transient Small Break And General Network Code", August 1985.
6-24 WCAP-10054-P-A (Proprietary), WCAP-10081 (Non-Proprietary), Lee, H., et al.,
" Westinghouse Small Break ECCS Evaluation Model Using the NOTRUMP Code",
August 1985.
6-25 WCAP-10054-P-A Addendum 2 (Proprietary), " Addendum to the Westinghouse Small Break ECCS Evaluation Model Using the NOTRUMP Code: Safety Injection into the Broken Loop and Cosi Condensation Model", August 1994.
6-26 WCAP-11145-P-A (Proprietary), WCAP-11372 (Non-Proprietary), Rupprecht, S.D., et al.,
" Westinghouse Small Break LOCA ECCS Evaluation Model Generic Study with the NOTRUMP Code", October 1986.
6-27 WCAP-14710-P-A, D. J. Shimeck, "1-D Heat Conduction Model for Annular Fuel Pellets",
May 1998.
6-28 NSD-NRC-96-4639,"1995 Annual Notification of Changes to the Westinghouse Small Break LOCA ECCS Evaluation Model and Large Break LOCA ECCS Evaluation Model Pursuant to 10 CFR 50.46 (a)(3)(ii)", February 9,1996.
6-25
(
(
6-29 WCAP-8301 (Proprietary), WCAP-8305 (Non-Proprietary), Borderlon, F. M., et al.,
"LOCTA-IV Program: Loss-of-Coolant Transient Analysis," June 1974.
6-30 WCAP-10266-P-A Revision 2 with Addenda (Proprietary), Kabadi, J.N., et al., 'The 1981 Version of the Westinghouse ECCS Evaluation Model Using the BASH Code" March 1987.
6-31 WCAP-8339 (Non-Proprietary), Bordelon, F. M., et al., " Westinghouse ECCS Evaluation Model- Summary", June 1974.
6-32 WCAP-8302 (Proprietary) and WCAP-8306 (non-Proprietary), F. M. Borderlon, et. al.,
" SATAN-VI Program: Comprehensive Space-Time Dependent Analysis of Loss-of-Coolant", June 1974.
6-33 WCAP-8170 (Proprietary) and WCAP-8171 (Non-Proprietary), Kelly, R. D. et al.,
" Calculation model for core Reflooding After a Loss-of-Coolant Accident (WREFLOOD Code)," June 1974.
6-34 WCAP-9220-P-A Revision 1 (Proprietary), WCAP-9221 (Non-Proprietary), Eicheldinger, C., " Westinghouse ECCS Evaluation Model - 1981 Version", February 1982.
6-35 WCAP-8345 (Proprietary), Hsieh, T., and Raymund, M., "Long-Term Ice Condenser Containment LOTIC Code Supplement 1," July 1974, WCAP-8355, Supplement 1, May 1975.
6-36 NTD-NRC-94-4143, Letter from N. J. Liparulo (Westinghouse) to W. T. Russel (USNRC),
" Change in Methodology for Execution of B ASH Evaluation Model", May 23,1994.
6-37 NTD-NRC-95-4520, Letter from N. J. Liparulo (Westinghouse) to W. T. Russel (USNRC),
" Change in Methodology for Execution of BASH Evaluation Model", August 29,1995.
6-26
6-38 WCAP-10484-P-A Addendum 1, " Spacer Grid Heat Transfer Effects During Reflood",
September 1993.
6-27
Table 6-1 l
Rod Ejection ARROTTA Results Parameter BOC HZP BOC HFP EOC HZP EOC HFP Time of peak power, sec 0.286 0.077 0.173 0.080 Peak power level, % of full power 1880 138 5139 155 Peak nodal power relative to core average 7.99 3.44 16.40 3.96 Time that trip setpoint reached, sec 0.246 0.061 0.155 0.057 Time for beginning of trip rod motion 0.746 0.561 0.655 0.557 Table 6-2 Rod Ejection SIMULATE-3K Results Parameter BOC HZP BOC HFP EOC HZP EOC HFP Time of peak power, sec 0.296 0.076 0.187 0.083 Peak power level % of full power 1884 133 5280 154 Peak nodal power relative to core average 7.127 3.508 12.997 3.605 Time that trip setpoint reached, sec 0.246 0.061 0.155 0.057 Time for beginning of trip rod motion 0.746 0.561 0.655 0.557 6-28
Table 6-3 Rod Ejection Transient Kinetics Input Parameters Parameter Computer Code BOC HZP BOC HFP EOC HZP EOC HFP Ejected Rod Worth, pcm ARROTTA 720 201 900 l 196 MTC(pcmrF) ARROTTA +7.% +0 05 -9.45 -9.73 DTC(pcmrF) ARROTTA -0.90 -0.90 -1.19 -1.19 Delayed Neutron Fraction, ARROTTA 0.0055 0.0055 0.0(M0 0.0040 Ejected Rod Worth, pcm SIMULATE-3K 721 203 900 197 MTC(pcmrF) SIMULATE-3K +7.00 +0.08 -10.09 -10.09 DTC(pcmrF) SIMULATE-3K -0.90 -0.90 -1.20 -1.20 Delayed Neutron Fraction, SIMULATE-3K 0.0055 0.0055 0.0040 0.0010 6-29
Figure 6-1 Reference Core Loading Information 11 G F E D C B A 14 17 15 17 IT 16 16 17 4.10 4.40 4.15 4.15 4.is 4.40 4.40 4.40 8 24/2.5 P 24/2.5 24/3.0 P 24/3.0 24/3.0 P 0 24/2.5 P 0 43.165 0 33.53 0 33.452 15.245 21.428 0 58.124 21.143 51.68 22.606 51.893 36.447 40.26 15.2M 14 17 16 17 16 17 16 4.40 4.15 4.15 4.15 4.15 4.40 4.15 9 0 24/3.0 24/3.0 P 24/3.0 24/2.5 P 24/3.0 24/3.0 P 36.418 0 22.295 0 22.495 0 22.41 53.838 22.308 43.375 22.364 42.515 19.812 33.733 16 17 15 17 16 16 4.15 4.15 4.15 4.15 4.40 4.15 10 24/3.0 P 24/3.0 24/3.0 P 24/3.0 12/2.0 P 24/3.0 P 22.583 0 32.3N 0 18.423 22.588 43.611 22.605 51.037 21.883 .36.579 32.353 15 17 16 17 15 4.15 4.15 4.40 4.40 4.40 11 24/3.0 P 24/2.5 24/2.5 P 12/2.0 24/3.0 P 35.265 0 19.774 0 38.018 53.466 22.485 40.535 18.483 44.405 16 17 16 4.15 4.40 4.40 12 24/3.0 P 24/2.5 24/3.0 P 21.824 0 19.812 41.811 19.727 32.032 16 15 4.15 4.40 13 24/3.0 P 24/3.0 P 21.823 32.088 35.106 38.42 Key:
Batch number Enrichment Nurnber of BP fingers / wt9c of boron in BP (P means bps pulled)
BOC exposure (GWD/MT)
EOC exposure (GWD/MT)
Batch 17 is fresh fuel Batch 16 is starting its second burn Batches 14 and 15 are starting their third burn 6-30
Figure 6-2 FSAR Section 15.4.8 - Control Rod Ejection BOC HFP Core Power vs. Time I
(
l 160
,- l l l f !
,, . I e ,,
j lN % j ll= 1
, l .
j l
1 I
i i
i
{
i\\ l !
t e ! l l t , !
, i
- i I
0 0 000 0.500 1.000 1.500 2.000 2.500 3.000 3.500 4.000 Time, seconds Figure 6-3 FSAR Section 15.4.8 - Control Rod Ejection BOC HZP Core Power vs. Time 10.0E+3
, a.3 300M+0 100E+0 --
g 1 E+0 --- l I s I j l
- .100.0E 3 --
10.00-3 -
10E-3 --
i :
100.0E4 -
i ' I 100E4 --
IDE-6 -
1000E-9 0.000 0.500 1.000 1300 2 000 2.500 3.000 3.500 4.000
. Time, seconds 6-31
Figure 6-4 FSAR Section 15.4.8 - Control Rod Ejection EOC HFP Core Power vs. Time 160 I !
i { [ !
140 -
i (
! ! l i20 l t j i i J.-anw,l; m 2:00 ,
I l r i I, '
Ee -
\ , .
i I l
4Q ~
I 20
+
l -
O i ! l '
0.000 0.500 lax) ISO 2 000 2.500 3.000 3.500 4 a10 Time, seconds Figure 6-5 FSAR Section 15.4.8 - Control Rod Ejection EOC HZP Core Power vs. Time iO0E+s 10E*3 -
100.0E+0 -
I I 10.0E+0 --
j - - - - -
g 1.0E 0 - '
= ,
1 8.100.0E 3 - '
h ! !
! ' Mtu g 10.0E-3 -
1 OE.3 .-
i0026 _
i !
l 10.0E-6 -
I 10E 6 -
100.0E-9 0.000 0500 lux) ISD 200 2.500 3DX) 3.500 4 0(10 Time, seconds 6-32
7.0 FUEL ASSEMBLY REPAIR AND RECONSTITlTflON The reconstitution of fuel assemblies is a routine occurrence during refueling outages in light 3 water reactors. This is due to the concerted effort on the part of utilities to maintain zero fuel defects dering cycle operation. This zero defect goal requires aggressive programs in two areas.
First, all reasonable measures must be taken in the design and manufacturing of fuel assemblies to prevent any type of known failure mechanism. Secondly, failures that do occur during operation should be identified and the failed fuel rods removed before subsequent cycles.
\
Duke Power's primary replacement candidate for use in reconstitution is a fuel rod that contains pellets of natural uranium dioxide (UO2). Aside from enrichment, this rod is the same in design and behavior as a standard fuel rod and is analyzed using standard approved methods. If local grid structural damage exists, the use of a natural UO2 replacement rod is not the preferred alternative and solid filler rods made of stainless steel, zircaloy, or ZIRLO* would be used.
The NRC-approved DPC-NE-2007 topical report, Reference 7-1, describes the methodology and guidelines Duke Power uses to support fuel assembly reconstitution with filler rods. The guidelines were developed to ensure acceptable nuclear, mechanical, and thermal-hydraulic performance of reconstituted fuel assemblies. Specific results were provided in the report for the Mark-B and Mark-BW fuel designs with licensed codes. As stated in DPC-NE-2007, the methodology would be applicable if different fuel designs or codes are licensed by Duke Power.
Duke Power will use the same licensing and analysis approach for reconstitution of the RFA design at McGuire and Catawba. The methodology described in Reference 7-1 will be used along with the licensed codes and correlations described in this report. These codes will be used to analyze reconstitution with filler rods for acceptable nuclear, mechanical, and thermal-hydraulic performance. For a reload core using reconstituted Westinghouse fuel, Westinghouse will evaluate the effects of the reconstitution on the LOCA analysis using the methodology given in Reference 7-2, 7-1
}
l l
As discussed in Reference 7-2, Westinghouse has reviewed the criteria specified in Standard Review Plan 4.2 (Reference 7-3) and determined that the only fuel assembly mechanical criteria impacted by reconstitution are:
- 1) fuel assembly holddown force, and 7
- 2) fuel assembly structural response to Seismic /LOCA loads.
l Westinghouse evaluated both of these criteria and concluded that the reconstituted fuel assembly l designs are acceptable for both normal and faulted condition operations.
7.1 References 7-1 DPC-NE-2007P-A, Duke Power Company Fuel Reconstitution Analysis Methodology, October 1995.
7-2 W. H. Slagle (Ed.)," Westinghouse Fuel Assembly Reconstitution Evaluation Methodology", WCAP-13060-P-A, July 1993.
1 7-3 "Section 4.2, Fuel System Design", Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants - LWR Edition, NUREG-0800, Rev. 2, US Nuclear Regulatory Commission, July 1981.
l l.
7-2
8.0 IMPROVED TECHNICAL SPECIFICATION CHANGES Since the RFA design will be first implemented for Catawba 2 Cycle 11, changes to the current McGuire and Catawba Technical Specifications are not necessary. However, the following changes to the Improved Technical Specifications (ITS), originally submitted to the NRC on May 27,1997 with numerous supplements submitted thereafter, are necessary to license the RFA design.
Figure 2.1.1-1 (Reactor Core Safety Limits - Four Loops in Operation) will be modified to delete the 2455 psia safety limit line. This line is the current upper bound pressure at which power operation is permitted and is dependent on the pressure range of the critical heat flux (CHF) correlation used in DNBR analyses. The critical heat flux correlation of the resident Mark-BW fuel is applicable up to a pressure of 2455 psia. Deleting the 2455 psia safety limit line is necessary due to implementation of the WRB-2M CHF correlation for the RFA design, which has an upper range of 2425 psia (Reference 8-1). The 2400 psia safety limit line will remain as the upper bound safety limit line because it is within the range of the CHF correlations for the RFA and Mark-BW fuel designs.
ITS 4.2.1 will be revised to add ZIRLO* cladding to the fuel assembly description. ITS 5.6.5 will be revised to add this topical report to the list of approved methodologies for McGuire and Catawba.
The nuclear design related Technical Specification limits were reviewed for transition and full core reloads comprised of the Westinghouse RFA design. The power distribution Technical Specifications for Fq and FAH have a 2rk factor in each specification's surveillance which is used to account for the possible increase in Fq and FAH between flux maps. This factor for IFBA cores will have to be burnup dependent because of the increased burnout rate of the integral bumable absorber relative to the lumped burnable absorbers. The technicaljustification for this proposed change is given in Sections 8.1 and 8.2.
8-1
8.1 Technical Justification for Surveillance Requirement SR 3.2.1.2 and SR 3.2.1.3 Fy(x,y,z) is measured periodically using the incore detector system to ensure that the value of the total peaking factor, F q-RTP, assumed in the accident analysis is bounding. The frequency requirernent for this measurement is 31 effective full power days (EFPD). To account for the possibility that F q(x,y,z) may increase between surveillance, a trend of the measurement is performed to determine the point where peaking would exceed allowable limits if the current trend continues. If the extrapolation of the measurement indicates that the F,(x,y,z) measurement would exceed the Fq(x,y,z) limit prior to 31 EFPD beyond the most recent measurement, then either the surveillance interval would be decreased based on available margin, or the F,(x,y,z) measurement would be increased by an appropriate penalty (cunently 1.02) and compared against the Fy (x,y,z) operational and RPS surveillance limits to ensure allowable total g peaking limits are not exceeded.
Technical Specification surveillance SR 3.2.1.2 and SR 3.2.1.3 currently specify that the Fq(x,y,z) measurement be increased by 1.02. This value was chosen because it bounded the maximum Fq(x,y,z) increase in typical reload cores. However, for reactor cores containing integral burnable absorbers, a larger penalty may be required over certain burnup ranges early in the cycle due to the rate of burnout of this poison. This penalty can be incorporated into either the M g(x,y,z) or M,(x,y,z) margin factors, or be provided in tabular form as a function of burnup.
It is proposed that this penalty factor be moved to the Core Operating Limits Report (COLR) in tabular form to facilitate cycle specific updates. Table 8-1 provides an example burnup dependent penalty factor that would replace the cunent 1.02 value. For bumup ranges where the increase in Fq over the 31 EFPD surveillance interval is less than 2.07c, the current 1.02 penalty factor will be maintained.
Relocation cf this penalty factor to the Core Operating Limits Report (COLR) was included in TSTF-98 (Technical Specification Task Force), Revision 2. This generic change to NUREG-1431 was approved by the NRC in April 1998.
8-2
8.2 Technical Justification for Surveillance Requirement SR 3.2.2.2 l
l The nuclear enthalpy rise hot channel factor, Fm(x,y,), is measured periodically using the incore detector system to ensure that fuel design criteria are not violated and accident analysis assumptions are not violate 1. The frequency requirement for this measurement is 31 effective full power days (EFPD). To account for the possibility that Fm(x,y) may increase between surveillance, a trend of the mec aement is performed to determine the point where peaking would exceed allowable limits if the current trend continues. If the extrapolation of the measurement indicates that the Fm(x,y) measurement would exceed the Fm(x,y) surveillance limit prior to 31 EFPD beyond the most recent measurement, then either the surveillance interval would be decreased based on available margin, or the Fm(x,y) measurement would be increased by an appropriate penalty (currently 1.02) and compared against the Fm(x,y) surveillance limit to ensure allowable peaking limits are not exceeded.
Technical Specification surveillance SR 3.2.2.2 currently specifies that the Fm(x,y) measurement be increased by 1.02. This value was chosen because it bounded the maximum Fm(x,y) increase in typical reload cores. However, for eactor cores containing integral burnable absorbers, a larger penalty may be required over certain burnup ranges early in the cycle due to the rate of burnout of this poison. This penalty can be incorporated into either the Fm(x,y) surveillance limit or be provided in tabular form as a function of burnup.
It is proposed that this penalty factor be moved to the Core Operating Limits Report (COLR) in tabular form to facilitate cycle specific updates. Table 8-2 provides an example burnup dependent penalty factor that would replace the current 1.02 value. For burnup ranges where the increase in Fm(x,y) over the 31 EFPD surveillance interval is less than 2.07c, the curmnt i.02 penalty factor will be maintained.
Relocation of this penalty factor to the Core Operating Limits Report (COLR) was included in TSTF-98 (Technical Specification Task Force), Revision 2. This generic change to NUREG-1431 was approved by the NRC in April 1998.
8-3
l 8.3 References
)
l 8-1 WCAP-15025-P, Modified WRB-2 Correlation, WRB-2M. for Predicting Critical Heat Flux in 17x17 Rod Bundles with Modified LPD Mixing Vane Grids, Westinghouse Energy Systems, February 1998.
b-4
L f
(
Table 8-1 F,(x,y,z) Margin Decrease Over 31 EFPD Surveillance Interval (Typical Values)
F,(x,y,z) Margin Burnuo (EFPD) Decrease Penalty Factor 4 2.00 %
12 2.28 %
25 3.31 %
50 3.45 %
100 3.24 %
200 2.00 %
EOC 2.00 %
Note: Linear interpolation of the penalty factors is adequate for surveillance performed at intermediate bumups.
8-5
Table 8-2 l
Fe(x,y) Margin Decrease Over 31 EFPD Surveillance Interval (Typical Values)
Fm(x,y) Margin Bumun (EFPD') Decrease Penalty Factor 4 2.00 7c 12 2.40 %
25 2.50 %
50 2.60 %
100 2.15 %
200 2.00 %
EOC 2.00 %
Note: Linear interpolation of the penalty factors is adequate for surviellances performed at intermediate burnups.
8-6
__