Tennessee Valley Authority, Post Office Box 2000, …Tennessee Valley Authority, Post Office Box...

Tennessee Valley Authority, Post Office Box 2000, Decatur, Alabama 35609-2000

September 19, 2008

TVA-BFN-TS-418

10 CFR 50.90

U.S. Nuclear Regulatory CommissionATTN: Document Control DeskMail Stop OWFN, P1-35Washington, D. C. 20555-0001

Gentlemen:

In the Matter of ) Docket Nos. 50-260Tennessee Valley Authority ) 50-296

BROWNS FERRY NUCLEAR PLANT (BFN) - UNITS 2 AND 3 - TECHNICALSPECIFICATIONS (TS) CHANGE TS-418 - EXTENDED POWER UPRATE (EPU) -SUPPLEMENTAL RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION (RAI)ROUNDS 3 AND 18 AND RESPONSE TO ROUND 20 FUELS METHODS RAIs(TAC NOS. MD5263 AND MD5264)

By letter dated June 25, 2004 (ADAMS Accession No. ML041840301), TVA submitted alicense amendment application to the NRC for EPU operation of BFN Units 2 and 3. Thepending EPU amendment increases the maximum authorized power level byapproximately 14 percent from 3458 megawatts thermal (MWt) to 3952 MWt.

On July 17, 2008, NRC issued a Round 18 RAI (ML081700102) regarding AREVA fuelmethods used in support of Units 2 and 3 EPU operations. Round 18 consists of 32 RAIquestions, SRXB-91 through SRXB-122. To facilitate the review of TS-418, a meetingwas held on August 7, 2008, with NRC staff to review draft responses to SRXB-91 throughSRXB-116. Subsequently, on August 15, 2008, TVA submitted a partial response(ML082330187) to Round 18; specifically to RAIs SRXB-92, 93, 95, 96, 97, 99, 100, and102 through 116. This submittal responds to the remainder of the Round 18 RAIs.

Additionally, NRC staff conducted an audit of AREVA fuel methods from August 18through August 28, 2008, at the AREVA engineering facilities in Richland, Washington.As a result of the audit, TVA agreed to provide supplemental responses to a number ofthe August 15, 2008, Round 18 RAI responses and also to Round 3 RAIs SRXB-A.34 andSRXB-A.42. The original TS-418 Round 3 response was submitted on March 7, 2006(ML060680583). Lastly, this submittal also provides responses to the five fuels relatedRAIs, SRXB-123 through SRXB-127, from the NRC Round 20 RAI datedSeptember 16, 2008.

Tennessee Valley Authority, Post Office Box 2000, Decatur, Alabama 35609-2000

September 19, 2008

TVA-BFN-TS-418

U.S. Nuclear Regulatory Commission ATTN: Document Control Desk

. Mail Stop OWFN, P1-35 Washington, D. C. 20555-0001

Gentlemen:

In the Matter of Tennessee Valley Authority

) )

10 CFR 50.90

Docket Nos. 50-260 50-296

BROWNS FERRY NUCLEAR PLANT (BFN) - UNITS 2 AND 3 - TECHNICAL SPECIFICATIONS (TS) CHANGE TS-418 - EXTENDED POWER UPRATE (EPU)SUPPLEMENTAL RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION (RAI) ROUNDS 3 AND 18 AND RESPONSE TO ROUND 20 FUELS METHODS RAls (TAC NOS. MD5263 AND MD5264)

By letter dated June 25,2004 (ADAMS Accession No. ML041840301), TVA submitted a license amendment application to the NRC for EPU operation of BFN Units 2 and 3. The pending EPU amendment increases the maximum authorized power level by approximately 14 percent from 3458 megawatts thermal (MWt) to 3952 MWt.

On July 17, 2008, NRC issued a Round 18 RAI (ML081700102) regarding AREVA fuel methods used in support of Units 2 and 3 EPU operations. Round 18 consists of 32 RAI questions, SRXB-91 through SRXB-122. To facilitate the review of TS-418, a meeting was held on August 7,2008, with NRC staff to review draft responses to SRXB-91 through SRXB-116. Subsequently, on August 15, 2008, TVA submitted a partial response (ML082330187) to Round 18; specifically to RAls SRXB-92, 93, 95, 96, 97, 99,100, and 102 through 116. This submittal responds to the remainder of the Round 18 RAls.

Additionally, NRC staff conducted an audit of AREVA fuel methods from August 18 through August 28,2008, at the AREVA engineering facilities in Richland, Washington. As a result of the audit, TVA agreed to provide supplemental responses to a number of the August 15,2008, Round 18 RAI responses and also to Round 3 RAls SRXB-A.34 and SRXB-A.42. The original TS-418 Round 3 response was submitted on March 7, 2006 (ML060680583). Lastly, this submittal also provides responses to the five fuels related RAls, SRXB-123 through SRXB-127, from the NRC Round 20 RAI dated September 16,2008.

U.S. Nuclear Regulatory CommissionSeptember 19, 2008

As discussed with the NRC Project Manager for BFN, Ms. Eva Brown, onSeptember 17, 2008, responses to remainder of the Round 20 RAIs related to steamdryers, along with supplemental responses to Round 19 RAIs EMCB.147 andEMCB.192/150 pertaining to steam dryer analyses will be provided at a later date.

Enclosure 1 is a proprietary response to the subject RAIs and contains information thatAREVA NP, Inc. (AREVA) considers to be proprietary in nature and subsequently,pursuant to 10 CFR 9.17(a)(4), 2.390(a)(4) and 2.390(d)(1), AREVA requests that suchinformation be withheld from public disclosure. Enclosure 2 is a redacted version ofEnclosure 1 with the proprietary material removed and is suitable for public disclosure.Enclosure 3 contains an affidavit from AREVA supporting this request for withholding frompublic disclosure.

TVA has determined that the additional information provided by this letter does not affectthe no significant hazards considerations associated with the proposed TS changes. Theproposed TS changes still qualify for a categorical exclusion from environmental reviewpursuant to the provisions of 10 CFR 51.22(c)(9).

No new regulatory commitments are made in this submittal. If you have any questionsregarding this letter, please contact me at (256)729-7658.

I declare under penalty of perjury that the foregoing is true and correct. Executed on this19th day of September, 2008.

Sincerely,

0 . T. L ngleSite Licensing andIndustry Affairs Manager

Enclosures:1. Supplemental Response to Request for Additional Information (RAI) Rounds 3 and 18

and Response to Round 20 Fuels Methods RAIs (Proprietary Information Version)2. Supplemental Response to Request for Additional Information (RAI) Rounds 3 and 18

and Response to Round 20 Fuels Methods RAIs (Non-Proprietary InformationVersion)

3. Affidavit

U.S. Nuclear Regulatory Commission Page 2 September 19, 2008

As discussed with the NRC Project Manager for BFN, Ms. Eva Brown, on September 17,2008, responses to remainder of the Round 20 RAls related to steam dryers, along with supplemental responses to Round 19 RAls EMCB.147 and EMCB.192/150 pertaining to steam dryer analyses will be provided at a later date.

Enclosure 1 is a proprietary response to the subject RAls and contains information that AREVA NP, Inc. (AREVA) considers to be proprietary in nature and subsequently, pursuant to 10 CFR 9.17(a)(4), 2.390(a)(4) and 2.390(d)(1), AREVA requests that such information be withheld from public disclosure. Enclosure 2 is a redacted version of Enclosure 1 with the proprietary material removed and is suitable for public disclosure. Enclosure 3 contains an affidavit from AREVA supporting this request for withholding from public disclosure.

TVA has determined that the additional information provided by this letter does not affect the no significant hazards considerations associated with the proposed TS changes. The proposed TS changes still qualify for a categorical exclusion from environmental review pursuant to the provisions of 10 CFR 51.22(c)(9).

No new regulatory commitments are made in this submittal. If you have any questions regarding this letter, please contact me at (256)729-7658.

I declare under penalty of perjury that the foregoing is true and correct. Executed on this 19th day of September, 2008.

Sincerely,

Site Licensing and Industry Affairs Manager

Enclosures: 1. Supplemental Response to Request for Additional Information (RAI) Rounds 3 and 18

and Response to Round 20 Fuels Methods RAls (Proprietary Information Version) 2. Supplemental Response to Request for Additional Information (RAI) Rounds 3 and 18

and Response to Round 20 Fuels Methods RAls (Non-Proprietary Information Version)

3. Affidavit


Enclosures:cc (Enclosures):

State Health OfficerAlabama State Department of Public HealthRSA Tower - AdministrationSuite 1552P.O. Box 303017Montgomery, Alabama 36130-3017

Ms. Eva Brown, Project ManagerU.S. Nuclear Regulatory Commission(MS 08G9)One White Flint, North11555 Rockville PikeRockville, Maryland 20852-2739

Eugene F. Guthrie, Branch ChiefU.S. Nuclear Regulatory CommissionRegion IISam Nunn Atlanta Federal Center61 Forsyth Street, SW, Suite 23T85Atlanta, Georgia 30303-8931

NRC Resident InspectorBrowns Ferry Nuclear Plant10833 Shaw RoadAthens, Alabama 35611-6970

u.s. Nuclear Regulatory Commission Page 3 September 19, 2008

Enclosures: cc (Enclosures):

State Health Officer Alabama State Department of Public Health RSA Tower - Administration Suite 1552 P.O. Box 303017 Montgomery, Alabama 3(3130-3017

Ms. Eva Brown, Project Manager U.S. Nuclear Regulatory Commission (MS 08G9) One White Flint, North 11555 Rockville Pike Rockville, Maryland 20852-2739

Eugene F. Guthrie, Branch Chief U.S. Nuclear Regulatory Commission Region II Sam Nunn Atlanta Federal Center 61 Forsyth Street, SW, Suite 23T85 Atlanta, Georgia 30303-8931

NRC Resident Inspector Browns Ferry Nuclear Plant 10833 Shaw Road Athens, Alabama 35611-6970


JEE:BCM:BDLcc (w/o Enclosures):

G. P. Arent, EQB 1B-WBNW. R. Campbell, Jr., LP 3R-CS. M. Douglas, POB 2C-BFNR. F. Marks, Jr., PAB 1C-BFND. C. Matherly, BFT 2A-BFNL. E. Nicholson, LP 4K-CL. E. Thibault, LP 3R-CR. G. West, NAB 2A-BFNB. A. Wetzel, LP 4K-CS. A. Vance, WT 6A-KE. J. Vigluicci, ET 11A-KNSRB Support, LP 5M-CEDMS WT CA-K,

s:licensing/lic/submit/subs/EPU/RAI Round 3 and 18 and Round 20 Fuels Methods FAIs/Supplemental Response toRequest for Additional Information (RAI) Rounds 3 and 18 and Response to Round 20 Fuels Methods RAIs

u.s. Nuclear Regulatory Commission Page 4 September 19, 2008

JEE:BCM:BDL cc (w/o Enclosures):

G. P. Arent, EQB 1B-WBN W. R. Campbell, Jr., LP 3R-C S. M. Douglas, POB 2C-BFN R. F. Marks, Jr., PAB 1C-BFN D. C. Matherly, BFT 2A-BFN L. E. Nicholson, LP 4K-C L. E. Thibault, LP 3R-C R. G. West, NAB 2A-BFN B. A. Wetzel, LP 4K-C S. A. Vance, WT 6A-K E. J. Vigluicci, ET 11 A-K NSRB Support, LP 5M-C EDMS WT CA-K,

s:licensing/lic/submitlsubs/EPU/RAI Round 3 and 18 and Round 20 Fuels Methods FAls/Supplemental Response to Request for Additional Information (RAI) Rounds 3 and 18 and Response to Round 20 Fuels Methods RAls

NON-PROPRIETARY INFORMATION

ENCLOSURE 2

TENNESSEE VALLEY AUTHORITYBROWNS FERRY NUCLEAR PLANT (BFN)

UNITS 2 AND 3

TECHNICAL SPECIFICATIONS (TS) CHANGE TS-418EXTENDED POWER UPRATE (EPU)

SUPPLEMENTAL RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION (RAI)ROUNDS 3 AND 18 AND RESPONSE TO ROUND 20 FUELS METHODS RAIs

(NON-PROPRIETARY INFORMATION VERSION)

This enclosure provides TVA supplemental responses to Round 3 RAIs SRXB-A.34 andSRXB-A.42, a supplemental response to NRC's July 17, 2008, Round 18 RAI, and a responseto the five fuels methods related RAIs, SRXB-123 through SRXB-127, from NRC'sSeptember 16, 2008, Round 20 RAI.

7


ENCLOSURE 2

TENNESSEE VALLEY AUTHORITY BROWNS FERRY NUCLEAR PLANT (BFN)

UNITS 2 AND3

TECHNICAL SPECIFICATIONS (TS) CHANGE TS-418 EXTENDED POWER UPRATE (EPU)

SUPPLEMENTAL RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION (RAI) ROUNDS 3 AND 18 AND RESPONSE TO ROUND 20 FUELS METHODS RAls

(NON-PROPRIETARY INFORMATION VERSION)

This enclosure provides TVA supplemental responses to Round 3 RAls SRXB-A.34 and SRXB-A.42, a supplemental response to NRC's July 17, 2008, Round 18 RAI, and a response to the five fuels methods related RAls, SRXB-123 through SRXB-127, from NRC's September 16, 2008, Round 20 RAI.

,/


NRC staff conducted an audit of AREVA fuel methods from August 18 through August 28, 2008,at the AREVA engineering facilities in Richland, Washington. As a result of the audit, TVAagreed to provide supplemental responses Round 3 RAIs SRXB-A.34 and SRXB-A.42. Theprevious Round 3 responses were originally submitted on March 7, 2006 (ML060680853). Arevised response to SRXB-A.34 was also submitted on May 11, 2006 (ML061360148).

NRC RAI SRXB-A.34 (From Round 3)

Describe qualitatively the cross-section reconstruction process incorporated in CASMO-4 andMICROBURN-B2. The response should reflect the information provided in the slides (1-35) ofthe August 4 presentations, including high void fraction effects and accuracy. Provide flowchart(s), road map(s) and any other means to demonstrate the process, starting from thegathered raw void fraction data, how that data is used by CASMO-4 to generate the requiredcross-sections. In addition, briefly describe the development of the void fraction correlation andassociated uncertainties.

Supplemental Response to SRXB-A.34

MICROBURN-B2 versions prior to 2003 treated cross section dependency on spectral historydifferently between the fuel nuclide depletion module and the neutron flux calculation module.The fuel nuclide depletion module used [

] while the neutron flux iteration calculation module useda []. Thisinconsistency was remedied starting in 2003 by changing the depletion module to the[ ]. Starting from 2006,both modules were converted to the [

].

These changes over the years were mainly due to code maintenance concerns and did notimpact any result due to the [

3. Unlike the cross section dependency on the instantaneous void, the [] is rather weak. This is shown in Figure SRXB-A.34.1 for Pu-239 and in

Figure SRXB-A.34.2 for Pu-240. The [

]. At the high end of [ ], the differencebetween the [

]. This kind of difference isentirely within the uncertainty of nuclear cross section measurement and its evaluation processincluding the CASMO-4 lattice code.' It has no observable effect on the reactor nodal powerdistribution and the reactor criticality evaluation as has been verified in the code maintenancerecord of MICROBURN-B2.

E2-1


NRC staff conducted an audit of AREVA fuel methods from August 18 through August 28, 2008, at the AREVA engineering facilities in Richland, Washington. As a result of the audit, TVA agreed to provide supplemental responses Round 3 RAls SRXB-A.34 and SRXB-A.42. The previous Round 3 responses were originally submittecj on March 7, 2006 (ML060680853). A revised response to SRXB-A.34 was also submitted on May 11,2006 (ML061360148).


Describe qualitatively the cross-section reconstruction process incorporated in CASMO-4 and MICROBURN-B2. The response should reflect the information provided in the slides (1-35) of the August 4 presentations, including high void fraction effects and accuracy. Provide flow chart(s), road map(s) and any other means to demonstrate the process, starting from the gathered raw void fraction data, how that data is used by CASMO-4 to generate the required cross-sections. In addition, briefly describe the development of the void fraction correlation and associated uncertainties.


MICROBURN-B2 versions prior to 2003 treated cross section dependency on spectral history differently between the fuel nuclide depletion module and the neutron flux calculation module. The fuel nuclide depletion module used [

] while the neutron flux iteration calculation module used a [ ]. This inconsistency was remedied starting in 2003 by changing the depletion module to the· [ ]. Starting from 2006, both modules were converted to the [

].

These changes over the years were mainly due to code maintenance concerns and did not impact any result due to the [

]. Unlike the cross section dependency on the instantaneous void, the [ ] is rather weak. This is shown in Figure SRXB-A.34.1 for Pu-239 and in

Figure SRXB-A.34.2 for Pu-240. The [

1 At the high end of [ ], the difference between the [

]. This kind of difference is entirely within the uncertainty of nuclear cross section measurement and its evaluation process including the CASMO-4 .lattice code. It has no observable effect on the reactor nodal power distribution and the reactor criticality evaluation as has been verified in the code maintenance record of MICROBURN-B2. .

E2-1


r"

Figure SRXB-A.34.1 PU-239 sigma-1 Dependenceon Spectral History at 20 Gigawatt-days per ton (GWd/T)

r-

-UFigure SRXB-A.34.2 PU-240 sigma-1 Dependence

on Spectral History at 20 GWd/T

E2-2

r

r

NON-PROPRIETARY INFqRMATION

. Figure SRXB-A.34.1 PU-239 sigma-1 Dependence on Spectral History at 20 Gigawatt-days per ton (GWd/T)

Figure SRXB-A.34.2 PU-240 sigma-1 Dependence on Spectral History at 20 GWd/T

E2-2

..J



In August 30, 2004, General Electric Nuclear Energy (GENE) issued a 10 CFR Part 21 report(ADAMS ML042720293), stating that using limiting control rod blade patterns developed for lessthan rated flow at rated power conditions could sometimes yield more limiting bundle-by-bundleMCPR distributions and/or more limiting bundle axial power shapes than using limiting controlrod patterns developed for rated flow/rated power in the SLMCPR calculation. The affectedplants submitted amendment requests increasing their SLMCPR value. The staff understandsthat Framatome did not issue a Part 21 reporting on the SLMCPR methodology that addressesthe calculation of the SLMCPR at minimum core flow and off-rated conditions similar to GENE'sPart 21 report.

Reference the applicable sections of the ANF-524P-A SLMCPR methodology that specify therequirement to calculate the SLMCPR at the worst case conditions for minimum core flowconditions for rated power. Demonstrate that the SLMCPR is calculated at different statepointsof the licensed operating domain, including the minimum core flow statepoint and that thecalculation is performed for different exposure points.


AREVA NP 1 performs the safety limit Minimum Critical Power Ratio (SLMCPR) analysis on acycle-specific basis. As discussed in the original response to SRXB-A.42 (ReferenceSRXB-A.42.1), the core power distributions used in the SLMCPR analyses are obtained fromthe MICROBURN-B2 cycle-specific design basis step-through calculation. SLMCPR analysesare performed with these power distributions at the minimum and maximum core flow allowed atrated power.

The SLMCPR analyses supporting the BFN Unit 2 EPU submittal were performed for anATRIUMTM-10 2 equilibrium cycle that assumed Maximum Extended Load Line Limit Analysisplus (MELLLA+) operation. The BFN EPU SLMCPR analyses considered the minimum andmaximum flow at rated power for planned MELLLA+ operation. The cycle-specific SLMCPRanalyses supporting current operating cycles for BFN Units 2 and 3 were performed consistentwith the currently allowed power/flow maps for these cycles and did not include the MELLLA+flow window. Future cycle-specific BFN SLMCPR analyses will be performed consistent withthe allowable power/flow map for the cycle.

The AREVA SLMCPR methodology uses a design basis core power distribution. The criteria forselecting the design basis power distribution are specified in Reference SRXB-A.42.2 and statethat analyses be performed with power distributions that "...conservatively represent expectedreactor operating states which could both exist at the MCPR operating limit and produce aMCPR equal to the MCPR safety limit during an anticipated operational occurrence." Candidatedesign basis power distributions are obtained from the cycle-specific design step-through. Thedesign step-through reflects the cycle design energy and operating strategy planned by theutility and is the best projection of how the cycle will operate.

The design step-through is required to have margin to the operating limit MCPR (OLMCPR).Flatter (less peaked) radial power distributions are conservative for the SLMCPR analysis. Theradial power distributions from the cycle step-through are flatter than the radial power

1 AREVA NP Inc. is an AREVA and Siemens company.2 ATRIUM is a trademark of AREVA NP.

E2-3



In August 30,2004, General Electric Nuclear Energy (GENE) issued a 10 CFR Part 21 report (ADAMS ML042720293), stating that using limiting control rod blade patterns developed for less than rated flow at rated power conditions could sometimes yield more limiting bundle-by-bundle MCPR distributions and/or more limiting bundle axial power shapes than using limiting control rod patterns developed for rated flow/rated power in the SLMCPR calculation. The affected plants submitted amendment requests increasing their SLMCPR value. The staff understands that Framatome did not issue a Part 21 reporting on the SLMCPR methodology that addresses the calculation of the SLMCPR at minimum core flow and off-rated conditions similar to GENE's Part 21 report.

Reference the applicable sections of the ANF-524P-A SLMCPR methodology that specify the requirement to calculate the SLMCPR at the worst case conditions for minimum core flow conditions for rated power. Demonstrate that the SLMCPR is calculated at different statepoints of the licensed operating domain, including the minimum core flow statepoint and that the calculation is performed for different exposure points.


AREVA NP1 performs the safety limit Minimum Critical Power Ratio (SLMCPR) analysis on a cycle-specific basis. As discussed in the original response to SRXB-A.42 (Reference SRXB-A.42.1), the core power distributions used in the SLMCPR analyses are obtained from the MICROBURN-B2 cycle-specific design basis step-through calculation. SLMCPR analyses are performed with these power distributions at the minimum and maximum core flow allowed at rated power.

The SLMCPR analyses supporting the BFN Unit 2 EPU submittal were performed for an ATRIUMTM-102 equilibrium cycle that assumed Maximum Extended Load Line Limit Analysis plus (MELLLA+) operation. The BFN EPU SLMCPR analyses considered the minimum and maximum flow at rated power for planned MELLLA+ operation. The cycle-specific SLMCPR analyses supporting current operating cycles for BFN Units 2 and 3 were performed consistent with the currently allowed power/flow maps for these cycles and did not include the MELLLA+ flow window. Future cycle-specific BFN SLMCPR analyses will be performed consistent with the allowable power/flow map for the cycle.

The AREVA SLMCPR methodology uses a design basis core power distribution. The criteria for selecting the design basis power distribution are specified in Reference SRXB-A.42.2 and state that analyses be performed with power distributions that" ... conservatively represent expected reactor operating states which could both exist at the MCPR operating limit and produce a MCPR equal to the MCPR safety limit during an anticipated operational occurrence." Candidate design basis power distributions are obtained from the cycle-specific design step-through. The design step-through reflects the cycle design energy and operating strategy planned by the utility and is the best projection of how the cycle will operate.

The design step-through is required to have margin to the operating limit MCPR (OLMCPR). Flatter (less peaked) radial power distributions are conservative for the SLMCPR analysis. The radial power distributions from the cycle step-through are flatter than the radial power

1

2 AREVA NP Inc. is an AREVA and Siemens company. ATRIUM is a trademark of AREVA NP.

E2-3


distributions that would result from adjusting the control rod patterns until the core OLMCPR isreached. These control rod adjustments would result in a more peaked radial power distributionand increased margin to the SLMCPR. The design margin to the OLMCPR ensures that thepower distributions from the cycle step-through are conservative relative to the powerdistributions that may occur during actual operation of the cycle.

Figure SRXB-A.42.1 provides a comparison of the core radial power distribution from the designstep-through and from actual operation for a BWR/4 at EPU. The power distributions are at thecycle exposure that was limiting for the SLMCPR analysis. The figure shows that the actualpower distribution had a higher radial power distribution and is less flat than the designstep-through power distribution. In addition, for actual operation there was still 5.1% MCPRmargin. These comparisons demonstrate that the radial power distribution used in the SLMCPRanalysis is conservative relative to the required SLMCPR design basis power distribution andbounds actual operation.

Reference:

SRXB-A.42.1

SRXB-A.42.2

Correspondence, W.D. Crouch (TVA) to U.S. Nuclear RegulatoryCommission, "Browns Ferry Nuclear Plant (BFN) - Units 2 and 3, Responseto NRC Round 3 Requests for Additional Information Related to TechnicalSpecifications (TS) Change No. TS-418 - Requests for Extended PowerUprate Operation (TAC Nos. MC3743 and MC3744)," March 7, 2006(ML060680583).

ANF-524(P)(A) Revision 2 and Supplements 1 and 2, ANF Critical PowerMethodology for Boiling Water Reactors, Advanced Nuclear FuelsCorporation, November 1990

E2-4


distributions that would result from adjusting the control rod patterns until the core OLMCPR is reached. These control rod adjustments would result in a more peaked radial power distribution and increased margin to the SLMCPR. The design margin to the OLMCPR ensures that the power distributions from the cycle step-through are conservative relative to the power distributions that may occur during actual operation of the cycle.

Figure SRXB-A.42.1 provides a comparison of the core radial power distribution from the design step-through and from actual operation for a BWRl4 at EPU. The power distributions are at the cycle exposure that was limiting for the SLMCPR analysis. The figure shows that the actual power distribution had a higher radial power distribution and is less flat than the design step-through power distribution. In addition, for actual operation there was still 5.1 % MCPR margin. These comparisons demonstrate that the radial power distribution used in the SLMCPR analysis is conservative relative to the required SLMCPR design basis power distribution and bounds actual operation.

Reference:

SRXB-A.42.1

SRXB-A.42.2

Correspondence, W.D. Crouch (TVA) to U.S. Nuclear Regulatory Commission, "Browns Ferry Nuclear Plant (BFN) - Units 2 and 3, Response to NRC Round 3 Requests for Additional Information Related to Technical Specifications (TS) Change No. TS-418 - Requests for Extended Power Uprate Operation (TAC Nos. MC3743 and MC3744)," March 7, 2006 (M L060680583).

ANF-524(P)(A) Revision 2 and Supplements 1 and 2, ANF Critical Power Methodology for Boiling Water Reactors, Advanced Nuclear Fuels Corporation, November 1990

E2-4


1.45

1.4 * Design Step-Through

* Actual Operation

1.35

1.3

U. 1.250

-a 1.2

1.15

1.1

1.05

250 300 350 4000 50 100 150 200

Assembly Rank

Figure SRXB-A.42.1 Design vs.Actual Radial Power Factors

E2-5

is u "'

1.45 r

1.4 r --135K-1.3 " -


• Design Step-Through

<> Actual Operation

u.. 1.25 +---'-' ... -~o;:------------------

; ~ c;; 1.2 =s "' a:

1.15 +---------------"""11 .. -------

1.1 --

1.05 f-==----------- -----------~iiI!ooo.

o 50 100 150 200

Assembly Rank

250

Figure SRXB-A.42.1 Design vs. Actual Radial Power Factors

E2-5

300 350 400


On July 17, 2008, NRC issued a Round 18 RAI (ML081700102) regarding AREVA fuel methodsused in support of Units 2 and 3 EPU operations. Round 18 consists of 32 RAI questions,SRXB-91 through SRXB-122. To facilitate the review of TS-418, a meeting was held onAugust 7, 2008 with NRC staff to review draft responses to SRXB-91 through SRXB-1 16.Subsequently, on August 15, 2008, TVA submitted a partial response (ML082330187) toRound 18; specifically to RAIs SRXB-92, 93, 95, 96, 97, 99, 100, and 102 through 116. Beloware responses to the remainder of the Round 18 RAIs. Additionally, NRC staff conducted anaudit of AREVA fuel methods from August 18 through August 28, 2008, at the AREVAengineering facilities in Richland, Washington. As a result of the audit, TVA agreed to providesupplemental responses to a number of the August 15, 2008, Round 18 RAI responses, whichare also provided below as indicated.

NRC Introduction to Round 18 RAI

Table 1.3 in Enclosure 5 to the letter dated June 25, 2004, indicates that the COTRANSA2Version AAPR03 computer code was used to evaluate the anticipated transient without scram(ATWS) - overpressurization event. The licensee cites a May 31, 2000, letter from the NuclearRegulatory Commission (NRC) to Framatome (now AREVA) to support the use of COTRANSA2for the ATWS-overpressurization abnormal operating occurrence (AOO).

NRC RAI SRXB-91

In Enclosure 1 of the letter dated March 7, 2006, Tennessee Valley Authority (TVA) providesinformation in support of the use of the Ohkawa-Lahey void quality correlation againstATRIUM-10 test data in response'to SRXB-A.35. The Ohkawa-Lahey void quality correlationappears to under-predict the void fraction for the majority of the thermodynamic qualities testedat 6.9 Megapascal (MPa). The void reactivity coefficient is sensitive to the instantaneous voidfraction, generally becoming more negative with increasing void fraction.

Provide a quantitative determination of the impact of the bias in the void fraction inCOTRANSA2 on ATWS overpressure analysis results for the bottom head peak pressure. Thisshould include a comparison of the impact of the void bias to the margin between the peakcalculated pressure and the American Society of Mechanical Engineers Boiler & PressureVessel Code (ASME) acceptance criterion of 1500 pounds per square inch gage.

In addition, address how known biases are taken into account for future cycle specificcalculations and for bundle designs other than ATRIUM-1 0.

Clarifications Provided by the NRC following a meeting on August 7, 2008

Address the void bias for both the anticipated transient without scram (ATWS) overpressure aswell as ASME overpressure.

Response to SRXB-91

AREVA performs cycle-specific ATWS analyses of the short-term reactor vessel peak pressureusing the COTRANSA2 computer code. The ATWS peak pressure calculation is a core widepressurization event that is sensitive to similar phenomena as other pressurization transients.Bundle design is included in the development of input for the coupled neutronic and thermal

E2-6


On July 17,2008, NRC issued a Round 18 RAI (ML081700102) regarding AREVA fuel methods used in support of Units 2 and 3 EPU operations. Round 18 consists of 32 RAI questions, SRXB-91 through SRXB-122. To facilitate the review of TS-418, a meeting was held on August 7, 2008 with NRC staff to review draft responses to SRXB-91 through SRXB-116. Subsequently, on August 15, 2008, TVA submitted a partial response (ML082330187) to Round 18; specifically to RAls SRXB-92, 93, 95, 96, 97, 99, 100, and 102 through 116. Below are responses to the remainder of the Round 18 RAls. Additionally, NRC staff conducted an audit of AREVA fuel methods from August 18 through August 28, 2008, at the AREVA engineering facilities in Richland, Washington. As a result of the audit, TVA agreed to provide supplemental responses to a number of the August 15, 2008, Round 18 RAI responses, which are also provided below as indicated.


Table 1.3 in Enclosure 5 to the letter dated June 25,2004, indicates that the COTRANSA2 Version AAPR03 computer code was used to evaluate the anticipated transient without scram (ATWS) - overpressurization event. The licensee cites a May 31,2000, letter from the Nuclear Regulatory Commission (NRC) to Framatome (now AREVA) to support the use of COTRANSA2 for the ATWS-overpressurization abnormal operating occurrence (AOO).

NRC RAI SRXB-91

In Enclosure 1 of the letter dated March 7, 2006, Tennessee Valley Authority (TVA) provides information in support of the use of the Ohkawa-Lahey void quality correlation against ATRIUM-10 test data in response to SRXB-A.35. The Ohkawa-Lahey void quality correlation appears to under-predict the void fraction for the majority of the thermodynamic qualities tested at 6.9 Megapascal (MPa). The void reactivity coefficient is sensitive to the instantaneous void fraction, generally becoming more negative with increasing void fraction.

Provide a quantitative determination of the impact of the bias in the void fraction in COTRANSA2 on ATWS overpressure analysis results for the bottom head peak pressure. This should include a comparison of the impact of the void bias to the margin between the peak calculated pressure and the American Society of Mechanical Engineers Boiler & Pressure Vessel Code (ASME) acceptance criterion of 1500 pounds per square inch gage.

In addition, address how known biases are taken into account for future cycle specific calculations and for bundle designs other than ATRIUM-10.

Clarifications provided by the NRC following a meeting on August 7. 2008

Address the void bias for both the anticipated transient without scram (ATWS) overpressure as well as ASME overpressure.

Response to SRXB-91

AREVA performs cycle-specific ATWS analyses of the short-term reactor vessel peak pressure using the COTRANSA2 computer code. The ATWS peak pressure calculation is a core wide pressurization event that is sensitive to similar phenomena as other pressurization transients. Bundle design is included in the development of input for the coupled neutronic and thermal

E2-6


hydraulic COTRANSA2 core model. Important inputs to the COTRANSA2 system model arebiased in a conservative direction.

The AREVA analysis methods and the correlations used by the methods are applicable for bothpre-EPU and EPU conditions as discussed in responses (ML060680583) to previous RAIs(SRXB-A.15, SRXB-A.26 through SRXB-A.29, and SRXB-A.35). The transient analysismethodology is a deterministic bounding approach that contains sufficient conservatism to offsetbiases and uncertainties in individual phenomena. For bundle designs other than ATRIUM-10,the void-quality correlation is robust as discussed in the response (ML082330187) to RAISRXB-93 for past and present fuel designs. For future fuel designs, the void-quality correlationwould be reviewed for applicability, which may involve additional verification and validation.

A sensitivity study was performed for the limiting ATWS pressurization event for BFN Unit 3Cycle 14 with EPU to assess the bias between the ATRIUM-10 test data and the void-qualitycorrelation. The event was a pressure regulator failure-open (PRFO), which is adepressurization event, followed by pressurization due to main steam line isolation valve (MSIV)closure. The neutronics input included the impact of the fuel depleted with the changes in thevoid-quality correlation. To remove the bias in the MICROBURN-B2 neutronics input, the[ ] void-quality correlation was modified. To address the bias in the Ohkawa-Laheyvoid-quality correlation for the COTRANSA2 code, the void-quality relationship was changed toa [ ]. Additionally, the sensitivity study was repeated without depleting thefuel with the changes in the void-quality correlation (the change in the void-quality correlationwas instantaneous at the exposure of interest).

The reference ATWS case had a peak vessel pressure of 1477 pounds per square inch gauge(psig). The change in the void-quality correlations resulted in a 10-psi increase in the peakvessel pressure. The results for an instantaneous change in the void-quality correlation showedthe same impact. A study was also performed for the ASME overpressure event for BFN Unit 3Cycle 14 with EPU. The event was the MSIV closure with flux scram. The change in thevoid-quality correlations resulted in a 7 psi increase in the peak vessel pressure. The impact ofa change in the bias of the void-quality correlations on peak pressure is expected to be morethan offset by the model conservatisms. However, until quantitative values of the conservatismscan be demonstrated, TVA has directed AREVA to include a 10-psi increase to the peak vesselpressure for the EPU ATWS overpressure analysis and a 7-psi increase to the peak vesselpressure for the EPU ASME overpressure analysis.

NRC RAI SRXB-94

The initial steam flow rate at extended power uprate (EPU) conditions is higher than at pre-EPUconditions, and the transient power pulse is expected to be higher during the pressurization.The suppression pool temperature for Units 2 and 3 is based on an analysis for GE14 fuel.Provide a discussion on the means used to confirm that the results of the GE 14 analysis arebounding for ATRIUM-10 fuel. This justification should contain qualitative discussion regardingthe impact of the differences in nuclear characteristics and should consider the timing andnature of the transient power response during pressurization, relief, and boration.

Response to SRXB-94

The higher initial steam flow at EPU conditions will result in a slightly higher power pulse duringthe initial relatively short pressurization phase of the ATWS event. However, the total energyreleased to the suppression pool is dominated by the later much longer phase of the eventwhere power is reduced after the recirculation pumps trip and the core power is slowly reduced

E2-7


hydraulic COTRANSA2 core model. Important inputs to the COTRANSA2 system model are biased in a conservative direction.

The AREVA analysis methods and the correlations used by the methods are applicable for both pre-EPU and EPU conditions as discussed in responses (ML060680583) to previous RAls (SRXB-A.15, SRXB-A.26 through SRXB-A.29, and SRXB-A.35). The transient analysis methodology is a deterministic bounding approach that contains sufficient conservatism to offset biases and uncertainties in individual phenomena. For bundle designs other than ATRIUM-10, the void-quality correlation is robust as discussed in the response (ML082330187) to RAI SRXB-93 for past and present fuel designs. For future fuel designs, the void-quality correlation would be reviewed for applicability, which may involve additional verification and validation.

A sensitivity study was performed for the limiting ATWS pressurization event for BFN Unit 3 Cycle 14 with EPU to assess the bias between the ATRIUM-1 0 test data and the void-quality correlation. The event was a pressure regulator failure-open (PRFO), which is a depressurization event, followed by pressurization due to main steam line isolation valve (MSIV) closure. The neutronics input included the impact of the fuel depleted with the changes in the void-quality correlation. To remove the bias in the MICROBURN-B2 neutronics input, the [ ] void-quality correlation was modified. To address the bias in the Ohkawa-Lahey void-quality correlation for the COTRANSA2 code, the void-quality relationship was changed to a [ ]. Additionally, the sensitivity study was repeated without depleting the fuel with the changes in the void-quality correlation (the change in the void-quality correlation was instantaneous at the exposure of interest).

The reference ATWS case had a peak vessel pressure of 1477 pounds per square inch gauge (psig). The change in the void-quality correlations resulted in a 10-psi increase in the peak vessel pressure. The results for an instantaneous change in the void-quality correlation showed the same impact. A study was also performed for the ASME overpressure event for BFN Unit 3 Cycle 14 with EPU. The event was the MSIV closure with flux scram. The change in the void-quality correlations resulted in a 7 psi increase in the peak vessel pressure. The impact of a change in the bias of the void-quality correlations on peak pressure is expected to be more than offset by the model conservatisms. However, until quantitative values of the conservatisms can be demonstrated, TVA has directed AREVA to include a 10-psi increase to the peak vessel pressure for the EPU ATWS overpressure analysis and a 7 -psi increase to the peak vessel pressure for the EPU ASME overpressure analysis.

NRC RAI SRXB·94

The initial steam flow rate at extended power uprate (EPU) conditions is higher than at pre-EPU conditions, and the transient power pulse is expected to be higher during the pressurization. The suppression pool temperature for Units 2 and 3 is based on an analysis for GE14 fuel. Provide a discussion on the means used to confirm that the results of the GE 14 analysis are bounding for ATRIUM-1 0 fuel. This justification should contain qualitative discussion regarding the impact of the differences in nuclear characteristics and should consider the timing and nature of the transient power response during pressurization, relief, and boration.

Response to SRXB·94

The higher initial steam flow at EPU conditions will result in a slightly higher power pulse during the initial relatively short pressurization phase of the ATWS event. However, the total energy released to the suppression pool is dominated by the later much longer phase of the event where power is reduced after the recirculation pumps trip and the core power is slowly reduced

E2-7


as boron injection occurs. The ATWS analyses performed for BFN Units 2 and 3 included theimpact of the higher initial steam flow at EPU conditions. As shown in Table 9-4 of ReferenceSRXB-94.1, the impact of EPU operation on the maximum suppression pool temperature is notsignificant (<1 OF). This supports the conclusion that the initial power pulse, which is higher forEPU operation, is not significant relative to the total energy transferred to the suppression pool.

The suppression pool temperature analyses were performed for BFN Units 2 and 3 with GE fuel(Reference SRXB-94.1). An evaluation was performed to compare fuel neutronic parametersimportant for the ATWS analysis (void coefficient, boron worth) for ATRIUM-1 0 and GE fuel.The boron worth characteristics of ATRIUM-1 0 were slightly better while the void reactivitycharacteristics were slightly worse relative to the impact on the ATWS suppression pooltemperature analysis.

Additional analyses were performed to assess the impact of the difference in fuel assemblyreactivity characteristics on the suppression pool temperature during an ATWS. [

IAll fuel types in the core designs including the GE fuel were explicitly modeled in the aboveanalyses consistent with the approved methodology. The GE fuel was modeled with a level ofdetail equivalent to that used for the ATRIUM-10 fuel. CASMO-4 analyses explicitly modeledthe water rod configuration of the GE fuel. MICROBURN-B2 was used to calculate the corereactivity characteristics provided to the COTRANSA2 analysis. The GE fuel assemblygeometric and nuclear characteristics (enrichment and gadolinia distribution) were based ondesign data provided to AREVA by TVA. The hydraulic characteristics for the GE fuelassemblies were based on GE fuel assembly pressure drop tests performed by AREVA.

The BFN ATWS analyses described above were performed for cycles operating at pre-EPUpower levels. However, as shown in Table 9-4 of Reference SRXB-94.1, the impact of EPUoperation on the maximum suppression pool temperature is not significant. Therefore, thetrends observed for ATRIUM-10 fuel in the above analyses are equally applicable for EPUoperation.

The analyses described above confirm that the suppression pool temperature analysisdocumented in Reference SRXB-94.1 is slightly conservative for ATRIUM-10 fuel. In addition,the analyses show that the difference in reactivity characteristics between ATRIUM-1 0 and GEfuel do not have a significant impact relative to the large margin to the suppression pooltemperature limit shown in Reference SRXB-94.1.

The conclusions of the Reference SRXB-94.1 suppression pool temperature analysis areapplicable for ATRIUM-1 0 fuel and the acceptance criteria will be met for BFN Units 2 and 3EPU operation with ATRIUM-10 fuel.

E2-8


as boron injection occurs. The ATWS analyses performed for BFN Units 2 and 3 included the impact of the higher initial steam flow at EPU conditions. As shown in Table 9-4 of Reference SRXB-94.1, the impact of EPU operation on the maximum suppression pool temperature is not significant «1°F). This supports the conclusion that the initial power pulse, which is higher for EPU operation, is not significant relative to the total energy transferred to the suppression pool.

The suppression pool temperature analyses were performed for BFN Units 2 and 3 with GE fuel (Reference SRXB-94.1). An evaluation was performed to compare fuel neutronic parameters important for the ATWS analysis (void coefficient, boron worth) for ATRIUM-10 and GE fuel. The boron worth characteristics of ATRIUM-10 were slightly better while the void reactivity characteristics were slightly worse relative to the impact on the ATWS suppression pool temperature analysis.

Additional analyses were performed to assess the impact of the difference in fuel assembly reactivity characteristics on the suppression pool temperature during an ATWS. [

]

A" fuel types in the core designs including the GE fuel were explicitly modeled in the above analyses consistent with the approved methodology. The GE fuel was modeled with a level of detail equivalent to that used for the ATRIUM-10 fuel. CASMO-4 analyses explicitly modeled the water rod configuration of the GE fuel. MICROBURN-B2 was used to calculate the core reactivity characteristics provided to the COTRANSA2 analysis. The GE fuel assembly geometric and nuclear characteristics (enrichment and gadolinia distribution) were based on design data provided to AREVA by TVA. The hydraulic characteristics for the GE fuel assemblies were based on GE fuel assembly pressure drop tests performed by AREVA.

The BFN ATWS analyses described above were performed for cycles operating at pre-EPU power levels. However, as shown in Table 9-4 of Reference SRXB-94.1, the impact of EPU operation on the maximum suppression pool temperature is not significant. Therefore, the trends observed for ATRIUM-10 fuel in the above analyses are equally applicable for EPU operation.

The analyses described above confirm that the suppression pool temperature analysis documented in Reference SRXB-94.1 is slightly conservative for ATRIUM-10 fuel. In addition, the analyses show that the difference in reactivity characteristics between ATRIUM-1 0 and GE fuel do not have a significant impact relative to the large margin to the suppression pool temperature limit shown in Reference SRXB-94.1.

The conclusions of the Reference SRXB-94.1 suppression pool temperature analysis are applicable for ATRIUM-10 fuel and the acceptance criteria wi" be met for BFN Units 2 and 3 EPU operation with ATRIUM-10 fuel.

E2-8


SRXB-94.1 NEDC-33047P Revision 2, Browns Ferry Units 2 and 3 Safety Analysis Reportfor Extended Power Uprate, June 2004. (ML041840301)

Table SRXB-94.1 Energy Release toSuppression Pool

NRC RAI SRXB-98

It appears that COTRANSA2 has two centrifugal pump models, the first pump model neglectsthe inertia and the second pump model is based on homologous input. Identify which modeloption is used. If the second model option is used, verify that it is used to model the dualrecirculation pump trip during ATWS evaluations. Verify that the homologous input for therecirculation pumps for the Unit 2 analyses have been benchmarked against operational data atUnit 2.

Response to SRXB-98

The second pump model based on homologous input is used to model the dual recirculationpump trip during ATWS evaluations. The homologous curves are from the pump manufacturer.The pump speed and flow are initialized from operational plant data. Frictional torque and pumpinertia are tuned to model the plant coastdown rate.

E2-9


SRXB-94.1 NEDC-33047P Revision 2, Browns Ferry Units 2 and 3 Safety Analysis Report for Extended Power Uprate, June 2004. (ML041840301)

Table SRXB-94.1 Energy Release to Suppression Pool

[

]

NRC RAI SRXB-98

It appears that COTRANSA2 has two centrifugal pump models, the first pump model neglects the inertia and the second pump model is based on homologous input. Identify which model option is used. If the second model option is used, verify that it is used to model the dual recirculation pump trip during ATWS evaluations. Verify that the homologous input for the recirculation pumps for the Unit 2 analyses have been benchmarked against operational data at Unit 2.

Response to SRXB-98

The second pump model based on homologous input is used to model the dual recirculation pump trip during ATWS evaluations. The homologous curves are from the pump manufacturer. The pump speed and flow are initialized from operational plant data. Frictional torque and pump inertia are tuned to model the plant coastdown rate.

E2-9


NRC RAI SRXB-100

Section 2.1 of ANF-913(P)(A) states that cross sections are interpolated based on bothcontrolled and uncontrolled states at [[ ]] void fraction.These void cases appear to not be consistent with the void cases used to develop cross sectionresponse surfaces for MICROBURN-B2 [[ ]], explainthis discrepancy.

Supplemental Response to SRXB-100

In order to produce the COTRAN neutronic parameters, a series of MICROBURN-B2calculations are performed. These successive calculations are:

(1) Nominal initial conditions

(2)

E2-10


NRC RAI SRXB-100

Section 2.1 of ANF-913(P)(A) states that cross sections are interpolated based on both controlled and uncontrolled states at [[ ]] void fraction. These void cases appear to not be consistent with the void cases used to develop cross section response surfaces for MICROBURN-B2 [[ ]] , explain this discrepancy.


In order to produce the COTRAN neutronic parameters, a series of MICROBURN-B2 calculations are performed. These successive calculations are:

(1) Nominal initial conditions

(2)

E2-10


iThe 1½ energy group diffusion equation in steady-state can be written as

':a2) eP1+ ± V fIN

+ E-2. V"f 2 (I),2:a2

=0keff

The first term is a leakage.following figure.

This equation is integrated over the cylindrical node depicted in the

H

H

H

01j+1

fý Dr"+1j11

D1,i

Dl1 ,i-1

The leakage term is approximated as:

3 2DI,iDI,j(0 1,i- l,j) A

j=1 (DO,i + OD,j) HV

E2-11


]

The 1 Y2 energy group diffusion equation in steady-state can be written as

The first term is a leakage. This equation is integrated over the cylindrical node depicted in the following figure.

H

H

H

The leakage term is approximated as:

3 2D] .0] .(r!J]. -r!J] -) A _ L ,I,} ,I,} __

j=] (Dl,i + D],j) HV

E2-11


where

D1 ,i = D for plane of interestDj = D for the nodes adjacent to the plane of interest

01,i = flux in the plane of interest

014j = flux in the regions adjacent to the plane of interest

A = surface area between nodes i and jH = distance between nodes i and nodes j

V = node volume

E2-12

where

[

NON-PROPRI ETARY INFORMATION

Do = 0 for plane of interest

01,j = 0 for the nodes adjacent to the plane of interest

= flux in the plane of interest ([J1,i

C/J1,j = A

flux in the regions adjacent to the plane of interest

H

V

= surface area between nodes i and j

= distance between nodes i and nodes j

= node volume

E2-12


E2-13


E2-13


E2-14


I ,

E2-14


These final one-group cross section and leakage parameters are used in a new 1-dimensionalflux solution and the axial power distribution is updated for the next thermal hydraulic solution.Iterations between the 1-dimensional flux solution and the thermal hydraulic solution arerepeated until converged results are obtained for core power, power distribution, temperaturedistribution, and density distribution.

E2-15


These final one-group cross section and leakage parameters are used in a new 1-dimensional flux solution and the axial power distribution is updated for the next thermal hydraulic solution. Iterations between the 1-dimensional flux solution and the thermal hydraulic solution are repeated until converged results are obtained for core power, power distribution, temperature distribution, and density distribution.

E2-15


r..

Figure SRXB-100.1 Comparison of Scram Bank Worth for[ I

E2-16


r

.J

Figure SRXB-100.1 Comparison of Scram Bank Worth for [ ]

E2-16


NRC RAI SRXB-101

The Doppler coefficient is stated to be dependent on the broadening of the fast group crosssection and to be a function of fuel temperature.

* MICROBURN-B2 calculates the nodal fuel temperature based on quadratic fitting function.Provide this function. Discuss how the initial nodal fuel temperature is calculated. Provide acomparison of the quadratic function predicted nodal fuel temperature to results predictedusing a more sophisticated thermal rod conduction model and heat transfer coefficient, suchas XCOBRA-T.

* Expand on the discussion provided in ANF-913(P)(A) and describe what combination ofcalculations is performed to determine the reactivity contribution from Doppler for ATWSoverpressure analysis, for example, specify if a lattice calculation is performed to determinea coefficient relating microscopic cross sections to average fuel temperature.

* Discuss whether the rod temperatures in Section 2.1.3 of ANF-913(P)(A) are calculatedbased on a nodal average rod or for each rod in the node. Clarify how the transient nodalaverage fuel temperature is calculated.

* Provide a description of any differences between the COTRANSA2 thermal conductionmodels, including material properties, and the RODEX2 models. Discuss whether theRODEX2 code was used to develop input for COTRANSA2 similar to XCOBRA-T.

Response to SRXB-101

E2-17

[


NRC RAJ SRXB-101

The Doppler coefficient is stated to be dependent on the broadening of the fast group cross section and to be a function of fuel temperature.

• MICROBURN-B2 calculates the nodal fuel temperature based on quadratic fitting function. Provide this function. Discuss how the initial nodal fuel temperature is calculated. Provide a comparison of the quadratic function predicted nodal fuel temperature to results predicted using a more sophisticated thermal rod conduction model and heat transfer coefficient, such as XCOBRA-T.

• Expand on the discussion provided in ANF-913(P)(A) and describe what combination of calculations is performed to determine the reactivity contribution from Doppler for ATWS overpressure analysis, for example, specify if a lattice calculation is performed to determine a coefficient relating microscopic cross sections to average fuel temperature.

• Discuss whether the rod temperatures in Section 2.1.3 of ANF-913(P)(A) are calculated based on a nodal average rod or for each rod in the node. Clarify how the transient nodal average fuel temperature is calculated.

• Provide a description of any differences between the COTRANSA2 thermal conduction models, including material properties, and the RODEX2 models. Discuss whether the RODEX2 code was used to develop input for COTRANSA2 similar to XCOBRA-T.


E2-17


E2-18


E2-18


E2-19


E2-19


I

The RODEX2 computer code provides initial input information relative to core averagefuel-to-cladding gap heat transfer coefficients for the COTRANSA2 computer code. Assuch, RODEX2 uses steady-state heat conduction models. The heat conduction modelemployed by COTRANSA2 includes transient terms.

The fuel thermal conductivity correlations used by COTRANSA2 are equivalent to theRODEX2 models.

COTRANSA2 computes a fuel temperature for each axial plane in the core. Based onthe assumption of a core composition primarily consisting of uranium dioxide,COTRANSA2 does not account for gadolinium in the fuel thermal conductivitycalculation.

Heat capacities of fuel components (uranium dioxide, gadolinium, and cladding) are notrequired for the RODEX2 steady-state calculations, but are used in the COTRANSA2transient calculations.

The fuel pellet-to-cladding gap heat transfer coefficient used in COTRANSA2 is theproduct of a RODEX2 calculation.

E2-20


• The RODEX2 computer code provides initial input information relative to core average fuel-to-cladding gap heat transfer coefficients for the COTRANSA2 computer Gode. As such, RODEX2 uses steady-state heat conduction models. The heat conduction model employed by COTRANSA2 includes transient terms.

The fuel thermal conductivity correlations used by COTRANSA2 are equivalent to the RODEX2 models.

COTRANSA2 computes a fuel temperature for each axial plane in the core. Based on the assumption of a core composition primarily consisting of uranium dioxide, COTRANSA2 does not account for gadolinium in the fuel thermal conductivity calculation.

Heat capacities of fuel components (uranium dioxide, gadolinium, and cladding) are not required for the RODEX2 steady-state calculations, but are used in the COTRANSA2 transient calculations.

The fuel pellet-to-cladding gap heat transfer coefficient used in COTRANSA2 is the product of a RODEX2 calculation.

E2-20


Figure SRXB-101.1 RODEX Evolution of theDoppler Effective Fuel Temperature for

SPC Fuel at Constant Power-

E2-21

r


Figure SRX8-101.1 RODEX Evolution of the Doppler Effective Fuel Temperature for

SPC Fuel at Constant Power·

E2-21

..J


Figure SRXB-101.2 RODEX Evolution of theDoppler Effective Fuel Temperature for

SPC Fuel vs. LHGR and Burnup

E2-22

r


Figure SRXB-101.2 RODEX Evolution of the Doppler Effective Fuel Temperature for


E2-22

..J


Figure SRXB-101.3 MICROBURN-B2 Correlation Evolution of theDoppler Effective Fuel Temperature for


E2-23

r


Figure SRXB-101.3 MICROBURN-B2 Correlation Evolution of the Doppler Effective Fuel Temperature for


E2-23


NRC RAI SRXB-103

Provide the relationship of the term Feff to the S-factor. If axial integration is required todetermine the S-factors, specify how this is performed. Address whether the S-factors aresensitive to the bundle void distribution. Describe how the S-factors are determined forconditions typical (or bounding) for operation at EPU conditions.


Evaluations were performed to assess the impact on ACPR of a change in Feff resulting from thevariation in the lattice void fraction during a pressurization event. MICROBURN-B2 analyseswere performed using the nominal void correlation and an adjusted void correlation to assessthe change in Feff as void changes. The MICROBURN-B2 cases were run to reflect aninstantaneous change in core average void fraction of +0.05. For the limiting MCPR bundle inthe core, the changes in void, local peaking factor (LPF), and Feff were:

Avoid = +0.0441 (node 24)Avoid = +0.0456 (node 23)ALPF = -0.0026 (node 24)ALPF = -0.0030 (node 23)

AFeff = 0.0000 (assembly)

For other potentially limiting bundles (10% highest powered bundles) in the core, the change inFeff was between -0.0002 and +0.0011 for a +0.05 core average Avoid. In general, an increasein void fraction resulted in an increase in Feff for high power, low exposure (end of first cycle)assemblies and a decrease in Feff for low power, high exposure assemblies.

A decrease in Feff during the transient will improve the CPR during the transient and result in areduced ACPR. The converse is true for an increase in Feff during the transient. The sensitivityof MCPR to Feff is about 2 to 1; therefore, the sensitivity of ACPR is about twice the AFeff duringthe transient. The change in ACPR would be between 0.000 and +0.002 for a +0.05 coreaverage Avoid.

During a pressurization event, the core void will initially decrease followed by an increase incore void. Therefore, the effect of the change in void on fuel rod peaking factors (and Feff) willtend to be offset during the transient.

The assessment above for the impact of a void change on AFeff and A(ACPR) is based onassuming the nuclear power is instantly converted to surface heat flux. Because the time ofMCPR (-1.25 sec) is less than the fuel rod thermal time constant (- 5 sec), the actual impact onFeff and ACPR from the void change will be much less. At the boiling transition plane, there isan insignificant change in void until after the time of peak power. Because the increase in voidand the corresponding increase in Feff occur close to the time of MCPR, the slight change in rodpower will not significantly change the rod heat flux at the time of MCPR. Therefore, the effecton ACPR will be much less than estimated based on the MICROBURN-B2 analyses.

In summary, the above results show that the effect of the variation in void fraction during atransient on the Feff has an insignificant effect on ACPR.

E2-24

r NON-PROPRI ETARY INFORMATION

NRC RAI SRXB-103

Provide the relationship of the term Feff to the S-factor. If axial integration is required to determine the S-factors, specify how this is performed. Address whether the S-factors are sensitive to the bundle void distribution. Describe how the S-factors are determined for conditions typical (or bounding) for operation at EPU conditions.


Evaluations were performed to assess the impact on ~CPR of a change in Feff resulting from the variation in the lattice void fraction during a pressurization event. MICROBURN-B2 analyses were performed using the nominal void correlation and an adjusted void correlation to assess the change in Feff as void changes. The MICROBURN-B2 cases were run to reflect an instantaneous change in core average void fraction of +0.05. For the limiting MCPR bundle in the core, the changes in void, local peaking factor (LPF), and Feff were:

~void = +0.0441 (node 24) ~void = +0.0456 (node 23) ~LPF = -0.0026 (node 24) ~LPF = -0.0030 (node 23) ~Feff = 0.0000 (assembly)

For other potentially limiting bundles (10% highest powered bundles) in the core, the change in Feff was between -0.0002 and +0.0011 for a +0.05 core average ~void. In general, an increase in void fraction resulted in an increase in Feff for high power, low exposure (end of first cycle) assemblies and a decrease in Feff for low power, high exposure assemblies.

A decrease in Feff during the transient will improve the CPR during the transient and result in a reduced ~CPR. The converse is true for an increase in Feff during the transient. The sensitivity of MCPR to Feff is about 2 to 1; therefore, the sensitivity of ~CPR is about twice the ~Feff during the transient. The change in ~CPR would be between 0.000 and +0.002 for a +0.05 core average ~void.

During a pressurization event, the core void will initially decrease followed by an increase in core void. Therefore, the effect of the change in void on fuel rod peaking factors (and Feff) will tend to be offset during the transient.

The assessment above for the impact of a void change on ~Feff and ~(~CPR) is based on assuming the nuclear power is instantly converted to surface heat flux. Because the time of MCPR (-1.25 sec) is less than the fuel rod thermal time constant (- 5 sec), the actual impact on Feff and ~CPR from the void change will be much less. At the boiling transition plane, there is an insignificant change in void until after the time of peak power. Because the increase in void and the corresponding increase in Feff occur close to the time of MCPR, the slight change in rod power will not significantly change the rod heat flux at the time of MCPR. Therefore, the effect on ~CPR will be much less than estimated based on the MICROBURN-B2 analyses.

In summary, the above results show that the effect of the variation in void fraction during a transient on the Feff has an insignificant effect on ~CPR.

E2-24


NRC RAI SRXB-105

Verify that the Unit 2 transient analyses were performed using input options for closurerelationships that are consistent with the NRC approval of XCOBRA-T. This includes specifyingthe Levy subcooled boiling model, the Martinelli-Nelson two phase friction multipliers, the twophase component loss multiplier, the wall viscosity model, and thermodynamic properties fromthe ASME steam tables.

Revised Response to SRXB-105

The BFN Units 2 and 3 EPU transient analyses used the default models of XCOBRA-T. Thedefault models include the Levy subcooled boiling model, the Martinelli-Nelson two phasefriction multipliers, the two phase component loss multiplier, and the heated wall viscositycorrection model. [

] as discussed in a meeting with the NRC onMay 4,1995, (Reference SRXB-105.1). Thermodynamic properties from the ASME steamtables were used. The code provides a message if the default models are not used. PerAREVA's licensing analyses requirements, use of default models is required.

Reference:

SRXB-105.1 Correspondence, R.A. Copeland (Siemens) to R.C. Jones (NRC), "ATRIUM-10Presentations," RAC:95:080, May 4, 1995 (38-9091703-000).

NRC RAI SRXB-107

Address how the wall friction and component loss coefficients were determined for Unit 2.Address whether these parameters were input in the analysis to account for friction. Providethese parameters and the technical basis for their selection. Relative to pre-EPU conditions,channel flow tends to redistribute at EPU conditions as there are fewer low resistance bundlesin the core. Address whether the friction parameters were selected to be consistent with thisexpected trend.


During the NRC audit of AREVA codes and methods in Richland, Wa., from August 18 throughAugust 28, 2008, the NRC requested additional information regarding the background andprocess that [

Spacer Pressure Drop Testing

The Portable Hydraulic Test Facility (PHTF) is used by AREVA to obtain single phase losscoefficients for the spacers. The friction factor correlation is based on previous tests performedat the PHTF that remain applicable for current fuel designs (rods and channel have a consistentsurface condition). The pressure drops across the spacers are measured in the PHTF for eachnew fuel design. The PHTF has pressure taps just upstream of the spacers so that the flow willbe fully developed. The component of pressure drop due to friction is calculated and subtractedfrom the total measured pressure drop. The remaining pressure drop is due to the spacers andis used to determine the spacer pressure loss coefficients.

E2-25


NRC RAI SRXB-105

Verify that the Unit 2 transient analyses were performed using input options for closure relationships that are consistent with the NRC approval of XCOBRA-T. This includes specifying the Levy subcooled boiling model, the Martinelli-Nelson two phase friction multipliers, the two phase component loss multiplier, the wall viscosity model, and thermodynamic properties from the ASME steam tables.

Revised Response to SRXB-105

The BFN Units 2 and 3 EPU transient analyses used the default models of XCOBRA-T. The default models include the Levy subcooled boiling model, the Martinelli-Nelson two phase friction multipliers, the two phase component loss multiplier, and the heated wall viscosity correction model. [

] as discussed in a meeting with the NRC on May 4,1995, (Reference SRXB-1 05.1). Thermodynamic properties from the ASME steam tables were used. The code provides a message if the default models are not used. Per AREVA's licensing analyses requirements, use of default models is required.

Reference:

SRXB-105.1 Correspondence, R.A. Copeland (Siemens) to R.C. Jones (NRC), "ATRIUM-10 Presentations," RAC:95:080, May 4, 1995 (38-9091703-000).

NRC RAI SRXB-107

Address how the wall friction and component loss coefficients were determined for Unit 2. Address whether these parameters were input in the analysis to account for friction. Provide these parameters and the technical basis for their selection. Relative to pre-EPU conditions, channel flow tends to redistribute at EPU conditions as there are fewer low resistance bundles in the core. Address whether the friction parameters were selected to be consistent with this expected trend.


During the NRC audit of AREVA codes and methods in Richland, Wa., from August 18 through August 28,2008, the NRC requested additional information regarding the background and process that [

].

Spacer Pressure Drop Testing

The Portable Hydraulic Test Facility (PHTF) is used by AREVA to obtain single phase loss coefficients for the spacers. The friction factor correlation is based on previous tests performed at the PHTF that remain applicable for current fuel designs (rods and channel have a consistent surface condition). The pressure drops across the spacers are measured in the PHTF for each new fuel design. The PHTF has pressure taps just upstream of the spacers so that the flow will be fully developed. The component of pressure drop due to friction is calculated and subtracted from the total measured pressure drop. The remaining pressure drop is due to the spacers and is used to determine the spacer pressure loss coefficients.

E2-25


Preliminary ATRIUM-10 Spacer Loss Coefficients

Development of the ATRIUM-10 fuel design took place in Germany. Because PHTF pressuredrop testing was not complete, single phase pressure drop data for ATRIUM-10 was obtainedfrom the German development effort. For the use in preliminary ATRIUM-10 designassessments, the German data was used to develop single phase spacer pressure losscoefficients appropriate for use with Richland hydraulic models. Analyses using these singlephase losses resulted in an under prediction of the pressure drop data as shown inFigure SRXB-107.1. The spacer loss coefficients (K) used to generate the results presented inFigure SRXB-1 07.1 are of the form

K=A +BReC

where A, B, and C are constants and Re is the Reynolds number based on local fluid conditionsand geometry.

Until PHTF data was available for the ATRIUM-10 design, a means of adjusting theGerman-based pressure loss coefficients to better predict the pressure drop data usingRichland methods was developed. [

] areshown in Figure SRXB-107.2. The spacer loss coefficients (K) used to generate the resultspresented in Figure SRXB-107.2 are of the form

where [ ] for the ATRIUM-10 design.

Further development of ATRIUM-1 0 spacer loss coefficients was subsequently performedbased on PHTF ATRIUM-10 pressure drop data.

PHTF ATRIUM-10 Based Spacer Loss Coefficients

The ATRIUM-10 PHTF tests form the basis for the single phase loss coefficients currently usedfor design and licensing analyses supporting domestic BWRs. PHTF data was reduced todetermine single phase losses for the spacers in the lower (fully-rodded) region of the bundle,the spacer in the transition (end of part-length rods) region of the bundle, and the spacers in theupper (partially-rodded) region of the bundle.

Assessments of the predicted pressure drop relative to measured two phase pressure drop dataconfirmed the applicability of the [ ] for use with spacer pressure losscoefficients based on PHTF data. Results of analyses for each region of the bundle (lower,transition, upper) when using the PHTF spacer loss coefficients [ ] areshown in Figures SRXB-107.3, SRXB-107.4, and SRXB-107.5.

NRC Interactions

On May 4, 1995, a meeting was held with the NRC to describe the ATRIUM-10 design and theapplication of the approved AREVA methodology for the design. Two view graphs extractedfrom those presented at the meeting are provided in Figures SRXB-107.6 and SRXB-1 07.7.A summary of the May 4, 1995 meeting was provided to the NRC in Reference SRXB-107.1.

E2-26


Preliminary ATRIUM-10 Spacer Loss Coefficients

Development of the ATRIUM-10 fuel design took place in Germany. Because PHTF pressure drop testing was not complete, single phase pressure drop data for ATRIUM-10 was obtained from the German development effort. For the use in preliminary ATRIUM-10 design assessments, the German data was used to develop single phase spacer pressure loss coefficients appropriate for use with Richland hydraulic models. Analyses using these single phase losses resulted in an under prediction of the pressure drop data as shown in Figure SRXB-1 07.1. The spacer loss coefficients (K) used to generate the results presented in Figure SRXB-107.1 are of the form

K=A + B Rec

where A, B, and C are constants and Re is the Reynolds number based on local fluid conditions and geometry.

Until PHTF data was available for the ATRIUM-10 design, a means of adjusting the German-based pressure loss coefficients to better predict the pressure drop data using Richland methods was developed. [

] are shown in Figure SRXB-107.2. The spacer loss coefficients (K) used to generate the results presented in Figure SRXB-107.2 are of the form

[ ]

where [ ] for the ATRIUM-10 design.

Further development of ATRIUM-10 spacer loss coefficients was subsequently performed based on PHTF ATRIUM-10 pressure drop data.

PHTF ATRIUM-10 Based Spacer Loss Coefficients

The ATRIUM-10 PHTF tests form the basis for the single phase loss coefficients currently used for design and licensing analyses supporting domestic BWRs. PHTF data was reduced to determine single phase losses for the spacers in the lower (fully-rodded) region of the bundle, the spacer in the transition (end of part-length rods) region of the bundle, and the spacers in the upper (partially-rodded) region of the bundle.

Assessments of the predicted pressure drop relative to measured two phase pressure drop data confirmed the applicability of the [ ] for use with spacer pressure loss coefficients based on PHTF data. Results of analyses for each region of the bundle (lower, transition, upper) when using the PHTF spacer loss coefficients [ ] are shown in Figures SRXB-107.3, SRXB-107.4, and SRXB-107.5.

NRC Interactions

On May 4,1995, a meeting was held with the NRC to describe the ATRIUM-10 design and the application of the approved AREVA methodology for the design. Two view graphs extracted from those presented at the meeting are provided in Figures SRXB-107.6 and SRXB-107.7. A summary of the May 4, 1995 meeting was provided to the NRC in Reference SRXB-107.1.

E2-26


Applicability for EPU Operation

The ATRIUM-10 hydraulic models have been verified over a range of conditions that bound bothpre-EPU and EPU operating conditions. The applicability of the models is described andsupported by data presented in the thermal hydraulics section of the response to RAISRXB-A.15 (Reference SRXB-107.2).

References:

SRXB-107.1 Correspondence, R.A. Copeland (Siemens) to R.C. Jones (NRC), "ATRIUM-10Presentations," RAC:95:080, May 4, 1995 (38-9091703-000).

SRXB-107.2 Correspondence, W.D. Crouch (TVA) to U.S. Nuclear Regulatory Commission,"Browns Ferry Nuclear Plant (BFN) - Units 2 and 3, Response to NRC Round 3Requests for Additional Information Related to Technical Specifications (TS)Change No. TS-418 - Requests for Extended Power Uprate Operation (TAC Nos.MC3743 and MC3744)," March 7, 2006 (ML060680583).

E2-27


Applicability for EPU Operation

The ATRIUM-10 hydraulic models have been verified over a range of conditions that bound both pre-EPU and EPU operating conditions. The applicability of the models is described and supported by data presented in the thermal hydraulics section of the response to RAI SRXB-A.15 (Reference SRXB-1 07.2).

References:

SRXB-107.1 Correspondence, R.A. Copeland (Siemens) to R.C. Jones (NRC), "ATRIUM-10 Presentations," RAC:95:080, May 4, 1995 (38-9091703-000).

SRXB-107.2 Correspondence, w.o. Crouch (TVA) to U.S. Nuclear Regulatory Commission, "Browns Ferry Nuclear Plant (BFN) - Units 2 and 3, Response to NRC Round 3 Requests for Additional Information Related to Technical Specifications (TS) Change No. TS-418 - Requests for Extended Power Uprate Operation (TAC Nos. MC3743 and MC3744)," March 7, 2006 (ML060680583).

E2-27


Figure SRXB-107.1 ATRIUM-10 Bundle Pressure DropI I

E2-28


r

.J Figure SRXB-107.1 ATRIUM-10 Bundle Pressure Drop

[ ]

E2-28


Figure SRXB-107.2 ATRIUM-10 Bundle Pressure Drop[]

E2-29


r

..J

Figure SRXB-107.2 ATRIUM-10 Bundle Pressure Drop [ . ]

E2-29


Figure SRXB-107.3 ATRIUM-10 Lower Region Spacer Pressure DropUsing PHTF Loss Coefficients

I I

E2-30


r

Figure SRXB-107.3 ATRIUM-10 Lower Region Spacer Pressure Drop Using PHTF Loss Coefficients

[ ]

E2-30

.J


Figure SRXB-107.4 ATRIUM-10 Transition Region Spacer Pressure DropUsing PHTF Loss Coefficients

I I

E2-31


r

..J Figure SRXB-107.4 ATRIUM-10 Transition Region Spacer Pressure Drop

Using PHTF Loss Coefficients [ ]

E2-31


Figure SRXB-107.5 ATRIUM-10 Upper Region Spacer Pressure DropUsing PHTF Loss Coefficient

I I

E2-32


r

Figure SRXB-107.5 ATRIUM-10 Upper Region Spacer Pressure Drop Using PHTF Loss Coefficient

[ ]

E2-32

..J


r-

Figure SRXB-107.6 Viewgraph From May 4,1995Presentation to NRC

Regarding ATRIUM-10 Fuel

r-

Figure SRXB-107.7 Viewgraph From May 4, 1995Presentation to NRC


rn-i

E2-33

r

r


Figure SRXB-107.6 Viewgraph From May 4,1995 Presentation to NRC


Figure SRXB-107.7 Viewgraph From May 4,1995 Presentation to NRC


E2-33

..J


NRC RAI SRXB-108

At EPU conditions there are a higher number of higher powered bundles. It is possible, andlikely, for large axial sections of these bundles to be in an annular flow regime. Calculatingpressure losses near bundle features such as fuel spacers can be important in the prediction ofcritical heat flux, which tends to occur below fuel spacers where the liquid film is typicallythinnest.

On page 25 of Exxon Nuclear Company's XN-NF-84-105(P)(A), XCOBRA-T: A Computer Codefor BWR Transient Thermal-Hydraulic Core Analysis, it is stated that "[t]his [Martinelli-Nelson]formulation was developed for horizontal flow, but is reasonably accurate for vertical flow whereboth phasic flow rates are high enough to ensure turbulent co-current flow." Justify why theMartinelli-Nelson two phase friction multipliers are applicable in annular flow regimes.


When applying the XCOBRA-T two phase pressure drop models implemented in the1-dimensional hydraulic model of the COTRANSA2 code, the local (spacer grid) pressurelosses are automatically adjusted to preserve the core pressure drop predicted by the moredetailed 3-dimensional hydraulic representation in MICROBURN-B2. The XCOBRA-T initialflow rate is defined by a hydraulic demand curve predicted by XCOBRA, which defines therelationship between assembly power and the initial flow rate and accounts for the lack of a corebypass model in XCOBRA-T.

The orifice loss coefficient is automatically adjusted in XCOBRA-T to preserve the COTRANSA2(and MICROBURN-B2) initial core pressure drop and the initial flow rate defined by thehydraulic demand curve. Therefore, adjustments made to the local (spacer grid) pressurelosses in COTRANSA2 appear in the adjustments to the orifice loss coefficient in XCOBRA-T.The hydraulic channel nodalization of each code is discussed in the previous response to RAISRXB-1 15 (ML082330187).

NRC RAI SRXB-109

Section 3.3 of the Technical Evaluation Report attached to the NRC's safety evaluationapproving XN-NF-84-105(P)(A) states that critical power calculations may be inaccurate if theinlet flow is negative or if the inlet quality is above zero. Verify that for the transient analysesthat the bundle inlet flow is positive and that the inlet qualities are less than zero.


The transient code, XCOBRA-T, evaluates Reynolds number for each node for each step of thecalculation. If the flow becomes negative at any node, the code stops the calculation.

NRC RAI SRXB-112

Some models may have been updated to conservatively bound experimental data collectedsubsequent to the NRC review and approval of RODEX2. The staff notes that certainassumptions may be conservative in the assessment of linear heat generation rate limits thatmay not be conservative when evaluating transient heat flux during AOO simulation due to the

E2-34


NRC RAI SRXB-108

At EPU conditions there are a higher number of higher powered bundles. It is possible, and likely, for large axial sections of these bundles to be in an annular flow regime. Calculating pressure losses near bundle features such as fuel spacers can be important in the prediction of critical heat flux, which tends to occur below fuel spacers where the liquid film is typically thinnest.

On page 25 of Exxon Nuclear Company's XN-NF-84-105(P)(A), XCOBRA-T: A Computer Code for BWR Transient Thermal-Hydraulic Core Analysis, it is stated that "[t]his [Martinelli-Nelson] formulation was developed for horizontal flow, but is reasonably accurate for vertical flow where both phasic flow rates are high enough to ensure turbulent co-current flow." Justify why the Martinelli-Nelson two phase friction multipliers are applicable in annular flow regimes.


When applying the XCOBRA-T two phase pressure drop models implemented in the 1-dimensional hydraulic model of the COTRANSA2 code, the local (spacer grid) pressure losses are automatically adjusted to preserve the core pressure drop predicted by the more detailed 3-dimensional hydraulic representation in MICROBURN-B2. The XCOBRA-T initial flow rate is defined by a hydraulic demand curve predicted by XCOBRA, which defines the relationship between assembly power and the initial flow rate and accounts for the lack of a core bypass model in XCOBRA-T.

The orifice loss coefficient is automatically adjusted in XCOBRA-T to preserve the COTRANSA2 (and MICROBURN-B2) initial core pressure drop and the initial flow rate defined by the hydraulic demand curve. Therefore, adjustments made to the local (spacer grid) pressure losses in COTRANSA2 appear in the adjustments to the orifice loss coefficient in XCOBRA-T. The hydraulic channel nodalization of each code is discussed in the previous response to RAI SRXB-115 (ML082330187).

NRC RAI SRXB-109

Section 3.3 of the Technical Evaluation Report attached to the NRC's safety evaluation approving XN-NF-84-1 05(P)(A) states that critical power calculations may be inaccurate if the inlet flow is negative or if the inlet quality is above zero. Verify that for the transient analyses that the bundle inlet flow is positive and that the inlet qualities are less than zero.


The transient code, XCOBRA-T, evaluates Reynolds number for each node for each step of the calculation. If the flow becomes negative at any node, the code stops the calculation.

NRC RAI SRXB-112

Some models may have been updated to conservatively bound experimental data collected subsequent to the NRC review and approval of RODEX2. The staff notes that certain assumptions may be conservative in the assessment of linear heat generation rate limits that may not be conservative when evaluating transient heat flux during AOO simulation due to the

E2-34


competing effects of reactivity feedback and heat flux flow mismatch. If a model is"conservatively bounding" in RODEX2, and translated to XCOBRA-T, provide a discussion ofthe performance of the model for thermal margin transient calculations.


The draft response for SRXB-1 12 deals with changes to the RODEX2 code in its first part, butrequests additional information regarding the use of conservative assumptions in the abnormaloperating occurrence (AOO) transient response. The discussion regarding the conservatism ofthe gap properties should be addressed in the response to the second part of RAI 112. See thesecond and third sentences:

The staff notes that certain assumptions may be conservative in the assessment of linearheat generation rate limits that may not be conservative when evaluating transient heat fluxduring AOO simulation due to the competing effects of reactivity feedback and heat flux/flowmismatch. If a model is "conservatively bounding" in RODEX2, and translated to XCOBRA-T, provide a discussion of the performance of the model for thermal margin transientcalculations.

Summary of staff concern:

The NRC staff considered the coupling of the neutron flux and fluid conditions for AOO transientevaluations for both a reduced thermal time constant and an increased thermal time constant.When the time constant is over predicted, the fluid response to changing neutron power islagged. A pressurization transient, therefore, would result in an increase in the reactor powerthat is not impeded by subsequent rapid void formation due to hold up of the heat flux in thepellet. An over prediction of the time constant will tend to increase the fission power for such atransient. However, the same effect of holding the heat up in the fuel pellet has the dual effectof reducing the cladding heat flux response; therefore, the ultimate effect on the transient criticalpower ratio (CPR) is a combination of the conservative prediction of peak neutron flux with thenon-conservative prediction of the transient cladding heat flux.

For the case where the time constant is under predicted the inverse is true, the gross reactorpower increase due to pressurization is limited due to more rapid void formation in response tothe increasing neutron flux, but this is countered by a prediction of higher cladding surface heatflux relative to the pin power throughout the transient.

The input assumptions regarding the gas gap may increase or decrease the thermal resistance,and similarly, an increase or decrease in the thermal resistance does not have a clear impact onthe transient predicted CPR due to competing effects in the cladding heat flux and voidreactivity.


A gap conductance sensitivity study was performed for the 100% power/1 05% flow BFN loadrejection with no bypass (LRNB) transient event from Reference SRXB-1 12.1. The purpose ofthe sensitivity study was to show the ACPR trend for changes in gap conductance forCOTRANSA2 versus XCOBRA-T. The gap conductance change considered was [ IThe results are provided in Table SRXB-1 12.1. As seen from the results, an increase inCOTRANSA2 core average gap conductance results in a decrease in ACPR; whereas anincrease in XCOBRA-T gap hot channel conductance results in an increase in ACPR. A

E2-35


competing effects of reactivity feedback and heat flux flow mismatch. If a model is "conservatively bounding" in RODEX2, and translated to XCOBRA-T, provide a discussion of the performance of the model for thermal margin transient calculations.

Clarifications provided bv the NRC following a meeting on August 7. 2008

The draft response for SRXB-112 deals with changes to the RODEX2 code in its first part, but requests additional information regarding the use of conservative assumptions in the abnormal operating occurrence (AOO) transient response. The discussion regarding the conservatism of the gap properties should be addressed in the response to the second part of RAI 112. See the second and third sentences:

The staff notes that certain assumptions may be conservative in the assessment of linear heat generation rate limits that may not be conservative when evaluating transient heat flux during AOO simulation due to the competing effects of reactivity feedback and heat flux/flow mismatch. If a model is "conservatively bounding" in RODEX2, and translated to XCOBRAT, provide a discussion of the performance of the model for thermal margin transient calculations.

Summary of staff concern:

The NRC staff considered the coupling of the neutron flux and fluid conditions for AOO transient evaluations for both a reduced thermal time constant and an increased thermal time constant. When the time constant is over predicted, the fluid response to changing neutron power is lagged. A pressurization transient, therefore, would result in an increase in the reactor power that is not impeded by subsequent rapid void formation due to hold up of the heat flux in the pellet. An over prediction of the time constant will tend to increase the fission power for such a transient. However, the same effect of holding the heat up in the fuel pellet has the dual effect of reducing the cladding heat flux response; therefore, the ultimate effect on the transient critical power ratio (CPR) is a combination of the conservative prediction of peak neutron flux with the non-conservative prediction of the transient cladding heat flux.

For the case where the time constant is under predicted the inverse is true, the gross reactor power increase due to pressurization is limited due to more rapid void formation in response to the increasing neutron flux, but this is countered by a prediction of higher cladding surface heat flux relative to the pin power throughout the transient.

The input assumptions regarding the gas gap may increase or decrease the thermal resistance, and similarly, an increase or decrease in the thermal resistance does not have a clear impact on the transient predicted CPR due to competing effects in the cladding heat flux and void reactivity.


A gap conductance sensitivity study was performed for the 100% power/105% flow BFN load rejection with no bypass (LRNB) transient event from Reference SRXB-112.1. The purpose of the sensitivity study was to show the ~CPR trend for changes in gap conductance for COTRANSA2 versus XCOBRA-T. The gap conductance change considered was [ ]. The results are provided in Table SRXB-112.1. As seen from the results, an increase in COTRANSA2 core average gap conductance results in a decrease in ~CPR; whereas an increase in XCOBRA-T gap hot channel conductance results in an increase in ~CPR. A

E2-35


decrease in gap conductance shows the opposite trend. The XCOBRA-T ATRIUM-10 hotchannel model is slightly more sensitive to the change in gap conductance than theCOTRANSA2 ATRIUM-1 0 average core model. When both COTRANSA2 and XCOBRA-T gapconductance are changed by an equivalent amount, the net impact is no significant change inACPR.

Reference:

SRXB-1 12.1 EMF-2982(P) Revision 0, Browns Ferry Units 2 and 3 Safety Analysis Report forExtended Power Uprate A TRIUMTM-1O Fuel Supplement, Framatome ANP,June 2004.

Table SRXB-112.1 Gap Conductance Study

Increase in Gap Conductance

Gap conductance condition A(ACPR)

Core average[ ] -0.011

Hot channel [ ] +0.012

Core average and hot channel [ ] 0.000

Decrease in Gap Conductance

Gap conductance condition A(ACPR)

Core average[ ] +0.015

Hot channel [ ] -0.016

Core average and hot channel [ ] -0.001

NRC RAI SRXB-116

Address whether XCOBRA-T was used to demonstrate acceptable fuel rod thermal mechanicalperformance during transients. If XCOBRA-T is not used for this purpose, address howacceptable thermal mechanical performance is demonstrated during transients. If the method isnot consistent with the models in RODEX2 or later NRC-approved thermal mechanical code,justify the approach.


Aside from describing the method for normalization of the transient LHGR to the initial LHGR,provide some additional minor clarifications:

(1) The decay heat contribution will remain essentially static during the transient, addresswhether the normalization capture the varying rod decay heat sources;

(2) Specify the source of the decay heat constants (i.e. ANS standard);

E2-36


decrease in gap conductance shows the opposite trend. The XCOBRA-T ATRIUM-10 hot channel model is slightly more sensitive to the change in gap conductance than the COTRANSA2 ATRIUM-10 average core model. When both COTRANSA2 and XCOBRA-T gap conductance are changed by an equivalent amount, the net impact is no significant change in flCPR.

Reference:

SRXB-112.1 EMF-2982(P) Revision 0, Browns Ferry Units 2 and 3 Safety Analysis Report for Extended Power Uprate ATRIUMTM-10 Fuel Supplement, Framatome ANP, June 2004.

Table SRXB-112.1 Gap Conductance Study

Increase in Gap Conductance

Gap conductance condition fl(flCPR)

Core average [ ] -0.011

Hot channel [ ] +0.012

Core average and hot channel [ ] 0.000

Decrease in Gap Conductance

Gap conductance condition fl(flCPR)

Core average [ ] +0.015

Hot channel [ ] -0.016

Core average and hot channel [ ] -0.001

NRC RAI SRXB-116

Address whether XCOBRA-T was used to demonstrate acceptable fuel rod thermal mechanical performance during transients. If XCOBRA-T is not used for this purpose, address how acceptable thermal mechanical performance is demonstrated during transients. If the method is not consistent with the models in RODEX2 or later NRC-approved thermal mechanical code, justify the approach.

Clarifications provided bv the NRC following a meeting on August 7, 2008

Aside from describing the method for normalization of the transient LHGR to the initial LHGR, provide some additional minor clarifications:

(1) The decay heat contribution will remain essentially static during the transient, address whether the normalization capture the varying rod decay heat sources;

(2) Specify the source of the decay heat constants (i.e. ANS standard);

E2-36


(3) The rod power distribution is flattened due to gamma smearing of the thermal power,address how these gamma smeared power fractions are calculated; and

(4) Address how the direct moderator heat is accounted for.

The response should also provide a detailed description of the rod heat flux calculation forbundles with part length fuel rods, and address the code change as well as items 1-4 for eachregion (fully rodded, plena region, above plena region).


For bundles with part-length fuel rods (PLFRs), the rod heat flux calculation begins bycomputing the time-dependent heat flux generation rate at each axial section in the fuel rod.The updated equation corresponding to equation 2.130 of Reference SRXB-1 16.1 is:

q"(t) = P(t) I (ff + fc )FriFli Fa7FDrodj LNaNn

where

P(t) = transient reactor powerff = fraction of power produced in the fuelfý = fraction of power produced in the claddingNa = total number of assemblies in the coreN,. = total number of heated rods for type i assembly at the axial planeF; = radial peaking factor of type i assemblyF,, = local peaking factor of type i assemblyFa = axial peaking factor at the axial planeDrodj = fuel rod diameter of type i assemblyL = axial heated length

This equation differs from that in Reference SRXB-1 16.1 by replacing the initial reactor power inthe denominator with Tr. In addition, the variable definitions have been modified to identify thatthe total number of heated rods is dependent on both the assembly type and axial elevation,and the definition of L has been corrected to the axial heated length of the assembly. Thisequation is substituted into equations 2.129a and 2.129b in Section 2.5.5 of ReferenceSRXB-1 16.1 to define the volumetric heat deposition rate for the fuel pellet and cladding,respectively. This volumetric heat deposition rate is used in the right hand side of equation 2.85of Reference SRXB-1 16.1 to iteratively solve the transient heat conduction equation and thehydraulic conservation equations for the new time step temperatures and surface heat flux. Theheat flux is introduced into the channel energy equation (2.2 of Reference SRXB-1 16.1) throughthe term q'. This linear heat deposition rate is a summation of the energy added by directenergy deposition and surface heat flux:

J= P(t) fcoolFriFa+Hsurf"(TNodesT-Tfluid)'%'Drod,i"Nri Niq'(t) 1 NaL I

E2-37


(3) The rod power distribution is flattened due to gamma smearing of the thermal power, address how these gamma smeared power fractions are calculated; and

(4) Address how the direct moderator heat is accounted for.

The response should also provide a detailed description of the rod heat flux calculation for bundles with part length fuel rods, and address the code change as well as items 1-4 for each region (fully rodded, plena region, above plena region).


For bundles with part-length fuel rods (PLFRs), the rod heat flux calculation begins by computing the time-dependent heat flux generation rate at each axial section in the fuel rod. The updated equation corresponding to equation 2.130 of Reference SRXB-116.1 is:

where

pet) f, fc

Na Nri Fri F/i Fa Orad';

L

= = = = = = = = = =

transient reactor power fraction of power produced in the fuel fraction of power produced in the cladding total number of assemblies in the core total number of heated rods for type i assembly at the axial plane radial peaking factor of type i assembly local peaking factor of type i assembly axial peaking factor at the axial plane fuel rod diameter of type i assembly axial heated length

This equation differs from that in Reference SRXB-116.1 by replacing the initial reactor power in the denominator with TT. In addition, the variable definitions have been modified to identify that the total number of heated rods is dependent on both the assembly type and axial elevation, and the definition of L has been corrected to the axial heated length of the assembly. This equation is substituted into equations 2.129a and 2.129b in Section 2.5.5 of Reference SRXB-116.1 to define the volumetric heat deposition rate for the fuel pellet and cladding, respectively. This volumetric heat deposition rate is used in the right hand side of equation 2.85 of Reference SRXB-116.1 to iteratively solve the transient heat conduction equation and the hydraulic conservation equations for the new time step temperatures and surface heat flux. The heat flux is introduced into the channel energy equation (2.2 of Reference SRXB-116.1 ) through the term q'. This linear heat deposition rate is a summation of the energy added by direct energy deposition and surface heat flux:

, {P(t) } q (t) = -- fcool FriFa + Hsurf . (TNodesT - Tf/uid)·ff· Drad i . Nri Ni NL ' a

E2-37


where

f. 0 1o = fraction of power produced in the coolantHSurd = film heat transfer coefficient at the axial planeTNodesT = cladding surface temperature at the axial planeTr7,ud = fluid temperature at the axial planeN, = number of fuel assemblies in channel i

In addition to axially varying number of heated rods, proper modeling of PLFRs also requiresaxial variations in the active flow area, the heated perimeter, and the wetted perimeter, andthese parameters are now defined as axially dependent quantities in AREVA methods.Consequently, all references to these parameters or parameters derived from the basicgeometry data in the approved topical reports should be interpreted as being axially dependentvariables. The pressure drop due to the area expansion at the end of the PLFRs (or anywherein the active flow path) is modeled using the specific volume for momentum as expressed inequations 2.78 and 2.79 of Reference SRXB-1 16.1. For current designs, area contractionsoccur in the single phase region, but the coding was generalized to address area contractions inthe two-phase region based on a solution of the two phase Bernoulli equation.

An XCOBRA-T deposited power fraction sensitivity study was performed for the 100%power/1 05% flow BFN LRNB transient event from Reference SRXB-1 16.2. The purpose of thesensitivity study was to show the impact on ACPR from using generic ATRIUM-1 0 powerfractions versus case-specific power fractions. The case-specific power fractions are used inCOTRANSA2 and are obtained from CASMO-4/MICROBURN-B2. AREVA is in the process ofautomating the transfer of the case-specific power fractions into XCOBRA-T such that thegeneric values will no longer be used. [

] The power that would have been deposited []. A review of an ATRIUM-10 power deposition study showed that the [

]. A study wasperformed by taking [ ]. The resultsare provided in Table SRXB-1 16.1. The study shows no significant change in ACPR. [

] This studydemonstrates that the ATRIUM-10 generic power fractions in XCOBRA-T are adequate.

References:

SRXB-116.1 XN-NF-84-105(P)(A) Volume 1 and Volume 1 Supplements 1 and 2, XCOBRA-T:A Computer Code for BWR Transient Thermal-Hydraulic Core Analysis, ExxonNuclear Company, February 1987.

SRXB-1 16.2 EMF-2982(P) Revision 0, Browns Ferry Units 2 and 3 Safety Analysis Report forExtended Power Uprate A TRIUM TM-1O Fuel Supplement, Framatome ANP,June 2004. (ML041840301)

E2-38


where

fcool = fraction of power produced in the coolant Hsurf = film heat transfer coefficient at the axial plane TNodesT = cladding surface temperature at the axial plane Tf/uid = fluid temperature at the axial plane Ni = number of fuel assemblies in channel i

In addition to axially varying number of heated rods, proper modeling of PLFRs also requires axial variations in the active flow area, the heated perimeter, and the wetted perimeter, and these parameters are now defined as axially dependent quantities in AREVA methods. Consequently, all references to these parameters or parameters derived from the basic geometry data in the approved topical reports should be interpreted as being axially dependent variables. The pressure drop due to the area expansion at the end of the PLFRs (or anywhere in the active flow path) is modeled using the specific volume for momentum as expressed in equations 2.78 and 2.79 of Reference SRXB-116.1. For current designs, area contractions occur in the single phase region, but the coding was generalized to address area contractions in the two-phase region based on a solution of the two phase Bernoulli equation.

An XCOBRA-T deposited power fraction sensitivity study was performed for the 100% power/105% flow BFN LRNB transient event from Reference SRXB-116.2. The purpose of the sensitivity study was to show the impact on L\CPR from using generic ATRIUM-10 power fractions versus case-specific power fractions. The case-specific power fractions are used in COTRANSA2 and are obtained from CASM0-4/MICROBURN-B2. AREVA is in the process of automating the transfer of the case-specific power fractions into XCOBRA-T such that the generic values will no longer be used. [

] The power that would have been deposited [ ]. A review of an ATRIUM-10 power deposition study showed that the [

]. A study was performed by taking [ ]. The results are provided in Table SRXB-116.1. The study shows no significant change in L\CPR. [

] This study demonstrates that the ATRIUM-10 generic power fractions in XCOBRA-T are adequate.

References:

SRXB-116.1 XN-NF-84-105(P)(A) Volume 1 and Volume 1 Supplements 1 and 2, XCOBRA-T: A Computer Code for BWR Transient Thermal-Hydraulic Core Analysis, Exxon Nuclear Company, February 1987.

SRXB-116.2 EMF-2982(P) Revision 0, Browns Ferry Units 2 and 3 Safety Analysis Report for Extended Power Uprate ATRIUMTM-10 Fuel Supplement, Framatome ANP, June 2004. (ML041840301)

E2-38


Table SRXB-116.1 Deposited Heat Study

Fuel Cladding Moderator BypassCondition Heat Heat Heat Heat A(ACPR)

Generic power fractions [ ] [ ] [ ] [ ] NA

Case-specific power fractions [ ] [ ] [ ] [ ] -0.0004

Case-specific power fractionsI ] ] [ ] [ ] [ ] +0.0008

E2-39


Table SRXB-116.1 Deposited Heat Study

Fuel Cladding Moderator Bypass Condition Heat Heat Heat Heat fl(flCPR)

Generic power fractions [ ] [ ] [ ] [ ] NA

Case-specific power fractions [ ] [ ] [ ] [ ] -0.0004

Case-specific power fractions [ ] [ ] [ ] [ ] [ ] +0.0008

E2-39


NRC RAI SRXB-117

Enclosure 4 of the letter dated June 25, 2004, references NEDO-32047-A. In particular it isnoted that operation at EPU conditions is generally achieved by flattening radial core power. Asa result of this flattening the second harmonic eigenvalue separation is likely to be greatlyreduced. Therefore, under non-isolation ATWS conditions it is expected that the core will bemore susceptible to regional mode oscillations that at pre-EPU conditions.

Given the information provided in the NRC's contractors' technical evaluation report attached tothe safety evaluation approving NEDO-32047-A dated February 5, 1994, Appendix C:"Consequences of Out-of-Phase Instability Mode Not Proven More Favorable than In-PhaseMode." Provide an evaluation of the likelihood of a regional mode oscillation to develop undernon-isolation ATWS conditions. It is acceptable to evaluate the regional and core wide modedecay ratios for these conditions for an equilibrium ATRIUM-10 Unit 2 core using STAIF torespond to this request for additional information (RAI). Based on the available analyses,determine if such an oscillation at BFN would result in a significant increase in the fuel damagerelative to the results in NEDO-32047-A.

The analyses in NEDO-32047-A were performed for General Electric (GE) fuel. The analysesare generally applicable for pre-EPU core designs since hydraulic stability of the fuel productshas improved or at least remained the same. Provide a comparison of the channel stabilitycharacteristics of ATRIUM-10 to GE 8x8 fuel. If ATRIUM-10 is less stable than GE 8x8 fuel,consider any impact on the projected consequences of a non-isolation ATWS instability event.

Response to SRXB-1 17

The pre-EPU stability analysis for BFN indicates that the global mode is dominant over theregional (out-of-phase) mode where relatively large subcritical reactivity values are calculatedwith STAIF. For EPU cores with flatter radial power distributions, the calculated subcriticalreactivity values are noticeably lower in comparison. The resulting regional decay ratioscalculated for the EPU core are larger than the corresponding global mode decay ratio in aminority of cases, which warrants the examination of the effect of regional mode oscillationsdominating postulated ATWS instability events.

The task of evaluating the impact of large regional versus global mode oscillations is firstaddressed below from an analytical point of view and calculations are presented using areduced order model. The calculations will also address the effects of the parameters ofinterest, namely the subcritical reactivity due to core radial power flattening for EPU, increase invoidreactivity coefficient due to increasing the fresh fuel batch size, and fuel geometry effects(part-length rods and reduced pin conduction time constant for an ATRIUM-10 compared withan 8x8 fuel bundle). These effects will be demonstrated to result in equivalent consequences ofa postulated ATWS event relative to the results in NEDO-32047-A (Reference SRXB-1 17.1).Furthermore, the mitigation of the ATWS instability by reducing the core inlet subcooling as aconsequence of water level reduction by operator action (Reference SRXB-1 17.2) will bedemonstrated to be as effective in suppressing regional mode oscillations as for global modeoscillations.

Analytical Considerations

Unstable global mode oscillations grow exponentially at a fixed rate (decay ratio) from a smallperturbation. As the oscillation magnitude increases, nonlinear effects become important. The

E2-40


NRC RAI SRXB-117

Enclosure 4 of the letter dated June 25,2004, references NEDO-32047-A. In particular it is noted that operation at EPU conditions is generally achieved by flattening radial core power. As a result of this flattening the second harmonic eigenvalue separation is likely to be greatly reduced. Therefore, under non-isolation ATWS conditions it is expected that the core will be more susceptible to regional mode oscillations that at pre-EPU conditions.

Given the information provided in the NRC's contractors' technical evaluation report attached to the safety evaluation approving NEDO-32047-A dated February 5, 1994, Appendix C: "Consequences of Out-of-Phase Instability Mode Not Proven More Favorable than In-Phase Mode." Provide an evaluation of the likelihood of a regional mode oscillation to develop under non-isolation ATWS conditions. It is acceptable to evaluate the regional and core wide mode decay ratios for these conditions for an equilibrium ATRIUM-10 Unit 2 core using STAIF to respond to this request for additional information (RAI). Based on the available analyses, determine if such an oscillation at BFN would result in a significant increase in the fuel damage relative to the results in NEDO-32047-A.

The analyses in NEDO-32047-A were performed for General Electric (GE) fuel. The analyses are generally applicable for pre-EPU core designs since hydraulic stability of the fuel products has improved or at least remained the same. Provide a comparison of the channel stability characteristics of ATRIUM-10 to GE 8x8 fuel. If ATRIUM-10 is less stable than GE 8x8 fuel, consider any impact on the projected consequences of a non-isolation ATWS instability event.


The pre-EPU stability analysis for BFN indicates that the global mode is dominant over the regional (out-of-phase) mode where relatively large subcritical reactivity values are calculated with STAIF. For EPU cores with flatter radial power distributions, the calculated subcritical reactivity values are noticeably lower in comparison. The resulting regional decay ratios calculated for the EPU core are larger than the corresponding global mode decay ratio in a minority of cases, which warrants the examination of the effect of regional mode oscillations dominating postulated ATWS instability events.

The task of evaluating the impact of large regional versus global mode oscillations is first addressed below from an analytical point of view and calculations are presented using a reduced order model. The calculations will also address the effects of the parameters of interest, namely the subcritical reactivity due to core radial power flattening for EPU, increase in void reactivity coefficient due to increasing the fresh fuel batch size, and fuel geometry effects (part-length rods and reduced pin conduction time constant for an ATRIUM-10 compared with an 8x8 fuel bundle). These effects will be demonstrated to result in equivalent consequences of a postulated ATWS event relative to the results in NEDO-32047-A (Reference SRXB-117.1). Furthermore, the mitigation of the ATWS instability by reducing the core inlet subcooling as a consequence of water level reduction by operator action (Reference SRXB-117 .2) will be demonstrated to be as effective in suppressing regional mode oscillations as for global mode oscillations.

Analytical Considerations

Unstable global mode oscillations grow exponentially at a fixed rate (decay ratio) from a small perturbation. As the oscillation magnitude increases, nonlinear effects become important. The

E2-40


average power level drifts to higher values as a consequence of the nonlinearity of the neutronkinetics, which results in a negative reactivity feedback due to the increase of void fraction. Thenegative reactivity superimposed on the oscillating reactivity results in damping the neutronkinetics (References SRXB-117.3. SRXB-117.4, and SRXB-117.5). [

The regional mode oscillations are well understood in the linear limit where the power oscillationis attributed to the excitation of the first azimuthal harmonic mode of the neutron flux.Compared with the fundamental flux mode excitation associated with the global oscillation, thesubcritical reactivity of the first azimuthal eigenfunction contributes a damping effect on theneutron kinetics feedback. The hydraulic response is less damped compared to the globalmode case due to bypassing the damping effects of the recirculation loop. The regional modeoscillations may become the preferred oscillation mode for large-orificed cores (hydraulicdestabilization) and for small radial buckling (large core diameter and radial power distributionthat is relatively flat or ring-of-fire with relatively low power in the center).

E2-41

[


average power level drifts to higher values as a consequence of the nonlinearity of the neutron kinetics, which results in a negative reactivity feedback due to the increase of void fraction. The negative reactivity superimposed on the oscillating reactivity results in damping the neutron kinetics (References SRXB-117.3. SRXB-117.4, and SRXB-117.5). [

The regional mode oscillations are well understood in the linear limit where the power oscillation is attributed to the excitation of the first azimuthal harmonic mode of the neutron flux. Compared with the fundamental flux mode excitation associated with the global oscillation, the subcritical reactivity of the first azimuthal eigenfunction contributes a damping effect on the neutron kinetics feedback. The hydraulic response is less damped compared to the global mode case due to bypassing the damping effects of the recirculation loop. The regional mode oscillations may become the preferred oscillation mode for large-orificed cores (hydraulic destabilization) and for small radial buckling (large core diameter and radial power distribution that is relatively flat or ring-of-fire with relatively low power in the center).

E2-41


I

Description of the Reduced Order Model

The phenomenological description of large power oscillations in the global and regional modesis supported by the results of a reduced order model, which is used here to simulate large globaland regional mode oscillations. [

E2-42


]

Description of the Reduced Order Model

The phenomenological description of large power oscillations in the global and regional modes is supported by the results of a reduced order model, which is used here to simulate large global and regional mode oscillations. [

E2-42


The reduced order model allows fast and robust simulation of both the global and regionalmodes and helps to resolve issues that were not apparent at the time NEDO-32047-A(Reference SRXB-117.1) was issued. Most importantly, it helps to explore and provide insightinto the differences between the global and regional mode oscillations and their commonultimate limiting mechanism.

Results

The results of several cases performed with the reduced order model are presented. All ofthese calculations represent unstable oscillations growing to large magnitudes with parametervariations to address the issues of global versus regional and the effect of EPU core loadingwith fuel design differing from the fuel type used in NEDO-32047-A (Reference SRXB-1 17.1).These cases are:

E2-43


]

The reduced order model allows fast and robust simulation of both the global and regional modes and helps to resolve issues that were not apparent at the time NEDO-32047-A (Reference SRXB-117.1) was issued. Most importantly, it helps to explore and provide insight into the differences between the global and regional mode oscillations and their common ultimate limiting mechanism.

Results

The results of several cases performed with the reduced order model are presented. All of these calculations represent unstable oscillations growing to large magnitudes with parameter variations to address the issues of global versus regional and the effect of EPU core loading with fuel design differing from the fuel type used in NEDO-32047-A (Reference SRXB-117.1). These cases are:

E2-43


I

Conclusions

* Large regional mode oscillations have [mode.

" ATRIUM-10 bundle design differences from an older 8x8 [I

] effects compared with global

EPU effects (lower subcritical reactivity and higher void reactivity coefficient) []

* [

IReferences:

SRXB-1 17.1

SRXB-1 17.2

SRXB-1 17.3

SRXB-1 17.4

SRXB-1 17.5

NEDO-32047-A, "ATWS Rule Issues Relative to BWR Core Thermal-HydraulicStability," June 1995.

NEDO-32164 Revision 0, "Mitigation of BWR Core Thermal-Hydraulic Instabilitiesin ATWS," December 1992.

Wulff, W., H. S. Cheng, A.N. Mallen, and U.S. Rohatgi, "BWR Stability Analysiswith the BNL Engineering Plant Analyzer," NUREG/CR-5816, October 1992.

March-Leuba, J., D.G. Cacuci, and R.B. Perez, "Nonlinear Dynamics andStability of Boiling Water Reactors: Part I -- Qualitative Analysis," NuclearScience and Engineering: 93, 111-123 (1986).

March-Leuba, J., "Density-Wave Instabilities in Boiling Water Reactors,"NUREG/CR-6003, September 1992.

E2-44


]

Conclusions

• Large regional mode oscillations have [ mode.

• ATRIUM-10 bundle design differences from an older 8x8 [ ]

] effects compared with global

• EPU effects (lower subcritical reactivity and higher void reactivity coefficient) [ ]

•

References:

SRXB-117.1 NEDO-3204 7 -A, "A TWS Rule Issues Relative to BWR Core Thermal-Hydraulic Stability," June 1995.

SRXB-117.2 NEDO-32164 Revision 0, "Mitigation of BWR Core Thermal-Hydraulic Instabilities in ATWS," December 1992.

SRXB-117.3 Wulff, W., H. S. Cheng, A.N. Mallen, and U.S. Rohatgi, "BWR Stability Analysis with the BNL Engineering Plant Analyzer," NUREG/CR-5816, October 1992.

SRXB-117.4 March-Leuba, J., D.G. Cacuci, and R.B. Perez, "Nonlinear Dynamics and Stability of Boiling Water Reactors: Part I -- Qualitative Analysis," Nuclear Science and Engineering: 93,111-123 (1986).

SRXB-117.5 March-Leuba, J., "Density-Wave Instabilities in Boiling Water Reactors," NUREG/CR-6003, September 1992.

E2-44


SRXB-1 17.6

SRXB-1 17.7

Farawila, Y.M., and D.W. Pruitt, "A Study of Nonlinear Oscillation and LimitCycles in Boiling Water Reactors - I: The Global Mode," Nuclear Science andEngineering: 154, 302-315 (2006).

Farawila, Y.M., and D.W. Pruitt, "A Study of Nonlinear Oscillation and LimitCycles in Boiling Water Reactors - I1: The Regional Mode," Nuclear Science andEngineering: 154, 316-327 (2006).

E2-45


SRXB-117.6 Farawila, Y.M., and D.W. Pruitt, "A Study of Nonlinear Oscillation and Limit Cycles in Boiling Water Reactors -I: The Global Mode," Nuclear Science and Engineering: 154, 302-315 (2006).

SRXB-117.7 Farawila, Y.M., and D.W. Pruitt, "A Study of Nonlinear Oscillation and Limit Cycles in Boiling Water Reactors -II: The Regional Mode," Nuclear Science and Engineering: 154,316-327 (2006).

E2-45


r-

r..Figure SRXB-1 17.1.1 Relative Power for Case 1

Base Global Oscillation

Figure SRXB-117.1.2 Relative Power for Case 2Base Regional Oscillation

E2-46

r

r


Figure SRXB-117 .1.1 Relative Power for Case 1 Base Global Oscillation

Figure SRXB-117 .1.2 Relative Power for Case 2 Base Regional Oscillation

E2-46

..J

..J


r-

Figure SRXB-117.1.3 Relative Power for Case 3Global Oscillation

r-

Figure SRXB-117.1.4 Relative Power for Case 4Regional Oscillation

E2-47

r

r


Figure SRXB-117.1.3 Relative Power for Case 3 Global Oscillation

Figure SRXB-117.1.4 Relative Power for Case 4 Regional Oscillation

E2-47

.J


r-

Figure SRXB-117.1.5 Relative Power for Case 5Regional Oscillation With Decreased Subcriticality

r-

a-Figure SRXB-117.1.6 Relative Power for Case 6

Mitigated Global Oscillation

E2-48

r

r


Figure SRXB-117 .1.5 Relative Power for Case 5 Regional Oscillation With Decreased Subcriticality

Figure SRXB-117 .1.6 Relative Power for Case 6 Mitigated Global Oscillation

E2-48

.J

.J


r-

Figure SRXB-117.1.7 Relative Power for Case 7Mitigated Regional Oscillation

r"

Figure SRXB-1 17.1.8 Relative Power for Case 8Late-Mitigated Global Oscillation

E2-49

r

r


Figure SRXB-117 .1. 7 Relative Power for Case 7 Mitigated Regional Oscillation

Figure SRXB-117 .1.8 Relative Power for Case 8 Late-Mitigated Global Oscillation

E2-49

..J


r-

r-Figure SRXB-1 17.1.9 Relative Power for Case 9

Late-Mitigated Regional Oscillation

Figure SRXB-117.2.1 Inlet Mass Flow Rate for Case IBase Global Oscillation

E2-50

r

r


Figure SRXB-117 .1.9 Relative Power for Case 9 Late-Mitigated Regional Oscillation

Figure SRXB-117 .2.1 Inlet Mass Flow Rate for Case 1 Base Global Oscillation

E2-50

.J

..J


r-

-U

Figure SRXB-117.2.2 Inlet Mass Flow Rate for Case 2Base Regional Oscillation

r-

JFigure SRXB-117.2.3 Inlet Mass Flow Rate for Case 3

Global Oscillation

E2-51

r

r


Figure SRXB-117.2.2 Inlet Mass Flow Rate for Case 2 Base Regional Oscillation

Figure SRXB-117 .2.3 Inlet Mass Flow Rate for Case 3 Global Oscillation

E2-51

r- NON-PROPRIETARY INFORMATION

Figure SRXB-117.2.4 Inlet Mass Flow Rate for Case 4Regional Oscillation

r-

Figure SRXB-117.2.5 Inlet Mass Flow Rate for Case 5Regional Oscillation With Decreased Subcriticality

E2-52

r

r


Figure SRXB-117.2.4 Inlet Mass Flow Rate for Case 4 Regional Oscillation

Figure SRXB-117.2.5 Inlet Mass Flow Rate for Case 5 Regional Oscillation With Decreased Subcriticality

E2-52

..J

..J


r-

Figure SRXB-117.2.6 Inlet Mass Flow Rate for Case 6Mitigated Global Oscillation

r-

Figure SRXB-117.2.7 Inlet Mass Flow Rate for Case 7Mitigated Regional Oscillation

E2-53

r

r


Figure SRXB-117.2.6 Inlet Mass Flow Rate for Case 6 Mitigated Global Oscillation

Figure SRXB-117.2.7 Inlet Mass Flow Rate for Case 7 Mitigated Regional Oscillation

E2-53

..J


r..

rn-

rFigure SRXB-117.2.8 Inlet Mass Flow Rate for Case 8

Late-Mitigated Global Oscillation

Figure SRXB-117.2.9 Inlet Mass Flow Rate for Case 9Late-Mitigated Regional Oscillation

E2-54

r

r


Figure SRXB-117.2.8 Inlet Mass Flow Rate for Case 8 Late-Mitigated Global Oscillation

Figure SRXB-117.2.9 Inlet Mass Flow Rate for Case 9 Late-Mitigated Regional Oscillation

E2-54

..J


r.

r..Figure SRXB-117.3.1 Exit Void Fraction for Case 1

Base Global Oscillation

Figure SRXB-1 17.3.2 Exit Void Fraction for Case 2Base Regional Oscillation

E2-55

r

r


Figure SRXB-117 .3.1 Exit Void Fraction for Case 1 Base Global Oscillation

Figure SRXB-117.3.2 Exit Void Fraction for Case 2 Base Regional Oscillation

E2-55

.J

.J


r-

r-Figure SRXB-1 17.3.3 Exit Void Fraction for Case 3

Global Oscillation

Figure SRXB-117.3.4 Exit Void Fraction for Case 4Regional Oscillation

E2-56

r

r


Figure SRXB-117 .3.3 Exit Void Fraction for Case 3 Global Oscillation

Figure SRXB-117.3.4 Exit Void Fraction for Case 4 Regional Oscillation

E2-56

..J


r-

rn-Figure SRXB-117.3.5 Exit Void Fraction for Case 5Regional Oscillation With Decreased Subcriticality

r"

rn-

Figure SRXB-1 17.3.6 Exit Void Fraction for Case 6Mitigated Global Oscillation

E2-57

r

r


Figure SRXB-117.3.5 Exit Void Fraction for Case 5 Regional Oscillation With Decreased Subcriticality

Figure SRXB-117.3.6 Exit Void Fraction for Case 6 Mitigated Global Oscillation

E2-57

.J


r.

r..Figure SRXB-1 17.3.7 Exit Void Fraction for Case 7

Mitigated Regional Oscillation

Figure SRXB-1 17.3.8 Exit Void Fraction for Case 8Late-Mitigated Regional Oscillation

-d

E2-58

r

r


Figure SRXB-117.3.7 Exit Void Fraction for Case 7 Mitigated Regional Oscillation

Figure SRXB-117.3.8 Exit Void Fraction for Case 8 Late-Mitigated Regional Oscillation

E2-58

..J

NON-PROPRIETARY IN FORMATION

r-

r-Figure SRXB-117.3.9 Exit Void Fraction for Case 9

Late-Mitigated Regional Oscillation

Figure SRXB-117.4.1 Void Fraction in Selected Nodes for Case 1Base Global Oscillation

E2-59

r

r


Figure SRXB-117.3.9 Exit Void Fraction for Case 9 Late-Mitigated Regional Oscillation

Figure SRXB-117.4.1 Void Fraction in Selected Nodes for Case 1 Base Global Oscillation

E2-59


r-

Figure SRXB-117.4.2 Void Fraction in Selected Nodes for Case 2Base Regional Oscillation

r-

Figure SRXB-117.4.3 Void Fraction in Selected Nodes for Case 3Global Oscillation

E2-60

r

r


Figure SRXB-117.4.2 Void Fraction in Selected Nodes for Case 2 Base Regional Oscillation

Figure SRXB-117.4.3 Void Fraction in Selected Nodes for Case 3 Global Oscillation

E2-60

.J

.J

NON-PROPRIETARY INFORMATIONr-

Figure SRXB-117.4.4 Void Fraction in Selected Nodes for Case 4Regional Oscillation

r

Figure SRXB-117.4.5 Void Fraction in Selected Nodes for Case 5Regional Oscillation With Decreased Subcriticality

E2-61

r NON-PROPRIETARY INFORMATION

Figure SRXB-117.4.4 Void Fraction in Selected Nodes for Case 4 Regional Oscillation

r

Figure SRXB-117.4.5 Void Fraction in Selected Nodes for Case 5 Regional Oscillation With Decreased Subcriticality

E2-61

..J


NRC Introduction

The following are related to the June 3, 2008 response to SRXB-88.

NRC RAI SRXB-118

In the supplemental response to RAI SRXB-88, TVA provided the results of sensitivity analysesto evaluate the impact of void fraction uncertainty on the calculated delta-critical power ratio(DCPR) and the safety limit minimum critical power (SLMCPR). In the void fraction reductioncase, the DCPR is apparently unaffected and is accompanied by an increase in SLMCPR.

If the void fraction were reduced throughout the core by a fixed bias, the result would be toredistribute the reactor power according to the change in reactivity associated with the voidperturbation. Since those bundles with the higher bundle average void fractions will have agreater reactivity response, a reduction in the void fraction will tend to increase, slightly, thepower in those bundles with a higher bundle average void fraction relative to the bundles thathad a lower void content prior to the perturbation. The bundles with a higher bundle averagevoid fraction are the high powered bundles. Therefore, a fixed reduction in void fraction willincrease the radial power peaking factor. The increased radial power peaking factor for a givensteady state power level would result in fewer rods entering boiling transition as a result of atransient initiated from this state.

When this effect is considered, it is the equivalent of increasing the radial power peaking andreducing the SLMCPR since fewer rods are at the limiting end of the pin power statisticaldistribution. In effect, the span of pin powers to account for the 0.1 percent of highest poweredpins increases. Results of the TVA sensitivity analysis demonstrate the opposite trend. It isexpected that the imposition of a fixed void fraction reduction would result in a lower SLMCPR.Explain this discrepancy.


It should be noted that the sensitivity analyses presented in the SRXB-88 response were notbased on "fixed" void fraction changes. Rather, the analyses were based on modifications tothe void-quality correlation that resulted in a new nominal fit and offsets that were on average+0.05 void. The discussion above in the RAI question for SRXB-1 18 is based on a comparisonof trends for an instantaneous change in void fraction. The RAI SRXB-88 response included theimpact of the fuel depleted with the changes in the void-quality correlation. The difference indepletion changes the sensitivity of void friction modifications considerably due to the feedbackof modified power distributions on exposure distribution.

For the RAI SRXB-88 case, the change in the void-quality correlation was imposed over all fuelin the core from beginning-of-life. No changes were made to the fuel loading and rod patterns.The result of SRXB-88 was that a reduction in void resulted in more assemblies at higherpower. The radial peaking factors of the high-powered assemblies that contributed to rods inboiling transition were slightly more "flat" and resulted in a slightly higher SLMCPR.Figure SRXB-1 18.1 shows the slight differences in radial distributions.

The sensitivity analysis of SRXB-88 was repeated for an instantaneous change in voids. For aninstantaneous change in voids, the SLMCPR trends were the same as SRXB-88; however, thechange is small for both depleted and instantaneous void change, i.e., an SLMCPR change of-0.003 for +0.05 voids and +0.002 for -0.05 voids. The sensitivity can be explained by the

E2-62


NRC Introduction

The following are related to the June 3, 2008 response to SRXB-88.

NRC RAI SRXB-118

In the supplemental response to RAI SRXB-88, TVA provided the results of sensitivity analyses to evaluate the impact of void fraction uncertainty on the calculated delta-critical power ratio (DCPR) and the safety limit minimum critical power (SLMCPR). In the void fraction reduction case, the DCPR is apparently unaffected and is accompanied by an increase in SLMCPR.

If the void fraction were reduced throughout the core by a fixed bias, the result would be to redistribute the reactor power according to the change in reactivity associated with the void perturbation. Since those bundles with the higher bundle average void fractions will have a greater reactivity response, a reduction in the void fraction will tend to increase, slightly, the power in those bundles with a higher bundle average void fraction relative to the bundles that had a lower void content prior to the perturbation. The bundles with a higher bundle average void fraction are the high powered bundles. Therefore, a fixed reduction in void fraction will increase the radial power peaking factor. The increased radial power peaking factor for a given steady state power level would result in fewer rods entering boiling transition as a result of a transient initiated from this state.

When this effect is considered, it is the equivalent of increasing the radial power peaking and reducing the SLMCPR since fewer rods are at the limiting end of the pin power statistical distribution. In effect, the span of pin powers to account for the 0.1 percent of highest powered pins increases. Results of the TVA sensitivity analysis demonstrate the opposite trend. It is expected that the imposition of a fixed void fraction reduction would result in a lower SLMCPR. Explain this discrepancy.


It should be noted that the sensitivity analyses presented in the SRXB-88 response were not based on "fixed" void fraction changes. Rather, the analyses were based on modifications to the void-quality correlation that resulted in a new nominal fit and offsets that were on average ±0.05 void. The discussion above in the RAI question for SRXB-118 is based on a comparison of trends for an instantaneous change in void fraction. The RAI SRXB-88 response included the impact of the fuel depleted with the changes in the void-quality correlation. The difference in depletion changes the sensitivity of void friction modifications considerably due to the feedback of modified power distributions on exposure distribution.

For the RAI SRXB-88 case, the change in the void-quality correlation was imposed over all fuel in the core from beginning-of-life. No changes were made to the fuel loading and rod patterns. The result of SRXB-88 was that a reduction in void resulted in more assemblies at higher power. The radial peaking factors of the high-powered assemblies that contributed to rods in boiling transition were slightly more "flat" and resulted in a slightly higher SLMCPR. Figure SRXB-118.1 shows the slight differences in radial distributions.

The sensitivity analysis of SRXB-88 was repeated for an instantaneous change in voids. For an instantaneous change in voids, the SLMCPR trends were the same as SRXB-88; however, the change is small for both depleted and instantaneous void change, i.e., an SLMCPR change of -0.003 for +0.05 voids and +0.002 for -0.05 voids. The sensitivity can be explained by the

E2-62


small radial power distribution shifts in Figures SRXB-1 18.2 and SRXB-1 18.3. It is concludedthat the radial distribution is not significantly changed for the SLMCPR analysis; therefore, theimpact of the prescribed void-quality correlation changes is insignificant on SLMCPR.It is very difficult to identify the expected direction of the radial power distribution change due toa modification of the void-quality correlation. In addition to the void coefficient dependency onvoid fraction, there is an even stronger dependency of the void coefficient on exposure. For thelimiting case of SLMCPR the highest radial powers come from a range of assembly exposures.The importance of void changes in different assemblies of different exposures cannot beanalyzed with simplified models and isolated trends.

Independent of the trend, the analyses demonstrate insignificant impacts on SLMCPR.

E2-63


small radial power distribution shifts in Figures SRXB-118.2 and SRXB-118.3. It is concluded that the radial distribution is not significantly changed for the SLMCPR analysis; therefore, the impact of the prescribed void-quality correlation changes is insignificant on SLMCPR. It is very difficult to identify the expected direction of the radial power distribution change due to a modification of the void-quality correlation. In addition to the void coefficient dependency on void fraction, there is an even stronger dependency of the void coefficient on exposure. For the limiting case of SLMCPR the highest radial powers come from a range of assembly exposures. The importance of void changes in different assemblies of different exposures cannot be analyzed with simplified models and isolated trends.

Independent of the trend, the analyses demonstrate insignificant impacts on SLMCPR.

r

E2-63


0.99

° 0.98

0.970z

0.96

0.950 100

Bundle Index

0.99

a. 0.98

• 0.970z

0.96

0.95

Figure SRXB-118.1 SLMCPR Radial Power DistributionHigh-Powered Assemblies

Depleted Voids

'\. n\

- - - Reference

Modified...... Modified +0.05

- - - Modified -0.05

0 100

Bundle Index

Figure SRXB-118.2 SLMCPR Radial Power DistributionHigh-Powered Assemblies

Instantaneous Voids

E2-64

0.99

; ~ 0.98 OJ :;; ~

] OJ E 0.97 o z

0.96 E--Reference --Modified - - - - - . Modified +0.05

- - - - Modified -0.05


0.95 L-________ ______________________ _ ___ ~

0.99

; ~ 0.98 OJ :;; ~

" .~

o

~ 0.97 ~ ~ Z

0.96

\ \

Bundle Index

Figure SRXB-118.1 SLMCPR Radial Power Distribution High-Powered Assemblies

Depleted Voids

---~~- -------

t---Reference~ --Modified

- - - - - -Modified +0.05 - - - - Modified -0.05 1--------------------

100

0.95 I---------------- -----------------------l o

Bundle Index

Figure SRXB-118.2 SLMCPR Radial Power Distribution High-Powered Assemblies

Instantaneous Voids

E2-64

100


0.8

0.7 f

0.6

0

a.0.5

0.4

0z

0.3

0.2

0.1

---- ReferenceModified

------ Modified +0.05- --- Modified -0.05

400 500 600 700

Bundle Index

800

Figure SRXB-118.3 SLMCPR Radial Power DistributionLow-Powered Assemblies

Instantaneous Voids

E2-65


0.8 ,-----------------------------------------------------------------,

0.7

0.6

! II. iii 0.5 '6 (i .., ~ 04 .. . E o z

0.3

0.2 -1---Reference

--Modified

- - - - - -Modified +0.05

- - - - Modified -0.05

0.1 ~--------------~--------------~----------------~------------~ 400 500 600

Bundle Index

700

Figure SRXB-118.3 SLMCPR Radial Power Distribution Low-Powered Assemblies

Instantaneous Voids

E2-65

800


NRC RAI SRXB-119

Continuing with the void fraction reduction case, the decrease in void fraction wouldsimultaneously result in a redistribution of the axial power. Since those higher void nodes wouldhave a greater reactivity response than low void nodes, the axial power distribution would shiftupwards in the core. The upward shift in the axial power distribution has the effect of increasingthe reactor adjoint in the upper portions of the core. As pressurization transients are typicallylimiting, the impact of an upward shift in axial power on the transient power prediction should beconsidered. The upward shift in reactor adjoint directly affects the core void reactivity coefficientand tends to increase the sensitivity of the core reactivity to a pressure wave, since the backpressure wave is dissipated by void collapse in the upper parts of the core. Therefore, the corewide transient power would be increased relative to the base case, which appears to result in anincrease in the DCPR.

The results of the TVA sensitivity analysis do not demonstrate this trend. Address whyimposing a fixed void fraction reduction does not result in a higher DCPR.


The trend described above in the RAI question for SRXB-1 19 is for an instantaneous change invoids. As discussed in the previous response to SRXB-1 18, the results of the SRXB-88sensitivity analyses were based on fuel depleted with the changes in the void-qualitycorrelations. AREVA concurs with the general trend as described above for an instantaneouschange in void. The analysis of SRXB-88 was repeated for an instantaneous change in voids.Relative to the Reference case, the change in ACPR was -0.002 for +0.05 voids and +0.01 for-0.05 voids, which is consistent with the staff's observations.

NRC RAI SRXB-120

The void increase cases exhibited opposite trends relative to the void reduction cases. Thestaff found that the void reduction cases were not consistent with the staffs expectations.Provide information similar to the information requested in SRXB-1 18 and 119 for the fixedincrease in void fraction sensitivity analyses.

For each case in Study 1 provide:

* The limiting bundle: core location, initial radial peaking factor and axial power shape" Plots of the perturbed axial and radial core power shape" Plots of transient limiting bundle peak rod heat flux and mass flow rate* Plots of transient critical CPR" A comparison of the predicted power pulse heights and widths.


For an increase in void fraction, the responses to SRXB-1 18 and SRXB-1 19 provide therequested information. That is, the void trend was explained and the change did not result in asignificant impact to the SLMCPR and the transient analyses are consistent with the staff'sexpectations when instantaneous voids are considered.

E2-66


NRC RAI SRX8-119

Continuing with the void fraction reduction case, the decrease in void fraction would simultaneously result in a redistribution of the axial power. Since those higher void nodes would have a greater reactivity response than low void nodes, the axial power distribution would shift upwards in the core. The upward shift in the axial power distribution has the effect of increasing the reactor adjoint in the upper portions of the core. As pressurization transients are typically limiting, the impact of an upward shift in axial power on the transient power prediction should be considered. The upward shift in reactor adjoint directly affects the core void reactivity coefficient and tends to increase the sensitivity of the core reactivity to a pressure wave, since the back pressure wave is dissipated by void collapse in the upper parts of the core. Therefore, the core wide transient power would be increased relative to the base case, which appears to result in an increase in the DCPR.

The results of the TVA sensitivity analysis do not demonstrate this trend. Address why imposing a fixed void fraction reduction does not result in a higher DCPR.

Response to SRX8-119

The trend described above in the RAI question for SRXB-119 is for an instantaneous change in voids. As discussed in the previous response to SRXB-118, the results of the SRXB-88 sensitivity analyses were based on fuel depleted with the changes in the void-quality correlations. AREVA concurs with the general trend as described above for an instantaneous change in void. The analysis of SRXB-88 was repeated for an instantaneous change in voids. Relative to the Reference case, the change in .'1CPR was -0.002 for +0.05 voids and +0.01 for -0.05 voids, which is consistent with the staff's observations.

NRC RAI SRX8-120

The void increase cases exhibited opposite trends relative to the void reduction cases. The staff found that the void reduction cases were not consistent with the staff's expectations. Provide information similar to the information requested in SRXB-118 and 119 for the fixed increase in void fraction sensitivity analyses.

For each case in Study 1 provide:

• The limiting bundle: core location, initial radial peaking factor and axial power shape • Plots of the perturbed axial and radial core power shape • Plots of transient limiting bundle peak rod heat flux and mass flow rate • Plots of transient critical CPR • A comparison of the predicted power pulse heights and widths.

Response to SRX8-120

For an increase in void fraction, the responses to SRXB-118 and SRXB-119 provide the requested information. That is, the void trend was explained and the change did not result in a significant impact to the SLMCPR and the transient analyses are consistent with the staff's expectations when instantaneous voids are considered.

E2-66


Below are the requested results for Study 1 and are based on depleting the fuel with the changein the void correlations (the results are not for an instantaneous change in the void correlations).

* The limiting bundle: core location and initial radial peaking factor (Table SXRB-1 20.1).

* Initial axial power shape (Reference case) (Figure SXRB-1 20.1).

* Plots of the perturbed axial power shapes (initial conditions) (Figure SXRB-1 20.1).

* Plots of the perturbed radial core power shapes (initial conditions) (Figure SXRB-120.2).

* Plots of transient limiting bundle peak rod heat flux at axial heights (x/L) of 25%, 50%,and 75% (Figures SXRB-1 20.3 to -120.5).

* Plots of transient limiting bundle mass flow rate at axial heights of 0%, 25%, 50%, 75%

and 100% (Figures SXRB-120.6 to -120.10).

* Plots of transient CPR (Figure SXRB-1 20.11).

* A comparison of the predicted power pulse heights and widths (Table SXRB-120.2).

E2-67


Below are the requested results for Study 1 and are based on depleting the fuel with the change in the void correlations (the results are not for an instantaneous change in the void correlations).

• The limiting bundle: core location and initial radial peaking factor (Table SXRB-120.1).

• Initial axial power shape (Reference case) (Figure SXRB-120.1).

• Plots of the perturbed axial power shapes (initial conditions) (Figure SXRB-120.1).

• Plots of the perturbed radial core power shapes (initial conditions) (Figure SXRB-120.2).

• Plots of transient limiting bundle peak rod heat flux at axial heights (x/L) of 25%, 50%, and 75% (Figures SXRB-120.3 to -120.5).

• Plots of transient limiting bundle mass flow rate at axial heights of 0%, 25%, 50%, 75% and 100% (Figures SXRB-120.6 to -120.10).

• Plots of transient CPR (Figure SXRB-120.11).

• A comparison of the predicted power pulse heights and widths (Table SXRB-120.2).

E2-67


Table SXRB-120.1 Limiting Bundle Data

Study 1Modified

V-Q(-0.05)

ReferenceCalculation

Study 1Modified

V-Q(0.0)

Study 1Modified

V-Q(+0.05)

Location in Core W,J 23,24 21,22 23,24 37,24

MB2 initial radialpeaking factor 1.314 1.295 1.329 1.346

XCT converged initialpeaking factor radial4 [ ] [ ] [ ] [

REGIONSM1 3579U13 15

JR: H17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59

60 1 2 3 4 5 6 7 8 9 10 11 12 13 1415 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 5051 52 53 54 55 56 57 58 59 6 61 62 63 64 65 66 67 68 69 70

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 9293 94 95 96 97 98 99100 101 102 103104 105 106107 108 109110 11132 113 114 115116117 118

119 120 121 122 123 124 125 126 371 28129 130 131 1 133 134 135 136 137 138 139 140 141 142 143144145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 16W 161 162 163 164 165 166 167 168 169 170 171 172

173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202203 204 205 206 207 206 209 210 211 212 213 214 215 216 217 218 219 220 221 m2 223 224 225 226 227 228 229 230 231 23M233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 26D 261 262263 264 265 266 267 26B 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 2W293 294 25 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 310 321 3323 324 325 326 3V7 328 329 330 331 332 333 334 335 M6 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 3M2353 354 355 356 357 358 359 36D 361 362 363 364 365 366 367 36B 369 370 371 372 373 374 375 376 377 378 379 380 381 382383 384 385 386 387 388 389 390 M1 3W 394 3 396 397 398 399 400 40142 403 404 405 407 408 409 410 41142413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 43V 433 434 435 436 437 438 439 440 441 442443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 47247 474 475 C76 477 478 479 480 481 482 483 484485 486 487 488 489 490 491.49Q 493 494 455 496 497 496 499 500 501 502503 504 505 506 507 508 509 510 Sf11512 513 514 W15 516 517 518 519 520 521 =2 52 524 525 526 527 528 529 530 531 531533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 55 554 555556 557 558 559 56W 561 562563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592

593 594 595 596 597 598 599 600 601 602 603 604 605 606 W7 608 6W9 610 611 612 613 614 615 616 617 618 619 60621 622 623 624 625 626 6V7 628 629 630 631 632 633 634 635 636 67 638 639 640 641 642 643 644 645 646647 648 649 650 61 652 653 604 655 656 657 658 659 66W 661 662 663 664 665 666 667 669 669 670 671 672

673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711712 713 714715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 73V 733 734

735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750751 752 753 754 755 756 757 758 759 760 761 762 763 764

IR: 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59

The XCOBRA-T radial that results in a [ ] during the transient event.

E2-68

4


Table SXRB-120.1 Limiting Bundle Data

Study 1 Study 1 Study 1 Modified Modified Modified

Reference v-a v-a v-a Calculation (-0.05) (0.0) (+0.05)

Location in Core I,J 23,24 21 ,22 23,24 37,24

MB2 initial radial peaking factor 1.314 1.295 1.329 1.346

XCT converged initial peaking factor radial4 [ ] [ ] [ ] [ ]

REGIONS lR: 1 3 5 7 9 11 13 15 17 19 21 23 25 'Z7 29 31. 33 35 37 39 41 43 45 47 49 51 53 55 57 59

~: WA

ro 58 56 54 52 50 48 46

1 2 3 4 5 678 9 ro 11 U 13 M 15 16 17 18 19 20 21 22 23 24 25 26 'Z7 28 29 ~

31.~33~35~37~39~41~43"45~~484950

5152535455565758~ro~~~~~~~m~w

n~UN~~TI~~~~~~~~~~~~~~~

~~~%~~~~www~~~m~Dillillrumillllillimm

mmmmmm~~m~~rnm~m~~~m~m~ill~ill~

~~rn~w~m~m~~5m55~m~m~~~~~~mrnm

" mrn~mrnrnmwm~~~~~w~~mm~m~~~m~m~D~

~ ~~~~w~~~mmm~~mm~mmmmm~~~m~w~m~

~ m~~~m~~wm~w~~~w~~~&~m~~~~~~~E~

~ ~~~~w~~mmmmmmmmmm~~~~~~~w~~mm~

~ m~~~m~mmE~m~~~Enmmmmm~mmm~m~m~

~ ~~~~E~~Em~m~~~m~~~~~m~~~~~~~E~

~ E~~~E~E~E~E~~~E~~mmmmmmmmmm~E~

~ E~~~E~~~E~m~m~m~m~G~~~~~~Q~mmm

~ m~m~m~~~m~~~~~mQ~~m~m~~~m~~~w~

~ ~~~~W~~~~~~~~6~~~~~~~~~~~~$mmm

~ mmmmmmm~~~~~~~~~~~m~~~~~~~~~D~

22 ~~~~~~~~~~~~~~m~~~2~~~~~~~~~m~

20 m~~~m~~~~~~~~~~~~~~~~~~~~~~~~~ 18 ~~~~~~~mmmm~~~m~m~~~~~~~~~~~s~ ~ ~~~~~~m~~~~~~~~~~~mrururum~m~~~

M ~~~~~~~~~~~~m~~~~~~~~~oo~~~

U ~~~~~~~~~~~~~~~~~~~~~~~~mrn

ro rnrne~m~~~~~~~~~~~~~~~~~

8 ~~~~~DUU~~~DwnDmmmmm

6 mmmmmmmmmm~mmmmmmmm~

4 mmm~mwm~ro~~_m~w~

2 EBm~~&m~sw~_~~

lR: 1 3 5 7 9 11 13 15 17 19 21 23 25 'Z7 29 31. 33 35 37 39 41 43 45 47 49 51 53 55 57 59

The XCOBRA-T radial that results in a [ ] during the transient event.

E2-68


Table SXRB-120.2 Power Pulse

Pulse Width Pulse Height

Void (seconds) (% rated)

Reference 0.53 318

Modified 0.56 330

Modified +0.05 0.58 328

Modified -0.05 0.54 330

E2-69


Table SXRB-120.2 Power Pulse

Pulse Width Pulse Height Void (seconds) (% rated)

Reference 0.53 318

Modified 0.56 330

Modified +0.05 0.58 328

Modified -0.05 0.54 330

E2-69

NON-PROPRIETARY INFORMATION2.00

1.80

1.60

1.40

. 1.20

o 1.00a,.

,• 0.80

0.60

0.40

0.20

0.00

-.-- Reference-Modified

------ Modified +0.05- - - Modified -0.05

0 5 10 15 20

Axial Node

25 30

Figure SRXB-120.1 Initial Axial Power Shape- Depleted Voids

1.6

1.4

1.2

1

1 0.8

E8 0.6z

0.4

0.2

0 100 200 300 400 500 600 700 800

Bundle Index

Figure SRXB-120.2 Initial Radial Power Distributionfrom Transient Analyses - Depleted Voids

E2-70

NON-PROPRIETARY INFORMATION 2.00 r------------------------------------------------------------------,

1.80 .1----------------------

1.60 +----------------------

1.40 -1----------------;

:8 1.20 -1---------

r:. ~ ..

- - - Reference

--Modified

- - - - - . Modified +0.05

- - - - Modified -0 .05

~ 1.00 -I--------------j'-J'-, <---~~~---------------------'7'__'<

"n; ~ 0_80 .1------

O.BO i----r~---/-·--

0.40 \-- '--r-.. "-----------

0.20

0.00 -I---------~--------~----------~--------~--------__r-----------J

o 5 10 15

Axial Node

20 25

Figure SRXB-120.1 Initial Axial Power Shape - Depleted Voids

30

1 . B ·r-------------------------------------------------------------------~

1.4 ·1----------------

~~;:--;..:...:~ 1.2 +---------=-..;~:::::.-:-- _-,,-. _-~-_-.------------

\ 0.8 ·1------------

O.B +---------

0.4 1-------------

- - - Reference

0.2 --Modified

- - - - - . Modified +0.05

- - - - Modified -0.05

O ~====~~-~--~--~--~-----~--~ o 100 200 300 400 500 BOO

Bundle Index

Figure SRXB-120.2 Initial Radial Power Distribution from Transient Analyses - Depleted Voids

E2-70

700 800


0.50

0.40

0.30

X

0.20

0.10

0.00

-.-.-.--.-.-...... -..... .

---- Reference

- Modified

...... Modified +0.05- - -. Modified -0.05

23 23.5 24

Time (Sec)

24.5 25

Figure SRXB-120.3 Heat Flux vs. Time25% x/L - Depleted Voids

0.50

0.40

0.30

K

0.20

0.10

0.0023 23.5 24 24.5 25

Time (Sec)


E2-71


O.~ ·r----------------------------------------------------------------'

OAO

~ .s: 0.30 :; l:i !. >< " ~

,. /,," ---. ~~~.=~" .:.. . .:. . ..::.'...0-.., '-' .... _ • • ",._ -------- --- ~ <;. -- -- - ----------------- --.-.... ,.

i 0.201--:r

0.10 - - - Reference --Modified

----

..... . Modified +0.05

- . - . Modified -0.05 0.00 l===:::::::;=::::;::=--~_-__ - _ _,_-,__----_--_-__ .......J

23 23.5 24

Time (Sec)

24.5

Figure SRXB-120.3 Heat Flux vs. Time 25% x/L - Depleted Voids

25

O .~ r---------------------------------------------------------------_.

OAO 1-------- __________ ",.c:::==~=

, f .s: 0.30 :; l:i !. >< " ~ : 0.20 -- -:r

0.10 - - - Reference

--Modified ..•... Modified +0.05

- . - . Modified -0.05 0.00 L--______________ .,--________________ ----------____ ~--~~--__ --__ ___l

23 23.5 24

Time (Sec)

24.5

Figure SRXB-120.4 Heat Flux vs. Time 50% x/L - Depleted Voids

E2-71

25

NON-PROPRIETARY INFORMATION0.50

0.40

.r 0.30

.0

0.20

0.10

0.00

.~ ~ ~ , .' .' ..... . . . . . . . . . . -------------

---- Reference

- Modified

------ Modified +0.05

. Modified -0.05

23 23.5 24

Time (Sec)

24.5 25


50

40

0

0LoI

30

20

10

023 23.5 24 24.5

Time (Sec)

25

Figure SRXB-120.6 Mass Flow vs. Time0% x/L - Depleted Voids

E2-72

NON-PROPRIETARY INFORMATION 0.50 .,..-----------------------------------,

0.40 ~-----------.-------- -

0.10 .-= =:====;---

- - Modified

- - - - - -Modified +0.05

[--- ..... ~~ - - - - Modified -005

0.00 .L=--=.--=:::==::=~~--_-~_-~_-~ ____ - __ - __ __J 23 23.5 24

Time (Sec)

24.5

Figure SRXB-120.5 Heat Flux vs. Time 75% xlL - Depleted Voids

25

50 r--------------------------------,

40 -------------------------------------------------------------~

10 ·~~=====>---------------------------- - - Reference

--Modified

- - - - - . Modified +0.05 - - - - Modified -0.05

o t=~~~~~--____ ~ __ ~ __ --____________ --__ --~ 23 23.5 24

Time (Sec)

24.5

Figure SRXB-120.6 Mass Flow vs. Time 0% x/L - Depleted Voids

E2-72

25


50

40

0

I0:-

U=

30

20

---- Reference-Modified

...... Modified +0.05

.... Modified -0.05

10

023 23.5 24 24.5 25

50

40

Time (Sac)


.2U.

'AC-

2

LUU=

30

20

10

0 eeec-Modified

....Modified +0.05- .Modified -0.05

23 23.5 24 24.5 25

Time (Sec)


E2-73


50 r-------------------------------------------------------------~

40 ./--- --------------------

.--.-

10 1;======

- - - - - . Modified +0.05

_- _- _- ~::~:ce ~ - - - - Modified -0 .05

O L===~~~~--________ ~------____ ~--______ ~ 23 23.5 24

Time (Sec)

24.5


25

50 r---------------------------------------------------------------,

40 - ---

~ 30r-~~~~ - - - _.-------g ~ o

u:: '" In co

::E 20 - ---

10 lr=======-, - - - Reference

--Modified

- - - - - -Modified +0.05 - - - - Modified -0 .05

23 23.5 24

Time (Sec)

24.5


E2-73

25

NON-PROPRIETARY INFORMATION50

40

0U-,T

30

20

10

23 23.5 24 24.5 25

Time (Sec)

Figure SRXB-120.9 Mass Flow Vs. Time75% x/L - Depleted Voids

50

40

30F

LL'aa-a2

20

10

23 23.5 24 24.5 25

Time (Sac)


E2-74

.!!! g ~ 0 u:: <II <II to

:IE

~ o u:: <II <II to

NON-PROPRIETARY INFORMATION 50 .-----------------------------------------------------------------~

40

30 - - ..... &

20

10 1r=======-, - - - Reference

--Modified

- - - - - -Modified +0.05 - - - - Modified -0.05

O t=~~~~~~~~~~~~~--~~~ __ ~~~~ 23 23.5 24

Time (Sec)

24.5

Figure SRXB-120.9 Mass Flow Vs. Time 75% x/L - Depleted Voids

25

50.------------------------------------------------------------------.

40 -1-----

:IE 20 +-------- - ---- --- --

10 t=====-~---------- - - Reference

--Modified

- - - - - . Modified +0.05 - - - - Modified -0.05

O t=~~====~~~~--~~~------____ ~ __ --______ ~ 23 23.5 24

Time (Sec)

24.5

Figure SRXB-120.1 0 Mass Flow vs. Time 100% x1L - Depleted Voids

E2-74

25


f-'

Figure SRXB-120.11 CPR vs. TimeDepleted Voids

E2-75


Figure SRX8-120.11 CPR VS. Time Depleted Voids

E2-75

.J


NRC RAI SRXB-121

It should be noted that the increase in the operating limit CPR (OLMCPR) for the increased voidfraction cases is substantial relative to the base case. The Study 1 increase in OLMCPR is0.014 and the Study 2 increase in OLMCPR is 0.027.

In response to RAI SRXB-88, TVA stated that COTRANSA2 includes a 110 percent multiplieron integral thermal power as a conservative assumption. However, XN-NF-80-19(P)(A) ExxonNuclear Methodology for Boiling Water Reactors, THERMEX: Thermal Limits MethodologySummary Description Section 4.4 states: "In developing the methodology for the COTRANSAcode Exxon Nuclear addressed uncertainties in the code through the integral power variable.The revised methodology uses a more conservative deterministic bounding value (+10 percent)for the integral power uncertainty." TVA's evaluation of the 110 percent conservatism found thatthe OLMCPR margin afforded by the conservatism is [[ ]].

While analysis of pre-EPU reactor conditions, such as the Peach Bottom turbine trip tests,indicate that the 110 percent multiplier is adequate. The response to RAI SRXB-88 appears toindicate that at EPU conditions, that the integral thermal power response to a 5 percentuncertainty in void fraction may not be bounded by the conservatism afforded by the 110percent multiplier. This is evidenced by an increase in the OLMCPR in Study 2 that exceedsthe conservatism afforded by the total 110 percent multiplier.

It should be noted that the intent of the 110 percent multiplier is to conservatively bound alluncertainties, including uncertainties in other important variables such as flow and frictionfactors.

Given that the OLMCPR increase exceeds the 110 percent multiplier margin, provide ademonstration that the integrated effect of all conservatisms in COTRANSA2 for Unit 2 at EPUconditions is adequate. This demonstration may be provided by qualification against relevantoperating plant transient data to ensure conservatism of the methodology for EPU or near-EPUconditions or by comparison against a rigorous statistical treatment of all uncertainties or bysome alternative quantitative and applicable means.


As discussed in the response to RAI SRXB-88 (Reference SRXB-121.1), the [ ]correlation in MICROBURN-B2 was modified to adjust the mean to match the measuredATRIUM-10 void fraction data for both high and low void fractions. The modified [ Icorrelation was then further modified to generate two bounding correlations for the ATRIUM-10data of ±0.05 mean void. The results of the modified correlations were shown inFigure SRXB-88.2 and were used in sensitivity Study 1 and Study 2.

The sensitivity studies described in the response to RAI SRXB-88 are somewhat artificial andonly capture the sensitivity of portions of the methodology to void correlation uncertainty.Study 2 is especially artificial in that the results of the study only reflect the increase in core voidreactivity coefficient. Study 2 does not reflect that a different change in void fraction wouldoccur for a given pressure change with the modified void correlation. Study 1 included thiseffect and resulted in a smaller effect on the OLMCPR. Other effects of using a different voidcorrelation uncertainty are not incorporated into Study 1 (e.g., pressure drop correlationcoefficients would be different). These sensitivity studies are not complete assessments of theimpact of void correlation uncertainty on OLMCPR.

E2-76


NRC RAI SRXB-121

It should be noted that the increase in the operating limit CPR (OLMCPR) for the increased void fraction cases is substantial relative to the base case. The Study 1 increase in OLMCPR is 0.014 and the Study 2 increase in OLMCPR is 0.027.

In response to RAI SRXB-88, TVA stated that COTRANSA2 includes a 110 percent multiplier on integral thermal power as a conservative assumption. However, XN-NF-80-19(P)(A) Exxon Nuclear Methodology for Boiling Water Reactors, THERMEX: Thermal Limits Methodology Summary Description Section 4.4 states: "In developing the methodology for the COTRANSA code Exxon Nuclear addressed uncertainties in the code through the integral power variable. The revised methodology uses a more conservative deterministic bounding value (+10 percent) for the integral power uncertainty." TVA's evaluation of the 110 percent conservatism found that the OLMCPR margin afforded by the conservatism is [[ ]].

While analysis of pre-EPU reactor conditions, such as the Peach Bottom turbine trip tests, indicate that the 110 percent multiplier is adequate. The response to RAI SRXB-88 appears to indicate that at EPU conditions, that the integral thermal power response to a 5 percent uncertainty in void fraction may not be bounded by the conservatism afforded by the 110 percent multiplier. This is evidenced by an increase in the OLMCPR in Study 2 that exceeds the conservatism afforded by the total 110 percent multiplier.

It should be noted that the intent of the 110 percent multiplier is to conservatively bound all uncertainties, including uncertainties in other important variables such as flow and friction factors.

Given that the OLMCPR increase exceeds the 110 percent multiplier margin, provide a demonstration that the integrated effect of all conservatisms in COTRANSA2 for Unit 2 at EPU conditions is adequate. This demonstration may be provided by qualification against relevant operating plant transient data to ensure conservatism of the methodology for EPU or near-EPU conditions or by comparison against a rigorous statistical treatment of all uncertainties or by some alternative quantitative and applicable means.


As discussed in the response to RAI SRXB-88 (Reference SRXB-121.1), the [ correlation in MICROBURN-B2 was modified to adjust the mean to match the measured ATRIUM-10 void fraction data for both high and low void fractions. The modified [ ] correlation was then further modified to generate two bounding correlations for the ATRIUM-10 data of ±0.05 mean void. The results of the modified correlations were shown in Figure SRXB-88.2 and were used in sensitivity Study 1 and Study 2.

The sensitivity studies described in the response to RAI SRXB-88 are somewhat artificial and only capture the sensitivity of portions of the methodology to void correlation uncertainty. Study 2 is especially artificial in that the results of the study only reflect the increase in core void reactivity coefficient. Study 2 does not reflect that a different change in void fraction would occur for a given pressure change with the modified void correlation. Study 1 included this effect and resulted in a smaller effect on the OLMCPR. Other effects of using a different void correlation uncertainty are not incorporated into Study 1 (e.g., pressure drop correlation coefficients would be different). These sensitivity studies are not complete assessments of the impact of void correlation uncertainty on OLMCPR.

E2-76


It should be noted that a ±0.05 perturbation of the void correlation used in the SRXB-88sensitivity studies is substantial. For example, the +0.05 void scenario is equivalent to a[ ]. Themeasure of void correlation uncertainty used in the sensitivity analyses was somewhat arbitrarilydefined as a value that would bound the ATRIUM-10 test data. In a BWR, the core power andpower distribution are tightly coupled with the void fraction and a large error in predicted corevoid fraction would have a significant effect on the predicted power distribution measurementsobtained from operating reactors. If the error in void fraction was as large as assumed in theSRXB-88 sensitivity studies, the effect would be observed in comparisons of predicted tomeasured power distributions obtained from operating reactors.

Additional calculations were performed [

]. These results confirm the conclusion stated above that theincreased void variation of +0.05 is not realistic.

Integral power is a parameter obtainable from test measurements that is directly related toACPR and provides a means to assess code uncertainty. The COTRANSA transient analysismethodology was a predecessor to the COTRANSA2 methodology. The integral power figureof merit was introduced with the COTRANSA methodology as a way to assess (not account for)code uncertainty impact on ACPR. From COTRANSA analyses of the Peach Bottom turbine triptests, the mean of the predicted to measured integral power was 99.7% with a standard

E2-77


It should be noted that a ±0.05 perturbation of the void correlation used in the SRXB-88 sensitivity studies is substantial. For example, the +0.05 void scenario is equivalent to a [ ]. The measure of void correlation uncertainty used in the sensitivity analyses was somewhat arbitrarily defined as a value that would bound the ATRIUM-10 test data. In a BWR, the core power and power distribution are tightly coupled with the void fraction and a large error in predicted core void fraction would have a significant effect on the predicted power distribution measurements obtained from operating reactors. If the error in void fraction was as large as assumed in the SRXB-88 sensitivity studies, the effect would be observed in comparisons of predicted to measured power distributions obtained from operating reactors.

Additional calculations were performed [

]. These results confirm the conclusion stated above that the increased void variation of +0.05 is not realistic.

Integral power is a parameter obtainable from test measurements that is directly related to ~CPR and provides a means to assess code uncertainty. The COTRANSA transient analysis methodology was a predecessor to the COTRANSA2 methodology. The integral power figure of merit was introduced with the COTRANSA methodology as a way to assess (not account for) code uncertainty impact on ~CPR. From COTRANSA analyses of the Peach Bottom turbine trip tests, the mean of the predicted to measured integral power was 99.7% with a standard

E2-77


deviation of 8.1%. AREVA (Exxon Nuclear at the time) initially proposed to treat integral poweras a statistical parameter. However, following discussions with the NRC, it was agreed to applya deterministic 110% integral power multiplier (penalty) on COTRANSA calculations forlicensing analyses. That increase was sufficient to make the COTRANSA predicted tomeasured integral power conservative for all of the Peach Bottom turbine trip tests.

COTRANSA2 (Reference SRXB-121.2) was developed and approved as a replacement forCOTRANSA in the AREVA thermal limits methodology (Reference SRXB-121.3). Initially it wasnot planned to use the 110% integral power multiplier with the COTRANSA2 methodology.COTRANSA2 predictions of integral power were conservative for all Peach Bottom turbine triptests. The minimum conservatism was [ ] and the mean of the predicted to measuredintegral power was [ ]. The comparisons to the Peach Bottom turbine trip testsdemonstrated that the 110% integral power multiplier was not needed for COTRANSA2.However, because the thermal limits methodology that was approved independently ofCOTRANSA2 included discussion of the 110% integral power multiplier, the use of the multiplierwas retained for COTRANSA2 licensing calculations. With the 110% multiplier, theCOTRANSA2 predicted to measured mean integral power is [ ] for thePeach Bottom turbine trip tests. Applying a [ ] integral power multiplier provides anOLMCPR conservatism of [ ] versus the [ ] reported in the response to RAISRXB-88 for the 110% multiplier alone.

To summarize the above paragraphs, the sensitivity studies described in the response toRAI SRXB-88 overestimate the potential impact of uncertainty in the void correlation and the110% integral power multiplier is just one part of the conservatism in the COTRANSA2methodology and application process that covers methodology uncertainties.

COTRANSA2 is not a statistical methodology and uncertainties are not directly input to theanalyses. The methodology is a deterministic bounding approach that contains sufficientconservatism to offset uncertainties in individual phenomena. Conservatism is incorporated inthe methodology in two ways: (1) computer code models are developed to produce conservativeresults on an integral basis relative to benchmark tests, and (2) important input parameters arebiased in a conservative direction in licensing calculations. Justification that the integratedeffect of all the conservatisms in COTRANSA2 licensing analyses is adequate for EPUoperation is provided below.

The COTRANSA2 methodology results in predicted power increases that are bounding] on average) relative to Peach Bottom benchmark tests. In addition, for

licensing calculations a 110% multiplier is applied to the calculated integral power toprovide additional conservatism. This approach adds significant conservatism to thecalculated OLMCPR as discussed previously.

Biasing of important input parameters in licensing calculations provides additionalconservatism in establishing the OLMCPR. The Peach Bottom turbine trips wereperformed assuming the measured performance of important input parameters such ascontrol rod scram speed and turbine valve closing times. For licensing calculations,these (and other) parameters are biased in a conservative bounding direction. Theseconservative assumptions are not combined statistically; assuming all parameters arebounding at the same time produces very conservative results.

Assessments such as the Peach Bottom tests indicate that the integrated effect of all theconservatism in COTRANSA2 is adequate for non-EPU reactor conditions (as stated inthe RAI). To demonstrate that the impact of the change in void-quality correlations is

E2-78


deviation of 8.1 %. AREVA (Exxon Nuclear at the time) initially proposed to treat integral power as a statistical parameter. However, following discussions with the NRC, it was agreed to apply a deterministic 110% integral power multiplier (penalty) on COTRANSA calculations for licensing analyses. That increase was sufficient to make the COTRANSA predicted to measured integral power conservative for all of the Peach Bottom turbine trip tests.

COTRANSA2 (Reference SRXB-121.2) was developed and approved as a replacement for COTRANSA in the AREVA thermal limits methodology (Reference SRXB-121.3). Initially it was not planned to use the 110% integral power multiplier with the COTRANSA2 methodology. COTRANSA2 predictions of integral power were conservative for all Peach Bottom turbine trip tests. The minimum conservatism was [ ] and the mean of the predicted to measured integral power was [ ]. The comparisons to the Peach Bottom turbine trip tests demonstrated that the 110% integral power multiplier was not needed for COTRANSA2. However, because the thermal limits methodology that was approved independently of COTRANSA2 included discussion of the 110% integral power multiplier, the use of the multiplier was retained for COTRANSA2 licensing calculations. With the 110% multiplier, the COTRANSA2 predicted to measured mean integral power is [ ] for the Peach Bottom turbine trip tests. Applying a [ ] integral power multiplier provides an OLMCPR conservatism of [ ] versus the [ ] reported in the response to RAI SRXB-88 for the 110% multiplier alone.

To summarize the above paragraphs, the sensitivity studies described in the response to RAI SRXB-88 overestimate the potential impact of uncertainty in the void correlation and the 110% integral power multiplier is just one part of the conservatism in the COTRANSA2 methodology and application process that covers methodology uncertainties.

COTRANSA2 is not a statistical methodology and uncertainties are not directly input to the analyses. The methodology is a deterministic bounding approach that contains sufficient conservatism to offset uncertainties in individual phenomena. Conservatism is incorporated in the methodology in two ways: (1) computer code models are developed to produce conservative results on an integral basis relative to benchmark tests, and (2) important input parameters are biased in a conservative direction in licensing calculations. Justification that the integrated effect of all the conservatisms in COTRANSA2 licensing analyses is adequate for EPU operation is provided below.

• The COTRANSA2 methodology results in predicted power increases that are bounding ( [ ] on average) relative to Peach Bottom benchmark tests. In addition, for licensing calculations a 110% multiplier is applied to the calculated integral power to provide additional conservatism. This approach adds significant conservatism to the calculated OLMCPR as discussed previously.

• Biasing of important input parameters in licensing calculations provides additional conservatism in establishing the OLMCPR. The Peach Bottom turbine trips were performed assuming the measured performance of important input parameters such as control rod scram speed and turbine valve closing times. For licensing calculations, these (and other) parameters are biased in a conservative bounding direction. These conservative assumptions are not combined statistically; assuming all parameters are bounding at the same time produces very conservative results.

• Assessments such as the Peach Bottom tests indicate that the integrated effect of all the conservatism in COTRANSA2 is adequate for non-EPU reactor conditions (as stated in the RAI). To demonstrate that the impact of the change in void-quality correlations is

E2-78


similar for EPU and non-EPU conditions, the RAI SRXB-88 sensitivity analyses (Study 1)were repeated for BFN Unit 3 Cycle 14 without EPU. The change in ACPR relative tothe reference cases for EPU and non-EPU are shown in the table below:

A(ACPR)

Case EPU Non-EPU

+0.05 void +0.016 +0.024

-0.05 void -0.001 -0.007

Based on these results for EPU and non-EPU conditions, it is concluded that EPUconditions do not increase the sensitivity to a change in the void correlation.

As discussed previously, the core axial power distribution is tightly coupled with the voidfraction. A large error in predicted void fraction would have a significant effect on thepredicted axial power distribution measurements obtained from operating reactors. Thevery good comparisons between predicted and measured axial power distributionsobtained from operating reactors indicates that the void distribution within the core isbeing predicted well.Minimal plant transient data at EPU conditions is available to benchmark transientanalysis methodologies. However, at the request of the NRC, a COTRANSA2 analysiswas performed for a recent event that occurred at a BWR/4 approved for EPU operation.The event involved a reduction in pump speed in one of the recirculation loops followedby a sudden increase in the pump speed approximately 40 seconds later. The event didnot pose a challenge to the fuel; however, the event did result in a significant change incore void fraction. Because of the tight coupling between core void fraction and corepower, a comparison of the predicted to measured core power response during theevent is a good way to assess the accuracy of the void correlation. For this analysis, abest estimate approach was used and event specific licensing conservatisms were notapplied (e.g., measured data used as boundary conditions, realistic control systemparameters, best estimate core neutronics data). The recirculation pump speed versustime from the plant data was used as a boundary condition for the analysis(Figure SRXB-121.13). The COTRANSA2 analysis predicted the core power andreactor pressure response very well (Figures SRXB-121.14 and SRXB-121.15). Thevery good agreement for the predicted core power reached following the pump runbackand the following pump runup indicates a good prediction of the core void fraction duringthe event.

Based on the above discussions, the impact of void correlation uncertainty is inherentlyincorporated in the analytical methods used to determine the OLMCPR. No additionaladjustments to the OLMCPR are required to address void correlation uncertainty.

References:

SRXB-121.1 June 3, 2008, TVA Letter to NRC, Browns Ferry Nuclear Plant (BFN) - Units 2And 3 - Technical Specifications (TS) Change TT-418 - Extended Power Uprate(EPU) - Supplemental Response To Round 16 Request For AdditionalInformation (RAI) - SRXB-88 (TAC Nos. MD5263 AND MD5264) (MI081640325).

E2-79


similar for EPU and non-EPU conditions, the RAI SRXB-88 sensitivity analyses (Study 1) were repeated for BFN Unit 3 Cycle 14 without EPU. The change in ~CPR relative to the reference cases for EPU and non-EPU are shown in the table below:

~(~CPR)

Case EPU Non-EPU

+O.OS void +0.016 +0.024

-O.OS void -0.001 -0.007

Based on these results for EPU and non-EPU conditions, it is concluded that EPU conditions do not increase the sensitivity to a change in the void correlation.

• As discussed previously, the core axial power distribution is tightly coupled with the void fraction. A large error in predicted void fraction would have a significant effect on the predicted axial power distribution measurements obtained from operating reactors. The very good comparisons between predicted and measured axial power distributions obtained from operating reactors indicates that the void distribution within the core is being predicted well.

• Minimal plant transient data at EPU conditions is available to benchmark transient analysis methodologies. However, at the request of the NRC, a COTRANSA2 analysis was performed for a recent event that occurred at a BWR/4 approved for EPU operation. The event involved a reduction in pump speed in one of the recirculation loops followed by a sudden increase in the pump speed approximately 40 seconds later. The event did not pose a challenge to the fuel; however, the event did result in a significant change in core void fraction. Because of the tight coupling between core void fraction and core power, a comparison of the predicted to measured core power response during the event is a good way to assess the accuracy of the void correlation. For this analysis, a best estimate approach was used and event specific licensing conservatisms were not applied (e.g., measured data used as boundary conditions, realistic control system parameters, best estimate core neutronics data). The recirculation pump speed versus time from the plant data was used as a boundary condition for the analysis (Figure SRXB-121.13). The COTRANSA2 analysis predicted the core power and reactor pressure response very well (Figures SRXB-121.14 and SRXB-121.1S). The very good agreement for the predicted core power reached following the pump runback and the following pump runup indicates a good prediction of the core void fraction during the event.

Based on the above discussions, the impact of void correlation uncertainty is inherently incorporated in the analytical methods used to determine the OLMCPR. No additional adjustments to the OLMCPR are required to address void correlation uncertainty.

References:

SRXB-121.1 June 3, 2008, TVA Letter to NRC, Browns Ferry Nuclear Plant (BFN) - Units 2 And 3 - Technical Specifications (TS) Change TT-418 - Extended Power Uprate (EPU) - Supplemental Response To Round 16 Request For Additional Information (RAI) - SRXB-88 (TAC Nos. MDS263 AND MDS264) (MI08164032S).

E2-79


SRXB-121.2 ANF-913(P)(A) Volume 1 Revision 1 and Volume 1 Supplements 2, 3 and 4,COTRANSA2: A Computer Program for Boiling Water Reactor TransientAnalyses, Advanced Nuclear Fuels Corporation, August 1990.

SRXB-121.3 XN-NF-80-19(P)(A) Volume 3 Revision 2, Exxon Nuclear Methodology for BoilingWater Reactors THERMEX: Thermal Limits Methodology Summary Description,Exxon Nuclear Company, January 1987.

E2-80


SRXB-121.2 ANF-913(P)(A) Volume 1 Revision 1 and Volume 1 Supplements 2,3 and 4, COTRANSA2: A Computer Program for Boiling Water Reactor Transient Analyses, Advanced Nuclear Fuels Corporation, August 1990.

SRXB-121.3 XN-NF-80-19(P)(A) Volume 3 Revision 2, Exxon Nuclear Methodology for Boiling Water Reactors THERMEX: Thermal Limits Methodology Summary Description, Exxon Nuclear Company, January 1987.

E2-80


r-

Figure SRXB-121.1 BFN 2D TIP Statistic Comparison forVariations of the Void Quality Correlation

r"

Figure SRXB-121.2 BFN 3D TIP Statistic Comparison forVariations of the Void Quality Correlation

E2-81

r

r


Figure SRXB-121.1 BFN 20 TIP Statistic Comparison for Variations of the Void Quality Correlation

Figure SRXB-121.2 BFN 30 TIP Statistic Comparison for Variations of the Void Quality Correlation

E2-81


Figure SRXB-121.3 BFN Core Average Axial TIP Comparison at9026 MWd/MTU for Variations of the Void Quality Correlation


E2-82


r

r Figure SRXB-121.3 BFN Core Average Axial TIP Comparison at

9026 MWd/MTU for Variations of the Void Quality Correlation

Figure SRXB-121.4 BFN Core Average Axial TIP Comparison at 1755 MWd/MTU for Variations of the Void Quality Correlation

E2-82


r-

U-


r

U-


E2-83



r


E2-83

.J

.J


Figure SRXB-121. 7 A BWR/4 at EPU 2D TIP Statistic Comparison forVariations of the Void Quality Correlation

r

Figure SRXB-121.8 A BWR/4 at EPU 3D TIP Statistic Comparison forVariations of the Void Quality Correlation

E2-84


r

Figure SRXB-121. 7 A BWRl4 at EPU 20 TIP Statistic Comparison for Variations of the Void Quality Correlation

r

Figure SRXB-121.8 A BWRl4 at EPU 30 TIP Statistic Comparison for Variations of the Void Quality Correlation

E2-84

.J

.J


r.

-JFigure SRXB-121.9 A BWR/4 at EPU Core Average Axial TIP Comparison at

2127 MWd/MTU for Variations of the Void Quality

r

Figure SRXB-121.10 A BWR/4 at EPU Core Average Axial TIP Comparison at10621 MWd/MTU for Variations of the Void Quality Correlation

E2-85


r

Figure SRXB-121.9 A BWRl4 at EPU Core Average Axial TIP Comparison at 2127 MWd/MTU for Variations of the Void Quality

r

Figure SRXB-121.10 A BWRl4 at EPU Core Average Axial TIP Comparison at 10621 MWd/MTU for Variations of the Void Quality Correlation

E2-85

..J

..J



r

(.--


E2-86


r

Figure SRXB-121.11 A BWR/4 at EPU Core Average Axial TIP Comparison at 18459 MWd/MTU for Variations of the Void Quality Correlation

r

Figure SRXB-121.12 A BWR/4 at EPU Core Average Axial TIP Comparison at 2054 MWd/MTU for Variations of the Void Quality Correlation

E2-86

..J


------- Measured

-4a- Analysis Input

0.80

A.0.

I.

0.40

0.20

0.000.0 20.0 40.0 60.0 80.0

Time (sac)

Figure SRXB-121.13 Pump Speed

120

-4

---- M

easured

100.0 120.0

100

80

a.LU

S 60

0

40

20

0 -0.00 20.00 40.00 60.00 80.00 100.00

Time (sec)

120.00

Figure SRXB-121.14 Core Power

E2-87


1.20 ,----------------------------------,

1.00 Il-------------------------r:rr--.;--:-----i'------j

- -- - - - - Measured

--<>-Analysis Input

0 . 80 +-~----------------------f------------1

~ Q.

~ 0.60 +-----\-----------------+--------------1

E :l D.

S-II. W

~ ~

~ ~ 0 II.

0.40 4---------' ~------------_/

0.20 +-----------------------------------1

0.00 -!------.,------,----------,------,---------,---------j

0.0 20 .0 40.0 60.0

Time (sec)

80.0

Figure SRXB-121.13 Pump Speed

100.0

120 r---------------------------------

---+ - - Measured

--Calculated

100 -·~-------------------_r1

80

60 ·

40

20 -~-------------------------------_i

0 +------.,------,----------,------,---------r--------1

120.0

0.00 20.00 40.00 60.00

Time (sec)

80.00 100.00 120.00

Figure SRXB-121.14 Core Power

E2-87


1100 -r-

1050

1000

9.

950

+

...- -. Reactor Pressure, Measured

x Dome Pressure, Measured

-- Dome Pressure, Calculated

900

0 20 40 60

Time (sac)

80 100 120

Figure SRXB-1 21.15 Reactor Pressure

E2-88

Ci 'iii .S:


1100 ,------------------------------------------------------------------.

1050 i-----------------------------------------------------------------~

~ 1000 +---------~-------------------------------~----------------------~ I/) I/)

e C1.

950 +-----------~

- - -+ - - Reactor Pressure, Measured

• Dome Pressure , Measured

--Dome Pressure , Calculated

900 +---------~----------~----------r_--------_r----------~--------~

o 20 40 60

Time (sec)

80

Figure SRXB-121.15 Reactor Pressure

E2-88

100 120


NRC RAI SRXB-122

The modified correlations are based on constant slip models. Provide a discussion regardingthe treatment of subcooled boiling. This discussion should address void fraction continuity atthe boiling boundary. Describe any impact on the transient analyses arising from SCRAMreactivity worth if significant differences are expected based on treatment of subcooled boiling.


The thermal hydraulic methodology incorporates the effects of subcooled boiling through use ofthe Levy model. The Levy model predicts a critical subcooling that defines the onset of boiling.The critical subcooling is used with a profile fit model to determine the total flow quality thataccounts for the presence of subcooled boiling. The total flow quality is used with thevoid-quality correlation to determine the void fraction. This void fraction explicitly includes theeffects of subcooled boiling. Application of the Levy model results in a continuous void fractiondistribution at the boiling boundary.

The major influence that the void-quality models have on scram reactivity worth is through thepredicted axial power shape. As discussed in previous responses (e.g., SRXB-121), thevoid-quality models used for ATRIUM-10 fuel result in a very good prediction of the axial powershape.

Below are reponses to the five fuels related RAIs, SRXB-123 through SRXB-127, from NRC's

September 16, 2008, Round 20 RAI.


The following RAIs are based on proprietary draft responses provided during a public meetingheld with the TVA regarding the BFN Units 2 and 3 EPU review on August 7, 2008. Thesequestions focus on the proposed response to SRXB 106.

The draft response states that the calculation terminates in the calculated pressure exceeds thecorrelation bounds ([[ fl). However, under anticipated transient without scram (ATWS)conditions the pressure is expected to exceed this value [[[ 1] pounds per square inchgage (psig)].

NRC RAI SRXB-123

Discuss what allows the code to continue its evaluation of the ATWS transient withoutterminating.


The response to SRXB-1 06 is relative to the XCOBRA-T computer code. The XCOBRA-Tcomputer code is not used in the ATWS overpressurization analysis. The COTRANSA2computer code is the primary code used for the ATWS overpressurization analysis. The ATWSoverpressurization event is not used to establish operating limits for critical power; therefore, theSPCB critical power correlation pressure limit is not a factor in the analysis.

E2-89


NRC RAI SRXB-122

The modified correlations are based on constant slip models. Provide a discussion regarding the treatment of subcooled boiling. This discussion should address void fraction continuity at the boiling boundary. Describe any impact on the transient analyses arising from SCRAM reactivity worth if significant differences are expected based on treatment of subcooled boiling.


The thermal hydraulic methodology incorporates the effects of subcooled boiling through use of the Levy model. The Levy model predicts a critical subcooling that defines the onset of boiling. The critical subcooling is used with a profile fit model to determine the total flow quality that accounts for the presence of subcooled boiling. The total flow quality is used with the void-quality correlation to determine the void fraction. This void fraction explicitly includes the effects of subcooled boiling. Application of the Levy model results in a continuous void fraction distribution at the boiling boundary.

The major influence that the void-quality models have on scram reactivity worth is through the predicted axial power shape. As discussed in previous responses (e.g., SRXB-121), the void-quality models used for ATRIUM-10 fuel result in a very good prediction of the axial power shape.

Below are reponses to the five fuels related RAls, SRXB-123 through SRXB-127, from NRC's September 16, 2008, Round 20 RAI.


The following RAls are based on proprietary draft responses provided during a public meeting held with the TVA regarding the BFN Units 2 and 3 EPU review on August 7,2008. These questions focus on the proposed response to SRXB 106.

The draft response states that the calculation terminates in the calculated pressure exceeds the correlation bounds ([[ ]]). However, under anticipated transient without scram (ATWS) conditions the pressure is expected to exceed this value [[[ ]] pounds per square inch gage (psig)].

NRC RAI SRXB-123

Discuss what allows the code to continue its evaluation of the ATWS transient without terminating.


The response to SRXB-1 06 is relative to the XCOBRA-T computer code. The XCOBRA-T computer code is not used in the ATWS overpressurization analysis. The COTRANSA2 computer code is the primary code used for the ATWS overpressurization analysis. The ATWS overpressurization event is not used to establish operating limits for critical power; therefore, the SPCB critical power correlation pressure limit is not a factor in the analysis.

E2-89


NRC RAI SRXB-124

Discuss how the core coolability under 10 CFR 50.46 is evaluated for this event.


The ATWS event is not limiting relative to acceptance criteria identified in 10 CFR 50.46. Thecore remains covered and adequately cooled during the event. Following the initial powerincrease during the pressurization phase, the core returns to natural circulation conditions afterthe recirculation pumps trip and fuel cladding temperatures are maintained at acceptably lowlevels. The ATWS event is significantly less limiting than the loss-of-coolant accident relative to10 CFR 50.46 acceptance criteria.

NRC RAI SRXB-125

Assuming that the pressure is out of bounds, address how does the code conservativelypredicts the fuel temperature.


As indicated in the response to SRXB-1 23, the pressure limit is for application of the SPCBcritical power correlation. The SPCB correlation is not used in the ATWS overpressurizationanalysis.

Dryout conditions are not expected to occur for the core average channel that is modeled inCOTRANSA2. Dryout might occur in the limiting (high power) channels of the core during theATWS event; however, these channels are not modeled in COTRANSA2 analyses. For theATWS overpressurization analysis, ignoring dryout for the hot channels is conservative in that itmaximizes the heat transferred to the coolant and results in a higher calculated pressure.

NRC RAI SRXB-126

If a fuel rod is predicted in dryout, address how the heat transfer is modeled.


Dryout conditions are not expected to occur for the core average channel that is modeled inCOTRANSA2 for the ATWS overpressurization analysis. Dryout might occur in the limiting(high power) channels of the core during the ATWS event; however, these channels are notmodeled in COTRANSA2 analyses. For the ATWS overpressurization analysis, ignoring dryoutfor the hot channels is conservative in that it maximizes the heat transferred to the coolant andresults in a higher calculated pressure.

E2-90


NRC RAI SRXB-124

Discuss how the core coolability under 10 CFR 50.46 is evaluated for this event.


The A TWS event is not limiting relative to acceptance criteria identified in 10 CFR 50.46. The core remains covered and adequately cooled during the event. Following the initial power increase during the pressurization phase, the core returns to natural circulation conditions after the recirculation pumps trip and fuel cladding temperatures are maintained at acceptably low levels. The ATWS event is significantly less limiting than the loss-of-coolant accident relative to 10 CFR 50.46 acceptance criteria.

NRC RAI SRXB-12S

Assuming that the pressure is out of bounds, address how does the code conservatively predicts the fuel temperature.

Response to SRXB-12S

As indicated in the response to SRXS-123, the pressure limit is for application of the SPCS critical power correlation. The SPCS correlation is not used in the ATWS overpressurization analysis.

Dryout conditions are not expected to occur for the core average channel that is modeled in COTRANSA2. Dryout might occur in the limiting (high power) channels of the core during the ATWS event; however, these channels are not modeled in COTRANSA2 analyses. For the ATWS overpressurization analysis, ignoring dryout for the hot channels is conservative in that it maximizes the heat transferred to the coolant and results in a higher calculated pressure.

NRC RAI SRXB-126

If a fuel rod is predicted in dryout, address how the heat transfer is modeled.


Dryout conditions are not expeCted to occur for the core average channel that is modeled in COTRANSA2 for the ATWS overpressurization analysis. Dryout might occur in the limiting (high power) channels of the core during the ATWS event; however, these channels are not modeled in COTRANSA2 analyses. For the ATWS overpressurization analysis, ignoring dryout for the hot channels is conservative in that it maximizes the heat transferred to the coolant and results in a higher calculated pressure.

E2-90


NRC RAI SRXB-127

Discuss whether the heat transfer modeling approach is conservative in terms of the figure ofmerit (vessel pressure).


Dryout conditions are not expected to occur for the core average channel that is modeled inCOTRANSA2. For the ATWS overpressurization analysis, ignoring dryout for the hot channelsis conservative in that it maximizes the heat transferred to the coolant and results in a highercalculated pressure.

For BWRs, the fluid heat transfer coefficients are high and the thermal resistance of the fluid filmis much smaller than the thermal resistance of the cladding, the cladding-to-pellet gap, and thefuel pellet. Variations in the calculated heat transfer coefficients will have an insignificant effecton the calculated peak vessel pressure.

E2-91


NRC RAI SRXB-127

Discuss whether the heat transfer modeling approach is conservative in terms of the figure of merit (vessel pressure).


Dryout conditions are not expected to occur for the core average channel that is modeled in COTRANSA2. For the ATWS overpressurization analysis, ignoring dryout for the hot channels is conservative in that it maximizes the heat transferred to the coolant and results in a higher calculated pressure.

For BWRs, the fluid heat transfer coefficients are high and the thermal resistance of the fluid film is much smaller than the thermal resistance of the cladding, the cladding-to-pellet gap, and the fuel pellet. Variations in the calculated heat transfer coefficients will have an insignificant effect on the calculated peak vessel pressure.

E2-91

ENCLOSURE 3

TENNESSEE VALLEY AUTHORITYBROWNS FERRY NUCLEAR PLANT (BFN)

UNITS 2 AND 3

TECHNICAL SPECIFICATIONS (TS) CHANGE TS-418EXTENDED POWER UPRATE (EPU)

SUPPLEMENTAL RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION (RAI) ROUNDS 3AND 18 AND RESPONSE TO ROUND 20 FUELS METHODS RAIs

AREVA AFFIDAVIT

This enclosure provides AREVA's affidavit for Enclosure 1.

ENCLOSURE 3

TENNESSEE VALLEY AUTHORITY BROWNS FERRY NUCLEAR PLANT (BFN)

UNITS2AND3

TECHNICAL SPECIFICATIONS (TS) CHANGE TS-418 EXTENDED POWER UPRATE (EPU)

SUPPLEMENTAL RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION (RAI) ROUNDS 3 AND 18 AND RESPONSE TO ROUND 20 FUELS METHODS RAls

AREVA AFFIDAVIT

This enclosure provides AREVA's affidavit for Enclosure 1.

AFFIDAVIT

COMMONWEALTH OF VIRGINIA )) ss.

CITY OF LYNCHBURG )

1. My name is Gayle F. Elliott. I am Manager, Product Licensing, for AREVA

NP Inc. (AREVA NP) and as such I am authorized to execute this Affidavit.

2. I am familiar with the criteria applied by AREVA NP to determine whether

certain AREVA NP information is proprietary. I am familiar with the policies established by

AREVA NP to ensure the proper application of these criteria.

3. I am familiar with the AREVA NP information contained in the Responses to

NRC RAI for Round 18 and Round 20 for Browns Ferry EPU, dated September 2008 and

referred to herein as "Document." Information contained in this Document has been classified

by AREVA NP as proprietary in accordance with the policies established by AREVA NP for the

control and protection of proprietary and confidential information.

4. This Document contains information of a proprietary and confidential nature

and is of the type customarily held in confidence by AREVA NP and not made available to the

public. Based on my experience, I am aware that other companies regard information of the

kind contained in this Document as proprietary and confidential.

5. This Document has been made available to the U.S. Nuclear Regulatory

Commission in confidence with the request that the information contained in this Document be

withheld from public disclosure. The request for withholding of proprietary information is made in

accordance with 10 CFR 2.390. The information for which withholding from disclosure is

AFFIDAVIT

COMMONWEALTH OF VIRGINIA ) ) ss.

CITY OF LYNCHBURG )

1. My name is Gayle F. Elliott. I am Manager, Product Licensing, for AREVA

NP Inc. (AREVA NP) and as such I am authorized to execute this Affidavit.

2. I am familiar with the criteria applied by AREVA NP to determine whether

certain AREVA NP information is proprietary. I am familiar with the policies established by

AREVA NP to ensure the proper application of these criteria.

3. I am familiar with the AREVA NP information contained in the Responses to

NRC RAI for Round 18 and Round 20 for Browns Ferry EPU, dated September 2008 and

referred to herein as "Document." Information contained in this Document has been classified

by AREVA NP as proprietary in accordance with the policies established by AREVA NP for the

control and protection of proprietary and confidential information.

4. This Document contains information of a proprietary and confidential nature

and is of the type customarily held in confidence by AREVA NP and not made available to the

public. Based on my experience, I am aware that other companies regard information of the

kind contained in this Document as proprietary and confidential.

5. This Document has been made available to the U.S. Nuclear Regulatory

Commission in confidence with the request that the information contained in this Document be

withheld from public disclosure. The request for withholding of proprietary information is made in

accordance with 10 CFR 2.390. The information for which withholding from disclosure is

requested qualifies under 10 CFR 2.390(a)(4) "Trade secrets and commercial or financial

information."

6. The following criteria are customarily applied by AREVA NP to determine

whether information should be classified as proprietary:

(a) The information reveals details of AREVA NP's research and development

plans and programs or their results.

(b) Use of the information by a competitor would permit the competitor to

significantly reduce its expenditures, in time or resources, to design, produce,

or market a similar product or service.

(c) The information includes test data or analytical techniques concerning a

process, methodology, or component, the application of which results in a

competitive advantage for AREVA NP.

(d) The information reveals certain distinguishing aspects of a process,

methodology, or component, the exclusive use of which provides a

competitive advantage for AREVA NP in product optimization or marketability.

(e) The information is vital to a competitive advantage held by AREVA NP, would

be helpful to competitors to AREVA NP, and would likely cause substantial

harm to the competitive position of AREVA NP.

The information in the Document is considered proprietary for the reasons set forth in

paragraphs 6(b) and 6(c) above.

7. In accordance with AREVA NP's policies governing the protection and control

of information, proprietary information contained in this Document have been made available,

on a limited basis, to others outside AREVA NP only as required and under suitable agreement

providing for nondisclosure and limited use of the information.

8. AREVA NP policy requires that proprietary information be kept in a secured

file or area and distributed on a need-to-know basis.

requested qualifies under 10 CFR 2.390(a)(4) "Trade secrets and commercial or financial

information. "

6. The following criteria are customarily applied by AREVA NP to determine

whether information should be classified as proprietary:

(a) The information reveals details of AREVA NP's research and development

plans and programs or their results.

(b) Use of the information by a competitor would permit the competitor to

significantly reduce its expenditures, in time or resources, to design, produce,

or market a similar product or service.

(c) The information includes test data or analytical techniques concerning a

process, methodology, or component, the application of which results in a

competitive advantage for AREVA NP.

(d) The information reveals certain distinguishing aspects of a process,

methodology, or component, the exclusive use of which provides a

competitive advantage for AREVA NP in product optimization or marketability ..

(e) The information is vital to a competitive advantage held by AREVA NP, would

be helpful to competitors to AREVA NP, and would likely cause substantial

harm to the competitive position of AREVA NP.

The information in the Document is considered proprietary for the reasons set forth in

paragraphs 6(b) and 6(c) above.

7. In accordance with AREVA NP's policies governing the protection and control

of information, proprietary information contained in this Document have been made available,

on a limited basis, to others outside AREVA NP only as required and under suitable agreement

providing for nondisclosure and limited use of the information.

8. AREVA NP policy requires that proprietary information be kept in a secured

file or area and distributed on a need-to-know basis.

9. The foregoing statements are true and correct to the best of my knowledge,

information, and belief.

SUBSCRIBED before me this 1"Th•'

day of September 2008.

Sherry L. McFadenNOTARY PUBLIC, COMMONWEALTH OF VIRGINIAMY COMMISSION EXPIRES: 10/31/10Reg. # 7079129

NotIry PublicCommonwealth of Vlrginla I

7079129My Commission Expires Oct 31, 2010 1

--- -- .---------- I

9. The foregoing statements are true and correct to the best of my knowledge,

information, and belief.

SUBSCRIBED before me this ----'-It'_"tb __ day of September 2008.

Sherry L. McFaden NOTARY PUBLIC, COMMONWEALTH OF VIRGINIA MY COMMISSION EXPIRES: 10/31/10 Reg. # 7079129

SHERRY L. MCFADEN Noto"ry Public

Commonwealth of V1rg.tnla 7079129

My Commission Expire. Oct 31. 2010 . -- ---~

Tennessee Valley Authority, Post Office Box 2000, …Tennessee Valley Authority, Post Office Box...

Documents

Transcript of Tennessee Valley Authority, Post Office Box 2000, …Tennessee Valley Authority, Post Office Box...