NUREG-1343, Forwards NUREG-1353, Regulatory Analysis for Resolution of Generic Issue 82, 'Beyond Design Basis Accidents in Spent Fuel Pools.' No Actions Should Be Required of Licensees

Forwards NUREG-1353, Regulatory Analysis for Resolution of Generic Issue 82, 'Beyond Design Basis Accidents in Spent Fuel Pools.' No Actions Should Be Required of Licensees

ML20247R044
Person / Time
Issue date:	03/21/1989
From:	Houston R NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	Fraley R Advisory Committee on Reactor Safeguards
References
REF-GTECI-082, REF-GTECI-NI, RTR-NUREG-1343, TASK-082, TASK-82, TASK-OR NUDOCS 8906070119
Download: ML20247R044 (1)

Text

, _

L. t,

,l NAR 21 1989 MEMORANDUM FOR: Raymond F. Fraley, DirectB Advisory Committei. on Reactor Safeguards FROM: R. Wayne Houston, Director

). Division of Safety Issue Resolution Office of Nuclear Regulatory Research 1

SUBJECT:

RESOLUTION OF GI-82, "BEYOND DESIGN BASES ACCIDENTS IN SPENT FUEL P0OLS" Enclosed is our Regulatory Analysis for GI-82, "Beyond Design Bases Accidents in Spent Fuel Pools." Our conclusion is that no actions should be required of licensees. The bases for this conclusion are provided in the enclosed studies performed by our contractors, BNL (NUREG/CR-4982 and NUREG/CR-5281) andLLNL(NUREG/CR-5176). The value/ impact studies show that the potential reduction in spent fuel storage risk from the selected modifications will neither result in a substantial safety improvement nor be cost-effective.

Enclosed for your information is a proposed resolution memorandum. If the Committee wishes to review this resolution, please inform Karl Kniel (x23950) by Apri1 21, 1989.

Cd% .*rd By

\'

IL R4colicdon R. Wayne Houston, Director Division of Safety Issue Resolution Office of Nuclear Regulatory Research

Enclosures:

As stated cc: G. Bagchi J. Craig P. Cota F. Gillespie A. Hodgdon, OGC DISTRIBUTION: w/o enc 1.

RES Circ RWHouston OFdi RES Chron WMinners A NUDOCS KKniel 'i DSIR C/F GMazetis RPSIB R/F EThrom

• PDR

• See previous concurrence ,

0FC: RPSIB/DSIR* RPSIB/DSIR* RPSIB/DSIR* DD/DSIR* D/h

• NAME: EThrom:skd GMazetis KKniel WMinners RWHouston DATE: / /89 / /89 / /89 / /89 /7 /89

'i

' 8906070119 890321 NREA PDR ORG PDC

w. m. _ _ _ - - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Enclosure 1 NUREG-1353

' REGULATORY ANALYSIS FOR THE RESOLUTION OF GENERIC ISSUE 82:

"BEYOND DESIGN BASIS ACCIDENTS IN SPENT FUEL POOLS" Manuscript Completed: February 1989 Date Published:

Prepared by EdwardD.Throm I

Division of Safety Issue Resolution Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, D.C. 20555 h?fby

g .,. 3

=- .

.. .4 i ABSTRACT Generic Issue 82, "Beyond Design Basis Accidents in Spent Fuel Pools," addresses the concerns

with the use of high density storage racks for the storage of spent fuel, and is applicable to all Light Water Reactor rpent fuel pools.

This report presents the regulatory analysis for Generic Issue 82. It includes (1) a summary of die issue, (2) a summary of the technical findings, (3) the proposed technical resolution, (4) al-ternative resolutions considered by the Nuclear Regulatory Commission, (5) an assessment of the benents and cost of the alternatives considered, (6) the decision rationale, and (7) the i relationships between Generic Issue 82 and other NRC programs and requirements.

Based on this evaluation, the NRC staff concludes that no new regulatory requirements are war-ranted concerning the use of high density storage racks.

iii

- g; . .g

.x q

._ .+ '

TABLE OF CONTENTS l l

Page ABSTRACT.................................................................................................................. iii

. TABLE OF CONTENTS ..........................................................................................

v LIST O F T AB L ES . . . . .. .. .. .. .. .. .. . . .. .. .. .. .. .. .. .. .. . . . . . . .. . . . . . ... .. .. .... .. ... . . . .. .. . . . . .. .. . . . . .. . . .... .... .. . . viii ABBREVIATIONS AND ACRONYMS ................................................................. x PREFACE ................................................................................................................

x1 EXE C UTIVE S UMM ARY ........ .. .. ...... ........ .... .. ................ ...... ........ ...... ........ ............ .. ES-1

1. STATEMENT OF THE PROBLEM ................................................................. 1-1 1.1 Historical Background .............................................................................. 1-1 1.2 Safety Significance .............................................................................. 1-1

2. OBJECTIVES ......................................................................................................

2-1

3. ALTERNATIVE RESOLUTIONS .................................................................. 3-1 3.1 Alternative 1 - No Action .................................................................. 3-1 3.2 Alternative 2 - Requim Use of Low Density Racks .............................. 3-1 3.3 Alternative 3 -Impmy Cooling /Make-Up Systems .............................. 3-1 3.4 Alternative 4 - Install Spray Systems ...................................................... 3-1 3.5 Alternative 5 - Modify Spent Fuel Storage Rack Designs .................. 5-1 3.6 Alternative 6 - Cover Fuel Debris with Solid Material . . . . . . . . . . . . .................. 3-2 3.7 Alternative 7 - Improve Ventilation Gas Treatment System . ................. 3-2

4. TECHNICAL FINDINGS ............................................................................ 4-1 4.1 Spent Fuel Pool (SFP) Review Guidelines and Requirements .................. 4-1 4.2 Spent Fuel Storage Pool Design Featums .......................................... 4-4 4.3 Spent Fuel Pool Structures .................................................................. 4-6 4.3.1 BWR Mark I and Mark II Plants .......................................... 4-6 4.3.2 PWR and BWR MarkIII Plants .......................................... 4-6 v

t TABLE OF CONTENTS Page

4. TECHNICAL FINDINGS (CONTINUED) 4.4 Spent Fuel Storage Rack Descriptions ...................................................... 4-6 4.4.1 Low Density Racks (Cell Pitches 20 to 30 Inches) .................. 4-6 4.4.2 Medium Density Racks (Cel' Pitches 9 Inches (BWR) to 13 Inches (PWR)) . . . . . . . . . . . . . . . . . . 4-6 4.4.3 High Density Racks (Cell Pitches 6 Inches (BWR) to 9 Inches (PWR)) .................................................................. 4-6 4.4.4 Consolidated Fuel Racks ...................................................... 4-7 4.5 Evaluation of Spent Fuel Cladding Failure ........................................ 4-7 4.6 Quantification Of Accident Sequences in Spent Fuel Pool .................. 4-13 4.6.1 Structural Failure Due to Missiles .......................................... 4-13 4.6.2 Structural Failure Due to Aimraft Crashes .............................. 4-14 4-14 4.6.3 Structural Failure Due to 4.6.4 Reactor Cavity and Transfer Gate Pneumatic Heavy Loads Drop (Seal FailumsShipping

...... 4-15 Cask) ..

4.6.5 Inadvertent Draining of the Spent Fuel Pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-20

. 4.6.6 Loss of Cooling /Make-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-22 4.6.7 Structural Failure from Beyond Design Basis Seismic Event ...... 4-28 4.7 Summary of Accident Sequence Quantification .............................. 4-36 4.8 Radiological Consequences Evaluation ...................................................... 4-37 4.8.1 Radionuclides Inventories ...................................................... 4-37 4.8.2 Radionuclides Potentially Available for Release .................. 4-37 4.8.3 Estimated Releases and Consequences For SFP Accidents ...... 4-39 4.8.4 Summary Conclusions On Fuel Damage and Consequences . . . . . . 4-42 4.9 Other Issues Concerning Use of High Density Storage Racks .................. 4-42 4.9.1 Gaps In Neutron-Absorbing Materials ............................ 4-42 4.9.2 Potential For High Radiation Fields .......................................... 4-43 4.9.3 Refueling Cavity Seal Failure During Fuel Assembly Handling (GI137)..................................................................................... 4-43

5. VALUE/ IMPACT ANA LYS IS .... .......... .. ...... ............ .. .. .............. .... ...................... 5-1 5.1 Alternative 1 - No Action .................................................................. 5-1 5.1.1 Occupational Exposure (Accidental) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2 5.1.2 Onsite Property Damage ..................................................... 5-2 5.1.3 Offsite Property Damage ................................................... 5-3 5.1.4 Potential Consequences and Cost of SFP Accidents . . . . . . . . . . . . . . . . . . 5-5 vi

l

-TABLE OF CONTENTS

5. VALUE/ IMPACT ANALYSIS (CONTINUED) 5.2 Alternative 2 - Require Use of Low Density Racks .............................. 5-7 5.2.1 Risk Reduction Estimate ...................................................... 5-7 5.2.2 Cost of Low Density Storage .................................................... 5-7 5.2.3 Value/ Impact Summary ...................................................... 5-11 5.3 Alternative 3 -Improve Cooling /Make-Up Systems .............................. 5-14 5.3.1 Risk Reduction Estimate ...................................................... 5-14 5.3.2 Cost of Improved Cooling /Make-Up Systems .............................. 5-15 5.3.3 Value/ Impact Summary ...................................................... 5-15 5.4 Alternative 4 -Install Spray Systems ..................................................... 5-17 5.4.1 Risk Reduction Estimate ...................................................... 5-17 5.4.2 Cost ofInstalling Spray Systems .......................................... 5-17 5.4.3 Value/ Impact Summary . . . . . . . . . . . . . ......................................... 5-18 5.5 Altemative 5 - Modify Spent Fuel Storage Rack Designs .................. 5-21 5.6 Alternative 6 - Cover Fuel Debris with Solid Material .............................. 5-21 5.7 Alternative 7 -Improve Ventilation Gas Treatment System .................. 5-21 5.8 Relationships With Other Requirements and Activities .......... ................... 5-22 5.8.1 Sevem Accident Policy Statement .......................................... 5-22 5.8.2 Seismic Design Margins Program .......................................... 5-23

6. DECISION RATIONALE ........................................................................... 6-1 6.1 Comparison to the Backfit Criteria (10 CFR 50.109) ............................. 6-2 6.2 Comparison to the Safety Goal Policy Statement .............................. 6-3 6.3 Other Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............................ 6-5

7. IMPLEMENTATION ......................................................................................... 7-1

8. REFERENCES ...................................................................................................... 8-1 APPENDIX A Spent Fuel Data and Storage Requirements .............................. A-1 vii

o - ,-

LIST OF TABLES Table Page 4.1.1 10 CFR 50 Appendix A, " General Design Criteria .......................................... 4-3 4.1.2 Regulatory Guides ...................................................................................... 4-3 4.1.3 Other Guidelines / References ..............................................................- 4-3 4.2.1 TypicalPool Dimensions .............................................................................. 4-5 4.5.1 Estimated Likelihood of Self-Sustaining Zircaloy Clad Oxidation For Various Spent Fuel Rack Configurations and Decay Heat Levels ...... 4-11 4.5.2 Estimated Likelihood of Propagation of Zircaloy Fire to Older Stored Fuel For Various Spent Fuel Rack Configurations and Decay Heat Levels (Perfect Ventilation Cases) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-12 4.5.3 Summary of Radial Oxidation Propagation Results For Various PWR Spent Fuel Rack Configuranons and No Ventilation .............................. 4-13 4.6.1 Events in Which Pneumatic Inflated Seals Have Failed .............................. 4-17 4.6.2 Refueling Cavity Seal Leak Rates Following Seal Failure .............................. 4-20 4.6.3 Heatup and Boil Dry Times for a Typical Spent Fuel Pool .......................... 4-25 4.6.4 HEP Estimates for Failure to Diagnose Loss of Cooling .............................. 4-26 4.6.5 Typical failure rates and Repair Times For Cooling System Components (Taken fmm EPRI NP-3365) ................................................................ 4-26 4.6.6 Failure Frequency of Generic SFP Cooling and Make-Up Systems . . . . . . 4-27 4.6.7 Estimated Median Factors of safety and Logarithmic Standard Deviations Associated with the Safe Shutdown Earthquake (SSE)

(Post 1973 Seismic Design Methods) ..................................................... 4-34 4.6.8 Annual Seismic Failure Frequencies For Two Representative Spent Fuel Pools ......................................................................................... 4-35 4.7.1 Summary of SFP Accident Frequencies ...................................................... 4-36 4.8.1 _ Estimated Radionuclides Release Fractions During a Spent Fuel Pool Accident Resulting in Complete Destruction of the Fuel Cladding .............................. 4-38 4.8.2 Offsite Consequences of Spent Fuel Pool Accidents - CRAC2 Results ...... 4-41 4.8.3 Offsite Consequences of Spent Fuel Pool Accidents - MACCS Results ...... 4-41 l' .

viii 1

L .i ,- .

LIST OF TABLES Table Page 5.1.1 Ons'te Property Damage Costs Per Accident (1988 $s) .............................. 5-4 5.1.2 Offsite Health and Property Damage Estimates (1988 $s) Per SFP Accident 5-4 5.1.3 Best Estimate Consequences of a Spent Fuel Pool Accident .................. 5-6 5.2.1 Range of Unit-Cost Estimates for Additional Storage Requirements (Costs in 1988 $s per kilogram of heavy metal) ........................................ 5-9 5.2.2 Additiona1 Incremental Storage Capacity Requirements for Alternative 2 ...... 5-9 5.2.3 Storage Costs Associated With Alternative 2 (1988 $s) .............................. 5-10 5.2.4 Cask Storage Cost Estimates as a Function of Facility Capacity .............. 5-11 5.2.5 Summary ofIndustry Wide Value/ Impact Analysis for Alternative 2 Based on 100% Risk Reduction (1988 $s) ...................................................... 5-13 5.2.6 Cost-Benefit Sensitivity Analysis for Alternative 2 (1988 $s) .................. 5-14 5.3.1 Value/ Impact for Generic Impmvements to the SFP Cooling Systems (5% Discount Rate - 1988 $s) ................................................................ 5-16 l

5.4.1 Offsite Health and Property Damage Estimates (1988 $s)

With Pool Spray System (DF = 45) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-18 5.4.2 Summary ofIndustry Wide Value/ Impact Analysis for Alternative 4 Based on Spray System DF of 45 (1988 $s) .......................................... 5-20 L 6.3.1 Summary of SFP Seismic Failme Frequency Estimates ............................. 6-8 1

A.1 Nuclear Power Plant Data ........................................................................A-2 L

A.2 Projected Cumulative Storage Requirements l

Maximum AR Capacity, Assemblies .................................................... A-5 A.3 Projected Cumulative Storage Requirements Maximum AR Capacity, MTIHM ................................................................ A-6 A.4 1986 Inventory and Projected Annual Reactor Discharges Assemblies ........ A-7 l

ix L-_____-_______-___________-________ _

I -

ABBREVIATIONS AND ACRONYMS ALARA As Low As Reasonable Achievable AR At-Reactor ASCE American Society of Civil Engineers

! BNL Brookhaven National Laboratory

, BWR(s) Boiling Water Reactor (s)

CDF Core Damage Fmquency CFR Code of Federal Regulations DBE Design Basis Earthquake DF Decontamination Factor DOE Department of Energy EPRI Electric Power Research Institute FSAR Final Safety Evaluation Report GDC(s) General Design Criterion (Criteria)

GI Generic Issue GL Generic letter GSI Generic Safety Issue HCLPF High Confidence of Low Probability of Failure HEP (s) Human Error Probability (ies)

IE Inspection andEnforcement IEB Inspection and Enforcement Bulletin kgm kilogram -

KgU kilogram of uranium kw kilowatt LLNL Lawmnce Livermore National Laboratory LPCI Low Pressure Coolant Injection LWR (s) Light Water Reactor (S)

MTIHM Metric Tons ofInitial Heavy Metal MTHM Metric Tons of Heavy Metal MTU Metric Tons of Uranium Mw(t) Megawatts-thermal NERC National Electric Reliability Council NRC Nuclear Regulatory Commiasion NRR Office of Nuclear Reactor Regulation, NRC ORNL Oak Ridge NationalLaboratory PRA(s) Probabilistic Risk Assessment (s)

PWR(s) Pmssurized Water Reactor (s)

RES Office of Nuclear Regluatory Research, NRC RHR R::sidual Heat Removal RWST Refueling Water Storage Tank SFDF Spent Fuel Damage Frequency SFP Spent Fuel Pool SNL Sandia NationalLaboratory SRP Standard Review Plan SSE Safe Shutdown Earthquake SSMRP Seismic Safety Margins Research Program X

- - _ _ _ _____----_.-_m--__._m _

PREFACE This report presents the regulstory analysis, including the decision rationale, for the resolution of Generic Issue 82, "Beyond Design Basis Accidents in Spent Fuel Pools." The objective of this regulatory analysis is to determine whether the use of high density storage racks for the storage of spent fuel poses an unacceptable risk to the health and safety of the public. As part of this effort, the seisnue hazards for two older spent fuel pools were evaluated. The risk change estimates, value/ impact and cost-beneSt analyses, and other insights gained during this effort, have shown that no new regulatory requirements are warranted in relation to this generic issue.

Edward D. Throm I Xi

.e . ,

EXECUTIVE

SUMMARY

The risk of beyond design basis accidents in spent fuel storage pools was examined in WASH-1400. It was concluded that these risks were orders of magnitude below those involving the reactor core because of the simplicity of the spent fuel storage :x>ol design: (1) the coolant is at atmospheric pressure, (2) the spent fuel is always suberitical anc; the heat source is low, (3) there is no piping which can drain the pool and (4) there are no anticipated operational transients that could mterrupt cooling or cause criticality.

The reasons for the re-examination of spent fuel storage pool accidents are twofold. First, spent fuelis being stored instead of reprocessed. This has led to the expansion of onsite fuel storage by means of high density storage racks, which results in a larger inventory of fission noducts in the pool, a greater heat load on the pool cooling system, and less distance between acjacent fuel assemblies. Second, some laboratory studies have provided evidence of the possibility of fire propagation between assemblies in an air cooled environment. Together, these two reasons provide the basis for an accident scenario which was not previously considered.

In addition, in recent years, increasing knowledge in the geosciences has led to a better under-standing that, although still highly unlikely, it is more likely that nuclear power plants in the Eastern United States (i.e., east of the Rocky Mountains) could be subjected to earthquake ground motion greater than for which the plants were designed. For this reason, interest has developed in demonstrating that nuclear power plant structures and safety-related systems can safely withstand earthquake ground motion larger than their design earthquake ground motions l (po:;t-1973 safe-shutdown earthquake, SSE, or pre-1973 design-basis earthquake, DBE).

l Nuclear reactor plants include storage facilities for the wet storage of spent fuel assemblies. The l safety fu".ction of the spent fuel pool (SFP) and storage racks is to cool the spent fuel assemblies l

and maintain them in a suberitical array during all credible storage conditions and to provide a safe means ofloading the assemblies into shipping casks.

The SFP and components are reviewed to assure conformance with the requirements of 10 CFR Part 50 Appendix A General Design Criteria (GDC) 2,4,5,61,62, and 63. The review is per-formed under Section 9.1.2, " Spent Fuel Storage," of the Standard Review Plan (SRP). The SFP water level control system, cleanup system and cooling system are reviewed to assure confor-mance with the requirements of GDCs 2,4,5,44,45,46,61 and 63 under Section 9.1.3, " Spent 1 Fuel Pool Cooling and Cleanup," of the SRP. In addition, a finding related to 10 CFR Part 20, paragraph 20.1(c) is made as it relates to radiation doses being kept as low as is reasonably achievable (ALARA).

The methods used to provide cooling for the removal of decay heat from the stored assemblies vary from plant to plant depending upon the individual design. The safety function to be per-formed remains the same: the spent fuel assemblies must be cooled and must remain covered with water during all storage conditions. Assuming that the water is drained, or boiled off, from the spent fuel pool, the fuel rods will heat up until the buoyancy-driven air flow is sufficient to prevent further heatup. If the decay heat level is high enough to heat the fuel rod cladding to about 900 oC (1650 0F) the oxidation becomes self-sustaining, resulting in a Zircaloy cladding fire. Pmpagation of the Zircaloy cladding fire to older adjacent assemblies is likely if the decay heat level in an older adjacent assembly is high enough to heat that assembly to within 100 to ES-1 i

200 oC (200 to 400 0F) of the self-sustaining oxidation temperature. Although propagation of a Zircaloy cladding fire to one to two year old fuel by only thermal radiation can occur, the older fuel would have to be next to the hottest assemblies.

l The conditional probability of a Zircaloy cladding fire given a complete loss of water was found i to be 1.0 for PWRs and 0.25 for BWRs. The PWR value is based on the use of high density i

storage racks and the BWR value is selected based on the use of directional storage racks, with the channel box in place. The conditional probability of a Zircaloy cladding fue given a com-plete loss of water in low density storage racks is estimated to be at least a factor of five less l than for the high density configurations. The PWR conditional probability of a Zircaloy fire would be reduced to 0.2 and the BWR conditional probability would be reduced to 0.05. The actual risk mduction achievable may be greater. Open frame racks or cylindrical racks with large inlet holes could msult in an greater reduction in risk. The cooling time to preclude a Zir-caloy cladding fire could be reduced to less than 20 days, for a conditional probability of 0.05 of a Zircaloy fire for both fuel types.

In addition to implementing the requirements contained in 10 CFR Part 50 Appendix A of the

" General Design Criteria," and 10 CFR Part 20, concerning radiation doses being kept as low as is reasonably achievable, licensees should have implemented additional or corrective actions based on the following guidance:

1. E Bulletin 84-03, " Refueling Cavity Water Seals," issued August 24,1984.

2. E Information Notice 84-93, " Potential for Loss of Water From the Refueling Cavity," issued December 17,1984.

3. Generic Letter 85-11, " Completion of Phase II of ' Control of Heavy Loads at Nuclear Power Plants' NUREG-0612," issued June 28,1985.

4. E Infonnation Notice 87-13, " Potential for High Radiation Fields Following Loss of Water from Fuel Pool," issued February 24, 1987.

5. IE Information Notice 87-43, " Gaps in Neutron-Absorbing Material in High-Density Spent Fuel Storage Racks," issued Sep-tember 8,1987.

6. E Information Nctice 88-65, " Inadvertent Drainages of Spent Fuel Pools," issued August 18,1988.

7. E Information Notice 88-92, " Potential For Spent Fuel Pool Draindown," issued November 22,1988.

The risk from the storage of spent fuel in the spent fuel storage pool at light water reactors is dominated by the beyond design basis earthquake accident scenario. The seismic capacities, or fragility, of two older spent fuel pools indicate that the high confidence of low probability of failum (HCLPF) is about three times the safe shutdown earthquake (SSE) design level. The HCLPF values are estimated to be in the 0.5 to 0.65 g range. The median peak ground accelera-ES-2

o tion needed to fail these pools is estimated to be in the 1.4 to 2.0 g range, nearly a factor of ten higher than the SSE design value. A repon prepared by the American Society of Civil En-gineers also concluded that, in general, the seismic design of nuclear facility structures result in median factors of safety on the order of 4 to 19 based on post-1973 design criteria.

The structural capacity of the elevated BWR pool is lower than that for the PWR pool located at the ground level, however the lower conditional probability of a Zircaloy fire for the BWR fuel assembly design offsets the higher seismic failure frequency. The probability of a Zircaloy clad-ding fire, resulting from the loss of water from the spent fuel pool, is estimated to have a mean value of 2x10-6 per reactor year for either the PWR or the BWR spent fuel pool. The seismic event contributes over 90% of the PWR spent fuel damage probability, and nearly 95% for the BWR.

The source term for the spent fuel pool accident is not the rame as the source term associated with core damage accidents. The consequences of a spent fuel pool accident which results in the complete loss of water are dominated by the long lived isotopes, such as cesium and strontium.

The health consequences are dominated by the risk of latent cancer fatalities due to long term exposures.

The best estimate of the consequences of a spent fuel pool accident which results in spent fuel 6

damage to approximately one-third of an equivalent reactar core is 8: 10 person-rem. This total dose translate to a public health risk from a spent fuel pool accident of 480 person rem over an average remaining lifetime of 30 years, based on a Zircaloy cladding fire probability of 2x10-6 per reactor year. The best estimate offsite property damage cost is $4,000 million (1988 $s).

The best estimate values are based on a population density of 340 people per square mile within a 50 mile radius from the site and result from the release of radionuclides from the last fuel dis-charge,90 days after being discharged. The best estimate of the onsite costs for a SFP accident is $1,180 million (1988 $s), including five years of replacement power to replace the damaged spent fuel pool. Based on an average remaining lifetime of 30 years and a 5% discount rate, the present value of the offsite propeny damage is estimated to be $124,300 and the present value of the onsite propeny damage is estimated to be $32,400, based on a Zircaloy cladding fire proba-bility of 2x10, perreactor year.

The value/unpact and cost-benefit evaluations for the proposed alternatives for Generic Issue 82 do not indicate that cost effective options are available to mitigate the risk of beyond design basis accidents in spent fuel pools. The option to use low density storage racks for recently dis-charged fuel has a best estimate value/ impact ratio of $32,000 per averted person-rem based on a reduction in spent fuel damage frequency of 2x10-6 per reactor year. Low density racks would decrease the consequences by a factor of five to ten, but the value/ impact ratio is based on 100%

reductionin public dose.

The use of post-accident spray systems to mitigate the consequences of a spent fuel pool acci-dent has a best estimate value/ impact ratio of $3,300 per avened person rem. This assumes that a post-accident spray system can be designed to withstand the beyond design basis canhquake which causes gross failure of the spent fuel pool structum and has a decontamination factor (DF) of at least 45.

The risks associated with a severe accident in the spent fuel pool are also compared to the objec-tives and guidance in the Safety Goal Policy Statement. The estimated frequency of a spent fuel pool accident,2x10-6 per reactor year, resulting in spent fuel damage meets a target objective of ES-3

a few percent of a lx10-4 to 5x10-5 per reactor year value for overall core damage frequency.

The target objective for a "large release" of 1x10-6 per reactor year is marginally met, within a best estimate factor of two, but subject to interpretation since the definition of "large release" is still under development. In meeting the societal risk objective of 0.1% of the normally occur-ring risk to the public given the release frequency of 2x10-6 per reactor year, the latent cancer fatality rate from spent fuel pool accidents is estimated to be less than 3% of the target value for the operation of a nuclear power plant.

Therefore, the backfit criteria (10 CFR 50.109) that (1) a substantial inemase in the overall protection of the public health and safety is achieved, and (2) the direct and indirect costs of im-plementation are justified are not met, and Alternative 1 "No Action" is mcommended for the resolution of GI-82.

The risk and consequences of a spent fuel pool accident appear to meet the Safety Goal Policy Statement objectives. They would also meet the proposed lx10-6 per reactor year large-release fmquency guideline, at least pending definition of a "large release" by the Commission. There-fore the recommended resolution, Alternative 1 "No Action," is justified.

Although these studies conclude that most of the spent fuel pool risk is derived from beyond design basis earthquakes, thic risk is no greater than the risk from core damage accidents due to seismic events beyond the safe-shutdown earthquake. Therefore, reducing the risk from spent fuel pools due to events beyond the safe-shutdown earthquake would still leave at least a com-parable risk due to core damage accidents. Because of the large inherent safety margins in the design and construction of the spent fuel pool, Alternative 1 "No Action"is justified.

1 ES-4

L% ..

=,

[

REGULATORY ANALYSIS FOR THE RESOLUTION OF l

GENERIC ISSUE 82 l

"BEYOND DESIGN BASIS ACCIDENTS IN SPENT FUEL POOLS"

1. STATEMENT OFTHE PROBLEM 1.1 HistoricalBackground The risk of beyond design basis accidents in spent fuel storage pools was examined in WASH-1400 (Ref.1). It was concluded that these risks were orders of magnitude below those involving the reactor core be::ause of the simplicity of the spent fuel storage pool: (1) the coolant is at atmospheric pressure, (2) the spent fuelis always suberitical and the heat source is low, (3) there is no piping which can drain the pool and (4) there are no anticipated operational transients that could interrupt cooling or cause criticality.

The reasons for the re-examination of spent fuel storage pool accidents are twofold. First, spent fuel is being stored instead of reprocessed. This has !cd to the expansion of onsite fuel storage by means of high density storage racks, which results in a larger inventory of fission products in the pool, a greater heat load on the pool cooling system, and less distance between adjacent fuel assemblies. Second, some laboratory studies have provided evidence of the possibility of fire propagation between assemblies in an air-cooled environment. Together, these two reasons provide the basis for an accident scenario which was not previously considered.

In addition, in recent years, increasing knowledge in the geosciences has led to a better understanding that, although still highly unlikely, it is more likely that nuclear power plants in the Eastern United States (i.e., east of the Rocky Mountains) could be subjected to earthquake ground motion greater than for which the plants were designed. For this reason, interest has developed in demonstrating that nuclear power plant structures and safety-related systems can safely withstand earthquake ground motion larger than their design earthquake ground motions (post-1973 safe-shutdown earthquake, SSE, or pre-1973 design-basis earthquake, DBE).

1.2 SafetySignificance A typical spent fuel storage pool with high density storage racks can hold roughly five times the fuelin the core. However, smce reloads typically discharge one thirti of the core, much of the spent fuel stored in the pool will have had considerable decay time. This reduces the radioactive inventory somewhat. More importantly, after roughly three years of storage, spent fuel can be air-cooled. The spent fuel need not be submerged to prevent melting, although submersion is still desirable for shielding and to reduce airborne activity.

1-1

If the spent fuel storage pool were to be drained of water the discharged fuel from the last one or i two refuelings, stored in high density storage racks, could still be " fresh" enough to melt under decay heat. The Zirealoy cladding of this fuel could be ignited during heatup. The resulting fire, m a spent fuel storage pool equipped with high density storage racks, might spread to other fuelin the pool.

The heat of combustion, in combination with decay' heat, would certainly release considerable gap activity from the fuel and would probably drive " borderline aged" fuel into a molten condition. Moreover, if the fire becomes oxygen-starved (quite probable for a fire located in the bottom of a pit such as the spent fuel storage pool), the hot zirconium would rob oxygen from the uranium dioxide fuel, forming a liquid mixture of metallic uranium, zirconium, oxidized zirconium, and dissolved uranium dioxide. This liquid mixture would allow a release of fission products from the fuel matrix. In addition, although confined, spent fuel storage pools are '

almost always located outside of the primary containment. Thus, a release to the atmosphere is more likely relative to a release inside primary containment.

The safety significance of "Beyond Design Basis Accidents in Spent Fuel Pools" has been designated as a medium prionty issue (Ref. 2), Generic Issue 82. GI-82 applies to all light-water reactor spent fuel storage pools.

l 1-2

I

2. OBJECTIVES f The general objective of GI-82 is to evaluate the need for additional protective measures for the safe storage of spent fuel in high density storage racks in the spent fuel storage pool at light-water reactor sites.

Both prevention and mitigation are considered. A preventive option is one intended to reduce the frequency of accident sequences potentially conducive to the release of fission products from the spent fuel assemblies. A mitigative option is one intended to reduce the magnitude of the consequences that would result from an accident (environmental radiological releases).

Given the diversity of plant-specific design and construction of spent fuel storage pools, the applicability of any generic analysis of risk reduction measures is limited both with respect to the characterization of risk, and to the cost of implementation for any one plant. The analysis performed for GI-82 is intended to provide a broad evaluation of the value/ impact attributes of a given proposed alternative. Plant specific analyses are used for the seismic risk evaluations. In this case, older plants have been selected for the analyses. In general the older plants are more vulnerable to seismic induced failures.

The risk from the storage of spent fuel in spent fuel pools should be a small contributor to the overall risk associated with the opemtion of a light-water reactor (LWR). On the core damage frequency (CDF) risk level, or more specifically in this case spent fuel damage frequency (SFDF), a target for the resolution of Generic Issue 82, based on the Commission's Safety Goal Policy Statement, is that the contribution from spent fuel pool accidents be a small part (a few percent) of an overall CDP target of 1x104per reactor year.* Since spent fuel pools are not.

within the primary containment structure, a target SFDF for spent fuel pool accidents on the order of 1x104 per reactor year may be considered to be compatible with the proposed general performance guidelines given in the Commission's Safety Goal Policy Statement, that is, that

.the pmbability of a large release from an operating nuclear power plant should be no greater than 1x104 per reactor year. A more direct comparison of a SFDF target with the policy guidelines requires a definition of a "large release" in the policy statement.

• More recently, a core damage frequency goal of 5x10-5 per reactor year has been proposed 4 under the safety goalimplementation pmgram. This is a factor of two (2) lower than the lx10 value used herein, but is within the uncertainty inherent in calculations and assumptions made 4

assessing compliance with either goal, and its adoption in lieu of the 1x10 goal would not affect the recommendations made in this Regulatory Analysis.

2-1

3. ALTERNATIVE RESOLUTIONS In reaching its proposed resolution of GI-82, the staff considered seven specific alternative courses of action. These are discussed below. The requirements would be applicable to all light-water reactor (LWR) spent fuel storage pools, both in the operating or planned construction stage of licensing. There are 108 spent fuel pools for the 119 operating or planned LWRs at 75 .

reactor sites in the U.S. The three shutdown units, Dresden 1, Indian Point 1 and Humboldt Bay, am excluded from this accounting.

3.1 Alternative 1- No Action This proposed altemative assumes that no additional requirements for the safe storage of spent fuel in the primary spent fuel storage pool am needed. It also assumes that all applicable requirements and guidance to date have been implemented, but no implementation is assumed for related generic issues or other staff requirements or guidance that are rtill unmsolved or still under review.

3.2 Alternative 2 - Require Use of Low Density Storage This proposed altemative would require the use of low density storage racks for the storage of recently discharged fuel. Also, some reracking from high density to low density racks would be required. As a result, it is expected that additional at-reactor storage of spent fuel would be required to accommodate the lost capacity in the spent fuel storage pool. The use of low density racks shortens the cooling time to preclude a Zircaloy cladding fire by promoting air cooling if water is lost from the spent fuel. The likelihood and the amount of fuel damage would both decrease. This alternative is directed primarily towards prevention of a large release from the spent fuel pool.

3.3 Alternative 3 - Improve Cooling /Make-up Systems This proposed alternative would require improvements in the spent fuel pool cooling and/or make-up systems, beyond the requirements curremly used to license the spent fuel storage pools.

Improvements in these systems would reduce the likelihood of fuel damage from loss of cooling events. This alternative is primarily directed towards prevention.

3.4 Alternative 4 - Install Spray System This proposed alternative would require licensees to install post accident spray headers to mitigate the consequences of a Zircaloy cladding fire if the spent fuel storage pool is drained and cannot be reflooded. The likelihood of fuel damage would not change, but the spray systems would remove fission products and lower the consequences of a spent fuel pool accident. This altemative is primarily directed towards risk mitigation.

3.5 Alternative 5 - Modify Spent Fuel Storage Rack Designs This proposed alternative would require the licensee to compartmentalize the spent fu:1 storage pool by installing partitions (and individual coolant supply diffusers for each compartment) to limit the extent of the accident, or modify the storage racks to improve air circulation, should the spent fuel storage pool drain. This alternative is directed both towards risk mitigation and prevention.

3-1

\ .

1

- a

. l 1

3.6 Alternative 6 - Cover Fuel Debris with Solid Materials ,

1 This proposed alternative would require the development of a contingency plan to dump masive amount of solid materials into a drained spent fuel pool to cover the rubble bed to a depth of several feet. The materials would not be necessarily stockpiled on site, but could also l' be obtained in a timely manner on an ad hoc basis. The materials (sand, clay, dolomite, boron compounds, lead, etc.) are commonly available in all parts of the country. This alternative would be directed at risk mitigation.

3.7 Alternative 7 - Improve Ventilation Gas Treatment System This proposed alternative would require the installation of a building ventilation and filter system capable of reducing the concentration of airborne radioactivity before discharge to the environment. This altemauve would be directed at risk mitigation.

3-2

l

4. TECHNICAL FINDINGS -

1 l

4.1 Spent Fuel Pool (SFP) Review Guidelines and Requirements Nuclear reactor plants include storage facilities for the wet storage of spent fuel assemblies. The safety function of the spent fuel pool (SFP) and storage racks is to cool the spent fuel assemblies and maintain them in a subcritical array during all credible storage conditions and to provide a j safe means ofloading the assemblies into shipping casks.

The SFP and components are reviewed to assure conformance with the requirements of 10 CFR Part 50 Appendix A General Design Criteria (GDC) 2,4,5,61,62, and 63. The review is per-formed under Section 9.1.2, " Spent Fuel Storage," of the Standarti Review Plan (SRP) (Ref. 3).

The facility and components are reviewed with respect to the following:

(a) The quantity of fuel being stored.

(b) The design and arrangement of the storage racks for maintaining a subcritical array during all condidons.

(c) The degree of suberiticality provided along with the analysis and associated as-sumpnons.

(d) The effects of extemal loads and forces on the spent fuel storage racks, pool, and liner plate (for example, safe shutdown earthquake, crane uplift forces, mis-siles, and dropped objects).

(e) Design codes, material compatibility, and shielding requirements.

The SFP water level control system, cleanup system and cooling system are reviewed to assure conformance with the requirements of GDCs 2,4,5,44,45,46,61 and 63 under Section 9.1.3,

" Spent Fuel Pool Cooling and Cleanup," of the SRP. In addition, a finding related to 10 CFR Part 20, paragraph 20.l(c) is made as it relates to radiation doses being kept as low as is reasonably achievable (ALARA).

The methods used to provide cooling for the removal of decay heat from the stored assemblies vary from plant to plant depending upon the individual design. The safety function to be per-formed remains the same: the spent fuel assemblies must be cooled and must remain covered with water during all storage conditions. The capability of the spent fuel pool cooling and cleanup system to provide adequate cooling to the spent fuel during all operating conditions is reviewed on one of two bases. The first basis requires the cooling portion of the system to be designed to seismic Category I, Quality Group C requirements. The second basis allows a non-seismic Category 1 Quality Group C spent fuel pool cooling system provided that the following systems are designed to seismic Category I requirements and are protected against tornadoes:

the fuel pool make-up water system and its sources; and, the fuel pool building and its ventila-tion and filtranon systc.m. The make-up, ventilation and filtration systems must also withstand a single active failure. The systems are reviewed with respect to the following:

4-1

(a) The quantity of fuel being cooled, including the corresponding requirements for continuous cooling during normal, abnormal and accident conditions.

(b) The ability of the system to maintain pool water level.

(c) The ability to provide alternative cooling capability and the associated time re-quired for operation.

(d) Provisions to provide adequate make-up to the pool.

(e) Provisions to preclude loss of function resulting from single e:tive failures or failures of non-safety related components or systems.

(t; The means provided for the detection and isolation of system components that cculd develop leaks or failures.

(g) The instrumentation provided for initiating appmpriate safety actions.

(h) The ability of tl e system to maintain uniform pool water temperature condi-tions.

Other functions performed by the system, not related to safety, include water cleanup for the SFP, refueling canal, refueling water storage tank and other equipment storage pools; means for filling and draining the refueling canal and other storage pools; and surface skimming to provide clear waterin the SFP.

Load handling in the SFP area is Icviewed to assure conformance with GDCs 2, 5, 61 and 62 under Section 9.1.4, " Light Load Handling System (Related to Refueling)," and with GDCs 2,4, 5 and 61 under Section 9.1.5, " Overhead Heavy Load Handling Systems," of the SRP. In addi-tion the requirements identified in the resolution of GSI A-36, " Control of Heavy Loads Near Spent Fuel," as specified in Generic Letter 85-11 (Ref. 4), " Completion of Phase II of Control of Heavy Loads at Nuclear Power Plants (NUREG-0612)," are reviewed to assure that the im-plementation of Phase I of NUREG-0612 (Ref. 5) "has provided sufficient protection such that the risk associated with potential heavy load drops is acceptably small." Adequate justification is provided by means of (1) a single-failure proof crane, (2) operator training and procedures, maintenance and inspection pmcedures, safe load paths and mechanical or electrical stops to prevent movement of heavy loads over irradiated fuel, and/or (3) load drop analyses. The staff concluded, in GL 85-11. "that satisfaction of the Phase I guidelines assures that the potential for a heavy load drop is extremely small."

For reference, the titles of the various review and acceptance criteria are provided in Tables 4.1.1,4.1.2 and 4.1.3.

4-2 l

l

[.

1

, . ;e t

l' Table 4.1.1 10 CFR Part 50 3 r M = A, " General Design Criteria" 2 " Design Bases for Protection Against Natural Phenoranna."

4 " Environmental and Missile Design Bases."

5 " Sharing of Structures, Systems, and components."

44 " Cooling water."

45 " Inspection of Cooling Water System."

46 " Testing of Cooling Water System."

61 " Fuel Storage and Handling and Radioactivity Control."

62 " Prevention of Criticality in Fuel Storage and Handling."

63 "Monitorine Nel and Waste Storage."

. n -.

Table 4.1.2 Regulatory Guides 1.13 " Design Objectives for Light-Water Reactor Spent Fuel Storage Facilities at Nuclear Power Stations."

1.26 " Quality Group Classification and Standards for Water , Steam , and Radioactive-Waste-Containing Ccaponents at Nuclear Power Plants."

1.29 " Seismic Design Classification."

1.52 " Design, Testing, and Maintan=nce Criteria for Engineering-Safety-Feature Atmospheric Cleanup System Air Filtration and Adsorption Units of Light-Water-Cooled Nuclear Power Plants."

1.115 " Protection Against Low-Trajectory Turbine Missiles."

1.117 " Tornado Design classification."

8.8 "Information Relevant to Ensuring That occupational Radiation Esposures at Nuclear Power Stations Will Be As Low As Is Reasonably Achievable."

Table 4.1.3 Other Guidelines / References ANS 57.1/ ANSI N208, " Design Requirements for Light Water Reactor Fuel Handling System."

ANS 57.2/ ANSI N210-1976, " Design objectives for Light Water Reactor Spent Fuel Storage Facilities at Nuclear Power Stations."

NUREG-0554, " Single-Failure-Proof Cranes for Nuclear Power Plants."

NUREG-0612, " Control of Heavy Loads at Nuclear Power Plants."

4-3

4.2 Spent Fuel Storage Pool Design Features At some multi-unit sites a single poolis used for both units. Table A.1 of Appendix A identifies 11 dual unit pools, eight sites which have transfer canals between pools and three sites which can use transfer casks to move fuel between pools. The estimated maximum storage capacities (allowing for a full core reserve and other non-fuel reserve areas) and other plant specific infor-mation for all spent fuel pools in operation or planned are also provided in Table A.I. There are 108 spent fuel pools for the 119 operating or planned plants at 75 reactor sites in the U.S.

(excluding Dresden 1, Humboldt Bay and Indian Point 1, which are now shutdown; and the Ft.

St. Vrain HTGR).

The spent fuel pool floor and walls are lined with 1/8 to 1/4 inch thick stainlers steel liner plates.

The plates are welded to each other by seam welds. Under the seam welds, leak detection (control) channels are provided.

The design features of spent fuel storage pools keep the likelihood of loss of pool water and recriticality small. These features are:

(a) The fuel building concrete structure, the spent fuel storage pool, the spent fuel storage racks, the SFP cooling system, and the supports for the spent fuel handling trolley are designed to withstand seismic forces so that an earthquake as large as the safe shutdown earthquake will not cause loss of water or recriticality.

(b) Fuel storage racks are designed to keep the fuel widely enough separated so that stored fuel will not achieve criticality or, in high density storage racks, poison l

materialis added to the rack structure for criticality control.

(c) The SFP is designed to prevent inadvertent loss of water from the fuel region by dramage through connected piping systems. Although a pool cooling system is connected to the pool for decay heat removal, it is designed to prevent siphoning of the water. A connection exists between the SFP and the reactor pressure vessel head region through the fuel transfer pathway (refueling canal) which is provided with physical barriers to prevent SFP drainage when not in use. The pools are generally sized so that the fuel remains nearly completed covered if the transfer pathway is inadvertently opened.

(d) Should the water inventory in the pool fall below a pre-set level or increase in temperature, multiple water level, water temperature and radioactivity monitors would actuate alarms in the control room. A make-up water system is provided to keep up with smallleaks.

(e) Procedures and interlocks are provided to keep the crane from passing over the pool with heavy loads.

(f) The fuel building and the SFP are designed to accommodate the forces which might result from winds and missiles that might be generated by a tornado. Fur-ther, the spent fuel storage racks and the SFP cooling system are protected by structures designed to withstand these forces.

4-4

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

f For reference, the physical parameters and crane capacities of some typical PWR and BWR spent fuel pools are provided in Table 4.2.1 (from Ref. 6). Some sites share a single pool for multiple units, such as North Anna 1 and 2, Sury 1 and 2, and Oconee 1 and 2. At other sites a transfer canal exists between pools to allow for spent fuel movement, such as Browns Ferry 1 and 2, Calvert Cliffs 1 and 2, and Hatch 1 and 2. At a few sites, a spent fuel shipping cask is used, or available, to transfer spent fuel between pools (San Onofre 1,2, and 3, and Turkey Point 3 and 4).

i Table 4.2.1 Typical Pool Dimansions Other F4n4="=

Pool Dimensions Cask Area Areas Crane Clearance Plant L(ft) W(ft) E(ft) (sq. ft.) (sq.ft.) Capacity (ft)

PWRs 43 22.2 41.7 116 29 25 T 32 Ginna Indian Point 3 33 27 37 n/a n/a 40 T n/a Maine Yankee 41 37 38 100 230 125 T 24 North Anna 1 & 2 56.5 29.3 46.5 12 x 12 96 125 T 29 Oconee 1 & 2 71.3 15 38 131 0 100 T 29 Oconee 3 47.1 13.9 39 192 117 105 T 28 Palisades 38.8 14.7 38 9x9 8 100 T 28 Robinson 2 31 33.5 38.3 8.8x8.8 0 125 T 32 San onofra 1 14 39 39 n/a n/a 100 T n/a San onofre 2 44 23 46 n/a n/a 100 T n/a 37 40.5 10 x 12 0 105 T 28 St. Lucie 1 33 Surry 1 & 2 72.5 27.3 38.5 12 x 12 0 125 T 29 Turkey Point 3 41.3 25.3 40 n/a n/a 105 T 28 n/a n/a 105 T 28 Turkey Point 4 41.3 25.3 40 Bhh 56 34 38.8 160 200 125 T 28 Brunswick 1 Brunswick 2 56 34 38.8 160 200 125 T 28 40 31 37.8 15 x 12 0 125 T 27 Fitzpatrick Millstone 1 30.5 40.3 38.8 53 310 110 T 22 Monticello 40 26 38 50 20 85 T 23 Oyster Creek 39 27 40 153 21 100 T 24 Peach Bottom 2 40 35.3 40 10 x 10 0 125 T 32 Peach Bottom 3 40 35.3 40 10 x 10 0 125 T 32 32.8 26.1 38.8 77 0 100 T 26 Pilgrim 1 vermont Yankee 40 26 37.8 49 0 110 T 18 4-5

. .. \

4.3 Spent Fuel Pool Structures 4.3.1 BWR Mark I and Mark II Plants The spent fuel pool is located at the operating floor level, about 100 to 150 feet above grade.

i The pool floor and wall are designed for dead load and live load, hydrostatic pressure load, seis-mic load, thermal loads and loads resulting from the accidental drop of heavy objects. The thickness of the pool walls and floor is on the order of 4 to 6 feet. The horizontal and vertical  ;

loads from the pool floor are transmitted to the two longitudinal walls which are designed as deep girders supported at the peripheral wall of the reactor building.

4.3.2 PWR and BWR Mark III Plants The spent fuel pool is located at the ground level. The physical dimensions and design loads of the the pool are similar to the BWR Mark I and Mark II designs. Due to the lower elevation, the seismic response is relatively low in comparison to the elevated pools in the BWR Mark I and Mark II plants. The vertical and horizontal loads from the pool floor are transmitted to the ,

ground for plants with free standing or floor mounted fuel storage racks. For plants with laterally braced racks, the horizontal seismic loads from the fuel storage racks are transmitted to the pool wall at either the base level or at the base and upper seismic bracing level (about 14 feet above the base level). .

4.4 Spent Fuel Storage Rack Descriptions The following descriptions of spent fuel storage rack configurations are provided:

4.4.1 Low Density Racks (Cell Pitches 20 to 30 Inches)

The suberitical configuration is achieved by the physical separation of the assemblics in an open-frame aluminum of steel structure. Structurally the racks can be laterally braced at the up-per and lower levels or they could be bolted to the floor at four corners with the upper and lower grids connected by cross bracing.

4.4.2 Medium Density Racks (Cell Pitches 9 Inches (BWR) to 13 Inches (PWR))

The subcritical configuration is achieved by the flux trap principle. The assemblies are sur-rounded by stainless steel cans or cells which prevent the neutrons in the water region between the cells from returning to the fuel assemblies. The typical wall thickness is 1/8 inch. Struc-turally the racks can be laterally braced at the upper and lower grid levels, bolted to the floor at the four corners with the upper and lower grids connected by cross-bracing, or they can be can-tilever cells (2x2 or 4x4 modules to reduce flexibility) welded to a base structure.

4.4.3 Iligh Density Racks (Cell Pitches 6 Inches (BWR) to 9 Inches (PWR))

Suberiticality is achieved by the addinon of neutron absorbing poison material between the fuel assemblies. The poison is in the form of boron containing material such as boron-carbide, borated stainless steel, or borated aluminum. The storage cell walls have poison containing pockets. Structurally the racks are mostly free standing or laterally braced at the lower level.

The honeycomb construction provides structuralintegrity. The cell walls are typical 0.09 inches thick. The cells are attached to each other by fusion or spot welds.

4-6

l' BWR high density configuration can also be in the form of directional storage racks. In this 1 configuration the BWR assemblies are stored in 6 inch center-to-center racks, with a 5.3 inch open space between rows. No additional neutron absorber material is required in the rack struc-ture for criticality control.

4.4.4 Consolidated Fuel Racks The fuel assembly is disassembled and stored in a fuel canister. The canister is then stored in high density racks. The consolidation ratio can be 2 to 1. Two fuel assemblies can be com-pacted into the same physical dimensions of a single assembly. The non-fuel bearing material (such as grid spacers, guide tubes, etc.)is also compacted and stored. The compaction ratios for the non-fuel bearing material are estimated to be 10:1 for PWRs and 20:1 for BWRs (Ref. 7).

Since the non-fuel bearing material can take up room in the spent fuel pool, the consolidation ratio may be as low as 1.5 with a weighted average consolidation ratio of 1.63.

4.5 Evaluation of Spent Fuel Cladding Failure The results of work performed by Sandia (Ref. 8, Ref. 9) suggested that in certain fuel racking configurations (a) a self-sustaining zirconium-air oxidation reaction can be initiated, and (b) this self-sustaining reaction can propagate from one region of the pool to another.

These results were based on both experimental simulation and computer modeling. A computer program was developed by Sandia, called SFUEL1W, to evaluate conditions under which a self-sustaining Zircaloy reaction would occur and under what condition the Zircaloy fire would propagate to older stored fuel assemblies. Large uncertainties, associated with the phenomenol-ogy of Zirraloy oxidation and its propagation in spent fuel assemblies, were identified in the Sandia studies.

The SFUEL1W computer program was partially validated by Sandia (Ref. 9) and was also fur-ther validated by BNL (Ref.10). The calculated results of the SFUEL1W models were com-pared with existing Sandia National Laboratory (SNL) small-scale experimental results. The CLAD computer program, a modified version of SFUEL (an early version of SFUELlW), was used. CLAD was developed by Sandia to model the experimental test results (Ref. 9). Sandia had perfonned some verification studies against the experimental tests, but did not complete the work before funding ended. The BNL calculations with CLAD tended to result in an over-prediction of the peak cladding temperatures.

The NRC staff performed an independent verification of SFUEL1W using the CLAD computer program (Ref.11). The BNL verification program included some modifications to CLAD.

These modifications included the addition of helium properties to model the initial test condi-tions and a switch from helium to air flow, and an energy balance model to force conservation of energy on each gas contml model. In the staff program, additional modifications were made to CLAD. The most significant modification was the inclusion of the SFUEL1W gas heatup model.

This NRC staff modified version of CLAD was verified against two of the Sandia air tests and the results compared favorable with the available experimental data. The peak cladding tem-peratures calculations were in good agreement with the data. It was concluded that the SFUEL1W fuel, cladding and gas heatup models are satisfactory.

4-7

l l

The reaction rate equation for the oxidation of Zircaloy cladding in air used in the SFUEL1W computer program was also subject to uncertainties. BNL (NUREG/CR-4982) performed a literatum scarch related to the oxidation rate of Zircaloy in air. Based upon the current state-of- ,

the-art understanding of the associated phenomena and by performing sensitivity studies on '.he Zircaloy-air reaction rate correlation, it was concluded by BNL that the oxidation rate mudel used by Sandia in the SFUEL1W computer program is acceptable for the evaluation of spent fuel damage. For temperatures in the 800-1150 oC range (1470-2100 0F) the available data indi-cates that the Sandia correlation is valid, for exposure periods of 30 minutes. For longer periods the correlation may be non-conservative. At a constant temperature the rate of oxidation may increase with exposure time. This dces not alter the findings concerning the initiation and/or propagation of a s":lf-sustaming Zircaloy fire. Initiation is not influenced by the oxidation rate equation, and propagation can occur before cladding failure and relocation of the fuel rods oc-Curs.

The uncertainties in the Zircaloy oxidation propagation calculations under inadequate room ven-tilation conditions (most typical of the spent fuel storage pool structmes) were further studied by BNL (Ref.10) using the SFUEL1W computer program. A sensitivity study covering hot spent fuel decay power in the 20 to 90 Kw/MTU range was performed. For reference,90 Kw/MTU is the decay heat generation rate of fuel five to seven days following shutdown of a 3000 Mw(t) reactor. After one year the decay power level for a BWR is about 6 Kw/MTU and 11 Kw/MTU for a PWR. After two years, the decay power levels are estimated to be 4 Kw/MTU and 6 Kw/MTU respectively.

The SFUEL1W computer program is a finite difference solution of the transient equation for heating of the fuel rods considering:

- The heat generation rate from the decay heat and oxidation of the cladding,

- Radiation to adjacent assemblies and pool walls, and

- Convection to buoyancy-driven air flows.

The key assumptions and limitations of SFUEL1W are:

- The water drains instantaneously from the pool,

- The geometry of the fuel assemblies and racks remains undistorted,

- Temperature variations across the fuel rods are neglected, l - The air flow patterns are one-dimensional, anu

- The spaces between adjacent holders are assumed to be closed to air flow.

With respect to the limitation concerning instantaneous draining of the pool, this assumption simplifies the heatup model in the SFUWLlW computer program and is not intended to be rep-resentative of any accident sequence other than perhaps the catastrophic failure of the spent fuel structure from a beyond design base seismic event.

4-8

r After the water is drained from the spent fuel pool, the rods heat up until the buoyancy-driven air flow is sufficient to prevent further heatup. If the decay heat level is sufficient to heat the rods to about 900 oC (1650 0F) the oxidation becomes self-sustaining.

BNL has concluded (Ref.10) that:

- The likelihood of cladding fhe initiation is not very sensitive to the oxidation rate equation,

- The oxidation rate equation in SFUELlW is a reasonable representation of the available dats, and

- The likelihood of cladding fire initiation is most sensitive to the decay heat level and the storage rack configuration (which controls the extent of natural convection cooling).

It was also concluded that the oxidation pmpagation to older adjacent assemblies is likely if the decay heat level of the older adjacent assembly is high enough to heat that assembly to within 100 to 200 oC (200 to 400 0F) of the self-sustaining oxidation temperature. The radiation heat transfer from the burning assemblies then could be sufficient to raise the temperature of the older adjacent assembly to the self-sustained oxidation limit.

The following descriptions of spent fuel storage rack configurations are provided and are repre-sentative of the geometries used by both Sandia (Ref. 8) and BNL (Ref.10) to determine spent fuel storage configuration which can result in Zirraloy fhes and propagation of the fire to older stored fuel:

(1) High density PWR configuration: In this configuration, the fuel assemblies are tightly packed with neutron absorber material used in the rack structure to replace the reduced water moderator for criticality control. The center-to-center assembly spacing is 10.25 inches, the open gap between assemblies is 0.7 inches. This con-figuration is in use in nearly all PWRs, and is referred to as high density storage.

(2) Cylindrical PWR configuration: This configuration is typical of the early rack designs, used before at-reactor storage of spent fuel was required. The center-to-center assembly spacing is 12.75 inches, in a closed cylindrical stainless steel rack.

The typical cross sectional area of a PWR assembly is 8.4 by 8.4 inches. This is referred to as low density storage.

(3) Cylindrical BWR configuration: This configuration is typical of early BWR spent fuel storage rack designs. The center-to-center assembly spacing is 8.5 inches. The typical cross sectional area of a BWR assembly is 5.3 inches. This is referred to as low density storage.

(4) Directional BWR configuration: In this configuration the BWR assemblies are stored in 6 inch center-to-center racks, with a 5.3 inch open space between rows.

No additional neutron absorber material is required in the rack stmeture for criticality control. This is considered to be a high density storage configuration for BWRs.

4-9

Because of limitations in the SFUELIW computer program, BNL limited the BWR spent fuel analyses to the low density cylindrical configuration. The SFUEL1W computer program does not account for air flow between adjacent holders, an assumption which was based on the storage rack design. The self-sustaining oxidation analysis is govemed by the BWR channel box design, the air flow through the assembly. The SFUEL1W results are not significantly in-I fluenced by the BWR rack design.

1 The estimated likelihood of self-sustaming oxidation for various spent fuel rack configurations is provided in Table 4.5.1, based on a 12 month fuel cycle which is typical for most PWRs.

BWRs typically operate on an 18 month fuel cycle. Additional calculated results from the ear-lier Sandia work (Ref. 8) are also provided. In Table 4.5.1 the Critical Cooling Time is defined as the decay time to reduce the internal heat generation rate to a low enough value, the Mini-mum Decay Power, to preclude the Zircaloy cladding temperature from exceeding the self-sustained oxidation limit under air cooling. Considering a fuel cycle of not less than one year, the cooling time can be convened to a conditional probability of fire initiation.

The conditional probability of propagation of the Zircaloy fire to older stored spent fuel was evaluated by BNL (Ref.10). The results are provided in Table 4.5.2. In Table 4.5.2 the High Power Level is the decay heat generation of the recently discharged fuel and the Adjacent Power Levelis the decay heat generation rate of the older fuel at the Appmximate Decay Time. Per-feet ventilation is assumed for these two studies. The impact of no ventilation, which would result in the depletion of the oxygen from the air, is summarized in Table 4.5.3. The High Power Level and Adjacent Power Level are the same as described for Table 4.5.2.

After an extensive review of the SFUEL1W computer code and comparison to the SNL small scale experiments, BNL concluded that the code provides a valuable tool for assessing the likelihood of self-sustaining clad oxidation for a variety of spent fuel storage configurations (Ref.10). Additional studies performed by the staff supported the BNL conclusion.

For the purpose of evaluating the risk from beyond design basic accidents in spent fuel pools, the conditional probability of a Zircaloy cladding fire given a complete loss of water will be as-sumed to be 1.0 for PWRs and 0.25 for BWRs. The PWR value is based on the use of high den-I sity storage racks. The BWR value is selected based on the use of directional storage racks, I with the channel box in place.

For the proposed alternative to require recently discharged spent fuel be stored in low density storage racks, the risk reduction is estimated to be a factor of five (Ref.10). This level of risk reduction, resulting from the decrease in the cooling time needed to preclude a Zircaloy fire, is l

seen to be equivalent to the use of low density cylindrical storage racks with three inch inlet l

I holes. The PWR conditional probability of a Zircaloy fire is reduced to 0.2 and the BWR condi-l tional pmbability is reduced to 0.05. The actual risk reduction achievable may be greater than assumed. Open frame racks or cylindrical racks with larger inlet holes would result in an in-creased reduction in risk. The cooling decay time could be reduced to less than 20 days, for a conditional probability of 0.05 for both fuel types.

4-10

l l'j *

r. .

Table 4.5.1 Estimated Likelihood of Self-Sustaining Eircaloy clad oxidation For Various Spent Fuel Rack Configurations and Decay Heat Levels Inlet Mininnna Critical Spent Fuel Rack Orifice Decay Cooling Conditional Configuration Diameter Power Time Probability (inches) (Kw/MTU) (days) (per year) l BNL (Raf.10)

Last Discharge Only High Density PWR 10 11 360 1.0 5 ( 700 1.0 Cylindrical PWR 5 90 10 -0.0 3 45 50 0.14 1.5 15 250 0.7 Cylindrical BWR 3 30 30 0.08 1.5 14 180 0.5 SNL (Ref. 8)

Full Core Discharge Directional BWR 5 n/a 90 0.25 with channel box Directional BWR S n/a 30 0.08 without channel box PWR Open Frame -

n/a 10 -0.0 Cylindrical PWR 5 n/a 20 0.05 3 n/a 120 0.33 1.5 n/a 250 0.7 Notes: Conditional probability estimated from NUREG/CR-0649 for ma=4== peak cladding temperatures less than 600 OC. Self-sustaining Eircaloy oxidation onset is approximately 900 OC.

n/a - Not Available. In the Sandia studies, the decay power level was not reported, however these analyses were performed for a full core discharge situation.

4-11

_____--___._m. . . _ _ . - _ _ _ _ _ _ _ _

(

j. 1 l

Table 4.5.2 Estimated Likelihood of Propagation of Eircaloy Fire to Older Spent Fuel For Various Spent Fuel Rack Configurations and Decay Heat Levels Results for High Density PWR Racks With Large Inlet Holes (10" Diameter) and Perfect Ventilation High Adjacent Approximate Propagation Power Power Decay Level Level Time (kw/MTU) (kw/MTU) (Days) (Yes/No) 11.0 5.9 365 .Tes 19.2 5.9 365 Yes 90 5.9 365 Yes 90 4.0 730 No Results for Cylindrical PWR Racks With 3" Diameter Inlet Roles and Perfect Ventilation High Adjacent Approximate Propagation Power Power Decay Level Level Time (kw/MTU) (kw/MTU) (Days) (Yes/No) 90 11 365 No 90 19 180 Yes (1)

Results for Cylindrical PWR Racks With 1.5" Diameter Inlet Holes and Perfect Ventilation High Adjacent Approximate Propagation Power Power Decay Level Level Time (kw/MTU) (kw/MTU) (Days) (Yes/No) 90 11 365 Yes 90 5.9 730 Yes 90 3 1100 No 15 11 365 Yes 5.9 730 No Not. this is unlikely situation, assumes recent spent fuel discharge six monthb lter previous discharge. Fuel cycles are typically 12 to 18 months.

4-12

.. i

(- . 1 I'

1' l Table 4.5.3 Summary of Radial oxidation Propagation Results For Various PWR Spent Fuel Rack Configurations and No Ventilation Spent Fuel High Power Adjacent Power Propagation Rack Level Level Configuration (kw/MTU) (kw/MTU) (Yes/No)

Cylindrical 1.5" holo 90 5.9 Yes 90 3 No(1)

Cylindrical 3.0" hole 90 5.9 No 19.2 11 Yes High Density 10" hole 90 4 NoII)

Note: (1) Without ventilation the fire becomes oxygen starved. Oxygen deple-tion prevents propagation.

4.6 Quantification Of Accident Sequences in Spent Fuel Pools 4.6.1 Structural Failure Due to iWissiles High energy missiles which might impact with the spent fuel pool might result in sufficient structural damage to prevent cooling of the spent fuel. Missiles generated by tornadoes or from a turbine failure are considemd during plant licensing. As indicated in Section 4.1, these acci-dent sequences am reviewed by the NRC staff to assure compliance with the General Design Criteria.

In WASH-1400, the pmbability of a turbine failure and missile generation has been estimated to 4

be on the order of 1x10 per mactoryear, and the limiting strike probability for the spent fuel pool has been estimated to be 4.1x10-3, given an energetic missile. The probability of a turbine missile hitting the spent fuel pool is therefore estimated to be 4.1x10-7 perreactor year.

The probability of a beyond design basis tornado stnking a reactor site has been estimated to have a mean value of about 5x10-0 per reactor year (WASH-1400, Ref.1), with a mean proba-bility for all tornadoes of 5x10 4 per reactor year. A typica1 reactor site is estimated to be about 620 acres (one square mile), or greater (Ref. 6). The plan area of a spent fuel pool (50 feet wide by 60 feet long) is 1x104square miles. The probability of a beyond design basis tornado strik-ing the spent fuel poolis therefore on the order of 5x10-10 per reactor year. The probability of a tornado missile striking the spent fuel pool has been estimated based on the Zion site to be on the order of 1x104per reactor year (Ref.16).

4-13

O The likelihood of a missile, turbine or tornado generated, damaging the spent fuel pool and resulting in an unrecoverable loss of water is estimated to be less than 0.01 per demand (Ref.

16). The missile would have to cause sufficient damage to prevent filling or repair of the spent 4

fuel pool. Given tt estimated combitsed likelihood of a missile strike on the order of 1x10 per reactor year, the timated probabihty of the structural failure of the spent fuel pool from a missile resulting in a. 's of cooling of the spent fuelis less than lx10-7 per reactor year, on the order of 1x10-8per reauor year.

4.6.2 Structural Failure Due to Aircraft Crashes The probability of an aircraft stnking the spent fuel pool is proportional to the vulnerable area of the structum, the aerial crash density of an aircraft and the number of operations on applicable runways. SRP Section 3.5.1.6 is used to derive the hit frequency for a reactor site.

The probability of structural failure of the spent fuel pool as a result of an aimraft crash has been estimated using Zion PRA results (Ref.16). The mean hit frequency is estimated to be 6x10-9 per reactor year with 5% and 95% confidence bounds of 5x10* to 2x10-8 per reactor tear. The probability of structural failure of the spent fuel pool resulting in significant spent fuel damage was estimated to be less than 1x10-10 per reactor year in NUREG/CR-4982 (Ref.10) and in EPRI NP-3365 (Ref.16).

4.6.3 Structural Failure Due to Heavy Loads Drop (Shipping Cask)

Load handling in the SFP area is myiewed to assure conformance with GDCs 2,5,61 and 62 under Section 9.1.4, " Light Load Handling System (Related to Refueling)," and with GDCs 2,4, 5 and 61 under Section 9.1.5, " Overhead Heavy Imad Handling Systems," of the SRP. In addi-tion, the requirements identified in the resolution of GSI A-36, " Control of Heavy Loads Near Spent Fuel," as specified in Generic Letter 85-11 (Ref. 4), " Completion of Phase II of Control of Heavy Loads at Nuclear Power Plants (NUREG-0612)," are reviewed to assure that the im-plementation of Phase I of NUREG-0612 (Ref. 5) "has provided sufficient protection such that the risk associated with potential heavy load dmps is acceptably small." Adequate justification is provided by means of (1) a single-failum proof crane, (2) operator training and procedures, maintenance and inspection pmcedures, safe load paths and mechanical or electrical stops to prevent movement of heavy loads over irradiated fuel, and/or (3) load drop analyses. The staff concluded, in GL 85-11, "that satisfaction of the Phase I guidelines assures that the potential for a heavy load drop is extremely small."

The estimated probability of structural failum of the spent fuel from a shipping cask drop was esdmated by BNL based on a cask handling assumption of two fuel shipments per week, sunilar to that used in WASH-1400. The estimated probability of a shipping cask being dropped on the spent fuel pool wall,4without consideration of the requirements from GSI A-36, was estimated by BNL to be 3.1x10 per reactor year (Ref.10). The likelihood of pool damage was estimated by BNL (Ref.10) to be 0.1 per demand, one-in-ten drops causing sufficient damage to com-pletely drain the pool, with an uncertainty range of 0.01 to 1.0. The estimated reduction in the probability of a shipping cask drop for a plant which complies with the resolution of GSI A-36 (Ref. 4) has been estimated by BNL to be a factor of 0.001, for a revised probability of 3.1x10-8 per reactor year, including the 0.1 conditional probability of failure given a shipping cask dmp.

4-14 I

_ - _ _ - _ _ - _ _ - _ _ - - _ _ _ . _ _ . _ - - _ _ _A

. i

.. j A more detailed analysis of the resultant damage to a spent fuel pool structure as a result of a j shipping cask drop was performed by LLNL (Ref.17). A BWR and PWR spent fuel pool were  ;

l analyzed for a variety of cask weights and dmp heights. The results of the LLNL analysis indi-cate that the pool wall could suffer severe damage as a result of a cask drop. The indicated l 1

regions of potential reinfoming steel yield am quite extensive and while the integrity of the pool liner is difficult to predict, it was concluded by LLNL that it seems likely that the liner would be severely damaged. 'Ihe estimated probability of the structural failure of a spent fuel poolresult-ing from a dropped shipping cask is therefore considered to be equivalent to the probability of dropping the cask,3.1x10-7 per reactor year, given a cask handling rate of twice per week (104 per reactor year).

At the present time, spent fuel is mostly being accumulated in spent fuel pools. At a few facilities, the older fuel assemblies are being transferred to dry storage areas on site. To es-timate the probable number of cask handling operations per year the following assumptions are made:

(1) The spent fuel pool capacity, with a full core reserve, has reached a licensing limit (either structurally or due to cooling capacity restrictions).

(2) The capacity, maximum number of allowable assemblies, is maintained and excess assemblies are transferred to either an onsite dry storage area er the DOE repository.

(3) Based on a 12 month fuel cycle in PWRs with 200 assemblies in the core, about 70 assemblies would have to be removed annually fmm the spent fuel pool to accommodate reloads.

(4) Based on a 18 month fuel cycle in BWRs with 800 assemblies in the com, about 130 assemblies would have to be removed annually from the spent fuel pool.

(5) The weight of a PWR assembly is approximately twice that of a BWR as-sembly, 657.9 kg (1450 lbs) versus 319.9 kg (700 lbs). l (6) Based on TN-24P (Ref.18) and MC-10 (Ref.19) dry storage cask designs, j twenty-four PWR, or 48 BWR, assemblies can be moved per cask. The cask j weight is in the 100 ton range.

The estimated number of transfers per reactor year is therefore estimated to be about a factor of ten lower as compared to the WASH-1400 rate of 104. The probability of structural 4 damage to l the spent fuel pool as a result of a dropped shipping cask is estimated to be 3.1x10 per reactor

{ year (best estimate) for a reasonable cask handing rate. The upper bound estimate is taken from NUREG/CR-4982 (Ref.10) as 3.1x10-7 perreactor year.

4.6.4 Reactor Cavity and Transfer Gate Pneumatic Seal Failures Inflatable, pneumatic seals are used during refueling operations in PWRs to seal the gap be-tween the reactor pressure vessel flange area and the biological shield walls. This permits flood-ing of the reactor pressum vessel cavity above the core to allow for the safe handing of the fuel.

In B WRs, the reactor cavity seals are typically permanent stainless steel expansion bellows, and not subject to the failure modes associated with the pneumatic designs. Pneumatic seals are also 4-15

used to partition areas of the spent fuel pool, for example, between the shipping cask handling area or fuel transfer tube and the main spent fuel storage area. Ten reported cases of pneumatic seal failures resulting in actual or potential loss of water from spent fuel pools are listed in Table 4.6.1, three involving the refueling cavity seal and seven involvmg other pneumatic seals.

Three of the ten reported events involved the failure of the refueling cavity seal. In one case, no fuel was in the spent fuel pool and the failure occurred during installation and testing. In the other two events, the fuel transfer canal was closed at the time and no actual dramage from the spent fuelpool would have occurred.

At Surry 1 in May 1988 following the water loss from the refueling cavity seal failure, the plant operator opened the fuel transfer canal path to aid in refloodmg the reactor cavity. Personnel in-side contamment, on the fuel crane bridge, had to leave containment as a result of high radiation levels. Operators did not enter appropriate procedures for a loss of refueling cavity level, and the existing procedures provided inadequate guidance to operations personnel on a rapid loss of cavity water (Ref. 20). Procedures had been developed at Surry to address cavity water loss in response to IEB 84-03 (Ref. 21), but were inadvertently omitted from revised procedures in 1987. A review of the design by the seal vendor (Presray), who has stated that they manufacture and continue to supply most of the refueling cavity seals used throughout the industry, deter-mined that the design at Surry is unique and inadequate. A seal backup plate should have been provided to prevent movement of the seal. The IEB 84-03 review by the licensee failed to iden-tify the weakness in the seal design, although procedures were developed to address seal failure.

In the remaining seven reported instances, pneumatic seal failures have occurred which could result in the draining of the spent fuel pool. The event at Hatch (see Table 4.6.1 ) is considered to be unique, and in two of these seven cases there was no spent fuel in the pool at the time of the event.

For the purpos: of evaluating the potential for spent fuel damage from pneumatic seal failures, the event frequency was initially estimated by BNL to be 0.01 per reactor year and is generally applicable to PWRs. Based on advances in seal designs, increased awareness and surveillance resulting from IEB 84-03, BNL estimated the present failure rate to be an order of magnitude less,0.001 per reactor year. In addition to the seven events identified in NUREG/CR-4892, two more events related to seal failures where identified (Ref. 22) for a total of nine events in about 900 years of experience. The Surry 1 event, which occurred after NUREG/CR-4982 was pub-lished, would not alter the event frequency (ten events in 1,000 years of experience).

Not all seal failures will lead to loss of water from the spent fuel pool. A failure of the refueling cavity seal must occur coincident with an open fuel transfer canal. Even under this assumption, the spent fuel pools are designed to preclude significant (a few inches) fuel uncovery due to the leakage. The transfer canalis either located above the top of the storage racks or a weir is used to prevent lowering the level below the top of the fuel.

Seal failures coincident with fuel handling operations are being addressed as a separate issue, Generic Issue 137, as discussed in Section 4.9.3.

4-16

1 .

. .+-

Table 4.6.1 l- Events in Which Pneumatic Inflated Seals Have Failed Date Plant Seal Location Cause- Result 9/72 Pt. Beach 1(1) Transfer gate Failure of air supply 11,689 gal leak 10/76 Brunswick 2 Inner pool Air leak in seals an 5" level drop gate compressor power supply fallure 6/80 Trojan Transfer gate Not inflated prior to 10" below T.S.

draining refueling cavity, level alarm also failed 2/81 Davis-Besse Transfer gate Low seal pressure 15" level drop 5/81 ANO-2 Transfer gate Maintenance error 1000 gian leak air supply shutoff 8/84 Haddam Neck (2) Cavity seal Design weakness 200,000 gal leak in 20 minutes 10/84 San Onofre 2(1) Gate seal Air compressor fails 20,000 gal leak 11/84 San Onofre 2(1) Cavity seal (3) Manufacturing defect 19.5" level drop seal ruptured 12/86 Hatch (4) Pool canal Air supply valve 141,000 gal leak flexible joint closed 5/88 surry 1(5) Cavity seal Maintenance error 30,000 gal leak and des gn weakness Notes: (1) No spent fuel in pool at time of event.

(2) Fuel transfer canal closed. No water loss from pool.

(3) Failure during installation and leak testing (4) Make-up system cycled to maintain level, undetected for about 7.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. Hatch has interconnected transfer canal between two pools. Seal re-quired for SSE considerations. Unique design.

(5) Licensee had reviewed seal design following Haddam Neck event, and detes. mined design to be adequate. Procedures developed to address seal failure subsequently omitted. Maintenance error isolated air supply to seal, combined with low backup nitrogen accumulator resulted in seal failure. Pas-sive backup seal failed due to improper installation and design. Design determined to be unique to Surry. Maximum leakage 6,500 gpm for 4 minutes.

4-17

___----_---_--_--____--__.-__.-----__--__--_----_r---- .

In response to EB 84-03, the analyses supplied by licensees indicated that the failure of a pneumatic refueling cavity seal in most PWR plants would not result in massive leakage be-cause of the relatively narrow gap to be scaled and the geometric shape of the seal (the Haddam Neck design was determined to be unique because of the large gap sealed). Also, leaks from seal faihtres in the transfer canal gates would be limited, in most cases, because the leakage would be into a confined volume, for example from the pool into a drained up-ender sump. This volume is small in comparison to the spent fuel pool volurne and the level in the spent fuel pool would decrease slightly.

Licensee responses to EB 84-03 have been reviewed by the NRC staff to determine the credible leakage that could result fmm pneumatic real failures,in particular the refueling cavity seal. Al-though BWRs do not generally use pnemnatic seals (xrmanent stainless steel bellows are used),

some licensees did provide estimates for credible leacage in the highly unlikely case these seals were to fail. The seal failure leakage rates provided by licensees are listed in Table 4.6.2, and are representative of the maximum flow rates achieved for a fully flooded refueling cavity. As the water level drops, the hydrostatic pressure decreases and the flow rate will also decrease.

For example, in the Summer submittal (Virgil C. Summer response to EB 84-03, October 16, 1984, Dociet No. 50-395), even though the maximum flow rate is 5,500 gpm, it was estimated that it would require 160 minutes to drain the cavity to the level of the seal ring with the transfer tube open. If the transfer tube is closed, the spent fuel pool will not drain. The Watts Bar sub-mittal (Watts Bar response to 1EB 84-03, December 6,1984, Docket No. 50-390) estimated the time to drain to the reactor vessel. flange region was about 95 minutes. The Catawba /McGuire submittal (William B. McGuire 1 and 2 and Catawba 1 and 2 responses to EB 84-03, November 11,1984, Docket Nos. 50-369, -370, -413, and -414) provided the most comprehensive assess-ment of postulated leak rates. The times to drain the refueling cavity, considering the hydros-tatic pressure changes as level decreases, ranged from 12 minutes for 100% gmss failure to 414 minutes for a 1/16 inch gap around the entire seal circumference. For the 25% gross seal failure, the time was estimated to be 65 minutes. This case was reported to be identical to the Haddam Neck seal failure event.

The catastrophic complete failure of the refueling cavity seal is not considered to be credible.

The Haddam Neck seal failure is considered to the most limiting case. Even if a catastrophic failme occurred, with one to two feet of water remaining above the fuel in the spent fuel pool.

there woula be at least two hours available for the operator to take emergency actions to provide cooling for the spent fuel before the water covering the spent fuel boils off. This time estimate is based on Table 4.6.3, which shows that the maximum rate of boiling following a full core dis-charge five days after shutdown is equivalent to one foot every two hours (based on 49 hours5.671296e-4 days <br />0.0136 hours <br />8.101852e-5 weeks <br />1.86445e-5 months <br /> to boil off 23 feet of water). For a normal refueling, assuming 1/3 core discharge, the recovery time would be five hours. An example of a procedure develoxd to address this highly unlikely situation was provided in the Catawba /McGuire response to MB 84-03. The alternate cooling method is to recirculate water from the containment sump to the refueling water storage tank and then to the spent fuel pool using a residual heat removal pump.

The fuel tninsfer canal structure, located between the fuel transfer canal in the reactor vessel refueling pool and the fusi storage pool in the fuel storage building, is typically equipped with a metal expansion joint to accommodate flexure. In the event the fuel transfer tube tube expan-sion joint failed, the outer sleeve slip joints would limit the leakage flow. The maximum calcu-lated leakage through a slip joint was estimated to be 400 gpm (Indian Point 2 supplemental response to EB 84-03, March 31,1987, Docket No. 50-247).

4-18 l

l

BNL estimated (Ref.10) that the frequency of a serious loss of pool water inventory resulting from a pneumatic seal failure would be on the order of one in one hundred events (0.01). The combined probability esdmate was 1x10-5 per reactor year for a serious loss of pool water event (Ref.10). This estimate does not include credit for recovery actions to mitigate or stop the draining event from resulting in spent fuel damage. A conditional probability of failure to recovery of 0.05 was used by BNL, based on previous studies (Ref. 27), resulting in a frequency estimate of 5x10-7 per reactor year for the seal failure accident sequence.

To assure that the BNL estimate is appropriate for this event, the NRC staff examined the human error probability (HEP) of failure to diagnose this event in sufficient time to take mitiga-tive action. As discussed above, licensee responses to IEB 84-03 demonstrate that there is con-siderable time to respond to a seal failure event, even for postulated leakage on the order of several thousand of gallons per minute. The refueling cavity seal would have to fail while the transfer canal was open. For the protection of the spent fuel, the transfer canal gate valves can be closed and make-up water provided to restore the pool level if the level in the refueling cavity falls below the reactor pressure vessel flange region. Using the nominal HEP screening model (Ref. 21) to estimate the HEP for failure to diagnose the event, the median joint HEP for a one to four hour time period is 1x104to 5x10-5. The error factor on the median HEP is 30, there-fore the HEP value for failure to diagnose a serious seal failure is estimated to be in the 3x10-3 to 1.5x104range.

On December 17,1984 an IE Information No"ce, IN 84-93, " Potential for Loss of Water From the Refueling Cavity," was issued (Ref. 24). In this notice the staff concluded that " Adequate emergency procedures and properly calibrated refueling cavity water level instrumentation are considered to be important in the mitigation cf any loss-of-cavity-water accident."

Based on the heightened awareness to refueling cavity seal designs, installation, testing and maintenance of the seals, and to the need for adequate procedures to address seal failures, as identified in IEB 84-03 and in IN 84-93, and considering the time available to diagnose a serious seal failure on the order of one hour, the staff updated the estimate of the frequency of loss of spent fuel pool water resulting in fuel damage. Given a serious seal failure fre0uency estimate of 1x10-5 per reactor year with an HEP conditional failure probability of 3x10 (median value for one hour with error factor) to diagnose the seal failure, the best estimated frequency is 3x10-8 per reactor year of a seal failure resulting in spent fuel damage. The upper estimate for this event is 5x104per reactor year, based on the BNL evaluation in NUREG/CR-4982 (Ref.10).

4-19

Table 4.6.2 Refueling Cavity Seal Leak Rates Following Seal Failure Plant Type Leak Rate Assumptions (9Pm)

Point Beach 1/2 PWR 62 break area of 0.5 square inches Turkey Pt. 3/4 PWR 50 Comanche Peak PWR~ 100 break area of 1.0" diameter Oconee PWR 50 total dislodged inner seal TMI PWR 4,700 marinnnn, major gasket failure Vogtle PWR 175 8" section of gasket Wat':s Bar PWR 3,176 1/16" gap around seal Summer PWR 5,500 60 mil gap around seal Prairie Island 1/2 PWR 4,200 2" gap Watts Bar PWR 3,200 Haddam Neck PWR 10,000 actual seal failure data Catawba /McGuire PWR 103,642 gross failure, 100% of seal 20,467 gross failure, 25% of seal 3,210 1/16" gap around seal LaSalle BWR 185 Vermont Yankee BWR 500 FitzPatrich BWR 370 inner bellow seal 7,100 outer bellow seal Note: Leak rates are sawinnna flows with full refueling cavity wkter level. As level decreases, flow rates will decrease.

4.6.5 Inadvertent Draining of the Spent Fuel Pool There are other mechanism for draimng the spent ftiel, in addition to the seal failures discussed previously. Pipe breaks in the cooling system or heat exchangers, or siphoning paths could ,

result in loss of water. There have been a number of recent events resulting in partial draining 1 of spent fuel pools. These have been identified in IE Information Notice 88-65, " Inadvertent Drainages of Spent Fuel Pools." (Ref. 25)

At Wolf Creek (12/22/87) a valve in a retum line to the refueling water storage tank (RWST) l was left o xn following use of the spent fuel pool cleanup system to clean the RWST. The spent fuel pool Level indicator and low level alarm were both inoperable in the control room. Succes-sive tripping of the spent fuel pool coolir.g system pump alerted the operators to a pmblem. The minimum level in the spent fuel pool was 22 feet above the fuel.

4-20 t

4 At River Bend (9/20/87) an antisiphoning device in the purification system suction line was plugged in the upper spent fuel pool. River Bend is a Mark III BWR with an upper spent fuel pool near the reactor pressure vessel. Fuel movement to the primary spent fuel pool is not necessary unless the spent fuelis being completely discharged from the reactor. The upper pool was intentionally dramed below the level indicator range to accommodate placement of the steam dryer in the pool. When using the condensate storage tank (CST) to refill the upper pool, valve misalignment result in a siphoning effect. High radiation alarms and a level increase in the CST alerted the operators to the problem. The manual valves in the purification line were closed to stop the drainage.

At San Onofre 2 (6/22/88) a siphoning path was present in the purification system, and ap-proximately 9,000 gallons were siphoned. Although siphon breakers, check valves, and locked valves were installed, the admmistrative controls were not established allowing alignment of the system which led to the siphoning event.

Operating procedures for the interconnected systems associated with the spent fuel pools either were not sufficiently detailed or were incorrect and failed to prevent alignments causing unin-tentional drainage. At Wolf Creek procedures did not exist. Also surveillance procedures were not implemented to ensure the operability of all instrumentation and control equipment at Wolf Creek.

At Turkey Point 4 (8/16/88) approximately 3,100 gallons of water was released from the spent fuel pool through a vent valve on a failed pump. This event occurred after IE Information Notice 88-65 was issued.

At Surry Unit 1 (10/2/88) a small leak developed in a pneumatic seal in the fuel transfer system as a result of the an accidental pinhole guncture in the single air supply line. The leak was promptly detected and stopped before sea integrity was lost. The reactor cavity seal was not in-stalled at the time and if the seal failed, the loss of water could have lowered the spent fuel pool level to within 13 inches above the top of the stored fuel. IE Information Notice 88-92,

" Potential for Spent Fuel Pool Draindown," was issued on November 22,1988 (Ref. 26) to alert licensees to potential problems resulting from failure of the fuel transfer canal door seal.

The inadvertent draining of a spent fuel pool should be precluded by design or admmistrative procedures. Antisiphoning devices, or approved system alignment procedures should be in place to assure that the primary safety function of the spent fuel poolis not compromised. To this end, instrumentation to alert operators to potential pmblems should be operable. IE Infor-mation Netice 88-65 identified these issues and all holders of OLs and cps are expected to review the information and consider appropriate actions to avoid similar problems.

The frequency of a siphoning event was estimated (Ref. 27) to be 0.001 per reactor year, based on a break in the cooling system. An 0.01 conditional failure probability of the cooling system to isolate was assumed. Further it was assumed that the conditional failure probability of the backup make-up system was 0.015. The frequency of a siphoning event was estimated to be 1.5x104 per reactor year (Ref. 27). Based on a conditional probability of an anti siphoning check valve failure of 0.08 (Ref. 27), the frequency of this scenario resulting in spent fuel damage is estimated to be 1.2x10-8 per reactor year.

4-21 l

l

The operational experiences concerning spent fuel pool component performance indicates that partial pool draining resulting from non-seal failure related causes, such as inadvenent siphon-mg, do not result in a significant loss of water from the spent fuel pool (Ref. 28). At San Onofre 2, for example, the 9,000 gallons is equivalent to about 19.5 inches of water. At Davis-Besse, on June 6,1980, an improperly calibrated level alarm did not actuate until the level was 10.25 inches below the Technical Specification limit. At Trojan, June 10,1982, misalignment of the spent fuel pool purification system as a result of not closmg the door between the spent fuel pool and the fuel transfer canal resulted in a lowering of the level to 21 feet 9 inches above the top of the stored fuel. The Technical Specification limit is 23 feet. The rate of draining was 132 gpm.

As a result of increased awareness concerning the development and use of proper administrative procedures for system alignments and the use of anti-siphoning devices and the need for spent fuel pool level indication through the issuance of IE Information Notice 88-65, the staff con-cludes that the frequency of spent fuel damage from inadvertent draining of the spent fuel pool is less than 1x10" per reactor year, and the best estimate value is 1.2x10-8 per reactor year based on previous estimates (Ref. 27).

4.6.6 Loss of Cooling /Make-Up The acceptance criteria associated with the general design and operation of the spent fuel pool and it's related support systems, primarily the cooling and make-up systems, are based on the long time interval available to the plant operators to diagnose and correct failures in these sys-tems. In WASH-1400, for example, it was assumed that the likelihood of failure to recover from loss of cooling was 1x10-6 per event. With an assumed loss of cooling event frequency of 0.1 per reactor year, the probability of damage to the spent fuel was judged to be extremely small,1x10-7 per reactor year.

In WASH-1400 the assumed spent fuel poolinventory was limited to about 2/3 of a full core. - A pool loading of 1/3 of a core with 150 days decay and 1/3 with three d . s decay was assumed for the limiting condition. Approximately nine days (216 hours0.0025 days <br />0.06 hours <br />3.571429e-4 weeks <br />8.2188e-5 months <br />) woulu .e available before the 50,000 cubic feet of water in the spent fuel pool would be completely boiled off. The average pool loading assumption resulted in 3.8 weeks (640 hours0.00741 days <br />0.178 hours <br />0.00106 weeks <br />2.4352e-4 months <br />) available for the repair of the cooling system and/or water make-up to be accomplished. In WASH-1400 it was believed that spent fuel shipping would be occurring on a weekly basis and the likelihood of failure to recover was estimated to be 1x10-6, However, spent fuel is not being shipped within the United States and the spent fuel pool inven-tories are larger than assumed in WASH-1400. To determine the available time for recovery, a simplified calculation was performed. A 3000 Mw thermal plant is assumed to have discharged 1/3 of a full core annually for a period of 20 years, for a spent fuel pool heat load of 3.5 million BTU /hr of decay heat. Older fuel would not increase the heat load significantly. Based on the pool data provided in Table 4.2.1, it is assumed that there is 32,000 cubic feet of water covering the spent fuel, a 30 foot by 40 foot surface with a depth of 23 feet over the spent fuel. Heatup of only the water is assumed. The results of the calculation are provided in Table 4.6.3 for the case when 1/3 of the core is recently discharged or the case when the full core is discharged. Also provided in the table is the needed make up rate, in gallons per minute, to match the boil off rate.

. 4-22

i Based on Table 4.6.3, the time available to recovery from loss of cooling can be estimated as at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for the most limiting case - full core discharge five days following reactor shut-down. At this time the water level covering the spent fuel (approximately ten feet) is adequate to provide shielding to maintain the radiation levels in the spent fuel storage area to a low enough value to permit limited operator access to the SFP area as required to establish make-up.

The make-up capacity is less than 100 gpm and the assumption that the fire system can be used to provide make-up appears to be reasonable. It is also noted that the frequency of this limiting condition is estimated to be less than 5% of the lifetime of the facility. Full com offloading is an unusual occurrence, cxcept during ten year inservice inspections of the reactor pressure vessel Assuming four inspections per life, and eight additional unanticipated offloads, the estimated i frequency of having a full core in the spent fuel pool (assuming the full core is in the pool for a period of 30 days) is 1 year per 40 year life or 2.5% of the time. For the typical expected condi-tion of a 1/3 core discharge five days after shutdown, the time available to restore cooling and/or establish make-up is three days, For recovery actions which would not require operators to enter the potentially high radiation area, the available time to recovery cooling would be between two days, the most limiting case, and five days, for a normal refueling case. If the loss of cooling event occurs 30 days after discharge, the recover time intervals nearly double.

The spent fuel pool is equipped with temperature and level instrumentation to warn the operator of a degrading condition. Although these instruments are not safety grade, they do alarm in the control room. The spent fuel pool storage area also contains radiation monitors to alert the operator to degrading situations. These instruments provide the operator with the information necessary to initiate appropriate safety actions to assure that the safety function of the spent fuel pool, to maintain the spent fuel assemblies in a safe and subcritical array during all credible storage conditions,is not compromised. The issuance of IE Information Notice 88-65 (see Sec-tion 4.6.6 above) has emphasized the importance of assuring that the spent fuel pool level in-strumentation is operable and that surveillance procedures are in place to assuru the the in-strumentation is properly maintained.

Operator performance, the human error probability (HEP), is estimated from NUREG/CR-1278 Chapter 12, " Diagnosis of Abnormal Events." (Ref. 23) The nominal screening model is used.

Table 4.6.4 lists the median joint HEPs and the error factors for a variety of assumptions for failure to diagnose a loss of cooling event. Typical repair time estimates, as well as failure rates, for the components of the cooling system are provided in Table 4.6.5.

The failure rates in Table 4.6.5 are for all failure modes. For a series system which has one pump, one heat exchanger, one level indicator and four valves, the estimated failure rate of the system is 0.15 per reactor year. The mean time to repair the system is about 34 hours3.935185e-4 days <br />0.00944 hours <br />5.621693e-5 weeks <br />1.2937e-5 months <br />. This is representative of the munmum cooling system allowed under current requirements, and supports the estimated failure rate of a single cooling system of 0.1 per reactor year previously used in WASH-14C0 and in NUREG-0933 (Ref. 27). Based on the mean time to repair the spent fuel pool cooling system,34 hours3.935185e-4 days <br />0.00944 hours <br />5.621693e-5 weeks <br />1.2937e-5 months <br />, it is seen that there is adequate tiu e to repair the cooling system before the spent fuel pool level decrease to a level where spent fuel damage would occur and the make-up rates needed to match the boil off rates are not excessive.

Four " generic" fuel pool cooling and make-up systems have been examined to estimate the pos-sible range of failure frequencies for these systems. The four representative systems, based on current SRP acceptance entena, are:

4-23

l; l

System A: Minimum cooling and make-up system requirements. One full capacity l cooling train with redundant active components (i.e., valves, pumps, etc.). One l i

Category I make-up system and one backup pump or system (not requued to be Category I) which can be aligned to a Category I water supply.

System B: Muumum cooling and make-up system requirements (System A) with credit for make-up from tne fire system for recovery.

System C: Typical older system. One cooling train with backup active com-ponents (but the backup components are required to supplement coolmg about 30%

of the time); and, one safety grade make-up system and one non-safety grade make-up system.

System D: Typical older system (System C) with a third make-up train available for recovery (i.e., the fire system).

Systems A and B are not intended to represent actual systems. Rather, they are representative of the minimum requirements in the current SRP.

The failure rates and systems failure frequencies are based on data from WASH-1400 and as-sumptions used in NUREG-0933 (Ref. 27). Specifically:

1. A 0.1 per reactor year frequency for the initiating event, loss of cooling, is based on WASH-1400 estimates. As discussed above, this is the expected failure fre-quency for a typical single train system based on typical component failure rates.

2. The conditional failure probability of 0.05 for the second cooling train (Systems A and B) represents a relatively high common mode failure probability. This value was used in NUREG-0933 to assist in the prioritization of tlus genenc issue.

3. The conditional failure probability of 0.3 for the second cooling train (Systems C and D) represents the assumption that both cooling pumps are necessary 30% of the time. Thus a failure of either pump represents a failure of the system.

4. Train 1 of the make-up system is assumed to be independent of the cooling sys-tem and is assigned a low common cause contribunon. The likelihood of a prolonged station blackout events is assumed to be low. This assumption is sup-ported by a recent study completed by Sandia (Ref. 28). For 63 recorded incidents of loss of off-site power, the longest recovery time reported was about nine hours (for severe weather related losses), with the sample mean recovery time for all causes of 1.2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. The conditional pmbability of 0.015 is used, based on RHR relir.bility in the LPCI mode (WASH-1400).

5. Train 2 of the make-up system is assigned a conditional failure probability of 0.05. This system is not powered by emergency power buses and may be put out of service by a common mode failure of the spent fuel pool cooling system.

l f The estimated failure frequency of the spent fuel pool cooling and make-up systems resulting in I a heatup of the spent fuel without recovery is summanzed in Table 4.6.6 for each of the four sys-l tems.

4-24

- _ _ _ _ ____ _ _ a

t Table 4.6.3 Heatup and Boil Dry Times for a Typical Spent Fuel Pool Based on 1/3 Core Discharge Based on Full Core Discharge Days Q-decay Esat Boil Make-Up Q-decay Heat Boil Make-Up Aftar 1/3 core to 212F Off Boil Full Core to 212F Off Boil Shut Discharge fm 125F Water Off Discharge fm 150F Water Off Down (BTU /hr) (hrs) (hrs) (gpm) (BTU /hr) (hrs) (hrs) (gpm) 5 1.51x107 11.' 125.0 31.9 3.82x107 3.1 49.3 81.0 10 1.22x107 1?. 154.9 25.8 2.95x107 4.1 63.8 62.6 30 8.87x106 it. 212.2 18.8 1.97x107 6.1 95.8 41.7 45 7.75x106 21.s 242.8 16.4 1.63x107 7.4 115.5 34.5 65 6.96x106 24.3 270.4 14.8 1.39x107 8.6 135.2- 29.5 100 6.14x106 27 5 306.5 13.0 1.15x107 10.5 164.2 24.3 150 5.27x106 32.0 357.1 11.2 8.58x106 13.6 212.6 18.8 200 4.81x106 35.1 391.2 10.2 7.47x106 16.1 251.9 15.8 250 4.54x106 37.2 414.5 9.6 6.66x106 18.1 282.6 14.1 300 4.39x106 38.4 428.3 9.3 6.22x106 19.3 302.4 13.2 350 4.30x106 39.2 437.5 9.1' 5.94x106 20.2 316.6 12.6 365 4.29x106 39.3 438.6 9.1 5.91x106 20.4 318.4 12.5 Note: Q-decay includes 3.5x106 BTU /hr from 20 years of accumulated discharges.

4-25

t Table 4.6.4 Nominal HEP Model Estimates For Failure to Diagnose Loss of Cooling l 1

l Following 1/3 Core Following Full Core Discharge Discharge Time Median Time Median Minutes Joint Error Minutes Joint Error HEP Factor EEP Factor Failure to Diagnosis Loss of Cooling Before Boiling occurs 600 0.00002 30 180 0.00005 30 i

Failure to Recovery i Before IIigh Radiation Field 1*a SFP Area >1440 0.00001 30 1440 0.00001 30 Note: The HEP probabilities are based on the nominal HEP model (Ref. 32).

Previous estimates of failure to recovery (NUREG-0933, for example) have used a REP value of 0.05 considering the high temperature and high radiation fields following loss of water from the spent fuel pool, and considering the location of the spent fuel pool in a BWR. Make-up was assumed to be from a fire hose.

Table 4.6.5 Typical Failure Rates and Repair Times for Cooling System Components

('faken from EPRI NP-3365)

Component Failure Rate Range Average Repair (per hour) (per hour) Time (hours)

Piping (per 10 ft section) 3x10-10 1xio-11 to 3x10-8 30 Pump 1x10-5 3x106 to 3x10-5 4o Heat Exchanger 3x10-6 1xio-7 to 1x2 0-4 30 valves (per valve) 1x10-6 3xio-7 te 3x10-6 24 Instrumer.tation (per channel) 1x10-6 3xio-7 t.O 1x10-5 6 4-26

4 v

Table 4.6.6 Failure Frequency of Generic SFP Cooling and Make-Up Systems Without Recovery System Cooling Cooling Make-up Make-up Frequency Train 1 Train 2 Train 1 Train 2 of Heatup (per R-y) (per demand) (per demand) (per demand) (per R-y)

A/B 0.1 0.05 0.015 0.05 3.8x10-6 C/D 0.1 0.3 0.015 0.05 2.2x10-5 The probability of the complete loss of the cooling and the make-up systems resulting in spent fuel damage is dependent on actions taken by the plant operators to either restore the cooling and/or make-up system or provide an alternative method for make-up or cooling, for example use of the stanon fire fighnng system or use of a portable backup pump. In NUREG-0933, a conditional failure probability of 0.05 was assigned to the operator failure to accomplish recovery actions. The resulting frequency of fuel damage was esumated to be 1.9x10-7 per reac-tor year for System B (System A with recovery) and 1.1x104per reactor year for System D (System C with recovery). The 0.05 conditional failure probability of recovery before fuel damage was based on the assumption that the spent fuel pool area environmental conditions would eventually become severe enough to make it difficult to setup the fire hose for make-up.

The NUREG-0933 (Ref. 27) evaluation did not consider the time available to diagnose and take corrective or recovery actions.

Based on the time available to diagnose a loss of cooling event, at least three hours before boil-ing occurs, the HEP value for failure to diagnose is 0.0015 (median value for full core discharge with error factor). For the 1/3 core discharge case, these values are lower by a factor of about two. The probability of a pool heatup event resulting in boiling of the water in the spent fuel pool is therefore estimated to be 5.7x10-9 per reactor year for System A (3.8x104 x 1.5x10-3) and 3.5x10-8 per reactor year for System C (2.2x10-5 x 1.5x10-3).

The spent fuel pool cooling and make-up systems are designed to meet one of two basic sets of requirements. The first basis requires the cooling portion of the system to be designed to seis-mic Category I, Quality Group C requirements. The second basis allows a non-seismic Category I, Quality Group C spent fuel pool cooling system provided that the following systems are designed to seismic Category I requirements and are protected against tornadoes: the fuel 1

pool make-up water system and its sources; and, the fuel pool building and its ventilation and i

filtration system. The make-up, ventilation and fibration systems must also withstand a single active failure.

Since Generic Issue 82 is concerned with risk from beyond design basis events, LLNL evaluated the probability of a beyond design basis seismic event resniting in a loss of cooling event and

~

subsequent pool heatup transient (Ref.17).

4-27

The cooling and make-up water systems for two pools were reviewed, one for a BWR system and one for a PWR system. Event and fault tress were constructed by LLNL to identify the ac-cident sequences that result from failure of these systems. For components appearing in these accident sequences, seismic fragilities were estimated based on design information and plant walkdowns. Boolean equations developed from the fault tree analysis were quantified using the seismic fragilities and preliminary hazard curves for the two sites. The dominant components to spent fuel pool system failures were found to be similar to components that have been found to contribute to seismic risk in several PRAs concerned with reactor core damage; poorly anchored electrical equipment and tanks. The components which contribute significantly to the seismic induced failure of the spent fuel pool support systems are the non-safety electrical systems in the plant; the motor control centers, switchgear and station service transformers which have rela-tivelylow seismic capacities.

The estimated mean probability of a beyond design basis seismic event resulting in loss of cool-ing, combined with operator failure to properly align normal make-up, was found to be on the order of 1.5x104 per reactor year for both systems. A sensitivity study without operator failure rates or random failures showed little change in the mean probability estimate (Ref.17). The median (50-th percentile) probability estimate was found to be about an order of magnitude lower than the mean value. For the seismic induced loss of cooling event to result in damage to spent fuel, the previously determined nominal HEP (with error factor) failure to diagnosis and failme of the second make-up train are used. Using the values of 0.0015 for failure to diagnose and 0.05 for failure of the second make-up train to recover, the estimated probability of fuel damage is therefore 1.5x104 x 0.0015 x 0.05, or 2.1x10-8 per reactor year. Based on the es-timated failure frequencies used in Table 4.6.6 above, it would appear that common mode failures in the non-safety electrical systems resulting from a beyond design basis seismic event, at least for the two plants studied, may have been underestimated by a factor of two for System A (0.1 x 0.05 x 0.015, or 7.5x10 4per reactor year without credit for make-up train 2 as com-pamd to 1.5x10 4 per reactor year). The combined best estimate probability of spent fuel damage from loss of cooling, from component failures and beyond design basis seismic events, is therefore estimated to be 4x10-8 plus 2x10-8 or 6x10-8 per reactor year. Previous estimates, based on a conditional probability of failure to recover from a loss of cooling event value of 0.05 without consideration for event diagnosis and using the systems data in Table 4.6.6, provide an 4

upper bound probability estimate of 1.4x10 per reactor year.

4.6.7 Structural Failure of SFP from Beyond Design Basis Earthquakes The probability of failure of a structure is related to the functional relationships between the various physical parameters and the variabilities in the parameters themselves. Two types of variability are considered. The first variability is that which is potentially reducible and is called the uncertainty. The additional component, which cannot be practically reduced,is called randomness.

The seismic analysis of a structure incittdes two parts. The first is the stuctural capacity and the response of the structure to a seismic event. This is referred to as fragility. The second is the seismic input and site response :o an earthquake, the seismic hazard analysis. 'Ihc hazard analysis if comprised of two parts. The site response (peak ground acceleration, response spectra, and frequcacy for both horizontal and vertical ground motions) and the annual fre-quency of exceeding a given peak ground acceleration, the seismic hazard curve. Because of uncertainty, both the fragility and seismic hazard analyses are described by families of curves.

Each curve representing different confidence levels. The resulting combination of these two 4-28

• l

. .. j l

sets of curves in the PRA can yield large differences in the estimated probability of structure failure dependent on the confidence level. While the seismic hazard curves and the parameters used to define the fragility are developed and expressed in terms of median values, the mean value is used in PRA applications. 1 The NRC Seismic Margins Program has developed an additional measure of importance for as-sessing seismic risk. This measure is the high confidence oflow probability of failure (HCLPF),

defined in terms of the peak ground acceleration (g value). This value is derived from the fragility analysis and is independent of the site seismic hazard curves, the annual frequency of j exceedance of a given g value. The HCLPF value is defined as the peak ground acceleration at which there is a 95% confidence that failure will not occur. This value may be compared to the safe shutdown earthquake peak ground acceleration used in the deterministic analysis to deter-mine the margin in excess of the SSE for which no structural failure is anticipated.

A compreher.sive assessment of uncertainty and conservatism in the seismic analysis and design of nuclear facilities has been prepared by the American Society of Civil Engineers (ASCE) (Ref.

29). The objectives of this study were to:

1. Identify sources of uncertainties present in seismic analysis and design,

2. quantify uncertainties, when possible, and recommend actions where data are missing, and

3. Identify the status of current analysis and design methods relative to the scatter of data for known sources of uncertainty.

The current practices employed to determine the seismic input and site response produce conser-vative (e.g.,84th percentile or greater), not median values, of the design seismic input. Empiri-cal pmcedures for structural design provide an additional margin of safety across the entire design response spectrum in earthquake-resistant designs. The ASCE (Ref. 29) concluded that a nuclear power plant having a design seismic input value of 0.25 g may actually be able to withstand much larger values of peak ground acceleration.

Some siting procedures differ in the western United State because there is a strong ground mo-tion data base available. Less extrapolation is required and tectonic faults and structures are also much easier to identify. The design seismic input is still conservatively evaluated in comparison to historical data.

The ASCE estimated the median factors of safety and logarithmic standard deviations associated with the safe shutdown earthquake (SSE), based on post 1973 seismic design methods. These factors are provide in Table 4.6.7 (from Table 9.1, Ref. 29).

The probability of gross stnictural failure of dre spent fuel from a beyond design basis earthquake was estimated by LLNL (Ref. l') fo a typical, although older, elevated llWR spent fuel pool ar.d a typical, older PWR spent fuel pool.

4-29

r-

. - .. l l In analyzing the failure of the spent fuel pool structures and systems, LLNL considered the fol-lowmg.

(a) Loss of liner integrity precipitated by structural failure of the spent fuel pool.

(b) Loss of function of the fuel pool support systems (e.g. pool cooling and make-up water capacity) msulting in loss of water through boil-off or drainage.

j (c) Damage to fuel racks caused by fuel rack motion.

The failure modes of the spent fuel pool structures were determined by LLNL. For the BWR Mark I and Mark II elevated spent fuel pools, the failure mechanisms which need to be con-sidered are:

1. The failure mode of the pool floor is that of a slab fixed at the four edges. The

' girders supporting the pool are in reality long walls acting as deep girders and are supported by the peripheral walls of the reactor building.

2. Compressive and shear stresses at the reaction points of the girders (onto the reactor building walls) for transmitting vertical and horizontal seismic loads from the storage pool to the foundation needs to be considered for stmetural adequacy.

3. Due to large concentrated loads (50 to 70 kips) at each foot of the storage rack, bearing and punching shear stress in the pool floor should also be investigated.

4. For laterally braced high density fuel storage racks, large concentrated loads are transmitted to the pool wall at either the base level or at the base level and the up-per seismic bracing level. The effect of concentrated load needs to be investigated.

Although thermal loads are important in the design of the spent fuel pools, their influence on the fragility of the pool is judged not significant by LLNL because the thermal loads are self-relieving.

For BWR Mark III and PWR storage pools, which are general on or below grade, the failure modes for the pool floor are:

1. Punching shear stress due to concentrated loads at the foot of the storage racks.

2. Foundation settlements for soil; soil settlement may only be an issue for piping relative displacements.

3. Failure modes for the walls are similar to that described for the BWR Mark I and II designs.

Possible failure modes for the liner plate identifled by LLNL are:

1. Tearing of the liner plate or seam welds at the leak channels due to vertical / horizontal loads from fuel storage racks; this is of concern only if the rack slides and the foot bears on the leak channel.

4-30

2. Tearing of the liner plate due to sliding of the rack over any floor depression or wrinkles in the liner plate.

3. For a laterally braced rack, puncturing of the liner plate at the knuckle in the vicinity of the pool floor / walls intersection.

The failure modes of free standing or sliding racks, and for laterally braced high density fuel racks were considered by LLNL. Based on information provided to the NRC staff conceming a reracking amendment by a licensee (Ref. 30), LLNL concluded that the peak ground accelera-tion would have to exceed 1.5 to 2.0 g before failure of the free standing racks would occur.

The median acceleration capacity of the racks for incipient impact with the pool wall is es-timated to be 1.0 g, and it would requim 1.5 to 2.0 times this value to cause impact and damage.

Even then, the fuel rack design is such that the assembly cannot be compressed into a critical mass. LLNL therefore concluded that crushing of fuel and assemblies is not a credible failure mode of the spent fuel pool system.

The actual potential failure modes of the BWR spent fuel pool studied by LLNL included the following:

(a) Out of plane shear failure of the pool floor slab.

(b) Out of plane bending failure of the pool floor slab.

(c) Punching shear failure of the pool floor slab under the fuel rack support pad.

(d) Out of plane bending failure of the south pool wall.

(e) Bending and shear failure of the girder under the south wall.

(f) Bending and shear failure of the girder under the east pool wall.

(g) Overall transfer of N-S and E-W inertial loads to the reactor building.

The controlling failure mode with the lowest seismic capacity was determined to be the out of plane shear failum of the pool floor slab. The slab was evaluated for out of plane loading result-mg from dead weight load plus seismic load. Sources of dead loads am the weights of the dab, grout, water, fuel racks, and attached equipment. Sources of seismic loads am (a) vertical seis-mic response of the slab and attached masses, (b) fluid impulse and convective mode responses induced by horizontal seismic excitation, and (c) horizontal seismic msponse of the spent fuel racks.

The PWR spent fuel has two features which are not typical. The storage racks are both low den-sity and high density in design. And to accommodate the region of high density racks, a support columu was added benesth the spent fuel pool floor,in the waste gcs holdup tank room below the rpent fuel structure. The pool fbor is actaa!!y six feet above grade at Rchinson.

)

4-31

. .. ]

The actual potential failure modes of the PWR spent fuel pool studied by LLNL included the following:

(a) Out of plane bending of the East or South wall.

(b) Out of plane shear failure of the East or South wall.

(c) Out of plane shear failure of the pool floor slab.

(d) Out of plane bending failure of the pool floor slab.

(e) Overall seismic stability of the spent fuel.

The out of plane bending failure of the east wall was determined to be the failure mode with the lowest seismic capacity. The East or South wall resists the lateral forces of the old fuel racks. It was modeled by LLNL as a slab fixed on thme sides and fme on top. The loads considered in the evaluation of the seismic capacity of this wall are (a) hydrostatic loads (normal water level),

(b) hydrodynamic loads induced by horizontal and vertical accelerations in earthquakes, (c) wall inertia force, and (d) reaction forces from the old fuel racks.

The fragility of a structure, or component, is expmssed in tenns of its median factors, A,, r, and %o. Amis the median ground acceleration at which the probability of faihue is 0.5. p and p, are the random variability and the uncertainty in the median capacity based on a logno,rm model. pr and are the logarithmic standard deviations of the median value. Using the lognor-mal model a hig51 confidence of low probability of failure (HCLPF) capacity factor is defined, HCLPF capacity = Am exp( -1.64 ( + ). The HCLPF value is defined as the peak ground acceleration at whichlhem is a 95% , cony)idence that failure will not occur. This value compared to the safe shutdown earthquake peak ground acceleration used in the deterministic analysis to determine the margin in excess of the SSE for which no structural failure is an-ticipated. This fragility model and development of fragility parameters have been utilized in over 25 seismic PRAs.

The median factors of safety and variabilities of the spent fuel pool structure for the elevated 0.39. The HCLPF value is 0.5 g.

BWR were found to be Am = 1.4 g, with For the PWR spent fuel pool, Am = 2.0 g, with r. = 0.28 r = 0.26 and "a=nd u = 0.40. The HCLPF v 0.65 g. The HCLPF value shows a design margm of a factor of at least three over the SSE design peak ground acceleration for either pool. Typical SSE design values for LWRs are in the 0.15 to 0.2 g range. The BWR used in this study has an SSE value of 0.14 g, and the PWR has an SSE of 0.2 g.

Preliminary seismic hazard curves for the two sites were used to estimate the probability of failure of the spent feel pool structures from beyond design basis earthquakes. These prelimi-nary curves have not been finally reviewed by the NRC and were used by LLNL only tc obtain a better understanding of the seismic mduced spent fuel pool failure. The hazard carves may change after NRC review and guidance for their proper use will be devehiped. They r.m however expected to be a reasonable representation of the seismic characteristics of tbc sites. A recently published report (NUREG/CR-5042, Supplement 1, Ref. 31) cotapares the NRC and EPRI preliminary estima*cs for the annual probabibty of exceedance at the SSE carthquake les el 4-32 4

for nine mactor sites. The differences between the NRC and EPRI estimates am reasonable, with no gmater than about one onier of magnitude difference between the estimates at any con-fidence level. The median (50%),85% and 15% confidence levels wem compared.

The resulting annual sc.ismic failure frequencies for the two pools are provided in Table 4.6.8.

The results are provided for a variety of confidence levels and for a variety of cutoff values.

The use of a cutoff value ofless than 100% demonstrates the sensitivity of the analysis to the ex-treme tails in the seismic hazard curves and fragility curves. LLNL recommends use of the 99%

cutoff value based on their experiences with seismic PRAs (Ref.17). The mean failum fre-quency at the 99% cutoff value is used for this Regulatory Analysis.

The LLNL study used two representative spent fuel pools. These pools have been designed to the seismic design criteria existing in the late 1960s. Their large seismic capacities lead LLNL to conclude that the pools designed to current seismic standards (post 1973) should have higher seismic capacities and should not contribute significantly to seismic risk. Based on the demonstrated relatively high seismic capacity (the HCLPF capacity of the pool structums are es-timated to be more than thme times the SSE value), LLNL also concluded that the risk contribu-tion from spent fuel pool structural failures is negligibly small (Ref.17). In addition, the results obtained for the two pools studied also fall within the margins estimated by the ASCE for nuclear power plant seismic structures. The mr.rgins, based on median capacities, are about 8 for the BWR and 10 for the PWR which fall within the estimated structural factor range of 4 to 19 (see Table 4.6.8).

l 4-33

i l

Table 4.6.7 Estimated Median Factors of Safety and Logarithmic Standard Deviations Associated with The Safe Shutdown Earthquake (SSE)

(Post 1973 Seismic Design Methods)

Item Median Factor Logarithmic of Safety Stad. Deviation Structures capacity Ultimate Strength vs Code Allowable 1.2 - 2.5 0.16 - 0.20 Inelastic Energy Abso ption Capacity 1.8 - 4.0 0.20 - 0.30 Total Capacity ractor (1) 2.5 - 6.0 0.28 - 0.34

Response

Design Response Spectra 1.2 - 1.6 0.25 - 0.40 Damping Effects 1.2 - 1.4 0.09 - 0.20 Modeling Effects 1.0 0.10 - 0.20 Modal and Component Combination 1.0 - 1.1 0.15 - 0.20 Soil-Structure Interaction 1.1 - 1.5 0.10 - 0.40 Total Pasponse Factor (1) 1.6 - 3.2 0.40 - 0.59 Total Structural Factor (1) 4 - 19 0.52 - 0.65 Mechanical Equipment Capacity Factor 1.5 - 8.0 0.28 - 0.34 Building Response Fa A.Jr 1.6 - 3.2 0.40 - 0.59 Floor Spectra Factor 1.4 - 1.6 0.25 - 0.35 Total Mechanical Equipment Factor 3.5 - 40 0.59 - 0.72 Notes:

(1) These total factors are the product of the preceding individual factors upon which they are based. However, a range is shown for each of the in-dividual factors and it is not reasonable to assume all the individual factors would concurrently be at either their lowest or highest values. Judgment was introduced in establishing the expected range of these factors. For example, the estimated range on the structural median capacity factors of 2.5 to 6.0 is less than would be obtained from concurrently using either the lowest or highest values of the strength and energy absorption which would produce a median capacity factor range of 2.2 to 10.0.

4-34

v .;

t I ..

L .

. ..- )

, Table 4.6.8.

Annual Seimnic Failure Frequencies For Two Representative Spent Fuel Pools Fragility Data Cutoff Failure Frequencies Mean 5% 50% 954 Pool Am hr hu RCLPF Value 1/R-y 1/R-y (g) (-) -(-) (g) (4) 1/R-y_ 1/R-y BNR 1.4 0.26' O.39 0.5 100 3.8x10-5 1.8x10-ll 7.7x10-8 3.6x10-5 99 6.7x10-6 3.1x10-ll 8.3x10-8 1.9x10-5 97 3.8x10-6 3.1x10-ll 7.7x10-8 1.4x10-5 PWR 2.0 0.28 0.40 0.65 100 8.6x10-6 6.1x10-12 1.3x10-8 8.6x10-6 99 1.8x10-6 g,ggio-12 1.5x10-8 5.0x10-6 97 ' 9.9x10-7 9.5x10-12 1,4 ,10-8 3.5x10-6 l

4-35

4.7 Summstry of Accident Sequence Quantification The frequency of spent fuel damage resulting from accident sequences which can result in the loss of water from the spent fuel pool through either drainage or through boiling as a result of loss of cooling are summarized in Table 4.7.1. HEP failure to diagnose values, including error factors, based on the nominal HEP model (Ref. 23) have been used to develop the best estimate accident frequencies and are based on the most limiting condition in the spent fuel pool - a full core discharge into a pool contammg 20 years of spent fuel. The upper bound frequency values represent previous estimates from WASH-1400, NUREG-0933 and NUREG/CR-4982. In general, these previous studies did not consider the time available for recovery actions but, in general, used intentionally conservative assurnptions regarding operator performance.

Table 4.7.1 Summary of SFP Accident Frequencies Accident Sequence PWR Frequency BWR Frequency Best Estimate Upper Bound Best Estimate Upper Bound (per R-year) (per R-year) (per R-year) (per R-year)

Structural Failures 1.0x10-7

1. Missiles 1.0x10-8 1.0x10-7 1.0x10-8

2. Aircraft crashes 6.0x10-9 2.0x10-8 6.0x10-9 2.0x10-8

3. Heavy Load Drop 3.1x10-8 3,1x10-7 3.1x10-8 3.1x10-7 Pnaumatic Seal Faf'aree 3.0x10-8 5.0x10-7 3. 0x10-8 (1) 5. 0x10-7 (1) 1.2x10-8 1,oxio-7 1.2x10-8 1,oxio-7 Inadvertent Draint a 1.4x10-6 Loss of Cooling /Make-up 6. 0x10-8 (2) 1.4x10-6 6. 0x10-8 (2)

Total 1.5x10-7 2.4x10-6 1.5x10-7 2.4x10-6 Seismic Structural Failure 1.8x10-6 6.7x10-6 Conditional Probability of Zircaloy Cladding Fire Given Loss of Water (High Density Storage Racks) 1.0 0.25 Notes: (1) BWRs do not, in general, use pneumatic refueling cavity seals, but other pneumatic seals are used in the transfer canal.

(2) Includes beyond design be.ois seismic indaced loss of cooling and make-up.

4-36

4.8 Radiological Consequences Evaluation The inventory of radionuclides contained in spent fuel assemblies depends on the operating his-tory and the size of the plant. During refueling, the freshly discharged fuel contains a large in-ventory of isotopes with short half-lives in the range of approximately one to thirty days. These isotopes decay over the course of a year, until the next refueling outage.

The older fuel contains radionuclides which have longer half-lives. The older fuel approaches a decay rate which is inversely proportion to the decay time. For example, after four years, the spent fuel contains approximately one-forth of the specific activity of one-year old fuel.

During each refueling outage approximately one-third of a PWR core and about one-forth of a BWR core are off-loaded to the spent fuel storage pool. It is noted that releases for an accident involving the reactor core are basically noble gases and halogens, while for a spent fuel storage pool accident the releases are primarily alkali metals (such as cesium, Cs) and alkali earths (such as strontium, Sr). Therefore,it may not be appropriate to directly relate the probability of a spent fuel storage pool accident to a core damage accident because of the different radionuclides involved. .

4.8.1 RadionuclideInventories The ORIGEN2 computer program (Ref.12) was used by BNL to determine the radionuclides in-ventory of the spent fuel as a function of decay time. Separate inventories were calculated for activation products in the fuel assembly hardware and cladding, and for the fissions products and actinides sealed in the fuel elements. The data was obtained for a reference BWR and a reference PWR. Millstone Unit I and R. E. Ginna were selected by BNL as the reference plants for the source term evaluation. A comparison of the radionuclides inventory at different decay times, up to the time when the spent fuel storage pool reaches a capacity load, to the equilibrium inventory of a reactor core is provided in NUREG/CR-4982 (Ref.10).

4.8.2 Radionuclides Potentially Available for Release The source term for any postulated accident sequence is defined in terms of:

- the amount (curies) of each radionuclides,

- the composition, physical and chemical form of each radionuclides, and,

- the time and the duration of the release of the radioactivity to the environment.

The physical snd chemical processes that would take place in a drained spent fuel storage pool are not well characterized at the present time. It was therefore necessary for BNL to use en-gineering judgment to estimate the source term. The SFUELlW computer program does not ac-count for rek> cation of the resetion prodacts (molten tm-oxidized cladding, fuel dissolved in molten zirconium, etc). Also the degree to which exposed UOy would oxidize to U 3Og and reduce the release of less volatile fission products has not been ' studied. The estimate of the fraction of each radionuclides release was determined by BNL based on available data and on engineering judgroent, and is provided in Table 4.8.1 (from NUREG/CR-4982, Table 4.2, Ref.

10).

4-37

h

. ~ .. ,

I l-1 Table 4.8.1 l Estimated Radionuclides Release Fractions During a Spent Fuel Pool Accident Resulting in Complete Destruction of the Fuel Cladding Release Fractions (1) r' ham 4 cal Family Elemer.t or Value Uncertainty Isotope Used Range Kr, Xe . 1.0 0 Noble gases Halogens I-129, I-131 1.0 0.5 - 1.0 Alkali metals Cs, (Ba-137m) Rb 1.0 0.1 - 1.0 Chalcogons Te, (I-132) 0.02 .002 - 0.2 Alkali earths Sr, (Y-90), Ba (in. fuel) 2x10-3 1x10 1x10-2 Sr, Y-91 (in cladding) 1.0 0.5 - 1.0 Transition co-58 (assembly hardware) 0.1 0.1 - 1.0 Elements Co-60 (assembly hardware) (2) 0.12 0.1 - 1.0 Y-91 (assembly hardware) 0.1 0.1 - 1.0 Mb-95, Er-95 (in fuel) 0.01 1x10 1x10-1 Mb-95, Er-95 (in cladding) 1.0 0.5 - 1.0 Miscellaneous Mo-99 1x10-6 1,10-8 _ 1xio-5 Ru-106 2x10-5 1x10-6 1x10~4 Sb-125 1.0 0.5 - 1.0 1.anthanides La, Ce, Pr, Nd, Sm, Eu 1x10-6 1x10-'8 - 1x10-5 Transuranic Np, Pu , Am , Cm 1x10-6 1x10 1x10-5 Notes: (1) Reler.se fractions of several daughter isotopes are dete=4nad by their precursors, e.g. , Y-90 by Sr-90, Tc-99m by Mo-99, Rh-106 by Ru-106, I-132 by Te-132, Ba-137m by Cs-137, and La- by Ba-140.

(2) Release fraction adjusted to account for 100% release of the small asount of Co-60 contained in the Eircaloy cladding.

4 38

4 4.8.3 Estimated Releases and Consequences For Accident Categories The dose equivalent of the release estimates depends on many factors including the location of the spent fuel storage pool and equipment operability (for example, with and without filters in .

l the fuel storage structure). Cesium, for example, is expected to be released as an aerosol and fil-ters may pmvide an effective removal mechanism. If the fuel storage building structure cracks or if fans fail to function due a seismic event, the release may be substantial. The predicted release to the environment was estimated for each of several accident categories.

The radionuclides inventories for both the BWR and the PWR spent fuel pools were calculated using the ORIGEN2 computer program and the actual operating and discharge histories for a BWR and a PWR. For both plants, the noble gases and halogens in the spent fuel inventory are a small fraction of the inventory in an equilibrium core at shutdown, except for the freshly dis-charged fuel. The cesium and stmntium inventories are more than three times the equilibrium inventory, as a result of the large inventories of spent fuel in the pool (the calculation were based on 11 fuel cycles for the BWR and 16 fuel cycles for the PWR).

A re-evaluation of the cladding fire propagation estimate (given the complete loss of water from the spent fuel pool) indicates that, with the use of high density storage racks, there is a substan-tial likelihood of propagation to adjacent fuel bundles that have been discharged within the last one or two years. Subsequent propagation to even lower power bundles by thermal radiation is highly unlikely, but with a substantial amount of fuel and cladding debris on the pool floor, the coolability of even these lower power bundles is uncertain.

The fission product release fractions were calculated for two limiting cases in which a Zircaloy cladding fire is assumed to occur. In the first case (1) the cladding combustion is assumed to propagate throughout the entire pool and the entire inventory is involved. In the second case (2) the inventory is limited to only the most recently discharged fuel batch. Parametric studies were performed by BNL to evaluate the consequences as a function of the time of the postulated acci-dent.

In calculating the consequences of a spent fuel pool accident, BNL has assumed no credit for the ventilation and filtration systems in the fuel storage building. While these systems are designed to mitigate the consequences of a fuel handling accident, the design of these systems does not consider the high temperature conditions of a Zircaloy cladding fire (in excess of 2000 0F). Fis-sion product retention under these conditions is questionable. A sensitivity study with a decon-tamination factor of 10 was performed by BNL (Ref.10), to demonstrate the possible affect of fission product retention on structures.

For a Zircaloy cladding fire to occur, the fuel must be recently discharged (between 30 and 180 days in a cylindrical BWR configuration, and between 30 and 250 days in a cylindrical PWR configuration). Since the spent fuelis stored in high density racks, the probability for a Zircaloy f~ue in a PWR is assumed to be LG, given a complete loss of water. For a BWR, which uses direedonal racks, the pmbability of a Zircaby fire is assumed to be 0.25, given a ccmplete loss of water. This value is selected from the full core discharge calculation performed by Sandia (Ref. 5) with the channel box attached (see Tabic 4.5.1, above), and is also representative of the average values for the cylindrical BWR configuration with different inlet orifice sizes.

4-39

For a less severe accident in which the fuel is exposed to air but does not reach temperatures high enough to ignite the Zircaloy cladding, fuel pm failure could occur resulting in a release of the noble gases and halogens. Two cases have been considered by BNL. In the first case the entire poolis assumed to be drained but the decay period is one year since the last discharge and 50% of the pins are assumed to perforate or rupture. In the second case it is assumed that only part of the fuelis uncovered 30 days after the last discharge and 111 the rods fail. The conse-quences of either of these two cases are small and result from the re4 case of noble gases.

Sensitivity analyses of the offsite consequences for the Zircaloy fire caser were calculated with the CRAC2 computer program by BNL to study the affect of the source term assumptions on the population dose and interdiction area (Ref.13). The results of this study are provided in Table 4.8.2, along with the results for a case which represents cladding rupture only. The following assumptions were used by BNL (Ref.10):

- a generalized site surrounded by a constant population density of 100 persons per square mile within a 50 mile radius; i

- generalized meteorology (a uniform wind rose, average weather conditions);

- the population in the affected zones are relocated after 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, persons expected to receive more than 25 rem from ground shine in seven days.

There are several unusual characteristics of a severe accident that cause somewhat unexpected 1 results in the radiation exposure calculations. The radiation exposure is relatively insensitive to fairly large variations in the estimated release. This is due principally to the health physics as-sumptions within CRAC2. This has also been seen in calculation related to fission product releases from core damage accidents. The CRAC2 code assumes that decontamination will limit the exposure of each person to 25 rem. For the long lived isotopes (predominately cesium), the exposure is dce mainly to exposure after the area is decontaminated and people retum to their home. Thus, for this type of release the long term whole body dose is limited by the population in the affected sectors (about 0.8 million people in three of the 16 sectors downwind of the release within a 50 mile radius) or about 3,000,000 person-rem (Table 4.8.2).

Another measure of the consequences of a spent fuel pool accident is the interdiction area (the area with such c high level of radiation that it is assumed that it cannot be decontammated).

The interdiction area is sensitive to the source term as shown in Table 4.8.2.

Additional consequences calculations were performed by BNL (Ref.13) using the MACCS computer program and are provided in Table 4.8.3. The MACCS computer program models are described in NUREG-ll50, Appendix 0 (Ref.14). The source term data developed by BNL in NUREG/CR-4982 (Ref.10) was used for these new calculations. These calculations were per-formed for tne value impact studies in Section 5 based on the a site population density of 340 people per square mile, the mean population density around nuclear power plant sites projected for the year 2000 (Ref.15), Case 1. In addition, the offsite property damage cost for a spent fuel

' racol accident was also calculated with MACCS for use in the value impact studies in Section 5.

'A worst case analysis was also performed, assuming the entire spent fuel pool inventory is released at a high population site (Zion, Illinois, 860 people per square mile population density), l l

Case 2.

I 4-40

1 as A direct comparison of the consequences of a sevem accident in a spent fuel storage pool to the consequences of a severe com accident can be misleading. For the spent fuel pool accident, there are no "early" fatalities and the risk of early injury is negligible. For a severe com damage accident, early fr.alities and early injury am part of the risk due to the pmsence of the shorter livedisotopes. ,

I Table 4.8.2 offsite Consequences of Spent Fuel Pool Accidents - CRAC2 Results Case Description Whole Body Dose Interdiction Area (Person-Rem (square miles) per-Event) 1A. Total inventory 30 days after 2.6x106 224 last discharge 1B. Total inventory 90 days after 2.6x106 215 last discharge 2A. Last discharge 90 days after 2.3x106 44 last discharge 2B. Last discharge 90 days after 1.1x106 4 last discharge, DF of 10 reduction

3. 50% of all fuel rods leak 4.0 0 1 year after last discharge Note: Sensitivity study based on population density of 100 people per square mile within a 50 mile radius of the site, from Reference 10.

Table 4.8.3 offsite Consequences of Spent Fuel Pool Accidents - MACCS Results Case Description Whole Body Dose offsite Property (person-Rem Damage per-Event) (1983 $s)

1. Best Estimate Consequences Last discharge (1/3 core) 8.0x106 3,4xio9 20 days after discharge 50 mile radius Based on 340 people / square mile

2. Woxst Case Estimate Consequences Total inventory 30 days after 2. 6x107 2.6x1010 discharge, 50 mile indius Based on 860 paople/squara raile LA1

~

4.8.4 Summary Conclusions On Fuel Damage and Consequences The conditional probability of a Zircaloy cladding fire, given a complete loss of water from a spent fuel pool, is estimated to 1.0 for PWRs with high density storage racks and 0.25 for BWRs with directional storage racks (wiU the channel box in place in the assembly).

The propagation of the fire to older stored spent fuel assemblies is predicted to occur for spent fuel that has been stored less than two years, under some conditions. Propagation can occur as a result of radiative heat transfer from the hottest fuel assemblies to the older assemblies if the decay heat level of the older assemblies is sufficient to heat the cladding to within 100 to 200 oC (200 to 400 0F) of the self-sustammg oxidation temperature of 900 oC (1650 oF).

The best estimate of the consequences of a spent fuel pool accident which results in fuel damage is 8.0x106person-rem with an offsite property damage estimate of $3,400 million (1983 $s).

The best estimate is based on a populan,on density of 340 people per square mile within a 50 mile radius from the site and is a result of the release of radionuclides from the last fuel dis-charge (1/3 of a reactor core),90 days after being discharged. Although propagation of a Zir-caloy cladding fire to one-to-two year old fuel by thermal radiation can occur, the older fuel would have to be next to the hottest assemblies. Subsequent propagation to even lower power assemblies by thermal radiation is highly unlikely, but with a substantial amount of fuel and cladding debris on the pool floor, the coolability of even these lower power bundles is uncertain.

A worst case estimate of the consequences is based on a population density of 860 people per square mile within a 50 mile radius from the site and is a result of the release of radionuclides from the entire pool inventory, with the last fuel discharge being 30 days old. The consequences are estimated to be 2.6x107 person rem with an offsite property damage estimate of $26,000 million (1983 $s).

4.9 Other Issues Concerning Use of High Density Storage Racks 4.9.1 Gaps In Neutron-Absorbing Materials Board Notification 87-011 (Board Notification Regarding Anomalies in Boraflex Neutron Ab-sorbing Material, June 15,1987, Ref. 32) and IE Information Notice 87-43 (Gaps in Neutron-Absorbing Material In High-Density Spent Fuel Storage Racks, September 8,1987, Ref. 33) have been issued by the NRC to alert licensees of anomalies in boraflex neutron absorbing material used in the construction of high density storage racks in the Quad Cities Unit 1 spent fuel pool. The gaps were inferred from anomalies in " blackness" testing results and confirmed by underwater neutron radiography. The material supplier and the Electric Power Research In-stitute (EPRI) have undertaken research programs to collect data and information on gap forma-tion, including the effects of rack fabrication methods and irradiation damage mechanisms.

Boraflex is also used in the Turkey Point spent fuel pool racxs and in the Point Beach spent fuel pools. At Point Beach, some deterioration of the samples inserts were noticed during surveil-lance testing.

These anomalies da not impact on the finding concerning loss of water from a spent fuel pool.

4-42

4.9.2 Potential For High Radiation Fields IE Information Notice 87-13 " Potential for High Radiation Fields Following Loss of water '

l From Fuel Pool," February 24,1987 (Ref. 34), was issued to alert licensees of the potential for high radiation fields following the inadvertent loss of water from the spent fuel pool or transfer canal. Following the seal leakage at Hatch, the licensee determined that potentially high radia-tion fields could exit in the spent fuel area as a result of irradiated control blades being stored in the spent fuel pool on short hanger. Some of the control blades could have been completely un-covered if the water level dropped to the bottom of the transfer canal.

In determining the frequency of loss of cooling events and in evaluating HEP diagnosis and recovery actions, the assumptions concerning loss of water from a spent fuel pool used conser-vative upper bound failure rates for failure to diagnose prior to pool boiling and therefore ad-dress this concern. In the highly unlikely situation of draining the spent fuel pool to the transfer canallevel, licensee responses to IEB 84-03 (Ref.19), concerning pneumatic seal failures, indi-cate that emergency procedures have been considered which would not require entry into the spent fuel pool area - the high radiation fields from the spent fuel alone would likely restrict ac-cess.

4.9.3 Refueling Cavity Seal Failure During Fuel Assembly Handling Generic Issue 137, titled " Refueling cavity Seal Failure," is considering the issue of spent fuel damage resulting from a reactor cavity seal failure during fuel assembly handling (Ref. 35). The risk of failure of a single fuel assembly and the potential risk to plant personnel, not the general public,is being addressed.

The likelihood of uncovering stoIed spent fuel as a result of a seal failure and the risk to the general public are addressed in this Regulatory Analysis.

l l

l l

l 4-43 l l

,o. - .

l

5. VALUE/LMPACT ANALYSIS

' 5.1~ Alternative 1 - No Action This altemative assumes that no additional action is necessary based on the evaluation of the current risk associated with the use of high density racks for the storage of spent fuelin spent fuel pools at LWRs. It is also assumed that all applicable requirements and guidance approved to date have been implemented.

In addition to implementing the requirements contained in 10 CFR Part 50 Appendix A of the

" General Design Criteria," and 10 CFR Part 20, concerning radiation doses being kept as low as is reasonably achievable, licensees should have implemented additional or corrective actions based on the following guidance:

1. IE Bulletin 84-03, " Refueling Cavity Water Seals," issued August 24, 1984.

(Ref.21)

2. IE Information Notice 84-93, " Potential for Loss of Water Fmm the Refueling Cavity," issued December 17,1984. (Ref. 24)

3. Generic Letter 85-11 " Completion of Phase II of ' Control of Heavy leads at Nuclear Power Plants' NUREG-0612," issued June 28,1985. (Ref. 4)

4. IE Information Notice 87-13. " Potential for High Radiation Fields Following

~

Loss of Water from Fuel Pool," issued February 24,1987. (Ref. 34)

5. IE Information Notice 87-43, " Gaps in Neutron-Absorbing Material in High-Density Spent Fuel Storage Racks," issued September 8,1987. (Ref. 33)

6. IE Information Notice 88-65, " Inadvertent Drainages of Spent Fuel Pools,"

issued August 18,1988. (Ref. 25)

7. IE Information Notice 88-92, " Potential For Spent Fuel Pool Draindown," issued November 22,1988. (Ref. 26)

No costs am usually attributed to a "No Action" alternative because the future cost of accidents are conventionally counted as benefits or averted costs in the assessment of the alternative actions. However, a spent fuel pool accident would msult in cleanup and repair costs. In addition, replacement power costs could occur during the cleanup and repair period. Reactor operations without a safe place to store spent fuel could not continue. If the accident also results in a large release of radioactivity offsite, the costs of relocating people, restricting food and water, cleanup of contamination, and health consequences would add to these costs.

Occupational exposum due to a spent fuel pool accident could also be considered on a monetary basis. BNL has evaluated these attributes (Ref.13). The following paragraphs summarize this assessment of risk and the cost associated with a repmsentative (base case) spent fuel pool accident.

5-1

5.1.1 Occupational Exposure (Accidental)

Exposure to plant personnel associated with post-accident cleanup of a major spent fuel pool accident is expected to be similar to those associated with the cleanup activities at TMI-2. For this accident, BNL estimates the occupational radiation dose from cleanup is less than 4,580 person-rem (Ref.13). Since the potential offsite dose impact (per accident) ranges from 8 to 26 l

million person-rem, further refinement of this estimate is not warranted.

5.1.2 Onsite Property Damage The spent fuel pool accident seqtience involves (1) failure of the pool due to seismic or load drop events resulting in the complete loss of water inventory, or loss of cooling resulting in boiling dry the pool, (2) Zircaloy fire initiation of recently discharged fuel and the potential

! propagation of the fut to older spent fuel assemblies stored in the pool, and (3) loss of l confmement since the spent fuel pool building is assumed to be breached as a result of a seismic

! event or as a consequence of the highly exothermic Zirraloy fire.

The consequences of these scenarios are expected to be similar to the Category II accident defined in NUREG/CR-3568 (Ref. 36),50% clad melting and contamination. For this case cleanup and decontamination are estimated to be approximately $192 million (1988 $s). Plant outage times were estimated by BNL based on the time estimates to license, construct and test a I replacement pool (Ref. 37), and range from five to seven years. '1he cost (1988 $s) for a replacement pool, $54 million for a 400 MTU pool, and the cost of permanent disposal of damaged fuel, $30 million, were also estimated from reference 37.

The cost of replacement power is approximated by (from NUREG/CR-3568):

C = (0.13

• R + 0.12)106$/MW-year where R is the fraction of replacement energy by oil fired or non-economical power purchases from a given National Electric Reliability Council (NERC) region. This formula includes credit l for the avoided variable fuel cycle cost of a shutdown reactor. An R value of 0.41 was used for the best estimate (national avenge) and a value of 0.9 (highest NERC region) was used for the worst case estimate for replacement power costs (based on 1981 $s). A plant capacity of 65% is factored into the above equation and a 1,000 MW(t) generic plant size was assumed.

The BNL best estimate replacement power cost (in 1983 $s) for a 1,000 Mw(t) plant for five years is $867 million. For the worst case estimate the seven year replacement power cost is

$1,660 million.

Using NUREG/CR-4568 (Ref. 38) data, the cost of replacement power (1984 $) for the national average cost of 0.026 $/Kw-hr is $740 million for the best estimate and, using a high hT.RC region cost of 0.035 $/Kw-hr, the worst case estimate is $1,400 million. Assuming a constant 5% inflation rate over a four year period (to 1988), the current values would be $900 million and

$1,700 million respectively. These values are used in this regulatory analysis, and indicated that replacement power costs are not sensitive to modeling assumptions (2% to 4% uncenainty in comparing BNL 1983 values to staff 1988 values).

l l

5-2 l

The onsite costs are calculated from (NUREG/CR-3568):

Von = N x Ar x von H, -V on - value of avoided onsite propeny damage N - number of affected facihunes AF - change in accident frequency and e -rti (Ce+Cr+C9) (11- e -r (tf-ti) ) (1. o-m)

'Un=

o a r2

-where:

Uon - present value of onsite propeny damage conditional upon release

-cleanup and decontaminauen costs

. - repair, replacement of spent fuel pool, disposal of damaged fuel

- replacement powe tf' average years remam,r costsing till end of reactorlife,30 years ti '- year plant stans operating, taken to be 0 years r . - discount rate m - years plantis out of service The onsite propeny damage costs per accident, Von, are summarized in Table 5.1.1.

5.1.3 Offsite Health and Property Damage The offsite health and propeny damage estimates were obtained by BNL from the MACCS computer program (Secnon 4.6, Table 4.6.2), and are summarized in Table 5.1.2.

The offsite costs are calculated using the NUREG/CR-3568 methodology and discounting the

' cost of the 30 year remaining life of a typical facility:

Vo gg = N x Ar x voeg where:

V ort- value of avoided offsite propeny damage N number of affected facilitities

-AF change in accident frequency and e (-rti) . o(-rtf)

Uo gg = Cogg x r

where:

U ort- present value of offsite property damage, conditional upon release Coff- offsite propeny damage cost tf - average years remaining till end of reactor life,30 years ti - year plant stans operating, taken to be 0 years r -discountrate The offsite damage costs per accident, Vort, are summari: ed in Table 5.1.2.

5-3

. ~ ..

Table 5.1.1 Onsite Property Damage Costa Per SFP Accident (1988 $s)

Item Units Best Estimate Worst Estimate Cleanup and $1,300,000 192 192 Decontamination Repair Pool and $1,000,000 84 84 Dispose of Fuei Replacement Power $1,000,000 900 1,700 Average Number of years 30 30 operating years remaining Years plant is out years 5 7 of service Present Value (Von) 1988 $s 17,600 27,000 At 10% r,iscount rate Present Value (Von) 1988 $s 32,400 51,800 At 5% Discount Rate Table 5.1.2 Offsite Health and Property Damage Estimates (1988 $s) Per SFP Accident Case Description Whole Body Dose Offsite Present Value (person-Rem Property (Voff) per-Event) Damage At 10% At 5%

($s) Discount Discount l ($s) ($s)

Best Estimate Consequences Last discharge 90 days after 8.0x106 4,oxto9 76,000 124,000 discharge, 50 mile radius Based on 340 people / square mile i Worst Case Estimate Consequences Total inventory 30 days after 2.6x107 3.0x1010 580,000 940,000 discharge, 50 mile radius l

Based on 860 people / square mile 1

1 5-4 l - .- _ - - - - - _ _

5.1.4 Potential Consequences and Cost of SFP Accidents The probability of a spent fuel pool accident which would result in spent fuel damage is estimated to be 1.5x10-' per reactor year, summed over all accident sequences except for the beyond design basis seismic failum of the spent fuel pool structure. The conditional probability of Zircaloy cladding fire is estimated to be 1.0 for a high density racked PWR and 0.25 for a BWR.

The seismic structural capacity of the spent fuel pool has been shown to have a median ground acceleration in the 1.4 to 2.0 g range. This is a factor of 10 above the typical SSE design values. t The high confidence low probability of failure value (0.5 to 0.65 g) shows a margin of safety of three over the SSE design value. That is, there is less than a 5% chance of failure at a confidence level of 95% for a peak ground acceleration three times the SSE design value. The estimated mean seismic probability of a Zircaloy fire in a PWR spent fuel pool is 1.8x10 4 per '

reactor year, and in an elevated BWR spent fuel pool the estimated value is 1.7x10 per reactor 4 year. The probability of a spent fuel pool accident which would result in spent fuel damage is therefom estimated to be on the order of 2x10-6 per reactor year, including the scismic hazard, for a typical LWR spent fuelpool.

For comparison purposes, in a review of the results of seismic core damage and large release frequencies from studies performed as part of USI A-45, " Decay Heat Removal Requurments,"

the poir.t estimates for seismic core damage frequencies ranged from 1x10-5 to 8.3x10-5 per reactor year and the seismic release frequencies ranged from 4.6x104 to 3.7x10-5 per reactor year (Ref. 31). The dominant contribution to core damage was found to be from earthquakes in the 0.2 to 0.4 g range (Ref. 31). It is therefore concluded that, in comparison to the probability and consequences of a reactor core damage accident from a seismic event, the likelihood and risk associated with the beyond design basis seismic induced spent fuel pool failure are only a ,

small part of the overall risks associated with the operation of a nuclear power plant.

The best estimate consequences of a spent fuel pool accident are summarized in Table 5.1.3, based on a plant mean probability of a spent fuel pool accident of 2x10 4 per reactor year. This  ;

value is representative of both the PWR and BWR pool and includes the beyond design basis  !

earthquake accident. ,

The cost estimate data available is generally based on 1983 values. The offsite property damage from MACCS, the EPRI study on alternative spent fuel storage options, and onsite property 1 l

damage (cleanup and replacement power) am all monetized to, or represent 1983 costs. These 1983 cost estimates were escalated to 1988 values by using the Gross National Produc* Price Deflator ratio between 1988 and 1983 (121.4 divide by 104.1, or a factor of 1.17), taken from NUREG/CR-4627 (Ref. 39), Abstract 6.4 " Time-Related Cost Adjustments." As seen in Table 5.13, the monetized present value of the offsite health effects, at $1,000 per person-rem, de,minates present value estimates when compared to property damage costs at the 5% discount rate. At a 10% discount rate, as recommended in NUREG/BR-0058 (Ref. 40), the onsite and offsite property damage present value cost estimates would be about one-half the values shown in Table 5.1.3. The 5% discount rate is used in this Regulatory Analysis because it is believed to be more representative of the current economical environment than the 10% value. It is i therefore concluded that additional refinements in cost estimates concerning onsite and offsite property damage costs are not required and will not affect the value/ impact or cost benefit analyses provided. In addition, first order approximation of costs are generally being used in this Regulatory Analysis and any additional adjustment to these estimates is not warranted.

5-5 1

)

f-1 The significance of the potential consequences of this base case spent fuel pool accident evaluation, and the related costs associated with an accident, are discussed in more detail under Section 6, " Decision Rationale." Alternatives actions, other than this "no action" proposal, are considered in the following sections to determine if cost beneficial options are available to reduce the risk or consequences of a spent fuel pool accident.-

Table 5.1.3 Best Estimate Consequences of a Spent Fuel Pool Accident Frequency of occurrence of Eircaloy Fire 2x10-6 per reactor year (Mean value for PWR or BWR SPF)

Consequences, over 30 years based on 340 480 person rem people per square mile population density within a 50 mile radius of the site Present Value of Offsite Health Risk $ 480,000 (1988 $s)

Based on $1,000 per person rem Present Value of offsite Property Damage $ 124,300 (1988 $s) 5% discount over 30 years of remaining life Present Value of onsite Property Damage $ 32,400 (1988 $s) 5% discount over a five year repair period Total Present Value of a SFP Accident $ 636,700 (1988 $s)

Note: Without the beyond design basis seismic failure, consequences are estimated to be an order of magnitude lower.

5-6

i l

5.2 Alternative 2 - Require Use of Low Density Racks The use of high density storage racks inemases the probability of a release of radioactive from stored fuel if water is lost from the spent fuel pool. Studies performed in 1979 (Ref. 8) concluded that the minimum allowable decay time for PWR spent fuel in a well-ventilated room varies from five days, for open-frame storage configurations, to a value of 700 days, for high density, closed-frame configurations with wall-to-wall spent fuel placement. The minimum decay time for BWR fuel varied from five days to 150 days for the configurations studied. In addition, it was determined that the high density storage configuration could allow for the propagation of fuel damage from the recently discharged fuel to older, adjacent stored fuel in the pool. The msults of these 1979 studies prompted the identification of Generic Issue 82.

5.2.1 Risk Reduction Estimate On:: of the potential means to reduce the risk from loss of water in the spent fuel pool would be to require the use of low density storage racks for the recently discharged fuel. This alternative would mduce the pmbability of fuel damage, or a Zircaloy fire. The estimated reduction in risk is a result of the decreased decay time required for low density storages racks to pmelude spent fuel damage if water is lost, as compared to the high density racks. BNL estimated the reduction in risk to be at least a factor of five, or about an 80% reduction in the consequences '

(Ref.10). For the purpose of evaluating this alternative, a 100% reduction in the consequences will be assumed.

5.2.2 Cost of Low Density Storage Additional storage for spent fuel at reactor sites is mquired. If the DOE spent fuel repository opens in 2003, it is estimated that the industry will need to provided for between 12,200 MTHM to 20,000 MTHM of additional storage capacity (Ref. 7). If no apository is made available the additional capacity is estimated to be between 32,090 MTHM and 42,450 MTHM (Ref. 7).

I The use oflow density storage racks would require the need for additional at-reactor-site storage capacity to accommodate the change in the storage configuration, from high density racks. For each PWR low density storage location, three high density assemblies would be displaced. For a BWR, two high density assemblies would be displaced for each low density storage location j (Ref.13).  !

The evaluation of the amount of additional storage capacity, resulting from a proposal to require low density racks, is provided in reference 13. The fuel cycle information and spent fuel projections where obtained from. DOE /RL-87-11, " Spent Fuel Storage Requirements 1987" (Ref. 41). The proposed alternative for low density storage could increase the additional capacity by about 17,000 MTHM.

The cost of additional storage depends on the type of facility to be used. EPRI evaluated four alternative storage facilities in NP-3365 (Ref.16), " Review of Proposed Dry-Storage Concepts Using Probabilistic Risk Assessment." These concepts are:

1. Additional Pool, separate from existing pool,

2. Cask storage, 5-7

- - - - - - _ - - - _ _ - - - a

j

. l 1

3. Caisson, or dry-well, storage, and j

4. Vault storage.

The risk associated with these alternative concepts was found to be generally acceptable since the spent fuel would be five years old, or older, and the likelihood of significant fuel damage was found to be low. The consequences were found to be negligible when compared to the operating reactor at the site (Ref.16).

The costs for alternative storage concepts were evaluated by EPRI in NP-3380, " Cost Comparisons for On-Site Spent-Fuel Storage Options" (Ref. 37). In addition to the primary four alternatives identified above, silo storage, reracking and rod consolidation were addressed in terms of cost. The cost estimates are provided in units of dollars per kilogram of uranium

($/kgU) and vary with the size of the facility. The cost estimates decrease as the storage capacity increases, per unit, because of initial licensing and engineering fees associated with the facility design.

In a recently completed study perfonned by DOE (Ref. 7), the cost of additional storage for the cask concept and the rod consolidation concept have been reviewed and updated to reflect the limited amount of actual er.periences with these methods of providing additional at-reactor-site storage. These methods are considered to be the most practical and represent demonstrated technologies. The costs estimates are provided below, in Table 5.2.1.

The estimated cost (1988 $s) of additional storage by the year 2003 is estimated by DOE to range between $945 million and $1,267 million for 12,200 MTHM, and between $1,545 million and $2,000 million for 20,000 MTHM. If rod consolidation can accommodate the 350 MTHM requimments, the costs estimates would be mduced to $468 million for the 12,200 MTHM case and $793 million for the 20,000 MTHM case. The mean cost estimate for the 17,000 MTHM additional storage which would be needed if low density stomge is required is $1580 million, based on the DOE cost estimates. The mean cost (for the 108 pools) is $14.6 million per pool, for a unit cost of $93/Kgm of heavy metal.

The cost estimates for the dry storage concepts were developed assuming that the cost of a low density requirement would be an inemmental, additional cost as virtually every spent fuel pool will reach it's capacity limit and out of pool stomge will be mquired befom end-of-license if a repository in not available (Ref.13). Rod consolidation was not considered because it is not known how many pools can accommodate the additional decay heat load and structural loads associated with the in-pool increase and still meet NRC requirements. The unit cost for at-reactor storage decreases with an increase in the total amount, or capacity, of the additional storage required. The total licensing, engineering and fixed facility costs for the initial at-reactor storage facility is estimated to be $0.6 to $1.8 million (1988 $s) per facility, based on metal cask storage (Ref. 7). These one-time costs would not be impacted by this alternative, and are costs which will be incurred by most licensees prior to the availability of the DOE repository. The cost of the additional at-reactor storage which would result from this alternative is also provided, for reference, assuming that the additional storage capacity cost is based on the unit cost estimates for a facility the size of only the additional storage requirement. This is referred to as the lead cost estimate. Table 5.2.2 provides a summary of the incremental, capacity increases which would result from this alternative. The associated cost estimates for dry storage alternatives based on these capacities are provided in Table 5.2.3.

5-8

.. - . l

, .. l-l The DOE cost estimates based ca the incremental cost assp.:nption for the 17,000 MTHM 1 additional storage are $1,510 million, or a mean cost of %4 million per pool (1988 $s). The 1 DOE cost estimates (Ref. 7) level off for capacities in excess of 300 MTHM (see Table 5.2.4 below).

Table 5.2.1 Range of Unit-Cost Estimates for Additional Storage Requirements (Costs in 1988 $s per kilogram of heavy metal)

Capacity Increase Storage Technology 100 MTEM 300 MTHM 1000 MTEM Rod Consolidation (1) 40 - 75 30 - 50 n/a Metal Cask (10 MTEM) 60 -115 55 -115 55 - 100 concrete Cask 50 -105 45 - 90 45 - 80 Horizontal Concrete Modules 45 - 65 40 - 55 40 - 55 Modular Vault System 105 -155 70 -105 45 - 70 Notes:

('. ) - The unit costs are based on the cost for an additional storage slot created in the storage pool. From 2.6 to 3 spent-fuel assemblies must be consolidated for each storage slot (Ref 7).

n/a - An increase of 1000 MTHM is not applicable to rod consolidation because at a typical reactor not much more than 300 MTHM of additional spent fuel can be gained through consolidation (Ref. 7) .

Table 5.2.2 Additional Incremental Storage Capacity Requirements for Alternative 2 Capacity Range PWRs BWRs i'otal (MTHM) Impacted Impacted Impacted 0- 50 0 2 2 50 - 100 8 10 18 100 - 150 15 18 33 150 - 200 23 8 31 200 - 250 15 1 16 250 - 300 3 0 3 300 - 350 4 0 4 350 - 400 1 0 1 5-9

~

l:

l F Table 5.2.3 {

l' Storage Cost's Associated With Alternative 2 (1988 $s) l Cost: Per Poo1 ~ For All 108 Pools )

Discount Rate: 0% 5% 10% 0% 5% 10%

($1,000,000) ($1,000,000)

Pool (BNL Incremental Costs) 25.3 19.5 14.9 2,720 2,100 1,610 Drywell (BNL Incremental Costs) 10.6 9.5 8.0 1,150 1,038 863-Vault (BNL Incremental Costs) 24.1 19.5 14.9 2,612- 2,100 1,610 Cask (BNL Incremental Cos,ts). 14.0 14.2 12.2 1,516 1,539 1,318 Silo (BNL Incremental Costs) 18.2 14.2 11.2 1,959 1,539 1,178 Cask (BNL Lead Costs) 19.7 20.1 17.3 2,134 2,169 1,866 Cask (doe Incremental Costs) 14.0 14.2 12.3 1,510 1,530 1,330 Cask (Dom Lead Costs) 14.6 14.9 12.8 1,580 1,610 1,380 Notes:

zero % discount rate corresponds to the case where additional storage capacity is built now. The 5% and 10% rates reflect discounted costs in delaying the building of additional capacity until needed.

The diffemnce between the estimated costs for cask storage, in comparing the BNL-based and DOE-based estimates, are due to the difference in the $/Kgm costs estimates based on facility capacity. In Table 5.2.4, the BNL point-estimate cost (based on EPRI NP-3380, Ref. 37) is compared to the DOE lower and upper bound estimates.

5-10

~

.- s .

e Table 5.2.4 Cask Storage Cost Estimates as a Function of Facility Capacity Capacity BNL Point doe Lower doe Upper Estimate Estimate Estimate (MTHM) ($/Kgm) ($/Kgm) ($/Kgm)

(1983 $s) (1988 $s) (1988 $s)

(

25 113.2 105 160 50 113.2 95 140 75 113.2 85 120 100 113.2 80 115 125 113.2 80 110 150 113.2 80 119 200 113.2 80 105 300 99.8 80 100 400 93.4 80 100 500 , 89. 80 100 600 85.8 80 100 700 84.2 80 100 800 82.9 80 100 900 81.6 80 100 1000 81. 80 100 1200 79.4 80 100 1400 78.8 80 100 1600 77.5 80 100 1800 77.2 80 100 2000 77.2 80 200 5.2.3 Value/ Impact Summary The value/ impact, cost-benefit analysis is provided in terms of the mean industry risk from spent fuel pool accidents in Table 5.2.5. The best estimate accident fmquencies are used and the best estimate consequences, based on fission product release from 1/3 of a reactor core, are used.

The conditional probability of the Zircaloy fire, given the loss of water from the spent fuel pool, is 1.0 for the PWRs and 0.25 for the BWRs. Since the amount of spent fuel which could become involved in the release is uncertain, a sensitivity study using the worst case consequences, full spent fuel pool inventory at a high population site, is also provided.

The risk is comprised of 69 PWR spent fuel pools with a spent fuel damage probability of 1.95x10-6 per reactor year (including seismic events and conditional Zircaloy fire probability of 1.0 given loss of water) and 39 BWR spent fuel pools with a spent fuel damage probability of 1.71x10-6 per reactor year (including seismic events and conditional Zimaloy fire probability of 0.25 given loss of water). The mean remaining lifetime for the PWR spent fuel pool is 29.8 reactor years, and 27.9 years for the BWR spent fuel pool.

5-11

Since this alternative addresses dry storage, and because nearly every utility will require some additional dry storage prior to the start-up of the DOE repository, the NRC development and implementation costs concerning the licensing of a 10 CRF Part 72, " Licensing Requirements for the Independent Storage of Nuclear Fuel and High-Level Radioactive Waste," (amended August 19,1988,53 FR 31651), facility are not included as additional costs to be included in the value/ impact analysis. These costs will be incurred and are not affected by this alternative.

The industry and NRC operational costs associated with dry storage are also costs which will be incurred and am not impacted by this alternative. The additional storage requirements which would result from this alternative would not affect these costs. Industry and NRC costs which could result from a re-racking amendment to replace high density racks with low density racks as part of this proposed alternative are expected to be small in comparison to the dry-storage costs and have not been quantified. Inclusion of these additional costs would result in an even less favorable value/ impact or cost benefit assessment.

Sensitivity studies were performed by BNL to test the assumptions used (Ref.13). The discount rate applied to property damage costs, the monetary conversion factor for health effects, plant site economics and meteorology, and industry implementation costs were evaluated.

The recommended discount rate, NUREG/BR-0058 (Ref. 40), is 10%, which results in a lower estimate of property damages by a factor of two. Public dose reduction is not affected since it is not discounted.

A major difficulty with the net-benefit method is the evaluation of health effects in monetary units. Sensitivity studies of $500 and $2,000 per person-tem are used to demonstrate uncertainty.

The base line calculation adopted by BNL used the economic factors of the Zion plant site as a best estimate. Zion is somewhat higher than the U.S. average. A sensitivity study based on West Virginia was performed by BNL to evaluate the sensitivity to economic factors. The economics of West Virginia are considered to be much below the national average. Zion weathers conditions were also employed in the base line calculations with MACCS. A review of several severe accidents calculation, by BNL, indicated that varying weather models have a small effect on the public health and offsite consequences.

Since the costs associated with spent fuel storage construction costs and overhead and main:enance are well documented in either the EPRI NP-3380 (Ref. 37) study or in the DOE Dry-Storage study (Ref. 7), the industry implementation costs appear to be well defined and no sensitivity study is performed.

The impact of these sensitivity studies are shown in Table 5.2.6. The best case analysis is used as the reference, base line case. The results of the sensitivity study shows that the dominance of the cost of dry-storage in comparison to the potential damage costs is overwhelming and therefore this alternative is not cost effective. A factor of 10 increase in the probability of a Zircaloy fire in a spent fuel pool would not alter this conclusion.

5 12

.o- - .

I 1

Table 5.2.5 Summary of Industry Wide value/ Impact Analysis for Alternative 2 j Based on 100% Risk Reduction (1988 $s) j Attribute Dose Reduction (person-Rem) Cost ($1,000)

Best Est. High Est. (a) Best Est. High Est. (a)

Publia Health (1) 47,000 153,000 47,000 153,000 Occupational Exposure (Accidental) negligible negligible Onsite Property Damage (5% discount) 3,500 5,600 Offsite Property Damage (5% discount) 13,400 101,000 Industry Implementation and operation Cask Assumption -1,510,000 -1,510,000 NRC Development / Implementation and operation negligible negligible Net Benefit -1,446,000 -1,250,000 Benefit ($) / Cost ($) (2) 0.042 0.172 l Dose Reduction (person-rem)/Million $s (3) 31 100 l

Value/ Impact Ratio (4)

($/ Person-rem reduction) 32,130 9,900 Notes: (a) High estimate based on worst case of entire pool inventory at site with 860 people per square mile population iensity and Zion land use factors.

(1) Cost of health consequences set at $1,000 per person-rem.

(2) Averted costs divided by NRC + Industry implementation and operational costs.

(3) Public dose reduction divided by NRC + Industry implementation and operational cost.

l (4) Cost of NRC + Industry implementation and operation divide by public dose reduction.

l 5-13 l

l i

w__

Table 5.2.6 f Benefit / Cost Ratio Sensitivity Analysis for Alternative 2 (1988 $s)

' Parameter Under Study New Value Net Benefit Benefit /

j ($1,000) Cost Ratio (1)

Baseline Best Estimate -1,446,000 0.042 Discount Rate 10% -1,453,000 0.038 Health Effects $ 500/ person-rem -1, 47 ti, 0 0 0 0.022

$2,000/ person-rem -1,406,000 0.069 Site Economics West Virginia -1,453,000 0.038 Note: _ (1) None of these assumptions changes the risk reduction in person-rem averted, which remains constant at 47,000 person-rem averted. The implementation cost also is not changed. The value/ impact ratio remains

$32,130 per averted person-rem.

5.3. Alternative 3 - Improve Cooling /Make-Up Systems Four generic spent fuel pool cooling and make-up systems have been described in Section 4.7.6.

The probability of failure of these systems to fail to provide adequate cooling and make-up to the spent fac1 ranges from 2.2x10-5 per reactor year to 3.6x10-6 per mactor year, without consideration for recovery. With mcovery, the probability of damage to spent fuel is estimated to be in the lx10-8 per reactor year range, with an upper bound value of lx10-6 per reactor year.

A beyond design basis earthquake, as a result of the low seismic capacity of non-safety grade electrical components (motor control centers and switchgear),8is estimated to result in a probability of damage to spent fuel on the order of about 5x10' per reactor year. Essential components to either the cooling system or the make-up system, or both systems, are designed to the SSE and deterministically demonstrated to perform the safety function of maintaining the spent fuelin a safe and subcritical configuration for all credible storage conditions.

5.3.1 Risk Reduction Estimate Although a loss of cooling and subsequent heatup is a very slow event (on the order of several days), analyses have shown that after the spent fuel is uncovered, the remaining water would block air circulation and cladding overheating would occur for fuel which had been cooled for one year (Ref. 8). However, because of the lack of air circulation within the spent fuel holders, the oxida' ion reaction would be oxygen starved and the cladding would not melt. Thus, BNL concludes that catastrophic failure of the spent fuel would not be expected. Consequence 5-14

estimates for ruptured fuel pins was performed in NUREG/CR-4982 (Ref.10), and the resulting offsite consequences were found to be minimal, about 4 person-rem given the accident (see Table 4.6.2).

The economics of such an accident appear to be important. Since the reactor could not operate until the spent fuel pool was available, the cost of replacement power, until the spent fuel pool building was decontaminated and the equipment repaimd, could be considerable. It is estimated by BNL (Ref.13) that repairs and decontamination would take one month to one year depending on the degree of fuel damage and contamination. Replacement power costs estimates were obtained based on the method presented in Section 5.1. The onsite costs range from $19 to $227 million (1988 $s) conditional upon a spent fuel pool accident (Ref. 36). Integrated over the remaining lifetime of a typical plant, 30 years (the industry average), the expected cost associated with the gmdual coolant loss sequences (without discount) could be as high as

$150,000, based on a 2.2x10-5 per reactor year event (without credit for ecovery actions). The low value is estimated to be $12,500, without discount.

5.3.2 Cost ofImproved Cooling /Make Up Systems Two alternative systems for improvement of the spent fuel pool cooling system were evaluated by BNL (Ref.13) to assess the potential cost-benefit for each of the four generic system types:

1. Provide another full capacity pump and associated valves to eliminate the need for running the cooling system without a backup pump (System C and D). The first order approximation of the cost of this option is estimated by BNL to be

$50,000 (1983 $), or $60,000 (1988 $s) based on Section 5.1.4 cost escalation factors.

2. Provide a completely independent make-up train, BNL assumed this system to be similar to the primary spent fuel pool supply train. The hardware requires include a Category I water storage tank (200,000 gallon capacity), pumps, controls, and piping. The first order approximation of the cost of the independent make-up train plus overhead and maintenance costs were estimated by BNL to be one million dollars (1983 $), or $1.2 million (1988 $s) based on Section 5.1.4 cost escalation factors.

5.3.3 Value/ Impact Summary The cost-benefit characteristics of these options are summarized in Table 5.3.1. For the additional make-up train, the analyses assume 100% reduction in the initiating frequency, that is complete recover of the potential loss, and the averted costs are calculated at a 5% discount over the remaining average plant life of 30 years. The addition of a full capacity pump to System C or D would result in a risk reduction to the equivalent frequency of System A or B. For example the change in the initiating frequency for System C would be from 2.2x10-5 o t 3.8x10-6 per reactor year, or a change of 1.8x10-5 per mactor year.

l The only system which might benefit from either of these options is System C without credit for recovery actions. The generalized presentation of the Standard Review Plan requirements indicate that additional requirements, to improve the cooling /make-up system, would not result in a cost-beneficial improvement. When recovery actions are considered, the most appropriate estimate for the cost-benefit ratio is System D.

5-15

I. .

l The current requirements for the design of the spent fuel pool cooling and make-up systems, when credit for operator action to diagnose and recovery from a loss of cooling event i considered, are judged to be satisfactory. This finding is based, in part, on the assumption that, as a result of IE Bulletins and Infonnation Notices (see Section 5.1 above), licensees are aware of the need to assure that adequate instrumentation is available and maintained to alert the operators to degradation in the spent fuel pool or its support systems.

Table 5.3.1 value/ Impact for Generic Improvements To The SFP Cooling System (1)

(5% Discount Rate - 1988 $s)

System Description Option Cost of Averted Cost Benefit / Cost (Frequency) Option Range (2) Ratio Range

($1,000) ($1,000)

Low High Low High A Minimum SRP 1. Add Pump 60 None -

(3.8x10-6/R-y) 2. Make-up train 1,200 1.1 13.2 0.001 0.011 B F 4 n 4 == SRP 1. Add Pump 60 None -

With Credit 2. Make-up train 1,200 0.1 6.5 0.001 0.005 for Fire Hose (1.9x10-7/R-y)

C Old System w/. 1. Add Pump 60 3.5 41.4 0.058 0.690(3)

Both trains 2. Make-up train 1,200 6.3 75.7 0.005 0.063 30% of time (2.2x10-5/R-y)

D Old System 1. Add Pump 60 0.2 2.1 0.003 0.035 With Credit 2. Make-up train 1,200 0.3 3.8 0.001 0.003 for Fire Hose (1.1x10-6/R-y)

Notes: (1) Spent fuel cladding ruptures and releases gaseous fission products,

.no Zircaloy cladding fire. The offsite consequences are small, 4 person-rem .

given the loss of cooling cooling /make-up. Value/ impact ratio, in $a per

. averted person-rem, is very large (well in excess of $1,000 per averted person-rem) , however economics of spent fuel cladding rupture could be important.

(2) Averted costs of replacement power and cleanup / repair of spent fuel pool. Low estimate is for one month outage, high estimate is for one year outage.

(3) Based on a 10% discount rate, the averted cost estime.te is reduced to $24,700 and the benefit / cost ratio is reduced to 0.41. Similar reductions apply to all options at a 10% discount rate.

5-16

5.4 Alternative 4 -Install Spray Systems Post-accident spray systems have been considered as a potentially significant mitigative measure for spent fuel pool accidents. A scoping value/ impact assessment was performed by BNL (Ref.13) to provide some insights into the potential cost effectiveness of installing spray systems. The guidelines outlined in NUREG/CR-3568 (Ref. 36) for "First Approximation of Benefits and Costs" were used.

BNL emphasized that this assessment is scoping in nature due to the many assumptions involved and large uncertainties in data and decontamination factors assumed for spray systems.

5.4.1 Risk Reduction Estimate The principle reduction effect of the spray systems is achieved by decontaminating radiological releases thus permitting greater retention of fission products in the pool and the pool building.

Results of analyses of severe reactor accidents in support of NUREG-1150 (Ref. 42), indicate that containment spray systems can be significantly effective in reducing source terms and severity of consequences of nuclear reactor accidents (Ref. 43).

In this assessment,it is assumed that the major benefit of spray systems results from reduction in the offsite consequences. The onsite propeny damage is not effected, that is cleanup and repair and replacement power costs would still be incurred due to spent fuel damage and a Zircaloy fi t.

The effectiveness of the spray system is measured by the decontamination factor (DF), the amount of radioactive species released to the environment without the spray divided by the amount released with the spray. Decontamination factors for a spent fuel pool spray system are difficult to estimate without detailed calculations, therefore BNL assumed that the DF would be 45 based on NUREG-1150 analysis for the Surry plant containment spray system effectiveness.

The effects of a DF of 45 on the results of MACCS consequence calculations are provided in Table 5.4.1, and compared to the previous case without sprays. The effects of a spray system with a DF of 45 has the effect of reducing the offsite consequences to a small fraction of their original levels, therefore this can be considered to be an upper bound measure of the potential benefit of a post-accident spray system.

5.4.2 Cost ofInstalling Spray Systems Preliminary construction and industry maintenance costs were estimated by BNL. Assumed hardware requirements included a Category I water storage tank (200,000 gallon capacity) and a spray system including pumps, spray nozzles and associated hardware. The cost, on a first approximation basis, is estimated to be $1.2 million (1988 $s) per spent fuel pool (Ref.13).

The NRC cost associated with this option is estimated to be $100,000 (1988 $s) per spent fuel pool, roughly equivalent to one staff-year review effort per pool at $75,000 per staff-year (NUREG/CR-4627, Abstract 5.2, Ref. 39) plus $25,000 for the development and approval of a Technical Specification for the control of the administration, surveillance and mamtenance of the spray system.

5-17

Table 5.4.1 offsite Health and Property Damage Estimates (1988 $s)

With Pool Spray System (DF = 45)

Case Description Whole Body Dose offsite Property (person-Ram Damage per-Event) (1988 $s)

Without With Without With Spray Spray Spray Spray

'Best Estimate Consequences Last discharge 90 days after 7.97x106 1.25x106 4,oxio9 7.2x107 dischargo, 50 mile radius Based on 340 people / square mile Worst Case Estimate Consequences Total inventory 30 days after 2.56x107 6.78x106 3.0x1010 5.2x108

-discharge, 50 mile radius Based on 860 people /squaro mile 5.4.3 Value/ Impact Summary The value/ impact, cost-benefit analysis is provided in terms of the mean industry risk from spent fuel pool accidents in Table 5.4.2. The best estimate accident frequencies am used and the best estimate consequences, based on fission product release from 1/3 of a mactor core, are used.

The conditional probability of the Zircaloy fire, given the loss of water from the spent fuel pool, is 1.0 for the PWRs and 0.25 for the BWRs. Since the amount of spent fuel which could become involved in the release is uncertain, a sensitivity study using the worst case consequences, full spent fuel pool inventory at a high population site, is also provided.

The risk is comprised of 69 PWR spent fuct pools with a spent fuel damage probability of 1.95x10-6 per reactor year (including seismic events and conditional Zircaloy fire probability of 1.0 given loss of water) and 39 BWR spent fuel pools with a spent fuel damage probability of 1.71x10-6 per reactor year (including seismic events and conditional Zircaloy fire probability of 0.25 given loss of water). The mean reraaining lifetime for the PWR spent fuel poolis 29.8 reactor years, and 27.9 years for the BWR spent fuel pool.

The dose reduction estimate is derived from the change in the offsite health consequences 6 7

! shown in Table 5.4.1,6.72x10 person-rem per accident for the best estimate case and 1.88x10 person-rem per acciden: for the worst case. The offsite property damage costs are estimated using the revised MACCS values (with sprays), discount at a 5% rate over 30 years. The onsite property damage costs are assumed to be unchanged, cleanup and repair and replacement power costs are incurred.

5-18 i

r ,. -- -. .

The best estimate value/ impact ratio for this altemative is estimated to be $3,340 per averted person-rem and exceeds the general guideline value of $1,000 per averted person-rem. While the high estimate is seen to be marginally cost effective ($1,200 per averted person-rem), the use -

of Zion site demography. for the high estimate evaluation results is an overly conservative

. estimate of the risk reduction properties of a given plant modification (860 people per square -

- mile).

e l

l 5-19 i

I

' - - " - - -.___m__m. ._____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ , _ _ _ _ _ _

g _

9 Table'5.4.2 Sununary of Industry Wide value/ Impact Analysis for Alternative 4 Based on.a Spray System DF of 45 (1988 $s)

Attribute' Dose Reduction (person-Rem) Cost ($1,000)

Best Est. High Est.(*) Best Est. High' Est. (a)

Public Health (1) 39,450 110,500 39,450 110,500 Occupational Exposure (Accidental) negligible negligible Onsite Property Damage 0 0

-Offsite Property Damage (5% discount) 13,000 100,000 Industry Implementation-and operation -130,000 -130,000 NRC Development / Implementation and Operation. - 10,800 - 10,800

'Not Benefit - 88,400 + 69,700 Benefit ($) / Cost ($) (2) 0.373 1.50 Dose Reduction (person-rem)/Million $s (3) 280 840 value/ Impact Ratio (4) 3,340 1,200

($/ Person-rem reduction)

~

Notes: (a) High. estimate based on worst case of entire pool inventory at site with 860 people per square mile population density and Eion land use factors.

(1) Cost of health consequences set at $1,000 per person-rem.

(2) Averted costs divided by NRC + Industry implementation and operational costs.

(3)'Public dose reduction divided by NRC + Industry implementation and operational cost.

(4) Cost of NRC + Industry implementation and operation divide by

.public dose reduction.

5-20 1

5.5 Alternative 5 - Modify Spent Fuel Storage Rack Designs This proposed alternative would require the licensee to compartmentalize the spent fuel storage pool by installing partitions (and individual coolant supply diffusers for each compartment) to limit the extent of the accident, or modify the storage racks to improve air circulation, should the spent fuel storage pool drain. This alternative is directed towards risk mitigation, and to a lesser extent prevention.

This altemative was not quantified as part of this value/ impact study. The results of the cladding heatup calculation suggest that the only rack geometry that would msult in mitigation is low density racks. The probability of a loss of water from the spent fuel pool would not be changed. Compartmental restructuring of the spent fuel is not judged to be feasible without a significant loss in the storage capacity and the resulting need for additional at reactor dry storage is expected to overwhelm any potential risk reduction.

5.6 Alternative 6 - Cover Fuel Debris with Solid Materials This pmposed alternative would require the development of a contingency plan to dump massive amount of solid materials into a drained spent fuel pool to cover the rubble bed to a depth of several feet. The necessary materials would not be stockpiled on site, but could be obtained in a timely manner on an ad hoc basis, the materials (sand, clay, dolomite, boron compounds, lead, etc.) being commonly available in all parts of the country. This alternative would be directed at risk mitigation, not prevention.

This alternative was not quantified as part of this value/ impact study. The contingency plan wc,uld be concerned with a low fmquency event (on the order of 1.0E-6 per reactor year),

potential high consequence event. The results at Chernobyl can be used as a rough gauge of the efficacy of this measure, when carried out on a strictly ad hoc basis with no appamnt advanced planning. However, since the dominant risk sequence for the spent fuel pool accident is a beyond design basis earthquake, BNL concludes that it is dubious that the measures could be implemented soon enough to pmvent the major release to the environment during the first few hours of the accident (Ref.13).

5.7 Alternative 7 - Improve Ventilation Gas Treatment System This alternative would require the installation of a building ventilation and filter system capable of reducing the concentration of airborne radioactivity befom discharge to the environment.

This alternative would be directed at risk mitigation, not prevention.

This alternative was not quantified as part of this value/ impact study. Again the dominant risk contribution results from the beyond design basis seismic failure of the spent fuel pool structure, l a low frequency high consequences accident. To be effective, the spent fuel pool bliilding structure would have to maintain its integrity and the system itself would have to be designed to survive the postulated peak gmund acceleration which result in the spent fuel pool failure.

Additional investigations into this alternative are not considered to be reasonable.

5-21

5.8 Relationships With Other Requirements and Activities 5.8.1 Severe Accident Policy Statement A recently published report by LLNL, " Evaluation of External Hazards to Nuclear Power Plants in the United States - Seismic Hazard," NUREG/CR-5042, Supplement 1 (Ref. 31), summarizes the msult of the study of the risk of core damage due to seismic mitiated events.

The overall objective of the LLNI; study "is to present information that assists the NRC staff in  !

deciding whether seismic vulnerability searches for nuclear power plants should be in the I implementation of the Sevem Accident Policy Statement." To accomplish this objective, the LLNL report:

1. Considers effects of the evolution of design requirements and design practices on plant seismic capacity.

2. Identifies other specific review area of potential seismic vulnerability, including seismically induced fires and floods, spent fuel pools and seismic common-mode failures.

3. Identifies programs which address item 1 and/or item 2, and assess the extent to which these programs provide useful information on seismic capacity of nuclear plants.

4. Recommends incorporating appmpriate items from above into the seismic margins program or other seismic vulnerability searches.

The LLNL mport considemd the results presented in NUREG/CR-4982, " Severe Accidents in Spent Fuel pools in Suppon of Generic Safety Issue 82" (Ref 10), and concluded that "A comparison of the results of the fuel pool analysis with the two figures of merit is difficult since the fuel pool failure does not constitute core damage and any potential release involves long lived radioactive material. In addition, it is difficult to draw conclusions concerning spent fuel pools based on only a single generic analysis. Therefore, any decision on the inclusion of spent fuel pools into the sevem accident policy implementation requires more data and analysis, and cannot be concluded at this time."

The first figure of merit considered by LLNL is the core damage frequency. In numerical terms LLNL uses a mean core damage frequency in the range of 1x10-5 (or less) per reactor year as meeting the Commissioners stated objective, in the Policy statement on Safety Goals, as:

"providing reasonable assurance, given consideration to the uncertainties involved, that a core damage accident will not occur at a U.S. nuclear power plant."

5-22

[

l .,

~ The second figum of merit is the frequency of a large release. In the Policy Statement on Safety Goals, the following guidance is given as a general performance guideline:

" Consistent with the traditional defense-in-depth approach and the accident mitigation philosophy requiring performance of containment systems, the overall '

mean frequency of a large release of radioactive material to the environment from a mactor accident should be less than 1 in 1,000,000 per year of reactor operation."

The current status of this guideline is that the NRC staffis giving detailed consideration to how such a performance guideline can be implemented, including how to define more precisely the definition of a "large release of radioactive material to the environment." In the LLNL study, a large micase of radioactive material to the environment has been defined as a release of a substantial fraction of the radioactive core in a time period relatively early in the postulated accident scenario. This definition was derived from Probabilistic Risk Assessment (PRA) literatum which has defined a "large early release." Further discussion on the applicability of such guidance to Generic Issue 821s presented in Section 6.2.

5.8.2 Seismic Design Margins Program The current objectives of the Seismic Design Margins Program are:

1. To develop and improve guidance for assessing the inherent capability of nuclear power plants to withstand earthquakes above the design level.

2. To provide an effective and efficient means to identify vulnerabilities of nuclear plants to seismic events.

The seismic margins approach has chosen as one ofits figures of merit a high confidence oflow probability of failure (HCLPF). The HCLPF is a conservative representation of capacity and in simple terms corresponds to the earthquake level at which it is extremely unlikely that failure will occur. Two approaches are recommended for estimating the component HCLPF values: the PRA fragility approach and the Conservative Deterministic Failure Margins (CDFM) approach.

l The CDFM HCLPF approach has been developed and used by EPRI in a trial review of the l Catawba Nuclear Station, with a seismic margms earthquake (SME) of 0.3 g. The resultant HCLPF for core damage sequences was found to be 0.24 g. An NRC sponsored review panel examined the EPRI work and found the methodology can accomplish its main objective and is reasonably accurate (NUREG/CR-5042, Supplement 1, Section 4, Ref. 31).

The PRA HCLPF approach has been used by the NRC to evaluate core damage sequences at l Maine Yankee. The SME was also set at 0.3 g for this study. The HCLPF was found to be 0.21 g, and later revised to 0.27 g after the licensee committed to upgrading the refueling water storage tank.

The HCLPF approach used in the seismic design margins program does not use the seismic hazard curves. That is, the probability of a core damage sequences due to a seismic initiator are not evaluated in the traditional terms of frequency per reactor year used in PRAs. Instead the HCLPF value can be compared to the SME value. A plant HCLPF value greater than or equal 5-23 1

L_-_____-__---_--

to the SME would be considered to have adequate capacity since the SME would be chosen to assure adequacy. If the HCLPF value is less than the SME, then the site specific hazard curve could be used to estimate the recurrence frequency for that level of earthquake.

In NUREG/CR-5042, LLNL concludes that plant HCLPF capacity represents a conservative estimate at which there is a high confidence of a low probability of core damage. A more realistic parameter is the plant median capacity which is more than a factor of two greater than the HCLPF. It has been suggested that two times the plant HCLPF capacity could be used in conjunction with the median site specific hazani curve to obtain a recurrence frequency for comparison with some evaluation criterion. In light of the screening approach used for seismic margin reviews, LLNL goes on to conclude that research is needed to address what may be the appropriate factor that can be used along with the plant HCLPF capacity and what would be an appropriate evaluation criterion.

Until more definitive guidance is developed and approved by the Commission for the assessment of the external seismic hazard risk, the currently accepted guidelines for a regulatory impact analysis are used to define the risk. The mean failure frequency is used. The mean frequency is currently used for extemal events based on the use of the mean frequency in evaluating risk from internal events. Component and systems failures are described by their estimated mean failure rates.

5-24

~

6. DECISION RATIONAL.E The risk from the storage of spent fuel in the spent fuel storage pool at light water reactors is dominated by the beyond design basis earthquake accident scenario. The seismic capacity, or fragility, of two older spent fuel pools indicate that the high confidence of low probability of failure (HCLPF) is about three times the safe shutdown earthquake (SSE) design level. The HCLPF values are estimated to be 0.5 for the BWR and 0.65 g for the PWR spent fuel pools studied. The safe shutdown earthquake (SSE) for the two plants are 0.14 g and 0.2 g, mspec-tively. The median peak gmund acceleration needed to fail these pools is estimated to be in the 1.4 to 2.0 g range, a factor of ten higher than the SSE design value. A report prepared by the American Society of Civil Engineers (Ref. 29) also concluded that, in general, the seismic design of nuclear facility structums results in median factors of safety on the order of 4 to 19 based on post-1973 design criteria.

The structural capacity of the elevated BWR pool is lower than that for the PWR pool located at the ground level, however the lower conditional probability of a Zircaloy fire for the BWR fuel assembly design (0.25 as compared to the PWR value of 1.0) offsets the higher seismic failure fmquency. The probability of a Zircaloy cladding fire, resulting from the loss of water from the spent fuel pool,is estimated to have a mean value of 2x10-6 per reactor year for either the PWR or the BWR spent fuel pool. The seismic event contributes over 90% of the PWR probability, and nearly 95% for the BWR.

The source term for the spent fuel pool accident is not the same as the source term associated with core damage accidents. The consequences of a spent fuel pool accident which results in the complete loss of water is dominated by the long lived isotopes, such as cesium and strontium.

The health consequences are dominated by the risk of latent cancer fatalities due to long term exposures.

The best estimate of the consequences of a spent fuel pool accident which results in spent fuel damage to approximately one-third of an equivalent reactor core is 8x10 6 person-rem. This total dose translates to a public health risk from a spent fuel pool accident of 480 person-rem over an average remaining hcensed lifetime of 30 years. The best estimate offsite property damage cost is $4,000 million (1988 $s). The best estimate values are based on a populanon density of 340 people per square mile within a 50 mile radius from the site and result from the mlease of radionuclides from the last fuel discharge,90 days after being discharged. The best estimate of the onsite costs for a SFP accident is $1,180 million (1988 $s), including five years of replace-ment power to replace the damaged spent fuel pool. Based on an average remaining lifetime of 30 years and a 5% discount rate, the present value of the offsite property damage is estimated to be $124,300 and the pmsent value of the onsite property damage is estimated to be $32,400. As an upper bound, worst case, the consequences of the release of the full fuel pool at a high population site (860 people per square mile within a 50 mile radius from the site),26x10 6 person-rem, was used to evaluate.the sensitivity of the consequences for proposed alternatives.

The corresponding estimate in offsite property damage is $30,000 million (1988 $s).

l The consequences, in person-rem, from a spent fuel pool accident are mlatively insensitive to the quantity of spent fuel assumed to be released during an accident, when the typical assump-tions regarding interdiction dose and decontamination are applied. In the MACCS consequence calculations, no planned evacuation was assumed, however, persons expected to receive more than 25 rem from ground shine in seven days were assumed to be relocated in one day. An addi-tional dose limit over 30 years of 25 rem was also used to determine the interdiction level.

6-1 L___ _ _ _______ _____ ____ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ - . _ _ _ _ _ _ _ _ _ _ _ _

MACCS also includes a separate interdiction criteria for crops: crops are interdicted if the resulting ingestion doses would exceed 25 millirem per year. This dose rate is the U.S. Envimn-mental Protection Agency allowable chronic environmental dose rate for normal activities.

The amount of contamination, or land interdiction area, is strongly influenced by the quantity of spent fuel assumed to be released. Sensitivity studies have been performed for the release from the last refueling discharge and for release from the fullinventory of a spent fuel pool which has accumulated the equivalent of about four cores in spent fuel assemblies. Sensitivity calculations to study the possible effects of fission product mtention on structures and to study the possible effects of a spent fuel pool post-accident spray systems were also performed. The results of these analysis (based on the last discharge assumption) indicated that a decontamination factor assumption of ten reduces the consequences by a factor of two, and the interdiction area by a factor of ten (Ref.10). A decontamination factor of 45 results in a reduction in consequences of a factor of six and a factor of about 55 in the value of offsite propeny damage (Ref.13).

6.1 Comparison to the Backfit Criteria (10 CFR 50.109)

The value impact evaluation, pmsented in Section 5, for the proposed alternatives for Generic Issue 82 does not indicate that cost effective options are available to mitigate the risk of beyond design basis accidents in spent fuel pools. The option to use low density storage racks for recently discharged fuel has a best estimate value impact ratio of $32,000 per averted person-rem. Low density racks would decrease the frequency of a Zircaloy cladding fire by at least a factor of five to ten, and the value impact ratio is based on 100% mduction in public dose. For the worst case, a high population site with the full fuel pool inventory being released, the value impact ratio is $9,900 per avened person-mm. When compared to the general guideline value of

$1,000 per averted person-rem, the low density option is not justified.

The use of a post-accident spray system to mitigate the consequences of a spent fuel pool acci-dent has a best estimate value impact ratio of $3,300 per avened person rem, with a worst case estimate of $1,200 per averted person rem. This assumes that a post-accident spray system can be designed to withstand the beyond design basis earthquake which causes failure of the spent fuel pool structum and has a decontamination factor (DF) of at least 45. Other structums and equipment within the spent fuel storage pool building (for example the afueling crane) would also have to be reviewed to assure that their failure would not compromise the proposed spray system. Under the worst case release assumption, full fuel pool inventory at a high population site, this option is marginally cost beneficial but still exceeds the general guideline value of

$1,000 per avened person-rem. However the complete spent fuel pool inventory being released is considered to be highly unlikely. Results of cladding fire propagation calculations indicate that only fuel which is one to two years old could be involved in the release. Also, the demographics are a high estimate of the attributes of a typical plant modification (860 people per square mile).

Potential improvements to the spent fuel pool cooling and make-up systems were also ex-amined. The potential risk to the general public is estimated to be very small, on the order of 3 to 4 person-rem, given a loss of cooling event which results in failure of the spent fuel cladding but not a Zircaloy cladding fire. The value/ impact ratios are very large, well in excess of the general guideline value of $1,000 per avened person-rem, however the economics could be im-portant if the spent fuel pool is unavailable and the reactor is shutdown until cleanup and repairs are completed. The cost-benefit ratios for either an additional cooling pump or an additional make-up train were found to be less than one.

6-2

Thme additional alternatives, (1) to modify the spent fuel storage rack designs, (2) to cover the spent fuel debris with solid materials, and (3) to improve the ventilation gas treatment system, were not explicitly quantified. Compartmental modification to the storage rack designs would result in the displacement of fuel from the spent fuel pool to at-reactor storage casks, a costly option as shown in Alternative 2. Considering that the risk from a spent fuel pool accident is a i result of a beyond design basis earthquake, it it highly unlikely that materials could be  :

transported to the site to cover the spent fuel debris in time to reduce the mleases of radioactive materials from the spent fuel pool. Finally, since the integrity of the spent fuel building struc-ture following a beyond design basis earthquake is questionable, improvements in the ventila-tion gas treatment system would be difficult to obtain.

Therefore, the backfit criteria (Ref. 44) that (1) a substantial incmase in the overall protection of the public health and safety is achieved, and (2) the direct and indirect costs of implementation are justifled, are not met for any of the alternatives considered.

6.2 Comparison to the Safety Goal Policy Statement The frequency of damage to the spent fuel is estimated to be on the order of 2x10-6 per reactor year, including the bje ond design basis seismic earthquake. This value, when compared to a target value of 1x10 (or 5x10-5) for a core damage e.ccident, represents a small part of the i overall frequency of core damage - 2% to 4%.

The frequency of a release of radioactive material to the environment is assumed to be the same as the frequency of spent fuel damage. The underlying assumption is that the spent fuel pool housing (refueling building, auxiliary building or secondary building) fails due to either the dominant seismic event or due the extmme temperature conditions which would acompany a Zircaloy cladding fire and fuel melting scenario. The spent fuel pool housing does not provide a containment barrier similar to the containment structure surrounding the reactor core, especially under the conditions postulated to dominate the release of radioactive materials.

It is difficult to compare the estimated 2x10-6 per reactor year release frequency due to a spent fuel pool accident to a target value of 1x10-6 per reactor year for a large release, particularly without a definition for "large release" . The spent fuel pool source term is not similar to the core damage (or melt) source term and the consequences of a spent fuel pool accident are dominated by latent cancer risks. A possible definition is used in current PRA studies; that is, a "large mlease" is considemd to be an "early, large release" associated with an environmental release within a few hours of a core damage accident (presumably from 100% power). Another defm' ition of a "large release" currently being considered by the staff is a release that has a potential for causing an offsite early fatality (see for example NUREG-1150, Ref. 42). Either of these definitions, in particular any consideration for early fatalities, appear to suggest that the spent fuel pool release is not a "large mlease."

Societal risk to the public is based on the statistically expected number of early and latent cancer fatalities. The Safety Goal Policy Statement (Ref. 45) curmntly defines the early fatality area calculation as that within one mile from the site boundary. A ten mile radius is defined for cal-culating latent cancer fatalities. The language of the Policy Statement also mquires that the risk from an accident at a nuclear power plant be 0.1% of that normally encountered by the public.

Based on recent data (Ref. 46) the total fatality rate from cancer in the U.S. is 189.3 per 100,000 6-3

persons, or a risk of 1.9x10-3 per year. Therefore it can be inferred that a latent cancer fatality rate for nuclear power phmt operations of 2x10-6 per reactor year, or less, is consistent with the safety goal.

To meet the general objective for societal risk, the probability of a latent cancer fatality from a spent fuel pool accident should not be mom than a mlatively small fraction of ar. overall target value for nuclear power plant operations. The best estimate MACCS calculation for the spent fuel pool source term, for 340 people per square mile over a 50 mile radius, predicts a conse-quence of 8 million person-rem per event. The dose conversion factor for latent cancer fatalities is in the 150 to 200 latent cancer fatalities per million person-mm range. The expected number of latent cancer fatalities is 1,600 per event, and the latent cancer fatality rate would be 0.0032 per reactor year (1,600 latent cancer fatalities per event times 2x10-6 events per reactor year) for the affected population.

The mean population within a 10 mile radius of a reactor site is 57,000 people (based on a mean density of 182 people per square mile), and 2,670,000 people within a 50 mile radius (based on 340 people per square mile) in the year 2,000 (Ref. 47). The expected number of cancer fataltues from all causes in the 50 mile radius is 0.2% of the population, or 5,340 per year. In a 10 mile radius, the expected number of cancer fatalities is 114 per year. Using 0.1% of 10 mile radius value, a target value for latent cancer fatalities from the operation of a nuclear power plant would be less than 0.114 latent cancer fatalities per reactor year. The 0.0032 latent cancer fatalities per reactor year associated with the spent fuel pool accident is less than 3% of the 0.114 per year target value based on the calculation area specified in the Safety Goal Policy Statement, even without correcting for the fact that only a fraction of the 50 mile radius latent cancer fatalities would occur within the 10 mile radius.

4 The estimated frequency of a spent fuel pool accident,2x10 reactor 4 to year,5resulting in spent fuel damage meets a target objective of a few percent of a 1x10 5x10 per reactoryear value for overall core damage frequency. The target objective for a "large release" of 1x10* per reactor year is marginally met, within a best estimate factor of two, but subject to interpretanon since the definition of "large miease" is still under development. In meeting the societal risk ob-b'ective of 0.1% of the normally occurring risk to the public given the release frequency of 2x10-per reactor year, the latent cancer fatality rate from a spent fuel pool accidents is estimated to be less than 3% of the target value for the operation of a nuclear power plant.

Therefore, the risk and consequences of a spent fuel pool accident appear to meet the Safety Goal Policy Statement public health objectives. They would also meet the proposed 1x10-6 per reactor year large-release frequency guidelines, at least pending definition of a "large release" by the Commission. Therefore, Alternative 1 "No Action"isjustified.

6-4

6.3 Other Considerations In addition to implememing the requirements contained in 10 CFR Part 50 Appendix A of the

" General Design Criteria," and 10 CFR Part 20, concerning radiation doses being kept as low as is reasonably achievable, licensees should have implemented additional or corective actions based on the following guidance:

1. IE Bulletin 84-03, " Refueling Cavity Water Seals," issued August 24, 1984.

(Ref. 21)

2. E Information Notice 84-93, " Potential for Loss of Water From the Refueling Cavity," issued December 17,1984. (Ref. 24)

3. Generic Letter 85-11, " Completion of Phase II of ' Control of Heavy Loads at Nuclear Power Plants' NUREG-0612," issued June 28,1985. (Ref. 4)

4. IE Information Notice 87-13, " Potential for High Radiation Fields Following Loss of Water from Fuel Pool," issued February 24,1987. (Ref. 34)

5. E Information Notice 87-43, " Gaps in Neutron-Absorbing Material in High-Density Spent Fuel Storage Racks," issued September 8,1987. (Ref. 33)

6. E Information Notice 88-65, " Inadvertent Drainages of Spent Fuel Pools,"

issued August 18,1988. (Ref. 25)

7. IE Information Notice 88-92, " Potential for Spent Fuel Pool Draindown," issued November 22,1988. (Ref. 26)

Based on compliance with the GDCs and licensees taking corrective actions identified as a result of reviewing facility designs and operations based on E Bulletins and Information Notices, the frequency of a spent fuel pool accident resulting in a Zircaloy cladding fim and the release of fission products to the environment from internal events, such as missiles, heavy load drops, loss of cooling or make-up, inadvertent drainage or siphoning and pneumatic seal failures, is estimated to be on the order of 2x10-7 per mactor year. Operator diagnosis and recovery are important factors considend in the development of the event frequencies for these events and portions of this evaluation are premised on licensees having taken appropriate actions in response to the concerns identified to prevent similar occurrences, or at least understand the potential consequences of these events and develop appropriate procedures to respond to them and to mitigate the consequences.

The overall frequency of a spent fuel pool accident resulting in a release of radioactive materials to the environment is estimated to be 2x10-6 per reactor year for a light water reactor spent fuel storage pool when the external seismic hazard is included. The beyond design basis earthquake dominates the risk,90% to 95% of the total. The HCLPF value is estimated to be three times the safe shutdown earthquake (SSE) value peak ground acceleration value,in the 0.5 to 0.65 g range. The median capacity is estimated to be in the 1.4 to 2.0 g range.10 CFR Part 100 Ap-pendix III.(c) defines an SSE as:

1 6-5

"that canhquake which is based upon an evaluation of the maximum eanhquake potential considering regional and local geology and seismology, and specific characteristics of local subsurface material. It is that canhquake which produces the maximum vibratory ground motion for which cenain structures, systems, and components are designed to remain functional. These structures, systems, and components are those necessary to assure: (1) the integrity of the reactor coolant pressure boundary, (2) the capability to shutdown the reactor and maintain it in a safe shut down condition, or (3) the capability to prevent or mitigate the conse-quences of accidents which could result in potential off-site exposures comparable to the guideline exposures of 10 CFR Pan 100."

In NUREG/CR-5042, Supplement 1 (Ref. 31), LLNL reviewed available PRA literature to determine the seismic hazard contribution to core damage accidents. A review of analyses for A-45, " Decay Heat Removal Requirements," indicates that the dominant eanhquake range for core damage falls within the 0.2 to 0.4 g range. A review of Zion and LaSalle Seismic Safety Margins Research Program (SSMRP) analyses also concludes that the 0.2 to 0.4 g range dominates core damage from seismic initiators, The dominant component failures, contributing to core damage, were found to be:

1. Yard Tanks - condensate storage tanks, refueling water storage tanks.

2. Electrical Equipment - batteries, buses, cabinet anchorage, contacts, relays, transformers.

3. Diesel Generator Peripherals - fuel oil tanks, lube oil tanks, coolers.

4. Structural failures - block walls, service water buildings, reactor internals.

5. Equipment Anchorages.

In other words, this type of spent fuel pool accident requires an canhquake larger than that which would result in core damage and the release of radioactive material to the environment.

The mean core damage frequencies due to the seismic hazard are in the 3x104 to 1.4x104 per reactor year range based on published PRA results, with seismic related release frequencies in the range of 2x10-7 to 1.4x104 per reactor year for peak ground accelerations in the 0.2 to 0.4 g range (NUREG/CR-5042, Supplement 1, Table 3-3 Ref. 31). The spent fuel pool accident is es-timated to have a frequency on the order of 2x10 gper reactor year for a peak ground accelera-tion in excess of 0.5 g.

In estimating the likelihood of a beyond design basis canhquake resulting in a failure of a spent fuel pool, uncenainty can be introduced iato the evaluation when attempting to characterize the seismic hazard of a site. The seismic hazard is a quantification of the probability of exceeding a given peak ground acceleration on an annual basis. As shown in NUREG/CR-5176, and also noted m an NRC memorandum dated December 29,1988 (Ref. 48), the uncenainty in estimat-ing the seismic risk is about an order of magnitude, and relates to how expen judgment is used in the development of the site characterizations.

6-6 l

For each of the two plants studied by LLNL in NUREG/CR-5176, a family of seismic hazard curves were convolved with a family of plant-level seismic fragility curves to obtain a probabil-ity distribution of the frequency of occurrence of the seismic initiated accident under study. At the time this work was performed by LLNL, the complete family of seismic hazards curves were not available for the two plants studied, other than at some selected percentile values. When the seismic hazard curves are gmuped in this manner, the specific features of the individual hazard curves (for exat p they may intersect one another) are lost. Median and 95 percentile hazard curves were used by lNL to develop a discrete set of seismic hazard curves for each of the two plants studied. A lognormal distribution was used for the purpose of obtaining approximate risk estimates. The resulting lognormal distribution was cutoff at different percentile values to judge the sensitivity of the results. A cutoff value of 99 percent was recommended for use by LLNL in NUREG/CR-5176. The resultant frequency estimate for spent "uel damage due to a beyond design basis earthquake is 2.0x10-6 per reactor year for the LWR spent fuel pools studied in NUREG/CR-5176.

More recently, EQE Engineering, Inc., the same subcontractor employed by LLNL for the NUREG/CR-5176 effon, re-evaluated the seismic risk for the same two plants based on true mean seismic hazard curve data (Ref. 48). EQE provided two sets of results based on the use of two sets of experts, the "5 G-Expens" and the "4 G-Expens." The resultant mean annual fre-quency of failure of the spent fuel pool structures decreases by a factor of 8.8 for the BWR spent fuel pool ard 2.8 for the PWR spent fuel pool by removing one seirmic ground motion expen, or

" outlier," frorn the seismic hazard characterization estimate (for example when going from the "5 G-Expen" to the "4 G-Expen" ground motion expen judgment). Similar results were ob-tained by LLNL in NUREG/CR-5176 in going from a cutoff value of 100 percent to 99 percent, by eliminating a small portion of the tails from the lognormal distribution curves. Since the tail of the legnormal distribution extends to infinity,it might be possible to get values of the proba-bility of exceedance greater than one. Truncation of the lognormal distribution curves at an ex-ceedance value less than one, at 0.99, was used in the LLNL study. The relative magnitudes are similar. For the "5 G-Expen" values, the more recent plant specific BWR seismic failure fre-quency from the EQE study could be a factor of 5.5 higher than the earlier LLNL evaluation.

SImilarly, the more recent plant specific PWR seismic failure frequency could be a factor of 2.6 higher than the earlier LLNL evaluation. Based on the "4 G-Expen" values, the earlier LLNL evaluation of the seismic failure frequency is slightly higher than the more recent EQE values for both spent fuel pools studied. The mean seienic failure frequencies for the two methods are summarized in Table 0.3.1.

Due to the skewed nature of the distribution of expert judgment, the mean is a highly unstable estimate of the seismic hazard . In these distributions the most extreme opinion weighs heavily when the mean is calculated. The mean, which is an arithmetic average of all inputs, frequently exceeds the 85th percentile of all the inputs. This problem created by the skewed distribution of expert judgment exists for either method, the actual true arithmetic mean or the lognormal dis-tribution.

1 A re-evaluation for Alternative 2, the use of low-density storage racks for recently discharged fuel, using these higher seismic failure frequencies esults in a best estimate value/ impact ratio of $9,500 per avened person-rem. Using the worst case assumptions, the value/ impact ratio is

$3,000 per averted person-rem. Alternative 2 is judged to be the most practical option forieduc-ing the risk and the implementation costs are well defined. While a re-evaluation for Alternative 4, the installation of a post-accident spray system, indicates a marginally acceptable best es-timate value/ impact ratio of $1,050 per averted person-rem, the uncertainty in the implementa-6-7

tion cost of this option is large. The implementation cost is based solely on the instJiation of the spray system and does not consider the potential for the need to reinforce other parts of the spent fuel storage building structmes to assure that their failure in a beyond design basis earthquake would not compromise the spray system. Therefore, even with the h.igher seismic frequencies, the staff would not conclude that any of the options considered would be cost-effective.

Although these studies conclude that most of the spent fuel pool risk is derived from beyond design basis earthquakes, this risk is no greater than the risk from core damage accidents due to seismic events beyond the safe-shutdown earthquake. Therefore, reducing the risk from spent fuel pools due to events beyond the safe-shutdown earthquake would stillleave a comparable risk due to co:e damage accidents. Because of the large inherent safety margins in the design and construction of the spent fuel pool, Alternative 1 "No Action"is justified.

When taken together, the discussions presented in Sections 6.1,6.2 and 6.3 form the basis for a decision that no corrective actions are justified. The risk due to beyond design bnis accidents in spent fuel pools, while not negligible, are sufficiently low that the 9dded costs involved with fur-ther risk reductions are not warranted.

Table 6.3.1 Summary of SFP "eismic Failure Frequency Estimates Pool Type NUREG/CR-5176 Results NRR True Mean Results Cutoff Value Frequency Expert Group Frequency (per cent) (per R-year) (per R-year)

Elevated BWR 100 3.8x10-5 5 G-Experts 3.7x10-5 99 6.7x10-6 4 G-Experts 4.2x10-6 on Ground PWR 100 8.6x10-6 5 G-Experts 4.7x10-6 99 1.8x10-6 4 G-Experts 1.7x10-6 Note: The NUREG/CR-5176 frequencies at the 99 per cent cutoff level were used in this Regulatory Analysis as being representative of the best est bate, generic values for an elevated BWR spent fuel pool and a PWR spent fuel pool located at the ground elevation.

6-8

a., _

.. . , 7. o. , . .

w .. :o - .. .

' d f' s :,,.[(N

. . .g 4 g.

p: t

'.i7. IMPLEMENTATION' No regulatory action.is necessary for the resolution of this issue. This regulatory analysis and i; & ' the supporting contractor reports will be made publicly available as part of their normal distribu-

' tions. ~

y, i

6'r

• i,. .,

i 4e'

'l h 7

'J.

'I 7-1

H l

8. REFERENCES - I

1. U.S. Nuclear Regulatory Commission (USNRC), " Reactor Safety Study - An Assessment of Accident Risk in U.S. Commercial Nuclear Power Plants," WASH-1400, October 1975.

2. Memorandum from H2. Denton to R.J. Mattson, " Schedule for Resolving and Completing Generic Issue 82 - Beyond Design Basis Accidents in Spent Fuel Pools,"

dated December 7,1983. PDR Accession No. 8312270117.

3. U.S. Nuclear Regulatory Commission (USNRC), " Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plan:s, LWR Edition," NUREG-0800 (fmerly NUREG-75/987), June 1987.

4. Generic I.ctter 85-11, from H.L Thompson, Jr., " Completion of Phase II of ' Control of Heavy Loads at Nuclear Power Plants' NUREG-0612," dated June 28,1985. PDR Accession No. 8506270216.

5. U.S. Nuclear Regule. tory Commission (USNRC), " Control of Heavy Loads at Nuclear

' Power Plants," NUREG-0612, June 1980.

6. . Nuclear Energy Services, "At-Reactor Storage Expansion Engineering Evaluation,"

September 1981, prepared for U.S. Department of Energy, Savannah Flyer Operations Office.

7. U.S. Department of Energy (DOE), " Initial Version Dry Cask Storage Study,"

DOE /RW-0196, Office of Civilian Radioactive Waste Management, August 1988.

8. U.S. Nuclear Regulatory Commission (USNRC), " Spent Fuel Heatup Following Loss of Water During Storage," NUREG/CR-0649, March 1979.

9. N.A. Piscano, et. al., "The Potential for Propagation of Self-Sustaining Zirconium Oxidation Following Loss of Water in a Spent Fuel Storage Pool," January 1984 (Draft Report). PDR Accession No. 8505090480.

10. U.S. Nuclear Regulatory Commission (USNRC), " Severe Accidents in Spent Fuel Pools in Support of Generic Safety Issue 82," NUREG/CR-4982, July 1987.

11. Memorandum from E.D. Throrn to K. Kniel, " Spent Fuel Pool Fire Heat Transfer Modeling (GI-82 Beyond Design Basis Accidents in Spent Fuel Pools)," dated August 11,1987. PDR Accession No. 8710280175.

12. A.G. Croff, "ORIGEN2: A Versatile Computer Code for Calculating the Nuclide Compositions and Characteristics of Nuclear Materials," Nuclent Technningv. Vol. 62, pp. 335-352, September 1983.

13. U.S. Nuclear Regulatory Commission (USNRC), "Value/ Impact Analyses of Accident Preventive and Mitigative Options for Spent Fuel Pools," NUREG/CR-5281, March 1989.

8-1

j e =

14. U.S. Nuclear Regulatory Commission (USNRC), " Reactor Risk Reference Document,"

Vol. 3, Appendix O, " Overview of MACCS and CRAC2 Offsite Consequences Models,"

NUREG-1150, Draft for C6mment, February 1987. PDR Accession No. 8703180080.

15. U.S. Nuclear Regulatory Commission (USNRC), " Technical Guidance for Siting Criteria Development," NUREG/CR-2239, December 1982.

16. Electric Power Research Institute (EPRI), " Review of Proposed Dry-Storage Concepts Using Pmbabilistic Risk Assessment," EPRI NP-3365, February 1984.

17. U.S. Nuclear Regulatory Commission (USNRC), " Seismic Failure and Cask Drop Analyses of the Spent Fuel Pools at Two Representative Nuclear Power Plants,"

NUREG/CR-5176, January 1989.

18. Electric Power Research Institute (EPRI), "The TP-24P PWR Spent-Fuel Storage Cask:

Testing and Analysis," EPRI NP-5128, April 1987.

19. Electric Power Research Institute (EPRI), "The MC-10 PWR Spent-Fuel Storage Cask:

Testing and Analysis," EPRI NP-5268, July 1987.

20. Docket Nos. 50-280,50-281. Augmented Inspection Team Reports Nos. 50-280/88-34 and 50-281/88-34, dated September 30,198S.

21. U.S. Nuclear Regulatory Commission (USNRC), Inspection and Enforcement Bulletin 84-03, " Refueling Cavity Water Seals," dated August 24,1984.

22. Electric Power Research Institute (EPRI), " Fuel and Pool Component Performance in Storage Pools," EPRI NP-4561, May 1986.

23. U.S. Nuclear Regulatory Commission (USNRC), " Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications - Final Report,"

NUREG/CR-1278, August 1983.

24. U.S. Nuclear Regulatory Commission (USNRC), Inspection and Enforcement Information Notice 84-93, " Potential for Loss of Water From the Refueling Cavity,"

dated December 17,1984.

25. U.S. Nuclear Regulatory Commission (USNRC), Inspection and Enforcement Information Notice 88-65, " Inadvertent Drainages of Spent Fuel Pools,"

dated August 18,1988.

26. U.S. Nuclear Regulatory Commission (USNRC), Inspection and Enforcement Information Notice 88-92," Potential for Spent Fuel Pool Draindown,"

dated November 22,1988.

27. U.S. Nuclear Regulatory Commission (USNRC), "A Prioritization of Generic Safety Issues," NUREG-0933, pp. 3.82-1 to 3.82-6, December 1983.

8-2 l

-c . .

28. U.S. Nuclear Regulatory Commission (USNRC), "Modeling Time to Recovery and Initiating Event Frequency for Loss of Off-Site Power Incidents at Nuclear Power Plants," NUREG/CR-5032, January 1988.

29. American Society of Civil Engineers, " Uncertainty and Conservatism in the Seismic Analysis and Design of Nuclear Facilities," 1986.

30. Docket No. 50-271, Vermont Yankee Nuclear Power Corporation, " Vermont Yankee Spent Fuel Storage Rack Replacement Report," April 1986,

31. U.S. Nuclear Regulatory Commission (USNRC), " Evaluation of External Hazards to the Nuclear Power Plants in the United States - Seismic Hazard," NUREG/CR-5021, Supplement 1, April 1988.

32. U.S. Nuclear Regulatory Commission (USNRC), Bciard Notification 87-011, " Board Notification Regarding Anomalies in Boraflex Neutron Absorbing Material (BN 87-11),"

dated June 15,1987.

33. U.S. Nuclear Regulatory Commission (USNRC), Inspection and Enforcement Information Notice 87-43," Gaps in Neutron Absorbing material In High Density Spent Fuel Storage Racks," dated September 8,1987.

34. U.S. Nuclear Regulatory Commission (USNRC), Inspection and Enforcement Information Notice 87-13, " Potential for High Radiation Fields Following Loss of Water From Fuel Pool," dated February 241987.

35. Memorandum from W. Minners to K. Kniel, " Refueling Cavity Seal Failure,"

dated April 1,1986. PDR Accession No. 8604080427.

36. U.S. Nuclear Regulatory Commission (USNRC), "A Handbook for Value-Impact Assessment," NUREG/CR-3568, December 1983

37. Electric Power Research Institute (EPRI), " Cost Comparisons for On-Site Spent Fuel Storage Options," EPRI NP-3380, May 1984.

38. U.S. Nuclear Regulatory Commission (USNRC), "A Handbook for Quick Cost Estimates," NUREG/CR-4568, April 1986.

39. U.S. Nuclear Regulatory Commission (USNRC), " Generic Cost Estimates,"

NUREG/CR-4627, Revision 1, February 1989.

40. U.S. Nuclear Regulatory Commission (USNRC), " Regulatory Analysis Guidelines of the U.S. Nuclear Regulatory Commission," NUREG/BR-0058, Revision 1, May 1984.

41. U.S. Department of Energy (DOE), " Spent Fuel Storage Requirennents 1987,"

DOE /RL-87-11, September 1987.

42. U.S. Nuclear Regulatory Commission (USNRC), " Reactor Risk Reference Document,"

NUREG-1150, Draft for Comment, February 1987. PDR Accession No. 8703100040 (Vol. ), 8703170207 (Vol. 2), and 8703180080 (Vo'. 3).

8-3 l

43. U.S. Nuclear Regulatory Commission (USNRC), " Evaluation of Severe Accident Risks and the Potential for Risk Reduction: Surry Power Station, Unit 1," NUREG/CR-4451, Draft For Comment, February 1987. PDR Accession No. 8703170262.

44. U.S. Nuclear Regulatory Commission (USNRC), "Backfit Rule",10 CFR 50.109 (a)(3),

Federn1 Repster, Vol. 50, p. 38097, September 20,1985.

45. U.S. Nuclear Regulatory Commission (Uf NRC), ," Safety Goals for the Operation of Nuclear Power Plants: Policy Statement, ' Federal Register, Vol. 51, pp. 28044-49, August 4,1986.

46. U.S. Department of Commerce, " Statistical Abstract of the United States,1987," 107-th Edition, Bumau of Census.

47. U.S. Nuclear Regulatory Commission (USNRC), " Demographic Statistics Pertaining to Nuclear Power Reactor Sites," NUREG-0348, October 1979.

48. Memorandum from G. Bag' chi to K. Kniel, " Comments on Draft Resolution for Generic Issue 82 - Beyond Design Basis Accidents in Spent Fuel Pools," dated December 29,1988.

J l

I 8-4 i

j l

1

---

__._____________i

~

l .

l Appendix A Spent Fuel Data and Storage Requirements The Department of Energy (DOE), Office of Civilian Radioactive Waste Management (OCRWM) is responsible for the management and ultimate permanent disposal of the civilian spent fuel and high level radioactive waste generated as a result of commercial nuclear power plant operations in the U.S.

This responsibility is prescribed under the provisions of the Nuclear Waste Policy Act of 1982 (NWPA) (as ammended).

The greatest portion of the radioactive waste covered under this government responsibility will be spent nuclear fuel discharged from commercial nuclear power plants. Because most of the spent fuel that will ultimately require disposal has not yet been generated, planning for the management and disposal of this spent fuel must be largely based on projections of future spent fuel discharges from commercial nuclear power plants.

The OCRWM plans for ranagement and disposal of spent fuel are based on the DOE Energy Information Administration (EIA) nuclear energy projections. These data are used for this Regulatory Analysis, in support of resolution of Generic Issue 82, "Beyond design Basis Accidents in Spent Fuel Pool," to estimate the additional cost of at-reactor storage of spent fuel.

The data source is DOE /RL-87'-ll, " Spent Fuel Storage Requirements 1987,"

September 1987, United States Department of Energy, Richland Operations Office. Data used by the NRC, taken from this report, are provided in Tables A.1 through A.4 of this Appendix.

A-1

e 8+ ..

o TAala A.1. suolanN beWor Plast Data massaN anaC usT stanver suPPoomN panssNT(a) anz. tal manovak come s23s Psamt ansa WPELITY maan sTass TYPa (astE) ansa ansB CmPAC. CataC. vamBcm (a) surzet ARE NUCLEAR 1 ARE PNR 6 LCT CO. AR PWR 8$0 1974 2000 960 968 BW 177 82 ARE NUCLEAR 2 1 ARE PNR & LCT CO. AR PER 912 1980 2012 900 988 CE 177 14 BEAVER VALLEY 1 DUQUESNE LIGHT COMPANY PA PNR 835 1976 2016 833 433 NE 157 73 BEAVER VALLEY 2 DUQUESNE LIGHT COMPANY PA PNR 957 1907 2026 1008 1988 NE 157 72

- BELLEFDNTE 1 - TENNESSEE VALLEY AUTHORITY

• AL PNR 1235 1992(f) 2028 1054 1050 BN 205 93 BELLEFONTE 2 TENNESSEE VALLEY AUTHORITY AL PNR 1235 199Stfl 2030 270 1950 BN 205 93 BIG ROCK 1 CONSUMERS PWR CO. MI BNR 72 1965 2001 441 441 GE 84 11 A BRAIDN000 1 COMMOWNBALTH EDISON COMPANY IL PWR 1178 1987 2026 1950 1050 NE 193 02 A BRAIDWOOD 2 COMMONNEALTN ED350N COMPANY IL PNR 1175 1988 2027 0 0 NE 193 42 B BRONNS FERRY 1 TENNESSEE VALLEY AUTHORITY. AL BNR 1065 1974 2014 3471 3471 GE 764 140 B BROWNM FERRY 2 TENNESSEE VALLEY AUTHORITY AL BNR 1065 1975 2014 3133 3471 GE 764 140 BRONNS FERRY 3 TENNESSEE VALLEY AUTHORITY AL BNR 1068 1977 2017 2353 3471 GE 164 140 BRUNSNICE 1 CAR 3 LINA POWER 6 LIGHT COMPANY NC BNR(c) 821 1977 2010 1767 1903 GE 560 103 BRUNSWICE 2 CAROLINA PONER 6 LIGHT COMPANY NC BNR(c) 821 1975 2010 1325 1839 CE 560 102 A BYRON 1 N'^'"*1TK EDISON COMPANY IL PNR 1120 1985 2024 1050 1050 NE 193 82 A BYRON 2 COMMONWEALTH EDISON COMPANY IL PNR 1120 1987 2026 0 0 NE 193 82 CALLANAY 1 UNION ELEC COMPANY MO PNR 1171 1984 2024 1340 1340 NE 193 86 B CALVERT CLF 1 BALTIMORK GAS 6 ELEC CD. MD PNR 845 1975 2014 030 430 CE 217 84 B CALVERT CLP 2 BALT190RE GAS 4 ELEC CO. MD PWR 845 1971 2016 1000 1000 CE 217 44 CATANBA 1 DUKE PONER COMPANY SC PWR 1145 1985 2025 1419 2615 NE 193 09 CATANBA 2 DUKE POWER COMPANY SC PNR 1145 1986 2026 1421 2615 NE 193 82 CLINTON 1 ILLINOIS PNR CO. IL BNR 933 1981(f) 2027 2672 2672 GE 624 114 8 COMANCHE PE 1 TEXA8 UTILITIES GENERATING CO. TX PNR 1150 1989 2030 260 1695 NE 193 39 8 COMANCHE PE 2 TEXAS VT!LITIES GENERATING CO. TX PWR 1180 1909 2030 0 1607 NE 193 03 A CODE 1 INDIANA 6 MICH ELEC CD. MI PNR 1030 1975 2009 2040 2270 NE 193 et A CODE 2 INDIANA 6 MICH ELEC CD. MI PNR 1100 1910 2009 0 0 WE 193 10 -

C00FF.R STN WEBRASEA PUB PNR DISTRICT NE BNR 770 1974 2008 2366 2366 GE $44 101 CRYSTAL RVR 3 . FIDRIDA PWR CORP FL PNR 82$ 1977 2016 416 1157 DN 177 82 DAVIS-BESSE 1 TOLEDO EDISON 00. OH PNR 906 1978 2017 735 735 BW 177 83 DIABID CANYON 1 PACIFIC GAS AND ELP.CTRIC CO. CA PNR 1006 1985 2025 270 1324 WE 193 89 DIABID CANYON 2 PACIFIC GAS AND ELECTRIC CO. CA PNR 1119 1986 2025 270 1324 - NE 193 89 DRESDEN 1 N'^"**!TS EDISON COMPANY IL bNR 200 1960 1984 120 720 CE 464 47 DRESDEN 2 COB 9ENNEALTH EDISON COMPANY IL BNR 794 1970 2008 3531 3537 GE 724 125 DRESDEN 3 "'^""**lTH EDISON COMPANY IL BNR 794 1971 2006 3537 35n GE 724 126 DUANE ARNOLD IONA ELEC LIGHT 6 POWER CD. 3A BNR 534 1975 2010 20$0 20$0 GE 368 67 ENRICO FERMI 2 DETROIT EDISON COMPANY. M! BNR 1093 1987(f) 2025 230$ 2305 CE 764 140 FARLEY 1 AIABAMA PONER COMPANY AL PNR 829 1977 2012 1407 1407 NE 157 73 FARLEY 2 ALARMA PONER COMPANY AL PNR 829 1981 2012 1407 1407 NE 157 13 FIT 1PATRICE PNR AUTHOR 3TY OF STATE OF NY NY BNR 821 1975 2015 2244 2454 CE 560 103 FORT CALHOUN OMAHA PUE PNR DIST NE PNR 486 1973 2006 729 729 CE 133 47 FT ST VRAIN PUB SVC CO OF COIDRADO CO HTG 330 1979 2007 504 504 GA 1442 16 GINNA ROCHESTER GAS 8 ELEC CORP NY PNR 490 1970 2P06 1016 1016 NE 121 43 GRAND GULF 1 SYSTEM ENERGY RESOURCES,1NC. MS BNR 1250 1985 2022 3124 3124 GE SCO 145 MADDAM KECK NORTHEAST UTIL2 TIES CT PWR 542 1968 2007 1168 1160 NE 157 64 MARRIS 1 CAROLINA PONER 6 LIGHT COMPANY NC PNR 940 1987(f) 2026 480 3351 NE 157 73 8 HATCH 1 GEORGIA PWR COMPANY CA BNR 777 1974 2009 3025 3181 GE $60 103 B HATCH 2 GEORGIA PNR COMPANY GA BNR 70. 1579 2012 2765 2045 GE $60 104 HOPE CREEK PUBLIC SERV. ELEC AND GAS CD. NJ BNR 1110 1987(f) 2026 1078 3976 GE 764 141 HUMBOLDT BAY PACIFIC GAS AND ELECTRIC CD. CA BNR 65 1963 1976 406 486 GE 104 13 INDIAN PT 1 CONSOLIDATED EDISON CO. NY PNR 265 1962 1980 756 756 BW 120 23 INDIAN PT 2 CONSOLIDATED EDISON CO. NY PNR 873 1974 2006 980 980 NE 193 $$

INDIAN PT 3 PNR AUTHORITY OF STATE OF NY NY PNR 965 1976 2015 840 1317 NE 193 09 EENAUNER WISCONSIN PUBLIC SERVICE CORP N1 PNR 533 1974 2014 603 963 NE 121 46 1ACROSSE DAIRY 1AND PWR COOP NI BNR 50 1969 2002 440 440 AC 72 0 B LASALLE CTY 1 COMMONWEALTH EDISON COMPANY IL BNR 1122 1982 2022 1080 1080 CE 764 140 B LASALLE CTY 2 COfe40NNEALTH EDISON COMPANY IL BWR 1122 1984 2023 1000 1000 GE 764 140 LIMERICE 1 PKf1ADEl#KIA ELEC CO. PA BNR 10$$ 1986 2024 2040 2040 GE 764 141 LIMERICE 2 PH11ADELPHIA ELEC CO, PA BNR 105$ 1990 2029 2040 2040 CE 764 140 MAINE YANKEE MAINE YANKEE ATOMIC PNR CD. ME PNR 825 1972 2008 1416 1476 CE 211 80 MCGUIRE 1 DUKE PONER COMPANY NC PWR 1140 1981 2021 1359 1463 NE 193 09 MCGUIRE 2 DUKE POWER COMPANY NC PNR 1100 1984 2023 1421 1463 NE 193 89 A-2 L

l

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

r;

.f.ee e b c . . ,

4 lL  : O-j; .' f l_.

sAa!,a 1.1. sealenE PNueN sleet Bees ses*E mas:sN haRc me erAn9st aswrBoiar Passawetal mRx.ta) BaRosen aces slas PsANr mRess WTILITY unas erAss rYPs (asun mRes nRsu CAPAC. CRPAC. vimman (a) aartet MILLSTONE 1 NORTHEAST UTIL SVC CD. CT BNR 660 1970 2010 2184 2104 GE $80 103 MILLSTONE 2 NORTHEAST UTIL SVC CO. CT PWR 870 1978 2015 1112 1112 CE 217 88 MILLSTONE 3. NORTHEAST UTIL SVC CJ. CT PNR 1150 1986 2028 156 1836 NE 193 89 MONT1CEL14 NORTHERN STATE 3 PWR COMPANY MN BNR $45 1971 2001 2211 2231 GE 484 06 l' NINE MILE PT 1 N!AGARA 383HANE PONER CORP NY BNR 620 1969 2005 2362 2776 GE $32 94

! . NINE MILE PT 2 NIAGARA MOHANE PCNER CORP NY BNR 1980 1987 2026 2530 4049 GE 164 140 l- A NORTH ANNA 1 VIRGINIA PONER VA PNR 907 1970 2018 1737 1737 NE 167 72 A NORTH ANNA 2 VIRGINIA POWER VA PMR 907 1980 2020 0 0 NE 157 73 A OCONEE 1 DUES PONER COMPANY sc PNR 887 1973 2013 1298 1312 BN 177 82 A OCONEE 2 DUKE POWER COMPANY SC PNR 887 1974 2013 0 f BN 177 02 OCONEE 3 DURE POWER COMPANY SC PNR 886 1974 2014 818 825 BW 171 82 OYSTER CRE 1 GPU NUCLEAR NJ BNR 650 1949 2004 2600 2600 GE 560 to PALISADES ' CONSUMERS PNR CO. MI PNR 008 1971 2011 798 798 CE 204 80 PA14 VERDE 1 AA18DNA PUBLIC SERY1CE CD. AS - PWR 1270 1906 2024 665 1329 ' CE 241 99 PA10 VERDE 2 ARIEONA PUBLIC SERVICE CO. As PWR 1270 1986 2023 665 1329 CE I41 99 PAID VERDE 3 ARISONA PUBLIC SERVICE CO. A8 PWR 1270 1987 2026 665 1329 CE 241 99 PEACHB01 TOM 2 PHILADELPHIA ELEC CO. PA BNR 1065 1974 2008 3014 3814 ' GE - 764 140 PP.ACMBOTTOM 3 PHILADEIJHIA ELEC CD. PA BNR 1965 1974 2000 3019 3819 GE 164 140 PERRY 1 CLEVELAND ELEC ILLUM CO. OH BNR 126$ 1987 2026 4020 4020 CE 748 130 FILGRIM 1 BOSTON EDISON CO. MA MNR 6$$ 1972 2008 2320 2320 GE 580 103 A POINT BJACH 1 NISCONSIN ELEC PWR CO. NI PWR 497 1970 2007 1502 1502 NE 121 46 A POINT BP.ACH 2 KISCONSIN ELEC PWR CO. M1 PNR 497 1972 2000 0 0 NE 121 46 A PRAIRIE ISL 1 NORTHERN STATES PWR CO. MM PWR 530 1973 2008 1386 1386 NE 121 44 A PRAIRIE 13L 2 NORTHERN STATES PWR CO. NN PNR 530 1974 2004 0 0 NE 121 40 B QUAD CITIES 1 ' " a M TN EDISON COMPANY IL BNR 789 1973 2007 3687 3657 GE 124 129 8 OUAD CITIES 2 COMMONWEALTH EDISON COMPANY IL BNR 749 1973 2007 3897 3897 GE 724 126 RANCHO SECO 1 SACRAMENTO MUNICIP UTIL DISTR CA PWR 910 1975 2000 1980 1080 BW 177 42 ROEIN3ON 2 CAROLINA PONER 4 LIGHT COMPANY SC PNR 100 1971 2007 $44 $44 NE 187 66 RVR BEND 1 Gulf STATES UTILITIES LA BNR 936 1986 2025 3172 3172 GE 624 114 SALEM 1 PUBLIC SERV. ELEC. AND GAB CO. NJ PNR 111$ 1977 2016 1133 1170 NE 193 89 8A1,EM 2 PUBLIC SERV. ELEC. MD GAS CO. NJ PWR lill 1901 2020 1140 1170 NE 193 89 E SAN ONOFRE 1 SOUTHERN CALIF ED1 SON CD. CA PNR 436 1968 1999 216 216 NE 157 $8 E EAN ONOFRE 2 SOUTHERN CALIF EDISON CO. CA PNA 1970 1983 2012 800 800 CE 217 91 E AAN ONOFRE 3 SOUTHEM CALIF EDISON CD. CA PNR 1080 1984 2013 000 800 CE 217 90 SEABROOE 1 NHY DIVISION OF P3NH NH PWR 1150 1987(f1 2031 660 1236 NE 193 09 A REQUOYAN 1 TENNES$1E VALLEY AUTHORITY TN PWR 1148 1981 2021 1381 1301 NE 193 89 A SEQUOYAN 2 TENNE 3SEE VALLEY AUTHORITY TN PNR 1148 1982 2022 0 0 NE 193 89 SHOREHAM 3DNG I5L LGT CO. NY BNR 849 1998 (b) 2027 2176 2685 GE 560 102 SOUTR TEIA$ 1 HOUSTON LIGHTING e PONER CO. TI PWR 1250 1987 2027 196 1969 NE 193 105 SOUTH TEIAS 2 HOUSTON LIGHTING 6 POWER CO. TX PWR 1250 1989 2028 0 1969 NE 193 104 ST LUCIE 1 FIDRIDA PWR & LGT CD. FL PNR 430 1976 2010 728 728 ' CE 217 81 87 LUCIE 2 FIDRIDA PWR 4 LGT CO. FL PNR 804 1983 2023 1976 1016 CE 217 81 SUMMER 1 SOUTH CAROLINA ELEC 4 GA5 CD. SC PNR 900 1984 2024 1276 1276 NE 157 72 A SURRY 1 VIRGINIA POWER VA PWR 784 1972 2012 1944 1044 WE 187 72 A SURRY 2 VIRGINIA POWER ,

VA PNR 788 1973 2013 1764 1764(es NE 187 72 B SUSQUEHANNA 1 PENNSYLVANIA PWR 6 LGT CO. PA BNR 1065 1983 2022 2840 2440 GE 164 137 8 SUAQUEhANNA 2. PENNSYLVANIA PWR 4 LGT CO. PA BWR 106S Ital 2024 2840 2440 GE 764 137 THREE MILE ISL 1GPU NUCIZAR PA PNR 919 1974 2004 752 1401(63 BW 177 82 TROJAN PORT 1AND GENERAL E!2C OR PNR 1130 1976 2011 1400 1400 NE 193 89 E TURREY PT 3 F10AIDA PWR 6 LGT CO. FL PWR 693 1972 2 0.'? 1376 1404 NE 157 12 E TUREET PT 4 FIDRIDA PWR 6 LCT CO. FL PNR 693 1973 2007 614 636 NE 167 72 B VOCTLE 1 GEORGIA POWER COMPANY CA PNR 1069 1987 2027 tot 1117(g) NE 193 49 8 VOGTLE 2 GEORGIA PUNER COMPANY GA PNR 1069 1988 2020 280 1111(g) NE 193 89 VT TANIEE 1 VT YANKEE NUC1&AR PNR CORP VT BWR 514 1972 2012 1690 2470 GE 360 68 NASH NUC14AR 2 NAEH PUB PNR SUPPLY SYSTEM NA BNR 1100 1984 2023 2658 2658 GE 164 140 NATERPORD 3 1DU!sIANA POWER 6 LIGHT 1A PWR 1104 1985 2024 1C88 1366 CE 217 09 A NATT5 BAR 1 TENNESSEE VALLEY AUTHORITY TN PWR 1165 1999(ft 2025 1294 1294 WE 193 89 A NATT5 BAR 2 TENNESSEE VALLEY AUTHORITY TN PWR 1165 1990 (f) 2027 0 0 NE 193 89 WOLF CREEK 1 WOLF CREEE NUCLEAR OPERATING CO. ES PNR 1150 1905 2025 1327 1340 NE 193 89 YANKEE

• ROME 1 YANKEE ATOMIC ELEC CD. MA PNR 175 1961 2001 440 721 NE 16 18 A EION 1 COMMONWEALTH EDISON COMPANY IL PNR 100$ 1973 2008 2079 2019 WE 193 89 A EION 2 COMMONWEALTH EDISON COMPANY IL PWR 1085 1974 2006 0 0 NE 193 89

~

A INDICATES COMMON POOL SHARED BY TWO REACTORS B INDICATEA POOLS CONNECTED BY TRANSFER CANALS CAPACITIES AND INVENTORIES ARE COMBINED N!?H ONLY ONE FULL CORE RESERVE E INDICATES POOLS REQUIR1NG CASE TRANSFERe CAPACITIES AND INVENTORIES ARE COMBINED WITH ONLY ONE FULL CORE RESERVE A-3

I' 4

. se e.

e v .e tamla A.S. smalenN pas N Plant nota (een't)

TYPs pass stamTur esproOus Paasmetta) askE. ta)

- stoma n sivas w1L1w maast swans stammD nass nars Capac, anPac.

t:; <

Ic) BAUNSN!CE 1 PNR CAROLINA POWER 4 LIGHT COMPANY NC PWR 1977 2010 160 160 (c) BRUNSNiCE 2 PWR- CAROLINA POWER 4 LIGHT COMPANY WC PWR . 1978 2010 144 144 COE ID ,(INEL) (EG4G) DEPARTHENT OF ENERGY 1D BWR (h) (h)

UOE ID (INEL) (EC&G) DEPARTMENT OF ENERGY 3D HTS (h) (h) erPs FUBL staATW SIBUTDCIEF PREsENT(4) SRE.(a) esonass szens w12.1er sauB seams secamD nass ansa cnPac. onPac.

DOE 2D (INELI (EG40) DEPARTMENT OF ENERGY ID PWR (h) (h)

DOE OH (BATTELLE) DEPARTMENT OF ENERGY OH PWR (h) '(h)

DOE WA (HANFORD) DEPARTMENT OF ENER'!Y MA BNR (h) (h)

DOE NA (HANFORD) DEPARTMENT OF ENERGY MA PNR (h) (h) id) HARRIS 1 BWR POOL CAROLINA POWER 4 LIGHT COMPANY NC BWR 1957 2026 0 2057lg)

MORRIS-BNR/PNR IIORRIS OPERATION (AFR) IL BWR 2002 3735(1) 3775(1)

MORRIS *BNk/PNR .SIORRIS OPERATION (AFR) IL PNR 2002 1660(1) 1660(1)

NEST VALLEY NEST VAL 12Y DEMONSTRATION PRJ. NY BNR (h) (h)

NEST VALI2Y NEST VAL 12Y DEMONSTRATION PRJ. NY PNR (h) (h)

(j) OTHER PNR th) (h)

BNR PLANTS TOTAL 39 CURRENTLY OPERATINGs 32 TOTAL MNE: 33708 CVRRENTLY OPERATING MNEs 26312 PNR PLANTS TOTAL s 90 CURRENTLY OPERATINGs 63 TOTAL MNEs 16800 CURRENTLY OPERATING MNEs $6378 HTGR FIANTS TOTAL e 1 CURR8NTLY OPERATINGs 1 TOTAL MNEs 330 CURRENTLY OPERATING MNE 330 OPERATING A PLMNED PLANTS TOTAL : 120 CURRENTLY OPERATING: 96 TOTAL IINEs 199835 CURFENTLY OPERATING sNIEs 63017 RETIRED P! ANTS TOTAL s 3 TOTAL aINEs $30 NOTE: UTILITY DATA AS OF 12/31/1986 (a) IN AASEMBLIES (D) SHOREHAM ISSUED A LICENSE IN 1983 BUT HAS NOT OPERATED. 1984 STARTUP EST! MATED BASED ON PROJECTED FIRST DISCHARGE IN 1909.

i (c) SOME ROBINSON 2 PNR FURL 18 STORED AT THE BRUNSNICE (BNR) REACTORS.

(d) IN 1985, HARRIS 1 IDENTIF3RD SPACE POR THE FUTURE STORAGE OF BNR FUEL. (HARRIS 1 IS A PNR.)

(e) INCLUDES STORAGE CAPACITY OF DRY STORAGE INSTALLATION (ISFS1) ,

(f) STARTUP DATE BASED ON PROJECTED YEAR OF FIRST DISCHARGE.

(g) CURRENT AS OF 12/31/86 (b) CAPACITY FOR STORAG3 UNENONW, (1) POOL CAN HOLD BOTI( FUEL TYPES. CAPACITY Sh0NN REFLECTS ENT!kE POOL IN USS FOR OWE TYPE OF FUIL ONLY.

(j) ONE ROBINSON ASSIMBLY HAS BEEN SENT TO A 3DCAT10N WEICH DOES NOT HAVE AN E1A 1D.

A-4

4

?2 % , ,

I<fep010i4 QL. -., 9 t

.., e, SRBla 'A.S. Pre $estm6 cmulative SEmeage AegsLeenemEs--senstaan AR naranMy, Assamh11es poet. - 1987 1980 1989 1990 1991 1998 1993 1994 1998 1996 1997 1998 1999 2000 8001 2002 3003 3004 3005 sf LUCIE 1, P WR 42 122 122_ 194! 270 270 346 422 422 49e $14 514 .650 726 726 602 078 818 954 -

MILLSTONE 1 BNR 124 128 324 324 $20 120 716 716 912 912 lies 11001304130415001500169616941892 PALISADES P NR 0 19 '19 87 87 155 155 223 291 291 359 359 421 -427 495 563 563 631 631 OCDNEE 142 PWR 0 0 42 102 242 302 422 482 602 662 782 842 962 1022 lost 1202 1322 1322 1442 l OCONEE 3 - PWR 0- 0 .1- 1- 61 121 121 101 241 - 241 301 361 421 421 481 541 541 601 661

' ROBINSON 2 P NR 0 0 0 - 40 et 149 149 197 245 245 306 354 402 402 463 511 511 $59 620 BRUN8 NICK 1 DNR 0 0 0 157 157 345 345 533 717 717 905 1093 1993 1277 1277 1465 1653 1653 1837 LASALLE CTY 162 DNR L 0 .0 0 144 364 504 1924 1244 1464 It)4.2124 2344 2184 3004 3224 3664 3844 4104 4544

. BRVNSNICK 2 DNR 0 0 0 0 37 31 225 413 413 $97 785 705 973 973 1157 1345 1345 1533 1717 CALVERT CLP 142 Pra 0 0 0 0 0 5 101 197 293 389 485 501- 677 773 869 -965 1961 1157 1253 LACROSSE - DNR -0 0 0 0 0 13 31 61 45 05 109 133 157 141 205 205 205 205 205 PILGRIM 1 BNR 0 0 0 0 0- 0 160 160 356 356 548 548 140 740 140 936 936 1128 1128 PkAIRIE 13L 142 PNR 0 0 0 0 0 0 . 22 102 182 222 302 382 462 542 542 662 742 822 902' BYRON 142 PWR 0- 0 0 0 0 0 0 99 99 267 435 435 603 771 771 939 1101 1107 1275 INDIAN F7 2 . PNR 0 0 0 0 0 0 0 17 17 85 153 153 221 221 209 357 357 425 425 OYSTER CRE 1 DNR 0 0 0- 0 0 0 0 68 68 200 348 348 488 628 6tt 768 912 912 .912 FORT CALKOUN P NR 0 0 0 0 0 0 0 8 e 53 . De to 143 1st les 233 278 278 323-EION 1&2 , PWR - 0 0 0, 0 0 0 0 0 71 143 207 359 431 575 647 719 863 935 1001 BIG ROCK 1 BNR 0 0 0 0 0 0 0 0 15 35 35 55 75 95 95 95 95 95 95 LIMERICE 1 BWR 0 0' 0 0 0 0 0 0 152 152 372 592 592 012 1032 1032 1240 1456 1456 SAN DNOFRE 1,2,43PWR 0- 0 0 0 0 0 0 0 139 300' 409 570 679 945 1054 1163 1272 1381 1490 SEQUOYAH 142 PNR 0' 0 0 0. 0 0 0 0 40 40 200 2s0 360 520 600 600 160 920 1000 DAVIS-BEssE 1 PWR 0 0 0 0 0 0 0 0 9 9 70 130 130 190 250 250 310 370' 370 POINT BEACH 152 P WR 0 0 0 0 0 0 0 0 49 113 177 241 305 369 433 497 561 625 649 ARE NUCLEAR 1. P NR '- d 0' O O 0 0 0 0 0 17 17 17 71 137 197 191 257 257 317 BRAIDN005 142 PNR 0 0 0 0 .0 0 0 0 0 99 99 267 435 435 603 771 771 939 1107' BEAVER VALLEY 1 PNR 0 0 0 0 0 0 0 0' .0 25 94 94 163 163. 232 132 301 370 370 KAINE YANKEE - PNR 0 0 0 0 0 0 0 0 0 48 121 121 194 267 267 340 413 413 ' 486 NINE MILE PT1 DNR - 0 0 0 0 0 0 0 0 0 152 152 332 332 516 516 700 700 484 484 KADDAM NECK' PWR 0 0 0 0 8 0 0 0 0 0 3 3 56 104 161 161 213 266 266 EKRICO PERMI2 DNR 0 0 0 0 0 0 0 0 0 0 135 135 427 427 115 715 1007 1295 1295 COOPER STN DNR 0 0 0 0 0 0 0 0 0 0 6 122 234 346 458 570 682 194 902 KILLaf0NE 2 PWR 0 0 0 0 0 0 0 0 0 0 $1 116 116 171 242 242 307 36e 36e PEACRBOTTON 2' DNR 0 0 0 0 0 0 0 0 0 0 216 216 444 672 672 900 1828 1128 1356 PEACKBOTTOM 3 DNR 0 0 0 0 0 0 0 0 0 0 41 41 261 est 451 701 701 921 1141 FITEPATRICK DNR 0 0 0 0 0 0 0 0 0 0 134 134 110 486 406 662 830 838 1014

$ALEM 1 PWR 0 0 0 0 0 0 0 0 0 0 47 47 127 207 207 207 367 367 447 DRESDEN 2 DNR 0 0 0 0 0 0 0 0 0 0 0 67 67 225 383 383 541 699 699 CODE 142 PWR 0 0 0 0 0 0 0 0 0 0 0 61 149 317 397 445 565 653 821 GRAND GVLF 1 . SKR 0 0 0 0 0 0 0 0 0 0 0 104 372 372 640 900 900 1176 1444 LINERICK 2 DNR 0 0 0 0 0 0 0 0 0 0 0 132 132 348 564 564 700 700 996 NASH KUCLEAR2 DNR 0 0 0 0 0 0 0 0 0 0 0 110 286 426 582 750 898 1066 1210 ARK NUCLEAR 2 PNR 0 0 0 0 0 @ 0 0 0 0 0 0 21 89 49 151 225 225 293 DUANE ARNOLD DNR 0 0 0 0 0 0 0 0 0 0 0 0 118 246 246 366 486 406 614 VT YANREE 1 BKR 0 0 0 0 0 0 0 0 0 0 0 0 12 12 144 276 276 4be 540 NORTH ANNA 142 PNR 0 0 0 0 0 0 0 0 0 0 0 0 119 183 247 375 439 503 631 KENAVNEE PNR 0 0 0 0 0 0 0 0 0 0 0 0 24 61 90 135 172 209 246 YANKEE-RONE 1 PWR 0 0 0 0 0 0. 0 0 0 0 0 0 36 36 36 36 36 36 36 DRE3 DEN 3 DKR 0 0 0 0 0 0 0 0 0 0 0 0 0 61 219 219 377 535 535 KATCH 142 DNR 0 0 0 0 0 0 0 0 0 0 0 0 0 130 326 718 914 1110 1502 SUSQUEKANNA 162 DNR 0 0 0 0 0 0 0 0 0 0 0 0 0 96 560 792 1024 1400 1720 GINNA P rR C 0 0 0 0 0 0 0 0 0 0 0 0 27 59 91 123 155 187 ST LUCIE 2 PER 0 0 0 0 0 0 0 0 0 0 0 0 0 0 46 114 118 190 262 SALEM 2 Fwn 0 0 0 0 0 0 0 0 0 0 0 0 0 0 71 155 155 239 323 BROWN 3 FERRY 3 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 161 161 389 617 617 NATTS Bl.R 162 PWR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 $1 211 til 311 451 TURKEY PT 3s4 l #R 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4e 144 192 240 BROWNS PERRY 1&2 DNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 34 34 490 718 TROJA4 PKR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 37 85 1KDIAN PT 3 PKR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 80 RANCHO SECD 1 PWR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 43 104 PALO VERDE I P NR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12 QUAD CITIES 162 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 23 MCGUIRE 1 PWR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 29 CRYSTAL avR 3 PNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 72 CALLAWAY 1 PNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 37 PNR ASSEMBLIES 42 204 740 1316 2704 5370 0390 11933 15943 20312 141 504 1002 1924 3747 6505 10299 14125 17848 DNR AssEMBLILS 123 324 1978 2507 4182 7018 11301 17011 23649 30996 12e 625 1499 3195 5110 4399 13858 20429 27497 TOTAL ASSEMBLIES 170 $20 1826 3823 6890 12380 19591 20944 39592 51304 269 1129 2501 5123 4557 14904 24157 34554 45345 A-5

7-- -. - - - - - - ,- - _ - _ - - - -- --

. 9' eo-

((

i s-hr e' Skata A.3. Projected w **4se Storage DagmLeameNEs ^^ AR capaakty, 8Erts POOL 1987 1980 1989 1990 1991 1998 1993 1994 1998 1996 1997 1990 1999 3000 2001 2003 3003 2004 2006 ST LUCIE 1 PNR - 15 46 46 73 102 102 131 160 150 les 218 21e 247 276 276 305 335 335 364 MILL 870KE 1 - BNR 23 ' 23 50 58 92 92 127 127 162 162 197 197 231 231' 266 266 301 301 336 FAL2sADEs PNR 0 1 7 34 61 61 to 115 115 142 142' 170 170 197 224 224 251 251 OCONBE 142 P KR 0 0 29 ' 84 lit 140 195 223 279 307 362 390 445 473 501 557 612 612 668 DCONEE 3 PWR 0 0 0, 0 as ' 56 54 84 112 112 139 161 195 195 223 250 250 278 306 ROBINSON 2 PWR 0 0 0 17 38 63 63 84 105 105 130 151 172 172 197 218 218 238 243 3 l- BRUNSWICE 1 ' BNR 0 0 0 29 29 64 64 99 134 134 169 204 204 230 238 274 309 309 343 LA&ALLE CTY 162 BNR 0 0 0 26 66 106 196 226 266 347 347 427 507 547 547 667 707 147 027 BRUN5 NICE 2 DNR 0 0 -0 0  ?  ? 42 77 77 112 147 147 le2 162 216 251 251 297 321 CALVERT CLP 142 P WR 0 0 0 0 0 3 30 74 110 145 181 213 249 245 321 357 393 429 465 LACROSSE BNR 0 0 0 0 0 1 4 7 9 9 L12 14 17 20 22 22 22 22 22

. P2LORIM 1 BNR 0 0 0 0 0 0 28 26 63 63 97 97 131 131 131 165 165 199 199 PRAIRIE 1sL 1st PNR 0 0 0 0 0 0 8 36 65 79 '107 135 164 192 206 235 263 291 320 BYRON 342 PNR 'O O 0 0 0 0 0 42 42 113 184 184 255 326 325 391 468 468 539 INDIAN PT 2 PWR 0 0 0 0 0 0 0 4 8 38 69 69 100 100 130 161 161 192 192 OYSTER CRE 1 BNR 0 0 0 0 0 0 0 12 12 36 52 62 87 112 112 131 163 163 163 PORT CALHOUN PNR 0 0 0 0 0 0 0 3 3 19 35 35 51 67 67 e3 99 99 118 220N 142 P NR 0 0 0 0 0 0 0 0 32 65 131 164 197 263 296 329 394 427 460 BIO ROCK 1 BWR 0 .0 0 0 0 0 0 0 2 5 5 7 10 12 12 12 12 12 12 LIMERICE 1L BNR 0 0 0 0 0 0 0 0 27 27 66 105 105 144 183 103 220 258 258 EAN ONOPRE 1,2,43PNR 0 0 0 0 0 0 0 0 56 119 163 226 270 372 416 460 504 '540 592

- SEQUOYAN ist ' P NR 0 -0 0 0 0 0 0 ' O 18 18 92 129 166 239 276 313 350. 423 460 DAVIB-BESSE 1 PNR 0 0 0 0 0 0 0 0 4 4 33 61 61 89 117- 117 145 173 173 e- POINT BEACH 142 P WR 0 0 0 0 0 0 0 0 le 41 64 87 110 133 156 179 233 226 249 ARE NUCLEAR 1 PNR 0 0' 0 0 0 0 0 0 0 8 e 36 36 64 91 91 119 119 147 BRAIDWDDD 142 PNR 0 0 0 0 0 0 0 0 0 42 42 113 184 184 255 326 326 397 46e BEAVER VALLEY 1 PNR 0 0 0 0 0 0 0 0 0 12 44 44 16 76 los 104 140 172 170 MAINE YANREE PWR .0 _0 0 0 0 0 0 0 0 18 de 46' 14 102 102 130 15e 154 106 NINE MILE Pf3 BNR 0 0 0 0 0 0 0 0 0 26 26 - 57 57 89 . 89 120 120 ta2 152 NADCAM NECK PNR 0 0 0 0 0 0 0 0 0 0 1 1 to 39 59 59 91 97 97 ENRICO PERM 12 BNR 0 0 0 0 0 0 0 0 0 0 25 25 78 78 130 130 103 236 236 COOPER STN BNR 0- 0 0 0 0 0 0 0 0 0 1 22 43 63 83 te4 124 145 164 MILLSTONE 2 PKR 0 0 0 0 0 0 0 0 0 0 19 44 44 6e 97 92 117 140 140 PEACKBOTTOM 2 BNR 0 0 0 0 0 0 0 0 0 0 36 38 79 119 119 160 200 200 240 PEACnBOTTOM 3 BNR 0 0 0 0 0 0 0 0 0 0 1 7 46 85 85 124 124 163 202 FIT 1 PATRICK' BWR 0 0' O O O O O O e 0 24 20 55 86 86 118 149 149 180 SALEN 1 PER '0' 0 0 0 0 0 0 0 0 0 22 22 54 95 95 132 168 168 208 DRE3 DEN 2 BNR 0 0 0 0 0 0 0 0 0 0 0 11 la as 64 - 64 91 117 117 CDOR 162 P KR 0 0 0 0 0 0 0 0 0 0 0 28 64 136 173 200 245 201 35,3 GRAND GULF 1 BNR 0 0 0 0 0 0 0 0 0 0 0 18 66 66 113 160 160 207 255 LIMERICE 2 BNR 0 0 0 0 0 0 0 0 0 0 0 23 23 62 100 100 las 138 177 NASH NUCLEAR 2 BNR 0 0 0 0 0 0 0 0 0 0 0 19 50 75 102 132 158 188 213 ARK NUCLEAR 2 PNR 0 0 0 0 0 0 0 0 0 0 0 0 9 37 37 46 94 94 122 DUANE ARNOLL BNR 0 0 0 0 0 0 0 0 0 0 0 0 21 44 44 65 47 87 10N VT YANREE 1 BNR 0 0 0 0 0 0 0 0 0 0 0 0 2 2- 26 49 49 73 96 NORTN ANNA 142 P KR 0 0 0 0 0 0 0 0 0 0 0 0 55 85 114 173 203 232 292 KEKAUNEE PWR 0 0 0 0 0 0 0 0 0 0 0 0 9 23 37- 51 65 19 94 YANKEE-RONE 1 P KR 0 0 0 0 8 0 0 0 0 0 0 0 4 8 e e a e 8 DRESDEN 3 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 10 31 37 63 90 90 MATCH 142 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 24 60 133 170 206 279 8UsQUEKANKA 142 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 17 97 137 177 257 297 GINNA - F NR 0 0 0 0 0 0 0 0 0 0 0 0 0 9 21 32 ' 43 54 66 ST LUCIE 2 P NR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 le 46 46 74 102 5ALEN 2 PWR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 33 11 71 111 150 BROWNS PERRY 3 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 29 29 71 112 112 NATT5 hAR 142 PNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 24 97 97 171 208 TVRKEY PT 364 P WR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 22 66 88 110 BRONN5 FERRY 142 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 6 et 130 TROJAN PNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 17 39 INDIAN PT 7 P KR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 34 kANCHO SECO 1 P KR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 20 4e P ALO VERDE 1 PKR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 QULD CITIES 142 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 MCGUIRE 1 PNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 12

' CRYSTAL DVR 3 P KR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 33 CALLRNAY 1 P KR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 16 PWR NTIRM 15 83 315 553 1126 2234 3409 4973 6665 852s

. 53 209 424 402 1550 2106 4279 5899 7476 L bKR NTNIM 23 58 194 452 752 1261 2005 3033 4221 5535 23 113 271 577 919 1505 2474 3646 4905 TOTAL NTERN 38 140 509 1004 1879 3495 5494 6006 10086 14063 76 ,322 695 1379 2469 4211 6753 9545 12382 A-6 l~

E L _ ___.._ _ _ _ _ _ _

7. -

. >.a .

~'*

l '.

<. ..g e

~

~'

w, TAB 15 A.4. 1996 2moestery and Projected ammal Ramster Disahargas, Ash 11** (esm8E)

ENV.(a)

REA050R 1M419871M81M91DM 19911998199319M 199619M 199719981998 2000 20013008 2003 2004 2005 FT ST VRAIN H70 - 0 0 240 0 282 0 240 0 240 .0 240 0 240 0 240 0 1492 0 0 0 RESEARCH SITES PWR 97 0 0 0 0 0- 0 0 0 0 0 0. 0 0 0 -0 0 0 0 0 R2.5EARCH 8ITES BNR 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

~RESEARCH $1TES HTG' 720 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 PNR A33EMBL!ts 20309 3155 3577 3873 4010 3593 3259 3974 3954 3599 2644 3200 '3677 3171 3647 3961 4121 3225 3502 '3935 BNR ASSEMBLIES - 30605 3394 4468 4408 4440 4788 4380 4276 4408 5292 3758 4990 4680 5000 4640 5024 5088 5032 4508 5192 8T0 ASSEMBLIES 720 - 240 282 240 240 240 '240 240 1482 0

,0 0 0 0 0 0 0 0 0 0 TDTAL ASSEMBLIES $1434 '6789. 5327 8521 0690 8621 7879 0490 9924 4091 6402 8190 8357 8171 8247 8991 9209 8257 0090 9127 (a) ' PERMANENTLY DISCN/ AGED SPENT FUEL. THIS INCLUDES 50HE SPENT F'JZL, APPROXIMATELY 140 MT!HM, PHYSICALLY RESIDENT IN THE REACTOR CORE ON DECEMBER 31, 1986 WHICH IS NOT PLANNED TO UNDERGO ANY FUTURE JRRADIAT10N.

+

A-9

.)

di ']

h(f" 445 g) ,- )

je w._

.  : .t dP TABIA A.4. 1861avembery ame Projected Ammual Beamter Diaaha*ges, Assemblies ENV.(a)

• - 'aa \

";;" 1D.7 1 6 1 8 1. 0 1 1 1 2 1 3 1.D4 1. 8 1. 6 1 ,1. 1. 2000 2001 2002 2003 28 4 2006 ARE NUCLEAR 1 Fra 448 0 JO O 60 so 0 60 0 60 60 0 60 0 60 60 0 60 0 60 ARE NUCLEAR 2 PWR 288 'O 68 60 0 68 64 0 es 68 0 68 0 68 68 0 68 60 0 68 SEAYER VALLEY 1 PNR 243 73 0 69 0 69 69 0 69 0 69 69 0 69 0 69 0 69 69 0 SEAVER VALLEY 2 Prm . 0 0 0 37 0 73 0 73 73 73 0 13 0 73 13 0 73 73 - 0 73 SELLEFONTE 1 PNR 0 0 0 0 0 0 0 0 64 72 0 84 to 0 84 0 84 44 0 84 BELLEFONTE 2 PNR . 0 0 0 0 0 0 0 0 0 0 64 12 0 84 84 0 84 0 44 84 BIG ROCK 1 DNR 188 22 22 20 20 20 20 20 20 20 20 '

O 20 to 20 84 0 0 0 0 BRAIDWOOD 1 PNR 0 0 0 as es 0 es - 80 0 84 84 0 84 84 0 84 84 0 84 84 BRAIDWOOD 2 PWR ' 0 0 0 0 es 0 48

• 88 0 se 04 0- 84 64 0 84 84 0 84 04 BROWNS FERRY 1 DNR 1320 0 0 0- 0 0 - 228 0 228 0 220 228 0 228 228 0 228 0 228 220 BROWN 3 FERRY 2 SNR 1192 0 0 204 ' 0 220 0 224 220 0 ' 228 0 228 220 0 228 220 0 220 0 BRDNNS FERRY 3 DNR 1004 0 0 0 268 228 0 228 0 228 22s 0 224 228 0 224 0 228 22e 0 BRUNSW3CK 1 DNR 840 les 0 > les 184 0 188 0 les 184 0 188 180 0 184 0 188 Its 0 184 BRUN8 NICK 1 PNR 160 0 0 0 0 0- 0 0 0 0 0 0 0 0 0 0 0 0 0 0

'SRUN8 NICK 2 RNR_ 756 0 188 180 0 184 0 les 100 0 184 180 0 108 0 184 188 0 ist 184 BRUNSWICK 2 PNR 144 0 .0 0 "0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 BYROM 1 F KR 0 88 88 0 48 88 0 84 84 0 e4 04 0 84 84 0 84 84 0 84 BYRON 2 PNR 0 0 to - O se s8 0 es 44 0 84 e4 0 84 04 0 84 e4 0 se CALLANAY 1 - PNR 84 96 0 80 to 0 84 84 0 84 84 0 84 84 0 84 e4 0 84 84 '

CALVERT CLF 1 P NR 618 0 96 0 96 0 96 0 96 0 96 0 96 0 96 0 96 0 96 0 CALVERT CLF 2 P WR 432 88 0 96 0 96 0 96 0 96 0 96 0 96 0 96 0 96 0 96 CATAKRA 1 P WR - 64 60 0 69 68 72 72 0 72 73 73 0 72 72 72 72 0 73 - 12 72 CATArnA 2 PNR 0 0 65 48 64 64 69 0 60 00 12 0 72 73 72 0 73 72 12 72 CLINTON 1 BNR 0 0 140 192 0 160 176 0 172 164 0 172 172 0 172 172 0 172 172 0 COMANCHE PK 1 P KR 0 0 0 0 64 64 64 0 6e 64 64 60 64 64 68 64 64 64 64 64 COMANCHE PK 2 PKR 0 0 0 0 0 64 64 64 68 64 64 68 64 64 68 64 64 68 64 64 COOK 1 P WR 546 80 0 80 00 0 00 80 0 40 0 to 80 0 to 80 0 to 0 80 -

COOK 2 PWR 424 0 to 0 0 es 88 0 se sa 0 08 0 88 80 0 se 0 se se COOPER STN DNR 648 0 136 116 116 116 116 120 112 116 116 lit 116 112 112 112 112 112 112 10s CRYSTAL kVR 3 PKR 302 93 0 41 0 72 0 72 0 72 0 72 0 72 0 72 0 72 0 72 Davis-BESSE 1 . PNR 197 0 65 61 0 61 - 61 0 61 61 0 61 (0 0 60 60 0 60 60 0 DIABLO CANYDN 1 P WR 51 0 ss 0 44 0 e5 0 85 0 e5 0 85 0 es 0 85 0 85 0 DIA8LD CANYON 2 PWR 0 51 68 0 84 0 85 0 85 0 85 0 85 0 SS 0 85 0 85 0 DRESDEN 1 DNR 683 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 DRESDEN 2 DNR 1606 0 168 158 0 158 154 0 158 158 0 150 158 0 158 150 0 158 158 0 DRESDEN 3 DNR 1456 152 0 160 0 158 150 0 158 158 0 158 15s 0 158 158 0 158 154 0 DUANE ARNOLD BNR 696 120 120 0 120 124 0 120 120 0 128 120 0 120 125 0 120 120 0 120 ENRICO FERMI 2 DNR 0 0 0 232 292 0 276 0 296 0 292 288 0 292 0 20s 0 292 288 0 FARLEY 1 PWR 410 0 60 65 0 60 45 0 6s 65 0 68 65 0 ss 65 0 68 65 0 FARLEY 2 ' PNR 256 64 0 60 65 0 is 65 0 68 65 0 68 65 0 68 65 0 64 65 FIT 8PATRICE BNR 1012 1st 164 0 176 100 0 116 176 0 100 176 0 174 176 0 176 175 0 176 FORT CALHOUN FwR 334 44 45 0 45 45 0 45 45 0 45 (5 0 45 45 0 45 45 0 45 GINNA FWR 470 36 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 GRAND GVLP 1 BKR 264 280 0 268' 260 0 268 260 0 268 268 0 264 268 0 268 260 0 268 260 hADDAM NECK PWR 594 57 0 48 53 52 0 53 52 0 53 52 0 53 51 53 0 52 53 0 RARRIS 1 BNR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 HARRIS 1 PNR 0 0 53 0 52 52 0 53 52 0 52 53 0 52 52 0 53 52 0 52 RATCH 1 DNR 1107 240 196 0 196 196 0 196 196 0 196 196 0 196 196 0 196 196 0 196 KATCH 2 DNR 745 0 104 184 196 0 196 196 0 196 196 0 196 196 0 196 196 0 196 196 HOPE CREEK DNR 0 0 232 232 0 232 232 0 232 232 0 232 232 0 232 232 0 *32 232 0 DNR 390 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 EUMBOLDT BAY PNR 360 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

- INDIRN PT 1 0 0 2KDIAN PT 2 PNR 464 60 0' 60 0 68 ss 0 60 0 68 es 0 64 0 6e 68 0 64 0 INDIAN FT 3 PWR 292 16 0 76 16 0 76 0 76 16 0 16 0 16 16 0 16 0 76 76 KENAUNEE PNR 349 45 37 33 41 37 45 37 37 31 37 37 37 37 37 37 37 37 37 37 LACROSSE DNR 261 24 24 0 24 24 24 24 24 24 0 24 24 24 24 24 72 0 0 0 LASALLE CTY 1 DNR 132 324 0 les 220 0 220 220 0 220 220 0 220 220 0 220 220 0 220 220 LA1ALLE CTY 2 BNR 0 224 232 0 220 220 0 220 220 0 220 220 0 220 220 0 220 220 0 220 LIMERICK 1 Bra 0 268 272 0 224 0 220 216 0 220 0 220 220 0 220 220 0 208 216 0 LINERICE 2 DNR 0 0 0 0 32s 220 0 20s 0 224 212 0 216 0 216 216 0 216 0 216 MAINE YARKEE P NR 793 13 76 0 73 73 0 73 13 0 73 73 0 73 73 0 13 73 0 73 NCCUlkE 1 Pwn 219 69 72 0 72 72 72 0 13 12 72 12 0 73 12 72 0 72 12 73 MCGUIRE 2 P KR 186 73 69 64 69 0 60 40 72 0 73 12 72 72 0 12 73 72 8 72 A-7 l ii _ _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ ___ __

~ q Ok $ Q s

^

ad . .

-B e ,

o TAsia A.4. 1986 Inventory and Projected Annual Resoter Discharges, Asseanlies (con't)

IsV.ts) AS$sunLzas REACTOR 1986 1987 1984 1989 1990 1991 1992 1993 1994 1998 1996 1997 1990 1999 2000 2001 2002 2003 2004 2008 es-. .

MILLSTONI 1 BWR 1536 196 0 196 0 196 0 196 0 196 0 196 0 196 0 196 0 196 0 196 1 MILLSTONE 2 PNR 474 0 77 05 52 0 68 0 65 61 0 64 65 0 61 65 0 45 61 0 MILLSTONE 3 Fwn 0 84 0 to 44 0 84 84 0 84 84 0 84 to 0 to 04 0 44 84 HONTICELLO BKR 428 116 0 its 116 0 124 116 0 120 0 120 120 0 120 120 0 120 120 0 i MORRIS BKR 2047 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 MORRIS PWR 350 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 l

i NINE MILE PT1 BKR 1444 0 200 0 192 0 196 0 176 0 188 0 140 0 144 0 184 0 184 532 NINE MILE PT2 BWR 0 0 0 296 0 276 0 264 0 284 0 260 0 200 0 276 0 276 0 276 NORTH ANNA 1 PWR 294 41 65 0 65 65 0 65 66 0 65 65 0 65 64 0 64 64 0 64 NORTH ANNA 2 PKR 235 69 0 65 65 0 65 65 0 65 65 0 65 65 0 64 64 0 64 64 OCONEE 1 PWR 590 57 0 60 60 60 0 60 60 60 0 60 60 60 0 60 60 60 0 60 OCOKEE 2 Pwn 381 49 0 60 60 0 60 60 0 60 60 60 0 60 60 0 60 60 0 60 DCONEE 3 PNR 529 0 60 60 0 60 60 0 60 60 0 60 60 60 0 60 60 0 60 60 OYSTER CRK 1 BNR 1392 0 128 135 0 128 172 0 152 0 132 140 0 140 140 0 140 144 0 556 PALISADES PWR 545 0 68 0 60 0 68 0 60 68 0 68 0 60 0 64 68 0 60 0 PALO VERDE 1 PKR 0 80 77 0 93 77 0 93 77 0 93 77 0 93 77 0 93 77 0 93 FALO VERDE 2 PKR 0 0 92 85 0 85 77 0 93 77 0 93 77 0 93 77 0 93 77 0 PALO VERDE 3 PWR 0 0 0 92 45 0 05 77 0 93 77 0 93 17 0 93 77 0 93  ??

PEACKBO??QM 2 BKR 1462 272 0 308 0 284 0 264 220 0 228 228 0 228 220 0 220 228 0 228 PEACKBOTTOM 3 BWR 1496 268 0 0 216 236 0 220 216 0 224 220 0 220 220 0 220 0 220 220 PERRY 1 BNR 0 0 220 0 244 200 0 220 224 0 224 224 0 224 224 0 224 224 224 0 PILGRIM 1 BKR 1320 0 196 0 ,0 192 0 192 0 196 0 192 0 192 0 0 196 0 192 0 POINT BLACH 1 PNR 446 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 PQ1NT BLACN 2 PKR 408 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 32 PRAIRIE ISL 1 PWR 306 45 40 40 41 to 0 40 40 40 40 40 40 40 40 0 40 40 to 40 PRAIRIE ISL 2 PWR 415 0 40 40 40 40 40 40 40 40 0 40 40 40 40 40 40 40 40 40 QUAD CITIES 1 BKR 1393 172 0 160 160 0 160 172 0 160 160 0 160 160 0 164 160 0 160 160 QUAD CITIES 2 BWR 3424 0 168 160 0 160 160 0 160 160 0 160 160 0 160 160 0 160 160 0 RANCHO SECO 1 F KR 267 0 57 65 0 69 69 0 61 El 0 61 0 57 61 0 61 0 57 61 ROBINSON 2 P KR 270 48 61 0 48 48 61 0 40 44 0 61 44 48 0 61 48 0 44 61 RVR BEND 1 BWR 0 164 0 224 180 0 192 204 0 108 196 0 192 192 0 196 192 0 192 192 SALLM 1 PWR 344 43 0 85 85 101 0 01 85 0 to 80 0 to 8. 0 to to 0 to SALEM 2 PKR 174 0 73 89 0 97 101 93 0 05 84 0 84 .. 0 84 84 0 84 84 SAN ONOFRE 1 6 di 146 0 52 52 0 0 52 0 52 0 52 0 52 0 157 0 0 0 0 0 SAN ONOFRE 2 P KR 147 109 0 109 0 109 0 109 0 109 109 0 109 0 10b - 109 0 109 0 SAN ONOFRE 3 PKR 147 0 109 0 109 0 109 0 109 109 0 109 0 109 0 ;v9 0 109 0 109 SEABROOK 1 P KR 0 0 64 0 64 64 64 64 64 0 64 64 64 64 64 0 64 64 44 64 SE3UOYAH 1 PKR 212 0 0 80 80 0 40 80 0 00 0 80 to 0 80 to 0 to 40 0 SEQUOYAM 2 Pun 136 80 0 40 0 83 80 0 80 80 0 80 0 to 80 0 80 0 to 80 SHOREKAM BNR 0 0 0 204 0 176 160 0 112 184 0 184 184 0 184 184 0 184 184 0 SOUTR TEKAS 1 PWR 0 0 0 56 54 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 SOUTH TEKAS 2 PWR 0 0 0 0 56 54 52 52 52 52 52 52 52 52 52 52 52 52 52 52 St LUCIE 1 PWR 444 109 to 0 12 16 0 16 76 0 16 16 0 16 76 0 76 76 0 76 ST LUCIE 2 P KR 164 93 0 72 72 0 72 72 0 72 72 0 72 72 0 72 72 0 72 72 SUMMER 1 PKR 112 64 68 0 60 68 0 68 68 0 68 68 0 68 6e 0 68 68 0 68 SURRY 1 FKR 480 0 69 49 56 0 53 53 0 53 53 0 53 53 0 $2 52 0 52 52 SURRY 2 PWR 385 0 65 49 0 57 53 0 53 53 0 53 53 0 53 52 0 52 52 0 SUSQUEKANNA 1 BWR 488 240 0 240 220 0 236 232 0 232 232 0 232 232 0 232 232 0 232 232 SUSQUEKANNA 2 BKR 324 0 236 220 0 232 232 0 232 232 0 232 232 0 232 232 0 232 232 0 THREE MILE ISL 1 PKR 284 0 73 0 73 73 69 0 73 0 73 73 0 73 73 0 72 72 0 72 TROJAA P KR 379 57 47 48 49 40 40 48 40 44 48 40 48 48 48 48 48 48 48 48 TURKEY PT 3 PKR 424 77 60 0 48 0 52 44 48 0 48 48 0 48 48 0 48 48 0 44 TURKEY PT 4 P KR 446 0 52 52 0 48 48 48 0 40 48 0 48 48 0 48 0 40 40 0 j VOGTLE 1 P KR 0 0 44 84 0 84 84 0 84 84 0 84 84 0 84 84 0 84 84 0 VOGTLE 2 P KR 0 0 0 to 0 84 84 0 84 44 0 84 84 0 84 04 0 84 04 0 VT YANKEE 1 BKR 1322 136 0 132 132 0 132 132 0 132 132 0 132 132 0 132 132 0 132 132 NASH NUCLEAR 2 BKR 128 144 164 160 148 156 156 156 172 140 156 172 144 176 140 156 164 140 168 144 NATERFORD 3 PMR 92 0 88 88 0 to 80 0 88 88 0 08 88 0 80 80 0 80 88 0 l WATTS BAR 1 PNR 0 0 0 0 64 72 0 80 to 0 80 0 40 00 0 80 to 0 to 0 NATTS bAR 2 P KR 0 0 0 0 0 64 72 0 to 0 to go 0 00 to 0 80 0 to 80 KLST VALLEY BKR 85 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 NEST VALLEY P KR 40 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KOLF CREEK 1 P KR 52 52 76 0 76 76 0 76 76 0 76 76 0 76 76 0 16 16 0 16 Y AN KEE- ROK5 1 P WR 341 36 40 0 36 0 40 36 0 40 36 0 40 36 0 76 0 0 0 0

!!DN 1 P WR 514 0 76 72 0 72 72 0 12 72 0 72 72 0 72 72 0 72 72 0 EION 2 P KR 503 80 72 0 76 72 0 72 72 0 12 72 0 72 72 0 12 72 0 72 A-0 l

l l

W. '

- - - _ - - - - - _ -