ML20084R337

Rev 3 to TDR-406, Steam Generator Tube Rupture Procedure Guidelines

ML20084R337
Person / Time
Site:	Three Mile Island
Issue date:	05/17/1984
From:	Hukill H GENERAL PUBLIC UTILITIES CORP.
To:	Stolz J Office of Nuclear Reactor Regulation
References
5211-84-2093, TDR-406, NUDOCS 8405220310
Download: ML20084R337 (82)

Text

a.

e 406 g TDR NO. _ REVISION NO. @%

I 01 TECHNICAL DATA REPORT IVNY NO. PAGE OF PROJECT: Dil-1 DEPARTMENT /SECTION RELEASE DATE 5,!20 b REVISION DATE I4 S D DOCUMENT TITLE:

SG Tube Rupture Procedure Guidelines ORIGINATOR SIGNATURE DATE APPROVAL (Si DATE TUR)

$% k r* w:rC '

I11'8 "

ff '

/ /2lj//]

Z[ Rm ta.2 sy APPROVAL FOR EXTERNAL DISTRIBUTION DATE hYmhk bl L -i 3 Does this TDR include recommendation (s)? Eyes O No if yes, TFWR/TR

• o DISTRIBUTION A8STRACT:

• R. W. Bensel This document provides technical guidelines for dealing D. J. Boltz with single and multiple tube ruptures. A significant T. G. Broughton improvement in procedures will result from reduction of 1

• M. Campagna the minimum subcooling margin and RC pump trip on loss

• P. R. Clark of subcooling margin, waiver of fuel-in-compression

• J. J. Colitz limits, and revised RCP NPSH limits. Other benefits can I R. Finfrock be derived from revision of the RC pump restart criteria T. L. Gerber and from additional guidance regarding OTSG steaming and

, R. J. Glaviano isolation. Finally, revised guidance is provided for

'

• R. W. Keaten preventing tube leak propagation. It is recommended G. Lehmann

• that the tube-to-shell delta T be limited to 70F0 D. T. Leighton

• during tube rupture events.

B. Leonard W. W. Lowe d Revision 2 to this TDR included the following J. G. Miller

• recommendations for procedural revisions, some

• T. Moran

• of which have already been incorporated in j M. Nelson

• EP 1202-5.

• S. Newton

• M. J. Ross

• 1. Isolate the OTSG's on a measured or projected i

• H. B. Shipman dose rate of 50 mrem /hr whole body or 250 mrem /hr D. G. Slear thyroid dbse.

l C. W. Smyth

)

• M. J. Stromberg

• 2. Stop the non-ES HPI pump if the RCS is cooling R. J. Toole more than 100F/hr.

P. S. Walsh -

• Dr. R. N. WhiteseL

• 3. Priorities should be spelled out in EP-1202-5:

• R. F. Wilson a. Minimizing SCM has a priority over mini-mizing cooldown time. (

l l

b. Keeping OTSG 1evel below 95% is less im-i portant than control of the RCS cooldown B405220310 840517 rate. k f PDR ADOCK 05000289 i P P1)R I  !

. ~ . . . . _ - 20 . 2

i TDR 406 Rev. 3 Page 2 of 81 i

l ABSTRACT (Cont'd)

• 4. Initiate the DHR system at 300F under tube rupture conditions.

• 5. Trip the reactor if pressurizer level cannot be  !

d maintained above 150 inches with two HPI pumps on.

• 6. Raise the unaffected OTSG level to 95% before raising the affected OTSG level to 95% unless incore temperatures are not decreasing and there is no OTSG heat transfer.

I

• 7. If RCPs are not tripped within two minutes of a loss of SCM, maintain 1 RCP in each loop runt.ing.

d Revision 3 of this TDR provides guidelines for possible reduction d of offsite doses under specific plant conditions. These considerations d address deviations from OTSG steaming and isolation criteria which can d be evaluated as part of the long term response to a tube rupture. It d also includes some additional guidance on the use of shell thermo-d couples and address conflicts in requirements for RCP NPSH and DHR

, d maximum pressure.

+ Rev.1

• Rev. 2

, d Rev. 3

TDR 406 R:v. 3 Fage 3 og 81 S G TUB E RU PTURE PRO CE3 tRS GJ IDE L IN ES TDR d406

+ dev. L

• Rev. 2 d Rev. 3

. l l

TDR 'o0o i Rev. 3 i Pa ge 4 of 81 l TABLE Ol C3NTEiTS Page 4

Table o f C on ten ts . . . . . . . . . . . . . . . . . .

Lis t o f Ft guces . . . . . . . . . . . , . . . . . . . . 7 Lis t of Tables ................ . . . 8 Summary of Changes .. . . . . . . . . . . . . . . . 9 1.0 INTRODJCTION AND BACKGR3dND . . . . . . . . . . . . . 14 2.0 TE01 FUNCf10N3 SGfa PROG 3URE GJIDELLii.S DE VE LO PMENT PRO GR AM . . . . . . . . . . . . . . . . . 15 2.1 Development o f Des ign Bas is Guidelines . . . . . 15 2.1.L Litera ture Search . . . . . . . . . . 15 2.1.2 Limiting OT3G Tube Stresses . . . . . 17 2.1. 3 Steaming. Isolation and Filling o f the Leaking OTSG . . . . . . . . . . 18 2.1.3.1 Steaming, Isolation and Filling wL th bo th OT3 G's Laa ch in g . . .. . . . .. 20 2.1. 4 Minimum Allowable Subcooling Margin . 20 2.t.5 Waive Fuel in Compression LLm!ts .. . 21

2. t . 6 Reactor Coolant Pump NPSH LLmles . . . 21 2.t.7 Procedure Entry Point Cond!tton ... 21

2. t .S SLmula tor Exper tence . . . .. . . . . 22 2.1.9 Emergency Ltmlts for Decay Hea t System Initiation . . . .. . . . .. 22 2.1.10 Reactor f rip on Low Pressur tzer Level. 23 2.' 2 Development of Multiple Tube Rupture P roce dur e Guidel tnes . . . . .. . . .. . . . . 23 2.2.1 Revis ion o f RC P f elp and Res tar t C rite r ia . . . . . . . . . .. . .. . 24 2.2.2 OTSG Steaming and Level Control ... 25 2.2.3 Criteria for Feed and Bleed Cooling . 25 2.2.4 Cooldown/De pr ess uriza tion .. . . .. 26

• 2. 3 Add i tional Wo r k Re q uir eme n ts . . . . . . . . . . 27

• 2.3.1 Analyses , . . . . . . . . .. . . .. 27

• 2.3.2 Issue Rasolutton . . . . . .. . .. . 27 3.0 MAJOR REVISIONS ID EXISTINO PROCEJURE . ... . . .. 29

3. t Bas ic Plant Sta te . . . . . . . .. . . . . . . . 29 3.1.1 Assumed Plant Conditions . . . . . .. 29 3.t.2 Tube Rupture Guidelines for Loss of Subcooling . . . . . . . . . .. . 29 3.1.3 Revised Equipment Limits & Operatlng Procedures . . . . . . . . . . . . .. 30

+ Rev.'1

• Rev. 2 t C

T01 406 Rsv. 3 Pa ge 5 of 81 TABLE OF (DNTENTS (Cont 'd)

Page

3. 2 Discussion of Guldelines . . . . . . . . . . . . 31 3.2.1 Imme dia te Ac t ions . . . . . . . . . . 31 3.2.2 Followup Actions - Subcooling Maintained and RCP's Avanable . . . . 31 3.2.2.1 Ma inta in s Mintmam o f 152

• Subcooling Margin . . . . . . . . . . 31 3.2.2.2 Steaming / Isolation Celteria for th e Affsetad OT5G . . . . . . . . . . . . 31 3.2.2.3 Shell to Tube Delta T . . . . . . . . 32 3.2.3 lollowup Actions ( Automa tte Rx Trip has Occurred) . . . . . . . . . . . . . 32

3. 2. 3.1 Trip with a Loss of Subcooling Margin. 32 3.2.4 Followup Actions for Loss o f Subcooling . . . . . . . . . . . . . . 32 4.0 *S 1YJIATOR TRAINI.4G EXPERIENCE , . . . . . . . . . . . 34

• 4.1 IntroductLon . . . . . . . . . . . . . . . . . . 34

• 4. 2 Resul ts. . . . . . . . . . . . . . . . . . . . . 34

• 4. 2.1 January 1983 reaining . . . . . . . . 34

• 4.2.1.1 Comments . . . . . . . . . . . . . . . 35

• 4.2.2 J une 1983 f raining . . . . . . . . . . 35

• 4.2.2.1 Control of RCS Cooldown Rate . . . . . 35

• 4.2.2.2 Plant Stabilization Before Cooldown .

36

• 4.2.2.3 RCP Restart Celteria . . . . . . . . . 37

• 4.2.2.4 Core Flood Iank Isolatton . . . . . . 37

• 4.2.2.5 RCP Trtp Celterton . . . . . . . . . . 37 5.0

• S IN GLE AN') MULT IPLE S GTR GJ ID ELIN ES . . . . . . . . . 38

• 5.1 Sco pe . . .. . . . . . . . . . . . . . . . . . . 38

• 5.2 Gu tdelines and Ltmi ts . . . . . . . . . . . . . . 38

• 5. 2.1 Subcooling Margin Requirements . . . . 38

• 5.2.2 Reactor Coolant Pump TrLp Critec ton .

38

• 5.2.3 Reactor Coolant Pump Restart Celteria. 38

• 5.2.4 Resctor Coolant Pump NPSH for

• Emergency Operations . . . . . . . . . 38

• 5. 2. 5 HLgh Pressure Injection Throttling

• Cet teria . . . . . . . . . . . . . . . 33

• 5. 2. 6 O TS G Le v el . . . . . . . . . . . . . . 39

• 5.2.7 OTSG Isolation / Steaming Celteria . . . 39

• 5. 2. 7 .1 Pressure Control of an Isolated OTSG . 39

• 5.2.8 Cooldown Rate During a Tube Leak

• Event . . . . . . . . . . . . . . . . . 39

• 5.2.9 OTSG Shell-to-Tube Dif ferential

• Temperature Ltmit . . . . . . . . . . 40'  ;

• 5.2.10 Cooling Mode When Both OTSG's are ,

• Unavailable for RCS Hea t Removal . . .

40

• 5.2.11 Core Flood Tank Isolation . . . . . . 40

• 5.2.L2 Guideline Flow Chart . . . . . . . . . 40

+ Rev. 1

• Rev. 2

. M a

TDR 406 Rev. 3 Page 6 of 81 TABLE OF CONTENTS (Cont'd)

Page 6.0 (DNCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . 41 4

7.0 REFERENCES

. .................... 41 Appendix A: Comparison of Cuidelines to INPO and NRC Recommendations Appendix B: Procedure Change Safety Evaluations Appendix C: Guidelines Flow Chart Appendix D: Simplified Event Tree Appendix E: Process Computer Output d Appendix F: Additional Steaming and Isolation Considerations for Reduction of Radiological Releases -

• There are a total of 81 pages in this report including Figures, Tables and Appendices.

e i

+ Rev. 1

• Rev. 2 4 Rev.-3

l l

TDR 405 Ray. 3 Page 7 of 81 LIST 0/ IIGURES Title FL gure No.

Steam Generator Tube Rupture Guideline Develo pment Activlty Ne twork . . . . . . . . . . . . . 1 Break Flow For S!ngle Ruptured Tube . . . . . . . . . . . . 2 Ef fect of RO Pump Operat ton on Integrated Sfstem Leaxage for Single Ruptured Tube . . . . . . . . . . . . . . . . . 3 Mass and Energy Capabilittes of HPI and PORV . . . . . . . . . . . . . . . . . . . . . 4 TLma Behavior of Subcooling Margin for Snectrum of Ruptured Tubes . . . . . . . . . . . . . . 3 Emergency Reactor Coolant Pump N PS H L Lat ts . .. . . . . . . . . . . . . . . . . . . . 6 Single and MultLple Tube Rupture Gu ide l in es . . . . . . . . . . . ... . . . . . . . . . C- 1 S'.mpli.f ted OTSG Event Tree . . . . . . . . . . . . . . . . . 0-1

.o e

e Rev.1

• Rev. 2 d Rev. 3 I

g _

t +

I TDR 405 Rsv. 3 l t Page 8 og 81 LIST OF TABLE 3 Title Table No.

Tabular Vatu es of RCP Emergency NPSH Requiremen ts . . . . . . . . . . . . . . . . . . . . . 1 Pressurizer Spray Flow for Various Pump Combinations . . . . . . . ... . . . . . . . . . . . 2 Comparison of TDR #406 to Other Sources . . . . . . . . . . . . . . . . . . . . . A-1 d Com par is on o f We i. gh te d vs . A r t th me t * : Averaga . . . . . . . 5-1 h of OTSG Shell Thermocouples Shell Thermocouple Substitution . . . . . . . . . . . . . . . . . . . . . E-2 4 Wide Range T cold Input Substitutions . . . . . . . . . . . E-3 h Summary o f Dose Reduction Considerations. . . . . . . . . . F-1 4 t

O l

• j . . _

+ Rev. 1 i-

• Rev. 2 f Rev. 3 L_

~

l l

l TDR NO.

"6 MUCl68r TITLE SG Tube Rupture Procedure Guidelines, Rev. 3 PAGE 9 OF 81

SUMMARY

OF CHANGE APPROVAL DATE REV l

3 Figure 6, Table 1 and Section 2.1.6 and 2.1.9 7(6p( I-2 .r ;

were revised to address simultaneous DHR and RCP operation, clarify #1 seal staging requirements and specify that NPSH curves are dependent on the number of operable pumps per loop.

it 2 er 3 Section 2.3.2 on issue resolution was updated. < (, g(

Additional work items have been noted.

3 Section 3.2 was updated to agree with the g, d ;( 112-f ;

guidelines of Section 5.

3 Appendix E was amended to provide guidance in g g ;( ,2 2 9. ?

calculating shell temperatures if the weighted average is not availabic from the computer; or if only thermocouple voltages are available in the control room.

3 Appendix F was added to address long term dose /*/ d ;( 11 A*fi reduction considerations. Under certain plant conditions variance from the OTSG steaming isolation criteria wiil allow a reduction in dose. This TDR recommends that those guides be appended to the tube rupture procedure.

l

?

I Aooo 0o17 9 80 l

tor uo.

ENuclear

~

406 TITLE SG Tube Rupture Procedure Guidelines, Rev. 2 PAGE 10 OF 81

SUMMARY

OF CHANGE APPROVAL DATE REV 2 Section 2.2.3. Expanded discussions on /s/ 5/12/8; feed and bleed cooling to include ADV and TBV mass and energy relief capabilities relative to decay heat and leak rate.

2 Section 2.3. Discussed additional work /s/ 8/12/8:

which will be addressed in a future revision to TDR 406.

2 Section 4.2.2. Added note regarding actions /s/ 8/12/82 to be taken if RCPs are not tripped within 2 minutes of a loss of SCM.

~

2 Section 4.2.3. Added note regarding loss of /s/ 8/12/83 SCM after RCPs are restarted.

2 Section 4.2.6. Revised guidance on raising /s/ 8/12/83 OTSG 1evels to 95%.

2 Section 4.2.7. Added isolation criterion on /s/ 8/12/83 250 mrem /hr. -

2 Section 4.2.11. Added criterion for Core /s/ 8/12/83 Flood Tanks Isolation.

2 Section 6.0. Added recommendations to isolate /s/ 8/12/83 CFTs using criterion provided and initiation of DHR system at 300F.

2 Section 2.2.1. Clarified that emergency NPSH /s/ 5/12/83 curves should be followed for both RC pump trip and restart.

2 Sections 3.2.4 and 4.2.3. Start one RCP per /s/ 5/12/83 loop or both RCP's in the same loop.

2 Section 3.2.2. Revised to be consistent with /s/ 8/12/83 the steaming criterion in Section 2.1.3.

Aooo 0017 9 60

TDR NO.

Nhg 406 TITLE SG Tube Rupture Procedure Guidelines, Rev. 2 PAGE 11 OF 81

SUMMARY

OF CHANGE APPROVAL DATE REV 2 Added detail to Table of Contents and reversed /s/ 8/12/83 the order of Sections 4.0 and 5.0.

2 Added Tables 1 & 2 which provide tabular data on /s/ 5/12/83 RCP NPSH requirements and on spray flow for various RCP pump combinations.

2 Section 2.1.3. Added recommendation for OTSG /s/ 8/12/83 isolation if iodine release rate exceeds 250 mrem /hr or whole body dose rates 50 mrem /hr, correcting error in previous revision.

2 Section 2.1.6 and Figure 6. Revised /s/ 5/12/83 emergency NPSH limits to account for cal-culated instrument errors during LOCA conditions.

2 Sections 2.1.8 and 4.2.2. Added dis- /s/ 8/12/83 cussion on the experience gained from the June 1983 Lynchburg simulator sessiohs.

2 Section 2.1.9. Added recommendation to /s/ 8/12/83 allow DHR system initiation at 300F instead of 275F.

2 Section 2.1.10. Recommendation to trip /s/ 8/12/83 reactor if 200 inch pressurizer level can-O not be maintained with two HPI pumps runn i.19 2 Section 2.2.2. Clarified guidance on /s/ 8/12/83 isolation criterion with leaks in both OTSGs.

2 Section 2.2.2. Addressed raising OTSG /s/ 8/12/83 level to 95% without causing an overcooling.

2 Section 2.2.2. Discussed why EFW should not /s/ 8/12/83 be used to control 0TSG pressure in an isolated 0TSG (chan ing previous recom- .

mendation of Rev. 1 .

l Aooo 0017 9 So

tor No.

ENuclear 406 TITLE SG Tube Rupture Guidelines, Rev.1 PAGE 12 OF 81 l

SUMMARY

OF CHANGE APPROVAL DATE REV l

1 Minor editorial changes and correction of /s/ 5/8/83 typographical errors on pages: 2,5,7,10,13, 19,20,22.A-1,A-3,A-4,B-1,A-2,B-2,6,18.

1 Revised cover page to show shell-to-tube /s/ 5/8/83 delta T can be controlled below 100F.

1 Added a List of Tables pp i & iii /s/ 5/8/83 1 Included use of MFW as means to cool OTSG /s/ 5/8/83 shell. p5 1 Indicated that continuous steaming of 0TSG is /s/ 5/8/83 simplest means of meeting OTSG level, pressure and differential temp. considerations' pp. 5,6,10,18 1 Eliminated reference to RAC for determining /s/ 5/8/83 when to isolate OTSG based on radiological conditions. p6 i Added Section 2.1.3.1 to discuss control when /s/ 5/8/83 both OTSG's are isolated. (Also p 10).

1 Provided discussion and Figurr. ror RCP NPSH /s/ 5/8/83 limits. Section 2.1.6 and Figure 6. Ref 25,26.

& Sections 4.2.1 and 4.2.4.

1 Revised explanation of ADV & TBV flow capability /s/ 5/8/83 relative to OTSG flooding (incorrect in Rev 0) pp 11,16 1 Added Reference to B&W guidance which allows /s/ 5/8/83 cooldown at 100F/hr during Tube Ruptures

, without a soak time even if cooldown rate is exceeded. p 11.& Ref 24 l

1 Section 4.2.3 revised to account for inability /s/ 5/8/83 to start either RCP in the A loop.

~

1 Added Section 4.2.7.1 and revised 4.2.7 to /s/ 5/8/83 l address Steaming Isolation of 0TSG considering the continuous steaming philosophy.

l 1 Simplified Section 4.2.8 on cooldown rate. /s/ 5/8/83 l

1 Added Section 4.2.9 on controlling GTSG shell-to /s/ 5/8/83 tube differential temperature.

AooO o317 9:8o

I TDR NO. l

@ $$f 406 j TITLE SG Tube Rupture Guidelines, Rev. 1 PAGE OF

, 13 gl REV

SUMMARY

OF CHANGE APPROVAL DATE 1 Added Section B.7,B.8, and B.9, which were left /s/ 5/8/83 out of Rev. O inadvertently.

1 Deleted redundant Section of Part 4 (guidelines) /s/ 5/8/83 1 Added Section C.2 through C.6 to discuss the /s/ 5/8/83 Guidelines Flow Diagram (Figure C-1) in words.

I Rewrote Appendix.E on Process Computer Output /s/ 5/8/83 and Alarms.

1 Revised Figure 4 to show decay heat levels /s/ 5/8/83 as a function of time.

e 1

l

! l 1

l Aooo 0017 9 80 l

9 T M 406 Rsv. 3 Page 14 of 81 1.0 I41R3DUCfl0N AND BACXGRO'JND In November 1981, primary to secondary sids leaks were discovered in the tubes o f bo th o f the 1MI-l Once Ihrough Steam Generators (O TS G). there are 13,531 tubes in each OTSC. The plant design basis for a steam generator tube rupture (SGTR) aceldent is the double ended offse t severence of a single tuba. Since extensive cLecumferentL41 cracking was discovered Ln approxLmately 1200 of the 31,000 tubes, Lt became clear that a revised set of procedures for dealing wL eh both single and multlpte SGras should be developed.

This report describes a program whlen has been formulated to Laprove existing procedures and operator training by providing imp' roved operator guldelines for dealing with tube leakage and tube rupture events. the guidelines development program will be described in detail, and the major revisions to the existing procedures whLeh have been (dentlfled as part of the program will bs discussed. The proposed guidelines will then be presented in tarms of their overall scope, with a step by step discussion of required operator actions. the analyttcal' evaluattons which are the basis for the reconnandations, consist of a series of simulations whLeh are ongoing and will be documentad in detail in a s ub seq uen t r epo r t. The guidelines in this TUR were tasted at the B&W simulator training cycle beginning in January,1983. The results of this training exper tenes are discussed. Finally, the overall conclus ions and major recommendations of the guldelines development program are documentsJ.

e G

G W

e-.

> Rev. 1

• Rev. 2 f Rev. 3'

TDR 406 Rsv. 3 Pa ge 15 of 81 2.0 TE CH FUN CT IO N3 S GTR GJ D E L IJ ES DE VE ID Rf EN T PRO GlM Elgure 1 shows the executLon of the sesam generator tube rupture gu'.deline development pro gram. The plan has three main paths: Pa th 1 is the development of design basis tube rupture guidelines. Path 2 is the development of multiple tube rupture guldelines; and , Pa th 3, is a benchmark ef fort to compare the RETRAN and RELAP 5 computar co d es. This last effort also includes an evaluation of the B&W AIDO analysis of a single tube rup:ure using MLitfRAP. The purpose of this TDR is to explain paths 1 & 2. The benchmarking and comparison efforts are discussed in a separate IDR describing all of the tube rupture analysis work. None o f the com pu ter analys is o f Pa th 3 h as been used to justlfy the recommendstLons of this report. The analyses were an aid in concaptualizing the pnysleal processes during a tube rupture.

2.1 Development of Des ign Bas is Guidelines (Path 1)

The major activittes involved in developing this part o f the guldeline were to:

1. Search exLstLng industry events and procedures for lessons to be learned about handling tube ruptures.

2 Define allowable steam generator stresses during cooldown (e t ther as cooldown rate or as tube / shell delta T).

l

3. Determine when OTSG's should be isolated and when they should be s teamed .

4 Revise the minimum allowable subcooling margLn.

1. Waive fuel in compression LLmits.

5.

!

• 6. Develop ~ emergency RCP N PSH limits.

7 Rede fine entry point c ondi tions .

• 8. Factor in experience from use of the guidelines on the B&W slaula tor .

Each of these items are Alscussed in de tail in the following sections.

2.1.1 Literature Search Several tube rupture leaks have occurred at various operating reactors wL thin the last four years. The experience gained froa these events has offered us an opportunity to improve tube rupture guldel ines. The major lessons learned from these events have been

+ Rev. L

• Rev. 2 d Rev. 3 x *l

4

, TOA 400 i Rev. 3 i Pa ge 16 of 81 a

sussar1:ed in various doeumenes fcon the NAC, IN PO, and plant i procadures and Lncluded Ln the BW ATOG tube rupture guidelines l

(Re ferences 1-10). The lessons includt the following: l 4

1. Subcooling margLn should be minimLzed to alnialze primary to secondary leakage. Subcooling is atintained by keeping the ACS temperature below the saturation temperature with OTSG cooling. <

Since the OTSG is in a saturated condition, Le is always lower tn pr es sur e than the RCS L f s ubcooling t s ma in ta ined, i Thera fore , keeping subcooling margLn a e or near L ts tainimma ,

4 acceptable value reducts leakag2.

) In order to maintain the ainlaan subcooling margin, several i

plant liatts have to be violated: fuel pin-!n compression

Itates and RCP NPSH Limits. the former is acceptable to violste j during emergancy condLttons, while the Latter has been ,

a reevaluated to determine acceptable emergency operation of ene pump.

2. RCP's should be maintained rieming for several ressons. Pump trLp on loss of subcooling margin allows the operator to maintain forced ftw for a leak size of up to several tubes

! Wile 1600 peig ES AS is mach more restrLettve. Forced RC flow .

provides s'everal benefles during a tube rupture. Flest, they ,

assure that stese volds do not form in the hot leg L! bends or j upper vessel head. Steam voids in these locatLons can interrupt.

! natural circulation or prevent RCS depressurization. Se cond ,

J RCP operatton results in a lower primary to secondary dif ferential pressure for a given subcooling margin (since core detta T ts samtler with' the RCP's running). FLnally, w Lth RCP's running, pressurizer spray is ava ttable and RCS pressure control

! is not dependent on the PORY or pressurizer vent.

! /

l Main feedveter can be used Lf RCP's are running; wLth pumps of f.

emargency feedwater mast be used, which is less ef fective in cooling the 015G shell, thereby increasing tubs' to sheLL delta -

T. (L.e., tube tensile loads). .

I

{ 3. RCS p essura should be asintainad 1ow enough to pcevent secondary s ide safe ty valves from lif t Ln,g. HPI flow was not 4

throttled suf fLeiently in the Ginna event of January 23, 1982 i

! and the s team generator fL Lled with water. SLnce the RC3

i.

• pressure was above the SG safety valve setpoint , the sa fetles opened resulting in an atmospheric release of radLoactLvity.

Moreover, the safeties were forced to pass liquid, *Leh mLght cause the open failure of the valves.

4. RCS Degassing

- RCP NPSM limits at Glana requ' t red shuttLng down of the reactor l

coolant pumps at Low pressures. Shutting the pumps allowed

p,ncondensible gases to collect Ln the top of the steam i t

l

! + Rev._t

• Rev. 2 l Rev. 3

TD1 406 Rev. 3 i y,

Page 17 of 81 f f generator tube U bends. These trapped gases prevented RC3 i depressurizatLon for many hours. An analogous situa tLon mLght occur at the hot leg U bends. the 1MI-l desLgn has always nad j capability of ventLng noncondensable gases from the U bends,  !

i however, whleh can be used Lf RCP's are not available . 7 I 5. WST Inventory l The Oconee tube leak of September 18, 1981 resulted Ln a 4 susta tned (17 hour1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br />) leakage from the RCS to the 013G's. Inis I leakage caused the. generator to f LLI. In order to prevent s team

Itne filling, the operatoes at Oconee transferred water out of j' the OTS G's . In effect there was a once through cooling path j from the WST through the core and out of the 3ISG's. fhis '

j experience illustrated the need .co assure adequate BWST .

Inventory for core cooling. Second , it hLghlighted the need to

. store cadloset tve water in the plant during a prolonged RC3

! - cooldown.

l l

• 6. Shett-to-Tube Delta T l- >

! A tube Leak at Rancho Seco in May 1981 yleided evidence of the

, Laportance of controlling 013G tube / sh' ell delta T. the axisting

! 11mits and precautLons at TAI-l is 100c '. Howevee, 'aefore

! tube / shell delta T exceeded 100f * , the leaking tube was placed

! under tensile stress and the tube was pulled Lato a

! circumferenttal tear. Ms tntaining tube / shell delta T limlts are t Laportant during tube rupture and are dLscussed in more detail below.

i ,

i i

2.1.2 Limittna 015G Tube Stresses .

j Stesa generator tube stresses are generatad as a result of tansite

[ loads placed on the tubes. Chese tens (le loads come from two load

componen ts . The. f test Ls the temperature differenttal between the

! tube and steam generator shell. . As the RCS temperature decreases ,

tube temperature decreases. At sons point the dL fferenca in ~

tempersture be tween the colder tubes and warmer shell. is suf fleient to result in tensite stresses' that pull apart a leaking tube. This 1 topte has been the subject of extensive analyses within GPJN in l

conjunction with BW, EPILI, and MPR and the subject of a separate t teport (see Re f . 15) .-

i ..

1 The second Load component is from OTSG pressure loedtng on. the

tube shee t whleh caus es alon ga t ion o f th e - shell. Isolation o f the

015G causes the tube / shell dLfference to increase while adding a tensile load on the tubes by elongating the shell v'ia pressure.

loedtag. Structurally there are compensattng a ffects involved Ln mittgeting these two load contributors. RapLd depressurisation olistnates the pressure induced stress but aggravates - the delta T

• f

+ Rev. 1 1

• Rev.,2-

.f Rev. 3'

,. .- . .- - -- .. - .L , . .,

__ ~. _ _ _ . . __ _ _ . . . _ _ _ ...._ _ - . . _ _ _ _ _ - . _ _ _ _ _

TDR 406 Rev. 3 da ge 18 of 81 I

Induced s tresses. The optimum OI3 G cooldown/da pressur tzation ra te has not been determined. However, Lt Ls known that isolatlng tne

) O T4G a t 1000 ps i s is no t the b es t me ans o f re du c ing s tr es s . i i Cooldown/ depressurization is the preferred me thod.

L There are three limits for tube / shell delta T that presently apply to i~

. TMI- 1. Plant "Llatts and P recauttons" (Re f. 22) limit delta T to

60i' during heatup and. to 1007' during cooldown with one OTSG i

isolatad. This value of 100i' assumed that tubes had no more than

40% through-wall cracks. In re ference 23, &SW established 142/

• for l a cooldown using both OTSG. The 70l* value in this TDA la proposed i as a gulds in determining an acceptable cooldown rate. If delta T '

l can be maintained at or below 70i", the operatoe has optimized the

~

plant cooldown rate. The 70F* llatt more conservatively assumes that i tubes in the OTSG are leaking below a detectable L Lait. A 70l

• value ,

[\

lletts propasstion of these cracks.

+ Control of sheLL to tube delta T ls accompilshed in several ways.

4

+ First by cooling the 0T30 Liquid (steaming) to allow the metal shell i + co cool. Second, by providing cool, main faadwater into the j

+ downconer. If net ther of these me thods works , tha RCS cooldown must j + be dscreased until the OT3G shell cools sufficiently. If reducing

+ the cooldown doesn' t woek, then .the cooldown aus t be terminated .

l 2.1.3 Steam'ng. Isolatton and Filling of the Lasking OTSG Isolacton of the leaking OTSG can result in the overfL1 Ling of that gen era tor. It is pra ferable to prevent overf t1 Ling , however, to l

allow plant cooldown in an expedttlous manner. If the OT3G f ;11s , I t i becomes a large pressurizar (as evidenend by the Gtana event). The l tLee Le took to cool down this mass of hot water greatly extended the

, cooldown of. the plant. Steaming also mainta tns some natural

" circulatton flow in the hot leg. This flow cools the hot leg U bend and decreases the chances of steem vold formation, j, ,

l

+ As dtseuased in Section 2.1.2, st aalng and depressurizatLon of the d OT3G also reduces OTSG tube stresses. However , de pr essur iza tion o f i the OTSG also increases leakage rate. As dlscussed in Saction 2. 2. 2, $

the OTSG pressure should be below RCS pressure to promote flow 1

through the hot leg. The opcLoum OTSG control results in j 1) depressurization of the OISG without causing large delta T's;

2) minimus RCi leakage; 3) promotion of natural eleculation flow in

' the hot leg; and 4) posttive Leakage from the RCS into the OI5G to assure hot les cooling Ln the absence of natural cleculation. The opetsua pressura control scheme to mest this celteria has not been  ;

4 determined analytically. ,

4 I + Rev.1

• Rev. 2 d Rev.;3 ,

_ __ ~~ _ . _ . _ ._ _

e j

TDA 605 s ., Rev. 3

,j* , m'a ge 19 of 81 l

i

+ Meeting all four of these crt teria vill probably result in a nearly

+ ' continuous s teaming o f the a ffected OTSG. Moreover , intermittent ,

& , s teaming o f- the OTS G's will g re.sul t in release of all'the noble gases '

& transportad into the OT3G f rom the RG. Therefore, the 101

& recommends celtinuous s teaming o f the OTSG's. The advantages of

& continuously steaming the affected OTSG's are: / 4

+ 1. All of the above OT3 G contvol cand.ltions tre :r.et .

s  !

+ 2. The operato( follows his normal cocaldown procedures.

+ 3. Pl an t : res ponse is symme tric . j

+ 4 Cooldown at low" pres bel eoperature t can. be a'ccomplished more quickly , allowing DH. sys tsa operation 9'o onev. n I' + * <

ContLnuous sea . ming thould result in a more p pi,d cootA un than intermittent,s teaming because of tube / sheir d_ elta T L LmLta t' ions.

Cooldown at 1001/hr us'hg the unaffected OI3G v1L L rest.it in a 70 i delta T llate, in 1-2 hours. From this time on, the OTSG would have to be steame:1 S:milarly,' the OT3G would have to be stasind to ma inta in nacurial circuladon. '

'v ,

s Ilthough it(ts h'ighly des irable to prevent s taam'Iiner f L LLing, there are certain circumstances whLeb ,dleiate that the(')T3&tshould be f L11 ed. The Engineering Mechani:se Section of GPMC has es tablished .

the capabiltty of the ' steam lines to sustaire the water hammer and deid load e ffects of. floodLng the s team L ines (Re f L L). This analysis shows that the loading 1.s acceptable wLthout pinning (except for - the dead load e ffects during, a design basis 4ar thquake). SLnce this combination of ' events is extremely gemote , the procedures- have been modLf ted to allow fLilingTI the CTSG.

( l .

Ifu guldelines have the operator , fill the O GG's only under two c '.r cums tance s . The flest coadition is that BWST level decreases below 21 ft. At this level, there is' still suf f Lclent inventory to flood both steam lines and put about 30,000 gs11ons of we.ter into the ,

containment building (Ref 12). Ihis amoun t o f wa ter is su f f L clent to-provida adequats NP3H in an I.PI to HPI "p tggyback" mode of core injection 'from the reactor building sump (Ref .13).

+ A 'second; reason to fLLt the OTSG ls for radlologLcal consideratLons.

+ The OTSG shoulo be tsolated Lf of fstte doses are approaching, levels which would requ' ire declaration of a SLte emergency. It'should.be noted that a St ea Emergency may already have -been declared based on

~

& OTSG 1eakage rate. , Nevertheless=, this level provides a ratLonale for dect ding that release rates are hLgh 'enough to warrant OI3G isolation.

\( ,

F e .

m

..y

1. ' + aav.1 A

.* Rev. 2

' d.Rev. 3-

~

u s .

'TDR 405 Rev. 3 lage 20 of 81 l

& ConsLderatton was given to def Lning isolation conditions based on AC3

+ act Lvity levels , me teorologf and s team line radia tion levels. RC3 activity level canno t be correla ted to of fs t te releases , since offsite dose will be a ffected by the location of the tube leak in ene OTS G, availabiltty o f the condenser and pla teout and decontamina tion f actor s. It ts also undesirable to isolate the OTSG based on assumed , me teorological cendL tions. Iha most desirable approach is l to isolate based on accust releases occurring during the event.

• Ihe existing Slts emergency limits are 50 mrem /hr whole body and

• 250 mrem /hr thyroid dose measure or projectad at the sita boundary

• ( Re f . 33). Section 2.l.9 discusses the length of tLme required to

' cool the plant down to DHR system condLttons. This length of time

• defines the integrated does allowed by tha guidelines (L.a. releasa

• rate for the s pecL f ted per iod o f t ime ).

2.l.3.L S t ea ming . Is ola t ion and F it t in g wi th Bo th O I3 G's Le ak ing

& Isolation and steaming of the OTSG's must be addressed for leaks in

& both OTSG's. Once RCS tamperature is below 5'4 0 F, a choice has to be F made regarding OTSG lsolatton. Both OISG's should be s teamed unless

+* elther the BWST level or of fstte release crLeeris is reached. If

+* both OTSG's are s teamed, then all s team loads from both OT3G's should b be isolated except for the TBV's/ADV's. ALI o ther s teaming ,

F tsolation and fLlling criteria should be followed.

• If the BJST level reaches 21 feet , then both OTSGs mus t be isola ted.

• If the dose criteria is reached, one OT3G should be isolatad and tae

• doses reevaluated. If the dose crL teria s till canno t ba met , tnen

• the second OTSG should be isolated.

2.1. 4 Mintmum Allowable Subcooling Margin

1. A primsey goal during a tube rupture is 'to minimize offsita dose.

Mintsizing leakage from the RCS is the first Line o f de f ensa .

Laakage from the primary to secondary Ls determined by the size of the leak, and by the dLf ferential pressure be tween the ACS and JISG.

Primary to secondary differentLal pressure is controlled by fixing

+ the degree of RCS subcooling. Once secondary pressure is fixad , cold leg temperature is de termined. Fo r any t Lme , d a cay hea t i s fixe d.

RCS flow (which is determLned by OISG 1evel or RCP operabiltty) tnen determines hot leg temperature. Reactor coolant pressure or ilPC flow

• then f Lxes the degree o f subcooling. Since the operator controls OTSG and RCS pressure and HPI flow, he is in control of the subcooling margin, hence pelmary to secondary detta P. FLgure 2, illustrates the effect of subcooling margin on primary to secondary l eaka ge .

Figure 3 ~illus trates the relattve ef fects of a cooldown with RC/'s ~

of f using 50?' and 2 52* subcooling. Even at the maxlmum cooldown of 100!*/ hr, the integrated leakage dt ffers by a factor of two.

+ Rev. L

• Rev. 2

/ Rev. 3

TDR 406 Rev. 3 Page 21 of 31 2.1.5 Waive Fuel in Compression Limits Fuel pin-in-compression limits are specified to assure that fuel pins are always in compression above 425F* in order to prevent detrimental orientation (i.e., radial orientation of hydrides) (Ref. 14). These limits require a high subcooling margin for RCS pressures ranging from 1350 psi to 550 psi. In correspondence dated January 20, 1983, (Ref.14) B&W confirmed that violation of these limits during tube rupture events is acceptable. When these limits are violated it is important that the pressure and temperature versus time be recorded so the effects on cladding can be evaluated. The evaluation must determine whether clad ballooning or incipient cracking has been induced.

2.1.6 Reactor Coolant Pump NPSH Limits i + RCP NPSH requirements place limitations on the minimum subcooling

+ margin. At low RCS pressures RC pump NPSH limits approach 100F* of

! + subcooling. However, general centrifugal pump test data have shown I + that NPSH requirements are substantially reduced at water tempera-

+ tures above 250F*. A review of THI.1 test data on the subject

~

+ reactor coolant pumps indicates a single loop flow of 98,500 gpm with

+ two loops in operation with one pump per loop. The pumps' manu-

+ facturer (Westinghouse) has provided required NPSH at various pump

+ suction temperatures (Reference 25) for the flow associated with two

• pump operation. The NPSH available, as indicated by the saturation

• margin monitor in the hot leg, is then calculated by considering the

+ total' pressure drop from the hot leg to the pump's suction (Reference

+ 26). The resulting NPSH requirements for 2 pump operation (actually, d either one pump or one per loop) are shown in Table 1 and Figure 6.

+ Note that the curves should be used as if the two loops of the RCS are d independent, i.e. either one or two RCPs are operating. Also shown

, d is the 4 pump operation (actually two pumps in one' loop or four pumps)-

+ NPSH curve which has considered the changed flow distribution in the

+ coolant loops. In addition, the normal NPSH curve and the 25F* sub-

+ cooling curve are shown for comparison purposes. RCP operation below 4 300 psig is only allowed if pressure differential across the No.'l d pump seal is maintained as 275 psi or greater (Ref. 32).

I + The emergency NPSH limits are intended for operation of RCP's during l + abnormal and emergency conditions such as small break LOCA, SG tube -

• rupture, station blackout and secondary side upset events. Pump

• limits and precautions must be adhered to while following the

• emergency NPSH limits (e.g., the pump should be tripped on high vibration).

l 2.1.7 Procedure Entry Point Condition r

! - The use of an emergency tube rupture procedure-should be limited to situations where normal limits (e.g. fuel ~ pin-in-compression and RCP -

NPSH) are being waived. The guidelines' entry point condition is chosen as 50 gps. A leak rate' of this msgnitude would be expected from the complete separation of one tube (as opposed to 385 gym for a double-ended offset of one tube). Less likely, (but more serious). _

+ Rev. l '

'

• Rev. 2 w

TDa 40o Rav. 3 da ge 22 of 81 would be leakage of this extent from a number of tubes. Bo th situations warrant entering the emergency procedure. Below this llatt, plant cooldown should be achievad wt thin normal llmLts unless addttlonal equipment failures occur.

2. l . 8

• Simula tor Experience

• Steam generator tube rupture proesdures were exercLsed during the

• January and June 1983 simula tor sessions, the experience gained from

• these two sessions has been factored Lnto this TOR. The principal

• lessons learned were that:

• 1. Controlling plant cooldown rate with 2 or 3 HPI pumps running

• is very dLf ficult a t best. Raising OISG level to 95% during

• this plant condleton may.not be possible.

• 2. Priorltization of plant control parame ters was not obvious

• to the operator in certain situatLons. The two situations

• whleh were encountered were:

• a. Minimizing subcooling margin has priority over mini-

• mizing the cooldown tLae , and ;

• b. Steaming to -control OT3G 1evel is less important

• than 2CS cooldown rata .

• 3. Plant response after RCPs are restaeted was unexpected to the

• o pera tion. Se c ond pum p r es tar ts ma y b e r equ ire d be f or e s ub-

• cooling margin stays above 25F3

• 4 Critaria for isolating core flood Ls required. Core flood

• tanks should be isolated in a subcooled system before they

• Ln itla te .

s

• 5. AdditLonal guidance is required Lf the RCPs are not tripped within two minutes of a loss of subcooling margin.

• These items are discussed in more detail in Sect'on 4.0.

2. l . 9

• Emergency LLmits for Decay Heat System Init lation

• Plant experiance indtcates that . Large portLon of tLee during

• cooldown occurs below the temperature of 350F. Staple analyses ,

• assuming only one ADV for. a loss of offsite power, us Eng

• the CSMP com pu ter co de (Re f 3 L ) indl ca te tha t the RCS can be cooled

• down below 300F in less than 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, and cooldown to _2 73i can be '

i

• accomplished in about ten hours. 275F is the normal DHRS initiation

• tempe ratur e. GPJNC has evaluated the capability of the DHA system to

• . operate a t a tamperature o f 300F (Re f 32) and concluded. that L t. is

• within the design capabilities of the system. Therefore we recommend

• tha t the tube rupture procedure allow initiation of the DHR system

• a t 300F LLstead o f 2 757.

+ Aav. L

• Rev. 2 d Rev.'3

TDR 406 Rev. 3 Page 23 of S1 4 As shown on Figure 6, the DHR system cannot be operated simultaneously d with reactor coolant pumps above a certain temperature. Therefore, a 4 decision will be required to evaluate the desirability of either: (1) d steaming the 075Gs but maintaining forced RCS; or (2) stopping d steaming but tripping the RCPs with the resultant potential for d' steam formation in the hot legs. This decision will be dependent 4 on the amount of radioaactivity release at that time. It is d technically acceptable to trip RCPs if the result is termination d of significant releases.

2.1.10

• Reactor Trip on Low Pressurizer Level

• Preventing a loss of subcooling margin has many advantages in

• contolling the plant. Before the spring of 1983, EP 1202-5 required

• the plant to be tripped if level could,not be maintained above 100

• inches with two HPI pumps running. This may not be a sufficient

• 1evel to prevent voiding of the pressurizer after a reactor trip.

• Emptying of the pressurizer causes a loss of subcoolin,g margin and

• the subsequent tripping of the RC pumps. In order to prevent this

• situation, the reactor should be tripped if level cannot be

• maintained above 150 inches or higher. This is sufficient volume

• (about 600 cubic feet) to prevent pressurizer voiding.

• There is a disadvantage to this recommendation since the safety

• valves will lift after the reactor trips. However, this situation is

• considered acceptable when weighed against the plant control

'

• advantages of having RCP's running. Also, only a certain window

• of break sizes will result in reaching 150 inches and not'100

• inches with full HPI flow. Outside of this window, both levels would

• be reached.

, 2. 2 Development of Multiple Tube Rupture Procedure

- Guidelines (Path 2)

The treatment of multiple tube ruptures relied on several sources of l information. The Ginna tube leak exceeded the single tube flow for a B&W plant and also resulted in a loss of subcooling. Therefore, that event legitimately represented a multiple tube rupture. The Oconee tube leak with a delay in getting onto decay heat removal,' prompted analysis of water inventories required to assure.a source of water for HPI-cooling.

Besides plant operating experience, this TDR investigated the following aspects of multiple tube ruptures:

1. Revision o' the RCP trip and restart criteria.

2. OTSG steaming and level control.

3. Establishment of criteria for going on feed and bleed cooling.

4 Cooldown/depressurization.

l

! + Rev. 1

• Rev. 2 d Rev. 3

_~ .

TDA 605 Rev. 3 da ge 24 of 81 2.2.1 Revision of RCP frip and Restart Celteria Based on initLal small break LOCA analyses reca Lved from PWR vendors in 1979, NRC concluded in NURCG 0623 that delayed trLp of reactor coolant pumps during a small break LOCA can lead to predletad fuel cladding temperatures in excess of current llcensing limits. At the ttee of RC pump trip the Liquid that was previously dispersed around the primary system through pumping action now collapsed down to low points of the primary system such as the bottom of the vesset and steam generators. Th is se para t Lon r es ul ts in s L gni fi can t uncov ery o f t'ie resetor core if system voiding Ls high enough, due to an insuf ficient amount of Liquid being availabla to provLde acceptable core cooling. Unacceptable consequences would result from delayad reactor coolant pump trLp only for a range of small breaks LOCA ( 023 to 0.25 f t2) and a range of trip delay times af:ar accLdent ini tia tion. Based on these findings , a mee ting of utility vendors and owners was held with NRC in September 1979. At this mae ttag L t was agreed that the 1600 pstg E3 AS signal provided CLmely Control Room LndLcacLon for manual actLon to prevent poss ible voiding scenarios.

GPU had B&W reevaluata these LOCA scenarios assumlag RCP's were trtpped on loss of subcoollng margin. Iha conclusion of that reanalysis was that loss of subcooling was an acceptable alternatLve to pump trip on 1600 ps tg E3 AS. In March 1983, the MtC Staf f required Ut L Litles to reevalua te their pump trip schemes (de f.17) .

GPUNC provided an evaluation of the pump tri.p critecton and a schedule for implementLng this criterion by May l,1983.

• The advantags of maintaining RCP's is that duhing Steam Generator Tube Ruptures in which alnimum subcooling margLn is maintained,

- continuous RC pump operatLon assures expedittous cooldown with a minimum primary to secondary differentLal pressure. This change in criteria for RCP trlp will allow RCP's to be operated for a grea ter spectrum of tube ruptures (LncludLng ruptures beyond the dasign

• basis) and to reduce tne of fstte doses for those events. Re ducing ~

• the allowable subecoling margin is not intended to reduce RCP

• equlpment pro te c tion. RCP's should be tripped Lf emergency APSd

• requiremants are not met, and should not be started untLL NRsH

• requirements are re establishad. If apptleable NP3H pump vibration

• limits are exceeded, then the RCP's should be trippad. Pumps should

• be restarted as indicated in the THI-l Small Break LOCA

• Procedure (EP 12 02- 68, At t a chmen t L). As noted Ln Section 2. 3,

• bumping celterion requires additional clarlfication.

FLgure 3 illustrates the reduced leakage possible with RCP's on.

Stailarily, restart of RCP's has a great advantage. During tuba ruptures , primary to secondary dLfferential pressure decreases raptdly since OTSG pressure is high. Leakage flow is exceedad by dPI j flow and subcooling margin should normsLly be restored within 20-60 minutes af ter larger tube ruptures. Res tarting RCP's provides pressurizer spray, and prevents void format *on -in the hot legs U bends and reactor vessel head.

F Rev. I

• Rev. 2 3

- - . . - .-- .~. -. . . . =- . - - _. -- .. - - . . -

, IDR 605 Rev. 3 Fage 25 of 81 k'

o 2.2.2 0TSG Steamina and Level Control

• The guidelines for OTSG steaming are nearly the same when either one

• - or both OTUGs are af fected. the OTSG pressure should be controllad l t o pr ev en t l if t in g o f s a fe ty v alv es ( L .s . s ta y b el ow 1000 ps ig). l

i. Level should be maintained below 95% on the operate' range. there are l several other issues to be considered for multLpie tube ruptures, l howev er. First, large tube ruptures may ' result in RCP trip. tha j- OTSG's should be steamed to maintain natural eleculation in the i af fected loop. Natural circulation flow will minlaize the

+ chancas of drawing a bubble In the hot leg U bend. Con tinuous

& steaming of ths OISG allows all of these constderations to be

• a ccommo ds ta d.

It Ls important .to recognize that a large tube ruptuce with loss of subcooling is a LOCA condLtion. fhera fore , L t is required to raisa OTSG 1evel to 95% to assure that Liquid Level in the tube region is

high enough to allow water to flow .into the core during boilec

• condenser cooling. If. Level is not raised to 95%,.then EiW flow must

• be at a high enough flow rate to penetrate the tube bundle

{

'

• sufficiently to provide auequate hest transfer. A flow rats

• of 450 gym total '225 gpa/0T3G) has been vertfled as acceptabla t

• by B&W (Re f 2 9) . This flow is the minimum available after a seismic

• event and worst case single failure, coincident with a small break i

• LOCA in which boller condenser cooling is required. It La important 3

• to recogr.ize that with two dPI's available , boiler ' condenser cooling

• ts not required. Procedures should therefore allow the operator to

• raise OT3G level to 95% tempered with the need to control the ACS

• cooldown rata. During tube rupture events with both' HPI-pumps available. . the unaf fected OTSG level should be raised first while

!

• the sffected OTSG 1evel should be preventad from botting dry

} ,

• (ma inta in a minlaum level. o f 3 0"). The operator can control'

'

• 1 OT3G !nstead of trying to raise level in both OT3G's i,

• s imul taneously. For the case with only one HPI pump , Lf incore j

• temperatures are not decreasing and the OTSG ls not ' removing RC3 '

• h ea t , then 'there will not be.a cooldown rate control constderation;

• moreover,- the plant may be in a'condLtton that requires boller i

• condenser cooling, there fore, 0T3G 1evels mus t be raised to 'the -93%

• 1evel- s tealtaneously in - this situatton '(Re f. 30) . .

l

• Section 2.L. 3.1 discusses steam generator isolation, steaming and.

l. + fitting criteria when both OTSG's are leaking.. This discussion also j +' appites .when the RCS subcooling margin ~has been lost. ,

, - 2. 2. 3 Criteria for Feed and Bleed Cooling t

. Analyses of multiple tube ruptures indleate that exLating plant procedures for establishing feed and bleed cooling are correct.

Faed l and bleed cooling should be' initLated when the OTSG heat sink is not

. ava ilable. If both steam generators 'are isolated during a tube rupture, the PGAV should he ' opened with futt.HPI turned on. An

+ Rev. 1-

• Rev. 21 l- d Rev.'3

. . _ _ . - - ._ a ~ _. .-. ...___2 . . . - , . . . _ .. _ ._..... . _ _

TDR.60o Rev. 3 ,

Page 26 of 81 l addt tional complication for tube ruptures , however, is the potential

• to flood the OTSG's and force open the safety valves under this c ondi t ion. If RCS pressure is below 1000 pstg , the PORV is capable

• of removing decay heat, even with itquid relief within two hours of

• resctor trLp assuming that there is no energy rellef out of the

• ruptured tube (see FL gure 4). Therefoce, tne operator can control

• RCS pressure by throttling HPI. Moreover, wLth RCS pressure below 1000 psig the OT3G safety valves will not lift.

If R CS pressure s tays above 1000 pstg , however , the operator must

• take actLon to prevent safaty valve lifts. A situation with RCS pressure above 1000 psts and netther OT3G available requires ths

• opening of the TSV or ADV's to control OT3G pressure below 1000 psig

• and level below the upper tube sheet. Elther tha A37 or T3V navs

• sufflelent steam espacLty at high OT5G pressure to remove decay

• heat. The TBV's also have suf fletent ca pacLey to prevent OT3G

• floodLng. However, Lf the leak rats is large enough, the staaming

• rate required to control level in the OT3G may result in an

• unacceptable RC3 cooldown rats. In this case, cooldown rate mus t be

• controlled and the OT3G allowed to flood. As discussed in Section

• 4. 2. 2.1, this s ituatLon does not seem likely (s e least a t high

• de ca y hea t l evels ). As decay heat decreases , s teaming can be

• teretnated when RC3 pressure goes below 1000 psig and Ls controlled

• by the PORV and 'dPI,

• The s teaming capacity of an ADV at 1000 psig exceads de cay hea t

• 1evels wLthin several minutes af ter reactor trip, dPI capacity

• exceeds the capacity of one AJV. Therefore, the RCS pressure can be

• controlled at 1000 ps L g in th is mode wL ehou t L L f e in g sa fet y. valv es .

• Subcooling margLn can be regained and the plant cooled dosn untLL an

• OISG heat sink can be restorad or until the plant can be put on decay

• hea t r emov al .

j y

• UntL1 OT3G 1evel is above the uppar tube sheet, pressure in the OI3G

• will remain below 1000 psts , since the RCS temperature is below

• 54 0F ' . WLeh level Less than 600 inches, however, tue operator stLLL

• must steam to keep pressure below 1000 pstg; therefore he should not

• have to s team to control. level on _ the a ffected OT5 G. When level goes

• above 600 inches, pressure in the OT3G is de termined by the steam

l.

• pressure in the steam line. If the Lines are leak tlght, then

• compression of the steam bubble can cause a pressure increase above

• 1000 ' ps ig. In this case,'the operator would steam the OTSG to reduce

• pr ess ure. However , L f there are s team leaks -in the sys tem (e .g. ,

• through sesam traps.) then the itnes could f Ltl with water before

• OT5G pressure increased. there fore to prevent this situation the

• OT3G's must be steamed to preclude this ' possibility.

' 2. 2. 4 Cooldown/De pr ess uriza tion Analyses of multlpte tube ruptures demonstrated that subcooling margin should be regained in 20-60 minutes '(see FLgure 5). ROP's can be started and a forced flow cooldown instituted. Even if RCP's are

+ Rev. L

• Rev. 2 i Rev. 3

IDA 605 Rev. 3 Pa ge 27 of 81 no t ava ilable , the cooldown during a multtple tube rupture can be accomplished within the single tube rupture guidelines. If equipmant

. failures prevent a normal natural circulation cooldown, then tha plant would be cooled doen with faed and blaed cooling. This maneuver would probably require inLtiation of feed and bleed cooling in the HPI/LPI '*ptggyback" mode . Existing plant procedures give correct guldance about when to initiate this mode (BJ3T level below 3 ft. ) .

+ Guidance from B&W on PfS/3rittle Fracture limits requires a " soak

+ tLee" to allow the vessel wall to reach the RCS temperature.

+ However, B&W has also recommanded that the "s oak tLms" La not

& required during tube rupture events in whLeh a rapid cooldown is

• necessary (Raference 24) .

Steam releases during multiple tube rupture events can be minimLzed by jud'.cious use of the EFW, HP1 and TBV's. Full HPI flow , in F conjunctLon with throttled EiW flow allows a 130/ */ he cooldown withou t having to s team either OTS G.

2. 3 Adit tLonal Work Requirements

2. 3.1

• Analys es

• As noted in SectLon 1.0, there is a program of ongoing computer

• analysis work simulating single and multtple steam generator tube

• ruptu r es . The effort includes the plant statas listad Ln Secetons

• 3.1.1 and 3.L.2. This list does not reflect cio s pec fle de ta lle d

• analysis efforts which are being undertaken as part of the tube

• rupture quantL tative development ef fort. the first analysis is a

• simulation of the vessel head reglon during natural cLeculation

"

• cooldown. This analysis ef fort will help in evaluating the ef fect of

'

• a vessel head bubble on the RC pressure response. It will also aid

• in evaluating the benef t c of the reactor vessel head vent.

• The second snalysis effort Ls being performed in conjunceton with

• Babcock and Wilcox Company. De tailed analyses are being performed to

• provide Laproved guidatines for OTSG flLitng after a loss of

• subcooling margin, wL th one , teo and three HPI pumps available. fna

• Intent of the analyses is to assure that the OT3G's are f Liled 4 without violating cooldown or shell to tuba differential camperatura

• limtes, while stL11 meettng core cootability requirements. This

• effort was constdered after the January 1983 simulator training

• session and further defined after the June 1983 simulator session.

2. 3. 2

• Issue Resolution

• A number of issues were identtfled which require further effort

• to resolve. these are tha following items: _ . _

• A. Operator guidance for identLfytng two phase natural cLre-

. A ulation cooling (boiler condenser). R&W provided this 4 guidanca to the B&W Owners Group (Re ference 37) and the 4 information is presently being evaluated by GP' JN.

V Aev. I w 1

4 TDR 406 4 Rsv. 3 Page 2Sof 81 I

• 5. Acceptability of excessive cooldown rates for very short time l 4 intervals. B&W has provided their response to this concern ]

d (Reference 39). Deviations of no more than 15F are allowed at i j -d. any given time.

l

• C. . Importance and technical basis of fuel-in-compression limits.

d This effort is still underway.

• D. Viability of DHR system initiation at temperatures above i d 3000F. This effort is still on going.

l

• E. Identification of pump vibration limits for various pump 4 combinations. This effort still requires resolution.  ;

l u

'

• F. Determine viability of. ATOG RCP " Pump Bump" crf terion.

4

• The bumping criterion would allow running pumps without j

• adequate subcooling or NPSH sargin as long as a steam generator

]

• heat sink is available. Determine whether the ATOG criterion

• anticipates that NPSH will be reestablished since the heat j

• sink is available. The existing bumping criterion in TM1-1

-

• energency does not require that a heat sink be established j

• and would allow continuous RCP operation in violation of NPSH

• limits.

l I d ATOG requires that RCP protective limits be observed for all

! d conditions except under certain inadequate core cooling i d situations. Therefore, the operator would only operate the

] 4 RCPs if normal limits allowed.

! d C. As noted in Section 2.1.9, simultaneous DHR and RCP operations

} d' are not allowed'above 275F. The basis for the'DHR pressure

. d limit and for the RCP seal staging pressure must be re- -

d evaluated for emergency conditions.-

{

I d H. Reactor coolant pump shaft vibration data were taken during 4 d the plant cooldown of 10/6/83 (Reference 38). The RCS was i d depressurized at a constant temperature of 400*F from a

d subcooling margin of 180F' to 60F*. Shaft vibration increased

4 to 26 mils. A plant test will be developed and executed to d determine if the pump response at this temperature is typical

! d or anomalous. If typical, then shaft vibration will have to i d be monitored by the operator when normal NPSH ' limits are violated.

d i

4' i i

- + Rev. 1 i,

• Rev. 2 4 Rev. 3'

-. c,-. ,- - , ,-,,,--e.-, - -- , ,,,,e- , ,,-m-.-n-e e,, v , ,rm. --- e

l i

TDR 40o Rev. 3 j Pa ge 29 of 81 3.0 D IS Cll33 D N O F MAJOA RE VIS D h3 ID E X13 r DG P10 CE3 tRE3 The development of the design basis guidat tnes discussed in Section 2.1 LdentLf ted a number of areas whleh were investLgated to de termine where spectfle changes should be incorporated Enta the nes guldelines. This section further explains whst areas of the guidat tnes should be revised.

3.1 Bas t c Plant Sta te 3.l.1

• As s ume d P lan t Condi tLons The following sssuspelons apply to the davelopment o f guide t ines as they apply s ingle tube leak / ruptures.

1. Subcooling margin (S CM) ts mainta ined.

2. Only one O TS G is a f fe cted . .

3. Condenser is available.

4 Reactor Coolant Pumps (RCP's) rensin on.

5. Decay heat ts removed by the intact OTSG unttL the Dacay Heat Rsmoval 'DH) systes can take over.

6. The a ffected OT3G can be steamed to maintain less than 35% Level (Opersting Range) and less than 1000 psig.

In addleton the revised guidelines wiLL have provisions to deal with ths following circumstances:

< l. RCP's not available.

2. Condenser not available.

3. HLgh radLation releases of fsite ,

4 Tube leaks in both OT3G's (but one OI3G remains espable of removing decay heat).

5. Steam lines associated wt th leaking OI3G flood.

• The consideratton of Ltems 1 and 2 are equivalent to an assumpcLon of

• t oss o f o f fs t te power .

3.1.2 Tube Rupture Guidelines For Loss of Subcooling Tube lasks in this estagory generally go beyond the Licensing bas ts, or are otherwise tsaarkable dua to plant condt tions (aside from ths tube leak) or equipment malfunction.

+ Rev. L

• Rev. 2 d Rev. 3

. 1 TDA 405 Rai. 3 Pa ge 30 a! 81 Tha. following conditions were assuesd in developing guidelines for this category of tube rupture event,

l. More than one tube leak.

2. S CM L s los t .

3. RCP's are unava ilable.

4 Pilot operated relief valve (P0dV) and Reactor Coolant S/ stas (RCS) hlgh point vents are available.

5. Unaffected OT3G can be steanad.

Contingencies The revised guideline will have provisions to daal with the fotLowing

, Additional circumstances:

1. Bo th OTS C' s ar e a f fe cte d.

2. Bo th OTS G 's a f fe c ted , bu t one O T3G co ma ins ca pab le o f RCS heat removal and eLther a) the FORV is unavailable or b) RC3 pressure s tays - above the ma in s team sa fe t y valve se t po int due to vold formatton in the RCS.

3. Na tther OISG ls capable of removing decay heat, and alcher a) the PORV is ava ilable , or b) the PORV is unava ttable.

3.t.3 levised Equipment Llatts & Operating Practleas During the course of the analyses leading to the guidal(nes providad I in Section 4.0. .It bacase apparent tha t certsin normal equipment lialts and operating practlces should be adjustad to effactively daal wL eh a tube leak / rupture. Ihess' changes will' help accomplish the following:

1. ML tLgate or prevent further OTSG damage.

2. ' Maximize tha cooldown eate to cold shutdown.

3. MLnialze SCM (thus sintalsIng priency to secondary Leanaga).

4 Maxtaize RCS pressure control options.

An Event Tree showing the various possible developments of an OT3G tube leak appears as AppendLx D to this report. -

+ Rev. L

• Rev. 2 4 Rev. 3

TDa 40o

< R:v. 3 Page 31 of 81 l

3. 2 Discussion of Guldelines 1 AppendLx C providas a logic dLagram of the tube rupture guidattnes (wlth a vettten discussion of thosa guLdelines). Ihis section of the d report provides descripelve text of the guidelines shown in that diesess and described in Section 5.0, the symptoms of the tube rupture procedure define the entry potnt conditlons wnen the emergency procedure is used. Ihis procedure nead only be entered for situations where a rapLd depressurizacLon of the plant Ls warranted.

When such condletons warrant , then the plant should be shut down and cooled down as expedLttously as possible and certain normal plant llatts (RCP NPSH, normal tube / shell deles T and fuel in compression liatts) are waived.

l 3.2.1 Iass d la te Actions s

the tube leak in questlon may not be large enough to cause a reactor telp. In such a case , the operstor begins a load reduction as raptdly as possible without causing a reactor trip (LO%/aln.).

Avolding a reactor tr'p prevents lL f ting of tha OT3G safety valves.

I 3.2.2 Followup Actions - Subcooling Mainta tned and RCP's Available

}

Once the load reductLon is initLated, the operator haa several as jor goals to schieve while bringing the plant to a cold shutdown condLtlon. Flest, he must prevent liftLng of the OTSG safety valves, 4 second , reducing petsary to secondary leakage by minimizing primarf 1 4 to secondary differentLal pressure; third, mLutatze stresses on

4 tha OT3G tubes by llatting shell/ tube delta f; and finally, sintsizing 4

d' releases by allowing the leaking OT2G to flood Lf offsite doses are l

Large enough (approaching levels at whlch a $lte Emergency would be j de clared) .

l 4 The following sections dLscuss the most Laportant control 4 considerstions recommendad by the guidelines.

4 3.2.2.1 Ma intain a MLnimum of 2 57

• Subcooling MLnialzing subcooling margLn means that primary to secondary

dLf ferential pressure is also sintsized, whleh reduces leakage and of fsite doses, saking the event reoes manageable.

3.2.2.2 Stessing/Isataeton CrL eerla for the Af feeted 0T3G 4 the operator used to be attowed (before 1983s to let the OTSG (LLL

, 4 anyttes that RCS pressure was below 1000 peig. the guldeline has 4 the operator steam the OTSG: fl e st , tio pr ev en t l if t L ng o f t he OT3 G 4 safety valves; second , to prevent the generator from filling; third, 4 to maintain OTSG 1evel on scale, fourth, to promote natural 4 circulation; and finally, to control shell/ tubs dLfferential 4 tempe r atur e. SLnce meettng these criteria w11L mean nearly constant 4 stessing, the guldelines require the operator to stess the OraG L 4 unless spectfle isolacton criteria are set.

+ Rev. 1 '

Sev. 2

EDA s05 Rev. 3

/a ge 32 of 81 3.2.2.3 Shell-to-Tube Delta f Plant limits and precaut'ons require an tntsining the OT3G tube 4 temperature wL thin 100/* of the shell temperature. De tailed analyses 4 lRef. 36) show that a shelt to tube detta T of 70/

• acceptably ilmits stresses and minlaizes the chances of increasing the leak from a pre exLsting r.hrough wall crack.

3.2.3 Followup Actions lktomstle Reactor felp has Occurrad)

All of the followup actLons dlscussed above stLL1 apply when the tube leak is large enough to cause an automette reactor trLp. In addt tton, the following procsdure changes would apply.

3.2.3.1 RCP TeLp W!th a I.oss of Subcooling Margin 6 Rupture of one or a few.0T3G tubes will likely result in RC3 4 depressurizatlon to the HPI setpoint , but may not result in a loss of 4 S ci. Tripping of RCPs on loss of subcooling sssures that the core 4 remains covered for all 14CA conditions. At the same tLaa , the 4 chance of pump trips during tube ruptures is reducsd.

3.2.4 Followup Actions for Loss of Subcooling h the second sect!on of the tube rupture procedure is encared enen RC3 subcooling is lost. Here , the operator must treat IDCA, as well as tube rupture symptoms. He is then able to pursue the followup cuoe rupture sctions. All of tha guldanca for followup actions without loss of subcooling apply.

The objective in this portton of the procaduce is to asintsin natural

• circulation (Lf possibla), reestabitsn subcooling margin, restart one reactor coolant pump per loop (Lf possible), and return to the

• section of the procedure for forced flow cooldown. If one pump can d no t be s ta r ta d pe r loop , th en b oth RCP' s in one L oop a r e s ta r te d.

d this will anxtalze the pressurizer spray flow for the given ROP 6 avattability, d Even if the OT3G raust be isolated, sesamlng ts sttil required to keep 4 011G pressure betov 1000 psL4 Pressure control is required to 4 prevent the MS safety valves from lifting.

If subcooling is regained in the RCS, then HPI is throttled, RCP's

• are started and the operator contLnues the cooldown. The desired RCP

• conf tguration is to start one pump in each toop. If the operator is

• unable to start an RCP in each loop then he should start both RCP's

• in one loop.

d the ressons for restartLng RCP's are similar to the reasons for not 4 tetyping them on low RCS pressure. the AC pump flow any cause volds 4 in the systas to cottapse, dropping pressutizee level. The f guldelines require the operator to watt 2 minutes for SCM to recover 4 to prevent Lncessant " pump bumpLng". ,

+ Rev. L

• Rev. 2 d Rev. 3

• l TDR 606 Rev. 3 Page 33 of 81 If subcooling cannot be restored, the operstor cools the plant down d on natural eleculatLon unless the OTSG heat s Ink is los t (e .g. due to loss of natural circulation in the unaf fected loop). ,WL th no s team generator heat sink, the operator must put the plant in a feed and bleed cooling mode. Feed and bleed cooling is initiated by isolsting

+

che OISG's, assuring full HPI is on and opening the 20RV. Witn RCi pressure below 1000 ps ig , wa ter ret te f ou t o f the

• P3RV is suff Lclant to keep the core cooled (See FLgure 4; a fter about

• 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. If the OISG heat sink is restored , the feed and blaad moda is term *nated sud a natural circulatton cooldown is reinitLatad.

If RCS pressure stays above 1000 pstg durind feed and bleed cooling (e.g., the head bubble prevents depressurizatlon or the FORV fa LLs closed) then the sacondary sLde safety valves have to be protected d f rom challenge . The operator controls OT3G pressure beide 1000 psig

• wlth whatever means are available (turbine bypass , or Atmospherte

• Dump Valves .). When the OTSG tube reglon is fliled wLen watar, tne

• operstor opens the A3V and leaves Lt open. This action minlaizes the

• chances that safety valves will be forced to relieve water and/or

• s team and fa il open.

o .

r i

k

+ Aav. I h Rev. 2 d Rev. 3

TDR 40s Rsv. 3

/a ge 34 of 81 4.0 S MJIATOR T1A14 L4G ZXPl!RIE4 CE

• 4.1 Inteoduct ton

• Most of the guidelines proposed in this TJA were incorporated Lata a lesson plan for the annual requallf tcation trainirig o f the til-1 Licensed operators at the BW simulator. A draft revis ion to til-l's EP 1202-5, incorporating the gutdelines , was also preparsd.

These documents were then used to inform the Licensed oparators of the changes contesplatad for EP 1202-5, and to demonstrate the combinad affects these changes would have. During the classroom session, eacn guldeline was described and the reasoning behind L t was explained.

During the s taatator session, their combined aff set was iL Lustrated oy running a large tube leak scenario twlce.

For the flest simulator run, the then exLsting revision of EP 1202-5 was used to deal wL eh the Leak. For the second run, tha draf t version was employed. It became apparent that the new guidelines made plant ,

control eas ter. ,

• As indtcatad in Sect ton 4. 2. 2, -the January L983 training cfcle was not

• ef fective in training operators on the baste concepts for treating

• tube ruptures. The June training cycle was sucesssful in

• communleating concepts.

4. 2 Results

• 4.2.t January 1983 frainin2

• Of all the suidetines proposed in this T0R ae that tLee, the two changes most useful (and obvious) to ths operators were:

n

1. Reactor Coolant Pump (RCP) trip as a followup to low subcooling margin (30() rathee than following autoastle HPI fron 'a low RCS pressure ES AS signati

2. HPI throttling when 504 requirements are met and pressuriser levet is back on scale, rather than walting for pressuriser Level ,

to r ea ch 100'* .

Another useful (but less obvious) change is the RCP restart criter ton based on regaining SCM rather than various combinations of prLeary and -

• secondary pressures.. This and Iten 1 above may be considered under the same general heading of increased 107 availabilley.

The exerclse of the draft RF 1202-5 was useful in critLquing ~the

+ conusplated changes. Slaulator experience also showed that L t is not

+ possible to raise 015G 1evet to 95% with full 3PI on, vnite sesamlns l

the 315G and anintaining a 100t */ he cooldown. the dLff Lculty was i

l

' + A41. .L

• sev. 2 I

d Rev. 3

IDA 405 Rsv. 3

/aga 35 of 81 cres tad by the s teaming of the OT3G's in this situation. 3P1 and throttled E/W flow can provida a plant coaldawn at neac 100i*/ hr if the O T3 G's ar e no t s t eamel .

During the simulator session, B&W revised the simulator to allow leakage of mors than 2 tubes and to allow leakage in both OT3G's.

4.2.t.L Commen ts

• This material was presentsd to two of seven groups by Tech.

Functions personnel, the remaining five groups received it fcom BSW training personnel who taught the material using cne sama lesson outline. B5W did not endorse the asterial. Comments from trainees indLcate that the training was of subious value. Ic will be necessary to repeat the training for all personnel.

4.2.2 June 1983 f ralning *

• A number of L tems were identlfiad during the BSW op3ratoc training

• slaulator sessions held from June 6 to June 29, 1983. The expec tence

• gained from using a revised tube rupture procsdure EP 1202-5. These

• L tems will be discussed below.

4.2.2.1 Control of RC3 Cooldown Rata

• Section 5.2.L dLscussed the dLfflcuttles in controlling

• cooldown rate while raising OI53 level to 955. At the tLaa,

• L t was the authoe's belief that s esamLng of the OI3G' 4 was

• causing tne excessive cooldown rate. dovever , fur thec

• dLscussion with B&W (Ref 29) Lndicatad a diffarent

• ex plans cion, the B&W dPI model calculates flow by

• Leeratively solving two equations of the form:

/

')

• E K (1)

Pd = 2 84 0 -

N and

• W= lPd-PRC3 ) A II l 2)

• where:

• Pd = pump dLscharga pressure

• Pgc3 = R C3 p r ess ur e .

• W = flow, Lba/ sec

• N = number of HPI pumps running

• A

• coeff tclent to account for numbec of HPI valves open.

• Che HPI flow is overpredleted for IMI-l wLth three HPI pumps

• running and/or low RCS pressure. The dL f ference casults f rom the

• physical arrangement at TMI-l, in whleh two pumps discharga into

• a common h ea de r . The esvitating venturies at Unit t also ceduce

• ths maximum flow of the RFI props cosparad to the valva

• cateulated by the s taulator.

• As a result of this understanding, sub:equent sleutstor drills

• were run with only two WPI pumps available and control of t

I

• cooldown rate was leproved. + Rev. I t

.

• Rev. 2

~

TOR 603 Rev. 3 dage 36 of 81

• The HPI inttiation rule has been reemphasized to the operator,

• namely that "futt" HPI means the full flos from two dPL pumps.

• It is acceptable to secura the third HPI pump when the RCS is

• saturated, and the OT3G heat s tak is avaliable or if cooldasa

• rate is 100? '/ hr. or more .

• A second consideratton in controlling cooldaan rats was in

• raistng the OT30 level incressa to 93% af ter a loss of subcooling

• margLn. Opecations believes that t t is an unnecessa.*y burdan on

• the operator to control cooldown race while raising level on both

• OT3 G's s inut tsneous ly. Therefore, the leaking generator level

• will not be raised until the unsf fectad OI33 has been raisad to

• 95% unless incore thermocouple temperatures are not dacreasing

• and the OTSGs are removing decay hest. fhe 95% level is

• tmportant ln establishing boiler condensor cooling duriag smati

• LOCA's in whlch only one HPI is availabla. However , ths ACS

• cooldown is only a concaen Lf both HPI pumps are runalag.

• Therefora tha two concerns are natually exclusive.

• Oparator training and EP 1202-5 have been revised to nave the

• operator control one OT3G tevet at a time as long as incore

• temperatures are decreasing. If RC3 camparatures are not

• sf fectsd by the secondary stde cooldown (L.e. , no secondary stda

• heat removal) then both OT3G's should be raised to the 93% level

• sleuttaneously.

• During the simulatoe session of June LL,1983 the operators were

• faced vlth a targe (about 1400 gpm) tube ruptura. thsy attemptad

• to control OT3G 1evet below 95% on the a f f actsd generator.

• Hasever, the cooldown este was too h!gh even wlth the unaf fectad

• OT3G tsolatad. The simulator response to the event was partL2LLy

• responsibts , but the procedure also naaded to ba more exptlett.

• The simulator leak modal currently uses the ortflee aquation to e

• pred!ct leak flow (Raf. 28). fhis model would inttially

• overpeedlet the bresk floe and would account for the very capLd

• flooding of the O r3G's compared to results of the RETAAN and

• RELAP5 computer codes. RETAA3 and RELAPS (de f. 35) analyses to

• date do not predlet such a responss. Nevertheless , the operacoe

• needs to recognize that cooldown cate control takes precedenca

• over OT3G tevel control anj !? 1202-5 was subsequently rettsad to

• prevent this conflict ln plant control requirements.

4.2.2.2 Plant Stab tL ization 5s fore Cooldown *

• The procedure used during the .traintng session nad ene

• operator inleiste plant cooldown and then establish minlaum

• subcooling margLn. The simulator sessions showed that the

• ACS could not bs dapressurized fast enough to maintain a

• minimum subcooling margtn. Therefore, training material was

• revised to emphasize the need to stabilize the plant and

• reach the etnimum allowable subcooling margLn. Plant

• cooldown should then be inttiated. the procedure was

• sodLf ted so that aLL four RCP's can be Left on untLL

• 500F ~ tas tead o f 540F. Based on the til-L AIJ3

• (Re f. 28), this change providas a dL fference of about 4 7%

• in the spray flow (see Tabte 2). thus pressurizer spray + Rev.1

• flow is maxlatzed for as long as possible. FLnally, the e gey, 2

_________-._______.__________.m.n.;

IDA 605 Rev. 3 Page 37 of 81

• operator is given the option of using tha pressurizer vent L!

• he is stL11 unable to reduce pressure suffleiently tu

• maintsin a minlaum subcooling margin.

4.2.2.3 RCP Res tart Criteria *

• Section 3. 2.1 obs erved tha t the A07 res tart cet teria on 25F

• subcooling sargLn (SCM) was very useful to the operator.

• Several areas required clartflestion or expansion, however.

• FLrst, RCP restart should not be attamptad unless pump

• emergency NPSH LLatts are met. the only axception is thst

• NPSH requirements are walved, with pump restart allowed

• during certain inadequsta core cooling condt tions as

• spectfled Ln plant procedures. Pump " bumps ," however, should

• be setemptsd even Lf NPSd requiraments are not sec. Sa cond ,

• Loss of subcooling asy occur after the RCP's are

• r es t ar ted. Collapsa o f scesa volds in tha ACS may cause

• voldtng of the pressurizer, resulting La a loss of

• S CM. RCS temperatures may be hotter in the tube ragion enaa

• in the enre if natural cLeculatLon has been lost. MLxing of

• this hotter water wLth the cooter core water will cause a

• decrease in the SCM. Several pump s tarts may be required

• before subcooling margin stabilizes above 25/*. Allowing two

• alnutss of RC3 flow is helpful in eliminattag both sesam

• volds and tesperature gesdients so that successive restarts

• wtLt be successful.

• 4.2.2.4 Core Flood Tank Isolation

• In several simulator runs, the operatLons were facad with

• tube rupture or small break L3CA condletons in whlch tha ACS

• was subcooled, but core flood tanks intcLated, providing

,

• cooling water whleh was not required , s Lnca SCM was

• ma in ts ined. Most signif tcant was that C!T inittsttoa

• delsyed RCS depressurization. NeLther the L3CA nor

• tube rupture procedure provides any guidanca about

• lsolation of the cors flood tanks. There fore , this TDR has

• been revised to provide guidance about when to isolata the

• core flood tanks (ses Section 5. 2.11).

• 4.2.2.5 RCP f rip Criterion

• QuestLons arose ressedLng the sectons to be taken Lf RCPs

• vere not trLpped within 2 minutes of a loss subcooling

• margLn. Clartf tcation was providad using the guidancs of the

• , TMI-t AT3C (int. 28) whlch requires the operator to keep the

• RCPs in each loop running Lf the two aluuta tLas Limit La

• excee ded. If RCPs are subsequently telpped, the ACS may be

• volded enough to uncover the core. RCPs should be run

• for at least 7000 seconds to assure that the core will not uncover (based on AppendLx K sssumptions). If pump Janas:

• may occur, then one pmep in each loop should be

• t r ippe d. If either of the running pumps fails, the tripped

• penp in that loop should be started. For s taptletty , the ,

• guidelines in this TOR call for i RCP to be run in each loop.

+ Rev.1

• This provides suf ficient flow to cool the core (Ref. 28).

}- _*h

TDa 605 Rev. 3 Pa ge 38 of 81 5.0 TUB 2 L2M/Rd PTURZ CU10ZLINE3 4 5.1 S:o pa

• The guidatines will deal wIth tube leaks la exesss of 30 gpa.

Primarf-to sscondary tube Leak rates less than 50 gpm will be nandlad in accordance with " Guidelines for /lant Oparations with S:aam Generator Tubs Leakage,* TDA 400 ( Ra f . 16) .

5.2 Guldet tnes & LLmt es

• This section provides plant 5,pect fle cachnical guida L lnes for tuoe rupture events whleh can be used to generste plant Emergency P roce du r es .

5. 2. t Subcooling Margin Rsquiremants

• Control Resctor Coolant System (AC3) subcooling margin (3Ci) bs tween 2 5/

• an d 50/ * . Maintain SCM as close to 2 5/

• as poss Lble consistent

+ wt th ths RCP NPSl! curve of FLgure 6 and while va tving fuel

+ pLn-tn compression ilmL ts .

This will alntatze primary to secondary dlffetentla! presaure, thus sintsizing the leak rats.

5 . 2. 2 Resctor Ooolant Pump Trip Critsrion

• felp Reactor Coolant Pumps (ROP's) when 3CM ts lost.

No ta :

• If RCP's are not trippad within 2 minutss of Loss of 23/

• SCM, enen

• run L 207 in each loop.

o

5. 2. 3 Assetor Coolant Pump Ras tart Critaria

• When the required subcooling margin '2 5/*) has been establishal.

• res tart L RCP per loop. If unable to statt an RCP in oae loop, statt

• both a0P's in the oppost te loop.

No te : If subcooling margLn is lost tamedLately after pump testart and does not return within 2 minutes, tha RCP's must be telpped sgain and not

• restarted untLt Sci ts regsined. Subcooling may be tort several class

• befort the pumps can be le f t running.

5. 2. 4 desetor Ooolant Pump NP3H for Essessney Oparattons *

• + the attsched curve 'FLgure 6) deplces the ACP NP3d Liste to be usej

+ during a cooldown with a tube leak.

5.2.5 Htgh Pressure Injection thrott11nt Ortteria

• throttle !!PI when SC4 requirements are met and pressurizet levei comis back on scale. '. No te tha t the other HPI throttling celteria remain tmchangsd.)

+ dev. L 4

, Rev. A

IDA 403 R:v. 3

/S8e 39 of 81

5. 2. 6 OT3G Level

• If SCH ts lost:

a) Raise level on the unaffected OTSG to 93% inite Leaving levet

• control on the unaf fected 3T3G a t 30 inches.

• b) Raise level on the affectad OT3G to 93%.

• NO TE : If incore thermocouple tamperatures are not decreasing and there is

• no heat transfer to the OT3G's, then both OI3G 1evels must be raised

• to 95% slmmuttaneously.

5.2.7 OT3G IsolatLon/ Steaming Critaria

• When the leaking OT3G is Ldentt fled , close all steam valves except ena A3 V' s an d TS V' s .

No te : Jo not close MS-VID untLL an alternate source of gland steam La ava ilable .

When RC3 Thot .ls less than 540/

• the a ffected OT3G sust be Isolated if:

(s) Borated Wstar Storags Tank level is below 21 f t., or (b) Of fst te dose projections approach the Level requiring a St ee

• Emsegency ( 50 mRea he whole body or 2 50 mRan/ he thyrold).

l No te : If both OT3G's are leaking and isolation is requitsd based on offstta dose projacetons, flest isolata the OT3G with the higher lestage. If such a distinceton cannot be made , isolate one OT3G and reevaluate e' offstte dose projections.

5. 2.7.L Pressure Control of an isolsted OT3G

• Stess the af facted OT3G's) only: ,

l. To keep OTSG pressure below 1000 pstg,

• 2. If the plant ts on feed and bleed cooling. OI3G 1evel mu.t be

• controlled below 600 inches on the wide range Indl ca t ion .

• If the OT3G sust be steam 3d opin the Turbine Bypass Valves or

• Atmospherte Dump Valve on the af fectad 0T3G.

5.2.8 Cooldown Rate During a Tube Leak Event 4 the cocidown este shall be limited to 2 asxtmum o f 1.6 7/ '/ min

'100/ */ hr) whether on forced or natural circulatLon.

M)te: Stesming of the 3T3G's may not be required if OT331.evel ts bein3 increased using EIN. ,

+ Rev.1

.

• Rev. 1 An _R_

Tod 600 Rev. 3 Pa ge 40of 81

5. 2.9 OT3G Shell-to-Tube Dif ferential Taiperstura Ltmit +*

+ Maintain OUG dlf ferenttal temperature less than 70?*. If this limit

+ is approsched, then:

+ 1. Reduce the cooldown rate in half.

+ 2. Continue s teastng on the af fected 3T3G.

F 3. Supply HN thru the startup control valve at about

+ 6 05 x 101bm/hr (if MN ls not being used).

& If the dif ferential tesperature approaches 100/ * , stop the cooldown F and asintain RG temperature constant. Remove decay heat by staamin3

+ the 073G(s) with the high dlf ferential camperature. Resume tha

+ :ooldown when the differenttal tempsrature drops below 70!*.

5.-2.10 Cooling mda when Both OTSC's are Unavailable foe AG dest Ra mova l *

(

Jse dPI '*f aed and bleed" to cool the RG vhen both OUG's are unava ilable. Open the Pilot Operated Rolle f Valva ; POAV), AC-RV-2, to

+ provide a cooling watar flow psch to the Reactor Bu tiding Sump.

i 5. 2.lt Oora Flood Tank Isolation *

• Isolats the Core Flood Tanks Lf:

• 1. Subcooling margin can be ma tntained above 25/', and

• 2. RG pressure is below 700 psig.

5.2.12 G Ldelina Flow Chart

• F . AppendLx C includes a flow chart and explanatory text showing the

" + tog!c path of the tube rupture guldelines.

+ da y . 1

• Rev. L d Rev. 3

TDR 406 l Rev. 3

Page 41 of 81 i l

! 6.0 00NCI,USIONS AND RECOMMENDATIONS I f

The ability of the plant to handle beyond design basis events can be

]

substantially increased and the RCS leakage can be reduced for design i basis leaks with the adoption of the following changes / additions to i tube rupture procedures.

! 1. Reduce ainlaus subcooling margin to 25F'

2. Replace tne existing RCP trip criteria with trip on loss of

! subcooling.

i 3. Adopt the steas generator isolation and pressure / level control guidelines of this guideline, l

i

! 4 Provide the RCP NPSH limits of Figure 6 for use during emergency j conditions.

5. Waive fuel pin-in-compression limits.

l 6. Control plant cooldown to limit the tube /shell delta T to 70F'.

I 7. Revise procedure entry point conditions to be leakage greater

than 50 spe.

, 8. Incorporate criteria for initiation of feed and bleed cooling into the tube rupture procedure.

l l 9. Adopt criteria for opening TBV's/ADV's if RG pressure stays j above 1000 pois during feed and bleed cooling.

i l , 10. HPI throttling should be allowed when subcooling is regained and

, pressuriser level is on scale.

I j

• 11. Core flood isolation triteria be incorporated into emergency j

• procedures dealing wu.h 14CA, tube rupture and steas line breaks j

• and in operating procedures dealing with dorced and natural i

• circulation cooldown. .-

i i

• 12. Decay heat removal system initiation under emergency conditions i

• be allowed ae 300F, i

! e 13. The additional considerations for dose reduction provided in i d Appendix F during tube ruptures being appended to the tube d rupture procedure.

f It is further recommended that these changes be implemented prior to

^

restart of 1MI Unit 1.

i I

+ Rev. 1

• Rev. 2

. d.Rev. 3

TDR 406

, Rev. 3

. Page 42 of 81 TABLE 1

• Tabular Values of RCP Emergency NPSH Requirements FOR 2 RCP PER LOOP OPERATION g INDICAT6D TEMP.(2) ALLOWABLE INDICATED PRESS.(3)

(F) (PSIG) d 94.4 203.5 (1) j t 194.4 207.2 (1) d 294.4 251.0 (1)

• 344.4 310.8

• 394.4 413.9

• 444.4 567.4

• 544.4 1187.1

, FOR 1 RCP PER LOOP OPERATION

• INDICATED TEMP. ALLOWABLE INDICATED PRESS.

• (PSIG)

(F)

• 94.4 115.5 (1) t 194.4 199.2 (1) t 294.4 243.0 (1)

• 344.4 302.8

• 394.4 405.9

• 444.4 559.4 l

• 544.4 1178.6 i .

NOTE:

d 1. RCP operation will not be limited by' NPSH requirements. Rather the re-d quirement to maintaina 275 psi differential on the #1 pump seal may be d limiting. Instrument error is not included in the 275 psig value.

d 2. Instrument error of 5.6F' for a small break LOCA condition has been d subtracted from the actual reading on the temperature instruments d (TI 959A and 961A).

t 3. An instrument error of 94.9 psi has been added to the actual reading i for the wide range pressure instrument (949A, B) based on errors d generated from a small break LOCA environment, o _ _ _ . __ b $ b_ _ _ __ _

TDA 60a Roe. 3 daga 43 of 31 TA812 2(1)

Pressurtzer Spray Flow for Various Pump Combinations IRMBER OF RC PUMPS RUNNING S PR AY F14W SPRAY IDOP OPPOSITE IDOP (% FULL FIK) 2 2 100 1

2 1 92 2 0 84 L (Spray LLne next to running pump) 2 60 1 (Spray line next to running pump) 1 ,

53 1 (Spray line next to Ldle pump) 2 50 1 (Spray line next to running pump) 0 41 1 (Spray line next to tdte pump) 1 38 1 (Spray line next to idle pump) 0 26 0 2 20

," 0 1 0 As a rule of thumb trlpping one pump in each loop will provLde a good balance between the spray flow rats and the heater capeetty. It will also provide good forced ctreutatt6n for cooldown.

1E)tE the fo11owing table will glve general guidance for the effects of running vartous pumps. This table was calculated for normal operating conditions. -

! IE)TES:

I

1. Esproduced f rom IM1-1 ATOG (Ref. 28), "Best Methods for Equipment Operatton", l Table 6. l

2. Entlee table was added with Rev. 2.

+ Rev. L

• Rev. 2 4 Rev. 3

TDA 60o Rev. 3 tage 44 of 81 7.0 REF ERENCE3

1. U. S. Nuclear Regula tory Commiss ion. NA C de po r t on the J anua ry 25, 1982 Stese Generator Tube bpture at R. E. GInna Nelear tower /taa:.

IfUa3G-0903 U. S.

• C.

2. U. S. Nuclear Regulatory Coastss ion. Sa fa ty EvaluatLon h port klated ta ths Restart of R. E. Ginna Nuclear Poser / tant. Ib e:ts t No . 50-2 4 4 HUAEO-0915. Ny 1982. O.S. 2 0.

3. C. Y. Cheng. Stess Generator fube Experience. NJAEG-0883. U. S. su C.

4 Roche' ster Gas & ElectrLe Corp. */coce durs E-1. 4. S /G fuba Apture. J una 23, 1982. GLnna StacLon.

3. "Proce dure !b . EL .O . Sa fety Inject Lon IntclatLon." Fe bruary 23, 1982.

Pt arle Island Nuclear Generating Station.

6. "Procadures E-1.3 and E-1.4 S/3 fube k pture." June 3, L982. Prarie Island Nuclear Genersting Station.

7. Duke Power Oo. "EP/0/A/1800/17. Stass Generatse tube Rup:ure." 3conee Nuclear Ststion,

8. INP3. "3 0E1 82 - 16 O r a f t . " January 4 L983.

9. IN PO. " Anstysis of Stess Generator Tube hptures at Oconee anj Ginas" INm 82-030. November 1983.

10. Babcock & WLtcox "Abnormat Trans tant Operstlng Gu ldelines. Three Mlle Island Nuclear Station - Unit One." June 1981.

' GFJ Nuclear Corp. "A:captablLLty of Intanttonally Wading Kain Steam i.Lnes 11.

During an OTSG Tube Ruptura Event." Jocument No.1101X-5320-Al$. July 30, 1982 - dngtneering Mechantes section.

12. L. C. Lan ese " WI- L O T3 G Tube k p ture Bas is f or Des L gn Leak Ra ta . " Au gust 31, 1982. SAP 0/118.

L3. M. Sanford "NPSH Requiresents for iiggyback Sefety InjactLoo Opracton -

N 1- 1. " Dece mbe r 21, 1982. M33-82 584 14 J. Veenstra. "Fuet itn Compress ton L!alts Durlug an SGra Event."

3 1-83-009. January 20, 1983. ,

L3 S. D. Le shnof f . "Machantest Integrity Analys ts of N1-1 OT3G Jnpluggad Tubes." TD4 338. G PJN C. Maech 9,1983. .

s

+ Aet. L

• Rev. 2 i Nov. 3

TDA 60o  !

Ret. 3

1. sge 45 of 81

15. P. S. Walsh. *TDA 400 Des f t Gutdetines for P1 ant Operaeton wt th ,

Stssa Generator tube kakage." PA -893, c'e br ua r y 15, 1983.

17 Darrell G. Eis e nhut , to H. D. Huktit. March 4, 1983. N cket !b .

50-289. March 4,1983. U. S. Nuclear Re gula tory Commission.

18. N. K. Savant , Dunn. 8. M. , J ones, R. C. , Res ponsa to GPd lettet ds ta d Au gus t . 2 8, 1980 ,' 1M I- 1/ 2 12 03 s . Do cume n t Lia n tl f', a r  !

12- t L! !718-00. Se ptembe r 26, 1980. Babcock 5 Wilcox Co. ,

Lynchburg, Virgints .

4

+ 19 H. D. Hukitt (GPJ!h to D. G. Eisennut (NAC). "aCP f rip on 2 3/ *

! Subcooting1 Marg in. " March 31,1983. 5211-83-017

• 20a. W. Drendall . "Saturatton Maratn Monitor Inaccuracy sNon-l Aceldent Conditions)". Ja t:. Ib. C-1101-6M- 53 50-002.

J un e 14, 1983.

!

• 20b. G. J. Ssdauskas . "tMI-1 Saturation Margtn Manitor i oop Error i

Analys ts." Cate. M). 1101X-3223-00 9 Rev. 2. J un e 14 , 1983. .

G pud C.

21. C. L. lehmann. "0T3G Lestags sad Operattng Llat es ." fJA 411 G PJN C. March 1983.

22. 1M1 Jnit 1. " Plant Llatts and e recsuttons. OP 1101-1. Ra v .

14. " Pebruary 2o ,1982. G PJd C.

(

23. "3etsestnatlon of Minimus Required rubo Walt Thtekness for 11/-/A Once through Steas Generators." 5 AW- 10146. October 1980, i e + 24 J. Veenstra to D. G. Slear. " Thermal Shock 01acif' catton For Smett Break Opersting Guldet tnes. " Septembe r 9,1982.

TM1 82-01 t.

I & 25. M. H. Ko s tr ey , News to L. C. U n es e "IM 1- 1 Rea cto r J oolan t P umps NPS3 List es Westinghouse Pumps" MC-1752. March 30,1983

+ 26. L.C. Pwu, GPJN Calculatton # 1101x-54 50-013. Rev. 2 " Minimum Subcooling Margin for RJP Operation." Augus t 5,1983, i

• 27. L.C. Pvu. GPJN'Calculatton # 1101X-54 50-015. "3a turattoa Mrstn Mont to r Ad jue tsen t on El ev a t ion Ot t f or enes Be twe en the t r ea s ur e Tap and the Tope of the Hot Lag." July 2 9, L943.

!

• 28. MW Abnorant f ransient Operating Guldet tnes ' ATOG).

4 74-1124158 00. Lrnchbars, VA, Aprit 24, (183.

• 29. Telecon vlth Ralph Roeser (MW) bt L. C. Lanese. J une 13, 1983.

• 30. Babcock 4 W!1cox Co. "Evalum*.lon of $5 LDCA Operattng Procadures ani Ef fectiveness of Isergency Tssdveter Spes/ for MW - Deslaned operstlns NBS$". MW Doc.10 77-11412 70-00 Lynchburg, VA.

Fe bruary , 198,3.

. + Rev. 1

1. h L__.

• TDA 606 Rev. 3

/a Se 46 of 81

• 3L. L. C. Pwu. GPJ Calc. E . 1101X- 54 50-014 "NI-1 Cooldown Ra te us Eng Atmospheric Dump Valves", Augaat 3. L983.

• 32. J. P. togstto.

• TMI-t De ca y ilas t Ramovat Sys tas".

June 1,1983. MS3 504

• 33. N. G. frikouros to R. J. Toole. *ihyrold Dose Limit for OT3G Isolatton during SG fube Rup:ure", SAPO #140, datad July 13, 1933.

• 34 TMI Unt t 1 O P-1103-6, Rev. 2 4 " Reactor Coolan t lump Opera tion".

S te p 3. 3, 2.

• 35. J. R. Wh i ta . REIAP5 Analysis of Stasm Ganeratat tube Ruptuce Translents in a Generic Lower oop Plant. R. P242 0-4 Nuclea r Se faty Analys ts Center. Palo Alto, CA. 1983.

d 36. G. Lehmann "TMI-1 OT3G Hot Tasting Results & Evaluat.on".

d TDA 488. Octobe r 25, 1983. 3 PJ.4 C.

I d 37. R. L. Black "3upplement to ON3-3 linal AT3G*. July a ,198J.

, d E3 C-2 93. Babcock & Wlico4 Co., Lynchburg, VA.

i 4 38. L. C. Lanes e "TMI- 1 90/ / hr C ooldswn Ta s t o f 0: t ob e r i L 981' .

d October 10,1983, 3APO f 191. G PJ3 C, ' P ars L ppan/, dJ.

l 4 39. G. T. Fa irburn "Hes tup /0ooldown Ra ta Ltalt fot 177/ A Plants".

d Decambec 11,198 L. GT!-81-039. Babcock & Wilcox Co.,

6 L/nchburg, VA.

t 9

/

4 I

i i

4

+ Aa v. I a Rev. - 2

/ Rev. 3

l

~

l l Dx ae

si;CE c j' g3 $ o v. 3

!!s-/

l Pac: 47 of s:

835hg M_i s j*'3rI I VkNhn; e Es> r- Eigg

~

!i!f'

~

ii l I alit l xglig

.s  :, I l l gi

I N,e , ,

ss s y  !!=

== g l _c.

"l g L

! i i' ,- - i=

_ _ _ p _ _ _ _ _;_ _ _ _ _ _ _ _ .

i gillie

" n.

I! l

. .i. l. ., .. . . _ .

i I t

z i I '

& =4 !!** = _it }II \TNy::1[j.

1i  !* =

! F 'N 5 s,i- gaja\-=~

qJE ! _

i ij is j hi l =i: h !E! h e l

E u- ____

i i

.___p__ ___j__. .______- -

o ' I qj l o _,g!!,

s , e. h~k,

!Ii *kI!'os; ~ I l' _.:. = 'eis)e .!! b_!.=a!!

a

.. __ N g

.s , I s,

o 1 i A is

a. M:p!

5: -  !, lg- Y - y% -

) 5 z  !  !!!il i o,s~' Jiir:: ,}!

i NY l sfgl !H,i, (38 "V

i  : _

I' = > >

$!!$3:

j ti glW,llWV g'P a i

=

i B

I i 1 l l

1 i

' = l l a

1 J

TDR 406 MGURE 2 Rev. 3 Page 48 of si Break Flow for Single Ruptured Tube 1050 , i i i 945 -

1 il ill 840 -

E 735 -

n.

N 630 - -

E

{

n.

525 -

$ 420 -

a -

315 -

210 - 1: 25'F SC, PUMPS ON -

II: 50*F SC, PUMPS ON 105 - 111: 50*F SC, PUMPS OFF -

0

/ 0 10 20 30 FLOW LBM/SEC l

l .

l-l

AGUM 3 Effect of RC Pump Operation on Integrated System Leakage for Single Ruptured Tube leo i I I I I I I I i I i n.

g 180

- 25 F SCM, RCP'S ON **.*...- -

= --- 60 F SCM, RCP'S ON **...

E

• 60 F SCM, RCP'S OFF 20 2:

140

~-

~

st) /*..' /

O

• .*.*/ /

/

120 -

s / ,,,

o g /

u.

...**..'. /

2

~.*'s **/ / /

100 - -

w q /

/

su o 80 -

/.**.*..- / -

1 uJ *,.*.'. y

/

G ,/

x 2 60

- * . . /. , -

l r /

g .*- /

~ *

/

/

l 4g -

,,.***/.*,/ -

/

/

/

. '*. *.* /

20 - o=

.$p./ 3*g w

,~

ruo*

0* I I I I I I ' ' ' I I '

20 40 60 80 100 120 TIME IN MIN $

l

g ' ' ..s ' , f 6

~ '

- [': '. ... ~ , ' f' $~

- ,c; .-. . , - 4 / -.,;. - - , ,'; ,, ,c . . ;n f.;. 3

, ~ , * , r: tg -7 ,, p .gt ; i ' , , ~ , ,' . j' ' * , * = . ? ' 1 *l , ' ..a  : / ; *.k. "2 T .j

. , ,_ , ,<< 7..;7 ,+ .

TOR 406 i Rev. 3 Page 50 of 81 FIGURE 4 Mass and Energy Capabbities of HPI and PORV l l

880 40

(.55)

/

/

800 - TWO HPI PUMPS j -

35 l.9 )

/

/

PORV ENERGY 720 -

R E M O V A L -> .. **"'""""""""' .y,/ - 30

,..- / ** .,, 0.8 >

l **

/

/ ..,***

'E 640 -

/ 25 m

• / .' -

a S j .,, (2.6) g Y / '

. m N PORV LIQUID REllEF 560 -

/ 20

$ ~ / (5.s ) E 2 / *

, /

480 -

15 l

/

! ONE HPI PUMP / 5-

/ -

r -

7 400 - / -

10 j

/

l

/

/

320 -

/ -

5 4

240 I I I I O

O 500 1000 1500 2000 2500 RC SYSTEM PRESSURE (psig)

• • Energy relief is product liquid relief capacity and *1.0 ANS enthalpy of 100F subcooled water. 2535 Mw (t)

\

1 FIGURE 5 E R # 406 Rev 3 Page 51 of 81 Time Behavior of Subcooling Margin for a Spectrum of Ruptured Tubes 100 j

~

. j 80 -

/

"" ~

/

= -

,G 60 - /

< /

E /

E I -

/

40 I 3

E ig l

8 TUBES ,

s' W .

g 5 TUBES N, '

N 1 20 4 -

I t "s'/

~

Y l 25 TUBES *,"- 26 TUBES a ' ' *' -

0 E 12 11 24 30 36 42 48 54 60 TIME (MINUTES)

TDM 408 Rev. 3 FIGURE s Page 52 of >

RCP NPSH Curves 1800

/

Minimum NPSH for RCP Normel 1500 Operations OP 1101 Figure 1.0-5.8

---

Emergency NPSH for 4 RCP operation

/ I or 2 RCP per loop.

1 00 / f f

- - - Emergency NPSH for 2 RCP operation f (one per loop) l 1300 - -

25'F Subcooling Margin Curve. j Note:  !

12*F and 50 psig instrument errors l 1200 have been incorporated in the normal i

NPSH curves, while 5.8*F and 94.9 psig j errors were for the emergency NPSH curves.

1100  ;

l 5

i 1

5'8"" l I #

E . Minimum pressure to maintain

$ compression forces on fuel clad g

3 900 (natural circulation) OP1101-1 Figure 1.0-5.11 #

I *[//

f j

l 800 -

= (Forced Flow, Figure 1.0-5.10) # /.[:

3 s //il

'E 700 ' "

' l*/:! '

l/l j*j' 800 ,

/.,:

500

! //l Minimum NPSH for RCP Normal /

Operations OP 1101 Figure l/'///

• l 400 1.0-5.6 .

/j/ '

d "

s.f.l ^#

300 --

- - - - -l -- - - - -l - -

• l l

Maximum pressure for operation /

200 of DH system (OP1101-1, ~

Figure 1.0-22.2, 22.3) /

l ~*.

! 100 l .,....... **

O 100 200 300 400 500 800 Indicated RC Temperature (*F) '

I I

1 TDR 606 l Rev. 3 f a ge 53 of 81 APPENJIX A.

IiI-L 3Gn PRO CE]W' GJ DELIN2 ANALr313

+ Re v. t

• Rev. 2 d Rev. 3

1

\

TDR 600 Rev. 3 i Page 54 of 81 '

l

, COMPAR133N 3/ GJDSLLNE3 AIO 22 1202- 5 RZV. 16 I3 TIE REQJIREMENIS OF VARDUS 30URG DOCUME3T3 A.0 S C3 PE The purpose of this Appand(x is to compare the guidelines of this TJd to the current revision (16) of EP 1202-3, OT3G Tuba Leak / Rupture , and guidat tnes , requirements , commitments or recommendattons from various sources. The sources reviewed were the I4I-1 Anticipated f ransient Operattng G2idelines (draft o f 15 Hay 1981; heretnafter referred to as AIDG), '01arlf tcation of IMI Action Plan Requirements" (raferred to as NUREG 0737), and the Safety Evaluation Report r elated to restar t o f Ginna (Raf. 2) (referred to as NURdG 0916) and the LIPO draft Significant Operating Event Raport of 04 Janucry 1983 concerning staam generator tube leaks (re ferred to as SO ER) .

Wi th r es pe ct to NUAEG 0737 and 0916, only those requiremants or comm* tments directly related to a tube leak emergency procedure were considered. With res pect to AIDG, only the followup guidance for tune leaks was considered, and then only Lf Lt dif fered from the guldance

+ in the latast approved tube rupture procedure (EP 1202-5 dev. L61.

+ Wt th respect to the SOER, only the recommenda tions rela ted to procedures were considered.

+ Ihe results of this comparison work are summarized Ln Tao te A-1.

I l

l l

l l

A-1

+ dev. L

• Rev. 2'

^

d Rev. 3

9 TDR 600 Rev. 3 Pa ge 55 of 81 TASLE A-1 Comparison of TDR 406 Guldelines to Other Sources j Sourca Addressed Bf L202- 5 406 Requ irement AIDG 0737 0916 30 ER Commen ts Run RCP's with X X X X low RC3 pressure RCP restart X X X X X do spactfic guid-anca providad for RCP restart with a "solld" pr es s ur iz er .

Subcooling margin X X X X HPI throttling X X X X Steam LLne X X X X ATOG does not flooding recognize E4I-l capability to flood steam Lines w L ehou t da ma ge .

Cooldown of X X X Guidel.ines provide dama ge d OT3 G f or continued s teaming o f af-fected OI3G for cooling' except when OT3G isota-

- tion is r equ ired .

Spectfy entry X- X X X Symptoms.

threshold Me thod for plant X X X X cooldown following SGTR Plant cooldown X X X ATOG refecs to following SGTR 2xcessive heat with stuc'c open transfar sect'on 3G safety valve dP 1202-3 and guidelines pro-vide means for minlaizing prob-ability of lif ting a 3G relle f valve .

A-2

+ Rav. L -

• -Rev. 2.

f Rev. 3 tM

TD.1 605 Rev. 3 22ge 56 of 81 TASLE A-1 (continued)

Source Addressed Bf Requirement AIDG 0737 0916 SOER 1202-5 406 Commen ts Af fected SG X X X peessure control OTSG tube to X X shell dL f f eren-tLal temperature

+

Criteria for X Ouida tines pre us ing ADV's in sume tha t preference to s teaming to main condenser condenser is always preferable to steaming to a tmospher'e .

ConsLder multiple X X Guldelines pra-tube ruptures pared in con-sideration of Cons Lder tube leaks X X thesa two cases.

In both OTSG's

+

HPI on inadequate X X do t s pec* flesily S CM stated in TDR

, 406, out L t is . an

' LaplicLt re-quirement of the HPI throttling celteria.

Constder excessive X X Overfaading con-primary to secon- sidered by daey heat transfer EP 1202-5.

Constder loss of X X. Guldelines make no offsite power partlcular dis-tinction be tween of fstta power available or unavailable , but they do provide guldanca Lf the equlpment d Ls-abled by L33/ is unava il able .

A-3

+ &a v . L

- -

• anym -

IDa 60a Rev. 3

' Page 57 of 81 TA3LE A-1 (continued)

Source Addressed Br Requ iremen: AIOG 0737 0916 30 ER 1202-5 406 Commen ts OTS3 level X X X control Primry pressure X X control wi th-

. out pressurizer .

sp.ay 4

Isolation of- X X X sf fected OTSG t

o I

i e

5

~~-

A-4 l'

+ dev. L~

• Rev. 2-4-Rev. . - . - .-,

TDA o00 Rev. 3 dage 58 of 81 A.1 In terpre ta tion A.1.1 "3aquirement" Columa L

These see paraphrased descript Lons of guidelines , requiremats, commitments , or cecommenda t tons from source documenes.

A.l.2 "3ource" Columns i

These columns define the origin of the requirement considared in this compar is on.

A.l.3 ' Add ressed B,r" Columns These colums de fine the documsat which answers the requiremaats. If a mark appears in the 1202-3 column withou t a c orres ponding mark in the TJ1406 column, it means that the guidanca in EP 12 02- 5, Ra v . i5 should be re ta ine d in the r ev is ion tha t incorpora tes tha guidelines of Section 4 of this T0A. If a mark app 2a r s ,in both columns , i t generally means tha t the guldelines in this IDR supercada the guidance in EP 12 02- 5, Ra v . 15. IE a mark appears in the TJA 406 column alone, it denotes a new guldet.ina to be incorporated into the revisel EP 1202-5.

A.l. 4 Commenes

• This column provides add;tional information if necassary. Ihe gu;,delines in this TDR supercada the guldance in d2 1202-5, Rev. 10 .

If a mark appears in the T3R 406 column alone, it denotas a nes guideline to be incorporated into the revisad EP 1202-5.

A-5

+ Re'v. 1

• Rev. .-.2 v

IDI 605 Rev. 3 da ge 59 o n 81 APPENJIX B P13 CE]URE CHMGE 3 A/E f Y E VAIJAfIJ E L

)

+ Aa v . 1.

f

• Rev. 2 I d Rei. 3

. l 1

TDA 400 Rev. 3

/a ge 60 of 81 B.O PROCEDURE CHA: IGE S AFETY ZVAWAfION3 The purpose of this AppendLx ts to address the safety tmpt teatLons of the key changes required to tsplement the tube rupture guldelines described Ln this T]R.

L. RCP trlp on loss of subcoolIng margin (3C4)

2. Change in SC3

• 3. Shell/ tube delta T of 702

• during amargencLes 4 Revisad RCS NP3R curve

5. Relaxation of fuel pin in compress ton limits

o. OT3G isolation celteria 7 RCP res ta r t ' cr ite r ia

8. H PI thro t tl ing a t 0" ins tea d o f 100 inches .

9. Leaving A3V open when no OT3G heat sinks and RC3 ls above L000 ps t g .

• 10. Isolatton of core flood tanks Esch of these L tems is addressed below: ,

8.1. RCP f rip on Loss of Subcooling MargLn

+ In a lettee dated Harch 4 L983 to H. D. HuklLt.(Aav 17), the K1C supercedsd the actLons requirsd in IE Bulletins 7 9-05C and 7 9-06C.

The s taff has instead concluded that "the need foe RCP trip foitowing a transient or accLdent should be determined by each app 1Lcattoa on a case-by, case bas is considering the Owner's G roup input." For several years , the B&*.i Owner's Group has supportad the concapt of ACP telp on 1.oss of subcoolEng margin. In Reference 19, GPJdC informed the NAC of their reasons for revising the telp criterion to. loss of subcooling

. ma r gLn . The safe ty evaluatLon for this change has been transmitted separately to tha I4I plant staf f for' review and approvat .

B. 2. Chang 2 in Subcooling MargLn

• GPJNC has evaluated the instrument error assoc *.ated with the subcooling margin monttor and alars. Under normal containment

• conditions , the loop error is + 10.1F0 (Ref. 20a S b). Under the temperature and radLatlon environment of a small bre.ak LOCA, this

, error is no worse than - 21.7F*. fhe basis for the ortginal 3CM was B-1

+ Rev. 1

-* Rev. 2

] Devn El

9 TJR 4J0 Rev. 3 da ga 61of 81

• 5/* geoma try correction plus 4 5/* s tr ing inaccuracy. Racent

• calcula tions (Ref. 2 7) have shown tha t only L. 30F geome try

• correctLon to the top of the hot Leg is required. Therefore safaty

• margins are not decreased by this change.

A completa safety evaluation has been prepared and transmitted to the site separately f or rev Lew and approval.

B. 3. Change in Shelt to Tube Dalta T The existing emergency limit for tube / shell de tta T a t TAI-l is 100/ ' . The LLmit is being revised to reduca tensile stress on leaking OTSG tubes . Previously the 11mte of l00!* was based on stresses to in ta c t tub es. Indeed, the LLmit has been Increased to 15a?' by Babcock & Wilcox; and, Lt ts valid in the absence of dagradad OTSG tubes.

The more restrictive 70!* limit tacreases plant safety Limits by reducing the likelihood of propogating a crack. This analytLcal work is documented La Ra ferenca 15. This change can be made undar the provisions o f 10 CFR 50.59 bacausa Lt does no t af fact te chnL cal spect f t ca tions. S*.nce tube s tresses ace reducad, plant safa ty L imits are increasad and the two addttional crLteria of 10 CFA 50.39 are also me t . Ramely, there are no new acctdants introduced into the plaut tha t have not been previousl y analyzad. Since shell to tube detta T has not been explict tly addressed in the F3 AR, exLsting p.aat safa ty margins have not been dacreased. 'n fact , plant safety margins have been increased sinca the allowable delts T has been decreased.

B. 4 RC P N PS H L*.m t es

- Reduced NP3H limlts bring the pump closec to a point of cavitation.

However, NP3d requirements have been reduced for lower temperatures as F de termined by the pump manufacturer Westlnghouse in Ra fecenca 2 5.

Margins have been mod'.f t.ed based on safety margins Ldentifled by taa pump manufacturer, therefore, the probabil.ity of pump cavitation has not been incrersed and plant safaty margins are protected. Matther are technical spacL fica tions af fected. The operatton of reactor coolant pumps a t low RCS pressures does not introduce any new accident or trans tent other than those already analyzed in the F3 AR. Pump operation is allowed at these lower RCS pressures but at a hlgher subcooling margin. Since real plant subcool*ng margin is stllt being maintained as discussed in item 2, there is no reduction in planc s a fe ty. Operation of the resctor coolant pumps increases plant safa ty margins ut.th res pect to thermal shock, increased 0NB ratios , ani

.mproved capabilitles for degassing the reactor coolant systam under tube rupture conditions when mintsam subcooling macgins are being ma in ta ined .

B-2

+ Rev. 1

• Rev. 2 d Rev. 3

.. - - . . - --- _ . - ~. __ . _ _ . . .- -

l TD1 40o Rev. 3 da ge 62 of 81 B.5. Fuel Pin and Compression LLmt ts

+ As addressed in Reference 14 B&W has cecommandad that fuei pin in compression, limits be waived during certain plant trans ien t condittons including steam generator tube rupture events. Fuel pin in i compression limtts have been established in order to ma(ntain cladding in te gr ity. Waiver of these limits does not reduce pLn integrity although reanalysis by B&W may be required when fuel compression 11mtts have been waived. SLnca cladding Latagrity will nave to be addressed each time these limits are violated, the damonstration of acceptable clad integrity will be mada. No new accidents or transients will be introduced then have been previously analyzed in the F3 AR. Similarly, plant safety margLns witL not be reducad, namely, cladding integrity will not be challenged.

B. 6. OT3G IsolatLon Oriteria The existing s team generator tube rupture procedure, EP-1202-5, at tows the operator to isolata the af fected steam generator anytime ACS pressure is below 1000 psi. The revised criteria would allow steaming o f the O TS G un til BWS T level is 2 L f t . or esdiation Limits approach s ite eme r ge ncy l iml es . Steaming of the OTSG introduces the potential 4 foe increasing of fstte radiation doses; however, these llatts will be maintained within the requirements of 10 Ci2 Part 20. It should be I noted that the isolation of the steam generator on high radiation is keye d towar ds ma in ta in in g Par t 2 0 lim L ts . Staaming of the generator when possible increases the chances of preventing major of fstte releases s ince floodtng of an OT3G can result in liquid retlef oat of the stesa safety valves with the possibiltty of safety valva failure.

The value of BWST level at 21 ft. Es suffletent to assure a source of water for the 2CCS pumps. The value o f 21 f t . allows suf f tcient inventory to flood both staam lines and allow the plant to be p.acad l-on feed and blee'd cooling in the racLeculation moda from the RB building sump (Re f. L2, L3). It should be further noted that tne doses associated wLth a steam generator tube rupture were increased when the requirement for maintaining subcooling margin was introduced Ento the plant procedures following tha I(1-2 accLdent. At tha t tima

' the issue was addressed in writing to tne NRC staff (Aef. 21) justlf tcation for the change was that Part 20 limits were being ma in ta Ined. This criterion is stL11 being maintained with the change in OT3G isolation criteria.

These changes can be msde under 100!150.59 because safe ty margins are not decreased. Technical Spect fleations are.not af fected by this ch ange . bb new accLdents ce trans tants are introduced wnich have not .

been previously analyzad since 'this guidance is intended to deal with events which are beyond the design basis of the plant-(L e., tube, rupture without condenser and 10P's) .

l l

3-3

+ Rev. 1

• dev. 2 d Rev. 3 .

TDA 606 Ree. 3 Pa ge 63 of 81 B.7 + RCP Restart Celteria

+ The RCP Restart Critacion assures that the pumps are not restarted

+ until the coca is adequately subcooled. ' No ta that there are other

+ ACP restart unrelated to this criterion). Re farence 19, and Seccioas

+ B.1 and B. 2 demons tra te tha t the core is adequa tely subcooled wt th

< + RCP's running and a 2 5?' subcooling margin. No Te chnical

+ Spect fications are af fected. No new acc! dents or transients are

+ introduced Lnto the plant; no safaty margins are decreased and no

+ accLdent consequences are increased. Allowing an earlier pump restart

+ glves the operator greater control ~over the plant since forced flos is

+ preferable to natural circulation coolind. this change can there fore

+ be mada under the provisions of 10CER50.59.

B.8 + HPI f hrottling at 0 inches ' Indica ted Level

+ Ihe safety aspects of throttling HPI on 2 5/

• subcooling margLn are

+ addressad in section B. 2. Core coolabiltty is not d? pendent on ene

! + pressurizer levet at which HPI ls thrott' led L.e., core coo 1 Lng Ls only

+ dependent on an indleation that the core coolant is subcoolad. the

! + bas'ts for requiring pressurizer level is so that the exLsting

+ pressurizer heaters are covered with wster so that they can De

+ energized. Energtzing the heaters before they are covered causes them

+ to burn out. On the other hand, there is no need to refill the

+ pressurizer to the 100 inch level at full HPI f tow. In fact , this

+ flow rate is undesireable for beo reasons. Rapid filling of the

+ pressurizer causes an RCS pressurizatLon during cond.tlons when

+ pressurizee sprays are unavailable. Insurges to the pressuelzer d

+ compress the s esam space. Pressurizer pressure must be ' reduced e ither

& by sprays (t f available) or pressurizer venting (vent line or POAV).

F Controlling the SPI flow mLntsizes the insurge rate, and hence, che

- + press ur iza t ion. This reduced pressuriza tion provides more margin to

, + the 100!* subcooling curve thereby minimizing challenges to the i + thermal shock / belttle fracture limit.

+ Ihis change does not represent an unreviewed safaty quest'on because:

l + l. No change to the Technical SpacLfications is requir'ed i

+_ 2. No new accLdents are introduced to the plant (the operator Is.

i still required to cover the pressurizer heaters before energizing them ) , and t

+ 3. The consequences of previously analyzed accLdents/ transients is not incre ase d. It is less likely that the operator will violate ~

the 100?

• subcooling margLn. Core cootability is not dependent on established pressurizer level, but only an adequate subcooling margin.

B-4 ,

. + Rev. 1

• Rev. 2

~ d Rev. 3

TDA h0h Rev. 3 Fage 64 of 81 B.9 + ADV's Open when RCS is above 1000 pst g with no OT3G Heat Sists

+ Ihts TDA provides guidance for certain situations well beyond the

+ design bas is. One su ch s L tua tion is the cas e wh ere the plan t is one

+ feed and bleed cooling, but RC: pressure is above 1000 psig. This

+ condt tion can result in L iqui d r et le f ou t o f the O TSG sa fe t y valv es .

- + Opening the AJV's is the preferred course of actLon because it

+ mintsizes the chance of an uncontrolled blowdown though tha OISG

+ safe ty valves. This condleton is well beyond the plant dasign basis ,

& Plant Tech Specs are not affected by this procedural s tep. fhere f ora F the change can be made under the provisions of LOC?R50.59.

+ Beyond the consLderatton of whether this changa can be made under tha

+ p r ov is ions , o f 10C?R 50.5 9, ; t L s b el ie v ed tha t o pe n in g t h e TS V' s/ AJ V' s

+ is prudent and reduces the risk of an uncentrollad release to the

+ env ironmen t.

B.10

• Criteria for Core Flood Tank isolatton

• The purpose of the core flood tanks is to assure core cooling for

• L3CA's in which: 1) ACS pressure is below 600 pstg, 2) HPI cannot

• provida core cooling, and 3) rcd pressure is too high for the LPI

• sys tem to operate. Ihe only sLtuatlons when these conditions occur

• see: 1) design. basis L3CA's , which HP1 does not initiata before the

• core begins to uncover and 2) core flood line break accidents with an

• HPL f ailure, and 3) sm2LL becak LO CA's in wnich the break is just

• Large enough to remove decay heat , but not to de pressur iza the ACS.

• For the large break LO CA s itua tion, C?T lsolation is not a princip2 L

• concern. The operator should isolate to prevent nitrogen introduction

• into the RCS once the tank is empty.

"

• For small break U3CA condttlons, a subcoolad RCS means that there is

' suf fic;ent heat remaval. In the pressure ranges in which core flood

• tank isola tion is o f interest , one HPI pump supplies suf flcient flow

• to keep the core covered (500 gpe). If the RC3 Ls 25? subcooled with

• the RCS below 700 pstg, then the C/T can be is ola ted .

• The core flood tanks also function in one non-LOCA situation - steam

• line brea'c accident. For large steam line break accLdents , the CIT's

• provide shutdown margin assuming the most reactive rod Ls s tuck out.

• There f oca , CPT isolation cannot occur untti elther HPI is operating

• and providtng a source of borated water to the core or unt'i ait rods

• have inser tad. If both these conditions ars met , then a plant

• procedural change can be made without introducing an unreviewed safety

• qu es tion .

3-5

+ Rev. 1

• Rev. 2 d Rev. 3

TDA 40s Rev. 3 Pa ge 65of 81 APPENJIX J GJ DSLINE3 /Lar4 Cl%U 0

k Rev. L

• Rev. 2 d Rev. 3

TDR 6J5 Rev. 3 Page 66 of 81 C.0 G313ELINE3 iLOJ CHA1T The f tow chart in this sectLon shows the major milestones and declaion po ints on the pa th from opera tion at full power through the development of an OT3G tube leak / rupture to inspection and repa;r of the damage. The flowchar t is not meant to be an exhaus tiva treatment

+ of all actions required to reach cold shutdown, rather it ia the

+ framework upon which a procedure can be cons truct2d.

C. L IdIERPRETATION

, Diamond boxes are dactsion points. The path taken out o f a diamond dapends on the answer to the question posed in the diamond. Boxes enclosed by a single line represent staps that take secands or minutas to a xa cu te. Boxes enclosed by double lines represent tasks tha t n2y require minutes to hours to accomplish. For the s ake of s implici ty, certain steps that wl.ll be required in the procedure have been omittad (e.g., confirming reactor trip, or making radiation surveys of tne secondary plant).

The dects ton points immediately folLowing a double-line 'cox are maant to force tha operator into a " thought-loop" so tha t L f conditions change , the oparator may select an alternate , more appropriata cooldown pa th. For instance , while cooling down on forced flow wL ea a tube leak in excess of 50 gpm, the operatoe should cont'anualty Laquire as to whe ther the Reactor Coolant Systas pressure and camperature are wLthin the capability o f the De cay Heat Ramoval S fstem. If s o , snen the operator should obv Lously change the RCS heat removat moda from s taamlng the OT5G's to using the DHR3. If not, then the oparator should continua to ask whe ther ths RCS conditions are suitabla f or force d flow cooling v La OT3 G's , L .e ., is subcooling inadequata , are tha OT3G's available/3K for use , are the AC pumps available. If the answers to these ques tions always no, fes , and yas , then continued forced flow cooldown is acceptable. If any o f the answers changa ,

th en the th ou g's t flow breaks out of the loop and presents tue operator vith new ce;teria for selecting an alternata cooldown mode.

+ rhis " thought loop" philosophy should be incorporated Lato the proce dure revision.

C. 2 + F:LOCEDUR AL OB JE CT IVE

+ The objectLve of the tube leak procaduce is to expeditiously coot down

+ and depressurize the plant so as to alntmize petmary to secondary

+ l eak ge and th us , i t t s h ope d , o f f s ite do s es . The process involves

+ recognition o f the event , shutting down the plant , and cooling down F the plant to the point where the Dacay heat Removal S stem f can remove

+ core heat.

C- 1

+ Rev. 1

• Rev. 2 f Rev. 3

IDA 405 i

Rev. 3 da ge 67 of 81 C.3 + ENT1f POINI

+ rhe proc 2 dure will be entered when a primary to secondary leak is

+ encountered that requires the plant to be shut down. the syuptoms of

+ a cube leak requiring shutdown are described in IDA 400 ;def. L6).

C.4 + PLANT SIUTDOWN

+ rhe rate ef plant shutdown from _100% power wlLL be datermined in pact

& by the magnitude of the RCS depressurization due to the leak. If the

& Leak ts small (the Mskeup Sfstem is able to keep up wLehit), tnen the

+ pl ant can be shutdown a t a cate commensurate wL th equtpment

+ =apabilittas and, to a certain extent, the leak rate . When the F reactor and turbine are of f line , the plant is ready to entee the F :ooldown phase.

5 + However , Lf the leak results in RC3 depressurizatLon to the trip

& se t po int , the reactor and turbine will be of f line immadiately. Tha

+ ensuing transient will have to be dealt wLth and the plant status wlLL

& have to be evatustad prior to the cooldown.

C. 4.1 + Preparatton for Cooldown ,

+ If the shutdown transient results in a loss of subcooling margta, HEL

+ must be Lamediately actuated and the Reactor Coolant Pumps (ACP's; F must be immedtately tripped. The OT3 G's mus t then be evalua ted far

& sultability as heat sidca for the RCS.

& If the shutdown transient does not result '.n a loss of suecooling

+ margin, tha OTSG's must set 11 be evaluatad for sultability as ACS heat

+ s ink s .

+ If neither OT3G can be used because of high offsite doses oc low BJST

+ tevel, then the cooldown will proceed directly using the HPI " feed ani

+ bleed" me thod.

+ For the balanca of the discussion in this section, assume that HPI

+ "fsed and bleed" is unnecessaey.

+ If the RCP's are off, Emergency Feedvater flow to the OTSG's must be

+ con f irmed, ths ICS will automattcally control OTSG level at 50% on

+ the Operating Range if the RCP's are off. If subcooling macgLn is

& 257, the operator must manually raise the level to 95% to promota

+ two phase nstural circulation in ths RCS,

+ Sines a forced eleculation cooldown is the most preferred moda, the

+ RCS conditions should be evalutted for RCP restart. If subcooling

+ ma r gL n is r ega in ed and the RCP J PS H lim'.ts ac e ma t , 2 ACP's should be F r es tar ted. If the pumps cannot be restarted , the cooldown must

& proceed by natural circulatton.

0- 2

+ Kav. L

• Rev. 2 d Rev. 3 j

I01 600 Rev. 3

/a ge 68 of 81 C.5 + PLANr C)3LDORN

& Ouring the cooldown, RCS condLttons must be contlnuously avaluated to

+ ensure that tha cocidown mode is appropriate and to datermine whathec

+ conditLons are suitable for the Decay Hest Removal S sf tem.

+ Regardless of cooldown mode , the following i tems , may be encounterad

+ while cooling down.

C.5.1

• HPI t hrot tling

• The exLstlng HPI throttling criteria are unchanged witn the foilowing

• excaption: HPI may be throttlad when subcooling is regained _ and

• pressurizee level comes on scale.

C.5. 2 + O TS G Steaming

& the affected OI3G may be steamed for RC3 heat removal purposes , but Lt

+ must be steamed to avold lif ting tha Main Steam safety valves , prevent

+ premature Steam line flooding, keep OT3G pressure less than RC3

+ pressure, and control OT3G tuba to shell dif ferential temperature.

C. 5.3 + OT3G Shell to Tube Dt f ferenttal Temperature

• It is necessary to mLaimize shell to tune dif farential camparature to

+ minimiza tensile s tresses on the OT3G tubes. As no tad above , s teaming

• ls one way to accomplish this; another is to decrease the cooldown

& rate; a third is to usa main Feedwater to cool the icwer downcomec.

C. 5. 4 + O TS G Press ur e C on tr ol Wh en RC3 Pres sur e is G rea te r i nan 1000 ps ig o + During a natural ctreulatLon cooldown or an HPC faed and biecd

" + cooldown, RCS pressure msy stay high. Emeegency Feedwatee can be used

+ to quench the s team s pace, if the OT3G is floodad, inventory can be F ret teved via tha Turbine Bypass Valves oe the Atmospheric Dump /alves.

C. S.5 & Cooldown Ra ta

+ rhe cooldown rate should be limited to Less than 1.61/hr to avoid

& resctor vessel brlette fracture concerns. It may not always ba F possible to observe this limit due to the effects d/t cooling and the

+ occassional necesalty to steam the damage d O T3 G.

C6 + EXII PO INT

+ The operators exlt the procedure when the RC3 heat s ink becomes the

+ Decay Heat Ramoval Systes.

l 1

C- 3 1

1

+ Rev. 1 i

• Rev. 2 d Rev. 3

a;lts a enia w o '

s FIGuilt C 1 Pn0Ceu 45 3 o SINGLE & MULTIPLE TUBE RUPTURE GUIDELINES tt t8641100 & a Orla411 a e

e ,

e

)( 095G a

a gasssilmle6444 test

- ivu vinu tt

' m SueC00tt4G tlAE , . TAC O - 4 %p thallall

• a IR8P vl1 A. A.G,. ,, 5 m e e

e 11 5

!a ja Jk no 8

Snut Smul0084 e IMI 085G 5 "'"#

1. 9 00nu o A WAH A84 C' 88 '

Ptatt a:0 0

, [, ,,

On to

(

a v5,i 18 5

!. S.E JL 9

.0 0.&p r ',

S C 00t 0,0wn etSG 5 j s 40 COOL 00we

> $. $. On 10 456

-W~ Ob GPM , lite 5 stinD vi5

. VI 5 j( me a

l RCP 5 40 JL AWAllAlti  % '

C006 00wn 0m a / aC5

$0aCIO f t0w a p 3 msg 40

~

050 h0At4At OMA PAOCEDURI5

$ CAP e

st5

l. JL a smr quipCINCVilW III a j 20

.C5 P I wish  :

a OmA B 1f CAP 8

, 5 C g VIS at5fARI v55 8 ', j On P055 e '

37 l a me Put RC5 0m 8

'OECAV#1Al e DineDWA4 5 C00s00wN - C00t DowW CocithG ON 04 e

f CACIO CoAC maluaAs CIRC e

645PtCl&AIPAth 8 *Q *A3 ")

= D) 40 C 0 565 oo < :o

$ .C5 Ac5 me P.I with *3 4-

a Saa

1. 8 38 v CAP ... $' @o

, o m

. ,gg S vt5 e , '

S M e

i I

I I

TDd 400 l Rev. 3 Page 70 of 81 APPENJIX D 3 DiPLIfID EVE'iI TR2S G

G G

+ day. 1

• Rev. 2 d Rev. 3

. .. _. . - - _ _ ~ - . - . - - . . - . - . _ _ - - - _.

t TDA 605 Rev. 3 dage 71 o! S1 D.0 3 NP: I7IE3 EVENT TRIE The event tree on the following page shows pass Ibie combinations of circumstances that vera constderad that resultad in the guldelines presentsd in this TOR.

The guidel tnes explic* tly statad in section ' +, wnen incorporatad into a revised OraG Tube Leak / Rupture Emergency Procedure, will enhance tne espab tLity o f til-1 to deal with an OraG tuoe leak. The pur pos e o f this section is to describe' tha features of the revised procedura.

The discussion which follows assumas that the logic presented ay the flowchart depletsd in Appendix D is adopted for the revissi proc 3 dure, f

l 1

-e r

9 i- D-1

+ Rev. L

• Rev. 2 d Rev. 3 j

R 406 E ENT FIGURE D-1 Page 72 Of 81 NO.

RCP'S SIMPLIFIED OTSG '

CONDENSER EVENT TREE AVAuBLE 1 LEAKING OTSG C ONDENSER AV Al( ABLE ,

s NO RcP's 3

RCS SUSC00 LED CONDENSER AVAILA8LE g RCP'S AV Aet A8L E SOTH OTSG'S LEAK A80vE 1050 *St CONOENSER AV AILA SLE SELOW 6 1050 PSI NO RCP's A80VE 1050 PSI BELOW 7 1050 PSt TUBE RUPTURE 'CONOENSER AV AILA BLE g

RCP'S AVAILABLE 9

' 1 LEAKING ofSG A80VE CONDENSER 1050 PSI AV AIL ABLE SELOW to 1050 PS1 NO RCP's 480vt 1050 PSI SUSC00 LING 8kW

MARGIN 105o PSI LOST CONOENSER AV AILA BLE RCP'S AVAILABLE SOTH 13 OTSG's LEA <

A80VE

.CONOENSER 1050 PSt AVASLASLE 1050 PSI NO RCP's agoyg j 1050 PSI BELOW 15 MORE iogo CAPActTY REQua#EO 1

J

lu 605 Rev. 3 Paga 73 of 81 1

l 1

APPENOIX 3 PROCESS CCMPJTE,1 OJrPJ r

/

i

& Aav. L

• Rev. 2 i Rev. 3

TDA 606 Rev, a Page 74 of 81 E.0 + PROCE33 COMPJTER OJIPJT AtO ALARMS E.1 + Sco pe

+ rhe process computee will have the follosing Lnformation available

+ wt th alarms as noted:

+ Subcooling margLn F OT3G Tube to Shell Dif ferentLal Temperature E.L . L + Subcooling Margin Alarm

& 3ubcooling margin will be computed for each hot leg and the average of

+ the five highest incore thermocouples, tha process computer shouti

+ triggar an alarm state (f: I l

F 3 CM 2 52

• E . L . 2 F O T3 G rub e to Sh ell DL f f a r en t L a l Tempe r a tu r e d the process computar calculatas shell temparature as follows for each

+ OTS'G t f all shell thermocouples are operable:

+ r ,g e t t = 0. 242 f t + 0.176 r2 + 0.201 I3 + 0.14 3 r4 + 0. 238 t' 5

'E-l)

+ LLmLttng the alarm statt to condL tions when Teotd Ls 535 inhloits

+ the alarm during normal operations.

d 6 If the process computer is not available, the arithmatic averag2 of 6 the five thermocouples can be used. Expertenes from tha Fall 1983 6 cooldown esses demonstratad that an arithmette averaga approxLmatas d the welghtsd averaga of equations E-1 wLehin several dagrees (see 6 fable E-1 for a comparison of the seithmette and weighted averaga),

d If a thermocouple is unavailable, then the arlthmette average should 4 be calculated using the substitutions of Table E-2 for snell 4 ther mo cou pl es. Table E-3 provtdes the substltution for cold leg 4 temperatures (Lndt es ting tube temp 2rature).

d It is also possible that thermocouple temper 2tures are not 6 available in any form except from via translation from voltaga 6 r ea din gs. In this s'.tuation, one thermocouple from the upper and 4 lower downcomers should be used and the shell temperature d calculated as:

E Ishel l = . 6 fe+.4f u

• 4 wnere T. - T 1, T2 or f ,3 and d Tu=T4 Or T$

+ Rav. L h Rev. 2 l Rev. 3

TDA 406 Rev. 3

Page.75 of 81 f.

E I FABLE d-1 ,

Oos par is on o f k *. gh ta d vs . A e t thme t t: Av era ge ,

of 075G Shell Thernocouples L. 902/ hr Cooldown 4:.th hiW"

• 9/19/ 83) ',,

TI:4E 'aln)'

s WEIGHIE3 AVdRAGd AAlf1MEflC AVERAGd 0 515.9 5 16 . 8 45 5 16 .1 5 15 .1 85, _ 481.7 482.5 12 5 t' 438.3 439.8 165 -I 391.'3 393.6 195 _

3 82. 0 3 83. 4 378.8 232 376.2 2 90/ hhr. _C otdown us 'ng E!W (LO/ 2/ 83) f ME (mini WEIGHTE3 AVE 1 AGE ARIfINECIC AVERAGE O 52 4. 1 '

524.3 49 . '9f.9-

~ .t 310 '.' 7/3. ,

89 0 483.8 - 4 83. '"

l- 452.3 454.4

- 12 7 16 7 401.2 401.1 "

e 197 385.1 .n 385.1 25)f' <

3 97.4 .

400.o.

4 G- n' di ,

, l'. i f,?

J'

(

t

. t-

+ Rev. 1.

• Rev. 2 d.Jh%__D

l l

TD2 406 Rev. 3 dage 76 o! S1 4

Table E-3 'Jide Range Teotg Input Substt tu tions 2 C- P- L Teold A 3 C 0 A toop 3 loop 0 0 0 0 Avg A Avg 3 0 0 0 X Avg A T2 4 53 0 0 X 0 Avg A T2 2 58 0 0 X X Avg A Avg B 0 X 0 0 T24 M Avg B 0 X 0 X r2 4 5A T3 4 Sa 0 X ~K 0 T34 M TE 2 58 0 X X X TE 4 5A Avg B X 0 0 0 T32 M Avg B X 0 0 X IE 2 M T2 4 53 X 0 X 0 TE2 M TS 2 53 X 0 X X C2 2 5A Avg B X X 0 0 Avg A Avg B X X 0 X Avg A TE 4 38 X X X 0 Av3 A TE 2 53 X X X X Avg A Avg B 0= Pump Running X= Pump Off A vg A = ( TE 4 SA + IE ' 2. M) /2 Avg 3 - l TE 4 % + TE ' 2 5A)/ 2 r3 959 May be s ub stL tu te d for - rE 2 M

~

TE % t May be s ubs titu ted ' f or TZ 4 58

+ tet . t

• Rev. - 2 w

TD1 food Rev. 3 dage 77 of 81 h

Table E-2 Shell Thermocouple Subst'.tu t~ on Failed l Subs ti tu ta l T/c  ! r/c i i i T5 i Ts i i i T4 l T5 i

! i T3 i T2 j l i T2 i 0.5 'rt + T311 l l Ti i T2  :

i I r4 & T5 l No Cal: l 1 i f3 &T2 j Tt l 1 i f3 &It j T2  !

I i r2 &Il i T3 i i i ft & r2 & T3 i No Cal l

+ Wide Ranga T:old should be used in determining 3T3G tube to shell

+ differenttal temperatures. Normally, use the wida rangs input from

+ TE-1-5ASB and TE-3-5A&B, although IE 953 and TE 901 can be ussd in

- & c a r ta in cas es . Table E. L.4. 2 de fines the da ta sources ,

a

+ Foc each loop, Calculata Shall to Tube delta T as follows:

+ I I-S = Tshell ~ fcold

& Ir-s should trigger an alarm s ta te if

+ Icold 53 52

• an d I T-3 70l *

+ Rev. L -

' A - Rev. 2 4 Rev. 3~

)

TDR 605 Rev. 3 ,

Page 75 of S1 b

APPdNOIX 2 A00If DNAL 3 TEAMING AN) ISOLArDN CAITERIA FOR RE0UCIDN O F R AD D LOGICAL REIZAS E3 e

+ Rev. L

• Rev. 2 f Rev. 3

TDR 406 Rev. 3 Page 79 of 31 4 F.0 ADDITIONAL STEAMING AND ISOLATION CRITERIA FOR REDUCTION OF RADI0IDGICAL RELEASES d- The steaming and isolation guidelines provided in Section 5.2.7 are d intended for the control room operator. Their development required d the balancing of diverse factors including ease of use and d equipment protection. The latter consideration in itself had d safety implications, since protection of one piece of equipment, d namely the steam generators, impacts offsite releases.

d Once the various emergency support groups assemble, opportunities d develop for minimizing releases for the specific plant condition.

d While this specific' event treatment is not appropriate for the d operator to deal with in the short term, it is appropriate for the d emergency support groups. This appendix provides guidelines for d varying from the isolation and steaming criteria. The d recommendations in the following sections spring frcm the following d considerations:

d 1. Isolation of one or both OISG's reduces the RCS cooldown rate d which increases the time to reach cold shutdown and to terminate d the primary to secondary leak.

d 2. Isolation of one OTSG when both are leaking may increase the '

d integrated dose since the release will continue from the d unisolated OTSG for a longer period.

d 3. Isolation of both OTSG's requires feed and bleed cooling which d could result in releases of steam 'or steam and water directly d to the atmosphere.

/ d 4. An isolated OTSG may flood, after which it may not be possible d to unisolate and return the OTSG to service.

d 5. Isolation of direct steam releases to the atmosphere is expected d to reduce the offcite thyroid dose by a factor of at least .

d 6. Isolation for dose reduction should be based on measured dose d rate to preclude premature isolation.

4 Table F-1 summarizes the guidance provided in Section F-1 through 4 F-3.

d F.1 . 0TSG ISOLATION SHOULD BE AVOIDED IF d 1. RCP's are not available - natural circulation cooldown may not d be possible with one OTSG since flow in one loop might stagnate d and a bubble could' form in the hot leg as primary pressure is d reduced.

d 2. Both OTSGs leak but the difference in leak rate is less than 4 a factor of eight - otherwise, the delay in cooldown may negate d the dose reduction from isolating one OTSG.

+ Rev. 1-

• Rev. 2-c Rev. 3

TDR 406 Rev. 3 Page 80 0f 81 4 F.2 OTSG ISOLATION MAY BE DESIRABLE IF ,

l

'only one OTSG l d 1. RCP's are operating, the condenser is unavailable d leaking and iodine dose rates are high - in this situation, d high iodine release rates could be terminated by isolation of the d leaking OTSG.

d Although cooldown time is increased, radioactivity releases will d be terminated. RCP operation enables control of the RCS, which d in turn allows cooldown of the leaking OTSG.

4 F.3 OTSG ISOLATION CRITERIA SHOULD BE RE-EVALUATED IN THE FOLLOWING d SITUATIONS d 1. RCP's operating, condenser unavailable, both OTSGs leaking, d iodine dose rates are high - isolation of one OTSG may be desirable d if the leak rate in one OTSG is significantly (about 8) greater d than in the other. The reduced dose rate from isolation of one 4 OTSG must be weighed against the shorter cooldown time with d steaming of both OTSGs. ,

d 2. Condenser available - isolation of one or both OTSGs greatly 4 increases cooldown times and increasec risk of an inadvertent or d uncontrolled release. A decision to isolate earlier than required d by procedural guidelines should be based on measured dose rates d if possible. In the absence of fuel failures, actual releases d under such conditions are expected to be quite low.

d 3. Only one OTSC is leaking and BWST level is 21 feet - if the good

, d OTSG is not expected to leak because shell/ tube delta T is being d controlled, then isolation is not required. Recall that the BWST d level isolation criterion was based on both steam lines being d flooded. If only one OTSG may be flooded, then BWST depletion d could not occur until level reaches 15 ft.

d F.4 TEMPORARY SHORT TERM MEASURES TO REDUCE OR TERMINATE RELEASES d Releases can be temporarily terminated without initiating feed and bleed cooling by -

4 a. Terminating steaming to the condenser -and not steaming to d atmosphere, d b. Not more than 1 RCP should be run, d c. If natural circulation is lost, steam again, d If steam generator pressure reached 1000 psi, steam again.

4 Heatup rate is 100-170F'/hr.

d These steps may provide enoughLtime to return an RCP to operation, d restore a condenser 'to service, or initiate protective actions, while d delcying the initiation of feed and bleed cooling. -i

+ Rev. 1

• Rev. 2 _

TDR 406 Rav. 3 Page 81 of 81 TABLE F-1

SUMMARY

OF DOSE REDUCTION CDNSIDERATIONS ONE BOTH OTSGs BOTH OTSGs OTSG LEAKING MORE LEAKING RCP CDNDENSER LEAKING THAN 8:1 DIFFERENCE EQUALLY On Available Avoid Isolation Avoid Isolation Avoid Isolation

! (F.3.2) (F.3,2) (F.3.2, F.1.2)

On No t Consider Consider Isolation Avoid Isola, tion Available Isolation (F.1, F.3.1) Of One (F.1.2)

(F.1) 0FF Available Avoid Isolation Avoid Isolation Avoid Isolation

Of One OTSG Of One OTSG Of One (F.1) (F.1) (F.1.2,-F.3.2) 0FF Not ~ Avoid Isolation Avoid Isolation Avoid Isolation Available Of One OISG Of One OTSG Of One (F.1) (F.1, F.2) o NOTE

Condenser available means main condenser or steaming to auxiliary condenser via MFP.

/

I

+ Rev. 1 l

• Rev. 2 4 Rev. 3

1 l

GPU Nuclear Corporation NUCIMr Post Office Box 480 Route 441 South l

l Middletown, Pennsylvania 17057 0191 717 944-7621 TELEX 84 2386 Writer's Direct Dial Number:

May 17, 1984 5211-84-2093 Office of Nuclear Reactor Regulations Attn: John F. Stolz, Chief Operating Reactors Branch No. 4 U. S. Nuclear Regulatory Ommission Washington, D.C. 20555

Dear Mr. Stolz:

Three Mile Island Nuclear Station Unit I, (TMI-1)

Operating License No. DPR-50 Docket No. 50-289 TMI-l Steam Generator Tube Rupture Guidelines TDR-406 Rev.3 Attached for your information is the nost recent revision to the TMI-1 Steam Generator Tube Rupture Guidelines. This revision incorporates changes irvlimted in previous correspondence. It should be understood that this TDR is a living document and as such will be revised frtxn time to time to include more recent information. The NRC staff will be kept informed as significant revisions occur.

Sincerely, H. . Hukill, k

Director, TMI-l IDI/07S/mle

• Enclosures CC: R. Conte H. Silver 00 i

l i j

/

GPU Nuclear Corporation is a subsidiary of the General Public Utilities Corporation