ML063190467

Transmittal of Approved Topical Report DOM-NAF-3 NP-A, Gothic Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment.

ML063190467
Person / Time
Site:	Millstone, Kewaunee, Surry, North Anna
Issue date:	11/06/2006
From:	Gerald Bichof Dominion Energy Kewaunee, Dominion Nuclear Connecticut, Dominion Resources Services, Virginia Electric & Power Co (VEPCO)
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
06-772 DOM-NAF-3, Rev 0.0-NP-A
Download: ML063190467 (158)

Text

Dominion Resources Services, Inc.

5000 Dominion Boulevard, Glen Allen, VA 23060 i' Dominion November 6, 2006 United States Nuclear Regulatory Commission Serial No.06-772 Attention: Document Control Desk NL&OS/ETS: RO Washington, D.C. 20555 Docket Nos. 50-280/281 50-33 8/33 9 50-33 6/42 3 50-305 License Nos. DPR-32/37 NPF-4/7 DPR-65/NPF-49 DPR-43 VIRGINIA ELECTRIC AND POWER COMPANY (DOMINION)

DOMINION NUCLEAR CONNECTICUT. INC. (DNC)

DOMINION ENERGY KEWAUNEE. INC (DEK)

NORTH ANNA AND SURRY POWER STATIONS UNITS 1 AND 2 MILLSTONE POWER STATION UNITS 2 AND 3 KEWAUNEE POWER STATION APPROVED TOPICAL REPORT DOM-NAF-3 NP-A GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO POSTULATED PIPE RUPTURES INSIDE CONTAINMENT In accordance with the NRC guidelines, Dominion, DNC and DEK are hereby submitting the published version of DOM-NAF-3-NP-A. The information requested in the guidelines has been incorporated into the published Topical Report DOM-NAF-3-NP-A. Under separate letter, proprietary copies of DOM-NAF-3-P-A have been provided to Mr. Siva Lingam, NRC Licensing Project Manager for North Anna and Surry Power Stations.

If you have further questions or require additional information, please contact Mr. Thomas Shaub at (804) 273-2763.

Very truly yours, Gerald T. Bischof Vice President - Nuclear Engineering Virginia Electric and Power Company Dominion Nuclear Connecticut, Inc.

Dominion Energy Kewaunee, Inc.

Attachment Commitments made in this letter: None

Serial No.06-772 Docket Nos. 50-280/281/338/339/336/423/305 Page 2 of 3 cc: U.S. Nuclear Regulatory Commission (w/o Att.)

Region 1 475 Allendale Road King of Prussia, Pennsylvania 19406-1415 U.S. Nuclear Regulatory Commission (w/o Att.)

Region 11 Sam Nunn Atlanta Federal Center 61 Forsyth Street, SW Suite 23T85 Atlanta, Georgia 30303 U.S. Nuclear Regulatory Commission (w/o Att.)

Region III 2443 Warrenville Road Suite 210 Lisle, Illinois 60532-4352 Mr. S. C. Burton (w/o Att.)

NRC Senior Resident Inspector Kewaunee Power Station Mr. S. M. Schneider (w/o Att.)

NRC Senior Resident Inspector Millstone Power Station Mr. J. T. Reece (w/o Att.)

NRC Senior Resident Inspector North Anna Power Station Mr. N. P. Garrett (w/o Att.)

NRC Senior Resident Inspector Surry Power Station Mr. D. H. Jaffe (w/o Att.)

NRC Project Manager - Kewaunee Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 0-7-0-1 Rockville, Maryland 20852-2738

Serial No.06-772 Docket Nos. 50-280/281/338/339/336/423/305 Page 3 of 3 Mr. V. Nerses (w/o Att.)

NRC Senior Project Manager - Millstone Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 8C2 Rockville, Maryland 20852-2738 Mr. S. P. Lingamn (w/o Att.)

NRC Project Manager - North Anna and Surry Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 8 G9A Rockville, Maryland 20852-2738 Mr. R. E. Martin (w/o Att.)

NRC Project Manager U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 8-1-12 Rockville, Maryland 20852 Mr. L. N. Olshan (w/o Att.)

NRC Project Manager U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 8-1-12 Rockville, Maryland 20852

Serial No.06-772 Attachment Topical Report DOM-NAF-3-NP-A GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO POSTULATED PIPE RUPTURES INSIDE CONTAINMENT Virginia Electric and Power Company (Dominion)

Dominion Nuclear Connecticut, Inc. (DNC)

Dominion Energy Kewaunee, Inc. (DEK)

Dominion 0R Topical Report DOM-NAF-3 Rev. 0.0-NP-A GOTHIC Methodology For Analyzing the Response to Postulated Pipe Ruptures Inside Containment Nuclear Analysis and Fuel Nuclear Engineering September 2006

DOM-NAF-3-O.0-NP-A GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment Nuclear Analysis and Fuel Department Dominion Richmond, Virginia September 2006 Prepared by:

Dana M. Knee Prepared by:

Albert Gharakhanian Reviewed by:

Joseph 0. Erb o~

Recommended for Approval:

Supervisor, Nuclear Safety Analysis App ~oed:

Director, Nuclear Analysis and Fuel

Topical Report DOM-NAF-3, Rev. 0.0-A includes:

" NRC Safety Evaluation Report, dated August 30, 2006

" Classification / Disclaimer

" Abstract

" Topical Report DOM-NAF-3, which was submitted to the NRC in a letter dated November 1, 2005 (serial number 05-745) and supplemented by a letter dated July 14, 2006 (serial number 06-544)

ATTACHMENTS:

1) NRC Request for Additional Information on DOM-NAF-3 and Dominion Responses, dated June 8, 2006 (14 pages)I

2) Supplemental Information, Replacement Pages and GOTHIC Nodalization Diagrams for DOM-NAF-3, dated July 14, 2006 (34 pages in the proprietary version, 28 pages in the non-proprietaryI version)

3) Original Pages Replaced by Attachment 2 of Dominion letter 06-544, dated July 14, 2006 (7 pages that were included in the original subm-ittal dated November 1, 2005)

DOM-NAF-3-0.0-P-A is a proprietary version of the topical report that is required because in the letter provided as Attachment 2 herein includes proprietary information. A non-proprietary version (DOM-NAF-3-0.0-NP-A) will be published without that Attachment 4. All other content is non-proprietary.

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 lop -,-August 30, 2006 SERIAL #L2(e22ZP-7 Mr. David A. Christian Senior Vice President W- AUG 312006 and Chief Nuclear Officer Virginia Electric and Power Company Innsbrook Technical Center NUCLEAR~ ue0ENStNO 5000 Dominion Boulevard Glen Allen, VA 23060-67 11

SUBJECT:

KEWAUNEE POWER STATION (KEWAUNEE), MILLSTONE POWER STATION, UNIT NOS. 2 AND 3 (MILLSTONE 2 AND 3), NORTH ANNA POWER STATION, UNIT NOS. 1 AND 2 (NORTH ANNA 1 AND 2), AND SURRY POWER STATION, UNIT NOS. 1 AND 2 (SURRY 1 AND 2) -

APPROVAL OF DOMINION'S TOPICAL REPORT DOM-NAF-3, "GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO POSTULATED PIPE RUPTURES INSIDE CONTAINMENT" (TAC NOS. MC8831, MC8832, MC8833, MC8834, MC8835, AND MC8836)

Dear Mr. Christian:

By letter dated November 1, 2005, as supplemented by letters dated June 8 and July 14, 2006, Dominion Energy Kewaunee, Inc., Dominion Nuclear Connecticut, Inc., and Virginia Electric and Power Company, (the licensees), requested approval for the generic application of Topical Report DOM-NAF-3, "GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment."

GOTHIC (Generation of Thermal-Hydraulic Information for Containments) is a general purpose thermal-hydraulics computer code developed by the Electric Power Research Institute for performing containment analyses. The licensees have developed an analytical method using the GOTHIC methodology to replace the current containment analysis at Kewaunee, Millstone 2

.and 3, North Anna 1 and 2, and Surry 1 and 2.

The enclosed Safety Evaluation (SE) documents the basis for the U.S. Nuclear Regulatory Commission (NRC) staff's conclusion's that Topical. Report DOM-NAF-3 is acceptable for the licensees' nuclear facilities. The SE defines the basis for the acceptance of the report.

In accordance with the guidance provided on the NRC website, the licensees are requested to publish an accepted version of this topical report within 3 months of receipt of this letter. The accepted version shall incorporate this letter and the enclosed SE between the title page and the abstract. It must be well indexed such that information is readily located. Also, it must contain, in appendices, historical review information, such as questions and accepted responses, and original report pages that were replaced. The accepted version shall include an

"-A" (designated accepted) following the report identification symbol.

D. Christian I If the NRC staff's criteria or regulations change such that its conclusions as to the acceptabilityI of the topical report are invalidated, then these licensees will be expected to revise and resubmit its respective documentation, or submit justification for the continued applicability of the topical report without revision of the respective documentation.I Sincerely, Ho K. Nieh, Acting Director Division of Policy and Rulemaking Office of Nuclear Reactor Regulation Docket Nos. 50-305, 50-336, 50-423, 50-338, 50-339, 50-280, and 50-281

Enclosure:

I Safety Evaluation cc w/encl: See next pageI

I Virginia Electric and Power Company cc:

Ms. Lillian M. Cuoco, Esq. Office of the Attorney General Senior Counsel Commonwealth of Virginia Dominion Resources Services, Inc. 900 East Main Street Building 475, 5th Floor Richmond, Virginia 23219 Rope Ferry Road Waterford, Connecticut 06385 Mr. Chris L. Funderburk, Director Nuclear Licensing & Operations Support Mr. Donald E. Jernigan lInnsbrook Technical Center Site Vice President Dominion Resources Services, Inc.

Surry Power Station 5000 Dominion Blvd.

Virginia Electric and Power Company Glen Allen, Virginia 23060-6711 5570 Hog Island Road Surry, Virginia 23883-0315 Mr. Jack M. Davis Site Vice President Senior Resident Inspector North Anna Power Station Surry Power Station Virginia Electric and Power Company U. S. Nuclear Regulatory Commission Post Office Box 402 5850 Hog Island Road Mineral, Virginia 23117-0402 Surry, Virginia 23883 Mr. C. Lee Lintecum Chairman County Administrator Board of Supervisors of Surry County Louisa County Surry County Courthouse Post Office Box 160 Surry, Virginia 23683 Louisa, Virginia 23093 Dr. W. T. Lough Old Dominion Electric Cooperative Virginia State Corporation Commission 4201 Dominion Blvd.

Division of Energy Regulation Glen Allen, Virginia 23060 Post Office Box 1197 Richmond, Virginia 23218 Senior Resident Inspector North Anna Power Station Dr. Robert B. Stroube, MD, MPH U.S. Nuclear Regulatory Commission State Health Commissioner 1024 Haley Drive Office of the Commissioner Mineral, Virginia 23117 Virginia Department of Health Post Office Box 2448 Richmond, Virginia 23218

I Millstone Power Station, Unit Nos. 2 and 3 I

cc:

I Edward L. Wilds, Jr., Ph.D. Mr. Evan W. Woollacott Director, Division of Radiation Department of Environmental Protection Co-Chair Nuclear Energy Advisory Council 128 Terry's Plain Road I

79 Elm Street Hartford, CT 06106-5127 Simsbury, CT 06070 Mr. Joseph Roy I

Regional Administrator, Region I Director of Operations U.S. Nuclear Regulatory Commission 475 Allendale Road Massachusetts Municipal Wholesale Electric Company I

King of Prussia, PA 19406 P.O. Box 426 First Selectmen Ludlow, MA 01056 I Town of Waterford Mr. David W. Dodson 15 Rope Ferry Road Waterford, CT 06385 Licensing Supervisor Dominion Nuclear Connecticut, Inc. I Building 475, 5th Floor Charles Brinkman, Director Washington Operations Nuclear Services Westinghouse Electric Company Roper Ferry Road Waterford, CT 06385 I 12300 Twinbrook Pkwy, Suite 330 Rockville, MID 20852 Mr. J. Alan Price Site Vice President Dominion Nuclear Connecticut, Inc.

I Senior Resident Inspector Building 475, 51h Floor Millstone Power Station c/o U.S. Nuclear Regulatory Commission Rope Ferry Road Waterford, CT 06385 I

P. 0. Box 513 Niantic, CT .06357 I Mr. J. W. "Bill" Sheehan Co-Chair NEAC 19 Laurel Crest Drive I Waterford, CT 06385 Ms. Nancy Burton I 147 Cross Highway Redding Ridge, CT 00870 I

I I

I I

Kewaunee Power Station cc:

Resident Inspectors Office Plant Manager U.S. Nuclear Regulatory Commission Kewaunee Power Station N490 Highway 42 N490 Highway 42 Kewaunee, WI 54216-9510 Kewaunee, WI 54216-9511 Regional Administrator, Region IIl Ms. Leslie N.Hartz U.S. Nuclear Regulatory Commission Dominion Energy Kewaunee, Inc.

Suite 210 Kewaunee Power Station 2443 Warrenville Road N 490 Highway 42 Lisle, IL 60532-4351 Kewaunee, WI 54216 David Zeilner Chairman - Town of Carlton N2164 County B Kewaunee,Wl 54216 Mr. Jeffery Kitsembel Electric Division Public Service Commission of Wisconsin PO Box 7854 Madison, WI 53707-7854, Mr. Michael G. Gaffney Dominion Energy Kewaunee, Inc.

Kewaunee Power Station N490 Highway 42 Kewaunee,WI 5421 6 Mr. Thomas L.Breene Dominion Energy Kewaunee, Inc.

Kewaunee Power Station N490 Highway 42 Kewaunee, WI 54216

UNITED STATES 0NUCLEAR REGULATORY COMMISSIONI WASHINGTON, D.C. 20555-0001 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATING TO TOPICAL REPORT DOM-NAF-3 KEWAUNEE POWER STATION (KEWAUNEE)

MILLSTONE POWER STATION, UNIT NOS. 2 AND 3 (MILLSTONE 2 AND 3)

NORTH ANNA POWER STATION, UNIT NOS. 1 AND 2 (NORTH ANNA 1 AND 2)

SURRY POWER STATION, UNIT NOS. 1 AND-2 (SURRY 1 AND 2Q DOCKET NOS. 50-305, 50-336, 50-423, 50-338, 50-339, 50-280, AND 50-281

1.0 INTRODUCTION

By lette r dated November 1, 2005 (Agencywide Documents Access and Management SystemI (ADAMS) Accession Nos. ML053060266 (pages 1 through 40) and ML053060273 (pages 41 through 85)), as supplemented by letters dated June 8 and July 14, 2006 (ADAMS Accession Nos. ML062070314 and ML-062020394, respectively), Dominion Energy Kewaunee, Inc.,I Dominion Nuclear Connecticut, Inc., and Virginia Electric and Power Company (the licensees),

requested approval for the generic application of Topical Report DOM-NAF-3, "GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment."I The licensees requested the U.S. Nuclear Regulatory Commission (NRC) staff's approval of this topical report to perform licensing basis analyses for the containment response for pressurized-water reactors (PWRs) with large, dry containments. The June 8, 2006, letterI responded to the NRC staff's request for additional information, dated April 28, 2006 (ADAMS Accession No. MLO6I 180146). The July 14, 2006, letter corrected a modeling error identified by the licensees, and provided additional information r.equested by the NRC staff.

GOTHIC (Generation of Thermal-Hydraulic Information for Containments) is a general-purpose thermal-hydraulics code for containment analysis developed for the Electric Power Research Institute (EPRI) by Numerical Applications, Inc. (NAI), for applications in the nuclear power industry. This safety evaluation (SE) addresses the licensees' proposed use of GOTHIC for licensing basis analyses. Specifically, GOTHIC methodology would be used to replace the evaluation methods in the updated final safety analysis reports (UFSARs) for the containment design requirements listed below:

1. Loss-of-coolant accident (LOCA) containment peak pressure and temperatureI

2. Main steam line break (MSLB) containment peak pressure and temperatureI

3. LOCA containment depressurization time (CDT) for Surry 1 and 2 and North Anna 1land 2

4. LOCA containment subatmospheric peak pressure (SPP) for Surry 1 and 2 and North Anna 1 and 2

5. Net positive suction head available (NPSHA) for pumps that take suction from the containment sump. For Surry 1 and 2 and North Anna 1 and 2, a time-dependent NPSHA is calculated from a transient containment response for the inside recirculation spray (IRS), outside recirculation spray (ORS), and low head safety injection (LHSI) pumps

6. Minimum and maximum sump water level and liquid temperature for input to other analyses (e.g. , strainer debris head loss and component stress analyses)

-7. Containment liner temperature verification

8. Equipment qualification (EQ) temperature validation, and

9. Transient performance of closed cooling loops for heat exchangers associated with the emergency core cooling systems (ECCS) and containment heat removal systems.

As stated in the licensees' application and discussed in Section 3.0 below, GOTHIC methodology for some of the above proposed design-basis ana *lyseshas been previously approved by the NRC staff for other licensees., Therefore, the primary focus of this SE will be on the proposed use of GOTHIC for applications that have not been previously approved by the NRC. staff; and, hence, could not be implemented by the licensees using the provisions of Title 10 of the Code of Federal Regulations (10 CFR), Part 50, Section 50.59.

2.0 REGULATORY EVALUATION

The General Design Criteria (GDC) contained in 10 CFR Part 50, Appendix A (as stated below),

establishes minimum requirements for the principal design criteria for water-cooled nuclear power plants. The NRC staff considered the following requirements for this review.

Criterion 4, Environmental and dynamic effects design bases. Structures, systems, and components important to safety shall be designed to accommodate the effects of and to be compatible with the environmental conditions associated with normal operation, maintenance, testing, and postulated accidents, including loss-of-coolant accidents. These structures, systems, and components shall be appropriately protected against dynamic effects, including the effects of missiles, pipe whipping, and discharging fluids, that may result from equipment failures and from events and conditions outside the nuclear power unit. However, dynamic effects associated with postulated pipe ruptures in nuclear power units may be excluded from the design basis when analyses reviewed and approved by the Commission demonstrate that the probability of fluid system piping rupture is extremely low under conditions consistent with the design basis for the piping.

Criterion 16, Containment design. Reactor containment and associated systems shall be provided to establish an essentially leak-tight barrier against the

uncontrolled release of radioactivity to the environment and to assure that the containment design conditions important to safety are not exceeded for as long as postulated accident conditions require.

Criterion 38, Containment heat removal. A system to remove heat from theI reactor containment shall be provided. The system safety function shall be to reduce rapidly, consistent with the functioning of other associated systems, the containment pressure and temperature following any loss-of-coolant accidentI and maintain them at acceptably low levels.

Suitable redundancy in components and featu res, and suitable interconnections,I leak detection, isolation, and containment capabilities shall be provided",to assure that for onsite electric power system operation (assuming offsite power-is not available) and for offsite electric power system operation (assuming onsite powerI is not available) the system safety function can be accomplished, assuming a single failure.

Criterion 50, Containment design basis. The reactor containment structure, including access openings, penetrations, and the containment heat removal system shall be designed so that the containment structure and its internal compartments can accommodate, without exceeding the design leakage rate and with sufficient margin, the calculated pressure and temperature conditions resulting from any loss-of-coolant accident. This margin shall reflect consideration of (1) the effects of potential energy sources which have not beenI included in the determination of the peak conditions, such as energy in steam generators and as required by § 50.44 energy from metal-water and other chemical reactions that may result from degradation but not total failure ofI emergency core cooling functioning, (2) the limited experience and experimental data available for defining accident phenomena and containment responses, and (3) the conservatism of the calculational model and input parameters.

The NRC staff used the guidance in the Standard Review Plan (SRP), "Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants - LWR Edition," NUREG-I 0800, Section 6.2.1, "Containment Functional Design," Section 6.2.1.1.A, "PWR Dry Containments, Including Subatmospheric Containments," Section 6.2.1.3, "Mass and Energy Release Analysis for Postulated Loss-of-Coolant Accidents," Section 6.2.1.4, "Mass and Energy Release Analysis for Postulated Secondary System Pipe Ruptures," and Section 6.2.2, "Containment Heat Removal Systems," for this review.

The NRC staff also used Regulatory Guide (RG) 1.82, "Water Sources for Long-TermI Recirculation Cooling Following a Loss-of-Coolant Accident," Revision 3, November 2003, and NUREG-588, "Interim Staff Position on Equipment Qualification of Safety-Related Electrical Equipment," Revision 1, November 1980 as additional guidance for its review.

3.0 TECHNICAL EVALUATION

GOTHIC solves the conservation equations for mass, momentum and energy for multi-component, multi-phase flow in lumped parameter and/or multi-dimensional geometries.

The phase balance equations are coupled by mechanistic models for interface mass, energyI

and momentum transfer that cover the entire fl ow regime from bubbly flow to film/drop flow, as well as single phase flows. The interface models allow for the possibility of thermal non-equilibrium between phases and unequal phase velocities, including countercurrent flow.

GOTHIC includes full treatment of the momentum transport terms in multidimensional models, with optional models for turbulent shear and turbulent mass and energy diffusion. Other phenomena include models for commonly available safety equipment, heat transfer to structures, hydrogen burn and isotope transport.

GOTHIC is maintained by EPRI under a 10 CFR Part 50, Appendix B quality assurance program, is widely used in the U.S. and worldwide, and has been extensively verified and validated by NAI, as documented in the GOTHIC Qualification Manual.' The licensees have indicated that they have participated in the EPRI GOTHIC Advisory Group since the late 1980s in order to ensure a solid understanding of the code capabilities and limitations, to monitor industry applications, and to guide the code qualification effort.

For Topical Report DOM-NAF-3, the licensees used GOTHIC Version 7.2dom, which consists of the EPRI-released Version 7.2 and two enhancements specific to the licensees that were implemented during testing of the GOTHIC containment model for Surry 1 and 2. As noted above, the NRC staff has performed similar reviews for GOTHIC methodology. Recently this included the use of GOTHIC Version 7.0 for Ft. Calhoun 2 and Kewaunee 3 , and GOTHIC Version 7.1 for Framatome Advanced Nuclear Power (ANP), Inc.4 The differences between GOTHIC 7.0, 7.1, 7.2, and 7.2dom, with respect to the analyses of the containment response to design-basis accidents (DBAs) as discussed in this SE are not significant. For the most part, the later versions correct coding errors and include user features to enable the user to apply models consistent with the NRC staff's limitations. For example, in GOTHIC Version 7.2, the Mist Diffusion Layer Model (MDLM) heat and mass transfer option was replaced with the Diffusion Layer Model (DLM) option and optional enhancement factors for mist generation and film roughening effects. The DLM option eliminated the boundary layer mist formation and the height dependent film roughness enhancements to address concerns identified during the NRC staffs review of the Kewaunee amendment (see footnote 3).

In Section 3.0 of DOM-NAF-3, the licensees provided the proposed methodology for constructing GOTHIC models to perform licensing basis analyses for large, dry containments.

The licensees stated that the methods are intended to provide realistic but conservative results based on previously accepted PWR containment methodologies and the extensive validation 1NAI 8907-09 Rev 8, "GOTHIC Containment Analysis Package Qualification Report, Version 7.2,"

published by EPRI, September 2004, 2 ADAMS Accession No. ML-033100290, letter from A. B. Wang, USNRC, to R. T. Ridenoure, Omaha Public Power District, "Fort Calhoun Station, Unit No. 1 - Issuance of Amendment (TAC No. MB7496)," dated November 5, 2003.

3 ADAMS Accession No. ML-032681 050, letter from A. C. McMurtray, USNRC, to T. Coutu, Nuclear Management Company, LLC, "Kewaunee Nuclear Power Plant - Issuance of Amendment (TAC No. MB6408),"

dated September 29, 2003.

4 DM Accession No. ML-052240302, Letter from H. N. Berkow, USNRC, to R. L. Gardner, Framatome, "Final Safety Evaluation for Framatome ANP Topical Report BAW-1 0252(P), Revision 0, 'Analysis of Containment Response to Postulated Pipe Ruptures Using GOTH IC,' (TAC No. MC3783)," August 31, 2005.

base for GOTHIC. In Section 4, the licensees documented GOTHIC containment analyses for Surry 1 and 2 that demonstrated the acceptability of the analysis methodology described in Section 3. Analyses were performed for LOCA peak pressure and temperature, MSLB peak pressure and temperature, containment depressurization, and NPSHA for the LHSI pumps.

Benchmark comparisons were made to the LOCTIC analyses described in the Surry 1 and 2 UFSAR. As described in UFSAR Chapter 14.B.2.3.3.1 for Surry 1 and 2, LOCTIC is a computer program used to calculate containment pressure and temperature transients.

Although not documented in Topical Report DOM-NAF-3, the licensees indicated that the bench-marking also included GOTHIC model adjustments to mimic the same physical behavior as [OCT10. For example, the GOTHIC droplet phase was effectively disabled to support aI comparison to the LOCTIC equilibrium flash model and the containment volume liquid/vapor interface area was set to zero. The licensees stated that these benchmarks used long-term mass and energy data calculated by LOCTIC. The licensees' objective was to demonstrate adequate modeling of containment components, nodalization of piping systems, and modeling of spray systems, with respect to another containment response code. The licensees confirmed that these benchmarks showed a successful comparison of the containmentI The licensees have also performed a sensitivity study for break locations, single failures, and design inputs to determine conservative assumptions for each required analysis for Surry 1 and 2. The results are contained in Table 4.7-1 of Topical Report DOM-NAF-3 and are consistent with the current [OCTI0 analyses for Surry 1 and 2 with the exception of the limiting single failure for the calculation of NPSHA for the ORS and IRS pumps. Since each plant has specific design criteria and engineered safety features that require sensitivity studies, the licensees have stated that they will perform similar bench marking and sensitivity studies to define the set of conservative assumptions for the other plants, as part of the licensing basisU The licensees' demonstration analysis and bench marking for Surry l.and 2 provided reasonable justification for the appropriateness of its proposed GOTHIC methodology. In the following sections, specific components of Topical Report DOM-NAF-3 methodology are discussed further beginning with features that have been previously approved by the NRC staff for similar applications.

3.1 Containment Response Methodology for DBAs As noted above, the NRC staff has previously approved GOTHIC methodologies for analyzing containment response to LOCA and MSLB events (see footnotes 2, 3, and 4). The analyses use models to maximize containment pressure and temperature using inputs to the GOTHIC methodology mass and energy release data that are generated by other NRC staff-approved methods. In response to the NRC staffs request for additional information, the licensees have confirmed that the DOM-NAF-3 methodology for maximizing LOCA and MSLB containmentI pressure and temperature uses NRC staff-approved models for the containment response (e.g.,

the DirectIDLM for heat transfer between passive heat sinks and the containment atmosphere in Topical Report DOM-NAF-3, Section 3.3.2, and the break release droplet model withI 100-micron droplets in Topical Report DOM-NAF-3, Section 3.5.1). This aspect of Topical Report DOM-NAF-3 (Applications 1-4, Section 1.0) is acceptable to the NRC staff and no further review is required.

3.2 Post-Reflood Mass and Energy Release Model The NRC staff has also previously reviewed and approved GOTHIC methodology for post-reflood mass and energy release calculation for Framatome ANP (see footnote 4).

However, in response to the NRC staffs request for additional information, the licensees stated that they were unable to make full comparison with Framatome's methodology because it contained proprietary information. The NRC staff has identified certain differences between Framatome's and the licensees' methodologies with regard to their approach for major component modeling, but the basic elements of both methodologies for long-term (post-reflood) mass and energy release calculation are similar. For both methodologies the transition time for GOTHIC generated mass and energy calculation starts at the end of reflood, once the core is quenched and has been fully covered with water, and ECCS injection maintains the core covered so that decay heat removal and sensible heat removal is assured at all times. Both methodologies account for all remaining stored energy in the primary and secondary systems in accordance with SRP 6.2.1.3 for the post-reflood phase.

The licensees' GOTHIC methodology for long-term mass and energy release acquires the energy for each source term at the end of reflood from the fuel vendor's mass and energy release analysis. The rate of mass and energy release is determined by a simplified GOTHIC reactor coolant system (RCS) model that is coupled to the containment volume. Thus, the flow from the vessel to the containment is dependent on the GOTHIC-calculated containment pressure. Lumped volumes are used for the vessel, down-coiner, cold legs, steam generator secondary side, up-flow portion of the steam generator tubes and down-flow portion of the steam generator tubes. Separate sets of loop and secondary system volumes are used for the intact and broken loops with the connections between the broken loop and containment as necessary for the modeled break location.

In Section 4.3.2 and 4.4.2 of Topical Report DOM-NAF-3, the licensees provided comparison of mass and energy release data calculated by the proposed simplified GOTHIC RCS model with data from the NRC staff-approved FROTH methodology in WCAP-8264-P-A 5 and WCAP-1 0325-P-A6 , as implemented using the Stone & Webster (SWEC) LOCTIC containment response code. For the hot leg break case, the GOTHIC integral mass release matches closely with the FROTH/LOCTIC generated mass release, while the GOTHIC integral energy release was slightly higher and more conservative than the FROTH/LOCTIC generated energy. For the pump suction break case, both the integral mass and energy releases match very closely with the FROTH/LOCTI C generated data.

Although this comparison shows that no margin was gained with the proposed methodology, with respect to mass and energy releases, the simplified RCS methodology provides a reduction in containment depressurization time and a less severe pressure increase following containment spray termination, as shown in Section 4.4 of Topical Report DOM-NAF-3. The 5 WCAP-8264-P-A, Rev. 1, 'Westinghouse Mass and Energy Release Data for Containment Design,"

August 1975. (WCAP-8312-A is the Non-Proprietary version).

6 WCAP-10325-P-A, "Westinghouse LOCA Mass and Energy Release Model for Containment Design -

March 1979 Version," May 1983. (WCAP-1 0326-A is the Non-Proprietary version.)

I licensees attribute this gain in margin to other mechanistic features of GOTHIC that were previously reviewed and approved by the NRC staff. The NRC staff concurs with thisI assessment and finds the methodology for post-reflood mass and energy release calculation acceptable. However, the modeling technique is highly complex and iterative (e.g. modeling of the primary metal stored energy); therefore, as a condition of approval for Topical ReportI DOM-NAF-3, conservative mass and energy release values calculated for Surry 1 and 2 shall be duplicated for North Anna 1 and 2, Millstone 2 and 3, and Kewaunee through appropriate bench marking and model adjustment prior to implementing this methodology in licensingI applications.

3.3 Methodology for Calculating NPSHA Section 3.8 of Topical Report DOM-NAF-3 describes the licensees' proposal to perform transient calculation of NPSHA through conservative model adjustment of the long-term containment response model. The calculation is performed internally in GOTHIC using an industry standard formulation for prediction of pump NPSHA. The same formula was used previously in the SWEC LOCTIC containment analysis methodology, which performed a transient calculation of NPSHA for the current licensing bases' at North Anna 1 and 2 and Surry 1 and 2. NPSHA is the difference between the fluid stagnation pressure and the saturation pressure at the pump intake. NPSHA depends directly on transient predictions of sump temperature, sump water level, and containment pressure.

The licensees intend to employ this methodology for North Anna 1 and 2 and Surry 1 and 2.

Both plants have subatmospheric containments that are required to be depressurized followingI a DBA in accordance with the assumptions in the dose consequence analyses. The current licensing bases for North Anna 1 and 2 and Surry 1 and 2 allow credit for containment over pressure to calculate NPSHA for the the IRS, ORS, and ILHSI pumps891. Although the proposedI methodology is applicable to any large, dry containment, it cannot be used for the other licensees' plants that do not credit containment overpressure to calculate NPSHA in their licensing bases.

In the licensees' proposed methodology, the GOTHIC simplified RCS containment model is used with a separate small volume for the pump suction. The pump suction volume elevation 7 ADAMS Accession No. 9811090068, Letter from J. P. O'Hanlon (VEPCO) to USNRC, "Virginia Electric and Power Company, North Anna and Suny Power Stations Units 1 and 2, Generic Letter 97 Assurance of Net Positive Suction Head for Emergency Core Cooling and Containment Heat Removal Pumps; Response to a Request for Additional Information," Serial No.98-546, October 29, 1998.

8 ADAMS Accession No. 9903030158, Letter from N. Kalyanamn (IJSNRC) to J. P. O'Hanlon (VEPCO),

"Completion of Licensing Action for Generic Letter 97-04, 'Assurance of Sufficient Net Positive Suction Head for Emergency Core Cooling and Containment Heat Removal Pumps'; North Anna Power Station, Unit Nos. 1 and 2I (TAC Nos. MA001 5 and MA001 6)," February 25, 1999.

9 ADAMS Accession No. 9904070170, Letter from G. E. Edison (US NRC) to J. P. O'Hanion (VEPCO),

"Completion of Licensing Action for Generic Letter 97-04, 'Assurance of Sufficient Net Positive Suction Head for Emergency Core Cooling and Containment Heat Removal Pumps'; Suny Power Station, Unit Nos. 1 and 2 (TAC Nos. MA0050 and MAOO51),"'April 1, 1999.

and height are set so that the mid-elevation of the volume is at the elevation of the pump first-stage impeller centerline. The volume pressure, with some adjustments for sump depth, is used in the NPSHA calculation. The temperature in the suction volume provides the saturation pressure. The junction representing piping between the sump and the suction volume reflects the friction pressure drop between the sump and the pump suction. A correlation is used to define the sump depth or liquid level as a function of the water volume in the containment. The correlation accounts for the sump geometry variation with water depth and accounts for the

.holdup of water in other parts of the containment.

The proposed methodology incorporated several adjustments to the simplified RCS containment model to ensure a conservative calculation of NPSHA. A multiplier of 1.2 is applied to the heat transfer coefficient for the containment heat sinks to compensate for the non-conservative values (with respect to NPSHA calculation) generated by the Direct DLM heat transfer Model. All of the spray water is injected as droplets into the containment atmosphere (nozzle spray flow fraction of 1). Analyses are performed using the largest Sauter spray droplet size and a confirmatory analysis is performed by reducing the Sauter diameter by 2, which sufficiently covers code and spray performance uncertainty without creating drops too small that may cause excess droplet holdup in the atmosphere. A conservative water holdup volume is subtracted from the containment liquid volume to reduce the sump water height. Other adjustments include use of upper limit for containment free volume and minimum initial containment pressure. The conservatism incorporated in this methodology meets the applicable regulatory positions in RG 1.82 In Section 4.5 of Topical Report DOM-NAF-3, the licensees provided benchmark results comparing GOTHIC calculation of LHSI pump NPSHA to LOCTIC analyses from the UFSAR for Surry 1 and. 2 for a pump suction break LOCA transient. The GOTHIC results showed good agreement with the LOCTIC case. The more realistic GOTHIC modeling of the RCS and steam generators resulted in slightly more energy being transferred to the containment at the time the LHSI pumps take suction from the sump. At the time of minimum NPSHA, the GOTHIC sump temperature is actually slightly higher than the LOCTIC value; however, the GOTHIC pressure is also higher, yielding a small, net increase in NPSHA. The licensees concluded that the higher sump temperature and containment pressure than LOCTIC is consistent with the additional energy addition from the RCS model, and is considered to be a reasonable and more accurate system response.

The proposed use of GOTHIC methodology to calculate NPSHA uses an industry standard formulation that was previously approved by the NRC staff and incorporates applicable conservatisms contained in RG 1.82. As such, the NRC staff finds this acceptable.

3.4 GOTHIC Application for Component Design Verification The NRC staff's previous acceptance of the GOTHIC containment response calculation methodologies for containment design limits does not explicitly cover the use of GOTHIC results for component design verification. As a result, in Section 2.3 of Topical Report DOM-NAF-3, the licensees included Applications 6-9 for the NRC staff to review and approve regarding the use of GOTHIC output for specific component analyses.

3.4.1 Application 6: Sump Data for Input to Other Analyses GOTHIC modeling assumptions can be biased to produce conservative results with respect to sump water level and liquid temperature. The licensees' requested approval to use these conservative results for validation against component design limits. As discussed in SectionI 3.3, the methodology for performing pump NPSHA calculations produces a higher sump water temperature profile than LOCTIC and is thus more conservative than LOCTIC. The licensees' plan to use this GOTHIC sump water temperature profile for validation against componentI design limits.

Because the licensees are using a sump water temperature profile that is more conservative than the NRC staff-approved LOCTIC code, the NRC staff finds the use of the GOTHIC generated sump temperature and level data for input to other analyses acceptable.

3.4.2 Application 7: Containment Liner Temperature Verification The licensees' proposed methodology for the containment liner temperature verification is a sightly modified version of the peak containment temperature model. A conservativeI containment liner response is obtained by adding a small conductor that has the same construction and properties as the liner conductor. A conductor surface area of 1 ft2 is used to minimize impact on the lumped containment pressure and temperature response. The insideI heat transfer option is the same as used for the actual liner conductor (Direct with DLM) with a multiplier of 1.2 for conservatism.

The DirectIDLM model has been previously accepted by the NRC staff and the 1.2 multiplier is a reasonable enhancement for conservatism; therefore, the NRC staff finds the proposed GOTHIC methodology for the containment temperature verification acceptable.I 3.4.3 Application 8: Equipment Qualification (EQ) Temperature Validation

.The licensees' proposed methodology for EQ temperature validation consists of adding a small conductor for the equipment in the containment response GOTHIC model with the appropriate break scenario and single failure consideration that fits the particular equipment's characterstics. The condensation option for the direct heat transfer package is set to Uchida with a constant multiplier of 4.0, consistent with NUREG-0588. Both the natural and forced convection heat transfer options are activated. The convective heat transfer coefficient is calculated using the blowdown rate and the containment free volume, consistent with NUREG-0588. A characteristic length appropriate for the particular equipment is input.

The proposed methodology is consistent with the NRC staff's guidance in NUREG-0588;I therefore, the NRC staff finds the proposed GOTHIC methodology for EQ temperature validation acceptable.

3.4.4 Application 9: Transient performance of closed cooling loops for heat exchangers associated with the ECCS and containment heat removal systems.

GOTHIC heat exchanger component modeling has been previously reviewed and approved by the NRC staff as part of the GOTHIC methodology for containment response to LOCA and

MSLB events. The proposed methodology for transient performance of closed cooling ioops for heat exchangers associated with the ECCS and containment heat removal systems is an incremental change to the LOCA and MSLB peak containment pressure and temperature analyses; therefore, this is acceptable to the NRC staff.

4.0 CONCLUSION

The NRC staff finds the licensees's GOTHIC computer code methodologies, as documented in Topical Report DOM-NAF-3, acceptable subject to the following conditions: (1) Prior to the implementation of the GOTHIC post-reflood mass and energy methodology contained in this topical report for North Anna 1 and 2, Millstone 2 and 3, and Kewaunee, the licensees shall perform bench marking similar to the one performed for Surry 1 and 2 to ensure conservative values are calculated; (2) The GOTHIC NPSHA methodology contained in this topical report cannot be used for other plants that do not credit containment overpressure to calculate NPSHA in their licensing bases.

The NRC staff concludes that sufficient conservatism has been incorporated in the licensees' methodologies to provide assurance that adequate margins to design values will be maintained to satisfy reguilatory requirements.

Principal Contributor: G. Tesfaye Date: August 30, 2006

Classification/Disclaimer The data, information, analytical techniques, and conclusions in this report have been preparedI solely for use by Dominion (the Company), and they may not be appropriate for use in situations other than those for which they are specifically prepared. The Company, therefore makes no claim or warranty whatsoever, expressed or implied, as to their accuracy, usefulness, or applicability. In particular, THE COMPANY MAKES NO WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, NOR SHALL ANY WARRANTY BE DEEMED TO ARISE FROM COURSE OF DEALING OR USAGE OR TRADE, with respect to this report or any of the data, information, analytical t~echlniques, or conclusions in it. By making thisI report available, the Company does not authorize its use by others, and any such-,use is expressly forbidden except with the prior written approval of the Company. Any such written approval shall itself be deemed to incorporate the disclaimers of liability and disclaimers of warranties provided herein. In no event shall the Company be liable, under any legal theory whatsoever (whether contract, tort, warranty, or strict or absolute. liability), for any property damage, mental or physical injury or death, loss of use of property, or other damage resulting from or arising out of the use, authorized or unauthorized, of this report Abstract As part of a continuing effort to develop and maintain in-house, thermal-hydraulic safety analysis.

capability, Dominion (including 'Virginia Electric and Power Company, Dominion Nuclear Connecticut, Inc., and Dominion Energy Kewaunee, Inc.) has developed a methodology for performing licensing basis analyses for the containme nt response to postulated pipe ruptures inside containment. The methodology employs the GOTHIC computer code and -is applicable for analysis of large, dry containments for pressurized water reactors. GOTHIC is a general-purpose,I thermal-hydraulics computer code developed by the Electric Power Research Institute for applications in the nuclear power industry. The NRC has approved GOTHIC for use in containiment analyses for several U.S. nuclear power plant licensees. The GOTHIC analysis methodology developed by Dominion is described in this topical report and received NRC approval in a Safety Evaluation Report dated August 30, 2006.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 2

Table of Contents Section Page Classification/Disclaimer............................................................................... 2 Abstract.................................................................................................. 2 Table of Contents ...................................................................................... 3 List of Tables ........................................................................................... 5 List of Figures ........................................................................................... 6 List of Acronyms and Abbreviations................................................................. 7 1.0 Introduction....................................................................................... 8 2.0 GOTHIC Overview and Applications ........................................................... 9 2.1 Overview of the GOTHIC Computer Code.................................................. 9 2.2 NRC-Approved GOTHIC Containment Analyses ......................................... 10 2.3 Dominion Licensing Applications of GOTHIC............................................. 11 3.0 GOTHIC Containment Analysis Methodology.............................................. 13 3.1 Containment Noding.......................................................................... 13 3.1.1 Free Volume........................................... .............. 14 3.1.2 ContainmentHeight................................. ...... ............. 14 3.1.3 HydraulicDiameter.................................. ............. 15 3.1.4 Liquid-Vapor Interface Area..........!........................................................... 15 3.2 GOTIHC Model Elements................................................................ I... 16 3.2.1 Junction Parameters............................................................................. 16 3.2.2 Accumulator Nitrogen............................................................................ 17 3.3 Passive Heat Sinks ............................................................................ 18 3.3.1 Heat Sink Geometry and Nodalization........................................................... 18 3.3.2 ConductorSurface Heat Transfer............................................................. 19 3.3.3 ContainmentLiner Thermal Response.......................................................... 20 3.3.4 Equipment Qualification.................................... ..............20 3.4 Containment Spray and Heat Removal ..................................................... 21 3.4.1 Spray Nozzles..................................................................................... 21 3.4.2 Heat Exchangers................................................................................ 25 3.4.3 Containment Air Recirculation Fans.......................................................... 25 3.5 Break Mass and Energy Release............................................................. 27 3.5.1 Treatment of Break Effluent..................................................................... 27 3.5.2 MSLB Mass and Energy Release ........................................ ....28 3.5.3 LOCA Mass and Energy Release................................................................ 28 3.5.3.1 Blowdown, Refill and Reflood Stages ................................................. 28 3.5.3.2 Post-Reflood Stage ..................................................................... 28 3.5.3.3 GOTHIC Long-Term Mass and Energy Release Modeling .......................... 29 3.5.3.4 Qualification of the GOTHIC Mass and Energy Release Models..................... 34 Topical Report DOM-NAF-3, Rev. 0.0-A Page 3

3.6 Containment Initial Conditions............................................................. 35 3.7 Run Control Options.......................................................................... 37 3.7.1 Revaporization Fraction........................................................................ 37 3.7.2 Fog Model ....................................................................................... 37 3.7.3 Maximum Mist Density......................................................................... 38I 3.7.4 Drop Diameterfrom Mist....................................................................... 38 3.9.5 Minimum Heat Transfer Coefficient .......................................................... 38 3.7.6 Reference Pressure.............................................................................. 39I 3.7.7 Force Entrainment Drop Diameter............................................................ 39 3.7.8 Vapor Phase Head Correction................................................................. 39 3.7.9 Kinetic Energy................................................................................... 39I 3.7.10 Phase Options...................................................................... 7'"*'"***.... 40 3.7.11 Force Equilibrium............................................................................... 40 3.7.12 Drop-Liquid Conversion........................................................................ 40I 3.7.13 Version 6.1 Formulations......................................... ............................. 40 3.8 Calculation of NPSH Available .............................................................. 41 3.8.1 Containment OverpressureLicensing Basis..................................................... 41I 3.8.2 GOTHIC Analysis of NPSH Available.......................................................... 42 3.8.3 Water Holdup.................................................................................... 44 3.9 Time Steps..................................................................................... 45 4.0 .GOTHLIC Demonstration Analyses for Surry Power Station ................................. 46 4.1 Surry Power Station Description.............................................................. 46I 4.2 Surry Power Station GOTHIC Model Overview ........................................... 47 4.2.1 Geometry ..............................................................47 4.2.2 Engineered SafeguardsFeatures................................................................ 48I 4.2.3 Mass and Energy Model ......................................................................... 49 4.2.4 Containment Heat Sinks ......................................................................... 49 4.3 GOTHIC Analysis of LOCA Peak Pressure ............................................... 50I 4.3.1 ContainmentResponse ........................................................................... 50 4.3.2 DEHLG Mass and Energy Release.............................................................. 50 4.3.3 Summary.of DEHLG Peak PressureComparison ................................51I 4.4 GOTHIC Analysis of Containment Depressurization...................................... 57 4.4.1 Containment Response........................................................................... 57 4.4.2 DEPSG Mass and Ener~gy Release .............................................................. 57I 4.4.3 Summary of ContainmentDepressurizationComparison...................................... 59 4.5 GOTHIC Analysis of LIISI Pump NPSH Available........................................ 63 4.5.1 Containment Response.........................o.................................................. 63I 4.5.2 DEPSG Mass and Energy Release ... .......................................................... 64 4.5.3 Summary of LHSI Pump NPSHa Comparison................................................. 65 4.6 GOTHIC Analysis of MSLB Event.......................................................... 72I 4.7 Sensitivity Studies............................................................................. 77 4.8 Summary of Demonstration Analyses....................................................... 77 5.0 Conclusions ............................................... ........................................ 81 6.0 References ........................................................................................ 82 Topical Report DOM-NAE-3, Rev. 0.0-A Page 4

List of Tables Table 3.5-1: Comparison of Experiment and Surry Conditions for Cold Leg Mixing 33 Table 3.6- 1: Containment Initial Conditions 35 Table 3.6-2: Containment Initial Conditions for Subatmospheric Plants 35 Table 4.3-1: GOTHIC Comparison to LOCTIC for DEHLG Peak Pressure 52 Table 4.3-2: Primary System Energy Release from 115.8 to 1500 Seconds 52 Table 4.4- 1: Sequence of Events for Containment Depressurization Analysis 59 Table 4.4-2: Comparison of DEPSG Break Mass and Energy 59 Table 4.5- 1: Sequence of Events for LHSI Pump NPSHa Analysis 66 Table 4.5-2: GOTHIC Comparison to LOCTIC for LHSI Pump NPSHa 66 Table 4.7-1: Matrix of Conservative Inputs for Surry GOTHIC Containment Analyses 78 Topical Report DOM-NAF-3, Rev. 0.0-A Pg Page 5

List of Figures Figure 4.3 -1: DEHLG Containment Pressure Comparison to LOCTIC 53 Figure 4.3-2: DEHLG Containment Vapor Temperature Comparison to LOCTIC 53 Figure 4.3-3: DEHLG Containment Liquid Temperature Comparison to LOCTIC 54 Figure 4.3-4: DEHLG Containment Sump Level Comparison to LOCTIC 54 Figure 4.3-5: DEHLG RSHIX Total Heat Rate Comparison to LOCTIC 55 Figure 4.3-6: GOTHIC Conductor Heat Transfer Coefficients Comparison to LOCTIC 55 Figure 4.3-7:

Figre4.3negyRelas 7:DEHG nteraedCopaiso t LOTI DEHLG Integrated Energ Release Compariso to LOCTIC 5

56 Figure 4.3-81: o DEHLGainmntegraedMssur freas ContimetDparedsuizto LCC560 Figure 4.4-1: Containment VporTessraure for Containment Depressurization 60 Figure 4.4-2: Containment Vapord Temperature for Containment Depressurization 60 Figure 4.4-3: CoPG ntanentLqigeprat for Containment Depressurization ureas 61 Figure 4.4-4: DEPSO Integrated Massg for Containment Depressurization 61

,Figure 4.4-6: DEPSG SG Secondary Temperatures for Containment Depressurization 62I Figure 4.5-1: Containment Pressure - LHSI Pump NPSH 6 Figure 4.5-2: Containment Vapor Temperature - LHSI Pump NPSH 67 Figure 4.5-3: Containment Liquid Temperature - LHSI Pump NPSH 68 Figure 4.5-4: Containment Sump Level - LHSI Pump NPSH 68 Figre t.55:Cotanmet ntgra BrakMas Figure 4.5-5: Integral Break Massg to Containment

-LHI PmpNPS

- LHSI Pump NPSH 6

.69 Figure 4.5-7: SG Secondary Side Liquid and Vapor Temperatures - LHSI Pump NPSH 70 Figure 4.5-8: Primary Metal (TA22) and Reactor Vessel Liquid Temperatures (TL15s1) -

.LHSI Pump NPSH 70 Figure 4.5-9: Intact Cold Legs (Volume 22) and Downcomer (Volume 23) Temperatures Pump NPSH 71

-LHSI Figre .6-: MLB OTHC Mdel wthonainentPresur Figure 4.6-1: MSLB Containment VporTessraure with GOTHIC Models 7

73 Figure 4.6-2: MSLB Containment VPorTe mpueusnPratsurewitsh GOTHICtModes 73 Figure 4.6-3: MSLB Containment VporTessraure using Pressure Flash Assumptions 74 Figure 4.6-4: CmprsLB ContainmentVaoTeprteusn o Pressure Flas Assumptionsu 74 Figure 4.6-5: Comparison of Containment VporTessraure for MSLB Spectrum 756 Topical Report DOM-NAF-3, Rev. 0.0-A Page 6

List of Acronyms and Abbreviations Term Definition AFW Auxiliary feedwater CAR Containment air recirculation fans CDT Containment depressurization time (time to reach subatmospheric pressure)

CLS Consequence limiting safeguards CS Containment spray CVTR Carolinas Virginia Tube Reactor DEHLG Double ended hot leg guillotine DEPSG Double ended pump suction guillotine DLM Diffusion layer model ECCS Emergency core cooling system EPRI Electric Power Research Institute EQ Equipment qualification ESF Engineered safeguards features HDR Heissdampfreaktor HHSI High head safety injection IRS Inside recirculation spray LHSI Low head safety injection LOCA Loss of coolant accident MSLB Main steam line break accident NPSH Net positive suction head NPSHa NPSH available NRC Nuclear Regulatory Commission NSSS Nuclear steam supply system ORS Outside recirculation spray RCS Reactor coolant system RMT Recirculation mode transfer RS Recirculation spray RSI-X Recirculation spray heat exchanger RWST Refueling water storage tank SG Steam generator SI Safety injection SPP? Subatmospheric peak pressure SW Service water SWVEC Stone & Webster TS Technical Specifications UFSAR Updated Final Safety Analysis Report Topical Report DOM-NAF-3, Rev. 0.0-A Pg Page 7

1.0 Introduction This topical report documents a methodology for performing containment analysis licensing calculations using the GOTHIC (Generation of Thennal-Hydraulic Information for Containments) computer code. GOTHIC is a general-purpose thermal-hydraulics code for containment analysis developed for the Electric Power Research Institute (EPRI) by Numerical Applications, Inc. (NMI).

GOTHIC has been approved by the NRC- for containment analysis applications at several U.S.

licensees (see Section 2.2). Dominion plans to use the methodology in this report to replace the evaluation methods in the updated final safety analysis reports (UFSARs) for Surry, North Anna, and Millstone Units 2 and 3 for the containment design requirements listed in Section 2.3.

The topical report is broken into five major sections:

Li Section 2 provides general information on the GOTHIC code, NRC licensing information, and the planned applications of GOTHIC for containment analyses at Dominion.

Li Section 3 documents the GOTHIC analytical methodology applicable to large, dry PWR containrnents. References are made to NRC-approved applications for the containment response models (e.g., Direct/DLM condensation). A post-reflood mass and energy release methodology that couples the reactor coolant system, steam generator secondary side, and the containment is also presented.

" Section 4 documents Surry demonstration analyses of the methodology in Section 3, withI comparisons to the Stone & Webster (SWEC) LOCTIC code, for LOCA peak pressure, containment depressurization, and NPSH available for the LHSI pumps. MSLB demonstration cases are also included. The results of Surry sensitivity studies on key parameters are included as a matrix to demonstrate that appropriate, conservative assumptions are specified in the analysis models.

o Section 5 documents the topical report conclusions.I

" Section 6 *includesthe reference list.

Topical Report DOM-NAE-3, Rev. 0.0-A Page 8

2.0 GOTHIC Overview and Applications 2.1 Overview of the GOTLUC Computer Code GOTHIC (Generation of Thermal-Hydraulic Informnation. for Containments) is an integrated, general-purpose thermal-hydraulics code for performing licensing containment analyses for nuclear power plants. The code has been developed for the Electric Power Research Institute (EPRI) by Numerical Applications, Inc. References 1-3 document the bases for GOTHIC Version 7.2, the most recent EPRI-released code version. The following code description is obtained from the GOTHIC Technical Manual [1].

GOTHIC solves the conservation equations for mass, momentum and energy for multicomponent, multi-phase flow in lumped parameter and/or multi-dimensional geometries. The phase balance equations are coupled by mechanistic models for interface mass, energy and momentum transfer that cover the entire flow regime from bubbly flow to film/drop flow, as well as single phase flows. The interface models allow for the possibility of thermal non equilibrium between phases and unequal phase velocities, including countercurrent flow. GOTHIC includes full treatment of the momentum transport terms in multidimensional models, with optional models for turbulent shear and turbulent mass and energy diffusion. Other phenomena include models for commonly available safety equipment, heat transfer to structures, hydrogen burn and isotope transport.

A complete description of the qualification of the GOTHIC code for use in containment analysis is provided in the code qualification report [3]. The reader is referred to this document for a discussion of the degree and type of code qualification performed. This topical report refers to Reference 3 to support the selection of specific models in Domuiniion analyses.

Dominion has participated in the EPRI GOTHI-C Advisory Group since the late 1980s to ensure a solid understanding of the code capabilities and limitations, to monitor industry applications, and to guide the code qualification effort. For the first licensing applications, Dominion plans to use GOTHIC version 7.2domn, which consists of the EPRI-released version 7.2 and two enhancements specific to.Dominion that were implemented during testing of the GOTHIC containment model for Surry Power Station.

The two code changes are summarized.

1) The first change improves the iterative solution for the heat exchangers so that a non-convergent condition is avoided. This condition was discovered during development testing of Surry containment models. in GOTHIC version 7.2.

2) The second change adds a user-specified multiplier on the Film heat transfer option to allow sensitivity studies on the mass and energy release model. This feature allows the core conductor to be initialized without using internal heat generation.

Topical Report DOM-NAF-3, Rev. 0.0-A Pg Page 9

Both changes were added to version 7.2 under the Dominion software quality assurance (QA) prograrn, which is part of the overall Dominion QA program under 10 CER 50 Appendix B. Dominion receives GOTHIC code error reports from EPRI and evaluates each error under the software QAI program. Dominion may upgrade to new versions of GOTHIC or install patches to correct code errors as they are made available. The methodology is not restricted to a specific version of the code.

Dominion develops and maintains in-house analytical methodologies in accordance with NRC Generic Letter 83-11, Supplement 1 [4]. Dominion informed the NRC of its formal Generic Letter 83-11 program in Reference 5. The Dominion procedure for controlling safety analysis computer codes and models was transmitted to the NRC in Dominion letter 02-280 in response to NRC Request for Additional Inform-ation item 4d regarding topical report VEP-FRD-42, Revision 2 [6]. Section 3.1 of the staff SER for VEP-FRD-42, Revision 2, found the Dominion approach to he an acceptable evaluation process [7]. Future upgrades to new versions of GOTHIC or installation of patches to correct errors will be evaluated under the same NRC-accepted program.

2.2 NRC-Approved GOTIUC Containment Analyses GOTHIC has been approved by the NRC for containment analysis applications at several U.S.

licensees. Recent NRC approvals are documented in References 8-13. The list is not meant to be exhaustive but merely to demonstrate the recent acceptability of NRC to a range of containment licensing applications with GOTHIC. For containment modeling, Dominion has selected correlations that have been previously approved by the NRC and has confirmed the applicability of the models toI large, dry PWVR containments. For calculation of post-reflood mass and energy release, a simplified GOTHIC model of the reactor coolant system (RCS) and steam generator secondary side has been developed and coupled to the containment. Section 3.5 describes the methodology for modeling the mass and energy release. Section .4 describes model qualification analyses that were performed for Surry pump suction and hot leg breaks with comparisons to NRC-approved Westinghouse methodologies for post-reflood mass and energy releases [14, 16]. Framiatome recently received NRC approval for use of a coupled mass and energy release model [30, 31].

Topical Report DOM-NAE-3, Rev. 0.0-A Page 10

2.3 Dominion Licensing Applications of GOTHIC This report documents the analytical methodology for performing containment analysis for Dominion's large, dry PWR containments with the GOTHIC computer code. Dominion plans to use GOTHIC to perform UESAR calculations for the following containment design requirements (some criteria are specific to subatmospheric containments):

1. LOCA containment peak pressure and temperature,

2. MSLB containment peak pressure and temperature,

3. LOCA containment depressurization time (CDT) for Surry and North Anna,

4. LOCA containment subatmospheric peak pressure (SPP) for Surry and North Anna,

5. Available net positive suction head (NPSHa) for pumps that take suction from the containment sump. For Surry and North Anna, a time-dependent NPSHa is calculated from a transient containment response for the inside recirculation spray (1]RS), outside recirculation spray (ORS), and low head safety injection (LHSI) pumps,

6. Minimum and maximumn sump water level and liquid temperature for input to other analyses (e.g., strainer debris head loss and component stress analyses),

7. Containment liner temperature verification,

8. Equipment qualification (EQ) temperature validation, and

9. Transient performance of closed cooling loops for heat exchangers associated with the ECCS and containment heat removal systems.

The LOCA peak pressure calculation is the simplest to perform because the maximum pressure occurs early in the transient (about 20 seconds for blowdown peaks), before the spray systems activate. The peak pressure is dependent on the containment volume, heat sink characteristics, the break energy, and how the break fluid is modeled. MSLB peak pressure calculations run longer because of the continuing release of high energy steam until the auxiliary feedwater (AFW) flow to the faulted steam generator is isolated.

Calculation of CDT and SPP is performed to demonstrate that the containment pressure is bounded by the assumption for contaimnment leakage in the dose consequences analyses. Currently, the North Anna and Surry licensing bases requires the containment to be subatmospheric in one hour and remain subatmospheric thereafter [27, 28]. The CDT and SPP analyses both assume the single failure of one emergency bus, but other assumptions for safety injection flow rates and containment initial conditions are different in order to produce the most conservative effect (e.g., minimum containment initial temperature is conservative for SPP because of the larger air mass that challenges the long-term heat.

removal of one train of recirculation spray).

Topical Report DOM-NAF-3, Rev. 0.0-APae1 Page I I

'For long-term analyses (e.g., the CDT, SPP, EQ and MSLB), the spray systems and other heat removal components are incorporated, and the depletion of the refueling water storage tank (RWST) liquid inventory is modeled in order to predict the time of recirculation mode transfer (RMT), when the safety injection system swaps suction from the RWST to the containment sump, and the time ofI containment spray (CS) pump termination.

The NPSHa calculations for Surry and North Anna present a challenge in that the appropriate pump suction conditions must be determined based on mixing of cold water that is injected to the PUMP suction with hotter sump water, the incorporation of suction friction and form losses, and the explicit inclusion of containment overpressure. In addition, conservative modeling of spray systems, condensation heat transfer, and other features is applied different from the CDT and SPP cases (seeI Section 3.8).

Topial ev.0.0-epor DO-NA-3, Pae 1

3.0 GOTHIC Containment Analysis Methodology This section provides the Dominions methodology for constructing GOTHIC models for performiing licensing basis analysis for large, dry containments. The methods are intended to provide realistic but conservative results. Justification is based on previously accepted PWVR containment methodologies and the extensive validation base for GOTHIC. Many of the input parameters required to construct GOT HIC containment model carry some uncertainty. The following sections provide the methods that have been adopted by Domi~nion to obtain conservative results for a given analysis objective. Some model components and parameters are not specifically listed, either because they have no impact on the analysis or the exact physical behavior or values are expected to be readily available.

3.1 Containment Noding Plant licensing analyses use a single volume (node) for the containment building with separate treatment given to the sump and containment atmosphere regions. Inherent 'in this lumped parameter approach is the assumption that within each region the fluid is well mixed. During a LOCA or MSLB, the mixing 'induced by the break jet is significant. Later in the transient, containment sprays and/or containment fan coolers continue to promote mixing in the containment. The degree to which well-midxed conditions are attained depends on the location and size of the break, major obstructions in the containment, spray flow rate. and pattern, and the location and ducting of fan coolers.

GOTHIC has the capability to model the containment in more detail and calculate the three-dimensional distribution of mass and energy within the containment. Three-dimensional GOTHIC models are referred to as subdivided analyses. To assess the impact of subdivided versus lumped parameter modeling, the CVTR (Carolinas Virginia Tube Reactor) tests were simulated with both types of models [3]. The CVTR tests were typical of a MSLB located high in the containment except that the steam was introduced through a diffuser that reduced the jet momentum and mixing. Results, from the subdivided simulations indicate near well-mixed conditions 'in the upper containment above the operating deck but significantly lower and varied temperatures and steam concentration in the region below the operating deck. The degree of mixing was similar during the steam injection and while the containment sprays were active. In the CVTR containment, the operating deck is a major obstruction between the upper and lower containment and certainly contributed to the nonunifonnity of the atmosphere. Experimental results for LOCA type conditions in the Marviken and Heissdampfreaktor (HDR) containments also indicate significant variation in conditions in the containment. While these test containments are more compartmentalized than a typical large dry containment, they 'indicate that some degree of non-uniformity is possible.

Topical Report DOM-NAF-3, Rev. 0.0-APae1 Page 13

Results from lumped and subdivided GOTHIC models for the CVTR tests *indicate that the predicted peak pressure and temperature from the lumped analysis are larger than in the subdivided analysis. Since the major energy removal mechanism during a blowdown is heat transfer to the containment structures due to convection and condensation, one might think thatI the maldistribution of steam and high temperature conditions would lead to less heat removal because less conductor surface area is exposed to the high energy conditions. H owever, the condensation rate is a strong function of the steam concentration, and the increased condensation rate mn the regions of high steam concentration more than compensate for the smaller effective heat transfer area. Were this not the case, it would be necessary to use subdivided models that consider local effects such as break location and orientation or to add extra conservatism to the lumped model to account for these effects.

The foregoing justification for a single-volume approach to predict peak containment pressure and temperature applies to LOCA and MSLB. conditions. For these accident scenarios, the high energy region in the containment is large even though the entire containment might not be fully mixed and the concrete structures are still absorbing heat when the short duration blowdown isI over. For long-term analyses, the spray systems are activated, the open regions of the containment are expected to be well-mixed [18], and the single-volume lumped model should be representative of the actual conditions.

3.1.1 Free VolumeI The containment free volume is the space occupied by the containment atmosphere. It can be difficult to calculate the free volume exactly because of the complex shapes of all the large and small equipment and structures inside the containment. For a given mass and energy release, a smaller free volume will typically give higher peak pressure and temperature. For containment pressure and temperatureI analysis, a low estimate is used for the containment free volume. For NPSHa calculations, an upper bounding value is specified to minimize the containment pressure.

3.1.2 Containment Height The containment height, H, is used for two purposes:

1. The nominal floor area is calculated as Af =H whrVistepcfedrevou .Te floor area is used in the calculation of the drop deposition rate due to gravitational settling.

2. The height is used to calculate the conductor film thickness in the DLM condensation options.

Refer to Section 3.3.2 for the assumptions regarding the DLM characteristic height.

The containment height is calculated using the guidance in Section 3.4.1.2 to ensure that the spray height properly accounts for the spray heat and mass transfer in the covered region.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 14

3.1.3 Hydraulic Diameter GOTHIC uses the hydraulic diameter, Dh, to calculate the surface area of thermal conductors in contact with the atmosphere using A,=-4 Equation 1

.The user-specified hydraulic diameter is calculated ftom this formula using the containment free volume and the total surface area of the conductors in contact with the vapor.

3.1.4 Liquid-Vapor Interface Area The liquid-vapor interface area is used to calculate the heat and mass transfer between the vapor and the liquid phase. It can be set to zero to prevent any heat and mass transfer at the interface or to a very large value to force thermal equilibrium between the vapor and liquid phases. The default value is the maximum of Af and A~, where A,, is the wettable area calculated from 4Vless any conductor Dh surface area that is too hot to allow a liquid film and Af is the nominal floor area defined in Section 3.1.2.

This gives a large area for interfacial heat and mass transfer under the assumption that during a LOCA or MSLB nearly all of the surface area will be wet due to condensation or deposited water from the break. The default value has been used for all of the GOTHIC validation against experimental data for simulated line breaks in containments [3]. The GOTHIC default value will be used for the containment lumped volume for containment integrity analyses. For NPSHa, a minimum sump pool surface area is used to minimize the evaporative heat and mass transfer with the net effect of leaving more energy in the sump liquid as the containment depressurizes (and the vapor temperature is less than the liquid temperature). The liquid/vapor interface area inputs for the simplified RCS model are described in Section 3.5.3.3.2.

Topical Report DOM-NAF-3, Rev. 0.0-APae1 Page 15

3.2 GOTHIC Model Elements.

As described in Section 3.1, a single lumped volume models the containment building with separate treatment given to the sump and containment atmosphere regions. Containment passive heat sinks are included in the lumped volume. The RWST is modeled with a volume to provide an accurate prediction of inventory drawdown for determining spray flow rate as a function of level,3 the time of safety injection recirculation mode transfer (RMT), and the time of containment spray termination. An atmosphere boundary condition is used to maintain the RWST pressure as the tank drains.

Other volumes serve as junction connectors to model piping for the safety inje ction and spray systems. The volumes allow accurate modeling of cold injection flow to pump suctions.

Additional details are included in Section 4 for the Surry demonstration analyses. Where appropriate, valves may be used to isolate components/volumes when not in use. Plant-specific models may be different because of design features that require a different treatment (e.g., pump start ramp times, pump heat, heat exchanger performance, and piping fill delays). These modelI details are not considered part of the analysis methodology 3.2.1 Junction Parameters For a single volume containment model with most of the flows specified, most of the junction parameters are not influential. The few influential parameters are discussed.

oi For junctions taking suction from the containment sump, the junction end elevation and end height are set so that the Junction end is fully submerged.

Li For a junction that models a suction line, the junctio n area, friction length and loss coefficients must be accurate and consistent so that the pressure drop from the sump to the pump will beI accurately predicted for NPSHa analysis.

o For the junctions used to connect the various volume components of the RCS for the long-term mass and energy release, the areas and loss factors need to be consistent and reasonably accurate so that the model will correctly predict the flow through the SG loops and the flow split at the two ends of the break.

The volume average velocity in the containment is determined by the junction flows and the junction parameters. The* volume average velocity is used in the calculation of heat and mass transfer coefficients and in the drop deposition models. Forced convection heat transfer is not credited so the only potential influence is the drop deposition. GOTHIC includes drop depositionI Topical Report DOM-NAF-3, Rev. 0.0-A Page 16

due to impaction. Impaction deposition increases with increasing velocity and increasing drop size. Impaction deposition will be significant only during the blowdown. This deposition will reduce the drop mass in the atmosphere and may cause a small increase 'in peak temperature and pressure.

The lumped volume velocity is calculated as Y LjujAj junctions u=attached to V -Euto where L is the junction inertia length, u is the junction velocity and A is the junction area. The intention of this formula is that the flow through junction area A expands to an area of VIL. To maximize the impaction deposition and maintain reasonable volume average velocities, the inertia length of the break junctions is set to the containment height and the junction area is set to the assumed break area. The lengths and areas of other junctions will have negligible effect on the impact deposition as long as physically reasonable values are used.

3.2.2 Accumulator Nitrogen The NSSS or fuel vendor LOCA mass and energy release data include the water injected from the ECCS accumulators. The accumulator nitrogen is a contributor to the total containmnent pressure and therefore can affect containment depressurization time and NPSHa. A boundary condition injects the nitrogen volume into the containment atmosphere consistent with the timing in the vendor mass and energy release calculation. GOTI-IC inputs for nitrogen pressu re, temperature, and volume are based on allowable operating ranges in the plant Technical Specifications with consideration of uncertainty.

Section 4.7 documents the conservative direction for these parameters from the Surry sensitivity studies for containment depressurization and NPSHa. The accumulators do not contribute nitrogen to the containment during a MSLB.

Topical Report DOM-NAF-3, Rev. 0.0-APae1 Page 17

3.3 Passive Heat Sinks 3.3.1 Heat Sink Geometry and Nodalization Thermal conductors are the primary heat sink for the blowdown energy. The conductors can be.

made of up any number of layers of different materials. One-dimensional conduction solutions are used to be consistent with the lumped modeling approach.

The thermal conductor is divided into regions, one for each material layer., with an appropriate thickness and material property for each region. GOTHIC accepts inputs for material density, thermal conductivity and specific heat. These values are obtained from published literature for the materials present in each conductor. Conductors with high heat flux at the surface and low thermal conductivity must have closely spaced nodes near the surface to adequately track the steep temperature profile. The node spacing is set so the node Biot number for each node is lessI than 0. 1. The B iot number is the ratio of external to internal conductance.

It is not practical or necessary to model each individual piece of equipment or structure in the containment with a separate conductor. Smaller conductors of similar material composition can be combined into a single effective conducto r. In this combination, the total mass and the totalI exposed surface area of the conductors is preserved. The thickness controls the response time for the conductors and is of secondary importance. The conductors are grouped by thickness and material type. The effective thickness for a group of wall conductors is calculated by Equation 3.

The heat sink material types, surface areas, and thickness are derived based on plant-specific inventories. Concrete, carbon steel, and stainless steel are the most common materials.

teff - EgopEquation 3 iEgroup

.If there is a small air gap or a contact resistance between the containment liner and the concrete, it is modeled as a separate material layer at the nom-inal gap thickness with applicable material properties. This overestimates the contact resistance because convection and radiation effects will be ignored. A maximum gap Iconductance of 100 Btulhr-ft2 -F is used, consistent with other recent containment analysis applications [20, 21]. The gap width is determined by dividing the gap thermal conductivity by the gap conductance.

A ll containment passive heat sinks are included in, the lumped containment volume. The primary system metal and SG secondary shells are included in the simplified RCS model that is used for the calculation of long-term mass and energy release (see Section 3.5); however, these conductors, are not used for condensation or convection heat transfer with the containment atmosphere.

Topical Report D OM-NAE-3, Rev. 0.0-A Page 18

3.3.2 Conductor Surface Heat Transfer The Direct heat transfer option with the DLM (Diffusion Layer Model) condensation option is used for all containment passive heat sinks except the sump floor. With the Direct option, all condensate goes directly to the liquid pooi at the bottom of the volume. The effects of the condensate film on the heat and mass transfer are incorporated in the formulation of the DLM option. Under the DLM option, the condensation rate is calculated using a heat and mass transfer analogy to account for the presence of nonconidensing gases. It has been validated against seven test sets [3]. It also compares well with Nusselt's theory for the condensation of pure steam where the rate is controlled by the heat transfer through the condensate film. As shown in the GOTHIC Qualification Report [3], the DLM option generally underpredicts the condensation rate and has previously been accepted by the NRC for LOCA and MSLB containment analyses [8, 9].

The opt ions for natural convection heat transfer for sensible heat transfer and radiant heat to steam are activated as allowed by N1JREG-0588 [22]. A natural convection option is selected consistent with the conductor geometry and orientation. Although the DirectIDELM validation basis includes tests with forced convection heat and mass transfer, forced convection has not been accepted for peak temperature and pressure analysis and is not used.

A characteristic height can be specified for each heat transfer option to estimate the film thickness that builds up on the conductor. For typical large dry containment conditions, the heat and mass transfer is controlled by the boundary layer in the vapor phase and the resistance through the film is relatively small so the specified height is of secondary or less importance. When using the DLM option, the characteristic height is set to the containment volume height. This gives thick liquid films that will slightly reduce the heat and mass transfer rates once the film is fully established.

This is conservative for containment pressure and temperature analysis. For NPSHa analysis, the heat transfer coefficient is multiplied by 1.2 for conservatism (see Section 3.8.2).

For a conductor representing the containment floor or sump walls that will eventually be covered with water from the break and condensate, the Split heat transfer option is used to switch the heat transfer from the vapor phase to the liquid phase as the liquid level in the containment builds. A quicker transition to liquid heat transfer is more conservative for containment analysis. The Split option is used with utjlmax,, the maximum liquid fraction, set to

=l d Equation 4 nn H where d is the transition water depth and H is the volume height. A reasonable value for d of 0. 1 inch switches the heat transfer from the vapor phase to the liquid phase as the liquid level in the containment reaches 0. 1 inch. Other values may be appropriate depending on the geometry of the floor and sump.

Topical Report DOM-NAF-3, Rev. 0.0-APae1 Page 19

For conductors with both sides exposed to the containment, the Direct option is applied to both sides. Alternatively, if the conductor is symmetric about the centerplane, a half-thickness conductor can be used with the total surface area of the two sides and an insulated back side heat transfer option. The conductor face that is not exposed to the atmosphere is assumed insulated.I The Specified Heat Flux option is used with the nominal heat flux set to zero.

Containment walls above grade and the containment dome have a specified external temperature boundary condition with a heat transfer coefficient of 2.0 Btulhr-ft2 -F to model convective heat transfer to the outside atmosphere. The GOTHIC heat transfer solution scheme allows for accurate initialization of the temperature distribution in the containment wall and dome prior to the transient initiation. This heat transfer coefficient is used 'inthe current LOCTIC licensing basis for North Anna [27] and Surry [28] and remains appropriate for the containment interface with the outside air. Framatome also used this value in Section 6.1.1 of Reference 30.

3.3.3 Containment Liner Thermal Response The containment liner temperature is verified to be less than the design limit by repeating the peak temperature analyses with one modification. A conservative containment liner response is obtained by adding a small conductor that has the same construction and properties as the liner conductor. A conductor surface area of 1 ft2 .is used to minimize impact on the lumped containment pressure and temperature response. The inside heat transfer option is the same as used for the actual liner conductor (Direct with DLM) with a multiplier of 1.2 for conservatism.

3.3.4 Equipment QualiflicationI GOTHIC can be used for verification of equipment qualification (EQ). Since both the maximum temperature and the time that the equipment is exposed to high temperature need to be considered, the particular break scenario and single failuLre for EQ may be different from that for the containment peak pressure analysis and will depend on the characteristics of the equipment.

The temperature response of the limiting equipment can be modeled by adding a small conductor for the equipment. The condensation option for the Direct heat transfer package is set to Uchida with a constant multiplier of 4.0 consistent with NUREG-0588. [22]. Both the natural and forced convection heat transfer options are activated. The characteristic velocity U (fi/sec) for calculating the heat transfer coefficient is specified using control variables as U =25 MED Equation 5 VI where MED is the blowdown rate *inlbm/Whr and V is the containment free volume consistent with NUREG-0588 [22]. A characteristic length appropriate for the particular equipment is input.I Topical Report DOM-NAF-3, Rev. 0.0-A Page 20

3.4 Containment Spray and Heat Removal Dominion nuclear stations include a range of designs for containment spray systems and long-term containment heat removal. This section covers the general modeling practices for spray nozzles, spray pumps, spray system delivery times including piping fill time and pump start delays, containment air recirculation (CAR) fans, and heat exchangers that are used for containment heat removal. The representative demonstration analyses for Surry in Section 4 exercises all of the models except the CAR fans, which Surry does not have. Each plant-specific application will ensure appropriate, conservative modeling for all applicable heat removal components.

3.4.1 Spray Nozzles GOTHIC includes models that calculate the sensible heat transfer between the drops and the vapor and the evaporation or condensation at the drop surface. The efficiency-the actual temperature rise over the difference between the vapor temperature and the drop inlet temperature-cannot be directly specified in GOTHIC. The efficiency is primarily a function of the drop diameter. The GOTHIC models account for the effect of the diameter through the Reynolds number dependent fall velocity and heat. transfer coefficients. A heat and mass transfer analogy is used to calculate the effective mass transfer coefficient, which is used to calculate the evaporation or condensation.

The method for modeling sprays is to inject the drops into the containment via a junction using a nozzle component. The drop size and the fraction of the water flow to convert to drops to account for the height of the spray header are input by the user. The determination of conservative inputs is described in the following sections.

3A4.1.1 Spray Diameter Spray nozzles typically deliver a spectrum of drop sizes. Smaller drops fall more slowly and reach equilibrium with the vapor more quickly than larger drops because of the larger surface area to mass ratio. GOTHIC does not directly model the drop size distribution. It is assumed that the specified diameter is the Sauter mean diameter. The Sauter mean diameter is calculated from its definition using Equation 6.

d2= ' ~~~x Equation 6 J f(X)X 2dX wheref is the frequency of drops of a particular size.

Topical Report DOM-NAF-3, Rev. 0.0-APae2 Page 21

A given mass of drops at the Sauter mean diameter has the same surface to mass ratio as the actual drop spectrum. The consistency of the surface to mass ratio ensures that the heat transfer rate to heat capacity ratio is correct.

A given mass of drops at the Sauter mean diameter also has the same total projected area to mass ratio as the actual drop distribution. Since the deposition rate is given by a balance of the body force and the drag force on the projected area, the fall velocity and deposition rate of the Sauter mean drops are representative of the full drop spectrum. GOTHIC accounts for the growth or shrinkage of drops due to condensation or evaporation.

The drop fall velocity is a function of the drop drag coefficient. The coefficients used in GOTHICI are those recommended by Ishii [23] and include the, effects of a large population of drops falling together.

The drop heat and mass transfer models have been validated using data from Spillman [24]. The GOTHIC predicted evaporation rate is in the middle of the range of evaporation rates fromI experimental data and rates from correlations. Since evaporation and condensation are controlled by the same mechanism (i.e., turbulent diffusion through the boundary layer), it is reasonable to expect that GOTHIC also fairly represents the condensation rate.

3.4.1.2 Spray Height The lumped parameter approach assumes that conditions are uniform throughout the volume. .

When sprays are injected into a volume, the drops are assumed to be uniformly distributed throughout the volume regardless of the specified elevation of the junction that carries the spray flow. However, in the actual containment there are typically some regions that are not directly covered by the sprays. The containment geometry parameters must be set to properly account for the spray heat and mass transfer in the covered region.

The heat and mass transfer at the spray droplet surface is determined by the drop and atmosphereI temperatures, the steam content of the atmosphere, the drop surface area and the heat and mass transfer coefficients. The heat and mass transfer coefficients depend on the fluid properties at the given temperatures, the drop diameter and pressure and the fall velocity of the spray droplets.

Appropriate heat and mass transfer coefficients will be applied if the drop diameter is consistent with the actual spray drop size and if the fall velocity is correct. Spray drops typically reach their terminal velocity within a few feet of the nozzle and the fall velocity is assumed equal to the terminal velocity for lumped modeling in GOTHIC. The terminal velocity depends on the drop diameter and the atmosphere properties. GOTHIC will calculate appropriate heat and mass transfer coefficients if the spray drop diameter is set to the Sauter diameter in Section 3.4.1. 1.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 22

From the definition of the Sauter mean drop diameter, the total drop surface area exposed to the atmosphere will be correct if the total drop volume suspended in the atmosphere is correct. The total drop volume in the modeled containment volume is Vd =Vd Equation 7 where V is the specified containment volume and ad is the drop volume fraction in the volume. In the actual containment, the suspended drop volume is Vd =~Vad Equation 8 where V, is the sprayed volume in the containment and crd is the drop volume fraction in the sprayed volume.

Since we want the modeled drop volume to be the same as the actual drop volume in the containment, combining the above two equations gives ad -AVS Equation 9 Neglecting the relatively small amount of condensation on the drops, under steady conditions the drop deposition rate equals the spray injection rate. In the containment, the drop deposition rate is

,Y =AadU-Pd =in, Equation 10 where A' is the floor area where the drops are deposited, U~is the terminal velocity, pd is the density of the water in the drops and m, is the spray rate.

In GOTHIC, the deposition rate is calculated from Y=AfadU-.Pd =M, Equation 11 From the three equations immediately above, the relationship for the floor area is derived in Equation 12. This floor area will give the correct drop volume and surface area exposed to the containment atmosphere.

Topical Report DOM-NAF-3, Rev. 0.0-APae2 Page 23

f adf saf

=-Ac =-VAc

.f Equation 12I Since, by assumption in GOTHIC, A =L Equationl13

'H where H is the specified height for the containment volume, the height of the containment volume should be set to H = V'Equation 14 f

Setting the containiment volume height as recommended above has some side consequences that must be considered:

1. It will increase the pool surface area for heat and mass transfer. However, since the effective area of heat and mass transfer is the maximum of the pool area and the surface area definied by the hydraulic diameter (4V/Dh), as long as 4V/Dh > Af, there is noI effect on peak pressure and temperature analyses.

2. For NPSH analysis, the water depth in the contairnment will have to be adjusted to account for the artificially increased pool area, A'.. Sensitivity studies have shown that NPSHa is not sensitive to a reduction in containment height, because the spray modeling assumptions applied in Section 3.8.2 ensure a conservative spray response that minimizes the containment pressure for NPSH analysis.

The spray volume, V,, is set to the total volume below the spray headers under the assumption that the region interior to the headers is adequately covered by the spray. The deposition area, A'f is set to the total horizontal area at the bottom of the sprayed regions where the sprays are expected to collect. For all calculations, the nozzle spray flow fraction is set to 1.0.

Topicai.Report DOM-NAF-3, Rev. 0.0-A Page 24

3.4.1.3 Spray Coverage The spray header arrangement may result in less than 100% coverage of the containment area below the nozzles based on the nozzle spray cone geometries. However, the sprays induce substantial mixing 'inthe containment [18]. Further, the sprays typically achieve 100% efficiency within a short distance from the nozzle [25]. The 100% spray efficiency assumption was approved in the Kewaunee licensing application of GOTHIC [8]. Therefore, unless the sprays are arranged so that isolated sections of the containment are not covered, the conservatism included by modeling the sprayed volume (Section 3.4.1.2) is sufficient to assure overall conservatism of the spray effectiveness.

3.4.2 Heat Exchangers Heat exchangers that remove energy from the containment sump are modeled with the available heat exchanger options in GOTHIC. Use of a GOTHIC heat exchanger option dynamically couples the heat exchanger performance to the predicted primary and secondary fluid conditions.

This can provide a small benefit compared to other codes (e.g., LOCTIC) that use bounding UA values to cover the fluid conditions predicted over the entire transient.

The GOTHIC heat exchanger type that closely matches the actual heat exchanger is selected. The inside and outside heat transfer areas are calculated from the heat exchanger geometry details. For tube and shell arrangements, the shell side flow area is set to the open area across the tubes at the midplane of the heat exchanger and the shell side hydraulic diameter is set to the tube outer diameter as recommended in Reference 17. The .GOTHIC option for built-in heat transfer coefficients is used to determine heat transfer coefficients that depend on the primary and secondary side Reynolds and Prandtl numbers. The heat exchanger models in GOTHIC are for basic heat exchanger designs and may not account for the details of a particular heat exchanger (e.g., baffling in a tube-and-shell heat exchanger). A forcing fuinction can be used on the primary and secondary side heat transfer coefficients to tune the heat exchanger performance to manufacturer or measured specifications. Alternatively, the heat transfer area can be adjusted to match the specified performance. Fouling factors and tube plugging are applied when conservative..

3.4.3 Containment Air Recirculation Fans Containment air coolers are modeled using a GOTHIC FAN COOLER type heat exchanger. This heat exchanger model has been validated against experimental data for LOCA conditions [3]. The fan cooler heat exchanger model calculates the condensation on the tubes and fins in the presence of noncondensing gases. A heat and mass transfer analogy is used to estimate the mass transfer coefficient from standard heat transfer coefficients for heat exchangers. The GOTHIC built-in heat transfer coefficients are used for both the primary and secondary side.

Topical Report DOM-NAF-3, Rev. 0.0-APae2 Page 25

The air/steam flow through the fan cooler can be specified with a volumetric fan. The mass flow rate through the fan is based on the specified volumetric flow rate and the vapor density in the upstream volume. If the actual cooler has the fan positioned upstream of the coils, the volumetricI fan can be placed on the same junction as the heat exchanger to get the correct mass flow through the cooler. If the fan is positioned downstream of the coils, then an additional volume must be added to the model between the cooler and the return to the containment. The conditions in this volume will be representative of the cooler outflow and the volumetric fan should be placed on the junction for air return to the containment.

Fan coolers vary widely in the arrangement of the cooling coils and the water flow circuits through the coils. The GOTHIC fan cooler model has the flexibility to reasonably approximate any coil and flow configuration. However, there are minor variations in design that result in slight differences in manufacturers stated performance and GOTHIC results. The performance of the GOTHIC model for the fan cooler is tuned to match the manufacturer specification or test data.

There are two forcing functions that can be used to tune the performance: a multiplier on the primary and secondary side heat transfer coefficients and a multiplier on the film thickness. The film thickness controls the resistance through the liquid film that builds on the fins and tubes. A multiplier on the film thickness will have an impact on cooler performance under design basis accident conditions but no significant affect at normal operating conditions. The cooler is first tuned by adjusting the multiplier on the primary and secondary side heat transfer coefficients to match the manufacturer specification or data for normal operating conditions. The film thickness multiplier can then be adjusted to match specified performance at the accident conditions.

Topial ev.0.0-epor DO-NA-3, Pae 2

3.5 Break Mass and Energy Release 3.5.1 Treatment of Break Effluent Flow boundary conditions are used to model the break release paths. The boundary condition pressure, together with the break enthalpy, determines the phase split of the break flow and the phase densities and, therefore, the junction velocity. A lower pressure will result in more steam injection (assuming two-phase conditions) at a higrher velocity. Since any inijected water will quickly flash to the saturation temperature at the containment pressure and the break pressure cannot be lower than the containment pressure, the boundary condition pressure is set to the containment transient pressure by capturing the containment pressure in a control variable and then assigning the control variable as a forcing function on a nominal pressure of 1.0.

When superheated water is released from the reactor vessel or steam generator, it flashes until the temperature drops to the saturation temperature at the containment total pressure. This flashing causes rapid acceleration of the liquid and breaks the flow up into small drops. Based on experimental data for superheated water discharges, a drop size of 100 microns (0.003937 inch) is used [19]. This value is on the high side of the experimental data and has been approved by the NRC for LOCA and MSLB applications at Kewaunee [8, 9] and Fort Calhoun [10]. For LOCA analysis, the GOTHIC model assumes a constant drop size of 100 microns for the liquid release from the break until the end of the blowdown phase, at which time a continuous liquid is assumed. This assumption is reasonable since the pressure difference between the RCS and containment after the blowdown will not be significant enough to force a break up of the liquid.

GOTHIC includes a drop break up model that can be activated for the break ju nction rather than specifying the drop diameter. The model generates drops from the liquid flow from flashing of superheated water and due to hydrodynamic forces on the water. For flashing conditions, the model generates drops that are approximately 80 microns, which agrees with ex perimental data.

The advantage of using the drop break up model is that the drop formation will automatically cease as the water temperature becomes subcooled. To make the drop break up model work properly, the upstream pressure must be approximately the actual pressure upstream of the break.

If -the containment pressure is used, the water will not be superheated and drops will not be formed. The drop break up model will not been used in any licensing calculations.

Topical Report DOM-NAE-3, Rev.*0.0-APae2 Page 27

3.5.2 MSLB Mass and Energy Release For MSLB, the mass and energy release data is obtained from the NSSS or fuel vendor using NRC-approved methods. The break junction uses 100-mnicron droplets for entrained liquid release, which was approved by the NRC staff in References 8-10. Consistent with current. plant licensing bases, a range of break sizes from small split breaks to the largest double-ended break size is analyzed over the range of 0% to 102% of rated thermal power. Analysis of this range ensures that the most conservative results are predicted for containment pressure and temperature. Plant-specific applications will specify the NRC-approved methodology used to generate the mass and energy release data.

3.5.3 LOCA Mass and Energy Release 3.5.3.1 Blowdown, Refill and Reflood Stages During a LOCA event, most of the vessel water will be displaced by the steam generated by flashing. The vessel is then refilled by the accumulators and the high and low pressure injection systems. GOTHIC is not suitable for modeling the refill period because it involves quenching ofI the fuel rods where film boiling conditions may exist. Current versions of GOTHIC do not have models for quenching and film boiling. Therefore, for the blowdown, refill and reflood stages, the mass and energy release rates are obtained from the NSSS or fuel vendor LOCA analysis using NRC-approved methods. The vendor release data includes the water from the ECCS accumulators, but the nitrogen release to containment is modeled separately in GOTHIC (see Section 3.2.2).

3.5.3.2 Post-Reflood Stage At the end of reflood, the core has been recovered with water and the ECCS continues to supply water to the vessel. Residual stored energy and decay heat comes from the fuel rods. StoredI energy in the vessel and primary system metal will also be gradually released to the injection water and released to the containment via steaming through the core or spillage, into the containment sump. In addition, there may be some buoyancy-driven circulation through the intact steam generator loops that will remove stored energy from the steam generator metal and the water on the secondary side. Depending on the location of the break, the two-phase mixture in the vessel may pass through the steam generator on the broken loop and acquire heat from the stored energy in the secondary system. For these conditions, GOTHIC is capable of calculating the mass andI energy release from the break into containment.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 28

3.5.3.3 GOTHIC Long-Term Mass and Energy Release Modeling The GOTHIC long-term mass and energy release accounts for the transfer of the decay heat and the stored energy in the primary and secondary systems to the containment after the end of reflood. The energy for each source term is acquired at the end of reflood from the fuel vendor's mass and energy release analysis. The rate of energy release is determined by a simplified GOTHIC RCS model that is coupled to the containment volume. Thus, the flow from the vessel to the containment is dependent on the GOTHIC-calculated containment pressure.

Lumped volumes are used for the vessel, downcomer, cold legs, steam generator secondary side, up-flow steam generator tubes and down-flow steam generator tubes. Separate sets of ioop and secondary system volumes are used for the intact and broken loops with the connections between the broken loop and containment as necessary for the modeled break location. The NSSS or fuel vendor's calculated mass and energy inventory at the end of reflood establishes the liquid volume fractions and the fluid temperatures in the primary and secondary systems.

The primary and secondary system geometries, including primary system resistances, are consistent with the models used for non-LOCA accident analyses. In order to predict the natural circulation through the intact loops and the correct water level in the vessel and downcomer, the volumes are modeled With the correct elevations. and heights. The vessel height may be adjusted so that the water and steam inventory at the end of reflood matches the vendor's boundary conditions, but this correction does not affect the hydraulic analysis.

Safety injection fluid is added to the downcomner volume (for the intact cold legs) and the broken loop cold leg. In both locations, the SI fluid mixes with the resident fluid and any vapor from the intact SGs.

The SI flow is taken from the RWST until a low-low level is reached, at which time the SI fluid is taken from the containment sump.

3.5.3.3.1 Energy Terms Thermnal conductors are used to model the core (stored energy plus decay heat), primary metal, secondary metal, and heat transfer across the SG tubes.

Primary Metal Stored Energy The distribution of energy throughout the primary system metal may not be provided in the vendor data at the end of reflood. Instead, a lumped metal energy is provided. The metal in contact with steam in the vessel and piping would be substantially hotter than the metal in contact with liquid. For example, the Surry lumped metal energy at the end of reflood corresponds to an average metal temperature about -170 F above the RCS saturated liquid for a pump suction Topical Report DOM-NAF-3, Rev. 0.0-APae2 Page 29

break. Initializing GOTHIC with this temperature. distribution leads to unrealistic boiling rates early in the post-reflood phase. The effective heat transfer coefficient 'inthe boiling regime is very high and it is expected that the metal in contact with liquid would not be substantially hotter than the RCS liquid at the end of reflood. However, the metal in the steam region of the vessel couldI be substantially hotter, accounting for the high level of stored energy at the end of reflood.

The primary metal is modeled conservatively such that all of its stored energy is released when the vessel is fully depressurized. One thermal conductor is used for the energy stored in the primary system metal. The Film heat transfer option is used on the conductor side in contact with the RCS liquid so that boiling heat transfer can be modeled. The other side of the conductor is insulated.

To consolidate conservatively all of the primary metal in one conductor in contact-, with liquid, theI metal temperature is initialized a few degrees hotter (typically 5 F) than the saturation temperature and an effective mass is calculated using Equation 15 (assuming 5 F hotter metal)..

M eff Mnom -(Iý,, - Tmi) Equation 15 where Mnom is the nominal lumped primary. metal mass, TO,,~ is the lumped primary metal temperature based on M,,,m and the vendor end-of-reflood primary metal energy, TLo, is the saturation temperature of the fuilly depressurized vessel, and Tsa, is the estimated vessel saturation temperature at the end of reflood. This approach gives reasonable boiling rates at the beginning of the post-reflood phase. It conservatively models the energy removal rate from the metal because it assumes that all of the primary system metal is in contact with water. This method ensures that allI of the stored energy in the primary metal is removed when the vessel is fully depressurized and the acceptance criteria for containment depressurization and NPSHa are challenged.

Core Stored Energy The fuel rods are modeled with a thermal conductor. The vendor's energy inventory at the end of reflood is used to set the initial temperature of the fuel rod conductor, consistent with the total heat capacity of the defined conductor. The Film heat transfer option is used on the rod surface inI contact with the. RCS liquid so that boiling heat transfer can be modeled. The other side of the conductor is insulated.

Decay Heat The decay heat is modeled by specifying a time-dependent internal heat generation for the fuielI conductor. The decay heat fractions are acquired from the .1979 ANSI/ANS Standard 5,.1-1979 with 2a uncertainty added [29]. These fractions are consistent with the current licensing analysesI performed for Surry Power Station [16] using the methodology in WCAP-10325-P-A [15]. The decay heat rate is based on 102% of rated thermal power to account for calorimetric uncertainty.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 30

Steam Generators A therm al conductor is used to model the transfer of energy stored in the shell side of the steam generator to the SG secondary fluid. The initial temperature is set to match the available stored energy specified at the end of reflood by the fuel vendor analysis. The up flow and down flow tubes on the steam generators are modeled separately with thermal conductors. This allows for the possibility of boiling in the up flow tubes and superheating of the steam in the down flow tubes. The heat transfer from the secondary side to the primary side is modeled using conductors with the inside connected to the primary system tube volumes. The Film heat transfer option is used on both sides of the tube. This option automatically accounts for heat transfer to the liquid or vapor phase as appropriate and includes boiling heat transfer modes.

3.5.3.3.2 Pump Suction Breaks Pump suction and cold leg breaks require special consideration because of the potential for significant energy transfer from the SG secondary fluid to the two-phase m-ixture leaving the core. The simplified RCS GOTHIC model is used to calculate the mass and energy release rate in the post-reflood phase. During the early part of this phase, there is substantial boiling in the vessel due to the decay heat and the release of stored energy in the fuel, vessel and internals. The boiling raises the two-phase level of the water in the vessel. The surface level affects the amount of. water that is carried into the steam generators with the steam produced 'inthe vessel and, consequently, the rate of energy removal from the steam generators. The stored energy in the steam generators will be released more quickly if there is significant water carried in to the steam generators. Another phenomenon that must be addressed is mixing in the cold legs and downcomer of steam from the intact SGs with cold SI. The GOTHIC treatment of both phenomena is addressed below.

Vessel Two-Phase Level GOTHIC includes the capability to model pool swell due to boiling and vapor flow through the pool, but the volume must be subdivided for these models to be effective. For lumped volumes, the pool surface level is simply determined by the product of the liquid volume fraction and the volume height. If the volume is subdivided, with multiple levels, the level of the water in the upper volumes will be raised by the vapor that displaces the liquid in lower volumes. In addition, the Yeh [34] model is used to estimate that effective liquid level and vapor fraction in the water for any flow paths (junctions) connected to the cells above the lowermost layer.

To activate the Yeh model, the volume representing the RCS vessel is subdivided into two cells.

The lower cell represents the lower plenum and core region up to the bottom of the upper plenum and the upper cell represents the upper plenum and upper head. All of the vessel heat sources are located in the lower cell so that all of the vapor flow is into the bottom of the upper cell. This will Topical Report DOM-NAF-3, Rev. 0.0-APae3 Page 31

maximize the pool level from the Yeh correlation and will maximize the pooi swell due to void formation in the lower cell. The hydraulic diameter for the vessel cells will affect the bubble rise velocities and the pool swell. The core hydraulic diameter is specified to capture both effects.

Steam Condensation in Cold Legs The mixing of cold safety injection (SI) water with the steam from the intact steam generators .

influences the condensation rate and therefore the flow rate through the steam generators. The release of stored energy in the steam generators will be accelerated if the flow through the steam generators is increased. Higher condensation in the downcomer and cold legs also increases the sump temperature for analysis of NPSH available. A subdivided model for the cold leg and theI water injection would provide a realistic estimate of the mixing and condensation rate inthe cold leg. However, with a simplified lumped volumne modeling of the primary system components, the condensation rate is largely control by the specified liquid/vapor interface area.

For Surry, the maximum SI flow into a single loop is 1527 gpm through a 6" injection pipe, givingI an injection velocity of about 19 ft/s. This jet enters the top of the cold leg. The jet momentum is expected to result in substantial mixing in the cold leg. Further, based on GOTHIC calculations, the steam velocity entering the cold legs is about 85 ft/s from early in the post-reflood phase up to

-800 seconds after the LOCA. These steam velocities are high enough to result in substantial drop entrainment and high condensation rates. Based on these conditions, it is expected that water in the cold leg would condense steam up to the point that it reaches that saturation temperature or all of the steam produced by the primary system, whichever is smaller.

Experimental evidence indicates that, during the post-reflood injection, the condensation rate is maximized. The tests documented in Reference 35 were for a 1/3 scale (10" diameter) cold leg.

Table 3.5-1 compares the test conditions to the post-reflood conditions from a Surry GOTHIC analysis of a double-ended pump suction (DEPSG) break. The Surry water injection flow rate isI for maximum SI. The range of conditions for Surry are from the end of reflood (i.e., activation of the GOTHIC simplified RCS model at -200 seconds) to the start of SI recirculation flow at

-3600 seconds. Early in this period, the Surry steam flow rate is substantially higher than the experimental range. The injection temperature is lower than the experimental conditions but it is not expected to significantly influence the mixing phenomena. The higher steam flow rates would give more entrainment than the experiments, which promotes maximum condensation. The smaller pipe diameter used in the tests would exaggerate mixing somewhat at lower velocities when gravitational forces become relatively more important.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 32

Table 3.5-1: Comparison of Accident and Experimental Conditions for Cold Leg Mixing Parameter Experiment 1[351 Surry DEPSG Pressure, psia 22, 50 12 to 50 Injection Water Velocity, ft/s .1-16 19 Injection Wa ter Temperature, F 80, 120, 150 45 (RWST)

Injection Angle, degrees 90, 45 90 Steam Temperature, F Sat, 500 Saturation - 400 Steam Flow, ibm/s 3.85, 8.25, 16.8 (full scale) 10 to 65 Based on the previous discussion, the GOTHIC downcomer volume uses a value of 1.OE+08 ft2 for the liquid/vapor interface area to promote thermal equilibrium conditions, consistent with the experimental evidence. This assumption conservatively maximizes the steam condensation rate and the energy removal rate from the steam generators. Sensitivity studies indicate that the specified value is sufficient to ensure thermal equilibrium conditions in the downcomer volume.

The complete mixing assumption between the steam from the intact cold legs and the SI water is also consistent with the NRC-approved methodology in the Westinghouse FROTH code [14].

The GOTHIC model can split the SI flow to the downcomer (for the intact cold legs) and to the broken cold leg based on the actual plant flow distribution. For the broken loop, equilibrium conditions are not assured and some of the injected water may exit the cold leg without significant interaction with the steam. Therefore, the GOTHIC broken cold leg volume assumes a value of 1.OE+08 ft2 for the liquid/vapor interface area only when it is conservative to assume thermal equilibrium (e.g., to maximize sump temperature for NPSH analyses). For containment depressurization, a liquid/vapor interface area of 0 is specified in the broken cold leg volume, because it is conservative to add to the containment any steam that exits the downcomer (alternatively, the broken cold leg can be modeled with a junction to the containment).

3.5.3.3.3 Hot Leg Breaks For a hot leg break, the core exit fluid preferably flows out of the broken hot leg, bypassing the steam generators. The flow to the intact SGs will be a very small fraction of the total core exit flow, calculated by the GOTHIC hydraulic model. Because all cold leg injection fluid must pass through the core, the core and vessel metal transfer all of their energy to the SI fluid which spills out of the break. The simplified GOTHIC RCS model is initialized consistent with the energy distribution provided by the vendor at the end of reflood (see Section 3.5.3.3. 1).

The simplified GOTHIC RCS model developed for the pump suction break is used for analysis of the hot leg break, with differences for the definition of flow paths that discharge to the Topical Report DOM-NAF-3, Rev. 0.0-APae3 Page 33

containment. The hot leg model is also simpler in that it does not include two-phase level swell in the core because there is no need to model liquid entrainment into the steam generator tubes. It also does not assume thermal equilibrium in the downcomer, as very little steam flows through the intact steam generators since all subcooled ECCS water must flow through the core.

3.5.3.4 Qualification of the GOTHIC Mass and Energy Release Models Section 4 describes a typical application of this methodology for Surry Power Station. The analyses demonstrate the conservative nature of the GOTHIC post-reflood mass and energy release rates for DEPSG and DEHLG breaks. This same type of benchmarking would be performed for each plant-specific application.

The integral mass and energy release rates for the two DEPSG break cases (i.e., containment depressurization and LHSI pump NPSHa) compare well to the current post-reflood methodology using the NRC-approved FROTH methodology in WCAP-8264-P-A [14] and WCAP-10325-P-A

[16], as implemented using the SWEC LOCTIC containment response code. The SG secondary energy is removed very quickly by the liquid entrainment from the vessel into the intact loop SG tubes. At the time of interest for both LHSI pump NPSHa (i.e., sWitchover to sump recirculation) and containment depressurization (i.e., containment pressure at 14.7 psia), the SG secondary energy has been released to the containment and the core and primary metal conductors are at or just above the primary system liquid temperature. The integral mass and energy results are similar to the FROTH/LOCTIC analysis methodology.

For the DEHLG break (Section 4.3), the GOTHIC model employs a conservative assumption for the core conductor temperature at the end of reflood. In addition, the simplified RCS model continues to remove primary metal energy when containment pressure decreases below 14.7 psia.

Compared to the Westinghouse methodology, more energy is removed from the primary system in the form of higher SI spillage temperatures. The mass release to the containment matches the LOCTIC analysis very closely.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 34

3.6 Containment Initial Conditions The initial containment atmospheric conditions are chosen consistent with the guidance in NUREG-0800, Sections 6.2.1 and 6.2.1L1LA [37]. The assumptions vary depending on the type of containment being analyzed. For atmospheric containments, the influence of the containment initial conditions was confirmed by running parametric studies using a typical GOTHIC model that assumes a Technical Specifications limit on total pressure and by varying one input while keeping the others constant. The most conservative settings for containment integrity analyses are summarized in Table 3.6-1. The term MAX indicates that the parameter is set to the largest allowable operating value (accommodating instrument uncertainty), while MIN indicates that the parameter is set to the smallest allowable operating value.

North Anna and Surry operate with subatmospheric containments with Technical Specifications limits on maximum and, minimum containment air partial pressure. Establishing the air partial pressure limits requires some different input assumptions from Table 3.6-1. For example, LOCA peak pressure cases use the maximum containment temperature and maximum relative humidity.

The maximum containment temperature provides the largest initial vapor pressure and the most stringent limitation on air partial pressure. Table 3.6-2 documents the containment initial condition assumptions for subatmospheric containments. Additional analyses are listed that affect the containment air partial pressure limits. These sensitivities are consistent with the current UFSAR LOCTIC analyses for Surry and North Anna.

Table 3.6-1: Containment Initial Conditions Analysis Pressure Temperature Humidity LOCA Peak Pressure MAX MIN MIN MSLB Peak Pressure MAX MLAX MIN LOCA Peak Temperature MAX MAX MAX MSLB Peak Temperature MIN MAX MIN Table 3.6-2: Containment Initial Conditions for Subatmospheric Plants Analysis Pressure Temperature Humidity LOCA Peak Pressure MAX MAX MAX MSLB Peak Pressure MAX MAX MAX LOCA Peak Temperature MAX MAX MAX MSLB Peak Temperature MIN MAX MIN Containment Depressurization MAX MAX MAX Subatmospheric Peak Pressure MA4X MIN MAX NPSH Available MIN MAX MAX Topical Report DOM-NAF-3, Rev. 0.0-APae3 Page35

Treatment of Instrument Uncertainties The containment analysis 'includes design inputs for plant parameters that are controlled by Technical Specifications (TS). Examples include containment air partial pressure, containmentI temperature, RWST temperature, and service water temperature. The GOTHIC analyses account for instrument uncertainty on the TS surveillance parameters in one of two ways, with an example for each application:

1) Surveil the TS limit and apply the instrument uncertainty deterministically to develop a GOTHIC input.

Example: A TS containment temperature maximum limit of 125 F with a 1 F uncertainty would be analyzed in GOTHIC at 126 F.

2) Set the plant surveillance limit with margin to the TS limit, which would be the GOTHIC input.

Example: With a. 1 F uncertainty, the plant surveillance would verify containment temperature is less than 124 F. GOTHIC input would be the TS limnit of 125 F.

Both options provide flexibility to accommodate differences in plant surveillance practices while ensuring that the GOTHIC containment analyses are bounding for operation at the TechnicalI Specification limits. Plant-specific applications can utilize either method.

Topial epot DO-NA-3,Rev.0.0A PgeI

3.7 Run Control Options The GOTHIC default settings for the run controls are used. The default settings were used for all of the GOTHIC validation against experimental data [3]. The default settings are listed with a brief discussion of the significance of the parameter for containment analysis. Only those parameters that may affect the calculated results are discussed. The remaining typically control code output and have no impact on the computed results.

3.7.1 Revaporization Fraction Default value: DEFAULT The revaporization fraction is the fraction of the condensation rate that can be vaporized if the steam in the containment is superheated. Superheat conditions typically o ccur only for a MSLB.

NUREG-0588 [22] allows a maximum revaporization rate of 8% corresponding to a revaporization factor of 0.08. If the atmosphere is superheated, the specified revaporization will be credited regardless of the degree of superheat.

Under the DEFAULT option, GOTHIC uses its built-in models for calculating the vaporization of the liquid in the containment. This model uses a heat and mass transfer analogy to estimate the mass transfer coefficient from the heat transfer coefficient. It accounts for the convective heat transfer and the evaporation in the presence of non-condensing gases. The rate of heat and mass transfer depend on the degree of superheating. That is, the vaporization rate will increase as the superheat 'increases.

For a MSLB simulation, the DEFAULT option typically gives containment pressures and temperatures, that are very close (within 0.1 psia and 10 F) to the same model with the

.revaporization factor set to 8%. The default models are more physically based and recommended for MSLB analysis. The DEFAULT option was used in the NRC-approved Kewaunee submittal for power uprate [8, 9] and is part of the basis for all validation of the DLM condensation option.

3.7.2 Fog Model Default value: OFF The fog model is used to generate fog when the containment atmosphere becomes supersaturated.

The fog model has been superseded by the mist model. The option is retained in the code to allow comparison with earlier versions. The fog model creates very small drops that, when combined with drops from the blow down or sprays, result in an average drop diameter that may not be representative of either the fog or the spray.

.Topical Report DOM-NAF-3, Rev. 0.0-APae3 Page 37

3.7.3 Maximum Mist Density Default value: DEFAULT With the mist model in GOTHIC, if the containment atmosphere becomes supersaturated, smallI water drops will be generated and the heat of vaporization is added to the vapor phase to eliminate the subcooling. If the vapor starts to become superheated, any mist will be evaporatedI using heat from the vapor phase. The mist droplets are assumed to be very small so that they move with the vapor and do not settle out. However, if the mist density exceeds the specified value for this control parameter, it is assumed that the mist begins to agglomerate and drops are formed that are added to the drop phase. The default value is 1 gm/in 3 based on meteorological data [26].

3.7.4 Drop Diameter from Mist Default value: DEFAULT As described in Section 3.7.3, when the mist density exceeds the specified maximum value, the excess mist is converted to drops at the diameter specified by thi~s control parameter. The default value is 200 microns. This is larger than, drops typically found in clouds (-20 microns [26]) but is purposely selected larger to avoid mixing very small drops with containment sprays. The drop formation at this diameter does not significantly affect the diameter of the drops from spraysI because the formation rate is much smaller than the spray rate. However, the GOTHIC predicted drop diameter in the containment should be monitored to confirm that the drops are close to the expected value from the sprays.

3.7.5 Minimum Heat Transfer Coefficient Default value: 0.0 This is the minimum heat transfer coefficient on the vapor side of the liquid vapor interface. There is also a corr esponding miinimumn mass transfer coefficient by the he at and mass transfer analogy.

For containment analysis, GOTHIC is allowed to calculate appropriate heat and mass transfer coefficients based on the vapor and liquid phase conditions.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 38

3.7.6 Reference Pressure Default value:. DEFAULT If a positive value is specified, the vapor density in the body force term in the momentum balances is calculated using the specified pressure and the local temperature. This option is a carry over from older code versions where the vapor pressure may have been inappropriately influenced by the presence of liquid pools. It is no longer useful and DEFAULT is specified so that the local pressure is always used for calculating the vapor density.

3.7.7 Force Entrainment Drop Diameter Default value: DEFAULT In subdiv ided models, the user has the option to force the conversion of liquid flow through a horizontal cell face to drops at the specified diameter. The default value is 0.1 inches. This parameter does not affect lumped models.

3.7.8 Vapor Phase Heat Correction Default value: INCLUDE In a cell with a liquid pool that extends above the cell midplane, the cell pressure will reflect the gravitational head of the water in the pooi that is above the cell midplane. If this option is set to INCLUDE, this pool gravitational head is subtracted from the cell pressure to calculate the vapor phase pressure.. This option is retained only for the purpose of comparing against older code versions that did not include the pressure adjustment. In all other cases the option is set to INCLUDE.

3.7.9 Kinetic Energy Default value: IGNORE For high speed flows the kinetic energy is significant and some of the fluid thermal energy is converted to kinetic energy as the flow is accelerated. This option will have minimal effect on containment analysis and is set to IGNORE.

Topical Report DOM-NAF-3, Rev. 0.0-APae9 Page 39

3.7.10 Phase Options Default values: INCLUDE For single phase problems, one or two of the phases can be ignored to speed the computation. For containment analysis, all phases are important and the option is set to INCLUDE for all phases.

3.7.11 Force Equilibrium Default value: IGNOREI If this option is set to INCLUDE, the phases will be forced into thermal equilibrium by using very large interface heat and mass transfer coefficients. The interphase drag coefficients for junctions will also be set to large values to force the phases-to travel at a common velocity. This option is available for subcompartment analysis and is not used for containment analyses. When this option is set to IGNORE, there is no assumption regarding phase equilibrium in GOTHIC and the phase temperatures and velocities are determined by the phase balance equations and the interfaceI 3.7.12 Drop-Liquid Conversion Default value: INCLUDE If this option is set to IGNORE, then drop deposition and drop entrainment will not be allowed.

Drops injected at the break or as sprays would remain suspended in the atmosphere indefinitely.I For containment analysis, this option should be set to INCLUDE, allowing GOTHIC to deposit and entrain drops based on the mechanistic model in GOTHIC. Drop entrainment is not expected but drop deposition is a significant contributor to the containment response.

3.7.13 Version 6.1 FormulationsI Default value: OFF The development of version 6.1 included changes to some fundamental models in GOTHIC that resulted in some small but significant changes in the calculated results for some models. TheI involved code revisions in 7.0 and later were determined to be improvements and more faithful to the physics. This option was added to allow comparison of the newer code results with those from previous versions. For all other purposes this option is set to Off.

Topical Report DOM-NAE-3, Rev. 0.0-A Page 40

3.8 Calculation of NPSH Available 3.8.1 Containment Overpressure Licensing Basis The licensing bases for North Anna and Surry both allow credit for containment overpressure in calculation of NPSH available (NPSHa) for the RS and LHSI pumnps. The overpressure is explicitly included in the SWEC LOCTIC containment analysis methodology, which performs a transient calculation of NPSHa. This methodology was reviewed recently by the NRC as part of Generic Letter 97-04. The following excerpt is from a letter from Virginia Power to the NRC dated October 29, 1998

[38]:

Virginia Power summarized the analysis methodology concerning the use of containment overpressure for the determination of NPSH for the emergency core cooling and containment heat removal pumps in the previous response to Generic Letter 97-04 noted above.

Furthermore, we have concluded from a review of the relevant correspondence that the methodology to credit containment overpressure is part of the licensing bases for both Surry and North Anna. A specific value for containment overpressure credit in the determination of NPSH for the emergency core cooling and containment heat removal pumps has not been previously provided to the NRC for review and approval. Rather, NRC approval has been directed at verification of the adequacy of the methodology used to determine that the available NPSH-is greater than the required NPSH for these pumps. to Reference 38 provided additional details on the specific analysis methodology that was used to maximize sump temperature and minimize containment pressure for calculating NPSHa for the RS and LHSI pumps. The NRC agreed with the, licensing basis for this transient methodology for calculating NPSHa in Reference 39 for North Anna and Reference 40 for Surry. Using the same approach as the current LOCTIC transient methodology, Dominion will use GOTHIC to predict conservatively the containment conditions following a LOCA in order to determine the NPSHa for the LHSI and RS pumps at Surry and North Anna. An overall conservative calculation is performed to minimize containment pressure and maximize containment sump temperature.

Currently, the other Dominion plants (Millstone Point 2, Millstone Point 3, and Kewaunee) do not credit containment overpressure in the calculation of NPSH available. The methodology in Section 3.8.2 is intended for use at North Anna and Surry. However, the methodology is applicable to large, dry PWR containments and may be used in future licensing applications for other plants.

Topical Report DOM-NAF-3, Rev. 0.0-APae4 Page 41

318.2 GOTHIC Analysis of NPSH Available NPSHa is the difference between the fluid stagnation pressure and the saturation pressure at the pump intake. To calculate NPSHa for a given pump, the GOTHIC containment model includes a

.separate small volume for the pump suction. The volume elevation and height are set so that the mid-elevation of the volume is at the elevation of the pump first-stage impeller centerline. The volume pressure (with some adjustments for sump depth) can then be used in the NPSHa calculation. The temperature in the suction volume provides the saturation pressure. The junction representing piping between the sump and the suction volume reflects the friction and form pressure drop between the sump and the pump suction. The pump suction volume also allows accurate modeling of the mixing of cold water that is injected into the suction of the RS pumps at Surry .and North Anna.

The single volume GOTHIC model does not account for geometry details of the sump or the liquid that is held up in other parts of the containment. GOTHIC does calculate the total amount of liquid in the containment. A correlation is used to define the sump depth or liquid level as a function of the water volume 'inthe containment. The correlation accounts for the sump geometry variation with water depth and accounts for the holdup of water in other parts of the containment, as discussed in Section 3.8.3. This correlation is installed in a GOTHIC control variable for use in the NPSI~a calculation.

With the above modeling features in place, the NPSHa is calculated via control variables as NPSH,,= P,+p,g[E, (ý) -E~c,]-HlP5 ,., ) Equation 16 where P, is the. GOTHIC calculated pressure 'inthe pump suction volume, p, is the liquid density in the sump, E5, is the elevation of the sump surface obtained from the installed correlation or table as a function of V,, (the water volume 'in the contaimnment), E, is the elevation of the containment volume, H is the height of the containment volume, a, the liquid volume -fraction in the containment, P,, 1(Ts) is the saturation pressure at the pump suction temperature, p, is the fluid density at the pump suction.

Worst case conditions for NPSHa depend on the time that the pumps take suction from the sump.

Therefore, the parameter settings that minimize NPSHa may vary depending on the timing for the operation of the pumps. In general, settings that reduce containment pressure and increase the sump water temperature reduce the N`PSHa. Section 4.7 lists the input parameter studies that provide the limiting set of conditions for Surry.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 42

The water in the sump comes from three sources: direct deposit of mass from the break, condensate from the conductors, and spray drops. The drops from the blowdown will be very small and at the saturation temperature at the containment steam partial pressure when they enter the sump. After the blowdown, the spillage water from the vessel is directly put in the sump with no heat transfer to the atmosphere or walls and equipment in the containment. This is a conservative approach for NPSH analysis. The condensate is generated at the saturation temperature at the steam partial pressure and added directly to the sump. The heat transfer between the conductors and the condensate on the way to the sump is conservatively neglected.

If the spray drops are modeled as recommended below, the drops will enter the sump at the maximum possible temperature. Heat and mass transfer at the sump surface is allowed.

GOTHIC's model for heat and mass transfer at a pool is in good agreement with experimental data (e.g., the Grout Mold evaporation experiments [3]). For NPSH analysis, the liquid temperature is greater than the vapor temperature for most of the event, so a minimum pool area is specified to minimize evaporation. With this overall approach, the predicted sump temperature is conservatively high for the duration of the simulation.

The following adjustments are made to ensure a conservative calculation of NPSHa:

1) The heat and mass transfer to the containment heat sinks are expected to be under-predicted using the Direct heat transfer model. This is non-conservative for NPSH analysis. A multiplier of 1.2 applied to the heat transfer coefficient was shown to provide adequate conservatism in the calculation.

2) All of the spray water is injected as droplets into the containment atmosphere (nozzle spray flow fraction of 1). Analyses are performed using the largest Sauter droplet size. A confirmatory analysis is performed by reducing the Sauter diameter by 2, which sufficiently covers code and spray performance uncertainty (i.e., variation in nozzle design and orientation, nozzle flow rate and different header elevations) without creating drops too small that may cause excess droplet holdup in the atmosphere. NPSH analyses are relatively insensitive over this range of droplet size, and the two cases together confirm that the effect of sprays on reducing containment pressure is maximized. The minimum NPSHa is reported from the case that provides the smaller NPSHa.

3) A conservative water holdup volume is subtracted from the containment liquid volume to reduce the sump water height. See Section 3.8.3.

4) The upper limit on containment free volume is used.

5) The minimum containment air pressure is used.

6) Conservative assumptions for spray and other system parameters are used in accordance with plant-specific sensitivity studies (Surry results are summarized in Section 4.7).

Topical Report DOM-NAIF-3, Rev. 0.0-APae4 Page43

The NPS14a result from GOTHIC is based on the conditions at the pump first-stage impeller elevation. The difference in elevation between the pump intake and the containment floor is included. Also, the pump suction friction and form losses (including the clean strainer) areI specified in. the junction between the containment and the pump. Therefore, the margin between the GOTHIC-calculated NPSHa and the required NPSH includes all essential elements of the problem except for strainer bed debris head loss, which is calculated external to GOTHIC and compared to the available margin betwcen NPSHa and required NPSH.

3.8.3 Water Holdup A conservative containment water holdup volume is subtracted from the GOTHIC-calculated containment liquid volume to reduce the sump water height. The holdup volume includes the following items:

1) water added to the spray system piping,

2) water trapped from transport to the containment sump in volumes (e.g., refueling canal and reactor cavity),

3) condensed films on heat structures,

4) films on platforms and equipment that form after spray is initiated, and

• 5) other losses (e.g., water absorbed in insulation).

Some of the parameters are taken as constant penalties against the water level early in the event.

Other items, such as fillfing the refueling canal, are dependent on spray actuation times and coverage. Control variables are used to calculate the total decrement to the GOTHIC containment liquid volume fraction .as a function of time. The corrected liquid volume fraction is then enteredI into a table of containment water level versus volume to determine the sump level to be used in the NPSHa calculation.

GOTHIC has a droplet field in each control volume. This droplet field explicitly models water drops in the atmosphere separately from the continuous liquid field. Thus, the mass of waterI droplets is not included in the containment liquid volume fraction that is used to determine the containment water level. No level correction is required for the water drops in the atmosphere.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 44

3.9 Time Steps Calculations are divided into a number of time domains to adequately control the output and time steps for the various phases of the transient. Small time steps and frequent graphics output is needed to accurately track the transient during the blowdown phase and to capture the peak temperature and pressure. Larger time steps and longer graphics intervals can be used for the long term analysis.

There are numerous internal controls on the time step based on numerical stability requirements and limits on the incremental change *inkey variables. These limits generally provide a good solution with a minimum number of time steps. However, user guidelines instruct the analyst to demonstrate that the automatically selected time step provides a converged solution or to impose additional time step limits to achieve a converged solution.

The recommended approach for -time step sensitivity studies is to first allow GOTHIC to select its own time step limits based on the internal controls. Plot the time step and then rerun the calculation with imposed limits that approximate the automatically selected time steps. Reduce the imposed limits by a factor of two and compare results. Repeat until there are no significant changes in key parameters (e.g., peak temperature and pressure). Alternate methods for time step sensi tivity may be followed as long as time step convergence is demonstrated.

Topical Report DOM-NAF-3, Rev. 0.0-APae4 Page 45

4.0 GOTHIC Demonstration Analyses for Surry Power Station This section documents GOTHIC containment analyses for Surry Power Station that demonstrate the acceptability of the analysis methodology described in Section 3. Analyses were performed for LOCA peak pressure and temperature, MSLB peak pressure and temperature, containment depressurization, and NPSH available for the LHSI pumps. Comparisons were made to the SWEC LOCTIC analyses described in the Sunry UFSAR. Two types of benchmarks were performed:

1. GOTHIC models were adjusted to provide the same physical behavior as LOCTIC. For example, the GOTHIC droplet phase was effectively disabled to compare to the LOCTIC equilibrium flash model and the containment volume liquid/vapor interface area was set to zero.

These benchmarks used long-term mass and energy data calculated by LOCTIC. The objective was to demonstrate adequate modeling of containment components, nodalization of piping systems, and modeling of spray systems, with respect to another containment response code.

These benchmarks showed a successfuil comparison of the containment response.

2. GOTHIC models were changed to implement the methodology in Section 3 and were run using the same plant design inputs (e.g., initial conditions, ECCS and spray flow rates, heat sinks) as in the LOCTIC analyses of record. The post-reflood mass and energy release is calculated using the GOTHIC RCS model. These comparisons show the modeling benefits from GOTHIC while demonstrating similar transient behavior to LOCTIC.

The second set of analyses is included in this section to demonstrate the GOTHIC analytical methodology. Each analysis includes a comparison to the LOCTIC containment response and the mass and energy release rates to justify the GOTHIC simplified RCS model for DEPSG and DEHLG breaks. Surry does not have a MSLB containment response analysis in the UFSAR. Analyses were performed using North Anna mass and energy data with the Sun-y containment model.

4.1 Surry Power Station Description S urry Power. Station is a three-loop Westinghouse PWR with a subatmospheric containment design.

The following plant description is taken from Chapters 5 and 6 of the Sunry UESAR .[28]. Surry's engineered safeguards features (ESF) that mitigate a LOCA or MSLB event include:

1. A safety injection (SI) system that injects borated water into the cold legs of all threeI reactor coolant loops.

2. Two separate low-head safety injection (LHSI) subsystems, either of which provides long-term removal of decay heat from the reactor core.

3. Two separate subsystems of the spray system-containment spray (CS) and recirculation spray (RS)-that operate together to reduce the containment temperature, return the containment pressure to subatmospheric, and remove heat from the containment. The RS Topical Report DOM-NAF-3, Rev. 0.0-A Page 46

subsystem maintains the containment subatmospheric and transfers heat from the containment to the service water (SW) system.

The CS system consists of two pumps that start on a Consequence Limiting Safeguards (CLS) containment pressure high high signal and draw suction from the RWST until the tank is empty. The RS system consists of four independent trains, each with one pump that takes suction from the containment sump. The RS pumps are started currently using delay timers that are initiated on the CLS signal. The delay time allows for sufficient water to accumulate in the sump. Each RS train has a recirculation spray heat exchanger (RSHX) that is cooled by SW (on the tube side) for long-term containment heat removal. The SI system consists of two LHSI and thrce HHSI pumps that draw from the RWST and inject into the RCS cold legs. The SI pumps take suction from the RWST until a low-low level is reached. Then the LHSI pumps swap suction to the containment sump and the HHSI pumps swap suction to. the LHSI pumnp discharge.

4.2 Surry Power Station GOThIC Model Overview This section contains a detailed discussion of plant-specific modeling details for Surry that are not the same for all GOTHIC containment models covered by this report. Differences between plant systems may require different model approaches for volumes, flow paths, trips, etc. For example, S urry and North Anna have slight differences in the recirculation spray systems that require a different number of volumes and flow paths. In addition, the modeling of other elements, such as piping fill times, pump start ramps, and pump heat addition, may vary between models without affecting GOTHIC results. Therefore, these model differences do not represent a change in the methodology, because the treatment does not affect the GOTHIC results.

4.2.1 Geometry The Sun-y containment is represented by a lumped, volume. Other volumes model the RWST and piping for the spray and safety injection systems. Ten volumes are used to model the primary system and secondary side of the SGs in accordance with Section 3.3.3. Separate conductors model the core, primary metal, SG tubes, and SG secondary metal. Twenty thermal conductors model the containment passive heat sinks. Flow paths model the break through the end of reflood using the vendor's mass and enthalpy data. At the end of reflood, the GOTHIC simplified RCS model is activated. The release from the first set of flow paths is stopped and different flow paths are activated from the RCS. For a DEPSG break, different flow paths model the release from the broken ioop cold leg and the broken loop pump suction during post-reflood. For a DEHLG break, different flow paths model the broken hot leg release from the vessel and the broken hot leg connection to the SG The design inputs for the physical plant (e.g., containment free volume and diameter, RWST available volume, piping volumes, RS timer setpoints) are consistent with the LOCTIC analyses of record in the Sunry UESAR [4]. The sump level in both codes is based on a 126-ft diameter cylindrical containment.

Topical Report DOM-NAF-3, Rev. 0.0-APae4 Page 47

4.2.2 Engineered Safeguards Features The GOTHIC model includes a flow boundary condition to model the CS pumps. Flow is variable as a function of the RWST level and downstream pressure. Pump heat is added via a coupled boundary condition. Pipe fill time and pump start delays are incorporated into a delay time that passes before the CS pumps deliver flow to the containment headers. A fraction of CS pump flow is diverted to the suction of the ORS pumps using boundary conditions.

Each RS pump is modeled with a flow boundary condition. Constant flow rates are assumed to bound the minimum and maximum delivered flow rates calculated from system analyses. RS pump heat is added with a coupled boundary condition. Trips are used io start the pumps after the required time delay has passed, including uncertainties and pump start delays. Control volumes model the filling of the RS pump discharge piping. Control volumes are used for the RS pump suctions to allow the mixing of bleed flow and the accurate calculation of NPSHa at the pump first-stage impeller. Suction friction and form losses are consistent with the LOCTIC analyses.

Each of the four recirculation spray lines contains a single-pass, shell-and-tube heat exchanger located inside containment between the RS pump and the spray header. Heat exchanger performnance must be modeled correctly to ensure a conservative prediction of heat removal from the sump for long-termI accident analysis. The RSHXs model selections *inGOTHIC were benchmarked to a detailed heat exchanger design code over the range of accident flow rates and temperatures in the RS and SW systems. The models include tube plugging and fouling for analyses where it is conservative.

Safety injection is modeled with flow boundary conditions that draw from the RWST and the containment sump. Before the end of reflood, sink boundary conditions remove mass from the RWST consistent with the vendor mass and energy calculation. At the end of reflood, the GOTHIC mass and energy model is activated and boundary conditions inject RWST water into the primary system. When the RWST reaches a low-low level, the boundary conditions are terminated and another boundary condition directs water from the containment sump to the primary systerr.

Section 3.4.2 specifies a nozzle spray flow fraction of 1 with a reduced containment height. To get a sump level comparable to LOCTIC in the benchmark analyses, the containment height was calculated from the free volume and pool area and a spray flow fraction of 0.9 was used. This preserved the sump level and was shown to be more conservative than the. methodology in Section 3.4.2. Plant designI analyses will implement the methodology in Section 3.4.2 and use a spray flow fraction of 1.0.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 48

4.2.3 Mass and Energy Model LOCA break mass and energy release data up to the end of reflood is obtained from WCAP-14083

[16], which is the current Surry licensing basis LOCA data. Two flow boundary conditions represent the two sides of the broken pipe through the end of reflood. Mass and enthalpy is specified for each break side based on the Westinghouse data. Accumulator nitrogen is injected to the containment with another boundary condition. During the post-reflood phase, the GOTHIC simplified RCS model described in Section 3.5 calculates mass and energy releases out of both sides of the break for the rest of the transient. The vessel and downeomer are initialized (pressure, temperature, liquid fraction) consistent with the WCAP-14083 data at end of reflood. Volumes are used for the secondary side of the intact loops and broken loop SG, respectively.

Surry does not have plant-specific mass and energy release data for MSLB containment response.

Instead, North Anna MSLB data was used after it was determined to be conservative for Surry.

The North Anna data was obtained from WCAP- 11431 [32], which is the North Anna licensing basis MSLB mass and energy data using WCAP-8822-A [33] methods.

4.2.4 Containment Heat Sinks The containment passive heat sinks geometry and thermal properties were set the same as the LOCTIC input. The modeling guidelines for nodalization of each conductor from Section 3.3 was applied. The MSLB analysis model *includesthe accumulator tanks filled with water as an additional heat sink. The containment heat sinks are grouped into the following categories.

" Containment structure shell below grade

• Containment structure shell above grade

• Containment structure dome and liner

" Containment structure floor above floor liner

• Containment structure mat below floor liner

• Internal concrete slabs

• Carbon steel inside the containment

• Stainless steel inside the containment

• Accumulator tanks filled with water (MSLB only)

Heat transfer options were set consistent with Section 3.3.2. The Direct heat transfer option with DLM condensation was applied to all containment heat sinks except the sump floor. The Split option was used for the floor to switch the heat transfer from vapor to liquid as the liquid level builds 'in the basement. The containment walls above grade and the containment dome used a specified external temperature of 95 F with a heat transfer coefficient of 2.0 Btulhr-ft2 -F, which is consistent with the current LOCTIC analyses. For the LHSI pump NPSI-a analysis, a multiplier of 1.2 is applied to the Direct heat transfer coefficient (see Section 3.8.2).

Topical Report DOM-NAF-3, Rev. 0.0-APae4 Page 49

4.3 GOTHIC Analysis of LOCA Peak Pressure 4.3.1 Containmnent Response The containment peak pressure is obtained from a DEHLG break. Table 4.3-1 compares the key results of a GOTHIC benchmark analysis to the LOCTIC containment peak pressure analysis from the Surry UFSAR. Plant design inputs for containment initial conditions (12.5 psia, 125 F, and 100% humidity) and passive heat sinks are the same. The only differences are related to the GOTHIC methodology selections described in Section 3 (e.g., 100-mnicron break droplet size).

Figures 4.3-1 through 4.3 -6 compare the GOTHIC containment pressure, vapbr temperature,I liquid temperature, sump level, RSHX heat rate, and four conductor heat transfer' coefficients to LOCTIC values shown as discrete points. The GOTHIC containment temperaturie and pressure profiles exhibit the same behavior as LOCTIC but with different magnitudes. The lower peak pressure from GOTHIC is attributed to the droplet phase and the DLM condensation model. The droplets provide more heat transfer area and tend to produce smaller pressures than a liquid release from LOCTIC. Figure 4.3-6 compares the DirectIDLM heat transfer coefficients for four different GOTHIC heat sinks to the Tagarn-i-Uchida model used on all LOCTIC heat sinks.

As expected, GOTHIC provides margin in containment peak pressure and temperature but produces a higher containment liquid temperature than LOCTIC. In the long-term, the GOTHIC RSHXs have higher heat rates to remove the energy from the sump, such that the liquid temperature and RSHX heat rates converge at 1200 seconds.

4.3.2 DEHLG Mass and Energy Release The methodology in Section 3.5.3.3.3 was used. Westinghouse mass and energy release data is used up to 115.8 seconds, the end of reflood for the DEHLG break. At that time, the GOTHIC simplified RCS model is activated with initial conditions that are consistent with the Westinghouse mass and energy distribution from WCAP-14083 [15]. At this time, the break release is SI flow heated by the core and primary metal conductors. Figures 4.3-7 and 4.3-8 compare the integral energy release and integral mass release, respectively, to the LOCTIC output (which uses the Westinghouse data without adjustment). The integral mass release matches closely. The GOTHIC integral energy release to the containment is about 6% higher at 1200 seconds. Table 4.3-2 compares the GOTHIC integral energy addition to the primary coolant from the core and primary metal conductors to the difference in Westinghouse energy over this period [15]. The energy difference is based on two modeling differences.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 50

For conservatism, the GOTHIC model was initialized with the core conductor at the primary system liquid temperature (235 F). In contrast, the Westinghouse methodology in WCAP-8264-P-A removes all of the core stored energy before the end of reflood. The GOTHIC assumption adds a small amount of additional stored energy to the primary system.

The second difference is due to the ability of GOTHIC to calculate realistically the vessel liquid subcooling in response to more than adequate SI flow that is available to remove the core and metal energy. At 1500 seconds, the GOTHIC vessel liquid temperature is 137 F. This is about 100 F less than the vendor value of 235 F at the end of reflood. The lower value is expected from 3300 gpm SI flow removing core and metal energy. In contrast, the Westinghouse methodology applied in WCAP-14083 reduces the primary system liquid from 235 F at the end of reflood to 212 F at 1500 seconds. During this phase, the Westinghouse method does not remove any thin metal energy but the thick metal releases -10 MBtu. Over the same period, the GOTHIC primary metal conductor (thin and thick metal) has added almost 20 MBtus to the break fluid.

4.3.3 Summary of DEIILG Peak Pressure Comparison The GOTHIC containment temperature and pressure profiles exhibit the same behavior as LOCTIC but with different magnitudes. GOTHIC produces a lower blowdown peak pressure because of the break droplet model and the Direct/DLM condensation model. In the long-term, containment pressure and liquid temperature results. converge as the RS heat exchangers remove the excess energy in the GOTHIC sump liquid. The GOTHIC simplified RCS model for post-reflood mass and energy release from DEHLG breaks has been shown to be more conservative than the Westinghouse methodology 'in Reference 14. The model removes primary system energy in accordance with the calculated subcooling of the RCS liquid in the form of higher SI spillage temperatures as the vessel depressurizes.

Topical Report DOM-NAE-3, Rev. 0.0-APae5 Page 51

I Table 4.3-1: GOTHIC Comparison to LOCTIC for DEHLG Peak Pressure I

Peak containment pressure, psia GOTHIC 57.53 LOCTIC 59.14 I Time of peak pressure, sec 18.2 18.0 Peak containment vapor temperature, F 273.4 275.6 I Time of peak vapor temperature, sec: 18.0 18.0 Peak containment liquid temperature, F Time of peak liquid temperature, sec 253.3 31.0 234.5 30.8.

I Integral energy release at 1200 sec, MBtu Integral mass release at 1200 sec, Mlbm 1 404.0 1.1920 1 380.6 1.19551 I

Table 4.3-2: Primary System Energy Release from 115.8 to 1500 Seconds I

GOTHIC Westinghouse I

Core Decay Heat, MBtu.

Thick + Thin Metal, MBtu 82.0 19.5 81.21 9.76 I Core Stored Energy Included in Core Decay Heat No change Ivalue above I I

I I

I I

I I

I Topical Report DOM-NAE-3, Rev. 0.0-APae5 Page 52 I

Figure 4.3-1: DEHLG Containment Pressure Comparison to LOCTIC 1 Containment Pressure PRi DC25T UC . .......

0.11 Time sec GOTHIC 7.2ijom(OA) SeD21!2OO5 10:29-36 Figure 4.3-2: DEHLG Containment Vapor Temperature Comparison to LOCTIC 3 Containment Vapor Temperature TV1 DC27T 0 -I .... .

C...........

0 .............

Time (sac)

GOTHIC 7.2dom((OA) Sevi2g/WaO 10:29:36 Topical Report DOM-NAF-3, Rev. 0.0-APae5 Page53

Figure 4.3-3: DEHiLG Containment Liquid Temperature Comparison to LOCTIC 2 Containment Liquid Temperature 0.1 1 10 100 1000I Time (Sec)

GOTHIC 7.2dom,(OAl Seorm9I205 jo029:36.

Figure 4.3-4: DEHLG Containment Sump Level Comparison to LOCTIC 4 Sump Level1 LL1 DC17T c'............ .. ...

0~

E 0.1 1 Time (sec)

GOTHIC; 7.2dom(OA) Sep2gi2WS00 10:2Q-:36 Topical Report DOM-NAE-3, Rev. 0.0-APae5 Page 54

Figure 4.3-5: DEHLG RSHX Total Heat Rate Comparison to LOCTIC 57 RSHX Total Duty cv57C DC1 9T

~0 coJ 0L - i I I I I I I I I II I 0 0.3 0.6 0.9 1.2 1.5 Time (sec) Xle3 GOTHIC 7.20CM(OA) SepIWn205 '-0:2W,36 Figure 4.3-6: GOTHIC Conductor Heat Transfer Coefficients Comparison to LOCTIC 58 HTC vs LOCTIC HAI HA7 HAll HA16 DC20T 0 - - ---- ------.

0 M Ij C\j C 0 II, CI CI Q) 0 co 0.1 1 10 100 1000 Time (sec)

GOTHIC 7.2dom(OA) S-02912OD 10:29:36 Topical Report DOM-NAF-3, Rev. 0.0-APae5 Page 55

I Figure 4.3-7: DEHLG Integrated Energy Release Comparison to LOCTIC I

52 Total Break Energy

<D-ov45C

• DC34T

--

I C)

I

.... . ... . .. . .. . .. . . ..... .. . . .. .. . . . ... .. .

-~ C............ . ...... .....

I CO . .. ... ..... :...... ........

21 I

0 . ............ ...... .............. .. ....

I 0J-0.1 1 10101000 Time (Sec)

I GOTHIC 7.2domiOA) SoY'21Y2OOS 10:29ý3 Figure 4.3-8: DEHLG Integrated Mass Release Compared to LOCTIC I

53 Total Break Mass Release CV50C DC1 8T I

I LO I

I I

0.1 1 110 100 l00w I

I Time (sec)

GOTHIC 7.2dom(OA) SepI/ZW2005 10:29:36 I

I Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 56 I

4.4. GOTIUC Analysis of Containment Depressurization Containment depressurization is analyzed for subatmospheric containment designs to demonstrate that the containment pressure becomes subatmospheric within the time that is assumed for containment leakage in the dose consequences analyses. The maximum containment depressurization time occurs for a DEPSG break with minimum safeguards and mirinimum flow rates for the safety injection and spray systems. The LOCTIC analysis of record from the Surry UFSAR was repeated using the same design inputs and the GOTHIC methodology selections described in Section 3. The GOTHIC simplified RCS model for mass and energy release is consistent with Section 3.5.3.3.2. The vessel volume is subdivided with two axial nodes to activate the Yeh model for two-phase level swell and liquid entrainment into the SG tubes.

4.4.1 Containment Response Table 4.4-1 compares the time sequence of events from the GOTHIC and LOCTIC analyses.

Figures 4.4-1 through 4.4-3 compare the containment pressure, vapor temperature, and liquid temperature to LOCTIC results (shown as a dashed line). During the early part of the transient, GOTHIC predicts lower containment pressure and vapor temperature than LOCTIC, but the sump temperature is higher. The RSHX duty increases and the sump temperatures converge after 1000 seconds. This containment response is consistent with the DEHLG model comparison in Section 4.3. However, the GOTHIC depressurization time is shorter and the subatmospheric peak pressure is less than LOCTIC. The difference in long-term containment pressure is explained by the GOTHIC post-reflood break energy distribution in Section 4.4.2.

4.4.2 DEPSG Mass and Energy Release Westinghouse mass and energy release data is used up to 200 seconds, the end of reflood for the DEPSG break with minimum SI flow. At that time, the GOTHIC RCS model is activated with initial conditions that are consistent with the Westinghouse mass and energy distribution in WCAP-14083 [15], which used the NRC-approved FROTH analysis methodology [14, 16] to calculate the post-reflood mass and energy release rates. The LOCTIC analysis modifies the FROTH mass flow rate by adjusting for differences in the SI flow rates versus those assumed by Westinghouse (bounding maximum flow rates are used so that the FROTH analysis does not have to be repeated if system improvements are realized). Thus, the FROTH/LOCTIC methodology for DEPSG breaks is the comparison standard in this section.

Figures 4.4-4 and 4.4-5 compare the GOTHIC integral mass and energy releases to LOCTIC.

Table 4.4-2 shows that the GOTHIC integral energy release to the containment is about 1%larger and the integral mass is very close to LOCTIC at 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. Table 4.4-2 also compares the integral energy at the end of reflood and at the time that GOTHIC predicts subatmospheric conditions (2201 seconds). Figure 4.4-6 shows the SG secondary liquid temperatures drop quickly with the Topical Report DOM-NAE-3, Rev. 0.0-APae5 Page 57

containment depressurization. The GOTHIC primary system energy release is more conservative but the distribution of the energy requires further discussion.

The GOTHIC depressurization time is less than LOCTIC because of difference in modeling the core exit steam distribution, which in turn affects the break energy distribution between liquid and vapor. The FROTH analysis methodology that generated the post-reflood data for LOCTIC assumes 95% of the core exit steam enters the broken loop SG [14]. This non-mechanistic assumption results in a higher mass release of superheated steam from the broken ioop SG than GOTHIC. FROTH still removes all of the intact loop SG secondary energy in 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />, effectively maintaining a large amount of liquid entrainmient into the intact SGs with only 5% of the core exit steam flow. As a result, the LOCTIC containment response using the FROTH inputs takes longer to deposit the additional superheat mass in the sump and depressurize the containment.

In contrast , the GOTHIC simplified RCS model is mechanistic, using hydraulics to determine the steam flow split to the intact and broken SG loops. The GOTHIC steam velocity determines the amount of liquid entrainment into the SG tubes. While biasing the loss coefficients in the intactI loop hot legs to force more steam into the broken loop SG will increase the superheated steam release,' the reduced liquid entrainment into the intact loop SGs will slow the secondary energy removal rate. In the aggregate, the total break energy from this bias is less than the amount when GOTHIC calculates the flow split and carries liquid into the intact loop SG tubes. A GOTHIC sensitivity case with a large loss coefficient in the intact hot legs confirmed this conclusion. The secondary side temperature on the intact loop SGs decreases very slowly. The integral energy release at 3600 seconds is 640.1 MBtu (compared to 683.3 MBtu) and the containment becomes subatmospheric 200 seconds earlier (-2000 seconds). In conclusion, it is conservative to use the GOTHIC hydraulic model (i.e., no bias on the steam flow), such that the secondary energy is quickly removed from all of the SGs during the system depressurization.

The GOTHIC subatmospheric peak pressure occurs after the CS pump, which sprays 45 F water,I is stopped on low RWST level at 4324 seconds. The RS system continues to spray warmer water (sump water passed through the RS heat exchangers) and the containment pressure increases until a peak occurs and the RS system reaches an equilibrium with the core decay heat that is spilled to the sump. The GOTHIC subatmospheric peak pressure is less severe than LOCTIC for two reasons. First, the containment pressure when the CS pumps stop is. about 2 psi lower than LOCTIC because of the aforementioned distribution of break energy between liquid and vapor.-

Second, as the containment pressure begins to rise after CS termination, the thermal conductors in the primary and secondary systems can absorb energy. In contrast, LOCTIC only discharges energy from the primary and secondary systems if pressure is decreasing during the post-reflood phase. Once the CS system stops and containment pressure starts to rise, LOCTIC does not have a mechanism to add energy into the primary system (i.e., no thermal conductors). Adding this energy back to the primary system is physically realistic and therefore appropriate.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 58

4.4.3 Summary of Containment Depressurization Comparison The GOTHIC containment response shows similar behavior to LOCTIC, with differences in the magnitude of pressures and temperatures. In the short-term, the lower peak pressure and higher sump temperatures are attributed to the droplet phase, the DirectIDLM condensation and break effluent models. In the long-term, GOTHIC's lower containment pressure is attributed to the smaller superheated steam flow rate from the broken. loop SG compared to the non-mechanistic Westinghouse analysis. However, the GOTHIC DEPSG model removes the energy in the primary and secondary systems and results in a more conservative energy release than LOCTIC (see Table 4.4-2).

Table 4.4-1: Sequence of Events for Containment Depressurization Analysis Event (seconds) GOTHIC LOCTIC High containment pressure reached to actuate CLS 2.38 2.3 Peak pressure occurs 20.0 19.4 Safety injection actuates 22.6 22.6 Containment spray actuates 99.4 99.4 IRS pump spray becomes effective' 223.5 216 ORS pump spray becomes effective' 415.9 415 Containment pressure reaches 14.7 psia 2221 2820 Switchover to SI recirculation mode 3699 3750 Containment spray terminates (low RWST level) 4324 4370 Subatmospheric peak pressure occurs 5500 5510

____________________________________ (-2.22 psig) (-047 psig)

1) Effective times include pump start delays and piping fill times.

Table 4.4-2: Comparison of DEPSG Break Mass and Energy GOTHIC LOCTIC Integral energy release at end of reflood (200 seconds), MBtu 313.9 313.3 Integral energy release at 2200 seconds, MBtu 599.5 595.9 Integral energy release at I hour, MBtu 683.3 673.4 Integral mass release at I. hour, Mlbm 2.097 2.10 Topical Report DOM-NAF-3, Rev. 0.0-APae5 Page 59

Figure 4.4-1: Containment Pressure for Containment Depressurization Containment Pressure PRI DC25T 10 100 1000 1le+04 Time sec rOTHIC 7.2dOnnIPA! Sanlý28121MS 15-51-29 Figure 4.4-2: Containment Vapor Temperature for Containment Depressurization 3

E) a)

F-10 le+04 Time (sec)

L~JI Fb~ ~ OCtXCV~ZAJJ IL~.fl Ca Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 60

Figure 4.4-3: Containment Liquid Temperature for Containment Depressurization 2 Containment Liquid Temperature TL1 DC26T 0- -

Cl) 0........................... ................... .......... .......

I.............

........... ... ........................

0.1 1 10 100 1000 le+04 Time (See)

GOTHIC 7.2donflGA) Sept28lOOS 15-51:29 Figure 4.4-4: DEPNG Integrated Mass for Containment Depressurization Total Break Mass ED CV50G DCl7TT 0

0.x 0 10 10 e0

...............

.... .......). i.m.

GO....... S ~ 12 ----------- .......

Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 61

Figure 4.4-5: DEPNG Integrated Energy for Containment Depressurization S2 Total Break Energy cv45C DC34T U')

CD C

,eW M

10 100 1000 le+04 Time (sec)

GOTHIC 7,2domfOA) Sei/28O00 15:51:2S Figure 4.4-6: DEPSG SG Secondary Temperatures for Containment Depressurization fso S.G. Secondary SieTmeaue

. . . . ............

5 ....

....

CL 0)

E 0 0*i

)1 2 3 4 5 (cli~kt~

Time (see) 1e3 GOTHIC,7 2dL1rn!OA 5ý1 t"0515:51:29 Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 62

4.5 GOTHIC Analysis of LHSI Pump NIPSH Available A GOTHIC calculation of LHSI pump NPSHa is compared to the LOCTIC analysis from the Surry IJFSAR for a DEPSG break with one train of safeguards and maximum SI flow. The minimum NPSHa occurs at recirculation mode transfer (RMT), when the LHSI pump swaps suction from the RWST to the containment sump. After RMT, NPSHa increases as the containment pressure stabilizes and the sump temperature decreases from the RS heat exchangers removing energy. Thus, it is important that.

the primary and secondary system energy be removed at a high rate to maximize the sump temperature before RMT. The DEPSG model for containment depressurization from Section 4.4 was biased in accordance with Section 3.8.2 to minimize NPSHa. The spray nozzle drop diameter was reduced by a factor of 10 (which produced the same minimum NPSH as the method specified in Section 3.8.2), the nozzle spray flow fraction was set to 1.0, a multiplier of 1.2 was applied to the conductor heat transfer coefficients, and the upper limnit on the containment free volume was used.

The containment initial conditions and design inputs were the same as the LOCTIC analysis.

Water holdup was excluded because it was not part of the LOCTIC analysis.

4.5.1 Containment Response Table 4.5-1 compares the sequence of events and Table 4.5-2 compares the results at the time of minimum NPSHa. Figures 4.5-1 through 4.5-4, compare the containment pressure, vapor temperature, liquid temperature, and sump level to LOCTIC results shown as discrete points. The distribution of the energy release 'into containment is indicated by the containment pressure and temperature response. During the early part of the event (<1000 sec), the GOTHIC sump liquid temperature is considerably less than LOCTIC, the Vapor temperature is slightly higher, and the pressure is higher. The LOCTIC pressure flash option models the break liquid as a continuous liquid addition to the sump. GOTHIC break modeling using droplets results in a different containment energy distribution. In general, the LOCTIC pressure flash option causes a very conservative amount of energy to be retained in the sump liquid with less vapor flashed into the air space. This is evident from the very high (> 250 F) LOCTIC sump temperatures that are maintained until almost 1000 seconds even while the RS heat exchangers are removing sump energy. The vapor temperature is slightly less than the GOTHIC values. LOCTIC assumes no interfacial heat transfer between the sump pool and containment atmosphere, which also explains the high liquid temperatures.

For the first few seconds, the LOCTIC vapor temperatures are much higher than GOTHIC. This is due to the lack of a droplet model in LOCTIC, which results in a brief period of superheat.

Once the IRS and ORS pumps become effective (200-400 seconds into the event) and the sump liquid is sprayed into the containment, the difference between the model responses becomes less noticeable. At the time of RMvT, the GOTHIC sump liquid temperature is about 1 F higher than LOCTIC and the pressure is about 0.7 psi higher. The higher sump temperature provides a relative adverse effect on NPSHa while the increased pressure is a benefit. The sump levels in Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 63

Figure 4.5-4 is very close, with GOTHIC slightly lower (4.12 ft vs. 4.2 ft) at RMT. The net result is that the GOTHIC minimum NPSHa is about 1.4 ft higher than the LOCTIC value.

4.5.2 DEPSG Mass and Energy ReleaseI The DEPSG model from Section 4.4.2 is used with thermal equilibrium in the broken loop cold leg using a liquid/vapor interface area of 1E+08 ft2 consistent with Section 3.5.3.3.1. This promotes thermal equilibrium between any vapor from the downeomer and the SI added to that cold leg, which produces elevated sump temperatures. The SI flow is split based on the plant configuration for flow to the downcomer (for the intact cold legs) and the broken loop cold leg.

Figures 4.5-5 and 4.5-6 show a good comparison of the integral mass and energy releases over the entire transient, with GOTHIC values about 0.5% higher at the time of RMTT (Table 4.5-2).

The SG secondary fluid temperatures in Figure 4.5-7 decrease rapidly early in the event as the vessel level swell model causes liquid to rise into the SG tubes, drawing energy from the SG secondary side fluid. At RMT, the SG secondary side temperatures are approximately 200 F.I Similarly, the primary side metal in Figure.4.5-8 follows the reactor vessel fluid temperature to a minimum of about 213 F at RMT before increasing from the increases in containment pressure and temperature after CS termination. The effect of the large liquid/vapor interface area in the downcomer and broken cold leg volumes is seen in Figure 4.5-9. Although superheated steam is delivered to the downcomer (volume 23) from the intact cold legs (volume 22), the downeomer liquid and vapor phases are in equilibrium. A similar effect occurs in the broken loop cold leg volume, except the temperatures are lower due to mixing with additional SI flow. In conclusion,I the simplified RCS model appropriately and conservatively removes the primary and secondary stored energy before RMT. In addition, the complete mixing that occurs in the downcomer and broken cold leg volumes ensures that the liquid discharged to containment is at the highest (most conservative) temperature.

Some of the primary system volumes demonstrate oscillatory beha vior, such as the temperatures shown in Figures 4.5-8 and 4.5-9. Oscillations in liquid flow are caused by oscillatory phase.

change, most likely in the, steam generators, which causes pressure perturbations throughout the primary system and corresponding flow oscillation. The oscillations are similar to those observed in the FLECHT SEASET tests, which had similar system depressurization and coo ling [36].

Liquid temperatures exhibit this type of behavior as a result of primary system conductors going in and out of boiling heat transfer mode in response to fluctuating pressures. The result is swingsI in the heat transfer coefficient and heat flux, which affect both the conductor surface and liquid temperatures.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 64

4.5.3 Summnary of LHSI Pump NPSHa Comparison The GOTHIC comparison case shows good agreement with the corresponding LOCTIC case.

The simplified RCS model has removed all of the SG secondary side energy when the vessel and SGs are frilly depressurized. The GOTHIC integrated mass and energy release into containment at the time of minimum NPSH is actually slightly higher. The primary difference between the two cases is due to the LOCTIC pressure flash option, which determines how break energy is partitioned between the containment liquid and vapor regions. The LOCTIC treatment of the break liquid as a continuous liquid forces more of the break liquid energy to be deposited in the containment sump with less flashing of vapor into the air space. GOTHIC uses more realistic models for the treatment of the break releases with some of the liquid being dispersed as droplets in the vapor space. It also allows for mass and heat transfer between the sump pool and air space.

These differences are much more pronounced early in the event, but become less noticeable as the vapor and liquid regions are mixed via the operation of sprays. The more realistic GOTHIC modeling of the RCS and SGs results in slightly more energy being transferred to the containment at the timne the LHSI pumps take suction from the sump. At the time of minimum NPSHa, the GOTHIC sump temperature is actually slightly higher than the LOCTIC value; however, the GOTHIC pressure is also higher, yielding a small, net increase in NPSHa. The higher sump temperature and containment pressure than LOCTIC is consistent with the additional energy addition from the RCS model, and is considered to be a reasonable and more accurate system response.

Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 65

I1 Table 4.5-1: Sequence of Events for LHSI Pump NPSHa Analysis Event (seconds) GOTHIC LOCTIC High containment pressure reached to actuate CLS 2.7 2.3 Peak pressure occurs 19.8 19.4 Safety injection actuates 22.6 22.6 Containment spray actuation 99.7 99.3 IRS pump spray becomes effective' 223.6 214.3 ORS pump spray becomes effective' 420.8 411.3 Switchover to SI recirculation mode transfer (RMT) 3230 3240

1) Effective times include pump start delays and pipe fill times.

Table 4.5-2: GOTHIC Comparison to LOCTIC for LHSI Pump NPSHa GOTHIC LOCTIC Time of SI recirculation mode transfer (RMT), sec 3230 3240 LHSI pump NPSH available, ft 18.4 17.0 Containment pressure, psia 10.57 9.89 Sump liquid temperature, F 161.1 160.1 Containment vapor temperature, F 111.3 98.1 Sump liquid level, ft 4.12 4.2 I Integral energy release at RMT, MBtu 674.2 670.7 Integral mass release at RMT, Mlbm 2.134 2.120 I I

I I

I I

I Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page66 I

Figure 4.5-1: Containment Pressure - LHSI Pump NPSH PRi1 D025T to ............

.......... . .......... ...............

........

.................

... ............

0~

03 U) to ............ .............

U) N~ .................................. ......................

2 0.

---

...................................... ....... ..... ................

Ai i ý i x L ; i -&i 3 j i 1 . . . .. . . ý o 1 10 100 1000 le+04 Time sec Figure 4.5-2: Containment Vapor Temperature - LUSI Pump N-PSH 1-VI DC39T 0O to __ __ _ _

N 0

E 0

U'.,

0.1 1 10 100 1000 1et04 Time (see)

Topical Report DOM-NAE-3, Rev. 0.0-APae6 Page 67

Figure 4.5-3: Containment Liquid Temperature - LHSI Pump NPSH TL1 DC26T o-N E

0 01 LO L

0.1 10 1 1e+04 Time (Sec)

I Figure 4.5-4: Containment Sump Level - LHSI Pump NPSH I LL1 DC42T I I

.................

-J M.

N CO 10 1e+04 Time (sac)

I Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 68 I

Figure 4.5-5: Integral Break Mass to Containment - LHSI Pump NIPSH cv500 DC40T x

............

........ ..... ...

cli

..

...

...

...

......

..

...

........ .............. . . *.

CO 0

0.1 10 100 1000 le+04 Time (sac)

Figure 4.5-6: Integral Break Energy to Containment - LHSI Pump NPSH cv45C DC34T x

If)

OS If) 0 5D U*)

N 0

Time (sec)

Topical Report DOM-NAF-3, Rev. 0.0-APae6 Page 69

Figure 4.5-7: SG Secondary Side Liquid and Vapor Temperatures - LHSI Pump NPSH (Volume 18 =Intact Loops, Volume 21 = Broken Loop) 1O18 TL21 TV1 8 TV21 0

.. . . . .. . . .. . . . . . . .. . . .. . . . .. . . . . . . .

C'L)

E Q

0 1-Time (sec) Xle3 Figure 4.5-8: Primary Metal (TA22) and Reactor Vessel Liquid Temperatures (TL15s1) - LHSI Pump NPSH TA22 TL15sl 0l - -

C> _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _

E 0.001 1.001 2.001 3 4 Time (Sec) Xe Topical Report DOM-NAE-3, Rev. 0.0-A Page 70

Figure 4.5-9: Intact Cold Legs (Volume 22) and Downcomer (Volume 23) Temperatures LHSI Pump NPSH TL23 TV23 TV22 0-0 E LO 0 0.8 1.6 2.4 3.2 4 X~a3 t iffe ikSeiG Topical Report DOM-NAE-3, Rev. 0.0-APae7 Page 71

4.6 GOTHI1C Analysis of MSLB Event Surry does not have an explicit MSLB containment response analysis in the UFSAR. However, an explicit analysis will be performned for Surry as part of the plant-specific implementation of this topicalI report. The North Anna UFSAR includes MSLB containment response analyses using LOCTIC. The North Anna MSLB3 mass and energy release data from Reference 32 was confirmed to be conservative for Sunry. This section describes MSLB containment response analyses performned with the Surry GOTHIC containment model and North Anna mass and enthalpy d ata.

Two cases were analyzed using data from a 1.4 ft2 break at 102% power because this break produces a superheated containment atmosphere and the benefits of the GOTHIC DLM condensation option can be demonstrated. The first case demonstrates the use of the GOTHIC modeling assumptions in Section 3. Figures 4.6-1 and 4.6-2 show the containment pressure and temperature predictions for this case. The atmosphere remains superheated for a very short time, returning to saturation within 10 seconds from the time of the break. The containment pressure peaks -200 seconds when the faulted SG reaches dryout and the mass release rate matches the AEW addition rate.

• The second case incorporated two changes to GOTHIC to mimfic LOCTIC and compares the Surry containment response to a North Anna UFSAR LOCTIC analysis. First, the droplet diameter was set to none for the break boundary condition to mimic the LOCTIC pressure flash model and force all break liquid to enter containment in the continuous liquid phase. Second, the condensation option was changed from DLM to UCHIDA for the containment heat sinks. North Anna has a larger I

containment free volume and a core power (2893 MWt versus 2546 MWt) than Surry. Because the Surry GOTHIC model (with different heat sinks, free volume, and spray flows compared to North Anna) was used, the mass release was reduced by the ratio of core power. The intent was to compare behavior to a North Anna LOCTIC analysis. Figures 4.6-3 and 4.6-4 compare theI containment pressure and temperature to the LOCTIC data shown with points. Simulating the LOCTIC pressure flash model allows the atmosphere to remain superheated longer (-200 seconds) and to reach a much higher peak temperature.

Once this difference between the codes was understood, the Surry GOTHIC model was run with North Anna mass and energy release data from seven different combinations of break size and initial power level. Break size ranged from split breaks to the maximum 1.4 ft2 applicable to Surry3 and North Anna. Power level ranged from 0% (limiting for peak pressure due to the larger SG liquid mass) to 102% power (limiting for superheat). The comparison was done to validate the GOTHIC response to the range of break conditions. Figures 4.6-5 and 4.6-6 compare theI containment pressure and temperature for all seven cases. The trends are consistent with the North Anna LOCTIC analysis results. As described in Section 3, GOTHIC MSLB analyses will use the DLM condensation option and the break droplet model. This section demonstrates that the GOTHIC modeling over the range of MSLB conditions is acceptable.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 72

Figure 4.6-1: MSLB Containment Pressure with GOTHIC Models 1 Containment Pressure PRI W ~ __

co.... .........

0.1.

GOHI 7.dmO)J0 31201:42 10 100 1000 le+004 Time sec Figure 4.6-2: MSLB Containment Vapor Temperature with GOTHIC Models 3 Containment Vapor Temperature TV1. STi 00 00 E. . ....... .............. ..........................

..............

0 0.1 1 10 t00 1000 1e+004 Time (sec)

GOTHIC 7.2dom(OA) Jari/31/2005 11:04:23 Topical Report DOM-NAF-3, Rev. 0.0-APae3 Page 73

Figure 4.6-3: MSLB Containment Pressure using Pressure Flash Assumptions Containment Pressure U)

0) co.............

.. L............

0.11 10 100 1000 le+004 Time sec GOTHIC 7.2dom(OA) Jan/31/2005 09:02:21 Figure 4.6-4: MSLB Containment Vapor Temperature using Pressure Flash Assumptions 3

....................

0) C\J E 0 .................

I-I

.t...............

0.11 10 100 1000 le+004 Time (sec)

GOTHIC 7.2dom(OA) Jant/31/200509:02:21 Topical Report DOM-NAE-3, Rev. 0.0-APae7 Page 74

Figure 4.6-5: Comparison of Containment Pressure for MSLB Spectrum 70 60 50 S40 30 20 10 0 50 100 150 200 250 300 350 400 450 500 Time, sec Topical Report DOM-NAF-3, Rev. 0.0-APae7 Page 75

Figure 4.6-6: Comparison of Containment Vapor Temperature for MSLB Spectrum 320 300 280 260 240 220 200 180 160 140 0 50 100 150 200 250 300 350 400 450 500 Time, sec Topical Report DOM-NAF-3, Rev. 0.0-A Page 76 M-- M M- M M M M-- M M M M M M-M M

4.7 Sensitivity Studies The conservative assumption for a particular analysis depends on the design requirement that is being verified. Sensitivity studies will be performed for break locations, single failures, and design inputs for each plant-specific GOTHIC containment analysis. Table 4.7-1 documents the results of the studies for Surry's containiment analysis criteria. The conclusions are consistent with the current LOCTIC analyses with the exception of the limiting single failure for the calculation of NPSHa for the ORS and IRS pumps. With LO CTIC, the minimumn NPSHa for the ORS and IRS pumps occurs for a case with full engineered safeguards (no single failure). The GOTHIC analyses produce the same minimum NPSHa for the full safeguards case and for other cases with single failures, which emphasizes the need to analyze the single failures for each design effort.

Table 4.7-1 illustrates the breadth of sensitivity analyses that were. performed for Surry to confirm the lim-iting assumptions for the current plant configuration. The results are specific to Surry's current configuration and are not intended to cover all Dominion PWRs, since each station has specific design criteria and engineered safety features that require sensitivity studies. Dominion will perform similar sensitivity studies to define the set of conservative assumptions for each PAIR application.

4.8 Summnary of Demonstration Analyses Based on the comparison to LOCTIC, it is concluded that the GOTHIC model selections identified in Section 3 appropriately model the containment response for LOCA and MSLB events. GOTHIC shows similar behavior for containment pressure and temperature to the SWEC LOCTIC code for a DEHLG break with maximum safeguards and a DEPSG break for containment depressurization and LHSI pump NPSHa. GOTHIC predicts lower peak containment pressures because of the DLM condensation model and the break droplet model. The GOTHIC liquid temperature is higher in the short-term, but the RS heat exchangers and the interfacial heat and mass transfer in GOTHIC bring the vapor and liquid phase temperatures close together.

GOTHIC predicts shorter depressurization times because of the simplified RCS model that mechanistically removes energy from all steam generators, while the FROTH methodology non-mechanistically biases superheated steam flow through the broken loop steam generator. For the LHSI pump NPSHa analysis, GOTHIC predicts a slightly higher sump temperature and containment pressure at the time of minimum of NPSHa. Overall, the long-term containment response is .comparable to LOCTIC. The analyses also demonstrate that the simplified RCS model is conservative for calculating post-reflood mass and energy release rates for both DEPSG and DEHLG breaks.

Topical Report DOM-NAF-3, Rev. 0.0-APae7 Page77

Table 4.7-1: Matrix of Conservative Inputs for Surry Demonstration GOTHIC Containment Analyses Note: This table is based on the current plant configuration. Plant modifications can change these results.

Table Key (also refer to the List of Acronyms and Abbreviations)

Min= Assume the minimum value for the range of the design input Max =Assume the maximum value for the range of the design input N/A =Not Applicable: the key analysis result occurs after this parameter becomes effective or the component is not part of the containment response (e.g., accumulator nitrogen does not discharge for MSLB).

N/S = Not Sensitive: the key analysis result is not sensitive to changes in this 'input parameter.

LOCA Peak j MSLB Peak Containment Subatmospheric LHSI Pump 1ORS Pump IRS Pump Pressure* jPressure/Temp # JDepressurization [_Peak Pressure NPSH ] NPSH J NPSH General Break Type DEHLG 1.4 ft2 for pressure DEPSG DEPSG DEPSG DEHLG DEHLG 2

0.6 ft for temp #

Reactor Power 102% 0% for pressure 102% 102% t02% 102% 102%

102% for temp #

Single Failure N/A 1 emergency bus 1 emergency bus 1 emergency bus 1 emergency None &None&

bus Containment Air Pressure Max Max / Min # Max Max Min Min Min Temperature Max Max Max Min Max Max Max Relative Humidity 100% 100%!/ 0% # 100% 100% 100% 100% 100%

Free Volume Min Min Min Min Max Max Max Heat Sink Surface Area Min Min Min Max Min Min Min Topical Report DOM-NAF-3, Rev. 0.0-A Page 78

- --- --- --------- -- -

LOCA Peak MSLB Peak IContainment Subatmospheric LHSI Pump ORS Pump IRS Pump Pressure* Pressure/Temp # IDepressurization Peak Pressure NPSH , NPSH NPSH Safety Injection HNSI Injection Flow Rate N/A N/S Min Max Max Min Min LHSI Injection Flow Rate N/A N/S Min Max Max Min Min LHSI Recirc Flow Rate N/A N/A Min Max Max N/A N/A LHSI Suction Piping N/A N/A N/S N/S Max N/S N/S Friction Loss Accumulator Nitrogen N/A N/A Max Max Min Min Min Pressure Accumulator Nitrogen N/A N/A Max Max Min Min Min Volume Accumulator Nitrogen N/A N/A Min Min Max Max Max Temperature RWST Temperature N/A Max Max Max Max Max Max Initial RWST Level N/A N/S Min Min Min Min Min SI Recirc Mode Transfer N/A N/A Late Late Early N/A N/A Containment Spray CS Flow Rate N/A Min Min Min Max Max Max CS Start Time N/A Max Max Max Max Min Min Bleed Flow to ORS Pump N/A N/S N/S N/S N/S Min Min Suction Topical Report DOM-NAF-3, Rev. 0.0-APae7 Page 79

LOCA Peak [ MSLB Peak Containment JSubatmospheric LHSI Pump ORS Pump TIRS Pump Pressure* Pressure/Temp # Depressurization jPeak Pressure NPSH NPSH NPSH Recirculation Spray RS Piping Volume N/A N/S Max Max N/S Min Min IRS Flow Rate N/A N/S Min Min Min Min Max ORS Flow Rate N/A N/S Min Min Min Max Min MRS Recirculation Flow to N/A N/S N/S N/S N/S Min Min Pump Suction RS Timer Delay N/A N/S Max Max Max Max Max IRS Suction Loss N/S N/S N/S N/S N/S Max Max ORS Suction Loss N/S N/S N/S N/S N/S Max Max Service Water SW Flow Rate N/A N/S Min Min Min Max Max SW Temperature N/A N/S Max Max Min Min Min HX Tube Plugging/Fouling N/A N/S Max Max Max 0 0

• LOCA peak pressure and temperature assumptions are the same since a saturated containment environment is maintained.

1. MSLB peak temperature occurs for small breaks and the spectrum is reviewed for any plant. operating parameter changes. The peak temperature is obtained by using minimum air pressure and 0% humidity (peak pressure cases assume maximum air pressure and 100% humidity).

& Sensitivity studies have shown that the full ESF case (no single failure) produces the same minimum NPSH as many single failure scenarios.

Design studies must evaluate single failure scenarios with the full ESF case.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 80

5.0 Conclusions Dominion has developed a containment analysis methodology using the GOTHIC computer code for application to large, dry PWVR containments. The GOTHIC model selections and techniques for the containment parameters (e.g., DLM condensation, lumped containment volume) have been approved previously by the NRC for containment analysis licensing calculations [8-13] and are specified in Section 3. Section 4 demonstrates that the GOTHIC containment modeling selections provide a reasonable comparison to the LOCTIC analyses for Surry Power Station and that some margin in containment peak pressure is gained with justification.

Dominion has developed a mass and energy release model for the post-reflood phase that couples the primary system and secondary system stored energy depletion to the containment pressure response.

The DEHILG break model for Surry was shown to provide more conservative energy releases than the NRC-approved Westinghouse methodology in WCAP-8264-P-A [14]. The DEPSG break model for Surry was compared to the NRC-approved Westinghouse FROTH methodology [14, 16] and was shown to provide as conservative mass and energy release rates. In addition, the timing of the GOTHIC energy release was consistent with the need to remove the SG secondary side energy in order to maximize containment depressurization or sump temperature, depending on the accident acceptance criteria of concern. Because of the complex model and plant-specific inputs requirements, Dominion will benchmark each new plant application of the post-reflood mass and energy methodology to ensure that the mass and energy release is as conservative as the plant's existing NRC-approved calculation.

In conclusion, the GOTHIC containment analysis methodology described in this report ensures a conservative calculation of the containment response for the containment analysis acceptance criteria listed in Section 2. Dominion plans to reference this analysis methodology for plant-specific license amendments starting in December 2005.

Topical Report DOM-NAE-3, Rev. 0.0-APae8 Page 81

6.0 References

1. NMI 8907-06, Revision 15, "GOTIHIC Containment Analysis Package Technical Manual, Version 7.2," published by EPRI, September 2004.

2. NMI 8907-02, Revision 16, "GOTHI-C Containment Analysis Package User Manual, Version 7.2,"

published by EPRI, September 2004.I

3. NMI 8907-09, Revision 8, "GOTIHIC Containment Analysis Package Qualification-Report, Version 7.2," published by EPRI, September 2004.

4. NRC Generic Letter 83-11, Supplement 1, "Licensee Qualifications for Performing, Safety Analysis," June 24, 1999.

5. Letter from David A. Christian (VEPCO) to NRC, "Virginia Electric and Power, North Anna Power Station Units 1 and 2, Sunry Power Station Units 1 and 2, Qualifications for Performing Safety Analyses, Generic Letter 83-11, Supplement 1," Serial No.00-087, March 15, 2000.

6. Letter from David A. Christian (VEPCO) to NRC, "Virginia Electric and Power, North Anna Power Station Units 1 and 2, Surry Power Station Units 1 and 2, Response to Request for Additional Information, Dominion's Reload Nuclear Design Methodology Topical Report," Serial No.02-280, May 13, 2002.

7. Letter from Scott Moore (NRC) to David A. Christian (VEPCO), "Virginia Electric and Power Company - Acceptance of Topical Report VEP.-FRD-42, Revision 2, 'Reload Nuclear Design Methodology,' North Anna and Surry Power Stations, Units 1 and 2," June 11, 2003.

8. Letter from Anthony C. McMurtray (NRC) to Thomas Coutu (NMC), "Kewaunee'.Nuclear Power Plant - Issuance of Amendment (TAC NO. MB6408)," September 29, 2003.

9. Letter from John G. Lamb (NRC) to Thomas Coutu (NMC), "Kewaunee Nuclear Power Plant -

Issuance of Amendment Regarding Stretch Power Uprate (TAC NO. MB 903 1)," February 27,I 2004.

10. Letter from Alan B. Wang (NRC) to R.T. Ridenoure (OPPD), "Fort Calhoun Station, Unit No. 1 -

Issuance of Amendment (TAC NO. M1B7496)," November 5, 2003.

Topical Report DOM-NAF-3, Rev. 0.0-A Page 82

11. Letter from L. Mark Padovan (NRC) to D.N. Morey (Southern Nuclear Operating Company),

"Joseph M. Farlcy Nuclear Plant, Units 1 and 2 - Issuance of Amendments re: Steam Generator Replacements (TAC Nos. MA4393 AND MA4394)," December 29, 1999.

12. Letter from Frank Rinaldi (NRC) to J.T. Gasser (Southern Nuclear Operating Company), "Vogtle Electric Generating Plant, Units 1 and 2 Re; Issuance of Amendments (TAG Nos. NMB5046 AND MB5047)," June 4, 2003.

13. Letter from M.S. Tuckman (Duke Power Company) to the NRC transmitting approved version of Topical Report DPC-NE-3004-P-A, Revision 1, "Mass and Energy Release and Containment Response Methodology," dated December 18, 2000.

14. WCAP-8264-P-A, Rev. 1, "Westinghouse Mass and Energy Release Data for Containment Design," August 1975. (WCAP-83 12-A is the Non-Proprietary version).

15. WCAP- 14083, Revision 0, "Virginia Power Surry Power Station Units 1 and 2 Contaimnment LOCA Mass and Energy Release Analyses for Core Uprating Engineering Report," May 1994.

16. WCAP-10325-P-A, "Westinghouse LOCA Mass and Energy Release Model for Containment Design 7 March 1979 Version," May 1983. (WCAP-10326-A is the Non-Proprietary version.)

17. Schlunder, E. (Ed.) "Heat Exchanger Design Handbook," Hemisphere Publishing, 1983.

18. Marx, K. D., "Air Currents Driven by Sprays in Reactor Containment Buildings", Sandia Report SAND84-8258, NIJREG/CR-4102, May 1986.

19. Brown, R., and York, J.L., "Sprays Formed by Flashing Liquid Jets", AIChE Journal Volume 8, Number 2, May 1962.

20. Letter from L. William Pearce (FENOC) to NRC, "Beaver Valley Power Station, Un~it No. 1 and No. 2, BV-1 Docket No. 50-334, License No. DPR-66, BV-2 Docket No. 50-412, License No.

NPF-73, License Amendment Request Nos. 317 and 190," June 2, 2004. (Enclosure 2 documents the IVIAAP topical report and analyses).

,21. WCAP- 16219-NP, "Development and Qualification of a GOTHC Containment Evaluation Model for the Prairie Island Nuclear Generating Plants," March 2004, submitted as Exhibit D in letter Serial No. L-PI-04-017 from Joseph M. Solymossy (Nuclear Management Company) to NRC, "License Amendment Request (LAR), Request for Use of GOTHIC 7 in Containment Response Analyses." September 1, 2004.

Topical Report DOM-NAF-3, Rev. 0.0-APae8 Page 83

22. NUJREG-0588, Revision 1, "Interim Staff Position on Environmental Qualification of Safety Related Electrical Equipment", November 1980.

23. Ishii, M., "One-Dimensional Drift-Flux Model and Constitutive Equations for Relative Motion Between Phases in Various Two-Phase Flow Regimes", ANL-77-47, October 1977.

24. Spillman, J. J., "Evaporation from Free Falling Droplets", Aeronautical J, 1200:5, pp. 181-185, 1984.I

25. Parsly, L. F. "Design Considerations of Reactor C ontainment Spray Systems - Part VI, The Heating of Spray Drops in Air-Steam Atmospheres," ORNL-TM-24 12, January 1970.

26. Pruppacher, H. R., and Klett, J. D., "Microphysics of Clouds and Precipitation", D. ReidelI Publishing Co., Boston, 1978.

27.NpdtedFinl orh nnaPowrafeytatonAalsisReprt

27. North nnPower Station Updated Final Safety Analysis Report.

29. ANSI/ANS-5.1-1979, "American National Standa rd for Decay Heat Power in Light-Water Reactors," August 1979.

30. Abdelghany, J. M., et al., "Analysis of Containm-ent Response to Postulated Pipe Ruptures UsingI GOTHIC," Framatomne ANP report BAW-10252(NP), Revision 0, July 2004.

3 1. Letter from Herbert N. Berkow (NRC) to Ronnie L. Gardner (Framatome), "Final Safety Evaluation for Framatome ANP Topical Report BAW- 10252(P), Revision 0, 'Analysis of Containment Response to Postulated Pipe Ruptures Using GOTHIC,' (TAC No. MC3783),"

August 31, 2005.

32. WCAP- 1143 1, Revision 0, "Mass and Energy Releases Following a Steam Line Rupture for North Anna Units 1 and 2," February 1987.

33. WCAP-8822, Revision 0, "Mass and Energy Releases Following a Steam Line Rupture,"

September 1976, with Supplements l and 2 dated September 1986.

34. Cunningham, J. P. and Yeh, H. C., "Experiments and Void Correlation for PWR Small Break LOCA Conditions", ANS Transactions, Vol. 17, 1973, pp. 3 69-70.I Topical Report DOM-NAF-3, Rev. 0.0-A Page 84

35. Lilly, G.P. and Hoebreiter, L.E., "Mixfing of Emergency Core Cooling Water with Steam: 1/3 Scale Test and Summary," EPRI 294-2, Electric Power Research Institute, June 1975.

36. Hochreiter, L.E., et al., "PWvR FLECHT SEASET Systems-Effects Natural Circulation and Reflux Condensation Data Evaluation and Analysis Report," WCAP-10415, FLECHT SEASET Program Report No. 14, February 1985.

37. NUREG-0800, "Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants," US Nuclear Regulatory Commission.

38. Letter from J. P. O'Hanlon (VEPCO) to USNRC, "Virginia Electric and Power Company, North Anna and Surry Power Stations Units 1 and 2, Generic Letter 97 Assurance of Net Positive Suction Head for Emergency Core Cooling and Containment Heat Removal Pumps; Response to a Request for Additional Informnation," Serial No.98-546, October 29, 1998.

39. Letter from N. Kalyanam (USNRC) to J. P. O'Hanlon (VEPCO), "Completion of Licensing Action for Generic Letter 97-04, 'Assurance of Sufficient Net Positive Suction Head for Emergency Core Cooling and Containment Heat Removal Pumps'; North Anna Power Station, Unit Nos. 1 and 2 (TAC Nos. XMAOO15 and MIAOO 16)," February 25, 1999.

40. Letter from G. E. Edison (USNRC) to J. P. O'Hanlon (VEPCO), "Completion of Licensing Action for Generic Letter 97-04, 'Assurance of Sufficient. Net Positive Suction Head for Emergency Core Cooling and Containment Heat Removal. Pumps'; Sunry Power Station, Unit Nos. 1 and 2 (TAC Nos. MA0050 and MAQOS 1)," April 1, 1999.

Topical Report DOM-NAF-3, Rev. 0.0-APae8 Page 85

I I

I Topical Report DOM-NAF-3, Rev. 0.0-NP-A I

GOTHIC Methodology for Analyzing the Response to I Postulated Pipe Ruptures Inside Containment I

Attachment 1 I NRC Request for Additional Information I

on DOM-NAF-3 and Dominion Responses dated June 8, 2006 I 14 pages after the cover page I I

I I

I I

I I

I I

Dominion Resources Services, Inc. E)

S00() Domninion Boulevard, Glen Allen, VA 2.3060 Dom..inuion June 8, 2006 United States Nuclear Regulatory Commission Serial No.06-408 Attention-. Document Control Desk NL&OS/PRW RO One White Flint North Docket Nos. 50-305 11555 Rockville Pike 50-336/423 Rockville, MD 20852-2738 50-338/339 50-280/281 License Nos. DPR-43 DPR-65/NPF-49 NPF-4/7 DPR-32/37 DOMINION ENERGY KEWAUNEE. INC. (DEK)

DOMINION NUCLEAR CONNECTICUT. INC. (DNC)

VIRGINIA ELECTRIC AND POWER COMPANY (DOMINION)

KEWAUNEE POWER STATION MILLSTONE POWER STATION UNITS 2 AND 3 NORTH ANNA POWER STATION UNITS 1 AND 2 SURRY POWER STATION UNITS 1 AND 2 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION REQUEST FOR APPROVAL OF TOPICAL REPORT DOM-NAF-3.

GOTHIC METHODOLOGY FOR ANALYZING THE'RESPONSE TO POSTULATED PIPE RUPTURES INSIDE CONTAINMENT In a letter dated November 1, 2005, Dominion Energy Kewaunee, Inc. (DEK), Dominion Nuclear Connecticut, Inc. (DNC) and Virginia Electric and Power Company (Dominion) requested the approval for the generic application of Topical Report DOM-NAF-3, "GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment," for Kewaunee Power Station (KPS), Millstone Power Station (MPS), North Anna Power Station (NAPS) and Surry Power Station (SPS), respectively. GOTHIC is a general-purpose, thermal-hydraulics computer code developed by the Electric Power Research Institute for applications in the nuclear power industry. The NRC has approved GOTHIC for use in containment analyses for several U.S. nuclear power plant licensees.

In Topical Report DOM-NAF-3, DEK, DNC and Dominion have developed an analytical methodology using GOTHIC for performing licensing basis analyses for the containment response for pressurized water reactors with large, dry containments. Plant specific applications of topical report DOM-NAF-3 will be implemented by DEK, DNC and Dominion according to the requirements of 10 CFR 50.59 for changes to USAR/FSAR/U FSAR evaluation methodologies.

In a letter dated April 28, 2006, the NRC requested additional information in order to complete its review of the submittal. The response to the request for additional information is provided in Attachment 2. As part of the response to NRC's question 2, DEK, DNC and Dominion have provided a CD-ROM that contains information DEK, DNC and Dominion consider to be proprietary. Therefore, Attachment 1 to this letter contains a request for withholding the information provided in the enclosed CD-ROM from public release under the provisions of 10 CFR 2.390. The associated affidavit attesting to the proprietary nature of the information is also included in Attachment 1.

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Page 2 of 4 If you have questions or require additional information, please contact Mr. Paul R.

Willoughby at (804) 273-3572.

Very truly yours, Eugene S. Grecheck Vice President - Nuclear Support Services Attachments: (2)

1. Application for Withholding and Affidavit of Eugen e S. GrecheckI 2.. Response to NRC Request for Additional Information: Topical Report IDOM-NAF-3

Enclosure:

CD-ROM that contains the electronic GOTHIC input and output files from the benchmark cases in Sections 4.3, 4.4, 4.5, and 4.6 of DOM-NAF-3 Commitments made in this letter: None

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/28 01281 Response to Reques~t for Additional Information Submittal of Topical Report DOM-NAF-3 Page 3 of 4 cc: U.S. Nuclear Regulatory Commission (w/o Encl.)

Region I 475 Allendale Road King of Prussia, Pennsylvania 19406-1415

.U.S. Nuclear Regulatory Commission (w/o Encl.)

Region 11 Sam Nun~n Atlanta Federal Center 61 Forsyth Street, SW Suite 23T85 Atlanta, Georgia 30303 U.S. Nuclear Regulatory Commission (w/o Encl.)

Region III 2443 Warrenville Road Suite 210 Lisle, Illinois 60532-4352 Mr. S. C. Burton (w/o Att.) (w/o Encl.)

NRC Senior Resident Inspector Kewaunee Power Station Mr. S. M. Schneider (w/o Att.) (w/o Encl.)

NRIC Senior Resident Inspector Millstone Power Station Mr. J. T. Reece (w/o Aft.) (w/o Encl.)

NRIC Senior Resident Inspector North Anna Power Station Mr. N. P. Garrett (w/o Att.) (w/o Encl.)

NRC Senior Resident Inspector Surry Power Station Mr. D. H. Jaffe (w/o Encl.)

NRC Project Manager - Kewaunee Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 7D1 Rockville, Maryland 20852-2738

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Page 4of 4 Mr. V. Nerses (w/o Encl.)

NRC Senior Project Manager - Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 8C2 Rockville, Maryland 20852-2738 Mr. S. R. Monarque (2 Encl.)

NRC Project Manager - North Anna Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 8-Hi112 Rockville, Maryland 20852-2738 Mr. S. P. Lingam (w/o Encl.)

NRIC Project Manager - Surry Power StationI U. S. Nuclear Regulatory Commission One White Flint North 11555 RockvilIlIe Pi ke MailI Stop 8 G9A Rcockville, Maryland 20852-2738

Serial No.06-408 Docket Nos. 50-305 336/423 338/339 280/281 ATTACH-MENT 1 APPLICATION FOR WITHHOLDING AND AFFIDAVIT OF EUGENE S. GRECIIECK DOMINION ENERGY KEWAUNEE, INC.

DOMINION NUCLEAR CONNECTICUT, INC.

VIRGINIA ELECTRIC AND POWER COMPANY KEWAUNEE POWER STATION MILLSTONE POWER STATION UNITS 2 AND 3 NORTH ANNA POWER STATION SURRY POWER STATION

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment 1 Page 1 of 2 APPLICATION FOR WITHHOLDING AND AFFIDAVIT OF EUGENE S. GRECHECK 1,Eugene S. Grecheck, Vice President - Nuclear Support Services, state that:

1. l am authorized to execute this affidavit on behalf of Dominion Resources Services, Inc. (DRS).

2. DRS is submitting a CD-ROM that contains the electronic GOTHIC input andI output files from the benchmark cases in Sections 4.3, 4.4, 4.5, and 4.6 of Topical Report DOM-NAF-3, for NRC review. The CD-ROM contains proprietary commercial information that should be held in confidence by the NRC pursuant to the policy reflected in 10 CFR §§ 2.390(a)(4) because:

a. This information is being held in confidence by DRS.

b, This information is of a type that is held in confidence by DRS, and there is a rational basis for doing so because the information contains sensitive commercialI information concerning DRS' containment analysis methodology.

c. This information is being transmitted to the NRC in confidence.I

d. This information is not available in public sources and could not be gathered readily from other publicly available information.I

e. Public disclosure of this information would create substantial harm to the competitive position of DRS by disclosing confidential DRS internal containment analysis methodology information to other parties whose commercial interests may be adverse to those of DRS. Furthermore, DRS has expended significant engineering resources in the development of the information. Therefore, the useI of this confidential information by competitors would permit them to use the information developed by DRS without the expenditure of similar resources, thus giving them a competitive advantage.

3. Accordingly, DRS requests that the designated document be withheld from public disclosure pursuant to the policy reflected in 10 CFR §§ 2.390(a)(4). f Eugene S. Grecheck tre00 Vic reient - Nuclear Support ServicesI

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment I Page 2 of 2 COMMONWEALTH OF VIRGINIA )

COUNTY OF HEN RICO The foregoing document was acknowledged before me, in and for the County and Commonwealth aforesaid, today by Eugene S. Grecheck, who is Vice President -

Nuclear Support Services of Dominion Resources Services, Inc. He has affirmed before me that he is duly authorized to execute and file the foregoing document in behalf of that company, and that the statements in the document are true to the best of his knowledge and belief.

Acknowledged before me this ~ 'day of 2006.

My Commission Expires: ay~~g:ýIz-~ o NoayPublic (SEAL)

I Serial No.06-408 Docket Nos. 50-305 336/423 I 33 8/339 280/281 I

I ATTACHMENT 2 I

I I

I RESPONSE TO NRC REQUEST FOR ADDITIONAL INFORMATION: TOPICAL REPORT DOM-NAF-3 I

I I

I I

I DOMINION ENERGY KEWAUNEE, INC. I DOMINION NUCLEAR CONNECTICUT, INC.

VIRGINIA ELECTRIC AND POWER COMPANY KEWAUNEE POWER STATION I MILLSTONE POWER STATION UNITS 2 AND 3 NORTH ANNA POWER STATION SURRY POWER STATION I I

I

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment 2 Page 1 of 6 RESPONSE TO NRC REQUEST FOR ADDITIONAL INFORMATION: TOPICAL REPORT DOM-NAF-3 NRC Request for Additional Information dated April 28, 2006 [Reference 1]

By letter dated November 1, 2005, Virginia Electric and Power Company, Dominion Nuclear Connecticut, Inc. and Dominion Energy Kewaunee, Inc. (the licensees),

submitted proposed Topical Report DOM-NAF-3 for the Nuclear Regulatory Commission (NRC) staff's review and approval. The licensees are requested to reply to the following questions.

NRC RAI#41:

In Section 2.2 of DOM-NAF-3, the licensees stated that "[flor containment modeling, [it]

has selected correlations that have been previously approved by the NRC and has confirmed the applicability of the models to large, dry PWR [pressurized water reactor]

containments. For calculation of post-reflood mass and energy release, a simplified GOTHIC model of the reactor coolant system (RCS) and steam generator secondary side has been developed and coupled to the containment ... Framatome recently received NRC approval for use of a coupled mass and energy release model..."

For all of the intended GOTHIC applications listed in Section 2.3, please identify those modeling techniques and assumptions (if there are any) that are different from what was previously reviewed and approved by the NRC staff, which requires the NRC staff's prior approval. For example, what makes your post-reflood mass and energy release model different (less conservative) from that approved for Framatome. Be specific and provide justification where appropriate.

Dominion Response:

Topical report DOM-NAF-3 [Reference 2] presents an analytical methodology for performing containment response design basis calculations with two components: 1) containment response model; and 2) simplified reactor coolant system (PCS) model for calculation of post-reflood mass and energy (M/E) releases. The containment response model is used for all applications in Section 2.3 of DOM-NAF-3. The NRC has approved GOTHIC for analyzing the containment response to loss of coolant accident (LOCA) and main steamline break (MSLB) events [References 3-8]. The analyses use models to maximize containment pressure and temperature using mass and energy releases that are generated by other NRC-approved methods and input to GOTHIC.

The DOM-NAF-3 methodology for maximizing LOCA and MSLB containment pressure and temperature uses NRIC-approved. models for the containment response (e.g., the Direct/Diffusion Layer Model for heat transfer between passive heat sinks and the containment atmosphere in DOM-NAF-3, Section 3.3.2, and the break release droplet model with 100-micron droplets in DOM-NAF-3, Section 3.5. 1).

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment 2 Page 2 of 6 To adequately evaluate all aspects o1 the containment design, a simplified RCS model is used to calculate long-term M!E releases and heat removal rates from the primaryI and secondary systems for all LOCA applications in Section 2.3. The Dominion post-reflood M/E release model is a new application that is different from other NRC-approved applications of GOTHIC. The Framatome GOTHIC methodology report

[Reference 9] is proprietary and a comparison to their long-term mass and energy release methodology (Section 5.1.2.3.2 in Reference 9) was not possible. The Framatome methodology was referenced on page 10 of DOM-NAF-3 only to point out that the Dominion method of coupling the RCS and containment models inside GOTHIC was not unique and that the NRC has approved the use of a coupled ýmethodology previously. Dominion believes the details below provide further explanationmof its use ofI GOTHIC which will facilitate the NRC review of Dominion's request.

Post-Reflood Mass and Energv Release ModelI Surry Power Station (SPS) and North Anna Power Station (NAPS) have subatmospheric containments that are required to be depressurized following a design basis accident in accordance with the assumptions in the dose consequences analyses.

The original design basis required a depressurization of the containment to subatmospheric conditions within one hour and subatmospheric conditions thereafter.I The GOTHIC simplified RCS model provides margin with respect to the NRC-approved Westinghouse post-reflood methodology. (WCAP-8264-P-A and WCAP-1 0325-P-A) that is the current licensing basis for SPS and NAPS. DOM-NAF-3, Section 4.4, shows thatI the GOTHIC methodology provides a reduction in containment depressurization time and a less severe pressure increase following containment spray termination, even though the integral energy release to the containment is similar between GOTHIC andI LOCTIC. Both of these effects represent. margin in the containment design relative to the current LOCTIC licensing basis analyses. This margin is attributed to how the post-reflood MIE release model distributes energy from the break.

The application of the post-reflood M/E. release methodology for SPS and NAPS is a

'Departure from a Method of Evaluation Described in the ESAR" because neither of theI two criteria specified in 10 CFR 50.59(a)(2) is satisfied:

i) The! method does not produce conservative or essentially the same results as the Westinghouse FROTH methodology that is the current licensing basis for SPS and NAPS. While the GOTH 'IC integral mass and energy releases are comnparable or more conservative, the distribution of energy released to the containment is different and provides margin in the containment depressurization time. NEI-96-07, Rev. 1, Section 3.4, states "Gaining margin by revising an element of a method of evaluation is considered to be a nonconservative change and thus a departure from a method of evaluation...". Further, a comparison to the proprietary Framatome methodology was not possible.

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment 2 Page 3 of 6 ii) The method has not been "approved by the NRC for the intended application."

Section 4.3.8.2 of NEI-96-07, Rev. 1, details a review process to identify if the methodology has been approved for general or specific applications. Application of the Dominion post-reflood M/E release methodology for subatmospheric containment depressurization calculations represents a new application of GOTHIC that has not been approved previously.

Calculation of NPSH Available As described in Section 3.8 of DOM-NAF-3, long-term containment analyses are performed to demonstrate adequate net positive suction head (NPSH) margin for the recirculation spray (RS) and low head safety injection (ILHSI)ý pumps that take suction from the containment sump following a LOCA. The calculation is performed internally in GOTHIC using an industry standard formulation for prediction of pump net positive suction head available (NPSHa). The calculation of NPSHa depends directly on transient predictions of sump temperature, sump water level, and containment pressure (SPS and NAPS credit containment overpressure in the NPSHa calculations as described in Section 3.8.1 of DOM-NAF-3).ý The calculation of NPSHa using the same formula was previously performed by Stone & Webster using the LOCTIC computer program.

The Dominion calculation method uses the simplified RCS model and applies specific assumptions (e.g., complete mixing in the intact loop cold leg for pump suction breaks) to the GOTHIC containment models to ensure a conservative response compared to a maximum containment pressure analysis. Dominion concluded that the assumptions in Section 3.8 of DOM-NAF-3 apply sufficient conservatism for a transient calculation of NPSHa with GOTHIC. Section 4.5 of DOM-NAF-3 shows that GOTHIC produces slightly higher NPSHa for the Surry LHSI pump compared to LOCTIC and attributes the differences to GOTHIC's liquid/vapor heat and. mass transfer model and the distribution of break energy between vapor and liquid.

The NRC has not reviewed previously this specific methodology for calculation of NPSHa. Further, the specific assumptions in Section 3.8 of DOM-NAF-3 are elements of the methodology that ensure a conservative calculation of NPSHa and these elements have not been reviewed. For example, DOM-NAF-3 specifies that a minimum containment pool area (specific to the plant being analyzed) is used to minimize evaporation for NPSH calculations, because GOTHIC's interfacial heat and mass transfer model provides a minor benefit in containment pressure compared to LOCTIC (which has no such model) and results in higher NPSH margin. In conclusion, the application of GOTHIC for NPSHa calculations is a "Departure from a Method of Evaluation Described in the FSAR" because neither of the two criteria specified in 10 CFR 50.59(a)(2) is satisfied:

i) The method does not produce conservative or essentially the same results as the Stone & Webster LOCTIC methodology. As shown in DOM-NAF-3, Section 4.5.3,

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment 2Page 4of 6I GOTHIC provides NPSH margin for the LHSI pump. NEI-96-07, Rev. 1, Section 3.4, states "Gaining margin by revising an element of a method of evaluation is considered to be a nonconservative change and thus a departure from a method of evaluation...".

ii) The use of GOTHIC with the specific assumptions in Section 3.8 of DOM-NAF-3 has not been "approved by the NRC for the intended application". Section 4.3.8.2 of NEI-96-07, Rev. 1, details a review process to. identify if the methodology has been approved for general or specific applications. The specific GOTHIC methods in 'Section 3.8 of DOM-NAF-3 with the coupled RCS/containment model are unique and require NRC review.

GOTHIC ArDlications for Comogonent Design Verification DOM-NAF-3, Section 2.3, specifies the use of GOTHIC for long-term containmentI anaiyses that verify that ESAR containment design limits are met (Applications 1-5).

The applications can be categorized into two types of containment analyses that use different model assumptions to produce either a maximum containment pressure profileI (Applications 1-4) or a maximum sump temperature (Application 5). As discussed earlier, the NRC has approved the GOTHIC containment modeling techniques in DOM-NAF-3 for calculating maximum containment pressure from LOCA and MVSLB events

[Referenc~es 3-8]. Dominion requests NRC approval of the DOM-NAF-3 methodology for calculating transient pump NPSHa.,'

I NRC acceptance of the GOTHIC containment response calculation methodologies for containment design limits does not explicitly cover the use of GOTHIC results for component design verification. As a result, Dominion included Applications 6-9 for NRCI to review and approve the use of GOTHIC output for specific component analyses. The methodology for performing pump NPSHa calculations (Application 5) produces a maximum sump water temperature, and Domin'ion plans to use the GOTHIC maximumI sump water temperature profile for validation against component design limits. For example, the predicted sump water temperature is confirmed to remain less than acceptable limits for the recirculation spray system piping following a LOCA (ApplicationI 6).

The renmaining GOTHIC applications implement assumptions that maximize containment pressure and vapor temperature, while minimizing sump water temperature. The methods for verifying that the containment liner temperature (Application 7) and equipment temperatures (Application 8) remain below their limits areI incremental changes to the LOCA and MVSLB peak containment pressure and temperature analyses (Applications 1 and 2). Again, since the containment modeling assumptions are biased to produce a conservative containment response, the GOTHICI results from these cases can also be used for component design verification. One example is the use of the minimum sump water temperature for determining the fluid viscosity for calculating the sump strainer head loss (Application 6).

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment 2 Page 5 of 6 References for Response #1

1) Letter from Stephen Monarque (USNRC) to David A. Christian (Dominion), "North Anna Power Station, Unit Nos. 1 and 2, Surry Power Station, Unit Nos. 1 and 2, Kewaunee Power Station, and Millstone Power Station, Unit Nos. 2 and 3 - Request for Additional Information (RAI) on Proposed Topical Report DOM-NAF-3 (TAO Nos.

MC8833, MC8834, MC8835, MC8836, MC8831, and MC8832)," April 28, 2006.

2). Letter from Leslie N. Hartz (Dominion) to USNRC, "Dominion Energy Kewaunee, Inc. (DEK), Dominion Nuclear Connecticut, Inc. (DNC), Virginia Electric and Power Company (Dominion), Kewaunee Power Station, Millstone Power Station Units 2 and 3, North Anna Power Station Units 1 and 2, Surry Power Station Units 1 and 2, Request for Approval of Topical Report DOM-NAF-3, GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment," Serial No.05-745, November 1, 2005.

3) Letter from Herbert N. Berkow (NRC) to Ronnie L. Gardner (Framatome), "Final Safety Evaluation for Framatome ANP Topical Report BAW-10252(P), Revision 0,

'Analysis of Containment Response to Postulated Pipe Ruptures Using GOTHIC,'

(TAO No. MC3783)," August 31, 2005.

.4) Letter from Anthony C. McMurtray (NRC-).to Thomas Coutu (NMC), ."Kewaunee Nuclear Power Plant - Issuance of Amendment (TAC NO. MB6408)," September 29, 2003.

5) Letter from John G. Lamb (NRC) to Thomas Coutu (NMC), "Kewaunee Nuclear Power Plant - Issuance of Amendment Regarding Stretch Power Uprate (TAC NO.

MB9031 )," February 27, 2004.

6) Letter from Alan B. Wang (NRC) to R.T. Ridenoure (OPPD), "Fort Calhoun Station, Unit No. 1 - Issuance of Amendment (TAC NO. MB7496)," November 5, 2003.

7) Letter from L. Mark Padovan (NRC) to D.N. Morey (Southern Nuclear Operating Company), "Joseph M. Farley Nuclear Plant, Units 1 and 2 - Issuance of Amendments re: Steam Generator Replacements (TAC Nos. MA4393 AND MA43,94)," December 29, 1999.

8) Letter from Frank Rinaldi (NRC) to J.T. Gasser (Southern Nuclear Operating Company), "Vogtle Electric Generating Plant, Units 1 and 2 Re; Issuance of Amendments (TAC Nos. MB5046 AND MB5047)," June 4, 2003. ADAMS Accession No. MLO031 600761.

9) Abdelghany, J. M., et al., "Analysis of Containment Response to Postulated Pipe Ruptures Using GOTHIC," Framatome ANP report BAW-10252(NP), Revision 0, July 2004.

Serial No.06-408 Docket Nos. 50-305/336/423/338/339/280/281 Response to Request for Additional Information Submittal of Topical Report DOM-NAF-3 Attachment 2 Page 6 of 6 NRC RAI #2:

Provide nodal diagrams that show the GOTHIC control volumes, junctions, etc.,I described in Section 4.2.1 for the demonstration analyses performed for Surry Power Station, LUnit Nos. 1 and 2.

Dominion Response:

Dominion did not provide the.GOTHIC nodal diagrams with DOM-NAF-3 because theyI are proprietary materials and Dominion desires to keep the topical report non-proprietary. Further, the nodal diagrams are difficult to interpret without the detailed system and component descriptions that are included in the GOTHIC input file. ToI answer the RAI, Dominion has provided the NRC with a proprietary CD-ROM that contains the electronic GOTHIC input and output files from the benchmark cases in Sections 4.3, 4.4, and 4.5 of DOM-NAF-3. One main steam line break GOTHIC modelI is included from Section 4.6 (the nodal diagram is the same for all cases presented).

The CD-ROM includes a Microsoft WORD file that explains the file formats and requirements for viewing (GOTHIC Version 7.2 is acceptable for the .GTH files).

Topical Report DOM-NAF-3, Rev. 0.0-NP-A GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment Attachment 2 Supplemental Information, Replacement Pages and GOTHIC Nodalization Diagrams for DOM-NAF-3 dated July 14, 2006 28 pages after the cover page NON-PROPRIETARY VERSION omits Attachment 4 from the July 14, 2006 letter

Dominion Resources Services, Inc.

5001) D~ominion Boulevard. Glen AlIim, VA 23060 0 Dominion Jul-y 14, 2006 United States Nuclear Regulatory Commission Serial No.06-544 Attention: Document Control Desk NL&OS/PRW RO One White Flint North Docket Nos. 50-305 11555 Rockville Pike 50-336/423 Rockville, MD 20852-2738 50-338/339 50-280/281 License Nos. DPR-43 DPR-65/NPF-49 NPF-4/7 DPR-32/37 DOMINION ENERGY KEWAUNEE. INC. (DEK)

DOMINION NUCLEAR CONNECTICUT. INC. (DNC)

VIRGINIA ELECTRIC AND POWER COMPANY (DOMINION)

KEWAUNEE POWER STATION MILLSTONE POWER STATION UNITS 2 AND 3 NORTH ANNA POWER STATION UNITS 1 AND2 SURRY POWER STATION UNITS 1 AND 2 SUPPLEMENT TO REQUEST FOR APPROVAL OF TOPICAL REPORT DOM-NAF-3. GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO POSTULATED PIPE RUPTURES INSIDE CONTAINMENT In a letter dated November 1, 2005 (Serial Number 05-745), Dominion Energy Kewaunee, Inc. (DEK), Dominion Nuclear Connecticut, Inc. (DNC) and Virginia Electric and Power Company (Dominion) requested approval for generic application of Topical Report DOM-I NAF-3, "GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment," for Kewaunee Power Station (KPS), Millstone Power Station (MPS),

North Anna Power Station (NAPS) and Surry Power Station (SPS), respectively. GOTHICI is a general-purpose, thermal-hydraulics computer code developed by the Electric Power Research Institute for applications in the nuclear power industry. The NRC has approved GOTHIC for use in containment analyses for several U.S. nuclear power plant licensees.I In Topical Report DOM-NAF-3, DEK, DNC and Dominion have developed an analytical methodology using GOTHIC for performing licensing basis analyses for the containment response for pressurized water reactors with large, dry containments. Plant specificI applications of topical report DOM-NAF-3 will be implemented by DEK, DNC and Dominion according to the requirements of 10 CFR 50.59 for changes to USAR/FSARIUFSAR evaluation methodologies.

While developing a plant-specific amendment request for the North Anna Power Station using the DOM-NAF-3 GOTHIC methodology, Dominion engineering personnel discovered that some GOTHIC applications produced less conservative results. After further evaluation, it was determined that a similar situation existed with the license amendment request for Surry Power Station, provided to the NRC in a letter dated

Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 Supplement to Submittal of Topical Report DOM-NAF-3 Page 2 of 4 January 31, 2006 (Serial Number 06-014). In a conference call of June 21, 2006, Dominion notified the NRC of the issues with the GOTHIC analysis methodology in the November 1, 2005 submittal and agreed to provide replacement pages for the affected sections with a description of the basis for change. In addition, Dominion agreed to provide copies of GOTHIC nodalization diagrams for DOM-NAF-3. Dominion considers the GOTHIC nodalization diagrams proprietary information in accordance with the provisions of 10 CFR 2.390(a)(4). Accordingly, Attachment 1 of this submittal contains a description of the changes to the November 1, 2005 submittal. Attachment 2 contains the replacement pages to DOM-NAF-3. Attachment 3 is the application for withholding and affidavit requesting withholding of proprietary information for the GOTHIC nodalization diagrams. The proprietary version of the GOTHIC nodalization diagrams are provided in and the, non-proprietary, redacted version of the GOTHIC nodalization diagrams are provided in Attachment 5.

Dominion continues to request approval of topical report DOM-NAF-3 by September 1, 2006 to support the implementation of license amendments during the Surry Unit 2 fall refueling outage. If you have questions or require additional information, please contact Mr. Paul R.Willoughby at (804) 273-3572.

Very truly yours, Gerald T. Bischof Vice President - Nuclear Engineering Dominion Energy Kewaunee, Inc.

Dominion Nuclear Connecticut, Inc.

Virginia Electric and Power Company Attachments: (5)

1. Description of changes to the November 1, 2005 submittal

2. Replacement pages for the November 1, 2005 submittal

3. Application for Withholding and Affidavit of Gerald T. Bischof

4. GOTHIC Nodalization Diagrams (Proprietary .version)

5. GOTHIC Nodalization Diagrams (Non-proprietary, redacted version)

Commitments made in this letter: None

Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/28 1 Supplement to Submittal of Topical Report DOM-NAF-3 Pae3o cc: U.S. Nuclear Regulatory Commission Region I 475 Allendale Road King of Prussia, Pennsylvania 19406-1 415 U.S. Nuclear Regulatory Commission Region 11 Sam Nunn Atlanta Federal Center 61 Forsyth Street, SW Suite 23T85 Atlanta, Georgia 30303 U.S. Nuclear Regulatory CommissionI Region III 2443 Warrenville Road Suite 210 Lisle, Illinois 60532-4352 Mr. S. C. BurtonI NRC Senior Resident Inspector Kewaunee Power StationI Mr. S. M. Schneider NRC Senior Resident Inspector Millstone Power Station Mr. J. T. Reece NRC Senior Resident InspectorI North Anna Power Station Mr. N. P. Garrett NRC Senior Resident Inspector Surry Power Station Mr. D. H. Jaffe NRC Project Manager - Kewaunee Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 7D1 Rockville, Maryland 20852-2738

Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 Supplement to Submittal of Topical Report DOM-NAF-3 Page 4 of 4 Mr. V. Nerses NRC Senior Project Manager - Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 802 Rockville, Maryland 20852-2738 Mr. S. R. Monarque NRC Project Manager - North Anna Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike Mail Stop 8-1-12 Rockville, Maryland 20852-2738 Mr. S. P. Lingam NRC Project Manager - Surry Power Station U. S. Nuclear Regulatory Commission One White Flint North 11555 Rockville Pike MailI Stop 8 G9A Rockville, Maryland 20852-2738

I Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 I

I ATTACHMENT 1 I

I I

I SUPPLEMENT TO REQUEST FOR APPROVAL OF TOPICAL REPORT DOM-NAF-3. GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO I

POSTULATED PIPE RUPTURES INSIDE CONTAINMENT DESCRIPTION OF CHANGES TO THE NOVEMBER 1. 2005 SUBMITTAL I I

I I

I DOMINION ENERGY KEWAUNEE, INC.

I DOMINION NUCLEAR CONNECTICUT, INC.

VIRGINIA ELECTRIC AND POWER COMPANY KEWAUNEE POWER STATION I

MILLSTONE POWER STATION UNITS 2 AND 3 NORTH ANNA POWER STATION UNITS 1 AND 2 SURRY POWER STATION UNITS 1 AND 2 I

I I

I

Serial No.06-544 Docket Nos. 50-305r336t423/33813391280/28 I Domin ion submitted topical report DOM-NAF-3, "GOTHIC M ethodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment" to the NRC for review in Reference

1. The report describes the analytical methodology to be used for licensing basis containment response analyses. Recently, Dominion identified an issue with the method for selecting sp ray drop size for NPSH calculations that requires a change to DOM-NAF-3.

Description of the Issue Section 3.8.2 in DOM-NAF-3 describes the Dominion GOTHIC analysis methodology for calculating available net positive suction head (NPSHa). This methodology was developed for the Surry and North Anna low head safety injection (LHSI) and recirculation spray (RS) pumps, but is an acceptable method for other pressurized water reactors with large, dry containments.

Adjustments are made to the GOTHIC containment models to ensure a conservative calculation of NPSHa. DOM-NAF-3, Section 3.8.2 includes the following statement:

All of the spray water is injected as droplet~s into the containment atmosphere*

(nozzle spray flow fraction of 1) and the Sauter droplet size is reduced by a factor of 10. These assumptions ensure that the maimurnheat is absorbed by the drops and the effect of sprays on reducing the- containment pressure is maximized.

Smaller drop size will increase the drop holdup in the atmosphere, which will further reduce the containment pressure.

This model assumption was confirmed to provide a conservative NPSHa for the Surry LHSI pumps for a double-ended pump suction guillotine (DEPSG) break during the development of the topical report methodology. However, the assumption was not validated for all break locations and single failure scenarios for Surry. The topical report was submitted to the NRC on November 1, 2005 [Ref. 1]I, and the methodology was used for Surry analyses that were submritted on January 31, 2006 [Ref. 2]. While preparing design analyses for North Anna using the DOM-NAF-3 methodology in June 2006, it was discovered that reducing the Sauter droplet size by a factor of 10 was conservative for LHSI pump NPSH analyses using the DEPSG break model but produced less conservative NPSHa results for the RS pumps for double-ended hot leg guillotine (DEHLG) breaks. A subsequent review of the Reference 2 Surry design analyses concluded, that the factor of 10 reduction in droplet size can produce less conservative results than the Sauter mean diameter for some, but not all, of the Surry NPSH analyses with GOTHIC.

Subsequently, Dominion performed a detailed investigation of this issue with .Numerical Applications, Inc. (NMI), the GOTHIC code vendor. NAT had provided support during the Surry GOTHIC containment model development and had recommended the droplet diameter reduction for NPSH calculations. Reducing the drop size by a factor of 10 gives very small drops, well beyond any uncertainty in the code or spray performance. These small drops lead to drop.

Page 1 of 5

Serial No.06-544 Docket Nos. 50-305/336/423/338f339/280/28 1 concentrations in the atmosphere that are much higher than expected and provide increases in NPSHa, from higher containment pressure, for certain breaks and spray assumptions. For Surry DEHLG breaks, a 10x reduction in spray drop size below the Sauter mean would increase NPSHa. Compared to the pump suction break, the hot leg break has less steam release to the atmo sphere with more heat going directly to the pool since all injection flow is forced to pass through the core. The higher steam flow to the atmosphere in the pump suction break results in a slower cooldown rate. A higher fraction of the spray cooling power is needed to absorb the condensation heat, leaving a smaller fraction for sensible heat reduction. It is the sensible heat reduction that is primarily responsible for the containment pressure reduction. The higher cooldown rate for the hot leg break cases make them more sensitive to the effects of increased drop concentration. In the cooldown situation, a high drop concentration from the small drops increases the containment temperature and pressure. The higher containment temperature deposits hotter drops in the pool, which reduces the NPSHa, while the higher containmentI pressure increases the NPSHa. In the Surry hot leg break analyses, the resulting increase in containment pressure is a more dominant effect than the increase in pool temperature, resulting in a net increase in NPSF~a compared to using the Sauter droplet size. Thi s sensitivity was not clear during the methodology development.

Based on our evaluation, a revised methodology for selecting spray droplet size in NPSH calculations is required for DOM-NAF-3. :Dominion advised the NRC of this development in a teleconference on June 21, 2006. Domrinion stated that it would submit a revised method for selecting spray drop size for NPSH calculations. In addition, Dominion stated that the equation for calculating NPSH would bc modified to use the fluid density at the pump suction in order toI recover some of the NPSH margin lost to the spray drop issue.

Change to DOM-NAF-3 With a better understanding of the impact of drop concentration on NPSH, for NPSH analysis the variation in drop size below the Sauter diameter will be limited to a factor of 2 to cover code and spray performance uncertainty. NPSH analyses will be performed using the largest Sauter droplet size. A confirmatory analysis will be performed by reducing the Sauter diameter by 2,I which sufficiently covers code and spray performance. uncertainty without creating drops too small that may cause excess droplet holdup in the atmosphere. The minimum NPSHa will be3 obtained from the case that provides the smaller NPSHa. The drop hold-up effect is small for typical, nominal spray drop sizes and very little variation is seen in the range of droplet size from Sauter to one-half Sauter. NPSH analyses are insensitive over this range of droplet size, and theI two cases together confirm that the effect of sprays on reducing containment pressu .re is maximized and that sufficient conservatism is included to address uncertainty in spray performance.

Page 2of 5

Serial No.06-544 Docket Nos. 50-3051336/'423/338/339/280/281I The following changes to DOM-NAF-3 are proposed to revise the spray drop diameter method:

ci Page 24: The factor of 10 reduction in spray drop size is described. The material is changed to address the spray model conservatism for NPSH calculations without a specific value.

" Page 43: Item 2 in the list of adjustments for NPSH analysis will be modified to state:

All of the spray water is*injected as droplets into the containment atmosphere (nozzle spray flow fraction of 1). Analyses are performed using the largest Sauter droplet size.

A confirmatory analysis is performed by reducing the Sauter diameter by 2, which sufficiently covers code and spray performance uncertainty (i.e., variation in nozzle design and orientation, nozzle flow rate and different header elevations) without creating drops too small that may cause excess droplet holdup in the atmosphere. NPSH analyses are relatively insensitive over this range of droplet size, and the two cases together confirm that the effect of sprays on reducing .containment pressure is maximized. The minimum NPSHa is reported from the case that provides the smaller NPSHa.

" Page 63: Section 4.5 documents the results of the Surry demonstration case for LHSI pump NPSH and mentions the factor of 10 reduction for spray drop size. Sensitivity studies have shown that this GOTHIUC case is not sensitive to drop size ranging from the analyzed smallest value (Sauter/lO) to the largest Sauter diameter. With the density change to Equation 16, NPSHa would actually increase by 0.4 ft. Because this case merely demonstrates the GOTHIC behavior versus LOCTIC and the reported minimum NPSHa is conservative, the results in Section 4.5 are not changed. The text is modified to address the difference between the assumed drop diameter of Sauter/lO and the revised method in Section 3.8.2.

a Page 77 and Table 4.7-1: With the revised method for selecting the minimum drop size for NPSH calculations, the RS pump NPSH sensitivity analyses in Table 4.7-1 were revisited.

With spray modeling maximized to reduce containment pressure, there is very little difference in minimum NPSHa for a range of single failures and the full engineered safeguards features (ESF) case that assumes no failure. The preyious analyses performed with the factor of 10 reduction had showed a close grouping of results also, with the failure

• of I emergency bus producing the limiting NPSHa. Table 4.7-1 is modified to identify the full ESF case as limiting with a footnote, regarding the importance of validating the limiting single failure for the RS pumps for each plant change. Text in Section, 4.7 (page 77) is modified also to describe the similarity of results for different scenarios.

Page 3 of 5

Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 Change to the NPSH Calculation Equation In addition to the change in spray drop modeling, Equation 16 is changed to use the fluid densityI at the pump impeller. The original methodology included the term ppin the denominator for the rated density for the pump at which NPSH required is specified. The fixed density of 62.3 ibmn/ft for 70 F water was used to add conservatism to the NPSH calculation methodology.

However, the transient pump suction fluid density is more appropriate and provides some NPSH margin to offset that consumed by the change to the spray drop size. Equation 16 is changed by replacing pp, with p,, which is defined as the fluid density at the pump suction. This value is taken from the GOTHIC pump suction volume at the impeller centerline. The following changesI to DOM-NAF-3 are proposed.

Page 42: Revise Equation 16 and the supporting text with the following insert.

NPSHQ,=~ +.p[E()E HIPa(Ts) Equation 16 gP.1 where P, is the GOTHIC calculated pressure in the pump suction volume, p, is the liquid density in the sump, E, is the elevation of the sump surface obtained from the installed correlation or table as a function of V,,, (the water volume in the containment), E, is the elevation of the containment volume, HI is the height of the containment volume, a, the liquid volume fraction in the containment, Pat(Ts) is the saturation pressure at the pump suction temperature, p, is the fluid density at the pump suction.

Method of Changing DOM-NAF-3 Dominion proposes to replace seven pages in DOM-NAF-3 based on the previous technical discussion. The replacement pages to DOM-NAF-3 are included in Attachment 2.

Pagc 4 of 5

Seria No.06-544 Docket Nos. 50-305/336/423/338/339/280/28 1 References

1. Letter from Leslie N. Hartz to USNRC, "Dominion Energy Kewaunee, Inc. (DEK),

Dominion Nuclear Connecticut, Inc. (DNC), Virginia Electric and Power Company (Dominion), Kewaunee Power Station, Millstone Power Station Units 2 and 3, North Anna Power Station Units 1 and 2, Sun-y Power Station Units 1 and 2, Request for Approval of Topical Report DOM-NAF-3, GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment," Letter Serial No.05-745, November 1, 2005.

2. Letter from Leslie N. Hartz to USNRC, "Virginia Electric and Power Company, Surry Power Station Units 1 and 2, Proposed Technical Speci fication Change and Supporting Safety Analyses Revisions to Address Generic Safety Issue 191," Letter Serial No.06-014, January 31, 2006.

Page 5 of 5

Serial No.06-544 I

Docket Nos. 50-305/336/423/338/339/280/281 I

I ATTACHMENT 2 I

I I

I SUPPLEMENT TO REQUEST FOR APPROVAL OF TOPICAL REPORT DOM-NAF-3. GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO I

POSTULATED PIPE RUPTURES INSIDE CONTAINMENT REPLACEMENT PAGES FOR THE NOVEMBER 1,.2005 SUBM ITTAL I

I I

1 I

DOMINION ENERGY KEWAUNEE, INC.

I DOMINION NUCLEAR CONNECTICUT, INC.

VIRGINIA ELECTRIC AND POWER COMPANY KEWAUNEE POWER STATION I

MILLSTONE POWER STATION UNITS 2 AND 3 NORTH ANNA POWER STATION UNITS 1 AND 2 SURRY POWER STATION UNITS 1 AND 2 I

I I

I

A, "ýA'f=-A'. Equation 12 Since, by assumption in GOTHIC, At, Equation 13 H

where HI is the specified height for the containment volume, the height of the containment volume should be set to H Equation 14 Setting the containment volume height as recommended above has some side consequences that must be considered:

L It will increase the pool surface area for heat and mass transfer. However, since the effective area of heat and mass transfer is the maximum of the pool area and the surface area defined by the hydraulic diameter (4V/D,1), as long as 4V/Dh > A1, there is no effect on peak pressure and temperature analyses.

2. For NPSH analysis, the water depth in the containment will have to be adjusted to account for the artificially increased pool area, A' . Sensitivity studies have shown that NPSHa is not sensitive to a reduction in containment height, because the spray modeling assumptions applied in Section 3.8.2 ensure a conservative spray response that minimizes the containment pressure for NPSI- analysis (Section 3.8.2).

The spray volume, Y,, is set to the total volume below the spray headers under the assumption that the region interior to the headers is adequately covered by the spray. The deposition area, A',

is set to the total horizontal area at the bottom of the sprayed regions where the sprays are expected to collect. For all calculations, the nozzle spray flow fraction is set to 1.0.

Topical Report DOM-NAF-3Pae2 Page 4:ý24

3.8.2 GOTHIC Analysis of NPSH Avail ableI NPSHa is the difference. between the fluid stagnation pressure and the saturation pressure at the pump intake. To calculate NPSHa for a given pump, the GOTHIC containment model includes a separate small volume for the pump suction. The volume elevation and height are set so that the mid-elevation of the volume is at the elevation of the pump first-stage impelecntri.ThI volume pressure (with some adjustments for sump depth) can then be used in the NPSHa calculation. The temperature in the suction volume provides the saturation pressure. The junction representing piping between the sump and the suction volume reflects the friction and form pressure drop between the sump and the pump suction. The pump suction volume also allows accurate modeling of the mixing of cold water that is injected into the sucti]on of the RS pumps atI Surry and North Anna.

The single volume GOTHIC model does not account for geometry details of the sump or the liquid that is held up in other parts of the containment. GOTHIC does calculate the total amount of liquid in the containment. A correlation is used to define the sump depth or liquid level as aI function of the water volume in the containment. The correlation accounts for the sump geometry variation with water depth and accounts for the holdup of water in other parts of the containment, as discussed in Section 3.8.3. This correlation is installed in a GOTHIC control variable for use in the NPSHa calculation.

With the above modeling features in place, the NPSHa is calculated via control variables as NPSHJ = P + p, g[E, (V,,)-E,. - Hac, I-~,I)Equation 16 gp, where P., is the GOTHIC calculated pressure in the pump suction volume, p, is the liquid density in the sump, E, is the elevation of the sump surface obtained from the installed correlation or table as a function of V,,. (the water volume in the containment), E, is the elevation of the containmentI volume, H is the height of the containment volume, a, the liquid volume fraction in the containment, .Psejr(Ts) is the saturation pressure at the pump suction temperature, p, is the fluidI density at the pump suction.

Worst case conditions for NPSHa depend on the time that the pumps take suction from the sump.

Therefore, the parameter settings that minimize NPSHa may vary depending on the timing for the operation of the pumps. In general, settings that reduce containment pressure and increase theI sump water temperature reduce the NPSHa. Section 4.7 lists the input parameter studies that provide the limiting set of conditions for Surry.

Topical Report DOM-NAF-3 Page 42

The water in the sump comes from three sources: direct deposit of mass from the break, condensate from the conductors, and spray drops. The drops from the blowdown will be very small and at the saturation temperature at the containment steam partial pressure when they enter the sump.. After the blowdown, the spillage water from the vessel is directly put in the sump with no heat transfer to the atmosphere or walls and equipment in the containment. This is a conservative approach for NPSH analysis. The condensate is generated at the saturation tem perature at the steam partial pressure and added directly to the sump. The heat transfer between the conductors and the condensate on the way to the sump is conservatively neglected.

If the spray drops are modeled as recommended below, the drops will enter the sump at the maximum possible temperature. Heat and mass transfer at the sump surface is allowed.

GOTHIC's model for heat and mass transfer at a pool is in good agreement with experimental data (e.g., the Grout Mold evaporation experiments [3]). For NPSH analysis, the liquid temperature is greater than the vapor temperature for most of the event, so A minimum pool area is specified to minimize evaporation. With this overall approach, the predicted sump temperature is conservatively high for the duration of the simulation.

The following adjustments are made to ensure a conservative calculation of NPSHa:

1) The heat and mass transfer to the containment heat sinks are expected to be under-predicted using the Direct heat transfer model. This is non-conservative for NPSH analysis. A multiplier of 1.2 applied to the heat transfer coefficient was shown to provide adequate conservatism in the calculation.

2) All of the spray water is injected as droplets into the containment atmosphere (nozzle spray flow fraction of 1). Analyses are performed using the largest Sautcr droplet size. A confirmatory analysis is performed by reducing the Sauter diameter by 2, which sufficiently covers code and spray performance uncertainty (i.e., variation in nozzle design and orientation, nozzle flow rate and different header elevations) without creating drops too small that may cause excess droplet holdup in the atmosphere. NPSH analyses are relatively insensitive over this range of droplet size, and the two cases together confirm that the effect of sprays on reducing containment pressure is maximized. The minimum NPSHa is reported from the case that provides the smaller NPSHa.

3) A conservative water holdup volume is subtracted from the containment liquid volume to reduce the sump water height. See Section 3.8.3.

4) The upper limit on containment free volume is used.

5) The minimum containment air pressure is used.

6) Conservative assumptions for spray and other system parameters are used in accordance with plant-specific sensitivity studies (Surry results are summarized in Section 4.7).

Topical Report DOM-NAF-3Pae4 Page Lý 43

4.5 GOTHIC Analysis of LHSI Pump NPSH AvailableI A GOTHIC calculation of LHSI pump NPSHa is compared to the LOCTIC analysis from the Surry UJFSAR for a DEPSG break with one train of safeguards and maximum SI flow. The minimum NPSHa occurs at recirculation mode transfer (RMT), when the LHSI pump swaps suction from the RWST to the containment sump. After RMT, NPSHa increases as the, containment pressure stabilizes and the sump temperature decreases from the RS heat exchangers removing energy. Thus, it is important that the primary and secondary system e nergy be removed at a high rate to maximize the sump temperature before RMT. The DEPSG model for containment depressurization from Section 4.4 was biased in accordance with Section 3.8.2 to minimize NPSHa. The spray nozzle drop diameter was reduced by a factor of 10 (which produced the same minimum NPSH as the method specified in Section 3.8.2), the nozzle spray flow fraction was set to 1.0, a multiplier of 1.2 was applied to the conductor heat transfer coefficients, and the upper limit on the containment free volume was used.I The containment initial conditions and design inputs were the same as the LOCTIC analy sis.

Water holdup was excluded because it was not part of the LOCTIC analysis.

4.5.1 Containment Response Table 4.5-1 compares the sequence of events and Table 4.5-2 compares the results at the time ofI minimum NPSHa. Figures 4.5-1 through 4.5-4 compare the containment pressure, vapor temperature, liquid temperature, and sump level to LOCTIC results shown as discrete points. The distribution of the energy release into containment is indicated by the containment pressure and temperature response. During the early part of the event (<1000 sec), the GOTHIC sump liquid temperature is considerably less than LOCTIC, the vapor temperature is slightly higher, and the pressure is higher. The LOCTIC pressure flash option models the break liquid as a continuous liquid addition to the sump. GOTHIC break modeling using droplets results in a different containment energy distribution, In general, the LOCTIC pressure flash option causes a very conservative amount of energy to be retained in the sump liquid with less vapor flashed into the air space. This is evident from the very high (> 250 F) LOCTIC sump temperatures that are maintained until almost 1000 seconds even while the RS heat exchangers are removing sump energy. The vapor temperature is slightly less than the GOTHIC values. LOCTIC assumes no inter-facial heat transfer between the sump pool and containment atmosphere, which also explains the high liquid temperatures.

For the first few seconds, the LOCTIC vapor temperatures are much higher than GOTHIC. This is due to the lack of a droplet model in LOCTIC, which results in a brief period of superheat.

Once the IRS and ORS pumps become effective (200-400 seconds into the event) and the sump liquid is sprayed into the containment, the difference between the model responses becomes less noticeable. At the time of RMT, the GOTHIC sump liquid temperature is about I F higher than LOCTIC and the pressure is about 0.7 psi higher. The higher sump temperature provides a relative adverse effect on NPSHa while the increased pressure is a benefit. The sump levels in Topical Report DOM-NAF-3 Page 63

4.7 Sensitivity Studies The conservative assumption for a particular analysis depends on the 'design requirement that is being verified. Sensitivity studies will be performed for break locations, single failures, and design inputs for each plant-specific GOThIC containment analysis. Table 4.7-1 documents the results of the studies for Surry's containment analysis criteria. The conclusions are consistent with the current LOCTIC analyses. With LOCTIC, the minimum NPSHa for the ORS and IRS pumps occurs for a case with full engineered safeguards (no single failure). The GOTHIC analyses produce the same minimum NPSHa for the full safeguards case and for other cases with single failures, which emphasizes the need to analyze the single failures for each design effort.

Table 4.7-1 illustrates the breadth of sensitivity analyses that were performed for Surry to confirm the limiting assumptions for the current plant configuration. The results are specific to Surry's current configuration and are not intended to cover all Dominion PWRs, since each station has specific design criteria and engineered safety features that require sensitivity studies. Dominion will perform simnilar sensitivity studies to define the set of conservative assumptions for each PWR application.

4.8 Summary of Demonstration Analyses Based on the comparison to LOCPIC, it is concluded that the GOTHIC model selections identified in Section 3 appropriately model the containment response for. LOCA and MSLB events. GOTHI-C shows similar behavior for containment pressure and temperature to the SWEC LOCTIC code for a DE1-LG break with miaximumn safeguards and a DEPSG break for containment depressurization and LHSI pump NPSHa. GOTHIC predicts lower peak containment pressures because of the DLM condensation model and the break droplet model. The GOTHIC liquid temperature is higher in the short-termn, but the RS heat exchangers and the interfacial heat and mass transfer in GOTHIC bring the vapor and liquid phase temperatures close together.

GOTHIC predicts shorter depressurization times because of the simplified RCS model that mechanistically removes energy from all steam generators, while the FROTH methodology non-mechanistically biases superheated steam flow through the broken loop steam generator. For the LHSI pump NPSI-a analysis, GOTHI.C predicts a slightly higher sump temperature and containment pressure at the time of minimum of NPSHa. Overall, the long-term containment response is comparable to LOCTIC. The analyses also demonstrate that the simplified RCS model is conservative for calculating post-reflood mass and energy release rates for both DEPSO and DEHLG breaks.

Topical Report DOM-NAF-3Pae7 Page 77

Table 4.7-1: Matrix of Conservative Inputs for Surry GOTHIC Containment Analyses Note: This table is based on the current plant configuration. Plant modifications can change these results.

Table Key Min= Assume the minimum value for the range of the design input Max Assume the maximum value for the range of the design input N/A Not Applicable: the key analysis result occurs after this parameter becomes effective or the component is not part of the containment response (e.g., accumulators for MSLB).

N/S = Not Sensitive: the key analysis result is not sensitive to changes in this input parameter.

LOAPeak 1 MSLB Peak 1 Containment Subatmospheric LHSI NPSH 0ORS NPSH IRS NPSH Pressure* jPressurei'Temp # jDepressurization [Peak Pesr General Break Type DEHLG 1.4 ft for pressure DEPSG DEPSG DEPSG DEHLG DEHLG 0.6 ft' for temnp #

Reactor Power 102% 0%for pressure 102% 102% 102% 102% 102%

102% for temp, #

Single Failure N/A 1 emergency bus I emergency bus I emergency bus 1 emergency None & None &

bus Containment Air Pressure Max Max / Min # Max Max Mini Min Mini Temperature Max Max Max Mini Max Max Max Relative Humidity 100% 100% / 0% # 100% 100% 100% 100% 100%

Free Vol ume Mini Mini Mini Mini Max Max Max Heat Sink Surface Area Min Mini Mini Max Min Mini Min Topical Report DOM-NAF-3 Page 78 M --- M- M---MM M-- M M- M M M M

LOCA Peak J MSLB Peak 1 Containment Subatmospheric 1LHSI NPSH 0ORS NPSH IRS NPSH Pressure* Pressure/Temp # JDepressurizationj Peak Pressure JI_______I______

Recirculation Spray ______

RS Piping Volume N/A N/S Max Max N/S Min Min IRS Flow Rate N/A N/S Min Min Min Min Max ORS Flow Rate IN/A N/S Min Mi Min Max Min IRS Recirculation Flow to N/A N/S N/S N/S N/S Min Min Pump Suction RS Timer Delay N/A N/S Max Max Max Max Max IRS Suction Loss N/S N/S N/S N/S N/S Max Max ORS Suction Loss N/S N/S N/S N/S N/S Max Max Service Water SW Flow Rate N/A N/S Min Min Min Max Max SW Temperature N/A N/S Max Max Max I Min Min HX Tube Plugging/Fouling N/A N/S Max Max Max 0 0 LOCA peak pressure and temperature assumptions are the same per Section 5.2.4.

1. MSLB peak temperature occurs for small breaks and the spectrum is reviewed for any plant parameter change. The peak temperature is obtained by using minimum air pressure and 0% humidity (peak pressure cases assume maximum air pressure and 100% humidity).

& Sensitivity studies have shown that the full ESF case (no single failure) produces the same minimum NPSH as many single failure scenarios. Design studies must evaluate single failure scenarios with the full ESF case.

Topical Report DOMI-NAF-3Pae0 Page 80

Serial No.06-544 I

Docket Nos. 50-305/336/423/338/339/280/281 I

I ATTACHMENT 3 I

I I

I SUPPLEMENT TO REQUEST FOR APPROVAL OF TOPICAL REPORT DOM-NAF-3. GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO I POSTULATED PIPE RUPTURES INSIDE CONTAINMENT APPLICATION FOR WITHHOLDING AND AFFIDAVIT OF I GERALD T. BISCHOF I

I I

I DOMINION ENERGY KEWAUNEE, INC.

DOMINION NUCLEAR CONNECTICUT, INC.

I VIRGINIA ELECTRIC AND POWER COMPANY KEWAUNEE POWER STATION MILLSTONE POWER STATION UNITS 2 AND 3 I

NORTH ANNA POWER STATION UNITS 1 AND 2 SURRY POWER STATION UNITS 1 AND 2 I I

I I

Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 Supplement to Submittal of Topical Report DOM-NAF-3 Page 1 of 2 APPLICATION FOR WITHHOLDING AND AFFIDAVIT OF GERALD T. BISCHOF I, Gerald T. Bisohof, Vice President - Nuclear Engineering, state that:

1. I am authorized to execute this affidavit on behalf of Dominion Energy Kewaunee, Inc. (DEK), Dominion Nuclear Connecticut, Inc. (DNC), Virginia Electric and Power Company (Dominion).

2. DEK, DNC and Dominion are submitting nodal diagrams associated with its GOTHIC containment analysis that contain proprietary commercial information that should be held in confidence by the NRC pursuant to the policy reflected in 10 CFR §§ 2.390(a)(4) because:

a. This information is being held in confidence by DEK, DNC and Dominion.

b. This information is of a type that is held in confidence by DEK, DNC and Dominion, and there is a rational basi~s for. doing so because the information contains sensitive commercial information concerning DEK, DNC and Dominion containment analysis methodology.

c. This information is being transmitted to the NRC in confidence.

d. This information is not available in public sources and could not be gathered readily from other publicly available information.

e. Public disclosure of this information would create substantial harm to the competitive position of DEK, DNC and Dominion by disclosing confidential DEK, DNC and Dominion internal containment analysis methodology information to other parties whose commercial interests may be adverse to those of DEK, DNC and Dominion. Furthermore, DEK, DNC and Dominion have expended significant engineelring resources in the development of the information. Therefore, the use of this confidential information by competitors would permit them to use the information, developed by DEK, DNC and Dominion without the expenditure of similar resources, thus giving them a competitive advantage.

Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 Supplement to Submittal of Topical Report DOM-NAF-3I

3. Accordingly, DEK, DNC and Dominion request that the designated document be

'withheld from public disclosure pursuant to the policy reflected in 10 CFR

§§ 2.390(a)(4).

Dominion Energy Kewaunee, Inc.

Dominion Nuclear Connecticut, Inc.

Virginia Electric and Power Company Gerald T. Bischof Vice President -NucIlea~gi neering COMMONWEALTH OF VIRGINIAI COUNTY OF HENRICO Subscribed and sworn to me, ANotary Public, in and for the County and State above named, this I/ day of L 0AQA 2006.

Notary Public My Commission Expires !111 (SEAL)I

Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 ATTACHMENT 4 SUPPLEMENT TO REQUEST FOR APPROVAL OF TOPICAL REPORT DOM-NAF-3. GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO POSTULATED PIPE RUPTURES INSIDE CONTAINMENT GOTHIC NODALIZATION DIAGRAMS (PROPRIETARY VERSION)

IWITHHOLD FROM PUBLIC DISCLOSURE PER 10 cfr 2.390(a)(4)1 DOMINION ENERGY KEWAUNEE, INC.

DOMINION NUCLEAR CONNECTICUT, INC.

VIRGINIA ELECTRIC AND POWER COMPANY KEWAUNEE POWER STATION MILLSTONE POWER STATION UNITS 2 AND 3 NORTH ANNA POWER STATION UNITS 1 AND 2 SURRY POWER STATION UNITS 1 AND 2

I Serial No.06-544 Docket Nos. 50-305/336/423/338/339/280/281 I

I ATTACHMENT 5 I

I I

I SUPPLEMENT TO REQUEST FOR APPROVAL OF TOPICAL REPORT DOM-NAF-3. GOTHIC METHODOLOGY FOR ANALYZING THE RESPONSE TO I

POSTULATED PIPE RUPTURES INSIDE CONTAINMENT GOTHIC NODALIZATION DIAGRAMS I

(NON-PROPRIETARY, REDACTED VERSION)

I I

I I

DOMINION ENERGY KEWAUNEE, INC.

DOMINION NUCLEAR CONNECTICUT, INC.

I VIRGINIA ELECTRIC AND POWER COMPANY KEWAUNEE POWER STATION MILLSTONE POWER STATION UNITS 2 AND 3 I

NORTH ANNA POWER STATION UNITS 1 AND 2 SURRY POWER STATION UNITS 1 AND 2 I I

I I

Serial No.06-544 GOTHIC Nodalization Diagrams for DOM-NAF-3 (Non-Proprietary Version)

In the below reference, the NRC requested that Dominion submit the GOTHIC nodalization diagrams for the Surry demonstration cases provided in Section 4 of DOM-NAF-3. This attachment presents the GOTHIC diagrams from the topical report LOCA cases in Sections 4.3 through 4.5. Tables are provided to summarize the model volumes and boundary conditions.

Reference:

Letter from Stephen Monarque (USNRC) to David A. Christian (Dominion), "North Anna Power Station, Unit Nos. 1 and 2, Surry Power Station, Unit Nos. 1 and 2, Kewaunee Power Station, and Millstone Power Station, Unit Nos. 2 and 3 - Request for Additional Information (RAI) on Proposed Topical Report DOM-NAF-3 (TAC Nos. MC8833, MC8834, MC8835, MC8836, MC8831, and MC8832),"

April 28, 2006.

Page 1 of 5

Serial No.06-544 GOTHIC Diagram for DEHLG Break (DOM-NAF-3, Section 4.3)

Page 2 of 5

Serial No.06-544 GOTHIC Diagram for DEPSG Break for Containment Depressurization (DOM-NAF-3, Section 4.4)

Page 3 of 5

I Serial No.06-544 GOTHIC Diagram for DEPSG Break for LHSI Pump NPSH (DOM-NAF-3, Section 4.5)

I I

I I

Page 4 of 5 I

Serial No.06-544 GOTHIC Diagram for DEPSG RCS Model for LHSI Pump NPSH (DOM-NAF-3, Section 4.5)

Page 5 of 5

Topical Report DOM-NAF-3, Rev. 0.0-NP-A GOTHIC Methodology for Analyzing the Response to Postulated Pipe Ruptures Inside Containment Attachment 3 Original Pages Replaced by Attachment 2I of Dominion letter 06-544, dated July 14, 2006 7 pages after the cover page

A ad~A' s - V c Equation 12 Since, by assumption in GOTHIC, Af =- Equation 13 H

where H is the specified height for the containment volume, the height of the containment volume should be set to H = V'Equation 14 Setting the containment volume height as recommended above has some side consequences that must be considered:

1. It will increase the pool surface area for heat and mass transfer. However, since the effective area of heat and mass transfer is the maximum of the pool area and the surface area defined by the hydraulic diameter (4 V/Dh), as long as 4V/Dh > 4k-, there is no effect on peak pressure and temperature analyses.

2. For NPSH analysis, the water depth in the containment will have to be adjusted to account for the artificially increased pool area, A . Sensitivity studies have shown that NPSHa is not sensitive to a reduction in containment height, because the conservative reduction in drop diameter by a factor of 10 makes the spray drops 100% efficient for NPSH analysis (Section 3.8.2).

The spray volume, V/, is set to the total volume below the spray headers under the assumption that the region interior to the headers is adequately covered by the spray. The deposition area, A'f is set to the total horizontal area at the bottom of the sprayed regions where the sprays are expected to collect. For all calculations, the nozzle spray flow fraction is set to 1.0.

Topical Report DOM-NAF-3Pae2 Page 24

3.8.2 GOTHIC Analysis of NPSH Available NPSHa is the difference between the fluid stagnation pressure and the saturation pressure at the pump intake. To calculate NPSHa for a given pump, the GOTHIC containment model includes aI separate small volume for the pump suction. The volume elevation and height are set so that the mid-elevation of the volume is at the elevation of the pump first-stage impeller centerline. The volume pressure (with some adjustments for sump depth) can then be used in the NPSHa calculation. The temperature in the suction volume provides the saturation pressure. The junction representing piping between the sump and the suction volume reflect s the friction and form pressure drop between the sump and the pump suction. The pump suction volume also allows accurate modeling of the mixing of cold water that is injected into the suction of the RS pumps at Surry and North Anna.

The single volume GOTHIC model does not account for geometry details of the sump or the liquid that is held up in other parts of the contaimnment. GOTHIC does calculate the total amount of liquid in the containment. A correlation is use d to define the sump depth or liquid level as a function of the water volume in the containment. The correlation accounts for the sump geometry variation with water depth and accounts for the holdup of water in other parts of the containment, as discussed in Section 3.8.3. This correlation is installed 'ina GOTHIC control variable for use in the NPSHa calculation.

With the above modeling features in place, the NPSHa is calculated via control variables as NPSH,a p9[PVwEP ,]s, Equation 16 where P, is the GOTHIC calculated pressure in the pump suction volume, p, is the liquid density in the sump, E, is the elevation of the sump surface obtained from the installed correlation or tableI as a function of V (the water volume in the containment), E, is the elevation of the containment volume, H is the height of the containment volume, a, the liquid volume fraction in the containment, Psat,(Ts) is the saturation pressure at the pump suction temperature, pp. is the rated density for the pump (density of the fluid for which the required NPSH is specified).

Worst case conditions for NPSHa depend on the time that the pumps take suction from the sump.

Therefore, the parameter settings that minimize NPSHa may vary depending on the timing for theI operation of the pumps. In general, settings that reduce containment pressure and increase the sump water temperature reduce the NPSHa. Section 4.7 lists the input parameter studies that provide the limiting set of conditions for Surry.

'rý ýal ýrtnC)4_TAP~q'1I F F 5

The water in the sump comes from three sources: direct deposit of mass from the break, condensate from the conductors, and spray drops. The drops from the blowdown will be very small and at the saturation temperature at the containment steam partial pressure when they enter the sump. After the blowdown, the spillage water from the vessel is directly put in the sump with no heat transfer to the atmosphere or walls and equipment in the containment. This is a conservative approach for NPSH analysis. The condensate is generated at the saturation temperature at the steam partial pressure and added directly to. the sump. The heat transfer between the conductors and the condensate on the way to the sump is conservatively neglected.

If the spray drops are modeled as recommended below, the drops will enter the sump at the maximum possible temperature. Heat and mass transfer at the sump surface is allowed.

GOTHIC's model for heat and mass transfer at a pooi is in good agreement with experimental data (e.g., the Grout Mold evaporation experiments [3]), For NPSH analysis, the liquid temnperature is greater than the vapor temperature for most of the event, so a minimumn pool area is specified to minimize evaporation. With this overall approach, the predicted sump temperature is conservatively high for the duration of the simulation.

The following adjustments are made to ensure a conservative calculation of NPSHa:

1) The heat and mass transfer to the containment heat sinks are expected to be under-predicted using the Direct heat transfer model. This is non-conservative for NPSH analysis. A multiplier of 1.2 applied to the heat transfer coefficient was shown to provide adequate conservatism in the calculation.

2) All of the spray water is injected as droplets into the containment atmosphere (nozzle spray flow fraction of 1) and the Sauter droplet size is reduced by a factor of 10. These assumptions ensure that the maximum heat is absorbed by the drops and the effect of sprays on reducing the containment pressure is maximized. Smaller drop size will increase the drop holdup in the atmosphere, which will further reduce the containment pressure.

3) A conservative water holdup volume is subtracted from the containment liquid volume to reduce the sump water height. See Section 3.8.3.

4) The upper limnit on containment free volume is used.

5) The minimum containment air pressure is used.

6) Conservative assumptions for spray and other system parameters are used in accordance with plant-specific sensitivity studies (Surry results are summarized in Section 4.7).

Topical Report DOM-NAF-3Pae4 Page 43

4.5 GOTHIC Analysis of LHSI Piup NPSH Available A GOTHIC calculation of LHSI pump NPSHa is compared to the LOCTIC analysis from the Surry UFSAR for a DEPSG break with one train of safeguards and maximumi SI flow. The minimum NPSHa occurs at recirculation mode transfer (RMT), when the LHSI pump swaps suction from the RWST to the containment sump. After RMT, NPSHa increases as the containment pressure stabilizes and theI sump temperature decreases from the RS heat exchangers removing energy. Thus, it is important that the primary and secondary system energy be removed at a high rate to maximize the sump temperature before RMT. The DEPSG model for containment depressurization from Section 4.4 was biased in accordance with Section 3.8.2 to minimize NPSHa. Specifically, the spray nozzle droplet diameter was reduced by a factor of 10, the nozzle spray flow fraction was set to 1.0, a multiplier of 1.2 was applied to the conductor heat transfer coefficients, and the upper limit on the containment free volume was used. The containment initial conditions and design inputs were the same as the LOCTIC analysis. Water holdup was excluded because it was not part of the LOCTIC analysis.

4.5.1 Containment ResponseI Table 4.5-1 compares the sequence of events'and Table 4.5-2 compares the results at the time of minimum NPSF~a. Figures 4.5-1 through 4.5-4 compare the containment pressure, vapor temperature, liquid temperature, and sump level to LOCTIC results shown as discrete points. The distribution of the energy release into containment is 'indicated by the containment pressure and temperature response. During the early part of the event (<1000 sec), the GOTHIC sump liquid temperature is considerably less than LOCTIC, the vapor temperature is slightly higher, and the

.pressure is higher. The LOCTIC pressure flash option models the break liquid as a continuous liquid addition to the sump. GOTHIC break modeling using droplets results in a diferent containment energy distribution. In general, the LOCTIC pressure flash option causes a very conservative amount of energy to be retained in the sump liquid with less vapor flash ed into the air space. This is evident from the very high (> 250 F) LOCTIC sump temperatures that are I maintained until almnost 1000 seconds even while the RS heat exchangers are removing sump energy. The vapor temperature is slightly less than the GOTHIC values. LOCTIC assumes no interfacial heat transfer between the sump pool and containment atmosphere, which also explains the high liquid temperatures.

For the first few seconds, the LOCTIC vapor temperatures are much higher than GOTHIC. This is due to the lack of a droplet model in LOCTIC, which results in a brief period of superheat.

Once the IRS and ORS pumps become effective (200-400 seconds into the event) and the sump liquid is sprayed into the containment, the difference between the model responses becomes less noticeable; At the time of RMT, the GOTHIC sump liquid temperature is about 1 F higher than LOCTIC and the pressure is about 0.7 psi higher. The higher sump temperature provides a relative adverse effect on NPSHa while the increased pressure is a benefit. The sump levels in I opical Keport 1JUIVI-NA-r-3 Page 6-3

4.7 Sensitivity Studies The conservative assumption for a particular analysis depends on the design requirement that is being verified. Sensitivity studies will be performed for break locations, single failures, and design inputs for each plant-specific GOTHIC containment analysis. Table 4.7-1 documents the results of the studies for Sunry's containment analysis criteria. The conclusions are consistent with the current LOCTIC analyses with the exception of the limiting single failure for the calculation of NPSHa for the ORS and IRS pumps. The LOCTIC assumption is full safeguards, while the GOTHIC analysis concluded that the limniting failure is one train of emergency power. This change is mainly because of the LOCTIC NPSHa analysis assumptions for pressure flash model and no interface heat transfer between the sump pool and the atmosphere, which creates lower containment pressures than GOTHIC for NPSH calculations.

This difference between the codes is accentuated when all six spray pumps are runining.

Table 4.7-1 illustrates the breadth of sensitivity analyses that were performed for Surry to confirm the limiting assumptions for the current plant configuration. The results are specific to Surry's current configuration and are not intended to cover all Dominion PWVRs, since each station has specific design criteria and engineered safety features that require sensitivity studies. Dominion will perform similar sensitivity studies to define the set of conservative assumptions for each PWR application.

4.8 Suimmary of Demonstration Analyses Based on the comparison to LOCTIC, it is concluded that the GOTHI1C model selections identified in Section 3 appropriately model the containment response for LOCA and MSLB events. GOTHIC shows similar behavior for containment pressure and temperature to the SWEC LOCTIC code for a DEI{LG break with maximumn safeguards and a DEPSG break for containment depressurization and LHSI pump NPSHa. GOTHIC predicts lower peak containment pressures because of the DLM condensation model and the break droplet model. The GOTHIC liquid temperature is higher in the short-term, but the RS heat exchangers and the interfacial heat and mass transfer in GOTHIC bring the vapor and liquid phase temperatures close together.

GOTHIC predicts shorter depressurization times because of the simplified RCS model that mechanistically removes energy from all, steam generators, while the FROTH methodology non-mechanistically biases superheated steam flow through the broken loop steam generator. For the LHSI pumrp NPSHa analysis, GOTHIC predicts a slightly higher sump temperature and containment pressure at the time of minimum of NPSHa. Overall, the long-termn containment response is comparable to LOCTIC. The analyses also demonstrate that the simplified RCS model is conservative for calculating post-reflood mass and energy release rates for both DEPSG and DEHLG breaks.

Topical Report DOM-NAF-3Pae7 Page 77

Table 4.7-1: Matrix of Conservative Inputs for Surry GOTHIC Containment Analyses Note: This table is based on the current plant configuration. Plant modifications can change these results.

Table Key Min = Assume the minimum value for the range of the design input Max = Assume the maximum value for the range of the design input N/A = Not Applicable: the key analysis result occurs after this parameter becomnes effective or the component is not part of the containment response (e.g., accumulators for MSLB3).

N/S = Not Sensitive: the key analysis result is not sensitive to changes in this input parameter.

LOCA Peak MSLB Peak J Containment ]Subatmospheric 1LHSI NPSH ORS NPSH IRS NPSH Pressure* Pressure/Temp # IDepressurization IPeak Pressure General Break Type DEHLG 1.4 ft2 for pressure DEPSG DEPSG DEPSG DEHLG DEHLG 0.6 ft2 for temp#

Reactor Power 102% 0% for pressure 102% 102% 102% 102% 102%

102% for temp #

Single Failure N/A 1 emergency bus 1 emergency bus I emergency bus 1 emergency Iemergency 1 emergency bus bus bus Containment Air Pressure Max Max /Min # Max Max Min Min Min Temperature Max Max Max Min Max Max Max Relative Humidity 100% 100% / 0%/ # 100% 100% 100% 100% 100%

Free Volume. Min Min Min Min Max Max Max Heat Sink Surface Area Min Min Min Max Min Min Min Topical Report DOM-NAF-3 Paae 78 M M - - - - - - M-

-- -M f -

- ------ --- ---------

LOCA Peak MSLB Peak 1 Containment TSubatmospheric TLHSI NPSH ORS NPSH IRS NPSH Pressure* JPressure/Temp # JDepressurization jPeak Pressurej________J_______

Recirculation Spray RS Piping Volume N/A N/S Max Max N/S Min Min

[RS Flow Rate N/A N/S Min Min Min Min Max ORS Flow Rate N/A N/S Min Min Min Max Min IRS Recirculation Flow to N/A N/S N/S N/S N/S Min Min Pump Suction RS Timer Delay N/A N/S Max Max Max Max Max IRS Suction Loss N/S N/S N/S N/S N/S Max Max ORS Suction Loss N/S N/S N/S N/S N/S Max Max Service Water SW Flow Rate N/A N/S Min Min Min MaxMa SW Temperature N/A N/S Max Max Max MinMi HX Tube Plugging/Fouling N/A N/S Max Max Max00

• LOCA peak pressure and temperature assumptions are the same per Section 5.2.4.

1. MSLB peak temperature occurs for small breaks and the spectrum is reviewed for any plant parameter change. The peak temperature is obtained by using minimum air pressure and 0% humidity (peak pressure cases assume maximum air pressure and 100% humidity).

Topical Report DOM-NAF-3Pae8 Page 80