ML20205F527
| ML20205F527 | |
| Person / Time | |
|---|---|
| Issue date: | 09/30/1988 |
| From: | Murphy E Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| REF-GTECI-A-03, REF-GTECI-A-04, REF-GTECI-A-05, REF-GTECI-SG, TASK-A-03, TASK-A-04, TASK-A-05, TASK-A-3, TASK-A-4, TASK-A-5, TASK-OR NUREG-0844, NUREG-844, NUDOCS 8810280082 | |
| Download: ML20205F527 (152) | |
Text
- -- - -
NRC Integrated Program for the Resolution of Unresolved Safety
) Issues A-3, A-4, and A-5 Regarding Steam Generator Tube Integrity 4
- Final Report l l l
i 1
U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation p' n %,
! [ t > ., (1 ' ^g
- a
%;;;,,/
l l i l C 10.7 % 000'/30
- 0044 f< l'I)li
NOTICE Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be available from one of the following sources: *
- 1. The NRC Public Document Room,1717 H Street, N.W.
Washington, DC 20555
- 2. The Superintendent of Documents, U.S. Government Printing Office, Post Office Box 37082, Washington, DC 20013 7082
- 3. The National Technical Information Service, Springfield, VA 22161
] Although the listing that follows represents the majority of documents cited in NRC publications, it is not intended to be exhaustive.
i F Referenced documents aveitable for inspection and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; NRC Offlee of Inspection and Enforcement bulletins, circulars, informa*lon notices, inspection and investigation notices; Licensee Event Reports; vendor reports and correspondence; Commission papers; and eglicant and licensee documents and correspondence.
The folbwing documents in the NUREG series are available for purchase from the GPO Sales Program: formal NRC staff and contractor reports, NRC sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of f Federal Regulations, and Nuclear Regulatory Commission Issuances, i Documents available from the National Technical Information Service include NUREG series !
, repons and technical reports prepared by other federal agencies and reports prepared by the Atomic !
Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
, Documents available from public and special technical libraries include all open literature items, such as books, journal and periodical articles, and transactions, federal Register notices, fJderal and
! state legislation, and congressional reports can u'vally be obtained from thete libraries,
, Documents such as theses, dissertations, foreign reports and tramlations, and non.NRC conference :
proceedings are available for purchase from the organization sponsoring the publication cited.
j S.ngle copies of NRC draf t reports are available free, to the extent of supply, upon written request t
to the Division of information Support Services, Distribution Section, U.S. Nuclear Regulatory Commission, Washington, DC 20555.
i Copies of industry codes and standards used in a substantive manner in the NRC re9 story process !
are maintained at the NRC Library, 7920 Norfdk Avenue, Bethesda, Marylanc'. t are available !
- there for reference use by the public. Codes and standards are usually copyhght and may be !
purchased from the originating organization or, if they are American National Standards, from tb1 r
- American National Standards Institute,1430 Broadway, New York, NY 10018. t
) '
. i l
1 4
i I !
NUREG-0844 NRC Integrated Program for the :
Resolution of Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Generator Tube Integrity l Final Report
' ' ' ' ' ~
__.._..n.. ~ . - -., -
ateYu c[SsptN tI198 Division of Engineering and System Technology !
Office of Nuclear Reactor Regulation !
U.S. Nuclear Regulatory Commission ,
Washington, DC 20555 l l
l p*"%,
hM6) .....
j l
l l
i l
j j
I
- ABSTRACT This report presents the results of the NRC integrated program for the resolu-tion of Unresolved Safety Issues (USIs) A-3, A-4, and A-5 regarding steam generator tube integrity. A generic risk assessment is provided and indicates that risk from steam generator tube rupture (SGTR) events is not a significant contributor to total risk at a given site, nor to the total risk to which the general public is routinely exposed. This finding is considered to be indf a-tive of the effectiveness of licensee programs and regulatory requirements for ensuring steam generator tube integrity in accordance with 10 CFR 50, Appen-dices A and B.
l This report also identifies a number of staff-recommended actions that the staff finds can further improve the effectiveness of licensee programs in ensuring
- the integrity of steam generator tubes and in mitigating the consequences of an SGTR. As part of the integrated program, the staff issued Generic Letter 85-02 encouraging licensees of pressurized water reactors (PWRs) to upgrade their i programs, as necessary, to meet the intent of the staff-recommended actions; i however, such actions do not constitute NRC requirements. In addition, this report describes a number of ongoing staff actions and studies involving steam generator issues which are being pursued to provide added assurance that risk ,
from SGTR events will continue to be small. l The staff will continue to monitor steam generator operating experiences as an indicator of the effectiveness of licensee programs for ensuring steau generator tube integrity. As has been true in the past, the staff may impose additional requirements (pursuant to applicable regulations) to continue to assure that licensees are implementing adequately effective programs where and if such '
action is determined to be necessary on the basis of operating experience or as a result of ongoing staff actions and studies.
The staff concludes that with final publication of this report, USIs A-3, A-4, i and A-5 are technically resolved.
I
[ r i
a a
i
,1 I
1 NUREC-0844 iii l
TABLE OF CONTENTS
.P_ age ABSTRACT .............................................. ............... iii LIST OF CONTRIBUTORS AND ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi ABBREVIATIONS ......................................................... xii 1 INTEGRATED PROGRAM OVERVIEW ...................................... 1-1 1.1 Background .................................................. 1-1 ,
- 1. 2 Development of Integrated Program ........................... 1-2
- 1.3 Scope of Value-Impact Evaluation ............................ 1-4 1.4 Industry Comments ........................................... 1-5 1.5 Risk from SGTR-Related Causes ............................... 1-6
- 1. 6 Disposition of Potential Industry Actions ................... 1-7 1.6.1 Staff Recommended Actions ............................ 1-7 1.6.2 Potential Industry Actions Warranting Further Staff Study .......................................... 1-8 i 1.6.3 Deleted Potential Industry Actions ................... 1-8 i
- 1. 7 Issuance of Generic Letter Regarding Staff Recommended Actions and Subsequent Follow-up Actions .................... 1-9
- 1.8 Staff Actions and Studies ................................... 1-10 1.9 Public Comments ............................................. 1-11 1.10 Implications of July 15, 1987 SGTR Event at North Anna Unit 1 ...................................................... 1-11 1.11 Conclusions Stemming from the Integrated Program ............ 1-12 2 VALUE-IMPACT EVALUATION OF POTENTIAL INDUSTRY ACTIONS ............ 2-1 2.1 Prevention and Detection of Loose Parts and Foreign Objects . 2-1
] 2.1.1 Secondary Side Visual Inspections and Improved ;
QA/QC Procedures ....................,................. 2-1 !
l 1 2.1.1.1 Staff Recommended Actions...................... 2-1 l
! 2.1.1.2 Basis for Initial Considaration .............. 2-2 l
{ 2.1.1. 3 Value-Impact ................'................. 2-3 l' 2.1.1.4 Conclusions .................................. 2-6 l
i 2.1.2 Loose-Parts Monitoring Systems ........................ 2-6 l 1 L
! 2.2 Inservice Inspection of Steam Generator Tubes................. 2-8 l I
2.2.1 Supplemental Tube Inspections ........................ 2-8 2.2.1.1 Potential Industry Action ................... 2-8 f' Basis for Initial Consideration .............
1 2.2.1.2 2-9
- I j !
NUREC 0844 y i 1
1 I
! L 1 l t
1 l
l CONTENTS (Continued) l l 1
Pa9e j l
1 2.2.1.3 Value-Impact ............................... 2-11 2.2.1.4 Conclusions.......... ....................... 2-14 >
2.2.1.5 Cost and ORE Value-Impact Assumptions. . . . . . . . 2 15 2.2.2 Full-length Tube Inspection........................... 2-17 l 2.2.2.1 Staff Recommended Action..................... 2-17 I 2.2.2.2 Basis for Initial Consideration.............. 2-18 2.2.2.3 Value-Impact................................. 2-18 2.2.2.4 Conclusions.................................. 2-19 2.2.3 Denting Inspections................................... 2-20 2.2.3.1 Potential Industry Action.................... 2-20 2.2.3.2 Basis for Initial Consideration.....!........ 2-20 2.2.3.3 Value-Impact................................. 2-20 2.2.3.4 Conclusions.................................. 2-21 2.2.4 Steam Generator Inservice Inspection Interval. . . . . . . . . 2-21 2.2.4.1 Staff Recommended Action..................... 2-21 2.2.4.2 Basis for Initial Consideration.............. 2-21 2.2.4.3 Value-Impact................................. 2-21 2.2.4.4 Conclusions.................................. 2-22 2.2.5 Inspections Following Shutdown for Repair of Leakage.. 2-22 2.2.5.1 Potential Industry Action.................... 2-22 2.2.5.2 Basis for Initial Consideration.............. 2-22 2.2.5.3 Value-Impact................................. 2-23 2.2.5.4 Conclusions.................................. 2-24 2.3 Improved Eddy-Current-Test Techniques....................... 2-24 2.3.1 Potential Indus try Action. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-24 2.3.2 Basis fer Initial Consideration....................... 2-25 2.3.3 Value-Impact.......................................... 2-25 2.3.4 Conclusions........................................... 2-26 2.4 Upper Inspection Ports....................................... 2-27 2.4.1 Potential Industry Action............................. 2-27 2.4.2 Basis for Initial Consideration....................... 2-27 2.4.3 Value-Impact.......................................... 2-28 2.4.4 Conclusions........................................... 2-29
- 2. 5 Seconda ry Water Chemi stry Program. . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-30 2.5.1 Staff Recommended Action.............................. 2-30 2.5.2 Basis for Initial Consideration....................... 2-30 NI' REG-0844 vi
CONTENTS (Continued)
Page 2.5.3 Value-Impact.......................................... 2-30 2.5 4 Conclusions........................................... 2-30
- 2. 6 Condenser Inservice Inspection Program....................... 2-31 2.6.1 Staff Recommended Action........................ ..... 2-31 2.6.2 Basi s for Initial Consideration. . . . . . . . . . . . . . . . . . . . . . . 2-31 2.6.3 Value-Impact.......................................... 2-32 2.6.4 Conclusions........................................... 2-35 l 2.7 Stabilization and Monitoring of Degraded Tubes............... 2-35 ,
! 2.7.1 Potential Industry Action............................. 2-35 l 2.7.2 Basis for Initial Consideration....................... 2-35 2.7.3 Value-Impact.......................................... 2-36 2.7.4 Conclusions........................................... 2-37 2.8 Primary-to-Secondary Leakage Limits.......................... 2 77 2.8.1 5taff Recommended Action.............................. 2-37 2.8.2 Basis for Initial Consideration....................... 2-37 2.8.3 Value-Impact.......................................... 2-38 2.8.4 Conclusions......................... ................. 2-39 2.9 Coolant Iodine Activity Limit................................ 2-39 l
2.9.1 Staff Recommended Action.............................. 2-39 l
2.9.2 Basis for Initial Consideration....................... 2-39 i l 2.9.3 Value-Impact.......................................... 2-40 l 2.9.4 Conclusions........................................... 2-42 l
l 2.10 Reactor Coolant System Pressure Contro1...................... 2-42 2.10.1 Potential Industry Action and Bases.................. 2-42 2.10.2 Value-Impact......................................... 2-42 2.10.3 Conclusions.......................................... 2'42 l l
2.11 Safety Injection Signal Reset................................ 2-43 ;
2.11.1 Staff Recommended Action............................. 2-43 2.11.2 Basis for Initial Consideration......... ............ 2-43 l 2.11.3 Value-Impact......................................... 2-43 2.11.4 Conclusions.......................................... 2-44 ;
2.12 Containment Isolation and Reset............................. 2-44 1 2.12.1 Potential Industry Action........................... 2-44 l 2.12.2 Basis for Initial Consideration..................... 2-44 l
i I
l NUREG 0844 vii t
1 l
l
_ _ - _ . . .~ _ _ - - . ._
i ^
4 CONTENTS (Continued)
Pans 2.12.3 Value-Impact........................................ 2-45 :
. 2.12.4 Conclusions......................................... 2-45 i
3
SUMMARY
OF RISK /NALYSES FOR STEAM GENERATOR TUBE RUPTURE (SGTR)
EVENTS.............................................................. 3-1 i 3.1 Single and Multiple SGTR Probabilities....................... 3-1 :
3.1.1 Initiating Event Probabilities........................ 3-1 3.1.2 Probability of SGTRs as Consequential Events.......... 3-2 -
3.1.2.1 Conditional Event Probabilities.............. 3-2 ;
3.1.2.2 Initiating Transients........................ 3-4 !
3.2 SGTR Events Challenging the Reactor Tri '
3
, Heat Removal Functions.................p and Decay
...................... 3-5 !
] 3. 3 SGTR Events Resulting From Loss-of-Coolant Accidents. .. .. . . . . 3-7 ;
- 3. 4 SGTR Events in Combination With Loss of Secondary Systee ,
1 Integrity or failure To Achieve Steam Generator .
?
Isolation.................................................... 3-9 l 3.4.1 SGTRs and Total loss of Secondary Integrity........... 3-10 '
j 3.4.2 SGTRs Occurring in con i SG Safety Va1ve. . . . ................................
. .. junction With a Stuck-Open 3-13 !
j
- 3.4.3 SGTRs and MSIV 3.4.4 Failures............................... 3-14 Tube Fuptures Affecting Multi
i Generators...................ple.........................
Steam 3-15 ;
j 3.4.5 RWST Depletion Time Calculations...................... 3-16 !
! 3.4.6 Event Sequences....................................... 3-18 !
i j 3. 5 Core-Melt Sequences.......................................... 3-23 i
3.5.1 Determination of Radionuclide Releases................ 3-23 !
! 3.5.2 Calculation of Consequences and Risks................. 3-25 :
l l 3. 6 Non-Core-Melt Sequences...................................... 3-25 f i
3.7 Conclusions.................................................. 3-30 (
)
4 NRC STAFF ACTIONS AND COMPLETED ITEMS............................. 4-1 l 1
4.1 Introduction................................................. 4-1 b
I i l i
! i 1 l l
i
! i 1 i j NUREG-0844 viii !
1 i
i I
i
CONTENTS (Continued)
Pale 4.2 Steam Generator Integrity.................................... 4-5 4.2.1 Steam Generator Tube S1eeves.......................... 4-5 4.2.2 Inservice Inspection Program for Denting.............. 4-6 4.2.3 Improved Eddy-Current Techniques...................... 4-8 4.2.4 Category C-2 Inservice Inspection Requirements........ 4-8 4.3 Plant Systems Response....................................... 4-9 4.3.1 Steam Generator Overfi11.............................. 4-9 4.3.2 Reactor Coolant System Pressure Control During an SGTR.................................................. 4-10 4.3.3 Pressurized Thermal Shock ............................ 4-11 4.3.4 Improved Accident Monitoring........... .............. 4-12 4.3.5 Reactor Vessel Inventory Heasurement.................. 4-12 4.4 Human Factors Considerations ..................... .......... 4-13 4.4.1 Reactor Coolant Pump Trip............................. 4-13 4.4.2 Control Room Design Review............................ 4-14 4.4.3 Emergency Operating Procedures Improvement............ 4-16 4.5 sad e'agical Consequences.................................... 4-22 4.5 1 Reassessment of Radiological Consequences Following a Postulated SGTR Event .............................. 4-22 4.5.2 Reevaluation of Design-Basis SGTR..................... 4-23 4.5.3 Secondary-System Isolation............................ 4-24 4.6 Organizational Response...................................... 4-24 4.6.1 Operations Center Ccmmunications and Notifications.... 4-24 4.6.2 Interaction Between Regional Base Teams and the Executive Team ....................................... 4-25 4.6.3 NRC Site Team--Location of Site Team Components and Public Affairs Information Flow.................... .. 4-25 4.6.4 Familiarization With NRC Response Plan...... ......... 4-26 4.6.5 Alternate Evacuation Routes and Sites................. 4-27 4.6.6 Deescalation of Emergency Classification.............. 4-27 4.6.7 Offsite Dose Assessments.. ........................... 4-28 LIST OF TABl.ES i 1 Disposition of pottntial industry actions ........................ 1-14 2 Related sta'f actions and f.tudies................................. 1-15 3 Summary of value-lenpact evalue. tion................................ 2-46 1
dUREG-0844 ix
4 i
I CONTENTS (Continued)
LIST OF TABLES (Continued) 4 P_a21 :
L r
4A Sumary of systems response to SGTRs with total loss of secondary ;
integrity (B&W, THI-1)............................................ 3-17 t
4B Sumary of systems response to SGTRs with total loss of secondary !
integrity (CE, Calvert Cliffs).................................... 3-17 r I
- 4C Sumary of systems response to SGTRs with total loss of secondary ;
j integrity (Westinghouse high heaa, Zion).......................... J-17 i 5 SGTRs and stuck-open SG safety valves (Zion)...................... 3-18 I
- 6 Sumary of systems response to single and multi tube ruptures with failure of the HSIV.......................ple ................... 3-Il j i
7 Summary of probabilities, consequences, and risks for SGTR events :
j leading to core me1t.............................................. 3-24 ?
- 8 Range of parameters considered.................................... 3-26
{
t 9 Consemuences of non-core-melt. SGTR sequences involving a stuck-j oper .afety va1ve................................................. 3-29
[
I 10 Sumary of probabilities and risks for non-core-melt SGTR '
sequences involving a stuck-open safety va1ve..................... 3-30 11 NRC staff actions and completed items............................. 4-2
)
1 LIST OF FIGURES Pa21 1 Maximum cladding temperatures obtained for cases with tube ruptures !
initiated at the start of refill and at the start of reflood..... 3-8 I t
2 Primary pressure decrease for MSLB with concurrent SGTRs......... 3-11 i
i 3 HPI flow rate for MSLB with concurrent SGTRs..................... 3-11 j 4 Tube rupture flow rate for MSLB with concurrent SGTRs............ 3-12 f 2
l 5 Pressurizer water level for MSLB with concurrent SGTRs............ 3-12 !
I APPENDICES 4
j A REFERENCES i '
I
! B EVALVATION OF SGTR EVENTS FOR PRIOR PERIODS OF VULNERABILITY TO l j RUPTURE UNDER POSTULATED MSLB ACCIDENT
{,
i !
j NUREG-0844 x
!s 4
i
) #
LIST OF CONTRIBUTERS AND ACKNOWLEDGEMENTS A. Akstulewitz S. Bryan ,
A. Busiik .
L. Frank M. Hawkins ;
G. Holahan l T. Ippolito T. Marsh R. Martin F C. McCracken ;
J. Mitchell ;
E. Murphy j P. Norian L. Phillips !
R. Riggs l R. Serbu !
J. Strosnider {
K. Wichman The authora wish to thank Electronic Composition Services for typing the many drafts.
I l
l l l
NUREG-0844 xi l
l l
l l
- , _ . ~ _ - _ . _ , . . _ . _ _ _ . . . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ , _ _ . _ _ _ . _ _ _ _ _ _
r ABBREVIATIONS ACRS Advisory Committee on Reactor Safeguards ,
ADV atmospheric dump valve AFW auxiliary feedwater AIF Atomic Irdustrial Forum, Inc.
AIT Augmented Inspection Team ALARA as low as reasonably achievable ASME American Society af Mechanical Engineers ATWS anticipated transients without scram AVT all-volatile treatment BAST boric acid storage tank B&PV boiler and pressure vessel B&W Babcock & Wilcox CE Combustion Engineering CFR Code of Federal Regulations CI containment isolation CISIP condenser inservice inspection program DHR decay heat removal DNB departure from nucleate boiling DST Division of Safety Technology EC eddy current ECCS emergency core cooling system ECT eddy current test ENS emergency notification system EOF emergency operations facility E0P emergency operating procedure EPRI Electric Power Research Institute FPL Florida Power and Light FSAR Final Safety Analysis Report GDC general design criterion GIMCS Generic Issue Management Control System gpd gallons per day gpm gallons per minute J
HHSIP high head safety injection pump HPI high pressure injection '
HPSJP high pressure safety-injection pump ID inside diameter l IC Office of Inspection and Enforcement INEL Idaho National Engineering Laboratory INP0 Institute of Nuclear Power Operations IRC Incident Response Center ISI inservice inspection LANL Los Alamos National Laboratory j LOCA loss-of-coolant accident LOOP loss of offsite power i LPMS loose parts monitoring system ,
LPSI low pressure safety injection i
MFLB main feedline break :
MPA multi plant action i
xiii !
l l
ABBREVIATIONS (Cont.)
MSIV main steam isolation valve MSLB main steam line break NDE nondestructive examination '
NSSS nuclear steam supply system OD outside diameter OL operating license ORE occupational radiological exposure OTSG once-through steam generator PASNY Power Authority of the State of New York PCI pellet-cladding interaction a PORV power-operated relief valve PTS pressurized thermal shock PWR pressurized water reacto" QA quality assurance l QC quality control
! RCP reactor coolant pumps j RCS reactor coolant system RES Office on Nuclear Regulatory Research RPS reactor protection system RG regulatory guide RME Rochester Gas and Electric Company RHR residual heat removal RPAM Regional Public Affairs Manager i i RWST refueling water storage tank <
RY reactor year SAI Science Applications, Inc.
SG steam generator SGGP/ Steam Generator Group Project /
SGTIP Steam Generator Tube Integrity Program SGOG Steam Generator Owners Group :
SGTR steam generator tube rupture !
! SI safety injection SIS safety injection system SMUD Sacramento Municipal Utility District SRP Standard Review Plan
! STS Standard Technical Specifications I
SV safety valve [
SWCP secondary water chemistry program
- TAP Task Action Plan ,
l
~l TD AFP turbine-driven auxiliary feedwater pump ;
TMI Three Mile Island TSC technical support center TV television i TVA Tennessee Valley Authority
- USI unresolved safety issue
W Westinghouse 90G Westinghouse Owners Group i WPSC Wisconsin Public Service Corporation i i
NUREG-0844 xiv l
l I
NRC INTEGRATED PROGRAM FOR THE RESOLUTION OF UNRESOLVED SAFETY ISSUES A-3, A-4, AND A-5 REGARDING STEAM GENERATOR TUBE INTEGRITY 1 INTEGRATED PROGRAM OVERVIEW
1.1 Background
Degradation of steam generators (SGs) manufactured by each of the three pres-surized water reactor (PWR) vendors has resulted from a combination of problems related to steam generator mechanical design, materials selection, fabrication techniques, and secondary system design and operation. To date, many different forms of steam generator tube degradation h&ve been identified including: stress corrosion cracking, wastage, intergranular attack, denting, erosion-corrosion, fatigue cracking, pitting, fretting, support plate degradation, and mechanical damage resulting from impingement of foreign objects or loose parts on steam generator internal components. One or more of these forms of degradation have i
affected at least 40 operating PWRs and have resulted in extensive SG inspections, tube plugging, repair, or replacement. A detailed description of steam generator i '
tube operating experience was provided in NUREG-0886 and, more recently, in NUREG-1063.
The majority of the SG tube failures that have occurred under narmal operating conditions were small stable leaks; some required plant shutdown, inspection, and corrective actions, but others were small enough (e.g., below the leak rate limit of the Technical Specifications) so that plant operation continued until a scheduled shutdown. However, four significant steam generator tube rupture (SGTR) events have occurred in domestic PWRs since 1975.* These events occurred on February 26, 1975, at Point Beach Unit 1; September 15, 1976, at Surry Unit 2; October 2, 1979, at Prairie Island Unit 1; and January 25, 1982, at R.E. Ginna.
SGTR events are defined by the NRC staff to be a primary to sero.idary leak in
- excess of the normal charging flow capacity of the reactor coolant system
- (NUREG-0651). l
! The first three of these events were evaluated in NUREG-0651. That report includes evaluations of systems response, operator actions, and radiological consequences during the three events. The event at the Ginna plant was addressed in NUREG-0909 and plant restart was evaluated in NUREG-0916. NUREG-0909 includes i descriptions of the event and significant staff findings; NUREG-0916 is an f evaluation of system responso, operator response, steam generator inspection and repair programs, emergency preparedness, and radiological consequences.
t Staff concerns which were raised relative to steam generator tube degradation i stem from the fact that the steam generator tubes are a part of the reactor coolant system (RCS) boundary and that tube failures result in a loss of primary coolant. In addition, the steam generator tubes constitute a particularly l
l
- This report was prepared before a fifth SGTR event that occurred on July 15, ;
1987, at North Anna Unit 1. The 19plications of this event with respect to the conclusions of this report are addressed in Section 1.10. l NUREG-0844 1-1 ;
i
1 important part of the RCS boundary since their failure allows primar/ coolant ,
into the steam generators where its isolation from the environment is not fully l ensured. The leakage of primary coolant into the secondary system has two major safety implications. The first is the potential for direct release of radio-active fission products to the environment, and the second is the loss of cool-ing water which is needed to prevent core damage. An extended uncontrolleo loss of coolant outside of containment would result in the depletion of the initial RCS inventory and emergency core cooling system (ECCS) water withcut the capability to recirculate the water.
l The NRC regulations (Title 10 of the Code of Federal Regulations (10 CFR))
establish the fundamental requirements relative to steam generator tube inte-grity. Specifically, the General Design Criteria (GDC) of 10 CFR Part J0, Appendix A, state that the reactor coolant system boundary shall "have an ex-tremely low probability of abnormal leakage" (GDC 14), shall "be designed with sufficient margin" (GDC 15 and 31), shall be of "the highest quality standards practical" (GDC 30), and shall be designed to permit "periodic inspection and testing...to assess... structural and leak tight integrity" (GDC-32).
10 CFR Part 50, Appendix B, is also pertinent to the maintenance of steam generator tube integrity. This appendix establishes quality assurance require-ments for the design, construction, and operation of safety-related components.
The pertinent requirements of this appendix apply to all activities affecting the safety-related functions of these components; these include, in part, inspecting, testing, operating, and maintaining.
- To ensure that the regulations are met, each applicant's steam generator design, 1 water chemistry, and inspection program are evaluated in accordance with i regulatory guidance in NUREG-0800, "Standard Review Plan," before issuing a i
) license and a safety evaluation report is issued. PWR applicants are also i
required to analyze the consequences of a d'. sign 4 asis steam generator tube 1
rupture. These analyses must show that the offsite radiological consequences, r
l considering the most limiting set of initial conditions and single failure, do l' j not exceed a small fraction of the limits of 10 CFR Part 100.
{ Once a plant is in operation, the licensee demonstrates continued compliance i with the regulations through periodic inspections, using the inspection criteria
,' in the plant Technical Specifications. Tubes which are found to be defective, as defined in the plant Technical Specifications, are repaired or removed from !
- service. To verify acceptable performance during operating periods between i
inspections, the plant Technical Specifications require shutdown if excessive .
primary to secondary leakage or excessive primary side activity occurs. Opera-j ting plants which experience primary to secondary leakage in excess of Technical i 1
Specification limits, or abnormal degradation as evidenced by the periodic ;
inspections, may be required on a case-by-case basis to implement additional mea- ;
i sures to provide added assurance of continued compliance with the regulations.
The additional measures may include items such as increased inspection frequen-l cies, lower Technical Specification limits for primary to secondary leakage, or l 1 improved water chemistry / corrosion control. '
1 j 1.2 Development of Integrated Program 3
Steam generator tube integrity was designated an unresolved safety issue (USI) f in 1978, and Task Action Plans (TAPS) A-3 A-4, and A-5 were established to i
NUREG-0844 1-2 i l
l l
4 evaluate the safety significance of
- gradation in Westinghouse, Combustion Engineering, and Babcock & Wilcox steam generators, respectively. These studies were later combined into one effort because many problems being experienced with steam generators supplied by these vendors were similar. The staff pre-pared a draf t report regarding this issue, which was originally intended for publication as NUREG-0844, "Resolution of Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Generator Tube Integrity." The draft report primarily considered corrosion-related failure mechenisms, including the "denting" mechan-1 ism, since those failures were the main concern during the period when most of
}
the technical studies were performed, s
I In May 1982, subsequent to the issuance of NUREG-0909 and NUREG-0916 regarding ;
) the Ginna SGTR event, the staff initiated an integrated program to consider the l lessons learned from the Ginna SGTR event and from the three previous domestic .
SGTR events (NUREG-0651), and to consider the recommendations identified in Section 9 of the draft USI report above. The draft USI report (now entitled "Initial Staff Pecommendations from the NRC Program for the Resolution of Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Generator Tube ,
Integrity") is available publicly as an enclosure to an NRC memorandum, ;
SECY 84-13, dated January 11, 1984. Theobjectiveoftheintegratedprogram ;
was to complete resolution of USIs A-3, A-4, and A-5, including identification ,
l of requirements that should be imposed on operating license (0L) applicants and licensees and identification of further efforts that should be undertaken by l
- the NRC. The program report herein describes the results of this integrated program and supersedes the draft USI report above as the technical resolution ,
l of these USIs. !
I j The staff initially considered a number of issues pertaining to steam generator j tube integrity and mitigation of SGTR events as part of the integrated program.
These issues were categorized as (1) those appropriate for consideration by l the staff as potential actions for all plants with steam generators (hereafter l referred to as potential industry actions), or (2) those issues warranting further staff study, already being considered as part of an existing regulatory q
, program, or those issues for which corrective actions have already been taken, i Issues within the latter three catec ries are discussed in Section 4 of this report.
l The issues within the first category, which had been identified as potential f j industry actions, were grouped into the technical areas listed below and were <
j subjected to a value-impact evaluation by the staff. Section 2 of this report l l
provides a description of each potential industry action, the initial basis i for its consideration, and the results of the value-impact analysis for each ;
item. These issues are:
l (1) Prevention and Detection of Loose Parts or Foreign Objects I (a) Secondary Side Visual Inspection and Improved Quality Assurance /
Quality Control (QA/QC) Procedures i l (b) Loose-Parts Monitoring System (2) Inservice Inspection of Steam Generator Tubes
]
) (a) Supplemental Tube Inspecti 's (b) Full-Length Tube Inspection.
NUREG-0844 1-3 i
1 4
l (c) Denting Inspections (d) Steam Generator Inservice Inspection Interval (e) Inspections Following Shutdown for Repair of Leakage (3) Improved Eddy-Current Test Techniques !
(4) Upper Inspection Ports l (5) Secondary Water Chemistry Program ,
(6) Condenser Inservice Inspection Program
! (7) Stabilization and Monitoring of Degraded Tubes 4
- (8) Primary to Secondary Leakage Limits (9) Coolant Iodine Activity Limit (10) Reactor Coolant System Pressure Control (11) Safety Injection Signal Reset (12) Containment Isolation and Reset 1.3 Scope of Value-Impact Evaluation i
The value-impact analysis of the potential industry actions, discussed in .
Section 2, addressed the potential for reductions in (1) the probability of core f i melt, (2) the probability of significant radiological releases comparable to ,
1 NUREG-75/014* PWR release categories 8 and 9 during SGTR events not leading to
- core melt, (3) public risk, and (4) occupational radiological exposure (ORE).
l Economic costs were also a consideration of this analysis. Het economic costs '
were considered in relation to the benefits to public health and safety. Net j economic benefits (deriving from the cost effectiveness of some of the poten-tial industry actions) were considered neither as a barrier nor a basis for the disposition of the potential industry actions under consideration.
1 The staff's evaluation considered input from a report by the staff's contractor, 1 Science Applications, Inc. (SAI), "Value-Impact At.alysis of Recommendations Con-l l cerning Steam Generator Tube Degradation and Rupture Events." The SAI report '
]
provided much of the basis for the staff's evaluation of the effectiveness of i
' the potential industry actions in reducing the incidence of steam generator tube i degradation, tube ruptures, and occupational radiological exposures and for the i
- staff's evaluation of the potential cost benefits and cost impacts.
The SAI report hcludes an assessment of the risk reduction potential associated <
j with prevention of SGTR events involving a single tube. The staff reviewed this
(
analysis and performed an independent analysis of the probability of various :
! scenarios involving ruptured tubes as described in Section 3. The staff's risk j analysis included consideration of risk from scenarios involving more than a 1
single ruptured tube.
l l
, *Formerly WASH-1400.
NUREG-0844 1-4 ;
]
i e
i I
The staff's value-impact analysis also considered industry comments as discussed in Section 1.4 below.
i 1.4 Industry Comments On December 9, 1982, the NRC staff transmitted Generic Letter 82-32, "Potential Steam Generator Related Generic Requirements," to all pressurized water reactor plant licensees for their information. This letter transmitted a draft value-impact report by SAI, and advised that any comments that the recipients chose to make on that report or on the probability and consequences of multiple tube ruptures would be considered in the staff's final value-impact report if the
- omments were provided in a timely manner. The SAI report was also sent to the nuclear steam supply system (NSSS) vendors: Westinghouse, Combustion Engineer-ing, and Babcock & Wilcox. The staff also invited industry comments on the program during two public meetings (July 29 1982 and July 5, 1983) with the Steam Generator Owners Group (SG0G), which Is sponsored by the Electric Power Research Institute (EPRI), and other interested parties.
Letters received from industry in response to the staff's initiatives above are listed in Appendix A. Industry comments specific to the potential industry actions were considered by the staff and are discussed in Section 2. General comments included the following:
(1) General agreement exists throughout industry that single SGTR events are not dominant contributors to risk.
(2) Westinghouse (letter, Feb. 15,1983) stated that its analyses have also demonstrated that multiple tube ruoture events are not a dominant contributor to risk.
(3) Mast respondents appear to be in agreement that many of the potential industry actions under consideration by the NRC staff involve the type of actions which should generally be followed by industry. Most often cited i by industry were actions pertaining to the prevention and detection of loose parts and foreign objects, full length tube inspections, maximum steam generator inspection intervals, and hiqh quality secondary watser chemistry. Specific comments by industry concerning these actions are discussed in Section 2.
(4) Some respondents question the need for regulatory action in these areas on the grounds that the benefits produced are primarily economic rather than safety related.
(5) Several respondents emphasized the need for ensuring that any requirements i are sufficiently flexible to accommodate plant-specific issues and differences in steam generator design.
(6) SG0G comments included the following (letters, September 30, 1982,
, April 24, 19( , and October 1, 1984):
1 (a) Steam generators and PWR plants were designed with the expectation that tubes in ste!1 generators would leak. PWRs have been designed and analyzed to accommodate such ruptura.
NUREG-0844 1-5
(b) Complete elimination of the potential for tube leakage or rupture is not a realistic or nemsary objective.
(c) Neither the utilities, the NRC, nor the public is well served by the uniform application of generic requirements to solve problems with complex site-specific causes. This is particularly true where the consequences of problems are economic rather than safety issues.
(d) Concern for steam generator reliability led the SG0G to voluntarily prepare secondory water chemistry guidelines based on extensive research rosults. The SGOG understands that non-member utilities also ara following the guidelines out of the same concern. With such evidence of utility action, imposition of the guidelines as an NRC requirement is unnecessary and would be counterproductive; i.e., it would be another indication that aggressive voluntary action by the industry does not prevent regulatior,with all its associated "red tape."
(e) The SGOG strongly endorses the position of the Advisory Committee on Reactor Safeguards (ACRS) documented in an ACRS letter to NRC Chairman Nunzio Palladino on October 18, 1983. This position is that the NRC proposals have considerable merit but that their implementation will not result in a significant reduction of risk to health and safety of the public. Rather, their implementation would result in a reduced rate ot' challenge to the safety systems and in economic benefits to the consumers of electric power. Therefore, the ACRS recommended ;
that the NRC proposals not be implemented as new regulatory require- i
- ments but instead as recommended industry actions.
T (f) Very substantfal progress has been made and :s continuing to be made by the PWR industry in the United States in improving the reliability of steam generator operation and USIs A-3, A-4, and A-5 can be considered resolved.
- 1. 5 Risk from SGTR-Related Causes l
The staff's risk analysis, which is described in Section 3, indicates that the core melt probability from SGTR-related causes is small, about 5.3 x 10 8/
reactor year (RY) for Babcock and Wilcox (B&W) plants and E 9 x 10 ?/9Y for j Westinghouse (W) and Combustion Engineering (CE) plants. These probabilities are a relativeTy small fraction (10% or less) of the overall probability of core l melt events from all causes based on probabilistic risk assessments that have been performed for a number of PWRs. The corresponding risk to the public is ;
estimated to be limited to 2.4 x 10 3 (B&W plants) and 1.7 x 10 3 (W and CE plants) latent fatalities /RY and 4.6 x 10 6 (B&W plants) and 4.3 x 10 6 (W and CE plants) early fatalities /RY from SGTR related accidents associated witii core melt (see Table 7, in Section 3).
Initiating event SGTRs, whose frequency is known from operating experience, contribute only 1.1 x 10 6/RY to the above core melt probability estimates. The balance of the above core melt probability estimates is associated with conse-quential event SGTRs whose frequency is more uncertain, particularly consequen-tial SGTRs involving multiple tube ruptures. The staff has made a number of i conservative assumptions to ensure that actual multiple tube frequencies are NUREG-0844 1-6 t
not grossly underestimated. However, this may have resulted in conservative estimates of the frequency of multiple tube ruptures and the corresponding probability of core melt.
4 The staff also evaluated the potential for significant radiological releases (comparable to NUREG-75/014* PWR release categories 8 and 9) during SGTR events
- not leading tu core melt. These events primarily involve SGTRs occurring in conjunction with a stuck-open steam generator safety valve so that primary coolant leaking into the steam generator can subsequently be released to the environment. The probability of SGTRs occurring in conjunction with a stuck-open safety valve is estimated to be 4.9 x 10 4/RY for B&W plants and 2.3 x 10 4/
RY for W and CE plants. Site-boundary doses would typically be expected to be small relative to the consequence limits of 10 CFR Part 100 based on best estimate assumptions regarding coolant iodine activity concentrations, iodine spiking, and meteorology. The use of conservative assumptions regarding these parame-ters, which are similar to design basis assumptions, leads to site-boundary dose estimates which may challenge or exceed the 10 CFR Part 100 limits; however, the frequency of such doses is significantly less than the above 2.3 x 10 4 to 4.9 x 10 4/RY estimate. The staff estimates public risk from non-core-melt releases to be very small: 7.0 x 10 7 latent fatalities /RY for B&W plants and l 3.0 x 10 7 latent fatalities /RY for W and CE plants.
On the basis of the evaluation above, the staff finds that SGTR events beyond the design basis do not contribute a significant fraction of the early and latent cancer fatality risks associated with reactor events at a given site. Further-more, the risk assessment indicates that the increment in risk associated with ,
SGTR events is a small fraction of the accidental and latent cancer fatality risks to which the general public is routinely exposed. These findings reflect in part the effectiveness of existing steam generator related requirements (see Section 1.1) for ensuring steam generator tube integrity. These include requirements which have been imposed on a case-specific basis in response to certain tube degradation problems experienced at one or more sites. However, these findings also reflect industry efforts since the mid-1970s to improve steam generator reliability. Steam generator reliability involves not just minimizing the potential for SGTRs, but also minimizing (1) the incidence of small tube leaks which may lead to unscheduled plan +. outages, (2) the need for j extensive steam generator repairs, and (3) the need for steam generator replace-i ment. Reliability improvements, therefore, have provided significant economic j benefits to the industry in addition to enhanced tube integrity.
1.6 Disposition of Potential Industry Actions
! 1.6.1 Staff Recommended Actions In view of the relatively low risk estimates associated with SGTR events, the staff has concluded that issuance of the potential industry actions in Table 1 l
- as generic requirements is not warranted at this time. However, the staff's ,
1 l *Formerly WASH-1400
! NUREG-0844 1-7
value-impact evaluation indicates that several of these potential industry actions as a group are effective and u st beneficial measures for significantly reducing (1) the incidence of tube degradation, (2) the frequency of tihe rup-tures and the corresponding potential for significant non-core melt releases, and (3) occupational exposures, and are consistent with good operating and engi-neering practice. As a group, these actions are also effective measures for mitigating the consequences of SGTRs. Adoption of these actions by licensees would further reduce public risk and provide added assurance that risk will continue to be small. These actions have been designated as staff recommended actions and involve the following topics:
(1) secondary side visual inspections and improved QA/QC procedures for prevention and detection of loose parts (2) inservice inspection of steam generator tubes: (a) full length tube inspections, (b) maximum steam generator inspection intervals (3) secondary water chemistry program (4) condenser inservice inspections (5) primary to secondary leakage limits (6) coolant iodine activity limits (7) evaluation of safety injection reset Details of the staff recommended actions and the staff's value-impact evaluation of these actions are provided in Section 2.
1.5.2 Potential Industry Actions Warranting Further Staff Study Apart from potential industry actions which the staff has dispositioned as staff recommended actions, the staff has concluded that others of these actions merit further study by the staf f (staf f actions). As indicated in Table 1, these actions involve issues relating to supplemental tube inspections, inspecting dents in tubes, improving eddy-current inspection techniques for steam generator tubes, and controlling pressure in reactor coolant systems during steam genera-tor rupture events. These potential actions and the bases for the disposition of these actions as staff actions are discussed in Section 2. In addition, these actions have been added to the total list of staff action items stemming from the integrated program for resolution of USIs A-3, A-4, and A-5, which is described in Section 1.8 below, and in additional detail in Section 4.
1.6.3 Deleted Potential Industry Actions The staff has concluded that the remainder of the potential industry actions identified in Table 1 are not appropriate as generic staff recommended actions, nor do these actions warrant additional study as a staff action. The potential industry actions which have been deleted involve the areas identified below.
I These actions and the bases for the staff's disposition of these actions are I discussed in further detail in Section 2. As appropriate, the staff may review l individual plants relative to these issues. These issues are 1
NUREG-0844 1-8
l (1) prevention and detection of loose parts (loose parts monitoring systems)
(2) inservice inspection program (inservice inspcctiont following shutdown for repair of leakage)
(3) upper inspection ports t
4 (4) stabilization and monitoring of degraded tubes (5) containment isolation and reset 1.7 Issuance of Generic Letter Regarding Staff Recommended Actions and Subsequent Follow-up Actions __
As part of the technical resolution of USIs A-3, A-4, and A-5, the staff issued NRC Generic Letter 85-02 to all PWR licensees and applicants to inform them of the staff recommended actions. In addition, PWR licensees and applicants were requested to describe their overall steam generator programs and how these
); programs compare with the staff recommended actions. ,
Licensee and applicant letters responding to Generic Letter 85-02 are listed in Appendix A. The staff's assessment of these responses was reported to the Commission in an NRC memorandum, SECY 86-97, dated March 24, 1986. The staff .
concluded on the basis of this assessment that the large mejority of licensees !
and applicants are following programs, practices, and/or procedures which are partially to fully consistent with or equivalent to the staff recommended actions.
Cases where licensee programs appeared to fall short of being consistent with i or equivalent to the staff's recommended actions were not considered indicative, in and of themselves, of significant risk or non-compliance with the regulations, and with one exception noted below for Oconee Units 2 and 3, and Arkansas Nu lear ;
i One, Unit 1 (ANO-1), none appeared to warrant regulatory action at this time. .
- It was stated in SECY 86-97 that the staff would inform licensees of its findings !
, relative to their plants as part of Multi-Plant Action (MPA) C-16. Because l this task has such low safety significance, the staff has dropped these plans.
It is implicit in the findings stemming from the integrated program herein that l licensees must continue to be vigilant against new or unusual problems which J may necessitate preventive, diagnostic, and/or corrective actions beyond the !
! licensee's normal practice. The staff will continue to monitor steam genera- !
tor operating experiences as an indicator of the effectiveness of licensee f
< programs. As has been the case in the past, the staff may impose additional !
requirements on a site-specific or generic basis if such action is determined i to be necessary on the basis of operating experience or the results of ongoing 1 staff programs discussed in Section 1.8 and in Section 4 of this report to i provide continued assurance of steam generator tube integrity in accordance i with 10 CFR Part 50, Appendices A and B. Any new requirements would be subject to the provisions of 10 CFR 50.55a, paragraph (g)(6)(ii), or 10 CFR 50.109, as ;
- applicable.
In SECY 86-97, the staff indicated its concern that the existing Technical l j Specification limit on allowable primary to secondary leakaCe at Oconee Units 2 and 3 may not be a sufficiently effective limit for preventing tube ruptures. ,
I i NUREG-0844 1-9 l
1 Since issuing SECY 86-97, the staff learned that this concern was also applic-
! able to ANO-1. Each of these three units has operating procedures which i
incorporate leakage limits that appear to be largely consistent with those recommended by the staff in Generic Letter 85-02. However, because of the i importance of primary to secondary leak rate limits in ensuring tube integrity, the staff believes that appropriate limits should also be included in the Technical Specifications for these units. At the staff's request, the licensees
- for these units have recently submitted proposed Technical Specification limits which are under review by the staff.
Finally, the staff acknowledges that the industry has made significant progress i in recent years in improving steam generator reliability. Industry-sponsored
. research by SG0G and EPRI has resulted in a number of improvements in steam l generator and secondary system design and in the availability to utilities of :
improved operating practices, non-destructive examination (NDE) methods, and I
( preventive and corrective measures pertaining to specific problems. Noteworthy achievements in this regard have included issuance of the SGOG/EPRI "PWR Secon- l l dary Water Chemistry Guidelines," Revision 1, June 1984, and the SGOG/EPRI "PWR l 4
Steam Generator Inspection Guidelines," Revision 1, July 1985. These improve-1 ments are gaining increasing acceptance and application throughout the industry, 1 tending to further reduce risk at the affected plants and to provide added .
assurance that risk from steam generator-related causes will continue to be '
small.
1
] 1. 8 Staff Actions and Studies 4 :
l The integrated program has identified a number of staff actions and studies a related to steam generator tube integrity, plant systems response, human factors l J considerations, radiological consequences, and organizational response to events, i These staff actions are discussed in Section 4 and are summarized in Table 2.
- A number of these actions have been completed, as noted in Table 2. Other staff l actions identified in Table 2 involve broad generic issues extending beyond *
{ strictly steam generator-related issues. These include issues being addressed 1 by another USI program (USI-49, "Pressurized Thermal Shock") and by currently d
approved staff implementation plans for on going gener;c issue reviews including
! "Improved Accident Monitoring" (NRC Generic Letter 82-33); "Reactor Vessel d
Inventor Measurement" (THI Task Action Plan (TAP) II.F.2, NRC Generic Letter i
82-28); y' Guidance on Reactor Coolant Pump Trip" (TMI TAP II.K.3.5, NRC Generic l
Letter 82-33); "Control Room Design" (TH! TAP I.0.1, NRC Generic Letter 82-33); :
1 and "Improved Emergency Operating Procedures" (THI TAP 1.C.1, NRC Generic Letter ,
j 82-33). Completion of these broad generic tasks is considered to be outside l the scope of the staff's integrated program to resolve USIs A-3, A-4, and A-5 t j regarding steam generator tube integrity. j j
- The remaining staff actions identified in Table 2 and discussed in Section 4 l
involve other issues related to steam generators. As noted in Table 2 and l
- discussed further in Section 4, a number of these remaining staff actions are >
- relatively low priority tasks which will remain inactive pending completion (
j of higher priority tasks and availability of staff resources. Others of these '
4 remaining staff actions, as indicated in Table 2, are actively being pursued [
! as part of Generic Issue 135, "Steam Generator and Steam Line Overfill Issues," ;
j and/or the Steam Generator Group Project / Steam Generator Tube Integrity l
f i NUREG-0844 1-10 !
Program 1ponsored by the NRC Office of Nuclear Regulatory Research. In view of the low risk estimates associated with SGTR events, the staff concludes that the resolution of USIs A-3, A-4, and A-5 is not contingent upon comple-tion of these tasks. However, these tasks will help ensure that risk con-tinues to be low and may lead to proposals fer revising existing regulatory guidance and possibly requirements concerning steam generator tube inspections and repairs, revisions to the Standard Review Plan concerning the design basis SGTR, and resolution of the steam generator / steam line overfill issue. The potential regulatory and safety benefits and cost of implementation will be assessed for any proposals stemming from these activities. If justified by this cost / benefit analysis, additional or revised regulatory guidance or requirements may be issued.
Of special note is the comprehensive assessment of steam generator inspection programs, including inspection sampling strategies and eddy current test prac-tices, being performed as part of the Steam Generator Group Project / Steam Generator Tube Integrity Program (SGGP/SGTIP) sponsored by the NRC Office of Nuclear Regulatory Research. There is increasing evidence from this program and from operating experience of deficiencies in the reliability of current field inspection practices, although the staff does not believe that signifi-cant risk to public health and safety is involved. The SGGP/SGTIP program is expected to lead to improved regulatory guidance which addresses these deficien-cies. This program may also lead to new augmented inservice inspection require-ments (pursuant to 10 CFR 50.55a) should it be determined from this program that added assurance of steam generator tube integrity is needed. In addition to staff efforts in this area, SGOG/EPRI will shortly be Ssuing Revision 2 to the "PWR Steam Generator Inspection Guidelines" to address inspection reliabil-ity deficiencies experienced to date.
1.9 Public Comments A draf t version of this report was issued for public coment in April 1985.
The comment letters received are listed in Appendix A. Most of the comments received focused on the merits of the staff recommended actions issued in NRC Generic Letter 85-02. These coments were generally within the scope of ecrlier coments discussed in Section 1.4, which were considered by the staff in the development and issuance of the staff recommended actions. None of the coments received regarding the draft version of this report took issue with the staff's major findings; namely that (1) SGTRs do not contribute significantly to the risks of nuclear plant operation nor to the risks to which the general public is routinely exposed and (2) USIs A-3, A-4, and A-5 can be considered resolved.
One utility, Carolina Power and Light Company, expressed the concern that the draft version of this report could be interpreted to mean that implementation of the staf f recomended actions in Generic Letter 85-02 is necessary in order to comply with 10 CFR Part 50, Appendices A and B (letter dated July 18,1985).
Such an interpretation would be contrary to the staff's intent as is now explained in Section 1.7.
1.10 Implications of July 15, 1987 SGTR Event at North Anna Unit 1 The material in tnis report generally predates the most recent SGTR event which occurred on July 15, 1987 at North Anna Unit 1. That failure has been attributed to a rapidly propagating fatigue crack caused by flow-induced vibration. The NUREG-0844 1-11
North Anna event was addressed in NRC Augmented Inspection leam (AIT) Report Hos. 50-338/87-24 and 50-339/87-24 and in the staff's safety evaluation (letter, December 11, 1987) authorizing North Anna Unit 1 to operate at 100 percent of licensed power. The AIT report includes a description of the event and staff findings pertaining to the systems response, operator response, and radiological consequences. The safety evaluation addresses the causes of the failure and the corrective actions taken to prevent a recurrence.
Generic implications of the North Anna event have been addressed in NRC Bulletin 88-02. The bulletin regeests that licensees and applicants perform specified inspections and analyses to determine whether their plants are susceptible to rapidly propagating fatigue cracks and that they implement corrective actions if necessary. These actions will ensure that steam generator tube integrity
' will continue to be maintained in accordance with 10 CFR Part 50, Appendices A and B, and that risk from SGTR related causes will continue to be icw. Thus, j
the conclusions of this report, as stated in Section 1.11 below, remcin valid.
1.11 Conclusions Stemming from the Integrated program (1) SGTRs do not contribute a significant fraction of the early and latent cancer fatality risks associated with reactor events at a given site.
The increment of risk associated with SGTR events is a small fraction of :
the accidental and latent cancer fatality risks to which the general '
public is routinely exposed. This finding is considered to be indicative ,
of the effectiveness of licensee programs and regulatory requirements for ensuring steam generator tube integrity in accordance with 10 CFR Part 50, Appendices A and B.
< (2) The Commission's curre,1t rules and regulations (i.e., 10 CFR Part 50, Appendices A and B; 10 CFR 50.55a; 10 CFR 50.109; and 10 CFR Part 100) provide the staff with adequate authority to ensure that licensees are implementing programs relating to steam generator tube integrity which provide adequate protection to public health and safety.
1 (3) The staff has identified staff recommended actions which can further l enhance the effectiveness of licensee programs for ensuring steam generator
]
tube integrity and for mitigating the consequences of an SGTR. The staff i
has issued Generic Letter 85-02 to inform licensees and applicants of the j staff recommended actions; however, these actions do not constitute NRC ,
1 requirements. l 1
I
, (4) It is implicit in the conclusions stemming from the integrated program
! that licensees must continue to be vigilant against new or unusual prob-lems that may necessitate preventive, diagnostic, and/or corrective actions beyond the licensee's normal practice. The staff will continue to monitor steam generator op. rating experiences as an indicator of the effectiveness of licensee programs. As has been the case in the past, the staff may impose additional requirements (pursuant to applicable regu-lations) to continue to ensure that licensees are implementing adequately ef fective programs where such action is determined to be necessary on the basis of operating experience or as e result of ongoing staff actions and !
1 studies.
, i
) NUREG-0844 1-12 ,
1
__ . _ . . _ . . _ . - _ _ . - - . . ~ . _ . . . - . . . . . . _ . .
(5) Section 4 of this report identifies a number of staff actions and studies involving steam generator issues that are being pursued to provide added !
l assurance that risk from SGTR related causes will continue to be small, j i
(6) The staff concludes that with final publication of this report (NUREG-0844), I USIs A-3, A-4, and A-5 are technically resolved. l i
I l
I l
i :
i
! I t
i -
] !
l }
l !
4 1
~
I l
i i l
(
i i
I t :
{
- I I l I !
i !
i l j l 1 i l NUREG-0844 1-13 i b ,
i l
~. . - _ -.- . . . . - - - -- . -- ,- . - .- ... - -
l t
i Table 1 Disposition of potential industry actions f Report
] Item Section Disposition
- 1. Prevention and Detection of Loose Parts l I
! (a) Visual Inspection of Secondary Side 2.1.1 Staff Recommended 1 q and QA/QC Work Procedures Action
. (b) Loose-Parts, Monitoring System 2.1.2 Dropped l 1 !
j 2. Steam Generator Tube Inservice Inspection j
) (a) Supplemental Tube Inspections 2.2.1 Staff Action (b) Full-l.ength Tube Inspections 2.2.2 Staff Recommended !
j Action l (c) Denting Inspections 2.2.3 Staff Action I
j (d) Steam Generator Inservice Inspection 2.2.4 Staff Recommended ;
Interval Action l
. (e) Inspections Following Shutdown for 2.2.5 Dropped j Repair of Leakage 1
J
- 3. Improved Eddy-Current Test Techniques 2.3 Staff Action ,
- 4. Upper Inspection Ports 2.4 Dropped f l
l 5. Secondary Water Chemistry Program 2.5 Staff Recommended j Action i 2
I
) 6. Condenser Inservice Inspection Program 2. 6 Staff Recommended ;
j Action [
! l 1 7. Stabilization and Monitoring of 2.7 Dropped l Degraded Tubes
! 8. Primary to Secondary Leakage limits 2. 8 Staff Recommended l 1 Action j i
- 9. Coolant Iodine Activity Limit 2. 9 Staff Recommended t~
Action 1
i 10. Reactor Coolant System Pressure Control 2.10 Staff Action 2
l
. 11. Safety injection Signal Reset 2.11 Staff Recommended j Action i
j 12. Containment Isolation and Reset 2.12 Dropped
~
l :
i
] NUREG-0844 1-14 4
i
Table 2 Related staff actions and studies
- i Report Item Section Action 1
- 1. Steam Generator Tube Integrity Guidance for Sleeve Repairs 4.2.1 **
Inspections for Denting 4.2.2 **
Improved Eddy Current Techniques 4.2.3 Generic Issue 135 l SGGP/SGTIP Category C-2 Inservice Inspection 4.2.4 Generic Issue 135 i SGGP/SGTIP
- 2. Plant Systems Responses Steam Generator Overfill 4.3.1 Generic Issue 135 !
Re&ctor Coolant System Pressure 4.3.2 USI A-45 Control THI TAP II.D.1 4
j -
Pressurized Thermal Shock 4.3.3 USI A-49 1
Improved Accident Monitoring 4.3.4 Regulatory
- Guide 1.97 Programs, Generic Ltr. 82-33 i
Reactor Vessel Inventory 4.3.5 TMI TAP II.F.2, '
1 Measurement Generic Ltr. 82-28 f
- 3. Human Factors l ,
L I -
Guidance on Reactor Coolant Pump 4.4.1 TMI TAP !!.K.3.5, Trip Generic Ltr. 82-33 i,
a t Control Room Design 4.4.2 THI TAP !!.D.1, l
Generic Ltr. 82-33 !
Improved Emergency Procedures 4.4.3 TMI TAP I.C.1, l Generic Ltr. 82-33 i
- 4. Radiological Consequences l r q Reassess Potential Consequences 4.5.1 ** I
i j See footnotes at end of table.
J
- NUREG-0844 1-15 1
Table 2 (Continued)
Report Item Section Action
- 5. Improved Organization Response to Events
- Seven actions are being taken 4.6 6 of 7 complete by staff l
, "These tasks are discussed in detail in Section 4.
- Not high priority. Will be scheduled pending availability of staff resources, i
l l
l NUREG-0844 1-16 .
r f
l i
l 2 VALVE-lMPACT EL LVATION OF POTENTIAL INDUSTRY ACTIONS l
E l This section discusses potential industry actions with respect to maintenance i of steam generator tube integrity and mitigation of steam generator tube ruptures f which were considered as part of the NRC integrated program. The basis for the '
consideration of these actions, the staff's value-impact findings concerning these actions, and the staff's disposition of these actions are also described. ;
The staf f's value-impact findings and the staff's disposition of the potential :
j industry actions are summarized in Table 3. t i f 2.1 Prevention and Detection of Loose Parts and Foreign Objects J
j Loose parts and foreign objects have caused two of the four domestic steam ;
! generator tube rupture (SGTR) events to date. In response, the staff identified i l for further consideration potential industry actions concerning (1) visual !
j inspections of the secondary side of steam generators and quality assurance [
procedures governing all work performed within steam generators, and (2) loose-parts monitoring systems. As discussed below, the potential industry actions in
) item 1 above have been categorized as staff recommended actions. Given imple- i 1
' mentation of these actions, the staff finds there is insufficient justification !
to support item 2 as a staff recommended action, i 2.1.1 Secondary Side Visual Inspections and Improved QA/QC Procedures l 2.1.1.1 Staff Recommended Actions f 4
l Visual inspections of the steam generator secondary side and improved quality 1 assurance / quality control (QA/QC) work procedures should be implemented for the {
l prevention and detection of loose parts and foreign objects, i j
l (1) Secondary-Side Visual Inspections
(
) Visual inspections should be performed on the steam generator secondary side in l
- the vicinity of the tube sheet, both along the entire periphery of the tube l
- bundle and along the tube lane, for purposes of identifying loose parts or I j foreign objects on the tubesheet and external damage to peripheral tubes just i above the tubesheet. An appropriate optical device should be used (e.g., mini-
! TV camera, fiber optics). Loose parts or foreign objects which are found should I i be removed from the steam generators. Tubes observed to have visible damage j i should be eddy current inspected and plugged if found to be defective, l l
l These visual inspections should be performed (a) for all steam generators at 1
{ each plant at the next planned outage for eddy current testing, (b) after any l I
- econdary-side modifications or repairs to steam generator internals, and I (c) when eddy current indications are found in the free span portion of j peripheral tubes, unless it has been established that the indication did not i result from damage by a loose part or foreign object. PWR operating license
] applicants should perform such inspections as part of the preservice inspection, i
i
! NUREG-0844 2-1 i
l l
l For steam generator models where certain segments of the peripheral region can l be shown to be inaccessible to an appropriate optical device, licensees and I applicants should take alternative actions to address these inaccessible areas, j as appropriate. l Licensees should take appropriate precautions to minimize the potential for corrosion while the tube bundle is exposed to air. The presence of chemical species such as sulfur may aggravate this potential, and may make exposure to the atmosphere inadvisable until appropriate remedial measures are taken.
(2) Improved QA/QC procedures -
f QA/QC procedures for steam generators should be reviewed and revised as necessary i to ensure that an effective system exists to preclude introduction of foreign ;
objects into either the primary or secondary side of the steam generator whenever it is opened (e.g., for maintenance, sludge lancing, repairs, inspection opera- l tions, modifications). As a minimum, such procedures should include (a) detailed accountability procedures for all tools and equipment used during an operation. l (b) appropriate controls on foreign objects such as eye glasses and film badges, j (c) cleanliness requirements, and (d) accountability procedures for components !
and parts removed from the internals of major components (e.g., reassembly of cut !
and removed components). I i
2.1.1.2 Basis for Initial Consideration l
Inspection methods and practices and material accountability controls at many i PWR facilities have not proven sufficiently effective in ensuring that loose i parts and foreign objects have been identified and removed before startup. For I example, the accountability controls in use at Ginna (NUREG-0909 and NUREG-0916) !
during the removal / modification of the downcomer resistance plate in 1975, while !
generally consistent with NRC guidance and industry, philosophy at that time, did not prevent parts of the plate from being lef t in the secondary side of the ,
steam generator. Deficiencies included (1) f ailure to perform a post-maintenance !
accountability inspection of the removed resistance plate to account fc.r all !
pieces, (2) no documentation to verify that an aaquate barrier existed to keep [
material from dropping into the steam generator, and (3) failure to perform ;
adequate post-maintenance inspection of the secondary side of the steam generator i for foreign parts. Foreign objects or loose parts have also been founo in the i steam generators at Zion 1 Prairie Island 1, North Anna 1, San Onofre 1. l Davis Besse, Rancho Seco, Oconee 3, Point Beach 7, Cook 1, and Turkey Point 4 ;
Several of these experiences are discussed in detail below. t Secondary-side visual inspections will also ensure that degraded conditions on '
the outer surface of peripheral tubes such as may be caused by loose parts or foreign objects have been adequately identified.
During the Ginna post-event activity, visual inspection with a remote TV camera revealed a number of foreign ubjects in the seccndary side of the B steam generator. The largest object appeared to be part of the steam generator's j downcomer flow resistance orifice plate which had been cut and removed in 1975 ;
during a modification of the steam generator. This large foreign object most s probably initiated the damage that led to the tube rupture. Post-rupture examina- l tion revealed that severe damage had occurred to 26 tubes in the periphery of l B steam generator. These tubes were so extensively damaged that the licensee j r
NUREG-0844 2-2 s
had to remove them to protect sound tubes. In addition, portions of two frac-l tured tubes found skewed between the tube bundle and the steam generator shell I
were also removed. Foreign objects were found in the A steam generator, althouCh no tube damage occurred that was attributable to these objects.
On October 2,1979, a tube ruptured in steam generator A of Prairie Island 1 (NUREG-0651) while the plant was operating at full power. The licensee estimated the leak rate at about 390 gpm. Visual and fiber optic inspections performed subsequent to the tube rupture incident revealed that the tube in
, Row 4, Column 1, had ruptured about 3 in above the tube sheet. The rupture was a classical tube burst with a "fish-mouth" opening about 1-1/2 in, long with a maximum width of about 0.5 in. The edges of the break were observed to be worn to a "knife edge." The tubes in adjacent positions (Row 3, Column 1
, and Row 2, Column 1) also showed signs of wear. All wear marks and the rupture i were on the outer peripheral side of the tube bundle at approximately the same
] elevation. A steel coil spring, 8.5 in long, 1.25 in. in diameter and of 3/32 in, wire diameter was found lying on the tube sheet adjacent to the defec-tive tubes. One end of the spring was wedged between the tube sheet and a flow-blocking device (the flow-blocking device diverts flow away from the open tube
, lane and into the tube bundle); the other end was free to move. A visible wear l pattern on the tube sheet indicated that the spring had moved back and forth i during plant operation.
On february 25, 1982, while preparing for eddy-current testing of the IB steam generator, Zion 1 station personnel discovered three pieces of a hinge about 30 in, long and 2 in, wide in the channel head plenum area of the steam generator.
These fragments were later determined to have come from an aluminum nozzle cover left in the 10 steam generator during tube testing in March 1981. It is believed that the aluminum cover dissolved during reactor operation leaving behind two stainless steel hinges that had held the cover. One hinge section was found bent, but in one piece, in the 10 steam generator. The other hinge j section was found in three pieces in the IB steam generator. The licensee attribt. led the presence of the three pieces of hinge to the reverse flow that occurred in mid-February 1982 when reactor coolant pumps A and 0 were shut
- down. More than 1100 protruding tube ends on the inlet plenum of the ID steam i generator had been severely damaged by the loose parts necessitating extensive repairs.
i j Visual inspections at Turkey Point 4 following plant shutdown in July 1982 to I repair a small tube leak, revealed that the leak was caused by damage to a tube !
at the periphery of the tube bundle by a loose part. The inspection was con- i ducted with a fiber optic device and included the entire periphery above the j tubesheet and the tube lane region. The loose part, identified as a check valve i a pin f rom the feedwater bypass line, rneasured 1 in. x 2-1/4 in. Damage was also ;
observed on other peripheral tubes in all three steam generators. All but one i of these other tubes had previously been plugged. Some of these tubes had leaked
! before being plugged in 1977. In addition to the pin, numerous other foreign
{ objects were found (e.g., bars, pieces of metal plate, bolts, wire, weld rod).
I 4
2.1.1.? Value-Impact Visual inspections as defined in Section 2.1.1.1 would result in initial inspec-tions of all steam generators (SGs) plus inspections each time the secondary side is opened for modifications or repairs. Assuming an average of three steam NUREG-0844 2-3 l
1 l
generators per plant, initial inspections of all steam generators during the next eddy-current test (ECT) outage, a remaining lifetime frequency of opening 1 the secondary side of all of a plant's steam generators for maintenance or repairs of once per five years, and an occupational radiation exposure of 5 to l 10 person-rem / inspection of one steam generator, the staff's consultant, SA1, i estimates this action to result in an estimated occupational exposure of from i 90 to 180 person-rem over an assumed 24 year average remaining life of the plant.
, For plants where an upgrade of existing QA/QC work procedures is needed in ordet' i to effectively preclude the introduction of foreign objects into the steam 1 generators, SA! estimates that implementation of these procedures could increase
! occupational exposures by 120 to 360 person-rem per plant over the assumed j 24 year average remaining life of the plant. Thus, the total ORE attributable
! to implementation of these actions is estimated by the staff to be from 210 to 540 person-rem per plant. SAI estimates on avoided occupational exposures due to implementation of the subject visual inspections and improved QA work proce-
, dures conservatively account only for the avoided dose attributable to preventing ,
1 ruptures in steam generator tubes and does not include avoided dose that accrues ,
d from preventing other types of damage, which must also be repaired, from loose '
i parts. When the avoided SGTR dose of 83 person-rem is considered, the net I i estimated ORE attributable to these two requirements is an increase of from 127 l j to 457 person rem over 24 years or about 5 to 19 person-rem per plant per year, j It should be noted, however, that these numbers are considerably smaller than l the avoided occupational exposures, 1060 person-rem (plant with medium degree ,
I of degradation) to 7500 person-rem (severely degraded plant), which can be !
! achieved through implementation of effective secondary water chemistry and t condenser inspection programs, which are discussed in Sections 2.5 and 2.6. i l
C SA! estimates that the cost required to implement these two actions (a' at $0.2M
! per plant ) is more than offset by the economic savings (> $3,1M) resulting from I implementation. Therefore, the staff finds that economic cost does not bar j implementation of these actions.
l These actions could potentially redece loose parts-related SGTRs by as much as an estimated 90%. Since two of the four SGTRs to date have been loose parts related, this translates to a 45% reducden in the overall SGTR frequency. l This reduction in SGTR frequency produces a corresponding reduction in t.he probabilities of core melt and significant non-core-melt releases. Using the criteria given in Table 3, these reductions correspond to a "low to medium" benefit in terms of reduced core-melt probability and a "high" benefit in terms i of reduced probability of significant non-core-melt releases. !
I t The following comments received from the SGOG (letter,, August 25, 1983) were j typical of many comments received from industry on this issue.
l l
- a. A one time visual inspection of the secondary side of a steam generator i through exisiting access ports is a reasonable method for finding loose ,
parts or foreign objects, providing the following points are recognized.
I (1) There are differences in steam generator geometry and access; !
1 therefore, the scope and type of visual inspection must be tailored l to the specific steam generator design. ;
6 i
i NUREG-0844 2-4 !
i i i
i
(2) Inspection should be balanced with awareness of the potential for tube corrosion when a steam generator is drained,
- b. Subsequent visual inspections of the secondary sides of steam generators should be performed only when the specific situation warrants, e.g., when nondestrw*ive examination suggests the presence of a foreign object or when QA/QC or -leanliness procedures employed during maintenance are judged to have been insufficient. When conducted, such a subsequent inspection should be restricted in scope and duration to the minimum required to resolve the specific question that prompted it.
Comments a.(1) and a.(2) above are considered to be consistent with the intent of the staff recommended actions. The staff also agrees in principle with SG0G's comment b above; specifically, that followup inspections should be performed only when the specific situation *arrants. The staff believes that the staff rec ^mmended actions are consistent with this goal, and are an acceptable approach to ensuring against loose parts induced damage.
Consumers Power Company (letter, September 1,1983) and Duke Power Company (letter, January 6,1983) expressed concerns regarding the potential for corrosion while the steam generators are drained and the tube bundles are exposed to the atmosphere. Consumers Power Company stated that it experienced damage requiring extensive tube plugging during a "dry-layup" condition similar to what would exist during the proposed visual inspection, and that recent corrosion events at Oconee (following auxiliary feedwater system modifications) and Arkansas 1 underscore the apparent sensitivity of steam generators to air. Consumers Power also commented that work by the Electric Power Research Institute (EPRI) also supports avoidance of a dry, moist layup environment for steam generator tubes, and concluded that proper followup of eddy current indications which are indicative of potential tube wear as well as appropriate procedural controls and QA/QC is sufficient for the prevention and detection of loose parts.
The staff acknowledges the desirab lity for minimizing exposure of tube bundles i
to air. This is particularly importsnt in cases where chemical species such as sulfur (which may be highly aggressite in the presence of oxygen even under ambient conditions) are present in the steam generator. Regardless of the
&mnstances under which the steam generator is being drained (e.g. , for purpose. M performing maintenance, repairs, modifications, and/or visual inspections), precautions must be taken to ensure that the potential for corrosion is minimized, and that the steam generators are restored to a wet layup condition as quickly as possible. The staff agrees that in certain cases, such as when significant amounts o' sulfur are present (as was apparently the case at Arkansas 1), exposing the tube bundle to air may be inadvisable until appropriate remedial measures are taken.
Visual inspections consistent with the staff recommended actions should not result in significant increases in exposures of the secondary side to the atmosphere. The staff recommended actions involve a single baseline inspection of each steam generator, with followup inspections only if modifications or repairs are made to the SG internals (which typically occur only once every five years) or when ECT indicates evidence of loose parts-induced damage. Wisconsin Electric Power Company, which has already performed such inspections, estimates approximately 2 days are required to inspect each steam generator (letter, August 10, 1983).
NUREG-0344 2-5 1
I l
1 With appropriate precautions, and remedial measures if necessary, visual inspections in accordance with the staff recommended action are not expected i to result in a significant detrimental impact on the steam generator "Aes.
! However, because of concerns related to corrosion, some licensees may elect
- to implement alternatives to visual inspections, including installation and operation of a loose parts-monitoring system as discussed in Section 2.1.4.
Any such alternative approach should include criteria for determining when appropriate remedial measures should be taken to address any suspected presence j ofloosepartsorforeignobjects, j l Duke Power Company also expressed the concern that visual inspections could themselves increase the potential for leaving foreign objects in the steam i generators. The staff disagrees with this comment since the staff recommended I j actions also call for improved QA/QC procedures in addition to visual inspections
- to preclude the source of Duke Power Company's concern.
2.1.1.4 Conclusions j Given that two of the four SGTR events to date have been caused by damage caused j
) by loose parts / foreign objects, effective licensee programs to prevent and detect j
] loose parts / foreign objects are an important element of overall licensee programs {
intended to ensure the integrity of the steam generator tubes. The staff concludes l
{ that visual inspections and improved QA/QC work procedures, as identified in i
) Section 2.1.1.1, would be effective measures for the prevention and uetection t l J of loose parts and foreign objects, and has therefore adopted these item', as l j staff recommended actions. Apart from benefits in the areas of reduced 3GTR j i frequency and the associated probabilities of core-melt and significant non- :
l core-melt releases, the staff's value-impact analysis indicates that trese i 1 actions will be cost effective. The net ORE impact of 127 to 457 pers Jn-rem 1 per plant associated with these actions is more than offset by the estimated j ORE reductions (1000 to 7500 person-rem) which can be achieved through implemen- ,
tation of effective secondary water chemistry and condenser inspection programs j (Sections 2.5 and 2.6). ;
i 2.1.2 Loose-Parts Monitoring Systems j In the ennt that loose parts or foreign objects are introduced into the steam !
generators, operation of a loose parts-monitoring system (LPMS) will increasa i q the probability of prompt detection. The staff considere1 as a potential !
< industry action that all PWRs should install and operate an LPMS. Such a system i should be capable of monitor'ng the steam generator prinas e and secondary sides, !
and should conform to Regulatory Guide 1.133. In additic , sufficient sensors l should be providec .n acoustically coupled regions of the steam generator to i I ensure adequate LPHS sensitivity for the detection of loese parts in the l l secondary side and in the channel head. !
l i
j It is likely that of the two SGTRs experienced to date that were attributed to loose parts, either visual inspection after maintenance or improved QA accountability would have detected the causes of one of the SGTRs before it l J occurred and either visual inspection, QA, or LPMS would likely have detected l
) the cause of the other SGTR before it occurred. SA! estimated a 70% effec- l tiveness for implementation of an LPMS in detecting loose parts, compared to a j l 90% ef fectiveness for secondary-side visual inspections and QA accountability. ;
i i
! l l NUREG-0544 2-6 !
I l
2
(
I !
I i
Give.n that two of the four SGTRs to date have been loose parts related, visual !
inspections plus QA would be expected to reduce the overall baseline SGTR fre- !
, quen,/ by 45%. Implementation and operation of an LPMS would incr;ase the !
amount of reduction to 48%. Thus, the incremental improvement that accrues from i the additit,n of an LPMS, given the existence of inspection and QA accountability, l 15 small. i The staff expanded on the SAI analysis to consider a range of effectiveness of 60% to 90% for steam generator visual inspection and QA/QC accountability and !
a range of 70% to 90% for LPMS.
i Given the above ranges of effectiveness, the overall SGTR frequency is reduced
! between 30% and 45% by implementing visual inspections plus QA/QC. If, in [
- addia. ion an LPMS is implemented, the overall SGTR frequency is reduced between ;
j 44% and 50%. Given the implementation of visual inspections and QA/QC, the !
added reduction in the overall baseline SGTR frequency attributable to imple- t j menting an LPMS is between 3% and 18%. There would be a corresponding addi- !
- tional reduction in the probability of core-melt and significant non-core melt [
releases. Based upon the value impact criteria otfined in Table 3, implementa- ,
tion of an LPMS would provide an additional "low" benefit in terms of reduced !
probability of core melt, and an additional "medium" benefit in terms of reduceo l i probability of significant non-core-melt releases. However, depending on the !
I effectiveness of visual inspections and QA, this "medium" benefit in reducing l 1 the probability of significant non-core-melt releases may be only a small >
l fraction of the corresponding benefits to be derived from visual inspections ;
- and QA/QC (Section 2.1.1) and from implementation of staff recommendations ;
I concerning improved secondary water chemistry control (Section 2.5) and ;
) condenser inspections (Section 2.6). l t
The net economic benefit of visual inspection plus quality assurance (QA/QC) !
is estimated by SA1 at $2.9 million (H). The net economic benefit of visual !
- inspection + QA/QC + LPMS was estimated to range from $2.2M to $2.6M where an I LPMS must be installed. The net benefit of visual inspection + QA/QC + LPMS !
- was estimated to range from $2.4M to $2.8M for plants having an acceptable LPMS. !
J f a The staff's consideration of a broader range of effectivenesses for the potential !
actions indicates a maximum net benefit nf $1.9M from visual inspections + QA/QC, !
] assuming these actions are 60% rather than 90% effective. Assuming a 90%
1 effectiveness for an LPMS, the net cost benefit for visual inspection + QA/QC +
) LPMS is from $2.3M to $2.EM. Thus, for the range of parameters considered, j the net cost saving attributable to an LPMS ranges between a negative $0.7M and
- a positive 50.9M. l l i i The SAI study indicates that the ORE required to implement an LPMS (10 to 15 ;
person-rem over plant life) is affset by the avoided ORE (6 person-rem over !
l plant life). As with the ecotomic impact, the staff has evaluated a broader (
! range of possibilities, but Sas found that any net ORE savings would be minimal j i (suimum of 33 pers ,n-rem ovtr an assemed 25 year re.saining plant lifetime), j The SGOG stated that an LPMS should not be required on the secondary sides of !
steam generators because visual inspection QA/QC and follow-up of ECT indica- !
tions should be adequate and because signals from an LPMS currently available {
have proven difficult to interpret leading to difficulty in determining when i action should be taken (letter, September 30, 1982). Wisconsin Public Service
) 2-7 NUREG-0844 l l ;
- o
Corporation (WPSC) stated that LPMSs have a low confidence level as illustrated by plant operating experiance and are not reliable enough to justify purchase and installation (letter, October 4.; 1982).
The SG0G comments or, " wcessity of an LPMS based on the adequacy of visual inspections + QA/~ a .istent with the staff's conclusions. The staff does not necessarily a. '
the remaining SG0G comments and the WPSC comments because it is als, from operating experience that a properly designed, maintained, and r m -
is can provide very useful information that can contribute to ave age to a plant. However, the benefits do not include significant net Gn. , or other benefits not already accounted for by visual inspections + QA/QC.
Depending upon the effectiveness of visual inspections and QA/QC in preventing loose parts-related failures, a secondary-side LPMS may provide only a small
, additional incr m ental reduction in SGTR frequency and in the probabilities of core-melt and significant non-core melt releases. Any potential for net cost benefits is also dependent on the effectiveness of the visual inspections and improvr1 QA/QC procedures. Future experience will indicate the actual effec-tiveness of the visual inspections and improved QA/QC work procedures and whether l additional backup provisions, such as an LPMS, are necessary. Given the imple- '
, mentatic . of appropriate secondary side visual inspections and improved QA/QC procedures, the staff concludes there is insufficient justification at this time to support implementation of a LPMS system as an additional staff recom-mended action. However, as discussed in Sections 2.1.1.1 and 2.1.1.3, some utilities may prefer to implement an LPMS in lieu of secondary-side visual
- inspections.
2.2 Inservice Inspection of Steam Generator Tubes i I
- 2.2.1 Supplemental Tube Ir.spections 2.2.1.1 Potential Industry Action i
4 The current Standard Technical Specifications (STS) (NUREG-0103, NUREG-0212, and NUTIEG-0452) specify that inservice inspection of the steam generator tubes be
! performed at periodic intervals. The minimum required inspection sample is 3%
~
of all SG tubes per plant. The results of this inspection sample are categorized as C-1, C-2, and C-3 depending on the severity of the results, with C-3 being i the most severo. It was initially proposed that the definition of Category C-2
! be revised to inccrporate Categories C-2 and C-3 as presently defined in the STS. Specifically, Category C-2 would have been redefined to include inspection results where 1 or more tubes is found te be defective or where 5% or more of '
the tubes inspected are found to be degraded. The definition of Category C-1 ;
- would have remained unchanged.
- For steam generators with Category C-1 results, no additional actions would have been required (unchanged from current STS). If steam generators with ,
j Category C-2 results were found, the following actions would have been required.
(1) In each stum generator where the results of the first sample inspection are Category C-2 (in accordance with the revised definition), additional 1
, tubes would be inspected. The sample would include either 100% of the :
, remaining tubes in tise steam generator, or a statistically based sample !
f NUREG-9844 2-8 i i i
I
which ensures with a probability of 0.95 that the number of defective ,
tubes that could remain uniispected is less than the tolerable number of ,
tube failures during design-basis accidents. Methods for determining !
these statistically based samples are detailed in NUREG/CR-1282. The :
tolerable number of tube f ailures would be determined by plant-specific l analyses (a description of these analyses is provided in a draft NRC i report, "Initial Staff Recommendations Stemming from Resolution of USIs
- Regarding Steam Generator Tube Integrity," which is enclosed with NRR memorandum, SECY 84-13 dated January 11,1984). The statistical methods and systems analyses would be approved by the NRC before the results could be used to develop and implement a statistically based sampling inspection program.
(2) Supplemental sample inspections could be limited to a partial-length ,
inspection af each tube, providing the inspection includes those l portions of the tubes where imperfections were found during initial !
sampling.
(3) Notwithstanding any inspection rotation schedule, any additional steam .
generators not yet inspected during the current inspection shall be
< inspected in accordance with the requirements in the Steam Generator Tube Sample Selection and Testing portion of the Technical Specifications.
(4) Prompt notification of NRC, in accordance with the Technical Specifications, l would be required,
, t 5 Under the proposed potential industry action, the staff would have considered i licensee proposals to change the Technical Specifications to permit st 31emental '
j sample inspections to be limited to subsets of tubes if it could be shc in fre-previous inspection results or from unique design or phenomenological rspects
- that the degradation is limited to well-defined areas contained within these -
1 subsets of tubes. l d
l Considering the results of the staff's value-impact analysis (described in ;
Section 2.2.1.3), the staff has concluded that this potential industry action '
is not appropriate in its present form for inclusion as a staff recommended action. As discussed in Section 2.2.1.4, the staff will evaluate the supple-mental tube inspection sampling issue as a staff action.
) 2.2.1.2 Basis for Initial Consideration The current requirements for inservice inspection frequency and scope are based primarily on experience, engineering judgment, and practicality. The required ,
frequency was based on the frequency of refueling outages so that regular ISI ;
- would not unnecess.rily affect plant availability and incur needless expense. '
The required scope of ISIS also was established primarily on the basis of experience and judgment with the goal of achieving safe operation of steam t generators by selacting a representative tube sample and minimizing personnel exposure. No analysis'has been performed which included (1) a system and ,
. accident evaluation to establish the limiting number of defective tubes that '
can be tolerated to fail during design-basis accidents and (2) statistical determination of the required scope of inspection to ensure that no more than the limiting number of defective tubes will not be inspected. Under the current l
ISI requirements, a minimum 3% initial sample of the steam generator tubes must ,
1 i l
5 NUREG-0844 2-9 i
! l 1
be inspected. Although there is no theoretical basis for the initial 3% sample i
size, 3% inspection sampling in conjunction with primary to secondary leakage
- rate limits in the Technical Specifications have been gener611y successful in
- identifying the existence of tube degradation problems, and serving as a basis
- for determining whether additional sampling should be performed. This success has been due largely to the fact that the primary modes of degradation affecting operating steam generators are mechanistic in nature. They result either from adverse chemical conditions, improper mechanical design or materials selection, or a combination of these parameters. The result is that when such conditions occur, the degradation is not generally isolated but affect 3 a large number of
, tubes. Thus, the initial 3% sample is sufficient to' identify the steam generators which are experiencing general degradation. Because of this, the 3% inspection i has also proved sufficient to determine if a steam generator tube leak was the j
- result of an isolated incident or if it was the result of a significant mode of '
q general degradation. Thus, the initial 3% inspection sample requirement would i remain unchanged under this potential industry action.
The inspection results for the initial 3% sample are currently categorized in
, the STS as C-1, C-2, or C-3 depending upon the number of defective tubes (i.e.,
tubes with imperfections which exceed the plugging limit) and degraded tubes >
(i.e., tubes with imperfections greater than 20% of the tube wall thickness, but less than the plugging limit) found as described below: '
i Category Inspection Results
[
C-1 Less than 5% of the total tubes inspected are degraded l tubes and none of the inspected tubes are defective. !
l C-2 One or more tubes, but not more than 1% of the total tubes [
] inspected are defective, or between 5% and 10% of the total !
tubes inspected are degraded tubes. #
l C-3 More than 10% of the total tubes inspected are degraded [
i tubes or more than 1% of the inspected tubes are defective. :
4 t
For results categorized as C-1, no additional sampling is required under the current sis. This would remain unchanged under the potential industry action.
For results categorized as C-2 (as currently defined in STS), the STS requires that an additional 6% sample (of the total number of SG tubes) be inspected.
, If the results of this additional sample are also categorized as C-2, then an !
4 additional 12% sample must be inspected. If the results of this third sample j y inspection are also categorized C-2, the current STS requires no additional !
The level of sampling required for Category C-2 has not been based sampling. ;
l on providing any specific statistical confidence level that requires that the 2
number of uninspected tubes with flaws exceeding the plugging limit will be ;
j less than the maximum tolerable number of tube failures for postulated accident ;
conditions.
To illustrate the limitations of the current requirement, one can consider a !
ll situation in which 1% of the SG tubes are defective, and that the. defective !
i tubes are uniformly distributed across the bundle. For a truly representative j
- initial 3% sample, the inspection results categorization would be C-2 (as ;
curr'ntly defined in the STS). If it is Category C-2, additional supplemental i l t N N G-0844 2-10 l
t
- - - - - , , - - _ - -. --------,---_n- , . - . + , . . . . , . , _ - - - - - - n-- - - -
samples ranging to a maximum additional 18% (6% + 12%) of the tube population would be inspected. Again, for a truly representative sample, the results of the supplemental inspection would also be STS Category C-2, and no additional sampling would be required. Under this situation, 1% of the remaining 79% of the tube population which was not inspected would be defective. This would amount to as many as 120 tubes (for a B&W steam generator) which would be defective but which would not be inspected.
Another limitation to be noted for this case is that under current requirements for C-2, the other steam generators would not have to be inspected. The STS permits the licensees to inspect the SGs on a rotating schedule, provided none i
of the SGs which are inspected are found to be Category C-3. Steam generators l at a given plant often (but not always) behave in a similar manner. When one SG is found to be in Category C-2, it is reasonable to infer that the uninspected steam generators may be similarly degraded.
Under the potential industry action for Category C-2, the supplementary sample size would be either 100% of the remaining tubes or would be based on plant-specific analyses as discussed in Section 2.2.1.1. The inservice inspection program would also be extended to include the other steam generators, in the event that they are not already included in the inspection program. In addi-tion, prompt notification of the NRC would be required if Category C-2 results are obtained.
For Category C-3 results (as presently defined in the STS), the current STS requires inspection of 100% of the tubes in the subject steam generator, exten-sion of the inspection into the other steam generator (s), and prompt notifica-tion of the NRC. Under the potential industry action, Category C-3 results as currently defined in the STS would be redefined as Category C-2. However, the additional actions to be taken under the potential industry action for inspec-tion results which are currently defined as Category C-3 would remain unchanged, except that a statistically based sampling plan could be employed in lieu of 100% inspection of the remaining tubes if justified by the statistical analyses and plant specific systems analyses discussed in Section 2.2.1.1.
Having determined from the initial 3% sample that a steam generator is experi-encing significant degradation, the intent of the potential industry action is to ensure that the number of uninspected tubes which may be defective is less than the tolerable number of tube failures during design-basis accidents. The tolerable number of tube failures would be based on (1) use of 10 CFR Part 100 to determine the maximum tolerable leak rate through failed steam generator tubes concurrent with a main steam line break (MSLB) outside containment, (2) use of the maximum containment design pressure limit to detec.iine the maximum tolerable leak rate through failed steam generator tubes during an MSLB inside containment, and (3) use of the 2200 F peak cladding temperature limit established by 10 CFR 50.46 to determine the tolerable leak rate through failed steam generator tubes concurrent with a loss-of-coolant accident.
2.2.1.3 Value-Impact Economic and ORE Value-Impact The staff estimates the plant average total present worth costs to implement the potential industry action over an assumed remaining 24 year plant life to be as follows:
NUREG-0844 2-11
- - - - - =. .- . . - - .
W: $2,900,000 CE: $3,540,000 B&W: $8,260,000 The bases for these estimates are provided in Section 2.2.1.5. The difference in costs between the NSSS SG types reflects differences in the number of tubes per SG and the number of SGs per plant. These estimates assume that licensees would submit proposed changes to the Technical Specifications to permit supple-mental inspections to be confined to subsets of tubes in which the degradation can be shown to be limited to these subsets. This assumption has reduced the estimated cost for implementing this action by approximately 50%.
l These cost estimates are dominated by added costs resulting from extended out-age times. These are plant average cost estimates. Depending on plant-specific circumstances, the implementation costs may be significantly higher at some plants.
However, these costs do not consider steps that could be taken by utilities to shorten the necessary inspection times (e.g., development of multiple prob.
inspection techniques and real-time ECT evaluation) since the feasibility of such steps have not been evaluated by the staff.
Based upon estimated avoided costs, the average cost benefit (impact) during the remaining plant life is as follows:
W: $[1,600,000] to $2,500,000 CE: $[2,240,000] to $1,860,000 B&W: $[6,960,000] to $[2,860,000]
Note: Unbracketed figures represent net cost benefit (savings);
bracketed figures represent net cost impact. .
The average per year ORE is estimated as follows:
W: 5.2 person-rem CE: 5.1 person-rem B&W: 9.0 person rem These ORES are only partially offset by the estimated avoided ORE of from 1.3 to 6.3 person-rem. However, when compared to the average ORE for inspection, maintenance, and repairs of steam generators of about 150 person-rem as reported in NUREG-0886, the net ORE impacts are relatively small. On this basis, the staff concludes that ORE is not a major factor in this value-impact evaluation.
Potential Reduction in Core-Melt and Non-Core-Melt Release Probabilities Based upon experience to date, the potential for reducing the baseline frequencies (Section 3) of tube rupture occurrences is estimated by SAI to be on the order of from 5% to 20%. This compares with an estimated 64% reduction from imple-menting the staff recommended actions for the prevention and detection of loose parts (Section 2.1), improved secondary water chemistry control (Section 2.5),
and condenser ISI (Section 2.6). Assuming that these staff recommended actions are implemented, the percentage reduction in baseline frequency of rupture directly attributable to the potential industry action for supplemental inspection sampling may be no greater than 6%. This estimate is somewhat uncertain, but NUREG-0844 2-12
does indicate a potential medium benefit (as defined in Table 3) in terms of reduced probability of significant but less than core-melt releases, but only a low benefit in terms of reduced probability of core melt.
These relatively low benefits reflect operating experience which has shown that when flaws escape detection during ISI sampling, the most likely consequence is a small leak. Allowable limits for primary to secondary leakage have been established in the plant Technical Specifications beyond which the plant must be shut down for appropriate corrective action, thus minimizing the potential for tube ruptures during normal and postulated accident conditions. It should be noted that the staff recommended action for PWRs to adopt the leak rate limits I ,
- specified in the Standard Technical Specifications (Section 2.8) would provide added assurance of adequate tube integrity. ,
Industry Comments ,
This potential industry action was commented on extensively by industry. The industry comments were generally negative, both as to the need for this potential action and as to potential costs and ORE impacts.
Comments received from SGOG (letter, August 25, 1983) were generally representa-tive of comments received from other industry representatives. Some of the major points made by SG0G included the following:
(a) While a 3% random sample is a reasonable, proven starting point for '
eddy current inspection of a steam generator, immediate escalation to inspection of all the tubes in the steam generator upon finding a single defective tube (or five percent degraded tubes) and inspection l of all other steam Jenerators are unwarranted and undesirable. l t
The current step-wise progression from 3% to an intermediate sized sample before launching a 100% inspection has ensured that steam generators with higher levels of degradation have been inspected completely while those with isolated instances of tube degradation have been able to avoid unnecessary additional costs '
and radiation exposure of 100% inspections.
l Utilities fully understand the need for adequate eddy current inspections and will continue to perform them; however, the requirement that all tubes be inspected if one in a sample is found to be defective may provide a disincentive to inspect more than the minimum sample size, i -
If eddy current inspection requirements are revised despite the 4 lack of demonstrated need, the revised requirements should retain an intermediate step between the initial sample and
- inspection of 100% of the tubes.
(b) The need to inspect all other uninspected steam generators should be evaluated on a case by case basis. When a mode of tube degradation can be isolater; to a single steam generator, costly and lengthy inspections of other steam generators are unnecessary.
NUREG-0844 2-13
_ _ _ ~ _ __ _ _ _ _ . - . . _ _ ._ _
(c) The cost estimates used in assessing the impact of the proposed actions appear to be low and unrealistic. They assume that the inspection can be performed off the critical path with no cost for replacement power. Additional unscheduled inspections will likely be in the critical path and replacement power would be required.
The cost of eddy current inspection, including replacement power costs, may easily reach $500,000 to $800,000 per day. [ Babcock &
Wilcox (letter, February 10,1983) and Duke Power Company (letter, January 6, 1983) noted that the impact on critical path outage time during unscheduled shutdowns to repair SG leakage would be particularly severe for plants with Babcock & Wilcox steam generators due to their large number of tubes.] '
(d) Occupational radiation exposure due to performing steam generator eddy current inspections under existing requirementsrhas been reported (NUREG/CR-1490) to be between 5 and 20 person-rem per steam generator.
Occupational radiation exposure could be expected to increase sub-stantially under the proposed requirements.
(e) Since inspection requirements and criteria will be defined, prompt notification of the NRC of results should not be required.
These comments and other similar comments from industry were considered in the staff's disposition of this potential industry action. Regarding comment (c)
Move, the staff acknowledges that consideration of additional outage time asso-ciated with this potential industry action was not included in the SAI value-impact analysis which was provided to industry for its review and comment. As discussed in Section 0.2.1.5, the cost estimates herein do consider the poten-tial for increases in outage time for Westinghouse, Combustion Engineering, and Babcock & Wilcox steam generators, respectively. Regarding comment d) above, the staff estimates of occupational exposures (see Section 2.2.1.5) are con-sistent with NUREG/CR-1490.
2.2.1.4 Conclusions Licensees should be aware of limitations of present supplemental inspection sampling requirements in the Technical Specification in cases where Category C-2 results are obtained during initial sampling. In such cases, licensees should carefully consider the need for additional inspections beyond minimum Technical Specification requirements as may be determined necessary to ensure that uninspected tubes will not be subject to rupture during normal operating or postulated accident conditions.
t The limitation of the current requirements notwithstanding, the potential for further reducing the baseline SGTR frequency, and thus the probabilities of .
core melt and significant non-core-melt releases, appears to be relatNely small t compared to the reductions that can be achieved through the implementation of visual inspection and QA for loose parts (Section 2.1), improved seconaary water ;
chemistry (Section 2.5), and condenser inservice inspection (Section 2.6).
Considering this and possibly significant net cost impacts, the staff concludes that the potential industry action as described in Section 2.2.1.1 is not appro-priate for inclusion as a staf f recommended action.
NUREG-0844 2-14 i
Ar, discussed in Section 4.2.4, the staff will undertake further evaluation of the supplemental tube inspection sampling issue as a staff-action. This effort will investigate the-need for alternative, more practical actions to upgrade the existing Technical Specification requirements pertaining to this issue which could be implemented on a case-by-basis as needed.
2.2.1.5 Cost and ORE Value-Impact Assumptions This section describes the basis for the' staff's cost and h ; value-impact >
evaluation for the potential industry action concerning N,,lemental tube inspections (Section 2.2.1.1). The staff's evaluation is based largely on the !
SAI unit cost and ORE estimates (letter, r, .uary 25, 1983 and SAI report, l "Value-Impact Analysis").
l l
The following assumptions were employed:
l l
(1) Westinghouse (W) steam generators: 3300 tubes /SG, 4 loops j Combustion EngTneering (CE) steam generators: 8500 tubes /SG, 2 loops ;
i Babcock and Wilcox (B&W) steam generators: 15,500 tubes /SG, 2 loops (2) The cost estimates herein assume a concurrent inspection of two steam generators. For W with 6600 tubes /SC plants, the four SGs have been idealized as two SGs, '
(3) For a given inspection, the staff assumed a 0.4 probability that one or more steam generators will be Category C-2 (as defined in current STS) .
- based on an informal staff survey of plant data.
(4) Given that one SG is C-2, the other SG was assumed to have 0.5 probability of also being C-2.
(5) Given that an SG is C-2, the probability of having to perform 100% (or .
close to 100%) inspection of the SG tubes was assumed to be 0.5. This is a highly judgmental estimate. It assumes that licensees would propose !
amendments to Technical Specifications to permit the supplemental inspec-tions to be limited to subsets of tubes in cases where it can be shown that the degradation is limited to well-defined groups of tubes because of !
unique design or phenomenological aspects. The effect of this assumption J is to reduce the estimated costs for implementing this potential requirement !
by approximately half. Although not directly considered, the probability l of performing 100% inspections could also be influenced by the use of ;
statistically based sample sizes based upon the tolerable number of tube i failures.
4 (6) Additional costs for 100% inspection were computed relative to cost of ;
inspecting 21% of the tubes per the current requirement.
i I (7) It is assumed that tubes will be inspected at a rate of 36 tubes /hr/SG. '
< This considers that 20% of the tubes will be inspected full length at
! 20 tubes /hr and 80% of the tubes over a partial length at 40 tubes /hr (see page 68 of the SAI report).
t l NUREG-0844 2-15 I
I
4 (8) From page 68 of the SAI report, ECT inspection costs are estimated to be
$22,000/ day for parallel inspection of two SGs with 4 crews in 10-hour shifts. Inspection costs are estimated to be $15,000/ day for inspection of one SG with 2 crews in 10-hour shifts.
(9) Additional ORE is estimated to be 5 person-rem for each additional SG entry (2 additional entries assumed for W, 1 additional entry assumed for B&W and CE) plus 2 x 10 3 person-rem for each additional tube inspected (page 37 of the SAI report).
(10) The staff estimates that in 80% of the cases in which at least one SG at a plant is found to be Category C-2, the licensees are already inspecting all SGs. This estimate is based upon an informal staff survey of plant data, j Based on the above, the annual costs and ORE to implement the potential industry
- action are estimated to be as follows
W: $34,000 4.6 person-rem CE: $40,000 4.5 person-rem i B&W: $68,000 8.0 person-rem To the above costs must be added the costs of substitute power (during scheduled outages) or replacement power (during unscheduled outages) in the event that the potential action results in an extended outage. SAI estimates a range in
! cost difference for substitute power from $0 to $15/MW-br or $0 to $360,000/ day, assuming that the utility can meet the demand with power from another of its
- nuclear or coal plants (letter, February 25,1983). If oil-fired plants are used, the cost difference is estimated to range from $25 to $45/MW-hr (Ibid).
The cost estimates given here are based upon an average cost difference of f
$360,000/ day for substitute power. .
( ,
The additional time necessary to perform the proposed additional supplemental j
inspections is estimated to be 7.5, 9.4, and 17 days for W, CE, and B&W plants, respectively. If the scheduled outage tinie is extended by the additional time 1 required to perform the additional supplemental inspections, then the cost for substitute power is estimated to be $2.7M for W, $3.4M for CE, and $6.15M for B&W. The additional annual cost for substitute power is determined by multi-
! plying these costs by 0.4 (probability of being C-2) times the probability that '
i the outage would be extended. SAI has estimated this latter probability to be ,
between 0.1 and 0.2 considering the range in durations of refueling outages i (Ibid). The SAI estimate was based upon 16 additional days being required to ;
perform the inspections. Considering the staff estimates of additional inspec-tion time (i.e., 7.5, 9.4, and 17 days for W, CE, and B&W, respectively), the staff assumed a 0.1 probability for W and CE plants and a 0.2 probability for '
B&W plants that the potential requiriment will result in an outage extension. !
Thus, the annual cost of substitute power is estimated to be as follows: i i
W: $54,000 -
CE: $68,000 J !
B&W: $245,000 !
i l NUREG-0844 2-16 ;
I .
f
Thus, the annual total cost to implement the potential action during scheduled outages is as follows:
W: $88,000 CE: $108,000 B&W: $314,800 If a plant is forced to shut down as a result of exceeding the Technical Specification primary to secondary leakage rate limit, an unscheduled SG inspec-tion is required. The industrywide average probability for SG-related forced shutdowns is 0.188 per year (SAI report). Implementation of the secondary water chemistry and condenser inspection recommendations (Sections 2.5 and 2.6) would be expected to reduce this frequency by an additional 32%. The probability that one or more steam generators will be categorized as C-2 is again assumed to be 0.4. If C-2, the probability of having to perform 100% inspection is again assumed to be 0.5. Cost of replacement power, assuming the outage is extended, is estimated to be $475,000/ day. On this basis, the plant's average annual costs and ORE to implement the potential industry action during unscheduled outages are as follows:
W: $91,000 0.6 person-rem CE: $119,000 0.6 person-rem B&W: $215,000 1 person-rem The total present worth costs to implement this potential action over an assumed 24 year remaining plant lifetime is obtained by increasing the "annual" figures (for scheduled and unscheduled outages) by a factor of 15.6 (letter, February 25, 1983) as follcws:
W: $2,900,000 CE: $3,540,000 B&W: $8,260,000 Present worth of avoided costs are estimated by SAI to be $1.3M to $5.4M. Thus, the present worth of the net savings (or costs) over a 24 year remaining plant lifetime is as follows:
W: $(1,600,000) to $2,500,000 CE: $(2,240,000) to $1,860,000 B&W: $(6,960,000) to $(2,860,000) 2.2.2 Full-Length Tube Inspection As discussed below, the potential industry action concerning full length tube inspections nas been dispositioned as a staff recommended action.
2.2.2.1 Staff Recommended Action The Standard Technical Specifications (STS) and Regulatory Guide 1.83, Part C.2.f, '
currently define a U-tube inspection as meaning an inspection of the steam generator tube from the point of entry on the hot-leg side completely around the U-bend to the top support of the cold-leg side. The staff recommends that tube inspections should include an inspection of the entire length of the tube (tube end to tube end) including the hot-leg side, U-bend, and cold-leg side.
NUREG-0844 2-17
This recommendation does not mean that the hot-leg inspection sample and the 4
cold-leg inspection sample should necessarily involve the same tubes. That is, it does not preclude making separate entries from the hot- and cold-leg sides and selecting different tubes on the hot- and cold-leg sides to meet the
- minimum sampling requirements for inspection.
Consistent with the current STS requirement, supplemental sample inspections (after the initial 3% sample) under this staff recommended action may be limited to a partial length inspection provided the inspection includes those portions of the tube length where degradation was found during the initial sampling.
d 2.2.2.2 Basis for Initial Consideration
- This recommended action involves a modified definition of a tube inspection such I' that the cold-leg side of the tubes from the upper tube support dnwn to the tube l outlet should also be in';1uded in the inspections. The basis for including the cold-leg sides of the tubes in the inspection is that operating experience has shown that the cold-leg side is also susceptiole to a variety of degradation
, mechanisms (e.g., wastage, pitting, denting, and fretting-induced wear).
A substantial fraction of PWR licensees currently recognize the importance of cold-leg tube inspections since approximately 70% to 80% of the plants with ;
U-tube-type steam generators are currently subjected to at least some cold-leg
, inspection during inservice inspections. Degradation (pitting attack) on the cold-leg side has been observed in hundreds of tubes at Indian Point 3 (meeting r
, summary, September 28, 1982) and Millstone 2 (letter, March 5, 1982). Prairie ,
! Island 2 (NUREG-0886), Salem 1, and Zion 1 have recently experienced cold-leg 5 corrosion-thinning degradation. New Westinghouse Model D steam generators experienced an early generic tube vibration problem in the preheater section 1 on the cold-leg side which could potentially cause accelerated wear on the tubes i 1 at the baffle supports. Examples of forced shutdowns as a result of cold-leg l j leakage include Indian Point 3 and Ringhals 3 in the fall of 1981. j l 2.2.2.3 Value-Impact i
In general, it will be possible to inspect the full length of tubing from either I the hot- or cold-leg side. The staff estimates that a single-entry full length inspection of a 3% sample of tubes will increase ORE by 0.4 to 1.0 person-rem
! per plant per annum. This compares with avoided annual ORE of 0.3 and 3 person- l
]'
rem for assumed reductions in forced outage frequency of 1% and 10%. !
j For approximately the inner five rows of tubes with small-radius U-bends, it will a not be possible to insert a standard full-sized probe through the U-bend to the t
- opposite leg. Smaller diameter probes can be used to inspect tM inner row tubes i
! on the opposite leg, but with some loss of sensitivity. Should circumstances ;
arise where additional test sensitivity is needed for these tubes, special- !
purpose probes (e.g. , probes utilizing surface riding coils) can be used. :
i Alternatively, the licensee may elect to make a separate entry into the opposite !
i leg to inspect with a full-sized standard probe. The licensee may also elect '
- to make a separate entry into the opposite leg in cases where the results of ;
- the first sample inspection indicate that degradation on the opposite leg is t l confined to elevations below, say, the first support. A separate entry !
] inspection will increase ORE by 1.6 to 12 person-rem. !
l !
i !
NUREG-0844 2-18 ,
t l l i, ,
The staff finds that separate entry inspections should generally not be necessary except perhaps in cases where degradation is found to be active in both legs.
This being the case, licensees would eventually find it necessary to inspect the cold leg (as a result of leaks or ruptures), regardless of whether such inspections were a requirement. The staff concludes, therefore, that the staff recommendation for full length inspections by all PWR licensees will have little net effect on ORE.
The economic costs of cold-leg-side inspections ($140,000 to $234,000 over the remaining plant lifetime), even if performed from entries on the cold-leg side, are more than offset by the benefits ($300,000) for even a 1% reduction in forced outage frequency. Therefore, cost does not appear to be a significant factor in this value-impact evaluation.
None of the four SGTR events to date has involved the cold-leg side. However, it is difficult to quantify the potential reduction in SGTR frequency associated with implementation of cold leg inspections, since 70% to 80% of Westinghouse and Combustion Engineering licensees already perform cold-leg inspections to some degree. Considering that the staff recomended actions pertaining to the prevention and detection of loose parts and foreign objects (Section 2.1) and improved secondary water chemistry and condenser inspection programs (Sections 2.5 and 2.6) are estimated together to produce about a 64% reduction in the baseline SGTR frequency, the staff has considered that implementation of full-length inspcctions would most likely provide an additional reduction in the range of 1% to 10% where such inspections are not now being implemented. This corresponds to a low benefit (as defined in Table 3) in tarms of reduced probability of core melt and to a medium benefit in terms of reduced probability of significant non-core-melt releases.
SGOG commented (letter, September 30, 1982) that inspection of cold-leg tubes is justified as noted by NRC; however, the scope should be flexible. Inspections should be plant specific with the extent and frequency based on each plant's history and experience wirh degradation in the cold leg. SG0G also commented that the hot-leg and cold leg inspectioris should not have to be performed on the same tubes. The staff's recommendation on full-length tube inspections is consistent with these comtrents.
Alabama Power Company commented that it supports the inspection of tubes through the entire tube length (letter, January 11, 1983). TVA proposed that the first cold-leg inspection be considered a baseline inspection and not be used to classify the results of the general inspection. Tube degradation detected during the first inspection would be addressed by taking subsets to bound the af fected tubes (letter, October 13, 1982). With respect to the TVA comment, the staff notes that extensive degradation problems have occurred on both the hot and cold legs. Thus, the staff believes that the cold leg should be inspected in the same manner as the hot leg. 7 2.2.2.4 Conclusions Considering the effectiveness of cold-leg inspections in minimizing the potential for tube ruptures, a potential medium reduction in significant non-core melt releases, and the small change in ORE and absence of any significant cost impact, the staff concludes that full-length tube inspections are justified as a staff NUREG-0844 2-19
recommended action for all PWRs with U-tube steam generators. This recommenda-tion addresses an obvious limitation of the current ISI requirement and is already being implemented to varying degrees by much of the industry.
2.2.3 Denting Inspections 2.2.3.1 Potential Industry Action Because denting has occurred in a number of steam generators, a proposal was developed to include criteria for denting inspections in all PWR Technical Specifications.
2.2.3.2 Basis for Initial Consideration The basis for initial consideration of this potential industry action was that operating experience has shown that surveillance of tube denting is necessary to preclude leakage from stress corrosion cracking caused by denting. There has been one instance (Surry Unit 2 in 1976) in which denting led to high stress ;
in the tube U-bend region resulting in an SGTR. Plant-specific criteria have I been established for plants.with extensive denting; however, generic or stan- '
dardized criteria have not been developed. Generic implementation of this potential industry action would ensure that effective criteria for denting are implemented for all plants.
I 2.2.3.3 Value-Impact '
i' 2
On the basis of its review of the SAI analysis, the staff finds that the ORE and economic impacts are offset by the avoided ORE and costs, and, thus, ORE ,
and economic costs are not major factors. Denting is a small to moderate !
contributor to the frequency of leakage (14 out of 140) and the largest single !
contributor (31%) to tube plugging rates. Although denting led directly to the tube rupture event at Surry 2, the potential for future occurrences of this !
type have been significantly reduced. This is because the precursor conditions !
that led to the Surry rupture are generally well understood, and licensees would ;
be expacted to recognize these conditions and to take appropriate corrective j action in a timely manner (see Section 2.4.3). In addition, plant-specific denting criteria have been established for plants that have experienced ;
extensive denting. Thus, the benefit of this potential industry action as a ger,eric requirement may be relatively low in terms of reduced frequency of }
rupture and of significant releases compared with other potential industry L 1 actions considered by the staff.
(
- The SG0G conunented (letter, September 30, 1982) that inspection for denting is [
] justified; however, (1) denting inspection requirements do not belong in the [
i Technical Specifications and (2) the requirements proposed are too broad. The !
l scope of inspection should really be based on the progression of denting. If ;
i denting has been arrested, a %% spread inspection upon finding a few dentid ,
i tubes is not warranted. TVA commented (letter, October 13, 1982) that denting l inspectiohs should be performed only on those tubes that would not allow passage !
! of the standard ECT probe. Alabama Power Company commented (letter, Jant.ary 11, !
, 1983) that it supports some type of inspection for denting but that further j evaluation is necessary before criteria are issued to licensees. !
NUREG-0844 2-20 t
i
2.2.3.4 Conclusions Generic implementation of generic denting criteria would not be expected to result in a significant reduction in SGTR frequency, core melt risk, or in the probability of significant non-core melt releases. However, as is discussed further in Section 4.2.2, the availability of generic denting criteria could result in a net cost savings to the NRC in terms of future review effort. There-fore, the staff will undertake further study and development of generic denting criteria as a staff action as discussed in Section 4.2.2.
2.2.4 Steam Generator Inservice Inspection Interval As discussed below, the potential industry action concerning steam generator inservice inspection intervals has been dispositioned as a staff recommended action.
l i
2.2.4.1 Staff Recommended Action The maximum allowable time between eddy current inspections of an individual steam generator should be limited in a manner consistent with Section 4.4.5.3 of the Standard Technical Specifications and, in addition, should not extend beyond 72 calendar months.
2.2.4.2 Basis for Initial Consideration The current Standard Technical Specifications (STS) and many plant Technical Specifications allow the regularly scheduled inspections to be limited to one steam generator on a rotating schedule if the results of the previous inspections indicate that all steam generators are performing in a like manner. The current STS inservice inspection frequencies require inspections at intervals from 12 to 24 months which may be extended up to 40 months if two consecutive inspections result in all ins ection s results falling into the C-1 Category or if two con-secutive inspections indicate that previously observed degradation has not continued and no additional degradation has occurred.
For two , three , and four-loop plants, this could result in an interval of 80, 120, and 160 months, respectively, between required inspections on an individual steam generator. Operating experience indicates that steam generators at a given facility do not necessarily perform alike, and thus inspection results for one steam generator are not necessarily representative of the condition of each other steam generator. Although the steam generators may be performing in a like manner for some period uf time following initial startup of the plant, new degradation mechanisms may develop that affect the steam generators in a non-uniform manner. To make the maximum allowed operating interval more con-sistent with actual degradation rates that have been observed in the ( bid, the staff initially proposed that the maximum interval be limited to 48 meiths.
This was subsequently increased to 72 months as discussed below.
2.2.4.3 Value-Impact SAI estimates on the basis of plant data that steam generator inspection inter-vals average between 24 and 36 months and that implementation of a 48-month maximum limit on inspection intervals would impact only a few plants, approxi-mately SL Beyond an initial adjustment of inspection intervals that might be NUREG-0844 2-21 l
.___. __ - .. - .- .- . . -- .. =. . . .
)
necessary at a few plants, generic implementation of this action would have a very small effect on the number of inspections being performed. SAI concludes,
, therefore, that although it is difficult to assess the exact cost and occupa-tional exposure impacts and benefits associated with this potential industry
- action, they appear to be quite small.
1 After further evaluation, the staff has concluded that a 72-month maximum j
inspection interval would be consistent with the initial objective of precluding
, excessively long inspection intervals ranging up to 160 months and would further 1
reduce potential cost or occupational exposure impacts. ,
a 1 The potential reduction in the baseline SGTR frequency, given generic implementa- I
) tion of this recommended action, was not specifically quantified by SAI, but is believed to be small relative to the potential reductions associated with the staff recommended actions for prevention and detection of loose parts and foreign objects (Section 2.1) and improved secondary water chemistry control and condenser ;
inspections (Sections 2.5 and 2.6).
1 j Industry comments generally supported a maximum 72-month inspection interval. l 4 SG0G commented there is general agreement with the proposed change to the maxi- !
mum inspection interval (letter, August 25, 1983). i t
2.2.4.4 Conclusions i A maximum 72 month inspection interval per steam enerator reflects accumulat.ed l operating experience, is consistent with good eng neering judgment regarding !
] the need for periodic inspections as part of an effective program to ensute l
! steam generator tube integrity, and involves minimal adverse impacts; thus, it '
has been incorporated as a staff recommended action.
l 1 2.2.5 Inspections Following Shutdown for Repair of Leakage i l
l 2.2.5.1 Potential Industry Action ;
The current STS and many plant Technical Specifications do not require that i
- unscheduled inspections be performed in the event of tube leakage unless the !
leak rate exceeds the allowable leak rate limits in the plant Technical Specifi- l cations. A proposal was made that this should be changed so that unscheduled .
l inspections pursuant to the Technical Specifications should be conducted if the l j plant is shut down to repair a primary-to-secondary tube leak, regardless of ;
. whether or not it exceeds the Technical Specification leakage rate. ;
I j 2.2.5.2 Basis for Initial Consideration l 1
i i The current provisions of the Technical Specifications allow plants experiencing
, tube leakage below the limit to shut down, repair the leak, and return to service
- without conducting further inspections. The concern prompting consideration of 4 this potential industry action was that since the occurrence of leaks during l service may be indicative of new phenomena or accelerated rates of Jegradation, i the corrective action to be taken upon shutdown should include tube inspections j in addition to plugging the leaking tubes.
a
- i !
1 i 1 l l NUREG-0844 2-22 {
! i i .
- r
2.2.5.3 Value-Impact SAI estimated that a generic requirement incorporating the subject potential industry action would result in a 17% increase in the number of required inservice tube inspections. ORE resulting from tube inspections would likewise increase by 17%. Economic costs would exceed 17% since a fraction of such required inspections will be in response to forced outages and the inspection will extend the required outage period. SAI estimates that the leakage tests performed as ,
an industry practice to identify leaking tubes can usually be performed in e n ;
day; a complete ECT inspection requires 3.5 days. The additional 2.5 days would result in considerable cost (replacement power of $1.8M) whenever the inspcctic.n ,
is on the critical path to plant restart. Accordingly, it would appear that this change has the potential to eliminate incentives for licensees to shut down to ,
repair tubes that leak at less than the Technical Specification limits. SAI could not estimate potential averted costs and ORE associated with this potential action.
The staff finds that the tiet increase in the number of inspections as a result of uniform implementation of this potential action would be significantly less than 17% since licensees usually elect to perform some level of tube inspection, >
even when the leak rate limit has not been exceeded prior to plant shutdown.
This would tend to reduce not only the net Screase in cost and ORE associated with uniform implementation of the subject potential action, but also the potential benefits in the areas of averted cost and ORE.
Tube inspections during unscheduled plant outages to repair steam generator tube leaks serve to minimize the potential for tube rupture. Indeed as previously F noted, licensees usually elect to inspect tubes during unscheduled outages to repair tube leakage, although tube sample sizes are sometimes less than what is l called for by the subject potential industry action. Even if not inspected at that time, the steam generators would eventually be inspected anyway, pursuant to current requirements, at the next scheduled inspection. SAI estimates that steam generator inspection intervals average 24 to 36 months, and do not extend beyond 48 months in the vast majority (95%) of the cases. In addition, operating limits on allowable primary to secondary leakage (see Section 2.8) minimize the potential for SGTRs during the interim period prior to the next scheduled inspection. SAI could not quantify the expected reduction in SGTR frequency associated with generic implementation of this potential action; however, the staff believes the potential benefit to be relatively small compared to the ,
potential benefits associated with the staff recommended actions for prevention 1 and detection of loose parts and foreign objects (Section 2.1) and improved secondary water chemistry and condenser inspections (Sections 2.5 and 2.6).
SGOG stated (letter dated September 30, 1982) that full-scale ECT inspections !
should not automatically be required during an outage to repair any leak on the j bases that (1) outages to repair tube leaks are all critical path time, and ,
! requiring a full-scale ECT during these outages penalizes a utility for con- j
> servative operating practice; (2) if the leak is associated with known generic l degradation, extensive inspection is not necessary; and (3) if the leak is not ;
associated with a known generic type of degradation, a minimum inspection should focus on whether other tube leaks are imminent. Similar comments were received from other industry sources.
l 4
NUREG-0844 2-23 l
,-e-- , - - . , +- - -
- - - , - - - - ,_- -,-m-- .,-------r-r - n- . ,,, --,-,r
2.2.5.4 Conclusions In consideration of the fact that (1) tube inspections are usually performed by industry during unscheduled shutdowns, even in cases where the Technical Specification leak rate limit has not been exceeded, (2) generic and uniform implementation of the subject potential action would not be expected to result in any significant further reductions in the industry baseline SGTR fraquency, and (3) potential cost and ORE benefits impacts could not be quantified, the staff concludes that the subject potential industry action should not be included as generic staff recommendation. The staff does believe, however, that licensees should continue to take appropriate action not only to repair observed leakage, but also any additional actions including inspections as needed to provide assurance that non-leaking tubes will not rupture before the next scheduled outage. In cases where the leak rate limits have not been exceeded, licensees should consider such factors as the nature of the degradation mechanism involved, i "leak-before-break" characteristics associated with subject degradation mechanism,
) and the planned date for the next scheduled inspection in determining the need for an unscheduled inspection. "Leak-before-break" characteristics refers to j the potential for the subject degradation mode to lead to tube rupture during j normal or postulated accident conditions.
- 2. 3 Improved Eddy-Current-Test Techniques
) 2.3.1 Potential Industry Action z To ensure that all licensees use appropriate methods and equipment in conducting !
eddy current tube inspections, a proposal was made for licensees to include several such practices in their eddy current test (ECT) program as follows: !
(1) Eddy-current testing techniques or data evaluation techniques which are capable of eliminating tube support plate, tube sheet, denting, or other ,
similar unwanted signal interferences and capable of discriminating among i multiple defects shall be used. t (2) Eddy-current prob u that can perform both absolute and differential coil inspections shall be used. Separate probes may be used to implement this l l dual capability. f l (3) Eddy-current data from both the differential and absolute channels soall j be evaluated as part of the overall data evaluation program. !
(4) In addition to the calibration standards required by Article IV-3200 of ;
] Section XI of the Boiler and Pressure Vessel Code of the American Society (
of Mechanical Engineers (ASME Code), a standard shall oe used with simulated j wear or fretting-type flaws to ensure a conservative interpretation of signals for which fretting or wear may represent a possible source of signals. Typical examples include absolute signals over a significant axial length of the tLbe, absolute signals for which there has been little or no corresponding differential signal, and signals that can reasonably be inferred as possible fretting or wear flaws on the basis of experience
] (e.g., indications at the tube-to-baffle plate inteisections in the preheater
, sections of Westinghouse Model D steam generators). The simuisted flaws !
! shall be sufficiently tapered and smooth so that they produce little or no {
j differential signal. 1 i
i NUREG-0844 2-24 i il l
l
- - - - - - - - - - - - - - - ,,n------- . . - - - - , - . - , , - - . - . - - - - - , - , - . , --,-----,- -r-,, ,~:- . n,, ,- , ,-
2.3.2 Basis for Initial Consideration ,
This potential industry action was proposed with the objective of improving the reliability and accuracy of CCT programs for purposes of detecting degradation in tubes. Regarding item (1) of the proposed industry action, laboratory exper-iments, and field experience have demonstrated the superiority of multiple-frequency ECT and other techniques to discriminate defect signals from background noise or interference signals. Sources of background noise include structural supports, tube diameter or geometry variation, and conductive deposits on the tubes. Multi-frequency or other equivalent techniques are essential to screen out these interference signals and thus permit accurate evaluation of the condi-tion of steam generator tubing.
Regarding items (2), (3), and (4) of the potential industry action, eddy-current inspections, defect detection, and interpretation capabilities can be enhanced through the use of absolute mode inspections in addition to differential mode inspections. A vall-thinning flaw that is gradually tapered at its edges, as may be the case for fretting-type wear defects, may not produce a detectable signal on the differential channels. Such a fretting-type wear flaw will generally produce a signal on the absolute channels. A tapered, localized radial fretting or wear standard, as opposed to the hole standards specified in the ASME Code, may be necessary to correctly interpret the amplitude of the signal.
The tube that ruptured at Ginna in January 1982 as a result of a long fretting wear defect had been inspected in April 1981 by both the differential and absolute modes. This tube exhibited no differential signal in April 1981 but did exhibit an absolute signal approximately 5 inches long, which was not recorded at that time. This April 1981 signal can be interpreted as less than a 20% indication using the calibration hole standard as specified in Section XI of the ASME Code. However, this signal can be interpreted as a slightly greater than 40% indication if a fretting or wear calibration standard is used, which is greater than the 40% plugging limit for Ginna. An evaluation of the absolute signal in April 1981 using a fretting or wear standard may have resulted in the tube being plugged before the wear had proceeded sufficiently through the tube wall to cause the rupture.
1 i Eddy-current testing has advanced significantly in recent years. Several aspects of this potential industry action, namely multifrequency ECT and use of absolute and differential mode inspections are already in widespread industry use. In addition, the industry has developed advanced technology, including specialized l eddy-current prebes and data evaluation techniques, that have been used at l numerous plants in response to specific tube degradation problems that have i challenged the detection capabilities of conventional ECT techniques.
2.3.3 Value-Impact The SAI analysis indicated that the ORE and economic impacts are small and are more than offset by positive ORE and economic benefits. As discussed above, improved ECT procedures could potentially have averted one of the four steam generator tube rupture events to date; namely the tube rupture event at Ginna.
Thus, the staff estimates that improved ECT techniques could potentially reduce the probability of rupture by up to 25% if implemented as a stand-alone action.
Implementation of the staff recommended actions identified in Sections 2.1, l
NUREG-0844 2-25
2.5, and 2.6 is expected to reduce the frequency of rupcures by a tot.al of 64%
as an industry average. Therefore, the additional potential incremental reduc-tion in rupture frequency from improved ECT techniques is about 10%. Thus, improved ECT techniques are considered to have medium potential (as defined in Table 3) for reducing the probability of significant but less-than-core-melt releases.
SG0G stated the following (letter, September 30, 1982): ECT techniques to be applied to specific steam generators should be established in 71 ant-specific programs prepared by utilities and submitted to 1RC rather that, through generic requirements uniformly applied. The factors that determine the proper ECT tech-nique for a specific application are plant and sometimes steam generator specific. ECT is currently evolving. It is uadesirable to require 'icensees to use more complex techniques than are necessary or to discourage experiments with new techniques as they are developed. SG0G also stated that use of a new "wear standard" should not be required. Special standards, where warranted, should be handled by plant-specific submittals. Development of new generic standards should be the result of actions by the appropriate ASME Section XI i Code Committee. l Tennessee Valley Authority (TVA) recommended deletion of the portion of the ECT techniques proposal addressing the provision of ECT probes to provide the l capability to perform both absolute and differentiti coil inspections (letter, October 13,1982). TVA based its recommendation on its finding that the pro-posed actions should address the capability of the inspection system to detect types of defects but should not specify the method or technique to be used since a number of different techniques now exist and others will probably be developed in the future which may constitute an advancement in N0E techniques.
These comments are representative of comments received from other industry sources.
2.3.4 Conclusions Use of appropriate ECT techniques as needed to detect defective tubes is a key element of an effective program for ensuring steam generator tube integrity. As noted by the SGOG, however, the factors that determine the appropriate ECT technique to be employed are often case-specific depending upon such factors as defect volume, orientation, and location and the presence of noise. Therefore, licensees should be aware of the limitations of the techniques they are employing such that more sensitive techniques are utilized as necessary in response to specific problems.
The staff has concluded that additional consideration of improved ECT techniques !
as a generic issue is warranted, but as suggested by the SGOG, this effort should !
be performed in parallel with ongoing ASME Code Committee activities to develop ,
updated ECT procedures for incorporation into the ASME Code Section V for NDE f and Section XI for ISI. Therefore, the staff has dispositioned '...e improved ECT issue as a staff action item which is addressed in Section 4.2.3.
NUREG-0844 2-26 ;
2.4 Upper Inspection , Ports t 2.4.1 Potential Industry Action i i
A proposal was made that for all PWR applicants, upper inspection ports.should t be installed before issuance of an operating license so that the upper support !
plates and inner row U-bend tubes can be visually inspected. Operating plants ;
were not included within the scope of this potential industry action based on a consideration of the ORE and economic impacts of installing poets in an ,
operating plant's steam generator.
2.4.2 Basis for Initial Consideration l A primary motivation for initial consideration of this potential industry action !
! is that upper inspection ports provide a direct means for monitoring the state i of deformation in the upper support plate which is caused by the buildup of a hard corrosion product (magnitude) in the annular regions between the support plate and tubes. This buildup of corrosion product is generally referred to as denting (NUREG-0886). As denting becomes trore advanced, it causes rectangular
- flow slots in the support plate, located between the legs of the inner row tube l U-bends, to deform into an hourglass shape. This "hourglassing" in turn causes the legs of the inner row tube U-bends to displace toward each other, thereby i leading to increased tube stress at the apex of the U-bend. This increased
stress creates the potential for stress corrosion cracking and ultimately rup- I J ture of the inner row tubes. This mechanism was the cause of the SGTR evont
, at Surry Unit 2 in 1976.
l Steam generators are generally equipped with inspection ports located between l the tube sheet and lower (first) support plate. These existing inspection ports
-(
provide only limited means for monitoring deformation of the upper support plate, i i including hourglassing of the upper sup p.at plate flow slots, due to the fact )
- that physical access and/or line of sight from the existing inspection port to !
l the upper support plate flow slot is through the flow slots of the lower five or i j more support plates.
{ Remote optical or camera techniques could be adapted to provide for improved !
I monitoring of the upper support plate from the existing inspection port. l However, experience indicates that centing generally tends to cause hourglassing l l in many of the lower support plate flow slots before it occurs in the upper i support plate. Thus, as denting becomes more advanced, it may not be feasible !
j to visually monitor upper support plate deformation directly.
1 Installation of upper inspection ports, located at one end of the tube lane !
l just above the upper support plate, would enhance the capability to effectively '
i and directly monitor the condition of the upper support plate, particularly for j any signs of early hourglassing of the upper support plate flow slots which I would indicate the need for taking appropriate preventive actions to preclude rupture of the inner row U-bends.
I j Apart from providing a means for visual monitoring of the upper support plate, i
i upper inspection ports also provide access to the inner row tube U-bends to facilitate the removal of U-bend tube specimens for laboratory investigation, j Laboratory examination of removed U-bend tube speciuens has been performed for i
I.
) NUREG-0844 2-27 i
i
a few plants to assist the licensees in their assessment of the nature and causes of stress corrosion cracks in the U-bends.
2.4.3 Value-Impact The above discussion provides the initial basis for consideration of this I
potential industry action. The staff's value-impact analysis indicates, however, i that generic implementation of this action would not be expected to reduce the
- frequency of SGTRs, nor the probabilities of core melt or significant non-core
- melt releases. The precursor conditions which ultimately led to the tube rupture j at Surry Unit 2 are now well understood. Alternative means are available for j detecting the onset of these precursor conditions, permitting timely implementa-
, tion of preventive and/or corrective measures. Preventive measures could i include the plugging of inner row tubes which are the tubes most susu ptible to I denting-induced rupture.
} Eddy current testing, tube gauging, and profilometry techniques are capable of
- detecting very early stages of denting, before deformation of the support plate
- flow slots would even be observable. As denting becomes more advanced, the i state of deformation of the lower support plates can be directly observed frcm the existing "lower" inspection ports. Experience has shown that any significant I "hourglassing" of the upper support plate flow slots is generally accompanied by l flow slot hourglassing of the lower support plates which can be monitored from :
the lower inspection port. l
[
Once early signs of denting have been detected, installation of upper inspection ?
- ports is one of the options licensees will have in terms of ensuring that the !
1 appropriate preventive measures are implemented on a timely basis. However, it 6
is the staff's judgment that eddy-current testing, tube gauging and/or profilo-j metry, and lower inspection ports can also be effectively utilized for this
! purpose. In addition, in view of the industrywide trend toward more effective secondary water chemistry programs, and the ongoing industry development of remedial measures (e.g., boric acid treatments) to control and/or reduce rates of denting, it is unlikely that most PWRs will be confronted with severe denting problems.
] l l Costs for installing upper inspection ports have been variously estimated to
- be in the $100,000 to $200,000 range by the SG0G (letter, September 30, 1982) I i and to be in the $300,000 to $450,000 range by SAI. There is no ORE impact i j associated with the installation of upper inspection ports in a preoperational ,
steam generator. There are no offsetting cost or ORE benefits except for those i
! plants where licensees install such ports in response to specific problems.
l SAI estimates the ORE associated with installation of these ports in an opera- i 1 tional steam generator to be about 100 person-rem. As previously noted, however, l j there is little evidence that there will be a widespread need for such ports in
' the future.
- SGOG commented that a generic requirement for upper inspection ports is not t i warranted for operating plants or plants under construction for the following '
reasons (letter, September 30, 1982):
l 4
(1) Visual inspection of the uppermost support plate or inner-row U-bends '
- ' is not normally necessary. Denting could bi detected earlier lower in the tube bundle and can be adequately characterized by eddy-current ;
l NUREG-0844 2-28 l 1 (
-- ----,--v . - , . - - - , , . - - - - - - n ,- ---r--n-,--n-----nv ----,,-----,----,,re-- --m e-,----e - - . ,---,-------,-r-
inspectior or profilometry of tubes. Tube cracking cannot be detected visually from the OD of tubes even if it does occur in the inner-row U-bends. Again, inspections from the tube 10 are more useful. !
(2) Ports are useful for removing sections of tubes or other steam generator internals to determine causes of degradation. For a given problem, sections are required from only a few steam generators.
Moreover, samples of U-bends have already been removed from steam generators +.o evaluate tube cracking, and upper support plate samples have already been removed to evaluate denting. -
l (3) Tube samples may well be required from selected steam generators in the future to evaluate other types of degradation. However, the area of interest will not necessarily, or even likely, be the uppermost supports. Additional ports installed now may well be in the wrong place or of the wrong size to be useful later.
I (4) In general, it is desirable to minimize the number of penetrations in i vessels and the number of mechanical closures that may leak. Moreover,
- adding a penetration provides the opportunity to introduce foreign
- objects. These considerations argue against adding penetrations
- unless there is a supportable use for them.
(5) The cost of adding ports in the field is estimated to be between
$100,000 and $200,000 per steam generator. As noted above, there are ',
no particular generic benefits. Experience has shown that ports can be added later, if, when, and where they are needed.
t (6) The need for additional inspection ports in a steam generator should be evaluated individually for each case. Additional ports should not ;
be required unless there is a demonstrated need, e.g., for use in solving or determining the cause of a problem.
TVA commented that generator internals can best be monitored by one or a !
4 combination of the following methods (letter, October 13,1982): (1) eddy-current i
test data analysis, (2) profilometry data analysis, and (3) flow slot measurements
, and remote IV camera inspection. If a sample has to be removed at an operating plant that has preinstalled inspection ports, the probability is low that the existing ports would be of the correct size and at the correct location.
1 WPSC commented that upper inspection ports constitute a costly investment that l l is not warranted. WPSC also commented that ports are not necessary if secondary water chemistry and condenser inspection programs are in effect as part of a i total management control program (letter, October 4, 1982).
l These comments are representative of comments from other industry sources and are generally consistent with the results of the staff's value-impact analysis.
i 2.4.4 Conclusions i' Generic installation of upper inspection ports on preoperational steam generators would not be expected to produce reductions in SGTR frequency, the probabilities I of core-melt and significant non-core-melt releases, or cost. Although imple-l mentation of thf s action could provido ORE benefits of about 100 person-rem in NUREG-0844 2-29
- ~ _ _ - - - - . . . . . ~ _ - _ . - -. - - -
cases where licensees later decide to install upper inspection ports in an operating steam generator, such benefits will likely be limited to a small number of plants. Thus, potential ORE reductions do not appear to be an important generic consideration. For these reasons, the staff concludes that this potential industry action should not be incleded as one of the generic
- staff recommended actions.
i 2.5 Secondary Water Chemistry Program Based on the staff's value-impact evaluation, this potential industry action has been dispositioned as a staff recommended action, j 2.5.1 Staff Recommended Action
! Licensees and applicants should have a secondary water chemistry program (SWCP) l to minimize steam generator tube degradation. The specific plant program should
, incorporate the secondary water chemistry guidelines in SGOG Special Report i EPRI-NP-2704, "PWR Secondary Water Chemistry Guidelines," October 1982, and l should address measures taken to minimize steam generator corrosion, including
! materials selection, chemistry limits, and control methods. In addition, the 2
specific plant procedures should include progressively more stringent corrective
- actions for out-of-specification water chemistry conditions. These corrective
) actions should include power redictions and shutdowns, as appropriate, when
' excessively cc rosive conditiont exist. Specific functional individuals should be identified as ' aving the responsibility / authority to interpret plant water I 1
chemistry informa. m and initiate appropriate plant actions to adjust chemistry, [
- as necessary. ;
i i 2.5.2 Basis for Initial ~onsiderat w l 3
} Secondary-side corrosion problems with steam generators have affected a large !
] number of PWRs to date. A description of the kinds, causes, and extent of l
! corrosion problems which have been experienced have been documented in j NUREG-0886 and NUREG-1063. These corrosion problems have caused substantial ;
- cost to the industry as a result of needed steam generator maintenance, repair, l 1 and replacement activities, and increased plant unavailability, and have resulted [
- in signifIcant occupational radiological exposure to workers. Two SGTR events, ;
) Point Beach 1 in 1975 and Surry 2 in 1976, are directly attributable to secondary-
{ side corrosion problems. NUREG-0651 provides a description of these events.
! Improved secondary water chemistry control has been recognized by the industry
- in general and by the staff as an important factor in reducing corrosion in '
! steam generators. The referenced SGOG guidelines were prepared by the Steam
! Generator Owners Group Water Chemistry Guidelines Committee and represent a consensus of a significant portion of the industry for state-of-the-art ;
I secondary water chemistry control. '
1 .
l 2.5.3 Value-Impact !
See Section 2.6.3 l 2.5.4 Conclusions ;
l See Section 2.6.4 I
- I
! NUREG-0844 2-30 !
1 L_- . , - - _-
l i
2.6 Condenser Inservice Inspection Program l Based on the value-impact evaluation in Section 2.6.3, the following potential !
industry action has been dispositioned as a staff recommended action. ;
I 2.6.1 Staff Recommended Action i Licensees and applicants should have a condenser inservice inspection program !
that addresses the following: i (1) procedures to implement a condenser inservice inspection program that will be initiated if condenser leakage is of such a magnitude that a power ,
reduction corrective action is required more than once per three-month '
period ,
(2) identification and location of leakage source (s), either water or air l (3) methods of repairing leakage f i
j (4) methodology for determining the cause(s) of leakage (
j (5) a preventive maintenance program l l 2.6.2 Basis for Initial Consideration i
- Condenser operating experience was summarized in EPRI-NP-481, "Steam Plant !
l Surface Condenser Leakage Study." and EPRI-NP-2062 "Steam Surface Condensers .
l Leakage Study Update." Tne studies assessed the leakage integrity of the l i condenser and the reliability and operability of the downstream components to I the contamination introduced from the recirculation water. Air and water l inleakage through the failed condenser tubing can contaminate the condensate,
- feedwater, steam generator water, and steam, which, in turn, degrades the ;
structural integrity of the steam generator tubes, reheater tubes, turbine, l'
- and other components in the cooling system. The tolerance to a given leak in I
a given plant is a function of the impurity content of the recirculation water, the presence or absence of condensate demineralizers, and the materials in the l
- condenser and feedwater trains. Many undesirable contaminants enter the !
i secondary system through condenser leaks, and condenser integrity is essential l
! to maintaining good water chemistry. Condenser tubes in the impingement, j 1 condensing, and air-removal sections of the condenser are subject to differeat l I failure rates and failure mechanisms. To some extent, failures resulting from l
{ vibration are related to operating conditions. Tubes in the impingement sector l i are susceptible to erosion by steam and to severance by missiles, tikewise, :
l when ammonia-sensitive alloy tubes are located adjacent to the air-removal J section of the condenser, a high incidence of ammonia-induced failures can be 1
anticipated with all-volatile treatment (AVT) coolant water control. Localized concentrations of ammonia can be orders of magnitude greater in the vapor phase !
than in the bulk condensate. A high concentration of ammonia in the condensate l
)
may induce failure of copper alloys by stress-corrosion cracking.
) Air inleakage into the condenser can cause corrosion of copper-containing i l condenser tubes and feedwater heater materials. When ammonia is also present, !
I stress-corrosion cracking of copper-based alloys, such as aluminum brasses or i j admiralty bronzes can also occur. The copper-nickel alloys are more resistant j l i NUREG-0844 2-31
to ammonia cracking, but can still be a source of copper ions. Copper ions entering the feedwater from these sources can trigger denting reactions in the steam generators. The utilities are taking st2ps to eliminate the use of ammonia-sensitive alloys from the air-removal section of the condensers and to replace them with more ammonia-resistant alloy tubing. Where denting is a concern, steps are being taken to eliminate all copper alloys from the condensers, by using such materials as stainless steels (for fresh-water service), Allegheny-Ludlum 6X, or titanium. Clearly, maintenance of a tight condenser will eliminate the primary source of the oxygen and chloride ions in the system and help to control denting in the steam generators.
2.6.3 Value-Impact This subsection addresses the staff's value-impact evaluation of staff recom-mended actions pertaining to SWCP and condenser inservice inspection programs (CISIPs). Because an effective CI. SIP is an integral element of an effective ,
SWCP, the values and benefits estimated for improved SWCP includes those benefits i
to be derived from an improved CISIP.
SAI divided existing plants into three groups to evaluate the potential effect of improved SWCP on the probability of steam generator related forced outages, i ruptures, and tube plugging, and steam generator replacement.
(1) Severe category plants with significant corrosion-induced degradation of the steam generator tubes (2) Clean category plants with little corrosion-induced degradation of the steam generator tubes (3) Medium category - the rest of the plants !
SAI estimated that approximately one-sixth of the operating units were considered to be in the severe category, one-half in the medium category, and one-third in the clean category. Units were grouped according to their history of plugged
, tubes as of 1982. Units in the severe category had multiple hundreds of tubes plugged; units in the medium category had around 100 tubes plugged; and units in I
the clean category generally had in the low tens of tubes plugged (after 6 to 8 years of service). c I
SAI estimated the following percentage reductions in SGTR, forced outage, and tube plugging frequencies associated with implementation of improved SWCP:
l EVENT FREQUENCY REDUCTIONS 1 Industry Sovere [
Event Average Plant Plant l ,
Corrosion-related SGTRs 38% 66% l All SGTRs 19% 33% '
i Forced outages 46% 68%
> Tube plugging 71% 75% [
, i The estimated reductions in overal) SGTR frequency would result in corresponding !
j reductions in the probabilities of core-melt and significant non-core-melt '
NUREG-0844 2-32 i l
i releases. Using the value-impact criteria defined in Table 3, these correspond l to a low benefit (industry average plant) and a low to medium benefit (severe l plant) in terms of reduced probability of core melt, and a medium to high ;
benefit (industry average and severe plants) in terms of reduced probability of i significant non-core-melt releases. '
The SAI analysis indicates that the occupational radiation exposure that would result from implementing the staff-recommended SWCP would be negligibly different from that incurred in present secondary water chemistry testing activities. ,
SAI estimates that up to 7500 person rem / plant can be avoided ever an assumed ;
24 year remaining plant lifetime for plants in the severe category as compared !
l to an estimated 1060 person-rem for plants in the medium category due to reduced !
or avoided SG inspection, maintenance, repair, and replacement. Thus, the staff -
recommendations pertainir,g to SWCP and CISIP appear to offer significant benefits in the area of reduced ORE.
SAI estimates the average plant lifetime costs to implement the proposed SWCP :
j to be $1.3M. Implementation of an improved CISIP would not add significantly
- to this cost. Avoided lifetime costs are estimated to range from $1.3M to l l
$12M for plants in the medium category to between $210M and $260M for plants in !
i the severe category. Thus, implementation of an improved SWCP and CISIP appears to be very cost effective.
i The SGOG stated that there is general agreement on the desirability of a SWCP l 1 (letters, September 30, 1982 and August 25, 1983). The SG0G also noted that !
j theEPRIguidelinesforSWCP(whichhavebeenincorporatedintothesubject !
- staff recommendation) are indeed guidelines and that strict adherence will not ;
) guarantee that steam generators will be corrosion free and also that exceptions l l to the guidelines may not lead to corrosion. Moreover,theguidelinesaresubject t to change as experience and more information are obtained. The guidelines were prepared by an SG0G committee for consideration and use by SG0G members, some
(
I j of whom disagree with some sections because of plant-specific considerations. ;
i i
Application of the guidelines must be flexible enough to consider plant-specific design features, operating requirements and philosophy, and steam ger.erator l history. The staff acknowledges the validity of these comments and concludes >
i j that they are consistent with the staff conclusions in Section 2.6.4.
i )
j The SG0G also commented that generic requirements for a CISIP are not warranted I and should not be a licensing condition on the following bases:
J j (1) Condenser inspection and leakage are not safety issues.
l (2) Maintenance of secondary water chemistry provides utilities with a strong j incentive to prevent excessive condenser leaks.
a l The SG0G also stated that utilities should remain free to establish the condenser 1 maintenance plans best suited for their plants. The frequency, extent, and type j of inspections can be based on:
1
) (1) history of condenser operation 3
(2) the presence of on-line leak surveillance techniques, e.g., cation conductivity monitors I
l
] NUREG-0844 2-33 I
u _ _ _ _ _ _ _ __ __
(3) plant design features, e.g., the role of po',ishers during cooling water inleakage (4) economic considerations Section 2.6.2 addresses the importance of maintaining the leak-tightness of the condensers from the standpoint of ensuring adequate secondary water chemistry control. The staff continues to believe that the subject CISIP should be imple-mented to ensure the leak-tightness of the condensers as part of a successful SWCP.
4 Other industry sources, including Duke Power Company (letter, January 26, 1983),
Florida Power Company (letter, January 10, 1983), Alabama Power Company (letter, January 11, 1983), and Wisconsin Public Service Company (letter, October 4,1982) stated their agreement with the value of a secondary water chemistry program and i condenser ISI program. However, there were several comments that such items should not be the subject of regulatory requirements. WPSC stated that (1) the SWCP does not warrant a license condition, (2) a firm management commitment such as that at WPSC can ensure a successful control program, and (3) it is prudent management to have an operational chemistry program, which management tupports, to protect the company's investment.
- WPSC's comments on the condenser inservice inspection program are similar to its comments on the SWCP with the addition that, because of the costs associated '
with a forced outage to repair a steam generator, management should consider implementing this requirement on its own; no NRC requirement is necessary. A similar comment was provided by Alabama Power Company. The staff's response to these comments is the same as the response to the SG0G CISIP comments. i Westinghouse commented that the costs associated with reduced power operation, l
, including shutdown, while chemistry is being corrected should be considered (letter, February 15, 1984). Further, the SAI estimate of $1M for hardware ,
j update does not consider that many plants will require significant condenser repair or replacement. The Power Authority of the State of New York (PASNY) I has commented that while it is endeavoring to reduce the concentrations of l impurities (at Indian Point 3) to levels within those of the SG0G guidelines,
) it is apparent that major plant modifications will be required to reduct all '
parameters to the desired levels. PASNY concludes that the costs for these modifications should be considered (letter, February 15, 1983).
The staff acknowledges that these costs have not been directly considered.
4 However, those plants that must perform frequent pcwer reductions or shut downs to correct out-of-specification secondary water chemistry, or that must perform (
extensive condenser modifications in order to avoid power reductions or shutdowns,
! are precisely the plants that stand to benefit the most in terms of avcided !
l economic cost, ORE, and tube ruptures. The staff further notes that for these
- plants, condenser modtfications should and are likely to be taken eventually as l a matter of economic necessity. The staff concludes that direct consideration ;
) of these costs would not have a significant effect on its value-impact findings -
l concerning the cost effectivenss of an improved SWCP and CISIP.
j :
NUWEG 0844 2-34 t
1 1
2.6.4 Conclusions The potential industry actions concerning secondary water chemistry control (Section 2.5.1) and condenser inservice inspection programs (Section 2.6.1) can provide a 19% to 33% reduction in overall SGTR frequency, a low to medium benefit (as defined in Table 3) in terms of reduced prooability of core-melt, a medium to high benefit in terms of reduced probability of significant non-core-melt releases, and substantial benefits,1060 to 7500 person-rem, in terms of avoided occupational radiological exposures. In addition, these actions are cost effective. The staff concludes that these potential industry actions should be incorporated as staff recommended actions.
l 2.7 Stabilization and Monitoring of Degraded Tubes i i 2.7.1 Potential Industry Action A proposal was made that licensees should conduct a study to develop criteria and procedures for plugging tubes which contain provisions for (1) the monitoring of further degradation in plugged nonleaking tubes for which the rate of further degradation cannot be reliably predicted and (2) the stabilization of degraded tubes that may be subjected to progressive degradation mechanisms having the potential to cause the tube to sever, subsequently damaging adjacent tubes.
2.7.2 Basis for Initial Consideration A plugged tube may continue to degrade after plugging so that it can become completely severed and subsequently cause damage to adjacent tubes. The most 4
important types of degradation, in this regard, are those that can potentially -
affect the entire circumference of the tube. Some of the more obv h t examples '
are fatigue-induced circumferential cracking and fretting wear frort f'sw-induced vibration. Monitoring the plugged tube's integrity would provide a warning of further degradation before severance or would prevent severed tubes from dcmaging neighboring tubes.
t Implementation of techniques beyond the current conventional inservice inspection practices to monitor the integrity of plugged tubes is less important for well understood and predictable degradation mechanisms than it is for newly encountered or unpredictable degradation mechanisms. An example of a recently encountered degradation mechanism that may benefit from monitoring is the tube fretting wear ,
from flow-induced vibration in the preheater section of Westinghouse Model 0 ;
steam generators. Such wear has led to a tube leak in the Ringhals 3 plant. :
i Operators at the Ringhals 3 plant chose to install limited leakage plugs in l tubes requiring plugging. Plant operators found that the tube wear rate and modes of plant operation causing excessive wear were well defined. However, the
- use of the limited leakage plugs was intended to provide an earlier indication thanm,Qhtotherwisebegainedofacceleratedtubewear. (Corrective inodifica-tions to the steam generator feedwater inlet sere subsequently installed at Ringhals 3 and at other Westinghouse Model 0 units to eliminate the source of the vibration problems.) The use of limited leakage plugs would also appear to be of interest in cases where there is uncertainty about the continued integrity of plugged tubes, s
)
i i NUREG-0844 2-35 I
The implementation of techniques to stabilize tubes becomes more important in cases in which the potential exists for complete and/or rapid circumferential cracking. Such degradation mechanisms may be driven solely by flow-induced vibration phenomena or by corrosion assisted by fatigue. An example of circum-ferential cracking propagated by fatigue because of vibration induced by high 4
fluid flows across the tubes has been encountered in the once-through steam generators (OTSGs) for tubes adjacent to the inspection lane in the vicinity of 4
the 15th tube support plate. It ;,os been the practice to stabilize such tubes with pluggable indications by inserting a rod inside the affected tube. The rod is attached to the plug, which is installed in accordance with conventional j practices, t
4 2.7.3 Value-Impact SAI estimates the cost of performing this study to be about $1H. Since much of the burden of conducting this study could be shared among owners groups, the costs per plant are likely to be substantially less than this figure, j Additional cost and ORE would accrue if these studies were to lead to installa-tion of stabilizing or monitoring devices based upon criteria derived fro- 5er>
studies.
I The SGOG comments (letter, September 30, 1982) are summarized as follows:
I t
(1) Of the large number of plugged tubes cited by the NRC, there has been only !
one isolated case of plugged tubes later causing damage to adjacent tubes. l This it because (a) there are few damage mechanisms that can cause a tube
, to sever where it is not rest *ained and then damage adjacent tubes and [
,' (b) when 5:ich mechanisms exist, they usually have been recognized and accounted for in corrective action:. Any proposals aimed at controlling ;
, a situation that is already rare need to be carefully focused and reviewed to ensure they will not create more problems than they are likely to j prevent.
i i
i (2) Monitoring the continued degradation of nonleaking plugged tubes may pro-i vide research data, but special actions to do so with the objective of
! preventing tube leaks are not warranted, i '
] (3) Limited-leakage plugs allow internal pressurization and slight internal !
j heating of a tube, conditions that support tensile stress and continued i j tube corrosion, respectively. Leaks from these plugs could not be '
distinguished from other small leaks. Use of such plugs would mean that some tubes might have to be plugged twice, with the attendant increase in !
radiation exposure. t t
(4) Stabilization of a degraded tube should be undertaken only when the damage !
! mechanism could cause the tube to sever in an unrestraired region and !
i damage adjacent tubes, and then only after the case has been individually j evaluated. j e J
(5) The reporting proposal is open ended in requiring identification of the l
- progressive degradation mechanisms "likely to occur" as well as those !
l that have ocurred. [
l t i
NUREG-0844 2-36 1
The SG0G comments are generally representative of other comments from the industry.
2.7.4 Conclusions The staff has considered the industry comments on this issue, In particular, the staff agrees that there is little evidence that severence of plugged tubes and resulting damage to adjacent live tubes is a widespread industry problem. l The Ginna tube rupture event did serve to point out the potential for large ,
foreign objects, in conjunction with high cross-flow velocities, to cause tubes to sever. Increased industry awareness nf the circumstances of the Ginna failure and of the need to have effective programs fer the detection and prevention of '
loose parts is expected to minimize the potential for further occurrences of this kind. In addition, the industry has, on its own initiative. installed tube i' stabilizing and monitoring devices in cases in which the potential exists for ;
flow-induced vibration to cause tubes to sever after they have been plugged. On the basis of these considerations, the St.:ff finds bere is insufficient basis [
for a staff position that the industry sidy this issue further. The staff will 3 continue to monitor industry practice in dealing with the severed-tube issue ,
and will take action on either a piant-specific or generic basis should it later i i
be determined to be appropriate. ,
J 2.8 Primary-to-Secondary leakage Limits j l Based on the value-impact evaluation in Section 2.8.3, this potential industry e I action is categorized as a staff recormended action. I l
1, 2.8.1 Staff Recommended Action f l All PWRs that have Technical Specifications (TS) limits for primary-to-secondary t
- leakage rates which are less restrictive than the Standard Technical Specification j j (STS) limits should implement the STS limits. ;
i I 2.8.2 Basis for Initial Consideration j The STS and many plant Technical Specifications limit the primary-to-secondary ,
) leakage through all steam generators not isolated from the reactor coolant system as well as the leakage through any one steam generator not isolated from the -
- reactor coolant system. These limits are based on two considerations. !
f
~
l (1) The total steam Generator tube leakage limit of 1 gpm for all steam ;
i generators ensures that the dosage contribution from tube leakage will be l j limited to a small fraction of 10 CFR Part 100 limits in the event of a ,
i single SGTR or steamline break. This limit is consistent with the assump- 1 j
tions used in the design-basis analysis of this accident.
l (2) The 500 gpd (0.35 gpm) leakage limit per steam generator provides added !
j assurance that ste.am generator tube integrity will be maintained in the !
- event of a postulated main steamline break (MSLB) or loss-of-coolant acci-l dent (LOCA). Permitting operation with leakage in excess of these limits I increases the potential that steam generatcrs may be vulnerable to tube j ruptures during postulated accidents such as an MSLB or LOCA.
I i
I NUREG-0844 2-37 l
L.,_-____ _ . _ _ _ _ _ - - _ _ _ _ - _ __ __a
The 500 gpd limit was originally developed on the basis of leak rate and burst test data for 0.875-in.-diameter x 0.050-in.-thick Westinghouse tubes, and has been shown to be a conservative limit for smaller diameter and smaller thickness tubes utilized in later model Westinghouse steam generators. The numerical
, value of the STS limit is intended to ensure that tubes with through-wall cracks which are leaking at less than the leak rate limit during normal operation have sufficient residual integrity to sustain postulated accident loadings without
, rupture.
The leakage rate limits provide a very important indication of the existence or rate of steam generator tube degradation. Experience has shown that some forms of degradation can develop in a period of time that is shorter than the routine inspection intervals or may be difficult to detect with current ECT techniques.
If such degradation occurs, the leakage rate limits act to indicate when plant I shutdown, ISI, and corrective actions to ensure tube integrity should be taken.
I 2.8.3 Value-Impact The STS limits are effective in minimizing both the likelihood and magnitude of offsite releases by minimizing the probability of the rupture and by limiting the transport of radioactivity into the secondary system. In addition, imple-mentation of the STS limits provides added assurance that the steam generators '
will not be operated in a condition in which they may be vulnerable to single or multiple SGTRs during pressure transients, such as an MSLB.
i SAI estimates that implementation of the STS limits at plants not currently ,
implementing these limits could potentially reduce the overall PWR (baseline) frequency of SGTRs by 15% if considered as a stand-alone improvement. Assuming
- a 64% reduction in this SGTR frequency associated with implementation of the j staff recommended actions concerning detection and prevention of loose parts
- and foreign objects (Section 2.1), improved secondary water chemistry programs 1
(Section 2.5), and condenser inspection programs (Section 2.6), the staff esti-1 mates that the additional incremental reduction in SGTR frequency attributable t
] to implementation of the STS limits to be approximately 5%. This corresponds to a low benefit (as defined in Table 3) in terms of reduced prubability of I core melt, and a medium benefit in terms of reduced probability of significant, ;
j non-core-melt releases. ;
i '
d Net cost impacts for plants not currently performing to the STS appear to be l
] relatively low. SAI estimates that lifetime impleiaentation costs could range l to a maximum of $1.3M for those plants for which it would be necessary to 1 incorporate revised sampling and analysis procedures in order to implement the ;
1 STS limits. Lifetime avoided costs would accrue due to the reduced potential i for SGTRs. SAI has estimated a potential benefit of $1.4M for affected pisnts, assuming implementation of the STS limits to be a stand alone improvement and assuming that SGTR reoairs will involve a 60-day plant outage. With implementa-tion of the staff recommended actions concerning loose parts and foreign objects, r
secondary water chemistry, and condenser inspections, however, that actual l j benefit would not be expected by the staff to exceed 50.5M.
SAI estimates that the ORE required to implement the STS limits are negligible. -
1 4 NUREG-0844 2-38
SG0G stated that there was general agreement with the proposal for implementing consistent primary-to-secondary leakage limits for all PWRs (letter, September 30, 1982). Wisconsin Public Service Corporation (WPSC) stated that Kewaunee is already in compliance with the proposed limits. Alabama Power Company stated that it supports the STS limits but that these should not be mandated for all plants without considering the past history and the design-basis-event analysis (letter, January 11, 1983).
The Sacramento Municipal Utility District (SMUD) commented that the STS limit of 0.35 gpm is based upon data for Westinghouse tubes, and that these data are not applicable to the smaller B&W tubes (February 5, 1983). The staff agrees that the STS leak rate limit was developed on the basis of data for Westinghouse 7 tubes, but believes that these limits can be reasonably applied to the B&W tube geometry. The staff notes, however, that limits higher than 0.35 gpm could be justified if backed up by appropriate test data indicating that such limits l adequately ensure the integrity of leaking tubes during normal operating and postulated accident conditions.
2.8.4 Conclusions ,
The staff finds STS leakage limits are an effective means for ensuring that the dosage contribution from tube leakage will be limited to a small fraction of 10 CFR Part 100 limits in the event of either a design-basis SGTR or a design-3 basis MSLB, In addition, the STS 500 gpd limit /SG is an effective measure for minimizing the probability of SGTR as a consequence of an MSLB. In consideration of the above, and the medium potential to reduce the probability of significant
- but less-than-core-melt releases, the staff has incorporated the STS limits on 1 allowable primary to secondary leakage as a staff recommended action. I i
2.9 Coolant Iodine Activity Limit
) f Based on the valu:-impact evaluation in Section 2.9.3, this potential industry ,
j action is categnrized as a staff recommended action.
2.9.1 Staff Recommended Action 1 PWRs that have Technical Specification limits and surveillance requirements for '
j coolant iodine activity that are less restrictive than the Standard Technical i Specifications (STS) should implement the STS. Those plants identified above I
that also have low-head high pressure safety-injection pumps should either: i (1) implement iodine limits that are 20% of the STS values or (2) implement
.I reactor coolant pump (RCP) trip criteria which will ensur9 that if offsite ,
I power is retained, no loss of O rced reactor coolant system flow will occur for ;
i SGTR events up to and including the design-basis double-ended break of a single !
I steam generator tube, and implement iodine limits consistent with the STS.
. 2.9.2 Basis for Initial Consideration !
) I The basis for the recommendation that all PWRs should be consistent in adopting [
l the STS derives from the staff's position that the STS coolant activity limits i j coupled with the STS surveillance requirements provide reasonable assurance l i that coolant activity will not contribute unacceptibly to offsite doses from a l l I !
1 t I I NUREG-0844 2-39 I i
steam generator tube rupture as severe as a design-basis event, thus ensuring that offsite doses from a design-basis SGTR remain within acceptable limits.
For example,11 PWRs do not have specific limits on radioiodine but do hr.ve limits on total gamma activity. Although the total primary coolant activ ity might remain substantially below the total Technical Specification act'.vity shutdown value, the actual radioiodine levels could be very high, Fwthermore, the potential for iodine spiking must be considered.
The basis for the staff position that some plants should adopt 20% of the STS activity limits derived from the Ginna experience. As stated in NUREG-0916, the amount of primary-to-secondary leakage and the total amount of water and steam released to the environment during the Ginna SGTR event were larger than would normally be predicted because of valve malfunctions and operatoa actions.
The staff found that the potential exists for doses to exceed 10 CFR Part 100 guidelines from a design-basis SGTR accident, and that these doses would occur only if a certain set of circumstances existed. These circumstances would be j (1) primary coolant with iodine concentration at the STS coolant iodine concen-tration spiking limit of 60 pCi/g dose-equivalent iodine-131, (2) maximum flow rate through a double-ended tube rupture, (3) flow through the tube rupture prolonged for 2 or more hours, (4) filling of the steam generator and steam line of the affected steam generator, (5) releases through the affected steam generator's safety or atmospheric dump / relief valves as a two phase mixture, and (6) conservative atmospheric dispersion factors. The actual radiological consequences of the Ginna accident were not severe because the reactor coolant's i
iodine concentration was very low. 0.057 pCi/g dose equivalent iodine-131 (about 1
2% of the plant's Technical Specification limit), and because the meteorologic conditions were far more favorable, with respect to offsite doses, than the i
conservative assumption used in the prior analyses.
The problems encountered with controlling and reducing reactor coolant system pressura following SGTRs appear to be more significant in plants at which the )
'j reactor coolant pumps would be manually tripped by the operator upon indication of an SGTR of the magnitude of a double-ended guillotine break. Since present )
l
- SGTR emergency operating procedures proposed by aany licensees require manual l
- RCP trip when pressure drops to a predetermined plant-specific value, plants 1 that have low-head high pressure safety-injection pumps (HPSIPs) are more likely i
1 to require RCP trip following an SGTR than are plants with n'gh-head pumps.
Therefore, plants with low-head HPSIPs and using the present RCP trip criteria 1 are more likely to have void formation, sustained leakage, and potentially, J steam generator overfill following an SGTR.
2.9.3 Value-Impact
) As part of the final safety analysis report (FSAR) safety analyses, all plants have been evaluated to ensure that a small fraction of 10 CFR Part 100 limits i are not exceeded during design-basis accidents. The subject proposal will provide l
added assurance that coolant iodine activity will not contribute unacceptably
^
to offsite dose during SGTR events as severe as a design-basis SGTR event.
However, the SAI analysis has indicated no significant risk reduction potential for this proposal.
] A small increase in ORE will probably result from increased primary coolant sampling in approximately tour plants which must upgrade their surveillance programs. With the application of ALARA (as low as reasonably achievable) techniques, this ORE increase is not expected by SAI to be significant.
NUREG-0844 2-40 l
An increase in ORE would result if imposition of the limit required additional shutdowns to replace leaking fuel elements. A shutdown and core unloading and reloading is estimated by SAI to result in from 25 to 60 person-rem. It is not likely that such additional shutdowns would be required to achieve compliance with the STS limits; however, it is likely that earlier shutdown and fuel re-placement would be required for some of the low-head HPSIP plants to meet 20%
of the STS limits. The SAI analysis indicated no identifiable benefit in terms of averted ORE because of the lower iodine limits. Therefore, the staff finds that ORE is not a significant factor for the STS limits but may be a significant factor for the 20%-of-STS limits for the low-head HPSIP plants.
Minor plant lifetime costs of $400,000 are associated with expanded surveillance programs for affected plants. A significant increase in costs would result if additional shutdowns to replace leaking fuel were required. As discussed above, it ir likely that such shutdowns would not be required by adoption of the STS limi *.s; however, it is very likely that such shutdowns may be required ,
by imposition of 20% of the STS limits. Therefore, costs are not a factor for '
the STS limiti but may be for the 20%-of-STS limits.
l ORE and cost impacts associated with 20%-of-STS limits would be unnecessary if !
changes in RCP trip criteria are implemented so that RCP flow is maintained 1
during SGTRs, and the 20% limit would not be applicable. The development of f i
reactor coolant pump trip criteria is being addressed as part of the resolution i of TMI Task Action Plan Item II.K 3.5 of NUREG-0737, "Automatic Trip of Reactor f Coolant Pumps During Loss of Coolant Accidents." TMI TAP II.K.3.5 is being !
implemented as Multi-Plant Action (MPA) G-1 and is discussed in additional l detail in Section 4.4.1.
SG0G (letter, September 30, 1982) stated that it had no comments on limiting coolant iodine activity and STS limits are a good starting point; hewever, there may be plant-specific reasons for exceptions. Such exceptions should be '
considered on a case-by-case basis.
WPSC (letter, October 4, 1982) stated that it did not believe that a generic '
requirement was necessary; rather a program designed to maintain fuel cladding integrity will attack the problem at its source, j
, l l Rochester Gas and Electric Corp. (RG&E) commenteu' that it submitted an analysis .
- by letter dated November 22, 1982 which demonstrated that the STS values for l I iodine are acceptable for a design-basis incident under the actual Ginna thermal i hydraulic transient, and, thus. the STS limits are acceptable for low-head high pressure plants (letter, dated January 24, 1983).
The staff finds that the SGOG comment is not inconsistent with the staff's disposition of this item as a staff recommended action for plants with less restrictive Technical Specifications than the STS. The staff also agrees with
! WPSC that licensees should be implementing effective programs to maintain fuel
- cladding integrity. However, the staff recommends that all PWRs should imple-l ment the STS limits and surveillance requirements for the reasons previously J stated. Regarding the comment from RG&E, the staff has requested RG&E to resub-j mit its analysis using more realistic assumptions regarding operator action j time.
I l
NUREG-0844 2-41 1
- - - - . =
2.9.4 Conclusions i
This potential industry action does not involve significant risk potential, but would provide added assurance of compliance with 10 CFR Part 100 guidelines.
The staff concludes that the potential industry action concerning coolant iodine activity (Section 2.9.1) should be incorporated as a staff recommended action.
The staff acknowledges the possibly significant ORE and cost increases associated 3 with the potential for additional refueling outages at plants implementing 20%
j of the STS limits in accordance with the staff recommended action. These impacts can be avoided at the subject plants by implementing appropriate reactor coolant pump trip criteria, in lieu of the 20%-of-STS limit, such that RCP trip for SGTRs i up to the design basis would not be required. Appropriate reactor coolant pump
- criteria are being developed and implemented as part of THI Action Plan j Item II.K.3.5 and Multi-Plant Action (MPA) G-1 as discussed in Section 4.4.1.
2.10 Reactor Coolant System Pressure Control 1
2.10.1 Potential Industry Action and Bases I l
Each of the four domestic SGTRs have resulted in longer times to depressurize the reactor coolant system (RCS) to below the secondary-side 9ressure than was previously assumed in plant analyses. In response, the staff considered a j potential requirement for licensees to evaluate further means of optimizing RCS pressure control with the objective of minimizing primary-to-secondary leakage.
2.10.2 Value-Impact ,
- There were no ORE changes attributable to this potential action since it proposed j only that a study be conducted with emphasis on the use of existing equipment. l 1.ikewise, there is no potential for a reduction in significant but less-than-core-melt releases attributable to a study. There would likely be changes in ORE and potential for significant but less-than-core-melt releases if beneficial
- results of such studies could be identified and were implemented. However, the value+ impact of such actions would be conducted at that time as a separate action
) beyond the scope of this potential action for an RCS pressure control study.
1 i
SAI estimates the economic cost per plant to perform such a study to be from ,
i $30,000 to $100,000. An assessment of the cost per plant indicates that it '
may be more cost ef fective to perform a smaller number of studies which would !
bound all PWRs.
SG0G stated that this potential industry action could result in an extensive analytical and procedure revision effort which is not warranted (letter, !
- September 30, 1982). SG0G also stated that the owners groups are currently 4
evaluating means of controlling reactor coolant system pressure during a steam j generator tube rupture and that NRC should not issue specific requirements i until these owners group ef forts are completed and reviewed. '
l
, 2.10.3 Conclusions ;
i The issue of RCS pressure control during an SGTR has been incorporated as an !
l ongoing staff action item as discussed in additional detail in Section 4.3.2.
I i
4 l NUREG-0844 2-42 I
_ _ - _ - _ _ _ _. __ - _ _. .. . _ - _ = --_.--_
i 4
i 2.11 Safety Injection Signal Reset ;
j Based on the value-impact evaluation in Sectica 2.11.3, this potential industry l action is categorized as a staff recommended action. ;
2.11.1 Staff Recommended Action L
The control logic associated with the safety-injection (SI) pump suction flow path should be reviewed and modified as necessary, by licensees, to minimize i the loss of safety function associated with safety-injection reset during an
- SGTR ennt. Automatic switchover of safety-injection pump suction from the boric acid storage tanks (BAST) to the refueling water storage tanks (RWSTs) j should be evaluated with respect to whether the switchover should be made en the basis of low BAST level alone without consideration of the condition of the SI signal.
r 2.11.2 Basis for Initial Consideration i
) Dur ng the Ginna event, a potential problem became apparent when the high head i
1 safety-injection pump suction had to be manually switched from the nearly ;
depleted BAST to the RWST after the SI signal had been reset. The safety- L injection logic was such that if the SI signal had not been reset, the pump l suction would have been automatically switched from the BAST to the RWST upon l low BAST level. However, if the SI signal has been reset, then manual operator 4
2 j actions are necessary to ensure that suction from the high-head SI pump is not i
- lost. Should the operator fail to effect this changeover, the SI pump could be t l damaged. (
5 1
l An improved design may be achieved if automatic transfer from the BAST to the !
RWST is provided on low BAST level under all operating conditions. This is a 3 l
desirable feature since in the event of a tube rupture, the contents of the BAST l may not be reduced to the low-level switchover set point for 20 to 30 minutes, l during which time the operators are precluded by procedures from resetting the !
SI signal. SI must be reset before the containment isolation (CI) signal can l
be reset. Resetting CI allows operation of equipment and systems that can aid l in mitigating the consequences of an SGTR, l 2.11.3 Value-Impact (
l No significant ORE changes were identified by SAI to implement the staff i recommended action. sal estimatas the total economic cost per plant to be l
$100,000. I Few comments relative to thi, proposed action have been received from industry.
SGOG stated that the specific example of safety injection pump suction is of such limited applicability that a generic requirement does not apply. The staff estimates that fewer than 10 plants would likely find that physical modifications are needed.
The value of the changes made pursuant to the staff's recommended actions are as follows:
(1) Since the probability of core melt as a result of SGTRs is relatively low, improvements in the safety injection legic are not warranted based solely NUREG-0844 2-43 l
on core-melt probabilities. However, it has not been shown that, from an accident mitigation standpoint, acceptable plant and offsite consequences would result from an SGTR without the high-head safety-injection pumps (HHSIPs). Current final safety evaluation report (FSAR) evaluations are conducted assuming the HHSIPs are available and injecting into the RCS to control and then regain adequate RCS inventory. The HHSIPs are then throt-tied, or secured, as the plant is cooled and depressurized to stop the leak. No analyses exist to support adequate plant and offsite consequences without the HHSIPs, which could happen if the operator failed to properly align suction after the safety-injection actuation signal has been reset.
(2) The staff's recommendation derives from the Ginna SGTR in which safety-injection flow was maintained after the termination criteria had been met because of operator concerns about core recovery. In the other SGTRs, elevated RCS pressure because of prolonged safety injection seems to be the rule rather than the exception. Therefore, SGTRs have resulted in situations in which damage to the HHSIPs (on plants with the logic in question) has an increased likelihood.
(3) From the standpoint of operator action, the SGTR event is a most challenging
- accident since a variety of diagnoses and manual actions must be effected in a relatively short time. The SGTR is also one of the more frequent
- accidents. From the standpoint of reducing the challenges to plant operators, the safety-injection system (SIS) logic should be improved so the operator
- would not have to take actions to protect the most vital RCS makeup capability in the plant.
2.11.4 Conclusions The staff concludes that the subject actions wili constitute an effective approach to ensuring that plants are in compliance with GOC 21, 23, and 35, and ensuring that plants do not have design futures that will, absent proper and timely a
' operator action during an SGTR event, result in damage to the safety injection system. In view of the minimal ORE and cost impacts, the relatively high '
frequency of SGTRs, and the complex operator challenges associated with this type of event, and the importance of maintaining defense-in-depth, the staff has incorporated the subject potential industry action as a staff recommended action.
2.12 Containment Isolation and Reset 2.12.1 Potential Industry Action '
A proposal was made that all licensees should review and evaluate the response !
of the letdown system to containment isolation and reset signals. Specifically, licensees should evaluate the containment isolation systems to ensure isolation of the low pressure portion of the letdown system inside contaiament (and its relief valve), thereby avoiding an unnecessary RCS leak during the event.
2.12.2 Basis for Initial Consideration During the Ginna event, the RCS letdown containment isolation valve closed, as !
d2 signed, on a containment isolation signal. However, as pressurizer level i
NUREG-0844 2-44 l
recovered later in the event, the selected letdown orifice isolation valve and the level control valve reopened as designed. Consequently, the letdown line was communicating with the reactor coolant system while 'e downstream portion of the letdown line remained isolated and theThis relief valve valve on thetoletdown relieves the pressure line i opened at a set point pressure of 600 psig.
relief tank and was the major contributor to the pressure relief tank level. l The Ginna containment isolation design therefore caused an unnecessary and i undesirable leak during an already complex event. ,
2.12.3 Value-Impact SAI estimates that the cost to utilities to evaluate their containment isolation I systems to be about $40,000/ plant. Modifications, if found to be needed, could run as high as $400,000. The potential cost savings associated with not having ,
j to replace the rupture disk, as a consequence of overfilling the pressure relief '
tank and bursting the rupture disk, is estimated to be minor in comparison to the cost of implementing any necessary modifications. Similarly, ORE associated l with containment cleanup activities at Cinna was estimated to have been less than [
0.6 person-rem. Thus, the proposed industry action vnuld not appear to have i significant ORE benefits. Although the proposed industry action could reduce l the complexity of the plant response during an SGTR, it is not estimated to result in any significant reduction in the probabilities of core-melt or signif- l l icant non-c9re-melt releases.
1 2.12.4 Couclusions f
! In the absence of identifiable benefits in the areas of reduced probabilities l
) of core-melt or significant non-core-melt releases, ORE, or cost, the staff ;
i concludes tiat this potential industry action should not be incorporated as a i
! staff recommended action.
I t I !
i I
i l
1 i i !
j 1 ;
l .
l I
l l NUREG-0844 2-45 i
! I
_. _ _ _ _ __ . m_ __ _ _ __ _ _ - _ _ . _ _ _ - -
l Table 3 Summary of valu rimpact evaluation Benefit in Benefit in Benefit in EE reducing reducing prevention El probability occupational Net of significant 5' of core radiation economic non-core-melt Item melt i2 exposure 1 benefit I releases 1 3 Remarks Disposition
- 1. Prevention & detection of loose parts (a) Secondary-side Low to medium Negative Medium High Can prevent 45% of Staff recom-l visual inspections (medium) SGTRs mended action and QA/QC work procedures (b) Loose parts Low Low Negative Medium Assumes secondary- Deleted monitoring systems (low) side inspections (LPMS) are 90% effective; LPMS is 70% effec-tive 2.
- 2. Inservice inspection of steam generator tubes (a) Supplemental tube Low Negative Marginal Medium inspections Staff action (Iow to (potentially medium) medium impact)
(b) Full-length tube Low Low Low to Medium inspections Eliminates obvious Staff recom-medium deficiency of cur- mended action rent TS (c) Denting inspections Lnw Low Medium Low to medium Staff action See notes at end of table l
l
x 55 Table 3 (Continued)
"c3 a Benefit in Benefit in Benefit in T reducing reducing prevention
- probability occupational Net of significant of core radiation economic non-core-melt Ites melt 1 2 exposure 1 benefit i releases 13 Remarks Disposition (d) Steam generator ISI Low Low Low Low to medium Precludes exces- Staff recom-interval sively long inter- mended action vals without inspections (e) Inspections follow- Low Prehably Probably Low to medium Deleted ing shutdown for negative negative repair of leakage (low) (low to med.)
- 3. Improved ECT tech- Low Medium Medium Medium Staff action l 7 niques m
- 4. Upper inspection Low Low Low Low Deleted l ports j S& Secondary water chem- Low to Medium to Medium to Medium to Can prevent 19 to Staff recom-
- 6. istry and condenser medium very high high high 33% of all SGTRs mended action inservice inspections
- 7. Stabilization and N.A. N.A. N.A. N.A. 4 monitoring of degraded tubes
- 8. Primary / secondary Low Low Low Medium Effective in pre- Staff recom-leakage limits (possibly venting SGTR mended action low impact) during MSLB and in mitigating radio-logical consequences I of SGTRs i
,.._ __._ _ _ __-_ ____ _ _ _ _ _ __ _-.~- ____ _ - _.-___-_ . _ ._.__.
l
)
i ,
t
! E Table 3 (Continued)
! E i
? Benefit in Benefit in Benefit in j S reducing reducing prevention i A probability occupational Net of significant i of core radiation economic non-core melt i Item melt 12 exposure 8 benefit I releases 2 2 Remarks Disposition
, 9. Coolant iodine activity I limit
- .STS for afi plants Low Negative Negative See next Effective in siti- Staff recom-l (low) (low) column gating radiologi- mended action j cal consequences !
of SGTRs 20%-of-STS for Low Negative
- Negative
- See next Effective in miti- Staff recom- !
j selected plants column gating radiologi- mended action **
m cal consequences i
{ j of SGTRs
- 10. RCS pressure control N.A. N. A. ' N.A. M.A. Staff action '
4 j 11. Safety injection Low Mone Negative See next Effective in pre- Staff recom-signal reset (low) column venting loss of mended action l
NHSIPs during !
1 SGTRs
- 12. Containment isolation N.A. N.A. N.A. N.A. Deleted and reset i N. A. - Not Applicable, study only
- - Impact can be avoided by changing reactor coolant pump (RCP) trip criteria so that RCP flow is maintained during SGTR.
1 l
) Notes (1), (2), and (3) are on next page.
1 ,
i 1
. - . . - - , - - . _ _ - , _ - - - - - . , - - , .-,..-~~=,n.-,---c-=-,-- . - ,- - - - _ , - - - - . _ . - _ - _ _ _
z ji Table 3 (Continued) 8 a Notes:
- 1 tow, Medium, and High are defined as follows:
Low Medjum i High Reduction in Core Melt < 10 8 10 4 to 10 6 > 10 4 Probability /RY Reduction in Significant < 10 6 10 4 to 10 6 > 10 4 Non-Core Melt Releases /RY Reduction in ORE (person- <5 5 to 60 > 60 rem /RY)
Net Economic Benefit <1 1 to 10 > 10 7 ($ Million over Plant Life) u 2 From Table 7 in Section 3 the baseline probability of core melt from SGTR-related causes is l estimated to be 5.1 x 10 g/RY for B&W plants and 3.6 x 10 6/RY for W and CE plants. Single SGTRs which occur as initiating events, and whose frequency is reasonably well established I from operating experience (1.5 x 10 2/RY), cont.-ibute approximately 1.1 x 10 6/RY to these baseline core melt estimates. The balance of the baseline core melt probability estimates derive from SGTR events (including multiple SGTRs) for which only highly judgmental event probabilities are available and which may be very conservative.
j Implementation of all the potential industry actions which have been dispositioned as staff-recommended actions would be expected to produce a low benefit (as defined in Note 1) given a baseline core melt probability of 1.1 x 10 6/RY, and a medium benefit given a baseline core l melt estimate of 5.1 x 10 6/RY for B&W plants and 3.6 x 10 6/RY for W and CE plants. The benefits shown in this table are the incremental benefits for each potential industry action.
3 The baseline probability of significant but less than core melt releases is given as 4.9 x 10 4/RY for B&W plants and 2.3 x 10 4/RY for W and CE plants in Table 10 of Section 3.
Single tube ruptures as initiating events contribute approximately 1.9 x 10 4/RY to the probability of these releases. Implementation of all the potential industry actions which have been dispositioned as staff-recommended actions would produce a high benefit per the definitions given in Note 1. The benefits shown in this table are the incremental benefits for each potential industry action.
I 3
SUMMARY
OF RISK ANALYSES FOR STEAM GENERATOR TUBE RUPTURE (SGTR) EVENTS This summary of the risk analyses addressing rupture of steam generator tubes was developed in the following manner. First, the staff's consultant, Science Applications, Inc. (SAI), performed a preliminary analysis of the risk asso-ciated with all events involving single tube ruptures. The staff reviewed that preliminary analysis as input to the final staff analysis. In addition, the
- staff developed independent analyses of events involving ruptures of single and multiple tubes. TheSAIandstaffanalyseswerethencombinedandsubjectedto ,
detailed review by a team of reviewers at SAI, by the staff's Reliability and i Risk Analysis Branch of the Division of Safety Technology, and by the Division [
of Risk Analysis of the Office of Nuclear Regulatory Research (RES). Various licensees and NSSS vendors also commented on the SAI analysis in response to ,
NRC Generic Letter 82-02. The following analysis considered the above anal- t yses, reviews, and comments. Although there will continue to be a diversity of j views on many of the details of the analysis, there appears to be a general ,
f consensus on the significant contributors to risk, and there is agreement that the risk of core melt from events involving steam generator tube ruptures is a relatively small fraction of the overall risk of core-melt accidents, j Section 3.1 provides an assessment of SGTR event probabilities involving single
- and multiple tubes. Various accident sequences involving SGTRs are examined in
- Sections 3.2, 3.3, and 3.4. Section 3.2 discusses SGTR events challenging the j reactor trip and decay heat removal functions. Section 3.3 discusses SGTRs that occur during loss-of-coolant accidents. Section 3.4 discusses SGTRs that occur
' concurrent with a loss of secondary system integrity. The overall probabilities, I consequences, and risks associated with SGTR accident sequences leading to core
! melt are discussed in Section 3.5. The probabilities, consequences, and risks i associated with non-core-melt releases during SGTRs occurring in conjunction l with a loss of secondary system integrity are discussed in Section 3.6. ,
3.1 Single and Multiple SGTR Probabilities i
In order to estimate the probabilities of various SGTR accident sequences, estimates of the probabilities of single and multiple steam generator tube t ruptures must be developed.
3.1.1 Initiating Event Probabilities !
For purposes of this analysis, the staff has assumed an SGTR frequency of 1.5 x 10 2/RY as an industry average value. This is consistent with actual i operating experience for Westinghouse plants; namely, four SGTR events during approximately 300 "mature" reactor years (i.e. , reactor operating years af ter the first two years of reactor life through mid-1986). Although Combustion Er.gineering (CE) and Babcock and Wilcox (B&W) plants have not experienced actual SGTRs to date, the staff has assumed the above SGTR estimate for Westinghouse i
l NUREG-0844 3-1 l
plants to be applicable to these plants as well.* This is judged to be reasonable in view of the limited number of mature operating years accumulated at CE and l B&W plants to date (i.e., 77 and 66 reactor years, respectively) and in view of the fact that extensive degradation problems have affected steam generators
, supplied by each of the NSSS vendors.
Each of the four SGTR events to date occurred as an "initiating" event. That is to say that each of the SGTR events occurred randomly while the reactor was being operated in a normal steady-state condition rather than occurring as a consequence of a plant transient or accident. In each case, the rupture occurred when the pressure-retaining capability of the subject tube degraded to a value less than the 1300 psid differential pressure across the tubes under normal operating conditions.
l Each of the SGTRs to date involved a rupture of a single tube. It is highly improbable that two or more tubes could rupture simultaneously as a true "initi-
)
ating" event during normal steady-state operation in view of random differences in flaw geometries, and therefore in pressure-retaining capabilities, which exist j from tube to tube. Rupture of two or more tubes are credible only as a con-sequence of a plant transient or accident when the loading on the tubes becomes more severe. However, as is discussed in Section 3.1.2.2, multiple tube SGTRs l
- involving certain initiating plant transients can be idealized as initiating '
i events for purposes of evaluating the corresponding potential for significant i
radiological releases and core melt. The potential for SGTRs as consequential i events is discussed in Section 3.1.2.
3.1.2 Probability of SGTRs as Consequential Events 3.1.2.1 Conditional Event Probabilities 1
SGTRs can potentially occur as a consequence of plant transients or accidents
) when loadings on the steam generator tubes are increased above normal operating i loads. Although there have been no SGTRs as consequential events to date, there ;
{ have been several instances in which steam generator tubes are known to have i 4
been sufficiently degraded such that they were vulnerable to rupture, given the
! occurrence of a severe plant transient or accident. These periods of vulnera-I bility can be inferred from the subsequent occurrence of SGTR events at the j subject plant during normal operating conditions. For a tube to be sufficiently l degraded to burst under normal operating conditions when the differential pres-i sure across the tubes is between 1300 and 1500 psid, it must first be degraded to the point where it is not capable of withstanding differential pressure load-ings ranging to as much as about 2600 psid for the spectrum of plant transients and postulated accidents such as a main steamline break (MSLB). The pressure I retaining capacity of a virgin, undegraded tube is generally between 9000 and
) 10,000 psid, depending on the tube diameter and wall thickness. During the
,l
{ *CE and B&W plants have experienced three tube failures involving relatively high primary-to-secondary leakage rates, but not sufficiently high to be classified '
I as SGTR occurrences. These incidents involved Fort Ca?houn in May 1984 (letter, i J oe 2;, 1984) and Rancho Seco and Oconee Unit 2 in May 1981 and September 1981, !
- respectively (INP0 Report 82-030). SGTR events are defined by the NRC staff to '
be a primary-to-secondary leak in excess of the normal charging capacity of the *
- *eacte coolant system. i
.l N W G-C.44 3-2 i
,._c- - , , , - - . _ _ - - - - - . - . . _ - _ . _ - _ - _ - _ _ _ . . . - . .
--..-_- _ _~ - - - - -. .- _ - . - __
1
- time period when the pressure-retaining capabilities of one or more tubes at u
- plant degrade to values that are 2600 psid or less, the plant can be said to be l
vulnerable to rupture given the occurrence of a postulated accident involving pressure loads of this magnitude. As the pressure retaining capability of the tube (s) continues to decline below 2600 psid, the plant becomes vulnerable to l rupture under less severe transients (involving pressure loadings between 1300 and 2600 psid) which may occur more often than postulated worst case accidents.
The staff has evaluated the circumstances leading to the SGTR events to date and estimated the periods during which each plant was vulnerable to rupture for postulated accidents involving a pressure differential of 2600 psid. These estimates are described in detail in Appendix B of this report. As seen in
] Appendix B, the staff estimates that the periods of vulnerability to rupture (under postulated accident conditions) preceding the SGTR events to date totaled .
1.2 years. These estimated periods of vulnerability correspond to 2.7% of the mature reactor years accumulated to dato at the four subject plants. Thus, the i conditional probability for rupture associated with these periods of vulnerability ,
i is 0.027 at the four subject plants.
Apart from periods of vulnerability which were terminated as a result of SGTRs, ;
i
- there may have been additional periods of vulnerability which were terminated as 1 a result of tube plugging following an inservice uddy current inspection and/or j i a small primary-to-secondary leak. To allow for these additional periods of ;
vulnerability, the staff has assumed an overall conditional probability of 0.05 :
I that one or more tubes will be vulnerable to rupture during postulated accident ,
conditions. This assumption is comparable to the above 0.027 conditional prob-l I ability estimate for the four plants which experienced SGTRs and is believed to I be conservative as an industry average for all PWRs. In view of the total of 440 l mature reactor years at PWRs to date, this 0.05 conditional probability estimate !
i allows for a factor of almost 20 over the total accumulated period of vulner-ability (1.2 years) estimated earlier as preceding the SGTR events to date. It ;
is implicit in the 0.05 conditional probability asu.mption, therefore, that :
periods of vulnerability to rupture during postulated accidents are almost t 20 times more likely to be terminated as a result of an inservice eddy-current inspection or small leakage event than as a result of an SGTR event.
The conditional probability estimate above is based on postulated accidents as initiating events which are assumed to involve an increase in pressure differential !
from the normal operating value of about 1300 psid to a maximum value of about i 2600 psid. For transients involving less severe pressure loadings, the condi-tional probabilities for causing a SGTR are reduced. The reduction factor will !
range between 0 and 1.0 for transients involving peak differential pressures i between 1300 psid and 2600 psid. A linear relationship between the reduction factor and the magnitude of the pressure transient can be assumed subject to the ,
following assumptions: (1) the rate of loss of pressure-retaining capability is t constant with time and (2) tubes that have degraded to a pressure-retaining I capability of 2600 psid will continue to degrade until they burst during normal !
l operating, transient, or, postulated accident conditions. It is implicit in the latter assumption that periods of vulnerability to rupture under transients or postulated accidents will be terminated by a tube rupture rather than as a result of an inservice inspection or small leak. However, consistent with the earlier assumption that periods of vulnerability to rupture under postulated accidents are 20 times more likely to be terminated as a result of a small leakage event or inservice inspection than by rupture under normal operating pressures, the NUREG-0844 3-3
probability that a tube will continue to remain in service as its pressure-l retaining capability is reduced to some given salue below 2600 psid is assumed l to decline in a linear manner from 1.0 at 2600 psid to essentially zero (actually i 0.05) at 1300 psid. Thus, the conditional probability for rupture during a given
, transient has been assumed to vary as the square of the ratio of the pressure
- increase associated with transient to the pressure increate associated with the l postulated worst case accident.* This is expressed in the following equation (hereafter referred to as the conditional probability equation).
~
2 APg - AP Cj = C, 3p , 3p Where: C
=Conditionalp'robabilityforoneorantetuberupturesduring 4
transient "i C* = Conditional probability for one or more tube ruptures during postulated MSLB accident (equals 0.05)
AP g = Pet.k differential pressure across tubes during transient "i" n = Normal operating pressure differential across the tubes (typi-AP cally 1300 psid) l AP* = Peak differential pressure across the tubes during postulated MSLB accident (about 2600 psid) l l
The staff estimates in Appendix B indicate that periods of voinerability preced- ;
ing two of the four SGTR events to date generally involved a single tube. The i evidence for the other two SGTR events is ambiguous with respect to wtather the subject plants were vulnerable to multiple tube ruptures. Accordingly, the staff has conservatively assumed that one-half of all SGTRs occurring as conse-quential events will involve multiple tube ruptures. It is very unlikely that as many as ten tubes could be vulnerable to rupture during a transient without some prior warning indication (e.g., a small leak or single SGTR during normal operation). Even in the unlikely event that more than ten tubes are vulnerable ,
to rupture for a given transient, rupture of one or a few of the weakest tubes {
[
would be expected to attenuate the pressure transient before sufficient pressure !
is reached to rupture the balance of the vulnerable tubes. The staff has !
assumed a conditional probability of 10 2 that SGTRs occurring as consequential i events will involve more than 10 tubes. !
3.1.2.2 Initiating Transients Plant transients and accidents that have the potential for causing SGTRs as consequential events have been divided into two groups for purposes of this analysis. The first group consists of initiating plant transierats or accidents ;
j wht:h significantly increase the likelihood that the subsequent SGTR will lead to significant radiological releases to the environment and/or core melt compared to SGTRs which occur as "initiating" events. The second group consists of plaat transients and accidents that do not significantly increase this likelihood.
"The exact relationship between the conditional probability of rupture and the magnitude of the pressure transient is complex. However, the assumption of a ,
"square" rather than a "linear" relationship reduces the resulting core melt I probability estiraate only by about 20%, and thus, does not have a significant effect on the outcome of this analysis.
NUREG-0844 3-4
The staff has considered the following plant transients and accidents as l belonging to the first group:
(1) anticipated transients without scram (ATWS)
(2) loss of coolant accident (LOCA)
(3) non-isolable loss of ser.ondary system integrity (e.g., main steam line break, stuck-open steam generator safety valve)
These plant transients and accidents, including their assumed frequency and severity, are discussed in Sections 3.2, 3.3, and 3.4, respectively. The corresponding potential for SGTRs as a consequence of these initiating events has Deen estimated from the conditional probability equation in Section 3.1.2.1.
Specific plant transients and accidents belongir.g to the second group (i.e. , not !
significantly increasing the likelihood that subsequent SGTRs will lead to sig- '
nificant radiological releases or to core melt) have not been evaluated in detail. ;
Consequential SGTRs falling into this second group are treated in this analysis !
as "initiating event" SGTRs. Single SGTRs in this category can therefore be i assumed to be accounted for in the 1.5 x 10 2/RY estimate for single SGTRs [
discussed in Section 3.1.1. The probability of multiple tube ruptures in this second group is assumed for best-estimate purposes to be 1.6 x 10 3/RY. This :
represents the 50% confidence value estimate for an event that has never occurred; L namely, zero multiple tube ruptures after 440 mature reactor years at PWRs to ,
date. Consistent with the estimates given in Section 3.1.2.1, this 1.6 x 10 8/RY probability for multiple tube ruptures can be broken down as follows: i 1.57 x 10 3/RY* for 2 to 10 tube ruptures 1 3.2 x 10 5/RY* for >10 tube ruptures ,
s! 3.2 SGTR Events Challenging the Reactor Trip and Decay Heat Removal Functions l l
1 from NUREG-0460, the conditional probability for failure of reactor trip is l
! estimated to be 3 x 10 5 (The Salem 1 ATWS events of February 22 and 25, 1983 j indicate that the unreliability of the trip system may have been six times higher i i than this value; however, corrective actions are expected to reduce the proba- <
) bility of failure to scram to close to the estimate of 3 x 10 5/ demand.) Based I 1
on an initiating event frequency of 1.5 x 10 2/RY for single tube rupture events j (Section 3.1.1), 1.6 x 10.a/RY for 2 to 10 tube ruptures (Section 3.1.2.2), and !
3 x 10 5/RY for >10 tube ruptures (Section 3.1.2.2), the probabilities of an ATWS l as a consequence of 1 tube rupture, 2 to 10 tube ruptures, and >10 tube ruptures r are 4.5 x 10 7/RY, 4.8 x 10 8/RY, and 9 x 10 1 /RY, respectively, j I
Recovery from SGTR(s) occurring in combination with an ATWS would be more
< difficult than for SGTR(s) not involving an ATVS. In addition, there are no
) emergenc) procedures for a combined ATWS/SGTR event. For purposes of this
] analysis, ATWS sequences which include SGTR(s) are assumed to result in core j melt. Core-melt sequences are discussed in Section 3.5.
i The stif f also considered SGTRs occurring as a consequence of ATWS events.
l Loss-ef-main-feedwater events are the most f requent of anticipated transients
! (abon 3/RY). The staff assessment specifically addresses events in which there I
]
- This analysis is actually based on rounded values of 1.6 x 10 3 and 3 x 10 5/RY, 4 respectively.
NUREG-0844 3-5 l
is a total loss of main feedwater, since these events produce a more severe primary pressure response during an ATWS than does a partial loss of main feedwater. On the basis of a limited survey of plant data, total-loss-of-main-feedwater events are assumed to occur at a frequency of 1/RY. This number is believed to be conservative for purposes of estimating SGTR probabilities for the entire spectrum of ATWS events.
Three ranges of moderator temperature coefficients producing pressure transients of varying severity were considered. For We3tinghouse plants, the estimated conditional probability of each range of peak primary pressures was taken from NUREG-0460. Unless action is taken to trip the turbine during the brief period of time before the peak primary pressure is achieved, the secondary-side pressure will remain stablo, perhaps decreasing by approximately 100 psi. If the turbine is tripped, the secondary side pressure will increase until the pressure relief set point is achieved, thus tending to partially offset the peak differential pressure across the steam generator tubes. Because of the short time interval involved (about 75 seconds), the conditional probability of turbine trip is 1
believed small, and thus no credit for manual turbine trip was assumed in this assessment.
Conditional probabilities for an SGTR as a function of the magnitude of the ATWS pressure transients were determined from the conditional probability equation in Section 3.1.2.1.
The following summarizes ATWS sequences leading to SGTRs:
ATVS Sequence Events Frequency or Probability l Total loss of main feedwater 1/RY i i l Failure to scram 3 x 10 5 Moderator coefficient in range such that peak primary pressure l may be in range of:
i
. (1) 2650 psig -
~ 0.5
) (ap = differential pressure q between primary and secondary = 4800 psid)
J (2) between 2650 and 3000 psig 3 0.49 (ap = 2150 psto)
(3) between 3000 and 3500 psig 0.01
- (ap = 2650 psid) 1 4 4
- Single SGTR 2.5 x 10 2 1,1 x 10 2 3,7 x 10 s
) 2 to 10 SGTRs 2.5 x 10 2 1,1 x 10 2 3,7 x 10.s j >10 SGTRs 5 x 10 4 2.1 x 10 4 7.4 x 10 5 I
1 NUREG-0844 3-6
l Taking the sum of the probabilities for the three peak primary pressure cases, the overall probability of SGTRs as a consequence of ATWS is as follows:
Single SGTR 2.3 x 10 7/RY '
2 to 10 SGTRs 2.3 x 10 7/RY
>10 SGTRs 4.4 x 10 8/RY To sum up, the probability of SGTRs occuring in conjunction with an ATWS (either as an initiating or consequential event) is an follows:
Single SGTR 6.8 x 10 7/RY 2 to 10 SGTRs 2.8 x 10 7/RY
>10 SGTRs 5.3 x 10 d/RY Core melt probabilities associated with other types of SGTR challanges to normal transient response functions are estimated by the staff to be 4.4 x 10 7/RY.
i This is consistant with SAI estimates for core melts caused by single SGTRs in conjunction with a loss of AC power (2.4 x 10 7/RY) and SGTRs in conjunction t71th a total loss of auxiliary feedwater (2 x 10 7/RY).
f
! 3.3 SGTR Events Resulting From Loss-of-Coolant Accidents The second category of SGTR events leading to core melt involves those sequences that include a LOCA and consequential tube failures. In these cases, the tube !
failures during the LOCA tend to inhibit the normal core reflood process and i thereby increase the estimated peak cladding temperatures and the core melt probabilities. ;
Analytical and experimental investigations into the influence of tube failures i on LOCA are documented in a series of reports from the Idaho National Engineer-ing Laboratory (INEL) (NUREG/CR-0175; letters, August 8, 1978 and November 3, l 1978; and INEL report August 1977). These reports indicate that the failure of 10 or fewer tubes would increase the peak cladding temperatures to approxi-l oately 1900*F for tube failure during refill and to 1800'F for tube failure I during reflood; and that failure of more than 20 tubes would have effects rang-ing only frnm slightly adverse to slightly beneficial. In all cases the lenk-age through the failed tubes would occur from the secondary system into the ,
primary system. These trend:: are shewn in Figure 1 taken from INEL report ,
CVAP-TR-78-015 (letter August 8, 1978). Therefore, these investigations indi- '
cate that the probability of a core melt following a LOCA would be increased for those events in which 10 to 20 tubes fail. A significant conservatism is that the experiments and calculations (described in NUREG/CR-0175; letters August 8, 1978 and November 3, 1978; and INEL report, August 1977) assumed the ,
tubes all completely severed at the worst time in the event (refill), all the intact loop's steam generators were equally and simultaneously affected, and the broken loop steam generator was not sinultaneously affected.
For purposes of the probabilistic assessment herein, a core melt probability of 0.1 has been assigned to LOCA events involving significantly elevated peak cladding temperatures, although the above analyses and experiments do not indi-cate that a core melt would occur. This assumption is intended to account for i plant-to-plant variations and for uncertainties in this and othe- aspects of ;
the ECCS response to a LOCA. In addition, the staff has conservatively assumed 1 l
i NUREG-0844 3-7 l
4 120 i i i I I O Tube ruptures at reflil 14 tube rupture A Tube ruptures at reflood 2372'F 1300 -
g a Calculated values for tube ruptures r ~
g at reflood (calculations required j due to core power trip) 1 16 tube rupture l
i i 1 1200 -
1 i -
l 20 tube g g rupture i 1150 - -
{ s \
t \
z jg \
5 1100 - -
1 i e I
$ g 12 tube g u 1060 -
rupture -
3 w I \
l' f I A
\
1000 -
g -
l 30 tube e Stube 1 rupture rupture to tube i 960 -
9 g rupture _
- \
\ '
\
m - \ -
36 tube rupture
- g I t 1 I I O 10 20 30 40 90 to ,
EQUIVALENT NUMBER OF TURE RUPTURES Figure 1 Maximum cladding ternperatures obtained for caser with tube ruptures initiated at the '
start of refill and at the start of reflood i
Source: Idaho National En9ineering Laboratory
- Report CVAP-TR-78 015 3-8
significantly elevated peak cladding temperatures to be associated with 2 or more SGTRs occurring during a LOCA rather than 10 to 20 SGTRs to allow for pos-sible uncertainties in the 10 to 20 SGTR estimate stemming from the INEL tests.
During a LOCA, the depressurization of the RCS may cause a reversal of the net pressure loading acting on the tubes from primary-to-secondary to secondary-to-primary. The conditional SGTR probability equation in Section 3.1.2.1 is not intended for specific application to transients involving a pressure reversal across the tubes. However, NRC-sponsored burst and collapse tests indicate that the reverse pressure differentials associated with LOCAs are substantially less limiting than a postulated MSLB from the standpoint of inducing an SGTR (NUREG/CR-0718). Therefore, the conditional probability of 2 or more SGTRs, given a LOCA, is assumed (very conservatively) to be 2.5 x 10 2/RY based on the i proba-bility estimates developed in Section 3.1.2.1 for M5LB.
The event sequence is as follows:
Event Probability (1) Large-break LOCA (NUREG-75/014) 10 4/RY (2) Failure of 2 or more tubes 2.5 x 10 8 l (3) ECCS ineffectiveness 10 1 2.5 x 10 7/RY
- Therefore, for large-break LOCAs, the core-melt probability due to concurrent i
SGTRs and steam binding-induced delay in core reflood is extremely low. The l staff is currently performing a more detailed assessment of the risks associated ;
i with LOCA events involving concurrent SGTRs as part of NRC Generic Issue 141.
This effort has been initiated to address comments submitted in a letter dated ;
April 20, 1987, from Dan L. Johnson of San Diego, California, in response to a ;
proposed revision to Appendix K of 10 CFR 50 concerning acceptanc,e criteria for ;
}; emergency core cooling systems which was issued for public comment in 1987 i (Proposed Rules; Federal Register Vol. 52, No. 41; March 3, 1987). l i !
3.4 SGTR Events in Combination With Loss of Secondary Systen Integrity or Failure To Achieve Steam Generator Isolation 1
1 This category of events includes single and multiple tube ruptures occurring i in conjunction with a non-isolatable loss of secondary system integrity '
i (outside containment) or failure of the main steam isolation valves (MSIVs). ;
j Events in this category involve escape of radioactive primary water into the
, secondary system and subsequently into the environment. The events are divided !
into three groups: those involving failure of the main steam line, those in-volving failure of the steam generator atmospheric dump valve or safety valves. l l and those involving failure of the MSIV to operate, thus preventing isolation i of the damaged steam generator. ,
t These groups of events are described in Sections 3.4.1 through 3.4.4 below.
Section 3.4.5 discusses the staff's analysis of the time available to the l l plant operators during the various event sequences to terminate the leakage i I of primary coolant into the secondary system before ECCS water from the refuel-ing water storage tank (RWST) is depleted. Probabilities for eash of the i event scquences considered are summarized in Section 3.4.5.
, NUREG-0844 3-9 l
1 !
l
., _ . _ _ _ _ _ . . _ _ _ . _ _ _ _ _ _ _.___m_ _ _ _ _ -- . _ _
l 3.4.1 SGTRs and Total Loss of Secondary Inw i - (Sequences 2, 5A, 58, 8A, 4 86,8C)
Steam line failure can be either an iu t v . w e ont leading to an SGTR or can
- occur as a consequence of an initiating f, A _,ent. The probability of an MSLB
, as an initiating event has been estimated to be 1 x 10 8/RY for the case of an MSLB in the portion of the steam line which is outside containment but before l the main steam isolation valve. A postulated worst-case peak pressure differ-1 ential across the tubes of 2600 psid has been assumed. Event sequences involv-
! ing SGTRs as a consequence of an MSLB have been designated as event sequences i 8A, 88, and 8C in Section 3.4.5.
I j The potential for a main steam line break as a consequence of an SGTR was suggested during the Ginna event when the steam lines filled with water. The staff has assessed the change in the probability of failure of the main steam j line from the increased stress levels associated with the deadweight of water i in the steam lines. Analyses have been performed of the increase in stress
{ 1evels tha'. would result from filling the steam lines in several plants. Infor-i nation extracted from analyses on the Ginna, Zion 1 Waterford 3, and Oconee 3 plants indicates that, although in some cases the spring hangers may be loaded 1 slightly beyond their operating range, they will not fail and that the stress levels in the main steam line will in all cases remain within the limits allowed
)' by the ASME Code. In addition to the analyses available, the steam lines were inspected after overfilling events at Oconee and Ginna and no indications of failures or incipient failures were found. Therefore, the staff concludes that the probability of failure of the main steam line is not increased by the deadweight loading. Nor is there considered to be a significant potential for j failure from waterhammer since the water in the steam lines will be essentially l I saturated. Accordingly, the estimates of risk in this report for event sequences (
{ that consider failure of the main steam lines are based on a conservativel I i determined conditional probability of main steamline failure of 1 x 10 3/ yover- i i fill event. Event sequences involving failure of the main steam line as a
- consequence of an SGTR event have been designated as event sequences 2, 5A, and SB in Section 3.4.5.
Main feedline break (MFLB) scenarios are not included in this analysis, but are thought to be no more likely to occur than main steam line breaks. There-fore, MFLB scenarios can be implicitly included by doubling the probability of an MSLB for those plants which do not have feedwater reverse flow check valves {
inside containment. For plants with feedwater check valves, an additional l
- failure of the check valve would be required to cause a non-isolatable loss of i
- secondary system integrity outside containment. '
]
l Analyses performed to support SECY 82-296 (memorandum, July 13, 1982) and J
NUREG-0937 have shown that in the event of an MSLB with accompanying SGTRs, the ,
- consequences with resper,t to reactor coolant inventory and core cooling are j j bounded by the spectrum of primary LOCAs. Figures 2 through 5 show the results )
{ of a double-ended MSLB and 1, 5, or 10 SGTRs at Calvert Cliffs. No credit for
] operator action was taken in the initial 2000 seconds, with the exception of the ,
j MSLB and one SGTR case where high pressure injection (HPI) was throttled at !
j 600 seconds to prevent filling the pressur4rer, f !n terms of stopping the loss of coolant through the broken tubes and out the i
- broken SG steam line, the primary pressure must be reduced to atmospheric l?
1 1 i NUREG-0844 3-10
) l J
usu/to-son 2 m m ee.
- usu/s-sen f.eeeese, usts/i-scu .
I y--
5esseeee. -
l 1
.se e. .
c '
<w{' . .: ~~~~ ~~. .-~
6 e
e ses me ese ese ses see w use ese seat "
fiut(e) l Figure 2 Primary pressure decrease for MSLB with concurrent SGTRs l (Primary pressure decrease did not depend upon number of tubes '
- ruptured until loop-A steam generator emptied at 90 seconds.)
i ,
- m. .- "' i .
{
{e u. f- usts/to-SCn !
. I i 5 ,, ! -
w$Ls/S-Sen !;
w 3 ; usta/t-Scn a
I s
t
[ s. i , ,
, k l l
< e. .
i i <
s
\--
I. .
~
- e. .
.e e ase as ese see see see w see .e. sees l 71 4 (e) l Figure 3 HPI flow' rate for MSLB with concurrent SGTRs (HPI flow throttled in 1- and 5-5GTR cases at 600 seconds and :
1800 seconds when primary system refilled. HPI was insufficient to reful system to normal level in the 10-SGTR case.)
t i
I i 3-11 .
see , ,
WSL3/10-5CTR 1 , see. -
l 4e ~.-- wSLS/3-SCTR l 5 usts/1-SCTR l y m. .
l
- I g me.j i
l t e e
! t - '~- - - -- -
l a.
L, __ . . -
.te a see a ese see .oe see w .se ese asse flut (s)
Figure 4 Tube rupture flow rate for MSLB with concurrent SGTRs l (Tube rupture flow equilitrated with HPI flow (compare j with Figure 3 which gives 1/2 th) total HPI flow))
t i
8- ;
?,,, ,
/
., .. , ~.: *- .
i y ,-
2 e s.
\
= ,..'..-
5 /. '<' , ,<\
1
, 3 \ , .' - usta/*-SCTR , !
g I i . , , , 'e ;
z i ,, ' '
wst.3/l-5ctR g '
- W5LS /l-SCTR i
,, L __
l
. u. .se e n .e use . e .se =se (
?twC (s) l I
(
Figure 5 Pressurizer water level for MSLB with concurrent SGTRs c (System refilled by HPI flow in 1- and 5-SGTR cases. HPI :
flow was insufficient to fill to normal level in 10-SGTR '
case.) ;
r 3-12 l
l a
j pressure. This is achieved utilizing the residual heat removal (RHR) system which requires the reactor to be first cooled to around 350'F. Once placed in operation, the RHR system would have to continue to cool down to the point where ;
- the reactor could be partially drained and depressurized. At this point, break i i flow would stop. This evolution takes considerable time, during which there ;
j would be a loss of coolant and RWST inventory out the break. ;
1 l 3.4.2 SGTRs Occurring in Conjunction With a Stuck Open SG Safety Valve r (Sequences 1, 4A, 48, 9A, 9B, 90) '
j Failure of the steam generator safety valves to fully reseat after being chal-i lenged during a plant transient occurs relatively frequently, but does not l j generally result in a significant pressure increase across the SG tubes which ;
cay challenge tube integrity. However, there are exceptions to this trend as '
} evidenced by a stuck-open SG safety valve occurrence at Davis-Besse Unit 1 in 4 March 1984 (NUREG 0090, Vol. 7 No. 1) which led to a complete blowdown of the affected steam generator and increased the pressure differential across the tubes from an initial value of 1300 psid to a maximum value of 2220 psid.
The staff has assumed a 10 2/RY frequency of stuck-open safety valve occurrences at B&W plants involving pressure transients comparable in seve,ii.y M the Davis- ,
i Besse event. The frequency of such events at W and CE ple.ts ;: .i.ved to be '
i substantially less since, unlike B&W plants, tfie SG safety valves are w ; rally i
, not challenged during a turbine trip. The staff considers the frequency of t such occurrences at W and CE plants to be conservatively enveloped by the estimated frequency T10 3/RY) of large MSLB occurrences discussed earlier in !
Section 3.4.1 and which have been designated as event sequences BA, 88, and 8C '
) in Section 3.4.5. Event sequences involving SGTRs as a consequence of a stuck- [
open steam generator safety valve have been designated as event sequences 9A, !
98, and 9C in Section 3.4.5. !
1 i In addition to being a potential initiating event for causing a subsequent SGTR, I a stuck-open safety valve can potentially occur as a consequence of an SGTR as .
evidenced during the Ginna SGTR event. During the Ginna event, the affected i steam generator filled up to the steam line safety valve as a result of primary-to-secondary leakage from continued operation of the safety injection pumns. :
The safety valve lifted five times at successively lower pressures and failed ;
to fully reseat (at least twice). The failure to completely reseat contributed ;
ts the overfill problem by lowering the damaged steam generator pressure, thus i mising the differential pressure across the broken tube and sustaining the !
It akage despite reduced primary system pressure. Although the leakage through l tM safety valve at Ginna was relatively small, the potential exists for a !
men serious safety valve f ailure given future SGTR/ overfill occurrences and -
su hequent challenges to the safety valves. I
{
In view of one such occurrence during the four SGTR events to date, the steff ;
has assumed that one in four SGTR events involving a single tube will result j in o m fill of the steam generator with subsequent challenge to the steam ,
genecator safety valves or atmospheric d ap valves. The staff has further ;
ass wed that all SGTRs involving multiple tubes will result in overfill and !
chaMenge to the safety valves or atmospheric d ap valves. These assumptions l art belieged to be conservative since they take no credit for improvements in j cmergency operating procedures which are being implemented under THI Task I Action Plan (TAP) I.C.1 and which are expected to result in a reduction in l 1
NUREG-0844 3-13
i i the potential for overfill. TMI TAP I.C.1 is discussed in additional detail in Section 4.4.3. In addition, the potential for SG ovcrfill and its conse-quences, and methods for preventing its occurrence are being evaluated as part of an ongoing staff program discussed in Section 4.3.1. This ongoing prog *am is expected to provide a basis for more realistic estimates of overfill potential and its consequences.
Given an SGTR event and subsequent overfill, the preferred approach is to uti-lize the atmot.heric dump valves (ADVs) to avoid challenges to the SG safety valves. If a . fety valve lifts and fails to reseat, there is no way to stop i the release of radioactive steam into the atmosphere until the RCS has been l
sufficiently depressurized to halt further primary to secondary leakage. On I the other hand, block valves are generally available to isolate the ADVs in l the event that they should stick open or leak excessively. Furthermore, the
- SG safety relief valves are designed to discharge steam rather than liquid.
Valve tests performed by EPRI have shown that safety valves have the potential
] for exhibiting severe disk chatter against the seat for cases where water is being discharged through the valves. Such chattering can cause damage to tbs i valve seat and valve internals and thus potentially may lead to a partiaM " or
- fully stuck open safety valve.
- Although the ADVT, constitute the preferred means for relieving secondary-side l 1 pressure during an SGTR, the ADVs may noc be available for this function as l I was the case at Ginna where the operators isolated the ADV by closing the
! blor.k valve because they apparently misinterpreted the SGTR procedures. Further-J more ADVs are gene
- ally not considered to be safety related and their avail- ;
i ability is not generally required by plant Technical Specifications. The staff i has assumed a 0.5 availability factor for purposes of this analysis, which is !
I' believed to be conservative, f
, t Given the noaavailability of the ADVs, the probability of a stuck-open safety (
i valve of sufficient magnitude to cause SG blowdown is not clear on the basis j of availat;1e data. The staff has assumed a failure probability of 0.1 per event involving a challenge to the safety valve ender SG overfill conditions. !
Failure in this context involves failure of the valve to close sufficiently to pitsvent SG blowdown and/or significant leakage. This assumption is not [
intended to include minor safety valve leakage such as occurred during the [
Ginna event. i l i
- SGTR event sequences involving a stuck-open safety valve as a result of steam t generator overfill are designated as event sequences 1, 4A, and 4B in f Section 3.4.5.
I i 3.4.3 SGTRs and MSIV Failures (Sequences 3, 6A, 6B) !
For single or multiple SGTRs, the damaged steam generator must be isolated from [
] the rest of the secondary system. Since the intact steam generators are used j to cool the RCS to allow RCS depressurization to stop the leat, the steam pressure in the intact steam generator will be decreased with time. If the
! damaged steam generator cannot be isolated from the intact steam generatars, ;
l then the damaged steam generator pressure will drop along with the intact steam [
! generator pressure. Primary system pressure would be lowered along with RCS j temperature in an effort to equalize pressure across the damaged tubes. However, )
! l 1
l NUREG-0844 3-14 !
i i l l i
erin.ary pressure must always be maintained high enough to maintain adequate RCS rubcooling and an ade wate temperature difference for decay heat removal. If the damaged ar.d intact steam generators are at the same pressure (because of failure to isolate), subcooling in the primary and pressure equalization across the damaged tubes cannot be attained, and break flow would continue until the ;
RCS reached atmospheric pressure.
If tne MSIV of the damaged steam generator fails to close, the damaged generator ccald be isolated by utilizing the intact SG MSIVs; the associated atmospheric relief valves would be utilized for RCS cooldown and decay heat removal. In the scenarios in which the MSIV of the damaged steam generator is assumed to fail, the staff has assumed that the steam bypass system continues to function to maintain the steam system pressure (hence the back pressure on the damaged SG tubes) at about 900 psig. If the bypass system failed to maintain the value because of a loss of offsite power, circulation water, or other failures, the SG safety valves would maintain pressure no higher than about 1000 psig.
Leakage into the damaged steam generator and the associated steam syrtem would continue until the RCS pressure was reduced to this value.
l SGTR sequences involving failure to close the MSIVs are designated as event sequences 3, 6A, and GB in Section 3.4.5.
3.4.4 Tube Ruptures \ffecting Multipie Steam Generators (Sequences 7A, 7B)
Should each aeam generator be affected by one or more SGTRs, the operator would be required to cool down and depressurize the RCS using at least one faulted steam generator. The use of the faulted steam generator would result in continuous leakage of primary coolant into the secondary system during the entire cooldown process. This would lead to increased releases of radioactive material into the environment. It is assumed in this analysis that the primary coolant would be cooled by opening the MSIV for one of the faulted steam gener-ators and dumping steam to the condensers in lieu of opening the atmospheric dump valve to minimize the offsite release.
If multiple SGTRs affecting multiple steam generators occer in combination with a loss of secondary system integrity, the sequence of events leading to core melt would be expected to occur over a longer period of time than if the same number of SGTRs occurred in only one steam generator. This gives the operator more time ;
to depressurize the RCS. This is due to the likelihood that the loss of secondary i system integrity (MSLB, stuck-open safety valve, etc.) will affect only one of the secondcry system loops, leaving the other secondary loop or loops with partial ur complete pressure control capability, respectively. Maintenance of elevated pressures in the other loops reduces or eliminates the leakage into these loops.
Thus, for a given number of SGTRs in combination with a loss of secondary sys-tem integrity, it is conservative to consider the SGTRs to occur in one steam generator.
Given a multiple steam' generator tube rupture occurrence, the probability that the ruptures will occur in more than one steam generator is less than one since the degradation will sometimes be more advanced or more widespread in one steam generator than in the other steam generators. The staff has assumed a proba-bility of 0.5 that a multiple tube rupture event will involve both steam generators of a two-loop' plant as a reasonably conservative estimate.
NUREG-0844 3-15
SGTR sequences involving tube ruptures in each steam generator of a PWR plant are designated as event sequences 7A and 7B in Section 3.4.5.
3.4.5 RWST Depletion Time Calculations In all SGTR scenarios, whether or not there are other compounding failures, there is a loss of RWST inventory out the break. This fluid is permanently lost for the function of maintaining RCS inventory since there would be no way of providing a recirculation flow path from the steam generator to the charging pump c Mfety injection pump sections.* Calculations have shown (memorandum, July 1 1982 and NUREG-0937) that in the case of large MSLBs with 1, 5, or 10 double anded tube breaks, the core remains covered and cooled, with adequate means of removing core decay heat, but the operator must cool and depressurize the RCS to stop the loss of RWST fluid. The time when the RWST would be depleted without this cooldown depends, to a large extent, on the specific plant. The calculations show that, in general, as expected, the time to deplete the RWST is lower for a larger number of broken tubes. Tables 4A, 4B, and 4C show the plant status at 2500 seconds. Those results were taken from calculations performed for the staff by Los Alamos National Laboratory (LANL) to support USI A-3, A-4, A-5 resolution, and are similar to those described in Section 3.4.1.
It should be noted that for W and B&W plants in the cases of five or more SGTRs with an MSLB, the RCS depressurizes to the point of accumulator injection, which helped to refill and stabilize the pressure. The CE plant pressure did not drop to the accumulator injection point because of the relatively low accumula-tor injection pressure. These calculations take no credit for operater action to affect RCS depressurization. The RCS cooldown rate shown indicates that the SI flow alone, in general, is sufficient to cool the RCS at an appreciable rate.
However, for CE, the system temperature was stable for 1 and 5 SGTRs.
' Table 5 shows the summary of systems performance for SGTRs with a partial loss of secondary integrity. These calculations (letter, December 20, 1982) were done for Zion 1 assuming the secondary safety valve stuck open upon initial opening. In these calculations, limited operator actions were assumed to man-ually trip all RCPs, to throttle HPI, and to initiate an RCS cooldown using the intact steam generators 15 minutes into the event. The results shown in Table 5 are taken before operator actions were assumed at 900 seconds, thus represen-ting a "no operator action" case.
- For cases involving a total or partial loss of secondary system integrity j this is obvious. For other cases, fluid may flow back into the primary system once RCS pressure is reduced below damaged SG pritsure.
4
! NUREG-0844 3-16 E
I
Table 4A Summary of systzs response l to SGTRs with tot:1 loss of secondary integrity (B&W, TMI-1)
No. of RCS pres- RCS RCS cooldown RCS leak- Hours to tubes sure, psig temp., *F rate, *F/hr age, gpm 2 deplete RWST3 1 2030 548 0 414 15.7 5 870 485 200 1275 5.1 10 625 485 160 1275 5.1 1 Comparisons are at 1000 seconds.
2 System is repressurizing, hence break flow is increasing.
3RWST capacity in 388,000 gallons.
Table 4B Summary of systems response 1 to SGTRs with total loss of secondary integrity (CE, Calvert Cliffs)
No. of RCS pres- RCS RCS cooldown RCS leak- Hours to2 tubes sure, psig temp., F rate, *F/hr age, gpm deplete RWST 1 5073 440 Stable 191 34.9 5 362 395 Stable 765 8.7 10 290 368 162 1200 5.6 1 Comparisons are at 1500 seconds.
2RWST capacity is 400,000 gallons.
3 Pressures are slowly rising in all cases.
Table 4C Summary of systems response l to SG1Rs with total loss of secondary integrity (Westinghouse high head, Zion)
No. of RCS pres- RCS RCS cooldown RCS leak- Hours to tubes sure, psig temp., F rate, F/hr age, gpm deplete RWST2 1 1522 467 65 350 18.5 l 2 1130 413 325 733 8.8 1 5 652 359*F 400 1200 5.4 l 10 NC NC NC NC NC 203 190 378 2500-60004 1 to 2 Note: NC - Not calculated 1 Comparisons are at 1000 seconds, i 2RWST capacity is 389,000 gallons. l 3These values do not come from specific transient calculations, but from scoping calculations based on HPSI and LPSI characteristics and estimated break characteristics for a W low-head plent.
4For scenarios involving a stiick-open SG safety valve, the flow would be limited to the valve capacity of approximately 2000 gpm.
l NUREG-0844 3-17 l
\
l
Table 5 SGTRs and stuck-open SG safety valves (Zion)
No. of RCS pres- RCS RCS cooldown RCS leak- Hours to tubes sure, psig temp., *F rate, *F/hr age, gpm deplete RWST' I 1 11601 540 -
320 20.3 l 2 11602 510 200 640 10.1 2 Pressure is dropping.
i 2 Pressures are equal at 15 minutes, although pressure is still decreasing
- for the 2-SGTR case and is stable for the 1-SGTR case.
Table 6 shows the systems response during events in which the MSIV fails to close, and the damaged SG pressure is maintained at about 950 psig by either the steam dump system or by the atmospheric dump valves.
Table 6 Summary of systems response to single and multiple tube rupture with failure of the MSIV Equilibrium Equilibrium Hours to Number of RCS pressure RCS-SG 1eak depletion tubes ruptured (psig) (gpm) of RWST 1 1370 380 13 '
2 1200 600 8 L
, 5 1000 760 7 -
10 930 810 6 '
i 20 or more s900 830 6 t
d 3.4.6 Event Sequences -
The following material describes P
'e event sequences for this category of SGTR failures. The probabilities that the operator would fail to take action are based on failure of the operator to cool down and depressurize the RCS within I
the estimated time periods to RWST depletion as given in Tables 4, 5, and 6.
Event sequences 4A, 48, 5A, 5B, 6A, and 6B involve multiple tube ruptures which i are assumed to occur in a single steam generator. The multiple tube rupture frequencies assumed in these sequences are 50% of the frequencies given in Sec-i tion 3.1.2.2, because it has been conservatively assumed in sequences 7A and 78 I
that the other 50% of the frequencies given in Section 3.1.2.2 involve at least one tube rupture in each steam generator at the plant. !
4
(
l l
4 NUREG-0844 3-18 i
Sequence 1 Event Probability (1) Single SGTR 1.5 x 10 2/RY '
(2) SG overfill 0.25 (3) SG safety valve challenge (i.e., ADV not available) 0.50 (4) SG safety valve fails open 10 1 (5) Failure to depressurize RCS to atmospheric pressure before 10 3 RWST is exhausted (operator error - s20.3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> available for a 340'F cooldown - from Table 5) 1.9 x 10 7/RY Sequence 2 Event Probability (1) Single SGTR 1.5 x 10 2/RY (2) SG overfill 0.25 (3) Main steamline failure consequential 10 3 (4) Failure to depressurize P.CS to atmospheric pressure before 10 3 RWST is exhausted (operator error - s)5.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> available for a 350 F cooldown - from Table 4A) ;
3.8 x 10 9/RY Sequence 3 Event Probability (1) Single SGTR 1.5 x 10 2/RY (2) MSIV failure to isolate SG 10 3 (3) Failure to depressurize RCS to 900 psig before RWST 10 4 is exhausted (operator error - 13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br /> available to depressurize RCS to 900 psig to stop break flow -
from Table 6) 1.5 x 10 9/RY Sequence 4A Event Probability (1) Multiple SGTR (2 to 10 tubes) 8 x 10 4/RY*
(2) SG overfill 1. 0 (3) % dafety valve challenge (ADV not available) 0.5 (4) :~ safety valve sticks open 10 1 (5) Failuro to depressurize RCS to atmospheric pressure bafore 10 2 RWST is exhausted (operator error - 5.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> available for a 285*F cooldown - from Table 4A) 4 x 10 7/RY
- This value represents 50% of the frequency estimate given in Section 3.1.2.2 to be consistent with the frequencies assumed for event sequences 7A and 7B.
NUREG-0844 3-19
S quence 48 Event Probability (1) Multiple SGTR (> 10 tubes) 2 x 10 5/RY*
(2) SG overfill 1.0 (3) SG safety valve challenge (ADV not available) 0.5 (4) SG safety valve sticks open 10 1 (5) Failure to depressurize RCS to atmospheric pressure before 10 1 RWST is exhausted (operator error - 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> available for a 170*F cooldown - from Table 40) 1 x 10 7/RY Sequence SA Event Probability (1) Multiple SGTR (2 to 10 tubes) 8 x 10 4/RY*
(2) SG overfill 1.0 (3) Main steamline failure 10 3 (4) Failure to depressurize RCS to atmospheric pressure before 10 2 RWST is exhausted (operator error - 5.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> available for a 285*F cooldown - from Table 4A) 8 x 10 8/RY Sequence 5B Event Probability (1) Multiple SGTR (> 10 tubes) 2 x 10.s/RY*
l (2) SG overfill 1. 0 l (3) Main steamline failure 10 3
- (4) Failure to depressurize RCS to atmospheric pressure before 0.5 RWSf is exhausted (operator error s1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> for a 170*F or less cooldown - from Table 40) l 1.0 x 10 8/gy Sequence 6A Event Probability (1) Multiple SGTR (2 to 10 tubes) 8 x 10 4/RY*
(2) MSIV fails to isolate affected SG 10 3 (3) Failure to depressurize RCS before RWST is exhausted 10 3 (operator error s7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> for depressurization to 900 psig - from Table 6) 8 x 10- "/RY
- This value represents 50% of the frequency estimate given in Section 3.1.2.2 io be consistent with ..le frequencier assumed for event sequences 7A and 78.
NUREG-0844 3-20
Seque ice 6B Probability E_vej .
(1) Multiple SGTR (more than 10 tubes) 2 x 10 5/RY*
(2) MISV fails to isolate SG 10 3 10 3 (3) Failure to depressurize RCS before RW5T is exhausted (operator error - 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> for depressurization to 900 psig - from Table 6) ,
2 x 10 'I/RY ,
Sequence 7A Probability Event (1) Multiple SGTRs (2 to 10 tubes) (affecting all SGs) 8s 10 4/RY (2) Intentional opening of MSIV and steam dump valve for 7.0 at least 1 SG i (3) Failure to depressurize RCS before RWST is exhausted 10 3 I (operator error - 5.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> to cool RCS from 485'F to 350*F where RHR system is activsted - from Table 4A which is extremely conservr'..ive for this case) 8 x 10 '
Sequence 7B Probability Event (1) Multiple SGTRs (10 or more tubes) (affecting all SGs) 1.5 x 10 5/RY (2) Intentional opening of MSIV and steam dump valve for at 1.0 least 1 SG.
(3) Failure to depressurize RCS before RWST is exhausted 10 2 (assumed value - depletion time estimates cannot be determined from Tables 4, 5, or 6) 1.5 x 10-'
Sequence 8A Probability Event (1) Large steamline break (containment wall to MSIV) 10.a/RY (peak op = 2600 psid)
(2) Consequential single SGTR 2.5 x 10 2 '
(3) Failure to depressurize RCS to atmospheric pressure before 10 3 RWST is exhausted (operator error - $15.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> available j for a 350*F cooldown - from Table 4A) 2.5 x 10.sfgy j i
i l
- This value represents 50% of the frequency estimate given in Section 3.1.2.2 to l
' be consistent with the frequencies assumed for event sequences 7A and 78. ;
i NUREG-0844 3-21
. _ _ __. . . = _ _ _
Sequence 88 Event Probability (1) Large steamline break (containment wall to MSIV) 10 8/RY (peak Ap = 2600 psid)
(2) Consequential multiple SGTRs (2 to 10 tubes) 2.5 x 10 2 (3) Failure to depressurize RCS to atmospheric pressure before 10 2 RWST is exhausted (operator error - 5.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> available for a 285'F cooldown - from Table 4A) 2.5 x 10 7/RY Sequence 8C Event Probability (1) Large steamline break (contairment wall to MSIV) 10 3/RY (peak op = 2600 psid)
(2) Consequential multiple tube ruptures (more than 10 tubes) 5 x 10 4 (3) Failure to depressurize RCS to atmospheric pressure before 0.5 RWST is exhausted (operator error - s1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> for a 170*F cooldown - from Table 40) 2.5 x 10 7/RY Sequence 9A Event Probability (1) Stuck-open SG safety valve (B&W plants only) 10 2/RY (peak op = 2220 psid)
(2) Consequential single SGTR 1.3 x 10 2 (3) Failure to depressurize RCS to atmospheric pressure 10 3 before RWST is exhausted (operator error - s20.3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> available for a 340*F cooldown - from Table 5) 1.3 x 10 7/RY Sequence 9B Event Probability (1) Stuck-open SG safety valve (B&W plants only) 10 2/RY (peak op = 2220 psid)
(2) Consequential multiple SGTRs (2 to 10 tubes) 1.3 x 10 2 (3) Failure to depressurize RCS to atmospheric pressure 10 2 before RWST is exhausted (operator error - 5.1 hours1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> available for a 285*F cooldown - from Table 4A) 1.3 x 10 8/RY NUREG-0844 3-22
Sequence 9C Probability Event- f (1) Stuck-open SG safety valve (B&W plants only) ,
(peak op = 2220 psid) 10 3/RY (2) Consequential multiple SGTRs (>10 tubes) 2.5 x 10 4 (3) Failure to depressurize RCS to atmospheric pressure 10 2 before RWST is exhausted (operator error - 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> available for a 170 F cooldown - from Table 40) 2.5 x 10.s/RY ,
3.5 Core-Melt Sequences A summary of all of the sequences contributing to the core-melt probability for SGTR events is shown in Tab'.e 7. This table presents the event sequences, ,
core-melt probabilities, and associated risk estimates. These risk estimates are based upon consideration of potential releases and calculations of the potential health effect consequences of such releases as disc > ed below.
It can be seen from Table 7 that initiating event SGTRs, wb e frequency is known from operating experience, contribute only L1 x 10 6/RY to the overall probability of core melt from SGTR-related cause,s. The t alance of the core-melt estimate is associated with consequent. lei event SGTRs whose frequency is more uncertain, particularly consequential SGTRs involving multiple tube l ruptures. The staff has made a number of conservative assumptions to ensure that actual multiple tube frequencies are not grossly underestimated. However, this may have resulted in conservative estimates of the frequency of multiple tube ruptures and the corresponding probability of core melt.
3.5.1 Determination of Radionuclide Releases None of the core-melt sequences listed in Table 7 has been subjected to detailed scenario-specific analysis of potential radionuclide releases to the atmosphere.
Qualitative considerations have, however, led to judgments about release poten-tials based upon analogies to sequences and release characteritations used in the Reactor Safety Study (NUREG-75/014). -
For events involving multiple tube ruptures and a total loss of secondary integrity, it is important to determine the amount of fluid above the broken tubes because when core damage commences and fission products are released, water in tSe stt.am generator above the broken tubes presents a mechanism for retainics tnose fission products. The water would significantly reduce the amount of fission products escaping to the environment, and would tend to i lo% r the predicted release. ,
For large MSLBs and large numbers of broken steam generator tubes, the system !
response can be broken into three phases: (1) initial blowdown of the fault u i SG; (2) dry SG, HPSI operation, cooling of the RCS, and loss of mass into he i SG which is gradually filling as RCS temperature drops; (3) RWST expended, heat !
transfer to the damaged SG, break flow to damaged SG. Using the results shown in Table 4C, the staff calculated the RCS temperature and faulted SG contents NVREG-0844 3-23 l
t
Tabla 7 Summary of probabilities, censequences, and risks fcr SGTR cy:nts lcading to c::re melt E
A Mean consequrces Risk
?
- $ Probability Release Deaths / event Deaths /RY
' ^
Sequence (1/RY) type Early Latent Early La wnt Les of Secondary System Integrity 1 Single SGTR + SGSV stuck open 1.9 x 10 2* 2 0 450 0 8.6 x 10 5 2 Sing;e SGTR + MSLB 3.8 x 10 9 2 0 450 0 1.7 x 10 8 3 Sing % SGTR + MSIV Failure 1.5 x 10 9 3 0 25 0 3.8 x 10 8 4A Mult. (2-10) SGTRs + SGSV stuck open 4 x 10 2 2 0 450 0 1.8 x 10 4 4B Mult. (>10) SGTRs + SGSV stuck open 1 x 10 7 1 11.7 1500 1.2 x 10 8 1.5 x 10 4 SA Mult. (2-10) SGTRs + MSLB 8 x 10 9 2 0 450 0 3.6 x 10 8 SB Mult. (>10) SGTRs + MSLB 1.0 x 10 8 1 11.7 1500 1.2 x 10 2 1.5 x 10 s 7A Mult. (2-10) SGTRs in mult. SGs 8 x 10 7 3 0 25 0 2.0 x 10 5 1 78 Mult. (>10) SGTRs in mult. SGs 1.5 x 10 2 3 0 25 0 3.8 x 10 8 i
y 8A MSLB + single SGTR 2.5 x 10 8 2 0 450 0 1.1 x 10 5 m 8B MSLB + multiple SGTRs 2.5 x 10 7 2 0 450 0 1.1 x 10 4 8C MSLB + multiple (>10) SGTRs 2.5 x 10 2 1 11.7 1500 2.9 x 10 8 3.8 x 10 4 9A SGSV stuck open + 1 SGTR 1.3 x 10 2** 2 0 450 0 5.9 x 10 5**
9B SGSV stuck open + 2-10 SGTRs 1.3 x 10 8** 2 0 450 0 5.9 x 10 4**
9C SGSV stuck open + >10 SGTRs 2.5 x 10 8** 1 11.7 1500 2.9 x 10 7** 3.8 x 10 5**
LOCA 1 LOCA + mult. (;0-20) SGTRs 2.5 x 10 7 2 0 450 0 1.2 x 10 4 Loss of Decay Heat Removal (DHR) 1 SGTR + lost of DHR 4.4 x 10 7 2 0 450 0 2.0 x 10 4 2 2-10 SGTRs + loss of DHR 3.5 x 10 8 2 0 450 0 1.6 x 10 5 3 >10 SGTRs + loss of DHR 7.0 x 10 18 1 11.7 1500 8.2 x 10 9 1.1 x 10 8 ATWS 1 ATWS + single SGTR 6.8 x 10 7 2 0 450 0 3.1 x 10 4 2 ATWS + 2-10 SGTRs 2.8 x 10 2 2 0 450 0 1.3 x 10 4 3 ATWS + >10 SGTRs 5.3 x 10 9 1 11.7 1500 6.2 x 10 8 8.0 x 10 8 TOTAL (W. CE plants) 3.9 x 10 8 4.3 x 10 8 1.7 x 10 3 (B&W plants) 5.3 x 10 8 4.6 x 10 8 2.4 x 10 3
- 1.9 x 10 7 = 0.00000015
- Estimates for event sequences 9A, 9B, and 9C are app icable to B&W plants only. 7
_ - _ - _ _ _ _ _ _ . _ ~ -.- - - - - . .
i when the RWST is depleted. Using conservative assumptions,* the staff calculated that there would be significant water inventory in the faulted SG at the time the core first uncovers. However, staff calculations could not assure the break would be covered in all cases for all plants. The staff has examined the event sequences and th .r associated release categories given in the' Reactor Safety Study (NUREG-75/014) and has classified the release characterizations for SGTR events shown in Table 7 on the basis of similarities of sequences with respect to radionuclide release transport, and deposition or decontamination mechanisms. l Release Type 1 i= a modified PWR-3 release category as described in NUREG-75/014.
The following factors have been changed in order to achieve a better estimate of actual consequences.
Factor PWR-3 Modified (Release Type 1)
- Time to release, hr 5 1 Warning time, hr 2 1 Release height, m 0 10 Release Types 2 and 3 represent NUREG-75/014 PWR-4 and PWR-6 releases, respectively.
3.5.2 Calculation of Consequences and Risks The risks of the accident sequences presented in Table 7 represent estimates for a typical PWR. Selected accident consequences were determined using "CRAC,"
a code provided by Sandia National Laboratories (latent and early fatalities).
The release fractions and release characteristics for all three scenarios, except as noted above, are taken from Table 5-1 of the Reactor Safety Study (NOREG-75/014). Byron Station was used in the evaluation as a representation of an "average" PWR site.** The consequences shown in Table 7 are conditioned upon the occurence of the event sequence and are shown in terms of early fatal-ities and latent fatalities. Using the probabilities of each sequence occurring, the total risk (i.e., consequences multiplied by probability of occurrence) to the public per year of plant operation is also shown in Table 7. It should be noted that the consequences and risks could be larger or smaller than shown in Table 7 because of such factors as population, meteorology, and protective action assumptions, but the relationships among impacts of various sequences :
should remain more or less constant.
i 3.6 Non-Core-Melt Sequences Since all the sequences that involve a loss of secondary integrity also involve a significant release of primary coolant to the environment, even without a l
. l
- The staff assumed no heat transfer to the intact SGs and no operator action i
. to reduce RCS pressure or to initiate auxiliary feedwater to add inventory to the faulted steam generator.
- Population distribution, protective actions, and meteorological input to the code were equivalent to those presented in the Byron Station Final '
Environmental Statement (NUREG-0848).
a 3-25 l NUREG-0844 1 !
t
- - - - ~ _ .
core melt, the probability and consequences of the corresponding non-core-melt sequences have been analyzed. Although these sequences assume correct operator action to depressurize the RCS to atmospheric pressure at a rate corresponding to a 100*F/hr cooldown, they also involve postulated failures beyond the current design basis. Therefore, the probability of these sequences is the same as that for sequences 1 through 9 above, except for the assumption relative to operator action. These releases did not include possible overpower-induced cladding
, failures (pellet-cladding interaction (PCI)), neither were undercooling-induced
- cladding failures (departure from nucleate boiling (DNB)) included. It is not
- expected that undercooling-induced cladding failures would occur; however, the staff is currently assessing the potential for overpower-induced fuel failures under a variety of accident and transient conditions.
The calculations below specifically address SGTR event sequences involving a stuck open safety valve (other event sequences are discussed later). For these events, the staff considered the source of radioactive iodine (initial coolant activity and iodine spiking); primary to secondary transport; partition and water transport after leaving the secondary system; and meteorology using best estimates and conservative estimates for each of these items, as summarized in Table 8.
Table 8 Range of parameters considered Most Best-estimate Parameter conservative value Initial coolant activity 1 pCi/gm 0.1 pCi/gm Iodine spiking Design basis 10 to 20% of (500x for 2 hrs) design basis Primary to secondary iodine ---
Best estimate transport calculations Partition (and plateout) 1.0 1. 0 in SG Partition and wau r transport 1.0 1.0 (iodine held after leaving SG in water, water transported in i small droplets)
Meteorology (x/Q) for site dose 1x10-3 2x10 4 (average calculations meteorology)
Meteorology for estimating ---
Best estimate health effects CRAC code calculations
, NUREG-0844 3-26
The bases for the estimates given in Table 8 are discussed b@ low.
Initial Iodine Activity The Standard Technical Specifications (STS) limit the primary coolant activity to 1 pCi/gm. Although some plants have higher limits and some exceed the STS value for a short period of time, the 1 pCi/gm appears to be a reasonable upper limit for an initial value. The NRC report on fuel performance, "Fuel Perform-ance Annual Report for 1981," NUREG/CR-3001, was consulted to determine how fre-quently the 1 pCi/gm value was exceeded. The staff found that 9 PWRs exceeded
, the limit on one or more occasions for a total of 33 times. However, in every case, the limit was only exceeded for a short period of time. The highest steady-state operating value was approximately 0.5 pCi/gm. The best estimate value of 0.1 pCi/gm was chosen as representative of a plant operating with sev-eral failed fuel rods. Although many plants run with lower coolant activity, some operate with values 2 to 5 times the best-estimate value.
_ Iodine Spiking The range of iodine spiking was assumed to vary from 10% to 20% of the design-basis spike (best-estimate assumption) to 100% of the design-basis spike of 500 times the steady-state iodine production rate for two hours following a reactor To establish a best estimate-value, primary coolant activity values were trip.
calculated as a function of time following a reactor trip using various assump-tions about the size of the iodine spike. The calculated increase in iodine concentration was then compared to the iodine spikes reported in the Fuel Performance Annual Report for 1981 (NUREG/CR-3001). Although there is consider-able variation from event to event and some uncertainty about exactly when the ;
post-trip samples were taken, it appears that an iodine spike of 10% to 20% of l the design basis spike is consistent with operating experience.
Primary-to-Secondary Iodine Transport ,
The transport of iodine from the primary system to '.,he steam generator was calculated with a best-estimate time-dependent (finite difference) code which accounts for dilution by ECCS, mixing in the RCS, leakage to the SG, radioactive decay, cleanup (letdown and SG blowdown), partition, carryover to the condenser, ,
feedwater addition and release to the atmosphere.
Partition in the Steam Generator For event sequences 1, 4A, and 4B involving an SGTR followed by SG overfill, and a subsequent stuck open-safety valve, the assumption of no partitioning (i.e.,
PF = 1.0) represents the best estimate for these conditions (NUREG/CR-2659).
For event sequences 9A, 9B, and 9C involving a stuck open safety valve followed by blowdown of the steam generator and rupture of one or more tubes, the effec-tive partition factor will depend on whether the steam generator water level is above or below the r'upture location (s). If the rupture (s) is not covered by water, then all primary liquid leaking into the secondary flashes into steam before it has a chance to mix with the secondary water. Thus, for uncovered tube ruptures, no partitioning (PF = 1.0) is assumed in the steam generator.
For purposes of assessing non-core-melt releases, the staff has conservatively assumed th tube rupture' location to be uncovered.
NUREG-0844 3-27
Partition and Water Transport After Leaving the Steam Generators The blowdown of a steam generator through a stuck-open safety / relief valve will be in the form of a two phase mixture of steam and water droplets. The size of the water droplets during a high pressure blowdown is expected to be small (10 pm) based on drop size predictions made for LOCA blowdown conditions (Gido, R. G., and A. Koestal, 1978). A mixture with such small droplets resembles a fog or mist. For the expected range of weather conditions, drop-lets of such a small size would be expected to stay in suspension ir, the atmos-phere and be carried away from the point of release. An analysis of iodine released during a steam line break with a concurrent steam generator tube rupture supports +,his conclusion (NUREG/CR-2659). In that analysis, the reten-tion on site of liquid released from the steam generator was neglected partially based on containment system experiments. These experiments showed that only approximately 20% of the mass of water exiting from the break into the atmos-phere was retained close to the point of discharge.
Meteorology The range of X/Q values in Table 8, which were used to calculate site boundary doses, represent the range from best-estimate conditions to typical FSAR 95%
meteorology. The best-estimate value was chosea to represent average conditions for a typical site. The offsite health effects (eean, latent deaths) were I calculated with the CRAC code. These were best-estimate calculations for the Byron site and were done in the same manner as for final Environmental Statement reports.
Results Table 9 summarizes the consequences of non-core-melt SGTR sequences involving a stuck-open steam generator safety valve. Table 10 summarizes the probabilities of these events and the corresponding public risk. SGTR sequences involving steamline breaks (event sequences 2, 5A, 58, 8A, 8B, and 80), stuck-open MSIVs ;
(sequences 3, 6A, and 6B), and intentionally opened MSIVs / sequences 7A and 78) will also cause radiological releases, but are not domit contributors to the t
- risk from non-core-melt releases. SGTRs occurring in conunction with a main steamline break have consequences similar to those indicated in Table 9 for event sequences 9A, 98, and 9C, but occur at a lesser frequency. SGTR events '
i involving an open MSIV (whether accidental or intentional) involve relatively i minor releases associated with the release of noble gas from the condenser, and i possible release of steam through the auxiliary feedwater pump turbine.
1 ,
l The site boundary doses in Table 9 are based on accumulated doses at the site '
- boundary over 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />. Thus, these doses are not directly comparable to the
) 10 CFR Part 100 limits, which are based upon accumulated dose for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> imme- ,
diately following onset of the postulated fission product release. For event j sequence 1, the 2-hour site boundary doses are approximately 20% of the 6-hour *
- values given in Table 9. For event sequences 9A, 98, and 9C, the 2-hour site boundary doses are approximately 60% of the 6-hour values given in Table 9 for -
single and multiple tube ruptures.
It can be seen that with best estimate assumptions on initial coolant icdine cencentration, iodine spiking, and meteorology, the of fsite boundary doses are (
! i 1 :
- NUREG-0844 3-48 !
I
-. - - ._ _. -- - .. __ _ _ _ . - - _ _ _ _ - = - _
1 i
Table 9 Consequences of noi.-7 rs 's- tGTP :. -
,vol ..u a stuck-epel S- afe v-
. L. , - i
.., -lea. - -
Over Sequence es+G^ si
- --2??3 1 Sing.e SGTR + overfill .06- f
+ SV stuck open ;
4A 2-10 SGT" 4 overfill . V 750
+ SV stuck open 4B > 10 SGTRs + overfill 130 . ' , T/50
, + SV stuck open j 9A Stuck-open SV + 122-SP 12-2600 1 SGTR 98 Stuck-open SV + 131-5600 13-2800 !
2-10 SGTRs e
i 1
9C Stuck-open SV + 133-5650 13-2800 [
> 10 SGTRs r
1 Range associated with best estimate and conservative estimate assumptions on initial coolant iodine act.ivity and iodine -
spiking. -
l 2 Range associated with best estimate and conservative estimate 1 4 assumptions on initial coolant iodine a.ctivity, iodine spiking, and meteorology.
8 Accumulated doses at the site boundary over 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, small compared to the consequence limits of 10 CFR Part 100. However, if un- ;
l favorable assumptions are made regarding tiiese parameters, the 2-hour offsite boundary doses may challenge or exceed these guidelines. l l
The risk estimates in Table 10 are very small and are based on the best-estimate i parameters given in Table 8. Events involving worst case values of these para- ;
meters are expected to occur at a mucn lower frequency than the overall fre- '
quency of SGTR events involving a stuck open safety valve (i.e., 4.9 x 10 4/RY ,
for B&W plants and 2.3 x 10 4/RY for W and CE plants) shown in Table 10 and, i therefore, are not expected to be dominant contributors to the risk from non- .
core releases. l l
)
l NUREG-0844 3-29 i
l
Table 10 Summary of proMilities and risks from non-core-melt SGTR sequences involving a stuck-open r safety valve v
Probability Risk of latent Sequenco (RY) fatality deaths /Rf 1 Single SGTR + overfill '1.9 x 10 4 2.4 x 10 7
+ SV stuck open 4B > 10 SGTRs + overfill 1 x 10 8 1.6 x 10 8
+ SV stuck open 9A Stuck open SV + 1.3 x 10 4* 1.9 x 10 7*
1 SGTR 98 Stuck-open SV + 1.3 x 10 4* 2.1 x 10 7*
2-10 fGTRs 9C Stuck open SV + 2.5 x 10 7* 4.0 x 10 10*
> 10 SGTRs i Total (W, CE plants) 2.3 x 10 4 3.0 x 10 7 j (B&W plants) 4.9 x 10 4 7.0 x 10 7
- Estimates for sequences 9A, 9B, and 90 are applicable to j B&W plants only.
i 3.7 Conclusions -
The foregoin 1 conclusions:g risk analysis carried out by the staff leads to the following (1) Although there are significant uncertainties inherent in the staff's i analyses, the analyses contain a number of conservatisms to minimize the potential for grossly underestimating risk. >
{
(2) Th staff's analyses indicate that the core-melt probability from all SGTR-related causes is small, about 5.3 x 10 8/RY for B&W plants and f i
3.9 x 10 8/RY for W and CE plants. These probabilities are a relatively f small fraction (10% or less) of the overall probability of core-melt events ;
from all causes based on probabilistic risk assessments that have been !
performed for a number of PWRs. The corresponding risk to the public is i estimated to be limited to 2.4 x 10 3 (B&W plants) and 1.7 x 10 3 (W and CE i plants) latent fatalities /RY and 4.6 x 10 8 (B&W plants) and 4.3 x 10 8 (W
melt based on calculations performed for a reprosentative PWR site-(Byron).
[
(3) SGTRs occurring in conjunction with a non-isolatable loss of secondary j system integrity can lead to significant offsite releases (comparable to !
NUREG-75/014 PWR release categories 8 and 9), even if core melt does not
[
NUREG-0844 3-30
)
t I
4
occur. The probability of SGTRs occurring in conjunction with a stuck- l open safety valve is estimated to be 4.9 x 10 4/RY for B&W plants and 2.3 x 10 4/RY for W and CE plants. Site-boundary doses would typically be expected to be smaT1 relative to the consequence limits of 10 CFR Part 100 based on best-estimate assumptions regarding coolant iodine activity con-centrations, ' adine spiking, and meteorology. Ths use of conservative assumptions regarding these parameters leads to site-boundary dose esti-mates which may challenge or exceed the 10 CFR Part 100 limits; however, ,
the frequency of such doses is significantly less than the above 2.3 x 10 4 to 4.9 x 10 4/RY estimate. The staff estimates public risk from non-core-melt releases to be very smail: about 7 x 10 7 latent fatalities /RY for ;
B&W plants and 3.0 x 10 7 latent fatalities /RY for W and CE plants.
(4) On the basis of the staff and SAI evaluations of the risk from SGTR acci-dents, as discussed above, the staff finds that SGTR events beyond the design basis do not contribute a significant fraction of the early and latent cancer fatality risks associated with other reactor events at a given site. Furthermore, the risk assessment indicates that the increment in risk associated with SGTR events is a small fraction of the accidental and latent cancer fatality risks to which the general public is routinely j exposed.
d t
J j
< t 1 I I
i NUREG-0844 3-31 1
4 NRC STAFF ACTIONS AND COMPLETED ITEMS 4.1 Introduction This section includes a discussion of those issues identified by the staff in 1982 (following the SGTR event at Ginna) as warranting further staff action or study. This includes a description of the initial bases and concerns leading to consideration of these actions and the current status of these actions.
Table 11 provides a summary listing of these actions and their status.
These NRC staff actions have been designated as Generic Issue 67. A prioriti-zation evaluation of these staff actions has been performed and incorporated into NUREG-0933, "Prioritization of Generic Issues," Supplement III. Resolution of these actions is being monitored by the staff's Generic Issue Management ,
Control System (GIMCS).
A number of the staff actions involve broad generic issues extending beyond i issues strictly related to steam generators. These include staff actions shown in Table 11 which relate to "Organizational Response" and which have been completed. These also include actions which are being addressed as part of other existing regulatory programs such as "Pressurized Thermal Shock" (USI l
A-49), "Improved Accident Monitoring" (NRC Generic Letter 82-33), "Reactor Vessel Inventorg Measurement" (THI Task Action Plan (TAP) II.F.2, NRC Generic
> Letter 82-28), Guidance on Reactor Coolant Pump Trip" (TMI TAP II.K.3.5, NRC Generic Letter 82-33), "Control Room Design" (THI TAP 1.D.1, NRC Ger.eric letter 82-33), and "% proved Emergency Operating Procedures" (TMI TAP 1.C.1, NRC Generic Letto 82-33). Completion of these broad generic tasks is considered !
to be outside the scope of the staff's integrated program to resolve "Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Generator Tube Integrity."
The remaining staff actions identified in Table 11 involve other issues related i to i, team generators. As noted in the table, a numbe* of these remaining staff actions are relatively low priority tasks which will remain inactive pending completion of higher priority tasks and availability of staff resources. Others of these remaining staff actions, as indicated in Table 11, are actively being Nrsued as part of Generic Issue 135, "Steam Generator and Steam Line Overfill
! Issues," and/or the Steam Generator Group Project / Steam Generator Tube Integrity Program (SGGP/SGTIP) sponsored by the NRC Office of Huclear Regulatory Research.
In view of the low risk estimates associated with SGTR events, the staff con-cludes that the resolution of USIs A-3, A-4, and A-5 is not contingent upon com-pletion of these tasks. However, these tasks will help ensure that risk contin-ues to be low and may lead to proposals for revising existing regulatory guid-j ance and possibly requirements concerning steam generator tube inspections and repairs, revisions to the Standard Review Plan concerning the design basis SGTR, 1
i and resolution of the steam generator / steam line overfill issue. The potential 4
safety benefit and cost of implementation will be assessed for any proposals stemming from these activities. If justified by this cost / benefit analysis,
- additional or revised regulatory guidance or requirements will be issued.
NUREG-0844 4-1 l
il
z Table 11 NRC staff actions and completed items E
h Report g section Subject Action Status 4.2 Steam Generator Integrity 4.2.1 Steam Generator Tube Sleeves Develop SRP guidance for Note 1 review of tube sleeving 4.2.2 Inservice Inspection Program Develop denting inspection Note 1 for Denting program 4.2.3 Improved Eddy-Current Develop improved ECT Techniques (ECT) Generic Issue 135 requirements SGGP/SGTIP 4.2.4 Category C-2 ISI Requiren.3nts Upgrade existing requirement Generic Issue 135 SGGP/SGTIP y 4.3 Plant Systems Response m
I 4.3.1 Steam Generator Overfill Determine potential for and l Generic Issue 135 l
consequences of SG overfill 1 4.3.2 RCS Pressure Control Determine optimized pressure During an SGTR USI A-45, control procedures TMI TAP II.D.1 ,
4.3.3 Pressurized Thermal Consider Ginna information Shock Program, A-49 USI A-49 in generic USI A-49 program i 4.3.4 Improved Accident Address Ginna experiences l Monitoring Regulatory Guide by implementing RG 1.97 1.97 Program, programs Generic Ltr. 82-33 4.3.5 Reactor Vessel Inventory Measurement Observation: Ginna information TMI TAP II.F.2, supports need for TMI ceneric Ltr. 82-28 TAP II.F.2 program See footnote at end of table.
Table 11 (Continued) 1 EE
- o El Report Subject Action Status E> sectica 2 -
4.4 Human Factors Considerations Reactor Coolant Pump Trip Issue requirements pursuant TMI TAP II.K.3.5, 4.4.1 Generic Ltr. 82-33 te TMI TAP II.K.3.5 program Consider Ginna information TMI TAP II.D.1, 4.4.2 Control Room Design Review in TMI TAP I.D.1 reviews Generic Ltr. 82-33 Consider Ginna information TMI TAP II.C.1, 4.4.3 Emergency Operating Procedures Improvement in generic TMI TAP I.C.1 Generic Ltr. 82-53 4.5 Radiological Consequences 4.5.1 Reassessment of Radiological Reassess potential Note 1 4 Consequences Following a radioiogical consequences a Postulated SGTR Event i 4.5.2 Reevaluation of Design-Basis Reassess adequacy of design Generic Issue 135 SGTR basis SGTR 4.5.3 Secondary-System isolation Reassess provisions for This task has isolating the affected steam been dropped generator 4.6 Organizational Response 4.6.1 Operations Center Communi- Improve coordination and Complete cations and Notification communication by issuance of procedures and other actions See footnote at end of table.
E Table 11 (Continued) 1 9 Report l g section Subject Action Status l $
4.6 Organizational Response
! 4.6.2 Interaction Between Improve coordination Complete
! Regional Base Teams and between t;as
! the Executive Teae I
4.6.3 NRC Site Team--Location Improve team deployment Complete of Site Team Components and the role of and Public Affairs the Public Affairs Office Information Flow 1
4.6.4 Familiarization With Provide training Complete NRC Response Plan
-i
- 4.6.5 Alternate Evnuation Ensure that alternates are Complete A Routes and Sites included in licensees' I plans 4.6.6 Deescalation of Emergency Issue the needed guidance on Complete by J
Classification deescalation plans and CY-89 i procedures 1
j 4.6.7 Offsite Dose Assessment Issued needed guidance on Complete
- offsite dose assessment I methods 1
2 1Not-high priority tasks. Completion of these tasks will be scheduled commensurate with the priority nature of the work and the availability of staff resources.
i j
)
-- - - - - , en anr-~ , , - ~ - _ - - . , - , - - - - -, ,
4.2 Steam Generator Integrity 4.2.1 Steam Generator Tube Sleeves Task NRC should develop guidance governing the design, installation, and inspection of steam generator tube sleeves. The guidance should be presented in a revision of the Standard Review Plan (NUREG-0800) and should include the following criteria:
(1) Sl> eves Intended To Repair Defective Tubes (a) Sleeves shall be designed in accordance with ASME Code requirements to act as the primary pressure boundary. The design shall be based on the assumption that all original defects in the sleeved tube have penetrated completely through the wall of the tube.
(b) Criteria for inspection of sleeved tubes shall be consistent with criteria for unsleeved tubes.
(c) Criteria for plugging tubes with degraded sleeves , hall provide the same margins of structural and mechanical integrity as those of RG 1.121 for unsleeved tubes.
(d) Sleeves shall be designed so that potential leakage from all installed steam generator tube sleeves during normal operating conditions shall be limited to the rate set in the Technical Specifications. Leakage during postulated accidents shall not result in offsite doses exceed-ing a small fraction of limits set in 10 CFR Part 100.
(e) The potential for wear and/or accelerated corrosion at the joints of tubes and tube sleeves shall be included in the materials selection and design evaluation of tube sleeves.
(2) Tube-Stiffening Sleeves (a) Criteria for inspection of sleeved tubes shall meet those of RG 1.83 and the Standard Technical Specifications (STS) for unsleeved tubes.
(b) Potential for wear and/or accelerated corrosion at joints between tubes and tuk vleeves shall be included in the design and evaluation of tube sleeves Initial Basis for Considecation For severely degraded steam generators, a large number of tubes may be removed from service as a result of tube plugging. This reduces the available primary-to-secondary heat transfer surface and may ultimately force a decrease in the rated output of the plant. Some utilities faced with the prospect of derating power have elected to replace the degraded steam generators. Such replacement, however, requires an extended outage and involves considerable cost to the utility and its customers. To prolong the life of severely degraded steam gen-erators, some utilities, with NRC approval, have elected to repair defective NUREG-0844 4-5
tubes with tube siceves. Tht. advantage of sleeving over plugging is that the
- repaired tube remains functional, reducing both the loss in heat transfer area and the increase in primary coolant flow resistance.
The tube sleeving procedure invrives inserting a. tube of smaller diameter (sleeve) inside the tube to be repaired. The sleeve is positioned to span the defective portion of the original tube and is then either hydraulically or mechanically expanded above and below the defective region. The expanded joints are sometimes brazed to ensure additional leaktightness. Tube sleeving repairs have been performed to restore primary coolant boundary integrity on straight 3
accessible portions of tubing degraded by wastage, intergranular attack, and t
stress-corrosion cracking.
3 Sleeves were first used at Palisades in 1976 and 1977 when approximately three 1
dozen sleeves were installed in tubes that had been degraded by wastage.
In 1980 and 1981, more than 6000 sleeves were installed in the San Onofre Unit I steam generators to repair tubes degraded by caustic stress-corrosion cracking. Sleeving programs have more recently been conducted at Indian Point Unit 3, Point Beach Units 1 and 2, Ginna, and Millstone Unit 2.
i Applicable regulations governing the design, installation, testing, and inser-vice inspectability of sleeves include General Design Criteria (GDC) 14, 15,
, 31, and 32 of 10 CFR Part 50, Appendix A. However, there are no generic regula-tory guidelines for implementing these regulations. To date the NRC staff has ,
reviewed and approved steam generator tube sleeving programs,on a plant-specific I basis. The staff anticipates numerous additional propused sleeving programs in the coming years. Criteria for the design installation, testing and inspection t of steam generator tube sleeves are necessa,ry to ensure adequate and uniform ;
i implementation of the General Design Criteria. !
i
) In addition, such critoria can ensure that any previous difficulties with !
sleeves (e.g., sleeve and tube degradation as a result of inadequate braze !
' procedures and controls) are recognized and appropriately accounted for in !
future programs.
Review Plan. The criteria should be included in a revision to the Standard
- I 1
! Status !
i k This task is not expected to result in a significant reduction in public risk, !
but has been classified as a regulatory impact issue based on reduced industry !
i and NRC costs (HUREG-0933, Supplement III). This task will be sche.1uled pend- r i
ing the availability of staff resources, after assigning resources to approved j high- and medium priority generic issues. {
i
},
{ 4.2.2 Inservice Inspection Program for Denting j Task l The NRC staff should propose a denting inspection program for inclusion in the
! Standard Technical Specifications (STS). The program should include criteria f for establishing the scope of the inspections and acceptance criteria (i.e.,
1 I f l NUREG-0844 4-6 !
i
- l
~ i i f I
denting iimit based upon tube restriction or strain) including %e following elements and definitions:
(1) gauging or profilometry inspection of any tube that restricts passage of the standard size eddy-current probe ( _ dia.)* as a result of denting (2) gauging and profilometry inspection of any tube that restricted passage of the standard size probe ( dia.)* during a previous inspection as a result of denting or that exhibiUd more than (_%)* strain if profilometry inspections were performed (3) criteria for gauging or profilometry inspection of additicnal tubes based on the results of initial inspections (4) appropriate plugging limits for dented tubes (5) for Westinghouse steam generators (a) criteria beyond which visual inspection of the support plate flow slots for potential deformation ("hourglassing") shall be performed (b) requirement to plug all row 1 tubes if detectable flow slot deforma- -
tion ("hourglassing") is observed, unless it can be directly verified that the upper support plate flow riots are undeformed Initial Basis for Consideration At present there is no specific mention in Regulatory Guide 1.83 and there are no specific requirements in the Standard Technical Specifications (STS) to r inspect tubes for denting. ,
Operating experience has shown that surveillance of tube denting is necessary to preclude development of stress-corrosion cracking induced by denting. There has been one instance (Surry Unit 2 in 1976) in which denting led to high stresses in the U-bend region resulting in an SGTR. Plant-specific criteria have been established for plants with extensive denting; however, generic criteria have not been developed. Generic criteria would ensure that minimum standards for denting inspections are available for application as needed.
Sta g
'l This task is not expected to result in a significant reduction in public risk, but M s been categorized as a regulatory impact issue since it would produce a small .' eduction in risk and would provide a net cost benefit to the industry and to the NRC (NUREG-0933, Supplement III). This task will be scheduled pend-ing the availability of staff resources, after assigning resources to approved h!qh- and med!'.im priority generic safety issues.
i
- *To be detemined.
NUREG-0844 4-7
4.2.3 Improved Eddy-Current Techniques Task The staff should evaluate, in parallel with ongoing ASME Boiler and Pressure Vessel (B&PV) Code activities, improved eddy-current test programs for incor-poration into the ASME B&PV Code,Section V for NDE and Section XI for Inservice Inspection.
Initial Basis for Consideration The need for this task was identified as a result of staff consideration of specific proposals for improved eddy current test procedures as discussed in Section 2.3 of this report.
Status This task is being addressed as par
- of a much broader program sponsored by the NRC Office of Nuclear Regulatory Re;earch (RES); namely the Steam Generator Integrity Program / Steam Generator Group Project. This task is also being tracked as part of NRC Generic Issue 135, "Steam Generator and Steam Line Over-fill." A detailed overview of the objectives, scope, and status of the RES program is presented in hUREG-0975, Vol. 5. This program includes a number of elements pertaining to eddy current testing. These include evaluation and development of advanced eddy current test methodologies and a detailed assess-ment of the capabilities of currently used and developmental eddy current methodologies to reliably detect, characterize, and size flaws in steam genera-tor tubing. Results from these studies are being used to assess needed changes to ASME B&PV Code requirements and Regulatory Guide 1.83 concerning eddy-current testing. Topical reports detailing results and recommendations stemming from this program are expected to be published before the end of calendar year 1988.
4.2.4 Category C-2 Inservice Inspection Requirements
. Tag The NRC staff should investigate more practical alternatives to the originally proposed potential requirement concerning "Supplemental Tube Inspections" which is discussed and evaluated in Section 2.2.1.
Initial Basis for Consideration Section 2.2.1 identified a number of limitations in the existing Technical Specification requirement for steam generator tube inspections when Category initial inspection sampling. The staff considered C-2 results a potential are found industry actionduring(Section 2.2.1) which addressed these limitatio but which was found to lead to potentially significant costs. The staff should investigate actual industry practice relative to the staf f's concerns and prac-tical alternatives to the staff's initial proposal for improved supplemental tube sampling inspection practices which could be implemented on a case-by-case basis as needed.
NUREG-0844 4-8
For reasons discussed in Section 2.2.1.3, this task is not expected to result i in significant reductions in risk. However, effective steam generator tube inspection programs are an important element of an effective overall program ,
to ensure steam generator tube integrity.
Status This task is being addressed as part of a much broad 9r program sponsored by the NRC Office of Nuclear Regulatory Research (RES); namely the Steam Generator i Integrity Program /5 team Generator Group Project. This task is also being l tracked as part of NRC Generic Issue 135, "Steam Generator and Steam Line Over- i fill." A detailed overview of the objectives, scope, and status of the RES i Vol As part of this program, the reli-program is presented in NUREG-0975abilityofsteamgeneratorinspectlonpr.5.ograms; of inspection frequencies and tube inspection sample sizes. Results from these !
studies are being used to assess needed changes to Regulatory Guide 1.83 concern- l ing steam generator inspections. Topical reports detailing results and recom- t mendations stemming from this program are expected to be published before the i end of calendar year 1988. ;
4.3 Plant Systems Response !
i 4.3.1 Steam Generator Overfill (
Task The NRC staff should select a small number of PWRs representing the PWR spec-
! trum of designs and determine the potential for and consequences of steam gen- t
- erator overfill. Further NRC or licensee actions should be determined on the 4 basis of this and other studies discussed in Sections 4.3.2, 4.5.1, and 4.5.2.
Initial Basis for Consideration During the Ginna SGTR, the affected steam generator filled up to the steam-line safety valve as a result of primary-to-secondary leakage from continued j operation of the safety-injection pumps. The safety valve lifted five times ,
at successively lower pressures and failed to f'111y reseat (at least twice). I i
The failure to completely reseat contributed t the overfill problem by lower- i ing the pressure in the damaged steam genera , thus raising the differential '
! pressure across the broken tube and sustaining the leakage despite reduced !
I primary system pressure. The probability of steat generator overfill can be '
reduced by improved emergency procedures. However, there are no programs or I i systems that can ensure that the steam generator will not overfill during l future steam generator tube ruptures. I 1
I l Failure of the valve to close could provide a direct pathway for the release J of radioactive primary water to the environment (releases at Ginna were very small for the reasons cited in Section 2.9.2). This sequence of events (i.e.,
j steam generator overfill followed by a stuck-open safety valve) is beyond the i
design basis for SGTR events in SRP 15.6.3 to establish that the radiological
- consequences meet 10 CFr. Part 100.
1 i
In addition, given the potential for overfill, the integrity of the steam lines under :ombined loads from excessive water mass in the steam lines and possible waterhammer should be investigated and confirmed.
NUREG-0844 4-9 i
1 1
Status This task is being performed as part of NRC Generic Issue 135, "Steam Generator and Steam Line Overfill Issues." The staff expects to issue a report describing the results of this effort oy the end of calendar year 1988.
4.3.2 Reactor Coolant System Pressure Control During an SGTR Task
?
The NRC staff should evaluate whether further improvements should be made by owners of PWRs, emphasizing the use of existing plant systems and equipment, i to control and reduce reactor coolant system (RCS) pressure following an SGTR in order to stop primary-to secondary leakage within the time frame assumed in the final saf9ty analysis report (FSAR). The spectrum of possible initial con-ditions, RCS thermal-hydraulic conditions, and break sizes should be considered.
The use of the pressurizer auxiliary spray system should be explicitly examined because its use may eliminate the need to use the pressurizer power-operated relief valve (PORV) in cases in which forced RCS flow has been lost. The study l should address the following objectives: (1) minimizing the primary-to-secondary ,
l leakage through the broken steam generator tube; (2) maximizing control over '
! system pressure; and (3) minimizing the chances of producing voids in the RCS and other complicating effects.
Initial Bases for Corsideration Without forced reactor coolant flow, which may occur as a result of reactor -
t couvant pump (RCP) trip or a loss of offsite power, the necessary RCS depres-surization followin pressurizer spray, gRCS an SGTR is more difficult fluid contraction causedbecause of the loss by the cooldown of the from normal dump-ing of steam to either the condenser or to the atmosphere will result in some reductiun in RCS pressure, but other measures must be taken to expeditiously reduce the RCS pressure to the point at which flow into the damaged steam generator stops. The pressurizer PORV was used during the Ginna and Prairie Island SGTR events to reduce RCS pressure. However, it is difficult to control pressure with the PORV because its use creates an additional loss of coolant.
The decrease in pressure can be so rapid that the steam voids may be formed in i the upper vessel head and in the top of the U-tuoes, further complicating the i depressurization. Formation of voids can lead to concerns about core cooling, j The Ginna operators were sufficiently concerned that they left the safety in-4 jection pumps operating, thereby overfilling the steam generator that had the ruptured tube and challenging the safety valve.
j It is not apparent that the auxiliary spray from the charging system could j have successfully lowered the RCS pressure to the point at which flow out the broken tube would have stopped. It may have been that, by spraying cold charg-ing fluid into the pressurizer, the decrease in pressure would have resulted in void formation, thus expanding RCS fluid and filling the pressurizer, and
! rendering further spray flow inetfective. This phenomenon as well as the thermal stresses on the spray nozzle itself should be examined.
1 NUREG-0844 4-10 4
Status The TMI Task Action Plan item I.D.1, NUREG-0737, has within its scope the developmtnt of emergency operating procedures (E0Ps) for accidents and tran-sients, including SGTRs. Likewise, in the USI A-45 study, "Shutdown Decay Heat Removal Requirements," the staff is also developing and studying the adequacy of current and alternate means of satisfying LWR shutdown decay heat removal reqd rements. The USI A-45 study will also be looking into shutdown require-ments imposed by SGTRs in PWRs. In its prioritization analysis in NUREG~0933, Supplement III, the staff concluded that the RCS pressure control issue is being addressed as part of these staff programs.
4.3.3 Pressurized Thermal Shock Task The difficulty in understanding the temperature transient experiencqd by tSe Ginna reactor vessel raises the question of the need for improved temperature monitoring in the cold leg and in the reactor vessel downcomer. This itsue and the effects of RCS flow stagnation associated with the isolation of a !
steam generator shou'd be addressed by the pressurized thermal shock program, [
A-49. !
Initial Basis for Consideration During the Ginna SGTR, the affected steam generator was isolated and the
- reactor coolant pumps were tripped. As a result, the flow in the B reactor coolant loop was reduced to a few hundred gallons per minute, while cold, high-pressure injection water was being injected into the loop. The cold-leg piping ,
apparently experienced a cooldown of approximately 260'F in 30 minutes. The j i reactor vessel apparently did not experience this rapid cooldown, since the flow in the cold leg was in the reverse direction, that is, from the reactor i
.j vessel toward the steam generator. Other events, as discussed in NUREG-0916, i resulting in steam generator isolation and continued safety injection could l result in adding cold water to the reactor vessel. j Status ,
j Studies of the probability, consequences, and resolutior. of such events are !
j within the scope of the pressurized thermal shock (PTS) program, A-49. This !
4 program has resulted in promulgation of a PTS rule (10 CFR 50.61) on July 23, ;
i 1985 which (1) establishes a screening criterion related to the fracture resis- !
l tance of pressurized-water-reactor (PWR) vessels; (2) require analyses and a schedule for implementation of neutron flux reduction programs that are reason- l ably practicable to avoid exceeding the screening criterfon; and (3) require
- detailed safety evaluations to be performed before plant operation beyond the l i screening criterion will be considered. Guidance on meeting the requirert.ents of the new PTS rule is provided by NRC Regulatory Guide 1.154 which was issued
!' in January 1987. The NRC staff is monitoring implementation of this rule in Multiplant Action (MPA) A-21.
i l
l J
NUREG-0844 4-11 J
4.3.4 Ieproved Accident Monitoring l Task The NRC staff should address accident monitoring weaknesses of the type observed at Ginna by implementation of Regulatory Guide 1.57, "Instrumentation *r Light-Water-Coole( Nuclear Power Plants To Assess Plant and Environs Conditions During and Following an Accident." and the safety parameter display system.
Initial Basis for Consideration During the January 25, 1982 event at Ginna, several weaknesses in accident monitoring became apparent; these included (1) nonredundant monitoring of RCS
> pressure, (2) failure of the position indication for the steam generator relief and safety valves, and (3) the limited range of the charging pump flow indica-tor for monitoring charging flow during accidents. Each of these areas is specifically addressed in Regulatory Guide 1.97. Assuming that the guide had been implemented on the Ginna plant before the January 25, 1982 event, the monitoring of the event would have been substantially improved and there would have been more assurance of correct operator actions in response to the event.
Improved accident monitoring also would have improved the NRC staff's ability to assess the plant status and the appropriateness of the licensee's actions and recommendations.
Status This issue ha: been resolved by Multiplant Action Item A-17 (Instrumentation to Follow the Course of an Accident). The resolution was issued in Supplement I to NUREG-0737 (Generic Letter 82-33).
4.3.5 Reactor Vessel Inventory Measurement Task The staff has observed that the implementation of THI Task Action Plan Item II.F.2, "Instrumentation fo.- Detection of Inadequate Core Cooling," would have substantially improved the Ginna situation by ensuring that steam bubble formation in the upper head of the reactor vessel could have been more accu-rately monitored.
Initial Basis for Consideration During the Ginna SGTR, the formation of a steam bubble in the reactor vessel '
upper head significantly complicated the course of the event. The uncertainty about the size of the steam bubble was a significant factor in the operators' decision to continue safety injection befand the point when termination is called for in the emergency procedures.
Status THI TAP Item II.F.2 is being implemented as Multiplant Action Item F-26. Letters to individual licensees (Generic Letter 82-28) and orders to Babcock & Wilcox licensees and M ansas Nuclear One, Unit 2 were issued on December 10, 1982.
i NUREG-0844 4-12
4.4 Human Factors Considerations 1
Human factors tasks are being addressed by ongoing programs in the control i ,om design review program (THI Task Action Plan Item 1.0.1), the emergency
! 'cedures review (TMI Task Action Plan Item I.C.1), and the reat. tor coolant 4 pump trip criteria review (THI Task Action Plan Item II.K.3.5).
- 4.4.1 Reactor Coolant Pump Trip Task The NRC staff should issue guidance for licensees regarding the development of reactor coolant pump (RCP) trip criteria that will ensure continued forced reactor coolant system (RC;) flow during steam generator tube breaks up to and
! including the design-basis tube rupture.
Initial Basis for Consideration
! Under the scope of TMI Action Plan Item II.K 3 in NUREG-0660, the NRC Bulle and Orders Task Force conducted generic reviews of the loss-of-feedwater ar. ,
j small-break loss-of-coolant events on all operating reactors. These revie.ws
- consisted of an evaluation of systems reliability analyses, guidelines for j
} emergency procedures, and operator training related to these events. i 1
l As a result of these reviews, a number of recommendations for improvements i
- were made and documented in NUREG-0565, NUREG-0611, NUREG-0623, NUREG-0626, and NUREG-0635. Included among these recommendations was the reevaluation of !
j '
, reactor coolant pump trip criteria.
4 j
The issue of reevaluation of reactor coo! ant pump trip criteria involves the ,
l potential improvement that might be achieved by establishing better criteria t on when to allow the operation of reactor coolant pumps and when to trip them. !
It was believed that better criteria might allow the use of rear. tor coolant l' pumps to aid in recovery from certain transients while still ensuring that these pumps could be tripped during a small-break LOCA. l Analyses indicate that continued operation of the RCPs following a range of >
small-break LOCAs could lead to excessive inventory loss for which the high- ,
pressure injection system would be unable to compensate. Generally, the range of break sizes of concern is from 0.02 to 0.2 ft: (2 to 5 in equivalent ;
diameter).
The interim position documented in NUREG-0623 requires manual tripping of the l reactor coolant pumps when the symptoms of a small-break LOCA (i.e., a safety-injection signal and low RCS pressure) are present. Although the interim post- <
tion appears to deal effectively with the problem of excessive inventory loss l for the small breaks of concern, it also has an effect on other more likely events. During the Ginna SGTR, the tripping of the reactor coolant pumps was the direct cause of the steam formation in the upper head of the reactor vessel i during RCS depressurization. The steam formation led the operator to quest 4n the adequacy of core cooling and to continue safety injection. This led ' i "he steam generator being overfilled and to the repeated opening of the stea* a er-ator safety valve, producing radiological release to the atmosphere. Tt 1 sp-ping of the reactor coolant pumps also eliminated the possibility of usi6 the NUREG-0844 4-13
normal pressurizer spray system to equalize the RCS and steam generator pres-sures. This led to the use of the pressurizer PORV, which subsequently failed open and had to be isolated. I Tripping the reattor coolant pumps during a tube rupture event is undesirable from the standpoint of managing the event, minimizing offsite radiologi:a1 releases, and avoiding a more serious consequence.
Status A resolution to this issue (THI TAP Item II.K.3.5) was identified in NRC Generic Letters 83-10(a) through 83-10(f) issued on February 8, 1933, to all PWR ,
licensees and applicants. The members of the W CE, and B&W Owners Groups have submitted generic analyses to satisfy the requirements of Generic Letter 83-10. NRC Generic Letters 85-12, 86-05, and 86-06 provided the staff's eval-uation of the Owners Group analyses. The utility for each plant is being required to submit plant-specific implementation information regarding the ilCP trip criteria. With respect to the Westinghouse Owners Group (WOG) analyses, the staff concluded that although the WOG had developed acceptable criteria for !
tripping the RCPs during small-break LOCAs and to minimize RCP trip for SGTR and non-LOCA events, the proposed RCP trip criteria may provide only margir21 assurance of preventing RCP trip for the design-basis SGTR event for plants with low-head HPI pumps. If a plant-specific analysis determines that the WOG alternative criteria are marginal for preventing unneeded RCP trip, the staff recommended that a more discriminating plant-specific procedure be developed.
i
- 4.4.2 Control Room Design Review (THI Task Action Plan Item I.0.1) 4 lisk As a result of a review of the Ginna control room following the tube rupture, several items related to the event were identified that are contrary to good human factors engineering principles.
These items, as described below (see l Initial Basis for Consideration) have been reviewed by the staff. Each of '
these items has been covered in the work to be done for the Item I.D 1 control room reviews, thus assuring that these items will be factored into all Item I.D.1 control room design reviews.
I Initial Basis for Consideration (1) Rotary Functional Identification 1
, This issue is based on the review of the power-operated relief valve (PORV) cycling activity described in NUREG-0909. The PORV rotary switches have discrete CLOSE, AUTO, or OPEN positions, but the associated block ' calve totary switches located immediately to the right of each PORV switch are spring-loaded return-to-center momentary-contact switches with CLOSE and OPEN posi-tions. The switches appear identical. In addition, the full-stroke cycle of the block valve is approximately 40 seconds, and it is not obvious whether the spring-loaded switch must be held for the full cycle or not, i
i' Rotary switches that appear identical but operate differently (e.g. , spring-loaded momentary vs. discrete posicion) should be provided with unique 1
! NUREG-084* 4-14 1
% ntifying features to indicate the operational differences. In addition, spring-loaded momentary-contact rotary switches that must be held by the operator in the activated position until the function is completed, versus activated and released (function continuous until complete), should also be provided with unique identifying features. Unique identifiers should be consistent throughout a control room, t
(2) Indicator Lights Burned Out The NRC task force reviewing the Ginna control room found that numerous indicator lights on the control room consoles had burned out, as was reported in NUREG-0909.
. This condition precluded an expeditious determination of equipment status.
Bulbs had burned out on both legend and nonlegend indicators. Since most nonlegend lights use a single bulb, indication was not availabl3 when the bulb was burned out. In addition, not all the bulbs were equally bright; the legend lights with dimmer bulbs were difficult to read. Apparently, the light intensity of the bulbs was related to age. If indicator lights are to be used to provide safety-system status information, they must operate properly.
I Administrative procedures should be inplemented that require a check, on a per-shift basis, of all indicator light bulbs in the control room except those that can only be tested by operating equipment (e.g., breaker controls). If the latter include multiple status (e.g., OPEN, CLOSED) in which each state is indicated by a light, a check should be made, on a periodic basis, to ensure that at least one light is il'uminated. Failed bulbs must be replaced immediately.
! As part of the detailed contrs i room design review required by the THI Action i Plan, a survey should be conducted to determine the number of bulbs that cr- be power tested (e.g. , push-to-test), have dual filaments, are long-life quahfied, or are operated with reduced voltage. This information should be used as the basis for a study to determine what hardware changes can be made to produce a more reliable indicator light system and a more effective and efficient test -
capability.
(3) Inconsistent Terminology Ouring the review of the Ginna SGTR event, several examples of inconsistent use of terminology on the control panels and between the panels and the plant '
procedures were observed. For example, it was observed that steam generator appeared as "STEAM GEN," "STM GEN," and "S/G." In addition, terminology i appearing on the panels was inconsistent with terminology appearing in proce- (
dures, and both were inconsistent with some commonly used terms. For example, j the commonly used term "PORV block valve" appears on the control panel label as "PRESSURIZER REL. STOP VLV " and is referred to in the SGTR procedure as "PORV backup isolation valve." This type of inconsistency may cause confusion and l i
may lead to operator error, All terminology and initialisms used in control rooms and in procedures should be reviewed for standardization and consistency, and modified accordingly.
NUREG-0844 4-15
Status The items identified at Ginna have been covered in the work to be done for the 1.D.1 control room design reviews, thus assuring that these items will be factored into all I.D.1 control room design reviews. This recommendation will be resolved as part of Task I.D.1 "Control Room Design Review," required by NUREG-0737, and implemented as Multiplant Action (MPA) Item F-08.
I 4.4.3 Emergency Operating Procedures Improvement (TMI Task Action Plan Item I.C.1)
Tasks and Bases The emergency operating procedures followed by the operators for coping with the Ginna SGTR event were basically sound. However, the NRC staff and plant personnel identified several areas as needing improvement. In addition, certain recommendations made in NUREG-0651 are also appropriate for further considera-tion. These items will be considered by the staff in conjunction with its ongoing work under TMI Task Action Plan Item I.C.1, "Short-Term Accident Anal-ysis and "'ocedures Revisions." A discussion of each of these items follows:
(1) Reactor Coolant Pump Restart Unambiguous guidance and criteria for RCP restart should be provided in emer-gency operating procedures (EOPs) for coping with an SGTR and other non-LOCA events. The guidance and criteria should be developed and updated recognizing the ongoing efforts to resolve the RCP trip issue, which is discussed in Sec-tion 4.4.1.
(2) Availability of Faulted Steam Generator Safety and Relief Valves During the Ginna SGTR, operators misinterpreted a procedural step intended to place the faulted steam generator atmospheric dump valve (ADV) in its manual mode by appropriately positioning its control. Instead, the: block valve was incorrectly closed, removing the ADV from service and forcing reliance on the safety valves to respond to overpressure conditions. During the event, the lowest set point safety valve opened five times.
Since safety valves will probably be required to pass mostly water instead of steam in the event the steam lines contain water, their reliability for reclosure at the desired lower pressure is questionable, and they may stick open. An SGTR with a stuck-open valve will allow relief from the primary to the secondary and to the atmosphere and can complicate core cooling.
Availability of the ADV avoids challenges to safety valves, and if the ADV sticks open it can be isolated to stop release to the atmosphere.
Plant-specific emergency operating procedures should, therefore, identify individually the appropriate valves for operation and make clear that remotely operated relief valves on faulted steam generators are not to be isolated but are to be made available for use.
NUREG-0844 4-16
l j (3) Multiple and Second-Order Failures r Emergency operating procedures (EOPs) should consider multiple failures and selected second-order failures. Task Action Plan Item I.C.1 requires that E0Ps consider multiple failures, and the SGTR event has highlighted the need to ensure that actions to cope with multiple failures are specified in the procedures. For example:
(a) Safety valves in the faulted steam generators were challenged because the f
~
ADV was isolated and because safety-injection (SI) pumps pressurized the 5 team generator through the break to the opening set point. Instructions !
in E0Ps could help mitigate this event (as discussed in the previous section).
(b) Multiple steam generator tube failures would compound the SGTR event. The leak is greater, the time required for isolation of the SG becomes crucial, ;
the steam lines can be filled, and the potential for saturated conditions in the primary system may be greater.
I (c) Inability to use the main condenser, coupled with the failure of the nonfaulted steam generator in a two-loop plant, can complicate residual j heat removal, j j (d) Inability to isolate the faulted steam generator can lead to increased of fsite radiological release and can complicate core cooling, f i
Procedures for coping with these types of failures can significantly mitigate !
,' the consequences of an SGTR. l l Some of the contingencies to be considered are plant specific; however, actions i specified in E0Ps will be supported by analyses, t
i (4) Steam Bubble Formation Operating events resulting in the inability to isolate small breaks require i I
that the reactor coolant system (RCS) be depressurized expeditiously to conserve inventory and minimize the release of radioactive coolant to the plant or environment. As a result of depressurization, steam can form in the RCS, particularly without a reactor coolant pump (RCP) in operation, so that the local metal temperatures could exceed the coolant saturation temperature.
Steam bubbles can form, even with core exit thermocouples, and the installed subcooling meter can indicate that the RCS is subcooled. Furthermore, if the depressurization is not adequately controlled, the RCS pressure may drop to that corresponding to the saturation temperature and cause the bubble to grow.
Events such as an SGTR, RCP seal failure, and small-break LOCA requiring rapid depressurization of the RCS can result in the formation of a steam bubble. The emergency procedure for an SGYR at Ginna did not address steam bubble formation.
Emergency procedures should cover this eventuality and provide guidance on identifying bubbles and on how to cope with and eliminate them.
NUREG-0844 4-17
(5) Cooling a Faulted Steam Genetator As a result of the tube rupture at Ginna, the faulted steam generator became overfilled and the written procedure in use did not address either cooling the
! steam generator or the precautions and care required for a steam line full of water up to the main steam isolation valve. The plant staff did develop, and management authorized, a cooldown method which involved several cycles of feeding the steam generator with unborated water, lowering the primary pressure, and causing backflow through the leak into the primary system. The plant staff
! was concerned with maintaining the integrity of the filled steam line and with l how this portion of the plant should be cooled.
I
] The bases for the PWR owners groups technical guidelines should contain an 4
analysis of the effects that caused the steam generator and steam lines to fill, j guidance to help prevent overfilling, and steps to cope with an overfilled steam generator. Alternative ways for cooling down the faulted steam generator j should be covered in some of the guidelines.
The E0P must also provide guidance on boron dilution and reactivity control l while cooling and dil F ' the RCS by any technique that allows unborated water
! to flow into the RCS. (ructions for sampling boron concentration or performing other de w tions must ensure against any encroachment on the required reactivity - -
i margin, j (6) Cooling an Intaa -
1erator i
1 The decision to secure ' e noanser and use only the atmospheric dump valve (ADV) relieving directly to the atmosphere during the Ginna event as the primary i
! means of decay heat removal may have reduced some inplant equipment contamina- !
l tion; however, it added to offsite releases and r3 moved the normal means of l
} plant cooldown. Also, after cooling with the ADV had begun, the decision to '
j break vacuum on the condenser removed the condenser as a backup means of energy
- q removal should a problem have occurred with the intact steam generator's ADV or i its control equipment. Licensees should evaluate the need for and consequences j of securing the condenser as a means of removing energy from an intact steam
! generator and using only the atmospheric relief valve on the intact steam j generator. The E0Ps should be structured to minimize offsite releases and to j avoid removal of backup heat removal equipment from service except when removal ;
- is necessary. ;
t
! (7) Safety-Injection Pump Termination and Restart Criteria f I l The EGP used at Ginna contained criteria for stopping and restarting the safety- ;
i injection (SI) pumps. Although a subcooling margin was not specified, the r operators' training and general knowledge equipped them with the understanding i of the need for maintaining a subcooling margin, which they did. E0Ps that ,
j specify the use of the 51 pumps should address the desired subcooling margin or l
! other appropriate indicators for stopping and restarting these pumps.
i I The E0P used at Ginna provided no guidance to make the operators aware of the l l likelihood of the formation .)f bubbles in the RCS, particularly when reactor '
coolant pumps were trippe:4 as they were. Neither did it advise the operators i
NUREG-0844 4-18
.' l 1
about how to cope with the large bubble that was formed when the PORV stuck !
open while they were trying to conduct a controlled depressurization. Because i of these plant conditions, the operators did not stop the safety-injection pumps ;
when they could have, ar.d thus they pressurized the secondary side of the faulted 1 l steam generator through the break to a safety valve opening pressure, and it opened, releasing radioactivity to the environment.
After the safety-injection pumps were stopped and when preparing to start an RCP, an SI pump was restarted as a precaution against an anticipated large ,
pressure transient that could collapse the bubble in the RCS. The large transient did not result, but again the faulted SG safety valve lifted. t i,
4 Unambiguous guidance and criteria on safety-injection pump operation and ;
termination should be provided in the E0P for coping with an SGTR and a bubble I in the RCS. These criteria should include a required subcooling margin to permit stopping and to require restarting. ;
l' i (8) Procedure Format and Clutter l Important space in Ginna's E0Ps was used up by nonrelevant instrument error i determination and guidance for reporting offsite authorities. This kind of l information should not be included in the caution or action portion of an E0P. i Tha NRC staff has prepared extensive guidance on the human factors element and !
j has disseminated it in NUREG v899, "Guidelines for the Preparation of Emergency l t Operating Procedures." i 1
t
! (9) Criteria for Natural Circulation Determination j j The desired means of core heat removal after tripping the RCPs is by estab-lishing natural circulation. If cooling is not by natural circulation, then i
- operators must confirm its existence and particularly under periods of abnormal RCS conditions such as the presence of a oubble other than in the presserizer, it must be repeatedly confirmed. Ginna has a procedure to confirm existence of natural circulation, and training and licensing examinations cover this area. !
Other plants should also have procedures containing explicit criteria for !
determining natural circulation. l, (10) Accommodation of Plant Differences From Reference Plant in Plant-Specifi_c Emergency Procedure Development The E0P generic technical guidelines for the most part are structured for a [
particular nuclear steam supply system design using a standard or typical plant. [
They also typically provide discussion, guidance, bases, and sample calculations !
for other designs to permit operators of those plants to have a smooth transi- i tion from the guidelines to plant-specific procedures. To illustrate the !
concern for accommodation of plant differences in transition from guidelines to f plant procedures, consider Ginna, a two-loop plant with a small pressurizer l compared with the large pressurizer in the four-loop plant. The Ginna RCS l operates at 2200 psi, about the same as most four-loop plants, and its steam t generator tubes are the same size as those in the four-loop plants. Thus, a '
1 large SGTR at Ginna would cause a leak of about the same magnitude as a similarly sized leak in a four-loop plant. This leak rate in a two-loop plant with a small pressurizer causes a much more rapid level and pressure decrease NUREG-0844 4-19 i
- . - _ . . _ _ . _-_ - _ - - _ _ -. - _ . _ _ . - - - _ , .-_ - _ _ . . . . . . _ . . .)
than in the four-loop plant. Thus, the operator must respond more rapidly to have the same influence as an operator on a four-loop plant. Similarly, a failure in the non-faulted steam generator on a two-loop plant leaves no backup for core heat dissipation, but in a four-loop plant two SGs would remain for heat removal.
Plant-specific analyses should be performed to account for dif ferences from the referenced plant.
(11) Rapid Determination of Faulted Steam Generator and Timely Depressurization of Reactor Coolant System To Minimize RCS Inventory loss and Releases -
Techniques should be developed to identify the damaged steam generator more rapidly, and these techniques should be included in the procedures and operator training.
Plant procedures should require timely securing of all feedwater to the damaged 1 SG as soon as it is identified. The procedures may allow intermittent feeding of the SG shculd its water level require such action.
l Procedures and operator training programs should emphasize the need to expeditiously secure steam flow from the damaged steam generator to the ;
turbine-driven auxiliary feedwater pump (TD AFP). The TD AFP should not be started (manually) unless the damaged SG has been identified and isolated, and steam from that SG to the TD AFP has been isolated. If the TD AFP has been automatically started, it should be secured if the other AFPs are operating and adequate feed flow exists. If the steam to the TD AFP is known to be from an l
undamaged SG, there should be no significant releases in the turbine exhaust and, therefore, no reason to secure the TD AFP. As a minimum, running times should be logged to enable the determination of radioactive release amounts.
The timely depressurization of the plant should be emphasized in the plant pro-J i
cedures and in operator training. Every licensee should adequately emphasize this important facet of the SGTR. In addition, RCS subcooling should be emphasized in plant procedures and operato. training.
(12) Main Steam Isolation Valve Closure During Plant Cooldown Procedure and operator training programs should address a possible MSIV closure in the unfaulted loop (s) during the cooldown following a safety injection signal.
l The operator should be given corrective actions to implement immediately.
(13) Use of Charging and letdown Systems 1 The emergency procedure should direct the operator to attempt to control the loss of RCS inventory caused by the SGiR by the use of charging and letdown systems and to initiate an orderly plant shutdown, if possible. Reactor trips
) from high power conditions could easily lead to lifting of SG safety valves I
and/or ADVs and result in direct radioactive release to the environment.
l If the decreases in pressurizer level and pressure are not controllable, then a j manual reactor trip shou)d be initiated before the indicated pressurizer level i goes off scale (low). This item should be identified in the plant-specific procedures.
NUREG-0844 4-20
(14) Operation of Loop Isolation Valves l i
Plants with loop isolation valves should investigate the use of these valves I following an SGTR and modify the plant procedures accordingly. Isolating the '
affected loop would almost immediately abate SGTR leakage, but may complicate plant cooldown. Licensees should, therefore, examine the advantages and disadvantages of loop isolation in their plant.'
(15) Use of Power-Operated Relief Valve 1 If the PORV is required to control RCS pressure on loss of normal spray cap-ability, guidance should be provided to the operator to monitor and control (if possible) the pressurizer relief tank parameters to minimize the potential of '
rupture-disk relief to containment.
(16) Potential Complicatina Events :
The operators should be provided with adequate procedural steps and training so that they can properly identify and correct the following situations during an j SGTR: [
(a) automatic opening of the pressurizer PORV(s) and/or the safety valves j l
(b) water solid or drained pressurizer I
l I
(c) saturation conditions in the RCS i l These situations may be postulated to occur either before or after the operator f f intervenes.
l (17) Site-Specific Operator Training ;
i i
Operator training using site-specific procedures on other than site-specific j simulators should be reviewed to determine whether the training received is j j realistic and practical. A determination should be made as to whether this (
j training would enable the operator to relate the site-specific emergency pro- >
cedures to the site-specific control boards and plant systems. I t
j (18) Steam Generator Level Control for Combustion Engineering Plants j
- The suggeeted procedure for CE guidelines to reduce the level in the faulted I steam ger, ator by letdown through the blowdown system to a holdup tank should i l be investigated to see that it provides an adequate level of protection for all l 1 possible break flows. !
l !
Status 1
'f Some of the above items are explicitly included in the review requirements of i i
TMI TAP I.C.I. Other items in the list are con',idered to be implicitly within the intent of TMI TAP 1.C.1 in that the availability of systems under expected )
r
- conditions (such as at Ginna) should be used in developing diagnostic guidance l 1 for operator and procedural development. The staff issued Generic Letter 82-33, !
i NUREG-0844 4-21 l
1 I
Supplement 1 to NUREG-0737, "Requirements for Emergency Response Capability,"
on December 17, 1982. This supplement restates the approved requirements and i
provides guidance on steps that have to be taken by licensees and the staff to
, accomplish this implementation. THI TAP I.C.1 is being implemented as Multiplant Action (MPA) Item F-05.
4.5 Radiological Consequences This section contains recommendations intended to reduce the potential radiological consequences of a steam generator tube rupture and to improve the ability to ac::urately measure the amount of radioactivity released from the
- plant.
' 4.5.1 Reassessment of Radiological Consequences Following a Postulated SGTR Event i
j Task I The NRC staff should reassess SGTR accidents to determine the effects of releases made for periods substantially longer and through release points other
! than those previously analyzed. These analyses should specifically address the 1 assumptions in Standard Review Plan Section 15.6.3 (NUREG-0800) and address the costs and benefits of requiring revised analyses by licensees.
Initial Basis for Consideration j
j Af ter the steam generator tube rupture at Ginna, plant equipment malfunctions j
and operator actions resulted in overfilling of the affected steam generator, which negated the ability of the moisture separators to retard the flow of j fission products to the environment. In addition, the water level in the sec- ,
! i l
ondary safety valves.
system reached the steam line and water entered the inlet of the stem l
This may have ultimately led one valve to fail to fully reseat. t Such failures allow an uncontrolled release to the environment and prevent 1
isolation of the affected steam generator. ;
As a result of these events it can be concluded that operating procedures and i
plant response limitations m,ay produce accidents different from those analyzed j l
previously in which accidents were assumed to be terminated in approximately !
30 minutes and proper water hvels were maintained in the affected steam gener-l ators (no overfilling). For these reasons, the staf f concludes that previous 1 i
{ analyses of t5e offsite consequence
- of SGTR accidents may not include the full range of conditions that may result from operator actions at specific plants, [
j
' Of primary concern are such factors as overfilling the affected steam Generator,
(
unanticipated relief valve releases, the accident duration, lack of liquid / gas phase iodine partitioning factors following overfilling and subsequent release of primary coolant in conjunction with iodine sp; king, and primary coolant ,
activity levels prior to such accidents. l Status i i
I This task will provide a better understanding and means to assess future SGTR events in operating plants relative to the consequence limits in 10 CFR Part 100 {
] v t
1 1 i i i l
NUREG-0844 4-22 i
i i
.,m__ ,.-y, , - , - - - - - - - - - _. . - - , _ . _ . , - - - _ ---,-._.,,,---m-,,,-w- ,._-- -- ,_
_m 9 . - - -
- and has been categot
- ced as a licensing issue (NUREG-0933, Supplement III).
This task will be sc.heduled pending the availability of staff resources, after assigning resources to approved high- and medium priority issues.
4.5.2 Reevaluation of Design-Basis SGTR 1 Task The NRC should consider, in conjunction with the tasks identified in Sec-tions 4.3.1 and 4.5.1 of this report, the necessity of reclassifying or '
ret.'efining the design-basis SGTR. The consideration should be based on observed aspe:ts of SGTRs that may be outside the bounds of the analyses currently requi.ed. The consideration should include, but not necessarily be limited to, the foilowing items: steamline flooding, stuck-open pressurizer PORV, steam
' formation. in the reactor vessel head, leaking steamline safety valves, steam ,
line and steam line support integrity under water-filled static and dynamic loading conditions, liquid rather than gaseous releases, criteria for terminat-ing safety injection, inadvertent isolation of steam lines, release of reactor
- coolant outside containment, and single-failure assumptions.
The staff should evaluate the lack of an explicit statement in the Standard Re- '
view Plan (SRP) regarding reliance on safety grade equipment on single failures along with a loss of offsite power. Other issues which should be considered include (1) whether it is appropriate to consider SGTR events involving more l
than a single ruptured tube, and/or SGTR events occurring as a consequence of a
- postulated MSLB or LOCA, (2) whether or not the loss of offsite power at some time after the identification of the event should be assumed, and (3) whether ,
! or not the offsite dose consequences must meet limits set in 10 CFR Part 20 rather than in 10 CFR Part 100. ,
i Initial Basis for Consideration !
The general basis for this recommendation is derived from the number of SGTRs that have occurred and the potential existing for SGTR doses exceeding the 3 guidelines of 10 CFR Part 100. However, these doses would occur only if there were an unlikely, but not impossible, set of circumstances as discussed in ;
detail in Section 8.1 of NUREG-0916. In any event. It is considered prudent to i
reconsider the SGTR event and the SRP assumptions and criteria. l The specific basis for consideration of the assumption of loss of offsite puwer !
l (LOOP) derives from a concern regarding whether or not a LOOP at some time after the initiation of an SGTR may place a more severe requirement on systems l and operators than a LOOP that is concurrent with the SGTR.
Status This task has been categorized as a licensing issue (NUREG-0933, Supplement III) and will be resolved as part of NRC Generic Issue 135, "Steam Generator and Steam Line Overfill Issues." The staff expects to issue a report of its findings by the end of 1988.
1 NUREG-0844 4-23
6 4.5.3 Secondary-System Isolation Task The NRC staf f should reevaluate the provisions for isolating the steam gen- '
erators in conjunction with the tasks identified in Sections 4.3.1 and 4.5.1 of this report. The evaluation should consider whether the current provisions -
for isolating the main steam and feedwater lines are adequate, with particular emphasis on isolation of the steam generator with RCS loop isolation valves, using closed-bonnet secondary safety valves, or containing the discharge from '
the steam generator safety and relief (atmospheric dump) valves.
- Status This task has been ranked as a low priority safety issue (NUREG-0933, Supple-ment III). Staff effort on this specific issue is not planned unless found to be warranted as a result of staff actions identified in Sections 4.4.3 and 4.5.1. i 4.6 Organizational Response I This section discusses actions taken or to be taken by the NRC staff that are related to matters discussed in NUREG-0909. Some of the following items do not result in additional requirements for further actions but do report the results of efforts taken in response to the Ginna event that have been completed or are not applicable as a generic task.
4.6.1 Operations Center Communications and Notifications ~
Task l
l Corrective actions, in response to the issues discussed below under "Bases" >
1 i
have been taken by the Office of Inspection and Enforcement (IE) relative to the NRC Operations Center. Steps include improving coordination and communication of Operations Center, regional office, and resident inspector response by pro-
! viding implementing procedures. The use of an additional telephone line has
- i been considered and was tried in a recent drill. In addition, the Telecommuni-cations Branch of the NRC has an ongoing contract with Calculon, Inc. .o study
] upgrading the emergency communications system.
) Bases 1
The results of assessments made during the Ginna event by the Region I base team i and the NRC headquarters tea s were not directly coordinated, and the communica-
) tions between the resident inspector and the Region I base team were not always tied into the NRC Operations Center. The health physics network, the Federal Telecommunications System, and commercial telecommunications systems functioned adequately; the emergency notification system (ENS) link was on13 marginally acceptable. Some problems encountered were attributable to the number of phones of f the hook and' to background noise in the room where the phones were located (e.g., Region I's Incident Response Center); other problems arose because of the phones' locations in relation to essential sources of informa-
} tion (e.g. , continuous midband frequency noise, intermittent high pitched tones, I
l NUREG-0844 4-24 4
l and unexplained brer.ks interrupted communications on the ENS lines). Tho staff of the NRC Headquarters Operations Center failed to make some notifications in a timely manner.
l Status The task has been completed. The results of this task will be implemented on a continuing basis.
4.6.2 Interaction Between Regional Base Teams and the Executive Team i
Task t The NRC, pursuant to NUREG-0845, should improve coordination between the t regional base team and the executive team, '
4 i Bases The conversations between the NRC Chairman and the Governor of New York and his representatives indic u d a lack of communication between the Region I Base t
- Team Radiological / Environmental Protective Measures Manager and the executive I team on NRC recommendations to the State of New York on field monitoring. The Chairman should have been briefed better on the State of New York's capabilities l
l and existing monitoring systems before he spoke to the Governor.
Status i
~
1 l ;
NUREG-0845 (Chapter III, Items I.4, J.4, and K.4) requires that all activities j conducted by technical teams be coordinated with the site team. Continued t participation by headquarters and regional base teams in exercise and training l sessions for implementing NUREG-0845 will improve this coordination. This ;
matter will be evaluated by the IE response coordination team in all future
' exercises in which IE and regional offices participate. Results of exercises l will be documented by lE in the exercise report, i This task has been completed. The results of this task will be implemented on a continuing basis. l I
4.6.3 NRC Site Team--Location of Site Team Components and Public Affairs Information Flow Task Upon arrival at the airport, if a site area or general emergency classification j exists, the NRC Site Team should be split, with the Public Affairs Coordinator ,
l going to the emergency news center, the State liaison officer and environmental / '
protective measures personnel going to the near-site emergency operations facility (EOF), and the rest of the initial site team going to the technical support center (TSC). As soon as possible, communication links should be estab-lished by the NRC personnel between the news center, EOF, TSC, base team, and Headquarters Operations Center.
NUREG-0844 4-25
i i
' The role of public affairs needs significantly improved definition, especially since NRC observations and information from headquarters, the regi(nal office, and the site may be relayed to the same individual. These recommendations must assume that a pui,iic affairs individual cannot gather and disseminate informa-tion simultaneously. With this in mind, the improvements needed for an effec-tive public affairs program are as follows:
(1) The Regional Public Af fairs Manager (RPAM) should be located in the In-i cident Response Center (IRC) so that information concerning the event can be gathered by visually observing status boards and by questioning the Base Team t.eader. If this is not por 'ble, a dedicated communicator or
, runner must be provided to brb f the MPAM not only on all of the informa-a tion received but also on the significance of that information.
(2) Upon arrival at the site, the Public Affairs Coordinator should be dis-a patched to the emergency news center to coordinate the release of NRC i information to the public. Communications must be established between i
]
the Public Affairs Coordinator and the site team leader as well as the counterpart public affairs individuals in headquarters and in the regional j IRC.
(3) All information released to the public from any location (site, regional office, or headquarters) should be coordinated so that all public affairs individuals are aware of what information is being released.
, Bases 1
l The bases for this task are problems experienced in deployment of NRC regional 4
staff as discussed in NUREG-0909, Section 6.4.2.2. !
i i
{ Status t
j Regional supplements and Revision 1 to NUREG-0845 have been published specify-ing procedures for improved deployment of the site team. The matter will be f
) implicitly reviewed as a part of the regional participation in exercises. ;
a Results of exercises will be routinely documented in the IE exercise report. ;
- This task has been completed. The results of this task will be implemented on j a continuing basis. !
I 4.6.4 Familiarization With NRC Response Plan !
1 i
3 Task I-a
} Sufficient training should be provided to all NRC personnel likely to be involved 1 in the implementation of the response plan to ensure familiarity with the plan f
- and the associated procedures, !
t l
l Bases I l
As discussed in NUREG-0909, Section 6.4.22, several examples of problems within L NRC (confusion with t e ENS and health physics network, failure to use IRC documentation forms, excessive noise in the IRC, and unavailability of public (
i i
NUREG-0844 4-26 !
I t
l affairs officers for prearranged briefings with the site team leader) aay be .
indicative of a lack of familiarity with the NRC emergency response plan and !
t procedures. NUREG-0845 was generally followed, although it had not been form-ally implemented and the indoctrination training had not been conducted at the time of the Ginna event. HUREG-0845 is new and fairly detailed so that all NRC I response personnel would benefit from additional training.
Status '
i Training is being conducted to ensure improved emergency response. The matter will be implicitly reviewed as a part of the NRC staff's participation in exercises. Results of the exercises will be routinely documented in the staff's exercise report.
This task has been completed. The results of this task will be implemented on i a continuing basis.
i 4.6.5 Alternate Evacuation Routes and Sites ,
lask The NRC staff should inspect the implementing procedures of the licensee's ,
emergency plan to ensure that they include plans for alternate evacuation ;
routes and sites to preclude evacuating personnel into a contaminated plume. ,
Bases
! As discussed in NUREG-0909, Section 6.1.2, during the Ginna event nonessential plant personnel were evacuated to an area, part of which was within the plume ,
released from the plant because of the wind direction. This resulted in sore ;
slight contamination of personnel, mainly on their clothing,
] i f Status 1
10 CFR Part 50.47(b)(10) requires licensees to provide a range of protective :
actions for emergency workers and the public. Section J.2 of NUREG-0654 depends (
l on this regulation to include provisions for evacuation routes and transporta-tion for onsite individuals to suitable offsite locations including alternatives l for adverse radiological conditions. The Office of Inspection and Enforcement l
Inspection Procedure 82202, "Protective Action Decisionmaking," requires that inspectors verify annually that licensees can assess an accident and make t recommendations for protective actions consistent with 10 CFR 50.47(b)-(10) and 1 NUREG-0654, Section J.
' This task has been completed. Results of these inspections are documented in i
- inspection reports.
4.6.6 Deescalation of Emergency Classification I Task '
j The NRC staff should evaluate the need for developing and issuing to licensees l
generic guidance regarding the licensees' deescalation of emergency classification, i
NUREG-0844 4-27 j
i 4
- , ~ . - _ , , - - - - - , . __ _ - . , - - . _ . _ - . . - - - - -__ -. - -- - _. -
Bases On the basis of the Ginna experience (see NUREG-0909, Section 6.1.1), specific plans and procedures addressing the technique for an orderly deescalation of the emergency response could have improved the selection of a plant status classification that more accurately described the condition of the plant in the recovery phase. The process of deescalation of emergency classification has not been addressed in NRC rules or in published criteria such as NUREG-0654.
' NUREG-0654 does address general plans for recovery and reentry, but does not provide generic guidance for deescalation. NUREG-0845 discusses the NRC pro-cedures for deescalation.
Status 4 r The staff, in coordination with the Federal Emergency Management Agency, intends j to incorporate guidance on this matter as part of an upcoming proposed revision to NUREG-06'4. Completion of this task has been delayed until 1989, pending completion of other higher priority work.
4.6.7 Offsite Dose Assessments
! Task The NRC staff should standardize the NRC method of assessing offsite doses and evaluate the need to make the assessment method available to all licensees.
Bases '
i Although the staff has concluded that the licensee's evaluation of the offsite i a doses during the Ginna event was, in general, consistent with the staff's !
evaluations (see NUREG-0909, Section 6.5.6), the staff has found that during
] licensee /NRC exercises, considerable effort was spent by the site team and the headquarters protective measures analysis team in resolving differences between licensee and NRC dose estimates. Confusion is created and NRC credibility diminishes when substantively different estimates are presented to State and local officials responsible for issuing orders on protective measures.
l Status The staff has developed an interactive rapid dose model for use by headquarters and the regions and has provided the model and portable computer equipment and i
training to all regional personnel. In addition, the staff is conducting regional workshops with State, licensee, and regional personnel on dose assess-ment and its role in protective action decisionmaking to ensure that all response personnel understand both the reasons for differences in dfsse projection made during act dents and their importance.
This task has been completed. The results of this task will be implemented on a continuing basis.
1 i
NUREG-0844 4-28
- APPENDIX A REFERENCES Bulletins NRC Bulletin 88-02, "Rapidly Propagating Fatigue Cracks in Steam Generator Tubes," February 5,1988.
Contractor Reports Idaho National Engineering Laboratory (EG&G), "Steam Generator Tube Rupture Study," PG-R-77-28, August 1977.
Idaho Nuclear Engineering Laboratory, "Code Assessment and Applications Program,"
CVAP-TR-78-015, November 1978.
Science Applications, Inc., "Value Impact Analysis of Recommendations Concerning
. Steam Generator Tube Degradation and Rupture Events," February 2,1983.
General Design Criteria Title 10 of the Code of Federal Regulations, Appendix A, "General Design Cri-teria for Nuclear Power Plants."
Industry Reports l
I EPRI-NP-481, Bechtel Corporation, "Steam Plant Surface Condenser Leakage Study,"
March 1977.
EPRI-NP-2062, Bechtel National Inc., "Steam Plant Surface Condenser Study Update," May 1982.
EPRI-NP-2704-SR, Steam Generator Ovners Group "PWR Secondary Water Chemistry Guidelines," October 1982.
EPRI-NP-2528-SR, EPRI, "EPRI PWR Safety and Relief Valve Test Program-Safety and Relief Valve Test Report," December 1982.
INPO Report 82-030, "Analysis of Steam Generator Tube Rupture Events at Oconee and Ginna," November 1982.
Gido, R. G., and A. Koestal, "LOCA-Generated Drop Site Prediction - A Thermal Fragmentation Model," Transactions of the American Nuclear Society, Volum, 30, Sections 371-372, 1978.
NUREG-0844 A-1
~
I Inspection Reports NRC Augmented Inspection Team (AITT Report Nos. 50-338/87-24 and 50-339/87-24, inspection conducted July 15-August 14, 1987, NRC Assession No. 8709040277.
Letters August 8,1978, from J. A.
Dearien,
EG&G, to R. E. Tiller, 00E, transmitting "The Capabilities of RELAP 4/ MOD 6, TRAC, and FLOOD 4 to Simulate a LOCA with Cor. current Steam Generator Tube Ruptures," CVAP-TR-78-015, July 1978.
November 3, 1978, from J. A.
Dearten,
EG&G to R. E. Tiller, DOE, Steam Generator Tube Rupture Effects on a LOCA Transmitting EG&G Idaho Report CAAP-TR-78-032, "Code Assessment and Appilcations Program," November 1978.
March 5, 1982, from E. L. Conner, NRC, to W. G. Counsil, Northeast Nuclear Energy Company, transmitting Amendment No. 73 to Facility Operating License for the Millstone Nuclear Power Station, Unit No. 2.
September 30, 1982, from A. D. Schmidt, SG0G, to D. G. Eisenhut, NRC, providing comments on proposed requirements.
October 4,1982, f rom C. W. Geiseler, WPSC, to T. A. Ippolito, NRC, providing comments on proposed requirements.
Octobcr 13, 1982, from L. M. Mills, TVA, to H. R. Denton, NRC, providing comments on proposed requirements.
October 19, 1982, from C. R. Hammond, Union Carbide Corporation, to K. R. Wichman, NRL,
Subject:
"Stress Analysis of Hydrotest in Main Steam Line A at Waterford S.E.S. Unit 3."
October 20, 1982, from L. C. Shiek, Lawrence Livermore Laboratory, to K. R. Wichman, NRC,
Subject:
"Maximum Stresses in Zion Unit 1 Main Steam Piping."
November 1, 1982, from H. B. Tucker, Duke Power Company, to H. R. Denton, NRC,
Subject:
"0conee Nuclear Station."
Nonmber 22, 1982, from J. E. Maier, RG&E, to Director, NRR,
Subject:
"Response to Safety Evaluation-NUREG-0916, Steam Generator Tube Rupture Incident.
R. E. Ginna Nuclear Power Plant," Docket No. 50-244.
December 9,1982, f rom D. G. Eisenhut, NRC, to All Pressurized Water Reactor Plant Licensees, Sub,)ect: "Potential Steam Generator Related Generic Requirements," (Generic letter No. 82-32).
December 10, 1982, from D. G. Eisenhut, NRC, to All Licensees of Operating Westinghouse and CE PWRs (except Arkansas Nuclear One-Unit 2 and San Onofre Units 2 and 3),
Subject:
"Inadequate Core Cooling Instrumentation System" (Generic Letter No. 82-28).
December 17, 1982, from D. G. Eisenhut, NRC, to All Licensees of C g rating Reactors, Applicants for Operating Licenses and Holders of Construction NUREG-0844 A-2
Permits,
Subject:
"Supplement 1 to NUREG-0737-Requirements for Emergency Response Capability" (Generic Letter No. 82-33)
January 6, 1983, from H. B. Tucker, Duke Power Company, to D. G. Eisenhut, NRC, responding to Generic Letter 83-32.
January 7,1983, froo C. W. Geisler, WPSC, to Director NRR, responding te Generic Letter 82-32.
January 10, 1983, from G. R. Westafer, Florida Power Corporation, to Director, NRR, responding to Genoric letter No. 82-32.
January 11, 1983, from F. l. Clayton, Alabama Power Corporation, to Director, NRR, responding to Generic Letter No. 82-32.
January 12, 1983, from M. R. Wisenburg, Houston Lighting & Power Company N to Director, NRR, responding to Generic letter No. 82-32.
January 12, 1983, from B. D. Withers, Portland Electric Company, to Director, .
NRR, responding to Generic Letter No. 82-32.
January 13, 1983, from E. A. Lider, Public Service Electric and Gas Company, to Director, NRR, responding to NRC Generic Letter 82-32.
i January 24, 1983, from J. E. Maier, RG&E, to Director, NRR, responding to NRC Generic Letter No. 82-32.
February 5, 1983, from J. Mattimoe, SMUD, to D. G. Eisenhut, NRC, responding l
to NRC Generic Letter No. 82-32.
l
. February 8, 1983, from D. G. Eisenhut, NRC, to all PWR Licensees, "Resolution of TMI Action Plan II.K.3.5, Automatic Trip of Reactor Coolant Pumps," Division l
of Licensing Generic Letter Nos. 83-10, a, b, c, d, e, and f.
February 10, 1983, J. H. Taylor, B&W, to D. G. Eisenhut, NRC, responding to NRC Generic Letter No. 82-32.
February 15, 1983, from J. Bayne, PASNY, to D. G. Eisenhut, NRC, responding to NRC Generic Letter No. 82-32.
February 15, 1983, from E. P. Rahe, Jr. , Westinghouse Electric Corporation, ,
to D. G. Eisenhut, NRC, with comments on SA'. Report (Letter No. NS-ERP-2706). I f
February 24, 1983, from A. Schere, CE, to D. G. Eisenhut, NRC, responding to .
February 25, 1983, from J. H. Morehouse, SAI, to T. Ippolito, NRC, providing additional information regarding the costs of implementing the potential ,
requirement for supplemental tube inspections.
August 10, 1983, from R. E. Uhrig, FPL, to H. R. Denton, NRC, "Generic Steam 1 Generator Requirements." ,
NUREG-0844 A-3
August 25, 1983, froo B. D. Withers, SGOG, to D. G. Eisenhut, NRC, regarding proposed generic steam genera +or requirements.
September 2,1983, from J. W. Williams, Jr. , AIF, to W. J. Dircks, NRC, "Generic Steam Generator Requirements."
January 3,1984, from 8. D. Withers, SGOG, to H. Denton, NRC, "NRC Proposed
- Generic Steam Generator Requirements."
April 24, 1984, from B. D. Withers, SGOG, to R. Purple, NRC, regarding the NRC Integrated Program for tne Resolution of the Steam Generator USIs. i i
June 22, 1984, from J. T. Collins, NRC, to W. C. Jones, Omaha Public Power j District, "Restart af the Fort Calhoun Station, Unit 1."
October 1, 1984, from B. D. Withers, SG0G, to H. Denton, NRC, to confirm i discussions with NRC staff on September 20, 1984.
April 17, 1985, from H. L. Thompson, NRC, to all PWR licensees of operatirg i reactors, applicants for operating licenses, and holders of construction per-f mits, and Ft. St. Vrain,
Subject:
"Staff Recommended Actions Stemming from NRC Integrated Program for the Resolution of Unresolved Safety Issues Regarding i
Steam Generator Tube Integrity" (Generic Letter 85-02),
June 11, 1985, from R. J. Rodriguez, SMUD, to H. L. Thompson, Jr. NRC, i
responding to NRC Generic ' otter 85-02.
June 13, 1985, from W. P. Johnson, Public Service of New Hampshire, to l H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02. l
)
June 13, 1985, from G. C. Sorensen, Washington Public Power Supply System, ;
j to Director, NRR, responding to NRC Generic Letter 85-02.
1 June 14, 1985, from J. W. Beck, Texas Utilities Generating Ccmpany, to H. L. Thompson, Jr. . NRC, responding to NRC Generic Letter 85-02.
4 June 14, 1985, from J. J. Carey, Duquesne Light Company, to H. L. Thompson, Jr., NRC, responding to NRC Generic Letter 85-02.
June 14, 1985, from K. W. Cook, Louisiana Power and Light, to H. L. Thompson, Jr., NRC, responding to NRC Generic letter 85-02.
June 14, 1985, from O. W. Dixon, Jr. South Carolina Electric and Gas l
Company, to H. R. Denton, NRC, responding to NRC Generic Letter 85-02.
June 14, 1985, from J. T. Enos, Arkansas Power and Light Company, to H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02.
June 14, 1985, from G. L. Koester, Kansas Gas and Electric Company, to l H. R. Denton, NRC, responding to hRC Generic Letter 85-02.
June 14, 1985, from R. P. Mcdonald, Alabama Power Company, to Director, NRR, responding to NRC Generic Letter 85-02, 1
{ NUREG-0844 A-4 i
i June 14, 1985, from B. D. Withers, Portland General Electric Company, to H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85 02. ;
June 17, 1985, from G. L. Alexander, Commonwealth Edison, to H. R. Denton, [
NRC, responding to Generic Letter 85-02.
June 17, 1985, from J. W. Cook, Consumers Power Company, to H. L. Thompson, Jr. , ;
NRC, responding to NRC Generic Letter 85-02. ;
June 17 1985, from J. A. Domer, TVA, To H. L. Thompson, Jr. , NRC, i respondIngtoNRCGenericLetter85-02. I June 17, 1985, from C. W. Fay, Wisconsin Electric Power Company, to i H. R. Denton, NRC, responding to NRC Generic Letter 85-02, j
' June 17, 1985, from D. C. Hintz, Wisconsin Public Service Corporation, to Director, NRR, responding to NRC Generic Letter 85-02. j June 17, 1985, from R. W. . Kober, Rochester Gas and Electric Company, to I J. A. Zwolinski, NRC, responding to NRC Generic Letter 85-02.
June 17, 1985, from C. A. McNeill, Public Service Electric and Gas Company, I to H. L. Thomson, Jr., NRC, responding to NRC Generic Letter 85-02. !
I.
, June 17, 1985, from J. D. O'Toole, Consolidated Edison Company, to i l H. L Thompson, Jr., NRC, responding to NRC Generic letter 85-02, l i !
1 June 17, 1985, from J. D. Shiffer, Pacific Gas and Electric Company, to j 1 H. L. Thompson, Jr., NRR, responding to Generic letter 85-02. r 1
June 17, 1985, from G. C. Sorenson, Washington Public Power Supply System, j to Director, NPR, responding to NRC Generic Letter 85-02. i I i l June 17, 1985, from W. L. Stuart, Virginia Electric and Power Company, to e H. L. Thompson, Jr. , NRC, responding to NRC Generic letter 85-02. [
l June 17, 1985, from M. R. Wisenburg, Houston Lighting and Power Company, to 2 H. R. Thompson, Jr., NRC, responding to NRC Generic Letter 85-02 '
}
- June 17, 1985, from S. R. Zimmerman Carolina Power and Light Company, to !
j H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02. !
I June 18, 1985, from R. L. Andrews, Omaha Public Power District, to H. L. Thompson, Jr., NRC, responding to NRC Generic letter 85-02.
) :
, June 18, 1985, from J. C. Brons, New York Power Authority, to :
- H. L. Thompson, Jr. , hRC, responding to NRC Generic Letter 85-02. j
! June 18,1985, f rom J. 1. Carey, Duquesne Light Company, to H. L. Thompson, Jr. , l NRC, responding to NRC r.,neric Letter 85-02, l' June 18, 1985, from J. W. Williams, Florida Power and Light Company, to j H. L. Thompson, Jr., NRC, responding to Generic Letter 85-02 regarding i St. Lucie Units 1 and 2.
I l NUREG-0844 A-5
i June 18, 1985, from J. W. Williams, Florida Power and Light Company, to j
H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02 regarding Turkey Point Units 3 and 4.
June 19, 1985, from D. Musolf. Northern States Power Company, to Director, NRR, responding to NRC Generic Letter 85-02.
June 19, 1985, from C. H. Poindexter, Baltimore Gas and Electric Company, to H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02.
, June 19, 1985, from D. F. Schnell, Union Electric Company, to H. L. Thompson, Jr., NRC, responding to NRC Generic Letter 85-02.
June 20, 1985, from H. D. Hukill, General Public Utilities Corporation, to H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02.
j June 21, 1985, from H. P. Alexich, Indiana and Michigan Electric Company, 3 to H. R. Denton, NRC, responding to NRC Generic letter 85-02.
1 June 21, 1985, from E. F. Van Brunt, Jr., Arizona Nuclear Power Project, j to H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02.
June 24, 1985, from R. P. Crouse, Toledo Edison, to Director, NRR, responding to NRC Generic Letter 85-02.
June 25, 1985, from J. F. Opeka, Ftrtheast Utilities, to Director, NRR, j responding to NRC Generic letter 85-02.
/ I
) June 26, 1985, from M. O. Medford Southern California Edison Company, to l
- H. L. Thompson, Jr. , NRC, responding to NRC Generic Letter 85-02.
a ;
) July 9, 1985, from G. R. Westafer, Florida Power Corporation, to -
j H. L. Thompson, Jr., NRC, responding to NRC Generic Letter 85-02.
l July 11, 1985, from H. L. Brey, Public Service Company of Colorado, to Regional Administrator, Region IV, NRC, responding to NRC Generic Letter 85-02.
i July 12, 1985, from G. D. Whittier, Maine Yankee Atomic Power Company, to j H. L. Thoepson, Jr. , NRC, responding to NRC Generic Letter 85-02. L i !
July 17, 1985, from H. A. Mahlman, B&W Owners Group Steam Generator ;
4 Committee, to H. L. Thompson, Jr. , NRC, providing comments on draf t t i NUREG-0844. '
i ,
i July 17, 1985, from H. B. Tucker, Duke Power Company, to H. 8!. Denton, NRC, '
reJponding to NRC deneric letter 85-02 regarding Oconee Units 1, 2, and 3.
July 17, 1985, f rom H. B. Tucker, Duke Power Company, to H. L. Thompson, Jr. , ;
NRC, respondir.g to NRC Generic Letter 85-02 regarding McGuire Units 1 and 2 (
and Catawba Units 1 and 2. !
t i l I
i ;
, I i !
l NUREG-0344 A-6 !
1 !
3
July 18, 1985, from S. R. Zimerman, Carolina Power and Light Company, to H. L. Thompson, Jr. , NRC, providing coments on draf t NUREG-0844. ,
July 29, 1985, from J. F. Opeka, Northeast Utilities, to Director, NRR, providing coments to draf t NUREG-0844.
July 31, 1985, from D. J. Van de Walle, Consumers Power Company, to Director, Olvision of Lictnsing, NRC, responding to Generic Letter 85-02.
August 12, 1985, from G. Papanic, Jr. , Yankee Atomic Electric Company, to Director, NRR, responding to NRC Generic Letter 85-02.
August 14, 1985, from J. A. Bailey, Georgia Power Company, to Director, NRR, responding to NRC Generic Letter 85-02.
August 14, 1985, from G. C. Sorensen, Washington Public Power Supply System, to D. G. Eisenhut, NRC, providing comments to draft NUREG-0844.
April 20, 1987, from Dan L. Johnson to the Secretary of the Comission, NRC, commenting on proposed rule concerning revisions to acceptance criteria for emergency core cooling systems, l December 1, 1987, from L. Engle, NRC, to W. L. Steward, Virginia Electric Power Company, providing Safety Evaluation of remedial actions taken in response to July 15, 1987 SGTR event at North Anna Unit 1. ,
Meeting Summary i
Summary of meeting held on September 21, 1982, to discuss Indian Point Unit No. 3 steam generator tubes, John 0. Thoma, September 28, 1982, NRC
- Accession No. 8210130063.
i Memoranda January 29, 1962, from F. Rowsome, NRC, to R. Tedesco,
Subject:
"Feed and !
Bleed Issues for CE Plants."
I March 5,1982, from R. Nttson, NRC, to R. Vollmer,
Subject:
"Decay Heat !
Sequence Precursor Program Draft Report." ,
July 13, 1982, f rom W. J. Dircks, NRC, to the NRC Commissioners, SECY-82-296, i "Resolution of AE00 Combination LOCA Concern." !
Cctober 13, 1982, from A. Buslik, NRC, to S. H. Hanauer,
Subject:
"Risk from i Steam Generator Tube Rupture Accidents."
l November 30, 1982, from W. J. Dircks, NRC, to the Commissioners, SECY-82-475, '
Subject:
"Staff Resolution of Reactor Coolant Pump Trip Issue."
March 24, 1986, f rom V. Stello, Jr. , NRC, to the Commissioners, SECY 86-97, "Steam Generator USI program - litility Responses to Staff Recomendations in Generic letter 85-02."
i NUREG-0844 A-7 I
i
1 i
NUREG Reports Issued By NRC
! NUREG-75/014 (formerly WASH-1400), "Reactor Safety Stud : An Assessment of j AccidentRisksinU.S.CommercialNuclearPowerPlants,y' October,1975.
l NUREG-0103, "Standard Technical Specifications for Babcock and Wilcox Pressurized Water Reactors," Revision 4, October 1980.
NUREG-0121. "Standard Technical Specifications for Combustion Engineering j Pressurized Water Reactors," Revision 2, December 1980.
) NUREG-0452, "Standard Technical Specifications for Westinghouse Pressurized l, Water Reactors," Revision 4, November 1981.
j NUREG-0460, "Anticipated Transients Without Scram for Light-Water Reactors,"
l April 1978.
l NUREG-0565, "Generic Evaluation of Small Break Loss-of-Coolant Accident i
Behavior in Babcock and Wilcox Designed 177-FA Operating Plants,"
January 1980.
NUREG-0611 "Generic Evaluation of Feedwater Transients and Small Break Loss-of-Coolant Accidents in Westinghouse-Designed Operating Plants,"
January 1980. I NUREG-0623, "Generic Assessment of Delayed Reactor Coolant Pump Trip During Small Break Loss-of Coolant Accidents in Pressurized Water Reactors,"
November 1979.
] f
' NUREG-0626 "Generic Evaluation of Feedwater Transients and Small Break !
Loss-of-Coolant Accidents in GE-Designed Operating Plants and Near-Term operating License Applications," January 1980.
) NUREG-0635, "Generic Evaluation of Feedwater Transients and Small Break j
Loss-Coolant Accidents in Combustion Engineering-Designed Operating Plants "
i February 1980. ,
L 1
I J
I NUREG-0651, "Evaluation of Steam Generator Tube Rupture Events," March 1980, f L
l NUREG-0654, "Criteria for Preparation and Evaluation of Radiological Emergency }
{ Respvase Plans and Preparedness in Support of Nuclear Power Plants," t i
February 1980. !
{ NUREG-0660, "NRC Action Plan Developed as a Result of the TMI-2 Accident,"
l May 1980. f NUREG-0700, "Guidelines,for Control Room Design Reviews," September 1981. I NUREG-0737, "Clarification of THI Action Plan Requirements," November 1980, t
NUREG-0799, "Oraft Criteria for Preparation of Emergency Operating Procedures," (
June 1981.
i j [
NUREG-0844 A-8 l 1
(
NUREG-0800, "Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants," LWR Edition, July 1981.
NUREG-0845, "Agency Procedures for the NRC Incident Response Plan," March 1982.
NUREG-0848, "Final Environmental Statement Related to the Operation of Byron '
Station Units 1 and 2," April 1982.
NUREG-0886, "Steam Generator Tube Experience," February 1982.
NUREG-0899, "Guidelines for the Preparation of Emergency Operating Procedures, Resolution of Comments on NUREG-0799," August 1982.
NUREG-0909 "NRC Report on the January 25, 1982 Steam Generator Tube Rupture atR.E.GInnaNuclearPowerPlant,"April 1982.
NUREG-0916, "Safety Evaluation Report Related to the Restart of R. E. Ginna Nuclear Power Plant," May 1982.
1 NUREG-0933, "Prioritization of Generic Safety Issues," Supplement III, July 1985.
"I NUREG-0937, "Evaluation of PWR Response to Main Steamline 8reak with Concurrent L Steam Generator Tube Rupture and Small Break LOCA," December 1982.
! NUREG-0975, Vol. 5 "Compilation of Contract Research for the Materials Branch, .
Division of Engineering Safety-Annual Report for FY 1986," March 1987. !
NUREG-1063, "Steam Generator Operating Experience Update, 1982-1983," June 1984. '
NUREG/CR-0175 (TREE-NUREG-1213), "Investigation of the Influence of Simulated Steam Generator Tube Ruptures During Loss-of-Coolant Experiments in the ;
Semiscale M00-1 System," Idaho National Engineering Laboratory, May 1978.
(
NUREG/CR-0718 (PNL-2937), "Steam Generator Integrity Program - Phase 1 Report,"
Pacific Northwest Laboratory, September 1979.
j NUREG/CR-1282 (SAND 79-2300), "Statistical Analysis of Steam Generator Inspection l Plans and Eddy Current Testing," Sandia National Laboratory, August 1980. I l
NUREG/CR-2659 (PNL-3794), "Iodine Transport Predicted for a Postulated Steam Line Break with Concurrent Rupture of Steam Generator Tubes," Pacific Northwest Laboratory, February 1983.
NUREG/CR-3001 (PNL-4342), "Fuel Performance Annual Report for 1981," Pacific Northwest Laboratory, December 1982.
NRC Regulatory Guides Regulatory Guide 1.45, Reactor Coolant Boundary Leakage Detection Systems,"
May 1973.
Rogulatory Guide 1.83, "Inservice Inspection of Pressurized Water Reactor Steam Generator Tubes," Revision 1, July 1975.
NUREG-0844 A-9
l
, Regulatory Guide 1.97, "Instrumentation for Light-Water-Cooled Nuclear Power
, Plants to Assess Plant and Environs Conditions During and Following an j Accident," Revision 2, December 1980.
I L Regulatory Guide 1.121, "8ases for Plugging Degraded PWR Steam Generator i Tubes," August 1976. '
a :
! Regulatory Guide 1.133, "Loose Part Detection System for the Primary System !
( for Light-Water-Cooled Reactors," Revision 1, July 1980. !
i I
Regulatory Guide 1.154, "Format and Content of Plant-Specific Fressurized i Thermal Shock Safety Analysis Reports for Pressurized Water Reactors," :
January 1987.
- 1 Regulatory Guide 8,8, "Information Relevant to Ensuring that occupr.tional l Radiation Exposures at Nuclear Power Stations Will Be As low As Is Reasonthly 1 Achievable," Revision 3, June 1978. !
l
[
l i
{ !
)
l I
i t
t i
)
1 i
NUREG-0844 A-10
, - - - ~ - ------a n , - - m.-- -
l l
i APPENDIX B EVALUATION OF SGTR EVENTS FOR PRIOR PERIODS OF VULNERABILITY !
TO RUPTURE UNDER POSTULATED HSLB ACCIDENT l Eh:.h of the four U.S. plants which have experienced SGTRs to date experienced .
a limited period, prior to the SGTR, during which it was vulnerable to rupture l under a postulated MSLB involving a peak primary-to-secondary pressure dif feren- [
tial of 2600 psid. Each of these periods of vulnerability was terminated after =
the degradation of the steam generator tubing had progressed sufficiently far to cause rupture under normal operating conditions involving primary-to-secondary i pressure differentials of about 1300 to 1500 psid. Staff estimates concerning !
the duration of these periods of vulnerability are provided below: i (1) R. E. Ginna Nuclear Power Plant [
j The failure sequence leading to the SGTR on January 25, 1982, involved imping; ment damage induced by a foreign object, plugging of degraded or [
leaking tubes, and breakage of previously plugged tubes with subsequent ;
camage to adjacent plugged and unplugged tubes (NUREG-0909, NUREG-0916). l This sequence of events is believed to have been occurring since about [
l 1975, t
} !
J The tube that ruptured in January 1982 had been inspected in April 1981, t but no indication was recorded for this tube at that time. A reevaluation !
! of April 1981 data performed subsequent to the SGTR event in January 1982 l revealed that the subject tube actually exhibited an absolute eddy current :
(EC) indication in April 1981 which was interpretable as a 40% indication i 5
utiliting a uniform wear scar calibraticn standard. At the time of the !
rupture, Ginna had accumulated 6.3 reactor operating months since April l
] 195L NRC-sponsored burst tests of steam generator tubing (NUREG/CR-0718) l l indicate that wr.il penetrations of 75% and 88% are necessary to cause rup- j l ture during an hiLB and during normal operating conditions, respectively. l l Assuming an 88% penetration at the time of rupture and a linear penetra- !
) tion rate, the amount of penetration would have exceeded 75% for 1.7 reac- :
i tor operating months. This represents the period of time during which l
) the subject tube could have been vulnerable to rupture during a postulated ,
MSLB.
i l
i Eddy-current inspection performed after the rupture event revealed no i j additional tubes sufficiently degraded so as to be vulnerable to rupture I
- during an MSLB. Thus, a postulated MSLB during the L7 reactor operating months prior to the January 1982 rupture would have resulted in no more than a single tube rupture, j Another tube was plugged with an 80% EC indication in April 198L The i degradation mechanism is believed to be the same as that which caused the
] January 1982 rupture event. No other tubes with similar indications were j found. Assuming no EC measurement error, and assuming the same rate of 1 B-1 i
degro m t as fr* the tube that eventually ruptured, the staff estimates that m tube could have been vulnerable to rupture during a postulated MSLB for a period of 0.7 reactor operating months before it was plugged.
If the EC error is considered and the assumption is made that the tube was degraded to just short of 88% (the point at which it would have ruptured during normal operation) at the time it was plugged, then the period of vulnerability would have been 1.7 months.
Additional tubes were plugged between 1976 and April 1980 as a result of EC indications and leaks attributed to the degradation mechanism described
,above. There is no direct evidence that any tubes experienced periods of
' vulnerabiltty (to rNture during MSLB) during this time. However, uncer-tainties cxist in this regard in view of possible large eddy-current measurement errors sometimes associated with long-wear scars and because of the ancertairity regarding the residual strength of tubes that leaked during normal operation. The staff has accounted for this uncertainty by assuming that one additional reactor month of vulnerability existed between 1976 and April 1980.
In summary, the staff conservatively estimates thet Ginna has experienced 4.4 reactor operating months during which it may have been vulnerable to rupture during a postulated MStB. There is no direct evidence that Ginna has ever operated during a period of vulnerability to multiple ruptures.
Any such period would have occurred between 1976 and April 1980 and would likely have been of very short duration (< 1.0 month).
(2) Prairie Island, Unit 1 The SGTR at Prairie Island Unit 1 on October 2, 1979 was the result of.
excessive wear caused by a foreign object rubbing against the tube (NUREG-0651). Th: staff has assumed that rupture occurred when the wear had penetrated 88% through the wall thickness. The foreign chject is believed to have been introouced during sludge lancing performed in the spring of 1976. Assuming that the wear degradation began then, and assuming a linear l wear penetration rate, and considering that 31 reactor months elapsed between spring 1976 and October 1979, the staff estimates that the wear exceeded 75% of the wall thickness for 4./ reactor operating meths. Thus, 4.7 months represents the period during which Prairie Island Ur vt 1 may have been vulnerable to an SGTR as a consequence of a postulated MSLB.
Investigations performed after the SGTR event of October 1979 indicated that no additional tubes were potentially vulnerable to rupture in the event of a postulated MSLB.
l (3) Point Peach Unit 1 i
, The Point Beach rupture on February 26, 1975 is believed to have occurred only after a crack had penetrated entirely through the tube wall and had grown sufficiently long to permit bursting. For calculational purposes, i 4 the staff has assumed that one tube was uniformly thinned to 88% through-wall /at the time of rupture) and that rupture could have occurred during an MSLB a? 75% throughwall. j t
i B-2 l t
I
The flaw is believed to have initiated at a high growth rate in September 1974 during an on-line conversion from phosphate to all-volatile treatment (AVT) secondary water chemistry. The staff has assumed that the tube was already degraded 50% throughwall in September 1974 either as a result of cracking or wastage. On this basis and the assumption that 5.5 reactor operating months elapsed between September 1974 and February 26, 1975, the staff estimates that the idealized flaw exceeded 75% throughwall penetration for 1.9 months prior to rupture during normal operation.
There is no direct evidence that Point Beach was vulnerable to multiple tube ruptures under postulated MSLB conditions before the February 1975 SGTR. Subsequent to the event, a sizable number of tubes were found to contain indications in excess of 90% throughwall. Experience has shown that cracks will generally produce small detectable leaks before degrading tube integrity enough to potentially rupture during normal operating or postulated accident conditions. However, sufficient uncertainties sur-round this event so that the staff cannot entirely dismiss the possibility 3
that a limited period of vulnerability to multiple failures under postu-lated MSLB conditions occurred. Such a period, if it existed, would have been less than the 1.9-month period during which Point Beach was vulner-able to a single tube rupture for a postulated MSLB.
(4) Surry Unit 2 Surry Unit 2 experienced an SGTR on September 15, 1976 as a result of j stress corrosion cracking in the U-bend (NUREG-0651). The staff does not have enough information to estimate the period during which this unit may have been vulnerable to rupture under postulated MSLB conditions. Sub-sequent investigation revealed cracks in other tubes; however, a limited period of vulnerability to multiple tube failures under postulated MSLB l conditions may or may not have existed.
Based on the above, the staff has concluded that Ginna, Prairie Island Unit 1, I and Point Beach Unit 1 together have operated for a total of 11 months during I which they were vulnerable to an SGTR as a consequence of a postulated MSLB.
Assuming that Surry Unit 2 experienced a period of vulnerability equal to the average period experienced by the three plants above, the total period of vulnerability experienced by these four plants was 1.2 years.
B-3
u aucuminuwo , cO-...O . E ,Om r -.. m-- ,,oc. .m., . .,,
g,POa- n.
E"?'" BIBLIOGRAPHIC DATA SHEET NUREG-0844 lit shtinucteowt 0*e TMs navgmsa 7187L8 ANO ,vettiLe J Lgave etahn NRC Integrated Program for the Resolution of Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Generator Tube Integrity a oAri ai> Oar cO uvio Final Report "ca'"
l
"^a
. Aur ,.Om ,,, September 1988 6 DAf f atPOmf ISSutD E. Murphy " " ' "
September l 1988
, Pintomus%Q OmGA4124T eOse gaug AhD waste %Q Acom4 55,,arbeele Coe, a Paostct.T ASE womE UNil 4 Wet a Division of Engineering and System Technology , , , _ , , , , , , , , , ,
Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, DC 20555 10 $PO%50meNQ ORGamigaTrose % Awt AND W Ask ee8Q ADDat h$ f,atwar le Caes ItoT,PtOfAtPORT l
l Same as 7, above. . ..m.u n o .miO ,, - . ,
l l
u .v a w..,Am,,,0,..
Technical Report This report presents the results of the NRC integrated program for the resolution of Unresolved Safety Issues (USIs) A-3, A-4, and A-5 regarding steam generator tube integrity. A generic risk assessment is provided and i- '3tes that risk from steam generator tube rupture (SGTR) events is not a significk .ritributor to total risk at a given site, nor to the total risk to which the generai public is routinely exposed. This report also identifies a number of staff-recommended actions that the staff finds can further enhance the effectiveness of licensee programs in ensuring steam generator tube integrity and in mitigating the e.onsequences of an SGTR. As part of the integrated program, the staff issued Generic Letter 85-02 encouraging licensees of pressurized water reactors (PWRs) to upgrade their programs, as necessary, to meet the intent of the staff-recommended actions; however, such actions do not constitute NRC requirements. In addition, this report describes a number of ongoing staff actions and studies involving steam generator issues which are being pursued to ,
provide added assurance that risk from SGTR events will continue to be small.
The staff concludes that with final publication of this report, USIs A-3, A-4, and A-5 are technically resolved, le DOCWt hf ANak enet e a t verOmDS C16(miPf 0al it avaitatikst ?
ST Af ttathf Steam Generators Tube Integrity PWRs Unresolved Safety Issues Unlimited Tube Degradation
,,,,,,,,,,,A,,,,,,,,
.icist,,,imi m sceo,imo.
'U2'assified
- n. . , -
Unclassified o ~ # . a o. ,Ac .
14 $m4 $
. . a w ., e. m m m m .,....,.........,,,
l UNITED STATES
,,aucun man NUCLEAR RE2ULATCRY COMMISSICN
. WASHINGTON, D.C. 20666 **"^*',0,",8 8 mo neua n.. a n OFFICIAL BUSINESS .
PENALTY FOR PRIVATE USE. 6300 120555139217 1 1hN1AI11A1131 US NAC-0APM-ADM DIV FOIA 4 PUBLICATIONS SVC R ft E S -POR NUREG P-210 WASHINGTON C 20555
, _ _ _ _ _ _ _ _ _ _ _ .