Text

Errata to Regulatory Analysis for Resolution of Generic Issue 94, Addl Low-Temp Overpressure Protection for Lwrs, Reprinting in Entirety Due to Printing Errors
	ML20055E089
Person / Time
Issue date:	12/31/1989
From:	Throm E; NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	;
References
	REF-GTECI-094, REF-GTECI-NI, TASK-094, TASK-94, TASK-OR NUREG-1326, NUREG-1326-ERR, NUDOCS 9007110011
	Download: ML20055E089 (95)
	v • d • e

-

iDk

/p m'c, UNITED $TATES

$- ~

,J.' i NUCLE AR_ REGULATORY COMMISSION

'E WASHINGTON, D. C. 2055t>

l

'%l,,,",, #

January 1990 ERRATA SHEET Report Number:

NUREG-1326 Report

Title:

Regulatory Analysis for the Resolution of Generic 1ssue 94,

" Additional I m -Temperature Overpressure Protection for Light-Wter Reactors" Prepared by:

Office of Nuclear Regulatory,Research (E. D. Thran)

U.S. Nuclear Regulatory Cannission Date Published:

December 1989 Instructions:

'Ihis document has been reprinted in its entirety because of printing errors. Please destroy the arevious copy of this NUREG you received and replace with tie correctly printed copy.

Division of Freedom ofInformation and Publications Services Office of Administration

~9007110011 891231 PDR NUREO 1326 R PDR J

NUREG-1326 Regulatory Analysis for the Resolution of Generic Issue 94,

" Additional Low-Temperature Overpressure Protection for Light-Water Reactors" U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research E. D. Throm 1.

A R(g l: '

kkk i

I i

e

'=

c AVAILABILITY NOTICE Availability of Reference Materials Cited in NRC Publications l'

Most 'ocuments cited in NRC publications will be available from one of the following sources:

1, The NRC Public Document Room, 2120 L Street, NW, Lower Level, Washington, DC j

20555 2.

The Superintendent of Documents, U.S. Government Printing Office, P.O. Box 37082, Washington, DC 20013 7082 3.

The National Technica; Information Service, Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publica-tions, it is not intended to be exhaustive.

Referenced documents available for inspection and copying for a fee from the NRC Public Document Room include NRC correspondence and internal NRC memoranda; NRC Office of inspection and Enforcement bulletins, circulars, information noticos, inspection and investi-i gation notices; Licensee Event Reports; vendor reports and correspondonce; Commission paperst and applicant and licensee documents and correspondence.

The following documents In the NUREG series are available for purchase from the GPO Sa!es Program: formal NRC staff and contractor reports, NRC-sponsored conference proceed-ings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regula-tions in the Code of Federal Regulations, and Nuclear Regulatory Commission issuances.

Documents available from the National Technical information Service include NUREG series reports 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 noticos, federal and state legislation, and congressional reports can usually be obtained from these libraries.

Documents such as theses, dissertations, foreign reports and translations, and non NRC conference proceedings are available for purchase from the organization sponsoring the publication cited.

Single copies of NRC draft reports are available free, to the extent of supply, upon written request to the Office of information Resources Management Distribution Section U.S.

Nuclear Regulatory Commission, Washington, DC 20555.

Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available there for reference use by the public. Codes and standards are usually copy-righted and may be purchased from the originating organization or, 'l they are American National Standards, from the American National Standards institu e,1430 Broadway, New York, NY 10018.

t 7

1

NUREG-1326 Regu atory Analysis for the Resolution of Generic Issue 94,

" Additional Low-Temperature Overaressure Protection for Lig at-Water Reactors" Manuscript Completed: September 1989 4

Date Published: December 1989 E. D. Throm Division of Safety Issue Resolution Ollice of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555 p* "%

s s

i 1'

[

ABSTRACT low temperature overpressure protection (LTOP) is Protection for Light Water Reactors." It includes (1) a required in pressurized water reactors (PWRs) to provide summary of the issue, (2) the proposed technical resolu-

[

protection against brittle reactor pressure vessel failure tion, (3) alternative resolutions considered by the Nuclear following an anticipated event. Typically these events are Regulatory Commission (NRC), (4) an assessment of the a result of either mass imbalance (excess charging in benefits and cost of the allematives considered with addi-companson to available letdown flow or inadvertent safety tional emphasis on the recommended resolution, (5) the injection) or energy input transients (restarting an idle decision rationale, and (6) the impacts and relationships reactor coolant pump causing an increase in the reactor between G194 and other NRC programs and require.

coolant system pressure as a result of rnixing cold water

ments, from the inactive loop with the remainder of the hot fluid and as a result of direct energy addition from a wanner secondary side heat sink). The significance of these

Ihe majority of the technical evaluations, and the develop-events is heightened during water solid operations, ment of the cost analyses, for the various alternatives considered were performed by Battelle Pacific Northwest low-temperature overpressure protection is required in Laboratories (PNL) under Technical Assistance to the the shutdown modes of operation, Mode 4 Hot Division of Reactor and Plant Systems, RES, FIN Number 1

- Shutdown, Mode 5 Cold Shutdown, and Mode 6 -

B 2998(NUREG/CR-5186).

Refueling with the reactor vessel head bolted down.

While operating in Modes 5 and 6 and with the reactor Additional considerations that could impact on tie recom-coolant temperature below 200oF, there are no technical mendations and conclusions regarding GI M have been specifications for containment integrity.

The conse-addressed by the NRC staff. Most notable are a reevalua-quences of an unmitigated low temperature overpressure tion of the consequence analyses to ensure that the event can be significant as a result of either containment estimated risk, in person-rem, is not overly conservative, bypass or failure of containment to isolate following and an adjustment in the NRC and industry implementa-reactor pressure vessel failure, tion cost estimates to account for plants not considered in the PNL risk evaluation. These are plants licensed after This report presents the regulatory analysis for Generic the end of 1986 or currently in the process of being Issue 94, " Additional Low Temperature Overpressure licensed for operation.

i iii NUREG-1326

1 I

i l

J CONTENTS l

Page ABSTRACT ill m-TABLES.....

vil ACRONYMS AND INITIALISMS..............

xi ACKNOWLEDGMENTS.

xiii-EXECUTIVE

SUMMARY

ES 1

1. STATEMENT OP PROBLEM 11

2. OBJECTIVES 21

3. ALTERNATIVERESOLUTIONS 31 3.1 Altemative 1 No Action Altemative

'31 3.2 Altemative 2 Change to Technical Specifications 31 3.3 Alternative 3 SiandRCPRestrictions 31 3.4 Alternative 4-Removalof RHR AutoclosureInterlock 31

- 3.5 Alternative 5 Safety Grade LTOP System 31 3.6 Alternative 6 Pressurizer Bubble 31

4. TECHNICAL FINDINGS 41

.... =

5. VALUE/ IMPACT ANALYSIS 51 5.1 Costs and Benefits of Altemative Proposed Resolutions 51 5.1.1 Alternative 1 - No Action Alternative 51 5.1.2 Altemative 2 Change to Technical Specifications 5-4 5.1.3 Alternative 3 SI and RCP Restrictions 59' 5.1.4 Altemative4 Removalof RHR AutoclosureInterlock 5 12 5.1.5 Altemative 5 Safety Grade LTOP System 5 15-5.1.6 Alternative 6 - Pressurizer Bubble 5 21 5.1.7 Summary of Best Estima'e Value/ Impact Ratios for All Altematives..........

5 25 v

NUREG-1326

Page 5.2 Relationships With Other Regulatory Issues 5 26 5.2.1 Regulatory Guide 1.99, Revision 2, " Radiation Embrittlement -

of Reactor Vessel Materials" 5 26 5.2.2 Generic Issue 99, "RCS/RHR Suction Line Interlocks in PWRs" 5 26 5.2.3 Generic issue 70, Power Operated Relief Valve and Block Valve Reliability" 5 27

6. DECISION RATION ALE 6-1 6.1 Conclusions Conceming LTOP Implementation 62 6.2 Improvements in LTOP Protection Syst:m Availability 64

7. IMPLEMENTATION..

71 REFERENCES..

R1 APPENDIX A

SUMMARY

OF OPERATING REACTOR EXPERIENCES A1 APPENDIX B -

SOURCE TERM EVALUATION B1 APPENDIX C INDUSTRY IMPLEMENTATION COST ANALYSIS DATA BASE C1-APPENDIX D NRCIMPLEMENTATION COST ANALYSIS DATA BASE......

D-1 APPENDIX E PRESENT VALUE COST ANALYSIS DATA BASE E-1 NUREO 1326 vi

TABLES Table Page ES.1 Summary of best estimate value/ impact (V/I) ratios for alternatives evaluated by NRC ES 3 4.1 Summary of PORY class operating reactor experiences 4-4 4.2 Summary of RHR SRV class operating reactor experiences 45 4.3

. Base case mean com damage frequencies from LTOP events 46 4.4 Consequences for various release categories 47 4.5 Base case consequences for various containment assumptions 47 5.1 Value/ impact summary for the proposed resolution of G194 (for 67 plants) 5-6 5.2 Estimated present value costs for avoided onsite damage (for 67 plants) 58 5.3 Estimated present value costs for avoided offsite health and personal property damage (for 67 plants) 58 5.4 Mean com damage frequency estimates for Alternative 3 5 10 5.5 Peak pressure spectrum summary for Alternative 3 5 10 5.6 Implementation cost estimates for Alternative 3 5 11 5.7 Value/ impact summary for Alternative 3 (for 67 plants) 5 12 5.8 Peak pressure spectrum summary for Altemative 4 5 13 5.9 Implementation cost estimates for Altemative 4 (for 40 PORV plants) 5-14 5.10 Value/ impact summary for Altemative 4 (for 40 PORV plants).....................................................

5 15

- 5.11 Mean core damage frequency estimates for Altemative 5 (PORV plants only) 5 17 5.12 Best estimate unit cost to upgrade OMS to be safety grade................................

5 18 5.13 Per plant occupational dose for safety-grade OMS upgrade 5 19 5.14 Implementation cost estimates for Alternative 5 (for 40 PORV plants) 5 20 5.15 Value/ impact summary for Alternative 5 (for 40 PORV plants).....................

5-20 5.16 Peak pressure spectrum for Altemative 6...................................................................................

5 22 5.17 Mean core damage frequency estimates for Altemative 6 5-22 vii NUREG 1326

Table Pege 5.18 Industry unit implementation cost for nitrogen bubble 5 23 2

5.19 Implementation cost estimates for Alternative 6 (for 67 plants).............

5-24 5.20 Value/ impact summary for Alternative 6 (for 67 plants) 5 24 5.21 Summary of best estimate value/ impact (V/I) ratios for alternatives evaluated by NRC 5 25 A.1 LTOP coding system A-7 A.2 Summary of NUREG-0224 LTOP data A8 A.3 Summary of NUREG/CR 2789 data.............

A.9 A.4 Summary of AEOD Case Study C401 data A.10 A5 Summary of LER Update Search data A4!

A.6 Summary of Babcock and Wilcox planis A 14 A.7 Summary of Combustion Engineering plants A 14

'i A.8 Summary of Westinghouse plants A43 i

A.9(a)

Combustion Engineering L'IDP events summary - total data base A 16 A.9(b)

Combustion Engineering LTOP events summary - without precursor data A 16 l

A.10(a) Westinghouse L10P events summary total data base A.17 A.10(b) Westinghouse LTDP events summary without precursor data......

A 17 A.11(a) Total LTOP events summary total W and CE data base A.18 A.11(b) Total LTOP events summary without precursor data - total W and CE data base........................

A.18 A.12(a) Combustion Engineering one LTOP channel unavailable summary A 18

.A.12(b) Combustion Engineering both LTOP channels unavailable summary..............

A 18 A.13(a) Westinghouse one LTOP channel unavailable summary A 19 A.13(b) Westinghouse both LTOP channels unavailable summary.................. -.......................................

A 19 A.14(a) Total for one LTOP channel unavailable summary A 19 A.14(b) Total for both LTOP channels unavailable summary.

A 19 NUREG-1326 viii

a j

,Li c

l i

Table

-Page i

)

A.15 Pressure /iemperature data summary,

A.20

=. +...............................

A.16 Pressure data by event initiator summary.....

A.20 j

B.1 Estimated environmental release fractions for a core melt accclent tesulting from a low. temperature overpressure event B14 B.2 Comparison of 50-mile radius consequences B.4

=........... ~...........................

B.3 Corisequences evaluation com[arisons (40 PORY plus 15 RHR SRV plants)

B.$

l B.4 Cormeque.ces estimates in person. rem for low. temperature overpressure events (40 PORY plus 15 RHR SRV plants).......

B5

=................

B.5 Comparison of consequences for various sites and releases B.5 C.1 Industry unit costs for technical specification and procedure revisions C4 E.1 Daily replacement power cost estimates by region E-4 E.2 EstWied present value costs for avoided onsite property damage (40 PORY plus 15 RHR SRV plants)

E.4

.

E.3 Comparison of offsite property damage costs E.5 E.4 Estimated present value costs for avoided offsite health and property damage (40 PORV plus 15 RHR SRV plants)

E.5

........................................-m............

E.5 Pnsent value cost summary for 40 PORY and 15 RHR SRV plants

[

(based on base case frequency. total value of averted damages)

E.6

=.,.................

i e

(

i ix NUREO.1326 i

i

E

?

7 T

a ACRONYMS AND INITIALISMS L5 ACI Autoclosure interlock (of RilR NRR Office of Nuclear Reactor suction line isolation valves)

Regulation,NRC ALARA As low as reasonably achievable NSSS Nuclear steam supply system ANSI American National Standards OMS Overpressure miugation system Irstitute PORY Power operated relief valve h

AOT Allowable outage time PNL Battelle Pacific Northwest ASME Americen Society of Mechanical Lateratories Engineers psi pounds per square inch B&W Babcock and Wilcox PTS Pressurized thermal shock BTP Branch technicalposition PWR Pressurized water reactor BV Block valve RCS Reactor coolant system BWR Bolling water reactor RCP Reactor coolant pump CDP Core damage frequency RES Office of Nuclear Regulatory K

r

[

CE Combustion Engineering Research,NRC r

CFR Code of Federal Regulations RiiR Residual heat removal COMS Cold overpressure mitigation system (used by Westinghouse) system (see LTOP)

RSB Reactor Systems Branch DliR Decay heat removal system (used RT(ndt) reference temperature, by B&W) nil ductility transition ECCS Emergency core cooling system SI Safety injection (refers to high-EPRI Electric Power Research Institute pressure pumps)

ODC Ocneral Design Criterion, SDCS Shutdown cooling system (used Appendix A,10 CFR Part $0 by CE) 01 Generic issue (used by NRC)

SP setpoint gpm gallons per minute SRP Standard Review Plan llPSI liigh pressure safety injection SRV Safety relief valve LER Licensec ovent report STS Standard technical specifications LCO Limiting cralitions of operation TS Technical specification LTOP Low temperature overpressurc TWC Through wall crack protection system (generic NRC USI Unresolved Safety issue term)

VFP Vessel fracture probability NRC Nuclear Regulatory Commission W

Westinghouse

~

a

=

r xi NUREG 1326

=

m

=

ACKNOWLEDGMENTS This report is the result of the efforts of both contractor Colburn EJ. Eschbach. M.S. Harris, and A.S. Talstabal and NRC personnel. The technical support services are noted. In addition, the contributions of NRC staff (W.

provided by Battelle Pacific Northwest Laboratories Minners, K. Kniel, O. Mazetis, and L. Gallagher) are also (PNL) through the efforts of B.F. Oore, T.V. Yo, AJ.

acknowledged by (ne Task Manager.

xili NUREO.1326

EXECUTIVE SUMM ARY General Design Criterion 15 of Appendix A to 10 CFR 0748 Ref. 2). The current staff guidelines for the Part 50 requires that *the reactor coolant system and LTOP system are found in Standard Review Plan 5.2.2, associated auxiliary, control, and protection systems shall

Overpressure Protection," and in its attached Branch be designed with sufDcient margin to assure that the Technical Position BTP RSB 5 2,

Overpressure Protec-design conditions of the reactor coolant pressure toundary tion of Pressurized Water Reactors While Operating at are not exceeded during q

adition of normal low Temperatures * (NUREG-0800 Ref. 3).

operation, including anticipu opera, 9 occurrences.*

Twelve overpressure transients, in PWRs, were reported o

es, as defined in during the period from 1981 to 1983 (Ref. 4) after comple.

Anticipated operational Appendix A to 10 Cm Par 0

,e "those conditions of Lion of USl A 26. Two of these events, at Turkey Point normal operation which are c?, L.ted to occur one or more Unit 4,

exceeded the technical specification times during the life of the nuclear power unit and include pressureActnperature limits, in addition, during this same but are not limited to loss of power to all tr. circulation timeframe there were 37 reported instances when at least pumps, tripping of the turbine generator set, isolation of one LTOP channel was out of service, ln 12 of these the main condenser, and loss of offsite power.*

cases, toth LTOP channels were inoperable.

General Design Criterion 31 of Append (x A to 10 CFR The continuation of overpressure transient events, and tt e Part 50 requires that 'the reactor coolant pressure unavailability of LTOP protection channels, suggested the boundary shall be designed with sufficient margin to need to reevaluate the current overpressure protection assure that when stressed under operating, maintenance, criteria, or their implementation, to determine whether testing, and postulated accident conditions (1) the bound-additional considerations are warranted.

ary tchaves in a nonbrittle mannet and (2) the probability of rapidly propagating fracture is minimized. The design Major overpressurization of the reactor coolant system shall reficct consideration of service temperatures and while at low temperature,if combined with a critical crack other conditions of the toundary material under operating, in the reactor pressure vessel welds or plate material, maintenance, testing, and postulated accident conditions could result in a brittle fracture of the pressure vessel. As and the uncertainties in determining (1) material long as the fracture resistance of the reactor pressure properties, (2) the effects of irradiation on material vessel material is relatively high, these events are not properties, (3) residual, steady state and transient stresses, expected to cause vessel failure, llowever, the fracture and (4) size of flaws."

resistance of the reactor pressure vessel materials decreases with exposure to fast neutrons during the life of Appendix G to 10 CFR Part 50 provides the fracture a nuclear power plant. The rate of decrease is dependent toughness requirements for the reactor pressure vessel on the metallurgical composition of the vessel walls and under certain condiuons. To ensure that the Appendix G welds. If the fracture toughness of the vessel has teen limits of the reacter coolant pressure boundary are no' reduced sufficiently by neutron irradiation, low-exceeded during any anticipated operational occurrences, temperature overpressure events could cause propagation technical specification pressure / temperature limits are of fairly srnali flaws that might exist near the inner provided for operating the plant.

surface. The assumed initial flaw might propagate into a crack through the sessel wall of sufficient extent to in the lat 1970s, it was noted that there were a large threaten vessel integrity and, therefore, core cooling number of events occurring at reactors while operating at capability, low temperatures (shutdown modes) where the technical specification pressure / temperature limits were being The safety significance of these continuing low-t exceeded. The frequency of these overpressure transients temperature overpressure transients was designated as was determined to be within the anticipated operational Generic issue 94,

" Additional Low Temperature occurrence definition.

Overprcssure Protection" (Ref. 5). 0194 applies to the design and opctation of all PWRs. BWRs have been ex-Low temperature overpressure protection (LTOP) was cluded from consideration because they do not normally designated as Unresolved Safety Issue (USI) A 26 in 1977 operate in a water solid configuration.

(NUREG 0371 - Ref.1). PWR licensees implemented procedures to reduce the potential for overpressure events The Babcock and Wilcox plants have also teen excluded and installed equipment mochfications to mitigate such from this evaluation because these units have not events based on staff recommendations from the USl A 26 experienced any low temperature overpressure transients evaluations, under Multi Plant Action item B-N (NUREG and, based on theoretical risk, do not contribute to the ES 1 NUREG 1326

1 Executive Summary overall risk of low temperature overpressure events.

cal specification for overgessure ps, ation to ensure that Batco;k and Wilcox plants do not operate in a water solid toth low temperature overpressure r i.5 ction channels are condition. A steam or nitrogen bubble is maintained in the operable, especially in a water sold.ondition; that is, to pressurizer. The bubble provides a minimum of 10 treat operationally the low temperature overpressure minutes for the operator to respond to an anticipated low-protection system as a system that performs a safety-temperature overgessure event. A single path is provided related function.

for pressure relief (PORV or RHR safety relief valve).

The specific action recommended is to reduce the allow-In reaching its proposed resolution for 0194, the NRC able outage time (AOT) for a single LTOp channel when staff considered six specific alternative courses of action, operating in Mode $ (cold shutdown) or Mode 6 (with the The requirements would be applicable to all Westinghouse reactor pressure vessel head tolted down) from the current and Combustion Engineering plants, toth operating AOT of 7 days to an AOT of 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months tefore remedial ac-reactors and reactors in the construction stage of licensing, tions to depressurite and to vent the reactor coolant sys-Fifty two Westinghouse plants and 15 Combustion tem would be required.

Engineering plants are considered in the determination of the industry and the NRC implementation costs.

The overall best estimate value/ impact, not including accident avoidance costs, is atout $160 pc': person rem The general objective of GI 94 is to evaluate the need for overted, if cost savings to the industry trom accident additional low temperature overpressure protection and to avoidance (cleanup and repair of onsite damage and examine alternatives to reduce the risk of core damage replacement power) were included, the overall accidents associated with low ternperature overpressure value/ impact ratio would improve significantly as the events in pWRs by reducing the likelihood of these events.

avoided costs for cleanup and repair and for replacement The basis for this is the need to ensure that there is a low power more than offset the combined implementation likelihood of brittle reactor pressure vessel failure costs for the industry and the NRC staff.

(through wall crack (TWC)). Such a failure, unlike most other accident scenarios that can lead to core damage, could result in the reactor pressure vessel's teing unavail-Table ES.) is provided as a summt.ry of the best estimate able for either subsequent recovery of the reactor core or dose reductions, occupatiomd exposures, industry as an additional barrier far fission product retention.

implementation costs, NRC implementation costs, and the value/impxt ratio for each of the attematives studied by To achieve these objectives, the staff's proposed resolu-the staff. The base case TWC frequency is 3.24x104 per tion for 0194 recommends a revision to the plant techni-reactor year.

NUREO 1326 ES 2

Executive Summary

. Table ES.! Summary of best estimate value/ impact (Y/l) ratios for alternatives evaluated by NRC.

T W C Preq Dose Occupational Industry NRC V/l Ratio (')

Alter.

Reduction Reduction Exposare Costs Costs

($ per averted native (1/R.yr)

(person tem) (person rem)

($1,000s)

($1,000s) person rem) 2 2.89xig6 14,500 n/a 1,370 950 160 3(a) 1.07xig6 7,000 rva 3,630 1,840 780 3(b) 0.21x106 1,400 n/a 1,290 950 1,600 3(a&b) 1.20xig6 8,400 tva 4,920 2,790 920 4(a) 0.16x10 4 700 n/a 770 650 1,900 4(b) 0.16xl&6 700 n/a 4,770 650 7,750 5

1.82x10-6 8,200 900 16,000 570 2,000 5(a) 3.00x10-6 13,400 900 16,000 570 1,200 6(a) 3.24x10-6 16,000 23,000 41.450 1,450 2,700 6(b) 1,74xIg6 9,300 23,000 41,450 1,450 4,600 Notes:

Sum of industry plus NRC implementation costs ($s) divided by dose reduction (person-rem).

2 Technical specification change,67 plants, proposed resolution.

3(a)

Siloc kout,67 plants.

3(b)

RCPrestart 67 plants.

3(a&b) Both Si and RCP,67 plants.

4(a)

ACI removal, w/o cost for disconnecting ACl,40 PORY plants.

4(b)

ACI removal, w/ cost for disconnecting ACl,40 PORY plants.

5 Safety grade OMS,40 PORY plants.

I 5(a)

Sensitivily Study, safety grade OMS,40 PORV plants.

6(a)

Pressurizer bubble, peak pressure less than 600 psi,67 plants.

6(b)

Pressurizer bubble,10% chance of reaching 2500 psi,67 plants.

ES 3 NUREG 1326

_~ _ - _

(;

y

{

'1, STATEMENT OF PROBLEM General Design Criterion 15 of Appendix A to 10 CFR evaluations, under Multi Plant Action liem B 04 (NUREO 4

Part 50 requires that "the reactor coolant system and 0748 - Ref.2). he current staff guidelines for the LTOP associated auxiliary, control, and protection systems shall system are found in Standard Review Plan 5.2.2, be designed with sufficient margin to answe that the

'Overpresswe Protection," and in its attached Branch I

design conditions of the reactor coolant pressure boundary Technical Position BTP R$B 52, "Overpresswe j

are not exceeded during any condition of normal Protection of Pressurized Water Reactors While Operating i

opemtion, including andcipated operadonal occurrences."

at Low Temperatures" (NUREO 0800 - Ref. 3),

)

Anticipated operational occurrences, as defined in Twelve overpresswe transients, in PWRs, were reported Appendix A to 10 CFR Part 50, am *those condidons of during the period from 1981 to 1983 (Ref. 4) after l

normal operation which are expected to occur one or more cornpletion of USl A 26. Two of these events, at Twkey times dwing the life of the nuclear power unit and include Point Unit 4, exceeded the technical specification t

- but are not limited to loss of power to all recirculation pressweAemperatwe limits, in addition, during this same I

b pumps, tripping of the turbine generator set, isolation of timeframe there were 37 reponed instances when at the main condenser, and loss of offsite power."

least one LMP channel was out of service, in 12 of these cases, both LMP channels were inoperable, i

General Design Criterion 31 of Appendix A to 10 CFR Part $0 requires that "the reactor coolant pressure De continuation of overpressure transient events, and the i

boundary shall be designed with sufficient margin to unavailability of LMP protection channels, suggested the

- assure that when stressed under operating, maintenance, need to reevaluate the current overpressure protection testing,, and postulated accident conditions (1) the criteria, or their implementation, to determine whether boundary behaves in a nonbrittle manner and (2) the additional considerations are warranted.

I probability of rapidly propagating fracture is minimized.

i De design shall reflect consideration of service Major overpressurization of the reactor coolant system l

temperatwes and-other ' conditions of the boundary while at low temperature, if combined with a critical crack L material under operating, maintenance, tesdng, and in the reactor pressure vessel welds or plate material, postulated accident conditions and the uncertainties in could result in a brittle fracture of the pressure vessel. As determining (1) material properties, (2) the effects of long as the fracture resistance of the reactor preature

{

irradiation on material properties, (3) tvaidual, steady state vessel material is relatively high, these events are not and transient stresses, and (4) size of flaws."

expected to cause vessel failme. However, the fracture t

resissance of the reactor pressure vessel-materials

Appendix 0 to 10 CFR Part $0 provides the fractme decreases with exposwe to fast neutrous during the life of i

toughness requirements for the. reactor pressure vessel a nuclear power plant. The rate of decrease is dependent under certain conditions. To ensure that the Appendix 0 on the metallwgical composition of the vessel walls and limits of the reactor. coolant pressure boundary are not welds, if the fracture toughness of the vessel has been exceeded during any anticipated operational occurrences, reduced sufficiently. by

.ncuuon

. Irradiation, technical specification pressureAemperature limits are

- low temperature overpressure - events could cause L

provided for operaung the plant.

propagation of fairly small flaws that might exist near the inner surface. The assumed initial flaw might propagate In the late 1970s, it was noted that there were a large into a crack through the vessel wall of sufficient extent to number of events occurring at reactors while operating at threaten vessel integrity and, therefore, core cooling l

low temperatures (shutdown modes) where the technical capability, Fallure of the pressure vessel could make it -

. specification pressureAcmperature limits were being impossible to provide adequate coolant to the reactor core exceeded. The frequency of these overpressure transients-and could result in major core damage or a core damage j

i

-was determined to be within the anticipated operational accident.

U f occurrence definition.

ne - safety significance of these continuing low temperature overpressure protection (LTOP) was low temperature overpressure transients was designated as j

' designated as Unresolved Safety Issue (USI) A 26 in 1978 Ocneric issue 94,

" Additional low Temperature (NUREG 0371 - Ref.1). PWR licensees implemented Overpressure Protection" (Ref. 5). GI 94 applies to the procedures to reduce the potential for overpressure events design and operation of all PWRs. ' BWRs have been and installed equipment modifications to mitigate such excluded from consideration because they do not normally

' events based on staff recommendations from the USI A 26 operate in a water solid configuration.

11 NUREO 1326 D'

W M

-i 7-w--w*Wy-w y ey w V www' et v--'ss war *=

-r-ttW---

a,-

yg-try9-r

'--"tV-*-NPMW"'T

'9%'WfP'F 4'FT'** - -W44W'--i'

""'d'N'r8-T"r4esu-'$-wT'*w-ererT-r-l%y-ee "ep-i-eg a--,-t4V-

2. OBJECTIVES The general objective of 0194 is to evaluate the need for initiate may arrest in the tougher sections of the vessel additional low temperature overpressure protection and to (farther away from the inside of Oc vessel where the examine alternatives to reduce the risk of core damage irradiation damage is attenuated). VISA accounts for Otis, accidents associated with low temperature ovemressure For cracks that may propagate through the wall, some may events in PWRs by reducing the likelihood of these events, not result in core damage, llowever, studies performed as

Ihe tesis for this is the need to ensure that there is a low part of the PTS effort indicate that a large fraction of likelihood of brittle reactor pressure vessel failure TWCs could result in large openings in the reactor vessel (through wall crack. (TWC)). Such a failure, unlike most for longitudinal welds, or complete opening of other accident scenarios that can lead to core damage, circumferential welds (Ref. 8).

could result in the reactor pressure vessel's being unavailable for either subsequent recovery of the reactor For the purpose of evaluating the risk from core or as an additional barrier for fission product low temperature overpressure events, the NRC staff retention, assumes that the probability of a through wa'I crack is in considering Oc risks associated with low temperature overpressure events, the NRC staff has identified specific in addition to minimizing the likelihood of brittle reactor characteristics related to these events that differ from most pressure vessel fracture (a through wall crack), the general core damage accidents. The concerns are related to the objective of the proposed requirements is to make the risk failure of Oc reactor pressure vessel itself, not the failure from LTOP transients during shutdown operations a small of emergency core cooling systems or decay heat removal contributor to the overall risk associated with the systems. In addition, low temperature overpressure events operation of a PWR, based on the guidance and objectives relate to shutdown modes of operation, Modes 4,5, and 6, of the Commission's Safety Goal Policy Statement (Ref, and the containment may te open during one of these 9). On the core damage frequency (CDF) risk level, a

events, target for the resolution of Ocneric Issue 94 is that the contribution from LTOP transients be a small part 4

Low temperature overpressure protection (LTOP) is a (a few perccnt) of an overall CDF target of lx10 per subset of the broader class of events related to reactor reactor year?

pressure vessel integrity, commonly referred to as pressurized thermal shak (PTS) events, llowever, the Since LTOP transients occur most frequently in Mode 5, severe thermal stresses due to overcooling of the reactor when containment may be open, an LTOP transient CDF pressure vessel are not present during LTDP events.

target of lx10 6 per reactor year may also te considerco to When ITS was being evaluated by both the industry and be compatible with the proposed general performance the NRC staff in the early 1980s, the requirements of USl guidelines given in the Commission's Safety Goal Policy.

A 26 had just teen imposed on the industry and i.e., that the probability of a large release frcm an consequently LTOP was not addressed in these studies. It operating nuclear power plant should be no gevater than was tulieved diat the resolution of USl A 26 had lx10 6 per reactor year. A more direct comparison of this adequately resolved LTOP concerns and these events were CDP target with the policy guidelines requires a definition not considered in the probabilistic risk assessments of "large release" in the policy statement.

performed (Ref. 6).

Reactor pressure vessel failure resulting from brittle fracture is generally defined as a through wall crack More rmndy, a core damase frequency seat of Sales.,,cacio, (TWC), resulting from the initiation and propagation of an p

assumed small flaw in the vessel. The probability of a year has been proioned under the safety soalimplementation presram.

TWC, or the vessel fracture probability (VFP), is

Ihis is a factor of iwo lower than the late value used herein. but is calculated with the VISA computer program (Ref. 7) for withm the uncertainty inherent in calculations and asiuid/.bi made an assumed transient. Crack, of flaw, initiation may not assening canphance with either goal, and Lt adaptim in heu of a i

always result in a TWC. Depending on the vessel material talM goal would not affect the recommendationi made in this characteristics and the assumed transient, some cracks that regulatory analysis.

21 NUREO 1326

3. ALTERNATIVE RESOLUTIONS la reaching its proposed resolution of G194, the staff for overpressure protection parallels that of the considered six specific attemative courses of action.

PORY group of plants.

'Ihese are discussed below. The requirements would le applicable to all Westinghouse and Combustion e

The current technical specification ACTION Engineering plants, toth operating reactors and reactors in staternent allows 7 days to restore an ineperable the construction stage of licensing, nfty two LTOp channel to oferable status or depressurize and Westinghouse plants and l$ Combustion Engineering vent the reactor coolant system (RCS) within the plants are considered in the detennination of the industry next 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . With both LTOP channels inoperable, and the NRC irnplementation costs.

Additional the ACTION statement requires the RCS to le discussions of each of these alternatives is provided in depressurized and vented within 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months .

Section 5.

e Under the current bases for Specification 3.0.4, 3.1 Alternative 1.No Action Alternative operations need not te restricted when corrective action should be ta'.en to obtain compliance with a This alternative assumes that no additional specification under certain situations even if the low temperature overpressure protection need be corrective actions are required within a limited provided. It also assumes that all applicable requirements period of time. Exceptions from Specification 3.0.4 and guidance to date have been implemented, but no have been provided for a limited number of implementation is assumed for related generic issues, or specifications when startup with inoperable other staff requirements or guidance, that are still equipment would not affect plant safety, unresolved or still under review.

3.3 Alternative 3. Si and RCP Restrictions 3.2 Alternative 2. Change to Technical Specifications Altemative 3 would require removal of all power to safety injection pumps and prohibit reactor coolant pump restart To achieve the otijectives stated in Section 2 above, this while in a water solid condition.

alternative calls for a modification to the plant technical specification for overpressure protection to ensure that 3.4 Alternative 4 Removalof RilR toth low temperature overpressure protection channels are Autoclosure Interlock operable, especially in a water solid condition; that is, to treat operationally the low temperature overpressure Alternative 4 would allow for crediting use of the RHR protection system as a system that performs a safety relief valves,in addition to the PORVs, for pressure safety related function. A summary of specific aspects of relief in mitigating an LTOP transient. Removal of the this alternative is as follows:

autoclosure isolation (ACl) interlock on the RIIR suction lines would be an additional requirement and has been e

The role of PORVs has changed such that PORVs evaluated as part of Generic Issue 99, *Ri!R/RCS Suction are now relied upon to perform one or more of the Line Interlocks on PWRs." For plants that rely on only the following safety related functions:

Ri!R safety relief valves for LTDP protection, no additional benefit would be obtained from this alternative, a.

mitigate a design basis steam generator tube rupture event, 3.5 Alternative 5. Safety Grade LTOP System b.

Iow-temperature overpressure protection of the reactor pressure vessel during startup and Alternative $ would require that the low temperature cooldown, or over;" essure protection system be upgraded to a fully safety-grade system.

c.

plant cooldown in accordance with Branch Technical Position BTP RSB 51 to Standard Review Plan Section 5.4.7,

Residual lleat 3.6 Alternative 6 Pressurizer llubble Removal (RIIR) System" (Ref. 3).

.\\lternative 6 would require that water solid operation te o

For plants that rely on safety relief valves in the prohibited by providing for a steam or nitrogen bubble in residual heat removal systern for low-temperature the pressuri7er at all times (other than during hydrostatic overpressure protection, the technical specification pressure tests).

31 NUREG 1326 m

i 1

4. TECilNICAL FINDINGS

%e PNL cvaluation of low temperature overpressure part of the ITS effort indicate that a large fraction of protection included a determination of die frequency of TWCs could result in large openings in the reactor vessel overpressure transients while at low temperatures, the for longitudinal welds, or complete opening of circum-failure of the overpressure protection system on a demand ferentialwelds(Ref 8).

basis, and the theoretical peak pressure that could te obtained, given failure of the overpressure protection in addition to evaluating the operating reactor experiences system. The probability of teactor pressure vessel fracture frequencies, the root causes for the overpressure transients due to brittle fracture was estimated based on the neutron.

and overpressure protection system unavailability were induced embrittlement of the limiting vessel material over also determined. To the extent feasible, the operator the remaining lifetime of each plant. The detailed evalua.

reactkins and responses to the actual events were also tion is found in NUREG/CR 5186 (Ref.10),

considered by PNL, thus establishing the test estimate, or base case, profile for the low temperature overpressure

%e frequency of overpressure transients was determined transient event frequency, overpressure protection system from actual operating reactor experiences, as reported in unavailability, and theoretical peak pressure spectrum.

the licensee event report (LER) system. Special reports and industry re[ orts were also reviewed to augment the in reviewing the actual operating reactor events as I

LER data tuse. Technical specification reporting require-reported in the LERs, it was noted that the licensees have ments provide for a 30 day report to the Commission always reported prompt operator action in response to whenever the low temperature overpressure protection low temperature overpressure transients, in 1 to 2 system is used to mitigate a pressure transient. Overpres-minutes, the operators have diagnosed the situation and sure protection system unavailability was also determined taken appropriate actions to terminate the effects of the from the same sourtes. The period from 1980 through the transients, usually resulting in only one or two cycling of end of 1986 is considered in this evaluation. The LER and the PORVs before stabilizing the reactor coolant system literature search was performed by the NRC staff and is pressure below the Appendix 0 limits. PNL therefore sammarized in Appendix A. More recent LTOP events assumed a 3. minute operator response time to develop the that have occurred in 1987 and the early part of 1988 have best c:.:imate peak pressure spectrum; somewhat been reviewed to determine if tbc base case evaluation conservative but not a worst case evaluation.

would have been altered by a detailed accounting of these events. PNL concluded that the base case would not be al.

Tne operating reactors were classified by overpressure tered by rnore than about 10%. None of these more recent protection system design. nree groups were identified.

events would have resulted in high reactor ca.5nt system pressures if the LTOP systems had failed.

1.

Two PORVs, water solid operation allowed: 32 The reactor pressure vessel fracture probability was Westinghouse units and eight Combustion Enginect-obtained by PNL from the VISA computer program (Ref.

Ing units (40 total units).

7). To account for the effects of neutron irradiation damage, the mean surface RT(ndt), reference temperature nil-ductility trar sient, shift was calculated using Revision 2.

Two SRVs in the RilR, water solid operation 2 to Regulatory Guide 1.W (Ref. I1),

allowed: nine Westinghouse units and six Combus-tion Engineering units (15 total units).

Reactor pressure vessel failure resulting from brittle frac-ture is generally defined as a through wall crack (TWC),

resulting from the initiation and propagation of an 3.

Single PORV with pressurizer bubble at all times, all assumed small flaw in the vessel. The probability of a Babcock and Wilcox plants (8 units).

TWC, or the vessel fracture probability (VFP), is calcu-lated with the VISA computer program (Ref. 7) for an assumed transient. Crack, or flaw, initiation may not nere are a small number of plants that have been licensed always result in a TWC. Depending on the vessel material with low-temperature overpressure protection systems that characteristics and the assumed transient, some cracks that do not fall into one of the above groups flowever, they initiate may arrest in the tougher sections of the vessel have been included in one of the three groups dependent (farther away from the inside cf the vessel where the it-on either the type of relief path (PORV or SRV) employed radiation damage is attenuated). VISA accounts for this or by determining the appropriate group based on the tech-For cracks that may propagate through the wall, some may nical specification requirements currently in existence at not result in core damage. Ilowever, studies performed as the plant.

41 NUREG 1326

,I~1

'1 Technical Findings low temperature. overpressure protection is required ing during this timeframe. Twenty three occurred in the

during shutdown modes of operation, Modes 4,5, and 6.

PORY class plants and seven in the RHR SRV class A review of the actual events concluded that virtually all plants. Two events in the PORY plants and one in the events are occurring in Mode $ with reactor coolant _

RHR SRV plants exceeded the Appendix 0 limits. The system temperatures ranging from 800F to 190oF. - The PORY plants accumulated 244 reactor years of experience characteristic transient used to determbe the vessel frac-and the RHR SRV plants accumulated $6 reactor years of ture probabilityc conditional on event occurrence, was experience.

developed based on the actual operating events. A vessel l

' wall temperstwe of 120"F and a heatup rate of 25'F per Essen% all the LTOP challenge events have occurred 1

how were used, ne 120oF wall temperature is represen-w%n the reactor coolant system was water solid -

tative of-the average ' temperature at which low.

(pressurizer filled). The events fall into two categories:

temperature overpressure events have occurred, nis those resulting from mass addition and those caused by l

temperature is also lower than that at which the reactor energy addition. Both of these types of events cause rapid.

- pressure vessel head may be removed at many plants and pressurization when the reactor coolant system is water-represents 1a reasonable limiting temperature for this solid.

evaluation. = A 25 F per hour heatup rate appears to be a j

. reasonable estimate based on heatup with decay heat and A representative pressurization rate for mass addition

residual heat removal system pump energy prior to reactor events was calculated by PNL. Using the compressibility J

coolant pumprestart. ~

of water of 0.0046% della volume per unit volume per j

atmosphere of pressure increase and based on a reactor -

A peak theoretical pressure was determined by PNL for coolant system volume of-11,000 cubic feet, the pres.

each event, assuming the low temperature overpressure surization rate was calculated to be 3.8 psi per gallon protection system failed to mitl ate the event.- If an addi.

injected into the reactor coolant system.

E

= tional pressure relief path were available, for example, the residual heat removal system safety relief valves, the peak Water expands when heated, increasing the pressure of a

[

pressure'was limited to the SRV setpoint, provided the water solid system. PNL calculated the pressurization rate '

rated relief capacity could accommodate the challenge, for energy addition events using the above 'value for i

compressibility and the specific volume of water in the For' events that occurred without another pressure relief 100 to 200oF temperature range. - The pressurization rate path, the peak theoretical pressure was limited to that was calculated to be 100 psi per F. increase in the reactor o

which could be achieved 3 minutes after the initiation of

~ coolant system average temperature. Slow heating of the l

. the event. nc 3 minute operator action response time reactor coolant system by operation of a single reactor a

was determined by PNL from the. review of the actual

. coolant pump generally does not exceed 250F per hour, events._ In all' cases the operators have recognized the

' which would conrespond to a pressure increase of about 42 overpressure transient and initiated actions to correct the psi per minute.

situauon within a 1. to 2-minute time period. The NRC-staff attributes this rapid response characteristic to training De mass addition events resulted in injection rates rang.

l

. and c procedure development in response to the ing from 20 gallons per minute (gpm) to 600 gpm. The implementation of Unresolved Safety Issue A 26. In addi-higher rates are associated with inadvertent safety injec.

tion, a rapid response will limit the number of times the tion events, while the lower values are typical of excess.

PORY is cycled while mitigating the event, charging without letdown. De energy addition' events are identified by the differential temperature between 'the The Babcock and Wilcox plants have been excluded from secondary side and primary side of the reactor coolant this evaluation because these units have not experienced systemi ne warmer steam generator is the heat source for

.any low temperature overpressure transients and do not these events.- The allowable temperature differential for m

. contribute to the overall risk of low temperature overpres.

reactor coolant pump restart is specified>in the current sure events. ' Babcock and Wilcox plants do not operate h technical specification. In the actual experiences data N

a' water solid condition. A steam or nitrogen bubble is base, the differential temperature is usually 'small,

'r

' maintained in ' the pressurizer. The bubble provides a ahhough in one case it was reported to be 85 F.

]

minimum of l_0 minutes for the operator to respond to an anticipated low temperature overpressure event. A single The operating reactor experictwes data base is summarized -

path is provided for pressure relief (PORY or RHR SRV).

in Table 4.1 for the PORY plants and in Table 4.2 for the ~

j RHR SRV plants, included in these tables are the mass -

For the period from 1980 through the end of 1986, there flow rates, temperature differences, the LTOP pressure F

were 30 challenges to the L10P systems at the 55 setpoints and the calculated peak (or hypothetical) pres.

(

- Westinghouse and Combustion Engineering plants operat.

sures assuming failure of the LTDp system. The actual H

LNUREO 13261

-4 2 N

Technical Findings peak pressure is provided for reference. Also provided are Operated Relief Valve and Block Valve Reliability." The the various pressure setpoints for the pressure relief estimated consequence for this event was found to be 9 systems as well as the specific feature that limited the million person. rems over a 30-year period for a typical peak pressure, either an alternative relief path or the as-eastern site with 100 persons per square mile population sumed 3 minute operator response time.

density. A 50-mile radius is used for all consequence evaluations as required by current NRC guidelines (Ref.

Table 4.3 summarizes the current frequency of low-13),

temperature overpressure events and overpressure protec-tion system unavailability. The vessel fracture probability Because of the wide variation in the plant specific vessel (or core damage frequency), as derived from the actual fracture probabilities and because of differences between operating reactor experiences, is also provided for the sites, the NRC staff did not select either a " typical" plant i

l mean plant in each group over the remaining plant lifetime or use the " average" plant for this analysis (for example.

l of the group. The mean core damage frequency for low-use the value and impact for the " typical" or " average" temperature overpressure events, as determined under plant and multiply the results by the number of plants l

0194, is 3.24x104 per reactor year. The mean core within a group). The variation in plant specific vessel damage frequency for the PORY plants is 3.04x104 per fracture probability, as well as the variation in site specific reactor year, and the mean core damage frequency for the consequences based on population density and environ.

RHR SRV plants is 3.76x104 per reactor year, mental factors, are considered in this evaluation of risk and consequence.

ne likelihood of a low-temperature overpressure event and the likelihood that the overpressure protection system will fail on demand, as well as the resulting theoretical De plant-specific vessel fracture probability, integrated pressure spectrum, are considered to be equal for all plants over the remainder of life, was obtained for the mean within each group. Dat is to say, no credit or penalty was surface RT(ndt) shift expected to occur for each plant (the given to any plant as a result of that plant's specific VISA probability results are provided in terms of the mean operating history. However, the resultant estimate of surface RT(ndt) value).

vessel fracture varies from plant to plant and from year to year. The probability of reactor pressure vessel fracture To account for site specific variables, population density, increases with plant life. Vessels become more brittle environmental conditions, and reactor size, the generic with age as a result of neutron irradiation. Frorn plant to consequence in Reference 12 has been scaled, by the NRC plant, the chemical composition of the limiting reactor staff, to the Siting Source Term data provided in pressure vessel material also varies, and the plant. specific Reference 14. Appendix B provides a discussion of the vessel fracture probability for each plant is unique to that technique employed and compares the results to tie composition. The estimated vessel fracture probability for generic source term evaluation. De objective of perform-each plant has been calculated and summed to obtain the ing this scaling evaluation is to ensure that the estimate of group total and mean values.

the consequences from low temperature overpressure transients properly accounts for the plant specific varia-A review of current standard technical specifications for tions in the source term (as a function of the power rating containment integrity in shutdown modes (Modes 4,5, and of a plant) and in the population density and environmen-

6) indicates that no containment integrity requirements are tal factors that affect the calculation of the consequences, imposed for reactor coolant temperatures less than 200T, ne scaled source term evaluation results in a 30% reduc-except during refueling operations when the reactor pres-tion in estimated dose, in person-rem, as compared to the sure vessel head is removed. Since the low temperature generic source term evaluation (based on Ref.12),

overpressure events of concern to this evaluation occur in Mode 5 at reactor coolant temperatures between 807 and The results of the scaling study are provided in Table 4.4 1907, the assumption that containment is open, at least for each plant category and for different release part of the time, is judged to be valid. Containment categories, based on the through wall crack frequencies, or integrity will be treated parametrically in this analysis.

core damage frequencies, presented in Table 4.3. The generic value overestimates the consequences for the RHR The consequence evaluation for a low temperature over.

SRV plants. Dese plants are the older, low power units pressure event, which results in reactor pressure vessel or the newer plants that tend to have better reactor pres.

fracture, was obtained for a late core meh sequence with sure vessel materials. He dominant contribution based on containment bypass (Ref.12). De studies were per-vessel fracture probability comes from the older, low-formed in conjunction with Generic issue 70, " Power-power plants.

43 NUREG 1326 i

f Technical Findings Table 4.1 Summary of PORY class operating reactor experiences.

Mass Addition Events Mass PORY RHR RHR RCS Peak Pres Hypo Actual Flow LTOP ACI SRV Initial Limited by Peak Peak Rate SP SP SP Pres All 3

Pres Pres Plant Yr (gpm)

(psi)

(psi) hth Min (psi)

(psi)

Calvert Cliffs 83 40 450 300 315 400 7

850 425 Ginna 83 40 435 450 600 310 X

760 435 North Anna 1 81 300 430 600 467 unk X

467 430 North Anna 1 83 530 430 600 467 350 X

467 430 North Anna 1 84 40 430 600 467 350 X

467 410 North Anna 1 85 20 430 600 467 350 X

467 430 Palisades 81 40 400 none 300 300 X

850 400 Salem 2 83 300 375 600 375 unk X

375 375 San Onofrc 1 83 600 522 370 500 300 X

25(0

$22 Surry 1 81 200 410 none 600 350 X

600 410 Surry 1 84 100 410 nor.e 600 325 X

600 412 Surry 1 85 200 410 none 600 350 X

6(O 410 Turkey Pt 4 81 90 415 465 600 310 X

14(0 1100 Turkey Pt 4 81 90 415 465 600 340 X

1400 750 Zion 1 84 100 435 600 450 unk X

450 450 Zion 2 85 190 435 600 450 unk X

2500 435 7 ion 2 86 150 435 600 450 unk X

450 450 Energy Addition Events Mass PORV RHR RHR RCS Peak Pres flypo Actual Flow LTOP ACI SRV Initial Limited by Peak Peak Rate SP SP SP Pres All 3

Pres Pres Plant Yr (gpm)

(psi)

Path Min (psi)

(psi)

North Anna 2 82 85 385 600 467 3M X

467 385 North Anna 2 82 35 385 600 467 350 X

467 365 Palisades 85 min 400 none 300 300 X

$00 375 Salem 2 -

84 min 375 600 375 325 X

375 350 Salem 2 85 min 375 600 375 325 X

375 380 Salem 2 85 min 375 600 375 340 X

375 380 NUREG 1326 4-4

Technical Findings Table 4.2 Summary of RHR SRY class operating reactor experiences.

1 Maas Additloa Events Mass PORY RHR RHR RCS Peak Pres Hypo Actual Flow LTOP ACI SRV Initial Limited by Peak Peak Rate SP SP SP Pres All 3

Pres Pres Plant Yr (gpm)

(psi)

Path Min (psi)

(psi) i Byron 1 85 600 450 700 450 unk X

450 450 Callaway 86 20 450 680 450 400 X

450 463 Farley 1 86 40 none 700 450 400 X

850 450 Farley 2 86 180 none 700 450 400 X

2500 700 Energy Addition Events Mass PORY RHR RHR RCS Peak Pres Hypo Actual Flow LTOP ACI SRV Initial Limited by Peak Peak Rate SP SP SP Pres All 3

Pres Pres Plant Yr (gpm)

(psi)

Path Min (psi)

(psi)

Farley 1 86 min none 700 450 400 X

550 450 Failey 2 83 min none 700 450 400 X

600 480 Summer 86 min none 700 450 400 X

$50 450 Since assumptions regarding containment are important to specific value ranges from about 40 to 3000 (Indian Point) l the estimation of public risk, and because LTOP transients with a median value of 185 persons per square mile. For predominately occur in Mode 5 when containment the plants considered in this evaluation, the mean popula-integrity is relaxed to allow for testing, maintenance, and tion density used to determine the consequences is es-the repair of equipment, three estimates for public risk are timated to be 280 persons per square mile (Refs.14 and used to study the effects of containment assumptions. The

16) based on 1982 population estimates. The projected best estimate evaluation is based on a 50% probability mean population density for the plants considered is es-weighting of a large release as a result of the timated to be 480 persons per square m'.se by the year containment's being open at the time of the core damage 2000, or about 50% higher, Though not tensidered in this.

accident and a 50% probability weighting that the release evaluation, if an adjustment to account for the projected is small, or similar to an SST2 release. The high estimate population growth is desired, then the consequences is based on the containment's being open at the time of the (person rem values) could be multipt'ed by 1.5 and the core damage accident, he low estimate is based on a value/ impact ratios ($ per averted penon rem) divided by 10% probability weighting of a large release as a result of 1.5 (or multiplied by 0.7). His assumes that the conse-the containment's being open and a 90% probability quences would be directly related to the increased popula-weighting of a small release. The base case consequences tion.

for the three containment assumptions are provided in Table 4.5.

As seen in Table 4.4, assumptions concerning containment integrity for fission product retention can result in an order The mean 50-mile radius population density around U.S.

of magnitude change in the estimated consequences of an nuclear power plant sites is estimated to be 340 persons unmitigated LTOP transient. This would suggest that per square mile by the year 2000 (Ref.15). The site-additional or modified containment requirements for 45 NUREO 1326 l

1

Technical Findings Modes 4,5. and 6 could achieve the risk reduction objec.

31 to provide assurance that the probability of a rapidly live of 0194. The NRC staff has not considered this as a propagating fracture of the reactor pressure vessel during proposed alternative for the resolution of 0194. The anticipated operation occurrences is minimized.

LTOP protection concern is related to GDC 15 and GDC Table 4.3 Base case mean core damage frequencies from LTOP events.

Challenge Overpressure Mean TWC Mean Core Plant Frequency L'IDP Spectrum Probability)

Damage Freg Over Life 0 (per R Y)(2)

Category (per R Y)

Unavailability psi 1rac 1.PORVs 0.094(3) 0.0870) 2500 0.09 3.92x!&3(4) 2.89x10 6 (40 plants) 1400 0.09 1.96x10-4 1.44x10-7 850 0.13 5.32x10 6 5.66x104 600 0.69 2.06x107 1.16x104 Total 3.04xl&6

2. RilR SRVs 0.125(5) 0,143(5) 2500 0.14 1.50xl&3 3.75x10-6 (15 plants) 850 0.14 2.00x10 6 5.01x104 600 0.72 2.21x10-7 2.84x109 Total 3.76x10 6 J

3. B&W 0.018(6) o,og7(6) 850 0.05 1.09x1&6 negligible (8 plants) 600 0.95 1.00x10 7 negligible Total negligible Industry Mean Core Damage Frequency 0) 3.24x1p6 f

Notes:

(1) Mean thruush wall crack OWC) probeMiity for all plants in each category, averaged over remaining lifetime from 1986 through end.

of. license. PORY plants: 969 total years,24 year:51 ant. RHR SRV plants: 452 total years,30 year:51 ant. B&W plants: 183 total years,23 yearstlant.

(2) Mean core damage frequency based on probability of readting peak pressure and total for each category of plan:s. Based on assump. -

tion that through wall cra:k leads to core damage.

Q) Frequency: 23 events in 244 reactor years. Unavailabihty: Based on the two events at Turkey Point 4 in 1981, unavailaN1ity is two out of 23 demands.

(4) Read as 3.92 times 10 to the minus 3, or 0.00392.

t (5) Frequency: 7 events in 56 reactor years. UnavailaMhty: Based on the event at Farley 2 in 1983, unavailability is one out of seven danands. '

l l

(6) Frequency:14:s then one event in 56 seactor years. Unavailability: Assume single PORY failure rate simDar to PORY class of plants.

G) - (40 x 3.0tx10' + 15 a 3.76x108y(40+15) l NUREG 1326 46 i

Technical Findings Table 4.4 Consequences for various release categories.

Generic Scaled SSTI SST2 Value Value Value Value (Person Rem)

(Person Rem) (Person. Rem) (Person-Rem) 40 PORY Plants 26,600 25,000 30,700 2,300 15 RHR SRV Plants 15,300 4,600 5,700 300 55 Total Plants 41,900 29,600 36,400 2,600 Notes:

Generic Wlue - bued on Reference 12 L1T>P transient. -

Scaled Value -

includes consideration of differenses between units based on populatice density, environmental canditions, and operating power levels.

SSTI Value.

band on SSTI reneue, similar to PWR.2 release fractions. Duvet broads of containman.

STD Yalue -

based on SST2 release, similar to PWR 5 relene fractions. Cantainment fission product mitigation systems func-tian, failure to isolate containment.

Table 4J Base case consequences for various contalament assumptions.

Best Estimate High Estimate Low Estimate (50% Scaled Value (100% Scaled (10% Scaled Value plus 50% SST2 Value)

Value) plus 90% SST2 Value)

(Person Rem)

(Person Rem) 40 PORY Plants 13,600 25,000 4,570 15 RHR SRV Plants 2,400 4,600 730 55 Total Plants 16,000 29,600 5,300 l

47 NUREG 1326

- - - +

e

_-y

4

5. VALUE/ IMPACT ANALYSIS 5.1 Costs and Henefits of Alternative For cracks that may propagate through the wall, some may Proposed Resolutions not result in core damage. However, studies performed as part of the PTS cffort indicate that a large fraction of 5.1.1 Alternative 1. No Action Alternative TWCs could result in large openings in the reactor vessel for longitudinal welds, or complete opening of circum-his alternauve assumes that no additional action is neces-ferential welds (Ref. 8).

sary, based on the evaluation of the crirent risk associated with low temperature overpressure events and on the The likelihood of a reactor pressure vessel through wall staff's review of the operating retetor experiences from crack or the mean core damage frequency estimate, based 1980 through the end of 1986. It is also assumed that all on the operating reactor experiences, is 3.24x104 per applicable requirere and pdance approved to date reactor year over the remaining licensed life of the PWRs have been implemented, but no implementation is evaluated. As a plant approaches the PTS screening i

assumed for related generic issues that are still currently criteria,10 CFR 50.61, the through wall crack probability

- unresolved, will increase to 7.4x104 per reactor year, assuming the LTOP event frequencies, unavailability, and peak pressure in considering the risk associated with low temperature spectrum profiles remain constant.

overpressure events, the NRC staff has identified specific characteristics related to these events that differ from most ne PNL cvaluation for the mean through wall crack core damage accidents. The concerns are related to the frequency, or mean core damage frequency, is based on failure of the reactor pressure vessel itself, not the failure the operating reactor experiences and represents a best of ernergency core cooling systems or decay heat removal estimate evaluation of the risk from low temperature over-systems. In addition, low temp:rature overpressure events pressure transiems.

There are uncertainties in the relate to shutdown modes of operation, Modes 4,5, and 6, estimated mean through wall crack frequency These are and the containment may be open during one of these addressed below.

t events.

Frequency of Eseats Low temperature overpressure protection (LTOP) is a subset of the broader class of events related to reactor ne estimated frequency of low temperature overpressure pressure vessel integrity, commonly referred to as events was obtained from a review of actual operating pressurized thermal shock (PTS) events. However, the plant events as reported to the Commission under require.

severe thermal stresses due to overcooling of the reactor ments contained in the technical specifications. For this pressure vessel are not present during LTOP events.

evaluation the events have been limited to actual transients When PTS was being evaluated by both the industry and that have challenged the low temperature overpressure the NRC staff in the early 1980s, the requirements of USl protection system and have occurred after the plant A 26 had just been imposed on the industry and conse-became operational, taken to be the date the unit first quendy LTOP was not addressed in these studies. It was generated electrical power. Pre-commercial events and

- believed that the resolution of USI A 26 had adequately precursor events are not included. Pre-commercial events i

resolved LTOP concems and these events were not are excluded because they pose no risk (no fuel in the considered in the probabilistic risk assessments performed reactor or the vessel has not experienced any irradiation (Ref,6).

damage).

Precursor events, events that could have challenged the low temperature overpressure system but Reactor pressure vessel failure resulting from brittle for other reasons did not, are also excluded because there fracture is generally defined as a through wall crack is no assurance that the reported instances are representa-(TWC), resulting from the initiation and propagation of an tive of actual total experiences. Consideration of pre-assumed small flaw in the vessel. The probability of a commercial and precursor events would increase the TWC, or the vessel fracture probability (VFP) is calcu.

estimated frequency of low temperature events from 0.1 lated with the VISA computer program (Ref. > u an as-per reactor year (30 events in 300 reactor years) to 0.183 sumed transient. Crack, or flaw, initiation may not always per reactor year ($5 events in 300 reactor years) for the result in a TWC. Depending on the vessel material period from 1980 through the end of 1986. Though not characteristics and the assumed transient, some cracks that considered in this evaluation, there were at least an addi-initiate may arrest in the tougher sections of the vessel tional six events in 1987, none of which would have (farther away from the inside of the vessel where the it.

changed the base case risk evaluation significandy (less radiation damage in attenuated). VISA accounts for this, than a 10% change).

5-1 NUREG 1326

Value/ Impact Analysis Oterpressure Protection System Unavailability SRVs, the plant technical specification wuld have allowed removal of one valve for up to 7 days wiu.9ut fur.

De overpressure protection system unavailability for the ther restrictions on plant operadons and this event may not PORY group of plams (0.087 per demand) is based on the have changed. Although LTOP unavailability in ROR two events at Turkey Point in 1981. Of the 23 events in SRV plants could be approximately a factor of two highet this group, overpressure protection was not available in than assumed in this evaluation ( two out of eight versus these two instances. In both events one of the redundant one out of seven), the LTOP event frequency and peak low temperature overpressure protection channels had pressure spectrum would not te changed significantly for been removed from service for maintenance. As a result the Ri!R SRV class of plants. De estimated base case of a single failure in die remaining charmel, the system mean through wall crack frequency for the RHR SRV -

was not able to mitigate the pressure tmnsient to the class of plants would increase from 3.76x10 6 to 6.0x104 Appendix 0 pressure / temperature limits. Prompt operator per reactor year. The mean frequency for all plants would action resulted in limiting the peak pressures to 1100 psi increase from 3.24x104 to 3.85x10+ per reactor year, and 750 ps1.

. In all three cases, one of the two low temperature over.

De overpressure protection system unavailability for the pressure protection channels had been removed from serv.

SRV group of plants (0.143 per demand)is based on the ice for maintenance. Plant startup, allowed under the cut-event at Farley 2 in 1983. Of the seven events in this rent technical specificadon, resulted in exceeding current group, overpressure protection was not available in this Appendix 0 pressure / temperature limits as a result of a one instance. In this event one of the redundant low-single failure in the overpressure protection system during temperature overpressure protection channels had been anticipated low temperature overpressurc events.

removed from service for maintenance. As a result of a single failure in the remaining channel, the system was not Since the technical specification for all plants would allow able to midgate the pressure transient to die Appendix 0 for plant openulons under similar circumstances, the low.

- pressure / temperature limiu.

temperature overpressure protection system unavailability based on the actual operating events is judged to be i

he actual failure was attributed to mechanical binding of appropriate for evaluating the potential rish from low-the passive spring loaded safety relief valve, which even-

. temperature overpressure transients.

tually opened at about 700 psi. While it may appear that the unavailability for this group of plants is high and based Yessel Failure Probabilities on an abnormal situation, it is noted diat in 1973 a similar event occurred at Zion 1.

he residual heat removal ne chemistry and fluence data used in this evaluation system safety relief valve failed to open during a were obtained from plant. specific submittals in response charging / letdown imbalance transient and the pressure to licensing acdvities related to US! A 49, Pressurized rose to 1300 psi after the autoisolation setpoint preasure of Thermal Shock." he mean surface RT(ndt) shifts for the 600 psi was reached, resulting in isolation of the residual limiting vessel material were calculated over the plant

-l heat removal system (Ref.17),

lifetime using Reference 11, considered by both the staff and the industry to be representative of the state.of the-art In addition, a potentially significant event occurred at knowledge concerning irradiation-induced damage.

Millstone 3 in January 1988. A combination of system interactions, plant personnel communication errors, and The characteristic transient used to determine the vessel inadequate procedures results in both the PORVs and the failure probability, conditional on event occurrence, was residual heat removal system safety relief valves being developed based on the actual operating events. A vessel unavailable to mitigate a maintenance induced low-wall temperature'of 120oF and a heatup rate of 25*F per temperature event. Prompt operator action prevented the hour were used. The IS ! wall temperature is represen-

' peak ' pressure from exceeding the Appendix 0 tative of the average temperature et which low-pressure / temperature limits. The licensee believed that temperature overpressure events have occurred. This LTOP protection was being provided by redundant temperature is also lower than that at which the reactor PORVs. In reality, LTOP protection was being provided pressure vessel head may be removed at raany plants and by redundant RHR SRVs up to the tin,e when one was represents a reasonable limiting temperature for this

< removed for maintenance. De remaining relief path was evaluation. A 25oF per hout heatup rate appears to be a

, lost as a result of unrelated maintenance activities, which reasonable estimate based on heatup with decay heat and produced inadvertent closure of the RHR suction line residual heat removal system pump energy prior to reactor isolation valves, resulting in a mass addition event--

coolant pump restart. A 500F change in the assumed charging without letdown. Had the licensee recognized vessel wall temperature is estimated to result in a factor of that LTOP protection was being povided by the RilR two change in vessel failure probability (cooler-a factor l'

NUREO 1326 5-2

Value/ Impact Analysis of two larger, warmer-a factor of two less), he vessel temperature overpressure protection is required for failwe probability is dominated by the pressure contribu.

shutdown modes of operation. ~ They. occur most

. Lion. At a heatup rate of 50"F per hour, the fauure proba-frequently in Mode 5, cold shutdown, with reactor coolant bility is estimated to be about 10% higher, as a result of temperatures less than 200 F.

A review of current W

, the increased thermal stress contribution.

standard technical specifications for containment integrity in shutdown modes (Modes 4, 5, and 6) indicates that no

%e staff recognizes that there have been concerns with containment integrity requirements are imposed for reac-

- the flaw distribution (crack size and frequency) assump-tar coolant temperatures less than 200*F except during

tions used in the VISA computer program, as identified refueling when the pressure vessel head is removed, during the rulemaking process for USI A 49,

Pressurized Dermal Shock." However, the use of VISA for this It is therefore reasonable to assume that containment evaluation is appropnate for the quantification of risk due integrity and containment isolation are questionable.

to brittle vessel faBure in that the results are consistent and Containment has been treated parametrically in this

- based on a previously developed methodology well known evaluation and, for the assumptions used,' the proposed

- to both the NRC staff and the industry. Sensitivity studies resolution is shown to be cost beneficial and well within point to uncertainties in the flaw size distributions as the the value/ impact ratio of $1,000 per person rem svened.

major unknown factor in predicting the failure probability of reactor vessels (Ref. 7).

The frequency of low temperature overpressure events has remained relatively constant. Excluding precursor events Sensitivity studies have been performed on the VISA and including pre commercial events, the frequency of-models used to determine the reactor pressure vessel events prior to 1980 was about 0.15 per reactor year, and through wall crack (TWC) probability in corQunction with after 1980 (through the end of 1986), 0.12 per reactor the pressurized thermal shock studies for USl A 49 (Refs.

year. In total, from 1%9 through the end of 1986, there 18 and 19). De effects of material property unconsinties, were 91 low temperature-related events (precursor and such as the fracture toughness and crack arrest toughness, pre 4.nd post commercial) occurring at 36 of the 55 plants and uncertainties in the assume l distribution of the copper evaluated in this study. It is therefore reasonable to.

content in the vessel material tend to increase the TWC assume that the event frequencies developed during this probability calculated from a factor of Iwo to about an or-study are applicable to all Westinghouse and Combustion der of magnitude. On the other hand, the effects of the Engineering PWRs. Further, the actual events all fall crack length to depth ratio assumption indicate a decrease within the design base for the low. temperature overpres-in the TWC probability calculated. For the L'!DP tran-sure protection system. mass and energy input imbalances sient, in going from an infinite length crack (the base case) resulting from charging / letdown flow mismatches, to a 6:1 length to-depth crack, the TWC probability inadvertent safety injection, and reactor. coolant pump decreases by a factor of about three. -The infinite crack

restarts, length as@ption is recommended for use because strong experimental data bases have not been established for jus.

From the above operating experience, the frequency of Lifying the use of a particular flaw length distribution (Ref.

- low temperature overpressure events is expected to remain -

19), ne assumptions used are consistent with the recom-constant, at about 0.1 per reactor year. It is therefore mondations of References 18 and 19.

appropriate for the NRC staff to consider that low-temperature overpressure events are anticipated transients he through wall crack probability obtained from VISA is and that the requirements of Appendix 0 in defining the used to obtain an estimate of the core damage frequency acceptable pressure / temperature limits for operation are and to allow the NRC staff to estimaic the relative impor-

'also appropriate, lance of low temperature overpressure transients. The impact attributes (cleanup and repair, replacement power he NRC staff does not recommend this alternative ("no -

costs, and offsite damage costs) are based on risk reduc-action").

Low temperature overpressure protection tion estimates of the core damage frequency for proposed requirements have been imposed on PWR licensees to W.

alternatives. In addition, the base case (or current) risk ensure that adequate protection against brittle reactor pres-

. estimate can be used to estimate the potci.tial costs of the sure vessel failure is provide <!, particularly for anticipated no action allemative, operational transients. To ensure that adequate margins-are maintained, the - Appendix O pressure / temperature ContainmentIntegrity limits are identified in the technical specifications to meet the requirements of General Design Criterion 31. While De consequence evaluation for low temperature over-the overall probability of a through wall crack is estimated

. pressure events is based on containment bypass or failure to be on the order of 3x104 per reactor year, the likelihood of containment to isolate following an event, low-of exceeding the Appendix 0 pressure / temperature limits q

h 53-NUREO 1326

-_m

Value/ Impact Analysis is once in cach ten events as a result of LTOP system Two groups of plants are used for this evaluation, as dis-unavailability, cussed previously, he grouping was based on the type of low temperature overpressure protection system ernployed No costs are usually attributed to a No Action" alterna-at the plant. Group I consists of those Westinghouse and Live because the future costs of accidents are convention.

Combustion Engineering plants that use PORVs for ally counted as benefits or averted costs in the assessment protection. Group 2 consists of those Westinghouse and of the alternative actions. Ilowever, a core damage Combustion Engineering plants that use safety relief accident resulti',g frorn a low temperature overpressure values in the residual heat removal system for protection, transient is er.tmated to result in $1.2 billion in cleanup Newer Westinghouse plants that allow either PORVs or and repair wts. In addition, replacement power costs RRR SRVs were placed in Group 2. The newer plants could occar during the cleanup and repair period, If the tend to have better reactor pressure vessel rnalerial acc%t also results in a large release of radioactivity properties and lower vessel failure probabilides due to the offsite, the costs of relocating people, restricting food and irradiation embrittlement. This grouping assumption will water, cleanup of contamination, and health consequences not have a significant impact on this evaluation, would add to these costs. Based on a 10-year period for the cleanup and repair of onsite damage, the present value The current technical specification for overpressure of these averted costs for 67 plants is estimated to be about protection allows one of the two channels to be out of

$5 million, based on a 5% continuous discount rate and a senice for 7 days with no restrictions on plant operations.

mean core damage frequency of 3.24x104 per reactor nis allows the low temperature overpressure protection year, ne present value of averted costs for offsite system to te degraded to a single channel system. The damage for 67 plants is estimated to be as high as 58 technical specification excludes the low temperature over.

million for a large release, discounted at 5% over the pressure protection system from being considered as a sys-remaining life of the plants included in this study. Thus, tem that performs a safety-related function.

An the convention of accounting for these averted costs in the operability check of a channel that reveals a failure before assessment of other alternatives should not obscure the a mode change does not (need not) prohibit proceeding possible costs associated with the *No Action" alternative, with less than the minimum number of channels in service (for LTOP, two channels are specified as the minimum).

5.1.2 Alternative 2. Change to Technical Specifications The proposed alternative would ensure that both channels ne benefit from implementing a proposed change to the are operable, when providing protection against brittle technical specification for overpressure protection would reactor pressure vessel failure while operating in Mode 5.

be a reduction in the frequency of core damage per reactor An inoperable channel needs to be returned to operable year due to a low temperature overpressure event. The status as soon as possible, and the NRC staff believes that risk reduction is based on improvements in low-it is not appropriate to continue with actions to retum to temperature overpressure protection systern availability, power operations with only one channel operable, as cur.

In addition to implementation costs associated with this rently allowed by the overpressure protection technical proposal, operational cost increases resulting from specification.

increased outage times are also wnsidered.

The reduction in core damage frequency expected would The industry implementation costs would be primarily be equivalent to the logical "and" of the estimated those incurred to revise the curant overpressure protection unavailability of the single channel system on which this technical specification and to modify the plant cooldown evaluation is based. For the Group 1 (PORVs) plants the and heatup procedures to reflect the revised technical reduction is estimated to be from 0.087 to 0.087'0.087, or specification ne NRC costs would be primarily those 0.0076. For the Group 2 (RHR SRVs) plants the reduction associated with the review and approval of the revised is estimated to be from 0.143 to 0.143*0.143, or 0.02. The technical specifications. These are discussed in the fol.

mean through wall crack frequency,per reactor y or core damage lowing paragraphs, frequency, is reduced from 3.24x10-3.47xtrM per reactor year. The core damage frequency 5.1.2.1 Risk Reduction Estimates reduction is 2.89x104per reactor year.

To estimate the change in the expected risk that the The actual benefit, in terms of LTOP protection proposed resolution could effect, both the postulated availability, may be less than assumed if undetected radioactive exposure (in person-rem) that would result in common cause failures are considered. Common cause the event of an accident and the reduction of the core failures, such as leakage past PORVs or air / nitrogen damage frequency must be estimated.

system failures, which are detectable, require immediate NUREG 1326 54

~

Value/ impact Analysis action to depressurize and vent the reactor coolant system per demand. If the common cause contribution is 0.0024 under the current technical specifications.

of the base case 0.1 per demand unavailability, then the resultant channel independent unavailability would te in the LER data base, there are 25 events that have been 0.0976 per demand (0.1 - 0.0024). The LTOP un-classified as teh L10P protection channels unavailable.

availability would then be 0.0024 for common cause Of these 25 cients,12 were related to IORY leakage or failures plus 0.0976'0.0976 for independent failures, or air / nitrogen problems. Of the remaining 13, four would be 0.012 per demand.

classified as common cause failures in two cases, proce.

dural error resu..cd in no LTOP protection for 14 knd 36 he estimation of risk (in person rem) is also dependent hours. In one case, the PORY block valves were found to on assumptions concerning containment integrity at the be closed for a period of about 5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months , in the second time of an accident. Since low temperature overpressure case, a blank Gange was installed in the vent line and events occur most frequently in Mode 5 at reactor coolant found about 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months later. No low temperature overpres-system temperature less than 20(FF, containment integrity sure transients occurred during these periods of time, is uncertain. No technical specifications for containment integrity exist below 200"F (except during refueling

%ere was one reported case where component failures operations when the reactor pressure vessel head has been were not detected by required surveillance the Millstone removed). Industry responses to NRC Generic Letter 87 3 event in January 1988. Unrecognized system interaction 12 (Ref. 21) indicate that containment integrity during between the solid state protection system and the overpres.

Mode 5 is often relaxed to allow for testing and sure mitigation systern (OMS) was not detected, rendering maintenance, and for repair of equipment (for example, both OMS channels inoperable for more than 2 months, containment penetrations, steam generators, and reactor he alternative relief paths via the RHR SRVs were avail-coolant pumps), nree risk estimates are employed to able until January, when one path was removed for address ct.ntainment integrity.

rnaintenance-related activities and the other relief path was lost as a result of unrelated activities, which inadvertently The best estimate vahie assumes that there is a 50% proba-caused isolation of the RHR system. The loss of the let-bility of the containment's being open at the time of the down flow path with continued charging resulted in an accident. The open containment consequence is based on LTOP transient that could not be mitigated by the LTOP the scaled source term evaluation presented in Appendix protection system. Prompt operator action limited the IL For the other 50%, the containment consequence is peak pressure to a value below the Appendix G limits. For based on the Siting Source Term evaluation for an SST2 this evaluation it is assumed that the common cause failure release (Severe core damage. Containment fails to isolate, existed for 1800 hours0.0208 days 0.5 hours 0.00298 weeks 6.849e-4 months (more than 2 months), even though Fission product release mitigating systems, e.g., sprays alternative relief paths were available during this period of and fan coolers, operate to tr duce release similar to time.

WASH 1400 PWR 5 release).

De average time plants spend in shutdown modes is about The high estimate value assumes that the containment is 30% per reactor year (110 days), including refueling open and the fission product release mitigating systems do outages (Ref. 20) with refueling outage times when the not operate. The low estimate value assumes that there is reactor is disassembled accounting for about 20 of the 110 a 10% probability of the containment's being open at the days. It is assurned that LTOP protectron is required 25%

time of the accident. The source terms are the same as the of a year per reactor year, about 2200 hours0.0255 days 0.611 hours 0.00364 weeks 8.371e-4 months in 340 reactor best estimate case, but weighted 10/90 instead of 50/50.

years (including accumuhted experiences through 1987 to properly account for Millstone 3), undetected common In addition to public health consequences, there are also cause failures resulted in LTOP protection unavailability occupational health consequerices associated with the acci-for roughly 1800 hours0.0208 days 0.5 hours 0.00298 weeks 6.849e-4 months during the 745,000 hou's when dent.

The short term and long term occupational LTOP protection is assumed to be needed. The probabil-exposures are taken from Section 3.3, Occupational ity of an undetected common cause failure resulting in llealth (Accidental),* of NUREG/CR 3568 (Ref. 22). The LTOP protection unavailability is therefore estimated to dose reduction estimate is based on an average reduction be 0.0024 per demand, as compared to the 0.1 per demand in core damage frequency of 2.89x104 per reactor year, (three failures in 30 total events) LTOP independent un.

based on the proposed resolution of GI 94 for 67 plahts availability used for this evaluation.

over an average remaining lifetime of 28 reactor years.

De L10P protection unavailability, including undetected The dose reduction estimates for the three cases are common cause failur:s, is estimated to be about 20%

provided in Table 5.1. The proposed resolution should not higher than used in this evaluation.

The assumed result in any additional operational exposure to plant per-improvernent in LTOP unavailability is from 0.1 to 0.01 sonnel.

55 NUREG-1326

Value/ Impact Analysis Table 5.1 Value/ impact summary for the proposed resolution of GI 94 (for 67 plants).

Dose Reduction (person rem)

Costs ($1,000s)

Best liigh Low Best liigh Low parameter Est.

Est.

Public Ilealth 14,500 26,700 4,700 Occupational Exposure 180 240 60 (accidental)

Occupational Exposure NA NA NA (routine)U)

Industry implementationG) 1,290 2,570 640 Industry Operational 0) 80 400 80 NRC Implementation 950 1,840 500 Value/ Impact RatioM) 160 180 260 (Sum ofindustry and NRCimplementation costs divided by the public dox reduction estimate 5/ averted person-rem)

Notes:

t (1) No significant routine exposure is anticipated as a result of the proposed resolution. L.ow temperature overpressure protection channel j

surveillance involves elecsrical circuit check without cantainment entry.

G) Costs associated with revisions to technical specifications and plant cmidown and heatup procedures.

Q) Present value of estimated replacement power costs resulting frum delayed startup caused tiy proposed resolution. Applicable to PORY plants only (40 of 67 units), see text.

(4) '!his decs not take into account the additional riegative impacts associated with avoided plant damage costs or replacement power costs resulting from reduced frequency of core damage (see text. Section 5.1.2.5).

5.1.2.2 Industry Implementrillon Cost Estimates would require immediate action to replace a failed channel to operable status before proceeding with planned startup The cost to the licensees to comply with the proposed operations.

requirements will vary depending on assumptions concern-t ing the level of effort necessary to revise the technical Since there are no additional surveillance requirements for specification for overpressure protection and to revise the the plants that rely on the residual heat removal system

)

plant cooldown and heatup procedures to reflect the safety relief valves, because these are passive devices, no change. In addition, there may be an extended outage additional surveillance checks for operability are resulting from the requirement to ensure both low-performed during shutdown. There will be no additional temperature overpressure protection channels operable delay in startup time and therefore no replacement power prior to startup. The cost would be associated with addi-costs for the RHR SRV plants, tional replacement power The replacement power costs r

would impact only plants using PORVs. The technical Appendix C provides the bases for the industry costs l

specification for PORV plants requires channel operability associated with the proposed recommendation for the checks every 31 days. The proposed recommendation resolution of GI-94.

s NUREO 1326 56

Value/ Impact Analysis ne industry implementation costs are provided in Table Appendix D provides the bases for the NRC costs 5.1. De costs are based on the level of effort required to associated with the proposed recommendation for de revise the technical specification and plant cooldown and resolution of GI 94.

heatup procedures. De best estimate costs are based on a level of effort associated with a simple change, $17,400 The NRC implementation cost estimates are calculated per plant for the technical specification changes and using the same assumptions as applied to the industry

$1.900 per plant for revisions to the plant procedures. The implementation costs. For the simple technical specifica-high estimate is based on a level of effort associated with a lion change, the best estimate cost is estimated at $14,200 complex change, $34,800 per plant for the technical per plant, and for the complex change the high estimate specification changes and $4,8(10 per plant for revisions to cost is $27,400 per plant. ne low estimate is based on the plant procedures. The low estimate assumes the level halving the simple, or routine, cost estimate. The NRC of effort is one half of that associated with the simple implementation costs estimates are provided in Table 5.1.

change.

5.1.2.4 Value/ Impact Summary -

industry operational costs are estimated based on the present value of additional replacement power resulting he value/ impact summary for the proposed resolution of j

from delayed startup as a result of the proposed resoiulion Generic Issue 94, " Additional Low temperature Overpres.

for GI 94 (see Appendix C) and we applicable only to sure Protection for IJght Water Reactors," is provided in plants that use PORVs for low temperature overpressure Table 5;1, The proposed resolution would impact all protection. Assuming a 4 hour4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months delay in startup (estimated Westinghouse and Combustion Engineering PWRs. Plants from Ref. 23) at $500,000 per day for replacement power, using either PORVs or the RHR SRVs can reduce the risk the estimated cost is $83,000, llowever, only a fraction of of britde reactor pressure vessel failure and reduce the startups will be delayed.

probability of exceeding Appendix G pressure / temperature limits by an order of magnitude by simply considering the It is assumed that diere are four nonrefueling shutdowns safety related role of these comporents, while operating at per reactor year per plant (Ref. 24). In most cases the low temperature, especially when water solid.

The shutdown mode will be exited prior to the need for repeat.

likelihood of a low temperature overpressure transient ing the surveillance. It is assumed that 5% of the time resulting in a peak pressure exceeding the Appendix G

.(once every 5 years) surveillance is required prior to re-pressure / temperature limits would be reduced from one-start. Further assuming that the probability of fixing the in ten to one in one hundred, the desired objective of 01 i

channel actually delays the startup 5% (one out of twenty)

94. The likelihood of brittle reactor pressure vessel frac-i of the time and that tic channel unavailability is 0.087 per ture (a through wall crack) is minimized.

demand, the frequency of delayed startup is estimated as:

The proposed resolution for Generic Issue 94 reduces the (4 shutdowns / year) x (0.05 delays) x (2 channels) mean core damage frequency to less than lx104 per x (0.087/ demand) x (0.05 repair delays) =

reactor year, from 3.24x104 o 3.5x107 per reactor year, t

1.74x10 3 delays per reactor year, meeting the target CDF objective stated in Section 2 For a plant that approaches the PTS screening criteria at the The average annual cost of a delayed startup, based on a end of license (with a CDF of 7x10-6per reactor year), the 4 hour4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months delay, is estimated to be (4 hr/24 hr) x target CDF objective would also be met (with the CDF

$500,000/ day x 1.74x10 3 per reactor year, or $145. At a reduced to 7x10-7 per reactor year),

discount rate of 5% the present value of the replacement power cost is $2,000 per PORY plant, over the average 5.1.2.5 Averted Damage Costs l

remaining lifetime of the PORY plants (24 years). At a 10% discount the present value is $1,400. If the need to

- Including the costs of averted plant damage, replacement return an inoperable channel to service occurs once per power, and offsite costs can significantly affect the overall reactor year, then the cost of replacement power would be cost benefit evaluation. In addition, the present value five times greater than assumed, or $10,000 per plant (at a associated with these factors can serve as a measure of the 5% discount rate),

worth of a proposed alternative, if two or more proposed alternatives could achieve similar risk reduction, but with 5.1.2.3 NRC Implementation Cost Estimates markedly different costs, then the present value estimates could be used to evaluate the relative worth of an alterna-The cost to the NRC to review and approve the revised

live, technical specification will vary depending on assump-tions concerning the level of effort necessary to perform The estimated present value costs for avoided plant the review, damage are summarized in Table 5.2. The cost for the 57 NUREG 1326

Value/ Impact Analysis cleanup and repatt of a plant following reactor pressure of accident avoidance could be estimated. From Appendix vessel fracture and fission product release is estimated to E, the estimated present value costs for offsite effects are be $1.2 billion (Refs. 22 and 25). The replacement power summarized in Table 5.3.

cost is based on the daily replacement cost typically used to estimate short-term power replacement costs. Should it

'the total present value of avoided damage is dependent on become necessary for a utility to provide replacement the release category assumption (containment integrity at power for a long period of time, for example,10 years as the time of the accident). Table 5.3 provides additional assumed herein, it is probable that alternative methods information that can be used to assist in evaluating the would be considered that may lower the actual costs. The benefits associated with any proposed ahemative for 01 detailed evaluation is provided in Appendix E.

94. To account for containment integrity assumptions, the SSTI and SST2 costs need to te properly weighted to

. If cost saving to the industry from accident avoidance develop the present value costs of avoided damage, it is (c!canup and repair of onsite damage and replacement noted that these cost estimates for offsite damage are prob-power) were to be considered, the overall valuchmpact ably conservative in that the fission product releases for a ratios would be improved significantly, At a 5% discount low ternperature overpressure event are lower than those rate, the present value of avoided damage would more associated with an SSTI release.

than offset the 52.32 million test estimate implementation cost.

The total avoided damage costs (both onsite and offsite) for the 67 plants are estimated to te 58.3 million for the If the add'xl savings to the industry from acddent best estimate case, $11.6 million for the high estimate t

avoidance resulting in reduced offsite health and property case, and $$.4 million for the low estimate case (assuming i

costs were to te considered, a measure of the total worth a 5% discount rate).

Table 5.2 Fatimated present value costs for svolded onsite damage (for 67 plants).

10% Discount 5% Discount Over 10 Years Over 10 Years Cleanup and Repair 51,200,000

$2,200,000 Replacement Power 51.300,000

$2,400,000 Total

$2,500,000 54,600,000 i

Table 5,3 Estimated present value costs for avoided offsite health and pers,onal property damage (for 67 plants).

Based on SST) Costs Based on SST2 Costs Over Plant Life Over plant Life 10% Discount 5% Discount 10% Discount 5% Discount Offsite Health

$ 640,000 5 970,000 S 23,000

$ 36,000 Offsite Property

$4,060,000 56,180,000

$ 63,000 S 86,(00 i

Total S4,700,000

$7,150,000

$ 86.000

$ 122,000 NUREG 1326 5-8

m s

Value/ Impact Analysis l'

5,1,3 Alternative 3. Si and RCP spectrum without these events. This is a scoping assump-Restriellons tion to maximize the risk reduction estimates. LTOP l

unavailability is assumed to be the same as for the base Like Alternative 2, Alternative 3 addresses risk reduction case. Bree of the PORY class mass addition events i

by evaluating changes to the technical specifications that would not have occurred, North Anna 1 in 1981 and in are intended to lower the frequency of occurrence of low.

1983, and San Onofre 1 in 1983 (see Table 4.1). he San temperature overpressure events. These revisions would Onofre 1 event would have eliminated one of the two 2500 apply to all 67 Westinghouse and Combustion Engineering psi cases in the PORY class of plants. Si lockout would plants. The technical specificadon changes here would have climinated one of the RHR SRV mass addition provide for additional administrative controls on the sys-events, the 450 psi event at Byron 1 in 1985 (see Table tems and components involved in causing a pressure tran.

4.2).

The frequency of low temperature overpressure sient while operating at low ~ temperatures, instead of events, based on the revised base case (the climination of increasing the availability of the mitigating system as events), would be 0.082 (20 cvents in 244 reactor years) proposed in Alternative 2.

for PORY plants with the Si proposal and 0.107 (six events in 56 reactor years) for the RHR SRV plants.

he first change would be to require that all high pressure safety injection (HPSI) pumps be locked out or have The risk reduction for not allowing RCP restart is power removed when operating in a water solid condition, estimated in a similar manner. For the PORY class of i

llowever, this alternative would also remove the SI path as plants, all six energy addition events would have been the normal means of supplying highly borated water for climinated. In addition, the two Turkey Point 4 events in reactivity control. A review of the Westinghouse and 1981 and one of the North Anna 1 (1984) mass addition Combustion Engineering standard technical specifications events would have been climinated (see Table 4.1). For indicates that boration requirements can te met with the RHR SRV class, all three of the energy addition events power removed to the SI pumps. Two additional boration would have been eliminated. %c frequency of low.

paths are identified for use in Modes 4 and 5.

One temperature overpressure events, based on the revised base contains the boric acid tank, boric acid transfer pumps, case (the elimination of events), would be 0.057 (14 events and a charging pump. The other contains the borated in 244 reactor years) for PORY plants with the RCP water stotage tank and a charging pump.

proposal and 0.071 (four events in 56 reactor years) for the RiiR SRV plants.

In addition, this alternative would not allow the restart of a reactor coolant pump while in a water. solid condition to Overall, if both proposed actions were to be implemented, provide additional risk reduction. Although there are the frequency of low temperature overpressure events is current restrictions in the standard technical specifications estimated to be reduced from 0.094 to 0.045 per reactor regarding restart of a reactor coolant pump, this alternative year for PORY plants and from 0.125 to 0.054 per reactor would be more stringent and a new technical specification year for RHR SRV plants. De frequency of events would would have to be developed for this requirement. Under remain high enough to still consider LTOP events as the current technical specifications, reactor coolant pump anticipated operational transients, at least once in the plant restart is allowed in a water solid condition provided the

lifetime, secondary side temperature is not hotter than the primary side by a specified value (as determined by analysis per.

The peak pressum spectrum would also change if these formed by each licensee),

proposed actions were implemented, as shown in Table 5.5.

5.1.3.1 Rbk Reduction Ihtimates

- The LTOP unavailability will not change from the base ne reduction in the mean core damage frequency case as a result of these proposed actions. The vessel associated with either an $1 or an RCP technical specifica.

fracture probability, at a given peak pressure, is a constant.

tion change, as well as the estimated change if both were ne risk reduction estimate in through wall crack

- to le implemented, are provided in Table 5 4 as Altema.

frequency for Si lockout is 1.07x104 per reactor year, tives 3(a),3(b), and 3(a) & 3(b), respectively, from 3.24x104 o 2.17x104 per reactor yearc The RCP t

restart proposal reduction is 2.12x10lper reactor year, The risk reduction for the proposed recommendation to from 3.24x104 o 3.03x104 per reactor year. The overall t

lock out all HPSI is estimated by evaluating the reduction risk reduction in through wall crack frequency is estimated in the frequency of LTOP events due to inadvertent safety to be 1.20x104 per reactor year, from 3.24x104 to injection and to determine the revised peak pressure 2Nx104 per reactor year.

59 NUREG 1326

Value/Itapact Analysis Table 5.4 Mean core damage frequency estimates for Alternatise 3.

Mean Core Damage Frequency (CDF)

CDF Before Alter Reduction Alternative 1/RY 1/RY 1/RY Ratio 3(a) 3.24x tp6 2.17x10 6 1.07x10-6 3,49 3(b) 3.24x106 3.03x t&6 2.12x10 7 1.07 3(a)&(b) 3.24x106 2.04x10-6 1.20x106 1.59 Ahernadve Desenpdons 3(a)

No S1 when w ster sobd, technical specificadon change, 3(b)

No RCP restan when water sohd, technical specificanon change.

3(a)&(b)

Both Si and RCP requirements, technical specificanon changes.

Table 5.5 Peak pressure spectrum summary for Alternatige 3.

Pressure Base Case SI Lock Out RCp Restart Si and RCP Plant Group (psi)

Fraction Fraction Fraction Fraction PORY Plants 2500 0.09 0.05 0.14 0.09 1400 0.09 0.10 0.

O.

850 0.13 0.15 0.21 0.18 600 0.69 0.70 0.65 0.73 RHR SRV Plants 2500 0.14 0.17 0.25 0.33 850 0.14 0.17 0.25 0.33 600 0.72 0.66 0.50 0.34 5.1.3.2 Industry implementation Cost Estimates

%e best estimate costs for the RCp restart technical specification change and plant procedure revisions are

%c industry implementation cost estimates, with high and based on the routine change cost estimates, $17,400 and low values, are provided in Tsble 5.6. De best estimate

$1,900 per plant, respectively, for a total cost of $19,300 costs for the Si technical specification change and plant per plant. The high utimate costs are based on the procedure revisions are based on the complex change cost complex costs estimates of $34,800 and $4,800, for a total estimates, $34,800 and $7,500 per plant, respectively, in cost of $39,600 per plant. The low estimate costs are addition, training costs are expected for the new Si based on halving the simple, or routine, change cost requirement to ensure that boration requirements will be estimates. The costs are 58,700 and $900, or $9,600 total met. The traininF cost is estimated to be $11,800 per per plant.

plant, fcrr a total cost of $54,100 per plant. De high estimate cost is based on doubling the cost to account for the additional costs associated with meeting boration The industry implementation cost for both actions is a requirements,$108,200 per plant. The low estimate cost straight forward addition of the individual costs. The best is based on the simple, or routine, change cost estimates, estimate cost is therefore $73,400 per plant. De high without the need for additional training. The costs are estimate and low estimate costa are $148,000 and $28,900

$17,400 and $1,900, or $19,300 total per plant.

per plant, respectively.

NUREG 1326 5-10

i Value/ impact Analysis Table 5.6 Implementatkm co6t estimates for Alternative 3.

t Unit Costs Total Costs 3

Alternative Plants Best High law Best High Low Costitem Affected Est.

Est.

Est,

($)

($1,000s)

($1,000s) 3(a)

Industry Tech Spec All -

34,800 69,800 17,400 Procedures All 7,500 15,000 1,900 Training All 11,800 23,600 0

Total 3,630 7,250 1,290 NRC Tech Spec All 27,400 54,800 14,200 1,840 3,670 950 Total 5,470 10,920 2,240 3(b)

Industry 1

Tech Spec All 17,400 34,800 8,700 i

Procedures All 1,900 4,800 900 Total 1,290 2,650 650 NRC Tech Spec All 14,200 27,400 7,100 950 1,840 t

Total 2,240 4,490 1,130 3(a)&(b)

IndustryTotal All 73,400 148,000 28,900 4,920 9,910 1,940 NRC Total All 41,600 82,200 21,300 2,790 5,510 1,430 Total All 115,000 230,200 50,200 7,710 15,420 3,370 r

Alternadve descriptions. See footnote to Table 5.4.

Plants affected: All. Applicable to 67 units. PORVs. Appticable to 40 units. Ri!R SRVs. Apphcable to 27 uniu.

5.1.3.3 NRC Implementation Cost Estimates a complex change, $27,400 per plant. The high and low estimates are $54,800 (double the complex cost) and The NRC implementation costs are based on the same

$14,200(routine change) per plantorespectively.

assumptions regarding whether the change to the technical l

specifications are judged to be complex or routine and are The NRC best estimate implementation cost for the RCP provided in Table 5.6. Assumptions used to estimate the restart technical specification change is $14,200 (routine) high and low cost estimates are consistent with those used per plant. 'Ihe high estimate is based on the complex cost to estimate the industry implementation costs. For the Si estimate, $27,400 per plant. The low estimate is based on change the NRC best estimate implementation is based on halving the routine cost, or $7,100 per plant.

5 11 NUREG 1326

o s

Value/ Impact Analysis he NRC impicmentation cost for toth actions is a based on the overall risk reduction and costs for toth straightforward addition of the individual costs. De best actions.

estimate cost is therefore 541,600 per plant. De high estimate and low estimate costs are $82,200 and $21,300 Alternative 3 is not recornmended because of the lower es-per plant, respectively, timated risk reduction and the high costs associated with these changes when compared to the proposed resolution, Alternative 2.

This alternative does not improve the 5.1.3.4 ' Value/ impact Summary LTOP system availability, nor does it appear to reduce the overall event frequency to a valt.e that could be considered De value/impacemmary for Alternative 3 is provided in low enough to exclude these transients fiom consideration c

Table 5.7, for each oPhe three containment assumptions, as anticipated operational transients.

Table 5.7 Valuellmpact summary for Alternative 3 (for 67 plants).

TWC Reduction Industry + NRC Person Rem Value/imptct Ratio let year Cost Averted

($/ Person Rcm)

Ilest Er.timate 1.20x104

$ 7.71 million 8,400 920 High Estimate 1.20x104

$15.42 million 15,600 990 Low Estimate 1.20x104

$ 3.37 million 2,800 1,200 5.1.4 Alternative 4. Removal of RHR While inadvertent isolation of the RilR can result in a Autoclosure Interlock low temperature overpressure event, the operating reactor experictces from 1980 through 1986 do not indicate that This alternative explored risk reduction from low-spurious ACI is a significant contributor to LTOP risk.

temperature overpressre events if the autoclosure inter.

%cre were no L' LOP transients (events during shutdown lock (ACl) on the residual heat removal suction line isola-operations that challenged the LTOP systems) directly re.

j tion valves is removed. It is expected that the frequency lated to spurious ACI of the RilR suction line valves.

1 of low temperature overpressure events will be reduced if Prior to the implementation of USI A 26 requirements, the ACI is removed tecause spurious closure of the about half of the low temperature 0,erpressure transients suction line isolation valves resulting in a loss of letdown occurred with the RIIR isolated. A large fraction of these will not occur. The base case risk analysis credits limiting were attributed to ACI actuation. The NRC staff twlieves the peak pressure of a low temperature overpressure event that the requirements of USl A.26 and the awareness of

- to the residual heat removal system safety relief valve LTOP concerns, combined with the slow closing time of setpoint if the residual heat removal system were fune -

the REIR suction line isolation valves (on the order of 2 tional and the ACI setpoint is above the safety relief valve minutes), have provided sufficient guidance and time to setpoint.

plant operators to mitigate these events before they chal-

_j i

lenge the PORVs.

An informal survey conducted by PNL of plants evaluated i

in this study found that 17 out of 26 plants surveyed would The risk reduction is therefore assumed to be related to the not be affected (the RHR SRV setpoint is actually below greater availability of the RFIR SRVs in the PORY plant the ACI setpoint). About 14 plants (out of the 40 PORY group to limit the peak pressure of a low temperature plants) would actually benefit from this alternative. Plants overpressure event if the PORVs fail to mitigate the tran-that rely on the residual heat removal system safety relief sient.

valves do not benefit from this alternative because the ACI feature is not permitted in the design. To maintain a 5.1.4.1 Risk Reduction Estimates common basis for evaluating alternatives, the NRC staff assumes that this alternative will be beneficial.to the

%is alternative would not reduce either the LTOP event PORY plant group, and the value/ impact evaluation is frequency or the unavailability of the PORV to mitigate based on these 40 PORY plants, toth in overall risk reduc.

Iow-temperature overpressure transients. Risk reduction i

tion and in implementation costs.

would be obtained by reducing the frequency of achieving l

NUREG 1326 5-12 1

l

\\

Value/ Impact Analysis high pressure from the base case value to 600 psi, the typi-In addition to the actual removal of the ACl, there are cal RHR SRV setpoint pressure.

additional costs related to technical specification chango and revised plant procedures. The best estimate values aro For the PORV class of plant,if the ACI feature had been based on a routine change, $17,400 for the tecnnical removed, three of the 23 operaung reactor events would specification change and $1,900 for changes to plant have resulted in lower peak pressures. The Ginna Si event operating and maintenance procedures, for a total of (1983) and the two Turkey Point 4 ewnts (1981) would

$19,300 per plant. The high cost estimate assumes that the have been limited tc 600 psi, the SRV setpoint. In the technical specification change is complex, $34,800 per base case,15 events were already ass imed to be limited to plant, and that the cost to revise procedures remains the the RHR SRV setpoint, three occurred when the RHR was same, for a total cost of $36,700 per plant. He low cost already isolated, one involved inadequate capacity to estimate is based on halving the routine cost estimate, relieve the mass flow sate, and one was a result of an 58,700 per plant.

ehetrical upset as opposed to un ACI actuation. The peak pressure spectrum that wouM result from this altemative is In addition to the costs associated with implementation of shown in Table 5.8.

this alternative, there may be additional costs to both the industry and the NRC that would reduce the overall he reduction in the through wall crack frequency for the benefit. Additional cos? would include plant specific PORY plants is estimated to be 1.60x10 7 per reactor year, studies to demonstrate that ucrall plant safety would not from the base case value of 3Nx104 per reactor year to be adversely impacted, and any additional features to 2.88x10 per reactor year. This allemative does not result protect against the inadvertent overpressurization of the 4

in a significant reductior. because the probability of high residual heat removal system (an interfacing loss.of-pressure events, combined with RHR isolation not related coohmt accident, Event V),

to ACl,is unchanged.

5.1,4,2 Industry implementation Cost Estimates 5.1.4.3 NRC Implementation Costs Estimates The industry implementation cost estimates, with high and The NRC implementation cost estimates are provided in low values, are provided i.n Table 5.9. Recently completed Table 5.9. Based on recently completed work on Genetic work on Generic issue 99 indicates that the cost for issue 99, the NRC cost associated with the review of the removal of the autoclosure interlock feature, including thc ACI removal is $2,000 per plant. The technical specifica-costs for a plant specific analysis, cable disconnecting, in-tion change cost is based on a routine change, $14,200 per terlock iugi: reprogramming, and radiation exposure, plant. The high estimate is based on a complex change ranges from 0100,000 to $150,000 per plant (Ref. 26).

cost, $27,400 per plant, and the low estimate is one half Rese values are used for this evaluation.

of the routine cost, $7,100 per plant.

Table 5.8 Peak pressure spectrum summary for Alternative 4.

Pressure Base Case ACIRemoval Plant Group (psi)

Fraction Fraction PORV Plants 2500 0.09 0.09 1400 0.09 0.0 850 0.13 0.09 600 0.69 0.82 RHR SRV Plants Not affected by Alternative 4 l

5 13 NUREG 1326

Value/ impact Analysis Table 5.9 Implementation cost estimates for Alternative 4 (for 40 PORY plants).

Unit Cost Total Costs Plants Best High Low Best High IAw Costitem Affected Est.

Est.

(S)

(5)

($1,000s)

Industry ACI removal PORVs 100,000 150,000 100,000 Total 4,000 6,000 4,000 Tech Spec PORVs 17,400 34,800 8,700 Procedures PORVs 950 950 450 Main. Proc PORVs 950 950 450 Total 770 1,470 380 NRC ACI removal PORVs 2,000 2,000 2,000 Total PORVs 80 80 80 Tech Spec PORVs 14,200 27,400 7,100 Total 570 1,100 280 Total Without ACI RemovalCosts 1,340 2,570 660 Total With ACI Removal Costs 5,420 8,650 4,740 5.1.4.4 Value/ Impact Summary SRVs. In addition, a plant specific analysis would be required to ensure that the overall plant safety would not he value/ impact summary for Alternative 4 is provided in be adversely impacted.

Table 5.10, for each of the three containment assumptions, based on the overall risk reduction. Two cost studies are The risk reduction estimate for this alternative is small, provided. In the first case, the costs are based on technical because the base case evaluation already credits the RHR specification and plant procedure changes only. This SRVs in limiting the peak pressure if the PORVs fail to would be equivalent to increasing the current ACI setpoint mitigate an LTOP transient. De resultant value/ impact to a value higher than the RHR SRV setpoint for addi-ratios are high, in excess of the $1,000 per averted tional LTOP protection (provided of course that the SRV person-rem guideline, could be shown to be adequately sized to prevent over-pressurization of the RHR). In the second case, the coss include the actual removal of the ACl, assuming no addi.

Alternative 4 is act recommended as a requirement to tional benefit for non-LTOP-related reduction in person-resolve GI 94, although in conjunction with GI-99, rem.

"RCS/RHR Suction Line Interlocks in PWRs," the removal of the autoclosure interlock appears to be benefi.

While a reduction in risk can be achieved for this alterna-cial. The low temperature overpressure event frequency tive for low temperature overpressure concems, it is not can be reduced (resulting from spurious closure of the known if the SRV setpoint can be modified to ensure that RHR suction line isolation valve and loss of letdown),

the requirements of Appendix G, the pressure / temperature De NRC staff believes that, although ACI removal is limits, are met over plant life, while providing margins to beneficial to 0194 concerns, a determination that overall ensure that the net pump suction head of the residual heat plant safety is not adversely impacted by its removal removal pumps is maintained following cycling of the would have to be further addressed.

l NUREG 1326 5 14

__n___

i-,

\\c i; '

Value/ Impact Analysis Table 5.10 Value/ impact summary for Alternative 4 (for 40 PORY plants).

f TWC Reduction Industry + NRC Person-Rem Value/ Impact Ratio

/R year:

Cost Averted (S/ Person-Rem) 1 1

Without ACI Removal Costs Best Estimate 1.60x10-7 S 1.34 million 700 1,900 High Estimate 1.60x10-7 5 2.57 million 1,300 2,000 Low Estimate-1.60x10"7 5 0.66 million 250 2,600

- With ACI Removal Costs Best Estimate 1.60x10-7 5 5.42 million 700 7,750-High Estimate 1.60x10-7

$ 8.65 million 1,300 6,650 a

Low Estimate 1.60x10-7 5 4.74 million 250 19,000 5.1.5 -

Alternative 5 Safety Grade LTOP gained by this alternative, a comparative evaluatio'n was System performed by PNL using the Sequoyah system as a model J

for a fully safety grade system (Sequoyah 1 and 2 are

-)

Alternativc 5 evaluated risk reduc 6. ased on requiring operating plants with a fully safety grade system the low pressure overpressure protection systems to be installed). In addition,Linformal interviews with vendor a

- upgraded to a fully safety grade system. The following personnel and NRC inspectors were performed by PNL '

changes wereidentified:

and used to gain additional insights for this alternative.

Industry studies (Ref. 27) and NRC studies (Ref. 28) were i

1.

Upgrade components to meet applicable enviionmen-also reviewed.

j 1

tal qualification _ design criteria required for safety-

. related equipment. (Regulatory Guide 1.26, " Quality It was concluded by PNL that upgrading the low.

Group Classification and Standards for Water,

temperature overpressure protection system to L safety-o L

Steam, and Radioactive Waste-Containing Com-grade status is not expected to result in significant changes

[

ponents of Nuclear Power Plants.")

to the PORV or RHR SRV hardware design or function -

I ing. Environmentally qualified valves are expected to be 2.

Upgrade actuation circuitry to meet redundancy and identical to existing valves. Little additional credit can be m

electrical separation criteria by NRC regulations and given for upgrades in actuation circuitry that provide j

IEEE standards. This does not apply to the passive, improved redundancy and -electrical separation. The

- spring. loaded RHR SRVs. (IEEE Standard 279, potential benefit associated with this alternative would be endorsed by Regulatory Guide 1.153, " Criteria for achieved through better surveillance requirements.

j I

Power, Instrumentation, and Control Portions of Safety Systems.")

5.1.5,1 Risk R@S Estimates 3.

Upgrade maintenance and surveillance testing As previously stated, from a risk sendpoint, the low.

{

- activities on valves and actuation circuitry such that temperature overpressure protection glem can bc

,j documentation, schedules, and testing methods considered to be a one-channel system since tnc addd satisfy the requirements set forth for safety grade specification allows one channel out of. service equipment. (ASME Boiler and Pressure Vessel continuously for 7 days at a time. This time is long com-

]

Code,Section XI, subsection IWV, " Inservice Test.

pared to that required to heat up the plant and exit tir j

ing of Valves in Nuclear Power Plants.")

low temperature overpressure protection operating move.

Surveillance and system startup activities during hee.up 1

To gain insight into the potential improvement in low-make this the most likely time for an upset to challenge the temperature overpressure protection system availability low temperature overpressure protection system.- All three 5-15 NUREG 1326

-)

i

i Value/ Impact Analysis of the actual overpressurization events, where the peak val is assumed to be 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months . A review of LER data pressure exceeded the Appendix 0 limit,in the data base between 1980 and 1986 identified 25 instances (at 12 resulted from a single channel failure when the other, plants) in 187 reactor years when both PORVs were redundant channel was removed from service for main-declared inoperable for low temperature overpressure tenance during startup, protection. IIit is assumed that low temperature overpres-sure protection is required 10% of the time (one third of The Millstone 3 event of January 19, 1988, peripherally the 30% per reactor year spent in shutdown modes), then supports the conclusion that safety grade low temperature the estimated failure rate, L, would be:

overpressure protection circuitry may provide minimal reduction in unavailability. The NRC Special Inspection L = 25 failures /187 reactor year Report 50-423/88 03 addressing this event states: "The

/ 867 hours0.01 days 0.241 hours 0.00143 weeks 3.298935e-4 months per year. or COPS (Cold Overpressure Protection System, the Millstone 3 OMS) is safety grade with redundant power L = 1.52x 104 failures per hour, supplier.

an channels and equipment trains."

Neverthe @ channels were disabled more than 2 The expected unavailability is therefore 0.063, which months befus w event when the solid state protection appears to be in good agreement with the 0.087 value used system was tag;d out, and this commen mode failure was in this evaluation, as derived from the LER LTOP data not detected until the event was analyzed in this case,

base, what was thought to be redundant and separate not only failed to prevent single channel failure but failed to in addition to the 25 times both channels were declared prevent simultaneous failure of both channels caused by inoperable, there were an additional 18 reports when one unrecognized system interactions.

channel was declared inoperable, excluding the SRV plants. For the accumulated 244 reactor years of PORY A study performed by Oak Ridge National Laboratory experience, the resultant failure rate would be 2.0x104 (Ref. 28) stated that "an assessment of the need to upgrade failures per hour, for an unavailability of 0.079.

PORVs and BVs (block valves) to safety related status concludes that such action would slightly improve PORY The optimum test interval may be obtained from the and BV reliability." A study performed by EPRI(Ref. 27) unavailability equation for Q aad is found to be 7 days.

provided the results of a reliability assessment, based on The corresponding unavailability would be reduced to fault tree analysis and operating reactor failure rate data.

0.025,if the test interval was reduced from 31 to 7 days.

No credit was taken for PORY reliability improvement However, other factors could affect this improvement, that might result from qualification of the PORVs. This such as additional maintenance errors or errors introduced also indicated that little improvement in availability would during testing.

be expected from an upgrade to safety grade.

PNL concluded that a reduction in the PORV LTOP Based on discussions with vendor and industry personnel unavailability may be obtained by more frequent surveil-and based on available literature conceming PORVs, the lance. A more thorough :malysis would be required to reduction in PORV unavailability due to hardware and cir-substantiate such an improvement, roughly a 60% reduc-cuitry upgrades is estimated by PNL to be less than 20%.

tion in unavailability (from 0.063 to 0.025 ) based on this study.

Upgrading the system to safety grade may lead to im-proved maintenance and surveillance attention for the The hardware and surveillance improvement; cannot PORV based group of plants, The unavailability of the simply be added. The hardware upgrades would reduce PORV as a function of the surveillance test interval, the failure rate. If it is assumed that the failure rate is failure rate, test duration, and mean time to repair is given reduced by 20%, from 1.52x104 failure per hour to

- as:

1.22x104 failure per hour, to account for improved Q = L x T/2 + L x T(r) + T(t)/T hardware, then the resulting unavailability would be 0.023, as compared tc d.025, where:

Q is the PORV unavailability L

is the failure rate, failures / hour T

is the test interval, hours This alternative would not reduce the frequency of LTOP T(r) is the mean repair time, hours events or result in any changes to the base case peak pres-T(t) is the test interval, hours.

sure spectrum. The peak pressure spectrum is calculated assuming failure of the LTOP system. The LTOP system The present test interval is 31 days, or 774 hours0.00896 days 0.215 hours 0.00128 weeks 2.94507e-4 months . The unavailability would be reduced from the base case value mean repair time is estimated at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , and the test inter-of 0.087 to 0.035, a 60% reduction. The mean core NUREG 1326 5-16 x -- - - - -

Value/ Impact Analysis darnage frequency estimates are provided in Table 5.11.

channels being available.

If the PNL estimate is Since there are uncertainties associated with the estimated considered to be a measure of improved LORY reduc 60n in PORY unavailabihty resulting from a require-availability based on the single channel assumption, then ment to upgrade to a safety grade system, a sensidvity Case 5(a) could also be considered as addmg the require-evaluation is provided, as Case 5(a), where the reduction ments of Alternahve 2 to this alternative. Ir, either case, in unavailabihty is estimated to be from 0.087 (the single the risk reducuon for Case 5(a)is nearly 100% of the base channel unavailability) to 0.035*0.035 to account for toth case PORY contribution.

Table 5,11 Mean core damage frequene) estimates for Alternative 5 (PORY plants only).

Mean Core Damage Frequency (CDF)

CDF Belore Alter Reduction Alternative 1/RY 1/RY 1/RY Rauo 5

3.h 10 6 1.22x 10 6 1.82x 10-6 2.5 5(a)

3. % 10 6 427 x 10-l' 3.00x 10-6 71.2 5,1,5.2 f ndustry implementation Cost Estimates of the ume (four umes over a 5 year penod), and that 20%

of the time the delays would actually result in a longer The cost for upgradmg the low temperature overpressure shutdown ume, that is, returning a channel to operability is protection system to a fully safety-grade system includes the only action that needs to be completed prior to restart, one time costs for technical specification revisions, proce-The failure rate of a ch nnel is assumed to be 0.087, the dure revision, environmental qualitication, PORV actua-same unavailability as the base case. Assuming the tion design, hardware replacement, valve installation, average delay is 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months , at a cost of 5500,000 per day, PORV actuation circuitry installadon (two per PORV),

then the annual average cost for replacement power is cal-additional ASME testing (ANSI /ASME), recurring analog culated as:

channel tests, and potential replacement power costs.

(4 shutdowns /yr) * (0.20 delays)

For this alternative it is assumed that there is no benefit to (4 channels) * (0.087 failure / demand)

upgrading the RHR SRV plants to safety grade. The NRC (0.20 repair delays) * (4 hr/24 hr)

staff believes that, since the RHR system is designed to (5500,000/ day)

ASME Section Ill, Class 2. Seismic Category I require-ments, no improvements in LTOP unavailability would bc

= S4,600 per plant per reactor year.

achieved. However, if it was found that on a plant-specific basis this alternative would result in an im-The present value of the replacement power, over the provement, or if it was found to be necessary to replace average 24-year remaining lifetime of the PORV plants,is the RHR SRV, there would be costs associated with this estimated to be 560,000 per plant at a 59 discount, or alternative. These cost.s are identified but not applied to 542,000 per plant at a 10% discount. The best estimate the this value/ impact evaluauon for PORV plants only, cost to upgrade the PORVs to safety grade is $387,000 per plant, or $16 million for the 40 PORV plants. The high Using the guidelines of NUREG/CR-4627 (Ref. 29),

cost and low cost estimates are obtained by doubling and supplemented by PNL discussions with licensees and ven-halving the best estimate cost, respectively, for va!ues of dors, the industry implementation costs per umt are es-532 million and $8 million. With respect to the upper timated in Table 5.12.

bound, as the necessary labor effort increases, considera-tion must be given to the impact of radiauon exposure As a result of increased surveillance requirements, it is (stay time) on resultant labor costs. Stay dme has been possible that delays in restart could occur for the PORY estimated to impact labor cost by a factor of three for the plants resulung in additional replacement power costs. In moderate radiation zones involved in this altemauve. This most cases restart will occur before repeating the surveil-is a result of delays associated with dressing out, ALARA lance test. Four forced shutdowns per reactor year are factors. interfacing with health physics, etc. The best assumed. It is also assumed that delays would occur 20%

esumate cost did not involve stay time because the hours, 5-17 NUREG 1326

Value/ Impact Analysis and hence the exposure, were judged by PNL to be on NUREG/CR-4627 Abstract 4.1, " Typical System-amenable to normal shift work practices. If the stay time Average Dose Rates." The per plant occupational dose is becomes an important factor in completing the work, labor summarized in Table 5.13. The RHR SRV values are

ost would increase by a factor of six instead of merely provided for reference only, the value/ impact evaluation doubling.

for this alternative addresses the PORY plants only, in addition to the financial costs associated with the

'Ihe total occupational exposure for the 40 PORV planti, is safety grade upgrade, there are also additional occupa-estimated to be 900 person rems for the installation of the tional exposures that would arise as a result of installation safety grade equipment and sdditional testing that would and new test requirements. The dose estimates are based be required for this alternative.

Table 5.12 Best estimate unit cost to upgrade OMS to be safety grade.

Per Plant Costs S/ Plant,1988 Item PORV RHR SRV Technical Specification Revision 17,400 17,400 Routine, NUREG/CR 4627,2.2.1,6.4 Maintenance Procedure Revision 900 900 Routine,NUREG/CR 4627,2.2.2,6.4 Equipment Environmental Qualification 140,000 90,000 PORV Analog Circuit Design 15,000 n/a Valve Hardware Package 60,000 20,000-Valve Installation 5,400 4,700 NUREG/CR-4627, 6.1,2.1.5.2.1.6,2.1.7,6.4,6.5 Analog Circuit Installation, PORV only cost of installing two circuits 98,000 n/a

- NUREG/CR-4627,6.1,2.1.5,2.1.6,2.1.7,6.4,6.5 Additional ASME Valve Test, cost per test Tests commence 5 years after installation 1,600 1,500 NUREG/CR 4627,6.1,2.1.5,2.1.6,2.1.7,6.4,6.5 Additional Analog Circuit Tests, PORV only, cost per year 3,200 n/a NUREG/CR-4627,6.3 and 6.4 Replacement power cost 5% discount, PORV only 60,000 n/a Total per Unit Cost 401,500 133,000 Note: 1he RIIR SRV costs are provided for reference only. The value/ impact evaluation for this alternative is based on risk reduction and costs for the PORV plants only.

NUREG-1326 5-18

-_.w..

M 4.,A-,

Value/ Impact Analysis 5.1.5.3 NRC Implementation Cost Estimates based on the overall risk reduction. Two cost benefit results are provided. in the first case, the risk reduction The best estimate NRC implementation cost is based on and the costs are based on LTOP upgrade to safety grade, the review and approval of a routine technical specifica-In the second case, the risk reduction can be considered to tion change,514,200 per plant. The high estimate is based include Alternative 2, the change to the overpressure on a complex change, $27,400 per plant, and the low protection system led.aical specification to ensure that estimate is obtained by halving the routine cost,57,100 per both channeis are operable when water. solid.

plant. The industry and NRC implementation costs are summarized in Table 5.14.

Because of the high costs associated with this alternative and because the decrease in risk is predominately a result 5.1.5.4 Value/ Impact Summary of considerations under Attemative 2, the NRC staff does not recommend this attemative. Even if the risk reduction The value/ impact summary for Alternative 5 is provided in were 100%, the value/ impact evaluation exceeds the Table 5.15, for each of the three containment assumptions, 51,000 per averted person-rem guideline.

Table 5.13 Per plant occupational dose for safety-grade ONIS upgrade.

Per Plant Occupational Dose (person-rem per plant)

Action item PORV RilR SRV Valve Installation, two talves 4.2 2.4 Analog ChannelInstallation, PORV only, two channels 13.3 n/a ANSI /ASME Valve Tests,3 tests over plant life,2 valves per plant 4.8 3.3 Total per plant 22.3 5.7 5-19 NUREG-1326

Value/ Impact Analysis

- 1 Table 5.14 Implementction cost estimates for Alternative 5 (for 40 PORY plants).

Unit Costs Total Costs Plants Best High Lew Best High Low Costitem Affected Est.

Est.

($)

(S)

($)

($1,000s)

Industry Total PORVs 401,500 803,000 200,800 Total RiiR SRVS 133,000 265,000 66,000 d')

Total PORVs 16,000 32,000 8,000 NRC Tech Spec PORVs 14,200 27,400 7,100 570 1,100 280 Total 16,570 33,100 8,280 Nae:

Altemative 5 is assumed to be aplicable to PORY plants only. In general, the RilR SRVs are ASME Section III, Class 2, Seismic Categwy I valves.11enefits from safety-grade considerations are expected to be minimal. Ilowever, should it be found that a plant-specific upgrade would be beneficial, then the unit cost summary provided for the RilR SRV plants can be used to estimate that cost ofimplementatimi to the

- industry. %e NRC unit cost will be the same Ia any RitR SRV plant.

Table 5.15 Value/ impact summary for Alternative 5 (for 40 PORV plants).

TWC Reduction Industry + NRC Person Rem Value/ Impact Ratio

/R-year Cost Averted

($/ Person-Rem)

OMS Upgrade to Safety Grade Best Estimate 1.82x106 516.57 million 8,200 2,000 liigh Estimate 1.82xig6 S33.10 million 15,000 2,200 L.ow Estimate 1.82x196

$ 8.28 million 2,700 3,100 OMS Upgrade Sensitivity Best Estimate 3.00xig6

$16.57 million 13,400 1,200 liigh Estimate 3.00xt&6

$33.10 million 24,600 1,350 tow Estimate 3.00x10-6

$ 8.28 million 4,500 1,850 NUREG-1326 5-20

Value/ impact Analysis 5.1.6 Alternative 6 Pressurizer Hubble plants has sufficient capacity to meet the stated needs.

Modification to the plant nitrogen delivery system would This alternative explored risk reduction from low-be required to route the nitrogen to the pressurizer, temperature overpressure events by evaluating a steam or nitrogen bubble in the pressurizer (e.g., no water solid PNL also assumed that the current waste gas system has operations in Westinghouse and Combustion Engineering adequate capacity to handle the stated needs. It is also plants). The risk reduction is achieved by providing more assumed thst no other plant modifications are needed to time for the operator to respond to a low temperature allow venting of the nitrogen from the pressurizer to the overpressure event. Based on the historical operator waste gas system, in current designs, the venting of the r

performance discussed previously, the best estimate for pressurizer is to the pressurizer quench tank, which operator action is assumed to be 3 minutes to recognize operates under a low pressure nitrogen blanket and is and mitigate an event. With a sufficiently sized bubble, already vented to the waste gas system.

the peak theoretical pressure would be limited to 600 psi.

If 10 minutes is assumed, then the peak pressure could PNL also assumed that operations can be conducted in reach the primary system safety relief valve setpoint value such a manner that no additional modifications to the heat of 2500 psi with the same frequency, about 10% of the removal system for the quench tank are required. Two time, as the base case analysis previously described. Even options appear reasonable for purging the nitrogen bubble with a pressurizer bubble, the mass addition events that at low temperatures that avoid overheating the quench can occur at Westinghouse and Combustion Engineering tank. One would require taking the pressurizer water solid plants (inadvertent Si and charging without letdown) are briefly, as the pressurizer heaters are energized to heat the -

sufficient to result in 2500 psi assuming the L10P system pressurizer to saturation conditions and produce a steam fails to mitigate the transient and the operator fails to bubble. The PORY would be opened to vent the pres.

respond in 3 minutes, surizer and charging flow would be increased so that nitrogen is pushed out as water fills the pressurizer. The Under this alternative _it is asa d that the current low-second is the procedure used at B&W plants, whereby temperature overpressure protection system would not be water in the pressurizer is heated with the nitrogen bubble modified. Redundant channels are still assumed. The in place, while primary pressure is maintained between 50 fiequency of events and consideration of the inadvertent and 100 psi. Once saturation is achieved, the pressurizer -

high pressure safety -injection and charging transients is vented by cycling the PORY while maintaining pressure indicate that operator actions cannot-be relied on to between 50 and 100 psi. Complete venting of the nitrogen mitigate > the low temperature overpressure events at bubble is indicated when the pressurizer pressure remains Westinghouse and Combustion Engineering plants; that is, constant when the PORV is opened, there may be less than 10 minutes available for the operator to diagnose and mitigate a low-temperature over.

An alternative mode of operation would be to heat the pressure event before the pressure would exceed the pressurizer to saturation after the reactor coolant system Appendix G limit.

pressure was raised to several hundred psi. The increased density of the vented steam would require modification to This ahernative would require the ability to deliver large the quench tank heat removal system to accommodate the quantities of high-pressure nitrogen (800 cubic feet at up added heat load. Though not considered for this evalua-to 500 to 600 psi, or about 10,000 standard cubic feet) to tion, the total estimated cost of the modified heat removal the pressurizer prior to plant cooldown and subsequently system would probably exceed one million dollars per vent the nitrogen and process it through the waste gas sys-

unit, tem on plant heatup. The high pressure is required to ensure adequate reactor coolant pump seal integrity prior 5.1.6.1 Risk Reduction Estimates to restart of an idle pump.

This alternative would not change the frequency of LTOP System requirements were developed by PNL based on a initiating events or reduce the unavailability of the LTOP comparison to the Rancho Seco, plant which is representa-system, as compared to'the base case. Risk reduction live of the B&W system for providing a nitrogen bubble, would be achieved by limiting the peak pressure to 600 psi PNL reviewed final safety analysis report data from or less if a 3 minute operator action time is assumed.

Sheron Harris, Byron, SNUPPS, Seabrook, and the CE There is a possibility that the peak pressure may still reach CESSAR 80 plan:s and compared the system descriptions 2500 psi if it is assumed that the operator fails to respond to the Rancho Seco information, in 3 minutes. It is possible to reach 2500 psi pressures in less than 10 minutes. The peak pressure spectrum for PNL assumed that the existing nitrogen delivery system Alternative 6 is provided in Table 5.16, for each assump-(for safety injection tank overpressure fill) at affected tion regarding operator response time.

5-21 NUREG 1326

Value/ Impact Analysis l

The estimated risk reduction for this alternative is 10% chance that the peak pressure will reach 2500 psi if provided in Table 5.i7. Two cases are shown. Case 6(a) is the operator fails to mitigste the transients within 3 based on limiting the peak pressure to 600 psi or less for minutes. The event frequency and LTOP unavailability

' all LTOP transients, and Case 6(b) assumes that there is a are assumed to be the same as the base case evaluation.

Table 5.16 Peak pressure spectrum for Alternative 6.

Pressure Base Case 3 Minutes 10 Minutes Plant Group (psi)

Fraction Fraction Fraction PORV Plants 2500 0.09 0.0 0.1 1400 0.09 0.0 0.0 850 0.13 0.0 0.0 600 0.69 1.0 0.9 RHR SRV Plants 2500 0.14 0.0 0.1 850 0.14 0.0 0.0 600 0.72 1.0 0.9 Table 5.17 Mean core damage frequency estimates for Alternative 6.

Mean Core Damage Frequency (CDF)

CDF Belore Atter Reduction Alternative 1/RY 1/RY 1/RY Ratio 6(a) 3.24x10-6 1.63x10-9 3.24x10-6

1990, 6(b) 3.24xig6 1.50x10-6 1,74xio-6 2.2 5.1.6.2-Industry Implementation Cost Estimates tive. This is a result of delays associated with dressing out, ALARA factors, interfacing with health physics, etc.

'Ihe best estimate industry unit costs associated with this The best estimate cost did not involve stay time because alternative, without the need to modify the quench tank the hours, and hence the exposure, were judged by PNL to heat removal system, are provided in Table 5.18 based on be amenable to normal shift work practices. If the stay NUREG/CR 4627 cost estimates. Additional details are time becomes an important factor in completing the work, provided in Section 11.2 of the PNL value/ impact analysis labor cost would increase by a factor of six instead of for Generic issue 94 (Ref.10),

merely doubling.

The industry implementation cost is doubled for the high in addition to financial costs, there are occupational cost estimate and halved for the low cost estimate. The exposures associated with this alternative. Based on best estimate industry implementation cost, for 67 plants, NUREG/CR-4627 Abstract 4.0, " Occupational Radiation is $41.45 million, with the high and low costs being $83 Exposure," and labor adjustments, the per plant occupa-million and $21 million, respectively. With respect to the tional exposure expected during installation of the nitrogen upper bound, as the necessary labor effort increases, system is estimated to be 216 person-rems. Based on consideration must be given to the impact of radiation ASME Section XI, manual valve operation tests, periodic exposure (stay time) on resultant labor costs. Stay time inspection and maintenance of this system, the additional has been estimated to impact labor cost by a factor of three occupational exposure, over the remaining plant life, is for the moderate radiation zones involved in this alterna-estimated to be about 125 person rems per unit. Based on NUREG-1326 5 22

Value/ Impact Analysis 67 units, the total occupational exposure is estimated to be 5.1.6.4 Valuellmpact Summary 23,000 person rems.

The value/ impact summary for Alternative 6 is provided in 5.1.6.3-NRC Impicmentation Cost Estimates Table 5.20 for each of the three containment assumptions, based on the overall risk reduction. Two risk studies are The best estimate NRC implementation cost is based on a provided. In the first case, the risk reduction is tused on routine technical specification review, $14,200 per plant.

limiting the peak pressure to 600 psi or less in the second in addition, the review, inspection and evaluation of the case, the risk reduction includes a 10% probability of licensee implementation is estimated to require 180 staff reaching 2500 psi a.: a result of a high mass addition tran.

hours per plant (20% of the industry engineering effort).

sient that is not mitigated by the operator, Based on Section 5.2 of NUREG/CR-4627, the cost is estimated as $7,500 per plant. The high cost estimate This alternative is not recommended because the best assumes the technical specification is complex and estimate value/ impact ratio, assuming a 100% reduction in doubles the review effort, $27,400 and $15,000 per plant risk, is well above the $1,000 per averted person rem

- respectively, $42,400 total per plant. 'Ihe low cost guideline, in addition, the resultant occupational exposure estimate is taken as one-half of the best estimate. The associated with this attemative is high, 23,000 person-industry and NRC implementation cost estimates are sum-rems, as compared to the estimated public dose reduction marized in Table 5.19.

estimates.

Table 5.18 Industry unit implementation cost for nitrogen bubble.

Per Plant Cost Action item

($/ plant,1988)

Technical Specification (Routine) 17,400 Waste Gas System Analysis 15,600 Nitrogen System Engineering 48,600 Nitrogen System Materials (valve, piping, instrumentation, etc.)

50,000 Installation (labor and engineering support) 270,000 AdditionalNitrogen Procedures 22,200 System Startup and Installation Tests 18,700 Licensing 9,300 System Quality Assurance 1 % 900 InitialTraining 4,200 Maintenance and Periodic Inspection Present value,5% discount over life 57,000 Total Cost Per Unit 618,600 5-23 NUREG-1326

Value/ Impact Analysis Table 5.19 Implementation cost estimates for Alternative 6 (for 67 plants).

Unit Costs Total Costs Plants Best High Low Best High Low Cost item Affected Est.

Est.

($)

($1,000s)

Industry Total All 618,600 1,237,200 309,300 41,450 83,000-21,000 NRC Tech Spec All 14,200 27,400 7,100 Q/A Review All 7,500 15,000 3,750 Total 1,450 2,840 730 Total 42,900 85,840 21,730 Table 5.20 Value/ impact summary for Alternative 6 (for 67 plants).

TWC Reduction Industry + NRC Person Rem Value/ impact Ratio

/R year Cost Averted

($/erson-Rem)

- Peak Pressure 600 psi Best Estimate 3.24x10 6

$42.90 million 16,000 2,700 High Estimate 3.24x10-6

$85.84 million 29,600 2,900 Low Estimate 3.24x10 6

$21.73 million 5,300 4,100 Peak Pressure 10% 2500 psi Best Estimate 1.74x10-6

$42.90 million 9,300 4,600 High Estimate 1.74x10-6

$85.84 million 17,200 5,000 Low Estimate 1,74x10 6

$21.73 million 3,100 7,000

-- NUREG-1326 5 24

Value/ Impact Analysis-5,1.7 Summary of Best Estimate dose reductions, occupational exposures, industry im-Value/ Impact Ratios for All plementation costs, NRC implementation costs and the Alternatives value/unpact ratio for each of the alternatives studied by the staff. The base case TWC frequency is estimated to be Table 5.21 is provided as a summary of the best estimate 3.24x104per reactor year.

Table 5.21 Summary of best estimate value/ impact (VII) ratios for alterna'ives evaluated by NRC.

TWC Freq Dose Occupational Industry NRC V/1 RatioM Alter.

Reduction Reduction Exposure Costs Costs

($ per averted native (1/R yr)

. (person rem) (person rem)

($1,000s)

($1,000s) person rem) 2 2.89x104 14,500.

n/a 1,370 950 160 3(a) 1.07x10 6 7,000 n/a 3,630 1,840 780-

-3(b) 0.21x104 1,400 n/a 1,290 950 1,600 3(a&b) 1.20x104 8,400 n/a 4,920 2,790 920 4(a) 0.16x10-6 700 n/a 770 650 1,900 4(b) 0.16x10-6 700 n/a 4,770 650 7.750 5

1.82x104

-8,200 900 16,000 570 2,000 5(a)-

3.00x104 13,400 900 16,000 570 1,200 6(a) 3.24x104 16,000 23,000 41,450 1,450 2,700 6(b) 1.74x104

.9,300 23,000 41,450 1,450 4,600 Notes:

Sum of industry plus NRC implementation costs ($s) divided by dose reduction (person rem).

2 Technical specification change,67 plants, proposed resolution.

3(a)

Silockout,67 plants.

3(b)

RCP restart,67 plants.

3(a&b) Both Si and RCP,67 plants.

4(a)

ACI removal, w/o cost for disconnecting ACI,40 PORY plants.

4(b)

ACI removal, w/ cost for disconnecting ACI,40 PORY plants.

5 Safety grade OMS,40 PORV plants.

5(a)

Sensitivity study, safety grade OMS,40 PORY plants.

6(a)

Pressurizer bubble, peak pressure less than 600 psi,67 plants.

6(b)

Pressurizer bubble,10% chance of reaching 2500 psi,67 plants.

1 5-25 NUREG 1326

i l

Value/ Impact Analysis 5.2 Relationships With Other Regulatory 1.

Better pressure instrumentation for low-temperature.

Issues operations to reduce the instrumentation uncertainties and widen the operatiorud window, 5.2.1 Regulatory Guide 1.99, Revision 2,

" Radiation Embrittlement of Reactor 2.

Direct temperature indication of the secondary side Vessel Materials" temperature to ensure that the secondary is not wanner than assumed in the safety analyses studies for the low temperature overpressure protection sys-Revision 2 of this regulatory guide (Ref.11) was used to tem setpoint analysis, determine the mean surface RT(ndt) shift resulting from irradiation induced embrittlement for each reactor vessel in the study. The plant specific chemistry data and 3.

Provide an accumulatar air or nitrogen bottle to the fluence projections to end-of license were obtained from air actuated normal letdown isolation valve to i

licensee submittals related to the " Pressurized Thermal prevent loss of letdown due to problems with the ak Shock" program, USI A 49. Revision 2 is considered by or nitrogen system, both the staff and the industry to be representative of the state-of the-art knowledge concerning irradiation damage.

These actions, if taken, may reduce the frequency of low-temperature overpressure events at some plants. However, Application of Revision 2 to the Appendix G the dominant risk is still associated with those events pressure / temperature limit analysis results in an earlier resulting from maintenance and testing errors, which result narrowing of the operational window for most plants. The in charging without letdown or inadvertent safety injection operational window is defined as the pressure envelope events. These events have the potential to drive the within which the operator must control pressure to ensure system pressure to the primary safety relief valve setpoint a high enough pressure to ensure reactor coolant pump pressure, 2500 psi, if the low temperature overpressure '

seal integrity and cooling and low enough to avoid exceed-protection system fails to mitigate the transient.

ing the Appendix G pressure / temperature limits following the actuation of the low temperature overpressure protec-In addition, as part of the implementation for Regulatory tion system during an anticipated operational transient (for Guide 1,99, Revision 2, the staff is also incorporating example, restart of an idle reactor coolant pump or an clarification into SRP 5.3.2,

" Pressure-Temperature inadvertent safety injection transient).

Limits," (Ref. 3) to ensure that whenever the Appendix G limits are revised the corresponding low temperature over-De impact of Revision 2 on the low temperature over-pressure protection system setpoints are also adjusted, as pressure protection system is considered in this evaluation, needed, to ensure that the Appendix G limits will not be ne operational concerns (the closing or narrowing of the exceeded as a resuh of an anticipated operational transient, operational window) have been addressed as part of the regulatory guide in its supporting regulatory analysis. As part of the implementation recommendation for Revision 5.2.2 Generic Issue 99,"RCS/RIIR Suction 2, Branch Technical Position RSB 5 2 to Standard Review Line Interlocks in PWRs" Plan 5.2 is being revised to include a definition of " low temperature." ne intended purpose of defining " low tem-perature" is to provide more operational flexibility near The findings related to removal of the autoclosure inter-the low temperature overpressure protection system lock (ACl) on the residual heat removal suction line isola-enable temperature for plants that rely on fixed setpoint tion valves in this study for 0194 are consistent with PORVs for protection.

preliminary information under Gl-99. While this action does not result in a significant reduction in risk from low-Industry awareness of the operational window concerns temperature overpressure events, the benefit to decay heat have prompted many licensees to review their plant removal concems may be sufficient justification to allow instrumentation, procedures, and low-temperature over-licensees to remove the interlock, in addition, each pressure event histories to determine what additional licensee would also have to perform a safety analysis to actions can be taken to reduce the occurrence of these demonstrate that overall plant safety is not adversely events. Specific actions under consideration are:

impacted by removal of the autoisolation closure feature.

NUREG-1326 5 26

Value/ Impact Analysis 5.2.3 Generic Issue 70," Power Operated schedule for the PORVs appears not to be justified as Relief Valve and Block Valve discussed in Section 5.1.5 and is not included as part of the Rellability" proposed resolution to 0194 A substantial reduction in risk can be obtained by simply The proposed resolution for 0194 is consistent with the considering the safety significance of these valves in preliminary findings of GI 70, although the risk reduction providing protection against brittle fracture failure of the for low temperature overpressure protection is significant reactor pressure vessel. G194 has determined that PORY when the PORVs (and the residual heat removal system unavailability resulting from not identifying these valves safety relief valves) are administratively treated as as being related to overall plant safety is the dominant con.

components that are used to perform a safety related func-tributor to their unavailability. This finding is equally tion. The cost benefit evaluation does not support an ef-valid for plants that rely on the RHR SRVs since these fort to upgrade the low temperature overpressure protec-valves are also treated as not being related to overall plant tion system to a fully safety grade system. Also for low-safety when providing protection for low temperature temperature overpressure protection, the marginal risk overpressure events in the overpressure protection techni-reduction that might be achieved by revising the testing cat specincation.

5 27 NUREG 1326

6. DECISION RATIONALE Major overpressurization of the reactor coolant system over the remaining licensed life of the 55 PWRs con-while at low temperature,if combined with a critical crack sidered, ranges from 5,300 to 29,600 person-rems, with in the reactor pressure vessel welds or plate material, the best estimate value being 16,000 person-rems.

could result in a brittle fracture of the pressure vessel. As long as the fracture resistance of the reactor pressure The present value, assuming a 5% discount rate, of core vessel material is relatively high, these events are not damage accidents resulting from L'IDP events is estimated expected to cause vessel failure. However, the fracture to be between $6.1 million and $13 million, with a best resistance of the reactor pressure vessel materials estimate value of $9.2 million, for onsite (replacement decreases with exposure to fast neutrons during the life of power and cleanup and repair) damage and offsite (health a nuclear power plant. The rate of decrease is dependent and property) damage.

on the metallurgical composition of the vessel walls and welds. If the fracture toughness of the vessel has been The proposed recommendation, to improve the LTOP reduced sufficiently by neutron irradiation, low-system availability, results in a change in the estimated temperature overpressure events could cause propagation core damage frequency of 2.89x10-6 per reactor year from of fairly small flaws that might exist near the inner 3.24x10 6 to 3.5x104 per reactor year, a factor of ten surface. The assumed initial flaw might propagate into a improvement.

He combined industry and NRC crack through the vessel wall of sufficient extent to implementation best estimate cost is $2.32 million. The threaten vessel integrity and, therefore, core cooling best estimate averted person rem is 14,500 for a best capability. Failure of the pressure vessel could make it estimate value impact ratio of $160 per averted person-impossible to provide adequate coolant to the reactor core rem. The present value best estimate for averted damage, and could result in major core damage or a core damage onsite and offsite, is $8.3 million.

accident.

Although the mean core damage frequency (CDF) for The mean core damage frequency, or reactor pressure LTOP events (3x104 per reactor year) is not large in vessel through wall cmck (TWC) probability, is estimated comparison to other typical CDFs associated with loss of-to be in the 3x104 to 4x104 per reactor year range over coolant accidents or loss of decay heat removal accidents (1x104 o 1x10 5 per reactor year range), with plant age the remaining licensed life of the 55 Combustion t

Engineering and Westinghouse PWRs considered in this the LTOP CDF increases because of reactor pressure ves.

evaluation. For a plant that approaches the IYrS screening set irradiation embrittlement. Also, though not quantified criterion (10 CFR 50.61) at end-of license, the TWC prob.

or explicitly considered in this evaluation, if life extension ability is estimated to be 7x104 per reactor year, assuming is considered, then the LTOP CDF could approach the that the frequency of LTOP events and the unavailability lx10-5 per reactor year range, assuming no additional ac-of the LTOP system remain constant.

tions are taken.

A review of current standard technical specifications for The LTOP core damage accident, resulting from the brittle containment integrity in shutdown modes (Modes 4,5, and fracture of the reactor pressure vessel, is a non-

6) indicates that no containment integrity requirements are recoverable event. Even with emergency core cooling sys-imposed for reactor coolant temperatures less than 200T, tem functional, failure of the pressure vessel could make it except during refueling when the reactor pressure vessel impossible to provide adequate coolant to the reactor core head is removed. Industry responses to NRC Generic and could result in major core damage or a core damage Letter 8712 (Ref. 21) indicate that containment integrity accident.

during Mode 5 is often relaxed to allow for testing, main-tenance, and repair of equipment (for example, contain-Further, since these events are most likely to occur in I

ment penetrations, steam generators, and reactor coolant Mode 5, the containment could be open or non-isolatable pumps). Three risk estimates were employed to address following vessel fracture, containment integrity. Since the low temperature over-pressure events of concern to this evaluation occur in While a requirement for containment integrity, during Mode 5 at reactor coolant temperatures between 80o and Mode 5 (water solid) with one LTOP channel inoperable.

190T, the assumption that containment is open, at least could reduce the public risk (in person-rem), the staff does part of the time, is judged to be valid. Containment not consider this mitigation action as an alternative for the integrity has been treated parametrically in this analysis.

resolution of GI 94. He likelihood of a core damage The resulting public risk from LTOP events, integrated accident would not change. The costs associated with 1

6-1 NUREG-1326

i l

b Decision Rationale onsite damage would not be reduced. The present value for low temperature overpressure protection is significant l

best estimate onsite damage cost is 55.2 million, based on when the PORVs (and the residual heat removal system I

a 5% discount rate, of the total 59.2 million cost with safety relief valves) are administratively treated as safety-offsite damage. Mitigation of public risk does not address resated components. The cost benefit evaluation does not the importance of preventing risk by ensuring that the support an effort to upgrade the low temperature overpres-likelihood of a rapidly propagating fracture of the reactor sure protection system to a fully safety grade system, pressure vessel due to embrittlement is minimized, as required by Appendix A," General Design Criteria " to 10 A substantial reduction in risk can be obtained by simpiy CFR Part 50.

considering the safety significance of these valves in providing protection agamst brittle fracture failure of the The staff proposal is to modify the current technical reactor pressure vessel. 0194 has detennined that PORY specification for overpressure protection to emphasize the unavailability resulting from not identifying these valves safety-related function of the PORVs, and the RHR SRVs, as being related to overall plant safety is the dominant con-for L10P protection, especially when water-solid. The tributor to their unavailability. This finding is equally reported LTOP transients have occurred in Mode 5 with valid for plants that rely on the RHR SRVs since these RCS temperatures ranging from 80 F to 190oF. Since this valves are also treated as not being related to overall plant temperature range includes Mode 6 RCS temperature less safety when providing protection for low-temperature than 140"F but with k,rr less than 0.95 as compared to k rr overpressure events in the overpressure protection techni-e i

less than 0.99 for Mooe 5, the staff concludes that the ad-cal specification, ditional administrative restriction for the single channel AOT is applicable to Mode 5 and Mode 6 (with the reactor pressure vessel head on).

6.1 Conclusions Concerning LTOP Implementation Technical specifications are required by the Atomic Energy Act of 1954 and are implemented in the Code of Low-temperature overpressure protection (LTOP) was Federal Regulations,10 CFR 50.36. Technical specifica-designated as Unresolved Safety Issue (USI) A 26 in 1977 tions are the rules for the normal operation of a nuclear (Ref. 1).

PWR licensees implemented procedures to i

power plant that ensure it is prepared to respond to acci-reduce the potential for overpressure events and installed dents. They define the operating conditions and limiting equipment modifications to mitigate such events, under i

conditions on which safety analyses are based. Since they Multi Plant Action B 04 (Ref. 2).

are a part of the nuclear plant's operating license, the tech-l nical specifications also have a legal basis.

He administrative controls and procedures that were iden-tified as part of Bo4 include the following items:

he proposed recommendation, that the low-temperature overpressure protection system be considered as a system that performs a safety related function-in particular 1.

Minimize the time the reactor coolant system (RCS) during Mode 5 or 6 operation, is consistent with the is maintained in a water-solid condition, intended function attributed to concerns with brittle reac-tor pressure vessel failure as defined in 10 CFR 50 Appen.

2.

Restrict the number of high-pressure Si pumps dix A and Appendix 0. It is not appropriate to consider operable to no more than one when the RCS is in the the function of this system as not being related to plant LTOP condition.

safety.

3.

Ensure that the steam generator to RCS temperature The staff therefore recommends that the technical difference is less than 50 F when a reactor coolant specification for overpressure protection be modified to pump (RCP) is being started in a water solid RCS.

ensure both channels of the low temperature overpressure protection system be operable, especially during Mode 5 4.

Set the PORY setpoint (if the particular plant relies or 6 operations. The allowable outage time (AOT) would on this component for LTOP) to a plant specific be decreased from 7 days to 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months with one channel in-analysis supported value and ha,c mrveillance that operable in Mode 5 or 6. This recommendation applies to checks the PORV actuation electronics and setpomt.

all holders of an operating license or holders of a construc-tion permit for Westinghouse and Combustion Engineer-ing PWRs.

The current staff guidelines for the LTOP system are he proposed resolution for GI 94 is consistent with the found in Standard Review Plan Section 5.2.2, preliminary findings of GI 70, although the risk reduction

" Overpressure Protection," and in its attached Branch 1

1 NUREG-13;.o 6-2

Decision Rationale Technical Position (BTP) RSD 5 2, " Overpressure Protec-reactor coolant pumps in Sections 3.41.3 (Hot Shutdown) tion of Pres;urized Water Reactors While Operating at and 3.4.1.4 (Cold Shutdown). High pressure safety injec.

Low Temperatures"(Ref 3).

tion pump operability restrictions are provided in Section 3/4 5.3 (ECCS Subsystems),

he implementation of the requirements for low-temperature overpressure protection (LTOP), the resolu-In addition to these administrative restrictions, the tran-

~ ion of USI A 26, has been found to be essentially uniform sient analyses are based on a dual channel system's t

for the Combustion Engineering (CE) and Westinghouse being operable to satisfy the single failure criterion of 10 (W) PWRs. With the exception of a few plants,' the CFR Part 50 Appendix A for a system that performs a LTOP protection systems consist of either redundant safety function. Therefore, the overpressure protection PORVs or redundant safety relief valves in the residual system technical specification is consistent with Criterion heat removal system (RHR SRVs) and in general meet the 2 of the Commission's Policy Statement on Technical guidance set forth in Branch Technical Position RSB 5 2 Specification improvements for Nuclear Power Plants.

"Overpressurization Protection of Pressurized Water Reac-De technical specification also satisfies Criterion 3 of the tors While Operating at Low Temperatures."

Policy Statement in that the LTOP system is the primary success path for the mitigation of low-temperature over.

Variability in meeting IEEE-279 requirements, equipment pressure transients that present a challenge to a fission environmental qualification, and the guidance of product barrier in this case, the reactor pressure vessel.

Regulatory Guide 1.26 (Ref. 30) exists. As part of the NRC staff acceptance of LTOP protection system designs ne standard technical specification action requirement for the implementation of the resolution of USl A 26, it for the LTOP system includes a 7 day AOT to restore an was concluded that the costs associated with upgrading ex.

inoperable LTOP channel to operable status before other isting systems to meet these requirements were not justifi-remedial measures would have to be taken (depressurize able. Further evaluations performed for GI 94 have also and vent the reactor coolant system). In addition, Action concluded that it is not cost beneficial to upgrade these

d. states that the provisions of Specification 3.0,4 are not systems to fully safety grade, applicable. Therefore, the plant may enter the modes for which the limiting conditions for operation (LCOs) apply, De section of the standard technical specification cover-during a plant shutdown or placement of the head on the ing the LTOP protection system is titled Overpressure vessel following refueling, when an LTOP channel is in.

Protection System, Section 3.4.10.3 for CE plants and Sec.

operable. In this situation, the 7 day AOT applies for res-tion 3.4.9.3 for W plants. The L'IOP system setpoints are toring the channel to operable status before remedial established based on additional restrictions for the restart measures would have to be taken. This is the same man-of an idle reactor coolant pump and on the number af ner in which the action requirements apply when an LTOP high pressure safety injection pumps and/or coolant charg-channel is determined to be inoperable while the plant is in ing pumps allowed to be operable when LTOP is re-a mode for which the LTOP system is required to be quired. Dese additional restrictions afine the initial con-

operable, ditions for the plant specific transient inalyses performed to establish the LTOP system setpoina. The additional Based on the NRC evaluation of the LTOP system un-restrictions are provided regarding the estartofinactive.

availability, it is concluded that additional restrictions on operation with an inoperable LTOP channel are warranted when the potential for a low temperature overpressure event is the highest, and especially when the plant is in a water-solid condition. The probabilistic risk assessment San Onofre Units 1 and 3 rely on a single RllR performed in support of the resolution of GI-94 is based CE (SDCS) SRV for LTGS. With the SRV inoperable, on the administrative controls and procedures identified as within B hours depressur.te and vent.

part of the Multi Plant Action Item B@ recommenda.

Maine Yankee relies en tso PORVs when pressure tions, it is therefore concluded that these additional is above 400 psig and two RM SRVs when pressure restrictions regarding, in particular, the restart of inactive is below 400 piig.

reactor coolant pumps and the operability of high-pressure safety injection pumps should be implemented in the tech-w -

oc Cook Units 1 and 2 rely on either two PORVs or nical specifications, as indicated in the standard technical one PORV and one Ri!R SRV.

specifications. Licensees should verify that these ad-Yankee Rowe relies on one PORV and two Ri!R ministrative restrictions have been implemented. Finally, SRVs.

it is concluded that these additional measures will help to Newer Westinghcuse plants allow either two PORVs emphasize the importance of the LTOP system, especially or two RilR SRVs.

while operating in a water-solid condition, as the primary 6-3 NUREG-1326

Decision Rationale success path for the mitigation of overpressure transients ments imposed du ing Mode 5, when the reactor coolant during low temperature operation, temperature is below 20TF.

Industry responses to Generic Letter 8712 (Ref. 21) also indicate that contain.

6.2 Improvements in LTOP Protection ment int:grity during Mode 5 is often relaxed to allow for System Availability testing, maintenance, and the repair of equipment.

De staff has determined that LTOP protection system The reported LTOP transients have occurred in Mode 5 unavailability is the dominant contributor to risk from with RCS ;emperatures ranging from 80oF to190 F.

Iow temperature overpressure transients, ne staff has Since this temperature range includes Mode 6, RCS further concluded that a substantial improvement in temperature less than 140 F but with k,rf ess than 0.95 as l

availability, especially during water solid operations, can compared to k rt less than 0,99 for Mode 5, the staff e

be achieved through improved administrative restrictions concludes that the additional administrative restriction for on the LTOP protection system.

the single channel AOT is applicable to Mode 5 and Mode 6 (with the reactor pressure vessel head on). The staff in developing the staff position on the resolution of the proposal is to modify the current technical specification low-temperature overpressure protection generic issue, a for overpressure protection to emphasize the safety related number of factors have been taken into consideration.

function of the PORVs, and the RHR SRVs, for LTOP protection in Mode 5 or 6, especially when water solid.

PORVs are relied on, by most Westinghouse designed plants and about one half of the Combustion Engineering The staff concludes that the LTOP system performs a plants, to provide LTOP protection. The NRC has deter-safety related function and inoperable LTOP equipment mined that over a period of time the role of the PORVs has should be restored to an operable status in a shorter period changed such that PORVs are now relied upon to perform of time. De current 7-day AOT is considered to be too one,or more,of tre following safety related functions:

long under certain conditions. The staff has concluded that the AOT should be reduced to 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months when operating 1.

Mitigation of a design-basis steam generator tube in Mode 5 or 6, when the potential for an overpressure rupture accident, transient is highest. The operating reactor experiences in.

dicate that these events occur during planned heatup 2.

LTOP protection of the reactor pressure vessel (restart of an idle reactor coolant pump) or as a result of during startup and shutdown or maintenance and testing errors while in Mode 5. The reduced AOT in Mode 5 or 6 will help to emphasize the 3.

Plant cooldown in compliance with Branch Techni-importance of the LTOP system in mitigating overpressum cal Position RSB 5-1 to SRP 5.4.7, " Residual Heat transients and provide additional assurance that plant Removal (RHR) System."

operation is consistent with the design basis transient

. analyses.

In addition to PORVs, the residual heat removal system safety milef valves (RHR SRVs) are also relied on to The evaluation performed for the resolution of this generic provide LTOP protection for some Westinghouse plants issue is based on plants being in compliance with their and for the Combustion Engineering plants that do not LTOP design bases analyses. Licensees should verify have PORVs. Newer Westinghouse plants have technical that the administrative controls and procedures identified specifications that require either two PORVs or two RHR in Section 6.1 have been implemented to ensure that the SRVs for LTOP protection.

plant is being operated within the design base. If it is determined that the design base was developed based on ne NRC staff has considered the conditions under which restricted Si pump operability and/or differential tempera.

a low-temperature overpressure transient is most likely to ture restrictions for RCP restart and that tb restriction occur. While LTOP protection is required for all shut-have not been impicmented as part c A-26 and down modes, the most vulnerable period of time was Multi Plant Action llem B-N, then restrictions found to be Mode 5 (Cold Shutdown) with the reactor should be implemented. This is not a new..quirement, coolant temperature less than 200oF, especially when water solid, based on the detailed evaluation of operating The proposed resolution for Generic Issue 94, " Additional reactor experiences performed in support of GI 94 LTOP Low-Temperature Overpressure Protection for Light-transients, which have challenged the overpressure protec-Water Reactors" is expected to reduce by an order of tion ' system, have occurred with reactor coolant magnitude the risks associated with operating at low temperatures in the range of 80oF to 190oF. In addition, a temperatures, he likelihood of an anticipated low-review of the standard technical specifications fo' contain-ten,perature overpressure transient's exceeding the ment integrity indicates that there are no specific require-pressure / temperature limits prescribed under Appendix 0 NUREG 1326 6-4

Decision Rationale

. to 10 CFR Part 50 is also expected to be reduced imm tive of G194. The likelihood of brittle reactor pressure one-in ten to one-in-one hundred, thereby minimizing th' fracture (a through wall crack)is minimized.

probability of a rapidly propagating fracture of the reactor pressure vessel in conformance with General Design The proposed resolution for Generic Issue 94 reduces the Criterion 31 and Oeneral Design Criterion l5.

mean core damage frequency to less than lx104 per reactor year, from 3.24x104 to 3.5x10-7 per reactor year,

[

meeting the target CDF objective stated in Section 2' i

'Ihe likelihood of a low temperature overpressure ~

above. For a plant that approaches the FTS screening transient's resulting in a peak pressure exceeding the criteria at the end of license (with a CDF of 7x104 per Appendix G pressure / temperature limits would be reduced reactor year), the target CDF objective would also be met

- from one-in ten to one in-one hundred, the desired objec-(with the CDF reduced to 7x10-7 per reactor year).

1 6-5 NUREG-1326

7. IMPl.EMENTATION 1he staff proposes to implement the recommendation for providing protection against brittle vessel failure, could the resolution of Generic Issue 94 by issuing a generic result in unacceptable consequences to the health and letter to all licensees and holders of a construction permit safety of the public During Mode 5 (and Mode 6) opera-for Westinghouse and Combustion Engineering designed tions, where LTOP transients occur, the containment nuclear steam supply systems. The content of the generic integrity requirements are often relaxed to permit main.

letter will address the staff concerns related to the current tenance, testing, and repair activities, administrative treatment of the low temperature overpres-sure protection system as not being a system that performs Each licensee will be requested to revise the overpressure a safety-related function, protection technical specification to ensure that both chan-nels of the system are operable when providing protection Plant operations in a degraded mode (one out of two chan-against brittle vessel failure, Sample standant technical nels removed from service), when the L'IOP system is specifications will be provided for guidance, 4

7-1 NUREG 1326

I REFERENCES 1.

U.S Nuclear Regulatory Commission, (USNRC),

12. C. lisu et al., " Estimation of Risk Reduction from

" Approved Category A Task Action Plans,"

Improved PORV Reliability in PWRs," Brookhaven NUREG 0371, Vol.1, No.1, November 1977, National Riboratory, NUREG/CR 4999, BNL-NUREG.52101, Final Report, March 1988.

2.

USNRC, " Operating Reactors Licensing Actions Summary," NUREG 0748, Vol. 4, Nos.1 11,1984,

13. NRR Office Letter No.16, Revision 2, " Regulatory Analysis Guidelines," dated October 3,1984. As 3.

USNRC, " Standard Review Plan for the Review of Amended. (See also NUREG/BR-0058.)

Safety Analysis Reports for Nuclear Power Plants, LWR Ldition," NUREG 0800, July 1981.

14. D.R. Strip, " Estimates of the Financial Consequences of Nuclear Power Reactor Accidents," Sandia 4.

Memorandum from CJ. llettemes, Jr., to H.R. Den-National Laboratories, NUREG/CR 2723, ton, " Case Study Report Low-Temperature Over.

S AND82 ll10, November 1982, pressure Events at Turkey Point 4," AEOD Case Study C401, dated March 16,1984.

15, USNRC, " Demographic Statistics Pertaining to Nuclear Power Reactor Sites. NUREG 0348, 5.

Memorandum from li.R. Denton to R.M. Bernero, November 1979.

Schedule for Resolving and Completing Generic issue 94 Additional Low Temperature Overpressure 16.

D.C. Aldrich et al., " Technical Guidance for Siting Protection for Light Water Reactors," dated July 23, Criteria Development,"

Sandia National 1985.

Laboratories, NUREG/CR 2239, SAND 81 1549, December 1982.

6, USNRC, " Pressurized Thermal Shock," SECY 82 465, November 23,1982.

17. USNRC, " Final Report on Reactor Vessel Pressure Transient Protection for Pressurized Water 7,

F.A. Simonen et al., "VIS A II A Computer Code for Reactors," NUREG 0224, September 1978.

Predicting the Probability of Reactor Pressure Vessel Failure,"

Pacific Northwest Laboratories,

18. E.P. Simonen et al., " Vessel Integrity Simulation NUREG/CR 4486, PNL 5775. March 1986.

(VISA) Code Sensitivity Study," Pacific Northwest

. Laboratories, NUREG/CR-4267, PNL 5469, Decem-

- 8.

F. A. Simonen et al., " Reactor Pressure Vessel Failure ber 1985.

Probability Following 'through Wall Cracks Due to Pressurized Thermal Shock Events," Pacific

19. E.P. Simonen et al., " VISA Il Sensitivity Study of Northwest Laboratories, NUREG/CR-4483, PNL-Code Calculations input. and Analytical Model

' 5727, April 1986.

Parameters,"

Pacific Northwest Laboratories, NUREG/CR 4614, PNL 5863. November 1986.

9.

USNRC, " Safety Goals for the Operation of Nuclear Power Plants; Policy Statement," Federal Recister.

20. Electric Power Research Institute (EPRI), "The In-Vol. 51, p. 30028, August 21,1986.

fluence of Fuel Cycle Duration on Nuclear Unit Performance," EPRI NP-5042, February 1987.

10. B.F Gore et al., "Value/ Impact Analysis of Ger cric Issue 94, ' Additional Low-Temperature Overpres-sure Protection for Light Water Reactors,'" Pacific

21. USNRC, Generic Letter 87-12, " Loss of RIIR While Northwest Laboratories, NUREG/CR-5186, PNL-RCS Partially Filled," dated July 9,1987.

6589, November 1988.

22. S.W.

IIcaberlin et al.,

"A Handbook for

11. USNRC, Regulatory Guide 1.99, Revision 2, Value/ Impact Assessment," Pacific Northwest

" Radiation Embrittlement of Reactor Vessel Laboratories, NUREG/CR 3568, PNL-4616, Decem-Materials," May 1988.

ber 1983.

R1 NUREG 1326

References

23. USNRC, Appendix 111. " Failure Data," to

Reactor

27. EPRI," Guidelines for PWR Pressure Protection Sys-Safety Study An Assessment of Accident Risks in tem Optimization," EPRI NP 3734, October 1984.

U.S. Commercial Nuclear Power Plants, WASH.

1400(NUREG/75 014),0ctober 1975.

28. G.A. Murphy and J.W. Cletcher, " Operating Ex.

perience Review of Failures of Power Operated Relief Valves and Block Valves in Nuclear Power

24. T.L. Chu et al., " Improved Reliability of Residual Plants," Oak Ridge National Laboratory and Profes.

Heat Removal Capability in PWRs As Related to sional

Analysis, Inc.,

NUREG/CR 4692, Resolution of Generic issue 99," Brookhaven Na-ORNIJNOAC 233, October 1987.

tional Laboratory, NUREG/CR 5015, BNL-NUREG-52121, May 1988.

29. Science and Engineering Associates, Inc., et al.,

Generic Cost Estimates: Abstracts from Generic

25. A.M. Rubin," Regulatory Analysis for the Resolution Studies for Use in Preparing Regulatory impact of Unresolved Safety issue A-44, Station Blackout,"

Analyses," NUREG/CR-4627, June 1986.

NUREG 1109, Draft for Comment, January 1986.

30. USNRC, Regulatory Guide 1.26, Revision 3,

26. Memorandum from B.W. Sheron, RES, to F.P. Gil-

" Quality Group Classifications and Standards for lespie, NRR, " Resolution of Generic issue 99 Includ.

Water, Steam, and Radioactive Waste Containing ing Actions Related to the Diablo Canyon Event,"

Components of Nuclear Power Plants," For Com-dated June 13,1988.

ment, February 1976.

I 1

l NUREG-1326 R-2

-l____..

~"'

9

- - - ----i-minim,i.mmu

.. i APPENDIX A

SUMMARY

OF OPERATING REACTOR EXPERIENCES-A1 NUREG 1326 m

Appendix A The purpose of this appendix is to document the results of taken to be the date when a unit first generated electricity Task I of the Task Action Plan for Generic issue 94 (NUREG-0020, " Licensed Operating Reactors").

" Additional Low Temperature Overpressure Protection for l

Light Water Reactors." The Task I objective was to up.

NUREG-0224 (Ref. A.1) listed 30 events that resulted in date the operational experiences data base to determine the significant overpressure transients while in shutdown root causes of_ low temperature overpressure events, to modes of operation. One event has been excluded from determine the unavailability of the protection systems this report because it occurred at a BWR (number 24 of installed during and after 1980, and to determine the peak Table 1, Peach Bottom Unit 2,3/6/74). In addition to the pressure and the initial temperature of actual low-remaining 29 cvents, three other unidentified events were i

temperature overpressure events.

his work was listed (Appendix C, I in 1977 and 2 in 1978). Since these performed by the NRC staff.

events pre date the implementation of LTOP systems, further identification was not made. The 29 identified he experiences data base has been developed from five NUREG 0224 events are summarized in Table A.2.

sources:

NUREG/CR 2789 (Ref. A.2) identified 15 events in shut-1.

NUREG 0224, " Reactor Vessel Pressure Transient down Modes 4,5, and 6, which were not identified in the Protection for Pressurized Water Reactors" (Ref, other data sources. Six events pre-date 1980 and nine oc.

A.1) curred after 1980. None of these events resulted in sig.

nificant overpressure transients, pressures over 500 psi.

2.

NUREG/CR 2789, " Pressure Vessel Thermal Shock All of these ever.ts are classified as precursors, or potential at U.S. Pressurized Water Reactors: Events and overpressure events.

Two other events located in Precursors, 1%31981"(Ref. A.2).

NUREG/CR 2789 were redundant, one from NUREG.

0224 and one from the AEOD Case Study C401, The 15 3.

AEOD Case Study C401, " Low Temperature Over-NUREG/CR 2789 cvents are summarized in Table A.3.

pressure Events at Turkey Point Unit 4" (Ref. A.3)

Ten events, excluding the two Turkey Point 4 events, are 4.

LER Update Search (Ref. A.4).

identified in AEOD C401 (Ref. A.3). Eight of these were also located in the LER Update Search. The two special 5.

Docket 50-275,"RHR System Autoclosure Interlock report events were not located in the LERs because the Removal Report for Diablo Canyon Nuclear Power ORNL data base does not include special reports. One Plant"(Ref. A.5) event is classified as pre-commercial. All of these 10 events were successfully mitigated by the LTOP protec-tion systems. In addition, one of the events (at Calvert The objective of Task I was to determine the post 1980 Cliffs) also reported both LTOP channels unavailable, data base regarding events and low-temperature overpres-Following actuation of the PORV, the operator closed the sure protection (LTOP) or overpressee mitigation system block valve. The operator thought the PORV actuation (OMS) performance data.

Operating Modes 4 (hot was spurious. The second PORV was out of service for shutdown),5 (cold shutdown), and 6 (cold shutdown with maintenance. De 10 AEOD C401 events are summarized reactor vessel head untensioned, or refueling) are in Table A.4. He Turkey Point 4 data are entered under -

considered for this data base Five categories are used to the LER Update Search because these two events include describe the cause of overpressure events. These cases where LTOP channels were also unavailable, categories are described in Table A.l. In summary they are:

The Sequence Coding and Search System (SCSS) data base includes data from 1981 through May 1986, and

- (1) Safety injection-related events..............................S about one-half of the 1980 LERs. Two searches were per-(2) Charging and letdown related events....................C formed to locate 91evant data:

(3) Residual heat removal (RHR) isolation.

4..............R

- (4) Reactor coolant pump restart events.......................P

(5) Other events, not related to above four categories.....Q l.

Find LERs coded with 'RCS' or pressurizer systems and with the abstract [ing] containing ' POP',

It is noted that some events have occurred prior to the

'OVERPRES', or 'LTOP'.

actual date of commercial operation at a given facility.

' nese data are included in the updated data base because 2.

Find LERs coded with the 'RCS' or pressurizer sys-they are infonnative with respect to the root cause of over-tems with an effect of high pressure and with a unit pressure events. The date each unit began commercial effect of hot or cold shutdown, hot standby, or refuel-operations is noted in the data tables and, for this study, is ing.

A-3 NUREG 1326

~ Appendix A Ninety.three LERs were located. These were screenul required to limit the overpressure to an acceptable value, and 57 were found to be applicable. Some of the LERs in accordance with Appendix G (10 CFR Part 50) require.

report multiple events. Seven were redursjant to the

- ments.

AEOD C401 data base.

'Ihe LER Update Search located 37 LTOP events, and 21 Additional data were located and included in the LER cases with one LTOP channel out of service and 25 cases

Update Search data. A preliminary report by AEOD," Air with both LTOP channels declared out of service. Sixteen Systems Problems at U.S. Light Water Reactors," Decem.

of the 37 LTOP events are classified as precursor events.

- ber 1986, identified additional LTOP data, not found else.

The LER Update Search data are summarized in Table where. Most notable is a special report for Farley Unit 2, A.S.

where the peak pressures during two LTOP events were 700 psi and 480 psi. In addition, a report prepared by The data have been compared to the date a piant was.

Westinghouse for Diablo Canyorilisted LTOP events not declared commercial. This date is taken to be the date the previously found (Ref. A.5),

plant first generated electricity (from NUREG.0020),

. This date selection screens out the fewest events for pre.

With the exception of the two Turkey Point Unit 4 events commercial data evaluation.

and the one event at Farley 2,'no other events resulted in significant overpressure transients, exceeding the Appen-dix 0 timits. The LTOP (OMS) systems functioned when The data base (excluding B&W) consists of the following:

Actual LTOP Events Precursor LTOP Events Post Pre.

Post Pre.

Source Total Commercial Commercial Commercial Commercial Pre.1980 Data NUREG.0224 28 14 14 NUREG/CR-2789 5

5 AEOD C401 LERs 3

2 1

Total 36 16 15 5

Post.1980 Data NUREG.0224 NUREG/CR-2789 8

4 4

I AEOD C401 10 9

1 LERs 37 21 15 1

. Total 55 30 5

15 5

NUREG.1326 A.4

l

/

Appendix A l

l One LTOP Unavailable Both LTOPs Unavailable Post Pre.

Post Pre.

Source Total Commercial Commercial Commercial Commercial AEOD C401 1

1 LERs 45 16 5

24 Total 46 16 5

25 0

I t

Actual LTOP Events Precursor LTOP Events

.i Post Prc.

Post Pre.

i Vendor Total Commercial Commercial Commercial Commercial Combustion Engineering 12 5

2 4

1 Westlighouse 79 41 18 16 4

Total.

91 46 20, 20 5

One LTOP Unavailable Both LTOPs Unavailable l

i l

Post Pre.

Vendor Total Commercial Commercial Commercial Commercial Combustion Engineering 6-1 5

Westinghouse 40 15 5

20 Total 46 16 5

25 There were 30 overpressure transients during the period ceeded the Appendix G pressure / temperature limits as 1980 through 1986. The two Turkey Point 4 events in specified in the technical specifications.

1981, at 750 and 1100 psi, and one of the two events at Farley 2 in 1983, at 700 psi (the other reached 480 psi) ex.

The LTOP events data base is summarized below:

A-5 NUREG.1326

Appendix A -

Safety Charging RHR RCP Others injection

/ Letdown Isolation Restart

/Op. Errs.

Vendor Post Pre Post Pre Post Pre Post Pic Post Pre 3

Combustion Engineering 2

1 2

2 2

Westinghouse 19 6 13 11 6

1 12 2

7 2

Total -

21 7

13 13 8

1 14 2

10 2

'Ihe LTOP unavailability data base is summarized below:

Mainte.

Operator Component PORV Leak Air /N2 nance Errors Failure Isolated Problems.

Vendor One Both One Both One Both One Both One Both Combustion Engineering i

2 3

4

-5 8

Westinghouse (Post Com) 4 4

3 1

Westinghouse (Pre-Com) 1 Total 6

6 3

2 6

4 6

8 There are a few cases where plants have contributed Salem 2 GVestinghouse) numerous events to the data base, four or more overpres-rure events and/or four or more reported ca,es of LTOP Salem 2 accounts for four cases of one LTOP channel un-unavailability. These plants are summt.rized below, available and six cases with both channels out. Eight of the 10 events resulted from isolation of leaky PORVs.

Indian Point 2 SVestinghouse)

Salem 2 also accounts for five LTOP events. All were Eight of 8%.51 NOxCO 0224 overpressure transients oc.

successfully mitigated by the LTOP system. One of the curred t.a Indian Point 2. Five of diese are classified as five is a pre-commercial event.

pre-commercial.

McGuire 1 GVestinchouse)

Surry 1 (Westinchouse)

McGuire 1 accounts for four pre-commercial safety injec.

Surry 1 accounts for sever LTOP events: one high.

tion events (Ref. A.2),

pressure event prior to 1980 cnd six after 1980. Three of these six are precursors, and the remaining three were suc.

North ' Anna 1 and 2 SVestinthouse) cessfully mitigated by the LTOP system.

North Anna 1 and 2 account for three cases with one

- LTOP channel unavailable and five cases with both chan.

Zion 2 nVestinchouse)

. nels out. Seven of the eight events relate to problems with the nitrogen pressure system used to actuate the PORVs.

Zion 2 accounts for five LTOP events: one high-pressure event prior to 1980 and four after 1980. Two of these four

' North Anna 1 and 2 also account for eight LTOP events, are precursors, and the remaining two were successfully All were successfully mitigated by tt OP system.

mitigated by the LTOP system, NUREG-1326 A-6

~

l Appendix A Tables A.6 through A.16 provide the details from the A.3 Mernorandum frorn CJ. Heltemes, Jr., to H.R. Den-literature searches.

ton, " Case Study Report 14w Temperature Over.

pressure Events at Turkey Point 4," AEOD Case REFERENCES FOR APPENDIX A Study C401, dated March 16,1984.

A.! U.S. Nuclear Regulatory Commission (USNRC),

A 4 Letter from O.T. Mays Director, Nuclear Safety In-

Final Report on Reactor Vessel Pressure Tnmsient formation Center, ORNL, to E.D. Throm, NRC, Protection for Pressurized Water Reactors,"

dated Septemler 2,1986.

NUREO-0224, Septemter 1978.

A.$ Westinghouse Electric Corporation, Power Systems, A.2 D.L. Phung," Pressure Vessel Thermal Shock at U.S.

RlIR System Autoclosure interlock Removal Pressurized Water Reactors: Events and Precursors, Report for Diablo Canyon Nucicar Power Plant,

1963 1981," Oak Ridge National Laboratory, Docket No. 50-275, WCAP 11117, Revision 2, Ap-NUREG/CR 2789.ORN1/NSIC 210, May 1983, pendix D, dated August 4,1987.

Table A.1 LTOP coding splem.

Event Sequence identifier:

S inadvertent safety injection as a result of operator error during Si testing, inadvertent 51 actuation signal. Could be C; pump or accumulators.

C Excess charging flow. Typically with letdown isolated but not caused by residual heat removal system isolation. Possible high CC llow.

R Residual heat removal (R}iR) system isolation resulting in charging without letdown.

P Restart of a reactor coolant pump (RCP).

Q Other events. Operator errors, procedure errors, or related to maintenance.

LTOP Unavailability identifier:

10 One low-temperature overpressure (LTOp) channel or relief path unavailable. Does not necessarily represent all planned maintenance that does not need to be reported.

20 Both L'IOP channels or relief paths unavailable. May include one out for maintenance when the second fails to mitigate an overpressure event.

Causes broken down into five categories:

(1)

Maintenance, (2)

Operator error, (3)

Component failure, (4)

PORY leakage and isolation,and

($)

Air or nitrogen system failures.

Pressure ColumnIdentifier:

N/A Data unavailable.

None No actual pressurc unsient, a precursor event.

Temp Overcooling event, no pressure data.

S.P.

Upper bound pressure limited to LTOP setpoint. Event mitigated.

Events used by PNL to define the operating reactor experiene data base are marked with F) in the summary data tables.

A7 NUREG 1326

Appendix A Table A.2 Summary of NUREG 0224 LTOP dHa.

Pressure romperature (peu (Doo r)

Desetor Type Date Trom To cause(I)

(*thmercial)

Beaver Valley 1 (W ) 02/24/76 400 1000 130 C-operator error, El. bus transfer, w/RHRS 06/14/76 (pre-com) 03/05/76 400 1150 150 s-operator error, E1. bus de-energized w/RHRS (pre com) 03/13/76 425 495 190 s-Inadvertent safety injection (pre-com)

D.C. Cook 1 (W ) 04/14/76 N/A 1040 110 C-Inadvertent letdown isolation, w/RHRS 02/10/75 R.E. cinna (W >

1969 95 2485 100 C Inadvertent letdosn isolation w/RHR$(pre-com) 12/02/69 Indian Point 2 (W ) 02/16/72 420 670 140 C Unknown cause, w/o RHRS (pre-com) 06/26/73 02/17/72 420 6LO 180 C-Operator isolated letdown, w/RHR$ (pre-com) i 03/08/72 400 640 115 P-koactor crolant pump restart -(pre-com) 04/06/72 422 600 170 C-Operator.;calated letdown, w/RHR$ (pre-com) 05/18/73 440 575 130 C Letdown isos..ed, w/RHRf (pre-com)

Da/23/74 425 525 190 P-Reactor coolant pump restart 02/22/74 150 560 115 S-Inadvertent safety injection, accumulator 09/12/76 400 515 110 C. Letdown isolated, air loss, w/RHRS Indian Point 3. (W ) 09/30/76 50 2250 185 R RHR isolation, spurious 04/27/76 Oconee 2 (BW) 11/15/73 000 1960 300 0-Procedure error, physics tests (pre-com) 12/05/73 Palisades (CE) 09/01/74 W/A 960 150 0-Operator error i

12/31/71

[

Point Beach 2 (W ) U.10/74 345 1400 170 8-Inadvertent safety injection 08/02/72 02/28/76 400 830 168 R RHR isolated, reduced letdown Prairie Island 1(W ) 10/31/73 430 1100 132 P-Reactor coolant pump restart (pre-com) 12/04/73 01/16/74 395 840 90 S-Inadvertent safety injection, accumulator Prairie Island 2(W ) 11/27/74 N/A 900 155 C-Inadvertent letdown isolation,w/RHR$(pre-com) 12/21/74 surry 1 (W ) 01/28/73 450 590 80 S-Insuvertent safety injection, accumulator 07/04/72 st. Lucio 1 (PE) 08/12/75 210 600 105 c-inadvertent letdswv. teolotion.w/sDCs(pre-com) 05/07/76 06/17/76 435 815 200 P Reactor coolant pump restert Trojan (2)

(W ) 07/22/15 400 3326 100 R RHR isolatit T, charging pumps (pre-Som) 12/23/75 Turkey Point 3 (W ) 12/03/74 50 000 105 R RHR isolattdn 11/02/72 tion 1 (W ) 06/13/73 110 1290 105 C-Operator erred (pre-com) 06/28/73 06/03/75 100 1200 115 R RHR isolation tion 2 (W ) 09/18/75 95 1300 88 R-RHR isolati a 12/26/73 t

.r_

(1) First one/two letters used to code data, see Table A.1.

- (2) Apparently pressuriger PORVs and SRV were isolated at tism of event.

i f

NUREG 1326 A8 i

1 Appendix A Table A.3 Summary of NUREG/CR.2789 data.

Pressure Tempe r ature (psi)

(Deg T)

Reactor Type Date From To causeIII (Commercial)

Davis Besse (8W) 04/19/00 N/A 140 6-Inadvertent $1 maintenance,3500 gals, to 170r 08/28/77

( L70P set-pressure is 330 poi )

D.C. Cook 1 (W ) 07/23/01 100 326 C-filling and venting 02/10/75 Farley 1 (W ) 10/24/79 N/A S Inadvertent SI, maintenance, precutsor event 08/10/77 03/30/01((2)

McGuire 1 (W )

N/A S-Inadvertent SI, maintenance,procursortpre-com) 03/30/81 2) w/A

s. Inadvertent SI, maintenance,precursortpre-com*

09/12/81 04/29/81 N/A S-Inadvertent S I, ma int erience, pre cur sor (pre-com) 05/07/81 None S-Inadvertent $1,cper. error, precursor (pre-com)

Millstone 2 (CE) 03/14/19 N/A R-RHP. (SDC) LPSI pump stopped, precursor event 11/09/75 03/14/79 Temp R-RHR (SDC) isolation, precursor event Oconee 3 (DW) 10/19/79 308 360 200 C-excess make-up, operator error 09/10/74 Sorry 1 (W ) 10/01/72 Temp 0-SG blowlown, valve failure, precursor event 07/04/72 04/26/80 N/A S-Inadvertent SI, maintenance, precursor event 04/30/60 N/A S-Inadvertent SI, maintenance, precursor event tion 2 (W ) 05/25/16 N/A S-Inadvertent safety injection, precursor event 12/26/73 09/03/B0 N/A S-Inadvertent safety injection 1 min, precursor (1) First one/two letters used to code data, see Table A.I.

(2) Two events reported.

1 A-9 NUREO-1326 i

Appendix A Table A.4 Summary of AEOD Caw Study C401 data.

Pressure 7emperature (psil (Deg T)

Reactor Type Date From To cause(II (Concercial)

Calvert Clif f 1 (CE) 04/2E/83 425 19B (')O-Orerator error, closes one PORv 01/03/75 04/26/6) 20-Operater error, 17 minutes R.E. Cinna (W ) 06/09/$3(2)

S.P.

(*)C Excess charging 12/02/69

( LTOP set-pressure 435 psi )

North Anna 1 (W ) 03/29/61 8.P.

(*)s-Inadvertent safety in9ection 04/17/78 l LTOP set-pressure 450 poi )

North Anna 2 (W ) 05/18/02 8.P.

(*)P-Reacce r coolant pump restart [ 1.70P 405 poi )

S.P.

(')P-React. coolant pump restart 08/25/B0 05/24/02(3) 05/23/03 387 115 (*)s-Inadvertent safety injection,3 min et 528 gpm Palisades (CE) 12/04/81(4)

S.P.

(*)S Inadvertent safety injection, 4 min 12/31/71

[ LTOP set-pressure 400 psi ]

salem 2 (W ) 06/17/B3 S.P.

( * ) S-Ina d ve r t ent safety injection 06/03/61

( LTOP set pressure 375 pai )

san Onofre 2 (CE) 05/07/62(2) s.P.

C-Letdown decreased w/ charging, /SDCstpre com) 09/20/;2

[ LTOP 7et-pressure 400 psi )

Surry 1 (W ) 07/02/01(4) s.P.

190 (*)C-Inadvertent charging, PCV failure 07/04/72

[ LTOP set-pressure 410 psi )

(1) First one/two letters used to code data, see Table A.I.

(2) All AEOD events, with the exception of these two special reports, were also located in the LER Update search performed 9/4/06.

(3) Listed under Unit 1 in AEOD report.

(4) Event date listed as 1982 in AEOD report, ccrrected to 1981 based on LER Update Search of 09/4/B6.

(*) Events considered in developing base case risk analysis.

NUREO 1326 A 10

t l

l Appendix A Table A.$ Summar) of LER Update Search data.

Pressure Temperature (psil (Deg F)

Reactor Type Date From 70 CauseIII (Commsrcial)

Dyron 1 (w ) 03/10/85 N/A

(*)s-Oper. error, Anadvertent s!, precursor 03/01/85 Ca11 sway 1 (w ) 07/16/04 350 4.5 140 C-Excess charging, loss of instr. air (pre-com) 12/19/84 07/16/84 150 425 140 C Excess charging, loss ef instr. air (pre com) 04/05/G6 380 463 104

(*)C-Excess chsteing, RCP seal injection valve failure Catawba 2 (W ) 04/09/86 260 399 C Letdown isolated by operator (pre-com) 5/18/86 Calvert Cliffs 1(CE) 10/20/85 20-ca11 bastion error, maintenance 01/03/15 D.C. Cook 2 (W l 01/08/83 20-Low Air supply pressure, isolated during test 03/22/78 07/03/83 20-Loss of Air supply, 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months 07/28/05 None S Inadvertent safety injection, precursor event Farley 1 (W ) 11/09/80 None C-More than 1 CC pump available, precursor event 08/18/77 11/07/86 400 450

(*)C-Operstor error on pressure control 11/15/86 400 450

(*)P-RCP restart, third pump Farley 2 (W ) 10/15/83(2) yon g7o g.3c. Excess charging, loss of instr. air 05/25/01 10/15/83(2) go.out for maintenance 10/15/83(2) 480 170

(*)P-RCP restart 10/15/83(2) lo.out for maintenance R.E. Cinna (W ) 05/06/80 20-Operator / procedure error, 14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months 12/02/69 05/22/94 None S-Inadvertent SI, low SG pressure, precursor Haddam Neck (W ) 11/30/03(?)

10-Loss of Air supply 08/07/77 11/30/83(2)

Io. Loss of Air supply 08/0*/84 315 380 325 0-Unknown, PORVs open, oper. closed no transient 08/03/84 20-Operator closed for 17 min.

01/05/86 20-Component failure, interlock Maine Yankee (CE) 07/17/81 10-Out for maintenance, 7 hr 20 minutes 11/08/72 12/01/83 None Q-Operator error, outdated procedures, precursor i

Mc Cuire 1 (W ) 03/10/02 10-Calibration error, timing, raintenance 09/12/91 He Guiro 2 (W ) 04/13/03 10-Instrument error, air in line, maintenance 05/23/83 08/21/86 360 368 180 0 Too close to set-point 11/15/06 350 368 180 Q-Too close to set point Millstone 2 (CE) 06/15/05 20-Procedure error, operator error 11/09/75 North Anna 1 (W ) 03/ /78 575 0-Electrical problem (pre-com) 04/17/78 03/ /B0 570 0-valve failure, RHR relief valve opens 03/18/81 20-Low N2 pressure, 1 out 8 hr then other falle 11/10/82 20-Block valve closed, 5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months , maintenance i

10/09/83 lo-Loss of N2, 4 days 09/14/84 350 410 88

(*)P-Reactor Coolant Pump restart 12/16/85 None 8-10,000 gal accumulator, vented. precursor event 12/19/05 350 395 135 (*)C-Charging and letdown control (1) First oneitwo letters used to code data, see $able A.I.

(2) AEOD report, ' Air Systems Problems in U.S. Light Water Reactors.*

(3) Two events over a two week pericJ.

(*) Events considered in developing base case risk analysis.

A ll NUREO 1326

Appendix A Table A.5 Summary of LER Update Search data (con'O.

Pressure 1emperature (psil (0., r)

Reactor Type Date From To cause(II (Commercial)

North Anna 2 (W ) 08/14/80 10-Low N2 pressure, vented 08/25/80 11/02/80 10-Low N2 pressure, 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months 06/20/81(3) 20-Low N2 pressure, 7 hrs 45 min 08/06/81(3) 20-Low N2 pressure, ' hours 08/17/81 20-Low N2 nrossure, 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months Palisades (CE) 08/13/03 20-Procedure error, 78 implementation op. error 12/31/11 08/26/85 350 375

(*)P-Reactor coolant pump restart, third pump 09/14/85-20-calibration errce, cal. 8/27/85, maintenance Palo Verde 1 (CE) 01/28/86 None S-Inadvertent safety injection, precursor event 06/10/85 04/06/85 s-Inadvertent aatety injection, precur.(precom)

Point Beach 2 (W ) 10/25/02(2) 10-Loss of Air, operator error 08/02/72 Prairie !aland 1(W ) 10/ /74 N/A S. Inadvertent safety injection 12/04/73 H.B. Robinson 2 (W ) 01/ /78 360 560 155 0 Heatup when RHR pump stopped 09/26/10 11/04/83 10-component, limit switch 12/15/04 20-Air /N2 1solated, 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months salem 1 (W ) 07/16/81 10-PORV leak, isolated, 48 hours5.555556e-4 days 0.0133 hours 7.936508e-5 weeks 1.8264e-5 months 12/25/76 01/06/02 10-PORV leaked, inclated, 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months 01/07/82 lo-Failed manuel test. component other open/ vent Salem 2 (W ) 04/19/81(3) 10-PORY leak, isolated, 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months 06/03/01 04/23/81(3) 10-PORV leak, isolated, 3 days 04/29/81(3) 10-PORV leak, isolated, 4 days 05/15/81(4)330 360 175 Q-Operator error, maintenance, PZR bub (pre-com) 06/12/81(4) 20-PORVs leak, isolated 06/18/81 20-PORVs leak, isolated 07/09/01 10-PORV leak, isolated 01/22/83 20-PORVs leak, isolated 01/26/83 20-PORV leak, other no Air 5

04/25/83 20-Blank in vent line, 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , maintenance 08/30/83 20-Position indicator fails, 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , component 02/15/84 325 350

(*)P-Reactor coolant pump restart 03/29/05 325 300

(*)P Reactor coolant pump restart 03/30/85 325 300

(*)P-Reactor coolant pump restart son onofre 1 (W ) 11/10/03 900 522 120

(*ls-Inadvertent safety injection 07/16/67 surry 1 (W ) 02/09/83 20-Low backup Air pressure. normal air available 07/04/12 06/01/84 325 412 145

(*)C-Excess charging, operator erro-05/12/85 350 410 150

(*)C-Letdown decrease w/ charging Summer (W ) 05/06/85 450

(*)P-Reactor Coolant Pump restart 11/16/82 (Notes RHR SRV failed to fully reseat]

Trojan (W ) 05/21/95 S SI signal 3 minutes, hear removed, precursor 12/23/75 10/21/06 none 0-PORV open in 8 sec, should be 0.28 sec, maintenance (procursor event)

(1) First one/two letters used to code data, see Table A.1.

(2) AE0D report,' ' Air systems Problems in U.S. Light Water Reactors.'

(3) Three events over a one week period reported.

(4) Two events over a one week period reported.

- (*) Events considered in developing base case risk analysis.

s NUREG 1326 A.12

Appendix A Table A.5 Summary of LER Update Search data (con't).

Pressure Temperature (psi)

(Deg r)

Reactor Ty;*

Date From to causeIII (Commercial)

Turkey Point 4 (W ) 11/28/81 310 1100 110

(*)P-Reactor coolant pump restart 06/21/73 11/29/81 340 750 110

(*)P-Reactor coolant pump restart 11/28/01 20-1 maintenance, other fails 11/29/81 20-1 maintenance, other fails 10/09/84 lo-component is11ure PORY opens,cperator closes Yankee Rowe (W ) 07/17/81 10 'eintenance error r

11/10/60 i

tion 1 (W ) 09/11/84 450

(*)C-Increased charging flow, op error 06/28/73 91/03/86(21 01/03/06(2)400 43$

(ein.RHR isolated letdown,w/ charging 190 gpia tion 2 (W )

12/26/73 435

(*)o-possible electrical bus problem (1) First nne/two letters used to code data, see Table A.I.

(2) Two events reported.

(*) Events considered in developing base ease risk analysis.

I f

1 3

A 13 NUREO 1326

'{.'

r Appendix A cx Table A.6 Summary of Babcock and Wikox plants.0)

Plant HW(e) Docket Concercial R Yrs & C R P Q Sum Events 10 1 Out 20 2 Out HP HP Date (12/06)

/R-Yr

/R*Yr

/R-Yr

/R-Yr Arkansas one-1 834 50-313 8/ 1/74 12.4 Crystal River 3 825 50-302 1/30/77 9.9 Davis-Desse 1 880 50-346 8/28/77 9.3 1----

1

.107 g

Oconee 1 860 50-269 5/ 6/13 13.7 g

Oconee 2 860 50-270 12/ 5/73 13.1

1 1

.077 1

.077 Oconee 3 860 50-287 9/ 1/74 12.3 --

1

.001 Rancho seco 1 916 50-312 10/13/14 12.2 Three Mile 1 792 50-289 6/19/74 12.5 (1) 8-safety injection C Charging / letdown R-RHR isolation P-RCP restart 0-Other/oper, error HP High pressure event. >500 psi 10-One LioP channel out 20-Both LTOP channels out Table A.7 Summary of Combustion Engineering plants.0)

Plant HW(e) Docket Commercial R Yrs $ C R P Q Sum Events 10 1 Out 20 2 Out HP HP Date (12/06)

/R-Yr

/R=Yr

/R-Yr

/R-Yr Arkansas one-2 858 50-368 12/26/78 8.0 Calvert Cliffs 1 845 50-317 1/ 3/75 12.0

1 1

.083 2

.1(7 Calvert Cliffs 2 845 50-318 12/ 7/76 10.1 Pt. Calhoun 1 478 50-285 8/25/73 13.4 Maine Yankee 190 50-309 11/ 8/72 14.1

1 1

.071 1

.071 Millstone 2 830 50-336 11/ 9/75 11.1

--2--

2

.180 1

.090 Palisades 198 50-255 12/31/71 15.0 1--11 3

.200 2

.133 1

.067 Palo Verde 18 1270 50-528 6/10/85 1.6 2----

2 1.250 Palo Verde 28 1270 50-529 5/20/06

.6 san onofre 2**

1070 50-361 9/20/82 4.3 --

1

.234 San Onofre 3**

1080 50-363 9/25/83 3.3 St. Lucie 1 810 50-335 5/ 7/76 10.7 1-2

.188 2.188 St. Lucio 2 810 50 389 6/13/83 3.6 Waterford 38 1165 50-382 3/18/85 1.8 (1)

S-Safety injection C-Charging /letdcwn R-RHR isolation P-RCP restart 0-Other/oper. error HP-High-pressure event. >$00 psi 10-One LTOP channel out 20-Both L70P channels out Notes

-LTOP with 2 SDCS SRVs

- -LTOP with 1 SDCS SRV NUREG 1326 A 14

Appendix A Table AA Summary of Westinghouse plants.0)

Plant MW(e) Docket Commercial R* Yrs 8 C R P Q Sum Events 10 1 Out 20 2 Out HP HP Date (12/06)

/R Yr

/R-Yr

/R-Yr Beaver Valley 1 833 50-334 6/14/76 10.6 21---

3

.284 3.204 Byron 18 1220 50 454 3/ 1/85 1.8 1-

-- 1

.556 Callaway 18 1150 50-463 10/24/04 2.2

-3*--

3 1.346 Catawba 1 1145 50 413 1/22/05 1.9 Catawba 2 1145 50-414 L/18/06

.6 -.

1 1.613 Cook 1 1030 50 315 2/10/75 11.9 --

2

.168 1.084 Cook 2 1090 50-316 3/22/78 8.8 1-+--

1

.114 2

.228 Diablo Canyon 1 1084 50-275 11/11/84 2.1 rarley 1* yon 2 Diablo Can 1106 50-323 10/20/85 1.2 860 50-348 8/10/77 9.4 12 4 426 Parley 2' 060 50-364 5/25/01 5.6

-1 1-2

.357 2

.351 1.179 Haddam Neck 582 50-213 8/ 7/67 19.4

-.--1 1

.052 2

.103 2

.103 Indian Pt. 2 873 50-247 6/.*f r 7 3 13.5 15-2 8

.592 8.592 Indian Pt. 3 965 50-286 4/27/76 10.7

--1-.

1

.094 1.094 kewaunee*

535 50 305 4/ 8/74 12,7 McGuire 1 1180 50-369 6/30/81 5.5 4--

. 4 727 1

.182 McGuire 2 1100 50 370 5/23/03 3.6

2 2

.554 1

.277 Millstone 3 1150 50-423 2/12/86

.9 North Ar.na 1 890 50-338 4/17/78 8.7 21 11 5

.505 1

.115 2

.230 North Anna 2 890 50-339 8/25/80 6.4 1-3

.472 2

.315 3

.472 Point Beach 1 497 50 266 11/ 6/10 16.2 Point beach 2 49' 50-301 8/ 2/72 14.4 1-1--

2

.139 1

.069 2.139 Prairie Island 1 507 50-202 12/ 4/73 13.1 2--1 3

.229 3.229 Prairie Island 2 507 50 306 12/21/74 12.0 --

1

.083 1.083 R.E. 01nna 470 50-244 12/ 2/69 17.1 12---

3

.176 1

.059 1.059 Robinson 2 655 50-261 9/26/10 16.3

1 1

.061 1

.062 1

.062 Salem 1 1090 50-272 12/25/76 10.0 3

.299 Salem 2 1115 50-311 6/ 3/01 5.6 1--31 5 1.030 4

.717 6 1.075 San onofre 1 436 50-206 7/16/67 19.5 1----

1

.051 Sequoyah 1**

1148 50 327 7/22/00 6.4 sequoyah 2**

1148 50-328 12/23/81 5.0 summer

900 50 395 11/16/82 4.1

-- 1

.244 Surry 1

??5 50 280 7/ 4/12 14.5 33--1 7

.483 1

.069 1.069 surry 2 775 50-201 3/10/73 13.8 Trojan 1130 50-344 12/23/75 11.0 111 3

.273 1.091 Turkey Pt. 3 128 50 250 11/ 2/72 14.2

--1--

1

.071 1

.071 Turkey Pt. 4 728 50-251 6/21/73 13.5

-- 2

.148 1

.074 2

.148 2.346 Wolf Creeki 1150 50-482 6/12/85 1.6 Yankee Rowe 175 50- 29 11/10/60 26.1 1

.038 tion 1 1040 50 295 6/28/73 13.5 -

3

.222 2.148 Elon 2 1040 50-304 12/26/73 13.0 k-2-1 5

.384 1.077 (1) s-safety injection C. Charging / letdown R-RHR istlation P-RCP restart 0-Other/oper. error HP-High-pressure event. >500 ps!

10-One 1.70P channel out 20-Both LTOP channels out Notest

'-LTOP with 2 RHR$ SRVs

- -Automatic LTOP l-LTOP 2 PORVs or 2 SRVs T

k A 15 NUREG.1326

Appendix A Table A 9(a) Combustion Engineering LTOP events summary. total data base.

SI CC RHR RCP Others Total Pest-com Total Pre-Com Total History y

N P

P P P P P P P P P

I O

R 0 R 0 R 0 R 0 R S Freq Frog 5 Frog Frog &

Frog Freq T

S E S E $

E $

E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T H

Unit R-Yr M Unit R-Yr M Unit R-Yr 69

.0 10

.0 11 1

.0 72-2 1.1 73 3

2.4 14 3

3.0 1

.333

.333 1

.333

.333 1

75 5

4.1 1

.200

.244 1

.200

.244 1

-+

16 7

5.1 1

.143

.175 1

1

.143

.175 17 7

1.0 78 8

7.1 2

.250

.250 79 8

8.0 2

2

.250

.250 60-79 8 38.4 4

.500

.104 1

.125

.026 5

.625

.130 1 2 1

1 80 8

8.0 1

.125

.125 1

.125

.125 81 8

8.0 1 82 9

8.2 1

1

.111

.122 1

.111

.122 83 11 9.9 2

.182

.202 2

2

.182

.202 84 11 11.0 85 13 12.3 1

1

.077

.081 1

.077

.081 2

.154

.163 1

86 14 13.6 1 1

.071

.014 1

.071

.074 80-86 14 11.0 2 1

1 5

.357

.070 2

.143

.028 7

.500

.099 1

2 60-86 14 109.4 2 1

2 2 3

- 9

.643

.082 3

.214

.027 12

.857

.110 2

Table A.9(b) Combustion Engineering LTOP events summary without precursor data.

51 CC RHR RCP Othere Total Post-Com Total Pre-Com Total History U

+= ------ ------ ------ ------ ---------------- ---------------- ----------------

N P

P P P P P P P P P

I O

R 0 R 0 R 0 R 0 R S Freq Freq S Freq Freq S Freq Freq T

S E S E S T S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T M

Unit R*Yr M Unit R-Yr M Unit R=Yr 69

.0 70

.0 71 1

.0 12 2

1.1 13 3

2.4

~ -

14 3

3.0 1

.333

.333 1

.333

.333 1

15 5

4.1 1

.200

.244 1

.200

.244 1

16 7

5.1 1

1

.143

.175 1

.143

.175 17 1

1.0 78 8

1.1 19 8

8.0 60-19 8 38.4 1 -

2

.250

.052 1

.125

.026 3

.375

.078 1

1 80 8

8.0 1

.125

.125 81 8

8.0 1

- 1

.125

.125 82 9

8.2 1

1

.111

.122 1

.111

.122

- 1

.091

.101 1

.091

.101 03 11

9. 's 84 11 11.0 1

.071

.081 85 13 12.3 1

.071

.001 1

86 14 13.6 -

80-86 14 11.0 1 3

.214

.042 1

.071

.014 4

.286

.056 1

1 1

60-86 14 109.4 1 2

2 2

5

.357

.046 2

.143

.018 7

.500

.064 NUREO 1326 A 16

Appendix A Table A.10(a) Westinghouse LTOP events summary. total data base.

El CC RHR RCP Others Total Post-Com Total Pre-Com Total History y

N P

P P P P P P P P P

I C

R 0 R 0 R 0 R 0 R S Freq Treg 8 Freq Trog S Frog Freq T

S E S E S E 8 E S E U Per Per U Per Per U Per Per Year S

R= Yrs T T

T T

T M

Unit R-Yr M Unit R-Yr M Unit R-Yr 1

.250

.323 1

.250

.323 1

69 4

3.1 10 6

4.4 71 6

6.0

=

1

.111

.141 4

.444

.563 5

.556

.704 3

1 1 72 9

7.1 1

.067

.088 3

.200

.263 4

.267

.351 13 15 11.4 1 2

1 6

.353

.380 1

.059

.063 7

.412

.443 1 1 1

74 17 15.8 4 2

.105

.112 1

.053

.056 3

.158

.168 2

1 75 19 17.9 5

.227

.248 3

.136

.149 8

.364

.396 16 22 20.2 1 2 2 1~2 77 g3 22.4 1

1 1

.040

.041 1

.040

.041 2

.000

.002 18 25 24.5 1

.040

.040 1

.040

.040 19 25 25.0 1 60-79 25 168.8 7 2 2 8 5 1 1 2 2 1 11

.600

.101 14

.560

.083 31 1.240

.184 4

.148

.155 4

.148

.155 80 27 25.8 3 1

1 5

.161

.174 5

.161

.174 10

.323

.348 81 31 28.1 1 4 2 2

2

.063

.064 2

.063

.064 2

82 32 31.1 6

.182

.184 6

.182

.1C4 83 33 32.6 3 2

1 6

.111

.180 2

.057

.060 8

.229

.240 2

2 84 35 33.3 1 2

1 9

.231

.239 9

.231

.239 85 39 37.6 4 2

3 8

.195

.198 1

.024

.025 9

.220

.222 2

1 1 86 41 40.5 1

4 5

1 40

.976

.174 8

.195

.035 48 1.171

.209 80-86 41 229.6 12 4 11 3 1

- 11 60-86 41 398.4 19 6 13 11 6 1 12 2 1 2 51 1.390

.143 22

.537

.055 19 1.927

.198 Table A.10(b) Westinghouse LTOP events summary without precursor data.

SI CC RHR RCP Others Total Post-com Total Pre-Com Total History y

N P

P P P P P P P P P

I O

R 0 R 0 R 0 R 0 R S Freq Freq S Freg Freq S Freq Freq T

S E S E S E S E S E U Per Per U Per Per U 7er Per Year S R-Yrs T T

T T

T H

Unit R-Yr M Unit R-Yr M Unit R-Yr 1

.250

.323 1

.250

.323 1

69 4

3.1 70 6

4.4 71 6

6.0 4

.444

.563 4

.444

.563 1

12 9

1.1 3 -

1

.067

.088 3

.200

.263 4

.267

.351 73 15 11.4 1 2 -

1 6

.353

.380 1

.059

.063 7

.412

.443 1 1 74 17 15.8 4 1

2

.105

.112 1

.053

.056 3

.158

.168 75 19 17.9 2

1 4

.182

.198 3

.136

.149 7

.318

.347 76 22 20.2 2 2 1 2 77 23 22.4 1

1 1

.040

.041 1

.040

.041 2

.000

.082 18 25 24.5 19 25 25.0 60 19 25 168.8 5 2 2 8 5 1 1 2 1 1 14

.560

.083 14

.560

.083 28 1.120

.166 80 27 25.0 1

1 1 4

.129

.139 1

.032

.035 5

.161

.174 81 31 28.7 1 1

2 2

.063

.064 2

.063

.064 -

82 32 31.1 2

6

.182

.184 6

.182

.184 83 33 32.6 3 2

1 4

.114

.120 2

.057

.060 6

.171

.180 84 35 33.3 2

2 2

6

.154

.160 6

.154

.160 85 39 3?.6 1 2

3 5

.122

.123 1

.024

.025 6

.146

.148 86 41 40.5 2

1 1 1

1 2

1 27

.659

.118 4

.098

.017 31 756

.135 80-86 41 229.6 5 9

3 1

- 11 60 86 41 398.4 10 2 11 11 6 1 12 2 3 2 41 1.000

.130 18

.439

.045 59 1.439

.148 A.17 NUREO.1326

Appendix A Table A.11(a) Total LTOP events summary. total W and CE data base.

81 CC RHR RCP Others Total Post-Com Total Pre-Com Total History y

N P

P P P P P P P P P

1 0

R 0 R 0 R 0 R 0 R S Frog Freq S Freq

' req 8 Freq Freq T

S E S E S E S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T M

Unit R-Yr M Unit R-Yr M Unit R-Yr 60-19 33 207.2 7 2 2 9 7 1 2 2 3 1 21

.636

.101 15

.455

.072 36 1.091

.174 80-86 55 300.6 14 5 11 4 3

- 12 1

1 45

.818

.150 10

.182

.033 55 1.000

.183 60 86 $$ 507.8 21 1 13 13 8 1 14 2 10 2 66 1.200

.130 25

.455

.049 91 1.655

.179 Table A.11(b) Total LTOP events Summary without precursor data. total W and CE data base.

S1 CC RHR RCP Others Total Post-Com Tota, Pre.com Total History U

N P

P P P P P P P P P

1 0

R O R 0 R O R 0 R S Freq Freq S Freq Freq S Freq Freq T

S E S E $

E S E $

E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

M Unit R-Yr M Unit R-Yr M Unit R-Yr 60 79 33 207.2 5 2 2 9 5 1 2 2 2 1 16

.485

.077 15

.455

.072 31

.939

.150 80-86 55 300.6 6 9

4 1

- 12 3

1 30 545

.100 5

.091

.017 35

.636

.115 60 86 55 507.8 11 2 11 13 6 1 14 2 5 2 46

.638

.091 20

.364

.039 66 1.200

.130 Table A.12(a) Combustion Engineering one LTOP channel unavailable summary.

Maint. Op Err Compnt Leaks Air /N2 Total Post-Com Total Pre-Com Total History y

N P

P P P P P P P P P

1 0

R 0 R 0 R v R 0 R S Freq Freq S Freq Freg S Freq Freq T

S E S E S E S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T H

Unit R-Yr M Unit R-Yr M Unit K-Yr 80 8

8.0 81 8

8.0 1

1

.125

.125 1

.125

.125 82 9

8.2 83 11 9.9 84 31 11.0 -

85 13 12.3 86 14 13.6 80-06 14 71.0 1 1

.071

.014 1

.071

.014 Table A.12(b) Combustion Engineering both LTOP channels unavailable summary.

Maint. Op Err Compnt Leaks A1!/N2 Total Post-Com Total Pre-Com Total History y

N P

P P P P P P P P P

1 0

R 0 R 0 R 0 R 0 R S Freq Freq $

Freq Freq S Freq Freq T

S E S E S E S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T M

Unit R-Yr M Unit R-Yr M Unit R-Yr 80 8

8.0 81 8

8.0 82 9

8.2 83 11 9.9 2

2

.102

.202 2

.182

.202 84 11 11.0 85 13 12.3 2

- 1

- 3

.231

.244 3

.231

.244 86 14 13.6 80-86 14 71.0 2

- 3 5

.357

.070 5

.357

.070

- NUREO.1326 A.18

Appendix A Table A.13(a) Westinghouse one LTOP channel unavailable summary.

Maint. Op Err Compnt Leaks Air /N2 Total Poet-Com Total Pre-Com Total History y

N P

P P P P P P P P P

I O

R 0 R 0 R 0 R 0 R S Freq Freq S Freq Freq S Freq Freq T

S E S E S E S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T M

Unit R-Yr M Unit R-Yr M Unit R=Yr j

1 1 1

.037

.039 1

.037

.039 2

.074

.078 80 27 25.8 3

.097

.105 3

.097

.105 6

.194

.209 2

3 81 31 28.7 1 4

.125 129 4

.125

.129 1

1 1

82 32 31.1 1 6

.192

.284 1

.030

.031 7

.212

.215 3

83 33 32.6 2 1

1 1

.029

.030 1

.029

.030 84 35 33.3 1

85 39 37.6 -

86 41 40.5

~~

3 3 5 1 15

.366

.065 5

.122

.022 20

.488

.007 3

80-06 41 220.6 4 1

j Table A.13(b) Westinghouse both.'. TOP channels unavailable summary.

Maint. Op Err Compnt Leaks Alt /N2

'otal Post-Com Total Pre-Com Total History U

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

N P

P P P P P P P P P

I O

R 0 R 0 R 0 R 0 R S Frey Freq S Freq Freq S Freq Frog T

S E S E S E S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T M

Unit R-Yr M Unit R-Yr M Unit R-Yr 1

.037

.039 1

- 1

.037

.039 80 27 25.8 8

.258

.279 8

.258

.279 2

4 81 31 28.7 2 1

.031

.032 1

.031

.032 82 32 31.1 1 7

.212

.215 7

.212

.215 1

- 2 3

03 33 32.6 1 2

.057

.060 2

.057

.060 1

1 84 35 33.3 85 39 37.6 1

.024

.025 1

.024

.025 1

86 41 40.5 20

.488

.087 8

- 20

.488

.087 2

2 4

80-86 41 229.6 1 Table A.14(a) Total for one LTOP channel unas allable summary.

Maint. Op Err Compnt Leaks Air /N2 Total Post-Com Total Pre-Com Total History U

N P

P P P P P P P P P

-I O

R 0 R 0 R 0 R O R S Freq Frog S Freq Freq S Freq Freq T

S E S E S E S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T M

Unit R*Yr M Unit R-Yr M Unit R-Yr 1

1 1

.029

.030 1

.029

.030 2

.057

.059 80 27 33.8 4

.103

.109 3

.077

.082 7

.179

.191 2

3 1

81 31 36.7 2 4

.098

.102 1

- 1

- 4

.098

.102 l

82 32 39.3 1 6

.136

.141 1

.023

.024 7

.159

.265 1

83 33 42.5 2 1 -

1

- 3 1

.022

.023 1

.022

.023 1

84 35 44.3 85 39 49.9 86 41 54.1 3

3 5 1 16

.291

.053 5

.091

.017 21

.382

.070 3

80-06 41 300.6 5 1

Table A.14(b) Total for both LTOP chEnnels unavailable summary.

Maint. Op Err Compnt Leaks Air /N2 Total F'..t-Com Total Pre-Com Total History U

N P

P P P P P P P P P

I O

R 0 R O R 0 R 0 R S Frog Freq S Freq Freq S Freq Freq T

S E S E S E S E S E U Per Per U Per Per U Per Per Year S R-Yrs T T

T T

T H

Unit R-Yr M Unit R-Yr M Unit R-Yr 1

.029 030 1

- 1

.029

.030 -

80 27 33.8 8

.205

.218 4

- 8

.205

.218 2

81 31 36.7 2 1

.024

.025

- 1

.024

.025 -

82 32 39,3 1 9

.205

.212 9

.205

.212 -

3 83 33 42.5 1

- 2

- 1 2

2

.043

.045 2

.043

.045 -

1 1

84 35 44.3 -

3

.058

.060 3

.058

.060 -

1 85 39 49.9 2 1

.018

.018 86 41 54.1 -

- 1

.018

.018 25

.455

.083 80-06 41 300.6 6

- 5

- 2

- 4

- 8

- 25

.455

.003 -

A 19 NUREO 1326

J Appendix A Table A.15 Pressure / temperature data summary.

Post Commercial Data Pre-Commercial Data P

T U

P T

U P

T U

P T U P

T U

P T

U Yr psi F

poi r

psi F

psi r

pai r

poi r

E9 2485 100 (W )

72 640 115 (W ) 650 180 (W ) 670 140 (W )

73 590 80 (W )

575 130 (W ) 1100 132 (W ) 1290 105 (W )

1860 300 (BW) 74 525 190 (W )

560 115 (W )

800 105 (W )

900 155 (W )

960 150 (CE) 840 90 (W )

1400 170 (W )

75 1100 115 (W ) 1300 80 (W )

600 105 (CE) 3326 100 (W )

76 515 110 (W )

815 100 (W )

830 168 (W i 495 190 (W ) 1000 130 (W ) 1150 150 (W )

1040 110 (W ) 2250 185 (W )

78 560 155 (W )

79 360 200 (BW) 80 330 140 (BW) 13 110 (W ) (2)1100 110 (W )(1) 360 175 (W )

81 410 190 (W )

150 170 (W )(

522 120 (W l 83 425 190 (CE) 170 (W )(2)400 100 84 380 325 (W )

410 BB (W )

412 145 (W )

425 140 (W ) 425 140 (W )

65 395 135 (W )

410 150 (W )

368 IBO (W )

86 3(8 100 (W )

463 104 (W )

(1) Turkey Point 4 events of November 28 and 29, 1981.

(2) Farley 2 events of October 15, 1983.

Table A.16 Pressure data by event initiator summary.

Safety Charging / Letdown RHR Isolation RCP Restart Other Events Injection

/No Letdown Op. Errors Post Pro. Post Pre-Post Pre-Post Pro.

Post Pre-Period Com Com com com com Com com Com com com 60-79 560 495 360 (BW) 575 000 3326 525 640 560 575 590 1150 515 600(CE) 830 815 (CE) 1100 960 (CE) 1860 (BW) 840 1040 t,5 0 1100 1400 670 1300 tB0 2250 900 1000 1290 2485 80-86 330(BW) 325 399 435 350 360 360 375 J76 400(CE) 375 (CE) 368 387 395 425 380 425(CE) 400 (CE) 410 425 380 435 430 410 405 522 412 405 435 410 450 450 450 450(1) 700(2I 400(2) 463 750 1100(2)

Notest Unless otherwise ind1:ated data are for Westinghouse plants.

(1) Farley 2 events of October 15, 1983.

(2) Turkey Pointt 4 evtnts of Noventer 28 and 29, 1981.

NUREG 1326 A 20

I.

APPENDIX B SOURCE TERM EVALUATION B1 NUREO 1326

Appendix B ne low temperature overpressure transient source term SST) and SST2 values were divided by three to adjust the was obtained from NUREG/CR-4999 (Ref, B.1), using the valucs to a 50 mile radius. To confirm this assumption, Source Term Code Package (STCP) for a late core melt cornparison to recently calculated results (NUREG/CR-with containment bypass from NUREG/CR 4551 (Ref.

5015, Ref. B.4) for a 50-mile radius are provided in Table B.2), he generic value was calculated to be 9 million B.2. The scaling approach is shown to te reasonable and person rems over a 30-year exposure period for a typical differs from the newer calculations by about a factor of castern site with an assumed population density of 100 two for the SSTI releasc. For the SST2 release, there is persons per square mile over a 50 mile radius.

good agreement between the two calculations.

De release estimates were obtained for a late core melt A comparison of the base case consequences, and averted with containment bypass.

Table B.1 provides a dose, for the five methods of source term evaluation are comparison of the release fra tions employed for this provided in Table B.3. The scaling approach, based on analysis as compared to PWR 2 and PWR 5 release fission product releases, is further demonstrated by com-categories, paring the results of scaling the generic release to both the higher SSTI and the lower SST2 release consequences.

Because of the differences in the vessel failure probability and because of differences between sites, it was not The generic value assumption overpredicts the source term reasonable for the staff to select either a

typical" plant or because site specific variables such as population density, use the " average" plant for this analysis (for example, use environmental conditions, and reactor power levels are not the values and impact for the

typical" or " average" plant accounted for. De scaled results yield similar results and and multiply the results by the number of plants within a indicate the source term is less than the SSTl release group). The variation in plant specific vessel failure prob-category, as is expected based on the fission product ability, as well as the variation in site-specific conse-release fractions (Table B.1). De SST2 value is used to quence based on population density and environmental estimate the consequences of a low temperature overpres-factors, are considered in this evaluation of risk.

sure event with containment isolation failure but with the fission product release mitigating sy tems (sprays and fan To account for site specific variables, population density, coolers) functional. Table B.4 prescals the results of the environmental conditions, and reactor size, the scaling study as used in this evaluation NUREG/CR-4999 generic consequence has been scaled to the Siting Source Term data provided in NUREG,CR-A comparison of the consequence es.imates for various 2723 (Ref. B.3),

plants and release categories is prov.ded in Table B.S.

The scaled SST1, LTOP, and scaled SST2 consequerwes, For the 55 reactors considered 'n this detailed evaluation, in person-rem, are listed. As seen,.nis evaluation adjusts the average value of the meta offsite health effect (in the generic source term from NUREG/CR 4999 to account person rem) was calculated fcr both the SSTI and SST2 for site variables, including power and population release categories. (SSTI tw ing similar to PWR 2 and densities.

SST2 similar to PWR 5.)

Each plant. specific value (person rem) was then scalei to the average value, in effect yielding a conversion factor to account for site-specific variability in populatien density and environmen-REFERENCES 70R APPENDIX B tal conditions with the average value being considered as the typical, generic result.

B.1 C. Hsu c. al., " Estimation of Risk Reduction from improved PORY Reliability in PWRs," Brookhaven he plant specific consequence for the low temperature Natior.at Laboratory, NUREG/CR-4999, BNL-overpressure event was then calculated by multiplying the NUREG 52101, Final Report, March 1988, NUREG/CR-4999 generic value by the scaling factor.

B.2 M. Khatib-Rahbar et al.," Evaluation of Severe Acci-ne results obtained were then multiplied by the power dent Risks and Potential Risk Reduction: Zion Power scale facter to obtain the final low-temperature overpres-Plant,"

Brookhaven National Laboratory, sure event value in person rem. He power scale factor NUREG/CR 4551. Vol. 5, Draft Report for Com-accounts for differences in the source term resulting from ment, BNL NUREG-52029, February 1987.

different power levels between units.

B.3 D. R. Strip, Estimates of the Financial Cr 3-SSTI and SST2 source terms were also evaluated as part quences of Nuclear Power Reactor Accidents," an-of this effort. Since the NUREG/CR-2723 results were dia Laboratories, NUREG/CR 2723, SAND 82-il10, obtained for an infinite radius, the NUREG/CR-2723 November 1982.

B3 NUREG 1326

Appendix B B.4 T. L. Chu et al., *1mproved Reliability of Residual B.6 D. C. Aldrich et al., ' Technical Guidance for Siting Heat Removal Capability in PWRs As Related to Criteria Development,"

Sandia Laboratories, Resolution of Generic Issue 99,* Brookhaven Na-NUREG/CR 2239, SAND 81 1549, Decernber 1982, tional Laboratory, NUREG/CR 5015, BNL-NUREG 52121,May 1988.

B.5 L. T. Ritchie et al., ' Calculations of Reactor Acci-dent Consequences, Version 2, CRAC2 Computer

Code, Users Guide,"

Sandia Laboratories, NUREG/CR 2326, S AND81 1994, April 1982.

Table 11.1 Estimated environmental release fractions for a core mell accident resulting from a low temperature overpressure event.

Species Category Kr I

Cs Tc Sr Ru La Ce Da PWR2 0.9 0.7 0.5 0.3 0.06 0.02 0.ON 0.004 0.05 LTOP 1.0 0.12 0.088 0.17 0.05 0.005 0.0014 0.003 0.005 PWR5 0.3 0.03 0.009 0.005 0.001 0.0006 0.00007 0.00007 0.001 Table 11.2 Comparison of 50 mile radius consequences.

G199 Data G1-94 Data NUREG/CR 5015 Scaled NUREG/CR 2723 Values Values MACCS CRAC2 CRAC2 Plant Category P-Rem P Rem P Rem Zion PWR2 2.37x107 3.Nx107 2.03x107 PWR5 1.99x106 2.14x106 1.85x106 Indian Point PWR2 7.10x107 3.59x107 PWR5 7.22x106 3.16x106 Generic Site PWR2 4.38x106 g,90x106 PWR5 6.10x105 6.40x105 i

MACCS MEl.COR Accident Consequence Code System (MACCS)," D.I. Chanm et al.. Sandia National Lahorntories, NUREG)CR-4691 (to be published).

CRAC2 ' Calculations of Reactor Accident Cmsequenes Versian 2 CRAC2,* Reference B.5.

l NUREG 1326 B4 i

Appendix B

/able 11.3 Consequences evaluation comparisons (40 PORY plus 15 RHR SRV plants).

Generic Value SSTI Scaled SST2 Scaled SST1 Yalue SST2 Value P Rem Value P Rem Value P Rem P Rem P. Rem Base Case 41,900 29,600 29,500 36.400 2,600 Averted 37,400 26,700 26,700 32,900 2,300 Generic Value Case All plants M 9.0x106penun tems.

Tt:- 11,4 Consequences estimates in person rem for low temperature overpressure events (40 PORY plus 15 RHR SRV plants).

Best Estimate Low Estimate (50% Scaled SST) liigh Estimate (10% Scaled SSTI plus 50% SST2)

(Scaled SSTI) plus 90% SST2)

Base Case 16,000 29,600 5,300 Before improvements Avened Dose 14,500 26,700 4,700 For Proposed Resolution Table 11.5 Comparls(m of consequences for various sites and releases.

Release Category 50 Mile Radius PopulationU)

Plant SSTI LTOP SST2 PM Sq. Mile Per Sq. Mile Site P Rem P Rem P Rem in 1982 in 2000 Byron 16.00x106 13.20x106 1.20x106 112 175 North Anna 9.39x106 7.63x106 0.35x106 109 185 Fort Calhoun 2.30x106 1.87x106 0.26x106 91 155 Zion 20.30x106 16.50x106 i.85x106 888 1369 Calvert Cliffs 12.00x106 9.80x106 0.58x106 310 501 Indian Point 31.60x106 25.60x106 2.78x106 2099 2998 Note:(1) Esumated frum NUREG/CR-2239 (Ref. It6).

B.5 NUREG 1326 a

API'ENDIX C INDUSTRY IMPLEMENTATION COST ANALYSIS DATA BASE C.1 NUREG-1326

Appendix C The proposed resolution for 0194 would require a revi-assumed that the failed channel can te repaired in parallel sion to the plant technical specification for overpressure with the required shutdown activity without extending the protection. It is also assumed that the cooldown and shutdown duration.

heatup procedures will be revised to reflect the changes to the technical specification.

It is assumed that there are four nontefueling shutdowns per reactor year per plant. In most cases the shutdown The basis for the industry cost estimates are Abstract mode will te exited prior to the need for repeated surveil-2.2.1,

Licensec Costs for Technical Specification lance. It is assumed that 5% of the time (once every five Change, and Abstract 2.2.2, ' Industry Costs for Writing years) surveillance is required prior to restart. Further, or Rewriting Procedures,* from NUREG/CR 4627 (kef, assuming that the probability of fixing the charmel actually C.1) Abstract 6.4," Time Based Cost Adjustments," from delays the startup 5% of the time and that the channel NUREG/CR-4627 was used to escalate the costs to 1988 unavailability is 0.087 per demand, the frequency of dollars.

delayed startup is estimated to te:

Table C.1 provides the current cost estimates for these (4 shutdowns / year) x (0.05 delays) x changes for each classification. Under the complex cost (2 channels) x (0.087/ demand) x (0.05 repair delays) assumption, it is also assumed that only one of the two procedure modifications will be costed as comple'.

= 1.74x10 3 delays per reactor year.

Modification to the second procedure will be considered as simple, based on completion of a similar task.

The average annual cost of a delayed startup, based on a 4 hout delay, is estimated to be (4 hr/24 hr) x in addition to costs associated with changes to the techni-5500,000/ day x 1.74x10 3 per reactor year, or $145. At a cal specification and cooldown and heatup procedures, the discount rate of 5%, the present value of the replacement required operability of both low temperature overpressure power cost is $2,000 per PORV plant, over the average protection channels could increase the duration of a non-remaining lifetime of the PORY plants (24 years). At a refueling shutdown. The cost is associated with providing 10% discount, the present value is $1,400, If the need to additiomd replacement power. The present value of the retum an inoperable channel to service occurs once per replacement cost is evaluated based on a 5% discounting reactor year, then the cost of replacement power would te over the remaining average plant lifetime, five times greater than assumed, or $10,000 per plant (at a 5% discount rate).

Specifically, replacement power costs would be incurred if an inoperable channel were discovered during shutdown

'lle number of Westinghouse and Combustion Enginect-and if reactor restart were delayed by the repair of the ing plants considered in the consequence evaluation is 55 inoperable channel. Replacement power costs are ap-units,41 Westinghouse and 14 Combustion Engineering plicable only for plants in the PORY category for low-units. The proposed resolution for GI 94 will also impact temperature overpressure protection, where these costs on plants in the construction phase or licensed af er the t

may result from potential fnilure of the PORY actuation end of 1986. These new units total 12,11 Westinghouse mechanisms detected by required surveillance. Safety plants and one Combustion Engineering plant. It is relief valve surveillances are not required except during a assumed that these new Westinghouse units will allow refueling outage and are assumed not to extend the dura-cither PORVs or SRVs for low temperature overpressure Lion of the shutdown. It is assumed that there are four protection and will use the standard technical specification nonrefueling shutdowns each year of reactor operation.

format. They are accordingly assigned to the RilR SRV According to the current technical specifications, PORY STS category.

actuation channel circuits must be tested within 31 days after entering a mode where the PORY is required to te REFERENCE FOR APPENDIX C operable for low-temperature overpressure protection and every 31 days thereafter. Since most shutdowns will te C.1 Science and Engineering Associates, Inc., et al.,

unscheduled, surveillance is assumed to occur

" Generic Cost Estimates: Abstracts from Generic immediately before entering the low temperature over.

Studies for Use in Preparing Regulatory impact pressure protection mode, if the surveillance fails, it is Analyses," NUREG/CR-4627, June 1986.

C-3 NUREG 1326

l 1

I Appendix C Table C.1 Industry unit costs for technical specincation and procedure revisions.

Simple Complex Technical Specification

$16,000(1985)

$32,000(1985)

$17,400(1988)

$34,800(1988)

Cooldown Procedurc

$ 900(1986)

$ 3,600(1986)

$ 950 (1988)

$ 3,800(1988) lleatup Procedure

$ 900 (1986)

$ 950 (1988)

$ 950(1988)

Total (1988 $s)

$19,300(1988)

$39,600(1988) i NUREO 1326 C-4

APPENDIX D NRC IMPLEMENTATION COST ANALYSIS DATA BASE D.1 NUREO 1326

l.

l l

Appendix D

'The proposed resolution for G194 would require a revi.

The simple cost estimate is $14,200 per unit and the com.

sion to the plant technical specification for overpressure plex cost estimate is $27,400 per unit.

protection. The NRC implementation cost is primarily associated with de review and approval of de revision and the costs incurred for Federal Register notices.

REFERENCE FOR APPENDIX D 1he basis for the NRC cost estimate is Abstract 5.1, *NRC Costs for Technical Specification Change," from D.1 Science and Engineering Associates, Inc., et al.,

NUREORR 4627 (Ref. D.1). Abstract 6.4, " Time-Based

Oeneric Cost Estimates: Abstracts from Generic Cost Adjustments," from NUREGER 4627 was used to Studies for Use in Preparing Regulatory impact escalate the cost to 1988 dollars.

Analyses,* NUREO/CR-4627 June 1986.

i j

D3 NUREG 1326

APPENDIX E PRESENT VALUE COST ANALYSIS DATA BASE E1 NUREO 1326

l l

Appendix E Including costs of averted plant damage, replacement tain the equivalent 1982 costs. A constant 7.5% inflation power, and offsite costs can significantly affect the overall rate is assumed to obtain 1988 values. The present values cost benefit evaluation. In addition, the present value are then obtained based on 10% and 5% discount rates associated with these factors can serve as a measure of the using the methodology described in Section 3.5, 'Offsite worth of a proposed alternative. If two or more proposed Property," of NUREO/CR 3568. Table E.3 summarizes alternatives could achieve similar risk reduction, but with the plant costs obtained, and Table E.4 summarizes the markedly different costs, then the present value estimates present value costs associated with avoided offsite could be used to evaluate the relative worth of an alterna-damage.

tive.

The present value costs for the 40 PORV and 15 RHR ne estimated cost for cleanup and repair of a plant fol-SRV plants are provided in Table E.5. Discount rates of lowing a core damage accident is estimated at $1.2 billion.

10% and 5% are shown. The cleanup and repair and the ne present value associated with cleanup and repair is replacernent power costs are discounted over a 10-year summed over each unit based on the core damage fre-period, assuming that the plant would be returned to quency reduction estimate for the proposed resolution for operation. The offsite costs, health and property damages, GI 94 and with a 10 year period for cleanup and repair, are discounted over the remaining life of the plants. These Discounts of 10% and 5% are assurned. The methodology data represent the estimated costs associated with the *No described in Section 3.6,

'Onsite Property," of Action" attemative, that is, they are calculated for a 100%

NUREO/CR 3568 (Ref. E.1) is used for this evaluation.

reduction in the current base case risk 3.24x106 per reactor year frequency of a through wall crack leading to ne present value for replacement power following a core core damage and fission product release, damage accident is estimated based on the summation of the discounted cost for each unit. The replacement power costs, by region, used in this analysis are provided in REFERENCES FOR APPENDIX E Table E.1 (taken from NUREG/CR 4568 Ref. E.2),

he methodology described in Section 3.6, "Onsite E.1 S.

W.

Heaberlin et al.,

"A Handbook for Property," of NUREG/CR 3568 (Ref. E.1) is used for this Value/ impact Assessment," Pacific Northwest evaluation. The estimated present value costs for cleanup Laboratorier, NUREO/CR 3568, PNL-4646, Decem-and repair and for replacement power are provided in ber 1983.

Table E.2. Discounts of 10% and 5% are assumed.

E.2 J. R. Ball, "A Handbook for Quick Estimates: A in addition to onsite property damage, offsite costs can be Method for Developing Quick Approximate Es-incurred as a result of the accider,t. Both offsite health and timates of Costs for Generic Actions for Nuclear offsite property damages are evaluated. De present value Power Plants,* Argonne National Laboratory, costs are obtained over the remainder of plant life, based NUREO/CR 4568, AN1/EES TM 297, April 1986.

on the expected reduction in accident frequency resulting from the proposed resolution for GI 94.

De plant.

E.3 D. R. Strip, " Estimates of the Financial Conse-specific costs are obtained from NUREG/CR 2723 (Ref.

quences of Nuclear Power Reactor Accidents," San.

E.3) by first calculating the damage values by removing dia National Laboratories, NUREO/CR 2723, the discount factor (4% used !n NUREG/CR 2723) to ob.

S AND821110, November 1982.

E-3 NUREO 1326

)

I

Appendix E Table E.1 Daily replaecment power cost estimates by region.

NERC Regions Rate ($/kW.hr)

DailyCost )

0 (1984 $s)

(1988 $s)

ECAR East Central Area Reliability 0.020 457,000 Coordination Agreement ERCOT Electric Reliability Council 0.035 801,000 ofTexas MAAC Mid Atlantic AreaCouncil 0.030 686,000 MAIN Mid-America Interpool 0.024 549,000 Network 1

MARCA Mid Continent AreaReliability 0.037 845,000 Coordination Agreement NPCC Northeast Power Coordinating 0.023 526,000 Council SERC Southeastem Electric 0.011 252,000 Reliability Council

{

SPP Southeast Power Pool 0.N0 815,000 WSCC Western Systems Coordinating 0.024 549,000 Council Average 0.026 594,000 Note: (1) Assurne $% year inflation in costs. Cost opplies to an 1120 M(e) unit with an average capacity factor of 0.70. Specific. plant costs are multiplied by power scaling factor.

Table E.2 Estimated present value costs for avoided onsite property damage (40 PORY plus 15 RilR SRV plants).

I 10% Discount 5% Discount Over 10 Years Over 10 Years Cleanup and Repair

$1,200,000

$2,200,000 Replacement Power

$1,300,000

$2,400,000 i

Total

$2,500,000

$4,600,000 I-l l

1 NUREG 1326 E-4

i Appendix E Table E.3 Comparison of offsite property damage costs, SST1 Offsite Cost SST2 offsite Cost Year of Heale Property Health Property Plant Operation

($)

Average 1980 4.42x108 2.78x109 1.68x107 4.30x107 Indian Point 1974 24.10x108 14.20x109 7.29x107 17.50x107 1

Zion 1973 16.40x108 7.40x10' 6.06x107 II.70x107 Palo Verde 1984 1.0lx108

!.30x109 0.70x107 1.57x107 i

Table E.4 Estimated present value costs for avoided offsite health and property damage (40 PORY plus 15 RHR SRY plants).

Based On SSTI Costs Bud On SST2 Costs Over Plant Life Over Plant Life 10% Discount 5% Discount 10% Discount 5% Discount Offsite Health

$ 640,000

$ 970,000

$ 23,000

$ 36,000 1-Offsite Property

$4,060,000

$6,180,000

$ 63,000

$ 86,000 Total

$4,700,000

$7,150,000

$ 86,000

$ 122,000 E.5 NUREG.1326

l l

Arnendix E Table 1:.S Present value cost summary for 4010kV and 15 RilR SRV plants (based on base case frequency.

total value of averted damages).

Present Value Present Value at 10% Discount at 5% Discount (51,0(0,0COs)

($1,000,0(Os)

Averted Cost Factor ICRV RiiR SRV Total IORY RIIR SRV Total l

~

Replace Power over 10 years 0.87 0.46 1.33 1.64 0.86 2.50 3

Cleanup / Repair over 10 years 0.98 0.41 1.39 1.87 0.79 2.66 l

SSTI ilcalth over plant life 0.65 0.06 0.71 0.98 0.09 1.07 SSTI Property over plant life 3.92 0.57 4.49 5.94 0.88 6.82 SST2 llealth over plantlife 0.023 0.003 0.026 0.036 0.004 0.040 SST2 Property over plant life 0.059 0.010 0.059 0.090 0.016 0.106 Total 11est Estimate 3.2 0.8 4.0 7.0 2.2 9.2 Totalliigh Estimate 6.4 1.5 7.9 10.4 2.6 13.0 Total Low Estimate 0.6 f2 0.8 4.0 2.1 6.1 i

NUREO.1326 E6

t

,N i

R FDN 335 U.S. NUCLE AR REGULQTORY COMMIS5lON 1 kl I,' btP EEN' BIBUOGRAPHIC DATA SHEET

[

~

ISee trutructrons on t'st reverse) r.m Lt ANo susin tt NUREO 1326 Regulatory Analysis for the Resolution of Generic issue 94, 3

DATE REPORT PUBLisHf D i

' Additional Low. Temperature Overpressure Protection for l

Light Water Reactors.

December 1989 1

4. FIN OR GR ANT Nuvele UlHORtS) 6 T YPE OF REPORT Edward D. Throm Regulatory Analysis

1. Pt RIOD CQvt R L D ronerne Dero 8 Pt,R..F O.,RMI.N.G..O.RG ANIZ AT ION ~ N AME AND ADDR L 5s ut Nac. preen

0rvwon. Otr ce er Repon, v.s hucker nepusocory comanwon, ans avaams.w,,s ri controcto

,,o

.,.. s,no

.Divisloc of Safety issue Resolution Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, D.C. 20555

' 9. sPONsO. RING.. orc ANIZA1lON - N AML AND ADDRESS no Nac, tre kn* *s above*'. n'ro*r**cror. orov*ar Nac posisoo$.

er Renoon, v s hacner orrevoarauw commaan.

e u

.aso a Same as above,

10. sVPPLEMENT ARY NOTES

{

i t. ABsT R ACT f?ou words or *p>

t<

h Low temperature overpressure protection (LTOt') is required in pressurized water reactors (PWRs) to i

. provide protection ~ against brittle reactor pressure vessel failure following an anticipated event. Typically these events are a result of cither mass imbalance or energy input transients. The significanca of thesc events is heightened during water-solid operations.

LTOP is required in the shutdown modes of lc operation, Mode 4 - Hot Shutdown, Mode 5 Cold Shutdown, and Mode 6 Rcfueling with the reactor

. vessel head bolted down. While operating in Modes 5 and 6 and.with the reactor coolant temperaturc below 200"F, there are no technical specifications for containment integrity. The consequences of an unmitigated low-temperature overpressure event can be significant as a result of either con:ainmei.t bypass or failure of containment to isolate following reactor pressure vessel failure.

i This report presents the regulatory analysis for Generic Issue 94, " Additional Low-Temperature Overpressure Protection for Light-Water Reactors.' It includes (1) a summary of the issue: (2) the proposed technical resolution, (3) alternative resolutions considered by the Nuclear Regulatory Co'r, mission (NRC),(4) an assessment of the benefits and cost of the alternatives considered with additional emphasis

.on the recommended resolution, (5) the decision rationale, and (6) the impacts and relationships between 0194 and other FRC programs and requirements.

it. u y won osnscR:e rOss a a -ore, or,~.- <= ~+,uar ~~a~ wr-e r- -r's

" ^ ' ' ' ' ' ' ' " ' ' " ' ' ' " '

Unlimited

-Low Temperature Overpressure Transient,' Ocneric Safety issue, a m.

,, m.s,a., s,,

Value/ Impact Analysis, Probab!'istic Risk Assessment, PWRs, a.e Overpressure Mitigation System, Pilot Operated Relief Valves, Unclassified Safety Relief Valves, Public Risk Unclassified a NuveesDeeAots It> PHILE 3

NRC FQHM JJ82 (? S 91 1

-.,_g. ;-

. UNITED STATES s:'aciat rounin ctass casi NUCLEAR REGULATORY COMMISSION toSicot 6 'EtS DA$

WASHINGTON, D.C. 20555 PERMtT % G g7 OFFICIAL BU$fNESS PENALTY FOR PRlVATE USE. 6300 i

~~.;

i4 i

l I'

i

--

ML20055E089

Contents

Text

Title:

SUMMARY

SUMMARY

.

SUMMARY

Navigation menu