~duke nergy. catawba, · 2013-09-11 · analyses are not impacted by this error, leading to an...

Robert J. Duncan~DUKE Sr Vice President

Nuclear OperationsNERGY. Catawba, McGuire

526 S. Church StreetCharlotte, NC 28202

Mailing Address:EC07H / P.O. Box 1006

Charlotte, NC 28202

o: 704-382-4098c: 919-812-7226f: 704-382-6056

Bob. Duncan(.duke-eoneroqy. com

10 CFR 50.410 CFR 50.46

August 29, 2013

U. S. Nuclear Regulatory CommissionAttn: Document Control DeskWashington, DC 20555-0001

Catawba Nuclear Station, Units 1 and 2Docket Numbers 50-413 and 50-414/Renewed License Numbers NPF-35 and NPF-52

McGuire Nuclear Station, Units 1 and 2Docket Numbers 50-369 and 50-370/Renewed License Numbers NPF-9 and NPF-17

Subject: Duke Energy Carolinas, LLC (Duke Energy): Report Pursuant to 10 CFR 50.46,Changes to or Errors in an Evaluation Model

References:

1) Letter, D. C. Culp (Duke Energy) to USNRC, Subject: Catawba Nuclear Station Units 1 and2, and McGuire Nuclear Station Units 1 and 2, Response to Information Request Pursuantto 10 CFR 50.54(f) Related to the Estimated Effect on Peak Cladding TemperatureResulting from Thermal Conductivity Degradation in the Westinghouse-Furnished RealisticEmergency Core Cooling System Evaluation and 30-Day Report Pursuant to 10 CFR 50.46,Changes to or Errors in an Evaluation Model," March 16, 2012. [ADAMS ML12079A180]

10 CFR 50.46 (a)(3)(ii) requires the reporting of changes to or errors in Emergency CoreCooling (ECCS) evaluation models (EMS). On July 31, 2013, Duke Energy received a letterfrom Westinghouse Electric Company identifying errors in the heat transfer multiplier uncertaintydistributions which affect the Best Estimate Large Break Loss of Coolant Accident (BELOCA)analysis of record for Catawba Nuclear Station (CNS) Units 1 & 2 and McGuire Nuclear Station(MNS) Units 1 & 2. Small Break LOCA analyses for CNS and MNS are not impacted by theseerrors.

The enclosed Attachment 1 provides a description of the errors, and the associated impact tothe Catawba and McGuire BELOCA analysis of record. Based on information supplied byWestinghouse, an assessment of this error results in a peak cladding temperature (PCT)decrease of 85"F for the limiting transient. These impacts to the BELOCA analyses arediscussed in Table 1, and are included on the PCT reporting sheets, Tables 2 through 4.

Aob6

U.S. Nuclear Regulatory CommissionAugust 29, 2013

Attachment 2 provides additional information on how Westinghouse performed the PCTevaluation of changes to the heat transfer multiplier uncertainty distributions.

Since the absolute value of the change in PCT is greater than 50 OF, this is considered to be asignificant change. 10 CFR 50.46(a)(3)(ii) states: " ... If the change or error is significant, theapplicant or licensee shall provide this report within 30 days and include with the report aproposed schedule for providing a reanalysis or taking other action as may be needed to showcompliance with 50.46 requirements

In Reference 1, Duke Energy has previously committed to submit by December 15, 2016 aLBLOCA analysis that applies an NRC-approved ECCS evaluation model that includes theeffects of fuel pellet thermal conductivity degradation. Since this report identifies a reduction inPCT, there are no adverse impacts to safety as a result of the ECCS evaluation model errorsdescribed herein, and all 10 CFR 50.46 acceptance criteria are met. Therefore, the existingLBLOCA reanalysis commitment discussed in Reference 1 is sufficient to address therequirements of 10 CFR 50.46(a)(3)(ii) pertaining to the most recent ECCS evaluation modelerrors described herein.

Several other changes were made to the WCOBRAITRAC computer code used within theBELOCA evaluation model. The specific details of these changes are also provided in Table 1,and were evaluated by Westinghouse as having no impact on the calculated PCTs. Since therewas no PCT impact due to these WCOBRA/TRAC code changes, they are not included in thePCT reporting sheets, Tables 2 through 4.

There are no new regulatory commitments contained in this letter.

Please address any comments or questions regarding this matter to Tom Byrne at(980) 373-3249 (Tom. Byrne@duke-energy. com).

Sincerel,

Senior Vice PresidentNuclear OperationsCatawba, McGuire

Attachment 1Table 1 - Errors/Evaluation Model ChangesTable 2 - Peak Cladding Temperature Summary - McGuire Units 1 & 2Table 3 - Peak Cladding Temperature Summary - Catawba Unit 1Table 4 - Peak Cladding Temperature Summary - Catawba Unit 2

Attachment 2 - Additional Information on the Evaluation of Revised Heat Transfer MultiplierDistributions for Plants Licensed with the CQD Evaluation Model

U.S. Nuclear Regulatory CommissionAugust 29, 2013

xc (with attachments):

V. M. McCree, Region II AdministratorU.S. Nuclear Regulatory CommissionMarquis One Tower245 Peachtree Center Avenue NE, Suite 1200Atlanta, GA 30303-1257

J. C. Paige, Senior Project Manager (CNS & MNS)U. S. Nuclear Regulatory Commission11555 Rockville PikeMail Stop 0-8C2ARockville, MD 20852-2738

J. Zeiler, NRC Senior Resident InspectorMcGuire Nuclear Station

G. A. Hutto, NRC Senior Resident InspectorCatawba Nuclear Station

ATTACHMENT 1

Table I - Errors/Evaluation Model Changes

Table 2 - Peak Cladding Temperature Summary - McGuire Units 1 & 2

Table 3 - Peak Cladding Temperature Summary - Catawba Unit I

Table 4 - Peak Cladding Temperature Summary - Catawba Unit 2

Attachment 1, Page 1 of 10

References:

A) Letter, G. R. Peterson (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46,Changes to or Errors in an ECCS Evaluation Model," April 11, 2001. [ADAMS ML01 1070266]

B) Letter, M. S. Tuckman (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46,Changes to or Errors in an ECCS Evaluation Model," April 3, 2002. [ADAMS ML021070672]

C) Letter, W. R. McCollum, Jr. (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46,Changes to or Errors in an ECCS Evaluation Model," July 291, 2003. [ADAMS ML032170639]

D) Letter, W. R. McCollum, Jr. (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46,Changes to or Errors in an ECCS Evaluation Model," May 26, 2004. [ADAMS ML041560349]

E) Letter, J. R. Morris (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46, Changes toor Errors in an ECCS Evaluation Model," June 21, 2005. [ADAMS ML051790210]

F) Letter, T. C. Geer (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46, Changes toor Errors in an ECCS Evaluation Model," March 13, 2007. [ADAMS ML070800546]

G) Letter, T. C. Geer (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46, Changes toor Errors in an ECCS Evaluation Model," May 22, 2007. [ADAMS ML071500297]

H) Letter, D. C. Culp (Duke Energy) to USNRC, Subject: Catawba Nuclear Station Units 1 and 2,and McGuire Nuclear Station Units 1 and 2, Response to Information Request Pursuant to 10CFR 50.54(f) Related to the Estimated Effect on Peak Cladding Temperature Resulting fromThermal Conductivity Degradation in the Westinghouse-Furnished Realistic Emergency CoreCooling System Evaluation and 30-Day Report Pursuant to 10 CFR 50.46, Changes to orErrors in an Evaluation Model," March 16, 2012. [ADAMS ML12079A180]

I) Letter, J. Thompson (USNRC) to K. Henderson and S. D. Capps (Duke Energy), Subject:Catawba Nuclear Station Units 1 and 2, and McGuire Nuclear Station Units 1 and 2, ClosureEvaluation for Report Pursuant to Title 10 of the Code of Federal Regulations, Part 50,Section 50.46, Paragraph (a)(3)(ii) Concerning Significant Emergency Core Cooling SystemEvaluation Model Error Related to Nuclear Fuel Thermal Conductivity Degradation (TAC Nos.ME8447, ME8448, ME8449, and ME8450)" November 16, 2012. [ADAMS ML12314A031]

J) Letter, M. J. Annacone, (Duke Energy) to USNRC, "Report Pursuant to 10 CFR 50.46,Changes to or Errors in an ECCS Evaluation Model," July 11, 2013. [ADAMS ML1 3199A279]

K) Westinghouse Electric Company Letter LTR-LIS-13-348; McGuire Units 1 & 2 and CatawbaUnits 1 & 2 - 10 CFR 50.46 Report for Revised Heat Transfer Multiplier Distributions, July 31,2013.

L) Westinghouse Electric Company Letter LTR-LIS-13-346; 10 CFR 50.46 Notification andReporting for WCOBRA/TRAC Changes and Error Corrections, July 30, 2013.


Table 1Errors I Evaluation Model Changes

Revised Heat Transfer Multiplier Distributions(Data provided in Reference K)

Affected Evaluation Model(s) Applicable to Catawba/McGuire:1996 Westinghouse Best Estimate Large Break LOCA Evaluation Model

Several changes and error corrections were made to WCOBRA/TRAC and the impacts of thesechanges on the heat transfer multiplier uncertainty distributions were investigated. During thisinvestigation, errors were discovered in the development of the original multiplier distributions,including errors in the grid locations specified in the WCOBRA/TRAC models for the G2 Refill andG2 Reflood tests, and errors in processing test data used to develop the reflood heat transfermultiplier distribution. Therefore, the blowdown heatup, blowdown cooling, refill, and reflood heattransfer multiplier distributions were redeveloped. For the reflood heat transfer multiplierdevelopment, the evaluation time windows for each set of test experimental data and each testsimulation were separately defined based on the time at which the test or simulation exhibiteddispersed flow film boiling heat transfer conditions characteristic of the reflood time period. Therevised heat transfer multiplier distributions have been evaluated for impact on existing analyses.Resolution of these issues represents a closely related group of Non-Discretionary Changes inaccordance with Section 4.1.2 of WCAP-13451.

A plant transient calculation representative of McGuire Units 1 and 2 and Catawba Units 1 and 2transient behavior was performed with the latest version of WCOBRA/TRAC. Using thistransient, HOTSPOT calculations were performed with both the original and revised heat transfermultiplier distributions. Based on the change in the 95 th percentile results, estimated PCT effectsof -40°F for Reflood 1 and -85°F for Reflood 2 have been established for 10 CFR 50.46 reportingpurposes for McGuire Units 1 and 2 and Catawba Units 1 and 2. For Catawba and McGuire, thelimiting PCT occurs late in the reflood phase (Attachment 2, Section 2.6, Late Reflood Limited).

Please see Attachment 2 for additional details on how Westinghouse performed the PCTevaluation of changes to the heat transfer multiplier distributions.


The following changes are described in Reference L, and are included herein for completeness,since these changes were incorporated into the WCOBRA/TRAC model used to assess the PCTimpact due to the revised heat transfer multiplier distributions described above.

ELEVATIONS FOR HEAT SLAB TEMPERATURE INITIALIZATION

Affected Evaluation Model(s) Applicable to Catawba/McGuire:1996 Westinghouse Best Estimate Large Break LOCA Evaluation Model

An error was discovered in WCOBRAITRAC whereby an incorrect value would be used in theinitial fuel rod temperature calculation for a fuel rod heat transfer node if that node elevation wasspecified outside of the bounds of the temperature initialization table. This problem has beenevaluated for impact on existing analyses and its resolution represents a Discretionary Change inaccordance with Section 4.1.1 of WCAP-1 3451.

Based on inspection of plant analysis input, it was concluded that the input decks for existinganalyses are not impacted by this error, leading to an estimated peak cladding temperatureimpact of 0°F.

HEAT TRANSFER MODEL ERROR CORRECTIONS

Affected Evaluation Model(s) Applicable to Catawba/McGuire: 1996 BELOCA

Several related changes were made to WCOBRA/TRAC to correct errors discovered whichaffected the heat transfer models. These errors included calculation of the entrained liquidfraction used in calculation of the drop wall heat flux, application of the grid enhancement factorfor grid temperature calculation, calculation of the Reynold's number used in the Wong-Hochrietercorrelation for the heat transfer coefficient from fuel rods to vapor, fuel rod initialization andcalculation of cladding inner radius with creep, application of grid and two phase enhancementfactors and radiation component in single phase vapor heat transfer, and reset of the critical heatflux temperature when J=2. These errors have been evaluated to estimate the impact on existingLBLOCA analysis results. Correction of these errors represents a closely-related group of Non-Discretionary Changes in accordance with Section 4.1.2 of WCAP-1 3451.

Based on the results of representative plant calculations, separate effects and integral effects testsimulations, it is concluded that the error corrections have a negligible local effect on heattransfer, leading to an estimated peak cladding temperature impact of 0°F.


CORRECTION TO HEAT TRANSFER NODE INITIALIZATION


An error was discovered in the heat transfer node initialization logic in WCOBRA/TRAC wherebythe heat transfer node center locations could be inconsistent with the geometric node centerelevations. The primary effects of this issue are on the interpolated fluid properties and gridturbulent mixing enhancement at the heat transfer node. This problem has been evaluated forimpact on existing analyses and its resolution represents a Non-Discretionary Change inaccordance with Section 4.1.2 of WCAP-13451.

Based on engineering judgment and the results from a matrix of representative plant calculations,it is concluded that the effect of this error is within the code resolution, leading to an estimatedpeak cladding temperature impact of 0°F.

MASS CONSERVATION ERROR FIX


It was identified that mass was not conserved in WCOBRA/TRAC one-dimensional componentcells when void fraction values were calculated to be slightly out of the physical range (greaterthan 1.0 or smaller than 0.0). This was observed to result in artificial mass generation on thesecondary side of steam generator components. Correction of this problem represents a Non-Discretionary Change in accordance with Section 4.1.2 of WCAP-1 3451.

This error was observed to primarily affect the mass on the secondary side of the steamgenerator. This issue was judged to have a negligible impact on existing LBLOCA analysisresults, leading to an estimated peak cladding temperature impact of 0°F.

CORRECTION TO SPLIT CHANNEL MOMENTUM EQUATION


An error was discovered in the momentum equation calculations for split channels inWCOBRAITRAC. This error impacts the (1) continuity area of the phantom/boundary bottom cell;(2) bottom and top continuity area correction factors for the channel inlet at the bottom of asection and for the channel outlet at the top of a section; and (3) drop entrainment mass rate perunit volume and drop de-entrainment mass rate per unit volume contributions to the momentumcalculations for split channels. This problem has been evaluated for impact on existing analysesand its resolution represents a Non-Discretionary Change in accordance with Section 4.1.2 ofWCAP-13451.

Based on the results from a matrix of representative plant calculations, it is concluded that theeffect of this error on the quantities directly impacted by the momentum equation calculations forsplit channels (velocities, flows, etc.) is negligible, leading to an estimated peak claddingtemperature impact of 0°F.


HEAT TRANSFER LOGIC CORRECTION FOR ROD BURST CALCULATION


A change was made to the WCOBRAITRAC coding to correct an error which had disabled rodburst in separate effect test simulations. This change represents a Discretionary Change inaccordance with Section 4.1.1 of WCAP-1 3451.

Based on the nature of the change and the evaluation model requirements for plant modeling inWestinghouse best estimate large break LOCA analyses with WCOBRA/TRAC, it is judged thatexisting analyses are not impacted by this change, leading to an estimated peak claddingtemperature impact of 0°F.

CHANGES TO VESSEL SUPERHEATED STEAM PROPERTIES


Several related changes were made to the WCOBRA/TRAC coding for the vessel super-heatedwater properties, including updating the HGAS subroutine coding to be consistent with Reference1 Equation 10-6, updating the approximation of the enthalpy in the TGAS subroutine to beconsistent with the HGAS subroutine coding, and updating the temperature iteration method andconvergence criteria in the TGAS subroutine. These changes represent a closely-related groupof Non-Discretionary Changes in accordance with Section 4.1.2 of WCAP-1 3451.

The updates to the calculations of the superheated steam properties had generally less than 1 OFimpact on the resulting steam temperature values, leading to an estimated peak claddingtemperature impact of 0°F.

Reference1. WCAP-12945-P-A, Volume 1, Revision 2, and Volumes 2 through 5, Revision 1, "CodeQualification Document for Best Estimate LOCA Analysis," 1998.

UPDATE TO METAL DENSITY REFERENCE TEMPERATURES


It was identified that for one-dimensional components in which heat transfer to stainless steel 304or 316 is modeled, the reference temperature for the metal density calculation was allowed tovary; as a result the total metal mass was not preserved. Correction of this problem represents aNon-Discretionary Change in accordance with Section 4.1.2 of WCAP-1 3451.

This change primarily impacts the reactor coolant system loop piping modeled in the large breakloss-of coolant accident (LBLOCA) WCOBRA/TRAC models. It was judged that the effect of thischange on the peak cladding temperature results was negligible, leading to an estimated peakcladding temperature impact of 0°F.


DECAY HEAT MODEL ERROR CORRECTIONS


The decay heat model in the WCOBRA/TRAC code was updated to correct the erroneouslycoded value of the yield fraction directly from fission for Group 19 of Pu-239, and to include theterm for uncertainty in the prompt energy per fission in the calculation of the decay heat poweruncertainty. Correction of these errors represents a closely-related group of Non-DiscretionaryChanges in accordance with Section 4.1.2 of WCAP-1 3451.

These changes have a negligible impact on the calculated decay heat power, leading to anestimated peak cladding temperature impact of 0°F.

CORRECTION TO THE PIPE EXIT PRESSURE DROP ERROR


An error was discovered in WCOBRAITRAC whereby the frictional pressure drop at the splitbreak TEE connection to the BREAK component was incorrectly calculated using the TEEhydraulic diameter instead of the BREAK component length input. This error has been evaluatedfor impact on existing analyses and its resolution represents a Non-Discretionary Change inaccordance with Section 4.1.2 of WCAP-13451.

Based on the results from a matrix of representative plant calculations, it is concluded that theeffect of this error on the pressure at the break and the break flow is negligible, leading to anestimated peak cladding temperature impact of 0°F.

WCOBRA/TRAC U19 FILE DIMENSION ERROR CORRECTION


A problem was identified in the dimension of an array used to generate the u19 file inWCOBRA/TRAC. The u19 file is read during HSDRIVER execution and provides informationneeded to generate the HOTSPOT thermal-hydraulic history and user input files. The array usedto write the desired information to the u19 file is dimensioned to 2000 in WCOBRAITRAC. It ispossible, however, for more than 2000 curves to be written to the u19 file. If that is the case, it ispossible that the curves would not be stored correctly on the ul 9 file. A survey of current BestEstimate Large Break LOCA analyses indicated that the majority of plants had less than 2000curves in their u19 files; therefore these plants are not affected by the change. For those plantswith more than 2000 curves, plant-specific sensitivity calculations indicated that resolution of thisissue does not impact the peak cladding temperature (PCT) calculation for prior analyses. Thisrepresents a Discretionary Change in accordance with Section 4.1.1 of WCAP-1 3451.

As discussed in the Background section, resolution of this issue does not impact the peakcladding temperature calculation for prior LBLOCA analyses, leading to an estimated peakcladding temperature impact of 0°F.


Table 2Peak Cladding Temperature Summary - McGuire Units 1 & 2

LBLOCA Cladding Temp Comments(OF)

Evaluation model : WCOBRA/TRAC, CQD 1996MNS/CNS

Analysis of record PCT (Reflood 2) 2028 Composite ModelPrior errors (APCT)

1. Decay heat in Monte Carlo calculations 8 Reference A2. MONTECF power uncertainty correction 20 Reference B3. Safety Injection temperature range 59 Reference C4. Input error resulting in an incomplete solution matrix 25 Reference D5. Revised Blowdown Heatup Uncertainty Distribution 5 Reference E6. Vessel Unheated Conductor Noding 0 Reference F7. Thermal Conductivity Degradation with Peaking 15 References H, I

Factor BurndownPrior evaluation model changes (APCT)

1. Revised Algorithm for Average Fuel Temperature 0 Reference F2. PAD 3.4 to PAD 4.0 -75 References H, I3. Peak FQ = 2.7 in bottom third of core 0 References H, I4. MUR Uprate to 101.7% of 3411 MWt 16 References H, I

Current Errors (APCT)

1. Revised Heat Transfer Multiplier Distribution -85 Reference KCurrent Evaluation model changes (APCT)1. None

Absolute value of errors/changes for this report (APCT) 85Net change in PCT for this report -85Final PCT 2016

SBLOCAEvaluation model: NOTRUMPAnalysis of record PCT 1323 2 inch break

Reference GPrior errors (APCT)

1. Evaluation of Fuel Pellet Thermal Conductivity 0 Reference JDegradationPrior evaluation model changes (APCT)

1. None 0Current Errors (APCT)

1. None 0Current Evaluation model changes (APCT)

1. None 0Absolute value of errors/changes for this report (APCT) 0Net change in PCT for this report 0Final PCT 1323


Table 3Peak Cladding Temperature Summary - Catawba Unit I

LBLOCA Cladding Temp Comments('IF)

Evaluation model : WCOBRAITRAC, CQD 1996MNS/CNS

Analysis of record PCT (Reflood 2) 2028 Composite ModelPrior errors (APCT)


Factor BurndownPrior evaluation model changes (APCT)

1. Revised Algorithm for Average Fuel Temperature 0 Reference F2. PAD 3.4 to PAD 4.0 -75 References H, I3. Peak FQ = 2.7 in bottom third of core 0 References H, I

Current Errors (APCT)1. Revised Heat Transfer Multiplier Distribution -85 Reference K

Current Evaluation model changes (APCT)1. None

Absolute value of errors/changes for this report (APCT) 85Net change in PCT for this report -85Final PCT. 2000


Reference GPrior errors (APCT)

1. Evaluation of Fuel Pellet Thermal Conductivity 0 Reference JDegradationPrior evaluation model changes (APCT)

1. None 0Current Errors (APCT)

1. None 0Current Evaluation model changes (APCT)



Table 4Peak Cladding Temperature Summary - Catawba Unit 2

LBLOCA Cladding Temp Comments(OF)Evaluation model: WCOBRA/TRAC, CQD 1996

MNS/CNSAnalysis of record PCT (Reflood 2) 2028 Composite ModelPrior errors (APCT)


Factor BurndownPrior evaluation model changes (APCT)1. Revised Algorithm for Average Fuel Temperature 0 Reference F2. PAD 3.4 to PAD 4.0 -75 References H, I3. Peak FQ = 2.7 in bottom third of core 0 References H, I

Current Errors (APCT)1. Revised Heat Transfer Multiplier Distribution -85 Reference K

Current Evaluation model changes (APCT)1. None

Absolute value of errors/changes for this report (APCT) 85Net change in PCT for this report -85Final PCT 2000


Reference G

Prior errors (APCT)1. Evaluation of Fuel Pellet Thermal Conductivity 0 Reference J

DegradationPrior evaluation model changes (APCT)

1. None 0Errors (APCT)

1. Evaluation of Fuel Pellet Thermal Conductivity 0DegradationEvaluation model changes (APCT)



ATTACHMENT 2

Additional Information on the Evaluation of Revised Heat Transfer Multiplier Distributionsfor Plants Licensed with the CQD EM

(Reference: Westinghouse Electric Company Letter LTR-LIS-13-406; Additional Information onthe Evaluation of Revised Heat Transfer Multiplier Distributions, August 14, 2013.)

1.0 Background on Error Identification and Reporting

As a result of code development and maintenance, several errors in the WCOBRAITRAC codeused for best estimate large break loss of coolant (BELOCA) analysis in the Code QualificationDocument (CQD, Reference [1]) and ASTRUM (Reference [2]) evaluation models (EMs) wereidentified. Some of the errors affected the WCOBRA/TRAC heat transfer models, the heattransfer node initialization or the heat transfer renoding logic, as well as other models. Thesechanges to WCOBRA/TRAC were described in Reference [3].

As a result of these changes, the following uncertainty distributions used in the CQD andASTRUM EMs were investigated for potential impact:

" Critical flow* Downcomer condensation* Upper plenum drain distribution (condensation and interfacial drag for upper plenum

injection)• Blowdown heatup heat transfer• Blowdown cooling heat transfer" Refill heat transfer" Reflood heat transfer

The results for the Separate Effects Tests (SETs) and Integral Effects Tests (lETs) used todetermine each of the potentially impacted uncertainty distributions were examined, comparingresults between the latest version of WCOBRA/TRAC (Version MOD7A Revision 8, with all ofthe errors listed in Reference [3] corrected) and WCOBRA/TRAC Version MOD7A Revision 6(which was used in the licensing of the ASTRUM EM in Reference [2]). It was determined thatthe results for the SETs and lETs used to develop the critical flow, downcomer condensation,and upper plenum drain uncertainty distributions were sufficiently similar; therefore, thosedistributions did not require changes. It was also confirmed that emergency core cooling (ECC)bypass predictions remain conservative. However, it was determined that the heat transfermultiplier distributions required additional investigation.

During the investigation into the potential impact on the heat transfer multiplier distributions,errors were identified in the development of the original multiplier distributions, including errorsin the grid locations specified in the WCOBRA/TRAC models for the G2 Refill and G2 RefloodSETs, and errors in processing test data used to develop the reflood heat transfer multiplierdistribution. These errors were also corrected and, using latest released version ofWCOBRAITRAC, the revised blowdown heatup, blowdown cooling, refill and reflood heattransfer multiplier distributions were determined.


2.0 Revised Distributions and Expected Effects

2.1 Background on Heat Transfer Multiplier Sampling

In order to sample heat transfer multipliers, a percentile for each time period heat transfermultiplier is sampled. That point is then converted to the heat transfer multiplier value based onthe cumulative distribution function (CDF) of the time period heat transfer multiplier. Figure 1illustrates this concept for a change from an old distribution to a new one (note that this CDFdoes not represent any actual CDF for the heat transfer multipliers, but is used simply fordemonstration). For example, if the 2 5 th percentile is sampled, Figure 1 shows that a multiplierof about 0.65 would be obtained for the old distribution. For the new distribution, the sampled25t" percentile would result in a multiplier of about 1.15.

2.2 Changes to the Heat Transfer Multiplier Distributions

The CDFs of the heat transfer multipliers changed as follows:

* Blowdown heatup heat transfer multipliers increased for low multipliers and across mostof the middle of the sampling range, and were mostly unchanged for the highestmultipliers.

* Blowdown cooling heat transfer multipliers decreased slightly from the top of the rangethrough the middle, and were mostly unchanged for low multipliers.

* Refill heat transfer multipliers decreased considerably at the top end of the range andgradually reduced to a slight decrease at the bottom end of the range. Although themagnitude of the change to the refill multiplier distribution was larger than that observedin the other distributions, the PCT impact is small because heat transfer rates are lowduring the nearly adiabatic refill time period.

* Reflood heat transfer multipliers increased at the bottom end of the range and themiddle, and then decreased at the top end of the range.

The implications of these changes are dependent on the behavior of plant transients. For theassessment, plants were classified as follows:

* Blowdown limited: A limiting PCT typically within the first 20 seconds of the transient." Early reflood limited: A limiting PCT after the end of the refill time period, but within about

the first 70 seconds of the transient." Mid reflood limited: A limiting PCT that is between the early and late reflood time

periods.* Late reflood limited: A limiting PCT generally after about 200 seconds.

The impact from the change to the heat transfer multiplier CDFs on each of these transient

types is discussed in the following subsections.

2.3 Blowdown Limited

Blowdown limited plants are only affected by the changes to the blowdown heatup heat transfermultiplier CDF. The increased heat transfer multipliers have a small benefit on PCT since theblowdown heatup time period is short.


2.4 Early Reflood Limited

Early reflood limited plants are affected by the changes to all of the heat transfer multiplierCDFs. The effects of the changes to the blowdown heatup and blowdown cooling heat transfermultiplier CDFs are limited since much of their effect diminishes through refill and the beginningof reflood. The effects of the changes to the refill heat transfer multiplier CDF are morepronounced since the early reflood PCT occurs shortly after the end of refill. The effects of thechanges to the reflood heat transfer multiplier CDF are limited since the run spends very littletime in the reflood time period prior to the PCT time.

2.5 Mid Reflood Limited

Mid reflood limited plants are affected by the changes to all of the heat transfer multiplier CDFs.The effects of the changes to the blowdown heatup and blowdown cooling heat transfermultiplier CDFs are very limited since most of their effect diminishes through refill and earlyreflood. The effects of the changes to the refill heat transfer multiplier CDF are limited sincemost of their effect diminishes through early reflood. The effects of the changes to the refloodheat transfer multiplier CDF are more pronounced due to the time over which the multiplier isapplied prior to the PCT time.

2.6 Late Reflood Limited

Late reflood limited plants are predominately affected by the change to the reflood heat transfermultiplier CDF. The effects of the changes to the blowdown heatup, blowdown cooling, andrefill heat transfer multiplier CDFs are negligible since their effect diminishes entirely throughoutthe lengthy reflood period. The effect of the change to the reflood heat transfer multiplier CDFcan be significant due to the longer time over which the multiplier is applied prior to the PCTtime.

3.0 Methodology for the Estimate of Effect

3.1 Selection and Description of Representative Transients

Representative PCT transients were used in determining the estimated PCT effect due to therevised heat transfer multiplier distributions. Heat transfer multipliers are applied in HOTSPOT;the HOTSPOT code performs a one-dimensional conduction calculation modeling the effect oflocal uncertainties on the hot rod, using thermal hydraulic boundary conditions taken fromWCOBRAITRAC. Plant characteristics determine the typical PCT transient behavior for theplant. Transients from different plants with similar PCT behavior tend to have fairly consistentthermal hydraulic characteristics around the hot rod. As a result, the choice of representativeplant was based on PCT transient behavior for the evaluation of the revised heat transfermultiplier distributions.

The representative transients discussed above were performed with the latest released versionof WCOBRA/TRAC, which incorporated correction of all of the errors listed in Reference [3].The representative transients were similar to Reference Transient calculations. Fuelperformance data which explicitly reflects burnup-dependent effects of thermal conductivitydegradation (TCD), calculated as described in Reference 4, was used for the representativecalculations.


3.2 Background of the CQD EM

Section 2.1 gives a high level description of sampling methodology. In the CQD EM, HOTSPOTruns 1000 calculations with randomly sampled local uncertainty attributes and produces 9 5 th

percentile results. For the CQD plant evaluations, two representative transients were executedto assess the early/mid reflood plants and late reflood plants; because the CQD EM individuallytracks the Reflood 1 and Reflood 2 PCTs, one representative plant was sufficient to representboth early and mid reflood plants.

3.3 Estimates of Effect

For each representative transient, the WCOBRA/TRAC calculation described in Section 3.1 wasexecuted. The results from this WCOBRA/TRAC calculation provided boundary conditions forexecution of the HOTSPOT code with the old and new heat transfer multiplier distributions. Theestimated effect for the Blowdown, Reflood 1, and Reflood 2 time periods for eachrepresentative plant calculation was determined from the 95th percentile HOTSPOT resultsusing the old heat transfer multiplier distributions, and the 95th percentile HOTSPOT resultsusing the revised heat transfer multiplier distributions. For these evaluation calculations, the twolatest released versions of HOTSPOT were used; the only difference between these HOTSPOTversions that affects the calculated results is the heat transfer multiplier distributions.

3.4 Applicability of the TCD Evaluations

It has been previously observed that explicitly considering TCD does not significantly impact thenature of the overall plant transient behavior and thermal-hydraulic response. In addition, theWCOBRAITRAC calculations described in Section 3.1 were performed using fuel performancedata which explicitly accounted for the effects of TCD. Therefore, the revised heat transfermultiplier distributions would be expected to have similar effect on the base and sensitivitycalculations executed to evaluate the effects of TCD and peaking factor burndown. The revisedheat transfer multiplier distributions do not invalidate the prior estimated effects for TCD.

3.5 Applicability of the Uncertainty Calculations

HOTSPOT runs are used in several steps of the uncertainty calculations in the CQDmethodology; thus, the changes in heat transfer multiplier distributions could have impact on thefinal Monte Carlo simulations. However, based on Section 28-3-2 of Reference [1], as long asan EM change does not substantially change the nature of the transient, an estimate of effectbased on a Reference Transient for a representative plant is sufficient. The heat transfermultiplier changes are only applied in HOTSPOT; therefore, the nature of the transient remainsunchanged. Because the representative calculations are meant to represent ReferenceTransient conditions (altered for the effects of TCD), the method used herein is consistent withthe approach described in Reference [1].

4.0 Summary of Effects and Observed Trends

As described in Section 3.2, in the CQD EM, each HOTSPOT calculation is comprised of 1000calculations where the local uncertainty attributes are sampled for each iteration from theirrespective distributions; the overall 95th percentile PCT results from the 1000 iterations are theresult of interest for the heat transfer multiplier evaluations. As such, the results are indicativeof generic trends due to the overall changes in the distributions.


For the early/mid reflood limited representative transient, the Reflood 1 PCT may be consideredrepresentative of early reflood PCT and Reflood 2 PCT may be considered representative ofmid-reflood PCT. For the early/mid reflood limited representative transient, the Reflood 1 PCTexperienced a small penalty, which is consistent with the expectations from Section 2.4 due tothe reduction in refill heat transfer multipliers. The Reflood 2 PCT of this representativetransient experienced a moderate benefit, which is consistent with expectations from Section2.5 due to the increase in reflood heat transfer multipliers over the low end of the range.

For the late reflood limited representative transient, the Reflood 1 PCT may be consideredrepresentative of mid-reflood PCT and Reflood 2 PCT may be considered representative of late-reflood PCT. For the late reflood limited representative transient, the Reflood 1 PCTexperienced a moderate benefit, which is consistent with expectations from Section 2.5 due tothe increase in reflood heat transfer multipliers over the low end of the range. The Reflood 2PCT of this representative transient experienced a large benefit, which is consistent withexpectations from Section 2.6 due to the increase in reflood heat transfer multipliers over thelow end of the range and the longer time for which the multiplier is applied.

5.0 References

1. WCAP-12945-P-A, Volume 1, Revision 2, and Volumes 2 through 5, Revision 1, "CodeQualification Document for Best Estimate LOCA Analysis," March 1998.

2. WCAP-16009-P-A, "Realistic Large-Break LOCA Evaluation Methodology Using theAutomated Statistical Treatment Of Uncertainty Method (ASTRUM)," January 2005.

3. LTR-LIS-1 3-346, "10 CFR 50.46 Notification and Reporting for WCOBRA/TRAC Changesand Error Corrections," July 2013.

4. LTR-NRC-12-27, "Westinghouse Input Supporting Licensee Response to NRC 10 CFR50.54(f) Letter Regarding Nuclear Fuel Thermal Conductivity Degradation (Proprietary/Non-Proprietary)," March 2012.


1.00 ~1 ~ I

0.80 - ---

0.7S . . . . . . ...

0.70 [----- . ..

0.65

0.60 ___ _

S0.55 .. . . .0 .50 -. . . . . . . . . .

0.40Ne

0.35 030l --- -d---0.25 • .... .. ..... ..

015

0.10

0.00_.. . . . . . .... .[.

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50

Hest Transfer Multipher [-1

Figure 1: Example Heat Transfer Multiplier Cumulative Distribution Function

(Note that this CDF does not represent any actual CDF for the heat transfer multipliers, but is used simplyfor illustrative purposes)


~duke nergy. catawba, · 2013-09-11 · analyses are not impacted by this error, leading to an...

Documents