Text

Interim Report, Regulatory Effectiveness of Station Blackout Rule
	ML20217D085
Person / Time
Issue date:	09/30/1999
From:	Raughley W; NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	;
Shared Package
ML20217D081	List: ML20217D078; ML20217D085;
References
	NUDOCS 9910140177
	Download: ML20217D085 (59)
	v • d • e

e

' :9 f

i e

Interim Report-

~ Regulatory Effectiveness of the Station Blackout Rule Prepared by:

William S. Raughley

-1 l

4 1

l 1

1 l

i

)

i Regulatory Effectiveness Assessment and Human Factors Branch Division of Systems Analysis and Regulatory Effectiveness l

. Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission I

i 9910140177 990930 )

PDR ORQ NREB l p

j ABSTRACT The NRC Office of Nuclear Regulatory Research is reviewing regulations, starting with the station blackout (SBO) rule, to determine if the requirements are achieving the desired outcomes and if the expected costs were realistic. This initiative is part of an evolving program to address regulatory effectiveness to make Nuclear Regulatory Commission (NRC) activities and decisions more effective, efficient, and realistic. This report presents an evaluation of the effectiveness of the SBO rule based on a comparison of the regulatory expectations to the corresponding outcomes. A set of baseline expectations was established from the SBO rule and related regulatory documents in the areas of coping capability, risk reduction, emergency ,

diesel generator reliability and unavailability levels, and value-impact. The report concludes i that although some issues may need attention, the SBO rule was effective and the industry and the Nuclear Regulatory Commission costs to implement the SBO rule were reasonable considering the outcome.

I 1

l l

I i

iii j

t! ,;

h }.

l i . ,

^)

CONTENTS LA B S T RA C T . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii s AB BR EVI ATIO N S . . . . . . . .. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vil EX E C U WE S U M M A RY . . . . . . . . . . . . . . . .x. . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . ix o

1 h INTRODU CTION . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .......1 J

1 72 B A C K G R O U N D .................. ....... . .... . ........ .. .. .........1 2.1 Station Blackout and the Station Blackout Rule . . , , . . . . . . . . . . . . . . . . . . . . . . . 1

' 2.2 . Studies Provide Technical Basis for Regulatory Requirements and Guidance . . . . . 2 2.3 Public and Industry Comment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.4 Resolution of Generic Safety issue B-56 . . . . . . . . . . . . . . . . . . . . .. . . . . . , , . . . . . 4 2.5 . Reg ulatory Follow-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3 ' ASSESSMENT OF THE STATIO'N BLACKOUT RULE , . . . . . . . . . , . . . . . . . . . . . . . . . 6 3.1i Method for Assessing Regulatory Effectiveness of the Station Blackout Rule . . . , . 6 I I

3.2 Comparison of Expectations and Outcomes . .. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . 8 3.2.1 ' Risk Reduction . . . . . . .. . . . . . . . . . . . . . . . . . . . ....................8 3.2.2L Emergency Diesel Generator Reliability and Unavailability . . . . . . . . . . . . . 11 3.2.3 : Minimum Acceptable Coping Capability, Plant Procedures, Training, and i M od ifica tio n s . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . < 13 1 3.2.4 ' . Value-impact Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . 14 3.2.5 Insights From Operating Experience Reviews . . . . . . . . . . . . . . . . . . . . . . 17 4 C O N C LU SION S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0

-5 REFERENCES .....................................................22 l

TABLES j 1 - Summary of Station Blackout Rule Expectations and Outcomes . . . . . . . . . . . . . . . . . . 7

.'2 ' Comparison of the Number of Plant Units With a Station Blackout Core.

Damage Frequency Before and After Station Blackout Rule implementation . . . . . . . . . 8 3 Probabilistic Risk Assessment / Individual Plant Examination Sensitivity Analyses . . . . . IC l 4 Industry Emergency Diesel Generator Trigger Failura) Rates . . . . . . . . . . . . . . . . . . . . . 12 I

5. : Station Blackout Rule Value-impact Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 6: Dominant Station Blackout Risk Factor Trends - Offsite Power System . . . . . . . . . . 19 !

L7 Analysis of Loss of Offsite Power Event Recovery 1968-1996 ............... .... 19 i

i I

v l

APPENDICES A -' Plant-Specific Station Blackout Information by Reactor Type and Operating Status

-.8, Comparison of Selected Station Blackout Characteristics

'^

Station Blackout Rule Activity and Modification Summary

, D . Jperating Events E : Emergency Diesei Generator Reliabilities and Industry Trigger. Failure Rates TABLES

.A-1; - Operating Pressurized-Water Reactors A-2: , Operating Boiling-Water Reactors A-3 Reactors No Longer Operating e ' D-1 Losses of Offsite Power Since 1990 Having Recovery Times Greater Than 4 Hours D-2 ' Station Blackout Challenges e

s N

I t

i 0

.vl ii h

~

9 ,y~;. :h .-

t ,. g, ll

.,+':- l0.~

4 "

. ' ABBREVIATIONS j

, Aac alternate ac 4

AOT! allowed outage time -

" ' '.CDF - icore damage frequency.

CFRL Code of FederalRegulations

? EDG~- . emergency diesel generator-EPS emergency power system '

. GSI'-' -l generic safety _ issue

~ lN 'information notice s :INEEL Idaho National Engineering and Environmental Laboratory (formerly INEL) - ;

,a llPE: o iindividual plant examination' l LER licensee event report a

-LOOP ; loss of offsite power - :

. l l

MOOS : maintenance out of service !

NEl Nuclear Energy Institute (formerly NUMARC and USCEA)

'NRC. .

. Nuclear, Regulatory Commission, U.S. ,

y =-NUMARC Nuclear Management and Resources Council (NEI) ll PRA probabilistic risk assessment ,

i RES ' Nuclear Regulatory Research, Office of (NRC) - !

! RG regulatory guide RY reactor-year ;

SBO. station blackout i

!' USl- Unresolved Safety Issue l

l 1

ij l < ,

- vil a

h _, ,

L ., l

.a _ .

rt EXECUTIVE

SUMMARY

Th6 NRC Office of Nuclear Regulatory Research is reviewing regulations, starting with the station blackout (SBO) rule, to determine if the requirements are achieving the desired outcomes and if the expected costs were sealistic. This initiative is part of an evolving program to address regulatory effectiveness to make Nuclear Regulatory Commission (NRC) activities and decisions more effective, efficient, and realistic. As part of ibis program, the staff noted in SECY-97-180," Response to Staff Requirements Memorandum of May 28,1997, Concerning Briefing on IPE [ individual plant examination) Insight Report," August 6,1997, that RES activities for assessing regulatory effectiveness would determine whether additional generic action is warranted, assess whether any new generic safety issues warrant attention, and assess the need for plant-specific actions based on results of individual plant examination. l l

The NRC designated the issue of SBO, which is a loss of all ac offsite and onsite power j concurrent with a turbine trip, as Unresolved Safety Issue A-44 in 198C to determine the need ;

for additional safety requirements since SBO can be a significant contributor to core damage frequency. In 1988, the Commission concluded that additional SBO safety requirements were justified and issued the SBO rule (10 CFR 50.63) to provide further assurance that a loss of both offsite and onsite emergency ac power systems would not adversely affect public health i and safety. !

l A regulation is considered to be effective if the expectations are being achieved and the costs are reasonable. To assess the regulatory effectiveness of the SBO rule, regulatory requirements were used as a baseline for comparison to the actual outcomes. The baseline provided a set of expectations in the areas of coping capability, risk reduction, emergency l diesel generator (EDG) reliability and unavailability levels, and value-impact. The companson of each outcome to a corresponding baseline requirement determined if the expectations were l achieved and if there were any differences between the expectations and the actual outcomes associated with the SBO rule that may require staff attention. The report contains the following detailed conclusions:

Comparison of the SBO rule expectations and outcomes indicates the SBO rule was effective, and industry and NRC costs to implement the SBO rule were reasonable considering the outcomes. Implementation of the SBO rule resulted in changes to procedures, training, EDG performance monitoring, hardware modifications, and the addition of diesel and gas turbine power supplies. These outcomes provide the plants with SBO coping capability, reduce the risk, and make the plants more tolerant to loss of offsite or onsite power as follows:

+ The reduction in the estimated mean SBO core damage frequency (CDF) was approximately 3.2E-05 per reactor-year (RY), slightly better than the 2.6E-05 per RY l

expected after implementation of the SBO rule. As a result of the improvements made due l to the SBO rule, the likelihood of core damage given an SBO was also significantly reduced, l in addition, after implementation of the SBO rule, mon plants have a lower SBO CDF than i expected. Further, SBO has been made a relatively small contributor to core melt as there are only five plants at which (a) SBO contributes more than 10 percent to the total CDF and l

(b) the CDF exceeds 1.0E-04 per RY. Also, the SBO rule resulted in the plants with the greatest number of losses of offsite power (LOOPS) from plant-centered events and ix

O I

extremely severe weather conditions having made the most improvement, such as access i to an alternate ac power supply. Consequently, these plants have relatively low SBO CDF.

Implementation of the SBO rule resulted in all plants having (1) a 4- or 8-hour coping capability; (2) established SBO coping and recovery procedures; (3) completed training for these procedures; (4) established an EDG reliability program +3t generally maintains EDG reliability levels at or above 0.95; and (5)imphmented modifications as necessary to cope with an SBO.

'= The operating experience indicates that the SBO rule provided additional defense in depth, given that for many licensees, SBO is a significant contributor to the overall risk, and SBO rule-related hardware and procedures have actually been used to mitigate event consequences and provide additional protection. Additional defense-in-depth is also provided by the SBO rule should the performance of the ac power system unexpectedly degrade as a result of deregulation c f the electric power industry or if it takes longer than expected to recover offsite power fohowing extremely severe weather conditions.

Comparison of the value-impact expectations to the corresponding outcomes indicates that the value-impact was within the expected range of reductions in public dose-per-dollar of cost. As expected, there was wide variation ni plant-specific values and impacts because the SBO rule provided flexibility. Not expected was the addition of costly power supplies that account for 75 percent of the industry cost impact. This factor explains why the NRC value-impact analysis underestimated the cost by a factor of 3. However, it appears licensees justified the cost for the power supplies based on offsetting monetary benefits, such as increased EDG allowed outage times for operating flexibility or meaningful risk reductions. Thus the value was also underestimated. The remaining 25 percent of the industry cost appears reasonable considering the outcomes: known coping capabilities, industry risk reduction from plant-specific procedural and hardware enhancements, and additional defense-in-depth.

Comparison of the SBO rule expectations to the corresponding outcomes concluded that the following issues warrant attention:

1. The 0.975 EDG reliability has not been maintained at levels equal to or above levels selected in RG 1.155, " Station Blackout," August 1988, that were used for the determination of the required coping time. In addition, EDG unavailability due to maintenance out of service was found to be 0.03, not 0.007 as assumed in RG 1.155, and indicates that plants will have great difficulty in meeting 0.970 reliability goals.

2. The 0.975 EDG target reliability goal may be unrealistic. Opeating data shows (a) the underlying EDG reliability appears to be 0.95 as the EDG tarp;t reliability groups of 0.95 and 0.975 are nearly the same being 0.954 and 0.956, respe;tively, and 2) achieving the EDG target reliability levels of 0.975 is not probable beine,17 percent and 54 percent with and without maintenance out of sentice, respectively.

3. There may be inconsistencies between the SBO P.G 1.155 and RG 1.160," Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," June 1993, reliability and unavailability criteria. The induury trigger failure rates recommended for EDG reliability performance monitor!ng and goa' setting in documents that are endorsed by x

c RG 1.160, do not reasonably assure the RG 1.155 reliability levels will be achieved. NRC processes for issuing regulatory guides need to ensure the absence of competing expectations. This will be important as the NRC continues to transition to a risk-informed performance based regulations.

4. Some licensee probabilistic risk assessment LOOP initiating frequencies appear to be a factor of 3 to 10 lower than the corresponding value determined from the operating experience.

l

5. Regulatory Guide 1.93, " Availability of Electric Power Sources," December 1974, forms the basis for technical specifications in the area of ac onsite and offsite power supply availability I by requiring shutdown of the reactor following extended ac power supply unavailability.

This criteria may adversely affect the SBO CDF.

6. Events continue to identified that the availability of some alternate ac pwer suoplies is

. limited by dependancies offsite or onsite power supplies that could lead to core melt during an SBO event.

l 4

l xi

1 1 INTRODUCTION l The NRC's Office of Nuclear Regulatory Research (RES)is reviewing regulations, starting with the station blackout rule (SBO), to determine if the requirements are achieving the desired outcomes and if the expected costs were realistic. This initiative is part of an evolving program to address regulatory effectiveness to make Nuclear Regulatory Commission (NRC) activities and decisions more effective, efficient, and realistic. I In 1980, the Commission designated the issue of SBO as Unresolved Safety issue (USI) A-44,

" Station Blackout," to determine the need for additional safety requirements since SBO can be

~

a significant contributor to core damage frequency (C ? -) and, with the consideration of containment failure, SBO can represent an important contributor to reactor risk. On June 21, 1988, the NRC concluded that additional SBO safety requirements were justified and issued the SBO rule as stated in Federa/ Register Notice 23203 (Title 10 of the Code of Federal Regulations Section 50.63 [10 CFR 50.63),

Loss of all alternating current power")[Ref.1].' The amendment was intended to provide further assurance that a loss of both offsite and onsite emergency ac power systems would not adversely affect public health and safety.

In May 1997, the staff briefed the Commission on the individual plant examination (IPE) insight report, NUP.EG-1560, " Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance," December 1997 [Ref. 2]. The report concluded that SBO remains a dominant contributor to risk at many plants, even after implementation of the SBO rule; that SBO frequencies are above 1E-05 per RY for many plants; and that, given the results, further ;

investigation of strategies for reducing SBO frequencies appears warranted, in a Staff Requirements Memorandum (SRM), " Briefing on IPE Insight Report,"May 28,1997, the Commission asked the staff to provide a scope and schedule to use the IPE results to assess regulatory effectiveness in resolving major safety issues. The staff responded in SECY-97-180, l

" Response to Staff Requirements Memorandum of May 28,1997, Concerning Briefing on IPE !

Insight Report," May 28,1997 [Ref. 3], noting that the Probabilistic Risk Assessment (PRA)

Implementation Plan was tracking activities to assess the regulatory effectiveness of major safety issue resolution efforts, including SBO SECY-97-180 noted that these activities would determine whether additional generic action is warranted, assess whether any new generic safety issues (GSis) warrant attention, and assess the need for plant-specific actions based on i i

IPE results.

'2 BACKGROUND' ,

I 2.1 Station Blackout and the Station Blackout Rule j i

L Federa/ Register Notice 23203 amended the regulations to define an SBO and add the SBC rule. In 10 CFR 50.2," Definitions," the staff defines SBO as the complete loss of ac electric l , power to the essential and nonessential electric switchgear buses in a nuclear plant (i.e., loss of the off.1te electric power system concurrent with a turbine trip and unavailability of the .;

emergency ac power system). In 10 CFR 50.63, the SBO rule requires that nuclear power plants be capable of withstanding an SBO for a specified duration and of maintaining core cooling during that period. The specified duration would be determined for each plant on the basis of a comparison of the individual plant design with factors that have been identified in 1

(

L f

NRC technical studies as the main contributors to the risk of core melt resulting from an SBO.

These risk factors are identified in the SBO rule as (1) the redundancy of the onsite emergency ac power sources, (2) the reliability of the onsite emergency ac power sources, (3) the frequency of loss of offsite power (LOOP), and (4) the probable time needed to restore offsite power.

The SBO rule requires licensees to propose and justify an SBO coping duration based on the four risk factors; maintain highly reliable onsite emergency ac electric power supplies; ensure that the plants can cope with an SBO for some period of time based on the probability of occurrence of an SBO at the site as well as on the capability for restoring power for that site;

- develop procedures and trinir 'o restore offsite and onsite emergency ac power should either become unavailable; and, n ner ory, make modifications necessary to meet the SBO rule requirements. The SBO rule also requires that the staff complete a regulatory assessment and notify the licensees of the staff's conclusions regarding the licensees' response to the SBO rule.

2.2 Studies Provide Technical Basis for Regulatory Requirements and Guidance The SBO rule evolved from the results of several plant-specific probabilistic safety studies; operating experience; and reliability, accident sequence, and consequence analyses completed between 1975 and 1988. In 1975, WASH-1400, " Reactor Safety Study," 1975, indicated that SBO could be an important contributor to the total risk from nuclear power plant accidents. This study concluded that if an SBO persists for a time beyond the capability of the ac-independent systems to remove decay heat, core melt and containment failure could follow, in 1980, the Commission designated the issue of SBO as USl A-44 and the staff completed several technical studies to determine if any additional safety requirements were needed.

NUREG-1032, " Evaluation of Station Blackout at Nuclear Power Plants," June 1988 [Ref. 4),

integrated the findings of the technical studies completed for USI A-44. NUREG -1032 presented the staff's major technical findings for the resolution of USl A-44, and provided the basis for the SBO rule and the accompanying RG.

The NUREG-1032 technical findings follow: (1) the dominant factors affecting the likelihood of SBO accidents at nuclear plants are (a) the LOOP frequency, (b) the time to restore offsite power following its loss, and (c) emergency diesel generator (EDG) reliability and redundancy; (2) the important characteristics of SBO are as follows: (a) the variability of estimated SBO event likelihood is potentially large, ranging from approximately 1.0E-05 to 1.0E-03 per reactor-year (RY) (a " typical" estimate is on the order of 1.0E-04 per RY), (b) the estimated frequency of SBO events that result in core damage or core melt can range from approximately 1.0E-06 to more than 1.0E-04 per RY (a " typical" CDF estimate is on the order of 1.0E-05 per RY), (c) the capability to restore offsite power in a timely manner (fewer than 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months ) can have a significant effect on accident consequences, (d) the redundancy of onsite ac power systems and the reliability of individual onsite emergency power supplies have a large influence on the likelihood of SBO events, (e) the capability of decay heat removal systems to cope with long duration blackouts (in excess of 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months ) can be a dominant factor influencing the likelihood of core damage or core melt for the accident sequence, (f) containment failure as a result of overpressure may follow an SBO-induced core melt. Smaller, low design pressure containments are most susceptible to early failure (possibly in fewer than 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months ). Some larger, high design pressure containments may not fail as a result of o"erpressure, or if they do fait, the failure time could be on the order of a day or more.

, 2

c c , 1 In March 1986, tha NRC issued draft Regulatory Guide (RG) 1.155, " Station Blackout," [Ref. 5) )

. which presented an acceptable method to comply with the SBO rule based on plant specific j

^

characteristics and the dominant risk factors from NUREG-1032. Specifically, draft RG 1.155 contains guidance on (1) selecting target reliability for EDG of 0.95 per demand or 0.975 per demand based on plant characteristics, (2) establishing an EDG reliability program to maintain the selected EDG target reliability level, (3) developing procedures and training to cope with an SBO, (4) selecting a plant-specific minimum acceptable SBO duration capability of either 2,4, l 8, or 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months based on plant-specific considerations, (5) evaluating SBO capability based on !'

- the selected duration capability, and (6) completing modifications to cope with an SBO. By analyzing the effect of variations in offsite and onsite ac power system design and plant location on risk from SBO accidents in NUREG-1032, the staff was able to develop an approach for ;

licensees to establish commitments on a reasonably consistent level commensurate with risk.

This provided flexibility for implementing the SBO regulatory requirements and consequently, !

on a plant-specific basis, the values and impccts associated with the SBO rule were expected to vary significantly.

RG 1.155 provides for a licensees to select and maintain an EOG target reliability level of either

- 0.95 or 0.975 per demand (measured as a probability of failures from test and unplanned l demands) and assumed EDG unavailability due to test and maintenance out of service (MOOS) :

while the reactor is operating is 0.007 (measured as a percentage of the hours the EDG is unavailable to the total hours in a time period of consideration). The EDG refers to an EDG train which includes a single engine, a generator, an output circuit breaker, and support subsystems necessary to power and sequence electricalloads on tne vital bus. The RG 1.155 reliability levels were based on information in NUREG-1032 that generally showed EDG reliability to be 0.98 or better. The RG 1.155 MOOS level was based on NUREG/CR-2989, ,

" Reliability of Emergency AC Power Systems at Nuclear Power Plants," June 1983, that found i MOOS to be an average of 0.006 with a range of 0-0.037. ,

, The technical basis of RG 1.155 characterizes the safety mission cf the EDGs as the ability to j perform upon demand. This is measured as an EDG reliability based on success per demand. i This reliability has an unavailability contribution from MOOS s'ince the EDG cannot perform its safety mission if demanded when out of service. Consequently, EDG MOOS is an important consideration since the plant risk is potentially higher due to the possibility of a demand while the EDG is unavailable. EDG unavailability measurement can be based on the hours the EDG is unavailable or on the number of failures while the EDG is in MOOS. Both measures are unbiased estimates of EDG unavailability and are comparable so long as both measures are given the same considerations, (i.e., both consider MOOS). High EDG target reliability requires that MOOS be small; RG 1.155 states that as long as the MOOS unavailabi!ity is not excessive, the maximum EDG failure rate (0.975) would result in overall reliability for the emergency power system (EPS).

2.3 Public and Industry Comment in March 1986, a Notice of Proposed Rulemaking on SBO was published in the Federal Register (51 FR 9829), " Station Blackout." That notice invited public comments on a proposed SBO rule, draft RG 1,155, and NUREG-1109," Regulatory /Backfit Analysis for the Resolution of Unresolved Safety issue A-44,' Station Blackout'," June 1988 [Ref. 6). Among the 53 letters that were received commenting on the SBO rule was a cr;tique of the backfit analysis that pointed out numerous errors, omissions, and inaccuracies. In addition, an industry comment 3

4 l

stated that the cests wuuld be much higher than those calculated by the NRC. The NRC addressed the cost concems in a report, " Response To Industry Comments On Station Blackout Cost Analysis (NUREG/CR-3840)," Science and Engineering Associates, Inc., and L Mathlech, Inc.. . November 1986, by increasing most of the cost items by 20 to 140 percent.-

The other comments were addressed in the final proposed resolution to USl A-44 that was reviewed and approved by the Committee to Review Generic Requirements in May 1987 and the Advisory Committee on Reactor Safeguards in June 1987.

In November 1987, the Nuclear Management and Resources Council (NUMARC)

(subsequently renamed Nuclear Energy Institute (NEI) ) submitted NUMARC 87-00, " Guidelines and Technical Bases for' NUMARC Initiatives Addressing Station Blackout at Light Water Reactors," No'vember 1987 [Ref. 7], as an alternative to comply with the SBO rule.

NUMARC 87-00 content followed RG 1,155 and added very prescriptive, practical guidance to help the industry implement the RG 1.155. NUMARC 87-00 addressed risk reduction by requiring the following: (1) taking action to reduce the risk if the licensees fell into the 8- or

. 16-hour coping category; (2) establishing procedures to cope with an SBO, restore ac power following an SBO, and to prepare for severe weather; (3) reducing EDG cold fast starts for testing; and (4) monitoring AC power unavailability by providing data to the Institute of Nuclear Power Operations on a regular basis. By reference in RG 1.155, the staff concluded that NUMARC 87-00 (dated November 1987) contains guidance acceptable to the staff for meeting the SBO rule. Since 1989, the industry has used the EPS unavailability as an industry safety system performance indicator. The industry defines and calculates unavailability as the

' average percentage of the total hours the EDGs trains of the EPS are unavailable.

The staff issued SECY-88-22, " Final Station Blackout Rule, USI A-44," January 21,1988,

--[Ref. 8] to obtain Commission approval for publishing the Notice of Final Rulemaking on the subject of SBO. In SECY 88-22, the staff recommended that the Commission issue the SBO

. rule, NUREG-1032, RG 1,155, and NUREG-1109which documented the evaluation of five

. attematives to close USl A-44 and the value-impact analysis of the proposed SBO rule. In SECY-88-22, the staff stated that USI A-44 was related to such other GSis as GSI B-56,

' " Diesel Generator Reliability"; USI A-45, " Shutdown Decay Heat Removal Requirements";

GS1 B-23, " Reactor Coolant Pump Seal Failures "; and GSI 128,

Electric Power Reliability." in SECY-88-22, the staff states that any additional requirements or guidance contained ~in the resolutions of these GSis must be consistent with the requirements of the SBO rule and were not expected to cause licensees to revise analyses, procedures, or equipment that were changed to comply with the SBO rule.

2.4 - Resolution of Generic $afety Issue B-56 The NRC planned that tha resolution of GSI B-56 would detail an EDG reliability program consistent with RG 1.155s RG 1.15.5'provides for a reliability program that includes reliability monitoring and a maintenance program that ensures that the EDG target reliability level is being achieved. In SECY-93-044," Resolution of Generic Safety issue B-56, Diesel Generator Reliability," February 22,1993, the staff recommended incorporating (by reference or example)

EDG maintenance unavailability operating experience and the NUMARC EDG reliability program (Appendix D and E of NUMARC 87-00, Rev.1," Guidelines and Technical Bases for NUMARC Initiatives Addressing Station Blackout at Light Water Reactors," August 1991

[ Pef 9]), into the RG and NUMARC guideline being developed for the Maintenance Rule.

4 I

y.

a Appendix D of NUMARC 87-00, Rev.1, employs the use of trigger values to indicate when EDG performance problems exist such that extensive corrective action is necessary.

Consistent with the resolution of B-56, RG 1.160, " Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," June 1993 [Ref.10], endorsed NUMARC 93-01, " Nuclear Energy Institute Industry Guideline for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," May 1993 [Ref.11], which in turn states that NUMARC 87-00, August 1991, trigger values may be used for performance monitoring and goal setting. In addition RG 1.160, states EDG reliability performance criteria or goals selected for implementing the intent of 10 CFR 50.63 could be monitored through use of triggers. RG 1.160, Rev.1, January 1995 [Ref.12], de!sted the text reference to the trigger values.

However, the RG resolution to B-56 was maintained through reference to NUMARC-93-01 in RG 1.160, Rev.1 and in RG1.160, Rev. 2, March 199~7 [Ref.13]. All revisions of RG 1.160 also stated that the RG 1.155 EDG target reliability levels could be used as performance criteria or goals.

2.5 Regulatory Follow-up The SBO rule requires that the NRC staff complete a regulatory assessment and notify the licensees of the staff's conclusions regarding the licensees' response to the SBO rule. The NRC completed safety evaluations for each plant that can be found in the NRC Public Document Room under each plant's docket number.

To assess the industry's compliance in implementing the SBO rule, the NRC completed eight pilot inspections by October 1994 using Temporary Instruction 2515/120 " inspection of Implementation of Station Blackout Rule." The objective of Tl 2515/120 was to verify the adequacy oflicensee programs, procedures, training, equipment and systems, and supporting documentation for implementing the SBO rule. The inspectors found that, overall, the licensees for the eight sites had satisfactorily implemented their commitments for conforming to the SBO rule. The inspections uncovered minor weaknesses in plant SBO documentation, primarily in the areas of electrical calculations and procedures that were considered to have no effect on !

SBO mitigation and to be insignificant. On the basis of these inspections, the staff concluded that it need not conduct additional team inspections to verify licensee implementation of the 1 SBO rule. However, additional discretionary SBO inspections could be conducted to verify that I the licensees have taken appropriate actions to comply with the CBO rule. I SBO rule assessments have been completed under other NRC programs and these were used in this report. For example, the NRC staff evaluated the impact of the SBO rule on CDFs in NUREG-1560. In SECY 97-180," Response to Staff Requirements Memorandum of May 28, 1997, Concerning Briefing on iPE Insight Report," dated August 6,1997, the staff also evaluated the industry's average cost per person-rem averted in satisfying the SBO requirements.

5

'3 ' ASSESSMENT OF THE STATION BLACKOUT RULE The ' scope of the SBO rule assessment is to determine if the SBO rule is effective, if costs

.~ attributed to the action were reasonable in view of the outcome, and if there are areas that may need the staff's attention to improve the effectiveness of the SBO rule.

1 Consistent with the original scope of the SBO rule, this assessment does not specifically address' other related GSis such as GSI B-23, " Reactor Coolant Pump Seal Failures." However, some licensees have essentially addressed GSis such as GSI B 23 in their PRA/IPE results that were used in this assessment.

3.1 Method for Assessing Regulatory Effectiveness of the Station Blackout Rule For the purposes of this assessment, the regulatory documents present the expectations (desired outcomes)in terms of specific objectives, requirements, and cuidance. The regulatory documents are considered to be effective if the expectations are being achieved. A difference between the expectations and the actual outcome may identify an area that needs staff attention.

To assess the regulatory effectiveness of the SBO rule, the expectations as defined by SBO regulatory requirements were used as a baseline for comparison to the actual outcomes. The baseline was established from objective measures as stated in the SBO rule, RG 1.155, 53 FR 23203, and NUREG-1109 in the areas of coping capability, risk reduction, EDG reliability and unavailability levels, and value-impact. The baseline requirements provide a set of

expectations as summarized in Table 1, " Summary of Station Blackout Rule Expectations and Outcomes."

No attempt was made to challenge or alter the existing baseline as stated in these documents, or to introduce new expectations, measures, or objectives. The comparison of each outcome to a corresponding baseline requirement determined if the expectations were achieved and if there were any differences between the expectations and the actual outcomes associated with the SBO rule that may require staff attention.

Data for plant-specific SBO rule actual outcomes relative to the expectations associated with the SBO coping capability, selected EDG target reliability, and modifications are shown in Appendix A,

Plant Specific SBO Information by Reactor Type and Operating Status." The data were only' collected from publicly available sources such as licensee PRA/IPE analyses as presented in the NRC PRA/IPE database; NRC/ licensee correspondence, particularly that

- related to the SBO rule safety evaluations; and operating experience. A 15-percent random sample validated that the values in the NRC PRAllPE database were consistent with the corresponding values in the. licensees' PRAs/IPEs.

. Completed NRC system risk and reliability studies, that analyzed the dominant SBO risk factors, were also a source of data to measure outcomes. Idaho National Engineering and Environmental Laboratory (INEEL) (formerly INEL) INEL-95/0035, " Emergency Diesel Generator Power System Reliability," February 1996 [Ref.14], presents an analysis of the reliability of EDG power systems at U.S. nuclear plants during the period from 1987-1993 and

. includes SBO insights relative to EDG target reliability levels. RES plans to reissue this report in 6

FY 2000 to reflect the operating experience through 1998. In NUREG/CR-5496 " Evaluation of Loss of Offsite Power Events at Nuclear Power Plants: 1980-1996," November 1998 [Ref,15],

the NRC contractor has analyzed LOOP events at U.S. nuclear plants from 1980-1996 to j determine their likelihood and duration. Appendix A also contains a tabulation of the ]

NUREG-1032 and NUREG/CR-5496 data of plant-specific LOOP events and LOOP event 1

' recovery times greater than or equal to 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months . l Table 1 Summary of Station Blackout Rule Expectations and Outcomes i

Station Blackout Rule Expectations l Actual Outcomes Observations i Area - Expected Result i 1

Minimum 100 plants analyze and ' Ali plants have 4- or 8-hour coping Expectations exceeded. l Acceptable select a 2 ,4 , 8 , and 16 capability. J Coping hour coping capability.

., Capability 100 plants develop 108 plants developed procedures, procedures and training. completed training.

39 plants complete I modifications. 72 plants completed modifications.

I Industry-Wide 2.6E 05/RY 3.2E-05/RY Expectations exceeded. 1 Risk Reduction I in Mean sBO overall risk reduced.

CDF l Most vulnerable plants completed initiatives to attain low sBO CDF, RG 1.93 actions to shutdown with unavailable power supplies increases the conditional core damage probability, Minimum EDG RG 1.155 plant target EDG MOOS is 0.03. Considenng MOOS (1) in view of the actual MOOS of Reliability and reliability 0.95 or 0 975, the EDG target reliability of 0.95 was 0.03, a target EDG reliability goat unavailability assuming EDG unavailability met in 41 of 44 reactor units examined, of 0 975 may be unrealistic. i due to MOOS is negligible at and (2) none of the 19 reactor units EDG's target re!! ability of 0.95 0.007. examined met the 0.973 EDG target appears to be realistic. Low reliability but roliability general!y 0.95 or EDG reliability may impact better, required coping times.

EDG unavailability monitored RG 1.160 guidance provides Regulatory guidance has per NuMARC 87-00 as alternative rehability and unavailability potentially conflicting measures endorsed by RG 1.155. performance entena and goals. for some performanco criterlon and goals.

Value-impact 2400 person-rem averted per '775 person-rem averted per $M. Outcome was near the low end ;

$M. Expected range was Based on averting 145,000 person-rem of the expected range. scope 700 to 5000 person-rem /$M. at a total cost of $187M. and consequently total were cost Based on averting 145K underestimated by a factor of 3 person-rem at a total cost of due to additions of power

$61.5M. supplies.

I 7 i l

L i l

3.2 Comparison of Expectations and Outcomes Table 1 summarizes the SBO rule (and related regulatory guidance and industry guidelines) elements, expectations, actual outcomes, and observations from each comparison. A more .

detailed discu'sion s follows.

3.2.1 Risk Reduction in NUREG .1109, the staff estimated that on an industry-wide basis the implementation of the SBO rule would result in an industry risk reduction of 2.6E-05 per RY. This expectation is based on a mean SBO CDF before and after the SBO rule of 4.2E-05 per RY and 1.6E-05 per RY, respectively., In addition, as indicated in NUREG-1109, the staff also expected a reduction in the range of the SBO CDFs after implementation of the SBO rule such that more plants would have a lower SBO CDF. The expected change in the range is noted in Table 2,

" Comparison of the Number of Plant Units With a Station Blackout Core Damage Frequency

. Before and After Station Blackout Rule implementation" by comparing the number of plants in each SBO CDF range category before and after SBO rule implementation. NUREG-1109 estimated a plant SBO CDF before SBO rule implementation from plant-specific dominant risk factor characteristics identified in NUREG-1032, various plant-specific NRC and licensee PRA/IPEs, and the assumption that all plants could cope with an SBO for at least 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months .

Table 2 Comparison of the Number of Plant Units With a Station Blackout Core Damage Frequency Beiore and After Station Blackout Rule implementation Parameter Number of Plants in SBO CDF Range (E-05 per reactor-year) sBo cDF <0.5 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 10 Range .99 0.49 1.99 2.49 2.99 3.49 3.99 4.49 4.99 9.99 35 Before sBo rule 5 13 14 7 13. 4 9 5 4 3 13 10 Implementation (Estimated)-

Expected Aher 23 23 14 9 6 5 6 5 4 0 5 0 sBo rule implementation Actual outcome 40 22 14 17 .1 3 1 2 0 2 1 0 After SBo rule implementation On the basis of the SBO CDF data for all plants in Appendix A, the mean SBO CDF outcome associated with SBO rule implementation is 1.0E-05 per RY. Therefore, the reduction in the estimated mean SBO CDF was approximately 3.2E-05 per RY, slightly better than the 2.6E-05 per RY expected. Information from Appendix A was used in Table 2 to show the actual outcome,(i.e., the changes in the SBO CDF range after implementation of the SBO rule). A comparison of the number of plant units in each SBO CDF range for that expected after SBO rule implementation to the corresponding values for the outcomes shows that the expectations were exceeded since more plants have lower SBO CDF and fewer plants have higher CDF than 8

j expected. The actual outcome is even more favorable when compared to the corresponding SBO CDF values before SBO rule implementation.

Appendix B,'" Comparison of Selected SBO Characteristics," was prepared from the data in Appendix A to facilitate comparisons and analyses of SBO characteristics. The comparison of the actual LOOP initiating frequencies while at power from 1968 to 1996 (obtained from NUREG-1032 and NUREG/CR-5496) to the values used in the PRA/IPEs shows that the j

- PRA/IPE LOOP initiating frequencies may have underestimated LOOP initiating frequencies by mn than a isc!or of 10 for three plants and by more than a factor of 3 for three plants.

. Anelysis of Appedix B indicates that the SBO rule was effective in addressing plants that have i greatest vubrabili'y to extremely severe weather and plant-centered LOOPS. As a result of the SBC rule, ths 21 plants that have the greatest vulnerability to a plant-centered LOOP (a l LOUP initiating frequency greater than 1.0E-01 per RY) have access to an alternate ac (Aac) I power supply, and 19 of the 21 plants have a relatively small SBO CDF being less than 1.0E-06 per RY.' In cddition, the eight plants that have the highest expected extremely severe weather '

frequency (Category 5) have access to an Aac power supply.

Table 3, "Probabilistic Risk Assessment / individual Plant Examination Sensitivity Analyses," was prepared from information in plant-specific PRAllPE sensitivity analyses to show the effect of typical SBO rule modifications on the overall risk. Table 3 lists the SBO rule modifications addressed in the PRA/IPE sensitivity analysis and the corresponding risk reduction associated with the modification as a percentage of the total plant CDF. Table 3 indicates that the overall risk (i.e., the risk from SBO and other initiators) was reduced for plants making major SBO rule modifications, such as adding a power supply, or even simple SBO rule procedural changes, such as shedding de loads. Further review of the data in Appendix A found that there are only five plants in which the percent SBO contribution to core melt exceeds 10 percent and the CDF exceeds 1.0E-04 per RY.

RG 1.93, " Availability of Electric Power Sources," December 1974 [Ref.16), was reviewed because it forms the basis for technical specifications in the area of ac offsite and onsite power supply availability. RG 1.93 requirements provides for the nuclear unit to shutdown following extended unavailability of either the offsite power source or emergency onsite power sources.

A shutdown with power supply unavailability completes accident sequence steps and increases the conditional probability of an accident. Other responses, such as assuring the immediate availability of coping systems or assuring adequate electric grid reserves when there are less than the required number of EDGs, may minimize the probability of an accident.

Reductions in the industry mean SBO CDF also reduced the probability that an SBO would result in severe core damage by a factor of 4. The probability that an SBO could result in severe core damage was calculated by multiplying the industry mean SBO CDF in RYs by 100 reactors and by the 25-year average remaining plant life expectancy at the time the SBO rule was implemented. Before implementation of the SBO rule, the probability of core damage given an SBO was approximately 0.105 (a 1 in 10 chance that an SBO would result in severe core damage), based on an industry mean SBO CDF of 4.2E-05 per RY. The corresponding actual outcome was approximately 0.0250 (a 1 in 40 chance tht an SBO would result in severe core damage), based on a mean risk reduction of 1.0E-05 per RY when compared to 0.105 prior to SBO rule implementation. This represents a factor of 4 reduction in the probability that an SBO would result in severe core damage.

9 I

m.

Table 3 Probabilistic Risk Assessment / Individual Plant Examination Sensitivity Analyses Effect on Overall Risk Description of Modification (Percent Reduction of Plant CDF)

Adding two safety EDGs

. Calvert Cliffs 24 Turkey Point 20 Adding DG one safety Diablo Canyon 14-18 Add' nonsafety EDG Arkansas Nuclear 1 23-36 Arkansas Nuclear 2 43-47 Procedural Arkansas Nuclear 1: EDG service water supply value 7 open 17 Monticello: Depressurize during SBO 17 Monticello: Battery load shed Credit of Combustion Turbine Generator Fermi 10 Extend battery life from 2 to 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months Arkansas Nuclear 1 16 Improve reliability of onsite Gas Turbine Generator Point Beach 13 AC Cross-tie Fermi 49 AC Cross connect and Automatic Depressurization System Monticello 38 Previous Assessment of Station Blackout Rule on Core Damaoe Freauencies The NRC staff evaluated the impact of the SBO rule on CDFs in NUREG-1560, NUREG-1560 concluded that SBO remains a dominant contributor to risk at many plants, even after '

implementation of the SBO rule; SBO frequencies exceed 1E-05 per RY for many plants; and

' that given the results, further investigation of strategies for reducing SBO frequencies appears warranted NUREG-1560 provides SBO risk perspectives based on comparison of the SBO CDFs for the 56 plants at which the SBO rule impact was addressed in the PRA/IPE to the SBO CDF at 51 plants on which the impact of the SBO rule on the SBO CDF was not known or credited, or was assumed in the PRA/IPE, to have no impact. NUREG-1560 observed that (1) the average reported reduction in total CDF is consistent with the average reduction in SBO CDF from the backfit analysis of the SBO rule;(2) the average SBO CDF for all plant units considered in the evaluation is comparable to a " typical " estimate from an evaluation of SBO accidents at nuclear plants; and (3) the large variability in the SBO CDF results for all the plant 10

units evaluated is also consistent with the variability in the SBO CDF results from the evaluation of SBO in other studies.

A_psessment The SBO rule was effective in achieving the desired reduction in the magnitude and range of the SBO CDF. The reduction in the estimated mean SBO CDF was approximately 3.2E-05 per RY, slightly better than the 2.6E-05 per RY expected. Also, more plants have lower CDF and fewer plants have higher CDF than expected, lhe SBO rule resulted in the plants with the greatest number of LOOPS from plant-centered events and extremely severe weather conditions having made the most improvement, such as access to an Aac power supply.

Consequently, these plants have relatively low SBO CDF. The SBO rule caused meaningful reductions in the risk at many plants and overall the probability of a core melt from an SBO was reduced from a 10-percent chance to a 2.5-percent chance. SBO has been made a relatively small contributor to core melt as there are only five plants at which SBO contributes more than

. 10 percent to the total CDF, and the CDF exceeds 1.0E-04 per RY. In addition, as a result of .

the improvements made due to the SBO rule, the likelihood of core damage given e.n SBO was l significantly reduced. However, mean LOOP initiating frequencies of some plants used in the PRA/IPE analysis may be underestimated by factors of 3 to 10 in comparison to the actual number of piant LOOPS experienced. In addition, RG 1.93, which forms the basis for technical specification in the area of ac offsite and onsite power supply availability, presently provides for shutdown of the reactor following unavailability of a power supply, and this action increases the conditional core damage probability.

3.2.2 - Emergency Diesel Generator Reliability and Unavailability In INEL 95/0035, an NRC contractor investigated the performance of EDG trains from 1987-1993 in comparison to the RG 1.155 EDG target reliability goals, with and without MOOS.

As discussed previously, the technical basis of RG 1.155 considered EDG unavailability due to MOOS. EDG MOOS is an important consideration since the plant risk is potentially higher because of the possibility of a demand while the EDG is unavailable. INEL-95/0035 found the following:

(1) The 0.95 and 0.975 EDG target reliability goal was generally met if MOOS was ignored.

The sample of 44 plant units with an EDG target reliability of 0.95, all 44 plants met the goal if MOOS was ignored. The actual mean reliability of the 0.95 target reliability goal population was 0.987 (if MOOS was ignored) with a corresponding uncertainty interval of 0.96-0.990 and the probability of meeting or exceeding the target goal of 0.95 is about 97 percent, in sample of 19 plant units with an EDG target reliability of 0.975, the target reliability goals were met for 18 of the plants if MOOS was ignored. The mean reliability of the 0.975 population was actually 0.985 (if MOOS was ignored) with a corresponding uncertainty interval of 0.95-0.99, the probability of meeting or exceeding the target goal of 0.975 is about 54 percent.

(2) Investigators found that the industry-wide average MOOS was actually 0.03 and not 0.007 as was assumed in RG 1.155. Considering the MOOS on a plant-specific basis,41 of the 44 plant units with an EDG target reliability of 0.95 met the goal. The mean actual reliability of this population was 0.956 with a corresponding uncertainty interval of 0.92-0.99, and the probability of meeting or exceeding the target goal of 0.95 is about 67 percent. Considering 11

. the MOOS on a plant-specific basis, none.of the 19 plants in the 0.975 EDG target reliability population met its target. The mean reliability of this population was 0.954 with an uncertainty interval of 0.91-0.98, and the probability of meeting or exceeding the target goal of 0.975 is about 17 percent.

~ RG 1.155 allows licensees to select the higher EDG target reliability levd of 0.97o to reduce the required SBO coping time. Consequently, failing to achieve the EDG target reliability levels

. may affect the plant coping duration. In addition, increased target reliability levels could effect PRA/IPE results if the PRAAPE is sensitive to an increase in the EDG failure rate. As an example, one licensee's PRA/IPE sensitivity analysis found that doubling the EDG failure rate would increase the CDF by 24 percent (8.23E-05/RY).

' The INEL-95/0035 data indicate that the underlying EDG reliability is 0.95 as the 0.95 and 0.975 EDG target reliability group actual mean reliabilities are nearly the same, being 0.956 and 0.954.- Further, the INEL-95/0035 data indicate the 0.975 EDG target reliability goal is not likely to be achieved, considenng the probability of meeting or exceeding the target goal of 0.975 is 17 percent with MOOS and 54 percent without MOOS failures.

Analysis of NRC and Industry EDG Performance Criteria RG 1.155 shows that plants select and maintain an EDG target reliability level of either 0.95 or 0.975 per demand. All licensees subsequently docketed commitments to maintain the selected RG 1.155 EDG target reliability level.

RG 1.160, which is used to evaluate the effectiveness of EDG maintenance activities under the

. SBO rule, presents EDG reliability performance criteria or goals that potentially conflict with those in RG 1.155. Failing to detect and correct degraded RG 1.155 EDG target reliability levels could impact the selected SBO rule plant coping duration. RG 1.160, which has been

- revised twice since June 1993,'has always stated that licensee RG 1.155 EDG reliability _

commitments could be used as a goal or performance criterion for EDG reliability. However, RG 1.160 also endorses the use of the trigger failure rates in Table 4," Industry Emergency Diesel Generator Trigger Failure Rates," for the purposes of goal setting by reference to NUMARC documents.

Table 4 Industry Emergency Diesel Generator Trigger Failure Rates l Selected Target ' Failures in 20 Failures in 50 Failures in 100 Reliability Demands Demands Demands 0.95 3 5- 8 0.975- 3 4 5 When compared to the RG 1.155 EDG target reliability levels, the trigger failure rates do not ensure the RG 1.155 reliability levels with a reasonable degree of confidence. Appendix E, l "EDG Reliability and Industry Trigger Failure Rates," which was prepared with INEEL assistance, indicates that the probability that the industry failure rates meets or exceeds the 12'

~

5 .

p ,

associated target reliability is at best less than 10 percent. Appendix E data also indicates that where a relatively high (95 percent ) exceeding confidence exists, the reliabilities are considerably less than the target reliabilities. Consequently, the EDG trigger failure rates may allow the EDG to be in a degraded condition.

It appears the weaknesses in the trigger failure rates was recognized in RG 1.160, dated June 1993, which noted that it is not practical to damonstrate by statistical analysis that conformance to the trigger values will ensure the attainment of high reliability, with a reasonable degree of confidence, of individual EDG units. Specific reference to the trigger values was deleted in RG 1.160, Rev.1, January 1995. However, their use remains an option in the current revision of RG 1.160, Rev. 2, March 1997, through endorsement of NUMARC documents.

Assessment Overall, and considering MOOS, the industry has achieved 0.95 EDG reliability, however, the 0.975 EDG reliability goal has not been met. An increase in the MOOS from an originally expected value of 0.007 to an actualindustry wide value of 0.03 indicates that individual plants will have great difficulty in meeting 0.975 EDG target reliability on a plant-specific basis.

Analysis of the INEL-95/0035 results indicates that a 0.975 EDG reliability goal may be unrealistic since there is only a 17 percent and 54 percent chance of meeting the 0.975 EDG target reliability with and without MOOS, respectively.

Five years of industry data indicates that the EDGs are inherently 0.95 reliable as the EDG target reliability groups of 0.95 and 0.975 are nearly the same, being 0.954 and 0.956; 0.95 may be a more realistic goal.

Comparison of the SBO rule and maintenance rule documents found that implementation of the maintenance rule and RG 1.160, in conjunction with the SBO rule and RG 1.155, and endorsement of supporting industry documents, did not provide the basis for ensuring that licensees establish as a goal and maintain a specific EDG reliability level as intended by RG 1.155.t RG 1.160 endorsement of industry guidance allows alternative EDG reliability goals and performance criteria that may allow degraded EDG reliability conditions to exist. Failing to achieve the EDG target reliability levels may impact the plant coping duration, and licensee SBO rule commitments to maintain specific EDG reliability levels. Analysis could determine if low EDG reliability compromises plant-specific FRA/IPE results.

3.2.3 Minimum Acceptable Coping Capability, Plant Procedures, Training, and Modifications in NUREG-1109, the staff expected that 100 plants would (1) be able to show a minimum acceptable coping capability of 2,4,8, or 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months based on plant-specific characteristics; (2) complete an analysis to show the plant's ability to cope with an SBO for the selected duration; (3) develop SBO-related procedures; (4) complete training on these procedures; and (5) complete modifications as necessary to cope. The SBO rule provided flexibility with a wide range of coping capabilities so plants with an already low risk from SBO were required to cope for the shortest time and need few, if any, modifications. Those plants with higher risk required longer coping times and possibly modifications to cope. Thirty-nine plants were expected to complete hardware modifications.

13

l Appendix C,4 Station Blackout Rule Activity and Modification Summary," was prepared to show the expected number of plants completing analyses, procedure development and training, various types of modifications in the licensees response to the SBO rule, and the estimated costs of the modifications as well as the corresponding outcomes. The costs are discussed in Section 3.2.4,"Value-Impact Analysis."

The outcome was that 108 plants selected a minimum SBO coping capability of 4 or 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , completed the coping analysis, developed procedures, and completed training; and 72 plants completed modifications. Not credited was the fact that some plants may have already developed adequate procedures and completed training before the SBO rule as a result of Generic Letter 81-04, " Emergency Procedures and Training for Station Blackout Events,"

February 25,1981.

Assessment

. The SBO rule expectations in the area of coping analysis, procedure development and training, and modifications were met. The scope and number of modifications to cope for specified l durations exceeded the expectations and may explain how risk reduction expectations were

{ also exceeded.

3.2.4 -Value-impact Analysis A comparison of value-impact expectations and outcomes was derived from NUREG-1109.

i The value-impact analysis in NUREG-1109 estimated the expected value based on the public dose reduction associated with the SBO rule.' The impacts were based on estimates of industry

( and NRC costs to implement the SBO rule. The ratio of the value to the impact was also derived as an indication of the cost effectiveness of the SBO rule. Each of these is discussed ;

l below. ]

Table 5, " Station Blackout Rule Value-impact Summary," was prepared from Appendix C to l summarize the expected impact, value, and value-impact ratio in comparison to the j correspondino outcomes. Appendix C used the expected values and impacts from l l

NUREG-1109. The outcomes were estimated using either the values and impacts from NUREG-1109 or from information submitted to the NRC by the licensees.

l In NUREG-1109, the staff estimated that the industry would spend approximately $60 million (M) and that the NRC would spend approximately $1.5M to implement the SBO rule. The total estimated SBO rule implementation cost of $61.5M represented a best estimate with a low of

$43M and a high of $95M. The total SBO rule implementation cost of $61.5M in NUREG-1109 l reflects significant increases as a result of industry comment preceding the issuance of NUREG-1109. In NUREG-1109, the staff recognized that there would be wide variation in plant specific costs, ranging from $0.35M for plants that only needed to make procedural changes to $4M if all.the anticipated modifications were completed. The outcome in terms of actual SBO rule costs was about $237M which, on the surface, exceeded expectations by about a factor of 4. The discrepancy is attributable to the addition of 19 power supplies with an estimated net cost of approximately $174M. These additions, although consistent with the SBO l rule were also motivated by monetary benefits that licensees could realize because these

[ power supplies result in operating flexibility. This suggests that it may be conservative to ascribe all these costs to the SBO rule and that the' NRC also underestimated the value.

14

[..

i ._

c i C

Table 5 Station Blackout Rule Valuo-impact Summary Value-impact Factors Expected ($) Outcome ($)

Impact-NRC and Industry implementation Cost Best estimate Miscellaneous SBO modificatioris 60M 63M NRC implementation 1.5 2.0M 19 additional power supplies 0 174M TOTAL 61.5M 237M Monetary savings attributed to adding additional power supplies for operating flexibility -50M TOTAL 187M Estimated range Total 43M- 95M 350K-20M Plant-specific 350K-4M -

Estimated Value Public dose reduction in person-rem 145,000 145,000 Value-impact Ratio (person-rem averted /$million)

Best estimate 2400 775 Range 700-5000 -

The NRC estW.ated that the added power supplies result in $50M added value from increased allowed outage times (AOTs) to provide operating flexibility. For example, Davis-Besse obtained NRC approval to change its technical specifications to increase the AOT on its EDGs from 3 to 7 days to provide flexibility in performing EDG maintenance while the reactor is at power. In submitting this change to the NRC, Davis-Besse noted that its SBO diesel generator installation would save $5.25M over the remaining plant life, including $3.15M for increased flexibility in performing maintenance on its safety EDGs and $2.1M in replacement power costs.

North Anna increased its EDG AOT from 3 to 14 days by crediting the non-Class 1E SBO EDGs, as did Peach Bottom by crediting the added SBO connection to Conowego Hydro, end 10 other licensees for crediting the excess EDG redundancy or Aac power supply documented M iheir SBO analyses. RES estimated that similar to Davis-Besse,10 reactor units received a total of $50M in benefit from increased EDG AOT to gain operating flexibility, it also appears that the addition of a power supply resulted in significant plant-specific reductions in risk. Turkey Point added two EDGs and cross-ties after determining these additions reduced the risk of an SBO after a LOOP from 7.6E-04 to 2.9E-06 per RY. Table 3 indicates that adding power supplies to Calvert Cliffs, Diablo Canyon, and Arkansas Nuclear One PRA/IPE resulted in reductions in the risk ranging from 14 to 47 percent.

15

9 In NUREG -1109, the staff expected that the SBO rule would result in value from public risk reduction by averting 145Kperson-rem from an accident; high and low estimates ranged from 216.5K person-rem to 65K person-rem, respectively. The estimated expected reduction in person-rem was calculated by multiplying the reduction in CDF per RY from an SBO by the l remaining life of the plant (assumed to be 25 years) and the estimated public dose based on !

the highest site source term that would result from an accident. The total reduction in person-rem for each plant was summed and divided by 10 to account for a smaller source term for an SBO at a 50 mile radius. The 145K person-rem derived in NUREG-1109 appears to be realistic, and may be low, in view of information that indicates the weighted population dose factor for the five NUREG-1150 power reactors ranges from 166K to 2M person-rem within 50 l miles of the plant (NUREG/BR-0184," Regulatory Analysis Technical Evaluation Handbook,"

February 1997).

In NUREG-1109, the staff calculated a value-impact ratio of 2400 person-rem per SM based on averting 145K person-rem at a cost of $61.5M with a range of 700 person-rem per $M to 5,000 person-rem per $M. The outcome was that the value-impact ratio was 611 person-rem per SM based on an estimate impact of $237M. However some plants received approximately $50M in monetary benefit from increased EDG AOT reducing the impact to $187M, resulting in a value-impact ratio of approximately 775 person-rem per $M, which is within the expected range.

Previous Assessment of the Cost Effectiveness of the SBO rule The staff also evaluated the industry's average cost per person-rem averted in satisfying the SBO requirements in SECY 97-180," Response to Staff Requirements Memorandum of May 28,1997 Concerning Briefing on IPE insight Report," August 6,1997. The staff used a different methodology and concluded that, on average, the SBO rule averted one person rem cost .f $4,750 (211 person-rem per $M), and that this result is likely overstated because it does not give full credit for other sizable economic benefits. Further, this result is skewed by a few plants whose SBO exceeded $10M. In comparison, most reactors incurred SBO costs of less than 51M. A supporting calculation suggested that about 70 percent of the reactors incurred costs of less than $1000 per person-rem averted (1000 person-rem per $M), and 75 percent incurred costs of less than $2000 per person-rem averted (500 person-rem per SM).

Assessment Comparison of the expected value-impact to the actual value-impact indicates that the actual value impact was within the expected range of reduction in public dose per dollar of cost.

l However, the staff underestimated the number of plants that would increase onsite emergency power sources. That is, the NRC did not anticipate that so many licensees would install additional safety-related and nonsafety-related power sources that account for 75 percent of the industry cost. Consequently, the expected "best estimate" cost was underestimated by a factor of approximately 3. However, it appears licensees justified the cost of the added power supplies basUd on reductions or offsetting monetary benefits, such as increased EDG AOT for operating fle>'5ility. Thus the NRC also underestimated the value. Not considering the costs of the power supplies, the costs to meet the SBO rule requirements were reasonable given that these plants have added SBO coping and recovery procedures, and have established an EDG reliability program to maintain EDG target reliability levels. Further, the SBO rule focused the NRC and industry resources on an area known to be important to the overall risk of an 16

I operating nuclear power plant, and the SBO rule analysis helps to ensure an acceptable level of safety is being maintained at operating nuclear power plants.

The industry and NRC costs to implement the SBO rule were reasonable considering the outcomes. As expected there was wide variation in plant-specific values and impacts because the SBO rule provided flexibility so that plants with higher risk required longer coping times and possibly more modifications to cope, and plants with already low risk from SBO were required to cope for the shortest time and needed fewer modifications to cope.

3.2.5 Insights, From Operating Experience Reviews Ultimately the performance during operating events provides a means to assess the effectiveness of regulatory requirements and guidance. Operating experience insights follow.

Station Blackout Rule Modifications The SBO rule modifications and procedures have been relied upon to provide protection during operating events. For example, the original Turkey Point Units 3 and 4 EPS design shared two 2850-kW safety-related EDGs and five onsite nonsafety-related EDGs that could be connected to the reactor unit EPS through a nonsafety-related switchgear. As a result of the SBO rule, the licensee added two 3095-kW safety-related EDGs that were estimated to reduce the risk of an SBO after a LOOP from 7.6E-04 per RY to 0.9E-06 per RY. A March 1993 report, "Effect of Hurricane Andrew on the Turkey Point Nuclear Generating Sation from August 20-30,1992,"

indicates that the five onsite nonsafety-related EDGs were unavailable because of water damage to the nonsafety-related switchgear; the total load on the four safety related EDGs was approximately 3400 kW, and about 3.5 days into the storm one of the two original 2850-kW EDGs tripped and was restarted in 2.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months . Had the two safety-related 3095-kW EDGs not been added as a result of the SBO rule, the load on the only remaining 2850-kW EDG would have exceeded its rating by 19 percent and most likely would have failed unless operators unloaded it quickly, leaving no ac power for 2.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months . However, the Turkey Point SBO rule implementation precluded the possibility of this occurrence.

Offsite Power System Dereoulation in SECY 99-129," Effects Of Electric Power Industry Deregulation on Electric Grid Reliability and Reactor Safety," May 11,1999 [Ref.17], the staff indicates that the risk significance of potential grid unreliability due to deregulation is likely to be minimal, although individual plants might have an increase in CDF due to deregulation should grid performance substantially degrade. Should grid performance unexpectedly degrade as a result of deregulation of the electric utility industry, the SBO rule provides additional defense in depth by requiring a plant to cope with an SBO for a specified duration.

Dominant SBO Risk Factor Trends The dominant factors affecting the likelihood of SBO accidents at nuclear plants are (1) the LOOP frequency, (2) the time to restore offsite power following its loss, and (3) EDG reliability and redundancy.

INEL-95/0035 (data from 1987-1993) compares the mean reliability per demand to the

- corresponding value in NUREG 1032 (EDG data from 1981-1983). The comparison showed 17

i the mean reliability of 0.98 in NUREG-1032, and th3 corr:sponding value from INEL 95/0035 was 0.956. INEL-95/0035 also analyzed the EDG train unreliability by year (from 1987 through 1993), and found that no trend of EDG train performance by year is evident.

Table 6," Dominant Station Blackout Risk Factor Trends - Offsite Power System," was prepared to compare the offsite power trends in NUREG-1032 (based on ',968-1983 operating experience) to the corresponding offsite power values in NUREG/CR-5496 ( based on 1980-1996 operating experience). The number of times a LOOP lasted 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months or more was also quantified in Table 6.

The results show overall improvement in LOOP frequency and duration. One exception is the increasing duration of grid-related LOOP events. This is being addressed by SECY 99-129.

The number of LOOPS in excess of 4 or more hours prompted RES to calculate the probability of not recovering from LOOP. Table 7," Analysis of Loss of Offsite Power Event Recovery 1968--1996," was prepared by counting the number of plant , grid , and weather-related LOOPS in NUREG-1032 and NUREG/CR-5496 and calculating the probability oi not recovering from the LOOP in each of the corresponding recovery time intervals in Table 7. Table 7 indicates that the probability of not recovering from weather-related LCOPs in 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months is high in comparison to either the plant- or grid-centered LOOPS. The data show there is a 47 percent chance of nonrecovery from a weather-related LOOP in 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months and an 18-percent chance of nonrecovery from a weather-related LOOP in 8 and 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months . In comparison the corresponding values from plant- and grid-related LOOPS are significantly less, being 0.05 and 0.06 in 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months and 0.0 and 0.02 in 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . Further, Table 7 shows the NUREG-1032 provided " enhanced" or expected probabilities of not restoring offsite power for plants with recovery procedures and at least one power source for prompt recovery. Comparison of the weather-related enhanced probability of nonrecovery to the corresponding calculated probability of nonrecovery indicates the enhanced levels have not been achieved.

Appendix D, Table D-1, Losses of Offsite Power Since 1990 Having Recovery Times Greater Than 4 Hours," summarizes six plant and three weather-related LOOPS from 1990 (when the SBO rule was implemented) to 1998 while the reactor unit was running that took 4 or more hours to recover offsite power. Three of the events were SBO-type events (i.e., involving a LOOP and subsequent or immediate unavailability or technical inoperability of one EDG).

Analysis of the plant events found weaknesses in communication and procedures that delayed recovery. These weaknesses were subsequently corrected and should significantly improve recovery from future losses of offsite power. Analysis of weather events shows that recovery of offsite power from a hurricane could take up to 4.5 days, from a tornado 28 hours3.240741e-4 days 0.00778 hours 4.62963e-5 weeks 1.0654e-5 months , and from ice accumulation or contamination from salt sprays approximately 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . Provisions for these types of events are addressed in NUMARC 87-00, Rev.1, which requires actions for achieving enhanced coping capability under hurricane and tornado conditions. Key features of NUMARC 87-00 are (1) actions to be taken in the 24-hour period preceding anticipated hurricane arrival and (2) a commitment to be in a safe shutdown condition 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months before the anticipated hurricane arrives.

18

L Tcble 6 Dominant Station Blackout Risk Factor Trends - Offsite Power System Source of LOOP Data NUREG 1032 NUREG/CR 5496 (1968-1983 LOOP (1980-1996 LOOP operating experience) operating experience)

Plant-related LOOP ~

Mean frequency 0.09 0.04

~ Median time te restore 18 minutes 20 minutes LOOPS > 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months .0 4 Grid-related LOOP Mean frequency (occurrence per 0.018 0.0019 year) . 36 minutes 140 minutes Median time to restore 1 (due to severe weather) 0 LOOPS > 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months Weather-related LOOP Mean frequency 0.009 0.0066 Median time to restore 4.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months 1.2-2.4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months LOOPS > 4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months 4 5

- Table 7 Analysis of Loss of Offsite Power Event Recovery 1968-1996 Number of LOOP Recovery Time in Mii i utes Events and Probability of Recovery 0 0 0 0 0 0 0 0 > 961 Total 30 60 120 180 240 480 960 Plant LOOP events 46 58 71 71 71 75 75 P (nonrecovery) 0.39 0.23 0.05 0.05 0.05 0 (in 388' min)

Weather LOOP events 4 4 6 6 9 14 14 17 17 P (nonrecovery) 0.76 0.76 0.65 0.65 0.47 0.18 0.18 0 (in 7950 min)

Enhanced P (nonrecovery) 0.65 0.40 0.25 .01 0.05 0.02 Grid LOOP events 7 9 11 15 17 18 18 P (recovery) 0.61 0.5 0.39 0.17 0.06 0 (in 388 min)

P (nonrecovery)

Potential SBO Alternate ac Power Source Unavailability Appendix D, Table D-2," Station Blackout Challenges," was prepared from a review of the 1990-1998 operating experience. That review found one event (licensee event report

[LER] No,335/98-007) that identified inadequate SBO recovery procedures during a simulator exercise in 1998 that could have complicated recovery from an SBO and four events (LER Nos.

346/98-006,247/98-007 and Information Notice [lN] 97-21 (Ref.18]) that identified the potential unavailability of an Aac power source (that was added as a result of the SBO rule) during an 19

p

.SBO cv:nt. Th31:tt:r is of concern cince th3 unav:ilibility of th3 Aac pow:r supply during an SBO could lead to core damage. Tv o of the four events resulted in IN 97-21. IN 97-21 described the dependencies of the Asa support system batteries and ac auxiliary power. The other two events were in 1998. One 1E98 event (247/98-007) noted that the SBO Aac power j source output circuit breakers were not capable of being closed onto a de-energized, bus as would be the case during an SBO. The licensee reported that the cause of the problem was insufficient comprehensive testing. In the second 1998 event (346/98-006), a tornado resulted in a LOOP and technical inoperability of one EDG, and offsite power was restored in 28 hours3.240741e-4 days 0.00778 hours 4.62963e-5 weeks 1.0654e-5 months .

An NRC analysis of the tornado event noted that a nonessential bus supplies power to the Aac, and if not powered, the batteries would deplete in approximately 28 hours3.240741e-4 days 0.00778 hours 4.62963e-5 weeks 1.0654e-5 months . Had the EDGs failed to start or run in the first 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months of the 28-hour event, the SBO-DG may have been unavailable and likely would have d to core damage.

Assessment implementation of the SBO rule has produced additional defense in depth. As demonstrated at Turkey Point, the SBO rule modifications and procedures have been relied upon to mitigate the consequences of, and provide protection during, operating events. In addition, the SBO rule provides for offsite power as a result of deregulation of the electric utility industry or changing offsite power system trends.

The probability of recovery in lesa than the SBO coping time for extremely severe weather LOOP events is lower than from other types of LOOPS. Because of the nature of these events,

! it is important to take actions before the occurrence of extremely severe weather conditions, consistent with industry requirements (that have been endorsed by the NRC).

Even after issuing an IN, an event revealed that an Aac power supply system availability was limited by dependencies on offsite or onsite power supplies that could lead to core melt.

4 CONCLUSIONS Comparison of the SBO rule expectations and outcomes indicates the SBO rule was effective, and industry and NRC costs to implement the SBO rule were reasonable considering the outcomes. Implementation of the SBO rule resulted in changes to procedures, training, EDG performance monitoring, hardware modifications, and the addition of diesel and gas turbine power supplies. These outcomes provide the plants with SBO coping capability, reduce the risk, and make the plants more tolerant to loss of offsite or onsite power as follows:

- The reduction in the estimated mean SBO CDF was approximately 3.2E-05 per RY, slightly better than the 2.6E-05 per RY expected after implementation of the SBO rule. As a result of the improvements made due to the SBO rule, the likelihood of core damage given an ;

SBO was also significantly reduced, in addition, after implementation of the SBO rule, more plants have a lower SBO CDF than expected. Further, SBO has been made a relatively small contributor to core melt as there are only five plants at which (a) SBO contributes more than 10 percent to the total CDF and (b) the CDF exceeds 1.0E-04 per RY. Also, the SBO rule resulted in the plants with the greatest number of LOOPS from plant-centered events and extremely severe weather conditions having made the most improvement, such as access to an alternate ac power supply. Consequently, these plants have relatively low SBO CDF.

20

-_ D

a implementation of the SBO rule resulted in all plants having (1) a 4- or 8-hour coping capability; (2) established SBO coping and recovery procedures; (3) completed training for these procedures; (4) established an EDG reliability program that generally maintains EDG 1 reliability levels at or above 0.95; and (5) implemented modifications as necessary to cope l with an SBO.

The operating experience indicates that the SBO rule provided additional defense in depth, given that for many licensees, SBO is a significant contributor to the overail risk, and SBO rule-related hardware and procedures have actually been used to mitigate event consequences and provide additional protection. Additional defense-in-depth is also provided by the SBO rule should the performance of the ac power system unexpectedly degrade as a result of deregulation of the electric power industry or if it takes longer than expected to recover offsite power following extremely severe weather conditions.

Comparison of the value-impact expectations to the corresponding outcomes indicates that the value-impact was within the expected range of reductions in public dose-per-dollt.: ;f cost. As expected, there was wide variation in plant-specific values and impacts because the SBO rule provided flexibility. Not expected was the addition of costly power supplies that account for 75 percent of the industry cost impact. This factor explains why the NRC value-impact analysis underestimated the cost by a factor of 3. However, it appears licensees justified the cost for the power supplies based on offsetting monetary benefits, such as increased EDG allowed outage times for operating flexibility or meaningful risk reductions. Thus the value was also underestimated. The remaining 25 percent of the industry cost appears reasonable considering the outcomes: known coping capabilities, industry risk reduction from plant-specific procedural and hardware enhancements, and additional defense-in-depth.

Comparison of the SBO rule expectations to the corrasponding outcomes concluded that the ;

following issues warrant attention: 1

1. TN O.975 EDG reliability has not been r.aintained at levels equal to or above levels i

selected in RG 1.155 that were used for the determination of the required coping time. In addition, EDG unavailability due to MOOS was found to be 0.03, not 0.007 as assumed in RG 1.155, and indicates that plants will have great difficulty in meeting 0.97E reliability l goals.

l

2. The 0.975 EDG target reliability goal may be unrealistic. Operating data shows (a) the underlying EDG reliability appears to be 0.95 as the EDG target reliability groups of 0.95 and 0.975 are nearly the same being 0.954 and 0.956, respectively, and 2) achieving the i EDG target reliability levels of 0.975 is not probable being 17 percent and 54 percent with and without MOOS, respectively.

3. There may be inconsistencies between the SBO RG 1.155 and RG 1.160 reliability and unavailability criteria. The industry trigger failure rates recommended for EDG reliability performance monitoring and goal setting in documents that are endorsed by RG 1.160, do not reasonably assure the RG 1.155 reliability levels will be achieved. NRC processes for issuing regulatory guides need to ensure the absence of competing expectations. This will .

be important as the NRC continues to transition to a risk-informed performance based l regulations.

l 21 .

1

r

4. Some licensee PRA LOOP initiating frequencies appear to be a factor of 3 to 10 lower than the corresponding value determined from the operating experience.

5. Regulatory Guide 1.93 forms the basis for technical specifications in the area of ac onsite and offsite power supply availability by requiring shutdown of the reactor following extended ac power supply unavailability. This criteria may adversely affect the SBO CDF.

6. Events continue to identified that the availability of some alternate ac power supplies is limited by dependancies offsite or onsite power supplies that could lead to core melt during an SBO event.

5 REFERENCES

1. U.S. Nuclear Regulatory Commission, "10 CFR 50, Station Blackout," Federal Register, Vol. 53, No.119, Page 23203, June 21,1988.

2. U.S. Nuclear Regulatory Commission, NUREG-1560, " Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance," December 1997.

3. U.S. Nuclear Regulatory Commission, SECY-97-180, " Response to Staff Requirements Memorandum of May 28,1997, Concerning Briefing on IPE insight Report," August 6, 1997.

4. U.S. Nuclear Regulatory Commission," Evaluation of Station Blackout Accidents at Nuclear Power Plants," NUREG-1032, June 1988.

5. U.S. Nuclear Regulatory Commission, " Station Blackout," Regulatory Guide 1.155, August 1988.

6. U.S. Nuclear Regulatory Commission, " Regulatory /Backfit Analysis for the Resolution of Unresolved Safety Issue A-44, ' Station Blackout'," NUREG-1109, June 1988.

7. Nuclear Energy institute, " Guidelines and Technical Bases for NUMARC Initiatives Addressing Station Blackout at Light Water Reactors," NUMARC 87-00, Rev. O, November 1987.

I f 8. U.S. Nuclear Regulatory Commission, " Final Station Blackout Rule, USl A-44,"

} SECY-88-22, January 21,1988.

9. Nuclear Energy Institute," Guidelines and Technical Bases for NUMARC Initiatives L

Addressing Station Blackout at Light Water Reactors," NUMARC 87-00, Rev.1, August 1991.

l l

10. U.S. Nuclear Regulatory Commission, " Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," Regulatory Guide 1.160, June 1993.

11. Nuclear Energy Institute, " Industry Guideline for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants, NUMARC 93-01, Revision 2, April 1996.

22

g

)

n. .,

j o

j

.12. U.S. Nuclear Regulatory Commission, " Requirements 'for Monitoring the Effectiveness of I Maintenance at Nuclear Power Plants," Regulatory Guide 1.160, Revision 1, January 1995.

13. U.S. Nuclear Regulatory Commission, " Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants," Regulatory Guide 1.160, Revision 2, March 1997,

14. U.S. Nuclear Regulatory Commission, " Emergency Diesel Generator Power System j Reliability 1987-1993," INEL-95/0035, February 1996. I

15. U.S. Nuclear Regulatory Commission," Evaluation of Loss of Offsite Power Events at Nuclear Power Plants: 1980-1996," NUREG/CR-5496, November 1998.

16. U.S. Nuclear Regulatory Ccmmission, " Availability of Electric Power Sources,"

Regulatory Guide 1.93, December 1974.

.~

17. U.S. Nuclear Regulatory Commission," Effects of Electric Power industry Deregulation on Electric Grid Reliability and Reactor Safety," SECY-99-129, May 11,1999. I

18. U.S. Nuclear Regulatory Commission, Info'rmation Notice 97-21, " Availability of Alternate l

AC Power Source Designed for Station Blackout," April 18,1997. Pace 21. l l

l 23

r, i

l l

i APPENDIX A ;

PLANT-SPECIFIC STATION BLACKOUT INFORMATION l BY REACTOR TYPE AND OPERATING STATUS I

I i

i i:

I

n a w ss o n ee t

_ d e

vmt t

u u h

^ eit in S Prym Oev0 Oo4 L c2 r

e 0 e w 3 r o 3 P

t al d s sa ir G'

s r o nc t i eer u t c

t a a vm e

t S O f

Pmn oio h r

e B Oct a t a 1 1 g S Oer Lcenp e i

n f W t oio s a r r r

e b ee w t n

p mo up la 2 1 2 1 2 3 O J P'

1

'. d f

n s a r o P y e

p t

c Og nt c 2 0-2 0-2 0-2 0-2 0-2 0-2 0-1 0- 3 y a Oin t n E E E E E E E E 0-T e Lait ev u e 8 4 4 4 3 3 0 6 E R Aieq e 5 8 4 5 4 6 3 0 r r Ri n r 3 5 6,. 2 t

o t e P f , 7 4 4 4 1 c a a W-e n c s Ge R d oydd la n i dG y e it r a a gh er ut o Dts s di e di e di e 1 n B

i z

r i n

cmisdaeic 1 dr o et s d s et s d s et s ds dad e 1 n u i ul f

d mdu aodd cf i ec d o Ar c do d o dGe s o s inlcorpo dd Ar c Ar c dDn 1 o s e M d n Aa AEo -

i t r m A a P m g e f

r n

o i t

n a

Oci BieGat sn cin Ae/er r

e v 1 1

/ '

1

/

1

/ 1 1 1 4

/ 1 I

t r

e SsmD ir detiEyi t t t yt/ms e se h

/

0 1

0 6

/

0 1

/

0 1

/

5

- 0 6

/

0 1

u p g/i U /

/ /

/

s ul 5 5 5

/

tetc nrsl i

ne a 5 5 5 7 5 o O caiu b 9 9 7 9

7 9 9 9 9 7 9 9 k era poicaemsim e ew /

4

/

4 / /

/

4

/

4

/

4 !

/

4 c 1 leh oh lec r 4 4 4 l

a A Sc( C r a tex B l e .

n b a

o T t F F i

t nD D a e 8 4 3 6 9 8 3 t cCf C 3 6

0 5 2 3 0 4

0 S eOotna r

PB 3 3 3 2 2 1 3 3 1 .

i c S l

P f

i c

e p

5 6 5 5 6 6 5 6 7 0 0- 0- 0 0- 0- 0- 0-S- OF E E E E E E E E 0-BD 8 3 1 6 0 0 0 2 E -

t SC 5 2 5 8 2 3 8 3 0 n 1 1 6 4 6 4 1 8 6 l

a P F 5 5 4 4 5 5 5 4 5 D 0- 0- 0- 0- 0- 0- 0- 0- 0-C E E E E E E E E E t

n 7 0 4 2 4 9 5 0 0 la 6 4 1 9 7 0 8 4 8 P 4 3 2 1 2 3 5 2 5 y y e e le le s n n la la d f

f sO i

sO V V o2 2 y Cl 2 2 ar sa ar sa r r o& & a t

& ba&1 t ne1 ne2 e1 v e2t v

w1 d s n s 1

w r 1 e s w s n al ot la vt at kalcit ai t it a kr citun ai ai n r n yi i t i l r un en en r la lan an P ANU ANU BU BU BU BU C CU CU u

n a w t ss o 1 7 n

e ee d

t u

61 2 9 vmtu i h et in S Prym Oev0 Oo4 Lc2 r

e 0 e w 8 r o 6 1

P t

al id s

r t sa ic~ G r

s o nr' u t c ee v

t a a e m t

S f

O Pmn oi o h r

e B Oct Qer a t

g S a 1 Lc np e e i

n f W t oio s a r r e

r bee w t n

p mo

) . up la 3 2 1 1 2 2 4 O t n N P o

_ . d n C a ( P~ y e

p s

r O o~

nt n c 1 0-2 0- 2 0-2 0-2 0-1 0-3 0-y o Oin t E E E E E t Laiveu e E E T c a Aieq t 5

3 0

5 0 1 0 7

7 1

0 5

r e Ri n e r 4 9 2 3 4 3 4 t

o R P f c r a e e t a n c l s

n d E G d e G r

eo d e

R W- ioyddar er o e 1 y D h D t t au g n h y

t a agh ui t hd sr se se 1 s 1 wa s B d e

ic m M >. eic da d dat d d a d e c,d iad d d a

f i ul md dcf u aod adl t

e e icalo e n

o i

z r

d o s inI p o cor i

loAca nb o

d d

lo e

d d

vA r

eo h

s lo C

2-i u M m d c n A d A St D A t

a s s

m e r

r P e f

o g SsieGac in cin r

e v

n n imD Ae/er t 4 2 2 3 1 i

t i

t S' s r detiEyi

/

ms e se h 1

4

/

4

/

5

/

0 1

4

/

5 1

J

/

0 1

2 4

/

4

/

5 u a r et cg/ sis l it t t ul a yt 5

9 7

/

5 7 5 9

/

5 5

9 7 o e t

caiunrbsine e 9 9 9 9 9 k p er poicaem mre w /

4 / / /

/

4 /

/

4 /

c a O leha oh lec Sc( C r a x t 4 4 4 4 4 l

e B 1 n A o e t F F i

t a

l b nD e D t a cCf Cn 2 6

5 1

3 1

8 8

6 4 -

4 1

S T eOotla r

PB 2 2 5 1 5 9 1 i

c S P f

i c

e p 5 6 5 5 6 5 e 6 0- 0- 0- 0 0- l OF - 0- b 0-4 t

BD E E E E E E ola E 8 0 3 2 t SC 5 2 5 1 0 2 nia v 0 n 1 3 3 1 5 1 a 1 l

a P F 5 5 5 5 D 0- 0- 5 0-5 0-5 0-4 0- 0- 0-C E E E E t

n 2 3 E E E E 6 4 7 5 6 2 8 3 3 7 l

a 6 6 8 1 P 5 1 1 8 2

8 n r

e o n 1

e s y

u est iv s 2 n

a2 o hi R e k 2 h cn B- o& C, 1A & la nU a

l o1 I

't a3 is C a n mk t

st Cs A, si : n oea v t it t n

la P CP CU ryi n

D a Ci n DU vU I

n v.

FU i

nt F

r o i G

n a w t ss o n

e ee d

t u

_ vmtu et n i h

_ i S Pym r Oe Ov0 Lc2 o4 r e 0 e w 9 0 r o 3 3 3 P

t el d ir 3

'sr sia G nrc t

s o u t c ee t

a a vm e

t S

f O Pmno e r

B Ooi ct a ht a

g S Oer Lce e 1

n f ep W it oio e a r r e

r bee p w t

)

. mo up n

!a 2 O t n N P 1 3 1 d

n o a C

(

P y e

p s

r Og c 2 2 2 2 2 1 y o O int n 'r t

0- 0- 0- 0- 0- 0-t Laive ue E E E E E E T c a Aieq t 1 9

0 4 0 0 2 r e Rin e r

8 4 7 1 1 t

o R P f 6 6 9 1 c r a e e t a n c s r e

R n n l de on G o l

yr c W t ioydd r era o igadi Dseie s t t

ee uo d

e2 e G

D y

d

a. a rg ha ui

misd i t

r r

eba t

e v aanr if d i t

B e i;t md d eic it v ,lola dergbliia i t

- as ss a1 s d e

d ul uaod cf r t s n i z

o s cor o i

hgs e eosaiv s df oux r

t u sn oo w er gino pu r d 3

o r u M inl p i

L r t d a r Cn op U c d A

i m a A -

t a s s A m r e

r P sn e f

o g Oci cin e

n BieGa t v 2 2 2 5 5 I n SsmD/msAe/er ir seh 3

/

0

/

0

/

0 1

/

0

/

0

/

0 t

i t

a detiEyi e t t yt /

6 6 6 1 6 6 u r et g/

t c nr sis ul it l

ne a 5

9

/

5

/

5

/

5

/

5

/

5

/

5 o e caiu bsim ew / 9 9 9 9 7 9

7 9

k c

p ler a poicaem e 4 /

8

/

8

/ /

O eh Coh lec r 4 4 8 8 la Sc( r a t x

e B 1 n A o e t F .F i

t l nD D a b e t a cCf Cn r

4 4

3 4

9 0

0 3 N M 6 S T eOot PB la 2 1 1 4 3 2 N i

c S P f

i c

e p 5 6 6 5 6 0 6 0- 0- 0- 0- 0- 0-S- OF E E E E 1

BD 7 0 E E E t SC 1 7 4 8 4

6 6

2 0 0 1

n 1 4 4 2 9 1 5

l a

P F 5 5 5 D 5 0- 0- 0- 5 5 5 0- 0-C E E 0- 0-E t E E E E n 0 3 0 6 0 2 1 la 1 4 4 6 7 6 4 P 3 4 3 5

(

t t n n e io io e 2 e e P P n e r& n n is n2 n3 u i1 u s o2 o3 t

n r ia t a t t a r i iat w Gi t lst lst cn i

a d n dn e l in lin i

l n

i P H InU iU K MU MU MU

e 4

n a w t

ss o n

e ee d t

u vmtu etin h

i S P ym r Oev0 Oo4 r 8 Lc2 e 8 3 66 4 e w 8 1 99 5 r o 3 1 22 4 P

t al d ir s

r t sia c G s o nr u t c ee t

a a vm e

S t

O f

Pmn oio h r

e B Oct a t a 2 g S Oer Lcenp e n f W i

t oio s a r r e

r ee p )

bw~

mo t

n

. up la 2 3 3 4 1 2 O t n N P

. d n

'o a C

(

P . y e sr Og' c 2 0- 2 2 2 0

2 0- 2 p o Oin nt n 0- 0- - 0-y t Lait ev e E E E E E E T

t c Aieq u 4 0 0 3 8

0 1

r a Ri n e 1 9 3 1

6 o e P f r 1 7 6 t R c r a e e t a n c s ds s e s o ryd dera ion a n R

l Ga e r,

eV gs in o dt inra y W it a a ngh ut Dei d

a ,sD e bi 2 s eiseut r gm B d cmisdaei c d h s gst l

odsA e er 2 n i u a r t dG e

i f

idn pco oh o e

i i

f md dcf ecor h pr db t ic d D doo n

o i

z r

d ul o s cor u aodi di d

As wc t

Csmf D

o er d ut si f ad dE A Mc uw t

pi r 4

it u M inl po m oir ca d

A Gom ss A

a s s

m e r

r P

e o s ci n r

f g Oc eGa Bt in v e 1 2 2 3 4 n 4

~

I i

n SsmD ir i Ae/er

/ms e se

/

0

/

0 1

1

/

0

/

0 6

/

0 1

/

0 t t det E yit t t yt h 6 / / 1 / / 6 u a et g/ i ul ae /

5 5 /

5 5 5 /

5 o

r e

t c sis ne l

5 9 7 9 9 7 7 9 k p caiunrbsim era poicaem er w /

4 9

/

4 /

4 9

/

9

/

8 c O sccoh lec leh r a t 4 4 4 l

a ( x e

B 1 n A o e t F F i

t a

l b nD' e D t a cCfotCn r

2 1

9 7

2 1

1 3

4 1

3 1

S_ T eO 1 1 1 2 1 6 8 PB la i

c S P f

i c

e p 6 6 0-6 0-5 0-5 0- 6 5 0- 0- 0-S- OF E E E E BD' E 7 0 1 1 E E t SC 0 5 1

9. ' 5 1 6 n 8 2 9 1 1 3 2 l

a P F 5 5 4 5 4 D- 0- 5 0- 5 0- 0- 0-C 0- 0-E E E E E E E t

n 6 3 7 0 5 S 0 la 1 0 1 0 2 7 2 5 9 1 5 3 P'

3 3 d a & e& hc 2 n2 n2 s dr 2 a la n An&1 2

e , d e e ,

B1 e& Is& 1 o s

t h s e1 ns a V1 s tnts ie r s n2 n tr it oi t is lot it i

bit l

a on cn la an i

ion i

ai r n on P NU OU P PU PU PU RU t

n

- 2 w ss o 5 n ee 7 t

d _

e t 6 vmt u 1 et n i u h _

S

_ Prymi Oev0 Oo4 Lc2 r

e 5 e w 5 _

r o 6 _

D t

al id s ia r 2 3 s G nrc t

r s o ee u t c

t a a evm S

t f

O Pmn oio h r

e B Oct a t a

g S Oer L c e e i

n f np W t oio s a r r be e r

e w t n

p )

mo up la 1 2 1 2 O t n N P d

n o a C

(

P y e

p s

r Og c nt n 2

0-2 0-1 0-1 0-1 0-2 0-3 0

2 0-y o Oin t e E E E

t L a ive u E E E E E 3 6 9 T c Aieq t 0 0 1 5 5 9 6 r a Ri n e 6 6 1

o e P f r 1 1 1 4 5 7 t R c r

- a e e t a n c s d d dei e d d R o yd d a l

o n e e t e ey G W- r sd sd h s e e h ddl D y a a gh er it ut i so Gso ss di et s di et s s ad p cmisda n Gem d o loapu d B d e i eic Dr Drem acr d s d s d a C , e i md u d cf d rs f

Epr Epr lo d d o d o d n z do ulcor aod o i

mia mia Ar c Ar c lo Dei d 5 o ir u M s. i nl p o c

o c

Cn C hs a A -

i t s m Da D A a s m e r

r e o P sn ci n r

e f g Oci BtieGa v 5 5 4 n n SimD/msAe/er s 2

/

2

/ 1

/

0

/

0 3

/

2

/

0 E yi e seh I - - / - -

- 1 t it a

det r i t t t yt 1

5

/

5 /

5 1

/

5 1

/

5

/

5

/

5 /

5 u r et t c g/

nrsis ne h ul a 7 7 9 7 7 7 7 7

k o e p caiu era polaem bsime ew i c 9

/

4 9

/

4

/

4 9

/

9

/

9

/

4 9

/

4 9

/

c O leh oh ec t r 4 4 4 l

a ' S c (C r a x e

B 1 n A o e F F i

t l

t nD D 7 b

t a a cCf Cn e 4 0

9 0

7 6

5 1

1 0

3 2

2 4 _

S T PB eOotla r

4 3 3 1 1 2 3 6 _

i c S P .

f i

c p

e 5 0-5 0- C 6 0-6 0-5 0-6 0-6 0-0-

S- OF E E E E E E E BD 0 0 E 5 4 3 2 9 t SC 1 7 0 6 6 5 3 0 -

n 2 1 2 2 2 1 5 8 l

a _

P F 5 5 5 5 4 4 _

D 0- 5 5 0- 0- 0- 0- 0- _

0- 0- _

C E E E E E E E _

t 0 E 0 2 6 0 5 n 2 5 0 3 6 8 7 2 a 5 3 6 P

l 5 2 2 1 1

~

f e3 r

k h 2 2 -

o& e ie o a& &

n 2 ic o y -

t m1 m2 O s u1 c2 u b r o1 us ys 1

n !et et nit Lt Lt a qi t r t ri a an i

lan i

an t

.in t

.in e en un _

SU SU SU SU l

P SU SU SU S

e 4

n a w t ss o- 5 d 3 n

e ee t

u 3 vmt i u h~

et in S Prym Oev0 Oo4 r 03 Lc2 e 59 e

r w

o 99 77 P

t al_ idr 7 s s ia~ G n c t

r s o ee r u t c vm t

a a e r t

f O Pmn oio h e

S B Oct a t a- 2 g S Oer Lc e e' n f np W i

t oio s a r r e

r bee -

w t n

p ) .

mo up la 4 O t n N P d

n o

a C

( P y 2 2 2 e s Og c 0-1 4 0-2 0- 0- 0-p y

r o Oinnt n-t e E 0-E E E E E t

c Laiv e u 8 6 4 2 T a Aieq t e 6 7 6 6 1 r e Ri n r '- 5 1 6 3 3 5 t

o R P f c r

'a e s d s e t n c s d ei rr a a n ng l

d n R io r yd d o n 2n atie 5 ag dd aos y W- t a a gh er i ut io it s t

d - d t

isni u

ad essG B d n cmisdaeic asG ci es ss e c eh CloAlbeD r t

.ap rE r e i xeD dGo d r ik g i

di ul md dcf i f

z u aod f d _c ai dt 6 n dDr c i

r o

i r o sinl corpm o d

oo AE A e r

l Deo h pof mo -

i u M M t

b s c A t

a s s

m e r e~

r P n r o Ocsi ci n e 3 e s v 1

f g BieGa 5 4 n n t

er se

/

0 / 2 /

/

?-

1 I

t it SsmD d et i i E Ayi r

/

m 'e t t t yt h 1

/

0 1

/

5

- /

5

/

5

/

/-

5 u a r

et cg/ i sis ne l

ul a 5 5 9 7 7 9 o e t

caiunrb sime e 7 9 9 9 1 era poic aem 9 /

/ 4 k p w /

/

8 4 4 /

4 c leh oh lec r 4 l

a O scc ( r a t x

e B 1 n A o e t F F D

i t

l nD e 7 6 8 a b cCf C 5 2 1 t a 1 4 1 4 S T eOot r

PB n 3 1 3 2 4 i

c S Pla W

c e

p -

5 0-6 0-7 6 0-5 0-5 0-0-

S- OF E E E E E E BD 7 0 4 3 8 t SC 5 7 4 2 7 8 n 1 4 4 6 1 1 l

a P F 4 4 5 5 5 5

D 0- 0- 0- 0- 0- 0-C E E E E E E t

n 9 3 9 0 0 2 4 7 8 la 4 3 4 1 8 4 P

t 1

n i4 2 k leit i n o dr r a e e

P& & o MU y3 B C

r e s 1 f 1 t

n ed en e it kr s lt t r

e t t 3 s t

t t f

r a- un gi on ain ain lo la hl P TI s- TU VU WU WU W m-

n a w t ss o 8 4 0 1 d

n ee e t u 5 8 vmt u etin h 1

i S Prym Oev0 Oo4 r Lc2 e 0 e w 4 r o 2 P

t al d i

r 1 s

r sia G n rc t

s o u t c ee v

t a e m a f Pmn r t

O oio e S B Oct h Oer a t

g S a 1 n L cnpe e i

t f

oio s W a r r r

e ee p bw mo t n

up la 3 3 O N P d

n s a r P y e t o Og nt n c 1 0-2 0-2 0-2 0-1 0-1 0-1 0-2 0-p c Oin t E E E E E E E E y a Lait ev u e 2 0 0 0 2 7 8 0 T e Aieq e 1 4 4 5 1 1 8 7 r R Rin P

r 1 7- 8 3 1 1 1 5 o r f

t c t e

a a s s nl n e n c s d n g G di.oor g , o s R W- o yd dera io n it r l

h e df oge r

fadn ei l D et t n hl ani dieta r e d o

e sco t htsnwom t

g a a gh s y n ut isni aeo sus oh t

cmisda is 2 cg d epy f l B in i f

imdu aod d cf eic d a dotss i

ri o

gl c e d d nnl ai i i

am dp l Mnoc ecr dsm l

o ot x e n

i d

i o o ul sinI corp o l

er o loCam mo ou np 7 o d es ausr l

e I deC B M d o o d o t -

i t m d df r A A R &r dn A a g i m n r i o

t a Oc si n ci n r e

f r BieGa e/yt h 5 2 n e SsmD/ms t

A 1

/ 1

/ 2 / 2 1

/ 1 ir i E yi ele a 0 0 0

/

0 I

t o det i t t mw t

e /

/

- 6

/ 1 4

/ 6 /-

5 6 4

1 u O t etc g/ sl s i u nr e 5 9

5 7

/

5 5

9

/

5 7 9

/

5 5

9 o 2-caiunrbsit era polaemx ee e r !

4 9 9

/

4 9

/ 1 9

/

4 k i c ve /

4 4 4 4 c A leh scc( r a ohec 4 la e s

B l n b o a -

F i T t nD F

D t

N N t

a cCf Cn e 7 2

7 6

8 6

8 4

3 0

2 4 M M S

PB eOotla r

6 3 3 5 2 N N i

c S P f

i c

e p 5 5 5 7 6 6 0- 0- 6 0- 0- 0- 7 0-S- OF E E 0-E E E 0-E BD 0 0 E 7 0 0 E 5 t

n SC 3 8 8 7 3 9 3 7 9 2 9 1 l

a 1 1 1 1 P

F 5 5 5 5 5 6 6 6 D 0 0- 0- 0- 0 0- 0- 0-C E E E E E E E E t

0 0 6 7 4 0 2 n 8 7 6 9 8 8 7 9 a 1 5

l P 4 2 2 7 7 1 y d r

r 3

lo _

e3 k 2 m k _

F& ic& n& A ic r

r s2 n w1 n e e2 e t a

s o p d n t

n wts nt s t n o st s a im r

P z

oi ui o ei u la r n r n i l r n e t i

P BU BU C C DU D F F _

w m

n a w s 's o n

e ee t

d t

u vmt i u h et in S Prym Oev0 Oo4 Lc2 r

e 5 w' 9 e 5 r o P

t al d i

s a r s G nric r t s o u

t c ee t a vm e

a f Pmn r t O oio e S. B Oct a h t

g S Oer a n Lcenp e i

t f

oio s W a

r r r ee e

p bmo w t up n 1 4 O ) .

N la t P d n n o a C P y 2 2 2 2 2 2 2 2 1

( Og c 0- 0- 0-e p s Oin ne nt 0- 0- 0- 0- 0- 0 E E E r t E E E E E E y

o Laive u 0 0 0 4 0 6 9 0 9

0 0

0 2

T t Aiet q e Ri n 8 2 2 3 9 5 7 5 1 r c r 6 2 2 o a P f t

c e a R ,

s d e r s s d. d e n e n d e n de d s d

e e ote e es l

R o yd d a o h d

eyr s d

eyr s io eir ee h hst wa h t

a it a a gh r er uit s cr e cr e et h et di t - s leir s y

cmisdaeicn lateg ateg vac st a

r s d da eytdr e d B d r r lai da gs a a W- i i

f md u dcf a

o paa t lpaa ebh c t

Vid f

ab por lo odt a t

o 8 n ebh c i

g dul o s aod low d eb l R o lea a l

o n M cor inl pom e R ee Uc c f

s e

A i

t i

l d m dN d d d a io m

r B r o g Osc in ci n e f

n n BtieGaAe/yt h 2 2

/

2

/ 2 1 1

3

/ 1 1

/

1 4

I i

t a SsmD/ms detr i E yi le i

e ma e /

/

- 0 6

0 6

4

/

5

- 0 6

1

/

- /

5

/

5 t

r et g/ sis t t t u e w 5 / / 5 7

/ 5 7 7 u e t c nr bsit er i

l nr 9 5 5 9 9 5

9 9 9 9 o p caiu ler poicaemx e

/

4 9

/

9

/

4 / /

/

4 /

4 f

4 k

c O a eh oh lec e ve 4 4 4 4 a Sc(C~ r a s l

2-B n A o e t F F i

t l

b nD D 7 6 N 2 N 7 a a e cCf C 8 8 3 M 6 M 7 t 6 4 3 7 0 S T eOotan r

3 1 1 8 N 4 N 1 c PB .l i S P f

ic e

p 6 6 6 5 5 7 5 6 6 0- 0- 0- 0- 0- 0- 0- 0-0-

S- OF E E E E E E E E E BD 6 0 3 8 2 0 0 0 0 t

n SC 4 3 2 3 8 1

2 5 5 7 3 3 3 3 1 3 5 l

a P

F 5 5 5 5 5' 6 5 6 5 D 0- 0- 0- 0- 0- 0- 0- 0- 0-C E E E E E E E E E t

7 3 6 3 4 0 0 0 0 n 7 2 3 6 7 3 6 5 1 a 4 2 5 3 l

P 1 2 2 4 4 k 1 2 f

e 2 2 lo lu e & ei li n t

ei t

G r k & le l

i n d

C le1 ic1 r c MU MU t

n h1 ct h2 ct e las es it n en t en t n a p Si t t o

a r t i an t i an o an min ni io ni io l

P G HU HU H LU iLU M NP NP

.l

n a w o 0 3 t ss n ee d 4 6 34 e t u 2 2 5 vmt u eit n h 1

S Pryim Oev0

- Oo4 Lc2 r

e 7 e w 7 r o 2 P

t al d s a' i

r s

r o tnr ic G s

u t c ee t

a f a vm e

S t

O Pmn oio r

e B Oct a h t

g S. Oer L ce a

e 5

n np i

t f

oio s W a r r r

e ee _

p bmo w t n _

up 3 O ) .

N la 1 2 1 1 7 9

t n P d

n o a C P y 2

( Og 2 2 2 e

p s nt n c 0- 2 0- 0-1 0- 0- 0-1 Oin 0-y r o La i e t

v e u E

6 E E 9

E 7

E E 0

- E T Aieq t 2 9 0 1 1

8 5 0 Ri n e t

r c r 3 5 6 6 4 3 1 t

o a e

P f c R a ie s l a

e r e n ,c s n t ot e e s r d es R G e l

sr l n u y

t a

ioydda t

a a gh r er utio so er io ot n te in u

db ea cc li oa c D t c h s

dlio iegt nl r

n cmisdaeic r c ut it ud n B W- i ca sa s or lad t s A 2 t r o d f

i n di o n g i ul f

d md o s cor u dcf aod i

de re eisc rd sd oy pe et mp r u d

e sm t

a lo omc o n M inl po d&pn i r

C h Rec la d

d in c e Mod cn 9 i

t i

l .m d A le A A d ina A a i o s M r

m B o g Oci sn ci n r

e f

n n BtieGa h 1 3 4 4 i

i t

a SsmD/msAe/yt ir detiEyi e l

e a e

/

0 6

/

0 6

1

/

0

/

0 1

1

/

0 2

/

2

/

/-

/

0 1

t 1 6 /

u em w 5 t t t / / / /

u r etcg/il si s nr 5 5 /

5 / 5 7 5 o e p

t caiunrbsit er 7 9

7 9

5 9 7 9

5 9 9 9 7 9

k era poicaemx e e / /

/

4 /

c a

O leh oh lec scc r a v

e

/

4

/

8 4 /

8 4 4 /

8 .

( s l

2- -

B n A o l e t F F i

t a b nD e D N N N N t

T a cCf Cn M 7 4 3 M M 5

7 M 3

1 S

PB eOotla r

N 8 4 N N 8 N 2 i

c S P f

i c _

e p 6 7 6 7 5 7 0- 0- 0- 0 0- 0- 1 0-S- OF E e E 1

E E 1

E -

BD 0 1 5 E 2 5 E 7 t

n SC 3 8 2 0 7 3 2 1 a 2 4 2 1

5 1 4 9 l

P _

F 6 6 5 5 5 6 D 0- 0- 0- 0- 6 0- 0- 5 0- 0-C E E E E E E t

0 3 0 0 E 5 E 0 n 9 5 3 8 2 5 7 3 _

la 3 5 5 1 1 4

P 1 1 k m a e o s n _

e t t 3 ie2 d n

n2 a -

Bo&

r .

C i&

t e h& t r

h 2 C 1 B e 1 u

n oee t e cst y im ds t r qs mk n t s ai r r

r g ai e si t r n la y en e li un v un ea i

P O PU P P QU R SU VY -

m 4

n a w ss o n

e ee t

d t

vmt u i u h.

et in S Prym Oev0 Oo4 r Lc2 e e w r o P

t ~

al d i

r s

r sia G n c t

s o r u

t c ee t a vm e

a f -

t O Pmn oio r

e S B Oct a h t

g S Oer a Lce e n f np W

i t oio s a

r r r ee e

p bmo' w t n

) up la O t N P d n n o a C P y 2 e ( Og c. 0-s nt n p Oin La e e' E t

y r u T t o Aieq' iv t

e 6

4 r c Rin r 2 o a P f t e c R a s ,

e r e n c l n d R t io r yd der ai o eds y a t a a gh n ut a hs ecter B cmisd eic dar e W- i fi md dcf u alpv o en 0 n

i g d ul aod 1 o n o s nlcor p o l e

r i i i M .i m d A t li a o m

r B r f

o g n

Osn BieGa c i ci n et yt .h e

n SsmD t

e ea detiEyi Amsl i 1 I t a ir

/

e /

/-

t t t t g/ isis u mw nre r r 5 u e tetc l

9 o p caiunrb era poicaemx sit ee /

4 k

c O leh Coh lec e ve a Sc( r a s l

2-B n A o e t F F i

t a

l b nD e DC 1 t a cCf 1 S T eOotla r

PB n 6 i

c S P f

i c

e p 5 0'

S OF E~

~- BD 7 t

n SC 0 l

a 1 P

F 5 D 0-C t

E n 3 7

la 1 P

t n

na ol g P t

n r ia t h e2 n slcit la a un P WNU

n a w t ss o n

e ee d t

u v!mtu h et in S Prym Oev0 Oo4

- r Lc2 e 0 e w' 0 r o 3 P

t al id s

r s a' tnr ic G r

s o u t c ee t

a f a vm r e -

t S

O B

Pmt oio r

e _

Oct a h t

g S Oer a 2 n Lc np e e -

i t

f oio s W -

a r r r

e ee p bmo w t n

up a 5 O N l P

1 1 1 d

n a g P - y e Og c 1 1

2 2 2 p n Oin nt n e 0- 0- 0- 0- 0-y i

t Laiv t

e u E E E E E T a r Aieq t 8 2 1 0 0 0 6

r e Rin e r 2 9 S 1 4 _

o p P f -

t c O a r r p e e n c s di.e om io r yd d a n f

R g l es t mpu n t a a gh er o h s sor d

ee y n ut e i dt B o cmisda eic da c t

se r

a r s s L i i ul f

md dcf yt go n d u aod i ode s a pr o o s cor l w

U c 1

o inl p o le u as 1

i N M m C dd -

t a s Da Fg A -

r m

r o -

o t

c s r _

f a Oc BieGa in' cin e h 3 5 I

n e SsmD/msAe/yt t

le ae 1

/

i l

2

/

0

/

0 1

/

0 -

R i

detiEyir e /

/- /

- 6 6 1 u em w t t t t 5

/ /

u 3- t et c g/isl s i nr 5

9 5

9 9 5

7 5

7

/

5 .

o caiunrb sit e 9 era poicaemx etr.v 9 9

/ / /

k c A leh( ohec l 4 4 4 /

4

/

8

/

4 _

l a l e scc r a e s

B b .

a n T F F -

o t nD D -

N N cCf C i

t e 7 6

5 2 a M 2 4

6 M S

t PB root e n la N 4 1

N -

c S P i

f i

c F e

p D 7

0-5 0-6 0-5 0-6 0- 7 0-C E E E E E S- 'O 0 0 0 1 0 E 1 3 7 1 0 4 t B 4 n S 5 1 8 1 7 l

a P

F 5 5 4 5 5 D 0- 0- 0- 0 0- 6 0-C E E E E E E _

t n 0' O 0 0 3 0 a 4 P. 9 4 1 _

l S 4 1 7 1 4 P

y r .

r e 2 k

c F

s m e e

n o n1 a ee o1 1

t n Rt n wt dk nk t s d c in lst nt la gi io oi r n ae aa l in i

ion i

P BP BU HN MY MU ZU

wwm--

-h , i

.s ,

f

. 1 I

APPENDIX B t

COMPARISON OF SELECTED STATION BLACKOUT CHARACTERISTICS l

1 i- I

C:mp:rirn of S:l:ct:d Stati:n Blickout Ch rcct:ri tics i PLANT LOOP initiating Frequency Extreme Severe Selected SBO Weather Group 5 characteristics SBO-CDF/ Coping PRA/IPE Actual frequency / number of time in minutes /

LOOP events at power since Aac access time in commercial operation minutes Pilgrim 6.17E-01 1.48E-01/4 E-10/8/10 Crystal River 4.35E-01 1.36E-01/3 E-06/4/-

Fort Calhoun 2.17E-01 0.76E-01/2 E-06/4/-

Fermi 2 1.88E-01 0/0 E-07/4/60 Turkey Point 3 1.7E-01 5.18E-01/14 X E-06/8/10 Turkey Point 4 1.7E-01 0.76E-01/2 X E-06/8/10 St. Lucie 1 1.5E-01 1.73E-01/4 X E-06/4/10 St. Lucie 2 1.5E-01 0/0 X E-06/4/10 Calvert Cliffs 1 1.36E-01 0.4E-01/1 E-06/4/60 Calvert C!iffs 2 1.36E-01 0.43E-01/1 E-06/4/60 South Texas 1.32E-01 0/0 X E-05/8/10 1&2 Nine Mile 2 1.2E-01 0/0 E-06/4/-

Daane Amold 1.17E-01 0.4E-01/1 E-06/4/-

Browns Ferry 2 1.12E-01 0/0 E-05/4/-

Dresden 2 1.12E-01 1.07E-01/3 E-07/4/60 Dresden 3 1.12E-01 0.35E-01/1 Millstone 3 1.12E-01 0/0 X E-06/8/60 San Onfre 2&3 1.1 E-01 0/0 E-06/4/-

Vermont Yankee 1.0E-01 0.38E-01/1 E-07/8/10 4

Millstone 2 9.1 E-02 12.5E-02/3 X E-10/8/60 Palo Verde 1 7.83E-02 15.3E-02/2 E-05/4/10 Brunswick 1 7.4 E-02 9.1E-02/2 X E-05/8/60 Brunswick 2 7.4 E-02 0/0 X E-05/8/60 Indian Point 2 6.91 E-02 20E-02/5 E-06/8/60 indian Point 3 6.80E-02 4.3E-02/1 E-06/8/60 Robinson 6.1 E-02 7.1 E-02/2 E-05/8/60 Point Beach 1 6.0E-02 10.3E-02/3 E-05/8/60 Point Beach 2 6.0E-02 3.7E-02/1 E-05/8/60 Salem 2 - 6.0E-02 11.1E-02/2 E-05/4/-

Oyster Creek 3.26E-02 10E-02/3 E-06/4/60 Ginna 3.50E-03 137E-03/4 E-06/4/-

Palisades 3.0E-03 71E-03/2 E-06/4/-

B-1

': 1 i

^

'4'.

g; 1

1 I

APPENDIX C STATION BLACKOUT RULE ACTIVITY AND MODIFICATION

SUMMARY

.. i l

i i

I l

I i

r; Ea

l i l j1 1 1]1jl{li l i ;l]);j ! ll l1l1i i b'I q;- ' .

+ -

t a t a

t a

t' s c t n s s a a io n' n' p s t it o' o it m t n a a a i

t a ic f ic s lp i fic fi otn 8) h d)o i

d' d Ce h o) o) dm 0 c 1

a mca mhc mhc ee (

e 5 e 2a 2a t

ap l

M

(

( e ( e mm 8. 0K M K' 0 MM MM

- ii t

so 75 33

0. 0 1 2 00 21 00 42 ET $$ 0 $$ $$ $$ 0 0

2 2 2

1 1 ~ 1 n k o n y

s o c lb io e r y ir ia Z r

a m e D o B m

~

c 2,

m t 3

i L )

y &

m u . it 4 1 u O & 2' c & s S 2 a a

& p 3 x n e 1 a t e

1 c n i

o m y k s io T

t a a r r ic2 s P h N e w& e t u

c F s1 c y o

i f

t n s n e x e S i

d la n ue t kr o P w r ic o u 2, o Bu ,L n T &

M B r

2 .

(

2

, 1 d t 2 h n 2

& S & '& t c

a &

1 y 2 1 1 a y

y s 2 d H r t 1 le& f i &

f n , r e

1 i

la1 2 v d o Ve l

C1 la & P C i

t o r n t n Is 1 ,

c r n

.A w id eo vt eyo ie r

y e o al s v n t e a ia l

r n l r ei l

laa r a il u B BM CC P F C R

t t u s o

o k C c d l

a e @

t s

B a t n

n m a it i

o s lp )h t E 0c t

a d 0 a S e 1 e t t cc ea (K

M0 pp 55 xm Ei 33 e' ie y h' .t c t

s

- n g a u s t

in n br o d

d' i n G n t

e .i a D G Gcd d u

e G

D D De i

r d E G e sdn

.t y E Ed G r E

D D s r

cs is t

g gar E s S ya E ss i

ic a n n ig e C

- ny las p it t p e o c P ol is s iu n x ia t n ne ar a

c x x r o t w e H a

c a gd u s e eo d d g g n

i f y ne s m md e e in' it d e i

pc e c o r

o rd e d d

d d

t sG it isx of o or x F Fd A iD x

Mas Cp E - - a A- - EE E

e e

g .

t c K a

p io00 u1 im d$

et t

s ca o

Ce on t r s pn 8i o deem t t

2 a

(

ic l

ap K K K ii t

mm 0 0 0 0M0 M 0 0

Mih f )

8. od c so 5 41 3 5 2 a ET 0$ $5$ 0S O $me r,

e n r v a e2.

) .

i R Smr t

n s l

ta t

r oile2

. me o e e2 sFM&uw So C

ke n m y ,P

( r ,e1 o no k n 2r y

r a

t c

u as Yi n Ccim&e

,irNe 3t a , la1 lc a

m O eo b

&P2 S u n

iR 2t iz & s, haN m a yn u e M3. 1 ,F1 eo o S m ydl ,ed ut aqg n a 3& r l e loas en N r o tn 2 i i t n io1 FmSl iaShsa t

a c Pe sAaP ne ,okW L ,1 i Pla ne wn f an oal,l o ,

r u fu leo cr 2 i

d 3 3 io o d c BDGib&na M

2 &

& 2 e&

e2 InO.

.d oe1 t

2 2, n, 2n t

d k 21 , kn m && i&aMSan o G2,3. 2 n e t n hce 1

at o P1 r n3 a e r

ioadr Yt o 1 1

&ad C y C Pe e t nB ae kk n&

co h r i

t v

2 ie r

n BV oh b r wiu oo u1 2e o RCot eqe ur f

i t mt r s t

ia no mc r a taG h inr sta f c ey d iloa ee ac igClaonu A FO InPP VP CM BDCPOSW e

lu t s

R o t C u d o e )

k t a s c n l

a it m io t

B E s a n ic fi o d e d i

t t t cc o t

a ea Mm S pp xm 50 2

. EI $ (1 e

h n t

o e i

n it i n y s br g t d u i u r n l

e b e t gt y o i i

b it d m it e snl i t a e d d ia e

i o st i

aibt gsa r

e it e le r :

cs c soss- i i n s s

G h r E

c or Exle reow D s G i

ss ny r e g o 1

s gc 1

s det ogc r

R d a D ola s si e vd E ix lo a an it nd r nd a aio e e ysd e f

ie t di t d d ic l esd l

d r n ye v fi y t Cnxd i Co pd hixd e r r ei o d e of a i

nbEA oru nMImd o

A eEA s

p p

t t

aqe u r p

Ms Nt N U A Br I m

f l ! i lllill! ll L l

y .

m '

t t a c s a

p n

_ io i

m t a

t c' s i f

Ce ont i d) oh dmee mca t l 0e ap 1 mm M' (K M ii t

so 3 1

M00 55 5

3 2

ET $ $$ $

L

, 21 k ,m

) c ,1 ele t

n irk t

aed a o s PerS e z r e C

( m ICV 2 t

y o c Feon l r t . pl ao a

m O u d loH o Psn i m . s, b m A 2eo S

u e m

e&dR n1 a ,2 1 a a isd n N u yla n&

t n

i t

o t n

DleP ,ra ,B e 1 e

io P

a la nF ke et rl e i

c P t o , eio vg l r

f i in srRV r iM e d l Car irC , , e m o ,.H e ia m s i n m M 2 t r N u 3 d &k, yl sge x , S &

r n 1 cOPiT e 2 e 2 2 3 a kc N o, y, h & v e y l r

r u t

1 i

R & r C w ac e o f

i t

v imle e la 1 n 2-s i P S O l

dt la h P i

t n d n; .

S t

s c n c u r ao32 a y

r t

a a N A BHM&& L C H S W l

e u t s

R o s t C n u d o . t o e it

.a a k t a ) .ci s c s f n a m n i

d o o

i l

B t

s o i t

E it a m a )h n d ic r ic c f a o e f o i e d

Mi n i

i t

t a

t t cc ea' Mo d

9. m oK m0 M S pp 6 m M 0 5 0 50 1

7 Eix m' '

5

'5( 2 1

$ (1 $ (1$

6 e o ht t

n s i n d o E y e it r I

it e a s e s r

d p~ ic s s t t

e e e G o fi e la b a it t r

r cs D c d i

r C s e

E' o o e - e s v ss i

n t m

t t n c o n ny s' a o la r i ol ia a d m b n pr s c e y ee c an t o e d d yr d ica f

i m t e ee r g e a fi t y la' u r s

y d dt t

/ r d a d lp e

L A

d e q o n

sep d A

d a Ab d h Ac d

A R T of Mas R e i M

Co Dc O

T

F 0

f l

APPENDIX D OPERATING EVENTS

,!l { -

j1l! {lji{ll .l1! 1l!

t r

a m, o

s s yroee mr ie t

r s e a d st h t

D olla e syie t

C t

6ct t

a wlar neian vnr e b he s u r

ahf et Gd sah rf cl eoe ,t o.e De oGt eetiw u u y g Eco s b nsa erum or r hDsl iaybt d wp u 8Eahtrh t nh a ,hdoa t

t o o 2 e w l gist dsn t f t sde8 e oe i

H lynGs 4 e oD laio nrt ieso r c r

eu nd arEiicnctr t

'na me n iswi o f o Ode ivedt xen et lel cu Gpet .y h

T xwoluea f

Dt - o hde t sela pot,s e net ym rl t r

e ph On of e nud a ;o er v dnu ne t

Bs v s is t

a S isun a e ma e rd fo a eno eurh t

u r

Pmh gai e aCmp hbrd n 0 w 3 r G ol u i

., t Ototgliwq n u nat l

o m 3e w s eirac t 6 r oo e Oanoae h nt 9 m Lf i rot a oucOa t

l r

wsess oh i 2 g

f p Pe r Oo i

T y

dr e

t o

Dme. oh v e t s haendd t t it n

Ots e t s

d ole n r spyot s L er, e uell r ot e nia a a a o v a r a fie ef le l

s, o cn oloegeer cd ih sb t

it dt et c dn upd n aw t nt Gla n sr e a e u m erl e. deI f

Dia u uo f s R dtr oa .Ev h af ce eb a t

n g n

dat eamd hig r t t imGen aDh u t

o t ch e

e ya r sfo nTl g b i

r -

duT t

v i v ht lyo a god n n E a c sl r ice.e n o rBae o s .

it lyfu PSHe l

f g H wl aib n ao r P usr s a ssd igr b 1 O bi n 0 ue es es Peh .e e ap D i

t 9 encg pn St rs v O s a

r 9 s ac adenioo s Ao ua oh L a er e

p 1

e Bh emusmsia s ausr t

h e rh y t

e0a d

n hr pe c t a d gr O n soad t b w an e f w2m e a

'v o cep tooplyG p

i - r o t ah S a ed dad oi sy r ir lo c r Dre ar n naeb asteD pea t

it isyr e t n ew omh d i mO n

w e h t spiB l

u ete o v t smt oetet ofo es apx a ht a

P E dGoh s suoS t e e eDrt h n er g isy s ph srpe d e inbo h aEl oc eTed r s st a bu a t u

l t r i

s t

f o mer in se .cda n nn t a o eu f aht al c t f

O n dTonsepy r t iai eve s l u

ad . y ec o

it o . cle u0ll t 4f s u ntnt e er d sl a f

d )s e ar b a gs e el n dr q u a o ip a eh r v spu i inb s

r c mtu epiene t ese o w e s e

o inmpoma k cc tsnled h - tg le pdee s t o u ooi8 l i ioaa s D Amf nems r it Pnw2 H Fiinn o

L -

8 1

- t n

9

/ iet l

n D e e 6

4 t

a hwve e v e P e t p 3 l

n E y li b

a e ?su t

- R9 2 h Oe p m O E9 w Oy T u Ota 6 B L9 & 2 6 t

.c Ot r 1 d 1 9 L 6t e 0 S o LS 0 9d 8 oyr 1 1 0

0 19d O D f r e f d -0 9a eB et eoo cayc t s89n s/91 a tr a o u n69l a91 e r 296eS 6 , r pb r t

/ - /

nDa r a e6 ,o ee ls2 ,r a4 yn ee r,

e B4 42 d I 8 9 e r w4 acw e b1 rt n mR s3 e at RF ie2 2h re ut e aw t

neevun emd i vRnmn P 6 , r iRne t

aRb ap r no aEuov e f

la S 0 a t PRESa rEu PLJWp o aEel t DLJTe A0 - CLFPa u

l r

j !I l 1jl ll lll:l\ ii o

g d nd ie a

t u

oye ssy yaa t

t e

d bla a n al le u a ae d2 d e q

) . e d d eid en n t

d et d en t r n i f

i hrUa a o l v m o Tne t i

C io r s . om m

( r ps r e spi r t e s s s e oOe p m awas r t u r o aOh t o o f s l e luLf s o t

g H n iat ina K n

a t a i 4 t r eu d q f dr u t o

o n onb _

a a .

oe n a e h s

h d n N d a tosh T e si o .t i

atuk et le r

e u ara t

si em i n no b u

o t

a co t r ue miomt r ds t

inp a g t

. e r ne t n 4 n e a r mto o8 1

v _

G c r in

' r i s

de 5 g a n _

s lew 8 n inoelc e e ia 3it g o f

t f p o d1 x ._

n o r

m e it o E _.

i rd it h t

T y

e mye ss lu n a sUl u t

s r r a lae l e r ns e e ol P b t

u f e u hpPo In v sd n Oor n o Ot i

s c atr tr fo Le mod m t

n e R e

af ,e aiv 0Oe 0L n 5 7

v g ep n s in 1

as ig 6 1

m.

E n cu n n - dt e to d e g 2 _

ex ped n g

i v a a t e ss -

i n a n e lut u ua it so n. D _

H el sa sacd _

t a t c n eh t laia t r 0 l r t s u a e Pr _

e i

p 9 iao gs gs t

Oto me nra _

9 r r O 1 rh sepe it doe nt p Os Lr e e oT atao c

f e

t p s us are ._

n c s ga n _

i ivt e ini t dsG n iw o S t r u r , i niD d e p r

+ e si n ur de v wdE e lt d e c ht pahe ue o 7m e f r w E o n sy r w n. _._

o e miT P h t

u1 o9 e hi r

oo 0

m. r eale p

5t a t t

a s id _

e f o sg nyr ,ne r t t t un a r o 3 ot s n n u e i i s o bt s iets 9 c u n .

f f ea icv n or e /eiy 3 om O

it p nl oP h c r 1 h mrev is r

us i

cr e / t pe e e e r

f o c hO 3 r8 o s s _

s w nte0 5 c e

t iO t o s _

s D WL At hp Ola1 er Aa _

e _

s s e o le le li L t i

h ih h ._

n w 1 e w w P _.

- v P P O P D t n E O O O 4O e e Ps u O L 9O l

m Ot a 5 O L 1 1 9 L b

a u 1 49 L 7 280 3 d n 1 T

c o

Ot LS n1 9d o01 e -

0 04 - 9r e d &0 99aoew t 08, 4 1 re d

e D f r e oo y5 n9 ,1 er , 4 9 9et 1 k 931 , rle t d

9rt n et cayc t a/2n t n 1 n r ic5/ 6 r- u

/ en r a C5r e w 2 1 1

/

1 ,

e e 1 eh 2 1 be wo nDa e e 7 eco- d c w2 s3hh s m3mc 1

d r t mR o2b m31l t o n - w ct ale leRvet nt n e nemd t

uRrrEaeih aEol h u a l

t f bRtont a a u a E pl leRinap r la evun iEc PRESa DLOPs l h SLAPa BLMWw SLNPs t l l ( ((lfl(lI <l1l illll i)I Il{lf

. jl. I I)!j jjli!lljljlIlllljj!lilj iliji 1IlIi

_. e '

y f eg ly e o nn e oi n - in a hekur d .-

t r g aa a . oa i n

meg st ood Ao Twl et a lu t

e m i

lt s ixh t o ge' ute oc

)

t n

rt pw pi t

.sh ed nnt eont h ide w t a re loh su e n dn ri a o as viof o rr a ea t

ao vt f o ems n e C

eficw

( gof e cl ot li0t e e r 4 u s it n s hi r dd nsW s sg e2 in s s ont oGk t

r h u lalo m mand 0 lota t o

inm o a ov l

P de o eE0' nd0 H f r eev 4 a re9 Os r df 4 Oua esivne3 ac eg 3 e eo s

t on r t i n L c vnih5h oert r nt u

a s s t esh ei t a at ct e x u giad ,

cl o mrn w h gh a t

T nt r eHnr me i t

t a a P . ned o ep e odf r

e ise ur t

nt t e o drl ue ep hd t

eO cb o t

at u olect eah e r sO mel e

,~

a e cs ppf ht o m e ma uL ov oa dv e eo umE snd cn . r ce i f r t bh G it ot h s m . ec d o* f s atf elds

t. wa eG nua s eo ied nh o aow r n ohoseD l r e h t e t

ps u rb E a r

t n cc r r

m r u lya e ot o g t dai o i

ed . i sr*,h n n aot a aer e T v v 2 'e ,w w o y ole yb s. aGla 9 5. s i eDtn9 nin f

osu l

r c op 1 2 np t

r e ae hee yi oa eid r r e at r e v

o lt cl s eau i

vb eym0ea m e lue r t eee3t c ai

,l t f h st n

eht e ao.

s c ir r ia nr t t s nde em l

da nty wfa p0 m d f

e erfao t f n e due e sp2i i h iat r it e R e u xot ht t mot o eh v g s s r s ea f luw o s rsomfo

. r t

t yt u on r n nt E n a e t Aupeg br ec o g i pwt n O d . ug pl f oeaf elp o n v ,sla BemAr a riun dmc e es e 3 a s r a t

hp Gp Sd e icnAc -

i t

a H mDe rdtsmfoOa oa yo B r

uf osr v e n D elicsl .

r esf ru nSW r .asi a e

p 0

9 5 t

r t h f de e k cna ap s eerba 9 4 tad s nn r e 0 o sr e f h O 1 1 sn t eniooh t t 5 at r wred oT du e ok - a et it t e u . eh l

c t cs pcbtuadd ea2 8 sh n osc t n pag ml irSiHe i

etot a eok ubD l

oet sgf ;a t uns et r oo i

S t s et chin .ht inae inup n n s t

,di t sWfo mf r l iyt nmc t

r e dt i r e v nst a ol ela aGk n s 4 oe o ne w E iw n- s - it ri a

uc rnE0 r

eD 0s e 5 em i

s1 s 6 gf eh t o e h ok c t 4 r r 2 re o P

e h

t i5 t

w _la eb icOceW3x f

di BleG k y e 4 c luo gias f im s ngmn t

f o , h oSer a0 lein s ear inf n t nieio i

n eTe me4 e5t an laht atsga r

t s o n h nfh ndlc8 e lant t

N a csn

.t t t u2 ie me t

e n O

it p Pdh ise ir iye r

ud aew r lao aNlaxb OnT nPum oOs e ut3tnir no e f c pl t o s h e siigpv Oa L ss

. hOrl iL up t

s e~ 5. te h e noipar uu e D A6b TrUPoah Abb Wahi om s

s o

le L . t e i le

_ n p~ h

_ 1

- e Pt y w ih v w D t n E OO t r P p e_ e Ps u OB o p O o l

b .m~Ot 4 LS e O o a u c Ot a 892d 3 09t 0 9 e e r, R 72L n 1 9 d 1 4 n0 L

d T o LS la o09 er o0 1

.D f r t n2 a w t s2 - 1 ,e y - 9 9r e

e oo i

- 1 ,

le o n 3 n91 1

e et caycr a t o/4 r 9 P 0 2 r- p e 99 in9 2tn r a/ t n

C57 7 encwo

/

m1 b 1 2 ,

nDa o6t e e e e y5t et c t r ne t

n mR e2 sh kRg utalen a t e h lpc R2 s - w o2h ct d ut o l a evu f emdn r uEu eie h v pr ua B.. ERgnap u bRr a

iEaa nt u h

TLAWwe l

SM DLMPs l l t

PRESa HLAPa

~

7<

- t.

o e e

h 0 t

md a6 e oronla f r f inn i u

s lt u eg dd eer ePs

) sen eke t

l uo kOao o t

r n rot n a Ye aw ee st s bLs o s r r e r sa no C atn N e s e ,i

(

d o e wis h a sst ur c s

r u

emh yrd t sTw e .r b ee ymps ae v se r H

o levan lar w ve o c

nf ion esg 4

dsloee t emp r gn n r ai aek goe er wo wvn i

uf t an aa le sis nf f e mtel h ntrY d vo l T o iot t n et aO r h

er f e s re r a t o h t ,o n e. t e

t e r s onm ri c tor mr a

e t

sie ec ar e wt a an t

rh o e ofs t

G r r i V g.

r n nn o

pye sn wbn a o

ehcet i r t pf r s

e wee s ot wt e t

ai t l a ain s se ot m Pyt inu ixm ie sv i

T a be r e uar o lplore p s srle sa k

uien r y edf r e emer h t neac s d atee t

e uceb h nvi t

n v nivbt l it ir hluo s o c ime our i wae t r sh t

n e 7 Sprr c m ms f e er g eeo ou i

R 7 r nc m oe e 2 e or t ht r v g r wie st no y u cr nh E n ooav t

s f

sat g i v

f Pk :

o r os nl a ad 4 n P d ns er a rb e it H a On ue t at a ew t ei mo r

D a

r 0 Olami L gmo r it w

e wG vesr e es t

e 9 n oir idr D r p 9 a Ecp f eE sa s O d eyn eh a 1

e ewd g se nn h a rA r pw c uNai ,m t

f p

. h ear in a c rd e osP t oew .e o nc t

sdO uo S e naO t t n df an rdl o pgP .

r e a t

e e yr eI sc ed L oi t v r a . g E hi ey cea df s ea w

o e c uqdi r t nn ae dn eat m l

P h it wrh e o oh l lip a lb t en ua t

eo a e f o s eRtu v r lyir p cmid mon f

t i

s n o

hem.A t t i cg e ut t sne r Vfo5 k 9 s f it info nein atrto s. 5 e5 a O p ir ;t hd ge ae rt 1 ugw s

f o c s

ut agvti h c nv idr nn eu 1 ectairne s e f n roiw uan ei h e se e D AlePS Digm Tblah t s

s o e e L t li lietn l i

h n h h e 1

e v w wve w D t n E P P e P e e Ps u O Op 0O l

b m Ot a O Oy t 1 9O 9

u e9 L 2 L - 3 1 L a Ot e0 0 1 t 0- 9dO c n 2 d LS kn 019 rde T o 9 eB i02 o e D f r e oo a1 91 e 1 1 r

eS P 9r 01 e r

et Y9 9 , t t cayc t t

/ 1 n . r e/en n73, nec er

/

r r a t 1 4 2ec e i2 0 b ee l

oDat n mR e o22 - w 23h -

0 w M2 mc - w rRinr p o t r et o nRr ct n o t

m eRv n ap ne evun emd l

la f

e E pla ioEaap l n

iEol t ZLMPa t t PRESa VLAPa NLNPa

~.

- ( . *f

_ e y n t o it ua t lnt o t n O oie uh n d aa ch ht o saS B r t icd gct lu nice ni r ,

seawso t

dre edweh t eat u

pin seh el nst d c n u mat t t vd onieo t l oer eo wsf uo enio t cail d a 3esieoa t s vr r in nt s o. d u el un u de oci n Nscpaari I ig t

t ei t et e e;e

.f er et a swh f

mn e o n st ht inrh blaTeTs mht c ic l

e wafono tum.s ri i sy . s a t

yer ei sl aa bl uow

'h rees u a t

s sus t ob rf t u ee

,d ciic af d

g e8 etne , dr d noce t

os i sceo coh h /9 i t 6 gcirh z oz r

ep rTm ta/nn1 eere r got m e, x .e 6 eheiof h i n es yt nh t t n - ebr e va r si o t dOeunoc e dt agr s

e v adt r na i n

muog s. a s a odrh e

e odop ce c s rbt t

r at o g

n remrOa t e

d ohkd neewe e OoB d ee t aoh l

l Br Se f

e deieir gs zd neo t n

a s bWidn i

Spmfi r er s elo h r t s C 't o bfeoe r v nnciuha eegact leie t

n t at yt ic enmio - r ,

e u luor o l d ie o ct e v k o m ud or nenve he ab nt u eu apus r

E c

,oogtn ud lco t f a s luce 5 g l 2oew mtneio sas n B cr eol -

n r e a t

i it t t

b cewu b D t

a n etah sd aen s r

e o Uwct aois yept ssa ebi r ud p

i t

a ppe l

clocd r t uba e la mu r act t

h Agn oof sa e d d t

O S iwc od za e n t

2- ir ce n i o s gi gr lyt D a cld eur c o io t

h bl t

f e se eref o t

a eht nd ml op . of osv e o e c yi s ip r g aigee r l

b t net r arohaev t

n c s newnieu sn e .

a n u e i

t l

eizdgr .

T e nt sb eap 2.heonaf ns i

u v a st ench oe d a

E gr o et r t

gacNpht or r ooo clc e noirt n o n mce s us h cOe u o en oyo ns a it it t f r t

asaBe 1 dnic sr yrsl wb f gd o it o dst e w

D t

s c

n sht cd et ateb e 8, aul ooi 9 wTnfed s ga t

io vepa e t rl oe t

p n imiuie lb /

2 r d c qt a 2 eaid sie c el oi t ir c leetin es T f k s i h ad s e /kGf 5 a neesoe uivat m nr r

D e ht Wfouic n naor S e

Obtrhi npb r cp r oe y

t e

p l t

i n y i b

e t t

la m n O i u e 7 6 B 7 a v

c' v 2 00 8 ge 0 0 o E ~

e 8 S 0 9d 208 n a D f & 89 9n 8 9u-et eo 1 /91 ,l le s89 n s91 a t

n i9 9 cay r e5 a e/6 4,o o/1,C nDa i3c 0h B4 P 7 42 A e

nef n m t r t u3 3c s3 2 dat n

n22 a

A la em evu LRneO ivReme n iRyO d

.EuB aEuov In EaB t

PRES SLJS DLJTe LMS

. s 9 O 3

1-t r s cdri a ee f

Od l eme ariit cAf f le t Bah S

ol gi s i

l ae eab Aht r r la ea t

ici a ui h h 3 s so a t

d Pa es f T ty ed r unin er eh nt oi r 1 a v .lp sah eop e P p' oie is et s t nt nb lsotaat Ou s noe i mbh o l

Osr a tot yr r M d e e n t

L e wst ll i o gn s r l l

awo

,Mb lin i

na eoi l it p

cl st m w h r t

aio it a nt eaicct eord e r pd e

vt lpa f f tr fo eo at a Enpc anor sdfere eafi tole t

u v r ic yceee c r oe o ep k f nt gnh ar ie h as t

in -y st cO r e laBon s ict uf a f r ch ird en os B nSio t an eeywa dt u r h n'

n a mr er, npt pee od 6o n

) io t go v igmtacr t G a t

n tanf ne sob ata D yP .

Sri u iw dec eAs o rd eo ahf e ObO C o BdO

( Fet hH la)TI.b it 1 SreL s d r c w s. n( r n o . gt e ee e

g euem nove i te or ein of wta a n h wh t

ot n gsel psor epec t

r pg a reinr rb e is r p c-l e eoro Dwt ai rl t s eu osp r de oo a l

s a r t

n h eptna ett r ano t t A t owdd e C c r cei l

eb v one d s o v t uaimpi os c st h t emi pee r

E u oAisc ce ri eya msp t

o t

g Sned isd kl k r dlucept a ph e yd 6 n e a e o ri c Cnr ud t se -

av D t

a i

t a

r l wfoh oy t Rib se t at r d sr open e B Pt dt o Nced re el p n Cb i

l i

er e ru n awp opx t a

r sue t

o Aaie, h ,o e O i t

t a t i utr e a qi a l

a av r ep 6,osppd t

r mCso eGa p

oDr o n

S mn1 o r 9 9 wpAb euAe mef 2- eu2p 1 ol h r tt p in pio el l irtg e D Aai 9 l

7 a c ,hiw t i ew l

f s yah Pso e t n o eI t

nNa t

i vAslaee ns up e hea l

b v yt

.s

, i t hi r cy a li o e ar cef e pwn o t

T E i b

ahd wt rt t

7 , bili h

e ael oi t nTle e a 9 rt x lih at b ec io i wb f

9 a u

,s a t

a .

c e. de ph oets t t f 1 o v ec Atode r e t

n pr c e t

,l o 7 ou o o ".eis sbid suat hc dh t

it ip 1

scnd r .

inoAiPes) sOa f r

r ei t

r 2ssoe 7 eeci r gr2 oOs f

alia nw c ine( Mf 9 d d d uq u wd mLG r

s nGee r

e ono D Nd aaar I

sn e Dpaf h rEeD t ODb y

t t i n .li e t im . b m n r g la u e li i

a c v P v o E 7 a

.D f & 9 n m et eo 3' 9u x cay r e1 1

,c nDa n28 a e

rt m o1 A t

n 7 l ne evuem

, t la f ls9iO r MIN AS pB li PRES p

p:.

4 'i .-p APPENDIX E EMERGENCY DIESEL GENERATOR RELIABILITIES AND INDUSTRY TRIGGER FAILURE RATES

~

Oh . f gA . > .

EDG Reliabilities and Industry Trigger Failure Rates Cindy Gentillon, Idaho National Engineering Laboratory, Statistics Unit ;

The Idaho National Engineering and Environmental Laboratory (INEEL) was asked to L investigate the statistical significance of industry emergency diesel generator (EDG) trigger f failure rates in ensuring that EDG reliability remains above target reliability levels (0.95 for some plants and 0.975 for others). The result of this investigation is that, when emergency diesel generator (EDG) failure rates reach the trigger levels represented by the bold data in the table below, the probability that their reliability meets or exceeds the associated target reliabilities is, at best, less than 6 percent.. The discussion below provides more detail about this finding.-

In the table, results for various sets of failure, demand combinations surrounding the Trigger l Failure Rates (in bold) are provided. The confidence bound section of the table gives

- examples of the level of information available from each set of data about the probability of

. success, i.e., the reliability, using a binomial distribution. For example within 20 demands !

and 2 failures, the confidence in exceeding a 0.958 reliability (between the two target I reliabilities)is only 5 percent . The confidence section shows that, with three failures in 20 1 demands, the confidence in meeting or exceeding a reliability of 0.95 is less than 2.5 percent

. In the right-most lower bound column where a relatively high (95 percent ) exceedance confidence exists, the reliabilities are considerably less than the target reliabilities.

Estimated Binomial distribution lower confidence Probability that reliability Failure Data ~ reliability bounds on estimated reliability > target reliability i Target Target reliability reliability Failures Demands 2.5% -5% 50 % 95% 0.95 0.975 2 20 - 0.90 0.968 0.958 0.869 0.717 0.029 0.021 3 0.85 0.943 0.929 0.819 0.656 0.056 0.044 4 0.80 0.913 0.896 0.770 0.599 0.087 0.072 3 50 0.94 0.978 0.972 0.927 0.852 0.022 0.017 4 0.92 0.967- 0.960 0.907 0.826- 0.034 0.028 5 0.90 0.955 0.946 0.887 0.801 0.047 0.039 6 0.88 0.942 0.932 0.867 0.777 0.061 0.052 4- 100 0.96 0.984 0.980 0.953 0.911 0.017 0.014 5 0.95- 0.978 0.974 0.943 0.898 0.023 0.019 6 0.94 0.971 0.967 0.934 0.885 0.030 0.025 7 0.93 0.965 0.960 0.924 0.873 0.037 0.032 8 0.92 0.958 0.952 0.914 0.860 0.044 0.039 9 0.91 0.951 0.945 0.904 0.848 0.051 0.045 10 0.90 0.944 0.937 0.894 0.836 0.059 0.053 Notes:

Bold entries correspond to the trigger failure rate levels.

Estimated reliability: Successes / Demands.

Lower confidence bounds: Based on data from a binomial distribution. The reliability is greater than the values cited, with the confidences stated in the column headings.

Probability of exceeding specified targets: computed as probability that a beta random variable with parameters (Successes +0.5, Failures +0.5) equals or exceeds the specified target.

E-1 I

l I

i: .

, i. ,

j u ** 4 The last two columns in the table contain results for the probability of meeting or exceeding the specific target reliabilities given each set of data. The calculations are based on a probability distribution for the success probability for each set of data, developed using the ,

" simple Bayes" method [see G. Box and G Tiao, Bayesian Inference in Statistical Analysis, Reading, MA: Addison Wesley,1973, Sections 1.3.4-1.3.5]. This method is widely used in risk assessment; for example, it is described in NUREG/CR 2300 (the 1983 PRA Procedures Guide published by the NRC). A beta distribution is a conjugate prior distribution for binomially-distributed data. In the simple Bayes method, a noninformative prior teta distribution (called the Jeffreys prior)is updated with observed data, resulting in a posterior beta distribution that describes the reliability for each set of data (each row in the table). The probabilities given in the table are tail probabilities for the resulting distributions. The probabilities highlighted correspond to the trigger failure data. All of these probabilities are less than 0.00.

1 i

l I

l l

l I

I E-2 1

,,

ML20217D085

Text

SUMMARY

SUMMARY

Navigation menu

Search