Text

E-mail Ennis, NRR, to Wu, NRR, Et. Al., VY EPU Supplement 30
	ML052220149
Person / Time
Site:	Vermont Yankee
Issue date:	08/02/2005
From:	Richard Ennis; Office of Nuclear Reactor Regulation
To:	Christopher Boyd, Thomas Scarbrough, George Thomas, Wu C; Office of Nuclear Reactor Regulation
References
	Download: ML052220149 (246)
	v • d • e

Paqe I I

I Rick Rick Engis Ennis - VY

- Supplement 30, EPU Sjupplernent VYEPU 30, Attachment 1111 Page 1 1 From: Rick Ennis To: Cheng-lh Wu; Christopher Boyd; George Thomas; Harold Walker; Thomas Scarbrough Date: 8/2/05 4:33PM

Subject:

VY EPU Supplement 30, Attachment 11 As followup to the email I sent earlier today transmitting portions of Vermont Yankee EPU Supplement 30, the licensee sent me Attachment 11 to Supplement 30 which contains 10 exhibits (files are attached).

The exhibits relate to the following RAls:

SRXB-A-18 (1 exhibit)

SPSB-C-52 (6 exhibits)

EMEB-138 (3 exhibits)

Please let me know if you have any questions. Thanks, Rick 415-1420

I c:\temp\GW}I00001.TMP Page Pacje iill1 c:temp\GW1OO001 .TMP Mail Envelope Properties (42EFD87D.BEF: 11: 372)

Subject:

VY EPU Supplement 30, Attachment 11 Creation Date: 8/2/05 4:33PM From: Rick Ennis Created By: RXE(nrc.gov Recipients Action Date & Time TWGWPO02.HQGWDO01 Delivered 08/02/05 4:33 PM HXW (Harold Walker) owf2_po.WFNDO Delivered 08/02/05 4:33 PM CIWI (Cheng-Ih Wu)

GXT (George Thomas)

TGS (Thomas Scarbrough) tvf5_po.TWFNDO Delivered 08/02/05 4:33 PM CXB5 (Christopher Boyd)

Post Office Delivered Route TWGWPO02.HQGWDO01 08/02/05 4:33 PM owf2_po.OWFNDO 08/02/05 4:33 PM twvf5_po.TWFNDO 08/02/05 4:33 PM Files Size Date & Time MESSAGE 1065 08/02/05 04:33PM BVY 05-072 Ex.SRXB-A-18-l.pdf 600742 08/02/05 03:21PM BVY 05-072 Ex. SPSB-C-52-l.pdf 591203 08/02/05 03:17PM BVY 05-072 Ex. SPSB-C-52-2.pdf 1159854 08/02/05 03:17PM BVY 05-072 Ex. SPSB-C-52-3.pdf 5083648 08/02/05 03:18PM BVY 05-072 Ex. SPSB-C-52-4.pdf 3401670 08/02/05 03:20PM BVY 05-072 Ex. SPSB-C-52-5.pdf 725499 08/02/05 03:20PM BVY 05-072 Ex. SPSB-C-52-6.pdf 282378 08/02/05 03:19PM BVY 05-072 Ex.EMEB-B-138-l.pdf 832848 08/02/05 03:21PM BVY 05-072 Ex.EMEB-B-138-2.pdf 972271 08/02/05 03:21PM BVY 05-072 Ex.EMEB-B-138-3.pdf 1256329 08/02/05 03:21PM Options Auto Delete: No Expiration Date: None Notify Recipients: Yes Priority: Standard Reply Requested: No Return Notification: None

I c:\temp\GW}00001.TMP Page 21 Concealcd

Subject:

No Security: Standard To Be Delivered: Immediate Status Tracking: Delivered & Opened

BVY 05-072 Docket No. 50-271 Exhibit SRXB-A-18-1 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263.-. Supplement. No. 30 Extended Power Uprate Response to Request for Additional Information

'RS-001, BWR Template SE" Total number of pages in this Exhibit (excludina this cover sheet) Is 8.

(2) draft GDC-9, Insofar as it requires that the reactorcoolant pressure boundary shall be designed and donstructed so as to have an exceedingly low probability of gross rupture or significant leakage throughout its design lifetime; AccidentandTransient Analyses 2.8.5.1 Decrease In Feedwater Temperature, Increase in Feedwater Flow, Increase in Steam Flow, and Inadvertent Opening of a Main Steam Relief or Safety Valve Regulatory Evaluation Excessive heat removal causes a decrease Inmoderator temperature which Increases core reactivity and can lead to a power level Increase and a decrease In shutdown margin. Any unplanned power level Increase may result Infuel damage or excessive reactor system pressure. Reactor protection and safety systems are actuated to mitigate the transient. The NRC staffs review covered (1) postulated Initial core and reactor conditions, (2)methods of thermal and hydraulic analyses, (3) the sequence.of events, (4) assumed reactions of reactor system components, (5)functional and operational characteristics of the reactor protection system, (8) operator actions, and (7)the results of the transient analyses. The NRC's acceptance criteria are based on (1)draft GDC-6, Insofar as It requires that the reactor core be desig to function throughout Its design lifetime without exceeding acceptable fuel damage limits,( draft GDC14 and 15, Insofar as they require that the core protection system be designed to act automatically to prevent or suppress conditions that could result In exceeding acceptable fuel damage limits and that protection systems be prWvided for.sensing accident situations and Initiating the operation of necessary ESFs; and (u) draft GDC-27 and 28, Insofar as they require that at least two reactrvity control systems be provided and be capable of making and holding the core subcritical from any hot standby or hot operating condition sufficiently fast to prevent exceeding acceptablefuel damage limits. Specific review criteria are contained In SRP Section 15.1.1-4 and other guidance provided In Matrix 8 of RS-001.

Technical Evaluation

[Unsert technical evaluation. The technical evaluation should (1) clearly explain why the proposed changes satisfy each of the requirements In the regulatory evaluation and (2) provide a clear link to the conclusions reached by the NRC staff, as documented In the conclusion section.]

Conclusion

The NRC staff has reviewed the licensee's analyses of the excess heat removal events described above and concludes that the licensee's analyses have adequately accounted for operation of the plant at the proposed power level and were performed using acceptable.

analytical models. The NRCstaff further concludes that the licensee has demonstrated thatthe

reactor protection and safety systems will continue to ensure that the nd the RCPB pressure limits will not be exceeded as a result of these events. Based on Ulis, the NRC sta concludes that the plant will continue to meet the requirements of draft GDC-6,414, 15, 27, and 28 following Implementation of the proposed EPU. Therefore, the NRC staff finds the proposed EPU acceptable with respect to the events stated.

INSERT 8 FOR SECTION 32 -MwRTEMPLATE SAFETY EVALUATION DECEMBER2003

(2)draft GDC-9, insofar as Itrequires that the reactor coolant pressure boundary shall be designed and constructed so as to have an exceedingly low probability of gross rupture or significant leakage throughout its design lifetime; 2.8.5.2 Decrease in Heat Removal by the Secondary System 2.8.5.2.1 Loss of Extemal Load; Turbine Trp; Loss of Condenser Vacuum; Cclosure of Main Steam Isolation Valve; and Steam Pressure Regulator Failure '(Closed)

Reaulatorv Evaluation I A number of Initiating events may result Inunplanned decreases in heat remov val by the secondary system. These events result in a sudden reduction In steam flow arid, consequently, result In pressurization events. Reactor protection and safety systems are acti,zated to mitigate the transient. The NRC staffs review covered the sequence of events, the anm3lytical models used for analyses, the values of parameters used Inthe analytical models, andIthe results of the translent analyses. The NRC's acceptance criteria are based on (1)draft GDG -6, Insofar as It requires that the reactor u'egnedcor to funelon throughout Its design life,time without exceeding acceptable tuedaamage lijIis, and (updraft GDC-27 and 28, insofar as they require that at least two reactivity control systems be provided and be capable of making and holding the pore subcritical from any hot standby or hot operating condition sufficiently fast to prevent exceeding acceptable fuel damage limits. Specific review criteria are contained In SRP Section 15.2.1-5 and other guldance provided rn Matrix 8 of RS-001.

-Technical Evaluation

[Insert technical evaluation. The technical evaluation should (1) clearly explain why the proposed chargges satisfy each of the requirements In the regulatory evaluation and (2)provide a clear link to the conclusions reached by the NRC staff, as documented In the conclusion section.]

Conclusion The NRC staff has reviewed the licensee's analyses of the'decrease in heat removal events described above and concludes that the licensee's analyses have adequately accounted for operation of the plant at-the proposed power level and were performed using acceptable analytical models. The NRC staff further concludes that the licensee has demonstrated that the reactor protection and safety systems will continue to ensure that the~tAD)and the RCPB pressure limits will not be exceeded as a result of these events. Baseionifs, the NRC staff concludes that the plant will continue to meet the requirements of draft GDC-6, 27, and 28 following Implementation of the proposed EPU. Therefore, the NRC staff finds the proposed EPUJ acceptable with respect to the events stated. .

I .

INSERT8 FOR SECTION 32 -8WR TEMPLATE SAFETY EVALUATION DECEMBER 2003

(2)draft GDC-9, Insofar as it requires that the reactor coolant pressure boundary shall be designed and constructed so as to have an .

exceedingly low probability of gross rupture or significant leakage throughout its design lifetime; 2.8.5.2.2 Loss of Nonemergency AC Power to the Station Auxiliaries Reaulatorv Evaluation The loss of nonemergency ac power Is assumed to result in the loss of all power to the station auxiliaries and the simultaneous tripping of all reactor coolant circulation pumps. This causes a flow coastdown as well as a decrease In heat removal by the secondary system, a turbine trip, an Increase In pressure and temperature of the coolant, and a reactor trip. Reactor protection and safety systems are actuated to mitigate the transient. The NRC staffs review covered (1)the sequence of events, (2) the analytical model used for analyses, (3) the values of parameters used In the analytical model, and (4)the results of the transient analyses. The NRO's acceptance criteria are based on (1)draft GDC-8, Insofar as It requires that the reactor coredesined to function throughout Its design lifetime without exceeding acceptable fuel maglimind (draft GDC-27 and 28, Insofar as they require that at least two reactivity control systems be provided and be capable of making and holding the core subcritical from any hot standby or hot operating condition sufficiently fast to prevent exceeding acceptable fuel damage limits. Specific review criteria are contained in SRP Section 15.2.6 and other guidance provided In Matrix 8 of RS-001.

Technical Evaluation

[Insert technical evaluation. The technical evaluation should (1) clearly explain why the proposed changes satisfy each of the requirements In the regulatory evaluation and (2)provide a clear link to the conclusions reached by the NRC staff, as documented In the conclusion section.]

Conclusion The NRC staff has reviewed the licensee's analyses of the loss of nonemergency ac power to station auxiliaries event and concludes that the licensee's analyses have adequately accounted for operation of the plant at the proposed power level and were performed using acceptable analytical models. The NRC staff further-concludes that the licensee h emonstrated that the reactor protection and safety systems will continue to ensure that thenQ1D-and the RCPB pressure limits will not be exceeded as a result of this event. Based on this, the NRC staf concludes that the plant will 'continue to meet the requirements of draft GDC-6,,iand 28 following Implementation of the proposed EPU. Therefore, the NRC staff finds the proposed EPU acceptable with respect to the loss of nonemergency ac power to station auxiliaries event.

INSERTS FOR SECTION 32 -BWR TEMPLATE SAFETY EVALUATION DECEMBER2003

(2) draft GDC-9, insofar as it requires that the reactor coolant pressure boundary shall be designed and constructed so as to have an exceedingly low probability of gross rupture or significant leakage throughout Its design lifetime; 2.8.5.2.3 Loss of Normal Feedwater Flow Regulatory Evaluation A loss of normal feedwater flow could occur from pump failures, valve malfunctions, or a LOOP.

Loss of feedwaterflow results Inan Increase In reactor coolant temperature and pressure which eventually requires a reactor trip to prevent fuel damage. Decay heat must be transferred from fuel following a loss of normal feedWater flow. Reactor protection and safety systems are actuated to provide this function and mitigate other aspects of the transient. The NRC staffs review covered (1)the sequence of events, (2) the analytical model used for analyses, (3) the values of parameters used Inthe analytical model, and (4) the results of the transient analyses.

The NRC's acceptance critoria are based on (1)draft GDC-6, Insofar as It requires that the reactor core be designed tQ function throughout Is design lifetime without exceeding acceptable fuel damage limitsand (%Nraft GDC-27 and 28, insofar as they require that at least two reactivity control systems be provided and be capable of raldrng and holding the core subcritical from any hot standby or hot operating condition sufficiently fast to prevent exceeding acceptable fuel damage limits. Specific review Qriteria are contained In SRP Section 15.2.7 and other guidance provided InMatrix 8 of RS-0D1.

Technical Evaluation

[Inserttechnical evaluation. The technical evaluation should (1) clearly explain why the proposed changes satisfy each of the requirements In the regulatory evaluation and (2) provide a clear link to the conclusions reached by the NRC staff, as documented In the conclusion section.]

Conclusion The NRC staff has reviewed the licensee's analyses of the loss of normal feedwater flow event and concludes that the licensee's analyses have adequately accounted for operation of the plant at the proposed power level and were performed using acceptable analytical models. The NRC staff further concludes that the licensee has demonstrated that the reactor protection and safety systems will continue to ensure that thKAWDM)and the RCPB pressure limits will not be exceeded as a result of the loss of norrnaTTlwaterflow. Based on this, the NRC taff -1 )

concludes that the plant will continue to meet the requirements of draft Gand 28

. following Implementation of the proposed EPU. Therefore, the NRC staff finds the proposed EPU acceptable with respect to the loss of normal feedwater flow event INSERT 8 FOR SECTION 3.2 -DWR TEMPLATE AFETY EVALUATION DECEMBER2003

(2)draft GDC-9, insofar as it requires that the reactor coolant pressure boundary shall be designed and constructed so as to have an exceedingly low probability bf gross rupture or significant leakage throughout its design lifetime; 2.8.5.3 Decrease In Reactor Coolant System Flow 2.8.5.3.1 Loss of Forced Reactor Coolant Flow Reaulatory Evaluation A decrease in reactor coolant flow occurring while the plant Is at power could result In a degradation of core heat train. An Increase in fuel temperature and accompanying fuel damage could then result lfoAf~. are exceeded during the transient. Reactor protection and safety systems are actuated to mitigate the transient. The NRC staffs review covered (1) the postulated initial core and reactor conditions, (2)'the methods of thermal and hydraulic analyses, (3) the tequence of events, (4) assumed reactions of reactor systems components, (5) the functional and operational characteristics of the reactor protection system, (6) operator actions, and (7) the results of the transient analyses. The NRC's acceptance criteria are based on (1) draf~t GDC6, Insofar as it requires that the reactor core be designed to function throughout Its esign letime without exceeding acceptable fuel damage limits and (udraft GDC-27 and 28, insofar as they require that at least two reactivity control systems be provided and be capable of making and holding the core subcritical from any hot standby or hot operating condition sufficiently fast to prevent exceeding acceptable fuel damage limits. Specific review criteria are contained In SRP Section 15.3.1-2 and Qtherguldance provided In Matrix 8 of RS-001.

Technical Evaluation Insert technical evaluation.' The technical evaluation should (1) clearly explain why the proposed changes satisfy each of the requirements In the regulatory evaluation and (2) provide a clear link to the conclusions reached by the NRC staff, as documented In the conclusion section.]

Conclusion The NRC staff has reviewed the licensee's analyses of the decrease In reactor coolant flow event and concludes that the licensee's analyses have adequately accounted for operation of the plant at the proposed power level and were performed using acceptable analytical models.

The NRC staff further concludes that the licensee has demonstrated that the reactor protection and safety systems will continue to ensure that the% pand the RCPB pressure limits win not be exceeded as a result of this event. Based on this, the NRC staff concludes that the pla will continue to meet the requirements of draft GDC-b27, and 28 following Implementation of

.the proposed EPU. Therefore, the NRC staff finds the proposed EPU acceptable with respect to the decrase inreactor coolant flow event. * * ',

INSERT 8 FOR SECTION 32 - DWRTEMPLATE SAFETY EVALUATION DECEMBER2003

! (2) draft GDC-9, insofar as it requires that the reactor coolant pressure boundary shall be designed and constructed so as to have an exceedingly low probability of gross rupture or significant leakage throughout its design lifetime; 2;8.5.5 Inadvertent Operation of ECCS or Malfunction that Increases Reactor Coolant Inventory Regjulatory Evaluation Equipment malfunctions, operator errors, and abnormal occurrences could cause unplanned Increases in reactor coolant Inventory. Depending on the temperature of the Injected water and the response of the automatic control systems, a power level Increase may result and, without adequate controls, could lead to fiel damage or ovdrpressurization of the RCS. Altematively, a power level decrease and depressurization may result. Reactor protection and safety systems are actuated to mitigate these events. The NRC staff's review covered (1) the sequence of events, (a)the analytical model used for analyses, (3)the values of parameters used Inthe analytical model, and (4)the results of the transient analyses. The NRC's acceptance criteria

ae based on (1)draft GDC-6. Insofar as it re uires th e reactor core be designed tg function thi Wthut acceptable fuel damage limitsand (%)draft GDC-27 and 28, Insofar as they require that at least two reactivity control systems be provided and be capable of making and holding the core subcritical from any hot standby or hot operating condflon sufficiently fast to prevent exceeding acceptable fuel damage limits. Specific review criteria are contained InSRP Setron 15.5.1-2 and other guidance provided in Matrix 8 of RS-001.

Technical Evaluation Dlnsert technical evaluation. The technical evaluation should (1) clearly'explain why the proposed changes satisfy each of the requirements In the regulatory evaluation and (2) provide a clear link to the conclusions reached by the NRC staff, as documented in the conclusion sectlon.)

Conclusion The NRC staff has reviewed the licensee's analyses of the Inadvertent operation of ECCS or malfunction that increases reactor coolant Inventory and concludes that the licensee's analyses have adequately accounted for operation of the plant at the proposed power level and were performed using acceptable analytical models. The NRC staff further concludes that the licensee hs demonstrated that the reactor protection and safety systems will continue to ensure that theVAnd the RCPB pressure limits will not be exceeded as a result of this event.

Based on ts,the NRC staff concludes that the plant will continue to meet the requirements of draft GDC-P , and 28 followiing Implementation of the proposed.EPU. Therefore, the NRC ('i) staff finds the proposed EPU acceptable with respect to the Inadvertent operation of ECCS or:'

malfunction that Increases reactor coolant inventory.

INSERT 8 FOR SECTION 3.2 - BWR TEMPLATE SAFETY EVALUATION DECEMBER 2003

(2)draft GDC-9, insofar as it requires that the reactor coolant pressure boundary shall be designed and constructed so as to have an exceedingly low probability of gross rupture or significant leakage throughout Its design lifetime; 2.8.5.4.3 Startup of a Recirculation Loop at an Incorrect Temperature and Flow Controller Malfunction Causing an Increase In Core Flow Rate Regulatory Evaluation A startup of an inactive loop transient may result in either an Increased core flow or the introduction of cooler Water into the are. Tphis event causes an Increase in core.reactiviy due to decreased moderator temperature and gore void fraction. The NRC staffs review covered (1) the sequence of events, (2) the analytical model, (3) the values of parameters used In the analytical model, and (4) the results of the transient analyses. The NRC's acceptance criteria are based on I draft GDC-6, InsofaraI requires that the reactor core be designed to function throughout its design lifetime without ecxiepdng acceptabl auel damage Jim is,(jdraft GDC-14 and 15, Insofar as they require that the pore protection systems be designed to act automatically to prevent or suppress conditions that could result In exceeding acceptable fuel damage limts and that protection systems be provided for sensing accident situations and Initiating the operation of necessary ESFs; (updraft GPC-32, Insofar as It requires that limits, which Include considerable margin, be placed on the rnaxlrnum reactivity worth of control rods or elements and on rates at which reactivity can be Increased to ensure that the potential effects of a sudden or large change of reactivity cannot (a) rupture the reactor coolant pressure boundary or (b) disrupt the core, its support structures, g other vessel intemals sufficiently to Impair the effectiveness of emergency core cooling; and (4)draft GOC-27 and 28, Insofar as they require that at least two reactivity control systems be provided and be capable of making and holding the core subcritcal from any hot standby or hot operating Condition sufficiently fast to prevent exceeding acceptable fuel damage.limits. Specific review criteria are contained In SRP-Sectlon 15.4.4-5 and other-guidance provided In Matrix 8 of RS-001.

Technical Evaluation insert technical evaluation. The technical evaluation should (1) clearly explain why the proposed changes satisfy each of the requirements In the regulatory evaluation and (2) provide a clear link to the conclusions reached by the NRC staff, as documented in the conclusion section.]

Conclusion The NRC staff has reviewed the licensee's analyses of the Increase In core flow event and concludes that the licensee's analyses have adequately accounted for operation of the plant at the proposed power level and were performed using acceptable. analytical models. The

NRC'staff further concludes that the licensee has demonstrated that the reactor protecti6n and' safety systems will'continue to ensure that theA snd the RCPB pressure limits will not be exceeded as a result of this event. Based on thisTIeNRC staff concludes that the plant will.

continue to meet the requirements of draft GDC-',414, 15, 27, 28, and 32 following implementation of the proposed EPU. Therefore, the NRC staff finds the proposed EPU acceptable with respect to the Increase in core flow event.

INSERTS FOR SECTION 32 - BWR TEMPLATESAEY EVALUATION DECEMBER2003

(2)draft GDC-9, insofar as it requires that the reactor coolant pressure boundary shall be designed and constructed so as to have an

-exceedingly low.probability of gross rupture or significant leakage throughout its design lifetime;

.5.6 Decrease In Reactor Coolant Inventory 2.8.5.6.1 Inadvertent Opening of a Pressure Relief Valve Reaulatory Evaluation The Ina~dvertent opening of a pressure relief valve results In a reactor coolant inventory decrease and a decrease in RCS pressure. The pressure relief valve discharges Into the suppression pool. Normally there Is no rdactor trip. The pressure regulator senses the RCS pressure decrease and partially closes the turbine control valves (TCVs) to stabilize the reactor at a lower pressure. The reactor power settles out at nearly the Initial power level. The coolant Inventory is maintained by the feedwater control system using water from the condensate storage tank via the condenser hotwell. The NRC staffs review covered (1)the sequence of events, ({) the analytical model used for analyses, (3) the values of parameters used Inthe analytlcal model, and (4)the results of the transient analyses. The NRC's acceptance criteria are based on (1)

GDCr uires that the reacor as Ift einsof e be desi ned to function throughout Its design Ietime without exceeding acceptable fuel damage limbts and (>iraft GDC-27 and 28, Insofar as they require that at least two reactivity control systems be provided and be capable of making and holding the core subcritical from any hot standby or hot operating condition sufficiently fast to prevent exceeding acceptable fuel damage limits. Specific review criteria are contained InSRP Section 15.6.1 and other guidance provided in Matrix 8 of RS-001.

Technical Evaluation UInsert technical evaluation. The technical evaluation should (1) clearly explain why the proposed changes satisfy each of the requirements In the regulatory evaluation and (2) provide a clear link to the conclusions reached by the NRC staff, as documented In the conclusion section.]

Conclusion The NRC staff has reviewed the licensee's analyses of the Inadvertent opening of a pressure relief valve event and conclud es that the licensee's analyses have adequately accounted for operation of the plant at the proposed power level and were performed using acceptable analytical models. The NRC staff further concludes that the licensee has demonstrated that the reactor protection and safety systems will continue to ensure that the DLs-and the RCPB pressure limits will not be exceeded as a result of this event. Based oneis, NRC staff concludes that the plant will continue to meet the requirements, of draft GDCS?7and28 .J following implementation of the proposed EPU.. Therefore, the'NRC staff finds the proposed EPU acceptable with respect to the inadvertent openlrib of ii pressure relief valve event.

INSERT 8 FORSECTION 32. DWRTEMPLATE SAFETY EVALUATION DECEMBER2003

BVY 05-072 Docket No. 50-271 Exhibit SPSB-C-52-1 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Calculation VYC-0886, Rev. 2 Total number of pages in this Exhibit I (excludina this cover sheet) is 13. l

VY CALCULATION CHANGE NOTICE (CCN)

CCN Number: 04 Calculation Number: VYC-0886 Rev. No. 2 Calculation

Title:

Station Blackout Documentation Analysis Initiating Document: EPU VYDI)M M Spec. Nol other Safety Evaluation Numnb.er:

TV_.

N/A

.a,,^

Superseded Calculation% NIA 0.---A-A oupjA4sucuy Lv MIA LlfU Implementation Required: 0 Yes E No Computer Codes: NIA Reason for Change:

The VYC-0886 Rev 2 is updated to assess the effect of Extended Power Uprate (EPU) on this calculation.

Description of Change:

This CCN updates VYC-886 Rev 2, for EPU.

Technical Justification for Change:

See Attachment A

Conclusions:

The results of Reference 1 were addressed at EPU conditions. The effect of EPU on VYC-886Rev2 are summarized In Attachment A.

Are there any open items in this CCN? 0D Yes E No Prepared By/Date Interdiscipline Review By/Date Independent Review By/Date Approved By/Date Liliane Schor 02/I/Zoo~

I__________ASLR3 0 , 6 ta'FJmsG.Rgr Final Turnover tcDDCC (Section 2):

1) All open items, if any, have been closed.

2) Implementation Confirmation (Section 2.3.4)

O Calculation accurately reflects existing plant configuration, (confirmation method indicated below)

E Walkdown a As-Build input review 0 Discussion with OR (print name) 0l N/A, calculation does not reflect existing plant configuration

3) Resolution of documents identified in the Design Output Documents Section of VYAPF 0017.07 has been initiated as required (Section 2.3.6,2.3.7)1 I I Print Name Signature Date Total number of pages in package including all attachments: 13 pages Note: VYAPF 0017.07 should be included immediately following this form. VYAPF 0017.08 AP 0017 Rev. 8 Page 1 of I

VY CALCULATION DATABASE INPUT FORM Place this form in the calculation package immediately following the Title page or CCN form.

VYC-0886/CCN04 2 N/A N/A VY Calculation/CCN Number Revision Number Vendor Calculation Number Revision Number Vendor Name: N/A PO Number: N/A Originating Department: DesnEnineering Critical References Impacted: El UFSAR 0 DBD ] Reload. "Check" the appropriate box if any critical document is identified in the tables below.

EMPAC Asset/Equipment ID Number(s): N/A EMPAC Asset/System ID Number(s): N/A Keywords: Decay Heat. SBO. Torus Temperature. Condensate Storage Tank. Ventilation For Revision/CCN only: Are deletions to General References, Design Input Documents or Design Output Documents required? 0 Yest ED No Design Input Documcnts and General References - The following documents provide design input or supporting information to this calculation. (Refer to Appendix A, sections 3.2.7 and section 4)

Significant Critical Difference Affected Reference

Reference # ** DOC # REV # ***Document Title (including Date, if applicable) Review ft Program (1) 1 VYCfl886 2 Station Blackout Documentation Analysis, 01/03/2001 _

2 TE 2003-064 Station Blackout PUSAR input 3 GE-VYNPS- N/A Letter, Michael Dick (GE) to Craig Nichols (ENOl), VYNPS EPU Task T0400: Decay Heat AEP-148 for Containment Analysis dated March 10,2003.

4 VYC-2282 0 Vessel Stored Energy with GEl4 Fuel at 20% Power Uprate, dated 519103 5 NUMARC 87- N/A NUMARC 87-00, dated 11/20/87, Including NRC accepted errata and 0 &A's from 00 MUMARC seminars and Topical Report F.

6 NW 91-98 N/A Letter, USNRC to VYNPC, NW 91-98, 'Vermont Yankee Station Blackout Analysis,'

June 5, 1991 7 NIA ASME Steam Tables 8 N/A W Technical Specification _

9 VYC-2270 0 VY GE 14 Appendix Rat 20% Power Uprate, dated 05/0912003 .

10 VYC-415 0 Appendix RIRCIC, HPCI &ECCS Room Cooling, dated 4/29/1986 11 VYC-415 0 Appendix R/RCIC, HPCI &ECCS Room Cooling, dated 910412002 CCN 0 _

VYAPF0017.07 AP 0017 Rev. 8 Page 1 of 4

Significant Critical Difference Affected Reference

Reference # ** DOC # REV # ***Document Title (including Date, if applicable) Review tt Program (/)

12 VYC-888 2 Station Blackout Documentation Analysis, dated 9/04/2002 CCN3 13 VYC-2279 0 Evaluation of EPU Impact on Ambient Space Temperatures During Normal Operation, dated 0411/2003 14 VYC-1347 0 Main Steam Tunnel Heatup Calculation, dated 11/1/96 15 OT-3122 19 Loss of Normal Power, 04/18/00 16 VYC-1628D 0 Torus Temperature Response to Appondix R and Station Blackout ScenarIos, dated CCN02 VYC-0886 Rev 2 CCN04, Page 3 of 7 VYAPF0017.07 AP 0017 Rev. 8 Page 2 of 4

VY CALCULATION DATABASE INPUT FORM (Continued)

Design Output Documents - This calculation provides output to the following documents. (Refer to Appendix A, section 5) tttCritical

- - - Affected Reference

Reference # ** DOC # REV # Document Title (including Date, if applicable) Program (M)

VYC-1432 4 VY Vessel Level for Appendix R, 05/17/1996 VYC-1458 0 Station Blackout Load Capacity Analysis, 10/15/1996 VYC-1628 0 Torus Temperature and Pressure Response to Large Break LOCA and MSLB Accident CCN3 Scenarios, 3/21/2002 VYC-1628D 0 Torus Temperature Response to Appendix R and Station Blackout Scenarios-dated November 5, 1998.

VYC-2159 0. VY-Cycle Independent Decay Heat-Comparison Between ORIGEN-2 and ANSIJANS 5.1-1979 Standard, 2/27/2001 VYC-2314 0 Minimum Containment Overpressure for Non-LOCA Events, 9/03/2003 DBD SADBD DBD V DBD CPS DBD /

DBD HPCI DBD DBD 1{VAC DBD /

DBD MS DBD DBD NBVI DBD /

DBD RCIC DBD DBD RHR DBD /

SSCA Vol 1 Appendix R VYC-0886 Rev 2 CCN04, Page 4 of 7 VYAPF0017.07 AP 0017 Rev. 8 Page 3 of 4 LPC#6

VY CALCULATION DATABASE INPUT FORM (Continued)

Reference #- Assigned by preparer to identify the reference in the body of the calculation.

- Doc # - Identifying number on the document, if any (e.g., 5920-0264, G191172, VYC-1286)

- - Document Title - List the specific documentation in this column. "See attached list" is not acceptable. Design Input/Output Documents should identify the specific design input document used in the calculation or the specific document affected by the calculation and not simply reference the document (e.g., VYDC, MM that the calculation was written to support.

- - - Affected Program - List the affected program or the program that reference is related to or part of.

t If "yes," attach a copy of WVY Calculation Data" marked-up to reflect deletion (See Section 3.1.8 for Revision and 5.2.3.18 for CCNs).

If the listed input is a calculation listed in the calculation database that is not a calculation of record (see definition), place a check mark in this space to indicate completion of the required significant difference review. (see Appendix A, section 4.1.4.4.3). Otherwise, enter tt "N/A" ltf If the reference is UFSAR, DBD or Reload (IASD or OPL), check Critical Reference column and check JFSAR, DBD or Reload, as appropriate, on this form (above).

Note: All calculations in the Design Output list were reviewed. No revision required.

Other Design Output were reviewed. The following revisions are required:

1) DBDs referencing VYC-886 Rev2 need to be addressed.

2) Calculation VYC-1347 should be addressed for EPU.

VYC-0886 Rev 2 CCN04, Page 5 of 7 VYAPF0017.07 AP 0017 Rev. 8 Page 4 of 4 LPC#6

Page 6 of 15 VY CALCULATION REVIEW FORM Calculation Number VYC-0886 Revision Number: 2 CCN Number:- 04

Title:

Station Blackout Documentation Analnsis Reviewer Assigned: Alan Robortshaw Required Date:.-February 2004 E Interdiscipline Review 3 Independent Revilew Comments* Resolution

1. Assumntions on Page 1 of Att. A need Reference. 1.Added

2. Need Reference for Table 1 in Att. A. 2 Done --

3. On page 5 of Att. A please state the TS CST Inventory. 3. Added TS CST inventory r90- - /-,

(

6W I L

_, 1k%:A4- I I/tkr) r Mo a 1% /

Sc6 vAS ,.IHo .

Or 14. Ad

7 Reviewer Signature Date / Calculation Preparer (Comments Resolved) Date Method of Review: 0 Calculation/Analysis Review El Alternative Calculation _ _ _

_- _ _ r_ e v_ _ ,_ l_

AGO_

Eli Qualification Testing Reviewer Signature (Comments Resolved) Date

Comments shall be specific, not general. Do not list questions or suggestions unless suggesting wording to ensure the correct interpretation of issues.

Questions should be asked of the preparer directly.

VYC-0886 Rev 2 CCN04, Page 6 of 7 VYAPF0017.04 AP 0017 Rev. 8 Page I of 1

Calculation VYC-0886 Rev2 CCN04 page 7 of 13 Page 7 of 7 VY CALCULATION OPBN ITEM LIST Calculation Number: VYC-886 Revision Number: 2 CCN Number: 4 Open Item Resolution Method of OI Tracking or Date Closed DBDs referencing VYC-886 Rev2 need to be verified for changes to torus temperatures.

VYC-1347 needs to be CCN for EPU VYC-0886 Rev 2 CCN04, Page 7 of 7 VYAPF0017.05 AP 0017 Rev. 8 Page 1 of 1

Calculation VYC-0886 Rev 2 CCN 04 Page 1 of 6 Attachment A Reason for Revision Revision 2 of VYC-0886 is updated to incorporate the EPU changes This CCN incorporates:

1) Condensate Inventory Requirements at EPU incorporating: -

- The decay heat at EPU from Reference 3.

- Vessel stored energy at EPU from Reference 4 (VYC-2282).

2) Loss of ventilation

3) Torus Temperature Assumptions (same as in reference 1)

1. No off-site power available (SBO)

2. The reactor depressurizes from 1095 psia to 100 psia during the SBO scenario. The 1095 psia is assumed to be an -average SRV setpoint. The 100 psia is assumed a low pressure setpoint where RHR system is deployed for shutdown cooling.

3. It is assumed that at about 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , the vessel pressure decreased to about 100 psia.

4. It is assumed that at 100 psia the fluid and solids in the reactor vessel are at the same temperature.

This is a reasonable assumption, since at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , most of the metal in the vessel will be at the fluid saturation temperature.

Condensate Inventory Requirements The inventory required for decay heat removal will be calculated using a formula given in NUJMARC 87-00 and also using the decay heat calculated in Reference 3.

Condensate Inventory to Remove Decay Heat From Reference 5 V = 35.55 gal/MWt = 35.55

1912

1.02 = 69331 gallons for 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months From the decay heat calculation Q decay at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months (interpolated in the integrated decay heat table - Table 1, Reference 3, next page) 20000 4.68951E+08 28800 6.11079E+08 40000 7.91978E+08

Calculation VYC-0886 Rev 2 CCN 04 Page 2 of 6 Attachment A Table 1 - Integrated Decay Heat for 20% power Uprate Time (sec) GE 2 sigma P/Po Integrated Integrated Kwsec Integrated, BTU 0.00000 1.00000 0.00000 0.00000 0.00000 0.10000 0.99210 0.09961 1.94254E+05 1.89980E+05 0.15000 0.96250 0.14847 2.89552E+05 2.83182E+05 020000 0.93280 0.19585 3.81959E+05 3.73556E+05 0.40000 0.74710 0.36384 7.09580E+05 6.93969E+05 0.60000 0.59080 0.49763 9.70503E+05 9.49152E+05 0.80000 OA9380 0.60609 I.18203E+06 1.15602E+06 1.00000 0.33880 0.68935 1.34440E+06 1.31483E+06 2.00000 0.15480 0.93615 1.82572E+06 1.78556E+06 4.00000 0.06073 1.15168 2.24606E+06 2.19664E+06 10.00000 0.05234 1.49089 2.90760E+06 2.84363E+06 20.00000 0.04546 1.97989 3.86127E+06 3.77632E+06 40.00000 0.03986 2.83309 5.52521E+06 5.40366E+06 60.00000 0.03687 3.60039 7.02163E+06 6.86715E+06 80.00000 0.03466 4.31569 8.41664E+06 8.23147E+06 100.00000 0.03321 4.99439 9.74026E+06 9.52598E+06 150.00000 0.03073 6.59289 1.28577E+07 1.25749E+07 200.00000 0.02909 8.08839 1.57743E+07 1.54273E+07 400.00000 0.02550 13.54739 2.64207E+07. 2.58394E+07

. 600.00000 0.02346 18.44339 3.59690E+07 3.51777E+07 800.00000 0.02197 22.98639 4.48290E+07 4.38427E+07 1000.00000 0.02079 27.26239 5.31682E+07 5.19985E+07 1500.00000 0.01861 37.11239 7.23781E+07 7.07858E+07 2000.00000 0.01707 46.03239 8.97742E+07 8.77992E+07 4000.00000 0.01370 76.80239 1.49783E+08 1.46488E+08 6000.00000 0.01209 102.59239 2.00080E+08 1.95678E+08 8000.00000 0.01114 125.82239 2.45384E+08 2.39985E+08 10000.00000 0.01047 147.43239 2.87529E+08 2.81203E+08 15000.00000 0.00986 198.25739 3.86649E+08 3.78143E+08 20000.00000 0.00918 245.86739 4.79500E+08 4.68951E+08 40000.00000 0.00775 415.22739 8.09793E+08 7.91978E+08 60000.00000 0.00699 562.63739 1.09728E+09 1.07314E+09 80000.00000 0.00647 697.19739 1.35970E+09 1.32979E+09 86400.00000 0.00633 738.15099 1.43957E+09 1.40790E+09 100000.00000 0.00608 822.54579 1.60416E+09 1.56887E+09 150000.00000 0.00539 1109.27079 2.16334E+09 2.11575E+09 172800.00000 0.00515 1229.46099 2.39774E+09 2.34499E+09 200000.00000 0.00492 1366A1299 2.66483E+09 2.60621E+09 259200.00000 0.00451 1645.27459 3.20868E+09 3.13809E+09 345600.00000 0.00406 2015.28259 3.93028E+09 3.84382E+09 400000.00000 0.00384 2230.16259 4.34935E+09 4.25367E+09 432000.00000 0.00373 2351.26659 4.58553E+09 4.48465E+09 600000.00000 0.00327 2939.18259 5.7321 1E+09 5.60601E+09 800000.00000 0.00290 3556.48259 6.93599E+09 6.78340E+09 864000.00000 0.00281 3739.39459 7.29272E+09 7.13228E+09 1000000.00000 0.00265 4110.81059 8.01707E+09 7.84069E+09

Calculation VYC-0886 Rev 2 CCN 04 Page 3 of 6 Attachment A Q = M (h(g) - hal)) to calculate the inventory requirement Where:

h(g) (Reference 1) = 1187 Btu/lbm (average between 1095 and 100 psia) [see note on page 22 of Reference 1]

ha) = 118 Btuflbm (150 TF conservative temperature of CST) v(l) = 0.01634 ft3ilbm @ 150 0F All properties are from Reference 7.

6.11079 E8

0.01634

7.48

= 69867 gal (1187 -118)

This inventory matches very well the NUMARC formula and it will be used.

Therefore the inventory requirement for 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months of decay heat is 69867 gallons.

Condensate needed to remove the vessel stored energy to depressurize from 1095 psia to 100 psia.

Stored Energy in Fluid The fluid energy at full power (EPU conditions is) (Reference 4)

Fluid (EPU, t=0.0) Mass Enthalpy Total Energy (ibm) (Btu/lbm) (BTU)

Liquid 386,971 525.54 2.03369E8 Steam 13,186.12 1191.05 0.15705E8 Total I 2.19074E8 The fluid energy at 100 psia is not changed from Reference 4. The level will be the same after depressurization for current licensed power (CLP) as for EPU. Hence the volumes of steam and liquid will be the same, as well as the enthalpy.

Fluid (depressurized at Mass Enthalpy Total Energy 100 psia) (Ibm) (Btu/lbm) (BTU)

Liquid 510322.4 298.4 1.5228E8 Steam 703.13 1187.2 0.008E8 Total . 1.5311E8 Thus, the difference in fluid energy: AEfluid = 2.19074e8 - 1.531 1E8 = 0.65964E8 Btu

Calculation VYC-0886 Rev 2 CCN 04 -Page4 of 6 Attachment A Stored Energy in Solid (From Reference 4)

Solid Total Solid Energy Heat Conductor Effective (EPU, time= 0.0) (BTU) Temperature (0F)

Liquid Exposed 0.9604399E8 601.24 Steam Exposed 0.25155507E8 609.18 Total 1.211995E8 602.83 Q = MCp AT = MCp (602.83 - 32) = MCp 570.83 MCp = 1.211995E8 /570.83 = 212321.53 Btu 1OF Tsat @ P= 100 psia = 328 0F (ASME Steam Tables- Reference 7)

At 100 psia:

Q= 212321.53 * (328-32) = 0.628472E8 Btu Total Energy removed from structures: AE struck 1.211995E8 - 0.628472E8 0.58352E8 Btu Total energy removed from the vessel during depressurization _

AEfluid + AE structue = 0.65964E8 + 0.58352E8 = 1.24316E8 Btu The inventory needed to remove this heat =

V = 1.24316E8*(.01634)*7.48 =14,214gallons (1187-118)

Hence, the total inventory requirements = 69867 gallons +14,214 gallons = 84081 gallons

Calculation VYC-0886 Rev 2 CCN 04 Page 5 of 6 Attachment A For the CLP (Reference 1) the total CST inventory requirements for removing the decay heat and vessel stored energy = 75,837 gallons.

The TS (Reference 8) CST inventory of 75000 gallons is exceeded at both EPU and CLP for the 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months coping duration.

The HPCIRCIC taking suction from CST has to make up for the vessel leakage (TS allowable and pump seal leakage of 61 gpm - Reference 6, page 17 of TER, also used in both Appendix R analysis (Reference 9, VYC-2270) and in Reference 1. The leakage amount does not change for EPU.

The needed CST inventory to account for leakage; V= 61 gpm

8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months *60 min/hour = 29, 280 gallons.

This inventory, added to that already calculated for decay heat and depressurization would total:

V = 29280 + 84081 = 113361 gallons, which would normally be available from CST but, if not could easily be made available from the torus. Therefore, the conclusions of VYC-886 Rev2 that the Technical Specifications CST inventory requirement of 75000 gallons is not adequate for an 8 hour9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months duration is valid at EPU. However, with Alternate .AC (Vernon Tie) and low pressure systems available, sufficient inventory is available from the torus for the required 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months .

The power uprate results in a need for more inventory, 113361 gallons versus 105117 gallons at CLP.

Reactor Coolant Inventory The depletion of the available inventory in CST will not jeopardize reactor coolant inventory because makeup inventory can be provided from the torus. When the torus temperature exceeds 140 0 F and RCIC and HPCI can not be used with suction from the torus, reactor inventory can be provided from the torus via low pressure pumps. Since VY is an Alternate AC plant, crediting use of the low pressure pumps is acceptable. This conclusion is unaffected by power uprate.

Loss of Ventilation The heat-up due to the loss of ventilation due to an SBO event for RCIC Room, HPCI Room, Main Steam Tunnel, Control Room, Switchgear Room, and Intake Structure is addressed in Reference 1.

RCIC Room The heat-up calculation is based on VYC415 RevO (Reference 10 modified by CCN 1 (Reference 11)). The heatup is based on the piping temperature, RCIC turbine and Switch Heat Loss. The RCIC Room Temperature calculated in Reference 12 is unaffected by EPU.

Calculation VYC-0886 Rev 2 CCN 04 Page 6 of 6 Attachment A HPCI Room.'

The heat-up calculation is based on heat loads from VYC-0415 Rev 0 (Reference 10). The heat loads are from the piping and the HPCI turbine. The heat loads are unaffected by power uprate. Therefore the calculation for HPCI room heat-up is not affected by EPU.

Main Steam Tunnel The issue is isolation of HPCI and RCIC on high steam tunnel temperature.

Reference 13 calculated an increase of 0.6 0F in the normal temperature of the steam tunnel, at EPU.

The conclusion of VYC-886 Rev 2, that the main steam tunnel heat-up is slow on loss of ventilation and the reactor will already be in the process of cool-down, is valid at EPU.

Furthermore, Reference 14 (VYC-1347) concluded that the heat-up in the main steam tunnel is less than that required to isolate BPCI and RCIC. For the case when the feedwater and main steam isolates (SBO conditions), the peak room temperature from Reference 14 is 174 0 F (isolation temperature assuming loop accuracy) at approximately 18 hours2.083333e-4 days 0.005 hours 2.97619e-5 weeks 6.849e-6 months . Based on the results of Reference 13, the change in Main Steam Tunnel Heatup will be very small at EPU. Furthermore, procedure OT-3122 (Reference 15) limits ?PCI & RCIC operation to 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months ; hence the reactor pressure after 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months should be low enough to permit operation of the Low Pressure Pumps (CS and RHR).

Therefore the impact of power uprate on heatup of the Main Steam Tunnel is negligible. It is recommended that calculation VYC-1347 be updated for EPU conditions.

Control Room Restoration of ventilation in the Control Room is governed by Procedure OT-3 122 and is unaffected by Power Uprate. Control Room Heatup for loss of ventilation is unaffected by power uprate.

Switchgear Room The heat loads in the switchgear room are unaffected by the power uprate.

Intake Structure The heatup of the Intake structure on loss of ventilation with only 2 Service Water available is unaffected by the power uprate since the heat loads in the intake structure are unaffected by power uprate.

Torus Temperature The Torus Temperature calculation for SBO at power uprate was performed in Reference 16 (VYC-1628D CCN02). The peak suppression pool temperature is 187.9 'F.

BVY 05-072 Docket No. 50-271 Exhibit SPSB-C-52-2 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Calculation VYC-1347, Rev.0 Total number of pages in this Exhibit (excludino this cover sheet) is 29.

DI.

7-X /7

,,- n ,-L ORIGINAL: PAGE 1of L11L PAGES Rev. 1: PAGE 1 of - PAGES Rev. 2: PAGE 1 of PAGES Rev. 3: PAGE 1 of - PAGES QA RECORD? IMS NO. M02.01.05 _

9' YES RECORD TYPE -09.C16.004

- NO W.OJP.O. NO. 4055 YANKEE NUCLEAR SERVICES DIVISION CALCULATION/ANALYSIS FOR TITLE Maln *team Tunnel Heatup Cakulation PLANT Vermont Yankee CYCLE N/A CALCULATION NUMBER Q1347 -

I U PREPARED BY REVIEWED BY APPROVED BY SUPERSEDES IDATE _,DATE,"y /DATE CALCJREV. NO.

ORIGINAL //I-f-301 REVISION I REVISION 2 REVISION 3 . I KEYWORDS GOTHIC: Room: Heat-up: RRU

(> IhUCLEAR SERVICES DIVISION OF YANKEEATOMIC ELECTRIC COMPANY 580MAIN STREET.

BOLTON, MASSACHUSETTS 01 740 I ;

YANKEE NUCLEAR SERVICES DIVISION CALC. NO. VYC1347 REV.._ DATE 11-1-96 TlTLE Main Steam Tunnel Heatup Calculation PREPARED BY- Jm Pnpoas REVIEWED BY PAGE 2 OF_

TABLE OF CONTENTS Section Descrintion Pa &e LIST OF TABLES .................. ,,,,,. 3 LIST OF FIGURES ........................................................................ ,. 4 1.0 PROBLEM DESCRIPTION ... 5 1.1 Objective .......... 5 1.2 Method of Solution. 6 1.3 Design Inputs .............. ;.;.-.6 1.4 Assumptions. 9 2.0 PROBLEM ANALYSIS .. 10 2.1 GOTkHIC Model Input ................................................... 10 2.1.1 ControlVolumes .......................... 10 2.1.2 Thermal 2~~~.1.2 Conductors ........ ........................ ;...........................,., "I TherlCndcos 1

2.1.3 Heaters

Main Steam Isolation .Vl.ve..........................................; ....... 1 2.2 GOTHIC Runs .. 20 2.2.1 Run MST .20 2.2.2 Run MST2 .24

3.0 CONCLUSION

. . .29 4.0 R F R N E ............................................. ........................ .. . 30

4.0 REFERENCES

.;;30)

Appendices A GOTHIC Run MSTI .35 B GOTHIC Run MST2 ..................... ,. 81 C Computer Code Evaluation ................... 112 D Reviewer's Comments ...................... 114

. i YANKEE NUCLEAR SERVICES DMSION CALC. NO. MYC-1347 REV. _ _ DATE 96 TITLE Main Steam Tunnel Heatup CaTulatfon PREPARED BY. Jim Pappas REVI EWED BY MAGE3 OF____

LIST OF TABLES Table No. Descrintion Pare 1.3-1 Thermo-Physical Properties..................................................................... 7 1.3-2 Outer Environment Temperatures ...........................................................

2.1-la M ST2 Dimensional Data.......................................................................... 15 2.1-lb. W est Wall Dimensional Data................................................................... 16 2.1-2a .Piping Design Data ............... 17 2.1-2b Piping Heat Transfer Surface Areas.......................................................... 18

YANKEENUCLEAR SERVICES DIVSION CALC. NO. YC-13347 REV. - DATE 11-1-96 TITLE Main Steam Tunnel Heatup Calculafton PP'PARED BY: Jm Papps REVIEWED BY PAGE 4 OF LIST OP FIGURES Eim=No. Descriotion EAge 2.1.3-1 MSIV Valve Outline Drawing ......................................................... 19 2.2-la MST1 Schematic .......................................................... 22 2.2-lb MST1 Temperature Profile ........................... .............................. 23 2.2-2a MST2 Schematic ......................................................... 26 2.2-2b MST2 Temperature Profile to Seven Days ............................... 27 2.2-2c MST2 Temperature Profile to Four Hours ............................... 28 A-1 MST1 Schematic .......................................................... 38 A-2 MSTI Input Tables ......................................................... 39 A-3 . MST1 Graphical Results ................. ;., ....................... 55 A-4 MST1 Output Verifying the MSIV Model ................................................... 70 A-5 MST1 Output Showing Condensation Heat Transfer Fluctuations ............. 73 A-i MST2 Schematic .......................................................... 82 A-2 MST2 Input Tables .................... ; ........... 83 A-3 MST2 Graphical Results .................... 98

YANKEE NUCLEAR SERVICES DIVISION CALC. NO. VYC1347 REV. DATE *1-1-96 TnTLE Main Steam Tunnel Heatut Calculation PREPARED BY Jim Pappas REVIEWED BY PAGE 5 OF 1.0 PROBLEM DESCRIPTION This calculation determines the temperature rise in the main steam tunnel with a loss of HVAC to document whether this rise will result in an automatic isolation of the HPCI and RCIC systems; The HPCI/RCIC excess air temperature switches, within the main steam tunnel, provide automatic isolation of the HPCI and RCIC lines if a temperature of 185-F

5F is sustained for longer than 30 minutes (References 23 and 24). The loop accuracy is 61F (Reference 36). Therefore HPCI and RCIC isolation can occur at a steam tunnel temperature as low as 174F. This high temperature isolation scheme is for line break protection and it is not intended for non-line break events, such as loss of main steam tunnel cooling under loss of normal power.

Normal ventilation in the main steam tunnel is supplied by the Reactor Building Ventilation System and by RRUs 17A and 17B, located in the tunnel. A Reactor Transfer Fan (RTF-1A/ 1B) exhausts air from the main steam tunnel at approximately 4200 cfm. Each fan has a total capacity of 14,400 cfm and takes inlet air from various locations in the Reactor Building. The RRUs circulate and cool air inside the main steam tunnel. The fan capacity of each RRU is 5000 cfm and serivice water, supplied to coils within the RRUs, provides the cooling.

RTF-IA and both RRUs are powered from 480v MCC 6A. This MCC is NNS and is supplied from 4160v Bus 1. RTF-1B is powered from 480v MCC 7A, which is also NNS, and is supplied from 4160ir Bus 2.

Controls are located on the Turbine Building HVAC control panel, with auxiliary indications on the Control Room 9-25 panel. Typically, one fan and both RRUs are operating, with the second fan in stand-by. The operating fan and the RRUs maintain the main steam tunnel environment temperature at a yearly average of 1252F, as described in the Vermont Yankee Environmental Qualification Program (Reference 25). A review of temperature data for the main steam tunnel indicates that the air temperature can peak at about 150'F during the summer months (Reference 30).

1.1 Oblective The objective of this calculation is to determine the temperature rise of the air in the main steam tunnel, during a loss of normal power and under the following conditions:

Loss of HVAC in the main steam tunnel;

Summer peak temperature for initial and boundary conditions.

Main steam and feedwater lines in the tunnel are:

- Isolated (steam and water are not flowing, GOTHIC run MSTI).

- Not isolated (steam and water are flowing, GOTHIC run MST2).

YANKEE NUCLEARSERVICES DMSION CALC. NO. 2W.Y- 347. REV. DATE 1IA-96 TITLE Main Steam Tunnel Heatup Calculation PREPARED BY Jim Pa6pas REVIEWED BY PAGE --6 OF 1.2 Method of Solution A lumped parameter GOTHIC model of the main steam tunnel is used to calculate the air temperature rise orn loss of HVAC. GOTHIC13 3 A 3

4) is a general purpose thermal-hydraulic computer program for design, licensing, safety, and operating analysis of nuclear power plant conitainment and other confinement buildings. See Appendix C for verification of the GOTHIC version used.

.The models consist of volumes, flowpaths, & themuz conductorsarranged and connected to represent the thermal-hidraulic response of the main Steam tunnel. The thermal mass ofeach conductor is included in the GOTHIC computatilon.

1.3 Design Innuts 1.3.1 The thermo-physical properties for the materials used are shown in Table 1.3-1.

1.3.2 The boundary temperatures for spaces surrounding the main steam tunnel are shown in Table 1.3-2.

IP IIIII!lp* I '.1 YANKEE NUCLEAR SERVICES DIVISION CALC. NO. WC-1347 REV. - DATE AI- ..

TITLE ;Mai Steam Tunnei Heatup Calculation . I R. BY ... R IEWEDBY PAGE -7 OF _

Tabl,- .3-Therngoehytcal Propertes Concreteo 5 1.05 142 0.156 Stwal Plpo'" 6825.0 467 0.11

.1000 0.073

.. o900. 0.067 800 0.060 700 0.055 Insulation 7 600 '

0.050 0.10 500 0.046 400 0.042 300 0.038 200 0.036 100 0.033 Notes a) Superscript numbers refer to References InSection 4.0.

b) Assumed.

I

I----

YANKEE NUCLEAR SERVICES DMSION CALC. NO. M4^347 REV. DATE 11.1-9 TITLE Main Steam Tne H calculation PREPARED BY Jim Papnas REVIIEWED BY PAGE 8 OF Table 1.3-2 Outer Environment Temperatures Normal 10F15F101 3F2 Operation 15F 125F IOOF 83Fs)

Peak Summer 160"F 130°F 120°F 90F 0 ) l Notes a) Vermont Yankee FSAR (Reference 32), Table 2.3.2. Highest mean daily maxirrum for summer months.

b) Vernont Yankee FSAR (Reference 32), Section 10.12, Summer design temperature.

YANKEE NUCLEAR SERVICES DIVSION CALC. NO. YC1347 REV. - DATE I1IM .

TITLE _Main Steam Tunnel HeatuD Calculaion PREPARED BY Jim Paopas REVIEWED BY PAGE 9 OF 1.4 Assumptions The critical assumptions used in the GOTHIC models are as follows:

1.4.1 Initial main steam tunnel air temperature is 1506F. This is based on Reference 30 and is considered conservative. Reference 30, describes that this value is derived from a temperature element that is close to hot process lines. Therefore, the corresponding ambient room temperature should be lower.

1.4.2 Initial main steam thermodynamic statepoint is saturated steam at 985 psia based on the heat balance shown on Figure 1.6-1 in the FSAR (Reference 32). Therefore, the temperature is 543F.

1.4.3 Initial feedwater thermodynamic statepoint is saturated water at 373F based on the heat balance shown on Figure 1.6-1 in the FSAR (Reference 32). Therefore, the pressure is 179.8 psia.

1.4.4 HPCI and RCIC turbine steam supply temperatures are 543'F.

.1.4.5 Both RRUs are inoperative for the analysis.

1.4.6 The air temperatures in the spaces surrounding the main steam tunnel are listed in Table .1.3-2 and are assumed to be constant throughout the transients.

1.4.7 In model MST1, where the main steam lines are isolated and the feedwater pumps are off, the four main steam lines (MS-1A through D) and the feedwater lines (FDW-14/ 15/16/17) dissipate the heat in the line volume, cooling down as they do so.

All other lines contain fluid at their respective constant temperatures, as listed in Table 2.1-2a.

1.4.8 In model MSM2, where the main steam lines do not isolate, all lines contain fluid at their respective constant temperatures, as listed in Table 2.1-2a.

1.4.9 Miscellaneous piping, steel, and equipment are left out of the models.

1.4.10 The floor is left out of the models to add conservatism to the room heat-up.

1.4.11 The west wall contains a metal section through which the main steam lines pass and which t*o blowout panels are installed. This metal section is modeled in the GOTHIC runs. However, other non-concrete wall sections are not. They include:

a) ventilation duct with blowout damper in the north wall, b) a blowout panel and a blowout door in the south wall, and c) various duct work and pipe sleeves.

This is assumed to be crnservative since It Inhibits natural circulation that would normally exist in the room.

YANKEE NUCLEAR SERVICES DMSION CALC. NO. 347 REV. DATE 11J146 TITLE Main Steam Tinnel Heatup Cal -laton PREPARED BY Jim Pappas REVIEWED BY PAGE 10 OF 2.0 PROBLEM ANALYSTS 2.1 GOTHIC Model Input The following sections describe the major input that was calculated for the GLTHIC models.

2.1.1 Control Volumes Main Steam Tunnel The main steam tunnel is Volume 1 in the models. The relevant dimensional data of the main steam tunnel for construction of the models are shown in Table 2.1-1. From these data, the wall surface areas and the room overall volume are obtained. The volume of the steam tunnel is:

Van = (North Wall) x (East Wall) x (Height)

Vusr = 36.25' x 24' x 25.5' = 22,185 ft The hydraulic diameter is:

D, =4A where A is the cross sectional area of the volume (i.e., the ceiling or floor area) and P., is the wetted perimeter. P. is defined by GOTHIC as S/h or the surface area of all structures divided by the height of the volume. S would, therefore, be the total surface area of all the walls and the ceiling. The floor is not modeled.

S V =-4(A~)h Ah Es+ At4a + A.ut + Asitl + Ace.a Dh = 4870X25.8) hY.4.38 + 924.38 + 612 + 612 + 870 D= 225 ft In the run where the main steam and feedwater lines are isolated, those lines are modeled as separate control volumes. The four main steam lines are lumped into one volume as are the two feedwater

.. I YANKEE NUCLEAR SERVICES DMSION CALC. NO. VYC-134Z REV. DATE 11-i-95 TlTLE Main Steam Tunnel Heatup CalculatIOn PREPARED BY Jim PaDOas REVIEWEDBY B PAGE_ 11 OF lines. The volumes are obtained from the pipe data in Table 2.1-2b, and follow. The hydraulic diameters are simply the pipe diameters.

Main Steam Lines VMSL 4 x -At)2 x L VlSL 4 X (8.062 1 12)2 x 46.5 VMSL a 263.75 ft Feedwater Lines VFDW = 4a)2t 1 4 1n6 + LI.T)

V. =(6.7811 12)2(45.7 + 47.9)

VFW - 93.90 ft3 2.1.2 Thermal Conductors The input for the thermal conductors that represent the steam tunnel walls and ceiling is taken from Table 2.1-1. The floor is left out of the model to add conservatism. Typically there is little heat transfer through the floor of a heated room.

The thermally significant piping found in the main steam tunnel are described in Table 2.1-2. The GOTHIC input for these conductors is also shown in the table.

2.1.3 Heaters

Main Steam Isolation Valves The main steam isolation valves (MSIV) in the steam tunnel have a substantial amount of un-insulated structure that makes up the yoke and actuator. Figure 2.1.3-1 (Reference 39) shows the outline of the valve. Heat will conduct through and out of this structure into the main steam tunnel There are four such valves in the tunnel.

The yoke of each MSIV consists of four 3" solid i-ods attached to the bonnet (Reference 40). The yoke acts as a support for the actuator and as a spring guide. Through the center of the yoke, the valve stem travels.

YANKEE NUCLEAR SERVICES DIVISION CALM. NO. VC1347 REV. - DATE j1-1-96 TITLE Main Steam Tunnel Heatup Calculation PREPARED BY Jim Pappas REVIEWED BY PAGE 12 OF The method of modeling the heat transferred to the main steam tunnel by this assembly will be to treat each 3' yoke rod and the stem as a fin. The stem will be modeled as though it were a fifth yoke rod.

Therefore, each MSIV will be modeled as having five 3' solid rod fins heated at one end. The heated end is that attached to the body of the valve. It is assumed that the actuator is far enough away from the valve body that any heat conducted to it is negligible.

GOTHIC cannot model this situation because it involves two-dimensional conductive heat transfer.

GOTHIC can only model one-dimensional conduction. Therefore, a formulation of the heat rate provided by the yokes will be derived here and input into GOTHIC as a 'heater'.

The general equation for such a fin is (Reference 29):

q = b-kA(T. - T,3)tanh(mL) where: q = heat rate (Btulhr) h = convective heat transfer coefficient (Btu/hrfPF)

P = perimeter of the fin (td ) (ft) d = diameter of the rod (at) k thermal conductivity of the rod material (BtulhrrtF)

A = cross sectional area of the fin (f?2)

T. = temperature of the heated end ('F)

Tc = ambient room temperature ("F) m= (ft)

L = length of the fin (ft):*

The values in the following table will be used. The value for h is taken from Reference 38 and is considered to be conservative. In a transient calculation, it would be expected to vary around a value of 0.5 Btu/hr ft2 .F to 1.0 Btu/hr ft2 -*F. The length, L, Is taken as the 'AC! dimension from Figure 2.1.3-1. This Is clearly much longer than the actual length of the yoke. However, the yoke dimension is not given. So, the more conservative, longer length is arbitrarily used. This presents hardly any penalty in heat rate to the room because the value is used in the tanho function which is barely sensitive to the length. For example, using the 9 ft value tanh(9) = 0.9999 and using half that value tanh(4.5J - 0.9998.

M_

YANKEE NUCLEAR. ,1 SERVICES DIVISION CALC. NO. Ct347L REV. DATE 11-1-96 I

TITLE Main Steam Tunnel Heatup Calculation PREPARED BY aim Pappas REVI EWED BY - PAGE 13 OF h 1.65 Btu/hre#F RpFerence38

=31n d Referetice 40

=0.25 ft P =x(0.25) 20.785 ft

k 25 BtulhrftF Table 1.3-1 2.

=

4 A = 025Y 4

= 0.049 ft2

. m = 1.85 x 0.785.

25 x 0.049

= 1.028ft1 L = 108 In Figure 2.1.3-1 L 9 ft (See discussion above)

The temperatures will be taken from the GOTHIC run using control variables. This will allow the temperature difference to vary with time to more accurately represent the changing heat transfer rate.

The source temperature, T., will be taken as the temperature inside the main steam line. This is highly conservative since the more appropriate value would be that rf the bonnet. Calculation VYC-660 (Reference 40) is a state-point calculation of the heat conduction through the same MSIV structure. For the state-point modeled In VYC-660, the steam Inside the pipe is modeled at 545'F and the bonnet temperature is calculated to be about 375'F. So, as expected, the bonnet is cooler than the steam inside the pipe. However, the assumptions in VYC-660 are not all consistent with those of the present calculation and a definitive correlation between these two temperatures is not readily derivable. Therefore, using the steam temperature is certainly conservative since.it is clearly bounding I.

- the bonnet can never be hotter than the steam.

YANKEE. NUCLEAR SERVICES DMSION CALC. NO. VYC.1347 REV. DATE 11--9

.TITLE Main Steam Tunnel Heatup Calculation PREPARED BY Jlm Panoas REVIEWED BY PAGE 14 OF The ambient temperature, T.,, is the bulk room temperature calculated by GOTHIC.

So, the heat rate generated by a single yoke rod is:

q =,J(t65 x0.785 x 25 x0.049X - T.0)tanht(028 x 9) q = 126( - T.,) Btulhr Each of the four MSIVs is to be modeled as having five such rods, and GOTHIC requires input in units of Btu/sec. So, the final input to GOTHlC is-q =4 MSIVs x 5 Rodsx 1 hre0 x 126(T- T.)

q 0.007(r. - To.) Btu/s In GOTHIC this will be represented as a heater with a heat rate of 0.007 Btu/s multiplied by a forcing function. The forcing function Is in turn equated to a control variable. And, the control variable represents the temperature difference between the main steam line and the room average of the tunnel.

I .

YANKEE NUCLEAR SERVICES DMSION CAMC. NO. MWC-347 REV. DATE 1141-96 TITLE Main Steam Tunnel Heatup Calculation PREPARED BY Jim Panpas REVAEWED BY PAGE 15 OF Tabm e 2-1a Mai Sta : Tune amesoniDi II Superscript numbers refer to References InSection 4.0.

YANKEE NUCLEAR SERVICES DIVISION CALC. NO. Wo1347 REV. - DATE 1-96 ITLE : Main Steam Tunnel Heatup Calculation PREPARED BY Jm Panoas REV IEWED BY PAGE X OF ak~2.,A-h West ll DiensioalDt Ln24d 24 .0l2' 24 Height I Wkith 25.5"1) 8.0(12) 255 Area 612.0 192 420 2)

(fet _ _ _ _ _ _ _

(f)M5r.3 Notes Superscript numbers refer to References in Section 4.0.

I

YANKEE NUCLEAR SERVICES DIVSION CALC. NO. McJ347 REV. - DATE 114-96 TITLE_ Main Steam Tunne Heatuo Calculation PREPARED BY Jampa=p- REVIEWED BY PAGE 17 OF abe2.J2a PEn Dzgn-at I'

a) Reference 27 says 2.5S b) HPCI Steam Supply C) RC1C Steam Supply d) FSAR (Reference 32), Figure 6.4-1 (Highest temperature at Location 2) e) Superscript numbers refer to References in Section 4.0.

f) RCIC Discharge g) HPCI Discharge h) FSAR (Reference 32), Figure 4.7-3 (Highest temperature at Location 3) i) Assumption 1.4.2 J) Assumption 1.4.3

YANKEE NUCLEAR SERVICES DIVISION CALC. NO. -VYI14Z7 REV.. DATE 1 TiTLE MarinSteamTunnell-leatuoCalculatn PREPARED .BY Jini Paopas REV IEWED BY PAGE - 18 OF Piping Heat Transfer Surface Areas MS41B 8.062 9.000 46.5(. 304.3 MS-IC 8.062 9.000 46.56 304.3 mS-ID 8.062 9.000 46.5(on 304.3 MS-4A0 4.781 5.375 17.6('") 81.8 MS-4B9' 4.781 5.375 352"'9) 135.9 FDW-14116 6.781 8.000 45.7(') 239.3 FDW-15M7 6.781 8.000 47.9M 250.8.

MS-SA(') 1.312 1.750 46.72) 91.7 I.

RCIC-1" 1.812 2.250 28.7(") 33.8 RCIC-2m 1.812 2.250 29.7"') 35.0 RCIC-BB 1.812 2250 6.5p1) 14.5 i

RCIC-BAv 1.812 . 2.250. 11.4c".) 25.4 3

HPCl-15B" 5.906 7.000 40.6cm 202.

HPCI-15A'a) 5.906 7.000 9.2m 45.8 Notes a) Area = 2is(Outer Radius + Insulation) x (1 ft12 In) x Length b) HPCI Steam Supply c) RCIC Steam Supply. .

e) Superscript numbirs refer to References In Section 4.0.

q RCIC Dlscharge g) HPCI Discharge

Ft c 2.

m.~'

0'AIFtOVM A"/C 1-IN TPVAJN C1N ill~

-YANKEE NUCLEAR SERVICES DMSION . CALC. NO. V 134Z REV. DATE 11 TITLE Main Steam Tunnel Heatup Caculatbn PREPARED BY <<JIm Pagoas REVIEVED BY PAGE 2 OF

2.2 GOTHIC Runs 2.2.1 Run MST1 This run of the main steam tunnel heat-up represents a typical loss-of-norm4 power event. On a loss-

'of-pe, the HVAC system trips and the MSIVS and feedwater pumps isolate. The room then heats up betause of the heat gain from the pipes within it. However, the major loads are from the main l

steam and feedwater lines and the fluids in those lines are not flowing. Therefore, their heat gain to

-the room diminishes as the transient progresses and the room eventually peaks out and then begins to

- ~drop tn'temperature. '

This run represents a typical heat-up of the tunnel following loss of ventilation, however many conservatisms are included so that the results are assured to bound a true event. These conservatisms include:

The initial main steam tunnel temperature of 1501F is based on the Reference 30 data and represents a peak room temperature as opposed to an average room temperature. A more representative average (initial) room temperature would be something lower.

Miscellaneous structures and equipment in the room are not modeled. They would act as heat sinks resulting in a temperature rise that is slower than that predicted by GOTHIC.

Miscellaneous 'cold' piping, such as service water piping, is not modeled as heat sinks.

Wall openings such as ventilation ducts/dampers or pipe sleeves are not modeled inhibiting cooling by natural circulation.

Natural circulation through the RRUs is not modeled. The RRUs trip on loss of power however they continue to receive cool service water and would contribute a small amount of cooling.

On a loss of power, HPCI and RCIC would automatically start resulting in flushing some of

the 373F water from the feedwater lines and replacing it with 140F water. This is not accounted for.

The MSIV.heat gain is conservative as described in Section 2.1.3. Most notably, the source temperature of the yoke, modeled as a series of fins, is the steam temperature itself as opposed to the bonnet temperature of the valve.

The flooras a heat sink Is not modeled.

The non-conservatisms in the model are:

Main steam line drains are not modeled. They would add heat to the room but only a small amount because the lines are about 2'A' NPS. This is believed to be counteracted by the lack of 'cold' piping being modeled as well.

There is no account for MSIV leakage that would continuously add a slight heat load to the main steam lines. (Run MST2 in Section 2.2.2 accounts for this).

YANKEE NUCLEAR SERVICES DMSION CALC.'N0 347 1W REV. DATE 11-X-96 TITLE Main Steam Tunnel Heatun Cafcuftion PREPARED BY Jim PaDDaS REVIEWED BY _ PAGE .21 OF

There is no spatial definition in the model, therefore an axial temperature gradient is not calculated. Because the HPCI/RCIC temperature switches are in the upper, ceiling, area they may experience a higher temperature thin that of the bulk room. This is believed to be counteracted by the initial room temperature of 150F which represents a temperature from a hot area of the steam tunnel. (See Assumption 1.4.1).

It is believed that.the conservatisms listed above far outweigh the non-conservatisms. Therefore, it can be assured that the true heat-up profile of the main steam tunnel will be a curve that is below -

and therefore bounded by - the GOTHIC result.

The GOTHIC model MSTl is shown schematically in Figure 2.2-la. It consists of the following:

a Volume 1 representing the main steam tunnel.

Volume 2 representing the main steam lines isolated at t 0 seconds.

Volume 3 representing the feedwater lines with pumps off and no flow at t -0 seconds.

Plow path 1 connecting Volume I with a pressure boundary condition I P. This flow path and boundary condition are used to maintain the pressure within'the main steam tunnel at 14.7 psia as the air heats up.

Thermal Conductors 1-6, 8, 9, 11-16 connecting the heat sources and sinks to Volume 1.

Thermal

. Conductor 7 connecting Volume 2 to Volume 1.

Thermal Conductor 10 connects Volume 3 to Volume 1.

Heater I representing the MSIVs.

Appendix A contains the detailed listing of the GOTHIC input for this run. Included are graphical results and calculations validating the run. The model is run for 7 days to determine the temperature rise profile of the air in the main steam tunnel. The heat-up of the main steam tunnel is shown in Figure 2.2-lb for the full 7 days.

The graph shows that the steam tunnel reaches a peak average temperature of 174*F after approximately 3/4/4 of a day. It then drops during the remainder of the transient. The peak is considered to bound the actual peak that would result during a true loss-of-ventilation scenario because of the conservatisms discussed above.

.. ... i

. . S .

Shm Tunnel - UiSTI F5;M. .Z -la W~OdatOtAD4 1696I'.

OThf Wiulon 4.1(OAoMa19

, _I j _

.1 ...

,I t1F

-I w t

Id I I

I

l. _ _ _ _ _ _ '_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I IAKEE ATOMIC ELECTRIC COMPANY CALCULATION NO Vyt-1147 ATTACHMENT PACE l

/O-31-36

Swain ToM i r1 - '.

WedOct ao7==9O9 10 GOTHIC Venorn 4.1(A).c May M I99S

. - -.Steam Tunnel Bulk Room Temperature

..' i)V1 co V -I I I l Time (days)

GOTHIC 4.1 (0A0c 10/29/96 15:56:41

YANKEE ATOMIC ELECTRIC COMPANY CALCULATON o VYC-1341 ATTACHMENT PACE f_ Xf Pc-5 0 ZI I

I

YANKEE NUCLEAR SERVICES DIVISION CALC.NO. WN1347 REV. DATE_11-1-M_

TITLE Mahi Steam Tunnel Heltup Calculation PREPARED BY Jim Papas REVIEWED BY . PAGE 24 OF 2.2.2 Run MST2 This run of .the main steam tunnel heat-up represents an extreme loss-of-normal power event. As stated in Section 2.2.1, on a loss-of-power, the MS!Vs and feedwater pumps normally isolate.

However, the possibility exists that one or more isolation function fails - such as a MSIV not closing, Furthermore, as stated in the non-conservatisms of run MSTI, it is more than likely that some leakage would exist past the MSIVs.

Justifiably quantifying such conditions is not straight forward. However, the situation can be bounded. ..The mfost bounding scenario is that none of the maiin steam and feedwater lines isolate and the steam/water continues to flow. The initial temperature of the fluids within these lines is, therefore, constant throughout the transient resulting in a much higher heat gain to the room. This modeling technique also clearly bounds any postulated leakage past isolated MSIVs.

The model, itself, is identical to MST1 except:

the volumes representing the main steam and feedwater lines are removed.

the conductors (7 and 10) that connected those volumes are moved into the main steam tunnel volume as internal conductors.

the boundary heat transfer coefficients on conductors 7 and 10 are fixed temperatures representing the steam and feedwater temperatures.

All other conservatisms and non-conservatisms listed for run MSTI remain in this run.

The GOTHIC model MST2 is shown schematically in Figure 2.2-2a. It consists of the following:

Volume 1 representing the main steam tunnel.

Flow Path I connecting Volume 1 with a pressure boundary condition I P. This flow path and boundary condition are used to maintain the pressure within the main steam tunnel at 14.7 psia as the air heats up.

Thermal Conductors 1 - 16 connecting the heat sources and sinks to Volume 1.

Heater 1 representing the MSIVs.

Appendix B contains the detailed listing of the GOTHIC input for this run and graphical results. The model is run for 7 days to determine the temperature rise profile of the air in the main steam tunneL The heat-up of the main steam tunnel is shown in Fig'ures 2.2-2b for the full 7 days and 2.2-2c for the first four hours. The first four hours is of particular interest for Appendix R scenarios.

I YANKEE NUCLEAR SERVICES DMSION CALC. NO. _ *1.347- REV. , DATE 11-1-96 TITLE M;an StamTunnel Heatuh Calculaion PREPARED BY. jlm PaRoas REVI EVED BY PAGE 25 OF The graphs slow that the steam tunnel reaches an average temperature of 172'F after four hours. It continues to rise until it is about 2071F at 7 days and still rising.

As with MST1, these results are considered to bound the actual temperature rise that would result during such a scenario. Furthermore, because the scenario itself is extreme by nature, the results greatly bound any possible steam tunnel heat rise that may be postulated.

-:-,: - -i... - 'I. '

I

' *d Oct 20 NTA4O I ,Pit;e 2.Z20-2c 001: V:an

_.1(CA).c -MAY I'

. j. ...

.. I i: . .

1 .,'

.X I I I-.

I t:. I 3 : .

.5 . I I .. I L a I EIIAr011C ELECTRIC CoBpANy CALCuLt.riogw. LVY9: _1 I ATUCHMENW ,Gf 2 IDF.l

/0- - 3

I -

Stum Timnd *UST2h auve s Z.z-Zi W~dOqt3OO7$42Z1998' i GOTHIC Vmb1W 4.1(qA)c . May1995 l.:

Steam Tunnel Bulk Room Temperature TVI 0

i cow P-lime (days)

GOTHIC 4.i(OA-c 1IV29/96 16:07:43-I, EE ATOMIC ELECTRIC CoMpANy CALCUtATIONO AlAChMfr No

_ AZ7 G

cA§42 l

.or_l fo-f 6

mmnw SlumTwme.STMN

'Wed OctSO OA21331906 PZ.a-2c.

OHIC Vdlcn'4.;1foA)-c ;May1995 AI ATOMIC ELECTRIC COMPANY AEE CALCULATI(0 -no ATTACNAMIT _ pAr,( ? E l to -.T)^9 e I

--- . I .--

YANKEENUCLEARSERVICES DMSION CALC. NO. UIC*347 REV. DATE .1.9 TITLE - Main Steam Tunnel Heahtu Calculation PREPARED BY' Panias REVIEWED BY PAGE 29 OF

3.0 CONCLUSION

The results of the two GOTHIC runs show the air temperature in the steam tunnel to approach about 172F. In the first four hours of each transient. Both transients show almost the same profile for that time hfranebecause the cooldown of the main steam and feedwater lines is not large enough in the first four hours to have a significant effect on the room temperature rise.

In the case wherethe main steam lines and feedwater lines isolate, the peak room temperature is about 174'F at approximately 18 hours2.083333e-4 days 0.005 hours 2.97619e-5 weeks 6.849e-6 months . In the case where these lines do not isolate, the room temperature rises to 17411 at about 6% hours. It continues to rise and is about 207' at the end of the seven day transient and still rising.

! As discussed in Section 2.2, the many conservatism included in the model offer a high degree of confidence that the GOTHIC results envelop any true heat-up profile of the steam tunnel. Therefore, the actual room heat-up is expected to be something less and it is concluded that HPCI and RCIC would not isolate under the conditions modeled.

The results of this calculation do not affect the FSAR, Technical Specifications, Technical Programs, or controlled drawings.

BVY 05-072 Docket No. 50-271 Exhibit SPSB-C-52-3 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Calculation VYC-1502, Rev.0 l Total number of pages in this Exhibit I (excludina this cover sheet) Is 43.

-*.. .'..'!-rr .I :. . -

Iy.j .

I, I .-

I I .

ORIGINAL: PAGE I OF LtPi3ES.. . -A REV. l: PAGE 1 bE .. *AGdS REV. 2: PAGE }l:OF _ PXG E :

REV.3: PAGE 1 OF_ ~P.A *

: ,'IM 1 - SN*0 . 0 .8.I

1 a... *:QA *.E~RD?- ':

ISN

. *4

EI3.C16 X13413 C091..! -

Y 9. A . . . . , .* , 0. 4055 II -

I.

YANIcE-NU[CLEAR SEkVlcEss DIIION CAU Ao A VSiS FOR.

- I II "

a-1

.. t a.

./... -

a-. '7-..'

...... s.d ..?t444t4+/-4I$?C.

'-4 __________

~ v-b ~ :

~J & Y C L~N A,

.:*-'AC 4.4 9'.

.5. . . 1 4 .

CALCUATIO 5*1.

w A.

- . 44 . . a * < . *

' 9 c 2 J 5 :

t~ 4 f !~v..:~ -.

tR7C' __3 T -

B. B Y . " - R RV1 1 r D

_ -e

ua of HVAQ, VYC-1502 Ftiv.- O..

'TA!LE.OF_,CONM-M'l LO Qbj . . . . . . . . . . . . . . . . . . . 3 3"'

-2A . . . . . . . . . . . . . . . . 4

-r2.2. .. . . . . . . ... .

.U CW,!O

-5 m64-d . . . . . . . . . . . .

v -'-jc X;4 A ni 4 Al 12V ZQ-

'FY 1..4 4

all VA 4 % R Histup:001MG l4w Diw,; '44 0; ..............

le:

OT 6dd';:,

J_ "JilTA31C -'bp.%d.!K-hW 4L

I

!Zr Control ~oom Hestap due to Loss of HVAC; VYC-I502 Rev. 0 1.0 Objective The ob.ective of'thii ca1u*lation is to determine the temperal -re of the control room as a functiop of tie as arijit of a los; of ventilation of the control room doe to an Appendix R fire o . The s criteria f,-,r this analysis is to have the control room maintaned in a habi P 1.one which ensures equipment opeblity. Based on published guideies, both are satifled by maitaining conirol room temperature at or below 120°F [1]

2.0'- MAWiod ztoe c it.itclosed room with concrete waIls, floor and rof. he roof£ north aiid -twallare poed to- ithe otdoor.etnvironment;wihiletlhe west wail is a common wa

.it tije tin bildib. and the south all is a commOn wl with the ractor building a ue's ,p b wl control m roof abve whichbventilation ductingis-rut dco ic - arehelddby the suspended ceiling strite.

The c>nfol .o ascoTry l panels .nd cabinets wit a variety of

.l~ctii~h j i~, tinglhts, p rsupplies and control 5ysstem.

Thc cIii~ t oa ts 4i1 as :3ther, an~inities .(computeessreiefiigeratopy

. i.I he is heis norii~llremoved by an HVAC s

utkia tlions may render the HVAC oan nIfd roo e a'Co4pl of.key parameters. The first key.-..

'. .. p.

Eethom.:, MIs hiaitt must be removed by conction,-

sconuucnn;, mass tfcr.~GiVc51 8&losso5~fve~tionf aiss transfer is-iot available to reifib :di.o;i tM far-e reurd to remove tie heatbthrobh 1-ftMdl h6Viali~ wfi~ec liven-

.nSincrne by convecon and fueU

.l . higher iiut iiebc hierthe iniide a"- temperatur.

en -v vn nis rlatnoftcofu vedog the uiig to the o: -- .- ,&iratof h:#6ul Is dep Kmtbn thetl uai-o

,'o- fhom fth aIrWinc .m tot he inside i 11..

Sf M-t . ,Y. suraeto tieithitd - fc make u te tr , :,c walL Ths th .si is -econd.key ta t.h bigher.

,..the ithquireddifcal temperature betwen the

, :. Zouxide x n l r

he. third key a is itrnheninl conuctors and aects only the transient hea of in odr. I" rs ' ieaiia and'i'al :' resistancecan: be usca it6

§ :p. ,.dict the sted st;stmer~n theoom siuii ,tey stat hieat input equals ie~t"

. ;: remoi.a 'iHo m a~~Iheml con huti can:scetocange the transienttepaur 1'

W601. - . ~~~~cn.

. - ; M:"A.tbang.the'trani.i

= - tem.p-.;w -.................

.1, . F-.,;

2*

I.-

I Conirbi Room Heatup due to Loss of UVAC; WC-1502 Rev. 0 nage S r:spcs by actitig as-a heat sink (as e roo heats up, some of the heat is absorbed by the intensI condcdtot as'it.heats up) or a beat source (if the heat sink started out hotter than the

.sumim w-" give up itbsbeat to the surroundings.)

In thrp key pazaraeteas need to be determined in order to solve the transiet liea. of the cotrol Ioi g a loss of ventia. Thesc parameters were dermined Vot' tobzn ;b1eriied *firndrawings and allows for determining the heat removal

. de .nve dondon T ii configuion of the room mith espect to

. :. I :sXuit-s - iiia is more .pzb~liatic and the-method of deternig these two iex.In.uin the subseent sectionss s co. .oon beatup will be calciilated using th G0TIC computer code [2].

AWb c 100s4cP4 ed in ordefr.Providec aive input for the code., In f4. *et o i udazyaxird it condiions ure cho . The GOTHIC code. :

~g' -. S'-.e-ofctr1 rvoon versus timie which can ten be used in

. . i any .ec~ssary .o rase of loss of entilaion to the control room.

The-GOTHiC comi. . gis themo-hydraulic problems and bas inimii l -ob.1sT(Z Iconuctois, hea sources,.control mecanims, air properties.as a Aem"etue t4

nte bas'beei

,, validae forfthis type of W-' '~~~ ' -'" ~ . -'.'1.

2.1'HeatSreas

Z.. ; sontrol an be classified as follows

beat in eipment. A variety of elerical niczitzl ui t:. rom . brl (compt moitor, relays, :A

. . - co ipbnvemiences such as -eft gcrirs *C machines, microwave ovens,

.a.

.,:ien . if, ...-,- ;,..*............

f&tir-lemt t3-i.o is g.lhger than the mside air temperatue, heat is

- -. S wna room; pWtpon het6d thatbit heai fc offtothe - surrundingp. The totil heat

.~p~ i :ds .. oh thenM er of jople i th coatrol roomat an one tme; Peifity~ n. more .d'(*h enWusIhe activity, the greaer th<<ieat3

{Z-Z.=v . : !:dy' ';.;'.-':.*

tcan :ev~culakd . tir dtifiitdto a fair degree of ccacy. . ..

baa be omie. >Rixw of HYA denotev~ at l@e~m sourc oft binform .i;i

-in.

~~~ -X,~~~~~~. .Ft.. a-V..-.,-t;L ,,,

... I

i III AC; VYC-l5 Rev. 0 ~

data~keidn ilatW -t~ ~ta11ea geniration in the control room. Thi data 4*;eli.

ofibicrnesurdth ikflow througb-the conitrol room Ti% ata an b usd to calcula tehe tatal heat rate ifurs.

cu mid nalsisan tquate beat source cmn be determined f-Kfcrheatfromtheindtifir to th outside ambienitar; h Wd by"Jifrnio ndrwns o&~z br7givc toffo-beatq-nd metc W4;C - ---------- P-q qgevrldfci meno,~ -roomi~es ha acnto om iazj W4, ,tnrued A~sciAD as,~b~.

"M i . ..

C tif. ;

'M .. 'ffl.ft. Si

!O caweMI~ -romst-below 12-0'FRfor. fthcprthfo

!perator , flnijqlrelimv dfltcu" ooYir the coto oMMth,

~i~if)P~e ~

V 'UP..

4 .ce j~O1:

I .

par e,:

Loss' f 'VAC; V'YClSOit2eif. O ovro; n the c 1 i& itis e4ccted hat shorty after for

-ja wifcn 120 'F.

caot trtsoom nomamhs ifingan te ioom - .

6w!room. It ws fouid that it roomIong-erm ItrO!om bes 3cst) twas .

Odnivith a contro oom li

'1room Ilongik' tamt

. . .r..- - ,.'"p_"-*i.

~~rm@ tep isltifiihisthkuatior These hcludrc to &coiznffr: t4z

.4.* '...si ml'c; si' aitl c&m' e yteti decreased I

. 'r'_..--'

itro loo .'.s Jc :' .;,'

5: -: ,.-- -

ri- -af~ - -;.

.. , *1 _-.

..- s* r

@Žr'hs .in'w.;l *ras the heal. '<

w(S*i~

ai tteradc I

.is

. . - . . . 4'

, t.

S¶ .. .~.-~ .. -

*I - ...- . .. .. ..

Cobtnrl Room Heatup due . Loss-of HVAC VYC-1502 Rev. 0 Pap 7 3 .0 putiasupin it The i~iput Wm as tios ised f'r the GOIMC model wil' be addressed step by step based WM .C*5rpD.rit tables;.Tis seon aildescribe and addess t ysical oi oeo r od e 4ld how te model was used aind tiihenlvdns iti'were i.e iD oider to aress pedfic riations oa post.iithe of tint con room hiat up. For .instace, du, due.e.te ; r r.

mdujiict hatup test will have specific initial

'The .OD d eltitttt .e:t~ld ie-tt ienftenrte .duigan

, 6e Ad4ciaaitiiit c~iditions. Tis stion &scribes the f sgt) of.:i- GO6IC input tables for the control X. *.* ;ro'i )etest T .inn S Ad Te input tables for the control room iirl -I3]43][6 1vu as ic pln view;of the control room envelope.

ffi . " JGiCti Cobntr6o1Rom GOTHLModdC

- ed., m ,a-am c

. isvokume. ipO The GOTHC model is

-hovcbcaU35in.-;i.;uie 2.AWbIle the Contrl Rombs a more complicated three a ..- *.a h--a

plan diniensions are: 80.063 fc~t (eastwest wlls) x 48.573 .

.~~~~~~

- Pm~s:,~isx

&,itl j d fi oe t th1j-Aw .o -i woie f ad eilnp thlD volnersomareat Xelevto ioos

. terefor-ethe i'rs"of ir29. -"t3 -

cior ~lit~o an*' sMO w - the

-on':tro-flboi .'.

psrleetn'bc s %:oitr 'aide ;R '~e6N a6bove XS:

ie t~mcwvvst srtef ofte~Is) i: x 4'.Te

~. U 2. t S i te3.

ib fluaniped-rooa~n ise9.1-

.g, .,

01.1 -~~

Mrex ffi *-P ngeQel.- 0t above thetotoSmllRodRss aapd rep eL m e. :.

a - arc ^ . iii fide o cg, oilspae a34 x3theg dr ioti exi X.

S~~~~~~- C .:.i ~ P .s@

:..C ni&

ffttoot d nib**

B . - W~i~dof 0R o

.____________________________$'__a. 3.7 _d _0

1flHenifndu&~,Los of HVAC ,V.YC-1 502 Rv. D.Pac

.2 J~

.. 4

so4& 4 1

4*4. .*

44.-

Hatpo i ~ . . B.C" O'*~w Vii,634'fil-f' 4 04 IC.

- ,4 NOW-.

A: ;2.<f303f

.s W

-t W V-

- -3

of HTVAC VYC-I502.Rev. O Off.Im ,. :- .:

iI S..... .. . .

.,.L *.9 1*. -

r 4* &

-Ii. -.

t. 1 C

C' * *.. ->

.'. ::j..p; ** a.,

A.

t*.

V \ .. I.

2V _

I..

.4§J.... 2

r . I o&*. ,.

7f 1-I

. j. . ..

C .*-****r.* . . *f C'!.

.V.i9 'Ca N'

1.

.4 .

4: .7

1 4'

t Eal ~

.... -C

.' .. . . . 4 . ..

cj.

.8

. -.1

'r ".

.-.... fl*.

14

Room HWetp de It'LEss ofHVAC; VYC-1502 Rev. 0 h.ige / .i id be Ed .hat-the cont cAoon.h -bMarge arrs of cab ts and control pamels. Most d fy tlos hu X fSee volume of the control .

Foi notw cAds tc Unaosd it is Xtedpeea

--- ADr,;r liX the cabinets do not reduce het cotrlrdm~rinlkanral, continezd witbn th

. ..a ......... 3 1 .*

dilaed -fioM 4Area/ Wetted Pirimeter.

80634-' 4853) = ,60.46 f I(2 M *.* .*,'.

.~1.... I~a4JwA iwkfc the *ta"o u of thb.rae-Th vlum beowthe drop ceifng -Calt

.ad sbtctm

- 003'487533

1583' -.41533.75 ft3 20267.4 Vt" i th eevai ig iouih of.t cifin in the control rom

~) adirlliigeol enl-0iof i67

&5i~583t ft.

A.

-. fa 90a h*i7wff Un.

ay~wtcr~p~rf p~l-Os No~141G Sfl~lc~pc~ccOvc

-7 &Tl i'h vlu t stii

R 1 de to Lom of HVAC; VYC-1502 Rev. 0

'control room modelf ill MundeMo only heating, no cooling, amd will try It Since th coition is set to mAintain a constant pressure, 2 Coni rodo Atxi4*tos-dion to maintain the constant

'bn' 't fm it:he boii y 'condition, the temperattm of the of no c~iscien.

ifiedin tlienodt6 T-woof~ftse.fl O paths connect the control Diunles to 1hti: v fluMd bounamay condition and the other.two Ivohe-.wit-..he ,o,.,rolr.omvboiI~iiC s ;;r;..,,..... .. , .... ....

tCd to th bieixdary conditions, the conm.cion bev~Iim s Ii&The conntion -heat is set-

areisset~ toa~suffciewnlylarge area to entuxe th:t the hydraulic i Xto m.e no esuzat. Ie loss .C bt nzotsu ciettcres'.ure the modeld'volume

,drOp ceiling volu' are itodded' iw air cihulation wih The rcp gs t ,allows b..

' - - i" - . ' - ... , . . .... ,

e> Si ~h ozyo ~e su~l~h cotrooom is ihibe -ContmVbL

th-e' cil lli ed'imconratkyiosa heights t lo thi 6¶ vlm24't oiitiat ifi heboto of the*6 eelb EN Z k I V, AAk' aE Thowlares is ns,oiut7~hbae 10 pnils'being reved and the panels am n liaii R~~*.

4 1-17; T'6he ydruiedian' iason -the-indM"vidubaal 2 ft 4 ft opening:

h& r bpWTh~aUe q imfoM s,1t c~a1e pizdo'W'vk~b *bz~h y c wI dea Mbectoa s £ctcie zui m eI

h."-t" . r .modbUb.avI4m Bo eh i lW 13 2a S.. ~f

**1

. gw.vow* - p I

'2 .- :.,- ;.

,cs6

VA,;VY-IsOr

o R. 1) > :'.l-"' 'Ptge

-o "2. '% . . . '

, ' ,.6 e-y -o iv the.bo 1ihe control ro( d

'41 cezngflc.) -. . ... . ..

-. ct

-i

4.

b4 loss co-~efficjcntof LO 42nd2S

-I Th-hbi3i" 4 &da &"to '& to -.a Wa

.41 1'7

4.

'h4*. '.: t. -

. 'at

" ,'" .'. .t.'

.' :: 'a'; . "

£ ... r,, ... ,-" .s

' ;. sr.

' l<; :' - .; . ,' ' ,

,.- ; . . . .' 4t,

-,. '4' * ,t4

~.4..

t : ;.:I ekei, 4'

~~~~~~.. :....J,.. p

,.,,t, ,.;V:

ccai*- , s obz *, **

-1 :;-.. .:

'.-.'i' -t

Page/i_

1m0 *Tz tieo Loss ofEHVAC; VYC-15(a2 Rzv. 0

.I

. of X: dro, clei n .

I.

of cSf the onfroTI room vbhiine,

,,'I

- "I e .

,stacal room heat I a.. .

.I lef..I.,....

,,4

'80.063-ft 9O.~i~80.63 fL2

-ol4"7ft

.4 1~ ti 485.73 f?

9ft 2 .

lii"h, oo .48A b f8t 388 9.

LI 80 63 33ft =467.0W I.omatch the test data ptu ziadeof4ie .

lc!Dpcrnue, .

'C...

eppzpriate ettiig Iaze Iu'tmi

.aee :.*..

9, ... t . .

. : _V . . ,. -, - .DDSI&?ed intvki . -

(.

mid aM V01*XLeb nl1oombside I &7.:IO) n{qIi1meAiside floor'(If5) andihe top 1 tonveclion rthe intemal.wom S

"he a , nk i s v'.

.'..-z 5.,.'Vt

~A

't"

,e x:'A

Control 90oom ductol 0oss of EVAC; VYC-1502 Rev. 0 Pane A Thbe I.

surfes (V~olume A-connemdonsIbr f m.conductors #1 - 5 & #7- #11) use an a ion }eat: trfecoeffic8 t:*epative of their oacutation ha sptifr bo ondM bd on tfe expected tempetes ung spa~ces edie; the ast- bi ldingtiirie bilding or outide air.

V it d~ o theica tri..nsfek . ficien ts sre x lain d furt e in thie Incx

~cci ee i ~ Ta N ~ .thi tg

.. .. : a

i. 2 & 3 arethe uiai convection correlation coefficients

[,faae ipcrf doum re ly. These wre used wih 5 &6 a '1 dAe mikr ecio heat tansfer coefficients.

jn eon ofe bat t ofcictt reiC snOt 0fllon to Ser' " i ff r e i ' ' ~ T ee t he :.Ta c mi f acfes a*RtfmAOT F"t7"',I- J. .

K-

1. .. ? I I ---

trip i cofeainsof intrs are given by; [10].::.

i. J.

for vrtd pla e inthe turblent range, forbo hicta i latcsfa#iz;g pp when heated in the wb .. .>

i rwhen heated in t turluit,, , ?awnge

~s ;.; . I.0 -.

ot T hr 1f , ,. insof 6i9'9of nd Lsis inmimts of-feeL.

h 0.406 lUMfl 2-.

h 0.470 .B.TU U/hr-1.-.F .

I-4`6do i ax 2 h- = 007i BJ/lhr fl2>'F. .. :

. - - 0 .

rcsrch6sen.

1o oom.i 1 ti d b-h ; - *

.5 , . *.'sA

'. . : N .t ^r*...x

I oa d~toLs f HVAC; (C4502 -Rv.o Pa 3ounidaly: conito for.te out& aeil(adrof floor) surfaces consist of-specified Tffdetswib a Coostant ~b~e&.!gj,Yeim u. For those Wals ivit an expoure-vergo th~ne SOLAIR, ieldftaldn

~ fth *ad6io to-i e asro 'm10 mkount

)'e ug.. [

LAIRTeeratu7e ae *7' _N .8.

. C.;

.I

.:L

.'v

.. 'ItI

.1.

.i,

.. - -`e-

.:I

. L 68t o.X l2b' ~Y7

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I .

WU Udg ib1C

- e.e. ,va. -

3-

.5.3. if .*.

. : I . -.

- 7 , ., . _. * ':'

i:

Coio1 Room lic Hetup dub to Loss ofIVAC; VYC- 502 Rev. 0 Page I'

1 he-ransient temperature for case of Mity) of the waUlls dd roof the effect bcontro room space is, in'essene

'emper:n:t.

om is Adj) zklb6thee tdbiiei building. From the FSAP , tXe of.f t&h JS1e b 5lb"FI0, F. lomwever, the hallway o0riSAn os not be sbet to ie'b the a.c ballWay outside the cbztro iieat li dnptw is t the otsid. I eotiioororoom andheas p amannr-in rhisi~iizeissat t~he control room rsligin a buildig.. From the FSAR, the I. -Thus, the teinp-

8Sectidon 10.2.3]. .

-. .I

^Lom The air te petue In

.c m ol r o . Th , he a h-' ..

an~d batery room loe HVAC. Th~e .

mi ~i ak t o tb b e te m e a ir oF-s njn o is th at the c able vaul

L ~ u~i'o~ru w~ i n ur=if

. . T r 'A -i J ,, _ 'r -,; '.. ;" '

pes required f,.r this model:

...he v re six ~ eih e cs: d

-con

1. *

(-:.: 7.kc nrts V. *.. **R . 2. 2 R

~.. c..'- .. ' . - .

i , ,. .-.- ---------

.. 1~'

- !Z Codmol Room epduc toLoss ofHVAC; VYC-1502 1v.0 PageRv

4. 4fft lbk ete slab (South W)

5. 5/8! ,.ic..-l fle

. 1 a

abstratht 1 fr rmal conductor F ;M-- ;tnCA is des-bied in the GOTIIllC User Manual. It

,,e .s, (ag and bc11t:eAtic.h boi -is .. c,..

- mtIiz sie auu'naicnaia ibiixmrvs iiu qumuue ases5:nt ci whe zan

';oeffin :LFozkaU slb a ct of he beat cmiditheqhlitive

.;s c; eiO L o . *I; :eiiividcd t 10 sbegions.

br t e CO l osouth wall, 2 fee thick [?] nd te adjaceit also 2 *eet tbc 5Tis rsfults in a total thicke'of;4 sfee.

-,-]

tI . - -Lo --;

.S-co ncr acoustic tile and abstract the :nit Fo*ell r- ir- conproperties anid for 1e rag A -~~jri~ (656F tos <200)ibiss a god aoss .m.rp-u&

.-4

.. <-'- -' e443 [13]. e **

Th 'C tj?.~I~sT-Wr;OZ tF [13ITabl 4,43

_> . . ic .101Ti ii:ie2.

i~lfbrwtfle

. . . * , *ci-ed

~ - '- L-j\wr- , igj tjY!UI f' OC DO8 -

cw i M * " Owet lftf~ X

'oiistic'le*' '*-S .

-- Z'.

Coutiol Room Hentup dae to LAM of HVAC; VYC-S5Z Rev. 0 Page/

Abs~ct Thcxmal-Conductor is Matcria Type 3 (steel was vsed 2s the basis fot this thermal condjxcor sitc Ja visizl inspecton of the v6ntrbl room Skives the 'unpremson that the majority of iih~intera ccducaddir is steel in-the~f~nihuof tk cabirets.)

(-13, pg 64LSI, eel-AS C1020]

ozih~d~i Tlin~jal 27BTU/hr-ft-17F,

2. Ocuais;1£3,BT/e 6th7l-imdZZ 4 rannt in question is' aswioprpersotnis ts.

is the~ xdiodl This omponent dosuotreesn cfl

~Xcinnt A ~l~~k fm in aii doe

~ their1~t-vaIes vl fri-fantables arc choscnri~bitil 1,~z WflrflyY~ tnsre hat he ir i th tw oline-is rMIxed thoroughl.

.CCU;...

inclired h odl Uc hy are wied to modeffth opening of thecilu

? ao at 4 aeae cnro OIies PMatcdz iiov& qcck fopnvfiw twof

~ Mtu

-r e ..

ocw We mrbtwe the 4

.~,.t.4*¶N o-Kwo'volu*m s#e.talothcxhuo f

.4.. 4 .. 0t*-

Rom Hcp dui to Loa of HVAC; VYC-1502 Rev. 0

.oi Pageff Ia~ qil close type, is includ d to p iit mMne Ihmn one pressme boumday oict n beigactv at any one tie. .hen.e flow paths between the control room he ' .tileV.oh Me c Cpen diFto one of the hxuday

. o . pevet AC o o m inteating.

ti ves as'i losse are ws includedmin the flw o~is~en~tobeink so as not t provideay rm e restrictive mnoeiigta

...l .sge'

- t wpeh the valves -andhdie volumetic fi,. It is-postlated that the operato ie ceilin tis t o l room r reaches II0F.

teoeA vesthe cnol ableito a form which can be used as poni e .. .,

prcv:°y.

u ble- is ued toet the citil xom tpatere so that can be.

used es a ofo.'e components dcinbed eiosy:

~~

~,- *, ~ ;. *

~. A.n e - 1tF initfibi iU-nnAu !a upon the given Scenario; Tq u5-.inenM- 'f-Iyi o ~ rat ae '"" vfct.but is. set tolhe

,-,;' *i airm MIR~i~f Seoti hert with respect to thIE iest In: the wia 1 on is solved l nve"gene...

II A f ortItra~~icit S tu r e 1

COutt? ROOm I}ntup diito Loss of HVAC; VYC-1502 Rev. 0

The a and'i-Idime int rvals de& theieuperiod for deMing the-pro.TflMSh room. Thwr chiosnto .bound te peniod of inters ..(e first

t~inte Ssedo det. za^psaxdoteady-ste tempetr for the

- .-. ~ seaztuclhre,D-R~ido.'t&

aoft*~)vt';ie*i drive the to coingec ai4 ai ion equatons toondxia A,

equ=i.ibium e While 1n ;i ;lb thi qaistoreach At- ;e ftt domain, it rts i a sted

-. vthua teaonabe ---

I:'.'d-'

. .' S.:-.- . . 5,

.sv

1

' *.t.

BYZ5.

d.

I . . .-.. ...

i'--

P; I Corjrojs Room Heatsp due to Loss of HVAC; VYC1502 Rcv. 0Page _LI 3.2 Assusptions tions i~S~ in This calculo are prested n se.wrate sections as the are some

~ions i apply to the GOTHIC model and those hat al to the

.t. .intsandcalcuhlaiobs associad with the test data.

32.i OTHC Model Assumpo Y

A*. * *. te .si tcontrol ioom mod AGOTHC inpit severa pcit asumpions were made. -fryare liste hee for refeence lay tihe contro roomand tht Q:ble vamlttbatty room lose HVAC at

.he tee.ss the control room and experience the same 6eperature transient as

~~ co.o.... *i..

r2- W.Ine m t conirl room reaches 107F the oprators have removed iR. 'hei , -ik 'h5 "0the wo -d volue e now o~~~~~ln o s i> CSv--Di-UM.

I- v -

t. .;.2.25 3=jjo loa Hea Cotaidculation Asswnptlons 'I cul;io .. otrr room haiti oad that follows, there are seve explicit and on-4add ree nce. wle negereor Aordztig .able.o..oi too b. ses bie F ylirciibn ,, ,'..,,

to-if'1, oelao~Sc -the pahii r'd' is cdiDsid -for-X;]'X* 'u -~- -; -d ua no~~der I~ored s mpercffe~en~s f~ee tlicr Iin he ducts and thie exteraird Hoevra iernt icontrort i.-to oaimht :load 9will be hken to atcoumt fi.

eany due t this [14J to I"'in Oo4 1OF ,;the operators take action to

. .oWair in t roo m air aid vice vosa.

ow. :.. '; '. ' :a M .

.. .- s t - ~ 9er#4 - .~e,

1?f,_s SN.

g ~~Control RoG }ij zudii 61La of HYAC:; VYP-1502:-R~v. 0 s_.<g N3';Aitvihbiplbii;sof- }3VtA,; :fionlis talsen to mitigate the lemperaue nse.-..:.

i g ~~~a - P , ,r is . pvid ed th at -b n atI east 66 00 c fin of bi ftid e§a ir...............................

ibts

'n , : . o. nl sles t 3 00 4 cf of o adir ...........................

b~'~ `ML:- ' id -uctleast 3 0 cis t~.(0,°;)hcd 6eue y5 e

g '~~~~~~~- '- g

--1Vt is ued-y4lO : 3

.nX

. _ S I

.1.

Controi Rom Heatup due to Loss of HVAC; VYC-1502 Rev. 0 Page Z3 1I ..

4.0 Calcutliton The clculon is %ivided into sevral sections. The objective of the calculation is to call mthe txaienA tem folloig a loss of VenalUaion due to an A' 2x Tevphykc caM. nfiCitoian~model of the control zoom bas been

I - in 1,3i Two Oz 0pic6es of ion are equired to complte the

.. . rom. lo.

fi]: d cos flroom intheat sik daactizston. Tests

e. on t.aVAC .te s ty i tlifdatat as ten tiai be used to chanca

-.: .1e- inet&al hj iCa . . c S. 4cd liei-d. des=4tion oe% datasedto cadcute the control room heat kld the l-calcition ofi ,S on uses the bea :od Information

4.2

.aswe as 1oI i&tlla o d the control mom "1Ht^.iiS in 4. use oXASif i to assss the ooD1 room.-

iidl a a loss e ,Sof C Ai'o't R Sioi.4p

.. ham"t potwo-otj acotios to inidpgae the.

W- :*fiH AtZ ::.

r ' -. " * .. ' 6 * ',_! '

^ td t csan be used to chaactiz WrU. at E~~N#took~Ei cojztrat'-themi lo'&,Bot eat t peatie;nasuet>S at gthe iiiilet andodtIt oh.

4he tomvnlA..n ikadn~if~ tzmn in ~the IACdcwz g z; ~dWka.3.' '-

.m. R ~ inixsrmet ionachi tio prgme.....nrwere the

efn~dung '-.ar~e-- QAe estz _rccfr 1ope, the ........................ pete by -"

the iwereprforjued

.-kt basisIr judging thre L

, .-.. . ,i, heatp . ,

hmnc 3

ystem ontrol 3 room.,. ie

.3 3..

3.

3. .

3. ,., 3

s of HVAC; VYC-1502Rcv. 0 Page AM uned for.air I oW through the control room:

p Fd P0F6 d

f~h~fo1~pwng 4abre S~the reslt of su--imple mss'anbd energybalance relatirns. They .ai of2tziqu Iitf TIVe-di-k iiii2BTUfibry

!PY alr-

~ .

I

Conto1l'Room lictupdvue ?~Los of HVAC; VYC-1502 Rev. 0 For b2i

=0.60 pa -iiO217psia (at t= 732)

Y F,, *P 1 ,= G.241302 psia p = i.i02 10,8 i i.-

.1.

1!11 IC It-I?

h2 28.8 U. y . A I.. J

,AT..,.. -"'4

-E

.1 assme ~'.tth ii.roadlnte cnrl room is ng~ligible.: I

,,,..,s ... .,,,,e ,,,,,m is .o .ta,.t;o the humidit matiol lie 1.

'ik..

11

- i. . K

. . rs a; r * -

h 6 o.44'4.ss)

;;S,

.1

,.;:.g

..: e :.-: :;H

-I

.c.'4 4-V the afo o rcorol room is gven by:

. I ..

.,E ,

24.4 ) - 91 .6 B2tJ z in 'o 1 5, 0 .B~ qhr or 4 &6 1 Bi /se rt, *, .:.' , ; . -. *. .. -~' **.

af .1. 10 %/based on ith i c l n y

- a : -

w-

- 1 -.. . v. -.

.-- . . t Coaitmo Room 1Ica*~ due to Lms of HVAC; VYC-1,502 Rev. 0 PgZ 4A3. Heit Loakd -2 To hdjpc~nfiri tfebea load that was calculated in-the pmcirious 6 cd1cudoofiedbfran data kiohided sectiozi the control room in ailem uraidm hat reported a control room HA bjzidh[lJ.AtacBm SOf

- tberfer n includedat t&ke -cm the Th~i~ni~i HVC sste ismor c injlestha the 'data takez for and used in the ilor~in~ibo c "'nk of~h air aro~zidthcsystem i indtdtcineteha Of-M Cot -.

oihebj e-i I bdemeterunc Mr and to theamthII of aznss in: .. 'Uw 1.22

'C

'4"-.4 5 A'

'on Frst W.aet&

7. -Sc 4-.

Corojloo Jic HA4u duc" LASS of HVAC VYC-ISOZ Rev. 0 PAgc Z7

> L 01.4.

r. --

-An qcr I

. .I

.. . .II i

...1j. u s 3*t

Ai r

.I

...: ...i I exLerrr- 1

.i .exh~uiuL,;,

V. - - :.,i -u-1 .:!'L i on .31

>Te,~etrieth niipics and (Wddidtyno of Thbe entceinit and leaving air' MM +4OtW(1O61.

A4t+

J';I,

4. .. w, amA~REab1 y a

1

cost a Comtrl rom Hcwp un tos of liVAC; '1yc-1s( Rev. 0 .5 specific~1y at the Wnet' I4 W0~.0081374

-056t*Vj-0.240-t*)

- 10934.0.149-1t (1093-5561G936) M44O4134A -O~5.2465.4 516

. 0.40:.(106:+O444t)

~ 157T

~flb.1.y 4j3 Ao 2 1093 0 II S *S *t

'S.

ControlRovm Hmabup due to Loss of HVAC; VYC-1502 Rev. 0 Page df and thvtotal beat addfition to the air stream from the control room is:

,.,qg.,A m4 (h 6 -h@

1i 668.85b dry alrice h6=2.582mT7J/lb dry air h,, 21-578BUIlb dry air r668.85%26582-21.578) = 3346.9 BTMrmin o- 200,815 BHTfhr or 55.78 M71sec t , ;

imnt n tthis ialue of control room heat load are not entirely Jcnowm since

.y values are not direcdy atibable to the instuments used.

4.13 .Determiutdiou olf Limiting Contrl Room hest Load

T" vi for control oom heat load have bben calculated usig independent ethods, n'and odne. he vialues-obtained ame

175,000 BTU/hr and i.J/. Thelaer Xvlue was'determined ivth data taken on a warm day in mid-1y.,. obtaine

a.cool dy eMay. Vaigtions in solar ht jv. ca toccupancy ea instrun.acucy can account for tbis diiterene, Eacb

--

insIaA iuc loa'coniductio d into or out of the control room through the U room wi11A 'lyiead d. -ius, tee is an inherent conservati inese

rva~siiihc

- thel i~C e hdic1i"wili be as electiital,16ads (conduction and OR.e0i being n tl]6 GOTIC model Input.)

Th;. f~orl.- to ensure a J din.. g value for control room beat load is used, the avetage of the

fwovas +0 .Yillb.I he control room heat load used for deterining the control room t ien:to:eip .tus a loss of HVAC duc to a Appndix Rfire event. Thiss

-oU difib -the information Teceived regarding the potential electrical icd ss.t.. .. s s . r -* s cif *17 -0O1 + 2O08) 1;2

1.1 = 206,700 BTUJhr or 57A2 BTUlsec N. .

L' -2.v___'_______________ ___ _ 1:

Controi Room Heamtp duc to Loss of IVAC; VYC-IS02 Rev. 0 Pagc 3 42 Characteriztion of Abstract Hleat Conductor As mntioned previously, the transient temperature of the control room due to a loss of HVAC depends on two important factors. the heat load in the control roor and the intermal

'weat absoirbutices. Ths section attempts to caractelize the intemal heat absorbing Structdres as a GOHIC heat conductor using test data obtained during a control room heup

> ;; test:

In a bted room, thetns ent temperature rise is predicatedion two major factors: the heat load , the heat .'Asingstnrctures. For the control room, the previous sections have

identd thheat load. to be able topredict p.rTus, tbe.tsi tempature profile, fhe heat abOrbi*ug structur mus be able to be modeled; However, the control room has a very co'n'i' ay of rple orbig structures (cabinets, pauil, ec.) niffing it dicult to model w ci u .Thus, a test was performed to gather data (temperature rise versus time and

hit lo)dMiat can beiked to characterie the hft absoibing structrs.

In te revio s sedions, 'Ae control-room heat load during the test is cGaeri Given t {his heat 1id and hown conh it i.ambient is possle to modify the -GOhIC model

....i de. previobsW to.atternpt to model the tesi conditions. An abstract internal keat b i 'I t-in the 1I model f the test By varying the surfae area of god.- i possIse to mith the tnwient test data and thus have a faily conducddr model for use in later control room heatup analyses.

. . Thle ia siesof tempertu at te locatidons within the control room tken W . .~.d;-\t4;;=r'-

i of bo~t3ints7 Becaus the mtrol oom is a three-dimensidnal stute hi O~ighiZtly

-dl s: its etc.) but iœ

-i,^afdn6the cdntrl TOom iS one d2iWnensionian (a lumped parmeter volume) itiajiot .p~te~ i .;O :C.it-ds w gllt attch .the et rsulcs exactly. However, to adftaev 'aih nit cond s matchg ihe slope of the temperature

Rt*. .se: ri w.illnb~ i..e.nt. tIas founda th :ilpe "ofthie cipiraww rise was fily constn 6verihje 'ent ' ,. '*r l idrs'of the Oei radings. fact, it was much more constant i--* than the X . ut. ,bl r oi*s.

T.he testt 'if inclued inAppendix D. The test data is ploted in FIgure 4, Figure 5 end Fr' 6. long ith 6,lis ng the least xqares linear data fit for each of the data sets.

eTM:he foliog eitons used to obtain the~l= t squrs fit Note that since the slope is the-D oi~lt iiabe of c e only the slope is calc.,ated.

I-,..,

.,p ., . . . :

0... .

"I-,.

___________________________ .* * *. . j... * '- . ;*'.-

r .

. I

? . . . . . . .

'.' . I I .I.. I -

I ---

r... ,. ".

I

i. .,%t-. r 77 7&-

'a.-.76 at pet .F Po' aI~ 2,

.)73 -

I C9 72p- 0 A

71

. ,.I I I I I a.

O 10.> 20 30 40 I... , - .. ..

'.. ..! ,.:I . -..' ..-

1 . ... I .. . .% .

Time (minutes)

'1 :

. . .. , .. * 'I . . S . . *.: .. 1 .. .. .

1 .I. ;...a. " ....

--., .."!-- ?,. i', !: .-.

... -;I . I.'.

.1 1.jtiSmy: .- -':' . ' .. . - .

.... -I'. , .. . - ,: I. ... : .:.. . .. . .. .

V

"4 .4...

.4 . . 4 . . . -

43/4. w- o

, f..~: ... ... .. .

. - .-4. . .

..r1.-,r,. 'K

.- ,e, " , -- e ,'

0-

4 ' . ' '. -;', , . ' . .

AA A.

,* ^ ~~~., , . *: .4: . . :'s'*' '

$1 la

1 1>....

g . 75 ' 8-8,,,

.a74' @.> '*.&' *.,,i:;,-!..._

!;, I'

. ni ..... . ; .. %.. 4j ' - ^ ~ . '

4':

.4 U'.

'4-.0 73- -

' . P, I . .

72 r7..:

r- --- .

I . ., I - . I,

'I ma.

.7 _

10 :. .I . .(20u 30 40 I' 1. " ... .....

4 ... ... ..

1/2

. -,, .. I.. . ,.,. .,

S.

. .. I I

'VI .4 . I -

.; . .... I I . .. . ..... . ..

, .; -. . .. )s

4I . *. s . . -4 s' * ... -, :" .: . c

I.

I 1t. 3'* X '. . :

In . .. . ;',

t*1- ..- **

I

- 1 Cl* s.-' "C; - -V'.

b V ,; t*

4 ,

/ . Zk -

v. I . , -,

., I I-4.'.-...

.9% .

.. t ... j .t.

.. t.

'9 z-C p.,

i I

p 2

I C

'p I

I 4I i.. .0 5 . .10 .: ;15 ' 20 25 30 A' Bel r#t EEN~;*tFls;S~ws; a->';i; *;i,.A a,,.a.0

,I, a,

.o i

-._.toX* l b ,-i-t ' :

'___ .~ - .'.' .. ' i'. '.. '.' Cf. i'. *. '. ' .'

_________,_______ Jo ;. t Ho 6 A;

I: - _t t t t Control jRoom Heatup due to Loss of HVAC; VYC-502 Recv. 0 Page 3LY y = mx+b N ( N 1N

'a-.

"A-NExjyj- SX £Y were in-m 11 Nf / \2

( m~ )(1 2(1 ('

aVti tS'J a)

N /H x 2 F, tx--XIx,1

'4.'

r ,Puting the data fobfite first data set in the equation for m gives:

r b --- !o I -

N-

.1 072. 0 2 5 73.6 25368

.:3 ; .* O.;.. 5.2 . 100 752

.5' -75.8 22 1137 5 20 76.2 400 1524

, ____.._._:25 76.4 625 1910 7 30 76.9 900 2307

.77.2 1225 2702 7.

SUm 140. -603A 3500 10700

-.: ' t- .- ;.. :.

.. . . . , I. .

£:

t . ^

Ml =

8 10700 'Fm z-140 'F6.4mM

S 2 _-(l4Omn)n2 835s00rn

= 0.133877minu

. *C .: ;,

,i., - 4'"* '-

Pr-' t.

V..

I ..'

f.

I CootwiiRoom Hcalp due to Loss of HVAC; VYC 1502 Rev. O Page jS The blance of the data can be similaly manipulated and the results arc:

DatasSt Slope

'1 ; 0.1338 F/Min 2 0i317 .:Fimn 3 0.08262°F/min 4 *O .1037 .. Eimlini 5 1 Q0.1114,./Min

.6 0:1 264u.. .Fimmn 7 0.3414 -Fimn 9 01350 MFznin Eaeprsl esentsri Seof temperature increase in a local region in the control room. As sch,. it' d*tetinedIn .iart by the heatload and the ixtdrn heat conductors in that local i tegion. zSince otli tS :. gi~veni'the average heat load in fte control room during The (S

-*-.-.test oh 43.1;),eaerage of these slopes will gnve a representation of the combined ovaage thi ht. coiducios in the control room.

g ~~~Avcrw~lol A'o. . cO

-01 iie°/aS . -*mi*.n

.:..This,;ih nis theIgtsloe of the t ient temperature curve for the GOTHIC nodel he

- ddifi~ca~iois~ Aid additiotis to the initial input model are:

6 . .- #12 absix heatsin.-: Area 2 55O00 if

.;iidii - all init temperatures at 73.16 °F6

p. ;

.i

.t-o~fflcients mud

- all teiperature boundary conditions set to 70°F to tail :didois in therooms adjacent to the control room and air.i .te..crd c(prlfy cloudly..-day basd on -the test data.) 7

* #1:(electrical1 48.61i13TU/= based on results 7' 2 A. * #2 (occupants) siet to 325 BU/sec based on 30 seated toicupant at 3.t DbG TU/ [14][l0j.

h .ol .ciFt .- O trip set t 106 seconds, off trip set to 0 seconds (no opeig of

-S & -On trip set to1O' seconds, offtrip set to0 seconds (no opening of Volifte IiiaCditions - Volume 1 (Coitrol Room) set to 73.16 °F Volume 2 V,

Oro CciiS ) set to 72.1 OF to match Time 0 test-data.

Awm- Tbwi O m= reY*A 4WitL The aulm waum; oIb' tnt We Ma an te easaas1Io

. itseIdta d 9 o te f hWs a krgenbe a w, a vtmld of only on bo. Il poe tusu IN x1 whkh woud ,A t xaUt dy mmftp at wau of PcOf OX COOM h Pasweed 31ihes ftft" condl:eiu. *bs -c * '.~ ';'t . ' .... g :'.

J : ,>" ... '. :-

~.1

-. . .. . _. . . . .. .1 -- __ ..-. . .. . - .. -

Control Hetutue to Loss of HVAC; VYC-1502 Rev. 0 W= Page 16b hRun CMurol areers - the tes lasted for 35 minutes; the GOTIUC problem wil

-

r.fior G6 tiint to be 'tle toWcover ith tst iod nd see any short tem trds.

le GTiC iu ta for this rdl areinudedin pcdix A. The GOTHIC outu je inclS on ncroflchiib .de the filename 'CONOLSO .

brgts .rronri~i rAiif'i*-k7td Fl> :The temperatme plot shows an peItuifoby h slower it steady inease. The initil ir o c~f~~ bte titibut:ca cbta he extplain~d by~tbe

-I. il& -Iniall,4be differential tt e dkv cclTthsh1acooductor ini1 I leads aery .small

&h ~s (th corrlaton beat va ase i ft:

6tilWvcir, he~atr.iT &aeaea fomth fbt heas; ru466

.d;.}..r. 9.-..o -

onT teizulr ufl sicess~ -

'pdftc fiitir toi t ant ri;n. ,:;

apeatreaii's calcultdrm- GOThICoutut inforrfi~oiL.-.

is F; At )W sec nds -787.6Th.F. : .

F)/(36O0f sec:>'1105X&s O60 c in 0. 1231 F/in.

_a..,12iWS 14la%oif th&dstsoe W

-co ,0 0:fl25 OD:g 000 m 40 i if .: t.,:.e. .or.:.

w.1

!. -§ J, *.i .. . .

. ; ,, ,,* .. '> . ,: o. .- .

.ZI

.. i.. ,

.. *. ... . . * ;, . ... l * *,,-

l . . l

' rS>. t-X

-d

_1 I

?

I Control loom Heatup due to Los of EVAC, VYC-I502 Rev. 0 . Page m7 1

SI i

II I

1 ':

'9*'

I 'WWAF4run xx mn x

x

.3

.9.. I..

.9..: .1...

I

.9.-. . .--i -.. .. S *9 VV?*m3

.. I

I 9.-. . .9 I I

...- **I V ,~ DC ea

.-..... I

.4* ..

49. .

9'.* 9.. -

tj---

I' 2* 4.z ,..

I -

Ii.. s' 9- 4 .. .... 0: .

-L 2*

-- U-.

m

... 9.... FIgure 8

' It, - . .

Conetrol RoM HestU due to Lfs of HWAC VYC-W502 Rev. 0 Page 3 4.3 Control Room 3ieatup During App. R Event Given 'the irdfomdion available from -the test data, it is now posile to adequately model the controjroom g a los of HVA.C followimg an Appenix R event The geneic control

' room yIC miput model descriled previously is used wiith modifications to the model

' made&

- 1Dfi.o o te bai f.the test results decri-bed-in pievioius sectons The mvdificatiols and atdditons to the in'stal input model are:

b-sCndi r - #12 abstct he-sinlc: Arca 22500 f*

6 s - alI intlim'c;am It 78 OF

. . *6effciefi&ts- for 3viill o to outside conditions, the temperatm

..ou-odtosset Ts> to, seHUa~c t apaufs

. -i Hr #l( iical lod) set to 57.'4.BlTU/sec based on zesh gFr.b . . ohhc;-i.p - t sedi tojllO 0 F control room teiperaue in accordance wih B* >vC.. - . , _p ouslydisc*i'

- .. OO omtmpeatur p in accordanc wvith oU=*]ii:i i es ai. 2 ake 7etto8OF;

-;z tjit~ p$-.~ -*` OTIC pobiem will hin for a four hour tansient (14466 Th .O I. t- b od ..,t in Appn=:ix B. The GOIC o.put rf -che" s r.

-. :. mn. -- "Airnu e the

,oom te : : eei20s i .i of the txa~iet Itis also nitl ~ sP wuld -bCe if t nieyd 4 s wih would.ecdssftati:sm

v..;.*t. s :cio t teapetur r~aJzsA1l °F. Theid .eilini spac min cooler W .- ;. ... It 'Xt -; dii ret

' 17of theitnicoom 1-t

-lttsmeasu since thebeet s~esi arelna e ap1hcrMngdkqi`ml control room..

., o.d j~d-;61

. esprlC -A<

W*,4&c .- blw Tha n~it .iermis o diArn' ifti~abl ac~~

n h t *)= a at loo levl odth teovpcfll"vlconataj romsne bAi room.

r-.r; . ....... . *. S .:*, - ,

4 .5 p.t4'sve that1he Y~QCP=A 6iii ~ smdequimen

&:S,* !*

bd..t

. ; en~tr ,-

tr

4. . UA2. .

ft>S,4 .'--,'**-, * *. t #*

s otSiieAij datn.

-. .U

~ ...

. - . ., - . . . .- .. I- - . .. - I ...

-a -

Ft.

4..

..-. ; .-r <

t:

III

-F If-.

I.-

Coanr Rom Hemp Sm Len of HVAC VYC-1 502 Rcv.O Page f1 E

1.

I .

Loss outHVAC- -- ..

- . r. *

--E

'3I

t2 *>. 10.8 14A.

T w.(t). --- XlOea

~61 .3 r 1.

Copn61Iiiom&atup,Loss

)f HVA'C Tcm eai Transent

-1 3 . p.

- .* t f 9 ' ' - .:

. ... :A -

X.-e ....

, 4v*g~i c~t~ r, :; .. ~ &~ s S. >. .- . . . ..

I:.

A z 1 5 t.

.i. *1* *

- ...:..'4t:.

tp..

'jj-c4..:

.......... . . _ .......... % s .

.:)f I I ... . .. . . I . ,

Coutr~l Room Heauip due to Los of HVAC: VYC-1502 Rev'. 0Pge' Psge va 4.4 , Control Room Heatup itgi Ion Options As by the Section 43, during the first four hours of the control room loss of HVA tansient th mperature h the control room stays under 120 "F. However, due to the beat load in the room- ad the insilating effect of the thick concrete walls, the steady state temp6ature is ej4m56W to be wel above 120 OF. To mitiate this temperatre rise, two

, . possible optfs are plor. Any option chosen will have to do one of two things, eithe

'singI~ or in ~combinationEither the heat 1oMd will have to be reduced or some alternate v:antitiozlcdoon wll have to be provded.

The frst option .6crd was to provide a means of alternate ventilation. Given some teni 3saryvenoilauion g tside air the control ro m tepm can be maintained at or beloi 120 0 F.?A GOTHIC model was used.to detee the-amount of air flow reqired.

he s od oEption was to reduce the heat.l6ad and p'id some means of alternate ti ion (at someduced level fi=-e first on;)

. ., ;1e I cinbrol room model dveloped previously was used with some minor

iascG

'mo.*I2ctiions. A~h~iic of the iC Go is .hown infigure 10. As seen on

euticW: nffo^I fl bobnday w u cofiMtion f a low path to that boundarj

.ere; . efL"I iai siii th rnndi ons are outli"ed below for each option.

e .-. .. ;,

. .ti:i The in~putf .tbles for this bptonrare included as A pendix The output for this

miincltded on microflche underthefilename "OTONi.SOT.

_I - M - It

'.ioi 2 Th Inut les for i option awr itiuded As Appendix F. The output for this rM -is 6flcfiichude uider thi filenan! "OPTION2SOT"

.Fl - a flow bowxdazy ondition was added to simulate a nt6i.---.' ffon air flow; in thi cas the low.Was set to a function vhich bad

. -... oflo ow)4400 di (f us) nd 110 6fs (6600 cfin for Option 1) or

.,50 f (300 fiifor Option.2) UhereaI&r (the flw.:is actly negative, intb the bo+b;jj- .; allowMng the cotol roop e boundary condition to provide

.he ihlet ii) .

IjlIlow h .

path. was modeled connecg e new fluid boundary condition Mith t w m t ntowpath b inendi to model any particular

-. aidh: .oitnitction e parameters as the odher

.cnlr cnionnel low Pam$

~ - .5&UD.I a forcing fh tion was assigned to the electrical heat iced; for Opinfim.tion wfs unity (n cang inheat load) but for Option 2, ionWasseo0.48 0thsfratifour ho trr u a tion inheat load The teprge t rcsls of these models wr shwwin Figure II and Figure 12.

7S ' 'r n~w*@@ .,z-.

II .4 - - . -

-:_ T 'k.. .. -

r -,,, -, - .

i

.. - . . 1. - .1 . -. Ies;=.. - ., -

Cona . . .

par ql IIi

. .. 11 I. I'll -II yzilnlw i

I I ., i :.

t I

i i

i I .- .-

DpCegSppc 1.*

j3

'212J II Ii.L -

10

.. I.

I- -

.2A AA

. .. e .. .,:. .:.

I conrotoom I,

I *

-1 I

K

1 - - . a -

.2 S

.. . .:1 I

. .. . . t

.1T Z .. ::

..: .I ...

r. ., I -, . " ,

A.. I- . .. ', , . . ... -- , 4. . .,:--

W .8.*C:l

- :. -4:

. .. : ... I

.fl..-

4. t.. . . I-... :-

-- -- - . -.- . I Control Hcawp dn~t15 4vs of JIVAC; VYC-1 502 Rey. 0P~~

room Io Hempde a1-I I .

Ccn F xm Hmgt Lons dHVAC, Opdmn II iI1'12 II I

]- J I 4

I 0 3.0 7. 10.8 . 14A te 1.

O0O 'fl { 0 5 . C C OXc 3

-1 1:4.. :s.

. T ' 1 Ve i1Ii . t o A G2-2C~i dRO~ m S~ h4 .1.OS ~ WC .pO po n2

- - .~d - r p e t m

I -- -- -

hmw b -i *

. -, - o- .

-- - c~;.!.4c 4 t;t~.

1.Lr~'~..-' .. *....

.}

t. .

I PAP 6 Loib ofHVAC; VYC-l502 Rev. 0 I .:

Conto koomHcR 3 d60 w .

I.W~

of HYAC Mngca, j i >X ewfvitlrdEhurs WIU> cartiol room loss aTptot o opmto I1 -gi5Zcn r w d f. ig Wowbd rcqnized-to minigacte 6 fz a;3i2 i11 ors-,is Ctofrl wor 2which ,src vj. i + : p git Ict:6M0 if:o~f oihsde air, or

, ., I.

by at leist 2% Ucmpie c52 lby at least 3000 cfm of outside asr.

b. l-oad dt

-.S

. ,' .a....... ;.f.j 4 .tt:

'4

. 4 4

-A V A'?

.19

.4 1 .: .

3/4,7. I.

W~~ - J ....;.r'....

:- :- ' 4j .. ,...:,

.4....

. -. ...*s**

.. . ..4

.5.

K'>

BVY 05-072 Docket No. 50-271 Exhibit SPSB-C-52-4 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Calculation VYC-2405, Rev.0 Total number of pages in this Exhibit (excludina this cover sheet) is 85.

CALCULATION COVER PAGE TIP-2 . [JIP-3 OJAF . OPNPS 0 VY Calculation No. lThis revision incorporates the following MERLIN Sh 1o8 VYC-2405 Rev. O DRNs or Minor Calc Changes: Sheet _ of 85

Title:

Drywell Temperature Calculation for a Station . 3QR CNQR Blackout Event at Extended Power Uprate.

Discipline: Fluid Systems Design Engineering Design Basis Calculation? 0Yes [ONo This calculation supersedes/voids calculation: N/A Modification Nofrask No/ER No: EPU El. No software used l Software used and filed separately (Include Computer Run Summary Sheet).

If 'YES', Code: GOTHIC VT.0p2 0 Software used and filed with this calculation. If "YES', Code:

System No/Name:

Component NolNanie:

(Attach additional pages if necessary)

' Print/Sign REV# STATUS PREPARER REVIEWER/ OTHIER APPROVER' DATE (Prel, Pend, DESIGN REVIEWERI A, V, S) VERIFIER DESIGN

._ _ .___ . VERIFIER 0 Prelim Liliane Alan L. James G.

Schor Robertshaw Roge ENN-DC-126 REV. 4 ATTACHMENT 9.2 CALCULATION COVER PAGE ENN-DC-126 Rav. 4 ATTACHMENT 9.2 CALCULATION COVER PAGE

_ _ _ _ _ _ _ _ _ _ . . . . . . *: -.-. . . S- ... , _,,

. I ., . i _ - - . .

sEntergy MI. f.

Calculation VYC-2405 Rev. 0 Page 2 Of 85 ENN-DC-126 REV. 4 ATTACHMENT 9.6 CALCULATION RECORD OF REVISIONS RECORD OF REVISIONS Calculation Number. VYC-2405 Rev 0 Page 2 of 85 Drywell Temperature Calculation for a Station Blackout Event at Extended Power Uprate.

Revision No. Description of Change Reason For Change R*

0 Original Issue N/A

Tntergy Calculation VYC-2405 Rev. 0 Pabe 3 of 85.

ENN-DC-126 REV. 4 ATTACHMENT 9A CALCULATION

SUMMARY

PAGE

. CALCULATION

SUMMARY

PAGE Calculation No. VYC-2405 Revision No.0 Drywell Temperature Calculation for a Station Blackout Event at Extended Power Uprate.

CALCULATION OBJECTIVE:

This calculation will address the VY drywell temperature for Station Blackout (SBO) at Extended Power Uprate (EPU) Conditions. The calculation will look into means to mitigate the drywell temperature for this event, such that there will be no need for Emergency Depressurization.

CONCLUSIONS:

See Section 7.0 ASSUMPTIONS:

See Assumption in Section 4.0. (see also list of open items - assumptions which need verification or implementation - Section 4.1)

DESIGN INPUT DOCUMENTS:

See Design Input Documents identified in References Section 8.0 AFFECTED DOCUMENTS:

See Assumption Section, Section 4.2 METHODOLOGY:

See Section 3.0

Entergy Calculation VYC-2405 Rev. 0 Page 4 of 85 TABLE OF CONTENTS Cover Sheet (ENN-DC-126 Attachment 9.2) ............................................. 1 Record of Revisions (ENN-DC-126 Attachment 9.6) ............................................ i2 Calculation Summary Page (ENN-DC-126 Attachment 9.4)............................................. 3 Table of Contents ............. ... 4 List of Effective Pages (ENN-DC-126 Attachment 9.15) .......................... .55.............

1.0 Background .6 2.0 Purpose o.nli..;.;.;.....6 3.0 Method of A nalysis................... 7 4.0 Inputs and Assumptionsi .................. ,

5.0 Input and A aysi Crteria.....................................................1....................................................5.......2 en 6.0 Calculation i Analyses ......... ;;:;26 7.0 Results and Conclusions ................. : 67 8.0 References.............;.;;.;.;.;.; 68 Computer Run Summary Sheet (ENN-DC-126 Attachment 9.10).71 Calculation Impact Review Page (ENN-DC-126 Attachment 9.7) .72 Calculation Design Verification and Review (ENN-DC-134) .77 Files on CD .......... 85 Attachments Attachment Al - A39 TOTAL NUMBER OF PAGE (Including Attachments) .124

. . . .. . .. - . . - - - .. 7 : . . .... .. . ". . . - .- .. . I . . I ... . : . .

. I. .

-TEntergy. Calculation VYC-2405 Rev. 0 Page 5 of 85 . Vw.

ATTACHMENT 9.15 LIST OF EFFECTiVE PAGES LIST OF EFFECTIVE PAGES Calculation Number VYC-2405 . Revision Number: 0 Page of 85 Torus Temperature Calculation for.a Station Blackout Event at Extended Power Uprate.

PAGE REV. PAGE REV. PAGE REV.

All 0

_. . ... ... .% A

'Entergy Calculation VYC-2405 Rev. 0 Page 6 of 85 N

1.0 Background

The Station Blackout (SBO) torus temperature calculation (Reference 1) accommodated a higher coping time of 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months versus 10 minutes previously assumed. In addition to an increased coping time, Reference I also eliminated the potential need for Containment Overpressure (COP) for the SBO event. In the process, it was determined that in order to implement coping strategies for the two hours, two additional parameters need to be analyzed:

- Drywell temperature and the coping strategy to accommodate an expected higher drywell

.temperature, and

- Procedural direction for the operators (if needed) to limit the drywell temperature while ensuring capability of HPCIIRCIC to maintain vessel level 2.0 Purpose This calculation will address the VY drywell temperature for. Station Blackout (SBO) at Extended Power Uprate (EPU) Conditions. The calculation will look into means to reduce the drywell temperature for this event, such that there will be no need for Reactor Pressure Vessel.

Emergency Depressurization (RPVED - Reference 30).

This analysis will address control of the drywell temperature by controlled depressurization (cooldown) and will show that RCIC/HPCI injection is maintained until power is restored and the low pressure pumps (RHR and CS) are available.

As indicated in Section 1.0 of Reference 1, for the SBO event, the Alternate AC (AAC) power source is restored at 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months into the event. After the restoration of power, torus cooling and drywell spray will become available.

2.1 Acceptance Criteria To evaluate the results the following criteria are applied:

1. The maximum allowable drywell bulk average temperature should remain below the EQ temperature (3400 F for the first 30 minutes and 325T'F for the next 270 minutes)

(Reference 19).

2. The maximum allowable drywell surface temperature is 281 TF (Reference 20).

3. The maximum allowable drywell air pressure is 56 psig, (Reference 27).

4 Maintain the torus pressure below PSP curve (Reference 30) during the 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months coping duration and the 10 minutes of low pressure pumps restoration period.

5. The analyses should provide assurance that there is no need to spray the drywell in the unsafe region of the DWSIL curve (Reference 30).

Entergy Calculation VYC-2405 Rev. 0 Page 7 of 85 3.0 Method of Analysis The model developed in Reference 1 is modified to accommodate changes related to the purpose described in Section 2.

The GOTHIC code (Reference 7), -Version 7.0p2 has been selected for use in this analysis. This code was used in the original suppression pool temperature calculation (Reference 6) and in the analysis for SBO at EPU conditions, Reference 1. This specific version of the code has been installed and complies with the ENVY SQA procedures ENN-IT-104 (it replaced VY procedure AP-6030) as documented in calculation VYC-2208 (Reference 8).

The following changes to the input SBO-NoLeak-80 to produce SBO-drywell2 are being added:

- drywell heat load

- drywell heat slabs

- leakage from drywell to wetwell

- modifications to the vacuum breaker modeling The GOTHIC input file for the case SBO-drywell2 is presented in attachment A.

4.0 Inputs and Assumptions The inputs and assumptions for the SBO event were developed in Reference 1. For completeness, they are added. to this calculation. The more important modifications to the model have been made, for this analysis, by the addition of the Drywell Heat Loads and Drywell Heat Slabs (see Section 5.0 for details).

The SBO scenario postulates a complete loss of onsite and offsite AC power. The vessel is assumed to be isolated at the start of the event.

The scenario is modeled as follows:

l) Scram occurs at time zero.

2) The MSIVs are isolated at time zero (this is a conservative assumption for the drywell temperature calculation since the energy transferred to the condenser while the MSIVs are opened will remain in the vessel).

3) The Reactor Vessel level is maintained with HPCI or RCIC in a band between 127-177 inches above Top of Active Fuel (TAF). Level is maintained with HPCI at a nominal flow of 4250 gpm. In reality HPCI flow will be adjusted to keep level in the band and to prevent excessive start/stop cycles. The HPCI (or RCIC) modeling in the GOTHIC input as a continuous flow (lower flow) or as intermittent flow has no effect on the drywell temperature analysis results.

The choice of RCIC or HPCI or the flow capacity has no effect on the analysis since HPCI injects intermittently to maintain inventory or can be throttled as required to maintain level.

If RCIC were used, it would injed more often.

+/-nter Calculation VYC-2405 Rev. 0 Page 8 of 85 f-f'*

4) HPCI takes suction from CST at 135TF. The CST inventory available for injection is 75000 gal.

The GOTHIC input value in Ibm = 75000 gals 1 ft317.48 gals /0.01627 ibm/ft3 (Reference 31 for the density a 135 0F) = 616271.7 Ibm

5) Power is restored at 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months . Torus cooling is initiated at 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes. Two RHR Service Water pumps are available at 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes, -delivering 4700 gpm. The second RHRSW pump is discontinued at 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months in the transient to maintain the Corner Room temperatures below the EQ limit (Reference I, Attachment B). The dryw'ell temperature analysis is performed for only 25000 seconds for the base case and for 14400 (4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months ) seconds for the sensitivity cases since, after 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes (7800 seconds),

the low pressure pumps are available to spray the drywell, if needed, hence there is no need to analyze the drywell temperature for a longer-duration.

6) An orderly reactor cooldown is initiated at one hour in. order to maintain the drywell.

temperature below the EQ limit (Reference 19) and the drywell shell metal below 281 TF (Reference 20). Two cooldown rates will be analyzed: 800F/hr and 456/hr.

7) The RPV level is controlled by HPCI until the CST is depleted or HPCI shutoff pressure is reached. Whei the pressure permissive is reached, one Core Spray pump-starts (after 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months ' and 10 minutes).to inject into the vessel ..After the level is recovered in the normal range, the Core:Spray system is used to maintain the level with the vessel pressure being controlled by an SRV cycling between 50 and 100 psig. The suppression pool is cooled continuously by the RHR system. The reactor vessel is maintained in this configuration. The RHR pump in torus cooling is also available for drywell spray after 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes.

8) The.HPCI turbine takes steam from the vessel to provide its motive power. It returns the exhaust steam to the torus. The steam to the turbine is not modeled since the model assumes SRY opening and closing to maintain pressure. Any steam not removed by the HPCI turbine will be removed through the SRV to maintain a certain pressure. The total flow through the SRV is increased, but the details of 'SRV flow are not important for this applications and the two (SRV flow and HPCI turbine) can be combined for model simplicity.

9) The liquid leak is modeled as a fixed flow of 8.4585 lb/sec (61 gpm, Reference 3) [(61 gpm

/60' s/min 7.4805 gal/ft 3 10.0161 ft3 /lb = 8.4585 lb/sec)] and it stays on for the entire transient. (Analyses will be performed with and without leak for one depressurization (cooldown) rate: 80 TF/hour). In reality, the leak is variable depending on pressure. Assuming a density of 62 lblft 3 and fixed flow is conservative for the drywell temperature analysis.

Analysis. of drywell temperature for a 45 /F/hourcooldown with no RPV leakage was not performed because for the case -with 80 0Flhr cooldown, for the period of interest the drywell temperature stays below 300 TF for both cases (with and without RPV Leakage). For the 45

Flhour cooldown, the temperature in the drywell for the no-leak case is expected to remain below 300 TF as in the 45 0F/hour cooldown case with RPV leakage for the analysis duration.

A ~Entergy Calculation VYC-2405 Rev. " Page 9 of 85

10) The analysis will assume a fouling of 0.0018 in the tubes and 0.0005 in the shell. This corresponds to an overall RHR Heat Exchanger (RHRHX)'fouling of:

Rf =RO ~d (d:L )+ R where:

2 Rfi and Rf = tube and shell fouling factors, respectively .(hr-fte-F/Btu) do = outside tube diameter (in) d= inside tube diameter (in) d 0.625 in (Reference 22) di= 0.527 in (Reference 22) 2 Rfo = 0.0005 br-fte-F/Btu (Reference 22) 2 Rgi = 0.0018 hr-ft -F/Btu (from 0.0020)

Overall RHRHX fouling 0.625 R. -=0.0018*(_ ___)+0.0005=0.0026 This number compares well with the maximum fouling calculated in Reference 23 6f 0.002307 and 0.002445 for the RBR HX E14-l A and RHR HX E-14-1B, respectively.

11) A variable SW temperature is used, consistent with Reference 1.

Since this change of depressurization (cooldown) function of service water temperature requires procedure changes, it is added in Section 4.1 as an unverified Assumption.

- For SW> 75TF, depressurize the vessel at 800F/hr or higher.

- For lower SW temperature (SW < 75TF no restrictions on depressurization) rates.

12) - Various assumptions made conceming the added Heat Conductors:

- The heat load decreases linearly when the temperature difference between RPV and drywell becomes smaller.

- For all conductors, only heat conduction is conservatively assumed in the air and concrete' layers.

- The outer surface boundary condition is conservatively assumed to be adiabatic (i.e., the heat transfer coefficient is set to zero)

,'*FEntergy Calculation VYC-2405 Rev. 0 Page 10 of 85 Table 1 - Vessel and Core Initial Conditions and Parameters, Primary Variables Parameter Nominal Analysis Basis

. Value Value Initial Reactor Power 1912 Mwth 1950 MWth 100% power +2 % uncertainty (per NEI-87-001, SBO can be performed at 100%

power, however this analysis used 102%

power, consistent with CLTP and the l Reference 1 analysis).

Core Decay Heat ANS 5.1 ANS 5.1 +2 a ANS 5.1 1979 standard+2 a uncertainty

-(Reference 24)

MSIV closure time 3.0-5.0 sec 0.0 sec (MSIVs Minimum value allowed retains the not modeled) maximum energy in the vessel.

RPV Pressure 1015-1025 psia 10452 psia Higher value, conservative, maximizes (Reference 28) _ the vessel energy.

Initial Vessel Level 162 inches 172 inches Analysis value conservatively accounts for 3 inches increase above normal (uncertainty and operational fluctuations) and 7 inches for dimensional uncertainties. These assumptions are LOCA assumptions and are judged conservative for SBO.

Core Flow Rate 48.0E6 lb/hr 51.36e6 lb/hr Includes ICF of 7%.

Initial Feedwater Flowrate 7.876e6 lb/hr 8.076d6 lb/hr Reference 25 (CE2003-20)

Initial feedwater 393.5-393.6 *F 393.9 OF See discussion in Reference 25.

temperature Feedwater is tripped at time 0, due to SBO. The feedwater is used only for the steady state initialization.

SRV Cycling 1080-1047.6 psi 1080-1047.6 psi The setpoints for the SRVs pre nominal.

.(between RPV (between RPV No additional as found allowable of 3% is and Drywell) and Drywell) added since it will have no effect on the drywell temperature since the SRVs open to remove the decay heat and, until the depressurization starts, indifferent of setpoints, the SRVs will cycle to remove the decay heat.

The operators will take manual control of the SRV and will cycle between 800 and 1000 psig (EOP Reference 37) to reduce the numbers of times the valves cycle. There is no effect on the calculation since the valves in any operational mode will open to remove decay heat.

Vessel Leak 61 gpm 61 gpm A constant 61 gpm leakage is assumed; (Reference 3). The analysis will be performed with & without leakage, since the drywell temperature will have a different profile for the cases with no leakage.

a-. ... ~~~~~. .. I , , .,,.

Entergy Calculation VYC-2405 Rev. 0 Page 11 of 85 Table 2- ECCS Initial Conditions and Parameters Parameter Nominal Analysis Value Basis Comments HPCI flow rate 4250 gpm 4250 gpm Tech Spec Flow Since the flow is intermittent there is (Reference 2) no need to use the ruin flow of 3570 gpm (uncertainty added) (References 2 and 5). In reality the HPCI flow will be adjusted to maintain level to prevent excessive pump stop/start.

HPCI pressure 1135-165 psia 1135-165 psia Reference S and 27. HPCI is shut off range if vessel pressure drops below 165 CST Temperature 120 F 135 OF OPEN Item CST available 75000 gallons 75000 gallons Available CST Per Reference 3, the Tech Spec value inventory (VY Tech inventory for IIPCI can be used.

Spec - injection An administrative limit for the CST Reference 2) level of 25% is required.

Core Spray Flow Curve of flow Same as The core spray The Core Spray System will be used vs. vessel- nominal. flow rate used in for level control only after the CST is torus AP. the SBO analysis depleted and/or the low pressure is

-of Reference I will reached.

be used. The flow rate is determined OPIA -Reference 5 as a function of the vessel-torus AP.

(consistent with the LOCA analysis)

RHR Flow 7000 gpm 6400gpm 6400gpm used in Consistent with Reference I (t-7800 seconds) . analyses limiting case and Reference 5.

RHR Hx Fouling 0.0005 shell, 0.0005.shell, Assumption input #10, supported by

_ 0.0018 tube .0.0018 tube . Reference 23.

RHR Hx Tube N/A 5% Allowable Design value providing margin above Plugging plugging margin the current plugging value of 3.6%

RHRSW Flow 470O 4700 gpm (2 4700 gpm (Reference 4)

RHRSW pumps) 4700 gpm =

650.98 lb/sec (at 85> . __

RHRSW Inlet 32-85 OF Variable, see If SW is > 75 T, Maximum Allowable Service Water Temperature assumptions, depressurize the Temperature (Reference 2) only for based on RPV with rates 80 depressurization rates> 80 F/hr depressurization 0F/hr or higher. For lower depressurization rates the (cooldown) rate SW has to be below 75 'F. This requirement is derived from the torus temperature calculation (Reference 1). The rate of depressurization was shown in this calculation to have minimal impact on the strategies to control the drywell temperature for SBO.

ILEntergy Calculation VYC-2405 Rev. 0

Page 12 of 85 . .

Table 3- Primary Containment Initial Conditions and Parameters Parameter Nominal Value Analysis Basis Comments Value Drywell 110-170 OF 170 OF Reference S The highest drywell Temperature temperature is used.

Drywell Pressure 16.4 psia 16.4 psia VY Tech Spec (Reference 2)

Wetwell 88 0F 90 OF Maximum Tech Spec A 2 F uncertainty is applied Temperature Value (Reference 2) via procedure to account for instrument uncertainty (Reference 26)

Wetwell Pressure 14.7 psia 14.7 psia Normal Torus operating pressure (vented to atmosphere via Standby Gas Treatment System)

Drywell Humidity 20 -100 % 100% (base Nominal Values: Use maximum drywell case) VY UFSAR humidity consistent with (Reference 27) Refeience 1 for the base Sensitivity . case. Sensitivities performed at performed at 20% drywell 20% humidity humidity.

Wetwell Humidity 100% 100% Nominal Values: Minimal to no impact on the VY UFSAR SBO drywell temperature.

(Reference 27)

Wetwell Water 6 068000 fte Minimum Tech.

Volume Spec. Value (Reference 2)

Drywell free 128,370 -131,470 131,470 ft Reference S The maximum value in volume ft3 (includes OPLAA is used for vents) Consistent with SBLOCA, IBLOCA and Reference 6, the Small Steam Breaks.

The values volume of the Vents: 16703 ft2 (VYC-proposed are drywell side of the 2306 -Reference 32) consistent with torus-drywell vacuum OPI-4A breakers of 372.3 fe Total Drywell Volume =

will be added to the 131470 - Vents Volume +

proposed value. Drywell side of Vacuum Breakers= 131470 -16703+

372.3 = 115139.3 fi' Wetwell free For the minimum Nominal Reference 5 The value at Dp>0 of volume water level of Values used. I _ _ 105,932.0 ft3 is used for a

'EntergW Calculation WYC-2405 Rev. 0 *Page 13 of 85 Parameter Nominal Value Analysis Basis Comments Value 68000 fte, the Consistent with total volume of 105,932 +

wetwell free The values Reference 6, the 68,000 = 173932 ft3 volume is proposed are volumes of the 107,104.8 ft3 for consistent with drywell side of the Used in calculation: 173932 Dp =0.0 and . OPL-4A torus-drywell vacuum + 99.4 ='174031.4 ft2 105,932.0 for breakers of 99A ft3 Dp>0.0 where Dp will be added the is the pressure proposed value.

difference between drywell and torus.

Vacuum Breakers- 05 psi 0.5 psi 0.5 psi Reference 2 pressure difference between wetwell and diywell for vacuum breakers to be fully open Drywell-to Max allowable Base case Reference S for the max Wetwell Bypass area =0.12 ft2 =0.12 ft2 leakage, Reference 33 for Leakage Tech Spec Allowable

.Tech Spec Allowable Sensitivity

=0.0033 f2 =0.0033 ft2

_ _- .- * . ..v ..a; . ; **

Entegy Calculation VYC-2405 Rev. 0 Page 14 of 85 .. .

4.1- Assumptions that need Verification or Implementation

1) Two (2) hour restoration of outside power (coping time).

2) Ten (10) minutes to start RHR flow through the RHRX, and the use of 2 RHRSW pumps and CS.

3) Acceptability of using 75000 gal from CST (change of level setpoint).

4) Maximum CST temperature of 135TF.

5) The depressurization rate function of Service Water temperature needs to be verified and proceduralized as follows:

For SW> 75IF;.depressurize the vessel at 800F/hr or higher.

- For lower SW temperature (SW < 750F)no restrictions on cooldown rates.

4.2 Affected Documents

1) DBD - for Residual Heat Removal - change the maximum tube side fouling resistance from 0.002 hr-ft 2 o-F/Btu to 0.0018 hr-ft2 -°F/Btu as well as the total fouling.

2) Change the description of the SBO event in the DBD for Safety Analysis.

3) Change all DBDs and documents that address the SBO coping time (identify and modify).

4) Change DBD Containment Pressure Suppression System to incorporate results of this calculation.

5) Review following documents for need of modification: VY UFSAR, and PUSAR.

6) Modify SBO procedure (OT-3122-Reference 36) to incorporate cooldown at one hour and provide guidance to the operators such that RPVED is precluded based on the results of this calculation.

Note: Section 4.1. & 4.2 items are being tracked via LO-VTYLO-2005-00135.

! - .: .. '--I - *%.

:.~ . J' k6i"-. - rI.

. Entergy Calculation VYC-2405 Rdv. 0 'Page 15 of 85 %

5.0 Input and Design Criteria The GOTIJC input from Reference I is modified to implement the features described in this section. The modified input is called SBO-drywell2.

The main features added to the SBO model are the drywell heat load and drywell heat structures.

A schematic of the system modeled is presented in Figure 1 Figure 1 - VY Containment and the Associated Systems Note: only RCIC pump is shown in this simplified model. Actually, BPCI is assumed to inject.

.. . 1Entergy Calculation VYC-2405 Rev. 0 Page 16 of 85 .

5.1 Drywell Heat Load Calculation The Drywell Heat Load Summary at Current Licensed Thermal Power (CLTP) is summarized in Reference10. The total.amount of heat given to the drywell at CLTP is 1,691,300 Btufhr. The drywell heat load was recalculated in Reference 18 for the Extended Power Uprate (EPU) as 1,700,675 Btu/hr. Since the Extended Power Uprate is performed at constant pressure, only the feedwater pipe and valves will be at higher temperatures (Reference 18), hence a higher Q for this component; (124,000 Btu/hr -Reference 10 versus 133,375 Btulhr at EPU -Reference 18) is calculated.

The total power to the drywell for EPU is presented in Table 4.

Table 4 Drywell Heat Load (Reference 18)

Item . Component Cooling Load No. (Btulhr) 1 Reactor Vessel 459,000 2 Recirc. Pumps, Valves and Pipe 278,000 3 Feedwater Pipe & Valves 133,375 (EPU Modified-

. _Reference 18) 4 Steam Pipe & Valves 212,000 5 Condensate & Instrument Lead Lines 82,000.

6 Control Rod Drive Pipe 50,400 7 Clean-up Pipe & Valves 17,800 8 Shutdown Supply Pipe 8,100 9 Steam Safety/Relief Valves 206,600 10 Biological Shield (Gamma Heating) 16,400 11 Safeguards System Piping 82,000 12 Steam Leak . 155,000 Total 1,700,675 Btu/hr

--- 7--

Entergy Calculation WYC-2405 Rev. 0

Page 17 of 85 5.1.1 Drywell Load Modeling The drywell heat is modeled as two heater #5H and #6H. The heater 5 represents the heat source which varies function of the liquid temperature in the vessel, while heater 6 represents the heat source which varies as a function of the vapor temperature in the vessel. See explanation of these two heaters in Section 5.1.2.

The heat loads which are exposed to the steam atmosphere (for Heater #6H) are:

Table 5 Heat Loads Exposed to Steam Item No. Component Cooling Load

. __._(Btu/hr)

. 30% of Reactor Vessel Heat Load 459,000*0.3 ='137700 4 Steam Pipe & Valves 212,000 9 Steam Safety/Relief Valves 206,600 12 Steam Leak 155,000 Total 711,300 The normal level is at about 0.3 of the total vessel height. Prom Reference 34 the distance from the.152 inches above TAF to the top of the vessel is 21.432 ft in the GOTHIC vessel model and to the vessel bottom is 41.193 ft.

The middle range of 152 inches is calculated as (177 inches + 127 inches)/2 = 152 inches.

The model assumes a normal level of 172 inches (Table 1) which is 20 inches above the 152 inches, hence from 172 inches above TAF to the top of the vessel there are -21.432 - 1.667 19.765 ft The liquid height = 41.193 + 1.667 ,42.86 ft Total GOTHIC vessel height = 62.625ft'(from Reference 34 =330.542 - 267.917 =62.625 ft)

Steam region = 19.765 /62.625 = 0.31 (used 0.3)

Heater 6 load = 711300/3600 = 197.58 Btu/sec Total Heat load = 1,700,675/3600 = 472.41 Btu/sec Thus, Heater 5 load = 472.41 - 197.58 = 274.83 Btulsec

- Entergy Calculation VYC-2405 Rev. 0 Page 18 of 85 5.1.2 Transient Heat Load Behavior It is assumed that the heat load decreases linearly when the temperature difference between RPV and drywell becomes smaller. In order to calculate the transient heat load, the following transient heat load procedure is used

1. When the difference between the vessel temperature and the drywell temperature is less than or equal to zero, the power of the heat source is zero.

2. When the temperature difference is greater or equal with to T,,t, the heater power will increase above the nominal value.

3. When a temperature difference exists between Tt.and zero, the power is linearly interpolated between'the nominal value and 0.0.

T= is defined in the way that the calculated power of the heat source is equal to -the nominal value at the beginning of the transient. Two GOTHIC control variable CV 41 and CV 42 are defined as the temperature difference between the vessel internal water temperature and the temperature inside the drywell (CV41) and between the vessel steam temperature and the temperature inside the drywell (CV42), respectively. The model shows higher steam temperature than saturation because of the heat slab exposed to steam which represents the heat structures in the Reactor Pressure Vessel .(RPV) exposed to a steam environment. Sensitivity studies which placed this heat structure in liquid eliminated the steam superheat, as expected. This is a' conservatism of the model. In reality all structures will be exposed to Tt = TuE Tyap The control variables are used as the independent variable of the functions, which gives the transient heat loads to the drywell, as described above.

5.2 Drywell Thermal Conductor Model Development The following thermal conductors are being added to the model.

There are several types of heat sinks and thermal conductors inside the drywell. The components included as heat sinks are the metal mass of 4 RRUs, vent pipes and the drywell liner.

Miscellaneous steel exists in the drywell, but has -not been previously quantified in detail.

Minimum heat sink components are considered conservative; therefore, miscellaneous steel is not included as heat sinks.

Dr~vell liner divided in (Reference 21):

1) Lower Drywell spherical portion,

2) UpperDrywell cylindrical portion, and

3) Drywell head.

The drywell wall consists of the concrete, the inner surface steel plate and the air gap. Zero heat flux boundary condition on the outside surface of the drywell wall is used.

- -_ x 7.F o . . ........ f ' ... . .-

, ., . . ,-_ ..: .. -. "I 1-4

`'Entergy Calculation WC-2405 Rev. 0 Page 19 of 85 ,.

The data on OPI4A (Reference 5) is used to model the steel liner. The surface area calculation for the liner was performed in Reference 9.

91 Table 6 - Drywell Steel Liner Elevation Steel Thickness (in) Surface Area (fte) GOTHIC Item #-from Ref. 9 . thermal conductor

_ __ __ _ __ __ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ N o.

2, page 47 of Ref. 9 El. 237.74'-E5 1.0 page 47 of Ref 9 1856.24 5 257.75' 3,page 47 of Ref. 9 El. 247.24'-El 0.8125 page 51 of Ref 9 2041.28 6 257.75' _ _ _ _ _ _ _ _ _

4.1, page 47,48 of Ref. 9 El. 257.75'-EL 0.6875 page 51 of Ref 9 1250A7 9 259.92' .

4.2, page 48 of Ref. 9 El. 259.92'-El 0.6875 page 51 of Ref 9 3802.73 7 28 3.69 ' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

5, page 48 of Ref. 9 El. 283.69'-EI. 2.5 page 51 of Ref 9 780.68 8 289.61' _ _ _ _ _ _ _ _ _ _ _

6, page 48 of Ref. 9 El 289.61'-El 0.635 page 51 of Ref 9 1898.24 10

_ _ __ __ _ _ _ _ _ _ _ __ _ _ 3 0 8.00 ' _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _

7,page 48,49 ofRef. 9 E.308.00'-El 1.25 page 51 of Ref 9 1114.72 11 318.50' _ _ _ _ _ _ _ _ _

8, page 49 of Ref. 9 El318.50'-El 1.25 page 5l of Ref 9 783.4 12 9, page 49 of Ref. 9 El. 327.75'-top 1.3125 page 51 of Ref 9 17183 13.

__ _ of drywell _

Total 15246.06 ftv The items 2 through 7 have 0.0025 inches of paint per Reference 13 and Reference 9, Appendix VI (for properties) and a 2 inches thick air gap (Reference 14) and a conservative low thickness of concrete of 24 inches is used from Reference 21.

Item 8 (side of drywell head -small cylinder) is modeled, with a 2.5 ft air gap outside the steel wall and conservatively low thickness of 1.5 ft of concrete (scaled from Reference 21). The thermal conductor has an adiabatic heat transfer boundary condition. Only heat conduction is assumed in the air and concrete layers. This is conservative.

Item 9 (top of drywell head) is modeled with a 6.7ft air gap outside the steel wall and a conservative low thickness of 24 inches of concrete (part of the concrete plugs) - (scaled from Reference 21).

RRps_(References II and 12)

ARRU = 1272.8 ft2, thickness =0.125 inches (used in OPL-4A-Reference 12).

. Entergy Calculation VYC-2405 Rev. 0 Page 20of 85 Vent Pipes Vent pipes surface area A vent pip=2 8 85. 7 ft2, thickness =0.125 inches (used in OPL4A-Reference 12)

NOTE: The total Surface Area of the steel components adds up to the value obtained from OPL-4a (i.e., Table 6 Total = 15246.06 ft2, RRUs = 1272.8 ft2 , Vent Pipes = 2885.7 ft2 , thus Total = 19405 ft2)

Total Surface area of Concrete Exposed to Drywell Air Space The surface area 'of the pedestal is the only concrete component quantified in OPL-4A. .The drywell floor is ignored because it may be covered with liquid and not directly exposed to the drywell airspace. Only the outer surface area of the pedestal was considered in OPL- 4A as well as Reference 9 because the inner surface has limited communication with, the drywell atmosphere. The biological shield wall (BSW) is a concrete structure surrounding the reactor pressure vessel and located above the reactor pedestal. Because of the proximity to the -reactor pressure vessel the BSW is at a temperature greater than the drywell (DW) ambient and thus a heat source (already incorporated into the drywell heater) and a heat sink only when its temperature drops below the DW temperature.. Because of the uncertainty of the BSW temperature and its limited value as a heat sink, the BSW is not considered here.

The OPL-4A value for the area is used and = 2068 ft2 (A value of 2108 ft2 was used in the model, addressed in Case 5).

Thickness of Concrete Exposed to Drvwell Air Space From Reference 5 = 4ft.

Properties of Materials Table 7 - Thermo physical properties of Passive Heat Sink Materials (Reference 5)

Material Density (Ibm/ft3) Specific Heat Thermal References (Btu/lbm-OF) Conductivity

__ (Btu/hr-ft OF)

Carbon Steel 489.0, 0.11 32 OF 31.8 15 68 F 31.2 212 F 30.0 392 T 27.8 572OF 26.0 Concrete 145 0.156 0.92 16 Paint 288 0.2 0.125 9

..V nesCalculation WYC-2405 Rdiv. 0} Page 21 of5 '

Air Ilermnal Properties From Reference 15.

T (OF) KX(Btu/hr-ft-0F) Cp (B3tu/lbm-°F) p(lbmn/ft3)

100 0.0157 0.24 0.07092 150 0.0167 0.241 0.06511 200 0.0181 0.241 '0.06017 250 0.0192 0.242 0.05593.

300 0.0203 0.243 0.05225 400 0.0225 0.245 0.04617 Heat Transfer Boundary Conditions On the inner surface of all the thermal conductors, the heat transfer coefficient is calculated by the GOTHIC code. The following options are used:

-Direct'Heat transfer Option.

-Sumrnation of the condensation and convection heat transfer.

-Max of Uchida and Guido-Koesfel condensation heat transfer option (sensitivity with Uchida for the limiting cases).

-Radiation heat transfer option is OPF for.all heat structures with exception of the drywell dome.(sensitivity with option OFF for the limiting case).

-The surface orientation is "FACE DOWN for the drywell domne", thermal conductor

13,and "VERT SURF' for heat conductors 5 through 12.

-All thermal conductors use 'VAP 'option...

The outer surface boundary condition is conservatively assumed to be adiabatic. The heat transfer coefficient is set to zero.

5.3 GOTHIC DrywelI SBO Model Development The following changes to the input SBO-NoLeak-80 (Reference l) to produce SBO-dryweZl2 are being added:

- drywell beat load

- drywell heat slabs

- leakage from'drywell to wetwell

- modifications to the vacuum breaker modeling The GOTHIC input is presented in Attachment A.

The GOTHIC model used for all cases is presented in Figure 2.

. 0-- -.- -. - .. -- - -- . . - .1. : .., :"S -".. '. - -, - '. .

.. 1.Enteqgy -

Calculation VYC-2405 Rev. 0 Page22 of 85W'

' SBO - S0P-Noleak-dryweqll Mar/15/2005 14:09:56

,GOTHIC Version 7.0p2(QA) - April 2002

File:. /home/schor/vyc-2120ccn/SBNSITIVITY/SBO/drywell-SBO/SBO-drywell2 LoacnmH - Long cem Contalrat UcOCA s, ; , . . I I

I I

.1 Figure 2 - GOTHIC SBO Model

>9 -.- ____ _._ ___ __ _,_ ___ __ ___.__ _5 .. ~ ,*, .;g. .sr e;.A 1- Z.-.

-*Entery CalculationVYC-2405ERe v.0 Page 23 of 85 Flow Paths

.Flow path.21 is added to model the vacuum breakers leakage path from the drywell air space to.

the wetwell air space. Per Reference 5, the maximum leakage area is 0.12 ft2. The elevations and the height of this junction were elected to be the. same as the vacuum breaker junction since the leakage is "aLround.the vacuum breakers.

The K rverse =K forward= 1.5 (expansion & contraction) (Reference 29)

Thermal Conductors Twelve new thermal conductors were added. The description of the thermal conductors was given in Section 5.2.

The temperatures of 11 of the thermal conductors were set at 160 TF, the pedestal thermal conductor is set at 152 'F. On page 73 of Reference 9 the average temperature for the middle and the top drywell node is calculated as 151.94 'F. Hence, the thermal conductors are set, conservatively at i 60 'F. The pedestal is in the lower drywell and middle drywell hence 152 'F is

.used (average for middle and top drywell is conservative). In Case 5, the temperatures of the heat slabs which represent the drywell wall were set at 170 OF (very conservative assumption).

Functions Two new functions are added Function 17 (FF1 7) represents the power to the drywell from the structures exposed to steam.

The function multiplies Q idit and represents (Tliquid - Tdzywdu).

The FF17 is:

AT (CV41)

-500 0 0 0 380 1 380000 1000 The independent variable is the temperature difference between (Tlquid T&yweU), CV4 1.

Function 18 is identical to the Function 17, but the independent variable is .CV42 (T vapr- Td&wCII)

Tref = Initial Vessel temperature - Initial Drywell temperature = 550'F -170c'F =380 OF where 550 'F is the initial vessel temperature, and 170 'F is the initial drywell temperature. (550 is determined from the GOTHIC model at time zero and 170 'F is the maximum drywell temperature, OPL4A-Reference 5).

' ~Enter - Calculation VYC-2405 Rev. 0 Page 24 of '85 Valves A new valve (V5) was added to represent the vacuum breaker. It opens on trip 33 (0.5 psi difference between wetwell and drywell (Reference 2) and it closes on trip 34 (0.3 psi-arbitrary, since the vacuum breaker valves.reseat when the pressure difference becomes less than 0.5 psi. A quick close valve is used for this component since the valve will close as soon as the 0.5 psi difference between wetwell and drywell disappears.

The vacuum breaker valve is modeled as Valve Type 3, with an area of 17.6737 ft2 (Reference 9). Note: the area of the valve from Reference 9 is slightly larger than the area of the flow path in which it is located. The valve area will have minimal impact on this analysis because the flow is limited by the area of the flow path. The area of the valve was changed to the area of the flow path in the final case analyzed, Case 5.

Materials Four new materials are added. The properties for the new materials are described in Section 5.2.

Trips and Controls Trip 18 is modified to ADS when the vessel pressure difference between RPV and drywell is lower than 100 psi instead of 50 psi in the original model. This trip is not used, however, the SRV valves will open at a AP of 100 psi, not 50 psi.

Trip 21 is modified to start depressurization (cooldown) at one hour (3600 seconds) in order to limit the drywell temperatures.

Trips 33 and 34 are added to open the vacuum breakers valves at 0.5 psi pressure difference between wetwell and drywell (trip 33) and close it on a AlP of 0.3 psi.

Coolers & Heaters Two new heaters are added, 5H and 6H to model the vessel heat to the drywell. These heaters are described in Section 5.1.

For heater 5H the heat rate of 274.83 is multiplied by the FF 17, while for heater 6H the heat rate of 1997.58 is multiplied by FF 18.

-Entry. Calculation VYC-2405 Rev.:0 Page 25 of 85 Heat Transfer Coefficients Types Two heat transfer coefficients are added.

The following options are used:.

-Direct Heat transfer Option.

-Summation of the condensation and convection heat transfer.

-Max of Uchida and Guido-Koestel condensation heat transfer option (sensitivity with Uchida for the limiting cases).

- Radiation heat transfer option is OFF for heat transfer coefficient type 6.

- Radiation heat transfer option is ON for heat transfer coefficient type 7.

The use of the radiation option has no effect on the results at these low temperatures

- The surface orientation is "VERT SURF' for heat transfer coefficient type.6.

- The surface orientation is "FACE DOWN for the drywell dome", heat transfer coefficient type 7.

The use of the surface orientation is appropriate since this is the thermal conductor physical arrangement.

-The.heat transfer coefficient types 6 and 7 use 'VAP" option since this is the drywell medium.

-Convection bulk T model: TgTf. The bulk temperature is the calculated vapor temperature. Tf is the .maximum between the calculated wall temperature and the calculated saturation temperature.

- Condensation heat.transfer Bulk T Model : Tb -Tw used. Tb is the minimum between the calculated vapor temperature and the calculated saturation temperature.

Control Variables Two control variables are added, 41 and 42 they represent the AT between Tiiq in RPV and. Tv drywell and between Tvap in RPV and Tv drywell, respectively. See Section 5.1.2 for additional information on the operation of these Control Variables.

. Entegy Calculation VYC-2405 Rev. 0 Page 26 of 85 ;

6.0 Calculation / Analyses Five (5) cases are analyzed:

Case I is called SBO-drywell2. It is the base deck, developed from Reference 1 and described in Section 5.3. Case I assumes no RPV leakage, depressurization (cooldown) with a rate of 80 TF/hour a 100% humidity and base deck inputs as described in Section 5.0.

Case 2 is called Case SBO-drywell2-80-sensy2-NoLeak Case 2 is identical to Case I with the change in humidity, changes in the leakage area-and minor changes in the heat transfer type 6 and 7. Case 2 assumes 20% humidity and minor changes in the heat transfer type 6 and 7. These changes are described in Section 6.2.1.

Case 3 is identical to case SBO-drywe112-80-sensy2-NoLeak but with leak. It is called SBO-drywell2-Leak-80-sensy. These changes are described in Section 6.3.1.

Case 4 is called SBO-drywell2-Leak-45-sensy and is identical to Case 3 but with a slower depressurization (cooldown) rate. It assumes a depressurization of 45 Manhour with leak in

order to show that with an early depressurization and slower cooldown rate the results are not changed and the drywell temperature is not impacted negatively by a slower cooldown. Consistent with Reference 1, the RHRSW temperature is changed to 75 'F.

Case 5 addresses changes found during documentation and as part of review. These changes are described in Section 6.4.1.

Case 5 is called SBO-drywell2-conmments. The following changes are made in Case 5 to address changes found during documentatfon and review:

. change the temperatures of the steel structures from 160 'F to 170 'F (very conservative assumption),

change the K reverse for the Junction 3 from to 3.93 from 3.964,

set the V3 Valve with the same area as the junction, and

. change the area for the pedestal from 2108 ft2 to 2068 ft2, consistent with OPL-4A.

Case 5 changes are described in Section 6.5.1.

~~~- a- . -~ . . . . ..<<....A t.s_2.

. Calculation WC-2405 Rev. 0 Page 27 6f 85 6.1.1 SBO-drywell2 Model Development SBO-drywell2 represents the base model for this calculation. The modification to the input are presented in Section 5.0 and the GOTHIC input deck is presented in Attachment A.

6.1.2 Case SBO drywell2 Results Figure 3 through Figure 11 present the main parameters for the base case SBO-drywell2. Figure 3 presents the drywell temperature. The drywell temperature increases to about 285 0F after one hour. The heatup is*arrested due to depressurization. At about 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months into the transient the temperature in the drywell starts increasing due to lower heat removal into the passive heat sinks (walls). The maximum drywell temperature is 290 OF. The air gap acts as an insulation and the steel liner is almost at 245 'F. However, after 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes the low pressure pumps are available so the operators can spray the drywell with the RHR pump, if needed. The results indicate that the temperatures in'the drywell stay below the EQ limit and the drywell liner is well below the 281 'F for the SBO coping duration.

Figure 4 presents the containment pressure. Due to the higher leak area the drywell and the wetwell are at the same pressure. At about 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months the pressure in the drywell is too low to spray the drywell, (unsafe area of DWSIL(EOP-3 -Primary Containment Control -Reference 30))

however the pressure increases to about 6 psig at about 12000 seconds at which point the operators would able to spray the drywell with the RHR pump, if needed.

Figure $ presents the RPV pressure. At one hour into the event it is assumed that the operators start depressurization (cooldown). The pressure drops to the HPCI shutoff pressure of 165 psia at about 12000 seconds. At that point only about 450000 lb were injected from CST (Figure 12). At this point the RPV is depressurized and the CS is available to inject.

Figure 7 presents the RPV level. The core stays covered. There is a dip in the normal level at about 12000 seconds when HPCI stops injecting and CS pump has not yet injected. This is due to the fact that the CS pump was set to inject at 14000 seconds; however CS is ready to inject at 7800 seconds.

Figure 8, Figure 9, and Figure 10 presents the drywell liner temperature. The drywell liner stays below 260 'F for the 7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months analyzed. After 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes the low pressure pumps are available for suppression pool cooling, drywell spray and maintaining vessel inventory.

Figui re II presents the suppression pool temperature, Since the vessel is depressurized early, the suppression pool temperature is below the maximum of 182.2 'F calculated in Reference 1, hence no containment overpressure is required.

-n - -.

'Enter .. Calculation VYC-2405 Rev. 0

Page 28 of 85 t SBO - SOP-Noleak-drywelli Mar/03/2005 13:28:01 GOTEIC Version 7.0p2(QA) - April 2002 File: /home/schor/vyc-2120Ccn/SRlNSITfITY/SBp/dryvell-SBO/SBO-drywell2 If__ __§l A"- _ _3 I..

toIA (Sac) 6<0 i.h2fefl3 k~d1)~2eS% S14,4S.

Figure 3 Drywell Temperature Case SBO-drywe!12 S9O - SOP-Noleak-drywelli Mar/03/2005 13:33:59 GCOlC-Version 7.0p2(QL) - April 2002 Pile: /bheiemch3orvyc-2120ccI/sVSIlTVrf/SBO/dryweell-SBO/SBO-dyviell2 3

Cmta1met Yrahzur 13.2 13

';TF17 71

I I

-1 It

'4

#4 - --- - -

LLLL LLLI LLL5-LLL5J-LLJ U 0 4 *JkJ.I 24 2 _81 TIM (see)

., I.,Uo) .JIngc% ll4At -

Figure 4 Containment Pressure - case SBO-drywell2

.. IIiEntergy Calculation VYC-2405 Rev. 0

Page 29 of85 .. .. . .

SO - 80F-Noleak-d ell. ..

Mar/03/2005 13:29:49 GOTaC Version 7.0p2 (A) - April 2002 Pilet /home/schor/vyc-2120cm/SENSITflTY/SBO/dryrell-SBO/S3Q-drywell2 24 rI Presssts a- -1 1 1 - 1 I I

.,..1

.I_ J

. N.2~. 1o 4 . j1 21.6 25.2 r 74 8

. .'. AC (sea) 4VM 1.41CM> f"163tZOOS UtI49.4 Figure 5 - RPV Pressure - Case SO -drywell2.

SEO - 80P-Nol&ak-dxywelll.

Mdr/03/2005 13:30:33 GOG C Version 7.0p2(QA) - April 2002 File: /home/rchor/vyc-2120ccn/SSEITIVIrTJSBO/drywell-SBOISBO-dzywelli 4eaator vessel ?TmeratIe TY4 YL4 ST4 TD4

s or Yeal T=Ree

Tin (.Ito)

..-. --. 1 ....... .... --

Figure 6- RPV Temperature - Case SBO-drywell2

-'Entergy Calculation VYC-2405 Rev. 0 Page 30 of 85

.WW SBO - 80C-Noleak-drywelll Mar/03/2005 13:01:07 GOTlMC Versicn 7.0p2IQA) - April 2002 File: /home/schor/vyc-2l2OccI/SINSITI TY/SBO/drywell-SBO/SBO-drywell2 21 sV Liqu14 14r1 '

.'M -.1- 'II -1 1 1v41 A- .~ .? I

_ Is .. Ci

's~2 1s 3-47tiI Sa~te Figure 7 - RPV Level - Case SBO-drywell2 SB0 - 80P-Noleak-drywelll Mar/03/2005 13:29:1 GOTHIC Versic .7.0p2(QA) - April 2002 File: /hone/schor/vyc-2120cn/SENSflIVrrT/S3O/dryweil-SB0/S3O-drywe12 1n g 75,7 7

tc 14 z)

, , re (Zen)

<2MV T.h204) NWO/2143 1t 49,42 Figure 8 - Surface Temperature for Heat Slabs 5,6 &7 - Case SBO-drwell2

-I '-

-1 ....; 4'. , -:.-r -- . ..

. Entergy Calculation VYC-2405 Rev. 0 Page3ltof 85 6 SBO - 80P-Noleak-drytelll Mar/03/2005 13:28:24 GOTHIC Version 7.0p2(QA) - April 2002 File: /home/scbor/vyc-21220.2ccn/52NSr MYT /C /drywe.1-SBO/SBO-drywell2 sI surface I"e trtzre

_I3 _

no TU 11

.4

- .5I .1nQ711.

Tint (sea)

6. #PY~2(#R)N.,jtP9 11,49,40 _

Figure 9 Surface Temperature for Heat Slabs 8,9,10,11 - Case SBO-drwell2 Sao - SOP-Noleak-dryieelll 14ar/14/2005 14:45:07 GOTC Version 7.0p2(QA) - April-2002 File: /houe/schor/vyc-2120cc1/SMSITIVf/SBO/drywell-SBO/SO-drywell2 52 surface teperature ThI1 D512 r 0

.I, i Z Tine (see) so" 7 ¢*}wvr~ 21,2 nml Figure 10 - Surface Temperature for Heat Slabs 12 and 13 - Case SBO-drwell2

IE5 Entergy Calculation VYC-2405 Rev./O Page 32 of 85; sBO - 80F-Noleak-drywelll Mar/03/2005 13:31:38 GO7IC Version 7.0p2 (GA) - April 2002 File: /home/schor/vyc-212occn/sRxsnlViTSIo/dryvell-SBo/sBo-drywen2 2 V tttu Smperautr S12 7L'2 b.

5.0 lin (seo)

-,,~zPn~wZwa*-w Figure 11 Suppression Pool Temperature case SBO-drywell2 8BO - SOP-Noleak-drywelll Mar/1412005 14:02:38 GOTOIC Version 7.0p2(QA) - April 2002 File: /bowe/schor/vyc-2120ccn/SzSrVrY/SfBO/dry/el-SBO/SBO-drywell2 40 r~tegratd Zc1 Pfa

° Sfl21

_ - -I

3 VI 1 1

10

.0 a

o a ..

0-F 77 I ZLLZL."

A a;

!L~-I 8 S.5 7.2 2@.1 14:4 i1 21.8 2 .i 21*3 lime (sea) n-e 1-.9f~ell -11412$ST 1s.13.s Figure 12- Integrated HPCI Flow - SBO-drywell2

.. Entery CalculationrVYC-2405 Rev. 0 Page 33 of 85 'V 6.2 Case SBO-drywell2-80-sensy2-NoLeak 6.2.1 Model modification Table 8 presents the modifications to the base deck SBO-drywell2 to produce SBO-drywell2 seny2-NoLeak.

The following modifications were made:

The Heat Transfer Coefficient Types 6 and 7 were modified to use Uchida-correlation for.

condensation heat transfer instead of MAX (maximum of Uchida or Guido-Koestel). For.this case since there is no RPV leakage, the choice of condensation correlation should have a minimal impact on results.

For the Heat Transfer Coefficient Type 7 the radiation option was turned off. Again, at these small temperatures, the radiation has a minimal impact on results.

The humidity in Volume 1 (Drywell) was modified from 100% humidity to 20% humidity to encompass all the humidity range in the drywell (Reference 5).

The reverse loss coefficient for the vacuum breakers was changed from lel8 to 3.964 (equal to the forward loss coefficient).

A coefficient of 3.93 should have been used. This is corrected in Case 5.

The vacuum breaker reverse coefficient is the weighted sum of the flow paths 7, 8 & 9 of Reference 9. (Same as the foriward loss coefficient)

K reverse = 1.168 (15.63/16'.23)2 + 2.528 (15.63/16.23)2 + 0.5 (1.53/1.53)2 = 3.93 (A K of 3.96 was used, less than a 1% difference)

The area of junction 21 is changed from 0.12 ft2, maximum leakage to 0.0033 ft2 (allowable Tech Spec leakage) -Reference 33.

Table 8; Input Modifications-SBO-drywell2-80-sensy2-NoLeakvs. SBO-dryweIl2

. EnterWy 0alcufation'VYC-2405'Rev.' O ." Page 34 of 85 t

C.

R Modifications In /hame/schor/vyc-212Occn/SENSITIVITY/SBO/drywell-SBO/SBO-drywel12-80-sen Mar/10I2005 10:42:03 GOTHIC Version 7.0p2(QA) - April 2002 File: /home/schor/vyc-2120ccn/SENSITIVITY/SBQ/drywell-SBO/SBO-drywell2-80-sensy2-NoLeak

Graphs (continued)

Graph Curve Number Title Mon - I 2 3 4 5 42 Heat to the sup CQ4H CQ2H 43 Leak Flow FL4 FLl9 FL20 44 Integrated Leak cv40 45 Title' cv39 46 FV18 FL18 FD18 47 cv38 48 Surface Tempera 1B5 TB6 T67 49 Surface Tempera .TA8 TA9 TA10 50 Surface Tempera . TAI1 TAI2 TA13 51 Surface Tempera TB8 TB9 TB10 TB11 52 Surface tempera TB13 TB12 53 TP8t600 TPRt600 TP1Ot6O TP1lt6O 54 TP13t50 TP12t50 55 Drywell Temera mY1 lUl Heat'Transfer Coefficient Types - Table 1 Heat Cnd . Sp Nait For Type Transfer Nominal :-Cnv Cnd Cnv m Cnv Cnv Rad

Option Value FF.Opt Opt HTC Opt Opt Opt 1 Correlat 0 VERT SURF PIPE FLOW OFF 2-Correlat 0 VERT SURF PIPE FLOW OFF 3 Correlat FACE DOWN PIPE FLOW OFF 4 Correlat . . FACE UP PIPE FLOW OFF 5'Sp Heat. 0.

6 Direct ADD ULCI VERT SURF PIPE FLOW OFF 7 Direct ADO UDII FACE DOWN PIPE FLOW OFF Run Control Parameters (Seconds )'

Time DT OT DT End Print Grajh Max. Dumo Phs Chn Int Min Max Ratio Time Int Int CPU Int Time Scale I le-06 1. 1. 100. 5. 0.1. leO06 0. DEFAULT 2 le-06 1. 1. 1200. 50. .1. le+06 0. DEFAULT 3 le-06 1. 1. 1300. 500. 10. let06 0. DEFAULT 4 le-06 1. 1..25000. 600. 10. le+06 --0. DEFAULT

.. .. Entergy.. -

Calculation'VYC-2405' Re'v.'0 ' Page 35 of '85 '

Modifications in /hame/schor/vyc-2120ccn/SENSITIVITYISBO/drywell-SBOISBO-drywel12-80-sen Mar/10/2005.10:42:03 GOTHIC Version 7.0p2(QA) - April 2002 File: /home/schor/vyc-2-20ccn/SENSITIVITY/SBO/drywell-SBO/SBO-drhel12-80-sensy2-NoLeak

,Voliune Initial Conditions Vapor Liquid Relative Liquid Ice Ice Vol Pressure Temo. Teop. Humidlty Volume Vol ume Surf.A.

(psia) (F) TF (1) Fractlo Fract. (ft2) def 14.7 80. . 80. 60. 0. 0. 0.

1 16.4 170. 170. 20. '0: 0. 0.

2 14 7 90. 90. 100. 0.39497 0. 0.

3 16.4 90; 90. 100. 0.00595 0. 0.

4 1045.2 549;97 533.12 100. 0.60794 0. 0.

Graphs 6rF h Curve Number Title Mon 1 2 '

3 .4 5 1 Qrwell Tenpera IV1 TL1 2 Wetwemllempera TV2 TL2 3 Containment Pre PRI PR2 4 Reactor Vessel rV4 TL4 ST4 TD4

5. FdiR Heat Exchan xqlH 6 Reactor Vessel AL4 7 Torus Water Vol AL2 8 Heat Exchanger tilH t21H 9 Wetwell. Vessel TL4 TL1 10 Conductor Teaipe TA1 .TA2 TA3 TA4M 1I Integral Vessel 'QV4 12 Vapor & Conduct: QL4 TA1.

13 Liquid & Conduc TL4 1A2 14 Vapor &Conduct- TV2 TA3 15 Liquid & Conduc TL2 TM 16 Vapor Heat Tran' 17 Liquid Heat Tra. hA2 18 Vapor Heat Tran HA3 19 Liquid Heat Tra hA4 20 '.Feedwater & Bre FL9 FL4 21 RPV Liquid Leve LL4 SRV and ADS Flo FV10 FY11 23 Feedwater Entha cY29 24 RPV Pressures PR4 25 Feedhater Contr cv27 CY28 26 Integrated Feed cv4 27 RHR Flow. FL5 28 Vessel Droplet. AD4 29 ECCS Injection FLU FL12 FL14 30 RPV Pressure PR4 VC2V FQ 31 ADS Valve Posit VC3V 32 SV Position VC2V 33 Cooldn FLow FY16 34 Vessel DropiDia D14 35 Reactor Vessel *PR4 36 Suppression poo TL2 37 Reactor Vessel PR4 38 Suppression Poo LL2 39 HPCI Flow Rate FL18 40 Integrated HPCI cv39 41 Core Spray Flow FL8

"Entergy . .

Calculation VYC-2405 Rev. 0 Page 36 0f 85 '

Modifications -in /hc~me/schor/vyc-2120ccn/SENSITIVITY/S8O/drywe 1-SB0/SBO-dryel 12-80 -sen Mar/10/2005 10:42:03 GOTHIC Version 7.Op2(QA) - kpr1l 2002 File: /hIme/schor/vyc-2120ccn/SENSITIVITY/SBO/drywl 1-SBO/SBO-drywel l 2-80-sensy2-NoLeak

? Flow Paths - Table 2 Flowi Flow Hd. Inertia Friction Relative Oep Mam Strat Path Area Dm am. Length Length Rough- Bend Trn Flow,

(ft2)- (ft) (ft) (it) ness (deg) Opt Opt 1 283.529 6.75 89.13 0. 0. - NONE

2 286.114 1.948 4.16' 0. -1. - NONE 3 15.63. 1.5625 44.925 28.72 0. - NONE 40.001005 0.03568 0.1 0. - NONE S5 3.14 2. 0.01 0. - NONE 6 3.14 2. 0.01 0. - NONE 7 3.14 2. 0.01 0. - NONE 0.01 0. - NONE 0.01 0. -NONE 10 0.09945 0.01 0. - NONE 0.35584 0.01 0. - NONE

2. 0.01 0. - NONE 13 3.14 2. 0.01 O0. - NONE.

14 3.14 2. 0.01 0. - NONE 15 3.14 2. 0.01 0. - NONE.

16 3.14 2. 0.01 0. - NONE 17 3.14 2. 0.01 0. - NONE:

18 3.14 -2. 0.01 0. - N0NE 19 0.5454 0.8333 0.1 0.1 - NONE 20 0.5454 0;5454 0.1 0.1 - NONE 21 .0.0033 1. 1. 0;- - NONE Flcw Paths - Table 3 Flow FWJd. *Rev. Critical Exit. Drop Paih Loss Loss Conp. Flow Loss Breakup Coe ff. Coeff. Opt. Model Coeff. MIodel 1 4. 243 4.2243 ON TABLES 1. OFF 2 1. 0.78 ON TABLES 1. OFF 3 a .964 3.964 ON OFF 0. OFF 4 0. OFF TABLES 1. OFF 5 OFF OFF O; OFF 6 OFF OFF .0. OFF 7 OFF OFF 0. OFF 8 OFF OFF . . OFF 9 OFF OFF O.. OFF OFF TABLES 0. OFF 11 OFF TABLES 0. OFF 12 OFF OFF 0. OFF 13 OFF OFF 0. OFF 14 10 OFF OFF 0. OFF 15 OFF OFF O.- OFF 16 letl8 OFF OFF 0. OFF 17 OFF OFF 0. OFF 18 OFF -OFF 0. OFF 19 OFF OFF 0: OFF 20 OFF OFF. 0. OFF 21 1.5 1.5 0FF OFF D. OFF

WEn y . Calculatidn VYC-2405 Rev. 0 Page 37 of 85 -

6.2.2 Case SBO drawell2-80-sensy2-NoLeak Results Figure 13 through Figure 20 present the main parameters for the case SBO-drywell2-80-sensy2-NoLeak. Figure 13 presents the drywell temperature. The maximum drywell temperature is about 289.4 OF and is reached after one hour and 30 minutes.. The heatup is arrested due to depressurization. At about 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months into the transient the temperature in the drywell starts increasing, due to lower heat removal into the passive heat sinks (walls) . The air gap acts as an insulation and the steel liner is almost at 255 'F. However, after 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes the low pressure pumps are available so the operators can spray the drywell with the RHR pump, if needed. The results indicate that the temperatures in the drywell stay below the EQ limit and the drywell liner is well below the 281 'F for the SB 0 coping duration.

Figure 14 presents the containment pressure. Due to a lower leak area the drywell and the wetwell are at not at the same pressure, the vacuum breaker opens to relieve the pressure difference at about 14000 seconds. At about 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months the pressure in the drywell is too low to spray the drywell, (unsafe area of DWSIL (EOP-3 -Primary Containment Control -Reference 30)) however the pressure increase to about 6 psig at about 10800 seconds at which point the operators would be able to spray the drywell with the RHR pump, if needed.

Figure 15 presents the RPV pressure. At one hour into the event it is assumed that the operators start depressurization. The pressure drops to the HPCI shutoff pressure of 165 psia at about 12000 seconds. At this point only about 450000 lb were injected from CST (Figure 17). The RPV is depressurized, and the CS pump is available to inject.

Figure 16 pump presents the RPV level. The core stays covered. There is. a dip in the normal level at about 12000 seconds when HPCI stops injecting and CS does not inject yet. This is due to the fact that the CS pump was set to inject at 14000 seconds; however CS is ready to inject at 7800 seconds, provided the pressure permissive is reached.

Figure 18, Figure 19, and Figure 20 presents the drywell liner temperature. The drywell liner stays below 260 'F for the 7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months analyzed. After 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes the low pressure pumps are available for suppression pool cooling, drywell spray and maintaining vessel inventory.

The suppression pool temperature for this case is very similar to the case SBO-drywell2 since the input changes results in minor changes to the drywell temperature and pressure but not in the suppression pool temperature since the heat transferred to the drywell is not subtracted from the vessel energy.

. . . . . .-. - . - .. . . . - : . _. .. "7 -..- - . . ..

'Enfergy g f.

M..

Calculation VYC-2405 Rev. O Page 38 of 85 <

5Do - 80F-Noleakw-dryell2-sensitivitieS-Set2 Mar/10/2005 10:36:15.

GOTC Version 7.Op2 (OA) - April 2002 Pile: I/hoslcr/vyc-212OcCa/SENSITlIVrY/SHO/drywel.-SDO/SBO-dryvell2-80 Ss . Hmyv.U T eratcre

s sI CS - - -

g 9g . I

.I Sise (ste)

AvMN F.tfa "W110.215 MMU 5 Figure 13 - Drywell Temperature -Case SBO-drwell2-80-sensy2-NoLeak LO -- BOF-Nol eak-drwell2.B-senitivitieS-set2 K~ar/10/200S 10:56:11

GOWIC Version 7.Op2 (0A) -'Apri 2002 .

File: /be/schor/vyc-2120ccn/SmNSrIVTY/SBO/dXewll-SDO/SBO-dryweli2- 8B 3 canof-t MISs=

a a

-4

- -- -L rlAC (sac) 6=k1.%tevl~a W.AfI9flee% 41,41,32 Figure 14 -Containment Pressure - Case SBO-drwell2-80-sensy2-NoLeak

~Entef Calculation'WC-2405 Rev;0 ' ' '-O Page 39 of'85' SBO - 801-zoleak-drywel2-:eiaitivitics-set2 l4ar/10/2005 11:11:35 GOTHIC Version 7.Op2CQA) - Jpril 2002 File: /eo/achor/vyc-2120OcC//SDSZITI T/SBO/drywell-SBO/SBO-drylIell2-O 24 Rfl ?zUStrt%

124. ;: Tress-rts V i-

-4 e _

. -IS sL* (si.)

a *,2¢#^>XZ~st..R.Xa Figure 15- RPV Pressure - Case SBO-drwel12-80-sensy2NbLeak SBO - 3OF-Noleak-dx.e112-sensitivities -set2 Mar/10/2005 11:15:46 GOTHIC verstio 7.Op2(QA) - April 2002 File: /bhoe/schor/vyc-212 cci/SZNSrTVIT'/SDo/drywell-SBO/SEO-dryvel2-80 Zi - _ _ __

2ROa Li-mid . .:

L.w1 .

?.L4 t

. .A IV (see) 9- - f ,9t Figure 16 - RPV Level - Case SBO-drwell2-80-sensy2-NoLeak

. 7ntergy - Calculation VYC-2405 Rev. 0 *Page4O6 o 85 -

SsO - s0P-2oleak-dryvell2-sensitivities-set2 Mar/10/2005 11:10:47 GOHIC Versioun 7.Op2(QA) .- April 2002 File: /home/schor/vyc-2120ccn/SRNSITIVITY/SBO/drywell-SBO/SBO-drywell2-80 40 =tgat 1 sv,.i

.1 1 1 1 14

t~~~~~~~ I 2tX"tt25l.tS

(St;)

Figure 17 Integrated HPCI Flow - Case SBO-drwell2-80-sensy2-NoLeak SBO - 80P-2oleak-4ryvell2-senuitiiities-set2 Mar/10/200S 10:33,22 GOMIC Version:7.0p2CQA) - 4pril 2002 Pile: /hcae/schor/vyc-2120ccn/SE4SITIVITY/sBO/drywellS-O/SOW-dxyuel12-8O

. . .4 e 0s .8 n7.2 30. 3.4 2.

' "Ib 1_

CJU (sea I

- - ------ 1-I.-

Figure 18- Surface Temperature for Heat Slabs 5,6 & 7 - Case SBO-drwell2 sensy2-NoLeak

.. .. I 'Entey .. Calculation VYC-2405 ReV. 0 Page 41 of'85 '

SBO - 807-Noleak-dxywefl2-sensitivities-set2 Mar/10/2005 10:34:30 GOTIC Version 7.0p2 (QA) - April 2002 File: /tome/schor/vyc-212oc=/SAS ITXVfl/SBO/drywell-SBO/SBO-dxywell2-80 51 Surface Temperature ra5s D110 ral g"Ye Ye th P F S.E ?.2 ;~ARE (so.4

) lX;[

Figure 19 - Surface Temperature for Heat Slabs 8, 9, 10 &11 - Case SBO-drwell2-80-sensy2-NoLeak SBO - 8OP-Noleak-drywell2-sensitivities-set2 Wxar101/2005 10 :32 :24 GOMXC Version 7.0p2(QA)I- April 2002 Pile: /home/schor/vyc-2120ccn/SNSIZTr TY/SBO/drywell-SBO/SBO-crywell2-80 2 Surface tbeperatur 1113 D112 I..

3 7..

d**3 7.,lco"I " 6l1241% 01M.92.

Figure 20 Surface Temperature for Heat Slabs 12 &13 - Case SBO-drwell2 sensy2-NoLeak !

-.-I -- . .... : . , , - .;-: ". .. .

.i X.. :_ . . . .

. -Entergy -' CalculatioiVYC-2405 ReV. 0 Page 42 of 85 Pe4o8 I.

s.- i.'V .

63 Case SBO-drvwell2-Leak-80-sensy 6.3.1 Model modification Table 9 presents the modifications to the deck SBO-drywell2-80-sensy2-NoLeak to produce SBO-drywell2-Leak-80-sensy The following modifications were made:

On BC 13, the ON trip is set to zero (0). This allows for a constant leak of 8.4585 lb/sec to leave the vessel.

The end time was changes to 14400 seconds (4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months ) since the purpose of this calculation was to show that the drywell temperature stays below the EQ drywell temperature and the drywell shell stays below 281 0F for the duration of 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and I0Oninutes.

Table 9 SBO-drywell2-Leak-80-sensy vs SBO-drywell2-80-sensy2-NoLeak

. - __ _ . -- -- -- ;. 11- , -:-.

... -FEntergy Calculation VYC-2405 Rev. 0 Page 43 of 85 I Modifications In /hane/schorlvyc-212OccnISENSITIVITY/SBO/drywell-SBO/SBO-drywell2-Leak-8 Mar1l412005. 15:38:40 GOTHIC Version 7.Op2(QA) - April 2002 Fl1e: /home/schor/vyc-2120ccn/SENSITIVITY/SBO/drywell-S8O/SBO-dr-yweIl2-Leak-80-sensy P.

Fluid. Boundary Conditions - Table 1 Press. Temp. Flow ON OFF BC# Desdript1on (psia) FF (F) FF (lIb/s) FF Trip Trip IF RHR/U'CI Suctio 20. 160 v-0.002 8 1 13 2C RHR/LPCI Dlscha 20, 160 3F LPCS Suction 20 160 v-0.002 7 30 31 4C LPCS Discharge 20. 160 5F Feedter 1000. el 5 1000 9 1 5 6F RWrTorus Sucti 20. 160 v-0.002 6 21 7C RHRITorus Disch 20. 160 0 1*

8F HPCI/RCIC Sucti 20. 160 -326.1 9 1 0 9C HPCI/RCIC Disch 20. 160 10P Cooldan Inlet 1. 10 1 11 21 .0 11F Cooldown Outlet 1. 10 111 112 12F CST Tank 14.7 135 587 13 28 27 13F Vessel Leak 1050. 554. -8.4585 13 0 14C Vessel leak to. 1050. 0 554 Run Control Parameters (Seconds)

Time DT DT (IT End Print Graph Max DunP Phs Chog Int ,Mn Max Ratio Time Int Int CPU Dnt Time Scale le-06 i. 1. 100; 5. 0.1 le+06 0 DEFAULT I 2 le-06 1. 1. 1200. 50. 1. le+06 0. DEFAULT' 3 le-06 .1. 1. 1300. 500. 10. 1e+06 0. DEFAULT 4 le-06 1. 1. 14400. 600. 10. 1e+06 0. DEFAULT Graphs Graph Cur2e Nur3er Title Mon 1 2 3 4 5 I Drywell.Tepera TV1 TL1 2 Webtell Tempera NV2 112 3 ,Contairert Pre PR1 PR2 II 4 Reactor Vessel TV4 TL4 ST4 TD4 5 RHR Heat Exchan xqlH 6 Reactor Vessel AL4 7 Torus Water Vol AL2 8 Heat Exchanger tIlH t21H 9 Wet'll. Vessel T12 TL4 TL1 10 Conductor TeTpe TA1 TA2 TA3 TM 11 Integral Vessel QL4 QWI 12 Vapor & Conduct 1V4 13 Liquid & Conduc TL4 TA2 14 Vapor & Conduct TV2 TA3 15 Liquid & Conduc 112 TA4 16 Vapor Heat Tran HUA 17 Liquid Heat Tra hA2 18 Vapor Heat-Tran HA3 19 Liquid Heat Tra hM 20 Feed6ter & ere FL9 FL4

'Enter gy. Calculation VYC-2405 Rev. 0 Page 44 of85 Modifications in /hcine/schor/vyc-212Occn/SENSITIVITY/SBO/drywell-SBOSBO-drywel12-Leak-8 Mar/14/2005 15:38:40 GOTHIC Version 7.0p2(QA) - April 2002 File: /home/schor/vyc-2120ccn/SENSITIVITY/SBO/drywell-SBO/SBO-drywell2-Leak-8O-sensy Graphs (continued)

Graph Curve Number Title Mon 1 2 . 3 . 4 5 21 RPV lnulid Leve LL4 22 SRVand ADS Flo FY10 FV11 23 Feed ateir Entha cv29 24 RPV Pres! ;sure5 YR4 25 FeedwateirContr o27 cv28 26 Integrati ;d. Feed cv4 ZJ RHR Flow FL5 28 Vessel Di AO4 29 ECCS njlection FL7 FL12 FLi4 30 RPV Pres.- sure PR4 VC2V FL7 31 ADS ValYWr Posit 32 SRV Posit Etion VC2V 33 Cooldown "FLOw, FV16 34 VesselDIrop Dia D14 35 Reactor. Vessel PR4 36 Suppress1[on Poo TL2 37 Reactor -Vessel PR4 38 Suppressl Ion Poo LL2 39 HPCI F1C4 xdRate FL18 40 Integrati :d HPCI cv39 41 Core .Spri y Flow FL8 42 Heat to.l the sup *CQ4H CQ2H .

43 Leak F1o4 FL4 FL19 FL20 44 .[ntegratj ;ed Leak cv4O 45 Title cv39 46 FY18 FL18 FDi8 47 cv38 48 Surface Tempera TB5 TB6 T7i 49 Surface 7Temp~era TA8 TA9 TAIO 50 Surface'j [eTpera TAll TAI2 TA13 51 Surface IFempera TB8 T89 TB9T10 111l 52 Surface 1tepera 1B13 TB12 B. ..

63 . TP8t600 TP9t600 TP10t60 TP11t60 54 TP13t5oTP12t5O X)OD S XC mxx==o~ x~oooo= Xooooc XXXoooc 300oo

3Entergy Calculation VYC-2405 Rev. 0 Page 45 of 85 63.2 Results Case SBO-drvwell2-Leak-80-sensy Figure 21 through Figure 29 present the main parameters for the case SBO-drywell2-Leak sensy. Figuqre 21 presents the drywell temperature. The maximum drywell temperature is about 290 F and is reached at about 12240 seconds (3.4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months ). The drywell heatup rate is arrested due to depressurization, however the leak brings enough energy from the vessel to continue the heatup. At 7800 seconds the drywell temperature is 285.8 TF, well below the EQ limit of 325 OF.

Figure 22 presents the containment pressure. The available water to spray the drywell (Reference

30) is the Diesel fire pump per Appendix M of OE 3107 (Reference 35) and it takes about one hour for aligning the fire pump for drywell spray. The drywell pressure is high enough to allow for drywell spray, if needed. The drywell temperature does not exceed the EQ drywell temperature limit and the drywell shell temperature stays below the limit of 281 TF hence the analysis shows that drywell spray is not needed for the coping duration. At about 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months and 30 minutes the wetwell pressure reaches equilibrium with drywell and slightly exceeds the drywell pressure. The vacuum breakers do not open during the time of interest..

At about 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months the wetwell pressure is about 26 psig, close -to the PSP limit of 27 psig.

However at this time the RHR pump is available for containment spray.

Figure 23 presents the RPV pressure. .At one hour into the event it is assumed that the operators start depressurization. The pressure drops to the HPCI shutoff pressure of 165 psia at about 12000 seconds. At that point only about 540000 lb were injected from CST (Figure 25). At this time the RPV is depressurized, and the CS pump is available to inject.

Figure 24 pump presents the RPV level. The core stays covered. There is a dip in the normal '

level at about 12000 seconds when HPCI stops injecting and CS does not inject yet. This is due to the fact that the CS pump was set to inject at 14000 seconds; however, CS is ready to inject at 7800 seconds, provided the pressure permissive is reached.

Figure 27, Figure 28, and Figure 29 presents the drywell liner temperature. The drywell liner stays below 280 F for the 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months analyzed. After 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes the low pressure pumps are available for suppression pool cooling, drywell spray and maintaining vessel inventory.

The suppression pool temperature is not a parameter of importance for this calculation. In Reference I it was shown that the suppression pool temperature is lower for-the cases RPV with leakage and lower for earlier depressurization hence the maximum suppression pool temperature will be lower that 182.2 A, calculated in Reference 1.

-. .., Entergy Calculation WC-2405 Rev. 0 Page 46 of 85' SBO - 8OF-Noleak-drywell2-Leak-80-senlaleivitles Mar/14/2005 11:02:51 GOTBIC Version 7.Dp2I(A) - April 2002 Pile: /1am2e/scr/Vyc-2l2OcCnISSITIVI ./SEO/drywell-SDO/SDO-dryweli2-Le mxpwen Uverature

?m nL1 I'

A-sh~e (see)

.-. . t1-f.A.S -. t a

.. n..t .,i.,,..t **,,*, gmg see Figure 21 -Drywell Temperatures - SBO-drywell2-Leak-80-sensy SBO - 80P-Noleak-dryvell2-Leak-80-senitievites Mar/14/2005 11:03:43 GOThIC Version 7. Dp2 (QA) - April 2002 File: Jboe/schor/vyc-2120ccn/S1TrflVTr/SBO/dryvell-SBO/SBO-d8Zywel2-Le I centaimtnt M~Sfurt "a V2

-4

'4 IM 96 sJa (see) darm 11.4201 "WISMI S tU.211_

Figure 22 - Containment Pressure - SBO-drywell2-Leak-80-sensy I

E..

Enters CalculationYVC-2405 Rev.'0 Page 47 6P85' SBO - 60P-Noleak-drywell2-Leak-80-sensitivities Mar/14/2005 11:05:31 GcTHIC Version 7.0p2(QA) - April 2002 File: /lome/schor/vyc-2120ccn/S tsITIrvITY/SBO/drywell-SBO/SBO-dxywell2-Le

24. )~ ??t~ax.

- t z

.99

I.

tare QStc

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

Figure 23 - RPV Pressure - SBO-dry*0eI2-Leak-80-sensy

.SO. -. 80F-Noleak-dxyvell2-Leak-80-sensitivities bar/14/2005 11:04:51 GOTHIC Version 7.OP21QA) - Jpril 2002 Pile: /home/echor/vyc-2120cu/rSBSITIV1TY/SBO/drywell-SBO/SBO-drywell2-Le it Liauid LrVel

=4 v\ _ \1 Ps (sr)

I I Figure 24 RPV Level - SBO-drywell2-Leak-80-sensy

~Entergy Calculation VYC-2405 ReiV. 0 "

Page 48 of 85 '

SBO - SOP-Noleak-dxywell2-Leak-so-sensitivities Mar/14/2005 11:09:57 GOTHIC Version 7.0p2(QA) - April 2002 File: /hbe/acbor/vyc-2120cyc =/SE2oSiiIv/SBO/drywell-SDO/SBo-dxywe112-Le 40 ntsgrattd. XP= 71w

'icfls i

O . .. . .

.1 f -.. ..

a w.Itsw ne5ttg I . )

Tim ({see I

e 25 - Integrated HPCI Flow - SBO-drywell2-Leak-80-sensy I SBo - aop-Noleak-dryve112-Lea-80-sensitivities Mar/14/2005 11t20l:57 GOTHIC Version 7.0p2(QA) - April 2002 rile: /home/scbor/vyF-2120cc/SrN5ITZVTY/SBO/drywell-SBO/Sso-dxrwe112-Le II I

i 43teak Flo Mg *L20

0~ -

/.

IA Som .42 .fV3120 .os$

i I

I I

71ke (set)

Figure 26-Leak Flow - SBO-drywell2-Leak-80-sensy i

~'Entergy Calculation VYC-2405 Rev:O Page49 6f55 SWO - SOP-N leak-drywell2-Leak-S0-sensitivities Mar/14/2005 11:07:22 GOTHIC Version 7.0p2(QA) - April 2002 File: /hboe/schor/vyc-2120ccn/SRSZIT2VITY/SBO/drywel1-SBO/SDO-dxywell2-Le 4t rfaet 7impraturt T'sS TC n7l Time (see)

$&I _. _ _

Figure 27 - Surface Temperature, Heat Structures 5,6,7 - SBO-drywell2-Leak sensy -

SBO - 80o-Noieak-dxywel12-Leak-6O-seuasitiities Mar/14/2005 11:06:34i GOTRIC Version 7.0p2(OA) - April 2002 Pile: /hce/schor/vyc-2120n1 /SXNS8IwI/STY/SO/dryell-SDO/SEO-drywell2-Le TIC Tss TlhU 511 I. C ft X T@t .c Figure 28 - Surface Temperature, Heat Structures 8, 9, 10, 11 - SBO-drywelJ2-Leak-80-sensy

.W.'-.

... Entergy Calculation"WC-2405 ReV. 0 Nage 50 of 85 A. '

SO - B0F-Noleak-dryve112-Leak-8O-sensitivitiea Mar/14/2005 11:09:22 GOTEZC Veruion 7.Cp2(QA) - April 2002 Pile: /home/schr/vyc-212Occn/SENSITIV+/-tW/SBO/drywell-SBO/SBO-drywell2-Le surf aoe ttPeratme

C sw i1 I.-

Co :

-Yl 4. 19 1.1 TIAe (see)

Figure 29 - Surface Temperature, Heat Structures 12,13 - SBO-drywell2-Leak sensy

.. Entergy CalculationVYC-2405Rev. 0 Page5 of 85 :

6.4 Case SBO-drvwell2-Leak-45-sensv 6.4.1 Model modification Table 10 presents the modifications to the deck SBO-drywell2-Leak-80-sensy to produce SBO-drywell2-Leak45-sensy.

Two modifications are made, the depressurization (cooldown) table, is changed from 80 TF/hour to 45 0F/hr (same cooldown curve as in Reference 1- Function 10).

The RHRSW temperature is changed from 85 TF to 75 "F consistent with Assumption 13 and Reference 1.

Table 10 SBO-drywell2-Leak-45-sensy vs SBO-dryweWl2-Leak-80-sensy

.Entergy X.. Calculation'WC-2405 Rev. 0 - Page 52 of 85 Modifications in /hcmelschor/vyc-2120ccn/SENSITIVITY/S80/drywell-SBO/SBO-drywell2-Leak-4 Mar/14/2005 17:49:07 GOTHIC Version 7.Op2(QA) - April 2002 File: /hone/schor/vyc-2I20ccn/SENSITIVlTY/SBO/drywell-SBO/SBO-drywell2-JLeak-45-sensy

' Functions FF# Description Ind. Var. Dep. Var. Points.

0 Constant-- 0 1 RHR K Tube Reynolds N Nusselt Nu 34 2 RHR Hx Shell Reynolds N Nusselt Nu 34.

3 Decay Heat Time (sec) Decay Heat 50 4 Pulp Heat Time (sec) Heat Rate 6 5 Feed Enthalpy cv4 Dep. Var. 34 6 RHRJTorus Timen(sec) Flow (gpm) 3.

7 LPCS Flow Curve 6 Flow (gpm) 13 8 LPCI Flow Curve cv26 Flow (gpm) 12 9 Feed Flow . cv28 Dep. Var. 3 10 Cooldown Pressu Time (sec) Pressure ( 39 11 Cooldown Temper cv33 Dep. Var. 3.

12 Cooldow Flow cv32 Dep. Var. 3.

13 Constant Ind. Var. Dep. Var. 6 14 'ECCS Pump Heat Time (sec) Heat Rate 6 15 Check Valve Ind. Var. Dep. Var. 6 16 SW. Time (sec) Service Wa 6 17 Dry ell Power C cv41 Dep. Var. 4.

18 Drywell Power C cv42 Dep. Var. 4 Heat Exchangers - Table 2 .

Heat Scndy Scnd Scndy Scnd Ext. Ext. Ext. Ext.

Ex. Flow Flow T amp TBmv Flow Flow Heat Heat (lbm/s) FF (F) FF Clbm/s) FF (Btuls) FF

.11H *1. 16 75.

Graphs Graph Curve Nuzter

Title Mon 1 2 3 4 5 1 Drywell Tenpera TV1 TL1 2 Wetwell Tempera TV2 TL2 3 Containment Pre PRI PR2 4 Reactor Vessel TV4 TL4 ST4 TD4 5 RHR Heat Exchan xqlH 6 Reactor Vessel AL4 7 Torus Water Vol AL2 8 Heat Exchanger tilH t21H 9 Wetwell. Vessel TL2 TL4 TL1 10 Conductor Tempe, TA1 TA2 TA3 TAM 11 Integral Vessel QW4 QV4 12 Vapor & Conduct T4 TAl 13 Liquid & Conduc TL4 TA2 14 Vapor & Conduct TV2 TA3 15 Liquid & Conduc TL2 TM 16 Vapor Heat Tran HW 17 Liquid Heat Tra hA2 18 Vapor Heat Tran HA3

._ .. Entergy. Calculation VYC-2405 Rev. D ' Page 53 of 85 `

Modifications in /hone/schorfvyc-2120ccn/SENSITIVITYJSBO/drywel 1-SO/SBO-dretl 12-Leak-4 Mar/14/2005 17:49:07 GOTHIC Version 7.Op2(QA) - April 2002 Fl 1e: /hcme/schor/vyc-2120ccn/SENSITIVITY/SBO/drwel 1-SBO/SBO-drwel 12-Leak-45-sensy Graphs (continued)

Graph Curve Number

. Title Mon. 1 2 3 4 5 19 Llqd Heat Tra hA4 20 Feeciiater & Bre FL9 'FL4 21 RPY LIcuid Leve .LL4 22 SRY and ADS Flo FVI FV11 23 Feedaater Entha cv29 24 RPV Pressures PR4 25 Feedater Corrtr cv27 cv28 26 Integrated Feed .cv4 27 RHR-Flcw . . FL5 28 Vessel Droplet AD4 29 ECCS InJection . FL7. FLU2 FL14 30 RPV Pressure PR4 VC2V FL7.

.31 ADS Valve Posit VC3V 32 *SRV Position VC2V 33 -Cooldcw FLow : FV16 34 Vessel Drop Dia DI4 35 Reactor Vessel.. PR4 36 Suppression poo TL2 37 Reactor Vessel PR4 38 SuDpression.P0 LL2 39 HPCI .FlwRate FL1B 40 Integrated HPCI cv39 41 Core.SprayFlcw FL8 42 Heat to te -sup .CQ CQ2H 43 Leak Flow - FL4 FU9. FL20 44 Integrated Leak cv40 45 Title CV39 46 FV18 FL18 FD18 47 cv38

48 Surface Teipera TB5 TB6 .187 49 Surface Ternpera TAB TA9 TA10 50 SurfaceTToipera TA11 TA12 TA13 51 Surface Tempera TB8 TB9 1810 .1B1 52 Surface tenpera 1T13 TB12 53 . TP8t600 TP9t600a .P1t60 TPllt60 54 TP13t50 TP12t50 55

.'n&tergy

'Calculation"WC-2405 Rev.-; O ' ' Page54f'85 '

Modi fi cati ons in /hanie/schorlvyc-2120ccn/SENSITIVITY/SB0/drywel l -SBO/SBO-dryel l 2-Leak-4 Mar/14/2005 17:49:07 GOTHIC Version 7;0p2(QA) - April 2002 File: /home/schor/vyc-2120ccn/SENSITIVYISBO/drYwel l -SBO/SBO-drywel l 2-Leak-45-sensy Function 10 Cooldown Pressure Ind. Var.: Tume (sec)

Dep. Var.: Pressure (psia)

Ind. Var.- Dep. Var. Ind. Var.. Dep. Var.

0. 1078.5 600. 1014.5 12008 953.5 1800. 895.3 2400. 839.8 3000. 787.

3600. 736.8 4200. 689.1 4800. 643.7 5400; 600.7 6000.- .559.9. 6600. 521.3 7200. 484.8 7800. 450.2 8400. 41L6 9000. 386.8 9600. .357.8 10200. *330.5 10800. .304.9 11400. 280.8 12000. 258.2 12600. .237.1 13200. ,217.,3 13800. .1.98.9 14400. -181.7 15000. 165.6 15600: 150.8 16200. 136.9 16800. 124'-2 17400. 112.3 18000. 101.4 91.3 19200. 82.1 VW88' -73.6 20400. 65.8 21000. 58.7 21600. 52.3 222D0. 46.4 1000000. 46.4

wEntegy Calculation VYC-2405 Rev. 0 Page 55 of 85 6.4.2 Results Case SBO-drWell2-Leak-45-sensy Figure 30 through Figure 38 present the main parameters for the case SBO-drywell2-Leak45-sensy. Figure 30 presents the drywell temperature. The maximum drywell temperature is about 293 TF and is reached at the end of the run (4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months ). The run was not extended beyond the 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months even though the drywell temperature continues to increase because at 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months the RHR pump is available for drywell spray, if needed. The mission time of 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes is.

achieved. -The drywell heatup rate is arrested due to depressurization; however the leak brings enough energy from the vessel to continue the heatup. At 7800 seconds the drywell temperature is about 290 0F, well below the EQ limit of 325 °'F.

Figure 31 presents the containment pressure. The drywell pressure is high enough to allow for drywell spray after one hour into the transient, if needed. The available water to spray is thee Diesel fire pump (Reference 30) per Appendix M of OE 3107 (Reference 35) and it takes about one hour for aligning the fire pump for drywell spray. The drywell temperature does not exceed the EQ drywell temperature limit and the drywell shell temperature stays below the limit of 281'

'F for the mission time of 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes hence spray from Diesel fire pump is not needed.

The vacuum breakers do not open during the time of interest.

At 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months the torus pressure is -about 25 psig and increasing, close to the PSP limit of 27 psig.

However at this time the RHR pump is available for containment spray.

Figure 32 presents the RPV pressure. At one hour into the event it is assumed that the operators start depressurizafion. The vessel pressure during the 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months of the run time does not reach the shutoff pressure for the HPCI pumps, so at 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months the iPCI pumps still inject to maintain inventory. At 4 hour4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months into-the event 6nly about 540000 lb were injected from CST (Figure 34), At this.time the RPV is not depressurized, and the HPCI pump continues to inject.

Figure 33 presents the RPV level. The core stays covered and HPCI maintains inventory for the duration of the analyses. There is no need to continue the calculation beyond 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months because the coping time of 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months was demonstrated.

Figure.35 shows that the leak is maintain constant for the duration of the transient.

Figure 36, Figure 37, and Figure 38 presents the drywell liner temperature. The drywell liner stays below 280 'F for the 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months analyzed. After 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months .and 10 minutes the low pressure pumps are available for suppression pool cooling, drywell spray and maintaining vessel inventory.

The suppression pool temperature is not a parameter of importance for this calculation. In Reference I it was shown that the suppression pool temperature is lower for the cases with leak and lower for earlier depressurization hence the maximum suppression pool temperature will be lower that 182.2 TF; calculated in Reference 1.

'-~Entergy. Calculation VYC-2405 Rev. 0 . Page 56 of 85 .

680 - dzyvell2 -Leak-45 -sefsltiv'ities ZMar/09/2005 1.8:54:07 GOTHIC Version 7..Op2(QA) - April 2002 P. File: /home/scbor/vyc-2120ccn/SENSITIVI7Y/SBO/dyell.6BO/SBdryvel12.Le Tlat (see)

Figure 30 -Drywell Temperature -Case SBO-drywell2-Leak-45-sensy

SBO -rell2-Leiak-45sensitivities l4ar/14/2005 17:59:16

&GOMICVersion 7.Op2(QA) - April 2002 File: /bohe/achor/vyc-2120ccn/SENSITIVITY/SEO/drywell-SBO/SBO-dryvell2-Le

. I .

.7a YR 7iat (see)

Figr

2tai) n 1- t tO e rae2$5 -2 a 4s4 Figure 31 -Containment Pressure - Case SBO-dryweW1-Leak-45-sensy

.'Entergy . .. .Calculation VYC-2405 Rev.O Page 57 of 85 SBO - drywe2I2-Leak-45-sensitivities Mar/09/2005 18:59:19' GOTHIC Version 7.0p2(QA) - April 2002 File: /hoine/Bcbor/vye-2220ce/SENSITIVIT/SBO/drywell -SO/SBO-drywe 2 -Le I24 M lrtssaxes I .

t t 1 1 I! I 7!2 (t) s (.e Figure 32 - RPV Pressure -Case SBO-drywell2-Leak-45-sensy SBO - 'drywell2-Leak-45-sensitivities Mar/09/2005 19:05:53 GOTHIC Version 7.0p2 (QA) - April 2002 Pile: /home/ecbor/vyc-2120c=/SENSITIVIrT/SBO/drywell-SEO/SBO-drywell2-Le 21 1 1 3 ; ___. N N _ _

Oo63 c '72 e*a .. s

Time (sea) 4riS 1.hp22#*) wfIt/as*e ii Figure 33 - RPV Level - Case SBO-drywell2-Leak-45-sensy

Entergy Calculation VYC-2405 Rev. 0 Page 58 of 85'M SBO - drywel12-Leakz45-senaitivities Mar/09/2005 19:14:51 GOaMC Version 7.Op2(QA) - April 2002 Pile: /hce/scbhor/vyc-2l2occn/sisrmrvxTY/SBO/drywell-SBO/SBO-drywel12- Le Figure 34 - Integrated HPCI Flow - Case SBO-drywell2-Leak-45-sensy S30 - drywell2-Leak-45-sensiti'itdies Mar/09/2005 19:16:54 GOTHIC Version 7.0p2(QA) - April 2002 File: Aome/schor/vyc-2120ccn/SHNSITIVITY/SB0/dryweel-SBO/SBO-drywell2-Le 43 taek o1w n4 711S rL20

- C

-4 m1

.U 1. 6 3.2i 4.8 6.4 8S s.S 1 . 2 .;2.8i 4

, .Tsle

- (sic)

Figure 35 - Leak Flow -Case SBO-drywell2-Leak-45-sensy

Entergy Calculation VYC-2405 Rev. -0 Page 59 of 85 0 - dryvell2-Leak-45-sensitivities Mar/09/2005 19:06:59 GOTUIC Version 7.0p2 QA) - April 2002 Pile: /home/schor/vyc-2120ccn/SBNSITIr~m/SBO/dryvell-8BO/SBO-drywell2-Le P.

4S Tg, STtz a--- _ ____ __ _ _ _ _ _

I..

r C. (see)

,' S (se;)

Figure 36 - Surface Temperature, Heat Structures 5,6,7 - SBO-drywell2-Leak sensy

' SBO - drywell2-Leik-45-nensitivities Mar/09/2005 19:08i37 GOTHIC Version 7 .0p 2 (QA) - April 2002 File: /hboce/schor/vyc-2l20c=n/SENSITIVTfr/SBO/d.ywell-SB0/SBO-dxywell2-Le o'xe DSrJIO ?mp Di

,S..

C

so 0tt~fs wltl St.

L..LLLI Figure 37 Surface Temperature, Heat Structures 8,9,10,11 - SBO-drywell2-Leak sensy

WM,

.'Entergy Calculation VYC-24051 Rev. 0 Page 60 of 85

.. Y BEEF ,.

t -t*.

.1 .. :

SBO - dqrywell2-Leak-45-sensitivitieG Mar/09/2005 19:11:15 GWtHIC Version 7.Cp2 QA) - April 2002 File: /hbce/schor/vyc-2L20ccn/SMBSITIVnIY/SBO/drywell-SBO/SBO-dryvell2-Le s2 Surface tmperatuw T01 Sa3,12,

'S 2- !2l1I

ret i

lt~

.4

~~~J "en ssee)

-4.SO-IS,12,.

(.2" Figure 38 - Surface Temperature, Heat Structures 12,13 - SBO-diyweII2-Leak sensy

~Entergy. Calculation VYC-2405 Rev: 0 Page 61 of 85 6.5 Case SBO-drywell-comments 6.5.1 Model Modifications This case addresses the reviewer comments and also some discrepancies found during the documentation. The following changes are being made:

-change the initial temperature for Heat Structures 14 from 160 TF to 170 'F.

-change the K,,vere injunction 3 to 3.93 from 3.964.

-change the flow area of the valve V3 to 15.63 ft2, same as the flow path flow area

-change the surface area of the concrete pedestal to 2068 ft2 The changes are made to case 2 but it could be done to any of the other cases.

Table 11 presents the modifications made to file SBO-drywell2-80-sensy2-NoLeak to create SBO-drywell2-comnments.

Table 11 Comparison between SBO-drywell-commints vs SBO-drywell2-80-sensy2-NoLeak.

-ME1nteW Calculation VYC-2405 RbV. 0 O

Page 62 of 85 -t Modifications in /home/schor/vyc-212kccn/SENStTIVITY/SBo/drowl I -SBO/SBo-drywell -caunent Mar/15/2005 14:07:44 GOTHIC Version 7.0p2(QA) - April 2002 File: /hane/schor/vyc-2120ccn/SENSITIYITY/SBO/drywell -SBO/SBO-drywel l -ccwrments Fl1o Paths - Table 3 Fltow Fwd. Rev. Critical Exit Drop Path Loss Loss Comp. Flow. Loss Breakup Coeff. Coeff. Opt. Model Coeff. Model 1 4.2?43 4.2243 'ON. TABLES 1. OFF 2 1. 0.78 ON TABLES 1. OFF 3 3.964 *3.93 ON OFF 0 OFF 4 0.

OFF TABLES 1. OFF 5 OFF OFF 0. OFF

6. OFF OFF 0 OFf 7 OFF OFF 0. OFF 8 OFF OFF 0. OFF

. 9 OFF -ALOFF 0. OFF 10 OFF TABLES 0. OFF 11' OFF TABLES 0. - OFF 12 OFF *OFF 0. OFF 13 OFF OFF . 0. OFF 14 OFF OFF

0. OFF 15 OFF .OFF 0. OFF 16 *le+18 OFF *OFF 0. OFF 17 OFF OFF . 0. OFF 18 OFF OFF 0. OFF 19 OFF OFF 0. OFF 20 OFF OFF 0. OFF 21 1.5 1.5 OFF -OFF 0. OFF Thermal Conductors - Table 1 Cond Vol HT Vol HT Cond S. A. Init.

- Description A Co B Co Type (ft2) T.(F) Or 1 Steamn Eposure 4 1 4 I 1 2965.72 609.23 I 2 LiquId Exposure 4 2 A4- 2 1 11521L8 647.4 I.

3 Torus, Vapor 2 3 2 5 2 13553.7 90. 1 4 Torus, Liquid 2 4 .2 5 2 13553.7 90. *1 5 Loer Drwell 1 5 1 6 3 1856.24. 170. I 6 Lower Orywell 1 5 1 6 4 2041;28 170. I 7 -Middle Dryell 1 151 5 .1 *6 5 3802.73 170 '.I 8 Middle Drywell 1 5 1* 6 *6 780.68 170. I 9 Middle Drll 1 5 1 6 5 1250.47 170.' I 10 Middle dry l1 1 5 1 6 7 1898.24 170. I 11 Middle Dr 1 1 5 1 6 8 1114.72 170.. I 12 Top orywell 151I 6 13 783.45 170. 1 13 Too Drywell 15 1 '7 14. 1718.3 170. 1 14 RRUs 1 51 6 11 1272.8 170. I 15 Vent Pipes 3 53 6 11 2885.7 160. I 16 Concrete Shield 1 51 .6 12 2068. 152. 1 Modifications in 8hanelschor/vyc-2120ccn/SENSITIVITY/SBO/drywell Mar/15/2005 14:07:44

-S80ISBO-drNel 1-carment GOTHIC Version 7.p2(QA) - April 2002 File: /hanefschor vyc-2120ccn/SENSITIVITY/SBO/drywl 1-S8OSBO-drywel l-c1n.ets Valve/Door Types Valve Sten Loss Flw Type Valve Travel Coeff. Area r Option Curve Curve (ft2) 1 QUICK OPEN 0 0 . 1.

2 CQECK VALVE 0 15 3.141 3 QWICK CLOSE 0 0 15.63

AdEntergy Calculation VYC-2405 Rev. 0 Page 63 of 85' 6.5.2 Case SBO-drvwell-comments Results

-The results of this case are presented in Figure 39 through Figure 40. Figure 39 presents the drywell temperature. Due to the fact that the initial temperature for the drywell thermal conductors increase by 10 TF, the drywell temperature is increased from 289.4 'F to 295.2 OF.

The containment pressure (Figure 40) is identical to the Case 2, hence the changes in the vacuum breaker inputs have no effect on results, as described in section 6.2.1.

Figure 41, Figure 42, Figure 44 presents the drywell liner temperature. The drywell liner stays below 260 TF for the 7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months analyzed. After 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes the low pressure pumps are available for suppression pool cooling, drywell spray and maintaining vessel inventory.

SW - 80F-Noleaik-drz~eil2-slSkitiVities-COcentsl Mar/15/2005 14:03f50 G0a3IC Version 7.0p2 (QA) -rell 2002 rile: /han/schor/vyc-2120ccn/SNSirlVITY/SW/drywell-SBO/SBO-rwll.-com S 3

]Dryreil Texperature .

3_ __ -

14 'CD D

_ TnwL (sac)

Figure 39- Drywell Temperature - Case SBO-drywell-comments

R. e.

Entery Calculatio'n"WC-2405 Rev. 0 ' ' Page'64 of'85' t SBO - 8OP-Noleak-drywell2-sensitivities-corcents Mar/15/2005 14: 05:35 2 GCOMIC VerSioa 7.Cp (QA) - Apr11 2002 Pile: /hoce/saor/vyc-2l20cen/SENSrTIVs/SBO/drywell-SBO/So-drywelU-CO

-- Cntainrnt Preasure PR.1FR2

,0 -

. .I . . .2 Tiste, (sac)

Figure 40 - Containment Pressure - Case SBO-drywell-comments SDO - EOP-foleak-drywell2-sensitivities-!coankts Y.ar/15/2005 14:04:06 GOTMIC Version 7.Cp2(OA) - April 2002 d File: /kome/nchor/vyc-2120c=/SMSTlrITY/SB/drywell-0SBO--

EIK Surface Ttzperatu.re THSTBGM87 C,'

C4 1-3

.2 144 109 19 1. .

Tiue (se)c 5 le3 Fl~igre 41 - Surface Temperature., Heat Structures 5,6,,7 - SBO-drywell-comments

!Entergy. ... Calculation WC-2405 Rev."O Page 65 of 85 SBO - 8OP-Noleak-dyvell2-eensitiivities-comefnts Mar/15/2005 14:04:52 GOTaIC Version 7.0p2(QA) -'April 2002 File: /bone/schor/vyc-212Occn/SSITEVITY/SBO/drywell-SBO/SBO-dryWell-com 51 -Surface Teaerabire 7BB Tog =10 7E11 4

k- .7 e CD .

i O

3,6 7.92 10.8 - 14.4.

, . .Ti. . (,

Figure 42 - Surface Temperature, Heat Structures 8,9,10,11 - SBO-drywell-comments SBO - SO-Noleak-drywell2-sensitivites-coalsents Mar/15/2005 14:05:12 GO1MC Version 7.0p2(OA) - April 2002 Pile: /hone/schor/vyc-2 Oc=/SZNS1T:VWrr/SBO/drywell-SBO/SBO-drywell-co 52 Surfacs teaperature CB13H12 14 N I. I Heat lab 13 4 3. 6.4 10.8 A 17 .2 Tiae (ec) Xte3 Figure 43 - Surface Temperature, Heat Structures 8,9,10,11 - SBO-drywell-comments

.- .- . - ' (&En t ' Calculation WC-2405 Rev. 0 Page 66 of 85' SBo - 2OF-Noleak-drywell 2 -snitivities-commeat .

Mar/15/2005 14:05:12 GOTHIC Version 7.Op2(QA) - April 2002 -

Pile: /hanelachwr/vyc-2120Cc=/SNS1ITI:VT/SBO/dxywell-SBO/SBO-dyrell-com 2 _Surfa tcaeraare

913Th12 C3 3.4 C4Heat Slab 12 j* -eat lab 13.

C.'

C 3.7.2 17.8 14.4 2 1' 2 1. .2 Te (see) 'xe3 Figure 44 - Surface Temperature Heat Structures 12,13 - Case SBO-dryMell-comments

Ete Y Calcudltion VYC-2405 Rev:;0 Page 67 6f'85 -

7.0 Results and Conclusions Assuminge Station Blackout with RPV depressurization (cooldown) at I hour after the event the following results and conclusions are found:

1) The drywell temperature -forall cases analyzed stays below the EQ drywell temperature profile for the entire SBO coping period of 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and the additional 10 minutes to power the low pressure pumps (i.e., the drywell temperature for all cases analyzed stays below 300 TF for more than 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months of transient).

2) The drywell liner temperature stays below the design temperature of 281 A: for more than 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months after the SBO event.

3) The drywell pressure stays below the design pressure of 56 psig.

4) For 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes the wetwell pressure stays below the PSP curve for all cases analyzed.

5) The analysis shows that there is no need to spray the drywell when in the unsafe region of DWSIL.

6) There is enough inventory in the CST to insure that the CST is not depleted before the time of low pressure pumps availability such that the core stays covered. A CST inventory of 75000 gallons was assumed.

7) The maxim suppression pool temperature for all cases stays below 182.2 'F.

8) The analysis predicts a conservatively high drywell temperature. Several factors contribute to this conservatism:

The heat transfer from the vessel to the drywell is based on a constant heat transfer coefficient at normal operating differential temperatures. However, this heat transfer coefficient will vary with the temperature difference to the /4 power based on the dependence on the Grashoff number.

The heat transfer to the drywell from the drywell heaters is not subtracted from the vessel.

The reactor building side of the thermal conductors are considered adiabatic.

A constant leakage is assumed; the leakage will decrease as the'vessel is depressurized.

9) No restriction on the rate of cooldown is applied to protect the drywell temperature beyond the restriction of depressurization (cooldown) function of RHRSW temperature (Reference 1).

Note: "Unverified Assumptions" and "Affected Documents" items are being tracked via LO-VTYLO-2005-00135 (also see Section 4.1 and 4.2).

. n

. tergy Calculation VYC-2405 Rev. 0 Page 68 of 85"1 8.0 References The References are divided into Section 8.1 and General References (Section 8.2). Section 8.1 includes all references.

8.1 Design Input References and General

References:

1 VYC-2398 Rev 0 Torus Temperature Calculation for a Station Blackout Event at Extended Power Uprate, dated March 2005.

2 VY Tech Spec.

3 NUMARC 87-00, dated 11/20/87, including NRC accepted errata and Q&A from NUMARC seminars and Topical Report F.

4 VYC-2397, Maximum RHRSW Flow to RHRHX for SBO & Appendix R, dated 1/2412005.

5 DRF 0000-0011-5646, OPL-4A (Containment Analysis Input Values) for Vermont Yankee Nuclear Power Station EPU/MELLLA+,dated 2/6/03.

6 Calculation VYC-1628 Rev 0, Torus Temperature and Pressure Response to Large Break LOCA and MSLB Accident Scenariosi daied 4127/98.

7 George, T. L., et. al, GOTHIC Containment Analysis Package, Version 7-0, July 2001.

8 VYC-2208 "GOTHIC 7.0 Code Installation Validation and Verification at VY", dated July 18,2002.

9 VYC-1457, VY Containment Heatup Analysis - Appendix R Application, dated 8/19/96.

10 GE Design Specification, No. 22A1 184, "Drywell Atmosphere Cooling System", Table I, Drywell Cooling Load Summary, #8 on Sheet 9.:

11 VYC-1254 Rev 3, Containment and RPV Volume Calculations, dated 5121/98.

12 VYC-1850 Rev 1, OPL-4A Input Preparation, dated 6/22/99.

13 GE Design Specification # 22A1182 RevI "Protective Coatings-Special" 14 Drawing G191526 Rev 2.

15 J. P. Holman, "Heat Transfer", McGraw Hill Book Company, 1981.

16 Standard Review Plan Branch Technical Position CSB 6-1.

17 VYC-1628D Rev 0 CCN02, Torus Temperature and Pressure Response for to Appendix R and Station Blackout Scenarios, dated 06/2312oo3.

-. Ente~y ' Calculation VYC-2405 Rev'Y0 Page 69 of'85`Nh I 18 VYC-2279, Evaluation of EPU Impact on Ambient Space Temperatures During Normal Operation, dated 8/26/2003.

19 EQ Manual Vol 1.

20 22A12W5, Rev 1, Reactor Containment Design Specification, September 1969.

21 VY Drawing G191489 Rev 2..

22 VYC-1290 Rev 0, Vermont Yankee Post-LOCA Torus Temperature and RHR Heat Exchanger Evaluation, approved Augustl, 1994.

23 VYC-2045 Rev 0 Residual Heat Removal Heat Exchangers Fouling Factors and Projected Heat Rates for Cycle 21, dated 12/1/99.

24 GE-VYNPS-AEP-146, Letter Michael Dick (GE) to Craig Nichols (ENOI), VYNPS EPU Task T0400: Decay Heat for Containment Analysis dated March 10, 2003.

25 TE-2003-020, Feedwater Parameters for Power Uprate, April14, 2003.

26 VY Memo VYS 2000/39, P A Rainey/T. P. Bowman to J. R. Lynch, 'Torus Temperature/SW Design Temperature Recommendations", April16, 2000.

27 VY UFSAR Rev. 19.

28 OP 0105 Rev 11, Reactor Operation.

29 Crane Technical Paper No. 410, Flow of Fluids through Valves, Flttings, and Pipe, 1976 Crane Co.

30 EOP-3 Primary Containment Control, Rev3, dated 10/19/02.

31 ASME Steam Tables -Third Edition, 1977.

32 VYC-2306, Torus Temperature for Appendix R events at EPU Conditions, dated 08/29/2003.

33 VYC-1850A Rev 0, "OPL 4A Input Preparation", dated 7128199.

34 VYC-1628B Rev 0, "Torus Temperature and Pressure Response to Small Break LOCA Scenarios, Model Development", dated 11/3/98.

35 OE-3107 Rev 17, EOPiSAG Appendices dated 04/29/2004.

36 OT 3122 Rev 19, Loss of Normal Power, dated 04/18/2000.

37 EOP-1 Rev2 RPV Control, dated 10/19/02

Entew- CalculationrVYC-2405 Rev. 0 Page 70 of 85 8.1 General References Georgejy. L., et. al, GOTHIC Containment Analysis Package, Version 7.0, July 2001 GOTHIC 7.0 Code Installation Validation and Verification at VY, dated July 18, 2002.

ASME Steam Tables -Third Edition,.1977 J. P. Holman, "Heat Transfer", McGraw fill Book Company, 1981 Crane Technical Paper No. 410, Flow of Fluids through Valves, Fittings, and Pipe, 1976 Crane Co.

Entergy . Calculation VYC-2405 Rev. 0 Page 71 ofn85 Mat ATTACHMENT 9.10 COMPUTER RUN

SUMMARY

SHEET COMPUTER RUN

SUMMARY

SHEET Page Iof I Calculation No. VYC-2405 Revision 0 Date 16 March 2005 Sheet I of I

Subject:

Drywell Temperature Calculation for a Station Blackout Event at Extended Power Uprate Code GOTHIC V7.0p2 Catalog No.02543 Version 7.0 SQA Classification Level A Run Title (variable, described in Section 6.0 and Att. A)

Run No. No., Run Date __By Output Use: Q Variable Values As Noted 0 Plot Attached

Disk. File No.

Description Of Output

Figures in text.

Input file on Disk.

Multiple cases were run, all are described in Section 6.0.

One case (base case) is attached in attachment A).

The Figures in text, for each case, have the date of the run & the run name.

Comments: None (Attached additional pages if necessary)

Review: 0 Information Entered Above is Accurate 0 Input Entry Accurate 0 Code Properly Executed (Based on User Manual)

Output Accurately Extracted or Location Specified Reviewer Comments A ,

Preparer (Print/Sign) Date Reviewer (Print/Sign) Date Liliane Schor Alan L. Robertshaw 3 q . 3//1/0 <

- . D

'Entergy. Calculation VYC-2405 ev;O " Page 72.6f 85 -

Calculation Impact Review Pages (ENN-DC-126 Attachment 9.7)

From System Engineering PAGE 3S OF57 ATTAcfWUrr9.7 - CALCULATM IUPACrREYM PARE CALCULATIO IMPACTREVE:W PAGE :

Date: 14l FebwarvX205 0 OR . 3 NOR Nte: X indicates required distibufion)

To: __Mechanical Engineering _X_Licesitg X Operations

_18C Engineering _ Elec Maintelan Ch iy

_: t Englneerng _ I&QMakienance _HP/RNid'oogipal

/ _ CN Engineering - Mech Maintenance _CompperApplications

,( System Engineering -Componert Engneig _Rad Engkneering

__Redctor Engneeving - Program Engineering ISI Engineoring

X DBD Owner RHR. P. Perez __Nudsear Engineerng [ ST Eng ierng X DBD Owner SA. P. Perez EO __PSAg (Name) (Other)

From: Lrsane Schor 802-451-3013

- ;(Originator Print Narne and Phone extenslon)

Calculaton No: VYC-2405 Revision No. Q

Title:

Drvwefl Temperature Calculaton fora Station Blackout Event at Extended PowereUprate.

Reference:

WA DateRelsponse Requtrect 16 February 2005 NESSAGE. Work organizatons are requested to rew the sLkiect calculaion (parls attached) to*

Identify Impacted cacuatons, procedures, Technical Speccons, FSAR soctins, 6.6idesign docutnerte (e.g.. EC Ifles, DBD, Appendx Rl. I5ltlST, PSA, UOVs'AO~s, etc.) and oter doumnts which must be updated because ot Ue. callation resuhsa. Albo provide the namne o e hIndIvIdual responsie for 2he action end the trackig nunber. Tho tracking IBem shaud udie a requrkement to

. ensure that any ER Implementation asssdated with the IRem Is completed prior to re4sing the Inpacled docuneni Sign and return the form to the originator.

IMPACT FEVIEW RESULTS:

Affected Documents Responsible Traddnhg IRemarks

. Indikidual Number _

£OP-3_ _ _ _ _ _ _ __

eo P ;:r.. C..1 e. ._ _ _ . _ _ _ _ _ _ __ _ _ _ _ _ _

D B9D -S VI 0. D __ _ __ _ __ _ .__ ._

Responding SupervlsoriManager (ordesignee): B-t &!eeuecK & . CAJ-,.A.

-v. _-- _

V//osa-ate NameSignature

7.-

... - En tergy. - I Calculation VYC-2405 Rev. .0 Pagie'736f'85 Calculation Impact Review Pages (ENN-DC-126 Attachment 9.7). Continued From System Engineering 9.

NUCLEARfN (JtL~TRRATE ENNDC424 REV. 4 Erery ANUAL Lvoxm~zo(ALtx PAGE 3.5 o57 ATTAorWEwIL97 CALCULATMIMON ACT REVM PAGE CALCULATION IMPACT REVIEW PAGE Date: 14 Febntw 2005 0 OR OElNOR (Note: X indicates requIred distrRbuticn)

To: _Mechanical Engineertng X Licensing '.. Y ..Operticns

_I&CEngineering =Elect Maintinartce.- .Chemlstgy Electrical Enqlneerlng ..... I&C Mairernance *HPIRad~okgicaI CMI Engtieering ......MechMairtenei *ComputerApplications

_XSystem Engineering - Cm ponengEgneerin~g - H ad Enrineering

_Reactor Engineering - ftrar mgeering - .IS!Englneertng

_X DBDOwnmer FllR.RPe rez - Nudear.Englneerkv g ,tIST Enqk"n X DBD Qwner SA P: Pere z ___EO ,PSA 4Name (Other)

From: Ulia-e Schor 802451-3013 (OngirtoraPrint Name and Phone edension)

Calculatic_ No-' VYC-2405 Revision No. O

Title:

Dmwell Temyoatura Calcuation fow a Saton E3lac1out Event a Extended Poww6rato .

Reference:

WA Date Response Sequired: i 6 February 2005 MESSAGEL Work organiztior arm requested to review the subject caiculaion (parts attached) to Identfy Impacted cabiatons. Pri~edures, Technial Specifcions, FSAR sections, other Sesign

&octiments (°-t EO Bies; DBD, Appendix K ISM. PSA, MOVs/AOV3, ec.) ad other &cumnenrs which must be updated because f the calcuation resuts. Also provide te nane of the ihrdvidual responsrle for the aion nd te raddng number. The tracking Item should indude a requirement to ensure that any ER Inpeentationagsocted wit the Rem Is crpleted prior to revising the impacted document Slin and retrn te form to the orknaor.

IMPACT REVIEW RESULTS:

Affected Docrimens Responsible Tracking Aemaks

__xgvkWa Number '

OP_4032 S. _cb_ See attacTed comments Responding SupervisorAoanager (or designee): Stewhen Joresch o Y160 Na~gnar Name/Si9mbure dOWAKal

' V Date

= t

EFEntergy Calculatioh VYC-2405 Rev;TO Page 74 6f95' Calculation Impact Review Pages (ENN-DC-126 Attachment 9.7), Continued From RHR and SA DBD Owner

.. . . 1.

.A., .- .

ArrAcmENT9.7 CAI.CULATnON IMPACT REVIEW PAGE CALCULATION IMPACT REVIEW PAGE 1i

.n~ nvtll -

L~dAU. EQW571FWV. - OR El NOR (Noe: X indicates required distribution)

To: Mechanical Engineering *.j icensing *XOperations

I&C Engineerng _Elect Mainteance _ ChOlsty

- Electrical Engineering = I&C Mainlenance __HPIRadoogical GChIEngineering . Mech Maintenance _ ComputerApplications

=System

_ Engineering . *= Component Engineering .'-RadEngineering Reactor Egineering -. Program Engineering -_ISI Engineering

_XDBD OrnerRHR. P. Peez NucearEST Engineering

_XDBD Owner SA. P. Perez _ EQ PSA (Name). ;(Other)

From: Lllane Schor O2-4511.3O1S (Originator Print Name and Phone extehsion)

Calculafon No., VYC-2405

R6in No. 0

Title:

ttelt Temperature Calulation foi' a Station Blackout Event at Extended Power Urmte.

. Re'-erence:.N/A* * * ,,* .

DateResponse Required 16 FebruarV 2005

ME$tAGE: Work orgarizatns are equested to review the subject calculation (parts attached) to kientffy Impacted calculations procedures, Technical Spedfications, FSAR sections. other design..

.dojrne ts:(e.g., EQ files, DiD, Apper C, tSiST, PSA, MOVs.AOVs. etc.) and other docunents.

which rnustbe updated because of the.cakdion resuts. -A provide the narme of tw Individual responsible for the action and the tracking nurmber. The tracldng item should Incdude a requirement to -

ensure that any ER rnpleamrtati6n associated whit the item is comnpleted prior to revising tie iJrnpacted document Sign and retumr the forri to the originator.

iMPACT REVIEW RESULTS:

Affected Documents Responsible Tracing Rernarlcs

.nivkidkal Number SA DBD P. Perez LOYVTYLO- May need update for Drywel -

2005.00135 SBO assmpIons Methodology of VYC-2405.

R~~~i nr i~n / U n n o r I n, dp i n

~~n1 T t t w

Date

. Name/Sfurd-_- - I

.. ';Entergy' - Calculation VYC-2405 Rdv.' 0. Pagb 756f'85 Calculation Impact Review Pages (ENN-DC-126 Attachment 9.7!, Continued From Operations - EPU Engineering NUCLEAR 6:1;ny ENN.DC-126 l REV.4 I &Erdergy MANAGEMIENT - I MANUAL tORMATIONAL113X P.AGE 35 0e57 ATrAcmmENTS.7 CALCULATIONthPACTREVIEWP/GE CALCULATION IMPACT REVIEW PAGE Date: 14 Februarv 2005 0 QR O3 NOR (Note: X dkcates required dstribution)

To: __Mechanical Engineering X Lcensing . LOperations

- I&C Engineering _ Elect Maintenance _Chemistry

_Electrical Engineering _ I&C Maintenance __HP/Radidogicai CIvU EngineerIng __Mech Maintenance -_ComputerApplications XSystern Engineering Component Engineering .=Rad Engineoring Reactor Engineering - Program Engineering ISI Engineering X C D8D Owner H -P.Perez_ _ Nuclear Enoneering IST Engineering X O:iD Ovwner SA. P.-Perei __EQ _ PSA (Name) (Other)

From: Liiane Schor 8024514013 (Originator Print Name and Phone extension)

Cakulation No.: WYC-2405 Revision No. e Tite: Drvwell Temperature Calculation for a Station Blackout Event at Extended Power Uprate.

Reference:

NIA Date Response Required: 16 Februarv 2005 MESSAGE: Work orgarfhtions are requested to review the subject cafdulaUon (parls attached) to Identify impacted calculations, p ocefres. Tedrical Specifications. FSAR sections, other design documents (e g.. EQ riles. DBD. Appendix R. i'VIST. PSA, MOVsJAOVs. etc.) and cter documents, whih must be updated because of the calculatlon results. Also provide the name of the nddual.

responsible for the action and thetrackng number. The tracking item should Include a requirement to ensure that any ER limpkementation associated with the ttem Is completed prior to revising the impacted documen. Sign and return the form to the originator.

IMPACT REVIEW RESULTS:

Affected Documents Responsible Tracking Remarks

._ ._. tIndividual Number EOP-3 Study' Gide OT.3122 .

Ltsson Plan for EOP-3 ON-3147 . .-

ON 314S .

Responding SupervisorManager (or desigiee) Bryan Crke & r */16/2005 NameSigiatbre Date.

!Entergy

Calculation VYC-2405 Rev. 0 C Page 766f85 'Y Calculation Impact Review Pages (ENN-DC-126 Attachment 9.7), Continued From Licensing

p. .

AEnNerU. MANAGEMINT QUIYRtZATXD . ,DC1* 4 MANUAL Lis 'rPAzan t'GE 35 OF57 ATrACWE 9.7 CALCULUION IDAPAcT REvEw PAGF cALCULATION IMPACT REVIEW PAGE Date: 14 Februanr 2005 . OQR ONQR

(Note: X hdicales requtred distribution)

To: _ Mechanda Engineering XUcensing XOperaUons

_ I&C Engineering _ ectMaintenance _ Chemistry

__2eclricl Engneerbg __I&C Maintonance _ HPRad-ciogical Civl Engineering Med Maintenance ComputerApplicatlons

_System Englneeng _ Component Eii~nerlng Rad Engineering Reactor Engeerig. __Program Eng1neering ISI Engineering XDBD Owneer Fj.erez __NiSdear Engineerino IST Engineerg X O80 Owner SA. P. erez __EQ =PSA (Name) (Other)

Fromr: Litione Schor' 10224M-14013 iOrlsfnalor

,_ . .;h..._ ._. .Prbit Name and Phone extension)

Calculaton No,: VYC-2405 Revision No. 0 TiUe: Drywell Temperature CalcuLbtion for a Station 8lad out Event at Extended Powef Uprate.

Reference. WA Date Response Requred: 16 February 2005 MESSAGE. Work organzations are requested to rerview the subject cilculation (parts ata~hd) to

$dent*f Impacted qaulations; procedures, Technical Spodiscatons. FSAR sections. other design

'documents (e.g..'Ea Iels, DBD, Appendix P. IsVIST, PSA, MOVs AOVs. 'et.) and other documents.

which riiust be updaled bikuse of the calculbnresubt. Also provde the name of the indMdual responsible for the adion and the traddcg numnber. The backing item should Indude a requiement to ensure that ary ER lrrqekm on associated wth the Rem Is complete prior b reviskg the Impacted documenL Sign and return the form to the originator.

IMPACT REVIEW RESULTS __

Affected Documents Rasponsalbe Tracking Remarks IndlivIdual. Number

- .~ -, I* _ __ _ _

Respondig SupemvisoriManager (or designee): 4.... Odi tc nk .A S' 7/-

Narn&lSkjature Daoe 1 -. .#6' ~-"-/ A- S'~

292 v7S&A_ /

/4< . 4- e° oeFL -fv

T',.;~Entergy~' Calculation VYC-2405 Rdv;0O NC& 77df 85' -

I.*.-.

Calculation Design Verification and Review (ENN-DC-134)

NN.134 Revi~oN 1 ATrACHMENr 9.1 DSINVERIFICATION COVER PAGE DESIGN VERIFICATION COVER PAGE O IP-2 QIP-3 DJAF []PNPS 01 VY Page I of 5 Document No. WC-2405 Revision 0

Title:

Drywell Temperature Calculation fora Station Blackout Event at Extended Power Uprate MQuality Related [lNonQualityRelated DV.Method: ' DesIgn Review ' QAlte mate Calculation []Qualiticatlon Testing VERIFICATION VERIFICATION COMPLET AND COMMENTS.

REQUIRED . RESOLVED (DVjndsig, and dalc)

O . Electrical O Mechanical El Instrument and Control

. Civil/Structural Desien Eneineirins,

.luid Systems ( a Alan L. Robertshaw.

Q a/r/or El Design Verification for WC-2405 Page 1 of 8

.- . Entergy Calculation VYG-2405 Rev."O Page 78of 95 '

PENN-DC-1 34 REVISION I ATTACHMENT 9.7 CALCULATION DESIGN VERWICATIONCHECKLIST IDENTCCATION: DISCIPL2NE.:

Document

Title:

Drywell Temperature Calculation for a Station El Civil/Strictural Blackout Event at Extended Power Uprale F] Electrical Doc. No.: WC-2405 Rev. 0 QA Cat. E]T&C Ver'fier*

Veni__er_.

SCgaauihosation for Alan L. Robertshaw

_j_ _ E OMechanical Nuclear ariou s .. Other:

~ifi~k~i : Design Eneineerink.

l JWA: . Fluld Svstems MEiTOD O;F VER1EICATION:

Design Review 0 Alternate Calculations El Qualification Testi Dealgr Inputs -Were the Inputs -correctly selected arnd Referrcen Incorporlted Into the desIgn? Pa~c No. Scakimn 5 oIJY 405 Design Irtfts Incilude deslgi bases, pfit~operationsi chdsuom. peforunce OR.-

riqi~renents, regulatory requirernents and comrnitments, codes, standards, Ileld dati, etc. Al Inornnalon used as design kotits shotd hays been k'brogmph N~o. __ _ __ _ __ _ __ _

rewlewed and!ppro~ed b~ the reisponsIble dessigi arganltaeon. asapplicable.

ARhifttnee to be retrevalfe or excerpts p of6uwents usedihotidbe Completion ofthe Refervhce 9oxes is .

Optional for at queslions..

lSt* aspteecific desilgn iipu procedures for gu~idarre InIdertlfftIrg ut.

Yes 0 -No E N/A [

Verifier Comments:

Section 5;0. Input and Design Criteria, of VYC-2405. has been satisfactorily reviewved. Any identified Design Input needing verification is listed in Section 4.1 of VYC-2405 and tracked via LO-VTYLO-2005-00135 (see Item #2 ofthis Calculaition Design Verification Checklist). All other Design Input has been verified in WC-2405.

Resolution: None needed.

Attachment 9.7 Calculation Design Verification Checklist Design Verification for WC-2405 Page 2 of 8

!EF tera ) CalculatioriWC-2405 Rest; D Page 79 6f 85 :

a_ Assumptions - Have the assumptions been vedfied? 'Refernce FcapNo. Uolcitio smmrv Pate Al Sea 41orVYc-2405 Yes E No 0 N/A - OR Pamrpih No.

Vcrifier Comments:

Section 4.1, Assumptions that need Verification orImnplementation, and LO-VTYLO-2005-00135

.CA02, document identified "Unverified Assumptions" from YYC-2405 Drywell Temperature Calculation for a Station Blackout Event at Extended Power Uprate. The following lists the various "Unverified Assumptions' fromr WC-2405:

1) 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months restoration ofoutside power (coping time).

2) 10 minutes to stait RHR through the RHRHX, '2 RT-TRSW pumps and CS.

3) Acceptability of using 75OCiO gal from CST ( ghaoge of lcvel setpoioi).

4) Maximum CST temperatmr ofi350 F. .

5) The depressurization iate function of Service Water temperature needs to be verified and proceduralized is follows:

- For SW> 75"F; depressurize the vessel at 807F/hr or higher.

- For lower SW temperature (SW < 750F) no restrictions on depressurization rates.

Upon verification ofthese assumptions. the calclation should be revised to convert the assumptions to Design Input and the calaulatiori Status ihould be changed.

'Resolution:

LO-VTYLO-2005-00135 CA02 hai bccn issued to track tcsc Unvcrificd Assumptions.

Upon verification of these assumptions, the calculation should be revised to convert the assumptions to Design Tnput and the calculation Status should be changed.

3. Quality Assurance - Isthe Quardy level correct?

Pqe No. Coer Sbeet of VYC2405 YesI5 No El N/A O OR Paranh No. __ _ _ _ _ _

Verifier Comments: WC-2405 is correctly designated 'Qualhy Related."

Resolution: None needed.

Attachment 9.7 Calculation Design Verification Checklist Design Verification for WC-2405 Page 3 of 8

Entergy . - P.8

Calculation WC-2405 RevWO P'age 80 of 85 E;-'.'

Codes, Stanidards and Regulatory Requirements - Reference Are the applicable codes, standards and regulatory Pae _ _

requirements. including Issue and addenda property OR Identilled and are their requirements for design met? Pgp~h No.

Yes0 No D N/A 5 Verifier Comments: Appropriate use ofrequirements (inputs, assumptions, methodology) set forth in the VY Technical Specifications and UFSAR have been Followed in VYC-2405 as needed.

Resolution: None needed.

5. Operating Experience - Have appricable construcbon and Refereirence operating experience been considered?

p e No. __ __ _ __ __-

-OR Yes ED

Nof D N/A 5E : araph}o.

Verifier Comments:

Consideration (discussions and reviews)has been given to various timelines for operat6 r actions, equipment start times, cooldown rates, etc. .Some.Unverified Assunipiions exists vhich will rely .

on, in part, to VY operating expeience(e.g., tfy hour restoration of outside power (coping time-of 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and 10 minutes), ten minutes to start flow through the .lRTRX, the use 6iHR of 2 RHRSW pumps and CS, and the acceptabilitj of uiing 75000 gal from CST. hange f leve setpoint). SeeItem #2, Assumptions,.ofthis CalculationDesign CVcrfication Checklist.

Resolution: Consideration has been given to VY operating experience. Some Uriverified Assumpiions exist (for which commitments have been issued). See Resolution of Item #2, Assumptions, ofthis Calculation Design Verification Checklist.

6. Interfaces - Have the design interface requirements been Ref reC -

satisfied and documented? Pa e Na._ _ _ _ _ _ _ _

OR-*

Yes 0 No E N/ .0 Par phNd. _ ** .

Verifier Comments:

The relationship between the Fluid Systems (design group") and other organizations within VY have been satisfactorily me using the Calculation Impact Review Page. The Calculation Impact Review Page was completed by persons in the Operations Department, System Engineering, Ucensing, and various DBD owners.

Resolution:

None Needed.

Attachment 9.7 . Calculation Design Verification Checklist Design Vedrfication for VYC-2405 Page 4 of 8

Ak

'Entergy.-.: Cfalculation VYC-2405 Rev. a Page 81 of85 A :

N7

.P..

7.a Mthos -Ws appop~ataanaltici mehodused Reen.

7. Methods-Wasanappropriateanaly0calmethodused?

P.

Pa-eWo. Section 3 VYC-2405

. 'Yes 0 'No.A N/A E OR PPPbliop~h.

Verfier Comments:

The GOTHIC code (Reference 7 & S of VYC-2405), Version i.op2 was selected for use in VYC-2405.

Resolution: NoneNeeded.

8 .Desfgn Outputs - Istheboutput reasonable compared to the Rdfcrr~uw IPa' No.

.Yes:19 NoD NIA ElxmOR Verifief Comments; The'output is reasonable compared to the inputs; Previous, similar Drywell SBO analyses are familiar with the preparer and reviewer of this SBO calculation and thus the output given in VYC-2405 was reasonable forthc various changes'and modifications made to the previous input.

Resolution: None Needed.

9.- Accepianc..Criteria -Are the acceptance crftedta . Ruefee incorporated inthe calculation suflident to allow venficaUon Pa.-CNO- s1io2or 0-2o that design requirements have been satisfactorily ' OR ,.

occompnshed? PubgmpNO.I

.Yes No l NIA Ei Verifier Comments:

Sccion 2.1 vas added to include the Acceptance Criteria.

Resolution:

None Needed.

Attachment 9.7 Calculaiion Design Verification Checkdist Design Verification for VYC-2405 Page S of S

. . . I. . . - .... . . . . _ . - _ .... - . . -__ .- . . .I . -.

Enteg Calculation VYC-2405 R-V. 0 ' Page 82 of 85 WY 1i. Records and Documentation -Are requirements for Referece Wecrd preparstion, review. aooroval- fetention. etr-. PeNo.& Seinm 4.2 nf wC-740ns adequately specified? OR Are all dCments prepared In' clear kgble mariner sWla r mcrorihking and/or PmNo.

oiler aocunnenton loramge r.ethod? Have l kpacted docuenhtsbeen identired for update?

Yes 0 . No Q N/AED .

.Veriier Comments: VYC-2405 was prepared in a clear legible manner suitable for microflIming and/or other ddcumentation storage method. All impacted documents (Design Output) have been identified for update in Section 4.2 of VYC-2405 and in LO-VTYLO-2005-00135 CAOI.

Resolution: LO-VTYLO-2005-00135 CAOI has been issued to track these affected documents.

lf any other documents Are identified during the Calculatioh Impact Review process, other commitments will be generated.

11. Software Quality Assurance- For a calculation that utilized R.Rerence software applications (e.g., GOTHIC, SYMCORD).was it ageNo. Sion30fWC-205 properly verified and validated Inaccordance with ENN IT- OR 104 or previous site SQA Program? Pao._*_*

Yes- NoD . N/A _

Verifier Comments: The GOTHIC code(Reference 7 & 8 of YC-2405), Version .0p2 was selected for use in VYC-2405. This code was used in similar SBO analysis (Referciices I of VYC-2405). This specific version of the code has been installed and complies with the ENVY SQA procedures ENN-IT-104 (replaced VY procedure AP-6030) as documented incalculation VYC-2208 (Reference 8 ofVYC-2405).

-Resolution: None Needed.

OTHJER COMMEPNTS See Attached list of General comnments from review of VYC-2405 RESOLUlTONS Attached General comments have been made to reviewer's satisfaction.

All comments for "NO' answers have been resolved satisfactorily.

Attachment 9.7 Calculation Design Verification Checklist Design Verification for WC-2405 Page 6 of 8

!:-Entergy -Calculation W C-2405 Rev. 0 Page 83 of 85 ;

General Comnients from Alan L. Robertshaw from Review of VYC-2405 (Attached to Calculation Design Verification Checklist)

Tn Section 2.0, Add addition Section (2.1) entitled "Acceptance Criteria" and add appropriate acceptance criteria.

Done

In Section .0, discussion of GOTHIC code and V&V calculation need appropriate references.

Done

In Section 4.0:

1. For Assumption #4, please correct the units for density (value iscorect).

2. ForAssumptionr#9, add that 61 gpm is from Reference'3.

3. Add VYC-2306 to Reference (from Drywell Free Volume of Table 3).

Added

In Section 5.0:

1. In Section 5.31,: show in more detail how total Heater #5 loads are calculated.

Done

2. In Section S.4, recommend addition of!simple" drawing of Drywell Done

3. In Section 5.4 show that the total Surface Area adds up to the Value obtained from OPL-4a (ie., Table A Total = 15246.06 fl, RRUs = 1272.8 ftW, Vent Pipes = 2885.7.

A, thus Total - 19405 ft2).

Added

4. Add to description of Item 8 'Side of Drywell Head."

5. Add to description of Item 9 'Top of Drywell Head."

Done, both items

6. Is RRU areaReference 11 or 12 or other? Please correct if needed.

Broth, added

7. Flow Path 21 Forward and Reverseloss coefficients should reference CRANE (or similar). data not found in steam table. Need to add CRANE to references Done

8. Please add additional inrormation on new valve added (vacuum breaker, Valve #5.

e.g., size, type, etc.).

Done Design Verification for VYC-2405 Page 7 of 8

-Entergy - Calculati66iVYC-2405 Rev.0 Pag084of 5

9. Also for Val e discussion, note that the valve area'is different from the line area used in Flow Path 21 Added

10. Initial Temperatures of the Thermal Conductors are not given (and reference).

Added I 1.Please further explain why a loss coefficient of3.964 was used instead of the calculated value of3.93 for the vacuum breaker flow path. Any impact on analysis?

.1added this discussion in the document I also saidthat the difference is I= than 1%and wil make no difference on tie analysis.Anys'ay, was correctedin Case S.

In Section 6.0:

I; List Five case individtially to add clarity.

Done

2. See minor grammatical I spelliigi concernsnoted in marked-up calc draft Done

3. Section.6.2.1 please add info on Flow Path 21 (iakagi) area change.

Done Id Section 8.0, Reference Sctions

1. Need to add Reference for ASME Stearn Tables and one for CRANE Technical Paper (if needed)

Added

2. AreReferences II and 13 used?.

Yes, Reference 71 and 12 areRRUreferences and 13 isforthepaint (addedsoremore discumion about the paint)

In the GOTHIC Input Deck, Thea Conductor #16, Concrete'Shield, a surface area value of 2108 ftwas used. A value of 2068 ft2 was given in the text of VYC-2405 (Section 5.4). Both of these numbers are found on page 35 of WC-1 850, Rev. I for OPL-4A preparation. The larger value (2108 fi2) is for total surface area, the smaller value (2068 fR) subtracts the

'Slots." The value used in VYC-1850, Rev. I CCN-O1 (OPL4A, Resolved for Analysis) uses the 2068 f value. Please discuss this in the analysis and determine which value to use, any sensitivities, etc.

Case addresses this

Tn Attachmerrt A, it appears that Figre on page A14 is a duplicate figure and should be removed.

Removed Design Verification for WC-2405 Page 8 of 8

-Entergy Calculation VYC-2405Rdv. O Page 85 of 85 :

Files on CD

S80*,mmed k~a"h H 4.50X8 CTHFib O3AW/t1IUW

t B*e,iwmwt41Mn * *6.fla'xGTHP16' *aYISmI.oF$

jM Capicr ~.i 52A W0Iww&Pt&MUO .0m713 Wednesdy. M~arch 15. 2W5 5V03.11PM

BW 05-072 Docket No. 50-271 Exhibit SPSB-C-52-5 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Calculation VYC-2279, Rev.0 l Total number of pages In this Exhibit (excludinc this cover sheet} Is 14.

r.

621 This document contains Vermont Yankee proprietary information. This information may not be transmitted in whole or in part, to any other organization without permission of Vermont Yankee.

VY CALCULATION TITLE PAGE VYC-2279_ _ 0 NA _ NIA VW Calculation Number Revision Number Vendor Calculation Number Revision Number

Title:

Evaluation of EPU Impact on Ambient Space Temperatures During Normal Operation QAStatu: 0 SC 0 NNS 0] OQA Operating CycleNumber' N/A

' The Operating Cycle Numbershould only be entered here if the results of the calculation onl apply d&ring a specific operating cycle otherwise enter WNA.

Calculation Supports ADesign Change/Specification? l Yes [ No VYDC 2003-008 VYDC/MMW WSpecxNo.

Implementation Required? 0 Yes OI No Calculation Done as a Study Only? O Yes 0No Safety Evaluation Number N/A Superseded Calculation Number, Title and Revision: NIA For Revisions List CCNs, LUs, or SAs inncoporatedtsuprseded by this revision: N1L Computer Code(s): None used Are there open itexs in this calculationlrevision? CD Yes 0 No Review and Approval: (Print and S M Preparr: Gene O'Brien5nM (JU ,

Interdisciplinc (Heat Balance, AtL B only) Preparer(s): N. Z 5vos Intcrdiscipline (Heat Balance, Att. B only) Reviewes): R Srin iv Final Turnover to DCC (Section 2):

1) All open items, if any, have been closed.

2) Implementation Confirmation (Section 23.4) Total No. Pages in Package 64 (including ill attachDents)

El Calculation accurately reflects existing plant configuration, (confinmation mnethod indicated below)

. E Walkdown l As-Built Input re-iew 0 Discussion with, (PrintName)

OR

[ N/A, calculation does not reflect existing plant configuration

3) Resolution of documents identified in the Design Output Documents Section ofVYAPF 0017.07 has been initiated as required (Section 2.3.6,23.7)

/

PrintedNarve Signature Dale Page 1of 28 Pages'

For calculations perfomed using AP 0017 tbis i the number ofpages in the body of the calculation. VYAPF 0017.01 Forvendorcalculationsthis is thenumberofpage of AP 0017 forms added. AP 0017 Rev. 8 (Ttic page, review forms, data sheets, 5059, etc.) Page I of I LPC#2

Calculation Number. VYC-2279 Revision Number. 0 CCN: 0 Page 6 of 28 TABLE OF CONTENTS SECTION PAGE NO.

VY CALCULATION TITLE PAGE .................................... 1 VY CALCULATION DATABASE INPUT FORM .................................... 2 TABLE OF CONTENTS................................................................................................................................6 1 Calculation .................................... 7 1.1 Objective .................................... 7 1.2 Summary of Results .................................... 7 1.3. Method of Solution .................................... 8 1.4 Inputs and Assumptions .................................... 10 1A.1 Inputs ................................... 10 1.4.2 Assumptions ................................... 14 1.5 Calculation ................................... 14 1.6 Results ................................... 26 1.7 Conclusion ................................... 27 Attachments ................................... 28 Attachment A ................................... Al Attachment B................................... BI Attachment C ................................... Cl Attachment D ................................... DI Attachment E................................... El

CALCULATION NUMBER VYC-2279 Rev. 0 CCN 0 Page 7 of28 (

1 Calculation 1.1 Objective The purpose of this calculation is to evaluate ambient temperature increases in several plant spaces following the increase in reactor power level to 102% of 120%, hereafter referred to as Extended Power Uprate (EPU). The EPU will increase core thermal power from the current licensed level of 1593 MWt to 1912 MWt. For bounding purposes, the 122% (1950.9 MWt) heat balances are used for EPU HVAC evaluations.

This calculation evaluates the EPU impact on ambient air temperature in the following buildings or areas during normal plant operation:

Reactor Building well

-a--t-ea un~nel oith~erReactor Building Areas Turbine Building

Reactor Feed Pump Room

Condensate Pump Room

HP Heater Area (including steam line shelf containing main steam lines)

LP Heater Area Note: Comments to this calculation provided by letters PUPVY-03-208 dated 7/16/03 and PUPVY 212 dated 7/18/03 have been reviewed and incorporated (see Attachment D).

Increases in area heat gain and ambient air temperatures, as a result of EPU, are predominately caused by increases in operating temperature of piping systems, and equipment, and air-cooled motors operating under increased loads. The preuprate piping system temperatures are selected or extrapolated from a PEPSE Heat Balance that is tuned to match preuprate (current) plant data. The EPU piping system temperatures are selected or extrapolated from a PEPSE Heat Balance that provides the most conservative results.

Affected areas are evaluated to determine the temperature gain due to increases in heat loss from piping and mechanical equipment.

1.2 Summary of Results The results of this calculation show the effects of the EPU in terms of increased ambient temperature and heat load are due to increased feedwater temperature, as well as increased horsepower from the condensate and feedwater pumps. The ambient temperature increases are specified in Section 1.6 (Results).

Calculation Number: VYC-2279 Revision Number: 0 CCN: 0 Page 8 of 28 1.3. Method of Solution This calculation evaluates the temperature increase in a specified area using current fluid and ambient air temperatures and EPU fluid temperatures to predict the EPU ambient air temperature and corresponding temperature rise. PEPSE Heat Balances at 100% CTP and EPU at 122% CTP (References 7 and 8) are used to obtain preuprate and EPU piping temperatures. The increase in heat loss from piping is determined by comparing the ratio of "temperature differential between EPU pipe and area air temperatures" to "differential temperature between pre-uprate pipe and pre-uprate area ambient air".

The basis for using this scaling approach to determine increased heat loss from piping and equipment can be obtained by referencing the ASHRAE Fundamental Handbook (Reference 12) Section 20.

Formula?(9) - ASHRAE 20.9 is used for flat surfaces qS, =ti. -t ..s)/R Formula (10) -ASHRAE 20.9 is used for cylindrical flat surfaces q., "(t1 , -t Ox) / [r. In (r1/r1)]/kj + [r. In (r.Vrl)]/k 2 Formula (11) -ASHRAE 20.9 for determining heat flow per area of pipe surface

q. =q,, (r, At )

Where q, = rate of heat transfer per unit area of outer surface of insulation qO = rate of heat flow per unit area of pipe surface, BtuI(hr)(ft 2 )

R = surface to surface thermal resistance k = thermal conductivity of insulation at calculated mean temperature.

tLS = Temperature of inner surface t = Temperature of outer surface r; inner radius of insulation rl, r2 = outer radius of intermediate insulation r,= outer radius of insulation In = natural or Napernian logarithm For the purposes of this calculation it can be assumed that there is one layer of insulation, therefore Formula 10 can be simplified as follows:

q, =(tL,-t O,) / [r., In (rj/r)-)/kj The increase or delta in heat transfer per unit area of insulation can be stated as follows:

Aq =EPU [(tot -S) / [r, In (r1/r,)]/kj] / pre-EPU[(ti,-t 05) / [r, In (r,/r,)]/k1 I There is no change in either: rb r,, or rl,.

Calculation Number: VYC-2279 Revision Number: 0 CCN: 0 Page 9 of 28 Based on the predicted temperature increases in the various process streams it can be assumed that there is no appreciable change between the preuprate and EPU values for k.

Therefore Aq=EPU[(t,-t es) / pre-EPU[(ti,-t )

Present station operating ambient air temperatures are used in the evaluation. If operating data is not available, plant design area temperatures are used. An iterative process using an Excel spreadsheet is utilized. First, an EPU ambient air temperature is estimated. Next, the EPU area heat gain multiplier is obtained using the ratio of the EPU pipe / ambient air temperature difference to the preuprate (current) pipe / ambient air temperature difference as shown below:

(EPU pipe temperature - EPU ambient air temperature) = EPU Area Heat Gain Multiplier (preuprate pipe temperature - preuprate ambient air temperature)

The EPU factor is obtained by subtracting I from the EPU heat gain multiplier.

The EPU factor is then multiplied by the preuprate temperature difference between the air in and out of a particular air handling unit to calculate the estimated EPU temperature rise. The temperature rise is then compared to the difference of estimated EPU and preuprate ambient air temperatures. If required, a new EPU ambient air temperature is estimated and the process repeated until the temperature rise is equal to the difference of estimated EPU and preuprate ambient air temperatures.

r Feedwater and Condensate pumps, flows from preuprate and EPU PEPSE Heat Balances along with appi pump curves will 'atebe used to determine horsepower changes in the respective pump motors to determi the heat gain increase to the room. The percentage increase in heat gain to the room will be utilized wit e temperature rise of ventilating/cooling air currently being supplied to the room or area being evaluated.

The heat load from thecensate and feedwater pumps is evaluated by calculating brake horsepower (BHP) at preuprate and EPU ws. BHP is calculated using the following equation from page B-9 of Reference 14:

Bhp (hp) = Q(gpm)*H(ft)* p(lb 1 Jft3 ) [2 0

pump efficiency]

The flow, Q, is calculated from the mass flow ratecified on the heat balance using the fluid density, p, calculated at the average of pump inlet and outlet tem atures. The pump head, H, and efficiency are obtained from the pump curve.

The heat generated is due to pump motor inefficiency and is ca ted using Chapter 26, Equation 21 from Reference 12:

q (BTU/hr) = BHP (hp)*2545 (BTU/hrhp) *[100-0% efficiency]/ % effici The heat load due to the pump motor inefficiency is calculated for preuprate and conditions. The overall heat removal capability of the coolers at preuprate conditions is determined an sensible heat load due to piping is obtained by subtracting pump heat load from cooler heat removal capac The piping heat load at EPU is scaled due to the increase in fluid temperature and added to the EPU p heat load to obtain total heat load for the room. The temperature increase across the coolers is calculated g Equation 39.6(b) from Reference 13:

Calculation Number: VYC-2279 Revision Number: 0 CCN: 0 Page 10 of 28 96

[Ti (tu mi/(R3hr OF))]

Both~~~auae anmcndnstepupr.

the thisate 1.4 Inputs and Assumptions 1.4.1 Inputs The inputs for this evaluation are the fluid temperature in system piping and ambient room temperature of the areas considered. The pre EPU fluid temperatures are obtained from a PEPSE heat balance based upon current plant operating data adjusted to 5.00" Hg. condenser backpressure to obtain maximum fluid temperatures (Reference 7). For those cases where the fluid temperature is not explicitly listed, the temperature is obtained based upon the pressure and enthalpy listed using ASME steam tables (Reference 9). The EPU fluid temperatures are obtained from the 122% heat balance with a condenser pressure of 5.00" Hg (Reference 8) which provided the highest, and therefore conservative, temperatures.

The current ambient steam tunnel and pump room temperatures are obtained from HVAC system design criteria (Reference 5). The main steam tunnel design temperature is 130 0F, the reactor feedwater pump room and condensate pump room design temperatures are 105 0F. The current HP and LP heater area temperatures are 125 0F as taken from the Environmental Qualification Program Manual (Reference 6, page 11). The 20'F Pre-EPU Vent/Cooling Air AT contained in the Table below for the HP and LP heater area spaces is based upon transfer air at 105 0F.

The design conditions for the air handling units are obtained from Reference 4, except for TRU-5 and TSFIA/]B, which have their design conditions specified in Reference 19.

Table 1.4-1 Area Equip. ID Flow, cfm Tin( 0F) Tout( 0F) AT Reactor Feedwater Pump Room TRU-1,-2,-3,-4 16,750 105 85 20 Condensate Pump Room TRU-5 21,400 105 85 20 Condensate Pump Room TSF-IA/lB 5,000 90 105 15 DrywellRRU-1,-2,-3,4 16,000 135 97 38 Main Steam Tunnel RRU-17A, .17B* 5000 130 105 25

Per Reference 18, the coils for these coolers are incorrectly piped as parallel flow rather than counter flow The design inputs for piping are summarized in the table below:

Table 1.4-2 Pre-EPU pipe Temp EPU pipe Temp Pre-EPU ambient air Pre-EPU Vent/Clg

'F (Ref. 7) 0F (Ref. 8) Temp F Air AT OF HP Heater Are ESS to FWHI____ _ _3 403.7 125.0 20.0 FWHI Shell 383.3 _ 125.0

__03.7 20.0 FW to FWHI 330.7 346.4 125.0 20.0 FW lvg FWH 1 374.4 392.6 _ 125.0 20.0 Drains lvg FWH1 343.6 356.9 125.0 20.0

Calculation Number: VYC-2279 Calulaio Revision Nuber Number:

VC-2790 CCN: 0 Paeiof2 Page 11 of 28 Pre-EPU pipe Temp EPU pipe Temp Pre-EPU ambient air Pre-EPU Vent/Clg F (Ref. 7)

I_________ 'F (Ref. 8) Temp 'F Air AT 'F ESS to FWII2 338.0 358.] 125.0 20.0 FWIH2 Sell 338.0 358.1 125.0 20.0 FW to FWHD 299.3 312.5 125.0 20.0 FW lvg FWH2 330.7 346.4 125.0 20.0 Drains lvg FWH2 309.7 322.6 125.0 20.0 LP Heater Area _ _ _ _ _ _ _ __ _ _ _ _ _ _ _

ESS to FWH3 N.4 321.2 125.0 20.0 FWH3 Shell 305.4 321.2 125.0 20.0 CND to FWH3 227.6 238.4 125.0' 20.0 CND lvg FW.H3 296.6. 309.8 125.0 20.0 Drains lvg FWI{3 235.0 24k4 125.0 20.0 ESS to FWH4 239.0 252.1 125.0 20.0 FWH4 Shell 239.0 252.1 125.0 20.0 CND to FWH4 167.9 180.9 125.0 20.0 CND lvg FWH4 227.6 238.4 12 0'O 20.0 Drains Iyg FWII4 175.1 189.1 125.0 20.0 ESS to FWH5 173.9 181.8 125.0 20.0 FWH5 Shell 173.9 181.8 125.0 20.0 CND to FWHS 135.1 134.7 125.0 20.0 CND yg FWH5 167.9 180.9 125.0 20 Drains lyg FWH5 145.1 141.8 125.0 20.0 CND Pump Room_____ ______

CND to CNP 133.8 133.8 105.0 20.0

Calculation Number. VYC-2279 Revision Number: 0 CCN: 0 Page 12 of 28 Pre-EPU pipe Temp EPU pipe Temp Pre-EPU ambient air Pre-EPU Vent/Cig

°F (Ref. 7) OF (Ref. 8) Temp °F Air AT °F C ;L__

CNP 133.1 _ 133.1 105.0 20.0 RFW Pump Room CND to RFP 296_._6 10S.0 20.0 FW lvg RFP 299.3 312.5 20.0 Main Steam Tunnel i FW Ivg FWH1 374.4 392.6 130.0 25.0 Main Steam 547.6 547.6 130.0 25.0

_ _ I__ _ _ _ I __ _ I _ I__ _ _ _

Pump information:

Reactor Feedwater Pumps Table 1.4-3 Preuprat2(Reference 7) EPU (Re 8 ce )

Flow 64l6_ 8048044 Ib/hr

of RF's 21_ 3 per pump 32037631 Ib/hr 2 6 826-8 Wf Tin 296. °F 309.8 °F _

Tout 299.31°F 312.51°F Condensate Pumps Table 1.4-4

-lPreuprate (Reference 7) l EPU (Refeece 8)

Flow -- 647526 Ib/hr 8076444 Ib/hr

of CN~s _ 9 3 per pump 2145842 Ib/hr __2692148 Ib/hr Tin 133.8 OF OF Tout 1331° 133.1 IT Drywell information (Reference 17):

Drywell Cooling Load Summary*

Table 1.4-5 Coinpon Btu/hr Reactor Pressu ureV's* 459,000 Recirc. Pumps, Valves, an 278,000 Feedwater Pipe & Valves 124,000 Steam Pipe & Valves 212,000 Condensate & Instrument Lead 82,000 Lines

Calculation Number: VYC-2279 Revision Number: 0 CCN: 0 Page 13 of28 Component Btu/hi Control Rod Drive Pipe ,0,400 Control Rod Drive Pipe 569,000**

Clean-up Demineralizer Pipe & 17,800 Valves Shutdown Supply Pipe 8,100 Steam Safety/Relief Valves 206,600 Biological Shield (Gamma 16,400 Heating)

Safeguards system Piping (RCIC, 82,000 LPCI, HPCI, and core spray) /

Sensible Heat Gain Total, Normal 1,536,300

-Operation Steam Leak, Valves /155,000 Latent Heat Gain Total, Normal 155,000 Operation Total Cooling Load, Normal Operation 1,691,300

Excluding allowance for drywell cooler motors

- Temporary initial load immediately following scram

Calculation Number VYC-2279 Revision Number: 0 CCN: 0 Page 14 of 28 1.4.2 Assumptions For the piping in the 4nas§temuel, it is assumed the feedwater piping contributes 1/3 to the total heat gain and the main steam piping contributes 2/3 due to their respective surface areas. This is based upon the fact there are four 18" main steam lines and two 16" feedwater lines. From Reference 14, the external surface area of four 18" pipes is 4

4.712 ft2 per foot of pipe = 18.848 f 2 per foot of pipe and the external surface area of two 16" pipes is 2

4.189 ft2 per foot of pipe 8.378 ft2 per foot of pipe. No confirmation is required.

It is assumed there is no appreciable change between the preuprate and EPU values for the thermal conductivity, k, of the pipe insulation based on the predicted temperature increases in the various process streams.

1.5 Calculation Piping:

A sample calculation for the estimated temperature rise from the feedwater piping in the steam tunnel is shown below.

Preuprate pipe temp. (F) 374.4 Reference 7 EPU pipe temp. (F) 392.6 Reference 8 Preuprate ambient air temp. (F) 130 Reference 5 Preuprate Air Handling Unit 25 Reference 4 (RRU-I7A, 130 - 105 = 25)

(AHU) AT CF) I I The EPU ambient air temperature is initially estimated at 131 'F. [used for initial iteration and checked later]

The EPU Area Heat Gain Multiplier is obtained using the equation in Section 1.3.

EPU Area Heat Gain Multiplier = (EPU pipe temperature - EPU ambient air temperature)

(preuprate pipe temperature - preuprate ambient air temperature)

EPU Area Heat Gain Multiplier = (392.6- 131) 0F (374.4 - 130) F EPU Area Heat Gain Multiplier = 1.070 The EPU factor is obtained by subtracting 1 from the EPU Area Heat Gain Multiplier EPU factor= 1.070 -1 = 0.070 From Assumptions section, the feedwater piping contributes 1/3 to the total heat gain in the main steam tunnel, so the EPU factor becomes:

EPU factor= 0.070 *1/3 EPU factor = 0.023

Calculation Number VYC-2279 Revision Number: 0 CCN: 0 Page 15 of 28 (

The temperature rise is then calculated by multiplying the EPU factor by the AHU temperature differential.

Temp. rise = 0.023

25.0 0 F = 0.586 The difference between estimated EPU and preuprate ambient air temperatures is:

131 - 130 =1 (this is not close enough to the calculated temperature rise of0.586 0 F)

Try a different EPU ambient air temperature of 130.6 and repeat the process.

Preuprate pipe temp. (°F)374.4 Reference 7 EPU pipe temp. (0 F) 392.6 Reference 8 Preuprate ambient air temp. (0F) 130 Reference 5 Preuprate AHU AT (F) 25 Reference 4 (RRU-17A, 130 - 105 = 25)

The EPU ambient air temperature is estimated at 130.6 'F.

The EPU Area Heat Gain Multiplier is obtained using the equation in Section 1.3.

EPU Area Heat Gain Multiplier = (EPU pipe temperature - EPU ambient air temperature)

(preuprate pipe temperature - preuprate ambient air temperature)

EPU Area Heat Gain Multiplier = (392.6 - 130.6)

(374.4 - 130)

EPU Area Heat Gain Multiplier = 1.072 The EPU factor is obtained by subtracting I from the EPU Area Heat Gain Multiplier EPU factor = 1.072 -1 = 0.072 From Assumptions section, the feedwater piping contributes 1/3 to the total heat gain in the main steam tunnel, so the EPU factor becomes:

EPU factor = 0.072 *1/3 EPU factor = 0.024 The temperature rise is then calculated by multiplying the EPU factor by the AHU temperature differential.

Temp. rise = 0.024

25.0 0 F = 0.6*F The difference between estimated EPU and preuprate ambient air temperatures is:

130.6 - 130 = 0.6 (this equals the calculated temperature rise of 0.6 0F)

The temperature rises for the remainder of the pipelines was calculated using the same method using an Excel spreadsheet.

The results are shown in Table 1.5-1.

CALCULATION NUMBER VYC-2275 Rev.O0 CCN0 Pge 16 of 28 Table 1.5.1 -Design Temsperature Differential v- -E .4" .... 3

211. Lh 4.h4.

P.(IV p.p" , Pe. - S.." L- w.a" .. a Vu,. .. utKmr -EuT ~ REV~_ _C,

________ Illis .3? l N I...an+/-. S.

1P-4-* I It m 1..I.l..

i1..Pdmgm INS INS 12 S. C....o am DID .16 -

Id234 211g IXSOn NI ill 6..t MI 4 *.1 = T:

.mI.P1r-Sr 2SAO 39SIS ,24 IMF Im' S.C-I 035W 301 6

IqIWI 3311 34k6 ISINS Iwo 61 S..C- two W0a.

orq ZIN kl4 126 114S.S NS 4C-1 5W mesiIw 4.33e2a 2121 135.1 SIW2 045 -.1. 03 5 .is .

m.57450 we 110 INS 'in4 IWO so 0426 21, 6.4# .AP or$ =12 1200' NS -- im". 1. O720 4.41 17 lal4 525. 13 4664 1 6 me54 4534 AS79444 52 £6... 02mJIT

. Ing3 13 1I'I lag INs __ I 06 .. W _ u3... 0.di

-IFY" 6489 0IS"aw 61 S.b.4. 21 - 2k -I SS.Mw wI31 3$ .. n .. &a . I*c41-1 _4554 _As. 2

.4w-6 ".- -- -. -- . .-.--

Calculation Number: VYC-2279 Revision Number: 0 CCN: 0 Page 20 of 28 1Ii The Total Cooling Load for Normal EPU Operation is: 1.700.675 B The ratio of cooling loads to differential temperatures is:

qepu / qp, = AT,. / ATp/

ATepu = (qq.

ATpe)/ qpft = (1,700,675 B (135 - 97)0 F) / 1,691,300 BTU/hr ATP, = 38.21 TF Since AT,,, = 135 - 97 = 38 The increase in d yw mperature due to the higher feedwater temperature is:

AT = ATe"u- -38.2 F - 380 F AT =0.2 a fmbient drywell temperature at EPU is 135 0F + 0.2 0F = 135.2 TF Main Steam Tunnel (RRU-17A, -17B) q (Btu/hr) = cfin

1.08 (Btu min/ft 3 hr TF) * [Tin - Tout]0 F q = 5000 cfm

1.08 (Btu min/ft 3 hr OF) * [130 - 105]°F q = 1.35 x 105 Btu/hr The main steam tunnel piping is evaluated using the Excel spreadsheet described above. As stated in the assumption section, the temperature rise from the feedwater piping is "weighted" at 33% based upon its surface area ratio when compared to main steam piping. The increased feedwater temperature will result in an ambient temperature increase of approximately O.6°F to 130.60 F. As previously noted, the service water to RRU-17A&B is piped backwards, such that there is parallel flow rather than counter flow (Reference 18). Because of this, both coils are in continuous operation rather than the initial design of one in operation and the other as a back-up. The minimal increase in ambient temperature due to EPU will not adversely affect current operation due to the fact the current peak allowable temperature in the tunnel is 150°F (Reference 6, pg. 10). Also, per Reference 6, Section 7.5.2.5, the high space temperature alarm set point is 160°F and the MSIV close/scram set point is 200°F.

Reactor Feedwater Pump Room C[RU-i.-2,-3.-4 T eat removal capability of the reactor feedwater pump room is:

q (Bt =cfm

1.08 (Btu min/ft3 hr F) * [Tin -Tout]°F q =4

167 fm

1.08 (Btumin/ft3hr°F) * [105 -85]°F q=1.45x OB B From above, the preuprate oad due to two pumps operating is 0.97 x 106 Btuthr and the EPU heat load due to three pumps operatin 1.47 x 0O6 Btuthr.

Using the heat removal capability with al oolers running, the preuprate hcat load due to the piping is:

1.45 x 106 Btu/hr - 0.97 x lO6 Btu/hr= 0.50 x 106 B The temperature scaling method using the preuprate and EPU he ances shows the ambient temperature will increase 1.2°F.

The EPU increase in piping heat load due to higher feedwater temperature can beap imated using:

qm = mCpATprc and q = CpATp Since mCp is the same preuprate and EPU, qpu / qpc = ATcpu / Tpm or qcpu = qplc *(ATcpu / ATpre)

Calculation Number: VYC-2279 Revision Number: CCN: 0 Page 26 of 28 Thoreas that contain condensate and feedwater piping are the only areas that will experience an ambient temperature increase during normal operation due to EPU. The following systems will not experience an ambient ter ature increase during normal operation due to EPU: Reactor Recirculation (RRS), Reactor Core Isolation *ng (RCIC), Residual Heat Removal (RHR), Reactor Water Clean Up (RWCU), High Pressure Coolant Inject CI) and Core Spray (CS) (References 20 through 25).

Reactor Building open areas Temperatures in the open areas of the reac iding will not increase during normal operation as a result of EPU (Reference 26).

Control Room As shown in Calculation VYC-1502 (Reference 28), Section 2.1, heat sou the control room are from electrical equipment, ambient outside air temperature, and personnel. None se sources will increase at EPU. Therefore, the Control Room HVAC system's ability to provide approp temperature and humidity conditions for personnel and equipment during all modes of operation and emergen condition is not impacted by EPU.

1.6 Results The results of this calculation are shown in Table 1.6-1 below.

Table 1.6-1 Area EPU Ambient Temperature Increase (TF)

Drywell 0.2 Main Steam Tunnel 0.6 LP Heater Area 4.1 HP Heater Area 1.7 Feedwater Pump Room 7.6 Condensate Pump Room 3.5 er tePrr i 5 inreae Th i man seamn tunnel ambient temperature due to the higher EPU feedwater temperature is Tte results of the piping evaluation are shown in Table 1.5-1. At normal operating EPU conditions, the am attemperature in the LP heater area will increase approximately 4.1UF to 129.1IF and the HP heater area will i e 1.7°F to 1263.7F The increase in feeda mp room ambient temperature due to the higher EPU feedwater temperature is 7.6 0 F to 112.6 This res aF. s achieved using both design and empirical information.

The increase in condensate pump room amDttemperature due to the higher EPU feedwater temperature is 3.5 'F. The ambient temperature in the conden ump room at EPU based upon design information 0 0 is 113.2 F and 122.5 F based upon empirical data.

It is noted that the temperatures obtained in this calculation are conse maximum temperatures for the purposes of obtaining bounding temperature increases within the subject ar Actual EPU maximum temperatures are anticipated to be lower than those calculated.

BVY 05-072 Docket No. 50-271 Exhibit SPSB-C-52-6 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Operation Procedure OP- 2192, Rev. 31 Total number of pages in this Exhibit I (exciudino this cover sheet) is 5. l

0 VERMONT YANKEE NUCLEAR POWER STATION OPERATING PROCEDURE l OP 2192 REVISION 31 IEATING, VENTILATING, AND AIR CONDITIONING SYSTEM USE CLASSIFICATION: REFERENCE I RESPONSIBLE PROCEDURE OWNER: Manager, Operations REQUIRED REVIEWS Yes/No E-Plan 10CFR50.54( ) No Security IOCFR50.54(p) No Probable Risk Analysis (PRA) I No Reactivity Management No LPC Effective Affected Pages

-No. Date 1 10/26/04 App A Pg 15 of 24 2 11/16/04 51 & ADDED 51A of 62 3 01/22/05 60 of 62 4 05/26/05 29 of 62 I Implementation Statement: N/A Issue Date: 07/06/2004 OP 2192 Rev. 31 Page 1 of 62

TABLE OF CONTENTS PURPOSE. 3 DISCUSSION ................ 3 ATTACHMENTS ................ ;

QA REQUIREMENTS CROSS REFERENCE ................................ 6 REFERENCES AND COMMMENTS ................................ 7 PRECAUI-IONS9LIATIONS ................................ 9 PREREQ ISITES........................................................................................................................................... 1 PROCEDURE .... ;12 Startup, Shutdown and Transfer of Service Building HVAC ................................................. 12 Startup, Shutdown and Transfer of Turbine Building HVAC ................................................. 15 Startup, Shutdown and Transfer of Reactor Building HVAC ................................................. 25 Startup, Shutdown and Transfer of Radwaste Building HVAC ................................................. 31 Startup and Shutdown of Intake Structure IVAC ................................................. 33 Startup and Shutdown of Discharge Structure HVAC ............................................... 34 Startup and Shutdown of Switchyard House HVAC .34 Startup and Shutdown of Stearn Heat HVAC .37 Startup, Shutdown and Transfer of Advanced Off Gas Building HVAC .38 Operation of Control Room and Cable Vault Battery Room HVAC ............................................ ;.42 Operation of Chiller Units SCH-1 and 2 ............................................ 44 Cross-Connecting SAC-1 and 2 Chilled Water.Systems ............................................ 46 Startup and Shutdown of UPS-2A-AC1 HVAC ............................................ 48 John Deere Diesel Generator Ventilation Controls ............................................ 50 Startup and Shutdown of Turbine Deck Office HVAC ............................................ 51 Startup and Shutdown of Switchgear Room HVAC .......................................................................... 52 Loss of Control Room Ventilation (UND98080) ................................. 53 Local Manual Operation of EDG Room A(B) Exhaust Fan TEF-2(3) (Use VYOPF2192.01) (ER990738_01) .................................................. ; 53 Startup and Shutdown of Steam Heat to the CST .................................................. 55 SCH-1-1 and SCH-1-2 Reset After Low Pressure Trip ................................................. 56 Temporary Space Heating...........................................................................................................5......57 FINAL CONDiTIONS . 62......

OP 2192 Rev. 31 Page 2 of 62

The Turbine Lube Oil Room fire dampers are controlled by a local hand switch located outside the Lube Oil Room door.

The air is exhausted to the atmosphere by wall exhaust fans, roof exhaust fans or through the station stack. Several areas, such as the Control Room and office spaces recirculate air on a continuous cycle.

The system serving the Control Room is designed to provide cooling during the summer and heating during the winter. Air is circulated through a chilled water cooling coil, steam preheat coil, a steam reheat coil and ductwork by one of two system fans. Fresh air is normally drawn into the system mixing with the recirculated flow. A humidistat, on the west hallway wall in the first fan room, controls the relative humidity between 20% and 50% with a humidifier located in SRHC-1. It is operated with instrument air and controls humidity by spraying steam into the air flow. Two exhaust fans in the North wall of the Control Room, kitchen and rest room, serve to exhaust these rooms. The "Control Room H and V" switch on CRP 9-25 is provided to allow the operator to isolate the Control Room and Computer Room by closing the fresh air dampers and the Control Room kitchen and bathroom exhaust vents during off normal conditions. This is accomplished by moving the switch from "NORMAL" to "EMER". If a report was received of a toxic gas release which could affect Control Room personnel, the operators would don the self-contained breathing apparatus located in the Control Room. If Control Room cooling is completely lost mergency cooling can be initiated using portable smoke ejectors. (UIND98080)

The Control Room and Service Building chilled water cooling coils are located in the SAC-1 and SAC-2 supply fan housings respectively. The cooling coils are cooled by dual compressor refrigeration units SCH-1 and 2 that cycle as necessary to maintain chilled water temperature. Dernineralized water from the chilled water pumps circulates though the chiller heat exchanger and gives off its heat to the chiller units. The cooled water passes from the chiller units to the cooling coils. The amount of chilled water flowing into the cooling coils is controlled by the mixing valves. Each mixing valve is controlled by a thermostat that senses supply air temperature. The Services Building chilled water piping system can be valved into the Control Room HVAC because both the NNS piping system and the current SC3 piping system, respectively, were designed and built to the same specifications (i.e., non-seisrnic), therefore, failure mechanism(s) are the same for each system. In addition, isolation of the NNS and SC3 systems can be accomplished because the valves are in a mild environment area. Continued operation of the Control Room HVAC by valving in the Service Building HVAC to supply chilled water is consistent with the Safety Class Manual and the HVAC DBD.

Air for air compressor cooling is drawn through a wall louver located in the outside wall.

This louver also supplies any required room ventilation air: Discharge dampers which exhaust into the room are located on the air compressor discharge duct to allow for air recirculation.

Two oil-fired steam boilers supply steam for the heating coils and some of the unit heaters. Other unit heaters are run electrically.

All RRUs and TRUs utilize service water as the cooling medium except the drywell RRU-1 through RRU-4 which use RBCCW. (M000D9502_14)

OP 2192 Rev. 31 Page 4 of 62

7. For minimum ECCS RRU corner room support operation, RRU 7 must be operable for the Northeast corner room and RRU 8 must be operable for the Southeast corner room.

RRUs 7 and 8 can be removed from service for maintenance and the associated Core Spray Pump/RHR Pump may remain operable. Refer to OP 2181, Service Water/Altemate Cooling Operating Procedure, Precautions and Limitations, for administrative requirements and actions necessary to maintain operability.

8. SP-1, SCH-1 and SAC-1 supply air conditioning for the Control Room. This is a Safety Class 3 system and requires special consideration for its timely repair.

9. To prevent the possibility of reverse air flow, building ventilation should be secured in the following order:

a. Radwaste Building

b. Turbine Building

c. Reactor Building

d. Service'Building

10. Minimize the time that TEF-2/3 are operated in manual. The UFSAR specified minimum Turbine Building design temperature, which includes the DG rooms, is 50QF and the AS-2 battery load calculation assumes an electrolyte temperature of 500F.

(ER960055_01)

11. Securing TRU-5 with the condensate pumps running will result in a condensate pump bearing temperature rise. Planned maintenance on TRU-5 should be coordinated such that the time the unit is out of service is minimized.

12. Control Room temperatures in excess of 780 F are indicative of a need for corrective action. Corrective actions need to be completed prior to exceeding 95°F to ensure the Control Room does not reach 120°F (upper temperature operability limit for Conto /

Room instrumentation). (U]ND98080)/

13. HS-139 and 115-140 in the Reactor Building must remain locked closed during plant operation due to House Heating Steam, High Energy Line Break concerns. These valves may only be opened with the permission of the Design Engineering Department.

(ER96-0482, TAGREVXEWL9703-26)

14. The Main Station Battery Room must be maintained at >600F. The main station battery load calculations are based upon this minimum temperature.

15. One of the two standby gas treatment trains should be placed in service whenever normal Reactor Building ventilation is secured.

16. In order to meet environmental qualification program requirements, the RCIC room fan or alternate ventilation from the Reactor Building must be operable and the RCIC room temperature must be less than 112 0F. However, to satisfy station blackout analysis, the RCIC Room temperature must be maintained less than 104 0F. (EPCQ9504)

OP 2192 Rev. 31 Page 10 of 62

2. Shutdown

a. Secure the West S oom exhaust fan SWGR-EF-1A.

b. erature cannot be maintained in the required range, notify the Operations Manager of the need to initiate actions to provide supplemental Q. Loss of Control Room Ventilation (UND98080)

1. If a loss of normal Control Room ventilation occurs, refer to Section J and place SAC-1B in service.,

2. If Control Room cooling is lost, perform applicable action:

a. Refer to Section L and cross-connect chilled water from the Service Building to the Control Room.

b. Perform the following:

1) In the Control Room back panel area, remove 11 full size ceiling tiles.

2) Open Control Room panel doors.

3) Notify Security and Shift Engineer that Control Room doors will be opened.

4) Open Control Room doors. /

5) Using two smoke ejectors or other portable cooling equipment to create temporary air flow paths, ventilate Control'Room./

6) Implement the administrative requirements of AP 0077.

R. Local Manual Operation of EDG Room A(B) Exhaust Fan TEF-2(3) (Use VYOPF2192.01) (ER99073801)*/

1. Ensure the selected EDG Room temperature RATS-1A(B)].

2. Obtain Shift Manager permiss manual operation of TEF-2(3).

3. Obtai ep ;

4. on a dedicated Auxiliary Operator at the selected Diesel Generator Room OP 2192 Rev. 31 Page 53 of 62

BVY 05-072 Docket No. 50-271 Exhibit EMEB-B-138-1 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Prediction of Unsteady Loading on A Circular Cylinder in High Reynolds Number Flows using Eddy Simulation.

l Total number of pages in this Exhibit l (excluding this cover sheet) is 9.

Proceedings of OMAE 2005:

2 4 h International Conference on Offshore Mechanics and Arctic Engineering June 12 - 16, 2005, Halkidiki, Greece OMAE 2005-67044 PREDICTION OF UNSTEADY LOADING ON A CIRCULAR CYLINDER IN HIGH REYNOLDS NUMBER FLOWS USING LARGE EDDY SIMULATION Sung-Eun Kim L. Srinivasa Mohan Fluent Incorporated, Lebanon, N.H. Fluent India 03766, U.S.A. Pune, India ABSTRACT of large-scale turbulent structure around a circular cylinder, which in turn affects unsteady loading on the cylinder.

Large eddy simulations were carried out for the flow Apparently, turbulence modeling is an issue here.

around a hydrodynamically smooth, fixed circular cylinder at There are largely three approaches being explored by CFD two Reynolds numbers, one at a subcritical Reynolds number practitioners to modeling high-Re turbulent flows around (Re = 1.4 x 105) and the other at a supercritical Reynolds circular cylinders and bluff bodies in general. One approach number (Re= 1.0 x 106). The computations were made using a employs unsteady Reynolds-averaged Navier-Stokes (URANS) parallelized finite-volume Navier-Stokes solver based on a equations supplemented by turbulence models as the governing multidimensional linear reconstruction scheme that allows use equations. URANS-based approach, despite its low of unstructured meshes. Central differencing was used for computational cost mainly due to less demanding mesh discretization of both convection and diffusion terms. Time- resolution requirement, is fundamentally ill equipped to capture advancement scheme, based on an implicit, non-iterative large-scale turbulent structure present in the flows. Results fractional-step method, was adopted in conjunction with a obtained using URANS computations typically underpredict three-level, backward second-order temporal discretization. the amplitudes of fluctuating forces. Obviously, this Subgrid-scale turbulent viscosity was modeled by a dynamic shortcoming has a serious negative implication to accurate Smagorinsky model adapted to arbitrary unstructured meshes prediction of VIV.

with the aid of a test-filter applicable to arbitrary unstructured Fundamentally more viable than URANS-based approach meshes. The present LES results closely reproduced the flow for bluff-body flows is large eddy simulation (LES). LES is features observed in experiments at both Reynolds numbers. designed to directly resolve large eddies, with the effects of The time-averaged mean drag coefficient, root-mean-square unresolved smaller eddies, namely subgrid-scale turbulence, on force coefficients and the frequency content of fluctuating the resolved large eddies accounted for by turbulence models.

forces (vortex-shedding frequency) are predicted with a However, LES is computationally very expensive, often commendable accuracy. prohibitively, especially when thin turbulent boundary layer is a predominant feature of the flow in question to be accurately INTRODUCTION resolved in a LES. Resolving near-wall turbulence with infinitesimal length and time scales requires extremely fine Unsteady loading on a circular cylinder caused by the flow mesh and small time-step size. For that matter, LES is more is a precursor to its vortex-induced vibration (VIV). The major feasible for free flows such as jets, mixing layer, and wakes, difficulty of computing flows around circular cylinders and massively separated flows.

originate from the fact that circular cylinder flows of practical There are very few LES studies published in the literature interest are high Reynolds number (Re) flows, often involving a of flows around circular cylinders at high Reynolds numbers.

laminar-to-turbulent transition in various regions of the flows Breuer [1] was the one of the very few who tackled the case of such as boundary layer, separated shear-layer, or near-wake. a high subcritical Reynolds number, Re = 1.4 x 105, at which The location of transition and the extent of laminar or turbulent the experiment of Cantwell and Coles [2] was conducted.

flow regime, in real flows, depend on such factors as Reynolds Using a multi-block, structured-mesh-based finite-volume number, freestream turbulence, surface roughness, span- solver with an explicit time-marching scheme, Breuer was able diameter ratio (LID), and blockage, among others. The state of to predict the global parameters of the flow, and the mean flow the flow in those regions dictates the formation and evolution and turbulence with a commendable accuracy. More recently, I Copyright © 2005 by ASME

Catalano et al. 13] attempted LES for three Reynolds numbers triangular, tetrahedral, pyramidal, prismatic, and hybrid in critical to supercritical Reynolds regime (Re = 0.5 x 106, 1.0 meshes. The solution gradients at cell centers that are needed x 106, 2.0 x 106). Their predictions of the global flow to compute convective and diffusive fluxes are determined parameters were in reasonable agreement with the experimental applying the Green-Gauss theorem [7]. Diffusive fluxes are data, although the skin-friction and the Reynolds number- discretized using central differencing. Discretization of dependency of the mean drag coefficient were poorly predicted. convective fluxes requires great caution in LES. Upwind-

- A more recent trend in turbulence modeling of bluff-body biased schemes such as second-order upwind scheme have been flows is to employ what may be called "hybrid" approaches in widely used for RANS computations. Unfortunately, numerical which one attempts to adjust turbulence models to local mesh diffusion introduced by upwind schemes, which might be resolution in one way or another. In one such approach acceptable in RANS computations for high-Re flows where referred to, in the literature, as detached eddy simulation eddy-viscosity is larger than molecular viscosity by orders of (DES), either a RANS-based or a subgrid-scale (SGS) magnitude, can overwhelm physical diffusion that is typically turbulence model is invoked depending on whether or not the much smaller in LES; subgrid-scale turbulent viscosity is filter-length (local mesh size) is larger than the local integral smaller than RANS-based eddy viscosity by orders of length-scale of turbulence. In DES, turbulence models magnitude. For this reason, a central differencing scheme [13]

essentially reduce to RANS models in near-wall region or when was employed for the discretization of convection terms in this the local mesh size is too coarse to explicitly resolve energy- study.

containing eddies. One fundamental criticism about DES comes down to the lingering question of how one can possibly TlME-ADVANCEMENT SCHEMES reconcile a RANS turbulence model and a subgrid-scale turbulence model, two very different models in terms of their An implicit fractional-step method. (FSM) [8] in spectral content, at the common interface. It should also be combination with a second-order accurate, three-level realized that, as a consequence of falling back to a RANS backward-differencing scheme for time-discretization was model in near-wall region, the fidelity of DES would be solely employed to advance the solution in time. In this algorithm, the determined by that of the embedded RANS model. Quite a few momentum equations are decoupled from the continuity studies employing DES have appeared recently on circular equation using an approximate factorization of the coupled cylinder flows. Among others, Travin et al.[4], Vatsa and Navier-Stokes equations. For incompressible flows, the FSM Singeri5], and more recently Pandya et al. [6] all employed preserves the formal second-order temporal accuracy without DES based on an eddy-viscosity transport model to study the having to perform, at each time-step, costly outer iterations to flow around a smooth *circular cylinder at subcritical and couple velocity and pressure. The FSM thus provides a highly supercritical Reynolds numbers. The results of these DES are efficient algorithm for CPU-intensive transient computations mixed, insofar as some of the global flow parameters such as like LES.

the r.m.s. force coefficients and the mean flow in the near-wake Consider a semi-discrete form of the Navier-Stokes were predicted poorly. equations in "pressure-correction" (&pI = p -+1p') form, The objective of this study is to assess the feasibility of LES for high Reynolds number flows around a fixed, smooth circular cylinder. To that end, two Reynolds numbers, one in' A G (u-1,)=(0r)

(I) subcritical (Re = 1.4 x 105) and the other in supercritical (Re = ID 0] rp-+

1.0 x 106) regime were deliberately chosen so that they straddle the critical Reynolds number (- 3 x 105). The main concern is where ut? and r are the velocity vector and the momentum how well LES can reproduce the known characteristics of the source vector, respectively, and the integer n the time level, flow at the two Reynolds numbers and the changes in the global with n+1 refers to the current time level. A is the coefficient flow parameters such as drag coefficient, r.m.s. force matrix defined in terms of the discrete convective and diffusive coefficients, and vortex-shedding frequency as the Reynolds operators and the time-advancement scheme chosen, and G and number increases. D the discrete gradient and divergence operators, respectively.

The results of the LES will be compared whenever The coupled system of equations in Equation (I) is possible, with the experimental data and other LES and DES extremely difficult to solve as it stands, since the matrix A has results reported in the literature. to be inverted for every iteration. In the FSM, the original coupled equations in Equation (I) are approximated by NUMERICAL METHOD The computations were carried out using the segregated solver in FLUENT, a general-purpose CFD software. The EA 0G][1J ID AtDG[ 0

&~G ( u.,(r')+[O(At2A 1 ] 8p' oL~J (2) details of the numerical method are described in the following.

Factorizing equation (2), we have a series of split operations SPATIAL DISCRETIZATION SCHEMES that consist of FLUENT employs a cell-centered finite-volume method A6 = -G P' +r'"

based on a multidimensional linear reconstruction scheme that AtDG p'= Dii (3) permits use of computational elements (cells) with arbitrary polyhedral topology, including quadrilateral, hexahedral, UnI =U6 - Aicfp-1f.

2 Copyright Q 2005 by ASME

On a per-iteration basis, the series of operations in Tu - 3Tlk -=~-2C, PijS (5)

Equation (3) closely resemble the SIMPLEC scheme. The difference is that, in the iterative SIMPLEC scheme, the series of operations in Equation (3) are repeated in a loop until the We determine the model constant, C,, using the dynamic solutions converge, while the FSM needs only one sweep. To Smagorinsky model (DSM) originally proposed by Germano et preserve second-order accuracy with the FSM, however, sub- a]. [12]. To implement the dynamic procedure for the present iterations are needed for the set of three momentum equations finite-volume solver requires a special test-filter applicable to and individual scalar equations to account for the nonlinearity arbitrary unstructured meshes. The dynamic procedure employed in and coupling among the individual equations and high-order in the present study is "local" in the sense that it does not require source terms. Yet, the non-iterative FSM is takes much less the existence of any statistically homogeneous directions, being CPU time than the iterative SIMPLEC scheme. applicable to arbitrary complex three-dimensional flows. The The system of discretized governing equations are solved details of the implementation of the DSM in the framework of using a point-implicit, Gauss-Seidel relaxation along with an unstructured mesh based finite-volume solver can be. found in algebraic multigrid (AMG) method to accelerate solution Kim [13]. The DSM has been validated for a number of wall-convergence: The solver and the subgrid-scale turbulence bounded flows such as a square cylinder and a sphere.

model are fully parallelized.

As a validation for the spatial discretization schemes and the transient algorithm, we computed laminar flow around a SOLUTION DOMAIN AND COMPUTATIONAL MESH circular cylinder at the Reynolds number of 100 with both non-iterative FSM and iterative SIMPLEC scheme. At this Our choices of the solution domain and the computational Reynolds number, the flow exhibits a periodic vortex-shedding. mesh were guided by the findings in Breuer's LES study [1].

A C-type structured mesh with 48,000 cells was used for the What one should keep in mind in determining the size of the computation. A dimensionless time-step size of At = 0.04 (non- solution domain is that the spanwise extent of the domain dimensionalized by DIUs D being the cylinder diameter, U0 the should be larger than the spanwise correlation length of freestream velocity) was used, with which one period of the turbulence. Fortunately, the spanwise correlation length is vortex-shedding was resolved with approximately 150 time- known to decrease as the Reynolds number increases. We took advantage of this fact, having decided to use a spanwise length steps. The predicted mean drag coefficient (ED), r.mis. lift of 2.OD for both the subcritical and the critical Re numbers.

coefficient (c,). and Strouhal number (So) are summarized in The top and bottom boundaries of the domain were placed at Table 1. The FSM gives practically the same predictions as the 10.5D from the cylinder axis. Thus, the blockage ratio of our iterative -SIMPLEC scheme, and the global parameters numerical model (HID, where H is the height of the domain) is predicted by both methods are seen to closely match the other approximately 4.8 %. The upstream inlet and the downstream predictions and the experimental data. exit boundaries are located at 8.5D upstream and 20.5D downstream of the cylinder axis, respectively.

A locally refined, hexahedral mesh with a total of 6.8 Table 1. Summary of the global flow parameters predicted million cells was used for the computations for both Reynolds for larminar flow (Re = 100) - Norberg [101 compiled numbers. The overall mesh resolution is comparable to case numerical results whose references can be found in ref. [10] "Bl" in Breuer [1]. The near-wall mesh resolution is such that the distance from the cylinder surface is 104D at the wall-CD C., St adjacent cells, which translates to y;(=u~y/1 well under 1.0 for Present - FSM 1336 0.28 0.165 the subcritical flow, and below 6.0 for the supercritical flow.

Present - SIMPLEC 1.345 0.28 0.166 For the subcritical flow, the boundary layer was resolved with 10-18 cells in the laminar region.

Park et al. [9] 1.33 0.33 0.165 Norberg [10] - 0.2310.29 - BOUNDARY CONDITIONS. AND OTHER DETAILS OF Williamson [t1] - - 0.164 COMPUTATION On the cylinder wall, we employ a law-of-the-wall that SUBGRID-SCALE TURBULENCE MODELING invokes proper wall-laws depending on the mesh resolution, namely, ye at the wall adjacent cells. The top and bottom wall For incompressible flows, the filtered Navier-Stokes boundaries were treated as a slip (zero-stress) wall. A equations can be written as periodicity condition was applied on the pair of lateral boundaries in the span-wise direction. At the upstream inlet boundary, the freestream condition was specified. At the I ap "} a (4) downstream exit boundary, we extrapolated the solution at ax, p aIx, ax, ax),I alx,~ variables from the adjacent interior cells in a mass-conserving manner.

where = u,u 1 -UIFJiis the subgrid-scale turbulent stress. In The time-step size (at) used for the computations is 0.005 this study, the subgrid-scale (SGS) stress is modeled using in a dimensionless unit for both Reynolds numbers. The time-isotropic eddy-viscosity as step size was determined based on the estimate of the 3 Copyright ©2005 by ASME

characteristic length and time scales of the smallest resolved eddies, r=t/u', where I was taken as 0.05D, and u as Table 2. Summary of the global flow Sparameters predicted by 0.2UQ From these rough estimates and with the dimensionless the present LES for Re = 1.4 x 10, compared with other time-step size of 0.005, one turnover time of the smallest numerical results (Breuer [1]; DES-TS - Travln el aL.[4J; DES-LS

- Travln el aL[4]; and experimental data (CC - Cantwell &

resolved eddies will be resolved with 50 time steps. Coles[2], WA - West & Apelt [14], SB: Szepessy & Bearman [15],

To shorten the initial transient period of the solution and to NO - Norberg [10J, ZD: Zdravkovlch [16])

quickly attain a statistically stationary state, we took, as the initial condition, a partially converged steady RANS solution with pseudo-random fluctuating velocity components E. CL -Co tr St superimposed on the mean velocity field. LES (Present) 1.13 0.59 1.20 0.67 0.205 The computations were carried out on a 16-CPU 1nx86 LES (Breuer) 1.45 - 1.76 0.34 0.204 cluster with AMD Opteron processors and Infiniband DES-TS 0.59 0.06 0.67 1.2 0.31 interconnects. With the non-iterative FSM, the computation DES-LS 1.08 0.29 1.04 1.1 0.2i took approximately 12 CPU-seconds per a time-step, which Exp.(CC) 1.24 - 1.21 0.5 0.18 translates into being able to complete roughly 7,200 time-steps Exp.(WA) 1.3 0.58 - .

in 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

Exp.(SB) 1.35 0.5 1.05 Exp.(NO) - 0.52 - - 0.18/0.195 RESULTS Exp. (ZD) 1.15 0.5/0.6 1.05/1.2 - 0.18/0.21 Subcritical Reynolds number (Re = 1.4 x10 5)

Figure I depicts the flow structure around the cylinder. The present LES gave the mean CD around 1.13, which Although the near-wake flow appears chaotic, the figure clearly falls within the scatter of the experimental data (1.1 - 1.35) for shows the existence of a large-scale motion, namely alternate smooth cylinders. Interestingly, our prediction came out vortex-shedding known to occur in the subcritical Re number substantially lower than Breuer's prediction despite the range. The color map of the resolved turbulent kinetic energy comparable mesh resolution, the same formal order of spatial.

in Figure I also indicates that the boundary layer remains accuracy (second-order) and the same dynamic SGS turbulence laminar up to the separation point, and transition occurs in the model used in both computations.

separated shear layer.

Figure 1. The flow structure behind the circular cylinder at Re = 1A X 105 - Iso-surface of the second invariant of Non-dimensional lime, to (= tUdD) velocity deformation tensor, colored by the resolved turbulent kinetic energy Figure 2. Time histories of the drag and lift coefficients predicted by the present LES for Re = 1.4 X 105 The global flow parameters predicted by the present LES are summarized in Table 2 along with the LES prediction of The DES-TS result [4] is shown to severely underpredict Breuer [I] and the DES predictions of Travin et a!. (4], and the the mean CD, better matching the data obtained with rough experimental data. . As for Breuer's LES result, among the cylinders [17], which is consistent with the DES predictions by cases involving different mesh resolution and SGS turbulence others [5, 6]. It has been found that DES, run in a normal models, we picked the result of case "Bl" whose mesh and mode with finite freestream turbulence, essentially yields a SGS turbulence model (dynamic Smagorinsky mode) closely tripped boundary layer, leading to a premature transition to match ours. In selecting the experimental data for the turbulence and an early drag crisis even at a subcritical comparison, we gave more weight to the ones measured on Reynolds number. The DES-LS result [4] based on the so-smooth cylinder surface, and with sufficiently large span-wise called "tripless" approach, in which laminar flow is essentially lengths (LID > 5) and smaller blockage-ratio (IDD < 0.1). enforced in the boundary layer, better predicts the mean CD.

(f 0

U)

--6.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 x/D St Figure 3. Power spectral density of CL time-history for Re Figure 5. Time-averaged axial velocity profile in the wake

1.4x O1 predicted by the present LES for Re = 1.4 x 105 The rum.s. lift coefficient predicted by the present LES is Figure 5 depicts the predicted time-averaged axial velocity within the scatter of the experimental data, although it is closer profile along the wake centerline at the mid-span. The overall to the upper bound. The CL history Figure 2 clearly shows agreement between the prediction and the measurement is quite that, despite modulation of the amplitude, there is a distinct good. The rapid relaxation of the mean axial velocity in x/D

frequency of vortex-shedding, which is consistent with what 1.0 - 3.0 shown in the measurement is an indication of has been known for smooth cylinders in the subcritical regime. vigorous mixing of momentum taking place in the near-wake The Strouhal number corresponding to the main shedding region. That our prediction closely captures the recovery of the' frequency was found to be 0.205 as shown in Figure 3. mean axial velocity verifies that the dynamics of the large-scale structure in the near-wake is well predicted in the present LES.

2.0 The length of the time-averaged recirculation bubble (see also Table 2 for 4,) from our LES was found to be slightly larger than what Cantwell and Coles [2] reported. Yet the present

-Present LES prediction is much closer to the data than the DES 1 c Exp. (Cantwell &Coles. 1983) predictions. Considering that hot-wire measurements become a Exp. (Sarioglu &Yavuz, 2002) increasingly difficult and less reliable near the recirculation region, our LES prediction is considered quite good.

c ) 0.0 ~\

l----- Upper surfacel 5.0- Lower surface 0 (deg.)

Figure 4. Mean static pressure (Cp) distribution on the cylinder surface for Re = 1.4 x 105 The time-averaged pressure distribution on the cylinder surface is shown in Figure 4, along with the experimental data

[2, 18] obtained at the same Reynolds number. The negative . .I . . I . . I . I . . I . .

peak predicted by the present LES is larger than the measured 0 30 60 90 120 150 180 0 (deg.)

ones. However, the pressure level in the separated region - the plateau after e = 90 degrees, and the mean base pressure (Cp,) Figure 6. Scaled Instantaneous skin-friction coefficient are all accurately predicted by the present LES. distribution on the cylinder surface at Re = 1.4 x IO' 5 Copyright 0 2005 by ASME

peaks at 3.0 around 0= 50 degrees, which is closely reproduced uV.W I I I I I I I I E - I l I I by the present LES.

0.004 - X l-- Uppe surface I-Low isuface] Supercritical Reynolds number Case (Re = 1.0 xl10' 0.003 \

0.002 0.001 ,

r . I "0 30 60 90 320 150 180 0 (deg.)

Figure 7. Resolved turbulent kinetic energy distribution at 0.0001D above the cylinder surface, predicted the present Figure 8. The flow structure behind the circular cylinder at LES at Re = 1.4 x 105 Re = 1.0 X 106 - an iso-surface of the second invariant of velocity deformation tensor, colored by the resolved As mentioned earlier, the experimental evidence [19] turbulent kinetic energy indicates that, in the subcritical regime, the boundary layer remains laminar until it finally separates from the cylinder surface. Therefore, it would be of much interest to see how Figure 8 portrays an impression of the instantaneous flow closely LES can reproduce the physics associated with the structure. Compared to the flow structure for the subcritical separation and the transition. Figures 6 and 7 depict the flow shown in Figure 1, the wake is much narrower and more circumferential variations of two quantities along the cylinder chaotic without any trace of a large-scale alternate vortex surface that shed some light on how the LES predicts the shedding, suggesting a delayed turbulent separation. This is separation and transition. typical of the flow in the supercritical regime. The color map Figure 6 shows a plot of the instantaneous skin-friction on the iso-suraface - representing resolved turbulent kinetic coefficient - rescaled to match the non-dimensionalization energy - also indicates that the laminar boundary layer is adopted by Achenbach [19] - against the circumferential angle. sustained farther downstream than the subcritical flow.

First, it is noticeable that the skin-friction distribution exhibits a Table 3. Summary of the global quantities predicted by LES for considerable top-bottom asymmetry. This is a clearly the Re = 1.0 x 10', compared with other LES result and the (upstream) influence of the large-scale, alternate vortex- experimental data; CA - Catalano et al.13]; SZ - Szechenyl.120];

shedding. What is most interesting in the figure is the SH - Shih et al. (211; ZD - Zdravkovich [16])

appearance of small transient separation bubbles on both upper and lower surface starting as early as 70 - 75 degrees. CD C2 - U St Interestingly, this range largely coincides with the range of LES (present) 0.27 0.12 0.28_

separation angles reported by many others [2, 19] deduced from LES (CA) 0.31 - 0.32 0.35 the inflection point of mean Cp curve; 77 degree by Cantwell Exp.(SZ) 0.25 and Coles [2], and 78 degrees by Achenbach [19] at Re = 1.0 x Exp.(SH) 0.24 0.33 105. Although not shown here, the separation angle determined Exp. (ZD) 0.210.4 0.1/0.15 0.2/0.34 0.18/0.5 based on the time-averaged velocity field obtained from the present LES was found to be much larger than the ones determined from Cp distribution, reaching around 98 degree, The global flow parameters predicted by the present LES which was also found by Breuer [1]. Based on the are summarized in Table 3. Our prediction of the mean drag instantaneous skin-friction distribution, we surmise that a coefficient (U.) closely matches the data of Szechenyi [20] and transient laminar separation occurs much earlier than the mean Shih et al. [211 interpolated at the present Reynolds number.

separation angle. We consider their data more reliable than others in view of the The distribution of the resolved turbulent kinetic energy sufficiently large spanwise lengths used (LID = 9.3 and 8.0, depicted in Figure 7 supports the possibility of an intermittent respectively), relatively low blockage ratio (7% and 11%,

laminar separation around 70 - 75. degrees, insofar as the respectively), surface smoothness; and relatively low turbulent kinetic energy is still quite low there. The complete turbulence intensity of the wind tunnels - 0.08 % for Shih et transition occurs a little downstream as indicated by the rapid al.'s data [21]. The r.m.s. lift coefficient prediction by the increase in the resolved turbulent kinetic energy between 75 present LES falls within the scatter of the experimental data and 85 degrees. The magnitude of the skin-friction coefficient compiled by Zdravkovich [16]. The base pressure is also seen predicted by the present LES is also in good agreement with to be in the range of the experimental data.

Achenbach's data [19] measured at Re = 1.0 x 10'.

.. .1 As before, we plotted the circumferential variations of the instantaneous skin-friction and the resolved turbulent kinetic energy in Figures I I and 12. The top-bottom asymmetry of the skin-friction distribution is seen to be much less pronounced than for the subcritical flow. Figure 12 suggests that the boundary layer sustains being laminar down to 90 degrees. The separation angle of the time-averaged velocity field from the present LES was found to be around 108 degrees, which is close to 106 degrees estimated by Shih et al.[21] using their experimental data on a smooth cylinder at approximately the same Reynolds number. All this taken together, our LES result suggests that the boundary layer undergoes transition starting as early as 90 degrees, before it finally separates - "turbulently" -

at around 108 degrees.

0 420 440 460 480 Non dmensional time, t* IUO/D)

Figure 9. Time histories of drag and lift coefficients for the 1---Upper surfacel circular cylinder at Re = 1.0 x I0' 5.0 - Lower surface The time histories of drag and lift coefficients are shown in Figure 9. Consistent with the flow structure portrayed in Figure 8, the CL-history at this supercritical Reynolds number is more U 0.

random than that for the subcritical flow. Indeed, no single frequency of vortex-shedding could be identified at this Reynolds number, as shown in the plot of the power spectral density of CL in Figure 10. This finding is consistent with the others' experimental observations, for instance, Shih et al. [21]

who found that coherent vortex shedding disappeared on a smooth cylinder beyond Re = 4 x 105. 0 30 60 90 120 150 180 e (deg.)

l0- ,,-1, I I I- . I - I an 7 - -Tr- 1 I ' Figure 11. Scaled instantaneous skin-friction coefficient distribution on the cylinder surface at Re =1.0 x 106 10 4

.1 0.

106 it

-4 . I °. I . I -I . I . I .A I Au 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Strouhal Number Figure 10. Power spectral density of the lift coefficient (CL) for the critical Reynolds number (Re = 1.0 x 106) 0 (deg.)

For a smooth cylinder, the experimental data collected by Figure 12. Resolved turbulent kinetic energy distribution at Zdravkovich [16] and Achenbach's data [19] indicate that, 0.0001D above the cylinder surface, predicted the present above the Reynolds number of 1.5 x 106 referred to as "TrBL4" LES at Re = 1.0 x 10 in [16], transition clearly precedes the boundary layer separation. The exact location of the transition depends on such factors as freestream turbulence level and surface The skin-friction prediction, including the location of the roughness. At Re = 1.0 x l06 falling in the range referred to as peak and the magnitude, is also in good agreement with the

'TrBL3", however, both sets of data seem 1o suggest that experimental data of Achenbach [19] measured at Re = 8.5 x transition occurs near the separation. 105 and Re = 3.6 x 106. At these two Reynolds numbers, 7 Copyright © 2005 by ASME

Achenbach's data show that, the scaled skin-friction University in Turkey for kindly allowing me to use his distributions have the maximum values in the range of 3.5 - 4.0 experimental data.

occurring around 0= 60 - 65 degrees. As shown in Figure I 1, our LES prediction closely matches Achenbach's data. This is REFERENCE in a sharp contrast with the LES prediction by Catalano et al.

[3]. which significantly overpredicted the overall level of the [I] Breuer, M., 2000. "A Challenging Test Case for Large skin-friction on the front half of the cylinder. Contrary to what Eddy Simulation: High Reynolds Number Circular is suggested by the experimental evidence, their LES seem to Cylinder Flow," Int'l J. Heat and Fluid Flow, 21, pp. 648 -

yield too early a transition, giving a skin-friction level typical 654.

of a turbulent boundary layer from immediately downstream of [2] Cantwell, B. and Coles, D., 1983, "An Experimental Study the front stagnation point. As they pointed out [3], one most of Entrainment and Transport in the Turbulent Near-Wake likely cause for the discrepancy is the substantially coarser of a Circular Cylinder," J. FluidMech., 136, pp. 321 - 374.

mesh (y around 50) used in their LES predictions. [3] Catalano, P., Wang, M., Iccarino, G. and Moin, P., 2003, "Numerical Simulation of the Flow Around a Circular

SUMMARY

AND CONCLUSIONS Cylinder at High Reynolds Numbers," Int'l J. Heat and Fluid Flow, 24, pp. 463 -469.

Turbulent flow around a smooth fixed circular cylinder [4] Travin, A., Shur, M., Strelets. M. and Spalart, P., 1999, was computed using large eddy simulation (LES) at two "Detached-Eddy Simulation Past a Circular Cylinder,"

Reynolds numbers (Re = 1.4 x 105, 1.0 x 106). The Flow, Turbulence aid Combustion, 63, pp. 293 - 313.

computations were carried out using a second-order accurate, [5] Vatsa, V. N. and Singer, B. A., 2003, "Evaluation of a parallelized finite-volume Navier-Stokes solver capable of Second-Order Accurate Navier-Stokes Code for Detached handling arbitrary unstructured meshes. An implicit, non- Eddy Simulation Past a Cricular Cylinder," AIAA Paper iterative fractional-step method was employed in favor of its 2003-4085.

high efficiency in LES for incompressible flows. [6] Pandya, M. J., Frink, N. T., Abdol-Hamid, K. S., and The LES results for both Reynolds numbers predicted, Chung, J. J., 2004, "Recent Enhancements to USM3D with a good accuracy, the global flow parameters such as mean Unstructured Flow Solver for Unsteady Flows," ATAA drag coefficient, r.m.s. lift coefficient, and the Strouhal number. Paper 2004-5201.

Furthermore, the salient features of the subcritical and the [7] Kim, S. E., Makarov, B., and Caraeni, D., 2004, "Multi-supercritical flows experimentally observed and measured on a Dimensional Reconstruction Scheme for Unstructured smooth cylinder, such as the separation angle, mean velocity in Meshes," AIAA Paper 2004-2548.

the wake, length of the recirculation bubble, and transition [8] Kim, S. E. and Makarov, B., 2005, "An Implicit location, are largely reproduced by the LES. Despite the use of Fractional-Step Method for Efficient Transient Simulation a large number of computational elements (6.8 million cells), of Incompressible Flows," To be presented at 17'h AIAA the solution turnaround time was quite reasonable. Computational Fluid Dynamics Conference, June 6 - 9, Toronto, Ontario.

To summarize what has been achieved through this study: [9] Park, J., Kwon, K., and Choi, H., 1998, "Numerical Solutions of Flow Past a Circular Cylinder at Reynolds

The spatial discretization method (finite-volume on Numbers up to 160," KSME Int. J., 12, pp. 1200 - 1205.

unstructured meshes) and the solution algorithm [10]Norberg, C., 2003, "Fluctuating Lift on a Circular (implicit fractional-step method) employed in this study Cylinder Review and New Measurements," J. Fluids and have been shown to be sufficiently robust and accurate Structures, 17, pp. 5 7 -96.

to be used in LES for high Reynolds number flows. [I I] Williamson, C.H.K., 1989, "Oblique and Parallel Modes of

The dynamic Smagorinsky model adapted to Vortex Shedding in the Wake of a Circular Cylinder at unstructured meshes using a novel test-filter [13] has Low Reynolds Numbers," J. Fluid Mech., 206, pp. 579 -

been shown to work robustly and accurately for 627.

complex, high Reynolds number turbulent flows. [12]Germano, M/, Piomelli, U., Moin, P., and Cabot, W. H.,

The present LES capability has been shown to predict 1991, "Dynamic Subgrid Scale Eddy Viscosity Model,"

the salient features of turbulent flow around a smooth Physics of Fluids A, 3, 19, pp. 1760 - 1765.

circular cylinder at both subcritical and supercritical [13]Kim, S. E., 2004, "Large Eddy Simulation Using regimes with a good accuracy and reasonable Unstructured Mesh and Dynamic Subgrid-Scale computational cost. Turbulence Models," AIAA Paper 2004-2548.

[14]West, G. S. and Apelt, C. J., 1993, "Measurements of It is planned to continue this work to include higher fluctuating Pressures and Forces on a Circular Cylinder in Reynolds numbers. the Reynolds Number Range 1 to 2.5 x 105," J. Fluids and Structures, 7, pp. 227 - 244.

[15]Szepessy, S. and Bearman, P. W., 1992, "Aspect Ratio and End Plate Effects on Vortex Shedding from a Circular ACKNOWLEDGEMENT Cylinder, J. Fluid Mech., 234, pp. 191 - 217.

[16]Zdravkovich, M. M., 1997, "Flow Around Circular The authors acknowledge that FLUENT and GAMBIT Cylinders, Fundamentals, Vol. 1, Oxford Univ. Press software were 'used for this study. The authors are also (Chapter 6).

thankful .to Dr. Mustafa Sarioglu at Karadeniz Technical 8 Copyright 0 2005 by ASME

L17]Roshko, A., 1961, "Experiments on the Flow Past a Circular Cylinder at a Very High Reynolds Number," J.

FluidMech., 10, No. 3, pp. 345 - 356.

[18]Sarioglu, M. and Yavuz, T., 2002, "Subcritical Flow Around Bluff Bodies," AIAA J. Vol. 40, No. 7, July.

[19]Achenbach, E., 1968, "Distribution of Local Pressure and Skin-Friction Around a Circular Cylinder in Cross-Flow up to Re = 5 x 106," J. FluidMech., 34, Pan4, pp. 625 - 639.

[20)Szechenyi, E., 1975, "Supercritical Reynolds Number Simulation for Two-Dimensional Flow Over Circular Cylinders," J. Fluid. Mech., 70, pp. 529 - 542.

1211 Shih, W. C. L., Wang, C., Coles, D., and Roshko, A., 1993, "Experiments on Flow Past Rough Circular Cylinders at Large Reynolds Numbers," J. Wind Eng. And Indus.

Aerodynamics, 49, pp.351 - 368.

BVY 05-072 Docket No. 50-271 Exhibit EMEB-B-138-2 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Large Eddy Simulation of Confined Swirling Coaxial Jets.

Total number of pages In this Exhibit (excludino this cover sheet) Is 10. I

Proceedings of ASME-FED:

2 nd Symposium on Measurements and Modeling of Large-Scale Turbulent Structures June 19 - 23,-2005 FEDSM 2005-77085 LARGE EDDY SIMULATION OF CONFINED SWIRLING COAXIAL JETS Sung-Eun Kim, Fluent Inc., Xuelei Zhu, Fluent Inc., Lebanon, Stefano Orsino, Fluent Inc.,

Lebanon, NH 03766, U.S.A. NH 03766, U.S.A. Lebanon, NH 03766, U.S.A.

ABSTRACT concentration were much less satisfactory. The major discrepancy was found in the core region, where the mean Large eddy simulations (LES) have been carried out for species concentration was grossly underpredicted by the RANS confined swirling coaxial jets discharged into a suddenly computations. Evidently, the RANS model employed gives expanded pipe, which was studied experimentally by Roback rise to an excessive mixing. Brankovic et al. [3] investigated and Johnson [1, 2]. The computations were made using a the sensitivity of their RANS predictions to such parameters as parallelized finite-volume-based Navier-Stokes solver that is the turbulent Schmidt number and the inlet turbulence level.

second-order accurate in time and space, and permits use of However, the influence of these parameters was negligibly unstructured meshes. The computational domain starts from an small. The poor prediction of the, species mixing has been inlet placed upstream of the swirl generator, which makes the attributed to the inability of RANS-based turbulence models to inlet boundary condition easy to specify. Subgrid-scale accurately represent the mixing by large-scale turbulent turbulent stresses were modeled using a dynamic Smagorinsky structure (or large eddies).

model applicable to complex three-dimensional flows without Playing a major role in mixing, large eddies are also any statistically homogeneous directions. Subgrid-scale responsible for undesirable yet somewhat subtler phenomena turbulent scalar flux is modeled using a constant Schmidt such as combustion instability and acoustic excitation inside number in conjunction with the dynamically computed subgrid- combustors. To address these issues properly, one has to turn scale turbulent viscosity. The LES predictions were found to to high-level turbulent simulation like LES. Akselvoll and closely reproduce the salient features of the flow and the Moin [5] used LES to compute the flow and the species species concentration downstream of the expansion. One of the concentration in non-swirling coaxial jets. Pierce and Moin [6]

conclusions was that a good resolution of the mean flow and attempted LES for the same confined coaxial swirling jet as is turbulence in the upstream region is crucial in accurately considered here. These studies demonstrated that LES can predicting the mixing downstream of the expansion. indeed predict the flow and the species concentration in confined coaxial jets with a commendable accuracy. LES also allows one to directly compute r.m.s. velocity fluctuations and INTRODUCTION r.m.s. species fluctuation which are important quantities in the context of turbulent combustion.

Confined swirling co-axial jets discharging into a suddenly The present study concerns a LES computation for the expanded pipe [1, 2] have been studied numerically by several swirling coaxial jets using a finite-volume solver that permits others using the Reynolds-Averaged Navier-Stokes (RANS) use of unstructured meshes. Unlike the study of Pierce and equations [3, 4] and large eddy simulation [6]. The flow in Moin [6], however, our computational domain starts from an question consists of a non-swirling jet in the center, and an inlet boundary placed upstream of the swirl generator. Despite outer annual jet with a swirl imparted by a 8-vane, variable- the larger solution domain and the implied increase in the angle swirl generator, with a swirl number of approximately computational resource, having the inlet boundary upstream of 0.41. the swirl generator makes it straightforward to specify the inlet Earlier numerical studies using RANS-based turbulence boundary conditions. Meshing the computational domain models [3, 4] have shown that the mean velocity field in the including the swirl generator, which could become a difficult mixing region - the shear layer between the jets and the inner and time-consuming task with structured meshes, is made recirculation region - was predicted with a reasonably good easier by the flexibility offered by unstructured mesh allowed accuracy. However, the predictions of the mean species by the present finite-volume solver.

This paper is, in many aspects, a progress report about an ongoing study whose ultimate goal is to find an optimal Ad = -GpT

+ r'"

strategy based on the technique of LES for modeling turbulent flow and species mixing in coaxial combustors with or without AtDGlp"' = Dai (3) swirl. uRA' =f&i- tG~p-'

The paper is organized as follows. First, we briefly discuss the numerical method and the subgrid-scale turbulence models On a per-iteration basis, the series of operations in for stresses and scalar flux. Special emphasis is laid on a Equation (3) closely resembles SIMPLEC scheme. The transient algorithm whose efficiency significantly benefits the difference is that, in the iterative... SIMPLEC scheme, the present study. This is followed by the details of the LES operations in Equation (3) are repeated in an outer loop until computations regarding the choices of the solution domain, the all the solution variables converge, whereas the FSM needs mesh (resolution), boundary conditions, time-step size, and the only one sweep. To preserve second-order accuracy with the overall solution strategy. The results are then presented. FSM, however, sub-iterations are needed for the set of three momentum equations and individual scalar equations to account for the nonlinearity in and coupling among the NUMERICAL METHOD individual equations and high-order source terms. Yet, non-iterative FSM is takes much less CPU time than iterative The computation were carried out using the segregated SIMPLEC scheme.

solver in FLUENT, a general-purpose CFD software. The system of discretized governing equations are solved FLUENT employs a cell-centered finite-volume method based using a point-implicit, Gauss-Seidel relaxation along with on a multi-dimensional linear reconstruction scheme that algebraic multi-grid (AMG) method to accelerate solution permits use of computational elements (cells) with arbitrary convergence. The N-S solver and the SGS turbulence model polyhedral cell topology including quadrilateral, hexahedral, are fully parallelized.

triangular, tetrahedral, pyramidal, prismatic, and hybrid meshes. The solution gradients at cell centers that are needed to compute convective and diffusive fluxes are obtained by SUBGRID-SCALE TURBULENCE MODELING applying Green-Gauss theorem [9]. Diffusive and convective fluxes are discretized using central differencing [8]. For incompressible flows, the filtered Navier-Stokes An implicit fractional-step method (FSM) [10] in equations can be written as conjunction with a second-order accurate, three-level backward- differencing scheme for time-discretization was employed to advance the solution in time. In this algorithm, the momentum equations are decoupled from the continuity axT at +a,."T, _ lap__ ar +_a a a.i1V_ . (4) equation applying an approximate factorization of the coupled at axJ- pax, axi-X- IxaX}J Navier-Stokes equations. For incompressible flows, the FSM preserves the formal second-order temporal accuracy without where T. =UJUJ -j#U is the subgrid-scale turbulent stress. In having to perform, at each time-step, costly outer iterations to this study, the subgrid-scale stress is modeled using isotropic couple velocity-field and pressure. The FSM thus provides a eddy-viscosity as highly efficient algorithm for CPU-intensive transient computations like LES. 6 Consider the semi-discrete form of the Navier-Stokes Tu~ 38X r =-2CR~MU, (5 equations in "pressure-correction" (p"Z = pn" -pn) form:

We determined the model constant, C,, using the dynamic Smagorinsky model (DSM) originally proposed by Germano et CA u"-J)= (l) al. [7, 8].

ID 0 Tipet to The filtered transport equation for a passive scalar is given where u"' and r are the velocity vector and the momentum by source vector, respectively, and the integer n is the time level.

A is the coefficient matrix defined in terms of the discrete a,P+2EE= a aa q) (6) convective and diffusive operators and the time-advancement ~T k k (axi )

scheme chosen, and G and D are the discrete gradient and divergence operators, respectively. The coupled system of equations in Equation (1) is extremely difficult to solve as it is, where qj is the subgrid-scale turbulent flux of the species since the matrix A has to be inverted for every iteration. In the concentration (up), and a the molecular diffusivity. The subgrid-fractional-step method, the original coupled equations in Equation (1) are approximated by scale flux is modeled using U A 0 ][i ID tiDGJ[0I 1Jv.1 )

IzG( U) 0 (0r)[

+~4 2 (2)

Sc, axi (7)

Factorizing equation (2), we have a series of split operations as 2 Copyright @ 2005 by ASME

where Sc, is the subgrid-scale turbulent Schmidt number. In this study, we used a constant value of 0.9.

To implement the dynamic procedure for the present finite-volume solver requires a test-filter applicable to arbitrary unstructured meshes. The test-filter finally adopted for this study is a top-hat filter involving a volume comprising the cell in the center plus the neighboring cells that share the faces with the center cell. To make the dynamic procedure tractable, an approximation was made that is tantamount to a non-varying C, over the test-filter volume. Thus, the dynamic procedure employed in this study is "approximately local" in the sense that, despite the ad hoc assumption, it does not require an existence of any statistically homogeneous directions, being applicable to Figure 2. A sectional view of the computational mesh complex three-dimensional flows. To avoid numerical instability (medium-sized mesh with 2.7 million elements) used In the likely to be caused by a large fluctuation of the model constant, computations.

we smoothed the model constant by applying the test-filter on it, and also clipped it so that the effective viscosity remains positive.

The additional details of the implementation of the DSM in the The computational meshes were built using Gambit. We framework of an unstructured mesh based finite-volume solver used three progressively refined, unstructured hexahedral can be found in the reference [8]. The DSM has been validated meshes for the computations. The three meshes have 1.1 for a number of wall-bounded flows such as fully-developed million (coarse), 2.7 million (medium), and 4.8 million (fine) channel flows and flows around bluff bodies such as a square elements, respectively. A sectional view of the medium-size cylinder and a sphere. mesh is shown in Figure 2. The resolution of the medium mesh is such that, in the mixing region, it can resolve the integral DETAILS OF COMPUTATIONS length-scale estimated around 20 mm based on the experimental observation [1, 2], which is equivalent to D16, with Solution domain, swirl vane geometry and meshes approximately 20 elements. It can resolve the smallest observable eddies of a size around 6 mm, which were observed A partial view of the computational domain is depicted in in the experiment [1,2], with 6 - 7 elements.

Figure 1, along with the coordinate system. The inlet boundary It is worth mentioning here that the mesh resolution in and is placed at 1.OD upstream of the swirl generator, where D is around the inner and the annular tubes and the swirl vanes is the downstream pipe diameter (D = 122 mm). The exit too coarse to resolve the energy-containing eddies originating boundary is at 15D downstream of the pipe expansion. The from and transported in that upstream region. We recognize computational domain thus has the swirl generator in it. The that poor resolution of the turbulent structure in the upstream entire inlet tube, 8 swirl vanes, and the downstream pipe were region will negatively impact the prediction in the downstream, modeled without using any periodic boundaries. particularly in the shear layers of the mixing region, inasmuch as the energy-carrying eddies from upstream feed the turbulent structure being developed in the downstream mixing region.

pipe expaslon Furthermore, none of the three meshes are fine enough to accurately resolve the viscous sublayer on the pipe wall downstream of the expansion. This was deemed justifiable in

,,, . .- f . a- : . - :

light of a relatively passive role played by the wall downstream swirl vane A 4>v 6'Gd--'1Qt.s of the expansion in the mixing occurring near the central core.

Boundary conditions, time-step size, and solution strategy On the annular jet inlet, a uniform axial velocity of 1.667 W/s was specified according to the data given by Roback and Johnson [1]. For the velocity boundary condition at the inner jet inlet, we used 0.797 Wm's, a value derived from the XLZ volumetric flow-rate (6.2 gallon per minute) given in Ref. [1],

Figure 1. A partial view of the computational domain instead of the inlet velocity mentioned in the same report (0.52 mis). The inlet velocity of 0.52 m/s quoted in the report does The swirl generator consists of 8 identical vanes mounted not match the given volume flow-rate, and is inconceivably too on the hub with an equal spacing in the azimuthal direction. low considering that the measured axial velocity immediately The blade-section has a camber and a vane-angle that change downstream of the expansion is around 0.8 mns.

with the radius. The blade geometry was taken from the On wall boundaries, we employed a blended law-of-the-design data given in the Ref. [1]. A NURB surface was built wall that invokes proper wall-laws depending on the local mesh based on the digitized surface geometry. resolution, namely yv at the wall-adjacent cells. The downstream exit boundary was treated as a "pressure-outlet" boundary offered in FLUENT. In essence, the solution 3 Copyright © 2005 by ASME

variables are extrapolated in a mass-conserving manner on this The LES was started using a steady RANS solution as the boundary. initial guess. To accelerate the solution to a statistically The influence of the exit boundary condition was stationary state, we superimposed a pseudo-random fluctuating discussed at length by Pierce and Moin [6]. They found that velocity-field on the mean velocity field taken from the RANS the usual convective outflow boundary condition applied on an solution. Before the statistics are collected, the LES exit boundary with a cross-section of the downstream pipe computation was run until the initial transients in the solution yielded a central recirculation zone that is far smaller than what are completely washed out, which typically took 2 - 3 flow-through times; one flow-through time is taken to be L'lb, where L is the length of the downstream pipe, and Ub the axial mean bulk velocity.

RESULTS Figure 4. Instantaneous velocity vectors in the r-z plane I

Figure 3. Instantaneous velocity vectors at four axial locations - top-left, z = 25 mm; top-right, z= 51 mm; bottom-left, z=102 mm, bottom-right, z = 203 mm ....... ..... ....

was observed in the experiment. They were able to bring their LES prediction much closer to the experimental data by putting .. . . .. . . .

a second expansion before the exit boundary, which was ,tLV conceived based on an argument that it better represents the settling chamber used in the experiment.

We also investigated the effect of the second expansion, comparing the result to what was obtained without he Figure 5. Time-averaged velocity vectors In the r-z plane expansion. However, we could not notice any appreciable difference between the two. The insignificantly small difference found in our computations is perhaps due to the way To obtain stable statistics of the solution, the transient the solution variables are extrapolated at "pressure-outlet" computations were continued for a sufficiently long period of boundary that could be different from their treatment of the time, typically for more than 7 - 8 flow-through times, until convective boundary. the time-averaged velocity and species concentration fields The time-step size used in the present LES was determined largely recover an axisymmetry.

based on the estimated characteristic time-scale of the smallest Regarding the mesh-dependency of the solutions, it was eddies to be resolved in the LES. We took v = 1.0 mls and I found that the LES results from the medium mesh (2.7 million

= 0.05D (D = 0.122 m) as the characteristic velocity-scale and cells) and the fine mesh (4.8 million cells) showed little length-scale of the smallest eddies. These estimates give an difference, while the coarse mesh result deviates a little farther eddy-tumover time of -r /V = 0.006 second. It was finally from other two. Unless stated otherwise, the results presented decided to use a time-step size of At = 0.0002 second, which in this paper are for the medium mesh.

will resolves one eddy-turnover time (i) of the smallest resolved eddies roughly in 30 time steps.

the penetration length of the center jet with high species concentration (colored pink in the figure in Figure 7) was found to be around 50 mm, which is closely reproduced by our LES results as shown in Figure 7.

Figure 6. Instantaneous velocity vector plots at two axial locations - left, z = 25mm; right, z= 51 mm Overall flow structure Figures 3 and 4 show the instantaneous velocity vectors projected on four crosiflow (r-0) planes and a r-z plane. The vector plots portray turbulent eddies with widely varying length-scales throughout the mixing region. Small eddies are Figure 8. Vortical flow structure in the mixing region shown to form in the shear layers between the inner jet and the visualized by the Iso-surface of the second-invariant of the outer annular jet, growing in size in the downstream direction. velocity deformation tensor, colored by velocity magnitude.

At z = 25 mm, small eddies are confined near the shear layer between the inner and the outer annual jets. Yet, larger eddies Figure 8 gives an overall impression of the turbulent are also seen to have formed in the annular recirculation region vortical structure in the mixing region. This figure indicates behind the step. One can visually tell from the figure that the that the flow in this coaxial jets are well mixed.

smallest structure at z = 25 mm resolved in the LES is roughly D/20 (D = 122 mm), which is close to the size of the smallest Velocity field eddy, 6 mm, observed by Roback and Johnson [I]. The size of the largest eddy at z = 25 mm, which is about two-thirds of the Figure 9 shows the mean axial -velocity profile along the step-height as shown in the figure, also closely matches the centerline of the pipe. The LES prediction quite closely experimentally observed value, 20 mm, quoted in ref. [11. reproduces the general trend such as the rapid drop of the mean Figure 5 depicts the mean velocity vectors on a r-z plane. axial velocity immediately downstream of the expansion and the gradual recovery further downstream. However, the recirculation zone predicted by the present LES appears to be shifted slightly downstream compared to what is indicated by the measurement.

LES (present) 1.0 o Measured (Roback and Johnson, 1983)

~. 0.5

>. 0 000 a 0r Figure 7. Contours of the Instantaneous species concentration in r-z plane Figure 6 and 7 depict the contours of the instantaneous Figure 9. Mean axial velocity along the centerline of the species concentration on two crossflow planes and a r-z plane. pipe.

The contours on the crossflow planes give an idea of the length-scale of the turbulent structure. The observations from these figures are consistent with those from the instantaneous velocity vector plots discussed earlier. In the experiment [1],

5s Copyright 0 2005 by ASME

Z = 5mm z=5 mm a

a

>e

'e c z = 25 mm Si Measured (Roback and Johnson, 1983)

-Present LES 1.01

'C.

0.51 a

4) W.

C O.0 I

. I I . I I .

-O.

.0.41 0.2 OA 0.6 0.8 r/r.

z=51 mm E

ZC a9 rlr, Figure 10. Mean axial and azimuthal velocity profiles at two axial locations (z = 25 mm, z = 51 mm)

Figure 10 shows the radial profiles of the mean axial and obtained by averaging the radial profiles taken at four azimuthal velocity components at three axial locations (z = 5 azimuthal locations (0 = 00, 900, 180°, 270°). The predictions mm, 25 mm, 51 mm). The profiles shown in the figure were are seen to capture the overall trends and the peak values of the 6 Copyright Q 2005 by ASME

velocity components. However, it is clearly noticeable that the used in this study, including the fine mesh (4.8 million cells),

locations of the peaks predicted by the LES are generally are fine enough to accurately resolve the energy-containing shifted toward the centerline. This implies that the outer eddies generated in the upstream region. ,....

annular jet expands or opens up less than it does in reality. The r.m.s. axial velocity fluctuation shown in Figure 11 This also means that the annular recirculation zone predicted by seems to support our reasoning. The r.m.s. axial velocity the present LES is longer (in the axial direction) than in the fluctuation at the centerline (r = 0) at z = 25 mm shown in the experiment. The underprediction of the jet angle is also top-figure, and the one at z = 5 mm (not shown here) are correlated with the downstream shift of the central recirculation severely underpredicted in the LES computation. As shown in zone discussed earlier. the bottom-figure for z = 51 mm, further downstream of the jet exit, the r.m.s. axial velocity fluctuation catches up with the z = 25mm data, as the turbulent eddies generated in the shear layer become full-fledged.

The impact of the incoming turbulent eddies on the mixing in the downstream as suggested above and the cost implication of using an extremely fine mesh in the upstream region beg a question; what would be the best practice that can be adopted to obtain a sufficiently accurate prediction of the flow and species mixing occurring in a coaxial jet combustor with LES? We

4) will ponder a little upon this question at the very end.

C4 Species concentration Figure 12 shows the profile of the mean species concentration along the pipe centerline. The LES prediction closely reproduces the trend - the plunge of the mean species concentration occurring near z = 50 mm. One noticeable discrepancy between the prediction and the measurement is in the length of the inviscid core for the species concentration. The LES yields an inviscid core for the species concentration almost down to z = 35 mm, whereas the experimental data indicates that the mean species concentration starts being diffused away almost immediately after the expansion. We think that this is again, in a i0.6 major part, due to the poor resolution of the incoming turbulent eddies which would take part in "tearing" the innerjet.

, 0.4

>e

'C 0.2 rIr Figure 11. r.ms. axial velocity fluctuation at two cross-flow planes (z = 25 mm, z = 51 m)

We surmise that, among others things, the most likely culprit for this discrepancy is the lack of mesh resolution in the region upstream of the expansion, in and around the inner and the outer annular pipes, and the swirl vanes. As mentioned Figure 12. Mean species concentration along the centerline earlier, turbulent eddies coming from the upstream region feed of the pipe the shear layers developed downstream of the expansion, enhancing the mixing of the momentum and the species concentration in the inner and the annular jets with the The radial profiles of the mean species concentration at two surrounding flow, which will lead to an increase in the jet axial locations are shown in Figure 13. The predictions are in angle. In our LES computations, these energy-feeding eddies good agreement with the measurements at both locations. The are almost missing, since none of the computational meshes mean species concentration profile at z = 25 mm predicted by the LES computation shows a sign of being "under-diffused", insofar 7 Copyright C 2005 by ASME

as it has a fuller profile than the measured one. The prediction at r.m.s. value at z = 25 mm, located in the shear layer between the z = 51 mm is somewhat lower than the measurement near the inner and the outer annular jets, is also considerably centerline. However, the overall agreemiient at this location is underpredicted. Apparently, the LES underpredicts the remarkably good in view of the steep change of the mean species entrainment of the ambient fluid occurring at this location. At z concentration near z = 51 mm, as can be seen in Figure 12. It = 51 mm, the prediction comes much closer to the should be noted, in passing, that the RANS predictions reported measurement, as the r.nms. velocity fluctuation did at the same in the literature [3] severely under-predicted the mean species axial location.

concentration at this axial location.

z=25 mm z =25mm r/r, z=51 mm 0

0 S

4)

C Ca ror rrO Figure 13. Mean species concentration at two axial Figure 14. rmrnps species concentration at two axial locations locations (z = 25 mm, 51 mm) (z=25 mm,51 mm)

Equally important - probably even more important than the The same remarks given regarding what might have caused mean species concentration in the context of modeling the discrepancy in the r.m.s. velocity fluctuation (Figure 11) turbulent combustion, is the fluctuation of species largely apply to the results for the r.m.s. fluctuating species concentration. The usual RANS-based turbulence models concentration.

cannot directly predict the rTm.s. fluctuating species concentration, unless the transport equation for the variance of fluctuating species concentration is explicitly solved. One PRELMINARY RESULTS WITH LOCALLY REFINED obvious benefit from LES is that one can directly compute it. MESHES Figure 14 depicts the r.m.s. fluctuating species concentration at the two crossflow planes. At z = 25 mm, as in the case of the The fact that the fine mesh (4.8 million cells) offers a r.m.s. axial velocity fluctuation, the LES result grossly meager improvement over the medium mesh (2.7 million cells) underpredicts the r.m.s. value in the core region. The peak in terms of the mesh resolution (spacing) warrants additional 8 Copyright 0 2005 by ASME iI

computations with substantially finer mesh, for instance, by halving the grid spacing in all three (axial, radial, and CONCLUSION azimuthal) directions. However, doubling the number of elements in the three directions results in an eight-fold increase *The flow and the species transport in confined swirling in the total cell counts, which becomes unwieldy. coaxial jets were computed using LES. An unstructured mesh-based finite-volume solver was employed for the computations.

A highly efficient time-advancement scheme was used in z= 5 nmu conjunction with second-order accurate temporal and spatial discretization schemes. A dynamic Smagorinsky model adapted to general three-dimensional flows was employed as the subgrid-scale turbulence model.

For LES, it is evidently a bold attempt and a costly

>i proposition to include the upstream components of a combustor like the swirl generator and the upstream tubes in the

.9 computational domain. Nevertheless, the present LES a

11 computations closely reproduce the salient features of the flow

> and the species concentration in the mixing region. For an X

accurate prediction of the mixing in the downstream (e.g.,

a combustion chamber), a good resolution of the mean flow and the turbulence in and around the inner and the outer annular tubes, and the swirl vanes turned out to be more important than originally thought. The numerical evidence found from this study indicates that the three globally refined meshes used in the present study, despite the largest cell counts reaching up to 4.8 million cells, still cannot provide a sufficient resolution of z=51 mnm the upstream region. We believe that the lack of mesh resolution is responsible for the discrepancy between the predictions and the measurements, most notably the overall shift of the velocity peaks toward the centerline.

Finally, a preliminary result was presented which was obtained with a new mesh locally refined in the upstream part of the domain. The significant improvement from this locally adapted mesh supports our conclusion, and at the same time, provides us with an avenue to improving the accuracy of LES prediction for the subject flow.

C a,

ACKNOWLEDGEMENT The authors acknowledge that FLUENT and GAMBIT software were used for this study. The LES-cluster in Fluent r/r. Inc. was used for the computations.

Figure 15. Mean axial and azimuthal velocity profiles at two axial locations - predicted with a locally refined mesh REFERENCES To keep the mesh size under a tractable limit, the medium [1] Roback, R. and Johnson, B. V., "Mass and Momentum mesh was locally refined only in the domain upstream of the Turbulent Transport Experiments with Confined Swirling expansion. Furthermore, only the cells within a specified Coaxial Jets," NASA CR-168252, 1983.

distance from the wall, roughly 0.07D in this study, were [2] Johnson, B. V. and Roback, R., 'Mass and Momentum refined to further save the cell counts. The local refinement Turbulent Transport Experiments with Confined Swirling resulted in a total of 6.5 million cells. The computation has Coaxial Jets - Part I," AIAA-84-1380, Presented at been carried out on this mesh. The results are shown in Figure AIAAJSAEIASME 20th Joint Ppropulsion Conference, 15, being compared with the medium mesh results, for the Cincinnati, OH, June II - 13, 1984.

radial profiles of the mean axial and azimuthal velocity [3] Brankovic, A., Ryder, Jr. R. C., and Syed, S. A., 1998, components at z = 51 mm. Clearly, the predictions with the "Mixing and Combustion Modeling for Gas Turbine adapted mesh are significantly improved over the predictions Combustors Using Unstructured CFD Technology:' AIAA with the medium mesh. The overall shift of the velocity peaks Paper 98-3854, Presented at 34t' AIAAIASME/SAE/ASEE is now far smaller than what we saw earlier, and the predictions Joint Propulsion Conference & Exhibit, July 13 - 15, 1998, capture the profiles of the mean velocity components much Cleveland, OH.

more closely.

[4] Lin, C. A., "Modeling a Confined Swirling Coaxial Jet,"

Annual Research Brief, 1998, Center for Turbulence Research., Stanford University, 1998.

[5] Akselvoll, K. and Moin, P., 1996, "Large-Eddy-Simulation of Turbulent Confined Coannualr Jets," J. Fluid mech.,

315, pp.387-411

[6] Pierce, C. D. and Moin, P., 1998, "Large Eddy Simulation of a Confined Jet with Swirl and Heat Release," AIAA Paper 98-2892.

[7] Germano, M/, Piomelli, U., Moin, P., and Cabot, W. H.,

1991, 'Dynamic Subgrid Scale Eddy Viscosity Model,"

Physics of Fluids A, 3, 19, pp. 1760 - 1765.

[8) Kim, S. E., 2004, "Large Eddy Simulation Using Unstructured Mesh and Dynamic Subgrid-Scale Turbulence Models," AIAA Paper 2004-2548.

[9] Kim, S. E., Makarov, B., and Caraeni, D., 2004, 'Multi-Dimensional Reconstruction Scheme for Unstructured Meshes," AIAA Paper 2004-2548.

[10]Kim, S. E. and Makarov, B., 2005, "An Implicit Fractional-Step Method for Efficient Transient Simulation of Incompressible Flows," To be presented at 17" AIAA Computational Fluid Dynamics Conference, June 6 - 9, Toronto, Ontario.

10 Copyright 0 2005 by ASME

BVY 05-072 Docket No. 50-271 Exhibit EMEB-B-1 38-3 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 30 Extended Power Uprate Response to Request for Additional Information Large Eddy Simulation Using Unstructured Meshes and Dynamic Subgrid-Scale Turbulence Models.

Total number of pages in this Exhibit I I (excludina this cover sheet) Is 17.

34th AIAA Fluid Dynamics Conference and Exhibit Juno 28 - Jity 1, Portland Oregon Large Eddy Simulation Using Unstructured Meshes and Dynamic Subgrid-Scale Thrbulence Models Sung-Eun Kimp FluentInc, Lebanon, New HampshIre, 03766, US.A.

This paper concerns development of a large eddy simulation (LES) capability based on a finite-volume solver that permits use of unstructured meshes. The solver employs a cell-centered scheme along with a multi-dimensional linear reconstruction. Convection and diffusion terms are discretized using a second-order central-differencing scheme. A three-level second-order scheme Is used for temporal discretization. Subgrid-scale turbulent stresses are closed using dynamic Smagorinsky model and dynamic turbulent kinetic energy transport model. A test-filter was designed for the dynamic procedure which Is applicable to unstructured meshes of arbitrary cell-topology. The dynamic procedure also avails Itself to three-dimensional flows without any statistically homogenous directions. Wall boundary conditions are Imposed using a wall-function approach that applies appropriate wall-laws depending on near-wall mesh resolution. The LES capability Is validated for a wide range of wall-bounded flows. We present here the results for a fully-developed channel flow and two bluff-body flows. The predictions are in good agreement with direct numerical simulation (DNS) results and the experimental data.

I. Introduction W E encounter many industrial applications of computational fluid dynamics (CFD) where Ihe flows are dominated W by unsteady, large-scale coherent structures. Those large-scale structures impact, to a great extent, various aspects of the flows such as energy consumption, safety, comfort, and noise. The ramification of whether or not one can harness the large-scale structures is therefore quite significant. Attempts to numerically predict such flows using unsteady Reynolds-Averaged Navier-Stokes (URANS) equations have been met with limited success. Some of the better RANS models seem to be capable of capturing the "largest" scale occurring often in the form of alternate vortex-shedding. However, the remaining coherent structures are left largely unresolved.

Large eddy simulation (LES) is fundamentally suited to the task of predicting coherent structures. The major obstacle in using LES for practical high Reynolds-number (Re) flows, from the practitioners' standpoint, is its high cost incurred by an unwieldy number of computational elements and painfully long solution time. Yet today's ever-increasing computing power is rapidly making LES feasible. Another difficulty often encountered when attempting LES for industrial applications comes from complex geometry. In CFD, meshing for industrial applications involving complex geometry by itself can become a grand challenge. It is hugely time-consuming or often impossible to generate high-quality structured meshes for complex configurations, which has led industrial CFD practitioners to opt for unstructured meshes. Although unstuctured meshes are routinely used today in RANS computations for industrial applications, attempts to conduct LES with unstructured meshes are just starting to appear in the literature. 1 -5 As yet the efficacy of unstructured meshes for LES for practical high-Re flows has not been fully established. Among the issues yet to be addressed are numerical accuracy, stability, and subgrid-scale turbulence modeling.

This paper is concerned with evaluating a LES capability developed using a finite-volume solver based on second-order numerics. Permitting use of unstructured meshes, the resulting LES capability can easily handle complex ge-ometries encountered in industrial applications. In addition, it lends itselt to local mesh adaptation that can be utilized

'Principal Engineer, Fluent Inc.. Lebanon N.H., Member AIAA.

Copyright 0 2004 by Fluent Inc.. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission.

I of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

to efficiently allocate computational cells, substantially reducing the computational cost. Adequacy of second-order spatial discretization for LES has often been questioned, and there are some misgivings about using it for LES. It will be shown in this paper, however, that the second-order central differencing scheme adopted in the present work yields a commendable accuracy for LES. For subgrid-scale (SGS) turbulence closure, we implemented two dynamic SGS vis-cosity models in the framework of unstructured meshes, namely, the dynamic Smagorinsky model originally proposed by Germano et a., 6 and Lilly 7 and the dynamic turbulence kinetic energy transport model of Kim and Menon. 8 9 For the dynamic procedure, a test-filter readily applicable to unstructured meshes was designed. The resulting dynamic SGS models can be used for three-dimensional flows without any statistically homogeneous directions.

The paper is organized as follows. We start with a brief description of the two dynamic subgrid-scale turbulence models and the details of their implementations. This will be followed by an overview of the numerical methods and algorithms adopted in this work. Finally, validations will be presented for a selected number of wall-bounded flows ranging from a fully-developed channel flow to a couple of bluff-body flows including one around a square-cylinder with salient edges and another past a smooth sphere.

II. Filtered Navier-Stokes Equation and Subgrid-Scale lurbulence Modeling!

A. Implicit filter with finite-volume discretlzatlon The governing equations for LES are generally obtained by filtering the Navier-Stokes equations in either Fourier (wave-number) or physical space. In the present work, the filtering operation (denoted by an overbar) is defined as a spatially convoluted integral of the variable in question as

¢(X = l6(y)G(x,y)dy (1) where D is the computational domain, y E D, and G is the filter function.

With the cell-centered finite-volume discretization and the linear reconstruction scheme employed in this work, the discrete solution variable at a cell-center (co) can be written as W(cO) = V (y)dy, yE V (2) where V is the volume of a computational cell.

The definition in Equation (2) of a discrete solution variable at cell center can be interpreted as a filtering operation

¢(x)--(co) = Vfl (y)dy, y E V (3)

The implied filter function, G(x,y), is then a top-hat filter G~~)-_I/IV for~x-y EV (4) 0 otherwise Using the filtering operation in Equation (3). the filtered Navier-Stokes equations for incompressible flows (as-sumed for brevity) can be written as at amj =_ a ?rijj a (v a\

a.ay p axi axJ axi ax}

ax,

=0 (6) where rij is the subgrid-scale stress defined by; aptj Wu -uuJrig (7) which is unknown and needs a closure.

Thus, we used in this work grid and finite-volume discretization as an implicit filter.

2 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

B. Subgrid-scale (SGS) stress models The SGS models based on the concept of isotropic eddy-viscosity compute the SGS turbulent stress from TO- 3kk~jJ = -2VSaj (8) where V,is the SGS eddy-viscosity, and SjJ the resolved rate-of-strain tensor defined by 1 (d~ ad-uI 2 (axj Dx,)

The task of SGS turbulence modeling is to express SGS viscosity, v,, as a functional of known quantities. In the present work, we employed two dynamic SGS eddy-viscosity models. They are described below.

1. Dynamic Smagorinsky Model (DSM)

The underpinning of DSM is the algebraic eddy-viscosity model originally proposed by Smagorinsky.10 In the original Smagorinsky's model, SGS eddy-viscosity is computed from V,= CV 2 II (9) where Cv is a model constant (Cv = 0.1 0.2), [l1 the modulus of rate-of-strain tensor of theresolved scales, and A = V 113.

The subgrid-scale stress can then be written as Bij-rkk = -2CA 2 11 ij (10)

Despite its simplicity, this model has several shortcomings. The most problematic one, from a practical standpoint, has to do with the model constant, Cv. There is no single value of the constant which is universally applicable to a wide range of flows. Another serious drawback is that the SGS viscosity model in Equation (9) with a constant value of Cv is not applicable to transitional flows where the flow in question is laminar either locally or intermittently, since Equation (9) always gives a finite SGS viscosity even in laminar region as long as there is velocity gradient.

Germano et al.6 and subsequently Lilly 7 conceived a procedure that resolves these problems. In this procedure, Cv is dynamically computed during LES, on-the-fly, using the information provided by the smaller scales of the resolved (known) fields. To separate the smaller scales from the resolved field, the dynamic procedure needs a so-called "test-filter" having a width (A) that is larger than the grid-filter width (A). Denoting the test-filtered variables by a tilde and putting the "grid-filtered" Navier-Stokes equations through the test-filter, we obtain "test-filtered" Navier-Stokes equations as at axj paxi axi ax, 0 0

, =0 (12) where T7j is now a "subtest-scale" stress defined by Tij-j -- (13)

The underpinning of the dynamic modeling is a similarity concept that Tij, the subtest-scale stress, can be written as a functional of the larger resolved scales in a manner similar to njj,6 which leads to 3 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

Ty - Tn2 = -2CvY2ISISqJ (14) where i is the test-filter width. Note that the same model coefficient, Cv, is used in the expressions for both a/j and Tyj.

Equation (14) alone is not helpful at all in determining Cv, because Tyj is not known in LES. The breakthrough came from ihe realization that tij and Tij are related to each other by6 7

jj-Nl = u-i (UB=JUtj) 1U stu U_ _1ii

-uj-u= iVUJ Lij (15)

The stress, Lij, which might be called subtest-scale Leonard stress, can be interpreted as the stress associated with the smaller resolved scales between the test-filter scale (A) and the grid-filter scale (A). Since 4j can be directly computed from the resolved scales in LES, the identity given by Equation (15) can be used to determine Cv. Thus, we have Lij - TL*L = aijCv - PijC (16) where aij = _232 S See (17)

P = -2 2 S21Sij (18)

One predicament that makes it difficult to determine Cv from Equation (16) is the fact that Cv in the second term on the right-hand side of the equation is under the test-filtering operator. As it stands, Equation (16) is an integral equation for Cv as discussed at length by Ghosal et al. This difficulty can be avoided by taking out Cv from the test-tilter operation as

- LJ - 'Lkk =Cv (aaj- Fi) (19)

We followed this rather ad hoc approach in this work despite its mathematical inconsistency, which amounts to as-suming that Cv remains constant in the computational cells associated with the test-filter.

Since there are more equations in Equation (19) than the unknowns, the model constant Cv is obtained by seeking for Cv which minimizes the error norm defined by E (L'= JLj-8 ... k-CvMeJ) (20) where Mj= - Dij = -2 ( ISlIj S2 _2 ilS)

Taking aElaC, and setting it zero, we obtain CV = LijMij (21)

Cv determined in this way varies with time and space. In fact, it varies in a wide range, often taking either large negative or positive value. Although negative Cv and consequently negative eddy-viscosity is often interpreted as representing "back-scatter" (flow of energy from smaller to larger scales), too large a negative eddy-viscosity causes numerical instability, ultimately leading to divergence of numerical solutions. The usual remedy for this numerical difficulty is to average Cv in statistically homogeneous directions. Obviously, this workaround can be exploited only when there are such homogeneous directions, which is a rarity in practical applications. Even if there are any statisti-cally homogeneousdirections, the averaging becomes extremely cumbersome with unstructured meshes. In the present 4 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

work, we simply "condition" the Cv computed by Equation (21) by test-filtering it. This simple volume-weighted aver-aging better preserves the locality of the model constant. To ameliorate the potential numerical difficulty caused by Cv being negative for an extended period of time, we evaluated two alternative approaches. In-one approach, we simply clip GC at 0 when it becomes negative. This option therefore rules out any chance for the model to mimic backscatter.

In another, effective viscosity (v +iv) instead of Cv is clipped at zero, permitting small negative SGS viscosity to happen. As yet we do not have any conclusive evidence that supports superiority of one approach to another in terms of prediction accuracy, except an indication that computations with the first approach appear numerically more robust.

For this reason, we used the first approach in the computations presented in this paper.

As mentioned earlier, the dynamic procedure described above requires a test-filter. The most important criterion the test-filter should satisfy specifically for this work is that it should be applicable to unstructured meshes of arbitrary cell topology without incurring unduly high cost. The test-filter finally adopted is a top-hat filter that involves a volume comprising the cell itself plus the neighboring cells that share the cell faces with the center cell. Thus, the test-filter operation amounts to a volume-weighted averaging of the variable in question, which is easily implementable and takes advantage of the data structure of the underlying finite-volume solver. With hexahedral meshes, the ratio of the test-filter to the gird-filter scale (A/-) is approximately 2.1 (= 91/3). The ratio for tetrahedra is smaller, being around I .7(= 51/3).

2. LocalizedDynamic Kinetic Energy Model (LDKEM)

TIhe dynamic Smagorinsky's model described so far is essentially an algebraic model in which subgrid-scale stresses are modeled using the resolved velocity field. A more elaborate SGS stress model would be the one which is directly based on SGS turbulence and can be used to parametrize SGS stresses. The most widely used ones among others are what can be called "one-equation" models in which SGS turbulent kinetic energy, k3 S, = (uk2- ri)/2, is explicitly computed by solving its transport equation.8 .9 t 1.12 In the present work, a localized dynamic kq,5-equation model of Kim and Menon8 '9 was chosen in favor of its overall efficacy. 13.14 In the LDKEM, the subgrid-scale eddy viscosity, v,, is computed from v, = Ck.&Sg/s a /(22 (22)

Consequently, SGS stress is can be written as Tsy- 3 ,J5Si=-2CAk/ ASjj (23) k3gs is obtained by solving its transport equation8 . 9 pig O- a, = tiaj-Ui-C, ^+ aa VI + )aks ] (24)

I5 1,ar TijaXJ aXj Ock I aXJJ The only difference between the original formulation and the present one is that the contribution from the molecular diffusion of kJ,, is included in this works As shown above, there are three model parameters appearing in these equations, namely, Ck, Cc, and at, which need to be specified. In the current implementation of the LDKEM, the first two are determined from the dynamic procedures to be described in the following, whereas cr is simply set to a constant value of 1.0. The underpinning of the dynamic procedure employed in the LDKEM is the hypothesis corroborated by theexperimental evidence 5 .16 that there is a strong correlation between the subgrid scale stress, wrij, and the subtest-scale Leonard's stress, L,>. In place of parameterizing Tj and utilizing the Germano's identity as was done in the DSM, the LDKEM models Lij directly as

- Ž!LLk = -2CkAkt4e/,jSJ (25) 5 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

where kno is the resolved turbulent kinetic energy associated with the scales between the test-filter (A) and the grid-filter (A). It can be directly computed from kest= (4~UkTUk) (26)

The model parameter, Ck, can then be determined from Equation (25) by minimizing the error norm as in the DSM.

Consequently we have Ck = LjMIj (27) with Miy defined by Mj = 2AkL/Sj (28)

The model parameter, CE, of the dissipation term in Equation (24) is also determined by a dynamic procedure, whose underpinning is the hypothesis that the dissipation-rate of kim (e) can be expressed in the same functional form as the dissipation-rate of k113.

3/2 e = Cc k;,g (29)

The dissipation-rate of kiess can also be computed from e =(v+v __ a-t _ u Z~ . (30) axjix ax, Xj xyj From Equation (29) and Equation (30), Cc is given in a closed form CC A(V+V,) ({au-,ait aui 1 (31) t 3/lf2 aXjax)(1 jax Xj The DKEM has several desirable attributes that the DSM lacks. First, as a consequence of parameterizing Lij directly, the dynamic procedure in the LDKEM does not involve any test-filter operation on the model parameter, C&.

Thus, unlike the DSM model, Ck is a genuine, local quantity free from any mathematical inconsistency. Secondly, Ck in the LDKEM behaves numerically more benignly than Cv in the DSM, having much less fluctuation. As a small premium, one can even save a small amount of computational effort with the LDKEM, inasmuch as the test-filtering on the SGS stress done on in the DSM is not needed. Lastly and probably most important, the LDKEM enjoys the benefits of a high-order turbulence model. Adopting SGS turbulent kinetic energy to parametrize the SGS stress renders the LDKEM better suited to non-equilibrium flows. By accounting the kg,-budget more rigorously, backscatter of kinetic energy is allowed in the LDKEM on a much sounder physical basis than in the DSM.

m. Boundary conditions Wall boundaries are the most crucial and yet difficult ones to handle in LES. In LES resolving all the way down to viscous sublayer, no-slip would be clearly the choice for wall boundary condition for the resolved velocity field.

However, the cost of such "wall-resolving" LES is prohibitively high for practical flows involving high-Re flows. A practical alternative is to use the law-of-the-wall bridging the wall and the first grid point (cell center) off the wall. The simplest implementation of the wall-law can done using

{ in(Ey+) fory+ >y+ (32) where E = 9.793, ic= 0.419, y+ -7ylv, u _ RI/V, y+ is the "cross-ovee'at which the two wall laws intersect. This demarcation of the entire inner layer into the two distinct layers is apparently a simplification which is at odds with the 6 oft7 Arnrican Insdtute of Aeronautics and Astrocautics Paper 2004-2548

presence of a buffer region in reality between the viscous sublayer and the log-layer. However, this is much better than exclusively relying on no-slip condition. One can mimic the presence of the buffer layer by blending the linear and the logarithmic laws using an adequate blending function. The blending has some merits, inasmuch as, besides giving a smooth transition between the two layers (numerically moie stable) and representing the mean velocity profile in the buffer layer more accurately, the blending also allows the respective laws in the two regions to be independently modified or extended to take into account other effects such as pressure gradient, surface roughness, and transpiration.

In the present work, we employed the blending function suggested by Kader.17 In the finite-volume discretization adopted in this work, the blended wall-law is employed to compute wall-shear, essentially diffusion flux at wall. To thatend, the wall-law is applied to the parallel components of the resolved velocity at wall-adjacent cells to compute the friction-velocity (oh)in an iterative manner.

The LDKEM needs a wall boundary condition for k5g,. To that end, either a Dirichlet-type boundary condition (ksg, = 0) or a Neumann boundary condition (Dk5 1,1 an = 0, in view of keg - y2) is conceivable. In the present work, we simply set the diffusion flux of kg, at walls to zero, which is essestially equivalent to the Neumann boundary condition.

IV. Flow solver The present work was carried out using FLUENT, a general-purpose CED code. FLUENT employs a cell-centered finite-volume method based on a multi-dimensional linear reconstruction scheme, which permits use of computational elements (cells) with arbitrary polyhedral topology, including quadrilateral, hexahedral, triangular, tetrahedral, pyra-midal, prismatic, and hybrid meshes. There are several choices of the solveralgorithms in FLUENT including coupled explicit, coupled implicit, and segregated solvers. For the computations presented in this paper, we used the segregated solver exclusively.

In the segregated solver, the governing equations are solved sequentially. Several ditterent solution algorithms are.

offered including SIMPLE, SIMPLEC, PISO, and fractional-step method (FSM). The temporal discretization in the segregated solver employs a fully-implicit, three-level second-order scheme. lime-accurate solutions can be obtained using either iterative time-advancement (ITA) scheme or non-iterative time-advancement (NITA) scheme. The NITA scheme greatly saves CPU time, since the costly outer iterations arc not needed. Unless stated otherwise, we used the fractional-step method in conjunction with the NITA scheme in this study.

Accurate spatial discretization is crucial in LES. The spatial discretization schemes employed in this work are based on a multi-dimensional linear reconstruction scheme.'1 20 Diffusive fluxes are discretized using central differ-encing. Discretization of convective fluxes requires caution in LES. Upwind-biased schemes such as second-order upwind, QUICK, and third-order MUSCL schemes have been most widely used for RANS computations. Unfortu-nately, numerical diffusion introduced by upwind schemes, which might be acceptable in RANS computations for high Reynolds-number flows, is detrimental to LES. This is because, in LES, numerical diffusion, however small it is, can easily overwhelm physical diffusion. This is the case even with high-order upwind schemes. For this reason, for LES, central-differencing schemes have been preferred for their meritoriously low - or zero in ideal conditions -

numerical diffusion. Thus we added a second-order central differcncing (CD) scheme for discretization of convective terms specifically forLES.2 ' Unfortunately, any pure CD schemes are susceptible to producing unphysical oscillations in the solution fields, which becomes especially pervasive in high Peclet-numbersituations - low diffusivity and coarse mesh which is almost a norm in LES for industrial applications. The usual remedy is to add a modicum of numerical dissipation, either explicitly or implicitly, to suppress the oscillations at the price of sacrificing spatial accuracy. In our implementation of CD, however, no numerical dissipation was explicitly added. And, unless stated otherwise, the CD scheme was used in this work. To back up the CD scheme in case it fails, we also developed what may be called a bounded central differencing (BCD) scheme that essentially detects in the solution fields any wiggles with a wave length of 2Ax or less (X< 2Ax) and suppress them by switching to upwind schemes of varying orders depending on the severity of the wiggles, while retaining the CD elsewhere. It should be emphasized that the BCD scheme significantly differs from the often-employed hybrid schemes blending central differencing and upwind schemes with a fixed ratio.

The BCD scheme is reserved for industrial applications involving high-Reynolds number flows and less-than-ideal 7 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

meshes.

The discretized algebraic equations are solved using a point-wise Gauss-Seidel iterative algorithm. An algebraic multi-grid (AMG) method is employed to accelerate solution convergence. The solver is fully parallelized, which is crucial in LES for industrial applications.

V. Validations The LES capability described so far has been validated for a wide range of wall-bounded flows from simple to complex ones. In this paper, we present the results of a fully-developed channel flow and two bluff-body flows. The channel flow case is a fundamentally important case whose subtlety offers an opportunity to critically evaluate various aspects of SGS turbulence modeling such as the dynamic procedure used to determine the model constants.

The bluff-body flows include the one around a cylinder with square cross-section at a moderately high Reynolds number and the one around a sphere at two Reynolds numbers. Deliberately chosen, both involve large-scale, co-herent structures around the bodies and in the wake, representing typical bluff-body flows encountered in industrial applications.

A. Fully-developed channel flow at Ret = 180 A fully-developed channel flow was computed for the Reynolds number of Rex = 180 (ReH = 3,300) using the two dynamic models. The computational domain is a box of the size [2itH x 2H x 7S] in the axial, normal, and spanwise directions, respectively. The computational domain is bounded by two walls on the bottom and the top of the channel, two pairs of periodic boundaries in the axial and the spanwise directions. The computations were carried out using two hexahedral grids; a coarse grid with 36 x 36 x 36 cells and a globally refined mesh with 72 x 72 x 72 cells. The resolutions of the meshes are such that, with the coarser mesh, y+ value at the wall-adjacent cells is approximately 0.6, and the cell size is Ay+ = 27 near the channel centerplane. The channel walls are treated effectively as no-slip boundaries due to sufficiently low y+ values at the wall-neighboring cells. On the pair of periodic boundaries in the axial direction, a pressure-drop across the pair derived using the given wall-shear (new = pu2) was specified, with the flow-rate determined as a part of the solution. The time-step size of At+ = 0.3 was used, where At+ =&tu2/v. The CD scheme was used for the discretization of convection terms.

The mean axial velocity (U+) and the three r.m.s.velocity components (il+, v'+, '+) predicted using the two dy-namic models are shown in Figure 1 on the following page along with the DNS results of Kim et aL22 The predictions with the coarse mesh are seen to overpredict U+ by about 8 - 12% near the center of the channel. The peak in the u+

profile is also overpredicted with the coarse mesh, whereas the peaks in the profiles of Vi+and w+ are underpredicted.

Our results with the coarse mesh show largely the same trends as found by others who employed grids of somewhat finer resolutions (32 x 64 x 32 mesh, 2 3 65 x 65 x 65 mesh 1), yet closely matching their predictions despite the coarser mesh employed in this work. However, our predictions of the r.ms. velocity components near the channel center is relatively poor. We surmise that the much larger grid spacing near the channel center (Ay+ = 27) is responsible for that.

The predictions improve greatly with the fine mesh. Both dynamic models reproduce the DNS results remarkably well. The mean axial velocity and the r.m.s. fluctuating velocity components are accurately predicted throughout the entire range of y+. Particularly noteworthy is the excellent agreement with the DNS data in terms of the peaks values and their locations of the r.m.s.fluctuating velocity components. Overall, the results obtained with the fine mesh compare favorably with other results mentioned earlier.1.2 3 For instance, our predictions are substantially closer to the DNS data than the results of Haworth and Janseni who computed the same channel flow using LES on a 65 x 65 x 65 node mesh using the Lagrangian dynamic Smagorinsky's model.

The present results are promising, inasmuch as they demonstrate that the second-order CD scheme in conjunction with the dynamic models is able to accurately predict this fundamentally important wall-bounded flow carrying an intricate newr-wall physics. Regarding the impact of SGS modeling, we did not find any significant difference between the results from the two dynamic models. This should not come as a big surprise, however, since the fully-developed channel flow is near equilibrium in the mean.

8 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

(a) Mean velocity (U*) wredictions (b) r.rnms. velocity (a'+) Predictions

y. y.

(c) r.m.s. velocity (v'+) predictions (d) r.m s. velocity (W+) predictions Figure 1. The results of LES using two dynamic SGS turbulence models for Reg = 180 9 of 17 American Institute of Aeronautics and Astrnautics Paper 2004-2548

B. Flow around a square cylinder The flow past a square cylinder measured by Lyn et aL24 was considered. The Reynolds number based on the freestream velocity (Uo) and the width of the cylinder (H) is 22,000. The subject flow is featured by a massive flow separation accompanied by unsteady large-scale structures of widely varying length scales. As such, it aptly represents turbulent flows around bluff bodies with sharp edges. Some others have also tackled this flow using LES.13 25 The domain size and the mesh resolution were chosen in reference to the earlier studies by others.t 3'2 5 A more comprehensive study using ditterent mesh resolutions and domain sizes (spanwise in particular) is aeterred tor a future study. Our objective here is to evaluate the efficacy of the present LES capability by comparing the predictions with other results based on meshes with comparable resolution. The computational domain is bounded by an upstream inlet boundary, top and bottom boundaries located 7.0 H from the center of the cylinder, and an exit boundary at 20 H from the cylinder axis. Freestream conditions were specified at the inlet. The top and the bottom boundaries were treated as symmetry planes (frictionless walls). The exit boundary was modeled as a pressure boundary where the solution variables are extrapolated in a mass-conserving manner. A pair of periodic boundaries separated with a span of 3.OH were used in the spanwise direction. The computational domain is filled with a hexahedral mesh with 660,000 elements. We took advantage of our unstructured mesh capability, embedding a block of locally refined mesh around the cylinder to better resolve the near-wall and wake regions as depicted in Figure 2 on the next page. The averaged wall-distance at the wall-adjacent cells is 0.012 H. The time-step size (At) used for the present computations is 0.02 time unit (H/Uo) which is comparable to that used by others.13 The CD scheme was used for the discretization of convective terms. The statistics were obtained during the LES for a sufficiently long period of time, typically for more than several scores of time units.

Figure 2 on the following page shows the profiles of the mean axial velocity (U/Uo) and the r.m.s.velocity fluctua-tions (u' and Vi)along the centerplane (y = 0) in the wake predicted using the two dynamic models. The time-averaged axial velocity distributions show that the length of the recirculation bubble behind the cylinder is predicted by the two dynamic models to be around L, = 0.9 which agrees remarkably well with the experimental value (Lr m: 0.9). The negative peak and the recovery of the mean velocity in the near-wake are also captured very closely. However, the predictions start to deviate from the measurement forx/H > 2.0, reaching 0.8Uo asymptotically in the far-wake. As shown in the figure, this value is considerably larger than what the measurement indicates (0.62 Uo).24 Interestingly, others who computed the same flowt3,25 also have grossly overpredicted the recovery of the axial mean velocity. Our LES predictions of the mean axial velocity in the far-wake with both dynamic models were found to be largely com-parable to the prediction by Sohankar et aL13 based on their dynamic one-equation model denoted by "OEDSMA" in their paper. However, the present LES predictions reproduce the mean velocity profile in the recirculation bubble more accurately than others. Particularly noteworthy is that our DSM yields somehow a much better prediction than the DSM model used by Sohankar et aL in terms of the recirculation bubble size and the asymptotic value of the axial mean velocity in the far-wake, which begs a question of what could possibly contribute to this sizable difference.

One possible cause is the effectively finer mesh used in the present computations which was made possible by the embedded region of fine mesh around the cylinder. It is also quite likely that the differences in the details of the DSM implementation is responsible. In this regard, it should be noted that Sohankar et aL average Cv in the spanwise (homogeneous) direction, whereas the DSM used here does not.

The r.m.s.fluctuating velocity components are also predicted with a reasonable accuracy by the present LES. The peak values are appreciably underpredicted. However, the locations of the peaks are closely captured. Another obser-vation worthy of mentioning is that V is relatively poorly predicted by both dynamic models.

Table I on page 12 summarizes other global parameters predicted by the present computations, along with the results predicted by others. 13 25 Our LES predictions of the mean drag coefficient, Strouhal number, and r.m.s.lift coefficient well match the measurements and the predictions by others.

10 of 17 Ameican Institute of Aeronautics and Astronautics Paper 2004-2548

(a) Mesh with an embedded fine mesh (b) Axial mean velocity (c) r.m s. axial fluctuating velocity (d)r.m.s. vertical fluctuating velocity Figure 2.Mesh used for the flow around a square cylinder and the mean axdal vreocity and normal stress distributions along the wake centerline II of 17

American Institute of Aeronautics and Astronautics Paper 2004.2548

Table 1. Summary of the LES predictions of the global quantities for the square.

cylinder case (RCH = 22,000)

Methods Lr St CD CL DSM (present) 0.9 0.133 2.19 1.19 DSM (Sohankar et al.13) z 0.6 0.126 2.03 1.23 DSM (Fureby et aL25) 0.83 0.132 2.0 1.34 LDKEM (present) 0.9 0.131 2.14 1.17 OEDSM (Sohankar etaL 13) s 0.6 0.132 2.32 1.54 LDKEM (Fureby et aL25) 0.74 0.130 2.10 1.32 24 Measured (Lyn et aL ) s 0.9 ; 0.13 m: 2.1 1.2 C. Flow around a sphere

1. DirectsimulationforReD= 300 Before tackling the turbulent flow cases, laminar flow at ReD = 300 was computed on a hybrid unstructured mesh using direct numerical simulation. At this Reynolds number, the flow exhibits a weak unsteadiness leading to oscillations in drag and lift forces. The hybrid mesh has a total of 860,000 cells, consisting of prismatic cells in the near-wall region grown from the surface triangles on the body surface and tetrahedral cells filling the rest of the solution domain. The time-step size of 0.04 DIU was used. Several others have computed this flow to validate their numerics. 2 8 Table 2. Suranmary of the prediction for the laminar flow over a sphere (Ren = 300)

Methods St CD Present 0.133 0.6671 1

Tomboulides etaL ' 0.136 0.671 Johnson and Patel2 6 0-137 0l 6S6 Measured 2 9 0.15 - 0.16 0.629 The results are summarized in Table 2 along with others' predictions. 26-28 Tomboulides et aL 27 and Johnson and Patel 26 used, respectively, a high-order spectral element method and a second-order upwind finite difference scheme on high-quality structured meshes. As shown, our predictions agree well with others' results, which is remarkable considering that a hybrid unstructured mesh was used in this work in conjunction with the second-order discretization scheme.

Z. Yurbulent7flows We considered two Reynolds numbers (ReD = 1.0 x 104, 1.14 x 106), one being in sub-critical and the other in super-critical regime. The subcritical flow case has been numerically studied by several others,2 '30 while the supercritical case was studied experimentally by Achenbach. 3 1 In this work, a hybrid unstructured mesh was deliberately used for both Re cases. The hybrid mesh has in total 2.46 million cells, consisting of 0.6 million prismatic cells in the near-wall region and tetrahedral cells filling the rest of the solution domain, with a large fraction of the total cell counts clustered in the near-wake region (see Figure 3).

The mesh quality is not exceptionally high and yet quite reasonable, except the rapid expansion of cell size around the 12 of 17 American Institute of Aeronautics and Astronautics Paper2004-2548

(a) Hybrid mesh consisting of prismatic and tetrahedral elements (b) Enlarged view of the near-wall mesh Figure 3. Hybrid unstructured mesh used for the flow around a sphere cell-clustered region. The average distance from the wall at the wall-adjacent cells is around 1.1 x 103 D. For the lower-Re case, the near-wall mesh is sufficiently fine to resolve the boundary layer, with the yO values at the wall cells below y+ = 1.0 for most part of the wall. For the higher-Re case, however, the near-wall mesh is far from being fine enough to accurately resolve the boundary layer which is much thinner than the lower-Re case, and the y+ values at the wall cells increase by almost two orders of magnitude. Thus, the wall adjacent cells are most likely to penetrate the fully turbulent region (log-layer) on a significant portion of the wall, especially near 0 = 90° where the skin-friction reaches a maximum. The mesh being not ideal, the higher-Re case offers a good opportunity to assess the wall-function based approach adopted in this work. Partial views of the mesh are shown in Figure 3.

Attempts to use the pure CD scheme have not been successful for this case. Numerical oscillations were observed sporadically in a few spots rather remote from the body where the cell size increases rapidly, being accompanied by abnormally large velocity magnitude. Although the oscillations were not catastrophic and affected the global quantities very little, the subsequent computations were carried out using the BCD scheme discussed earlier. A time-step size of 0.02 D/U was used for both Reynolds numbers. The data were collected for more than hundreds of time units.

Figure 4 shows the time-histories of CD for the lower-Re case recorded during the LES using the two dynamic models.

Table 3. Sumrnary of the LES prediction for turbulent flow past a sphere (ReD = 10,000)

Methods CD St 4i' LES with DSM (present) 0.438 0.182 86 - 87 86 - 88 LES with LDKEM (present) 0.433 0.185 86 - 87 86 - 88 DES (Pelaez et al.)3 0 0.430 - _

DES (Constantinescu et a!.?)8 0.397 0.200 84 - 87 93 - 108 LES (Constantinescu et aL)28 0.393 0.195 84 - 86 86 - 88 Measured 31l 32 ; 0.40 0.185 - 0.19 Correlation 3 3 x: 0.46 -

13 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

The global quantities predicted for the lower Re case are summarized in Table 3 on the page before along with others' results. The predicted CD values (0.438 and 0.433 for the DSM and LDKEM, respectively) are in fair agreement with the often-quoted experimental value of 0.4 measured in 1920's34 and other predictions. Constantinescu et aL28 predicted CD at around 0.4 using LES and detached eddy simulation (DES) on a structured hexahedral mesh having 450,000 nodes. Our predictions are closer to the value obtained by Pelaez et aL.30 (CD : 0.43) who carried out a DES on an unstructured mesh with 770,000 nodes. The Strouhal numbers predicted by the two dynamic models came out very close to each other, matching the measured one quite closely in view of the scatter in the experimental data.

The data of Achenbach31 and Kim and Durbin 3 5 favor lower Strouhal number around St = 0.15, whereas Sakamoto's data3 2 suggests a higher value between 0.18 and 0.19. The locations of flow separation (,) predicted by the two dynamic models are nearly identical and were found in 860 - 87°. The locations of laminar-to-turbulent transition (i,), which were obtained in this work by reading off the angle beyond which VI or krs, increases rapidly, were found in 86° - 880 for both dynamic models, which compare well with the LES predictions by Constantinescu et aL28 103 I

I* I '

I_ . l. __ Experimental .:3 lDSM Il \ 11 o LES with DSM 1LDKEM I 10LES with LDKFM 0.50 : : _ : [ = *Laminar caic.:

10 Ij CL) 0.45 -~j2 0.40 : . 1 \ ~~~e.DIEHEi 0.3 . . . .2

rD 200 300 400 500 600 10o _lop.1 0 I01 le _lee t UJD (a) Time histories of CD for ReD = 10,000 (b) Comparison of the predicted time-averaged CD values (ReD =

300,10.000,1.14 x 10o)and the experimental mean CD-curve for a range of Reynolds number FIgure 4. Time histories of drag coeficient (CD) and and the predicted mean drag coefficients for the sphere With the higher-Re case, the flow in reality has already undergone the "drag crisis", and the flow structure has changed drastically from those of subcritical regime. The change in the flow structure can be seen from Figure 5 and Figure 6. Depicted in these figures are the iso-contours of the second-invariant of the deformation tensor, (fljjQ,} -

SjjSj)/2. Both figures aptly portray the hairpin-like vortical structures inthe wake observed in experiments. The wake for the higher-Re case (Figure 6) is much narrower than in Figure 5, which is the consequence of the delayed onset of flow separation. The CD values predicted by the DSM and the LDKEM are 0.139 and 0.142, respectively. These values arc fairly close to the range of values (CD = 0.12 - 0.14) measured by Achenbach. 31 For the higher-Re case, the predictions of the locations of the separation and the onset of transition were much less satisfactory. At Re = 1.14 x 106, the experimental results 31 show that the transition occurs near 97° - 98° well before the separation occurring near 1200. The present predictions failed to reproduce this experimental finding. The present results exhibit too early a separation at around 100° and a delayed transition. This discrepancy is most likely due to the use of too coarse a mesh in this work to resolve the very thin boundary layer for the high-Re case.

The drag coefficients predicted in this study for the three Reynolds numbers are plotted in Figure 4 along with the mean experimental curve.

14 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

Figure S. Vortical structure In the near-wake of the sphere for ReD = 10,000 . visualized using the iso-contour or the second.invariant of the velocity deformation tensor Figure 6. Vortical structure In the near-wake of the sphere for ReD = 1.14 x 106 - visualized using the Iso-contour of the second-Invarlant of the velocity deformation tensor 15 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

VI. Summary and Conclusion A large eddy simulation capability based on a finite-volume solver has been developed and validated for a num-ber of wall-bounded flows. The finite-volume solver employs second-order numerics and permits use of unstructured meshes, thus being able to easily handle industrial applications involving complex geometry. Turbulence closure for subgrid-sale stresses is effected using two dynamic subgrid-scale viscosity models, namely, the dynamic Smagorin-sky model (DSM) and the localized dynamic k-equation model (LDKEM). These two dynamic models allow one to compute arbitrary three-dimensional flows without any statistically homogeneous directions. The validations demon-strated that the present LES capability is capable of predicting the wall-bounded flows of varying complexity with a commendable accuracy, having a potential to provide a practical tool for high-level simulation of turbulent flows encountered in industrial applications.

VII. Acknowledgments The author gratefully acknowledges use of Fluent Inc.'s software and thanks the members of the development group at Fluent Inc. Special thanks go to Sunil Vijay Unaune at the Fluent India who contributed the mesh used for the computation of the flow past a sphere.

References lHaworth, D.C. and Jansen. K., "Large-Eddy Simulation on Unstructured Deforming Meshes: Toward%reciprocating IC Engines:' Compuiers Fluids. Vol. 29.2000. pp. 493-524.

2 K.Mahesh, G.Constandnescu, P. M., "Large Eddy Simulation of Gas Turbine Combustors:' Annual Reserach Briefs, Center for Turbulence Research. NASA Amesl8tanford University. 2000. pp. 219-229.

3 Urbin, 0. and Knight, D, 'Large Eddy Simulation of a Supersonic Boundary Layer Using an Unstructured Grid,' AIAA Journal, Vol. 39, No.7. July 2001, pp. 927-935.

4 K. jnsen. T. M. and Reba, R., 'Finite-Element Based Large-Eddy Simulation of the Near-Nozzle Region of a Compressible Round Jet,"

Aiaa paper 2002-2358,2002.

5 S. Benhamadouche, K. M. and Constantinescu, G., 'Colocated Finite-Volume Schemes for Large Eddy Simulation on Unstructured Grids,'

Proceedings of the Surmer Program, Center for Turbulence Research, NASA Aames/Stanford University, 2002, pp.143-154.

6 M. Germano, U. Piomelli. P. M. and Cabot, W. H.. 'Dynamic Subgrid Scale Eddy Viscosity Model." Physics of Fluids A, Vol. 3. 1991, pp.1760-1765.

7 Lilly, D. K.. -A Proposed Modification of the Genmano Subgrid Scale Closure Method." Physics of Fluids A. Vol. 4. No. 3, 1992, pp. 633-635.

lKim, W.-W. and Menon, S., "A New Dynamic One-Equation Sub-Grid Scale Model for Large Eddy Simulations,' Aiaa paper 95-0356, 1995.

9 Kim, W.-W. and Menon. S.. "Application of the Localized Dynamic Subgrid-Scale Model to Turbulent Wall-Bounded Rlows:' Alaa paper 97-0210, 1997.

0

' Smagorinsky, 1., 'General Circulation Experiments with the Primitive Equations. part !: The Basic Experiment:' Monthly WeatherRNeview.

Vol. 91, 1963, pp.99-165.

t*S. Ghosal. T. Lund, P. M. and Akselvoll, K., "A Dynarnic Localization Model for Large Eddy Simulation of Turbulent Flows.' Journal of FluidMechanics, Vol. 286.1995, pp.229-255.

2 1 Davidson, L, "Large Eddy Simulation: A Dynamic One-Equatio Subgrid Model for Three Dimensional Recirculating Rlow:' Proceedings 26 1 26 6 of the 11th InternationalSymposium TurbulentShear Flows, Vol. 3, 1997, pp. . - . .

13 A. Sobankar. L D. and Norberg. C, "Large Eddy Simulation of Flow Past a Square Cylinder Comparison of Different Subgrid Scale Models." Journalof Fluids Engineering, Vol. 122. No. 1,2000. pp. 376-404.

4 1 Krajnovic, S. and Davidson, L. "Large Eddy Simulation of the Flow Around a Bluff Body,' AIAA Journal Vol. 40, No. 5. May 2002, pp.927-935.

t5S. Liu, C. M. and Katz. J.. 'On the Properties of Similarity SUbgrod-Scale Models as Deduced from Measurements in a Turbulent Jet,"

Journal of FluidMechanics, Vol. 275, 1994, pp. 83 -1 1 9 .

16S. Liu. C M. and Katz, J., Experimental Study of Similarity Subgrid-Scale Mode Ls of Turbulence In the Far-Fkildof a Jet, Direct and Large Eddy Simulation, Kluwer, 1994.

t7 Kader, B. A., "Temperature and Concentration Profiles in Fully Turbulent Boundary Layers,' InL J. IearMass Transfer, Vol. 24, No. 9, 1981, pp. 1541-1544.

1SMathur. S. R. and Murthy.J. Y.,"A Pressure-Based Method for Unstructured Meshes:'NumericalHeat Transfer, Vol.31. 1997. pp. 195-215.

16 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

I 9S.-E. Kim, S. R. Mathur. J. Y. M. and Choudhury. D, 'A Reynolds-Averaged Navier-Stokes Solver Using Unstructured Mesh-Based Finite-Volume Scheme." Aiaa paper 9840231, 1998.

20S.-E. Kim, B. M. and Caraeni, D., "A Multidimensional Linear Reconstruction Scheme for Arbitrary Unstructured Grids," Aiaa paper 2003-3990. Jan. 2003..

2 IL Davidson, D. CokIjat. J. F. M. A. L C. M. W. R.. 'LESFOIL: Large Eddy Simulation of Flow Around a High Lift Airfoil.".Notes on Nunerical FluidMechanics and Mutldisclplinary Design - Volume 83, Springer, 2002.

2J. Kim, P.M. and Moser. R., Turbulent Statistics in Fully Developed Channel Flow at Low Reynolds Number, JournalofFluldMechanics, Vol. 177, 1987, pp. 133-166.

23D. Choi, D. Prasad, M. W. and Pierce, C., 'Evaluation of an Industrial CFD code for LES Applications," Proceedingsof the 2000 Summer Program,Center for Turbulence Research, NASA AamenisStanford University, 2000, pp. 221-228.

24 D. A. Lyn, S. Einav, W. R. and Park, 1. H., 'A Laser Doppler Velocitimetry Study of Ensemble Averaged Characteristics of the Turbulent Near-Wake of a Square Cylinder." Journalof FluidMechanics, Vol. 304,1995, pp. 285-319.

25C Fureby, G. Tabor, H. G. W. and Gosman. A. D.. "Large Eddy Simulation of the Flow Around a Square Prism." AIAA Journal,Vol. 38, No. 3, May 2000, pp. 442-452.

26Johnson. T. A. and Patel. V. C. "Flow Past a Spbere up to a Reynolds Number of 300:' Journal of Fluid Mechanics. Vol. 378, 1999, pp. 19-70.

"A. G. Tomboulides, S. A. Orszag, G. A. K., "Direct and Large-Eddy Simulations of Axisymmetric Wakes." Aiaa paper 93-0546. Jan. 1993.

a2G. Constantinescu, M. C and Squires, K.. Turbulence Modeling Applied to Flow over a Sphere:' AIAA Journal, Vol. 41, No. 9, 2003, pp. 1733-1742.

29Roos. F. W.and Williams, W. W., 'Some Experimental Results on Sphere and Disk Drag:'AIAA Journal, Vol. 9. No. 2, 1971, pp. 285-291.

WJ. Pelacz, D. J. M. and Kandil, O., 'Unsteady Analysis of Separated Aerodynamic Flows Using an Unstructured Multigid Algorithm," Aiaa paper 20Jt1-Uo , Jan. 2001.

3t Achenbach. E., "Experiments on the Flow past Spheres at Very High Reynolds Numbers,"JournalofFluidbMechanics.Vol. 54, No.3.1972, pp.565-575.

32Sakamoto, H. and Haniu, H.. "A Study on Vortex Shedding from Spheres in an Uniform Flow,' Journal of Fluids Engineering, Vol. 112, 1990. pp. 386-393.

33 White, F. M., Viscous FluidFlow, McGraw-Hill, 1974.

1 3 Schlichting, H.. Boundary-LayerTheory, McGraw-Hill, 1979.

35 Kim. K. J. and Durbin, P. A.. "Observation of the Frequencies in a Sphere Wake and Drug Increase by Acoustic Excitation,". Physics of Fluids,Vol. 31, No. 11, 1988. pp. 3260-3265.

17 of 17 American Institute of Aeronautics and Astronautics Paper 2004-2548

ML052220149

Text

Subject:

Subject:

Subject:

Title:

Conclusions:

Title:

2.1.3 Heaters

3.0 CONCLUSION

4.0 REFERENCES

2.1.3 Heaters

3.0 CONCLUSION

Title:

SUMMARY

SUMMARY

1.0 Background

References:

SUMMARY

SUMMARY

Subject:

Title:

Reference:

Title:

Reference:

Title:

Reference:

Title:

Title:

Title:

SUMMARY

Navigation menu

Search