ML20237K200

Intervenor Exhibit I-SC-1,consisting of 870327 Board Notification 87-05 Re Encl Draft BNL Rept, Beyond DBAs in Spent Fuel Pools (Generic Issue 82). BNL Determines That Risk from Spent Fuel Pools Comparable to Core Melt Risks

ML20237K200
Person / Time
Site:	Diablo Canyon
Issue date:	06/16/1987
From:	Novak T Office of Nuclear Reactor Regulation
To:	Asselstine, Roberts, Zech NRC COMMISSION (OCM)
References
CON-FIN-A-3786, REF-GTECI-082, REF-GTECI-NI, TASK-082, TASK-82, TASK-OR OLA-I-SC-001, OLA-I-SC-1, NUDOCS 8709040249
Download: ML20237K200 (187)

Text

{{#Wiki_filter:__- ,go-2 75l3 2 3 - Nf 6/}s/g7 ff UNITED STATES _7.^ 5W/ g NUCLEAR REGULATORY COMMISSION r. wAsMNCTON, D. C. 20555 ,/ '87 ALS 26 P 4 :02 - March 27, 1987 Docket Nos. 50-275 iF-bocei + and 50-323 MEMORANDUM FOR: Chainnan Zech Commissioner Roberts Commissioner Asselstine Commissioner Bernthal Consnissioner Carr FROM: Thomas M. Novak, Acting Director Division of PWR Licensing-A

SUBJECT:

BOARD NOTIFICATION REGARDING BNL DRAFT REPORT ON SPENT FUEL POOL ACCIDENTS (BN 87-05) \\ In accordance with NRC procedures for Boaro Notification (BN) we are providing for your infonnation the following enclosed Brookhaven (BNL) draft report: "Beyond Design-Basis Accidents in Spent Fuel Pools (Generic Issue 82),"- Department of Nuclear Energy, Brookhaven National Laboratory, Draftt, January 1987, transmitted by letter from K. R. Perkins'(BNL) to E. Throm (NRC), dated February 5,1987 The report addresses the risk of spent fuel storage at nuclear power plants and includes an evaluation of accident initiating events and their probabilities. fuel cladding failure scenarios arising from uncertainties in Zircaloy~ oxidation l cladding failure scenarios. reaction rate data, and the potential for releases of ra perfonned by BNL for the staff on Generic Issue B2, "Beyond Desig Accidents in Spent Fuel Pools." within the staff for a peer review.The draft report has been distributed widely. Like most of the Generic Issues and Unresolved Safety Issues currently being studied by the staff, the likelihood of beyond design basis accidents is being examined to assess whether vulner-abilities could exist which might require further consideration of changes in design or operations. P oliminary staff opinion is that substantial portions of the report will need more critical review because~some assumptions appear to be oversimplified. expected to be issued this susiner. Comments will be provided to BNL and the final re The contractor states, based on these preliminary results, that the risk from spent fuel pools is comparable to the risk from core melts in general. The exidation of the Zircaloy cladding of thespent fuel pool risk is primarily du racks if water is lost from the pool. The dominant causes of water loss arespen due.to structural failures of the pool caused by droppino of a spent. fuel cask or by an earthouake.. Howeyer, all of the estimates of the factors affecting the risk are preliminary and the report is less an estimate of the risk than identification of the factors that need further study in order to make a

• f-f reasonably accurate. estimate of risk.

8709040249 870616 " g{ PDR ADOCK 05000275 g PDR

NUCl[AR RlutlLATORY COM]t$$10N 3c41 Docket es.50 ms-o a on;,,,,,,3.,,. b*b b A@A vi in the statter of StaN IDINTWlED - Applicant RfCliflD - Interveno' - REIECTID Cont't Off's Citntracto' DAll _h"I b ~ Other Wiuen 9IY' M Mf (I / Reporter 1 I L 1 f l i l l i I

2 Based on a brief review, the staff expects that the final result of the study l is likely to provide a much lower estimate of risk. The risk estimate in the report is not based on an analysis of the structural capability of fuel pools, but of similar concrete structures. Furthermore. the report effectively assumed that failure of the pool would result in the loss of integrity of the pool liner without considering whether the strain would be sufficient to fail the liner. Based on other seismic risk studies, the fuel pool could require a much stronoer, and therefore less probable, earthquake than that required to fail the systems necessary to control reactivity and cool a core. Failure due to a cask drop is not likely now since casks are used infrequently at only a few plants. The report also assumes that all of the casium would be released from the equivalent of three cores if a Zirceloy fire occurred. While the spent fuel in a pool has less of the short half-life fission produc.ts than a fresh core, this would primarily only affect estimates of early fatalities. However, in the analysis of the fraction of cesium released from the spent fuel, it was assumed that all of the fuel was at the maximum calculated clad temperature. If the radial and axial temperature distribution were cerisidered, the cairu-lated release fraction would likely be less, possibly by a factor of two. The report also assumes no retention of cesium in the fuel pool building. Even if an earthouake destroyed the building, the cebris would likely cause much of the cesium to be retained. Other studies of the cesium retention in similar buildings and the large fraction of cesium retained at Chernobyl indicate that the report may underestimate retention by a factor of ten. Overall the report may overestimate the risk from fuel poolt by a factor of about 200. The draft report does not pertain directly to currently ongoing licensing efforts for spent fuel pool expansion amendment requests by utilities, including hearings. However, we believe that the subject may involve substantial public, press or Congressional interest. For this reason we are providing you with the enclosed report. We also are providir.g the report, by copy of this Board Notification, to the Boards and Service Lists for the Ciablo Canyon Plant and Yement Yankee Station. We will inform you of further developments. D Thomas H. Novak, Ac irector Division of PWR Licensing-A

Enclosure:

As stated cc: See next page

Mr. J. D. Shiffer Pacific Gas and Electric Company Diablo Canyon c.c: Richard F. Locke, Csq. NRC Resident Inspector Pacific Gas A Electric Company Diablo Canyon Nuclear Power Plant Post Office Rox 7a47 c/o U.S. Nuclear Regulatory Commission San Francisco, California 94120 P. O. Box 369 Avila Peach, California 93t?4. Janice E. Kerr, Esq. California Public Utilities Commission Mr. Dick Plakenburg 350 McAllister Street Editor & Co-Publisher San Francisco, Califcrnia 94102 South County Publishing Company P. O. Gr 460 Arroyt brande, California 93420 Ms, Sandra A. Silver 660 Granite Creek Road-Bruce Norton. Esq. ~ Santa Cruz, California 95065 c/o Richard F. Locke, Esq. Paci'ic Gas and Electric Company Post Office Box 7442 Mr. W. C. Gcrcloff San Francisco, Celifornia 94120 Westinghouse Electric Corporation P. O. Box 355 Pittsburgh, Pennsylvania 15230 Dr. R. B. Fercuson Siera Club - Santa Lucia Chapter Rocky Canyon Star Route Managing Editor Creston, California 9343? San Luis Cbispo County Telegram Tribune 1321 Johnson Avenue Chairman 1726 M Street, N.V. San Luis Obispo County Board of l Suite 1100 Supervisors Washinoton,'DC 20036-4502 Room ??O County Courthouse Annex San Luis Obispo, California 93401 1 Mr. Leland M. Gustafson, Manager Federal Relations Fecific Gas and Electric Company Director 1726 M Streat, N. W. Energy Fecilities Siting Divisien Washington, DC 20036 4502 Energy Resources Conservation and Development Commission 1516'9th Street Sacramento, California 95814 Dian M. Grueneich Esq. Edwin F. Lowry, Esq. Mc. Jacquelyn Wheeler Grueneich & Lowry 2455 Leona. Street 345 Franklin Street San Luis Obispo, California 93400 San Francisco, California 94102 m .._._..__..__________.________________.____.__.d

y, Pacific Gas & Elettrir. Company Diablo Canycn s CC* Ms. Laurie McDermott Coordinator Ms. Nancy Culver Consumers Organized for Defense 192 Luneta Street of Environmental Safety San Luis Obispo, California 93401 731 Pacific Street, Suite 42 San Luis Obispo,-California 93401 President California Public Utilities Mr. Joseph 0. Ward, Chief Comission Radiological Health Branch California State 9uilding State Department of Health 350 McAllister Street Services San Francisco, California 94102 714 P Street Office Building #8 Sacramento, California 95814 Regional Administrator, Pegion V U.S. Nuclear Regulatory Commission 1450 Maria Lane Suite 210 Walnut Creek, California 94596 B ' ?\\ g. / m

Pacific Gas and Electric Company 3-Diablo Canyon ,cc: r Glenn 0.. Bright Administrative Judge Atomic-Safety and')atory Commission Licensing Snard U.S. Fuclear Regul Washinoton, DC 20555 / Dr. Jerry Harbour Administrative Judge Atomic Safety and Licensing Board U.S. Nuclear Regulatory Commission Washington, DC 20555 ' R. Paul Cotter, Jr., Chairman Administrative Judge Atomic Saff.ty and Licensing Board U.S. Nuclear Regulatory Commission Washington, DC 20555 ( i d i l ) ) j e I mm

Mr. R. W. Capstick Vermont Yankee Nuclear Power Vermont Yankee Nuclear Power Corporation Station cC* Mr. J. G. Weigand Mr. W. P. Murphy, Vice President & President & Chief Executive Officer Manager of Operations Vermont Yankee Nuclear Power Corp. Vermont Yankee Nuclear Power Corp. R. D. 5. Box 169 R. D. 5, Box 169 Ferry Road Ferry Road Brattleboro, Vermont 05301 Brattleboro, Yemont 05301 Mr. Donald Hunter, Vice President Mr. Gerald Tarrant, Commissioner Vemont Yankee Nuclear Power Corp. Vermont Department of Public Service 1671 Worcester Road 120 State Street Framingham, Massachusetts 01701 Montpelier, Yemont 05602 New England Coalition on Public Service Board Nuclear Pollution State of Vermont Hill and Dale Farm 120 State Street R. D. 2, Box 223 Montpelier, Vermont 05602 Putney, Vermont 05346 Vermont Yankee Decommissioning Mr. Walter Zaluzny Alliance Chairman Board of Selectman Box 53 Post Office Box 116 Montpelier, Vermont 05602-0053 Vernon, Yemont 05345 Resident Inspector Mr. J. P. Pelletier, Plant Manager U. S. Nuclear Regulatory Comission Vermont Yankee Nuclear Power Corp.- Post Office Box 176 Post Office Box 157 Vernon, Yemont 05354 Vernon, Vermont 05354 Vemont Public Interest Mr. Raymond N. McCandless Research Group Inc. Vermont Division of Occupational 43 State Street & Radiological Health Montpelier Vemont 05602 Administration Building 10 Baldwin Street Regional Administrator, Region I Montpelier, Vermont 05602 U. S. Nuclear Regulatory Comission 631 Park Avenue Honorable John J. Easton King of Prussia, Pennsylvania 19406 Attorney General State of Vermont Ellyn R. Weiss 109 State Street Harmon & Weiss i Montpelier, Yemont 05602 2001 S. Street, N. W. { Suite 430 J Washington, D.C. 20009-1125 John A. Ritscher, Esquire Ropes & Gray Office of the Attorney General 225 Franklin Street 1 Ashburton Place Boston, Massachusetts 02110 19th Floor Boston, Massachusetts 02108

s 4 7 -. s 3 ' o- .k ,.s. Yermoct Yankee Nuclear Power Corporation Vermont Yankee Nuclear Power Station ec: i . Charles Bechhoefer Chairman. 4 l. Atomic Safety and Licensing Soard Panel U.S. Muclear Regulatory Comissio"- Washingten,:DC 20555 p ' Glenn 0. Ertght Atomic Safety and Licensing Board Panel U.S. Nucl6ar Regulatory Comission Washingtoa, DC 20555 3 ( Jarres H. Carpenter Atomic Sefety and Licensing Board Panel ~ - U.S. Nuclear Regulatory Commission Washington, DC 20555 1 b 4 '\\. t i l ~ '. i ____N_-___-_-___. X ~

o k)l}h BROOKHAVEN NATIONAL LABORATORY Q]} ASSbCIATED UNIVERSITIES, INC. l Opton, Long Island. New York 11973 . (516) 282< 2147 - Depcrtment of Nuclear Energy FTS 666/ I . February 5,1987

• 1 l

Mr. Edward Throm Division of Safety. Review and Oversight U.S. Nuclear Regulatory Commission Phillips Bldg., Mail Stop 244 7920 Norfolk Avenue Bethesda, MD 20814 RE: FIN A-3786

Dear Ed:

I have enclosed four copies of the revised draft to the report entitled, "Beyond Design-Basis Accidents in Spent Fuel Pools (Generic Issue 82)." We have incorporated your comments' on the rough draft. The multidisciplinary nature of this effort required _ input from several organizations within -BNL. e The management review is still ongoing. This report satisfies the milestone for the draft report. We will com-plete the formal NUREG/CR after receiving the NRC review. Sincerely, M ~ K.R. Perkins, Group Leader Containment & Systems Integration Group cc: A. Benjamin, SNL H. Connell W.T. Pratt M.-Reich V.L. Sailor T. Teichman A. Tingle J. Weeks R. A..Bari (w/o encl.) W.Y. Kato ( " )

r o q p NUREG/CR-F., fg; > }" BNL-HUREG-BEYOND DESIGN-BASIS ACCIDENTS IN SPENT FUEL POOLS (GENERIC ISSUE 82) V.L. Sailor, K.R. Perkins", and H. Connell Containment & Systems Integration Group and J. Weeks Materials Safety Application Group ~ Department of Nuclear Energy Brookhaven National Laboratory Upton, New York 11973 (DRAFT) ~ January 1987 i Prepared for Division of Safety Review and Oversight Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commissien Washington, DC 20555 Contract No. DE-ACO2-76CH00016 FIN A-3786

l s. s l iii ABSTRACT-This investigation has provided an integrated assessment of the risk of beyond design basis accidents in spent fuel pools for two surrogate plancs (a PWR and BWR). The investigation included an assessment of initiating fre-quency, analyses of the accident progression including the fission product releases and health consequences. The. estimated health corisequences were found to be about 12 person-rem /Ry snd 130 person-rem /Ry for the BWR and PWR plants, respectively. These estimated risk results are comparable to the estimated risk posed by severe core damage accidents and appear to warrant further attention. However, the uncertainty in this estimate is large (greater than a factor of 10) and plant specific features may change 'the results considerably. Preventive and mitigative' measures have been evaluated qualitatively. It is suggested that for plants with similar risk potential to the two surrogate plants, the one measure which fs likely to be effective in reducing risk is utilization of low density storage racks for recently discharged fuel. How-ever, before such preventive measures are implemented a complete plant spe-cific risk assessment for pool related accidents should be performed including a structural fragility analysis of the pool itself. WSD 4 I e .I o

.p. 6 iv l l D a 4 - t 4 I t i i ) \\ l 4 + 6 ED O e .w,-,=-

v TABLE OF CONTENTS Page ABSTRACT................................................................. iii LIST OF TABLES.................... ...................................... ix LIST OF FIGURES.......................................................... xiii ACKN0WLEDGEMENTS......................................................... xy

SUMMARY

S-1 1. INTRODUCTION........................................................ 1-1 1.1 Previous Investigations........................................ 1-1 1.2 Related Events................................................. 1-3 1.3 Risk Potentia 1................................................. 1-4 1.4 Discussion of Spent Fuel Storage Pool Designs and Features..... 1-4 1.5 Selection of Surrogate Cases for More Detailed Studies......... 1-5 1.6 Report Content..................................................~1-6 1.7 Refe rences of Secti on 1........................................ 1-6 2. ACCIDENT INITIATING EVENTS AND PROBABILITY ESTIMATES................ 2-1 2.1 Loss of Water Ci rcul ati ng Capabil ity........................... 2-1 2.2 St ructu ral Fai l u re of Poo1................,..................... 2-2 2.2.1 Structural Failure of Pool Resulting from Seismic Events.................................................. 2-3 2.2.1.1 A Review of Seismic Hazard Data................ 2-4 2.2.1.2 Seismic Hazard Estimates for the Millstone and Ginna Sites................................ 2-8 2.2.1.3 Seismic Fragility of Pool Structures........... 2-9 2.2.1.4 Seismically-Induced Failure _ Probabilities...... 2-10 2.2.2 Structural Failures of Pool Due to Missiles............. 2-12 2.3 Partial Draindown of Pool Due to Refuelin F a i l u re s................................. g C a v i ty Se a l ...................... 2-12 2.4 Pool Structural Failure Due to Heavy Load Drop................. 2-16 2.5 Summary of Acci dent Proba bili ti es.............................. 2-18 2.6 References for Section 2....................................... 2-19 3. EVALUATION OF FUEL CLADDING FAILURE.................................. 3-1 3.1 Suma ry of SFUEL Results....................................... 3-1 3.1.1 Summa ry Model Des c ri pti on............................... 3-1 3.1.2 Clad Fire Initiation Results............................ 3-2 3.1.3 Cl ad Fi re Propa gat i on................................... 3-3 3.2 Val i dati on of the SFUEL Computer Code........................... 3-10 3.3 Concl usi ons Rega rdi ng SFUEL Analyses........................... 3-12 3.4 R ef e rences for Secti on 3....................................... 3-13 L______.____.

2 + I' vi Page I-4. CONSEQUENCE EVALUATION.............................................. 4-1 4.1 Radi onucl i de Inv ento ri es........................................ 4-1 4.2 Rel e a s e Est i mat e s.............................................. 4-1 4.2.1 Estimated Re10ases for Self-Sustaining Cladding Oxida-tion Cases (Cases 1 and 2).............................. 4 4.2.2 Estimated Release for Low-Temperature Cladding Failure .(Cases 3 and 4)..................'........................ 4-4 4.3 Of f-Si te Radi ol ogi cal Consequences............................. 4-5 4.3.1 Scena ri os for Consequence Cal cul ati ons.................. 4-5' 4.3.2 Con sequence Res ul ts...................................... ' 4-5 t 4.4 Re ferences fo r Secti on 4....................................... 4-6 5. RISK PR0 FILE........................................................ 5-1 5.1 Fail u re Frequency Esti mates..................................... ' 5-1 5.1.1 Spent Fuel Pool Failure Probability..................... 5-1 l 5.1.2 Spent Fuel Fail u re Li kel i hood........................... 5-2 5.2 Concl usi ons Regardi ng Ri sk..................................... 5-2 5.3 References for Section 5....................................... 5-2 ) 6. CONSIDERATION OF RISK REDUCTION MEASURES............................ 6-1 l 6.1 Ri s k P r ev enti on................................................ 6-1 l 6.2 Acci d ent Mi ti gati on............................................ 6-2 j 6.3 Conclusions Regarding Preventive and Mitigative Measures....... 6-3 J 6.4

• Refe rences fo r Secti on 6........................................ 6-3 APPENDIX A - RADI0 ACTIVE INVENTORIES.....................................

A-1 A.1 INTR 000CTION..........,............................................. A-1 A.2 SIMULATION OF OPERATING HISTORIES................................... A-1 A.2.1 Thermal Energy P roducti on vs Ti me............................ A-1 A.2.2 ' Fuel. Burnup Cal cul ati ons..................................... A-1 1 A.2.3 Cal cul ation of Radi oacti ve Inventori es....................... A i A.3 DATA FOR MILLSTONE 1................................................ A ~4 3 A.3.1 Reactor and Fuel Cycl e Pa rameters.................. 4........ A-4 l A.3.2 Hi sto ry of Ope rati ons........................................ A-4 A.3.3 BWR Fuel Assembly Model Used in ORIGEN2 Calculations......... A-5 l A.3.4 Cal cul ated Radi oacti ve Inventori es........................... A-6 l A.3.5 D e c ay He a t................................................... A-6 r

L I e o 1 vii 1 Page { i l A.4 DATA FOR GINNA...................................................... A-7 { A.4.1 Reactor and Fuel Cycl e Pa rameters............................ A-7 A.4.2 Hi sto ry of Ope rati ons........................................ A-7 A.4.3 PWR Fuel Assembly Model Used in ORIGEN2 Calculations......... A-8 A.4.4 Cal cul ated Radi oacti ve Inventori es........................... A-8 A.4.5 De c ay H e a t................................................... A-9 A.S REFERENCES FOR APPENDIX A........................................... A-9 APPENDIX B - IMPACT OF REVISED REACTION ON THE LIKELIHOOD OF ZIRCONIUM FIRES IN A DRAINED SPENT FUEL P00L.......................... B-1 B.1 INTRODUCTION.................................................. DISCUSSIONS................................................... ..... B-1 i B.2 CONCLUSIONS.................................................... .... B-1 B.3 RE CO MME ND ATI ON S................................... B.4 B-3 B.5 REFERENCES FOR APPENDIX B........................................... B-4 e eh e 4 l -b

1 i I i Viii. t R I I n / O \\ ) l l l ene l l l

s s ix LIST OF TABLES Table Page 5.1 Estimated Risk for the Two Surrogate Spent fuel. Pools' from the TWo Dominant Contributors......................................... S-6 1.1. Data on Spent Fuel Basins (as of December 31, 1984).............. 1-8 7 2.1 Typical-Spent Fuel Pool Dimensions and Water Inventories......... 2-24 2.2 Decay Heat as a Function-of Time Since Last Refueling (D'ata from Appendix A)................................................. 2-24 2.3 Examples of Thermal-Hydraulic Transient Parameters Assuming Compl ete Loss of Pool - Cool ant Ci rcul ati on........................ 2-24 4 Fragility Parameters' Assumed.in'This Stud 2.4 Sto rage Pool s............................y for Spent Fuel ........................ 2-25 2.5 Weighting Factors Assigned to the Various Hazaid and Fragility Cu rves fo r the Mill stone Case..................................... 2-25 2.6 Events in Which Infl ated Se al s Ha ve Fail ed....................... 2-26 2.7 Estimated Distribution of Human Error in Heavy Crane Operations.. 2 2.8 Assumptions Used in Calculating the Hazard of Catastrophic Struc-tural Damage to Pool Resulting from the Drop of a Shipping Cask.. 2-28 2.9 Sumary of Estimated Probabilities for Beyond Design Basis Acci-dents in Spent Fuel Pools Due to Complete Loss of Water Inventory. 2-29 3.1 Sumary of Critical Conditions Necessar Su stai ni ng 0xidati on...................y to Ini ti ate Sel f-3-14 3.2 Sumary of Radial Oxidation Propagation Results for a High Densitiy PWR Spent Fuel Rack with a 10 Inch Diameter Inlet and Pe rfect Venti l ati on.............................................. 3-15 3.3 Sumary of Radial 0xidation Propagation Results for a Cylin-drical-PWR Spent Fuel Rack with a 3 Inch Diameter Hola and Perfect Ventilation.............................................. 3-16 3.4 Sumary of Radial 0xidation Propagation Results for a Cylin-drical PWR Spent Fuel Rack with a 1.5 Inch Diameter Hole and Pe rfect Venti l ati on.............................................. 3-16 3.5 Sumary of Radial 0xidation Propagation Results for Various PWR Spent Fuel Racks wi th No Venti l ati on............................. 3-17 3.6 Comparison of SNL Small Scale Oxidation Tests to Calculations with CLAD........................................................ 3-18 L__-_______--__--_

s X Table Page 4.1L Comparison of Radioactive Inventories of Equilibrium Core with ) Spent Fuel Assemblies for Selected Isotopes (Millstone 1)........ 4-8 4.2 Estimated Radionuclides Release Fraction During a Spent Fuel Pool ' Accident Resulting in Complete Destruction of Cladding (Cases 1 and 2)................................................... 4-9 4.3 Estimated Releases of Radionuclides for Case.1 in Which a Zirconium Fire Propagates Throughout the Entire Pool Inventory ( W o rs t Ca s e )..................................................... 4-10 4.4 Estimated P.eleases of Radionuclides for Case 2 in Which Only the Last Discharged Fuel Batch Suffers a Zirconium Fire.............. 4-11 4.5 Estimated Releases of Radionuclides for Cases 3 and 4 in Which Low-Temp e ratu re Cl addi ng Fa i l u res 0ccu r.......................... 4-12 4.6 Comparison of Radioactive Inventories of Equilibrium Core with Spent Fuel Assemblies for Selected Isotopes (Ginna).............. 4-13 4.7 CRAC2 Results for Various Releases Corresponding to Postulated Spent Fuel Pool Accidents with Total Loss of Pool Water.......... 4-14 5.1 Estimated Risk for the Two Surrogate Spent Fuel Pools from the Two Domi nant Cont ri buto rs........................................ 5-4 } A.1 Reactor and Fuel Cycle Parameters for Millstone 1................. A-11 A.2 Sumary of Operational Milestones for Millstone 1................. A-12 A.3 Summary of Spent Fuel Batches in Millstone 1 Storage Basin (With Projections to 1987)........................................ A-13 A.4 Comparison of Cumulative Gross Thermal Energy ProductioW with Calculated Fuel Burnup from Start of O April 1,1987 (Millstone 1)...........perations in 1970 to ............................ A-14 A.5 Comparison of Radioactive Inventories of Reactor Core an Fuel Basi n (Mill stone 1)................................d Spent .......... A-15 A.6 Comparison of Radioactive Inventories of Most Recently Dis-charged Fuel Batch (Batch 11) with Longer Aged Dischar Batches (Batches 1-10) (Millstone 1)..................ged ............ A-16 A.7 Decay Heat Released from Spent Fuel Inventor charged Fuel Batches (Millstone 1)..........y for Various Dis- ...................... A-17 A.8 Radionuclides Contributions to Deca Batches (Millstone 1).............y Heat.for Various Spent Fuel ................................ A-18 A.9 Reactor and Fuel Cycle Parameters for Gi nna....................... A-19 i A.10 Summa ry of Operational Milestone for Ginna........................ A-20

s s Xi Table Page A.11 Summary of Spent Fuel Batches in Ginna Storage Basin (With Projections to 1987).............................................. A-21 A.12 Comparison of Radioactive Inventories in Reactor Core and S Fu el B a s i n ( Gi n n a ).........'................................ pe nt A-22 A.13 Comparison of Radioactive Inventories in Most Recently Discharged Fuel Batch with Longer Aged Fuel Batches (.Ginna).................. A-23 A.14 Decay Heat Released from Spent Fuel Inventory for Various Dis-ch a rged Fuel Batch es (Gi nna )...................................... A-24 A.15 Radionuclides Contributions to Decay Heat for Various S Batches (Ginna)....................................... pent Fuel A-25 em M i.______________ _ _. ---__2_-

Xii 4 I ) / -r .-. + -.

xiii LIST OF FIGURES Figure Page 2.1 Sei smic Hazard Curve for the Mill stone Site...................... 2-30 2.2 The 15, 50 and 85 Percentile Hazard Curves for the Millstone Site............................................................. 2-31 2.3 Seismic Hazard Curves for Millstone of Each of the Individual Experts Participating in the SEP Studies and/or the SHC Studies.. 2-32 2.4 Comparison of the Millstone Site Hazard Curves Generated from the Data Input of the SHC Experts, with Those Generated from the USGS Data and from the Historical Record of the Past 280 Years............................................................ 2-33 2.5 Se i smi c Haza rd Cu rve fo r Gi nna................................... 2-34 2.6 Fragility Curves for the Oyster Creek Reactor Building........... 2-35 2.7 Probability Density as a Function of Annual Failure Frequency (Millstone 1).................................................... 2-36 2.8 Cross Secti on of a Typical Pneumatic Sea 1........................ 2-37 2.9 Cross Section of Inflated Pneumatic Seal Seated in the Reactor Vessel Flange and Inner Surface of Cavity Wall................... 2-38 2.10 Uninflated Pneumatic Seal with Steel Hold-down Ring.............. 2-39 3.1 Compari son of CLAD to SNL data fo r Test 4........................ 3-19 A.1 Millstone 1: Operating history 1976-1984......................... A-26 O

• en

I b x1y \\ W i \\ e 1 G 0 'em t l 4 e. .w ~ ~ " " " - - - _ - - _ _ _. _ _ _ _, _, _ _, _ _ _ _ _ _ e ~

ACKNOWLEDGEMENTS This work was performed for the Reactor Safety Issues Branch of the Division of Safety Review and Oversight, NRR/NRC. The NRC Managers for the program were Mr. E. Throm and Dr. M. Wohl who provided considerable input and technical direction to the program. As with most integrated programs technical contributions were provided by many people with and external to BNL. In particular, the authors are in-debted to Drs. A. Benjamin (SNL) and F. Best (Texas A&M) who provided consid-erable assistance in implementating and understanding the SFUEL code. The authors are also grateful for several technical contributions from the DNE staff at BNL. Dr. K. Shiu provided considerable assistance in evaluating the ~ seismic hazard. Dr. T. Teichman assisted in several statistical evaluations. Dr. M. Reich was especially helpful in the interpretation of pool structural fragility results and Dr. L. Teutonico provided an evaluation of the oxidation rate data. Drs. A. Tingle and W. pratt helped set up and interpret the conse- , quence calculations with the CRAC2 code. Mr. A. Aronson implemented the ORIGEN2 code and provided the calculations for spent fuel pool fission product inventories for the actual discharge histories. e The authors are especially grateful to Ms. S. Flippen for her excellent typing of this report and for cheerfully accepting the numerous additions and revisions to this manuscript. I

h g XVi e f I I I 0 , g 5 j

~

e r. b s c) e se s' b e a e

o S-1 BEYOND DESIGN-BASIS ACCIDENTS IN SPENT FUEL POOLS (GENERIC ISSUE 82) SUPJ1 arf S.1 INTRODUCTION Generic Safety issue 82, "Beyond Design Basis Accidents in Spent Fuel Pools," was assigned MEDIUM priority in November 1983.1 In its prioritira-tion, the NRC staff took account of two factors that had not been considerad in earlier risk assessments:2 1. Spent -fuel is currently being stored rather than shipped for repro-cessing or repository disposal, resulting in much larger inventories of spent assemblies in reactor fuel basins than had previously been anticipated; and, 2 A theor'etical mode 1 .4 suggested the possibility of catastrophic 3 Zircaloy fire, propagating from assembly to assembly in the event of complet,e drainage

• water from the pool.

S.I.1 previous Investigations The Reactor Safety Study 2 (which did not take account of^the two factors above) concluded that the risks associated with spent fuel storage were ex-tremely small in comparison with accidents associated with the reactor core. That conclusion was based on design and operational features of the storage pools which made the loss of water inventory highly unlikely. Subsequent to the Reactor Safety Study, A.S. Benjamin et al.S " inves-s tigated the heatup of spent fuel following drainage of the pool. A computer code, SFUEL, was developed to analyze thermal-hydraulic phenomena occurring when storage racks and spent assemblies become exposed to air. i Calculations with SFUEL indicated that, for some storage configurations and decay times, the Zircaloy cladding could reach temperatures at which the

S-2 exothern.ic oxidation would become self-sustaining with resultant destruction l of the cladding and fission product release. The possibility of propagation to adjacent assemblies (i.e., the cladding would catch fire and burn at a hot enough temperature to heat neighboring fuel assemblies to the ignition point) was also identified. In such cases, the entire inventory of stored fuel could I become involved. Cladding fires of this type could occur at temperatures well below the celting point of the U0 fuel. The cladding ignition point is about 2 900*C compared to the fue? melting point of 2860*C. 5.1.2 Related Events t There is no cass on record of a significant loss of water inventory from a domestic, commercial spent fuci storaga pool. However, one recent incident ct.currM,it th8s Waddam Neck reactor tilat raised concern about the possibility ~ of a petia; draindm of a s::ortge ; col as & result of seal failure in the refueling cavity at a time whers the transfar tube gates to the poo'i were open, of when transfor of a spect fuel assembly was in progress.5 The Haddam Neck incident occurred during preparations for refueling. An inflatable seal bridging the annulus between the reactor vessel flange and the reactor cavity bearing plate extruded into the gap, allowing 200,000 gallons of borated water to drain out of the refueling cavity into the lower levels of the containment building in about 20 minutes. Gates to the transfer tube and the fuel storage pool were in the closed position, so no water drained from the pool.6 More recently a pneumatic seal failure in the Hatch spent fuel basin which released approximately 141, 000 gallons of water resulted in a drop in water level in the pool of about five feet.7 S.I.3 Report Objective The objective of this report is to provide an integral assessment of the risk potential of beyond design basis accidents in spent fuel pools. The risks are defined in terms of 1

S-3' 'I the probabilities of. Various= initiating events that might compromise the structural integrity of tha pool or its cooling capability. the probability of a system failure, given an initiating event, fu:t1 failure mechanisms, given a system failure, potential radionuclides releasas, and consequences of a specified release. This study generally 'follows the logic of a typical probabilistic risk analysis (PRA); however, because of the relatively limited number of potential-accident sequences, the analyses are greatly simplified. S.I.4 Spent Fuel Storage Pool Designs The configurations of spent fuel storage pools vary from plant to plant. 3 In BWR's, the pools are lor.ated within the reactor building with the bottom of the pool at about the same elevation as the upper portion of the reactor pres-sure vessel. During refueling the cavity poove _the top of the prec&ure vessel is flooded to the same elevation as'the storage pool, so that fuel assemblies can be transferred directly from the reactor to the pool via a gate which sep-arates the pool from the cavity. In PWR plants, the storage pool is loented in an auxiliary building. In some cases the pool surface is at about: grade level, in others the pool bottom is at grade. The refueling cavities are usually. connected to the storage pool by a transfer tube. During refueling the spent assembly is removed from the reactor vessel and placed in a contain-er which then turns on its side, moves through transfer tube to storage pool, set upright again and removed from the transfer container to a storage rack. Various gates and weirs separate different sections of the transfer and stor-age systems. More details concerning various configurations are given in Section 2.3 and Table 1.1. i S.1.5 Selection of Surrogate Cases for More Detailed Studies Two ' older vintage" plants were selected to serve as BWR and PWR surro-gates.for more detailed studies 5 The choices, Millstone 1 and Ginna, were based ' primarily on such factors as availability of data and the relative familiarity of_ the project staff with the various -candidate sites. The ) \\

0 ,S-4' o7erating histories of 'the two surrogate p.lants ~were modeled to obtain a realistic radioactive inventory in the various spent fuel-batches. )j S.2 ACCIDENT INITIATING EVENTS AND PROBABILITY ESTIMATES Accident initiating events that have 'been considered include pool heatup due to loss of cocif eg water circulation capability' structural failure of pool due to seismic events or missiles, partial dratndown of pool due to pneumatic real failure, and 1 structural failure of pool due to a heavy =1oad drop. Estimates o'f the likelihood for each of these initiators are provided in i Section 2. It is concluded that the dominani, initiators are structural fail-I ures resulting from a seismic events (-2x10-3/Ry) and - heavy load drops (-3x10-3/Ry). Sncertaintics in the probability estimates are quite large, being at least an order of magnitude iri either' direction. In the case of seismic events, the seismic hazard and structural fragilities both contribute ) to the uncertainty range. fop heavy load drops, human error probabilities '&nd structural damage potentials are the primary sources ~ offncertainties. l l S.3 EVALUATION OF FUEL CLADDING FAILURE The SFUEL computer code developed at Sandia National Laboratories (SNL) l by Benj amin et al.,3 analyzes the behavior of spent fuel assemblies after an accident has drained the pool. The analyses predict that self-sustaining oxi-dation of the Zircaloy cladding (i.e., a cladding fire) would occur for a wide I range of decay heat levels and storage geometries. Several limitations in the SFUEL analyses had been recognized in Reference 3 and have been addressed 'in a modified version of the code, SFUELIW." The BNL evaluations of SFUELIW have led to the conclusions that the modi-fied code gives a reasonable estimate of the potential for propagation of a 1 cladding fire from high power to low power spent fuel and that the code prol I vides a valuable tool for assessing the likeliholod of a catastrophic fire.for y a variety of spent fuel configurations in the event that the pool is drained. s -u

S-5 S.4 ' CONSEQUENCE EVALUATION Radioactive releases were estimated for the two surrogate plants for five cladding failure scenarios predicted by SFUEL calculations. S.4.1 RadioactpeInventories The ' radioactive inventories contained in the spent fuel pools (as of April 1987) for Millstone 1 and Ginna were calculated using the ORIGEN2 com-puter code,s based on the oparating histories of each of the plants (Appendix A). The calculated data included the 1987 inventories for each fuel batch discharged at each refueling over the operating history. i S.4.2 Release Estimates Fractional releases for various groups of radionuclides were estimated based on the physical parameters characterizing the SFUEL failure scenario. Thus, four source terms were estirrated corresponding to the four accident scenarios. S.4.3 Off-Site Ragological Consequences Off-site radiological consequences were calculated using the CRAC2 com-puter code.8 Becaun 'of several fe6tures in the health physics modeling in the CRAC2 code, the popuiation dose results appear to be of limited valut.. The most meaningful measure of the accident severity appears to be the inter-dir. tion area (contaminated land area) which in the worst cases was about two onfers of magnitude greater than for core-melt accident. No " prompt fatali-tids" were predicted and the risk of injury was negligible. S.5 RISK PROFILE The likelihood and consequences of various spent fuel pool accidents have been combined to obtain the risks which are sumarized in Table S.1. As noted above, the population dose results are of limited value because they are driven by decont. amination levels as' signed within the CRAC2 code. Thus the

y S-6 land interdiction area is included in Table 5.1 as a more meaningful reprs sentation of severity. The uncertainty in each of these risk indices is esti-mated to be an order of magnitude in either direction and is due principally to uncertainty in the fragility of the pools and uncertainty in the seismic bazard. Table S.1 Estimated Risk for the Two Surrogate Spent Fuel Pools from the'Two Dominant Contributors I Spent Fuel Interdiction Accident Pool Fire Health Risk Risk Initiator Probability /Ry (Man-rem /Ry) (Sq.Mi./Ry) Seismic induced PUR pool failure 1.6x10-5 37 8.4x10 4 Seismic induced BWR pool failure 1.8x10-6 4 7.6x10-5 Cask drop

• induced PWR pool failure 3.1x10-5 7'

.001 Cask drop

• induced BWR pool failure 2.5x10-6 6

1.1x10 4 )

• After removal of accumulated inventory resumes.

(Note that many new plants have pool configurations and administrative procedures which would preclude this failure mode.) The overall risk due to beyond design basis accidents in spent fuel pools for the PWR surrogate plant is about 130 person-rem /Ry and About 12 person-rem /Ry for the BWR surrogate. These estimates are comparable to present esti-mates for dominant core melt accidents and appear to warrant further attention on this basis alone. However, the unique character of such an accident (sub-stantial releases of long lived isotopes) makes it difficult to compare to reactor core melt accidents. The exposure calculations are driven by assump-tions in the CRAC modeling and the results are not sensitive to the severity of the accident. In terms of interdiction area this type of accident has the potential to be much worse than a reactor core melt accident. Note that the risk results are calculated for two surrogate plants and may not be applicable to generic pool types.

i S-7 S.6 CONSIDERATION OF MEASURE WHICH MIGHT REDUCE CONSEQUENCES A number of potential preventive arid mitigative measures have been pro-posed but the only one which is judged to provide a substantial measure of risk reduction is a modification of the spent fuel storage racks themselves. For those plants that use a high density storage rack configuration., Improve-ment in the air circulation capability is estimated to result in risk reduc-tion up to a factor of ten. S7 References for Summary c 1. "A Prioritization of Generic Safety Issues," Division of Safety Technolo-gy, Office of Nuclear Reactor Regulation, U.S. Nuclear Regulatory Commis-sion, NUREG-0933, December 1983, pp. 3.82-1 through 6. 2. " Reactor Safety Study, An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," U.S. Nuclear Regulatory Commission, NUREG-75/014 (WASH-1400), October 1975, App. I, Section 5. 3. A.S. Benjamin, D.J. McClosksy, D. A. Powers, and S.A. Dupree, " Spent Fuel Heatup Following Loss of Water During Storage," prepared for the U.S. Nuclear Regulatory Commission by Sandia Laboratories, NUREG/CR-0649 (SAND 77-1371), May 1979. 4 N.A. Pisano, F. Best, A.S. Benjamin and K.T. Stalker, "The Potential for Propagation of a Self-Sustaining Zirconium 0xidation Following Loss of Water in a Spent Fuel Storage Pool," prepared for the U.S. Nuclear Regu-latory Commission by Sandia Laboratories, (Draft Manuscript, January 1984) (Note: the project ran out of funds before the report was pub-lished.) 6. IE Bulletin No. 84-03: " Refueling Cavity Water Seal," U.S. Nuclear Regu-latory Commission, Office of Inspection and Enforcement, August-24, 1984 l 6. Licensee Event Report, LER No. 84-013-00, Haddam Neck, Docket: No. 50-213, " Failure of Refueling Pool Seal.," 09/21/84.

S-8 7. Nucleonics Week, December 11, 1986,.pg. 3-4. ) 8. A.G. Crof f, "0RIGEN2: A Versatile ' Computer Code for Calculating' the Nuclide Compaition and Characteristics of Nuclear. Materials," Nuclear. ] 1 Technology, Vol. 62,.pp. 335-352, September 1983.. j

9..L.T. Ritchie, J.D. Johnson and R.M. Blond,, Calculations of Reactor Acci i

dent Consequences Version 2, CRAC2: Computer Code User's Guide, prepared ' by Sandia National Laboratories for the U.S. Nuclear RF.gulatory Commis-sion, NUREG/CR-2326 (SAN 081-1994), February 1983. ) 1 I 1 ../ i l I i' e o j l l >j

l[ e n 1-1 j

7. n 1.

INTRODUCTION s J, Generic Safety Issue 82, "Beyond Design Basis ' Accidents in Spent Fuel Pools," was assigned MEDIUM priority in.. November '1983.1 In its prioritiza-tion, the NRC staff. took account of two factors that had not been considered. in earlier risk assessments:2 1. Spent fuel 'is currently being stored rather than shipped for repro-- cessing or repository disposal, resulting in much larger inventories ' of spent assemblies in reactor fuel basins than had previously-been anticipated; and, N suggests 1 the possibil'ity of catastrophic. Zirca-. . ( 2. A theoretical model 3 loy fire, prop"agating from assembly to assembly in the event of com - plete drainage of water from the pool. t 4' t,{ 1.1 Previous Investigations The Reactor Safety Study 2 (which did not take account of the two factors above) concluded that the risks associated with uspent fuel storage were ex-tremely small in comparison with accidents associated with the reactor core. Thst. conclusion was based on design and operational features of the storage pools which made the loss of water inventory highly unlikely, e.g., <? The pool structures were designed to withstand safe -shutdown earth-

quakes, The fuel racks were designed to preclude criticality, Pool design and instrumentation precluded inadvertent and. undetected

( loss of water inventory. Procedures and interlocks prevented the' drop of heavy loads on stored assemU; snd s Th. .w..ve structures were designed to accommodate the forces and missiles generated by violent' storms. Probabilities of pool failures due to external events (earthquakes, mis-siles) or heavy load drops were estimated to p in the range of 10-6fyg p, j X m 'l l

1-2 ) y Radioactive release estimates were based on melting of 1/3 of a core for var-ious decay periods, with and lwithout, filtration of the building atmosphere (see Ref. 2, Table I 5-2). i i Subsequent to th'e Reactor Safety Study, A.S. Benjamin et al.3 investigat-ed the heatup of spent fuel following drainage of the pool. A computer code, SFUEL, was, developed to analyze thermal-hydraulic phenomena occurring when storage racks and spent assemblies become exposed to air. The computer model takes into. account decay time, fuel assembly desigh, storage racks design, packing denshy, room ventilation and other variables that affect the heatup of the fuel. i Calculatidm.with SFUEL indicated that, for some storage configurations and decay times, the Zircaloy cladding could reach temperatures at which the extthermic oxidation would become self-sustaining with resultant destruction of the cladding and fissien product release. The possibility of propagation to adjacent assemblies (i.e., the cladding would catch fire and burn at a hot enough temperature to heat neighboring fuel assemblies to the ignition point) was also identified. In such cases, the entire inventory of stored fuel could ) become involved. Cladding fires of this type coul'd occur at temperatures well below the melting point of the UO fuel. The cladding ignition point is about 2 900'.C compared to the f e1 melting point of 2880*C, 9 Uncertainties in the SFUEL calculations were primarily attributed to un-certainties in the zirconium oxidation rates. Further work was done to refine th'e SFUEL computer mooel and to compare calculated results with experimental data." These more recent results have generally confirmed the earlier concepts of a Zircaloy fire which, given the rt,ht conditions, will propagate to neighboring assemblies. However, compari-sons to out-of-pile heat-up data have not shown good agreement with the code. The authors noted.that more work in several areas was needed to define more precisely the conditions and configurations which allow or prevent propaga-tion.

1-3 1 . Several studies have been conducted on alternative spent fuel storage concepts. Among these is a report published by the Electric Power Research Institute (EPRI), which applies probabilistic risk assessment techniques to several storage concepts.5 While this study does not directly address Generic-Safety Issue 82; however, it does provide useful' insight on appropriate analy. tical methodology as well as useful data on an in-ground (on-site) storage pool. 1.2 Related Events There is no case on record of a significant loss of water inventory from t a domestic, commercial spent fuc1 storage pool. However, one recent incident occurred at the Haddam Neck reactor that raised concern about the possibility of a partial draindown of a storage pool as a result of seal failure in the refueling cavity at a time when the transfer tube gates to the pool were oped, or when transfer of a spent fuel assembly was in progress.6 The Haddam Neck incident occurred during preparations for refueling. An inflatable seal bridging the annulus between the. reactor vessel flange and the reactor cavity bearing plate extruded into the gap, allowing 200,000 gallons : of borated water to drain out of the refueling cavity into the lower levels of the containment building in about 20 minutes. Gates to the transfer tube and the fuel storage pool were in the closed position, so no water drained from the pool.7 However, had these gates been.open at the. time-of the leak, and had they not been closed within 10 to 15 minutes., the pool would have drained to a depth of about 8.5 feet, exposing the upper 3 feet of the active fuel re-gion in the spent fuel assemblies.7 Also, had the transfer of spent fuel been in progress with an assembly on the refueling machine, immediate action would have been necessary to place the assembly in a safe location.under water to limit exposure to personnel. The NRC Office of Inspection and Enforcement required all licensees to promptly evaluate the potential for refueling cavity seal failures.6 Re-sponses indicated that the refueling cavity configuration at. Haddam Neck is un.ique in that the annulus between the reactor flange and the cavity bearing plate is more than 2 feet wide. In most plants this gap is.. only 2 inches

T s' 1 y i 1 q l wide.s About 40 operating (or soon to optrate) reactors ute' inflatable .<.Y seals. However, because of dusign differences, the 'Haddam Neck. failure does / not appear t3 te directly applicable?to the other planti., It is noted that 2 BWR plants hive permanent steel bellows seals to fill' the gap between the L 7 l(, reacter flitsge' and' the ravity bearing ' plate. This issue is discussed 'more { fully in Soction 2.3. i 1.3 Rfsk' Potential ~ i Tb risk potenti:als of "Beyond Design Basis Accidents in Spent Fuel l 4 l Pools" are defined in tenns of \\ the probabilities.s Lof various initiating events-that might compromisa the structural integrity of the pool or its cooling capability, j the probabf Mty-of a syste.n failure, given an initiating event, fuel faildre mechanf sms,.given a system failure, potential radionuclides relseses, and consequences of a specified release. - This study generally follows the logic of a typical probabilistic risk analysis (PRA); however, because of the relatively liiaited number of potential accident sequences, the analyses are greatly simpTified. 9 1.4 Liscuss' ion of Spent Fuel' Storage Pool Designs and Featur(ts. 1 The general design criteria.: for soeat fudl storage facilities are stated E in Appendix A of 10 47R 50,8 and are dis,cassed more full'y in Regulatory Guide 1.13.10 l The pool styctures, spent fuel racks arad overhead cranes must be design-ed to Seismic Category I standards. It is reqJf nd that :the systems be de-signed (1) with capability to permit appEcgristi periodic inspection and test-ing of components important to safety,'(2) with r,uitable ' shielding for. radia-l tion protection, (3)- with appropriate containment,) cerfinement, and filtering systws, (4) with a residual heat removal capability having reliability and } testability that reflects the importance to sty of ' decay heat.and other - l 9 - .h -( -w

1-5 residual heat removal, and (5) to prevent significant reduction in fuel stor-age coolant inventory under accident conditions.8 The configurations of spent fuel storage pools vary from plant to plant. Table 1.1 if sts various information about the pools for licensed plants. In BWRs, the pools are located within the reactor building with the bot-tom of the pool at about the same elevation as the upper portion of the reac-tor pressure vessel. (For example, at Oyster Creek the bottom of the' pool is at elevation 80'6", and the top at 119'3". The water depth is 38 feet.) Du r-ing refueling, the cavity above the top of the pressure vessel is flooded to j the same elevation as the storage pool, so that fuel assemblies can be trans-ferred directly from the reactor to the pool via a gate which separates the pool from the cavity. In PWR plants, the storage pool is located in an auxiliary building. In some cases the pool surf ace is at about grade level, in others the pool bottom is at grade. The refueling cavities are usually connected to the storage pool by a transfer tube. During refueling the spent assembly is removed from the reactor vessel and placed in a containt:r which then turns on its side, moves through transfer tube to storage pool, set upright again and removed from the transfer container to a storage rack. Various gates and weirs separate dif-ferent sections of the transfer and storage syst' ems. More details concerning various co6 figurations are given in Section 2.3. 1.5 Selection of Surrogate Cases for More Detailed Studies Two " older vintage" plants were selected to serve as BWR and PWR surro-gates for more detailed studies. The choices, Millstone 1 and Ginna, were made somewhat arbitrarily, based primarily on such factors as availability of data and the relative familiarity of the project staff with the various candi-date sites. The operating histories of the two surrogate plants were modeled to obtain a realistic radioactive inventory in the various spent fuel batch-es. Details of the modeling procedures and a listing of the calculated radio-nuclide content are presented in Appendix A. 4 _ _ - _ - _ _ _ _ _ _ _. - _ - - _ - _ _ - _ - - - ~ _ _ - - _ _. _ - - - _

1-6 It should be noted that both surrogate plants have relatively large in-ventories of spent fuel assemblies in their spent fuel basins. ,I 1.6 Report Content Accident initiating events and their probabilities are covered in Section 2. Fuel cladding failure scenarios based on the SFUELIW Computer Code are evaluated in Section 3. Included are sensitivity analyses of the failure scenarios arising from uncertainties in Zircaloy oxidation reaction rate data, and hardware configuration assumptions. Section 4 presents data on the poten-tial for releases of radionuclides under various cladding failure scenarios and compares the projected releases with releases associated with severe core accident sequences. In Section 5, risk profiles are developed in tems of person-rem population doses for several accident sequences. Section 6 considers measures that might mitigate beyond design basis accidents. 1.7 References for Section 1 1 1. "A Prioritization of Generic Safety Issues," Division of Safety Technolo-gy, Office of Nuclear Reactor Regulations U.S. Nuclear Regulatory Comis-sion, NUREG-0933, December 1983, pp. 3.82-1 through 6 2. " Reactor Safety Study, Ar. Assessment of Accident Risks in U.S. Commercial Nuclear Power plants," U.S. Nuclear Regulatory Comission, NUREG-75/014 (WASH-1400), October 1975, App. I, Section 5. 3. A.S Benjamin, D.J. McClosksy, D. A. Powers, and S. A. Dupree, " Spent Fuel Heatup Following Loss of Water During Storage," prepared for the U.S. Nuclear Regulatory Commission by Sandia Laboratories, NUREG/CR-0649 (SAND 77-1371), May 1979. 4. N.A. Pisano, F. Best, A.S. Benjamin and KoT. Stalker, "The Potential for Propagation of a Self-Sustaining Zirconium Oxidation Following Loss of Water in a Spent Fuel Storage Pool," prepared for the U.S. Nuclear Regu-latory Commission by Sandia Laboratories, (!1rafi: Manuscript, January 1984) (Note: the project ran out of funds before the report was pub-lished.)

1 5. D.D. 'Orvis, C. Johnson, and R. Jones, " Review of Proposed Dry-Storage Concepts Using Probabilistic Risk Assessaent," prepared for the Electric Power Research Institute by the NUS Corporation, EPRI NP-3365, February 1984 6 IE Culletin No. 84-03: " Refueling Cavity Water Seal," U.S. Nuclear Regu- -latory Commission,. 0ffice of Inspection and Enforcement, August 24, 1984.. l 7. Licensee Event Report, LER No. 84-013-00, Haddam Neck, Docket No. 50-213, "Failure of Refueling Pool Seal," 09/21/84 8. Licensee Responses to NRC IE Bulletin No. 84-03. Code of Federal Regulations, Tit [e 10, Part 50; " Domestic Licensing of-9. Production and Utilization Facilities, Append'x A, ' General Design Cri-teria for Nuclear Power Plants,' General Design Criterion 61, ' Fuel Stor-age and Handling and Radioactivity Control'." 10. U.S. Nuclear Regulatory Commission, Regulatory Guide 1.13, " Spent Fuel-Storage Facility Design Basis," December 1981. ( I. .____m______._.__ .m__-_._____ _ --- _ _ m_-_mW

g ne legs 0 2 2 2 1 1 2 2 2 2 1 1 1 sii 0 0 0 0 0 0 0 0 0 0 0 0 0 l ss = = = = = = = = = = = = = eea t E E E E E E E E E E E E SDB 8 B 8 B 8 8 8 8 B B B 8 8 0 D 0 D D D D D D D D D 0 4f 8 o 9 l 1 s o n od ro P n d e e e e e e e e e e e ei e r l l l l l l l l i l l bt ei g e e e e e e e e s e A e ma gt / ec aa N co rc B S S S S B 8 S 8 S B 8 el oo A 'R R R R R R R R R R R D tt S f y oro st an C e e sv y s ea t oa) v i itCt r-v n oe ieee 6 h td t vcc 9 1 2 1 0 2 9 3 3 0 ~6 0 1 ni cin el at er 4 6 8 6 6 0 2 3 C 6 5 0 6 vc oare 4 3 7 4 4 4 7 2 3 nu il ep i n d ef( o aR e ei R R gd aa rr o tf b = so ye rr e "e oo l tC us n f a ef h c vo 5 0 6 1 9 5 0 8 8 7 6 0 5 t n 0 4 1 3 8 6 8 4 7 5 4 0 2 ne I s ec n 2 1 1 2 1 1 1 0 2 1 1 0 0 pn do se ei r rt ee oc rf t a ae 5r r. F d" s e l deo uh o l t p ) c a s nhf ye I t o rl i oi ws l tb e enm f 9 h Sss uee 2 8 g 8 6 4 5 1 4 6 6 0 Nna F vs 7 6 8 6 5 2 8 2 1 7 1 0 4 I ob ns 1 0 8 7 0 9 9 2 0 5 8 1 Ss tIA 1 1 1 2 Ain n Brg edf ai peo . Eps Sr Gme o. s A od t o Ec SN O c ( T ,l See j gs Lal E re s Uos e Ft i sd fla T n ob e Nna mr i Ei reo 4 4 4 1 0 0 8 4 4 4 8 0 A 0 P esc 8 6 6 6 6 5 4 6 2 2 5 6 / 6 Ses bs 7 7 7 5 5 5 4 7 7 3 5 N 5 rn uAn Noi s i h Ocs Ii a t Afb u A l To F Dsenu of

i l

stt ar) 0 3 3 3 6 6 1 0 7 7 8 6 3 6 'R a e met 4 9 9 g 3 3 8 0 2 2 5 3 3 3 cn r wW 2 2 2 2 4 4 3 7 5 5 6 4 8 4 Wrp eoM 3 3 3 2 2 2 2 2 1 2 3 2 Bf s hP( T 1 1 e t 1 2 3 l b n i y y y a d I T' o r r r 1 2 l k t P r r r o c f n e e e k k 1 2 3 n l a k F F F c c r ir u l c i i n n n A t G 1 P o s s s w w r e e e a R n n n s s e d d d e p d h w w w n n p s s s n z n c g o o o u u o e e e a t a t I r r r r r o r r r u i r e S B B B B B C D D D D F G H

k s 1*.%i t 1 1 l - = e e m e e,. ... e e e Q-sh a N a a m - N N N N sn e w 5 &* N N 7 7 7 7 9 7 7 N N " * ~ ~ a a a a a a o a o a. o a-o o a 3 8": aaaaaaaaaaaaaaa E 8E N aE E 3 3 8 88 8 8-13

8..

% =

4. % 55t % % %.51 %

4 s g.3 ..u... to5w M * * *

• 9 a.
• M **
• a
• ~

- -. t. 5 a * * * *

• A 0 3 3 3 R S:

stu lat.- i a be. .v g 2. 3. 3.. S. S.. 8.. M.

2.. A. 3 2 3..*

R. . ~. oo.~ ~ ~ ~ -.- - ~o I:s t 8

• t

~ m. g* s.' -.. - ~ ~ ne... a R_.- .~ 5,- At* i .U. l 4*. 1 ',I 3 g# # E. E E. g *. M g3 3gg z i _ ~ l l 7 tr. 2 2 O

R SR 2.a I =n n=.

['g g,]O a a 2 E - ~- a a 2 ~ ~ ~ m - - - n n - ~. n. ~

• 2 t 5

.8 5

a. a.

-a - ~ . -. u.. 3 4 aat .E % ='-w a 2-4 = = ~ o a 6 -o o -m.. 6 3 1 a g g g h w 6 E .v y a 1 9 l y 1 2 3 E..N E &2 2 2 & &' ~ a

n l. imgs 0 0 0 0 a o sii = = = = se r i ss E E E E h g o eea 5 5 S 5 t a f S D 8, 5 5 B 5 nr oo d mt g e s n s . o n i u in d .l si l s. i at od el u n pn e e e ge b t e o l l l ru nn ei e e e A af a eo gt / h aa . m cf. f l pam rc t t S 5 so 'n vo oo R R R 8 i e ic tL 9 de t u S 1 g t ql ta e ea y s v t r ar e rn a l o l ee u '.f e r '.t e n y s a sd h er t oa) J ee g t e i tCt gt i ah v n rc h u ieee 0 0 8 0 1 ae qe t vcc hr t hh cin 0 0 L O sa r o cr a tt at er eare N ic l as ll ep d e ei def( n e aRe 9 l e p sn R R ae i w ub = so o l n ah s . w o nt e bs e s t V ao l. i e k n e ny l l e ys n gr b t h rr 0 es it m b t oo 2 t a l sn e m tC 0 h e eE s e f n 0 m v d s s o ef o* e vo 8 0 9 0' G rm l =;

• a s

n 0 0 1 0 a l E fr A R a ) I s R e e E W R t d n 0 0 3 O U yT d S. O W o de N r a Bp B t 't el oe r r 6 n rt tc g A 5 4 e o oc s nr 8 2 r C t a r eu t ) 1 9 a ( 5r e vo a g0 F t nS (0 e o d 1 c i l 1 t t e t 1 t en p tF n n c a e e n t e dc e eR t t i e ) re tC e e l a s g or -- a l l a b ye n t e re a a t a ri i sf d el v v a T ol t e yl e i i d l tb a ,R in u u enm r e "o t ci q q e uee 4 e f , c e e h F vs 0 0 7 0 p g ad T l es 8 'O t n e tlA 1 8 i l n s s n d 8e d ai e e edf e 0v i nf t f e peo s 3i t ae i i r 5 r n t u l d b u e fa b. t m d o. t t o c ol as a e c SM t e y t e s s u ( L rr r s r i ved e a a t wm l ak s n or l ra R R o pe i gu W W e s i t a q O B n e s l v eh o f s ) ae a ht 6 4 f i"e i e mc t r 5 6 a ob m r r r = a 6 5 l mr m e eu f e reo 4 4 8 A o C ho B o + esc 6 6 6 / C tS A n _ + s bs 7 7 3 N n eu R R e l man y i a ' g. ne W h e u I nl r id p P p o s ss n tt e u t e ee i cu 0 4 o F a i mi M nh 6 4 w l l ul us 1 1 t u b sb t f g m sm u e e e e e e ae b af g g v l R s s a a a a e s r ss r r h l s 's a ar) 3 3 3 3 r A e a o o b eet 9 9 3 2 a ( r t = t t 3 a r wu 2 2 5 3 e / e0 c s s s l eoM 3 3 1 3 l i T5 a iE d hp( c s 7 e sS a n n a i T u e e1 r aS t i a v N i c b a l ry =

s s

2 b ur ns a a t 5 m oo t gt h h s o e St R i n t n 2 s n sa 1 e U s ee ei i 1 2 e 2 A cv d p k k U t n a a a n k nn n c c a d eI e ce n d t n n a l e e r t i g i l w u e n n n Y c s c r el t ma s s d = a a a u e r e f a a st n n s l h h t N t u t et c i n u u e A p e e n o o S R o o ei r r r / u u o n S ( " t L S v B B D N q q m h t s s r s o u u e a o ) ) ) ) ) ) ) ) ) S S V W F a b c d e f g h 1 I t j

n le g's 0 6 9 0 0 0 0 0 0 0 0 0 0 0 sii = = c = = = = = = = = = = = ss E E E E E E E E E E E E E E lee a B 8 5 5 5 8 8 5 5 5 5 O D 5 SDB D 0 5 5 D 9 5 5 5 D D 5 4f 8o 9 1 s n 6 ro P e d d d d d d d d d d 'd d d d ei e r r r r r r r r r r r r r r bt el g g 3 g g 3 g g 3 3 g 3 3 3 ma gt ec aa co rc B S O S S S B B B B B S S B el oo A A F A A A A A A A A A A A D tl 5 f y oro stan C e e sv y s en t oa) ii itCt r v n e g oe i eoe 3 7

6. e 0

0 1 6 2 3 td t v A A A ciwc ni 6 6 7. 0 / 8 4 / 3 4 1 / g el at r 5 2 1 N 0 N 3 2 3 I 1 I vc otme 1 nu ii e p in d et( o aR e ei R - R gd aa rr o tf b so ye rr l "e oo ~ e tC es n fa ef g* g c vo g 5 5 0 0 7 7 2 3 t n 1 3 8. 0 A A 8 g 0 1 A ec n 2 9 0 0 N I 7 ne is / / / pn do '4 I 2 9 1 I 0 I se ei r rt ee oc rf t a ae Sr r. F d" s e l deo oho l t p - ) c a s nhf ye I t o ri i ol ws ltb e enm Sss uee g g' Nna Fvs 8 8 4 A 8 A 3 1 g A 4 I ob es 8 6 0 C / 6 - / 5 7 g / 1 S s tlA 3 1 1 N 9 I 5 1 I N 1 I Ain n Brg edf ai peo Eps Sr G me o. A od t o S c SI l O e ( T ,l j See - gs Lat Ere s II cs e Ft i sd fla T n ob e Nna mr Ei reo 7 7 7 A A 7 7 A 3 3 7 7 A 7 P esc 7 7 5 / / 1 1 / 3 3 7 7 / 5 Ses bs 1 1 1 N N 2 2 N 1 1 1 1 N 1 rn man 1 1 oi u I 0cs Il l a e Afb e To a F A l Dse - ne of

i l

stt ar) 8 5 0 ( 1 0 0 0 1 4 2 8 2 cn met 5 1 6 ' / 4 7 7 / 2 4 5 7 3 6 A 1 0 0 A 5 1 4 7 3 5 'R a e r wf 5 8 6 i l rp eon 2 2 2 N 3 2 2 II 3 3 2 2 3 2 iPf s hP( T ) d 't 1 2 n 1 3 1 o s s C y f f r l n ( e f f e o e l i i v e y 1 t 1 2 l 1 l l t s n n a C C 1 R s a 1 a s s v y e C 1 l a a 1 a t t a l B e P s s r w r r b 1 2 a e y l n n e n a e e w t s l e b a a v o l v v a k k s i b l a 'k k a r l l l t o o y v a r T r r e y a a a a o o r a l a A A S S C C C C C C C D O F

g g g g ig g 9 g 3 3 g o i 0 7 0 5 5 2 t 5 5 7 2 2 0, c I' ne l. l. 2 1 1 1 1 l. 1 1 1 1 1 I. iags 0 0 0 0 '0 0 0 0 0 0 0 0 0 0 0 sit = = = = = = = = = = = = = = l ss E E E E E E E E E t E E E E E eea 5 8 8 8 8 8 8 8 8 5 5 8 5 5 8 SDB 5 0 D 0 D 0 b 0 0 5 5 0 5 5 0 l o od Pa d d d d d d d d d d d 1 d d d o r r r r r r r r r r r r r r el g g g g g g g g 3 3 3 g g g 3 gt aa rc B S S S B S B B B B S B B e B oo A A A A A A A A A A A A A ~. A tL S 6 e y s t oa) i tCt v n 9 9 i eee 5 4 7 4 4 3 5 g 1 8 g 5 t vcc b cin 0 4 2 3 7 1 6 g 6 / 6 8 0 A at er 1 3 4 6 4 2 3 6 1 N 4 3 5 oare il ep 1 def( aR e R R b ye rr oo tCnef 9 g, vo 3 g 1 7 2 3 1 6 7 3 0 6 3 2 8 4 T7 7 2 6 4 A n ) 7 4 8 I s d / n 0 2 2 3 1 0 2 2 0 II 't do 1 1 5 ei n rt o oc C t a ( S r F 1 1 e ) l a s b ye a ri T ol l tb enm eee g g F vs 2 5 0 5 0 2 0 8 7 1 A 6 0 7 ns 6 0 4 4 6 3 4 6 7 3 / 7 2 3 tIA 3 3 5 1 3 1 2 5 N 3 2 0 n edf 1 peo Sr o. t o SN ( se - i fla ob e mr reo 7 3 1 7 h 3 3 1 7 3 A 7 7 7 7 esc 5 3 2 5 0 3 3 2 1 3 / 1 5 5 7 bs 1 1 1 1 1 1 1 2 1 I 7 1 1 1, I man u i Nl eu F lar) 2 s 0 5 8 5 0 0 1 1 0 5 5 8, met 5 0 2 2 h 5 2 5 3 1 1 0 7 7 4 r wW 6 5 5 8 7 0 6 6 4 4 7 1 7 5 eoM 2 1 1 1 2 3 1 2 3 3 2 2 2 2 hP( T b 1 3 2 n t t t e u k n n n e 2 1 2 t o c i i i k a e n h e o o o n 2 e n n 1 a 2 I P P P e a I l l n n n 1 a e T e e o A A P y C m n n n n r r t e a a a a a n e i i s h h ee l t n d i i i w l G G i r r e a e u u l t t n r r n F F i d d d d a o 4a n n n a a c c 1 I I I K M M M M I N O Io o c G

0 0 0 0 c 1 1 1 1 2 2 2 2 5 6 6 l. l. 2 2 i ne mgs 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 sit = = = = = = = = = = = = = = = iss E E E E E E t E E E E E E E E eea 8 8 8 5 8 0 5 5 5 8 8 8 8 5 5 SDe D D 0 5 0 0 5 5 0 0 D D 5 log Pn d d d d d d d d d d d d d d d o r r r r r r r r r r r r r r r ei g 3 g g g g 3 3 g g g g 3 3 g gt aa rc k S S S S S S S S S S S S S S oo A A A A A A A A A A A A A A A tL S C e. y s t oa) itCt v n g 3 i eee t vcc 6 5 7 0 7 3 2 8 1 1 0 A cin 1 9 / 5 2 0 2 1 6 8 4 0 et er 3 5 N 6 8 4 2 5 4 3 aare ll ep d ef( aRe R R ye rr oo tCnef g g , 0 0 vo 3 5 3 7 7 7 3 7 0 n 2 3 A 3 g 4 g 5 3 6 0 0 I s / n 1 2 R 4 4 1 0 1 1 0 1 0 ) do d ei rt 't ec n ta C Sr o F ( 1 1 ) a s e ye l ri b ol a ltb T enm uee g g Fvs 8 0 A 4, 1 0 2 6 5 M 7 ns 1 8 / 2 0 6 5 3 6 1 3 tiA 2 4 N 5 6 2 1 2 2 2, nedf peo Sr o. t o SN i ( s e - i f l* eb e mr reo 7 7 4 A '1 1 1 1 7 7 3 3 7 7 7 esc 7 7 0 / 2 2 2 2 7 5 3 3 5 1 1 bs 1 1 2 N 1 1 1 1 1 1 1 1 1 2 2 man h I il e. u F lar) 8 8 0 8 8 0 0 2 0 8 1 7 0 0 met 6 6 3 A 1 1 5 5 7 0 3 1 4 1 9 r wW 5 5 5 / 5 5 6 6 7 3 3 4 3 4 3 eoM 2 2 2 N 1 1 1 1 2 2 3 3 1 3 3 hP( T 1 2 d d 1 2 n n 1 1 a a 1 2 3 h h l l o t e c c s s c 2 e e e n s d a a I I e r r r a 2 3 e r e e S n f f f l d e B B e e o 1 2 e o o P, e e a V i i o s e e s t t r r h n s e h n n O O n n i e n n i i c i e e e l l i i a a n b l l n n n o 'c a a o o r r a o a a e a a c O O P P P P P P R R S S s S S ~ l

g g g g g 9 g g g 9 g g g 8 8 o 0 5 5 5 2 2 5 5 5 0 c iae 1 k. l. 1 1 1 1 1 1 2 1 1 1 mgs stt 0 0 0 0 0 0 0 0 0 0 0 0 0 ne = = = = = = = = = = = = = o l ss E E E E E E -E E

• E E

E E E N eea 5 3 B 5 5 5 5 B 5 B B B 5 SDB 5 5 D 5 5 5 D 5 D D D lyn d d d d d d d d d d d d d d o r r r r r r r r r r r r r r ei g g g g g g g g g g g g g g gt aa rc B B B B B B B 8 B B B B B B oo A A A A A A A A A A A A A A tl S Ce y s t oa) itCt v n g ieee t vcc 5 0 8 2 5 8 0 1 4 3 7 A M cin 1 3 3 / 9 4 9 0 5 2 0 / 9 A at er 1 2 4 N 9 2 5 6 6 N 1 oare il ep d ef( aR e R R b ye erroo tCn ef g vo 4 7 2 3 7 8 0 2 3 4 9 W n 3 6 6 A ) 3 8 I s 1 0 6 8 h. k d / 2 n 0 0 1 N 0 3 1 0 1 2 '2 't do /M 3 ei no rt oc C t a ( SrF 1 1 e ) l a s ba ye T ri ol l tb enm uee g Fvs 5 0 2 A 2 6 8 2 5 0 k 0 ns 6 3 5 / 5 0 0 0 1 4 3 / 5 tIA 1 3 N 6 2 3 4 4 N 2 nedf peo Sr o. t o SN( se i fla ob emr reo 3 3 7 A 7 7 7 7 7 3 7 7 A 5 esc g g 1 / 5 5 5 7 7 9 5 5 / ? b s I I 2 M 1 1 1 1 1 1 1 man Ni 1 N u i cu .- F l ar) 1 1 0 0 5 1 1 5 aet 1 1 0 6 7 4 4 3 i 1 0 0 A G rwW 4 4 7 5 7 4 4 5 i 4 2 2 N 6 1 0 0 / r eon 3 3 2 2 2 2 2 2 3 2 2 t hP( T 3 4 1 2 t. t t 1 2 r n 3 e n e e e e lo o d n i w a h h lc c 1 Z l 1 l l l 1 l a a l2 P P r a M M - P y y u u r d d n y y fo n. o o L L e y y en en a e e r u u m r r ea ea j k k e q q . m r r rl rl o r r t t e t t u u u hs h s r u u a 4c. 5 s 5 S S S S TI TI T T T W V

g 0 0 i sii s. r o iss E E he n o eea 5 5 t t g o f sDB 5 na i e or t d n mo c e o t v s d ns e u i l l s o sn e s. a i at w g e h n P n d d gl g t e s o r r re i nn ei g' g i gt au h eo s hf l p aa c t am y rc B S 5 sf a vo a l oo A A 8 io i c n tL S 9 d l u a 1 e o ql t g o ea e y sa p t t r a rs a a l r = oo r C u o r o e n .f e ,t y s b a s er t oa) J ed k o a i tCt ge l ah e l v n g rt e u i eee 3 1 ac v qe e ~ t vcc ,he e hh r cin 5 cr l tt o at er 4 o sr r m oare 1 N o e as i il ep dc d ei def( a r aRe g l n r sn e R-R ae g iw t ue so a l nb t ah L o n a bs V at b o l ey ye nn o gr E rr 0 e o t t S oo 2 ts p sn D tC 0 a eE n 0 mh = d 2 ef g o vo 7 G r g I . f "s =A n = 4 E I s E 'n r n 4 NR ye g O. do I rT n Dp - i o ) ei d rt o i p t t e d A a 't oc s nc l t a r er i ). r n Sr e o o vu u g0 C - F l t eo b (0 e ( c lS 1 c a l n e de e 'eR c 1 R ec e lF a ) rn f tC E. 1 a s g oe s Dt ye n t r = r0 s Dl e ri i se e1 l u l ol b t ,f B l a ltb a ".Fa e F en t a enm r tR ci o e T uee g p It' e i c t ki Fvs 3 g ad ar ns 6 O o n e e ug tIA 8 0t i l n r qs n d 0 d ai a h o edf e 0e l nf tH peo s 3v i oe n r Sr n i u id w ae o, e ft b t o Eh t o c oa as h Sn t i l y t a s n ( L re r i gh wr a ve s it l at e si n om i ra i ew o pr t gu r D s i e u q t d e s l t v eh n ee

• f s

) a a ht fta l e me t r 'E l t ob e e r rc = a n ui ba mr m o er f e . u oc reo 3 3 a C h u 8 o s

• o Do esc 9

9 C t o A n t d " s bs 1 1 y I a n S nw mAa i t s oo n u ea u l r g id u h h lIl o s s. n tt s tg e t e es i cu o n u a i me d ah w y w d 5 F l l ui l rs t l o e0 u b sl i f g m sb u e y n t uf t d s e e am b af . b e u o R s e a e n h y l s "m s s r ss l d a s l e ar) 0 0 r A o a b e m lt met 5 5 a ( ra t = a r r y aa r w 2 2 e / e c s l a e l eoW 3 3 l ) T0 a iE i h p e i q nu hP( c s 0 e sS a s t gh T 2 u e e7 r aS v s i it ~ N i c b a n i n rr e l ry = i i oa l b ur ns t s 1 f e b 5 m oo t gt o a e n a e St R in n b t d on T s n I sa n n ya r I 's ee el a l i I n-A cv dp t e o ae o nn ng a u P s Ct f d eI on ce d f i a t e e r ii i g g ol n s c r el td ma = t a a e r o f a al st n l 2 l u l t u t et ci i n A e d I as bt p 1 2 o o S R o ou ei / p a M n S ( " t Lb S v N S l T Dp i o n n t oio oo ) ) ) j ) ) ) ) ) ) i Z Z F a b c l e f g h f j ~ !l l

2-1 2. ACCIDENT INITIATING EVENTS AND PROBABILITY ESTIMATES 2.1 Loss of Water Circulating Capability The spent fuel basins of U.S. nuclear power stations contain a large in-ventory of water, primarily to provide ample radiation shielding over the top of the stored spent fuel.. Some typical pool dimensions and water inventories are shown in Table 2.1. The heat load from decay heat of spent fuel depends on decay time since the last refueling. Heat loads for the entire spent fuel inventory of the two older vintage surrogate plants are shown in Table 2.2 (data extrapolated to the 1987 scheduled refuelings). The cooling systems provided for spent fuel pools typically have a capacity in the range of 15 to 8 3 20x10 Btu /hr (4.4 to 5.9x10 kw). In the event that normal circulation of the cooling water is disrupted, e.g., due to station blackout, pump failure, pipe rupture, etc., the water temperature of the pool would steadily increase until bulk boiling occurred. I' a situation where the stored inventory was small, an equilibrium (Note: n temperature, below the boiling point, vould be reached at which surface evap-oration balanced the decay heat load). Thermal-hydraulic analyses of the consequences of partial or complete loss of pool cooling capability are a routine part of the safety analysis re-ports required for licensing and amendments thereto. Generally, these analy-ses consider several scenarios ranging from typical to extremely conservative conditions. A sampling of conservative results for several plants is given in Table 2.3. The data clearly demonstrate that the time interval from loss of circulation until exposure of fuel to air is quite long. Even in the most pessimistic case cited in Table 2.3 (Docket No. 50-247), the water level in the' p' ol would drop only about 6 inches per hour. Thus, there appears to be o considerable time available to restore normal cooling or to implement one of several alternative backup options for cooling. For licensing purposes, it has been accepted that the time interval for restoring cooling manually from available water sources is adequate >dthout requiring active (autcmatic) redundant cooling systems.

2-2 However, in considering the prioritization of Generic Issue 82, "Beyond Design Basis Accidents in Spent Fuel Pools," the NRC staff recognized that there is a finite probability that cooling could not be restored in a timely manner.2 The case treated in Ref. 2 was for a BWR. The estimated frequency for the loss of one (of two) cooling " trains" was taken to be 0.1/Ry (the l value assumed in WASH-1400).3 This combined with the coreditional probabili-ties of failure /non-availability of the second " train" yielded a combined fre-quency of a pool heatup event of 3.7x10-2/Ry. (This estimate appears to be somewhat conservative since no " pool heatup events" are on record after -103 reactor years of accumulated experience). To escalate from a " pool heatup event" to an event which results in fuel damage requires the failure of several alternative systems that are capable of supplying makeup water (the RHR and condensate transfer systems, or, as a last resort, a fire hose). Estimated frequencies of failure for each of the alter-natives, combined with the frequency of a pool heatup event, resulted in an estimated frequency of 1.4x10-6/Ry for an accident initiated by loss of spent fuel pool cooling. ) Originally, the spent fuel pool at the Ginna plant had only one installed cooling train with a " skid-mounted" backup pump and heat exchanger.

However, a second cooling train was to have been installed in 1986."

Because of the third option for cooling at Ginna (the skid-mounted system) the probability estimate for an accident initiated by a pool heatup event should be reduced to 5x10 7/Ry, i.e., about a factor of 3 smaller than for the BWR case analyzed in Ref. 2. 2.2 Structural Failure of Fool Because of the massive reinforced concrete structure of LWR spent fuel storage pools, designed to Category I seismic criteria, initiating events that would lead to a structural failure are extremely unlikely. On the other hand, a structural failure that resulted in rapid and complete draining of water from the pool would have serious consequences. Probabilities of events that might result in loss of structural integrity are estimated in the following two subsections.

f. 2-3 2.2.1 Structural Failure of Pool Resulting from Seismic Events Procedures and conventions for a detailed probabilistic risk assessment (PRA) of seismically-induced core damage accident sequences have been present-ed in Ref. 5. The recommended methodology could be applied to spent fuel pools as a separate plant component, or could be coupled to a core damage se-quence that might occur simultaneously during a severe earthquake. To date, the seismic PRA methodology has not been rigorously applied to spent fuel pools. Seismic risk analyses consist of three basic steps: 1) portrayal of the seismic hazard in terms of annual frequency of ex-ceedance as a function of some ground motion parameter (e.g., the peak ground acceleration); 2) assessment of the probability that the capacity of a structure or component can survive the seismic event, often expressed in the form of a fragility curve which is the inverse of the capacity for survi-val; and, finally, 3) a logic model, eag., an event tree, which relates a seismic-induced , failure to a higher order event that results in some category of ra-dioactive release. In principle, an appropriate convolution of the probability functions de-rived in steps 1) and 2) yields a probability function for seismic-induced failure. It is recognized that large uncertainties exist in the two input probability functions which are reflected in the function expressing the prob-ability of failure. The three steps and the treatment of the uncertainties have been summar-ized by Reed,6 who notes that the largest uncertainties are associated with step 1), i.e., the probabilities of occurrence of severe earthquakes having correspondingly very large ground accelerations. Reed makes the assertion that "due to the large uncertainties in the ground shaking hazard, it is

2-4 unproductive to refine the structure and equipment capacity calculations to accuracies which are it. consistent with the hazard uncertainty."6 The specific ) applicability to spent fuel pools of Reed's assertion is discussed in Section 2.2.1.3. 2.2.1.1 A Review of Seismic Hazard Data The primary difficulty in characterizing the seismic hazard at specific sites in the Eastern United States (EUS), i.e., sites to the east of the Rocky Mountains is that severe earthquakes are rare events in the EUS. A systematic analysis of recorded earthquakes and their relationship to geological features has yielded seismic zonation maps of the EUS.7 However, such information can-not readily be translated into the type of seismic hazard functions needed as input for pRA. Consequently, available historical data are insufficient for obtaining meaningful site specific estimates of the frequency of severe events. A different approach to seismic hazard analysis has been developed at Lawrence Livermore National Laboratory (LLNL) by D.L. Bernreuter and his col- ) leagues under NRC sponsorship. The initial study was a part of the NRC's Sys-tematic Evaluation Program (SEP).s The methodology has been further refined in a subsequent

study, "EUS Seismic Hazard Characterization Project" (SHC).8.10 Since the SEP and SHC results will be used for the seismic hazard esti-mates, some further discussion of the Bernreuter methodology is appropriate.

Three basic steps are involved: 1. the e11 citation of expert opinion to delineate and characterize seis-mically active zones in the EUS; 2. using the input data of each expert, the computation of the seismic hazard functions at specific reactor sites using several alternative ~ ground motion attenuation models with site corrections, and integrat-ing over each of the delineated seismic zones; and, finally,

2-5 3. the combining of the separate expert data accompanied by the genera-tion of uncertainty limits from the spread in expert opinions and from the self-evaluations of each expert on the degree of confidence in the input opinion. The various steps are carried out in a highly disciplined and systematic Provision is made at various stages for peer review of the methods manner. and input opinion, feedback to the experts and critical evaluation of the re-sults. In step 1, each expert prepares a "best estimate" map which delineates the seismic zones. Each zone is characterized by a set of parameters that give the maximum earthquake intensity to be expected for that zone (upper mag-nitude cut-off), the expected frequency of earthquakes, and the magnitude re-currence relation. For each input (zone boundaries, seismic parameters), the expert provides a measure of his degree of confidence. Also each expert is ' given the option of submitting alternative maps of differing zonations and characterizations (up to as many as 30 maps). The data from each expert are evaluated separately through step 2. In step 2, the contribution at a given site from each zone is integrated over the zone area and then over all zones. This requires the use of ground motion models for which a range of alternative models are employed to yield a set of alternative hazard curves. A " Ground Motion Panel"- of experts have selected several alternative models to be used, each having a weighting factor (see Ref. 9, App. C). Also each ground motion model incorporates a site spe-cific correction to account for local geology. in step 3, the results of the individual experts are combined to obtain a "best estimate" hazard curve and the uncertainty bands are computed in several alternative ways. It is obvious that the methodology requires a massive data collection and computer effort. In its present state, the final results are not in a form to be. easily applied to a specific *pRA by a non-expert in seismology. Further work is needed to develop a more convenient format for presenting the final

2-6 results.n In particular, numerical tabulations of the sets of hazard curves (such as those shown in Figs. 2.1 and 2.2) and their derivatives, dH/da I 4, for each reactor site would be helpful. Also, it appears that the local site geology needs more rigorous consideration in the derivation of the hazard l curves (see below). l Members of the Peer Review Panel have suggested several ways in which the methodology could be refined (see Ref.10, Section 7 and Appendices D.1-D.4). Many of these suggestions were implemented in the final feedback proces's and were included in the final results reported in Ref.10. In general, the re-viewers agreed that the results are " credible and as good as present scientif-ic understanding of eastern U.S. (EUS) seismicity probably allows" (Ref. 10, App. D.1). Comments from NRC licensees and their consultants indicated objections to application of the results to specific sites, noting that the site specific correction factors in the ground motion models were too simplified to ade-quately take local geological factors into account (Ref.10, App. 0.6). Al so, the criticism was made that the results were not adequately tested against ) recent historical records. In order to illustrate the hazard curves, their range of uncertainties and comparison with other studies, a series of figures taken from Ref.10 for the Millstone site is reproduced in Figs. 2.1-2.4 Figure 2.1 is the hazard curve obtained from combining the "best esti-mate" results for all experts in the SHC study (including the seismic and the ground motions panels). The curve plots frequency of exceedance per year vs. peak ground acceleration. Figure 2.2 illustrates the uncertainties in the hazard curve (15, 50, and 85 percentiles) derived from the spread in expert opinion and the self-confidence factors in the input parameters. It can be seen that the spread between the 15 and 85 percent 11es is about a factor of 20 at low PGA increasing to. about 350 at the high PGA. Comparison of Figs. 2.1 and 2.2 shows that

2-7 the "best estimate" curve is considerably higher than the 50 percentile, i.e., the mean > median. Figure 2.3 illustrates the spread in the "best estimate" hazard curves for all of the experts participating in either the SEPs or the SHC9 studies, or both (6 experts participated in both studies). The spread ranges from about one order of magnitude at lower PGA to about 1.5 orders of magnitude at the higher PGA. The curve marked "A," which falls considerably below the main grouping, was derived from data input in the SEP study by one of the ex'perts who participated in both studies. This revised input for the SHC project raised the derived curve by an order of magnitude at the low accelerations and by about two orders of magnitude at the higher PGA, this raises the obvious question of whether the experts were somehow influenced by the opinions of their colleagues, or whether the revision resulted from -a more careful consid-eration of the various geological factors that were taken into account in pre-parihg the input parameters. The question of testing the results for inadver-tent biases of this nature was addressed by the Peer Review Panel members, but their recommendations could not be fully implemented in the final report due to limited time and budget (Ref.10, pg. 7-3). Figure 2.4 compares the "best estimate" hazard curves for the individual SHC experts with curves generated from zonation maps prepared by the U.S. Geo-logical. Survey (USGS)12 and historical data of the past 280 years. As can be seen, the USGS hazard curve (denoted by "X") lies above-the SHG data. Bernreuter et al. attribute the difference between the SHC and the USGS curves to the variations in the equations used for conversions from intensity to mag-nitude and in the values for the rate of earthquake recurrence (Ref.10, pg. B-1 et seq.). As would be expected the 280 year historical hazard curve (de-noted by "H") falls below the SHC data because it does nct include postulated stronger earthquakes with return times much greater than the time span of the historical record. It should be noted that recent research has raised significant questions concerning the frequency of strong earthquakes in the coastal zone of the EUS.13 The speculation has arisen from paleoseismic field studies originally focused on the region of the strong earthquake near Charleston, SC, in 1886,

[ 2-8 which produced many " sand blows." lie,1s. These i result from the liquefaction. and venting to the surf ace of'sub-surface water-saturated sediment. Several sand blow craters have been found for which radiocarbon dating indicates that moderate to large earthquakes have recurred in the Charleston region on an - average of about every 1800 years.16 The lates't (prior to 1886) occurred' about 1100 years ago.18' Sand blows from prehistoric earthquakes have been un-earthed recently in the region extending from near Savannah, GA as far north as Myrtle Beach, SC.17 The broad extent of sand blows suggests that Charleston-type earthquakes might be associated with some tectonic feature which extends for some distance along the east. coast and not uniquely centered near Charleston. Up to the present time, no systematic field search'has been. made for sand blows outside of the Savannah to Myrtle Beach region.18 Recently Thorson et al. reported the existence of apparent sand blow craters in eastern Connecticut.18 These craters were recently examined by a USGS field team and assessed as not being of. the same nature as those obst.ved in South Carolina.18 2.2.1.2 Seismic Hazard Estimates for the Millstone and Ginna Sites The "best estimate" and the median,15 and 85 percentile seismig hazard curves developed by the SHC project for the Millstone site are shown in Figs. 2.1 and 2.2.18 These four hazard curves were used to develop the estimates of the seismic failure probabilities of Millstone 1, as described in Section 2.2.1.4 below. Hazard curves, such as shown in Figures 2.1 and 2.2, are expected to have some upper limit cutoff, i.e., PGA's which would never be exceeded. We have assumed that the upper limit cutoff for the Millstone site occurs at approxi-2 mately 1 g (980.7 cm/sec ), but a different cutoff would give a substantially different pool failure frequency. Seismic hazard curves for the Ginna site were not generated in the SHC project;20 however, the' SEP project included. data for Ginna.s Unfortunately, ^ the format of the SEP results, which were cirected primarily at obtaining site specific spectra cannot re&dily be translated into a "best estimate" hazard ') curve. In want of a better procedure, we have synthesized a hazard curve for

2-9 Ginna from the Millstone curve, using ratios of PGA's for 200, 1000, and 4000 year return times, tabulated for the two plants in the SEP ttudy.8. The hazard curve resulting from this synthesis is shown in Fig. 2.5. Because of the higher upper magnitude cutoff at the Ginna site, as perceived by experts, j (Millstone: kMI = 8.0 vs. Ginna: MMI e 8.2), we have assumed the upper cut-t off PGA of the hazard curve to be 1.25g. Although this is recognized to re a somewhat pessimistic assumption, it serves the useful purpose of illustrating the sensitivity of the calculated seismic risk to the upper cutoff of the haz-ard curve. 2.2.1.3 Seismic Fragility cf Pool Structures 1 Fragility curves specifically for spent fuel pools have never been devel-oped.21 It is necessary therefore, to rely on fragility assessments for other structures which appear to be of similar construction to spent fuel storage pool s. It must be recognized that this procedure introduces an additional element of uncertainty in the final risk estimstes -- an uncertainty that is difficult to quantify. Another source of uncertainty is the degree to which the stainless steel lining of a pool would enhance the seismic strength capac-ity (i.e., reduce the fragility). Conceivably, the reinforced concrete struc-ture of the pool could crack without loss of integrity of the pool lining. The dilemma of selecting an appropriate fragility for a BWR plant is aggravated by the fact that the pool structure extends typically from the 60 to the 100 foot elmtiocs above grade with the resultant amplification of the seismic bending st.rssses relative to the lower elevations of the structure.22 For the present analyses, two, somewhat diverse sets of fragility esti-mates, have been used:

1) the fragility curve developed by R.P. Kennedy et al.23 for the Oyster Creek reactor building; and

2) the fragility of the Zion plant auxiliary building shear walls (north-south ground motion).2s.

r q 2-10 In each case,' the fragility curve is defined by the following equation: )d F(a) = e [(in a/ /8R3 6 (2.1) { where F(a)' is the probability of structural failure given a peak ground accel-eration, PGA = a. 4(.) is the normal distribution function, is the median fragility level (i.e., the acceleration at which there is a 50% probability of f failure) and SR is the logarithmic standard deviation expressing the random-ness in the value of A third parameter, Su, is used to express the un-certainty in the median value and is used to generate upper and lower confi-dence limits. For example, it can be shown that the substitution for in X e+ 8u and - 8u Eq. 2.1 of generate respectively the 84 = = e and 16 percentile curves. Thus, a set of fragility curves can be generated free three parameters, , SR and s. The data used for generating the " Kennedy" and the " Zion" u curves are given in Table 2.4. Kennedy notes that the estimated median fragility value of about 0.75 g is considered applicable to plants designed in the U.S. in the mid 1960's. The Kennedy fragility curve is shown in Fig. 2.6, with the 84 and 16 percen-tile limits. The corresponding Zion curves appear in Fig. 22, pp. 3-35 of Ref. 24 2.2.1.4 Seismically-Induced Failure Probabilities The convolution of the derivative of a seismic hazard curve (e.g., Fig. 2.1) with a fragility curve, yields the annual probability of a seismically-induced failure. This can be expressed by the equation: amax dH P, g = g 4 F(a)3 da, .(2.2) where Pj,j is the ' failure probability obtained from the convolution of hazard curve i with fragility curve j., dH/da j is the derivative of the hazard curve 1 (i.e., the annual frequency of occurrence of peak ground accel-

eration, a,

and F(a)j is failure probability at acceleration, a, for. .=.

2-13, fragility curve J. The integration is cut off at the upper limit expected for the PGA. Since the seismic hazard curve is not an analytic function, the derivative dH/da and the integration are carried out numerically. Given many hazard and fragility curves from which to choose, and there being no a priori basis for choosing a particular pair, the convolution ex-pressed in Eq. 2.2 can be catried out for each pair of curves with weighting factors assigned to each o'f the curves in each set. The resultant collection of Pj,j gives a probability distribution which expresses the uncertainties in the analysis. The probability density distribution obtained for the Millstone site is shown in Fig. 2.7. At least in principle, the various hazard and fragility curves (sets i and j) do not have an equal likelihood of being correct. Therefore, a weight-ing factor (wi or ej) should be assigned to each curve which reflects an " engineering judgement" of its relative validity. The mean probability for failure is then derived from the following expression, F = {.3.j j,3 / {wj,3, (2.3) P f where {wi 1, Imj 1 and [mi,j {wtej = = 1. The weighting factors = = assigned by GNL for the Millstone case are given in Table 2.5. As can be s6en from the taoie, the "best estimate" hazard curve has been assigned a weighting factor of 0.5 with the remaining 0.5 distributed among the median,15 and 85 percentfle curves. The " Kennedy" set of fragility curves were assigned a total weighting factor of 0.75 with the remaining 0.25 distributed among the " Zion" set. Assuming an upper limit cutoff of 1.09, the mean probability of failure, ~P, derived from the 2a sets of Pj,j, usin2 the weighting factors f listed in Table 2.5 and Equation 2.3, was Tr = 2.2x10-5/ year (Millstone). In the case of Ginna, only a single hazard curve (Fig. 2.5). was used, I there being insufficient data to generate median,15 and 85 percentile curves for this site. Because of the structure of the Ginna spent fuel pool, the "Zi on" fragility curves are more appropriate, than the " Kennedy" curves.

2-12 i Therefore, higher weighting factors were assigned to the " Zion" curves as shown in Table 2.5. Based on an upper limit PGA cut-off of 1.25, the mean 9 probability resulting from the convolution of the single hazard curve with the six weighted fragility curves was T = 1.6x10-5/ year (Ginna). f The difference between the estimates for Millstone and Ginna, 2.2x10 5 vs.1.6x10-5, sh6uld not be regarded as highly significant, but more as an in-dication of the sensitivity of the results to the weighting factors assigned to the fragility curves. 2.2.2 Structural Failures of Pool Due to Missiles Missiles generated by tornadoes, aircraft crashes or turbine failure could penetrate the pool structure and result in structural failure. The probability of tornado missiles depends on the frequency of tornadoes at the site, the target area presented to the missile and the angle of im- ) pact. An analysis made by Orvis et al.25 for an average U.S. site derives a probability of <1x10-s/ year for structural loss of pool integrity due to a tornado missile (Ref. 25, pg. 4-44). Similarly, the analysis for structural failure.of a pool-from an aircraft crash yielded a probability of <1x10-10/ year (Ref. 25, pg. 4-58). The damage caused by Missiles generated by turbine failure depends on the orientation of tne turbine axis relative to the structure, as well as the fro-quency of turbine failure. An analysis by Bush yields a probability of -4x10-7/ year for spent fuel pool damage from a turbine failure missile.26 In the case of Ginna, the probability would be several orders of magnitude smaller (i.e., essentially zero) because the spent fuel pool is shielded from turbine missiles by the primary containment. 2.3 Partial Draindown of Pool Due to Refueling Cavity Seal Failures ) 1 On August 21, 1984, the Haddam Neck Plant experienced a failure of the refueling cavity water seal, while preparing for refueling. The water level l J

2-13 in the refueling cavity dropped by about 23 feet to the top of the reactor vessel flange within 20 minutes -- 4 loss of approximately 200,000 gallons, or a leak rate of about 10,000 gallons per minute.27 At the time of the event, refueling had not begun. The gates of the transfer tube connecting the re-fueling cavity to the spent fuel storage pool were closed. Although the seal failure did not result in an accident or in the release of radioactivity, the incident raised the question of whether similar failures might occur while spent fuel was being transferred or while transfer gates to the spent fuel basin were open, either case of which might result in exposure of spent fuel to air and possible clad failure. All licensed plants were instructed to evaluate the potential for and consequences of a refueling cavity seal failure.27 Refueling cavity seals, seal the gap between the reactor vessel flange and a flange on the inner periphery of the reactor cavity, or the floor of the cavity. BWR's have a permt.iently installed stainless steel bellows to seal the gap, and are, thus, not subject to failure of the Haddam Neck type. Many PWR's seal the gap with gaskets held down by a bolted flat steel ring. Such systems have experienced difficulties,in achieving tight seal be-cause of surface irregularities and small vertical and concentric offsets in the two flanges. Consequently, many plants have converted to inflatable (pneumatic) rubber seals. Also, it should be noted that pneumatic rubber seals are often used to seal the gates in transfer tubes or canals. Licensee responses to the IE Bulletin indicate that the Haddam Neck cavi-ty configuration is unique in that the width of the annular gap between the reactor flange and the cavity flange is about two feet, whereas, in most plants the gap is of the order of <1" to -3". As of sumer 1985 some 45 units used pneumatic seals in the refueling cavity.28 4

2-14 Typical pneumatic seals are illustrated in Figures 2.8-2.10. There are many variations in the details of the designs, e.g., some. plants have various types of retainers to support the rubber seals (e.g., see Figure 2.10), others rely on the rubber seal alone (e.g., see Figure 2.9). Acnording to the re-sponses of the licensees, even if a pneumatic seal should deflate, the leakage would be expected to be sanll or negligible, because the wedged shaped upper section would maintain a good seal (refer to Figure 2.8).- i.e., the deflated seal would not distort enough under the hydrostatic head to extrude through the gap. 2 Aside from the Haddam Neck 1984 incident, a few cases have been reported in which inflated seals have failed, either in the refueling cavity or trans-fer gates. None of these events had significant radiological consequences. Several such events are listed in Table 2.6. It is likely that this list is not exhaustive.. To the best of the authors' knowledge no data base has been compiled (or is available) of the failure rate of pneumatic seals and their pressurizing systems of the types used in nuclear power plants, or of similar seals used in non-nuclear industries. \\ Based on the limited experience cited in Table 2.6, the historical fail-ure rate in seals / systems is in the range of -1x10-2/Ry. Because of ad-vances in design, increased awareness and surveillance, the present failure rate is. estimated to be an order of magnitude smaller, i.e., -1x10-3/Ry. As is obvious from Table 2.6, a seal failure does nut necessarily result in the rapid loss of water inventory from spent fuel transit er storage loca-tions. The limited experience indicates that the most probable time for a refueling cavity seal to fail is shortly after installatica, while the cavity volume is being filled with water. According to the analyses supplied by a licensees in response to IE Bulletin No. 84-03, the failure of a pneumatic refueling cavity seal in most PWR plants would net result in massive leaks because of the relatively narrow gap to oe sealed and the geometric shape of the seal. Also, leaks from seal failures in t.ransfer tube / canal gates would be limited, in most cases, because the leakage would be into a confined ~

volume, e.g., from the storage pool into a drained up-ender sump.

Taking these factors into consideration, it is estimated that the frequency of a N ~

? - y, "p. 35 i y 5-15 1 serious loss of pool water inventory resulting from a pneumatic seal failure to be in the range of -1x10 5 g, /y Even a large loss of water inventory from the spent fuel pool does not .w cessarily result in uncovering and subsequent failure of fuel. Most spent

• .nel basins are constructed with weirs below the transfer gates which preclude complete drainage of the pool, even in the event of a catastrophic Haddam Neck type failure with the transfer tube / canal gates open.

In most cases, the wa-ter level would remain a foot or more above the active zone of the spent fuel assemblies. In a few cases, the upper several inches of the fuel could un-cover. (Note: Licensee responses to IE Bulletin 84-03 did not always provide information about the elevations of weirs and tops of stored assemblies.) In the event of h draindown of the pool to near the top of the fuel as-semblies, there.would still be time (1/2 to I hour) to close gates and restore a supply of water before the residual water inventory reached the boiling point. However, as noted in one licensee response, even if the fuel remained covered " dose rate in the vicinity of the spent fuel pool would, however, be high, complicating recovery from the event.=29 A pool heatup event similar to the partial draindown scenario described j above was considered by the NRC staff in Ref. 2. A conditional probability for failure to restore adequate makeup water was taken to be 5x10-2, based purely cn / dgement. Because of higher radiation levels in the partial drain-down scenario, it is estimated that the probability of failure to restore adequate makeup water to be somewhat larger, i.e., -1x10-1 Given all of the above, the pro'bability of a pneumatic seal failure which results in exposure to air of stored spent fuel with resulting clad failure is estimated to be of the order of P - 1x10 8/Ry. 1 4 e

2-16 2.4 Pool Structural Failure Due to Heavy Load Drop WASH-1400 considered the probability.of structural damage to the pool due to the dropping of a fuel transfer cask (Ref. 3, pg. I-97). In the analysis. it was anticipated that one spent fuel shipment per week would be the equilib-rium shipping rate. The estimated rate for a drop resulting in pool failura (for a single unit plant) was 4.5x10-7/Ry. The above frequency was based on a crane failure probability of 3x10 6 per operating hour. It was further assumed that each lift was of 10 minutes duration and for a 10 second period per lift the cask would be in a position to cause gross structural damags to the pool wall if a crane failure oc-curred. Human error was not considered. Since spent fuel is not currently being shipped, this hazard does not ex-ist at the present time. Hwever, at some point in the future, spent fuel will have to be removed from the reactor pools, either to some onsite storage facility, or eventually to a high level waste repository. At that time, the frequency of removal of spent fuel will be correspondingly greater. Orvis et al.2s have reexamined the cask drop probability and have used the following probabilities: Mechanical failure of crane = 3x10 6/ operating hour Electrical control failure of crane = 3x10-6/ operating hour Human error = 6x10-"/11ft. As can be seen, human error dominates the Orvis estimates for probability of a cask drop. The Orvis datum for human error was based on a study by Garrick et al. so which concerned human reliability in the positioning of heavy objects. The applicability of the Garrick study to crane operations is not obvious. Nevertheless, a human failure rate in the range of 10-3 to 10-" per operation appears to be consistent with data listed in the NRC handbook on human 31 reliability analysis for cases in which the operation has one or more people who serve as " checkers" and involves some degree of personal risk to the operating personnel. l

[ X% .a M V) J-9 y', q y f, ' l I~A y v;

9 3

L ( ' Obviously, dot all ' human failures associat'ed with de lifting _ and moving of a spent fuel dipping cask would result in strur.tural damage to the pool. The section of the pool where the cask is set dowa DAs an impact pad to absorb .the impulse obia' dropped. cask. Accidents in unloading the cask from or - reloading r.n,tka 'transpohc vehicle would not' involve the pool. /bnlyho zontaimovementsofthecaskIb

e. a structurally critical. sec' J

tiofof the pbol would poce the threat of structural damage. As.noted_above, J WASil-f400 assumed; that the sensitive section tis. the vertical wall 'at the pool i edgey It was'.ishlicitly assumed that all la$d drops on' the pool' edge woi:1d I result in ttructural failure. This ' assumption appears. to be too simplistic and conmquerttly too ccenservative for the f5110 wing reasoM: many " loa drcps" would be partially attenuated byicrane mechanisms whicti 1heit descent rates, and reduce impact energy, T L in case of some "off-center" hits, the full potential impt.ct energy ~ would not.' be absorbed by the pool edge NI:k tilted, one end strikes floor first), ind I l account should be taken of exterior cask fittings (e.g., cooling vanes) which absorb some imoact energy. y ?lo rigorous structural analyser have been performed to sc6pe the range of 6 mage to a rool odge from a cask drop. In the abssm.e of such analyses, it has been necessary to estimate th conditional proubility' of catastrophic S ' structural damage 'given a cask 1 drop in the vicinity os the pool edge. It is J estimated that thO conditional probability 2is less than 100% and greater than E A conditional probability of 10% has been(arbitrarily selected for the i hazard calculation and '100% cnd"1% used. for deftaing the range of uncertain-y ties. [ s ji Since human error,{ rather than mechanical or electrical failure, appears to dominate the hazard arising from shipping cask movements, the various steps in the crane operation have been identified in Table 2.7, which also lists the types of human error asnkiated with each step. The distribution of failure e o ( k Il I c ..:T 1 'l s

2-18 frequency in the various steps has been estimated and listed in the last col-umn of Table 2.7. (This distribution was subjected to " peer review" by ~ BNL rigging personnel and managers who oversee operations of this type.) It will be noted that most steps in the crane operation do not jeopardize the struc+" al integrity of the pool. Only in steps 5a and 5b (see Table 2.7) ) could t"z cask strike the pool edge. An accident of the type listed in Sa (horizontal movement with cask not high enough to clear the pool edge) would probably not cause serious damage because of the limited kinetic energy of the cssk associated with the slow velocity of horizontal crane movements.

Thus, only step 5b in Table 2.7 is considered in the hazard calculation.

Fcr purposes of calculating the cask drop hazard, i.e., the probability of catastrophic structural damage to the pool resulting from a cask drapping g on the pool edge, the assumptions listed in Table 2.8. were used. Table 2.8 { also lists the uncertainty ranges for each of the parameters. The results are as follows: Probability of structural failure due to cask drop on pool edge caused by mechanical or electrical failure of crane = 3.5x10 7/Ry. Probability of structural failure due to cask drop on pool edge caused by human error = 3.1x10-5/Ry. If the failure rates summarized in Table 2.8 are assumed to be statis-tically independent, then the uncertainty in the overall failure rate is domi-nated by the uncertainty in the probability of pool failure. Thus the overall uncertainty is about a factor of ten in either dire: tion. 2.5 Summary of Accident Probabilities The probability estimates made in Sections 2.1-2.4 are sumarized in Table 2.9. These include only those accidents that result in the complete loss of pool water inventory. It will be seen that shipping cask drop result-ing from human error and seismic induced failures dominate in the hazards. As -e o me,s em m

2-19 previously discussed the uncertainty in both of these probabilities is quite large and has been estimated to be an order of magnitude in either direction. 2.6 References for Section 2 1. The data cited in Table 2.3 were culled from submissions by the licensees of the respective dockets in support of license amendments for expanded spent fuel storage limits. 2. "A Prioritization of Generic Safety Issues," Division of Safety Technolo-gy, Office of Nuclear Reactor Regulation, U.S. Nuclear Regulatory Commis- } sion, NUREG-0933, December 1983, pp. 3.82-1 through 6. 3. " Reactor Safety Study, An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," U.S. Nuclear Regulatory Commission, NUREG-75/014 (WASH-1400), October 1975, App. I, Section 5. 4. Rochester Gas and Electric Corporation, Docket No. 50-244, " Design Crite-ria, Ginna Station, Spent Fuel Cooling System," EWR 1594, Revision 1, October 10, 1979; U.S. Nuclear Regulatory Commission, Safety Evaluation by the Office of Nuclear Reactor Regulation Supporting Amendment No. 65 ,to Provisional Operating License No. DPR-18, Rochester Gas and Electric Corporation, R.E. Ginna Nuclear Plant, Docket No. 50-244, November 14, 1984 5. M. McCann, J. Reed, C. Ruger, K. Shiu, T. Teichmann, A. Unione, and R. Youngblood, Probabilistic Safety Analyses Procedures Guide, Vol. 2, pre-pared for the U.S. Nuclear Regulatory Commission by Brookhaven National Laboratory, NUREG/CR-2815 (BNL-NUREG-51559) Vol. 2, Rev.1. August 1985. 6. John W. Reed, " Seismic Probabilistic Risk Assessment for Critical Facili-ties," preprint of paper presented at the 1985 ASCE Spring Convention, Denver Colorado, April 29 - May 3, 1985.

2-20 7. N.L. Barstow, K.G. Brill, 0.W. Nuttli and P.W. Pomeroy, An Approach to Seismic Zonation for Siting Nuclear Electric Power Generating Facilities in the Eastern United States, prepared for the U.S. Nuclear Regulatory Comission by Rondout Associates, Inc., NUREG/CR-1577, May 1981. 8. U.S. Nuclear Regulatory Comission, NUREG/CR-1582, Seismic Hazard Analy-sis, 5 Volumes: Volume 1, Overview and Executive Summary, D.L. Bernreuter and C. Minichino, April 1983. Volume 2, A Methodology for the Eastern U.S., Lawrence Livermore National Laboratory / TERA Corporation, August 1980. Volume 3, Solicitation of Expert Opinion, Lawrence Livermore National Laboratory / TERA Corporation, August 1980. Volume 4, Application of Methodology, Results and Sensitivity Studies, D.L. Bernreuter, Lawrence Livermore National Laboratory, October 1981. ) Volume 5, Review Panel, Ground Motion Panel and Feedback Results, D.L. Bernreuter, Lawrence Livermore National Laboratory,1981. 9. D.L. Bernreuter, J.B. Savy, R.W. Mensing and D.H. Chunn Seismic Hazard Characterization of the Eastern United States: Methodology and Interim Results for Ten Sites, prepared for the U.S. Nuclear Regulatory Comis-sion by Lawrence Livermore National Laboratory, HUREG/CR-3756 (UCRL-53527), April 1984 10. D.L. Bernreuter, J.B. Savy, R.W. Mensing, J.C. Chen and B.C. Davis, Seis-mic Hazard Characterization of the Eastern United States, Vol.1, "Meth-odology and Results for Ten Sites," Lawrence Livermore Mational Labora-tory, UCID-20421, April 1985. 11. This matter is under consideration but has not yet received NRC sponsor-ship (private communication, D.L. Bernreuter, Sept.1986).

2-21 2. S.L. Algermissen, D.M.

Perkins, P.C.

Thenhaus, S.L. Hangen and B.L.

Bender, Probabilistic Estimates of Maximum Acceleration and Velocity in r Rock in the Cartiguous United States, U.S. Geological Survey, open. file report 821033 (1982). 13. E.g., see R.A. Kerr, "Chcrieston Quakes are Larger or Widespread,", Sci-ence, 233, p. 1154, September. 12, 1986; and R.A. Kerr, " Eastern Quakes Pinned Down?," Science, 227, January 25, 1985. 14 S.F. Obermeier, G.S. Gehn, R.E. Weems, R.L. Gelinas and M. Rubin, "Geo-logic Evidence for Recurrent Moderate to Large Earthquakes Near Charles-ton, South Carolina," Science, 227, pp. 408-411, January 25, 1985. 15 E.g., see P. Talwani and J. Cox, "Paleoseismic Evidence for Recurrence of Earthquakes near Charleston, South Carolina," Science, 229, pp. 379-381, July 26, 1985. 16 R.E. Weems, S.F. Obermeier, M.J. Pavich, G.S. Gohn, M. Rubin, R.L. Phipps and R.8. Jacobson, " Evidence for Three Moderate to Large Prehistoric Holocene Earthquakes Near Charleston, SC," Proceedings of the Third U.S. National Conference on Earthouak::s. Engineering, August 24-28, 1986, Charleriton, South Carolina, pp. 3-14, (published by the Earthquake Engi-neering Research Institute). 17. S.F. Obermeier, R.B. Jacobson, B.S. Powars, R.E. Weems, B.C. Hallbick, G.S. Gohn and H.W. Markewich, " Holocene and Late Pleistocene (?) Earth-qtiakes Induced Sand Blows in South Carolina," Proceedings of the Third U.S. National Conference on Earthquake Engineering, August 24-28, 1986, Charleston, SC, pp. 197-208 (published by the Earthquake Engineering Re-search Institute). 18 S.F. Obermeier, private communication, Sept. 26, 1986 l i 19. R.M. Thorson, W.S. Playton and L. Seeber, " Geological Evidence for a Large Prehistoric Earthquake in Eastern Connecticut," Geology, 14, pp. 463-467 June 1986

2-22 ' 20. D.L. Bernreuter, private communication. Sept. 1986.-

21. A proposal for such -studies has been submitted to the ' NRC by the Structural Analysis Group of BNL.

22. E.g. see Figs. 12 and 13, pp.'3-22 and 3-23,'of Ref. 24.- 23. R.P. Kennedy, C. A. Corne11',, R.D. Campbell, S. Kaplan ' and H.F. Perla,. "Probabilistic Seismic Safety Study of an' Existing Nuclear Power Plant," Nuclear Engineering and Design, 59, (1980) 315-338.* 24. L.E. Cover, M.P. Bohn, R.D. Campbell and D.A. Wesley, Handbook of Nuclear Power Plant Seismic Fragilities, prepared for the U.S.- Nuclear Regulatory l Commission by Lawrence Livermore National Laboratory, NUREG/CR<3558. ^ (UCRL-53455), June 1985. 25. D.D. Orvis, C. Johnson and R. Jones, Review of Proposed Dry-Storage Con-l cepts Using Probabilistic Risk-Assessment, prepared.by NUS Corporation for the Electric Power Research Institute, EPRI NP-3365,.Feb 1984 ) 26. S.H. Bush, " Probability of Damage to Nuclear Components Due to Turbine Failures," Nuclear Safety,14, No. 3, May-June,1973. 27. U.S. Nuclear Regulatory Commission, Office of-Inspection-and Enforcement, IE Bulletin No. 84-03: " Refueling Cavity Water Seal," August 24, 1984 (This was subsequently foilowed up by more detailed. instructions for evaluation: ' Inspection Requirements for IE-Bulletin' 84-03, " Refueling - Cavity Water Seals," Temporary Instruction 2515/66, Inspection and' En-forcement Manual, USNRC, Office of Inspection and Enforcement, issue date -12/17/84.)

28. The total count was based on licensee responses to IE Bulletin No.

84-03. As of mid-1985 a few licensees had not yet filed responses. 29.. Response of Baltimore Gas and Electric to IE Bulletin No. 84-03. e .__.-._--.-.--._____u_.- .----_-_-__.___l__-__._____

2-23 30. B.J. Garrick et al., "The Effect of Human Error and Static Component Failure on Engineered Safety System Reliability," Holmes & Narver, Inc., HN-194, November 1967. 31. A.D. Swain and H.E. Guttmann, Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications, prepared for the U.S. Nu-clear Regulatory Commission by Sandia National Laboratories, NUREG/CR-1278 (SAN 080-0200), A'ugust 1983. 9 e se l

2-24 Table 2.1 Typical Spent Fuel Pool Dimensions and Water Inventories Length / Width / Depth Pool Volumes Nominal Water Inventory (feet) (cubic feet) (cubic feet) l 40/26/39a 4.1x104 3.5x104 43/22.25/40.25b 3.4x104 3.3x104 aBWR, Vermont Yankee. bPWR, Ginna. Table 2.2 Decay Heat as a Function of Time Since Last Refueling (Data from Appendix A) Decay Heat Load (106 Btu / hour) Decay Time Since Last Shutdown for Refueling Plant 30 days 90 days 0.5 years 1.0 year Millstone-1 4.43 3.10 2.38 1.76 Gi'nna 2.62 1.96 1,59 1.25 ') Table 2.3 Examples of Thermal-Hydraulic Transient Parameters, Assuming Complete Loss of Pool Coolant Circulation Rate of Time of Boil-Off Temp. Increase Boilingb Rate Docket No.a (*F/hr) (hours) (gpm) 3 (ft /hr) 50-325 5.0 13.5 28 262 50-250 9.7 9.3 N.A. 50-271 <3 >20 14 131 50-247 13.0 4.8 57 534 50-344 <6.3 >11 34 318 aSee Ref b ours a. 1. H fter complete loss of cooling capability.

2-25 Table 2.4 Fragility Parameters Assumeo in This Study for Spent Fuel Storage Pools / A Structure (g) 83 s Ref. u a Oyster Creek Reactor Building 0.75 0.37-0.38 24 Zion Auxiliary (N-S motion)b Building Shear Walls 1.1 0.12 0.20 25 aDesignated as the " Kennedy" fragility curves in the text. b esignated as the " Zion" fragility curves in the text. D Table 2.5 Weighting Factors Assigned to the Various Hazard and Fragility Curves for the Mill-stone Case MILLSTONE GINNA Seismic Hazard Curves: g g "Best Estimate" 0.50 1.00a 15% Confidence Curve 0.10 Median Curve 0.30 85% Confidence Curve 0.10 I1 M W" Fragility Curves: g g -- " Kennedy", Median 0.45 0.15 , 16% 0.15 0.05 , 84% 0.15 0.05 " Zion", Median 0.15 0.45 , 16% 0.05 0.15 , 84% 0.05 0.15 I.j - M T3r a" Synthesized" Curve. i

2-26 Table 2.6 ' Events in Which Inflated Seals Have Failed Seal Total. Date Plant location Cause Leakage 9/72 Pt. Beachl Transfer Gate Failure of air supply 11,689 gal. 10/76 Brunswick 2 Inner Pool Gate Air leak in seal plus (Pool level compressor power supply dropped 5") failure 5/81 Arkansas Nuclear _ Transfer Gate Maintenance error.. air 1000 gal / min One - 2 supply shutoff 8/84 Haddam Neck Cavity Seal Design weakness, seal 200,000 gal. shifted in 20 min. 10/84 San Onofre 21 Gate Seal Air compressor power 20,000 gal. failure 11/84 San Onofre 21 Cavity Seal ' Manufacturing defect, 2 seal rupture 12/86 Hatch Pool-Canal Valve to compressed 141,000 gal. Flexible Joint air _ supply closed 1No spent fuel was in the storage pool. I 2Failure occurred during installation and leak testing. t e %g-

2-27.. Table 2.7 Estimated Distribution of Human Error in Heavy Crane Operations. These Estimates, Made by BNL Staff, are Based _on Engineering Judgement and are Not Supported by Actuarial Data. Estimated Fraction of Total. Error a Operational Step Possible Human Errors Frequency (Per Cent)' 1. Install rigging Wrong slings (e.g., hoist rigging not 10 qualified for task) Improper installation (shackle, pins, 10 etc.) 2. Positioning of Crane hook not over center of gravity 15 trane over load, (load upset as tension applied) apply tension 3 Lift load Control error (wrong hoisting speed 10 unintentional reversal of direction) 4 Start horizontal Control error (move wrong direction, 10 travel lift or lower instead of move) Sa. Horizontal travel Control error (unintentional reversal of motion, overshoot stopping point) e Load not high enough to clear obstacles 4 Sb. Horizontal travel Control error or delayed rigging failure 1 resulting in load drop 6. Lower load Control error (wrong direction, descant 10 too fast) 7. Positioning of Inaccurate positioning cradle capsizes 20 crane over re-during set-down ceiving cradle and set-down Set down too rapid 10 load "It is assumed that the movement of. a spent fuel shipping cask is carried out by a qualified rigging crew consisting of a foreman, two or more riggers, and a crane operator. The foreman and riggers check each step and crane move-ments are signaled to the operator by the foreman who stands in a location providing adequate surveillance of the load, and can be clearly seen by.the operator. 9 l

2-28 Table 2.8 Assumptions Used in Calculating the Hazard of Catastrophic Structural Damage to Pool Resulting from the Drop of a Shipping Cask Assumed Uncertainty Item Value Range l Number of fuel shipments (eventual rate to reduce 2 accumulated inventory) per week Number of passes over pool edge per shipment 2a 0 Fraction of horizontal movement when cask is 0.25 0.1 to 0.5 above pool edge Total operational time in each movement, 10 8 to 30 I minutes per lift Time over pool edge per lift, seconds per lift 10 5 to 20 Mechanical failure rate of crane, per 3x10-6 10-6 to 10-5 operating hour Electrical failure rate of crane, per 3x10-6 10-6 to 10-5 operating hour f Total accident rate from human error, 6x10-" 10-4 to 10-3 ) failures per lift Fraction of human error cesk drop accidents 0.01 5x10-3 to occurring during horizontal motion of 5x10 2 crane, fraction of total Conditional probability of structural failure 0.1 10 2 to 1.0 of pool given a cask drop at pool edge loca-tion, failures per drop aSome spent fuel pools have a special section for the shipping cask sepa-rated from the main pool by a wall with a wier or gate. For such a configur-ation the number of passes over the " pool edge" would be zero and hence the risk to the main pool from a cask drop would be zero. 5 4 e

\\ 2-29 Table 2.9 Summary of Estimated Probabilities for Beyond. Design Basis Accidents in Spent Fuel Pools Due to Complete Loss of Water Inventory Estimated Probability /Ry Accident Millstone .Ginna Loss of Pool Cooling Capability 1.4x10 6 5.7x10 7* Seismic Structural Failure of Pool 2.2x10 5 1.6x10 s Structural Failure from Tornado Missiles <1x10-s <1x10 s ~ Structural Failure from Aircrash <1x10-10 <1x10-10 Structural Failure from Turbine Missile 4x10-7 -0** Loss of Pool Water Due to Pneumatic Seal Failure 0 1x10 8 1 3.1x10-5 3.1x10 5 Structural Failure from Cask Drop 1After removal of accumulated inventory resumes.

• With credit for third cooling system.

Other PWRs which typically have two spent fuel cooling systems would have an estimated fuel uncovery frequency of about 1x10 6/Ry.

• • Typical PWRs may have a failure frequency due to turbine missiles on the order of 4x10- but Ginna's pool is shielded from the turbine.

t 9 9 6 4 9 -__._._.._______m______._____._.--m__

2-30 l E l l 1 t i i i i I E l MILLSTONE "BEST ESTIMATE" ~ id ~E w a 2 4 C -3 d 10 r E u x W 1 O -4 D 10 5-5 z v1 8 m E 16 ) 0 =- -= _a = = 3 e z z I T l =_ =_ -7 I I l l 1 1 I I I 10 0 200 400 600 800 1000 2 PEAK GROUND ACCELERATION (cm/sec ) L/ Figure 2.1. Seismic Hazard Curve for the Millstone Site. The curve shown is the mean of the hazard curves generatsd from the "best estimate". input data of the ten experts par-ticipating in the SHC study combined with the "best es-timate" model of the ground motion panel. Site correc-tions are included (Source: Ref.10,pg.5-43). ~ ++------**e .+=

2-31 -l 10 g i i i i i I i 1 5 MILLSTONE -2 15, 50, 85 PERCENTILES __ 10 =- E. ? w = z ~_ g $e w s i E = 5 E E g U Id E-l z 85 w ~. w 50 '_J.id 1 = = 5 15 z 2 l r 5 i i6 O 200 400 600 800 1000 2 PEAK GROUND ACCELERATION (cm/sec ) Figure 2.2. The 15, 50 and 85 Percentile Hazard Curves for the Millstone Site. The data are based on confidence levels in the input seismicity data of the experts and uncertainties in the best choice of ground mo-tion models (Source: Ref.10,pg.5-45).

2-32 k -1 to -2 10 + 10 4 m I -4 10 ] NN -5 10 h 4 to -7 10 ~ -n n m v c e n a e b ACCELERATION CM/SEC**2

• MILLSTONE s

Figure 2.3. Seismic Hazard Curves for Millstone of Each of the Individual Experts Participating in the SEP Studies (Ref. 8) and/or the SHC Studies (Ref. 10). The curves.give, an indication of the spread in expert opinion (Source: Ref.10,pg.209). +a- ++

2-33 l' -1 10 -2 to -s to w to 1 -5 to SY -7 to + -n n n c e v. u -~ e 3 ACCELERATICH CWsEC**2 Figure 2.4 Comparison of the Millstone Site Hazard Curves Generated from the Data Input of the SHC Experts, with Those Generated from the USGS Data (Curve "X"), and from the Historical Record of the Past 280 Years (Curve H) (Source: Ref. 10, pg. 6-7 ). A

2-34 ) l l i l 10 -3 : i i i, .i i i 5 GINNA 10-2 3 j t l W i j ~ O z <a I W $ 10-3 r j x W i \\ D 10-4 r 1 z g cy. W a: ' 10-5, a e 3 E 10 -6 7 y l .i i l 10-7 200 400 600 800 1000 1200 PEAK GROUND ACCEL ERATION (cm/sec2) Figure 2.5 Seismic Hazard Curve for Ginna. This-curve was Synthesized from the SHC "Best Estimate" curve 1 for Millstone (see Figure 2.1), and PGA ratios for Millstone and Ginna given in the SEP studies.

2-35. ,1 I I I I i 1.0 FRAGILITY CURVE '8*'* t; o.8 2 MEDIAN-j O E a6 m w 84% m 3d 0.4 2 t 0.2 / i I~~ I O O.2 0.4 0.6 0.8 I.O - 1.2 PEAK GROUND ACCELERATION (g) -i r Figure 2.6 Fragility Curves for. the Oyster Creek Reactor Building. The curves generated by R.P. Kennedy et al. (Ref. 23) give 'the frequency of structural failure as a function of peak ground acceleration during an earthquake. l o

J 2-36 ,i - 1 I I i i j O.3 -- MEAN =2.15 x 10-5 yr.-l -{ Ymz $O,2 - =! E 8 E ) 0.1 E 10-8 10-8 io-7 IO-s 10-5 go-4 ANNUAL FAILURE FREQUENCY Figure 2.7 Probability Density for Seismically-Induced Fail-ure as a Function of Annual Failure Frequercy. l The histogram was obtained from 24 convolutions l of four hazard curves with six fragility curves-3 and includes the weighting factors assigned to each curve. l

2-37 1 4* i g '. ? C S k .EEI' 'N f V u f 3 j e / w s SEAL SYM Anour k l E / ^ y jl Y i l /, I dS' 83-83 a *. v.o - 1 .I Figure 2.8 Cross Section of a Typical Pneumatic Seal (Source: submission by Sequoyah Nuclear P1 ant. Docket No. 50-327,10/26/84 in response to IE Bulletin 84-03

2-38 2 /dd.99 *# VfsstL fLM ffAL (144 03'4 93 *m/) $t/Mont RIME (Jff &l/Jfj93) s $MARP KDGAS HANO l amoro or uus m 6 natur m.ua m xAr. - s

> aumer,uAcuinto se ava ra A xou-aurs f

SA9 C# 2* 647W2M PRISMY.94AL funwnsN

y? * *

• fLANM ANO CAVit"Y

>>MLL/MCAr RIN4) gv / %) ^ ,,, --CA VITY WALL (4&N951) MMNhW3F, INSULATIOy (ggf gejsmt) RKONott'/OvfD AMD Batk!!LLED worn ocwow s rror A* rusne W# # $719L (WNifM MAD Lf56 THAN d M LfMMABL/ CMLed/M6)10 ^#9701. stAL N sostruw & yw'LaffD Dd!TN O' SIN

• TO '*0V'M g *, t A.c o MflSP.KTDR"Y MAL. '34MrKM1 Figure 2.9 Cross Section of Inflated Pneumatic Seti Seated on the Reactor Vessel Flange and Inner Surface of Cavity Wall (Source:

submission by Sequoyah Nu-clear Plant. Docket No. 50-327, 10/26/84 in re-sponse to IE Bulletin 84-03.

2039 rq , a. - ] o N N\\\\\\ \\ A !=== = _ _ = _ =.= = _ _ =. _ = _=== ,re.es,. -swx4a 444 m t- -- _-_ __ = ~ asacrea unas. was pw_ 1, 7% / ~ = / i ,W S 5 5 g s..w 4,.. e i h '5i[f39-tY .k r, 5:u.- ~ = 5 5 . aam - EA.yy.epggge.i $ ga _,_ _ _ - ~ ~ - - ".O l : -, m.,D-

• *=

Figure 2.10 Uninfl ated Pneumatic Seal with Steel Hold-down Ring (Source: submission by Indian Point Station. Unit 2, Docket No. 50-247,11/30/84 in response to IE Bulletin 84-03). a

3-1 3. EVALUATION OF FUEL CLADDING FAILURE 1 Two previous studies,2 have analyzed the thermal-hydraulic phenomena assuming a complete drainage of the water from a spent fuel pool. The pre-vious section addressed the possible mechanisms for such an accident and pro-vided estimates for the accident frequency. This section provides a reevalua-tion of the basis for the SNL results.1,2 3.1 Summary of SFUEL Results The SFUEL code was developed by Benjamin et al.1 to analyze the behavior of spent fuel assemblies after an accident has drained the pool. The results reported in Reference 1 indicated a wide range of decay power levels for which self-sustaining oxidation of the cladding would be predicted. Several l'oita-tions in the SFUEL model were identified and addressed in a subsequent inves-tigation.2 But comparisons to small scale experiments were not very success-ful. ~ 3.1.1 Sunnary Model Description The SFt'EL code was developed at SNL and is described in Reference 1. -Basically it is a finite difference solution of the transient conduction equa-tion for heating of the fuel rods considering: The heat generation rate from decay heat and oxidation of the clad-ding. Radiation to adjacent assemblies or walls. Convection to buoyancy-driven air flows. The key assumptions in the analysis are: 1) The water drains instantaneously from the pool.

2) The geometry of the fuel assemblies and racks remains undistorted.

1 i

l 3-2

3) Temperature variations across the fuel rods are reglected.

1 4 i

4) The air flow patterns are one-dimensional.

I

5) The spaces 'between adjacent basket walls are assumed to be closed to I

air flow. After the water is drained from the pool the fuel rods heat up until the buoyancy driven air flow is sufficient to prevent further heating. If the decay heat level is sufficient to heat the rods to about 900*C, (1650*F) the oxidation becomes self-sustaining. That is the exothermic oxidation reaction provides sufficient energy to match the decay heat contribution and the temperature rise; rapidly. Reference 2 modified the SFUEL code to increase calculational stability and assess propagation of Zircaloy " fires" from high power to low power ass.em-blies. The SFUEL code was also modified by Pisano et al.2 to eliminate un-realistically high temperatures

• by non-mechanistically removing each node as it reaches the melting point of Zircaloy dioxide (2740*C or 4963*F).

In the present investigation, the oxidation cutoff has been reduced to 1900*C (The melting point of Zircaloy 3450*F). 3.1.2 Clad Fire Initiation Results An extensive review of the cladding oxidation models used in SFUEL,1,2 is given in Appendix B: 1. The likelihood of clad fire initiation is not very sensitive to the oxidation equation. 2. The oxidation equation used in SFUEL is a reasonable representation of the data. "Since the code does not explicitly treat melting of the cladding, tempera-tures as high as 3800*C were predicted.2 .w--

'3-3 3 v: 3. The likelihood of clad-fire initiation is most sensitive to the decay ' heat level and the storage rack configuration (which controls thefex-tent of natural convection' cooling). The critical conditions for clad fire initiation are summarized in Table 3.1. Note that. for the old style cylindrical fuel - racks with a large in1'et orifice (3 inch diameter) the natural conwetion cooling -in air'is predicted to be adequate to prevent self-sustaining oxidation (cladding " fires") after 10 days of ^ decay for BWR assemblies and 50 days for PWR assemblies. Howsver for the new high density fuel racks, natural' convective flows are so restrict-ed that even 'after cooling for a year there is pot'ential for self-sustaining oxidation. As: pointed out by Benjamin et lal.1 there'are a number of modifica-tions to the fuel rack design which would enhance convective cooling and re-duce the potential for cladding fires. However, the limited flow areac of the high density designs make it difficult to ensure adequate cooling by natural convection of air. For the assumption of annual discharges, the critical decay time can be-translated into a likelihood of cladding fire for a complete loss of pool water inventory. For the critical cooling times given in Table 3.1 the proba-bility of self-sustaining oxidation is approximately: 0,0 to 0.5 for BWRs with low density storage racks. 0.0 to 0.7 for PWRs with low density storage racks, and f 1.0 for PWRs with high density' storage racks. 3.1.3 Clad Fire Propagation 1 The: SNL investigations,2 of spent fuel behavior after a loss of pool integrity accident (assumedio result in complete drainage of the pool), iden-tified a range of power levels necessary for the initiation of self-sustaining clad oxidation and substantially lower power levels at which adjacent fuel bundles would oxidize once oxidation had been initiated. However, the phenom-enology of propagation is not well understood and there is considerable k 5 .i

3-4 uncertainty in these estimates. Benjamin et al.,1 Pisano et al.,2 Han3 and Johnsen" have pointed out a number of limitations in the previous analy-ses.1,2 In order to put the pre..it results in perspective it is worth mentioning the most important limitations: 4 l 1. The oxidation equation allows oxidation to continue beyond 1900*C (3450*F) where clad melting and relocation is expected. PBF and KfK tests show clad ' relocation at temperatures in the range of 1900* to 2200*C but the analyses have calculated temperatures as high as '\\ 3500*C (6330*F) without accounting for clad and fuel melting. At such high temperatures the radiation heat flux becomes very large and it is believed that the potential for propagation to adjacent bundles will be overestimated.. In order to provide more realistic estimates of the potential for oxidation propagation, BNL has chosen to terminate oxidation at the Zircaloy melting point. 2. The SFUE3, code had not yet been validated successfully against fuel ) 2 rod oxidation data, A preliminary comparison against SNL data was only partially successful. The SFUEL code has been compared to the SNL data in a separate sec-tion (3.2) and key portions of the code have been validated. Specif-ically, the axial. heat up (without oxidation) and the temperature at which self sustaidng oxidation is reached has been validated. If a low power spent fuel bundle heats up to within one or two hundred 'C of self-sustaining oxidation due to its own internal energy there is a high likelihood that the additional energy from an adjacent high power bundle will be sufficient to bring it to the initiation point. 3. The reaction rate equation has been criticized as being too low for long tem exposure, at low temperatures (when oxide layers may flake off and expose fresh Zircaloy). However, Appendix B has shown that the SFUEL calculations are not very sensitive to the low temperature oxidation rate. ~~ ~ ~~

3-5 4. The lack of a fuel and clad melting and relocation model has also been criticized. Development of realistic degraded fuel models is beyond the scope of the present investigation. However, we believe that the modified 2 SFUEL code (SFUEL1W ) has sufficient flexibility to estimate the im- } portance of oxidation propagation. 5. Johnson" criticized the clad failure criteriori used in the SNL ana-lyses.1,2 He noted that the clad failure could occur at tempera-tures as low as 650*C if the thermal loading in sustained for several hours. In view of the large uncertainty in the thermal behavior, we agree that a prediction of temperatures in excess of 650'C should not be viewed as successful cooling of the assembly. At these temperatures cladding failure and fission product release is very likely and the potential for cladding " fires" is high due to the effects of asymmet-ric heating (from adjacent high powstr bundles). Propagation of cladding " fires" by particulate (i.e., spallation) or zir-conium vapor transport has been investigated and eliminated in an approximate separate effects study by Pisano et al.2 However, propagation due to the heat flux (radiation and convection) from adjacent bundles is predicted to occur even to very low power assemblies (at power levels corresponding to 3 years of decays). The purpose of this section is to establish the range of conditions for which propagation is predicted to occur. Both the power of the initiating bundle and the power of the adjacent bundles have been varied as well as the ventilating conditions of the spent fuel building. It should oe emphasized that SFUEL does not address the question of Zir-caloy oxidation propagation after clad melting and relocation. For recently l iiischarged fuel (less than 90 days), or for severely restricted air flow i (e.g., nigh density PWR spent fuel racks) the oxidation reaction is predicted

777) [g 3 gyy y ;. s 4 3-6 ' (' a i r 24. f. 4 to bl. very vigocous Srd -failure: of both thef fvel rods and the fuhlir 'iseypected. Thus a large fraction n,f thel fuel ' rods would be ardected to fall! '{ to the bottom of the pool forming a'large debris bed. If. water!is not present ., a - /,1r.: the bottom of the pool, th,e ilebris bed _ will remainj hot [and will' tend to .7 heat the adjacent assemblies from below.; The investiostion >of debris bed for-z p.i t 1 [ mation is beyondlthe scope of the present' study, but itt' appears \\ c be an addi- { t r tional mechanism f sp oxidation propagation., -l ~ l. i / ,i r .[,' Results (,f, j As, pointed out in. Section 1.1.1 self-sustaining, uidation inivi.a. tion is'

j

. not ve"y sensitive to the oxidtt.lon' rate equatiopbut it is dependent upon' the

calculatert air flow (related to flow resistance) and. the power.1kM.

BWR i spent fuel with its low power density and' openjflow configuration must' be) re-cently discharged (within about-3. months) for self-sustaining oxidation to'be initiated and unless it;is a very high power bundle (discharged 'within 10 days or less) there is on1p'a slight chance of: propagation'to older low power fuel .j l-bundles. .) However, PWR spent 'fudl racks typically have a higher power deris'ityc and j U more flow # restriction, thus'salfwustaining oxidation can be initiated'in fuel-that has been discharged for one.ye w or more. Two fuel building ventilation conditions have been irlyestigatedi as de-d scribed below but it must be recognized that both of these assumptions corre-spond to.very idealized conditiions,that are unlikely: to' be' duplicated, in an actual accident. Rather these idealized conditions are provided. to demon-strate the importance of the various 3 assumptions. For a beyond design basis seismic event,;that catastrophically fails' tha pool,: it seems likely the faii-- ure of the fuel, building'may also occur.. However, Benjamin et al.( have sh'own 2 that'a very Isrge hole (at least 77 ft ) must be: opened in order to approxi-mate the perfect ventilation case. ~ 't q h

• k pm pas.

37 ) Perfect Ventilation / Under the perfect ventilation condition it is assumed that the fuel building is maintained at ambient conditions by a high powered ventilation system (note that the flow rate must be much higher than typical gas treatment 2 systems) or by a large opening (greater than 77 ft ) in the building. Oxygen is not depleted and the air entering the pool is assumed not to be heated by the hot gases exiting the fuel assemblies. The conditions necessary to initi-ate self-sustaining oxidation under perfect ventilation conditions were sum-marized in Table 3.1 for three typical fuel rack configurations. Note that these are " borderline" conditions in that a slightly lower power level or a larger inlet hole size would be predicted to prevent self-sustaining oxidation from occurring. Note that the " critical" conditions outlined in Table 3.1 do not imply that fuel rod failure is not predicted for power levels below these conditions. The power level must be reduced substantially (about 20%) to en-sure that the predicted clad temperature is below 650"C (the minimum tempera-l ture at which clad fdlure and fission product release is likely 5 to occur). For power and flow conditions that are only slightly below the "criJical" l i conditions it should be obvious that the b3at flux from a much higher power adjacent bundle would have the potential to push the "non-critical" fuel over the self-sustaining oxidation threshold. Thus the only real propagation ques-tion is. whether recently discharged (high power) spent fuel will radiate suf-ficient energy to initiate self-sustaining oxidation in low power fuel bundles that have been cooled for one or more years. In this context two limitations of the SFUEL1W2 code should be noted: l 1. The fuel storage racks are assumed to be imediately adjacent so that no air flow between racks is allowed. (The numerical approach used to calculate the heat transfer is numerically unstable if flow is allowed). 2. All fuel storage racks are assumed to be identical so that the q'ues-tion of propagation from high power cylindrical racks to low power high density racks cannot be addressed. 1

3-8 The first limitation probably repr7sents current storage practices where a number of fuel pools are approaching their design capacity. However, the ]' question of providing deliberate cooling channels between recently discharged fuel and the older fuel cannot be directly addressed. Based on engineering insight, it appears that, under the idealized perfect ventilation conditions, the provision of an air space of 6 to 12 inches around the periphery of re-l cently discharged fuel would minimize the likelihood of oxidation propagation to low power spent fuel assemblies. (Note that the code does allow for an air space adjacent to the pool walls and 6 to 12 inches is found to be adequate if flow through the bundles is not restricted.) Since high density fuel storage racks are predicted to cause self-sustaining oxidation even after storage for one or more years, it seems clear that it would be undesirable to store spent fuel in high density storage racks if it has been discharged within the last two years. (It may be worth noting that current practice restricts the storage density of low burnup fuel due to nuclear criticality considerations.) Thus the question of propagation from cylindrical fuel racks to high density fuel racks should be addressed, but the second limitation mentioned above precludes intermixing of the storage rack ') configurations. t The propagation results with perfect ventilation are summarized in Table 3.2 for the high density rack configuration described in Reference 2. Note that the lowest power (11.0 kW/MTU) for self-sustaining clad _ oxidation corres-ponds approximately to fuel that has been discharged for one year, but the oxidation reaction will generate sufficient energy to propagate to a fuel bun-die that is about 2 years old (6.0 kW/MTU). For a fuel assembly that has been discharged for about 10 days (90. kW/MTU) the high decay heat level causes ex-tensive clad oxidation in the high power bundle and a somewhat higher propen-sity to propagate to low power fuel assemblies (as low as 5 kW/MTU which cor-responds roughly to a 2-1/2 year old discharge). The propagation results for a low density fuel rack (cylindrical) with a 3 inch diameter inlet hole is sumarized in Table 3.3. Note that the range of power for the high power assembly is limited due to the improved. free convec-tion within this type of fuel rack. Thus self-sustaining clad oxidation is

3-9 initiated at decay power levels at or above 30 kW/MTU (corresponding to about 90 days of cooling). Assuming that more than one discharge per year is un-likely, the adjacent low power assembly must be less than or equal to about 19 kW/MTV (180 days of cooling). Thus propagation only occurs for fuel that has been discharged less than 1 year with initiation from fuel that has been dis-charged within 2 weeks. For a PWR cylindrical' fuel rack with only a 1.5 inch diameter flow hole, the air flow is much more restricted and the possibility of propagation is stronger as indicated in Table 3.4. For the 1.5 inch hole size propagation is predicted to occur' for cooling times as long as two years. Inadequate Ventilation As previously mentioned the case of perfect ventilation implies a very high ventilation rate that is not normally possible. Benjamin et al.1 extend-ed the SFUEL code to consider limited heat removal to just keep the spent fuel building at constant pressure. Details of the modeling are described in Ref-erence 1, but the main result of the model is that the fuel building atmos-phere heats up (due to decay heat and the chemical energy of oxidation) and the oxygen is depleted. Benjamini found that the heat-up of the buildi$g in-creased the likelihood of self-sustaining oxidation (i.a., decreased the decay power level necessary to initiate self-sustaining oxidation). This section is intended to address the question of whether limited ventilation also increases the likelihood of propagation to low power bundles. Table 3.5 provides a samnary of propagation runs under inadequate venti-lation conditions. For the analyses the high power assemblies are modeled to represent approximately 1/3 of the core for 1000 MWe plant and the fuel build-3 ing is taken to have a volume of 150,000 ft. The results given in Table 3.5 indicate that propagation is no more likely with inadequate ventilation than with perfect ventilation. In fact propagation does not occur for several con-ditions listed in Table 3.5 for which propagation was predicted with perfect ventilation. Although this result is somewhat surprising, it is simply a re-sult of the oxygen depletion calculation. That is, the oxidation of the 1

m 3-10 recently discharged assemblies uses up the oxygen supply before ' the lower power assemblies can be heated to'the point of seif-sustaining oxidation.. [ H In view of the ' potential. for fuel building 1 failure due to either ~ the assumed initiating ' event (e.g., a beyond design basis earthquake) or the rapid- ) building pressurization from_ Zircaloy combustion and decay heat,. BNL considers the oxygen depletion calculation to be unrealistic. Thus,. in spite of the-many uncertainties, the perfect ' ventilation model is expected to give the best approximation for the potential for propagation. Conclusions Regarding Propagation Based on the previous results we leave concluded that the modified SFUEL 2 code (SFUEL1W ) gives a reasonable estimate of the potential for propagation-of self-sustaining clad oxidation from high power spent fuel to low power spent fuel. Under some conditions, propagation is predicted to occur for spe'nt fuel that has been stored as long as 2 years. 1 l The investigation of the effect of insufficient ventilation in' the fuel ') ! building indicated that oxygen depletion is a competing factor with heating of 1 the building atmosphere and propagation is not predicted to occur for spent fuel that has been cooled for more than three years even without ventilation. y These results are in general agreement with the earlier _SNL studies.1,2 3.2 Validation of the SFUEL Computer Code 1 The SNL investigations,2 of spent fuel behavior. after a loss of pool integrity accident (assumed to result in complete drainage of the pool), iden-tified a range of power levels necessary for the initiation of self-sustaining clad oxidation and substantially lower power levels at which adjacent: fuel bundles would oxidize once oxidation had been initiated. However, an attempt 2 to validate the code was only minimally successful in that the post-test ana-lyses were able to match the heat-up rate in helium (without oxidation)L but' the SFUEL code over-estimated the temperature transient after air was intro-duced. .m,,,,

3-11 The objective of this section is to use the revised 3 oxidation rate equa-tion in SFUEL to analyze the SNL small scale tests to aid in validating the SFUEL code. The SNL tests are described in Reference 2, but in order to put the test results in perspective several important conditions should be high-lighted: 1. Tne test was of a small bundle of electrically heated rods (9 rods) ^ with a short length (38 cm). 2. In order to achieve self-sustaining clad oxidation f>850*C) the rods l were heated with a very low flow rate of helium before air wa: admit-ted to the test assembly. Under these test conditions the dominant heat loss is via radiation whereas for the postulated accident the dominant heat loss is via free convec-tion. These test conditions lead to laminar flow (a Reynolds number of about 100) in which oxygen diffusion to the cladding surface limits the reaction rate. Only one test (6) had a sufficiently high air flow rate to allow vig-ourous oxidation. Since the free convection and radiation calculations in SFUELb2 were inappropriate to the test configuration, Pirano et al.2 created a stripped down version called CLAD 2 which used a matrix inversion routine to calculate radiation losses. I After several preliminary attempts to analyze the helium portion af the tests we concluded that there were several errors which led to underestimation of the convection portion of the heat losses. Since helium has a much higher heat capacity and conductivity than air it appears to contribute to establish-ing the initial conditions. In order to provide an adequate simulation of the initial steady-state portion of the test we made two modificat' ions to the CLAD code: 1. Include helium properties with a switch to air properties at the i start of the transient. l i l l l \\

3-12 2. Include an energy balance on each gas control volume to force conser-vation of energy. With these changes we were able to obtain an adequate simulation of the initial portion of the tests. Using this revised version of CLAD with the Weeks' oxidation correlation,3 analysis of both the helium and the air por-tions of.the test looked reasonable, but still tended to over-predict the peak temperatures during oxidation. In order to bring the calculations into rea-sonable agreement with the small scale data the Weeks' correlation has been reduced by a factor of four (note that this corresponds approximately to the datascatter). Results The revised CLAD code has been used to analyze the SNL small scale exper-iments Tests 4, 5 and 6. The other three tests were intended to simulate propagation with nonuniform heating and structures that CLAD was not capable of modeling. The CLAD results for Test 4 are compared to the data in Figure 3.1. These results still tend to overpredict the temperature in the center of ) the test rod, but give reasonably good agreement at the top of the rod where radiative heat losses are large. The peak temperatures calculated by CLAD are summarized in Table 3.6 and compared to the peak measured temperatures for the three tests. Note that CLAD still overpredicts t'he peak temperature for the low flow rate test (4 and

5) but gives good agreement with the high flow rate tests where adequate oxy-gen is available.

It'should be noted that this " oxygen starvation" phenomenon appears to be a result of the extremely low laminar flow where o'xygen must diffuse to the clad surface. CLAD includes an oxygen depletion calculation but assumes that all the oxygen in each volume is immediately available at the surface. 3.3 Conclusions Regarding SFUEL Analyses After an extensive review of the SFUEL code and comparison to the SNL small scale experiments, BNL concludes that the code provides a valuable

3-13 tool for assessing the likelihood of self-sustaining clad oxidation for a variety of spent fuel configurations assuming that the pool has been drained. e The SNL small scale data provide a reasonable degree of validation for the heat-up and oxidation models, but the results are extremely sensitive to the natural convection calcu.lation which has not been validated. When oxidation is terminated at the Zircaloy melting temperature (assum-ing that the molten Zircaloy is relocated), oxidation propagation only occurs for spent fuel bundles which are already approaching the " critical" conditions for self-sustaining oxidation (see Table 3.1). However, this finding dees not mean that oxidation propagation is unlikely. On the contrary, for some high density storage configurations the " critical" conditions are approached fur spent fuel that has decayed for two to three years. Thus clad " fire" propaga-tion appears to be a real threat but the basic question remains as to what are the " critical" conditions for initiation of oxidation and what the uncertainty is for a given spent fuel configuration. The critical conditions are summar-ized in Table 3.1 for several typical spent fuel racks. While the heat-up and oxidation models have been validated to a limited extent by the SNL data (see Section 3.2), the authors believe that the largest source of uncertainty is in the natural convection flow rate. It is recommended that these free convec-tion flow calculations be verified against large scale data. Preferably the data would be obtained from spent fuel assemblies in typical storage racks (both high and low density). 3.4 References for Section 3 1. A.S. Benjamin, D.J. McCloskey, D.A. Powers, S.A. Dupree " Spent Fuel Heat-up Following Loss of Water During Storage," NUREG/CR-0643, March 1979. 2. N.A. Pisano, F. Best, A.S. Benjamin K.T. Stalker, "The Potential for Propagation of a Self-Sustaining Zirconium 0xidation Following Loss of Water in a Spent Fuel Storage Pool," Draft Report, January 1984. I 3.. J.T. Han, Memo to M. Silberberg, USNRC, May 21, 1984 4. G.W. Johnsen, Letter to F.L. Sims. EG&G, Idaho, April 4,1984.

3-14' Table 3.1 Summary of Critical Conditions Necessary toj Initiate Self-Sustaining Oxidation. Inlet Orifice Minimum Approx. Critical (l) l Spent Fuel Rack Diameter Decay Power Decay Time Configuration (inches) (kW/MTU)- . (days). i High Density PWR l (6 assemblies per. rack)' 5 6 700 High Density PWR (6' assemblies per rack) 10 11. 360 Cy1indrical P.WR - 5 - 90. 10-Cylindrical PWR' 3 45 50(2) Cylindrical -PWR. 1.5 15 250(2) Cylindrical BWR 1.5 14 180 Cylindrical BWR 3.0 70 <10 II)Crh. cal cooling time is the' shutdown time necessary to reach a decay power level below the minimum decay power for self-sustaining oxidation. l The cooling time to prevent cladding failure. is at least 20% longer.. s 1: (2) Note that these critical cooling times are somewhat lower than that found by Benjamin et al. since the orifice lots coefficient was modified at BNL 4

---

T-"

3-15 -1 Table 3.2 Summary of Radial Oxidation Propagation Results for a High Density PWR Spent Fuel Rack with a 10 Inch Diame-ter Inlet and' Perfect Ventilation Approximate High Power Level Adjacent Power Level Decay Time (kW/MTU) (kW/MTU) (days) Propagation 11.0 5.9 365 Yes 19.2 5.9 365 Yes 90 5.9 365 Yes 90 4.0 730 No i eh

i l 3-16 4 Table 3.3 Summary of Radial 0xidation Propagation Results for a- ) Cylindrical PWR Spent Fuel Rack with a 3 Inch Diameter ') ~l Hole and Perfect Ventilation High Power Level Mjacent Power Level._ Approximate Decay. Time (kW/MTU) (kW/MTU) (days) Propagation 90 11.0 365 No ~ 90 19 180 Yes*

• Note that this is an unlikely situation in that the. conditions. imply a six month period between discharges.

e Table 3.4 Summary of Radial Oxidation Propagation Results for a: Cylindrical PWR Spent Fuel Rack with a 1.5 Inch Diame-l ter Hole and Perfact Ventilation l l Approximate High Power Level Mjacent Power Level Decay Time .f (kW/MTU) (kW/MTU) (days) Propagation ? 90 11.0 365 Yes 90 5.9 730 -Yes 90 3.0 1100 No ' 15 11.0 365 Yes 15 5.9 730 No l i

3-17 Table 3.5 Summary of Radial Oxidation Propagation Results for Various PWR Spent Fuel Racks with No Ventilation f High Power Level Adjacent Power Level Spent Fuel Rack (kW/MTU) (kW/MTU) Propagation Cylindrical with 1.5 inch hole 90 5.9 Yes Cylindrical with 1.5 inch hole ' 90 3.0 No (0 depletion) 2 Cylindrical with 3 inch hole 90 5.9 No 1 Cylindrical with 3 inch hole 19.2 11.0 Yes High Density with 10 inch hole 90 4.0 No (0 depletion) 2 h O I

3-18 Table 3.6 Comparison of SNL Small Scale Oxidation ' Tests to' Calculations with CLAD' Peak Temperatures Data CLAD Air Flow Rate ("c) ("C) Test ' (1pta) Mid Top 4

12-1570 19(10 1400 5

28.3 ~1850-1960 1660' 6 56.6 >2003* 2100 1800

• Thermocouple failure.

I ) eh I I _D

w. 0 3 r e en 0 i 2 l q n i f p o 6 2 r e ~ e d t i n n s 4 i e n 2 ) p c i s 4 e t r f f t 2 s u e e o o 2 t n T n t t i e h h m r ( o c g g 0 f i i 2 f e e n a o h h e t + g a p d d y d l l o i i x I t m m oL 0 k N f S A O D 6 o O k o 1 D n t O oD i A O 4 t L 1 cC D u AOM f d o l o f L 2 r n C 1 t o n s I i r ' 0 ma 1 p ~ o m ~ r o f C g 8 e m1 O iT3

• 6 erug i

F

• 4

' 2 ~ ,c O D 0 0 0 0 0 0, 0 0 \\ C 0 0 0 0 0 0 0 0 6 5 4 3 2 1 0 9 8 1 1 1 1 1 1 1 U* if. f s a

4-1 4. CONSEQUENCE EVALUATION : A PWR and a BWR reactor were selected for risk evaluation based on a. pre-l liminary screening of perceived vulnerability and the spent fuel puol inven-tory. The reactors selected were Ginna 'and Millstone 1. Both. are older plants that were built before the current seismic design criteria were promul-gated and have relatively large inventories of spent fuel. 4.1 Radionuclides Inventories The radionuclides inventories for both the PWR and BWR pools were calcu-2 lated using the ORIGEN2 Computer Code for the actual operating and discharge histories for Ginna and Millstone 1. The ORIGEN2 program in ese at BNL was verified by comparison with results obtained at ORNL for.-1'dentical cases.3 A description of the -assumptions and methods of. analysis is given in Appendix A along with the detailed results for each species. The results for 'the risk significant species are summarized in Table 4.1 (Millstone.1) and Table 4.6 - (Ginn3). t For both plants, the noble gases and halogens in the spent fuel inventor-ies are a small fraction of the inventory !in an. equilibrium core at shutdown - except.for freshly discharged fuel, but cesium and strontium are more than three times the equilibrium inventory (see Tables 4.1 and 4.6). 4.2 Release Estimates The fission product release fractions have been calculated for two limit-ing cases in which a Zircaloy fire occurs: In Case 1, the clad combustion is assumed to propagate throughout the pool and the entire inventory is in-volved.- In Case 2 only the most recently discharged fuel undergoes clad com- )

bustion, The release calculations for Cases 1 and 2 make the assumption that if the spent fuel pool suffers a structural failure, coolant inventory will be totally drained, i.e., the leak rate will greatly exceed makeup capability

..___s_ -_______m .m__ ______-m__ ________m

4-2 even if the coolant systems are still available. The probability of Zircaloy fire and fission product release has been determined from BNL calculations de- ) scribed in Section 3. In order for a cladding fire to occur the fuel must be recently discharged (about 10 to 160 days for a BWR and 30 to 250 days for a PWR). This leads to a conditional probability for a Zircaloy fire of.28 for e BWR and.40 for-a PWR. If the discharged fuel is put into high density racks the critical cooling time is increased to one to three years and the conditional probability of a Iircaloy fire is increased to a virtual certain-ty. A reevaluation of the cladding fire propagation estimates indicates that there is a substantial likelihood of propagation to other fuel bundles that have been discharged within the last one or two years. Subsequent propagation to low power bundles by thermal radiation is highly uniikely, but with a sub-stantial amount of fuel and cladding debris on the pool floor, the coolability of even low power bundles is uncertain. 4.2.1 Estimated Rele'ases for Self-Sustaining Cladding 0xidalion Cases (Cases 1 and 2) e As discussed in Section 3.1 there are a broad range of spent fuel storage conditions for which self-sustaining oxidation of the cladding will occur if the water in the pool is lost. For Ginna with high density racks the condi-tional probability of a cladding " fire" is predicted to be nearly 100% while for Millstone 1 the probability is about 20%. If self-su:,taining oxidation occurs the fuel rods are predicted to reach 1500 to 2100'C over a substantial portion of their length. At these temperatures, the release fraction is pre-dicted to be substantial. Rough estimates of the fractional release of various isotopes have been presented in an attachment to Ref. 4 Included in the estimates were noble gases (100%), halogens (100%), alkali metals (100%), tellurium (2 to 100%), barium (2%), strontium (0.2%) and ruthenium (0.002%). Estimated release fractions of other isotopes are given in Table 4.2. These estimates are based on various considerations, including experimental

= i data (tellurium), location of the isotopac (whether in cladding as activation ' products 'or in fuel pellets as fission products), and melting / boiling points of the element or oxides' of the element. Comments or, the estimates listed in Table,4.2 follow: Tellurium:. The releases shown assume the lower limit 'of.Ref. 4 based on the tellurium release model recently propsed by Lorenz, et-al. 5 The low release-value assumes that a fraction of the Zircaloy cladding relocates (melts and flows downward) before oxidation is complete.s Alkali Earths: Because of. the high boiling points - of. the oxides of ' Sr and Ba, it is estimated that only a very small fraction (2x10 2) of these elements of fission product origen in the fuel pellets es-cape. It is estimated that 100% of the activation product Sr-89.and Y-91 contained in the Zircaloy cladding are released as aerosols. Transition Elements: it is estimated that 100%- of the transition l element activation products contained in the cladding' are. levitated-as aerosols of the oxides (smoke).. Note that the small release frac-tion of Ir-95 (0.01) takes into account:the large-inventory of fis-sion product Zr-95 trapped in the fuel pellets.. 1 It is assumed that only 10% of the activation products-in the assembly hardware escapes (see Table 4.2, Fe-55, Co-58, Co-60 and.Y-91). le Co-60 fraction is corrected for its small content in the cladding. Antimony: It is estimated that '100% of the' SB-125 is roasted out' of ' the fuel pellets, because of its high mobility. Lathanides and Actinides: A negligible release of the oxides of the latitanides and actinides is estimated because of their enorm 'l sta-- bility, low vapor pressures and ceramic characteristics. O

p q ' 44 : ] Case 1: Case 1 the " worst" case, assumes an accident that'results in 1 a Zircaloy fire that propagates throughout theLentire ' spent fue1Lin-I[ ventory in the pool, and that ' the accident occurs 30' days after the ] reactor. was shut down for discharge of the lastL fuel. batch..The est1. : mated. releases of radionuclides are listed in Table 4.3. These were obtained by combining.the "30-day" inventory given in column ' 3 of - Table 4.1 with the release fractions listed in Table 4.2. Case 2: Case 2 assumes an accident that results in a Zircaloy fire that involves only the last fuel batch to be discharged, and that the y accident occurs 90 days after the reactor was shut down for. feel dis-charge. The estimated releases of radionuclides'are listed in Table 4.4. These were obtained by hombining the inventory in the last fuel ) batch (data tabulated in Table A.6 of Appendix A) with the release: i fractions in Table 4.2. l 4.2.2 Estimated Release for Low-Temperature Cladding Failure (Cases 3 and 4) i t j For a less severe accident in which fuel is exposed to air but does-not. reach temperatures at which a Zircaloy fire, ignites, it is -assumed that the j cladding on many fuel rods will fail (i.e., develop leaks): resulting in a're-lease limited to the noble gases and halogens. Two limiting cases have been considered: Case 3: in which the entire pool is drained but the decay time since the last discharge is one year, and 50% of the fuel rods suffer clad rupture. Case 4: in which the pool drains to a level that exposes.the upper' por-1 tion of the fuel assemblies, the decay time for the list discharged fuel i bctch is 30 days, no Zircaloy fire. occurs but all of the fuel rods in the last discharged batch rupture. The estimated releases for Cases 3 and 4 are given in Table 4.5. i s 1

4-5 4.3' Off-Site Radiological Consequences ( 4.3.1 Scenarios for Consequences Calculations The off-site radiological consequences have been calculated using the CRAC2 computer code.' The scenario used in the CRAC2 calculations consisted of the following conditions: a generalized sita surrounded by a constant population density of'100 persons per square mile; generalized meteorology (a uniform wind rose, average weather condi-tions); and the population in affected zones was relocated after 24 hours. The radiological effects were calculated out to E a distance of 50 to 500

miles, CRAC2 calculations were made for, a range of possible re16ases as de-l scribed in Section 4.2.

The consequences are sumcarized in Table 4.7. 4.3.2 Consequence Results There are several unusual characteristf es ~ of a spent fuel accident that cause somewhat surprising results in the radiation exposure calculations. Specifically, the radiation exposure is insensitive to fairly large varittions in tha etitimated releasc. This is.due principally to the health' physics assumptions within CRAC. For the long lived isotopes (predominantly cesium), the exposure is due mainly to exposure after the area is decontaminated and people return to their hmes. The CRAC code asnwes that decontar.ination will I 1 limit the exposure of each person to 25 rem. Thus, for this type of release the long term whole ' body dose is. lin;ited by the population in-th'e affected sectors (about 0.8 million people in the 16 sectors for a 50 mile radius) to about 3x106 person-rem (only 3 of the 16 sectors are~ downwind). 1 ~ e __i

!a6 The extreme cases (1A; immeof ately after refueling and 18 and IC; with the total fuel pool inventory involved) result in much higher releases but no significant change in population dose. A more sensitive inoication of the consequences for a spent fuel accident is the interdiction area (the area with such a high level of radiatio 3 that it is assumed that it cannot ever be decontaininated). As indicated in Table 5 the worst spent fuel accident is calculated to result in an interdiction area o'f 224 sq. niiles. This is about two orders of magnitude higher than the interdiction area computed for ructor core melt accidents (about i to 10 2 m1 ). 4.4 Rgences for Sectior: 4 1. BNL Memorandum, V.L. Sailor and K.R. Perkins to W.T. Pratt, " Study of Be-yond Design Basis Accidents in Spent Fcei Pools," May 8 1985. 2. A.G. Croff, "0RIGEN2: A Versatile Computer Code for Cal'culating the Nu[ clide Compositions and Characteristics of Nuclear Materials," Nuclear Technolocy_, Vol. 62, pp. 335-352, September 1983. 3. Internal Memorandum, Brookhaven National Lahoratory, from V.L. Sailor to R.A. Bhri, " Comparison of BNL ORIGEN2 Calculations with CRNL," May 27, 1986. 4. Memorandum of J.T. Han to M. Silverberg, " Response to a NRR Request to Review SNL Studies Regarding Spent Fuel tieatup and Burning Following Loss of Water in Stortge Pool." U.S. Nuclear Regulttery Commission, (May 21, 1984). 1 { 5. R.A. Lorenz, E.C. Beshm and R.F Wichner, "'feview of Tellivrium Release j Rates from LWR Fuel Elements Under Accident Conditions," Proceedings of the International Meeting _ on Light Water Reactor Severe Accident Evalua- ) tion, August 28-September 1,1983, pg. 4.4-1, Anerican Nuclear Society j Order 700085, ISBN 0-89448-1117-6. 1 L_____._____...___ i

' 4-7 6. L.T. Ritchie, J.D. Johnson and R.M. Blond, Calculations of Reactor Acci-dent Consequences Versior. 2, CRAC2: Computer Code, tiser's~ Guide, pre-pared by Sandia National Laboratories for_ the U.S. Nuclear. Regulatory Comission, NUREG/CR-2326 (SAND 81-1994), February 1983. e 991D 0 e '91 4 9

c 4-8 Table 4.1 Comparison of Radioactive Inventories of Equili-brium Core with Spent Fuel Assemblies for Select-ed Isotopes (Millstone 1) . Equilibrium toont Fuel Poole (time after last d!6eharge) I sotope core 30-days 90-ceys 1 year (Totel Redicactivity. Cueles) Co 58 8.81 E+4 2.29E+4 1.26E+4 8.54 E+2 s Co 60 1.64E+5 3.72 Ee 3.15E6 2.85E+3 Kr 85 5.3 5E+5 1.41 E+6 1.39E+6 1.35E+6 Rb 86 6.22 E+4 1.01 EM 1.05E+3 3.64Ew2 sr 89 4 #1E+7 8.39E46 3.63E+6 8.33E+e sr 90 4.25E+6 8.42E+7 1.42E+7 1.39E+7 Y 90 4.37E+6 1.43E+7 1.42E+7 1.39E+7 Y 91 6.06E+7 1.18E +7 5.75E+6 2.21E+5 Zr 95 8.70E+7 1.94E+7 1.00E+7 5.10Ee Nb 95 8.91E+7 2.54 E+7 1.70E+7 1.11 EM leo 99 8.78E+7 1.49E+4 3.12E-3 neg. Tc 99m 7.69E+7 1 A3E+4 3.01 E-3 neg.D Ru 103 7.23E+7 1.53E+7 5.21E+6 4.07E+4 Ra 106 '2.48E+7 1.72E+7 1.53E+7 9.13EM Rh 106 2.63E+7 1.72E*7 1.53E+7 9.15E+6 sb125 9.07Ee 1.19E+6 1.14E+6 9.48E+5 Sb 127 4.97E+6 8.21 E+3 1.39E-1 neg.g is125m 1.93Ee 2.84Ee 2.76E+5 2.31E+5 To127 4.92E+6 2.21 E+5 1.45E M 2.52E+4 To127e 6.6t E6 2.18E +5 1.48Ee 2.57E+4 To129 1.49E+7 2.74E e 7.79E+4 2.68E+2 Ta129m 2.24E+6 4.21 E+5 1.20E+5 4.12 E,+2 To 132 6.72 E +7 3.74EM 8.64E-2 neg. 1129 ).75E+o 7.1 SE+0 7.1 SE+o 7.1 SE+0 1 131 4.74E+7 1.22E+6 6.35E6 neg.b ) I 132 6.83E+7 3.85E+4 8.90E-2 neg.b Xe 133 9.72E+7 7.29E+5 2.30E+2 f.eg.b Cs 134 6.10E+6 7.90E+6 7.4?E+6 5.80EM Cs 136 2.10E+6 2.0'EM 8.13Ee 3.91E-3 p Cs 137 5.84E+6 2.02E+7 2.01 E+7 1.97E +7 Be 137m 5.53E+6 I.91 E+7 1.90E +7 1.87E+7 Ba 140 8.36E +7 5.19EM 1.90E6 6.41E-2 Le 140 8.54E+7 5.97E+6-2.19EM 7.37E-2 Co 1 di 7.94E+7 1.32E+7 3.61 E+6 1.03EM Co 144 6.05E+7 2.64E+; 2.27E+7 1.16E+7 Pr 143 7.37E+7 5.44E+e 2.41E+3 ~~1.90 E-1 Pr 144 6.08E+7 2.64E+7 2.27E +7 1.16E+7 Pd 147 3.16t+7 1.54E+6 3.36E +4 1.10E-3 Sm 151 2.44EM 8.22 EM 8.21 E+4 8.16E +4 Eu 154 4.61E+5 1.34E+6 1.32E+6 1.25EM Eu 156 5.61 EM 8.26Ee 5.1 0E M 1.80E-1 Np 239 9.98E*6 5.59E+4 2e*ME+3 2.88Ee Pu 238 9.33E+4 4.515+5 4.53E+5 4.54 E+5 Pu 239 2.49EM 0,MK+4 8.89EM 8.89E+4 Pu 240 3.14EM 1.3E +5 1.30E+5 1.30E+5 Pu 241 7.19EM 2.29E,7 2.2 7E+7 2.19E+7 ha 241 8.86E+3 2.88Ee 2.94E+5 3.21E+5 On 242 2.09E+6 1.45EM 1.12 E +6 3.50E+5 l On 244 6.72 E+4 2.27E+5 2.2 SEe 2.19E +5

• Spent fuel pool Irwentory incNdes discharges from 11 refuelings cover
• Ing the portos from August 1972 through the projected refueling of 4rtl 1987.

bneg. = less than l o"3 Curles. i I '1 _a a_ _ - - - - - -. -

l. 4-9 Table 4.2 Estimated Radionuclides Release Fraction During a ipent Fuel Pool Accident Resulting in Complett Destruction of Cladding (Cases -1 and 2) Release Fractiona Element or val ue. Uncertainty t Chemical Family Isotope Used Range Noble gases Kr, Xe 1.00 0 Halogens I-129, 1-131 1.00 0.5-1.0 Alkali Metals Cs, (Ba-137m) Rb 1.00 0.5-1.0 Chalcogens Te,(I-132) 0.02 Alkali Earths Sr (Y-90), Ba (in fuel) 2 x 10-3 10-4-10-2 Sr, Y-91 (in clad) 1.00 0.5-1.0 Transition Co-58 (assembly hardware)b 0.10 0.1-1.0 El ements Co-60 (assembly hardware) 0.12 0.1-1.0 Y-91 (assembly hardware) 0.10 0.1-1.0 Nb-95, Zr-95 (in fuel) 0.01 10-3 10-1 Nb-95, Zr-95 (in clad) 1.00 0.5-1.0 Miscellaneous Mo-99 1 x 10-6 10 8-10.s Ru-106 2 x 10.s 10-6-10-4 e Sb-125 1.00 0.5-1.0 Lanthanides La, Ce, Pr, Nd, Sm, Eu 1 x 10-6 10-s-10-5 Tra'nsuranics Np, Pu, Am, Cm 1 x 10-6 10-e-10-5 aRelease fractions of several daughter isotopes are determined by their precursers, 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-140 by Ba-140. bRelease fraction adjusted to account. for a 100% release of the small amount of Co-60 contained in the Zircaloy cladding. l

4-10 Table 4.3 Estimated Raleases of Radionuclides for Case 1 in Which a ZirCa10y Fire Prop 8 gates Throughout the Entire P001 Inventory (W0rst Case) Time after Last 01scharge isotope 30-days 90-days I yeet" (Radioactivity, Curlest l Co 58 2.74E4 I.51 E+3 1.02 E42 Co 60 4.46E*4 3.78E+4 3.42 E +4 Kr 85 1.41 E+6 1.39E+6 1.33E46 Rb 86 1.01 E+4 1.0$E+5 3.84E-2 Sr 89 1.68EM 7.26E+3 I.67E+2 Sr 90 2.84 E +4 2.84 E44 2.78E +4 Y 90 2.64 E+4 2.84 E *4 2.78EM Y 91 1.18E +6 5.75E+5 2.21 E44 Zr 95 1.63E+6 8.39E45 4.26E M Nb 95 2.13E+6 1.42 E 46 9.27E +4 No 99 1.49E-2 neg.a nog.a Tc 99m 1.43E-2 neg.a neg.a Ru 103 3.06E +2 I.04E+2 8.14 E-1 Ru 106 3.44E 42 3.06E+2 1.83E+2 Rh 106 3.44 E+2 3.06E+2 1.83E*2 Sb I 25 1.19E +6 I.14E+6 9.48E+5 Sb 127 8.21 E+3 1.39E-1 nog.a 7a 125m 5.68EG 5.52 E +3 4.62 E +3 To127 4.42 E+3 2.90E+3 5.04 E+2 To127m 4.36E +3 2.96E+3 5.14 E+2 To 129 5.48E+3 1.56E+3 5.36E+0 To 129m 8.42 E +3 2.40E+3 8.24 E+0 To 132 7.48E+2 1.72 E-3 neg.a 1129 7.15E40 7.15E +0 7.15E +0 1 131 1.22E+5 6.3 5E+3 neg.a ) i 1 32 7.70E+2 1.78E-3 neg.a Xe 133 7.29E+5 2.30E +2 n*0.8 Cs 134 7.90E+6 7.47E +6 5.80E+6 Cs 136 2.05E+5 8.13E+3 3.91E-3 Cs 137 2.02 E+7 2.01 E+7 1.97E+7 Ba 137a 1.91E+7 1.90E+7 1.87E+7 Ba 140 t.04EM 3.80E+2 neg.a La 140 1.19E+4 4.38E+2 neg.a Ca 141 1.32 E +4 3.61 E+0 1.03E-2 Co 144 2.64E+t 2.2 7E+t 1.16E+1 Pr 143 5.44E+0 2.41E-1 neg.a.- Pr 144 2.64E+t 2.27E+t 1.16E +t iti147 1.54E40 3.36E-2 neg.a Se 151 842 E-2 8.21E-2 8.16E-2 Eu 154 1.34E40 t.32E+0 t.2 5E+0 Eu 156 8J 6E-1 5.1 OE-2 nog.a No 239 5.59E-2 2.88E-3 2.88E-3 Pu 238 4.51 E-1 4.53E-l 4.54 E-1 Pu 239 8.89E-2 8.89E-2 8.89E-2 Pu 240 1.30E-1 1.30E-1 1.30E-l Pu 241 2.29E+1 2.27E+t 2.19E+1 Am 241 2.28E-1 2.94E-1 3.21 E-1 Ca 242 1.45E+0 1.12 E +0 3.50E-1 On 244 2.27E-1 2.25E-1 2.19E-1 neg. = less than 10~3 Curles. e

~. 4-11 i Table 4.4 Estimsted Releases of Radionuclides for Case 2 I in Which Only the Last Discharged Fuel Batch Suffers a Zirca10y Fire Tlee efter I.est Discharge isotope 30-ceys vo-eays 1 year (Redioectivity, Curles) Co 58 2 J 8E+3 1.26E +3 8.49E4 Qo 60 9.17E+3 8.68E+3 8.12 E+3 Kr 85 2.39E+5 2.36E+5 2 J 5E+5 Rb 86 1.01 E4 1.05E+3 3.84E-2 'Sr 89 1.79E+4 7.75E+3 1.78E4 Sr 90 .3.04E+3 3.82E+3 3.78E+3 Y 90 3.86E +3 3.84E+3 3.78E+3 Y 91 2.66E*4 1.30E44 4.99E4 Zr 95 1.62E4 8.37E+5 4.2 5E +4 le 95 2.11 E*6 1.41E+6 9J4EM Mo 99 1.49E-2 neg." neg.a Tc 99m 1.43E-2 neg.a p g,a Ru 103 3.06EC 1.04E4 8.14E-1 Ru 106 2.24E4 1.99E4 1.19EC Rh 106 2J4E4 1.99E4 1.19E4 Sb 125 4.17E+5 4.00E+5 3.31 E+5 Sb 127 8.21 E+3 1.39E-t nog.a To 125m 1.88E+3 1.38E+3 1.61 E+3 To127 4.28E+3 2.80E+3 4.86E4 To127e 4.20E+3 2.86E +5 4.96E +2 To129 5.48E +3 1.56E+3 5.36L40 To129m 8.42 E+3 2.40E+3 8 J4E+0 To 132 7.48E 4 1.73E-3 neg.a 1129 8.84E-1 8.86E-1 8.86E-1' i 1 31 1.22E+6 6.35E+3 neg.a i 1 32 7.70E+2 1.78E-3 neg.a Xe 133 7J 9E+5 2 JOE +2 neg. Cs 134 3.53E 4 3.34E46 2.59E+6 Cs 136 2.0SE+5 8.13E+3 3.91Ee3 Cs 137 2.83E*6 2.81E+6 2.77E +6 8e137e 2.67E*6 2.66E+6 2.62 E+6 Se 140 1.04E+5 3.80E+3 1.28E-3 La 140 1.19E+4 4.38E+2 neg.a Co 141 1.32 E+1 3.61E+0 1.03E-2 Co 144 I.91 E*1 1.65E+1 8.43E+0 Pr 143 5.44E+0 2.41E-1 neg.a -- Pr 144 . 1.91 E +1 ~ 1.65E*1 8.43 E+0 141 1 47 1.54E +0 3.36E-2 nog.a $m 151 9J1 E-3 9.30E-3 9.2 5E-3 Eu 154 2.89E-1 2.85E-1 2.69E-1 Eu 156 8.37E-1 5.82 E-2 neg.a ?$ 239 - SJ6E-2 neg.* neg.a Fu 238 6.73E-2 6.87E-2 7.18E-2 Pu 239 9J 8E-3 9.2 BE-3 9.2 BE-3 i Pu 240 1.55E-2 1.55E-2. 1.55E-2 l Pu 241 3.73E+0 3.70E*0 3.56E+0 Am 241 6.01E-3 7.00E-3 1.14E-2 On 242 1.31E40 1.01E*0 3.16E-1 On 244 5.88E-2 5.84E-2 5.68E-2 neg. = less then 10*3 Curles. a 1 2 ~l 1

a I I 4a12 Table 4.5 Estimated Releases of Radionuclides'.for Cases' 3 and 4 in Which Low-Temperature Cladding Failures-Q Occur-9 i-Isotope Case 3a. Case 4 _ d b (Radioactivity, Curies)= u A> 'Kr 85 6.65E+5 "2.39E+5< ' % ' ' qj I 1 129 3.58E+0 8.84E-14 ~'I 131 -neg.C 1.22E+6 1 132. .neg.C 7.70E+2 Xe 133-neg.C 7;19E+5 acase 3 assumes:

1.. last fuel discharge has decayed for 1_ year.-

2. no Zircaloy fire occurs. 3. 50%'of the fuel rods develop leaks. 4 100%i release of noble gasas 'and halogens from '\\. leaking fuel

• rods. -

A b ase 4 assumes: C s last fuel batch ' ischarged has.. decayed for -30, days.. - ) 1. d w 2. no Zircaloy fire occurs.. 3. 100% of fuel rods in last discharge devlop leaks.. nj a. 100% release of noble gases.and halogens from' ' N,

61., vl leaking fuel rods.

N;'/;l p m cneg = less than 10-3 curies. ~~ 5 1 1 4 p k& '\\ p

1..

( } i i [- \\t' q 1., w 1 r'

1 m 443 Table 4.6 Comparison of Radioactive Inventories Of Equili-brium Core with Spent Fuel Assemblies for Select- ~< ed Isotopes (Ginna) Ecullibrium Soent Fuel Poole (time af ter last disc. heros) i Isotope Core 30-devs ' vo-days i year (Total Radioactivity, Qarles) Co 58 3.57E+5 5.93E+4 3.26E +4 2.21 E+3 Co 60 3.20E +5 5.97E+5 5.84E+5 5.29E+5 Kr 85 3.73E+5 9.64E+5 9.74E+5 9.27E+5 Rb 86 6.53E+5 7.22E+3 7.48E +2 2.74E-2 Sr 89 3.55E+7 3.53E+6 1.53EM . 3.50E+4 Sr 90 2.95E+6 1.02E+7 1.01E+7 9.95E4 Y 90 3.15E +6 1.02E47 1.01 E*7 9.95E46 Y 91 4.57E+7 5.11 E M 2.48E+6 9.54I+4 2r 95 6.41 E+7 8.64f 4 4.46E4 2.27E+5 Nb 95 6.34E+7 1.12E+7 7.51 E+6 4.93Eg5 Mo 99 s 6.83E+7 7.03E+3 1.48E-3 neg.b 1 Tc 998 5.89E+7 6.77E+3 1.42E-3 neg. Ru 103 5.85E +7 7.80E+6 2.88E+6 2.09E44 Ru 106 1.95E+7 1.09E+7 9.71 E+6 5.78E+6 Rh 106 2.15E+7 1.09E+7 9.71E+6 5.78E+6 Sb 125 0.04E+5 7.11 E+5 6.32E6 5.65Eg5 Sht27 4.12E+6 4.33E+3 1.35E-2 neg. .T9 125m 1.27Ee 1.70E+5 1.65E+5 1.37E+5 Te127 4.05E+6 1.19E+5 7.79E44 1.36E+4 Te 127m 5.19E+5 1.17E+5 7.95E+4 1.38E+4 s To U.9 1.21E+7 1.38E+5 3.93E+4 1.35E+2 To129m 1.80E+6 2.12 E+5 6.03E+4 2.07Eg2 To 132 5.33E+7 1.83 E+4 4.23E-2 naj. I129 1.27E+0 5.32E+0 5.32E40 ' 5.32 Ego i 1 31 3.76E+7 6.00E+5 3.12E+3 neg.b I 1 32 5.42 E +7 I.89E44 4.36E-2 nog.b ) Xe 133 7.64E+7 3.52 E+5 1.11 E+2 neg. Cs 134 5.82 E+6 6.35E*6 6.00E4 4.66E+6 Cs 136 1.87E+6 1.26E+5 4.99E+3 2.40E-3 Ca 137 4.21 E+6 1.48E+7 1.47E+7 1.44E+7 8e 137m 4.00E+6 1.40E+7 1.39E+7 1.37E+7 Be140 6.55E+7 2.47E+6 9.07E+4 3.05E-2 La 140 6.74E+7 2.85E+6 1.04E+5 3.51E-2 ( Co 141 6.28E+7 6.34E+6 1.72 E+6 4.91 E+3 Co 144 4.24E+7 1.38E+7 1.19E +7 6.09E+6 Pr 143 5.71E+7 2.54E46 1.12E+5 B.86E-2 Pr 144 4.27E+7 1.38E+7 1.19E+7 6.09Eg6 Nd 147 2.48E +7 7.42E+5 1.62 E +4 neg. Em 151 1.42 E+4 5.14E44 5.13E44 5.10E +4 Eu 154 4.09EM 1.09E+6 1.07E+6 i.01 E+6 Eu 156 7.22 E+6 7.58E M 4.68E +4 1.66E-1 Np 239 7.81 E+8 3.02 EM 3.26E+3 3.26E+3 Pu 238 1.01E+5 4.46E+5 4.46E+5 4.46EM Pu 239 1.3 5E +4 5.25E44 5.25E+4 5.25E+4 Pu 240 2.01E44 8.60E+4 8.60E44 8.6!E+4 Pu 241 4.852 4 1.52E+7 1.51 E+7 1.46E +7 Am 241 4.99E +3 2.10E+5 2.14E +5 2.32 E+5 On 242 1.91 E4 9.33E+5 7.20E+5 2.2 5E +5 Ca 244 1.25E+5 3.59E+5 3.56EM 3.46E+5 85 pent fuel pool inventory includes discherges from 15 refuelings covar* Ing the period frenr April 1983 through the projected ref ueling of April 1987 bneg. = less than 10~3 Curles. s k ' _/.

4-14 Table 4 7 CRAC2 Results for Various Releases Corresponding to Postulated Spent Fuel Pool Accident; with Total Loss of Pool Water Whole Body Dose Interdiction Area Case Description (Man-rem) (sq. miles) 6 1A. Total inventory 2.6x10 224 30 days after discharge 50 mile radial zone

18. Total inventory 2.6x106 215 90 days after discharge 50 mile radial zone 1C.* Total inventory 7.1x107 224 30 days after discharge 500 mile radial zone 2.

Last fuer discharge 2.3x106 44 90 days after discharge 50 mile radial zone

• Note that the consequence calculations in NUREG'-1150 are based on a 50 mile radial zone. Case IC is given as a sensitivity result.

4 ,/

. y , y-y s 5-1 1 l 5. RISK PROFILE' ' The likelihood and consequences of various spent fuel pool accidents has bcen ' estimated in the previous sections. The risk is sumarized in Table 5.1. is preriously mentioned, the exposure results are tied to the health physics assumptions regarding decontamination and maximum allowable exposure. Thus'tha land interdiction. area is included in Table 5.1 as a more meaningful representation of severity. The uncertainty in each of these risk indices is estimated to be an order of magnitude in either direction and is due princi-pally to uncertainty in the fragility of the pools and uncertainty in the 'shismic hazard. i Note that the risk results are calculated for two surrogate plants and may not' be applicable to generic pool types. / 5.1 Failure Frecuency Estimates 5.1.1 Spent Fuel Pool Failure Probability The likelihood of the var ous postulated spent fuel pool accidents was developed in SecticL2 and sumarized "iy Table 2.9., The probability is simi-lar to the frequency of dominant core melt sequences for many PRA's. The major contributors are: 1. Cask drop accidents, 2. Seismic induced pool failure, 3. Loss of pool cooling, and 4 Pneumatic seal failure. Note that all of those potential accidents are plant specific and their frequency will vary widelyifnxn plant to p1~ ant. In particular, BWR's do not have pneumatic seals so their failure frequency is zero. )

5-2 5.1.2 Spent Fuel Failure Likelihood Previous investigations,2 of spent fuel behavior after a loss of pool 1 integrity accident focused on the conditions necessary to initiate cladding "fi res" after a spent fuel pool has drained. The present project has reevaluated these conditions using the SFUEL code 2 developed by SNL. The likelihood of such cladding fires has been assessed in Section 3. For a PWR with high density storage racks, the conditional probability of a clad fire was found to be 1.0 while for a BWR with low dtnsity storage racks the proba-bility of a clad fire was found to be 0.08. 5.2 Conclusions Regarding Risk ~ The overall risk due to beyond design basis accidents'in spent fuel pools for the PWR surrogate plant is about 130 person-rem /Ry and about 12 person-rem /Ry for the BWR surrogate. These estimates are comparable to present esti-3 mates for dominant core melt accidents and appear to warrant further atten-tion on this basis alone. However, the unique character of such an accident ) (substantial releases of long lived isotopes) makes it difficult to compare to reactor core melt accidents. The exposure calculations are driven by assump-tions in the CRAC modeling and the results are not sensitive to the severity of the accident. In terms of interdiction area this type of accident has the potential to be much worse than a reactor core melt accident. The uncertainty in risk in terms of person-rem /Ry is driven principally by the uncertainty in the likelihood of complete draining of the spent fuel pool which is estimated to be at least an order of magnitude in either direc-tion. 5.3 References for Section 5 1. A.S. Benjamin, D.J. McCl oskey, D.A. Powers, S. A. Dupree, '" Spent Fuel Heat-up Following Loss of Water Duri19 Storage," NUREG/CR-0649, March 1979.

l 5-3 ' 2. ' N. A. Pi sano, F. Best, A.S. Benjamin, K.T. Stalker, "The Potent'ial for Propagation of a Self-Sustaining Zirconium Oxidation following loss of Water in -a Spent Fuel Storage' Pool," Draft Report, January _1984 3. " Evaluation of Severe Accident Risks and the Potential for Risk Reduc-tion," NUREG-1150 (To be published). t 4. e O Om e

5-4 Table 5.1 Estimated Risk for the Two Surrogate Sper<t Fuel Pools from the Two Dominant Contributors 3 pent Fuel Interdiction Accident Pool Fire Health Risk Ri sk Initiator Probability /Ry (Man-rem /Ry) (Sq. Mi./Ry) Seismic induced PWR pool failure 1.6x10-5 37 8.4x10 " Seismic induced BWR pool failure 1.8x10-6 4 7.6x10-5 Cask drop

• induced PWR pool failure 3.1x10-5 71

.001 Cask drop

• induced BWR pool failure 2.5x10-8 6

1.1x10-4

• After removal of accumulated iny'entory resumes.

(Note that many new plants have pool configurations and administrative procedures which would preclude this failure mode.) O ~~ ~

-l 6-1 1 CONSIDERATION'0F RISK REDUCTION MEASURES 6. Due to diversity in the nature of initiating ev'ents for beyond design basis accidents in spent fuel pools,. there appear to be several possible ways to reduce the risks. It must be emphasized that each of the contributors to risk are plant specific and one or more of the risk significant sequences identified in Section 5 may not be important-at other plant sites. The following sections discuss the advantage and disadvantages of a number of proposed risk reduction strategies. A cost benefit analysis has. not been l perfomed but the estimated risk appears to be large enough to justify further investigation of risk reduction measures. 6.1 Risk Prevention 1. Reduction of Stored Radioactive Inventory - Most of the consequences of a release of radioactivity from a catastrophic pool accident is associated with the large inventory of isotopes of intermediate half-lives, e.g., Cs-137, Sr-90. The potential release increases approxi-mately in proportion to the number of fuel assemblies in the storage inventory. One obvious measure for risk reduction is to transfer part of the inventory to altePnitive 5torage locations (e.g., see Ref. 1). 2. Air Circulation - The one universal prevention measGFe is.to promote air cooling in the event of loss of water cooling of the spent fuel. The new high density fuel storage racks restrict air flow and make - I even old spect fuel (one to two years) susceptible to heat-up and self-sustaining oxidation. The older style fuel baskets with large inlet holes (3 inch diameter or more per assembly) allow much freer air circulation. If all recently discharged fuel (less than two years) is kept in low density fuel baskets and they are separated from the wall and the older fuel by a one foot gap the'n the likeli-hood of self-sustaining oxidation would be reduced by a factor of 5 or more compared to the high density storage configuration. O _. _ _ _ _ _. _ _ _ _ _ _. - _ _ _. _ - - - _ _ - -. _ - -. - - - - _ - - - - _. - - _ - _ _ _ _ - _ - - - - - - _ _ - _ _. _ ~ - __-a

'l 6-2 3. Additional Cooling' Systems - Although loss-of-pool cooling appears to be' risk significant, an additional. cooling system is unlikely to be cost beneficial (unless the cooling system was 'substantially more unreliable than the two surrogate systems).- An additional. cooling: system would not affect the risk from pool' failure events (seismic or-cask drop accidents). Thus the not risk reduction would.be minimal unless loss-of-cooling were the dominant event. 4. Improved Prncedures and Equipment - The likelihood of cask drop acci-dents can be reduced by improving procedures, administrative controls and/or installing more reliable equipment. However, none of these improvements would reduce the -risk from the.other ' dominant 'se-quences. Thus the net risk reduction would be difficult ~ to quantify on a ' plant specific basis. It would appear to be usefulto conduct a complete risk evaluation before spent fuel shipment is begun at each site. A key piece of such an evaluation would.be a structural analy-sis of the pool respr.ase to the loading from a dropped cask. 6.2 Accident Mitigation l. Post-Accident Spray - Water s' pray has the potential to terminate the progression of a spent fuel pool accident whether or not the pool. is , intact. However, large quantities of water must be available (it would be necessary to continue spraying until the' pool could be repaired and reflooded) 'and the equipment would have to be seismical-ly qualified to a higher g' level than the pool' structure (in order for the sprays to have a high likelihood of ' surviving). Some pools may have fire sprays available in the spent. fuel pool l building. For those plants without sprays available, it Fams unlikely that the ex-pense of a new safety grade spray system could be justified consider-ing the large uncertainty in the risk. Temporary fire' hoses were suggested by Benjamin et al.,2 but. the radiation levels would make such ad hoc measures extremel'y difficult. Furthermore, if the spray is not initiated before the rods reach 900 C or there is insufficient flow, the water may aggravate the reaction by providing additional j oxidation. (The steam /Zircaloy reaction is also highly exothermic.) 1

6-3 2. Filtering For thors plants with a standby gas treatment system available, operation of the system has the potential to substantial".y reduce the fission product release from the building. However, the high ' temperatures and large aerosol production rate would tend to rapidly degrade the effectiveness of the system. The performance of such a filtering system would be difficult to characterize under fuel pool accident conditions. It is unlikely to be cost effective to install a new system large enough to handle the worst case spent fuel pool accident scenarios. 6.3 Conclusions Regarding Preventive and Mitigative Measures For those plants which have a similar spent fuel pool risk potential to the two surrogate plants, the one preventive measure which appears to have a substantial effect on risk (t risk reduction of 5 or more) is to maintain recently discharged fuel in low density storage racks that are isolated from the rest of the fuel racks by a foot or more of space (to provide free air circulation). However, there may be plant specific features which make a sub-stantial difference in the order of the dominant contributors to risk. There-fore plant specific risk evaluations should be performed before any changes are implemented at a given plant. 6.4 References for Section 6 1. D.D. Orvis, C. Johnson, and R. Jones, " Review of Proposed Dry-Storage Con-cepts Using Probabilistic Risk Assessment," prepared for the Electric Power Research Institute by the NUS Corporation EPRI NP-3365, February 1984

2.. A.S. Benjamin, D.J. McCloskey, D.A. Powers, S.A. Dupree. " Spent Fuel Heat-up Following loss of Water During Storage," NUREG/CR-0649, March 1979.

I

f G 5 i e 4 I O \\. Oh e ?. us e m a =9M6

A-1 APPENDIX A RADIOACTIVE INVENTORIES A.1 INTRODUCTION Two older-vintage plants, a BWR and a PWR, were selected to serve as sur-rogates for estimating the risks associated with "Beyond.. Design Basis Acci-dents in Spent Fuel Pools." The purpose of this' appendix is to describe the methods used to simulate the operating history of the two plants and. to sum-marize the calculated radioactive inventories contained in the fuel assemblies stored in the spent fuel basins. The surrogate plants were Millstone-l'- (BWR) and Ginna (PWR). A.2 SIMULATION OF OPERATING HISTORIES A.2.1 Thermal Energy Production vs Time The operating history of each surrogate plant was reconstructed from sev-- eral sources. The early history, prior to December 1,1975 was reconstructed from monthly summaries contained in Refs.1-3. Data for the period December 1,1975 through April 30, 1986 were taken from Ref. 4 Data from May 1, 1986 to April 1,1987 were extrapolated, based on recent average capacity factors and scheduled shutdowns. ~~ During each operating cycle (the period between successive refuelings), the average thermal power was calculated from the total thermal energy pro-duced during the cycle. No attempt was made to model variations in power lev-els during an operating period. (Fluctuations in. the monthly energy produc-tion are illustrated in Fig. A.I.). A.2.2 Fuel Burnup Calculations The number of fuel assemblies discharged at each refueling and their spe-cific burnup was obtained from a data base maintained by R.A. Libby of Pacific Northwest Laboratories (PNL) for the U.S. Department of Energy.5 It should be l -_-- -----~ ~

I_ A-2 noted that the inventory of spent fuel assemblies stored in the spent fuel basins at various points in time listed in the Libby data base differ from the data listed in Ref. 4 It is apparent from the operating histories that the data in the earlier volumes of Ref. 4 are less accurate. In general, the burnups listed in the Libby data base differ by a few l percent from the burnups calculated by the methods de cribed in the following i paragraphs. These discrepancies do not have significant effects on the over-all inventories of radionuclides, but only on the distribution of the inven-tories among the older fuel batches. In order to model the burnup of the various discharged batches of spent fuel, the following method was used. It was assumed that all fuel assemblies in the core during a given operating cycle provided the average specific powe-r, i.e., (MW /MT)$ = (W /D (MT) core ' th )$ D th ) where for operating cycle,1 MWth/MT is the average specific power per met-ric tonne of initial heavy metal, (Wth )j is the total thermal energy D produced in Dj days of the cycle, i, and MTeore is the metric tonnes of initial heavy metal in the core. The average specific burnup for each fuel batch, j, at discharge was cal-culated from the formula, (WthD/MT)) = (W /MT)$ $, 0 th where is the summation over the several operating cycles,1, that the fuel was in the reactor. (As noted below, ORIGEN2 also calculates the specific burnup which provides a check on internal consisten'cy of the data). The total burnup in the discharged fuel plus the burnup of assemblies re-maining in the core at the time of the April 1,1987 refueling equaled the

y A-3 total thermal energy production over the preceding history of the plant (e.g., I see Table A.4). l A.2.3 Calculation of Radioactive Inventories The average radionuclides content in each metric tonne of discharged fuel was calculated using the ORIGEN2 Computer Code.6 The code' treats the reactor core as a homogeneous body operating at an ' average specific power.. Account is taken of radionuclides decay during and following ' irradiation, decay chains, and successive neutron captures. The BNL version.of ORIGEN2 was benchmarked against the version in use at Oak Ridge National Laboratory by calculating an identical case, which yielded identical results.7 The results obtained from an ORIGEN2 calculation are slightly sensitive to the size of the time steps used in the irradiation calculation. Several preliminary calculations were made to select an appropriate set of time steps for which the sensitivity was negligible. (Shorter time steps give higher precision results, but at the expense of increased computer time. The crite-rion adopted was that the time-step sensitivity be less than 0.1% in the cal-- culated concentration of several key nuclides.) In a menry operating nuclear power plant ' fuel management strategies are complicated (e.g., see Ref. 8). Most fuel assemblies remain in the core for several operating cycles and are often shifted in location during refueling so as to optimize-burnup. Also, U-235 enrichment is varied. ORIGEN2 as used.at BNL did not take account of such detail, nor of the axial and radial distribu. tion of the power density. Thus, the radioactivity calculated for a particu-lar assembly would not correspond exactly to an actual assembly..Neverthe-less, the total calculated radioactivity in a discharged batch should be iden-tical to total in a real batch (in so far as the sirecision of ORIGEN2 allows). The calculations.do take account of the irradiation times in each operat-ing cycle and the decay that occurs during shutdowns for refueling or - pro-longed shutdowns for maintenance and repair.

r 'A As used at BNL,' the input for each irradiation cycle is the average spe-i cific power-and the length of the cycle. ORIGEN2 calculates the total average burnup of. each fuel batch over the irradiation cycles during.which it was in. the core..This calculated burnup was cross checked against " hand" calcula-tions for each batch, the " hand" calculations being based on the operating history (see Section A.2.2). The input for ORIGEN2 requires the ' specification of the elements con-tained in the fuel including trace impurities, the U-235 enrichment and the composition and amount of alloys used in the' fuel cladding and assembly hard-ware. For each plant, BWR and PWR, only a single. fuel 'and assembly composi-tion was modeled which is typical of fuel of recent vintage for the. respective reactors. Data for the fuel models were taken from Reference 9. The output of ORIGEN2 includes isotopic concentrations (of stable as well as radioactive isotopes), activity of radionuclides, ard thermal power prode-tion.of each radionuclides. These are given at specified decay times for acti-vation products (in cladding, hardware and trace elements in the fuel pel- ') lets), fission products and actinides. 1 The BNL calculations were made for each fuel batch from the date of the end of irradiation to the projected dates of May 1,1987, July 1,1987, Octo-ber 1,1987 and April 1,1988. A.3 DATA FOR MILLSTONE 1 A.3.1 Reactor and Fuel Cycle Parameters I Table A.1 summarizes several of the major reactor characteristics and fuel cycle parameters for Millstone 1. A,3.2 History of Operations-Several milestones in the operation of Millstone-1 are summarized in Table A.2. Monthly gross thermal energy production from 1976 through 1984 is . plotted in Fig. A.1. During the first 10 years of operation.the. plant ~

A-5 experienced two prolonged outages, i.e. Sept. 1972 to March.1973 (198 days) and October 1980 to June 1981 (254 days). Otherwise the refueling / maintenance outages have ranged from 35 to 76 days in duration avertging about 57 days. A more detailed narrative of the plant operating history from 1970 through 1981 appears in Ref.10, f.ppendix F, pp. F-31 through F-70. The only unusual experience with fuel cladding failures that has been noted occurred in 1974 when some 25 assemblies were found to have leaking fuel elements which forced a temporary power derating to stay within off-gas release limits.ll Since mid-1981, the plant has operated with nearly 100% unit service factor except for scheduled refueling outages. There have been 10 refueling campcigns since beginning of conmercial op-erations on March 1, 1971 (see Table A.3)". The next scheduled refueling will .be about April 1987 During the first 10 years, refueling occurr'ed at some-what irregular intervals, being dictated by unscheduled forced outages. Since 1981, refueling has been scheduled for approximately 18 month inttrvals, oc-curring in April or September. During the lifetime of the plant the average fuel burnup has generally increased from about 20,000 MWD /MT in 1972 to about 28,000 MWD /MT at present. A.3.3 BWR Fuel Assembly Model Used in ORIGEN2 Calculations A nominal BWR fuel element has been modeled, based' on dita presented in Ref. 9. This is an 8x8 element assembly of 2.75% U-235 enrichment, containing 1.5873 kg of gadolinium burnable poiron per metric tonne of uranium. The fuel cladding is Zircaloy-2. Other alloys c eent in the fuel assembly hardare include Zircaloy-4, Inconel X-750, SS302 and $S304. The alloy contents of the assembly hardware are included with weighting factors to take account of the axial variatio, of neutron flux which results in lower neutron activation at the ends of the assemblies. In addition to the fuel, the cladding and the as-sembly hardware, an allowance was made for the presence of " crud" composed of Fe, Co, and Hi on the outer surfaces of the cladding and assembly hardware.

A-6 A.3.4 Calculated Radioactive Inventories [ The calculated inventories of selected radionuclides

• are listed in Table A.5 for the reactor c' ore at the end of operating cycle number 11 projected to be on April 1,1987.

Also listed are the inventories in the spent fuel basin on May 1 and July 1,1987 and April 1,1988 assuming that 167 assemblies will be discharged in the April.1987 refueling. It should be noted that many of the isotopes that are of considerable im-portance in a core melt accident are those of short half-lives which are no longer present in the spent fuel after a few days of decay, e.g., Rb-91, Rb-93, Sr-93, Sr-95, Y-94, Y-95, Tc-104, I-134,1-135, I-136, Cs-138, Cs-140. On t the other hand, the spent fuel inventory contains much larger quantities of several long-lived isotopes than does the equilibrium co' re. N"oteworthy among these are H-3, C-14, Sr-90 (Y-90), I-129, Cs-137, Ba-137m, Eu-154, Pu-239, Pu-240, Pu-241, Am-241, and Cm-244. Table A.6 gives a comparison of the radionuclides inventories in the last ) fuel batch to be discharged with the summation of the. inventories contained in the ten batches discharged in the period from 1972 through 1985 A.3.5 Decay Heat Table A.7 sumarizes the decay thermal production in t7e various dis-charged batches. The data shown is for the whole batch, i.e., the specific thermal power (kilowatts per metric tonne) multiplied by the metric tonnes in the batch. Table A.8 summarizes the fraction of the decay heat contributed by vari-ous isotopes, The main contributors change with decay time, e.g., in the old-est fuel' (batches 1, 2, etc.) the largest contributors are Y-90 and Ba-137m, whereas the last discharged batch 11 is dominated by Cs-134, Rh-106, and Pr-144. The actinides are relatively small contributors.

• The selection of radionuclides was based on several criteria including poten-tial for biolugical concern, thermal power, and total curies of activity.

,mm... ~s ~*

.o Memorandum To: K. Perkins. Page 15 March 27, 1985 REFEf.ENCES (CONT'D) 120. Charlot, L. A. 'and Johnson, A. B., Jr., " Reaction of Zircaloy-2 with Water Vapor at 900 to 1200*C," on BNWL 120, July 15,1965. 121. Bostrom, W. A., "The 'High Temperature Oxidation of Zircaloy in Water," WAPD-104 Westinghouse El'ectric Corporation. March 1954 122. Pensler, J. P., " Studies of the Oxygen Gradients in oxidizing Metals, III i. Kinetics of the Oxidation of Zirconium at High Temperatures," J. Electroches. Soc., vol.112, pp. 447-484. 1965. 123. Wallwork, C. R., Rosa, C. J., and Smeltzer, W. W., " Breakaway Phenomena in the oxidation of Zirconium at 850 and 950*C," Corrosion Sci., 'vol. 5, pp. 113-119. 1965. 124. White, J. H., reported in AEC Fuels and Materials Dev'elopment ' Program, Progress Report No. 67, CEMP-67, General

• Electric Co.,-151, 1967.

125. Leistikov, S., et al., "S tudy on High Temperature Steam.0xidation L of Zircaloy-4 Cladding Tubes," Nuclear Safety Projecc Second Semiannual Report, 1975, KfK-2262, Karlsruhe,, 233, 1976. 126. Kendall, L. F., " Reaction Kinetics of Zirconium and Zircaloy-2 in Dry Air at Elevated Temperatures," Hanford Atomic Products Operation, Washington, i Contract No. W-31-109-Eng.-52, September 1955. 1 127. Urquhart, A. W. and Verai'. yea, D. A., " Characterization of.Zircaloy 0xidation Films," Zirconium in Nuclear Applications, ASTM STP 551, American Society for Testing and Materials,1974, pp. 463-478. 128. Sabol, C. P., Mcdonald, S. C. and Airey, C. P., " Microstructure of the - 0xide Films Formed on Zirconium-Based Alloys," Zirconium in Nuclear Applications, ASTM 'STP 551,; American Society for Testing and Materials, 1974, pp. 435-448. 2 129. H111ner, E., " Corrosion of Zirconium-Base Alloys-An overview," Zirconium in the Nuclear Industry, ASTM STP 633, A. L. Lowe, Jr. and C. W.-Pnrry, Eds., American Society for Testing and Materials, 1977, pp. 211-235. 130. Urbanic, V. F., " Oxidation of Zirconium Alloys in Steam at 1000 to y 1850*C," Zirconium in the Nuclear Industry, ASTM STP 633, A. L. Lowe, Jr. and C. W. Parry, Eds., American Society for Testing and Materials, 1977, pp. 168-181. ] l e t 4

a s Memorandum '"o: K. Perkins March 27, 1985 Page 16 REFERENCES'(CONT'D) 131. Dobson, W. G., _ Biederman, R. R., and Ballinger,.R. G., '"Zirealoy Oxidation in Steam Under' Transient Oxidizing Conditions," Zirconium in the Nuclear Industry,. ASTM STP - 633. A' ; !- Love, Jr. and G. W. Parry, Eds., American Society for Testing and Materials..1977, pp. 150-167. 132. Sawatzky, A., Ledoux, .G. A., and Jones, S., ~ " Oxidation of Zirconium During a High-Teagarature Transient," Zirconium in the Nuclear Industry, ASTM STP 633, A. L. Lowe, Jr. and G. W. Parry, Eds., American Society for Testing and Materials, 1977, pp. 134-149. 133. Pawel, R.- E., .Perkins, R. A.,

McKee, R.

A.,

Cathcart, J.;

V., Yurek, G. J., and Druschel, R. - E., " Diffusion.of Oxygen in Beta-Zircalov. and the High Temperature Ziresloy-Steam Reaction," Zirconium in the Nuclear Industry, ASTM STP 633,. A. L. Lowe, LJr. and G. W. Parry. Eds., American' Society for Testing'and Macarials,1977, - pp.119-133. 134. Hofmann, P. and Pol' itis, C., " Chemical Interaction Beeveen Uranium oxide and Zircaloy-4 in the Temperature Range Between 900 and 1500'C," Zirconium in the. Nuclear Industry (Fourth Conference), ASTM STP 681, American Society for Testing and Materials,1979, pp. 537-560. 3' J 135. Ocken, H., Biederman, R. R.,

Hann,

.C. R., and Uesterman, R. E., " Evaluation Models of Zircaloy Oxidation in Light of Recent Experiments," Zirconium in the Nuclear Industry (Fourth Conference). ASTM STP 681, American Society for Testing and Materials,1979, pp. 514-536. 136. Pawel, R. E. and Campbell, J. J., "A Comparison of the High Temperature oxidation Behavior of Zircaloy-4 and Pure Zirconium," Zirconium in tha Nuclear Industry (Fif th Conference), ASTM STP 754, D. G. Franklin, Ed., American Society for Testing and Materials,1982, pp. 370-389 137. Hofmann, P., Kerwin-Peck, D., and Nikolopoulos, P., " Physical and Chemical Phenomena Associated with the Dissolution of Solid UO 2 by Holten Zircaloy-4," Zirconium in the Nuclear Industry (Sixth Conference), ASTM STP 824, D.. G. Franklin and R. B. Adamson, Eds., American Society for Testing and Materials,1984, pp. 810-834 k38. Chung, H. M. and Thomas, G. R.,"Righ-Temperature oxidation of Zircaloy in Hydrogen-Steam Mixtures," r Zirconium in the Nuclear Industry (Sixth Confarenee), ASTM STP 824, D. G. Franklin and R. B.. Adamson, Eds., American Society for Testing and Materials,1984, pp. 793-809. 139. Leistikow, S., " Comparison of High-Temperature Steam oxidation Kinetics Under LWR Accident Conditions: Zircaloy-4 Versus. Austenitia Stainless , Steel No. 1.4970," Zirconium in the Nuclear. Industry (Sixth Conference), t ASTM STP 824, D. G. Franklin and R. B. Adamson, Eds., American Society for Testing and Materials,1984, pp. 763-779. e em .N ee-4+w

l e Memorandum To: K. Perkins Page 17 March 27, 1985 REFERENCES (CONT'D) 140. Turek, C. J., Cathcart,.J. V., and Pavel, R. E., " Microstructure of the Scales Formed on Zircaloy-4 in Steam at Elevated Temperatures," 0xidation of Metals _1,,0,, 255 (1976). 141. Pawel, R. E.,

Cathcart, J.

V., and McKee, R. A.,

• • ihe Kinetics of Oxidation of Zircaloy-4 in Steam at High Temperatures," J. Electrochem.

Soc._126, 1105 (1979). 142. Pavel, R. E., Cathcart, J. V. and Campbell, J. J., "The Oxidation of Zircaloy-4 at 900 and 1100'C in High Pressure Steam," J. Nucl. Hacer. 82 129 (1979). 143. Pavel, R. E., " Oxygen Diffusion in the Oxide and Alpha Phases During Reaction of Zircaloy-4 with Steam from 1000*C to 1500*C," J. Electrochem. Soc. I_26,, Illi (1979). 2 144. Pawel, R. E., Cathcart, J. V., and McKee, R. A., " Anomalous, 0xido Crowth During Transient-Temperature oxidation of Zircaloy-L," 0xidation of Metals 14,, 1 (1980). 145. P2 vel, R. E. and Campbell, J. J., "The Observation of Effects of Finite Specimen Geometry on the Oxidation Kinetics of Electrochem. Soc. E, 2188 (1980). Zircaloy-4," J. 146. Pawel, R. E. and Campbell, J. J., "The Oxidation of Pure Zirconium in Seemm from 1000* to 1416*C," J. Electrochem. Soc. 128, 1999 (1981). 147. Pavel, R. E. and Campbell, J. J., "The Effect of StructGral Changes in the Oxide on the oxidation Kinetics of Zir*: onium Corrosion," HACE-6,1983, p.162. in High Temperature 148. Dravnieks, A., Temperatures," J. Am. Chem. Soc."The Kinetics of the Zirconium-Nitrogen Reaction a 7,2,(1950) 3568 149. Rosa, C. J., Smeltzer, W. W., "The Nitriding Kinetics of Zirconium in the Temperature Range 750-1000*C," Electrochem. Technol.,4,(1966) 149. 150. Mallett, M. W.,

Baroody, E.

M.,

Nelson, H.

R., and Papp, C. A., "The T Surface Reaction of Nitrogen with Beta Zirconium and Nitrogen in the Metal," Report BMI-709 (rev.), Dec. the Diffusion of 12, 1951. 151. Evans, E. B., Tsangarskis, N.,

Probst, H.

B., and Caribotti, N. J., " Critical Role of Nitrogen During High Temperature Scaling of Zirconium " 1972, p. 248.in High Temperature Cas-Metal Reactions in Mixed Environments, TMS

d +8 i 1 1 1 1 1 -, 'i +6 +4 ~ 4 +2 : /.20 x 10 exp (-29077/RT) 6 1 1 o 0 o .E 'I } I 5.76 x'10 exp (-52990lRT) 8 -2 A o N cn 3 - MONO-TETRAGONAL PHASE CHANGE OF Zr0 3 2 3 E 4 1.15 x 10 exp (-27340lRT) 5 cr-B PHASE CHANGE 0F Zr - 0 SOLID 2 48 SOLUTIONS -10 o LElSTlK0W (1975) 0 WHITE (1967)~

.12

> HAYES AND ROBER50N (1945) I i ~1 1 I I t t 5 6 7 8 9 10 11 - 12 13 hM G lt54 g 97' L4 10 IT ( K) FIGURE 1 CORRELATIONS FOR ZIRCONIUM OXIDATION DI AIR (FROM REF. 1) 4Zy-4

e

• p i

TEMP. (*C) e,00 tS001400 1300 #200 8000 4000 to.s i, # } i ,, l, i

• \\

S ~, S g a ALPHA L AYER, 2 \\

• • • /*

2 ~ \\ N t0-a 4 \\ \\ e0 ~ \\ \\ y 5 \\ \\ \\ S 5 ~ \\ \\ \\ ~ { \\.g \\* \\ ~ v g ,\\ 2 eM g \\ \\ \\ E go-7 \\ u \\ \\

• 0' '

\\ Q \\ \\ ~ \\ \\ \\ q m S N N \\ 5 g N \\ \\' p \\ 4 g % la. 2 OxiCATION OF Ze (~) \\ ANO 2r 41---I tN STE AM \\ , o. s N 0* Oxs0E Lav!R, ,cm /s I 5 S e OxvGEN CONSUMED, ,lg/cm ) I/s 3 ~ 2 2 ~ 40 8 I 40* S.O SS so 6.5 7.0 7.S 3.0

• t000/T!*K)

FIGURE 2 PARABOLIC RATE CONSTANTS FOR OX1DE LAYER GROWTH, Y a-LAYER GROWTH, AND OXYGEN CONSUMPT1,0N FOR THE REACTION OF ZIRCONIUM (SOL'O LINES) AND ZIRC (DASHED LINES)WITH STEAM.HE RATE CONSTANTS FOR i OXYGEN CONSUMED (WEIGHT GAIN) WERE DETER MODELING ANALYSES (FROM REF. 136), g 8

j ~ q l I ~ BROOKHAVEN NATIONAL LABORATORY MEMORANDUM DATE: January 27, 1986 TO: Kenneth Perkins N M: John Weeka .SusJECT: Parabo11'e Fate Constanta for Oxidation of Zirealoy-4 in Dry Air On March 27, 1985, Lou Teutonico (1) provided you vich an assessment of the knowledge of oxidatica kinetics of zircaloy-4 in air and r. team based on the literature available at that time, as part of our overall assessment of the Sandia report, NUREC/CR/0649, under our NRC contract FIN A-3786. Among j i other things, Lou.showed significant differneces in the oxidation kinetics of zirconium metal and zircaloy-4 in steam, as evidenced by the work of Pawel and Campbell of Oak Ridge National Laboratory (2). Except'for this work, there were little available data at temperatures above_1100*C, where rapid reaction l rates ar's expected, exeept for the 1967 data by White on unallpyed zirconium (3) which show nignifican'ely higher rates than would be expected by extrap-olation of results obtained at lower temperatures. 1 4" *, l When work was initiated on this program, NRC offered to obtain for us some more recent unpublished German data. These were never received through that source. After Teuconico left Brookhaven, I attempted to reevaluate the ~ work he hed done in the context of comparing the Pawel and~Campbellidata with the Sandia curve. It is immediately apparent that the Pawel and Campbell parabolic. rate constants.are considerably lower than the curve used by g ~.__.-_L._c._L..___ _..L.... . ~....

P v Sandia. Figure I shows this comparison. Subsequently, while at the International Conference on Environmental Degradation of Nuclear Materials in September, I discussed the subject further with Dr. Hee Chung of"Argonne National Laboratory and Dr. Friedrich Carzarolli of KWU, through whom I requested the unpublished German, data. Chung pointed out that, while the ratc controlling step in the high temperature oxidation of zirconium or zircaloys is the diffusion of oxygen through the oxide and/or through the solid solution of oxygen in zircaloy that underlies it in both steam and air oxidation, there is a significant decrease in the oxidation rate observed in a steam environment due to an effect of the hydrogen produced during this oxidation on these diffusion constants. He pointed out that, while this effect has been observed by several workers, it is not suffic'lently quant'ified to permit us to use, high temperature steam data (such as some of his own, those of Prater and Courtright at PNL (4) and those of Pawei nd Campbell at Oak Ridge) to estimate oxidation rates under our fuel pool accident scenario. This leaves us, therefore, with only the White data in the high temperature range. Carzarolli advised me that most of the German data were generated by Siegfried Leistikov at Karlsruhe. Following the conference, I wrote both to Carzarolli and Leistikow, and from both sources received copies of Leistikov's more recent data. In particular, Leistikow sent me, not only unpublished curves in air and steam for oxidation kinetics of zircaloy-4, but several internal reports, in German, that contain the results of a few short-term r experiments above 1100*C. Appendix I to this memorandum gives the cover ) letter from Leistikow and his recent unpublished data. You vill note from the letter that data at temperatures above 1100*C may be svailable in a year or k

l '. 'l so. I also receivsd the report KFK 2587 dated Mar 3 19, 1978, which shows some high temperature oxidation rate measurements on zircaloy-4 in air, oxygen, and steam. I've included this figure as Appendix II. The few data available above 1100'C show that the oxidation rate _s of zircaloy-4 are much greater in air than they ere in either. oxygen or steam. Leistikow's new data show roughly a parabolic enrrosion rate behavior (slope of 1/2 on the log log plot) fer the first 30-60 minutes in both air and steam. They also show that the differenca between the air and the steam rates increases with temperature. Af ter 30-60 minutes, however, the rate at all but the highest temperature increated dramatically, especially in air. This may be due either to difficulty in contro11 tog the temperature of the highly exothermic zirconium / air oxidation, or to some " breakaway" type phenomenon in the surface oxides exposing the bare metal underneath. 1,eistikow drew his curves to suggest a new leveling off, at least at 950 and 1000*C af ter long times (t > 90 min.). At l'ower teciperatures, zirconium and zircaloy are know to oxidize according to the cubic law, which would mean a slope of 1/3 on a log log plot. The high temperature data used by Sandia were all approximated using

.7 ~

parabolic growth, which is more typical of diffusion controlled phenomenan such as are believed to occur at high temperatures. The new German data show a slope somewhere between 1/3 and 1/2 for the first 30 minutes or so. In an' attempt to compare these data with the Sandia curve, I drew lines with a slope of 1/2 through the data for the first 30 minutes in air, and from them calculated a parabolic rate constant which I have compared with the Sandia curves and the Pawel and Campbell data in steam in figure 1. I also used the

same approach on the much (approximately 10 x) higher long-term oxidation 950,1000, and 1100*C, on several of his curves in the first 60 races at minutes or so obtained in steam, and on the short-term data in Appendix II at 1160*C. b These rate constants.are also shown in figure 1. It is apparent, ' therefore, that the German starte data and those of Pawel and Campbell for zirealoy-4 in steam coitsistent in the temperture range in which they overlap. The new German air data are consistant with some of their own work (at short exposure times) published some years earlier (5). From the new Gernttn data, I suggest the rate equation * ) = 3.09 x 100 56 00) (g) exp (- where W is in mg 02 reacted per square cm, e is in seconds, and T is in *K. The instantaneous rate, g at time t and temperature T is given by de h= 3.09 x 108,,, g,56 00) The Sandia curve shows an abrupt increase in oxidation ra g at 10/T = 7, which they attribute to the mono-cetragonal phase change of Zr0. As can be 2 c seen in figure 1, the Pawel and Campbell data do not show such an abrupt change at this temperature; howe r, they were obtained in steam. The recent results of Fther and Courtright (4) (whf ch were presented at the 1985 Syg osium on Zirconium) show that for reactions in steam they find a similar jump at temperatures as high as 1500*C (1/T is 5.'5 x 10-4). This may be due to effects of the hydrogen produced by steam reaction on the oxide structurd on the circaloy. Unfortunately, Prater and Courtright plotted their data in . ~.. - - -

^' terms of thickness of the Zr02 film, and thus these could not readily be transferred to figure I which is in we. of 02 reacted. Since a considerable amount of the oxygen that reacts either from air or steam exists,in high concentration solid solutions in the zircaloy, and since we are concerned in our accident scenario with the heat generated by this reaction, I think it is important that we consider the total oxygen consumed rather than just the thickness of the layer. I would antici'pate the free energy of formation per gram atom of oxygen reacted be approximately th'e same for the zirconium oxygen solid solution as for Zr02 at these high temperatures. I have included Prater and Courtright's figure as Appendix III. Cohelusions Based on the information available to date it appears impossible to justify any major changes to the Sandia equation; in particular, the curve from the work of' White at temperatures above 1150'C appears to be all we have. However, this was obtained on unalloyed Zr, not zircaloy, and the higher rates for zirealoy-4 over those for unalloyed Zr observe,d by Pavel and Campbell in steam may also exist in air. For temperatures from 800-1150*C, I think the new German data fit in we11 with what was previously observed, and suggest using equation 1 given above. However, if the exposure is for periods ~ greater than 30 minutes, this curve may not be' conservative, as shown in the new German data plotted in figure 1. cc: W.Y. Kato W.T. Pratt V.L. Sailor L.J. Teutonico

-)1 1 i i 1. L.J. Teutonice, Memo to K. Perkins regarding FIN A-3786, March 27,1985. 2. R.E. Pawel and J.J. Campbell, Zirconium in the Nuclear Industry, ASIM STP 754, pg. 370, 1982. 3. J.H. White, GEMP-67, pg.151,1967. 4. J.T. Prater and E.L. Courtright, oxidation of Zircaloy-4 in Steam at 1300 to 2400*C, Presented at Seventh International Conference on Zirconium in the Nuclear Industry, June 24-27, 1985. 5. S. Leistikow et al., KFK-2262, pg. 233, 1976. j l l \\/i i j i

• l l

e e e .wp

r .l ~ L a 1 T, 'C r 1727 1394 1156 777 838 .727 636 560 8 I '7 i 1 I I i 1 k! l J 6' h Sandia 4 Curve 1 2 A 0 ~ N D ++ ss 2 'Ds ]-2 -(Powel et al) N o s 4 Zr-4 in steam N 's cE Zr in steam N o g-4 g., Sandia ~ Curve 1 38 -6 5 l Leistikow -1985 g i -8 e Zr-4 in Air, t < 30 min \\ 'l + Zr-4 in Air, t > 60 min 1 o Zr-4 in Steom, i < 30 min i -10 e Zr-4 in Air, t < 15 min (1978) f'] 1 56600 1

• 2

-12 8 T z ~ , = 3.09 x 10,xp Zr-4 in Alt, t<30 min I I I I i 1 I 5 6 7 8 9 10 11 12 [ 4 10 /T, 'K-l L -i Figure 1 -- d

44 ' .y .T Appendix 1. ( Kernforschungszentrum Karlsruhe 4. Gesenschaft sme bescruaanier Mettung U x- ~ x.,w.e. ca peau.eus.o o.nco n.,wone i In:,titut fur Dr. John Weeks f.interini und Festkorperforschung' 11 Katerials Technology Division j ,,,,g,,,,,,,..... Brookhaven Netional La ratory Upton, Iong Island, N.Y.'11973 Datune 21.10.1985 - Th. 1 USA esereener Dr. S.141stikow Teeesen:on4 net-2915 Swe Metemung f

Dear Dr. Weeks,

Dr. Garzarolli was right in telling you about our Zircaloy-4 oxidation experi-ments in air which in ficP. are not yet published. Sorry that we did not work above 1100 C which in facs could be done next year. So I send our curves Zircaloy-4 in a.ir 800 - 1100 C Zircaloy-4 in steam and air 800-1100 C and add some other, more general publications we wrote on Zircaloy-4 behavior i under IMR accident conditions. I was aware of the problem (normal operating / accident conditions) when.I applied for presentation of my paper to the Monterey conference. only your first pcsitive-reacthn gave me hope. A participation after the rejection of the paper was then excluded because we are contrib0 ting to IMR problems only under the safety aspect - which in fact have their own series of conferences. In case you have further questions don't hesitate to ask me. We dispose in case. of the air oxidation experiwetis about a lot of other information. + Very sincerely, P t \\ \\ A Y.Ak h L' YAC)Q +, Q r _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _ __ _ b. r i - '+ n ____._i_____. _ _ _ _ _

Am/A 2 (mg/dm ) 5 ~~ ~ Luft *C 4-f ~~ Dampf ::"" / l '/ 3-X -3 es, / s 2

e. '7 /

,/ * /, 2

• s ,,4.'

es' s', g- ?./r s s se I ^ l 1$ ', / 3o' - - o. /,,3 / /,. 8-a a -g 6- / /./ '. Ai. ,.e ***,.- 1100*C 5- , < ',..a 'p.g a -6 f n. **,s '.,.<**

• e.. -

..e'*.... s .5 7 1000 ,,. ' " 'e '/", -;' 3- -3 950*,r..p' y, ./ 900*C

• 2 800*C 10 I

_g 0,1 0,25 0,5 1 1,5 2 ,h I, 5 Ih1 3 5 1'O 1'S 3'O M N 360 [ min] g isotherme Zirca ov-4/0xidation Vernimich efer Mnecen ninnhmen. i onn* 44nn*r % s

d w _J .f .a f 4 ,r ~ y d** =J / W ? g% / j 1 m / ~ o e== 'l / ~ .e. 4: i 'f s r I -c, t. b g g .s_% i N ,6' ~ \\ k m. ~. l ( l . -. se :. ~- ^ l N c' ,j ' e W. I O =N !): M W .e ee e 9 \\ s 1 ..= O, ' .W S.S 9 .\\

g...

O e e...

e.

+ <f' 'p t } 4 N d. 3 ~ 4"* N y et [' e N n 1:b ~ l l

ZlRC A LOY 4 - HDCHTEMPERATUR-0XIDATION " ' ' ~ " *

• DURCH WASSERDAMPF.5AUERST0FF UND LUF1 IN ABHANGlG-20s. t,e..cntuuneo, KEli VON REAK110N51EMPERATUR UNO 2Eli

,,,,,, 73 5-15 min. 900-1300 *C.1 bar ( i o. A ? I j 5_ min gl 1000 j g g#! j g-900 1000 1100 1200 lac 1 gg.,,6ewentuunonme lag /cm2) . w,

• Souersteil a

2000 & WmetdomDI 1 10 min I I1 .7 ioco /' g,}y i 900 1000 nao uoDlot! M ~Gewchtuenohne o !ag/amil J a jg@ 1/' /:4.' i 15 min l icz .5 s' i g g. l l0

) .sp)NDQ 1'. 4. l s l .] Temperature ('C) 10-1 ~ - PNL-Present Study

---

O R NL - Cathcart, et. al.

~ - AECL-Urbanic and Heidrick - - ANL-Baker and Just ~n s. T 10-2 j E i 3 8 \\. N \\ 1 / { N. N.'\\. e e E N..\\ a: 5 \\* 3-c 10 3 s t N w s c ..N~ k-h, i g g i s l 1 10 ' 1-1 I t 3.5 4.0 4.5 5.0 5.5 6.0 6.5-7.5 10,000/ Temperature (K) Figure 5

A-7 r A.4 DATA FOR GINNA A.4.1 Reactor and Fuel Cycle Parameters Table A.9 summarizes several of the major reactor characteristics and fuel cycle parameters for Ginna. A.4.2 History of Operations Several milestones in the operation of Ginna are summarized in Table A.10. A narrative of the operating history from 1969 through 1979 can be found in Ref. 12, Append!x F. ~ Reconstruction of the refueling history during the early years of opera-tion has been difficult using data readily accessible to BNL Staff (direct ac-cess to the Licensee for information was precluded). Table A.11 lists the re-a fueling data used by BNL for the ORIGEN2 calculations, which were carried out in 1985. Subsequently, additional information has been located that would permit a revision of the data in Table A.11, but repeating the ORIGEN2 calculations did not seem worthwhile since only minor changes in the spent fuel radioactive in-ventories would have resulted. At the time Table A.11 was constructed, no data on the first refueling in February,1971 was available.' Also, some 84 fuel assemblies from early refuelings could not be accounted for. Later, it was learned that 81 assemblies had been shipped for reprocessing at the West Valley facility. These apparently were returned in 1985 to Ginna for storage in the spent fuel pool.13 At the time of the April 1972 refueling, cladding distortions due to fuel densification was discovered and 61 assemblies were replaced (Ref.12, pg. i F-56). Thus, the entry in Table A.11 for the second discharge is incorrect. I The total burnup not accounted for in the ORIGEh2 calculations amounts to 4.2% of the total thermal energy production from 1969 through April 1,1987. l The missing 4.2% burnup is for fuel discharged on or before April 1972.

A-8 'A.4.3 PWR Fuel Assembly Model Used in ORIGEN2 Calculations A nominal PWR fuel element has been 'modeled based on' data presented in Ref. 9. This is a 17x17 element assembly '(264 fuel _ cMents per assembly) of- - 3.2% ' U-235 enrichment containing - 461.4 kg' of uranium. The cladding is Zircaloy-4 Other alloys - present-- in the fuel assembly _ hardware include Inconel-718, Nicrobraze 50, SS-302 -and SS-304. The alloy contents of the assembly hardware are included with weighting factors to take account. of the axial variation of the neutron flux which results in lower' neutron flux which-' results in 1 ewer neutron activation at the ends of the assemblies.- In:addi-tion ' to the fuel, the cladding and the assembly hardware, an allowance was made for the presence of " crud," composed of Cr, Fe, Co and Ni, on the outer surfaces of the cladding and hardware. No corrections were made in the ORIGEN2 calculations to -account for~ stainless steel clad fuel that was used in the early history of the plant. i A.4.4 Calculated Radioactive Inventories ') The calculated inventories of selected radionuclides

• are listed in ' Table l

A.12 for the. end of operating cycle number 16 projected to be on April 1, 1987. Also listed are the inventories in the spent fuel basin on May 1 and July 1,1987 and April 1,1988, assuming that 24 assemblies will be discharged. in the Apr11 1987 refueling. It should be noted that many of the isotopes that are of considerable importance in a core melt accident are those of 'short half-lives which are no longer present in the spent fuel after a few days of decay, e.g., Rb-91, Rb-93, Sr-93, Sr-95, Y-94, Y-95, Tc-104, I-134, I-135, I-136, Cs-138, Cs-140. On the other hand, the spent fuel. inventory contains much larger quantities of several long-lived isotopes than does the equilibrium core. Noteworthy among these are H-3, C-14, Sr-90 (Y-90), I-129, Cs-137, 8a-137m, Eu-154, Pu-239, .i Pu-240, Pu-241, Am-241, and Cm-244

• The selection of radionuclides was based on several criteria including poten-tial for biological concern, thermal power and total curias of activity.

)

A-9 Table A.13 gives a comparison of the radionuclides inventories in the last fuel batch to be discharged with the sumation of the inventories contained in batches 2-15 discharged between 1976 and 1986. A.4.5 Decay Heat Table A.14 sumarizes the decay heat production in the various discharged batches. The data shown is for the whole batch, i.e., the specific thermal power (kilowatts per metric tonne) multiplied by the metric tonnes in the batch. Table A.15 sumarizes the fraction of the decay heat contributed by various isotopes. The main contributors change with decay time, e.g., in the oldest fuel (batches 2, 3, etc.) the largest contributors are Y-90 and Ba-137m, whereas the last discharged batch 16 is dominated by Cs-134, Rh-106 and Pr-144. The actinides are relatively small contributors. - t A.5 REFERENCES FOR APPENDIX A 1. Nucleonics Week, a weekly newsletter published by McGraw-Hill Publishing Co., New York, NY. 2. Nuclear Safety, published bimonthly by the Nuclear Safety Information Center, Oak Ridge National Laboratory. 3. Nuclear Engineering International, published monthly by IPC Business Press, Ltd., Sussex, England. 4. U.S. Nuclear Regulatory Comission, Licensed Operating Reactors, HUREG-0020, Vols.1-10, published monthly. 5. U.S. Depart:nent of Energy, Richland Operations Office, Program Office, Comercial Spent Fuel Management, (private communication from P.A. Craig, i Director, June 11,1985). Q .___.______________.__.____u

A-10 6. A.G. Crof f, "0RIGEN2: A Versatile Camputer Code for Calculating the Nuclide Compositions and Characteristics of Nuclear Materials," Nuclear Technology. Vol. 62, pp. 335-352, September 1983. 7. Internal Memorandum, Brookhaven National Laboratory from V.L. Sailor to R.A. Bari, " Comparison, of BNL ORIGEN2 Calculations with ORNL, May 27, 19eo. 8. T.G. Piascik, L.E. Fennern, S.R. Specker, K.L. Stark, R.E. Brown, J.P. ~ Rea, and K.T. Schaefer, "BWR Operating Experience at Millstone 1 with Control Cell Improved Design," Transactions of the American Nuclear Society, Vol. 32, pg. 706, June 1979. 9. A.G. Croff, M.A. Bjerke, G.W. Morrison, and L.M. Petrie, Revised Uranium-Plutonium Cycle PWR and BWR Models for the ORIGEN Computer Code, Oak Ridge National Laboratory, ORNL/TM-6051, September 1978. ~ 10. U.S. Nuclear Regulatory Commission, Integrated Plant Safety Assessment Systematic Evaluation Program, Millstone Nuclear Power Station, Unit 1, NUREG-0824, February 1983. 11. U.S. Nuclear Regulatory Commission, Nuclear Power Plant Operating Experi-ence, 1974-1975, NUREG-0227, April 1977, pg. 65. 12. U.S. Nuclear Regulatory Commission, Integrated Plant Safety Assessment Systematic Evaluation Program, R.E. Ginna Nuclear Power Plant, NUREG-0821, December 1982. 13. Rochester Gas and Electric Corporation, " Application for Amendment to Operating License to Amend Appendix A to Increase Spent Fuel Pool ~torage Capacity," submitted to NRC April 2,1984, Docket No.50-244.

A-11 Table A.1 Reactor and Fuel Cycle Parameters for Millstone 1 (Sources: Rcfs. 1-4) Assemblies in core: 580 Licensed thermal power: 2011 MWth (gross) Th( rmal power corresponding to maximum dependable capacity: 2006.5 MWth (gross) Nominal initial metric tonnes of heavy metal (IMTHM) per assembly: 0.1833 MT Average refueling cycle interval (since initial commercial operation): 21 to 22 months Recent refueling cycle intarval (since April,1979): about 18 months Average number of assemblies per discharge: about 173 Average IMTHM per discharge: about 31.7 MT \\ Average number of fuel cycles per assembly: about 3.35 Average period of irradiation (including downtime): about 72 months 1 Authorized Storage Pool Capacity (as of 1985): 2184 assemblies O e 0

t A-12 Table A.2 LSummary of Operational Milestones for Millstone 1 (Source: Ref. 4) ) Date of Initial Criticality: October 26, 1970 Date of First Electricity Generation: November 29, 1970 Date of Commercial Operation: March 1, 1971 Lifetime Cumulative Data: (January 1, 1971 - March 31, 1986) Hours, Generator on Line: 100,307.9 hours Gross Thermal Energy: 184.83 x 108 MWh Capacity Factor (MDC net): 67.4% 6 e l l l o s mism % =a-e-. 9>- A me, weagans

1 ~ l b t e ehu u v gF l mu ii o 5 7 2 8 3 6 9 9 0 2 3 f c t et o) aWnPT 9 4 5 1 8 1 8 2 3 2 6 l o cy e at u s p n (M l e 8 5 1 1 0 8 1 3 8 5 8 us i 7 2 6 0 4 0 6 1 7 2 fe ec msSi 1 1 2 2 3 3 4 4 5 n ha uo Crf gn tp Go no a it ,c d g u6 ne n l4 i g i es c2 l a s ve n0 er i uo a iil d 0. ft B tl o 8 6 0 4 8 6 4 6 8 6 3 t s es abo 2 3 8 0 2 7 4 3 0 8 5 o ,s r e l mP 2 3 5 6 7 9 1 3 4 6 nl o d g ue 1 1 1 1 ar 7e a msn tt g 8z r usi o CA ue /i t bms 1 r S e 0o eyn /h 1 rvn 4t o aao 0u e at7) wet a dh e n y 8s 6 6 0 5 9 6 4 3 4 7 0 r c he o as/y 5 2 5 6 3 2 9 9 1 7 3 afi th t cy1 a 3 6 2 8 3 9 3 6 1 5 hor t s ea/d 5 4 4 3 3 2 2 1 1 l DD5( t r l dse a nem

i. de i

M an fe n8 Ac go9 x n ) i pT nt1 .e 3 i .uM 6 5 1 0 0 4 8 0 3 2 3 dc sd s gn/ 8 9 8 9 9 5 9 7 6 5 6 di0 el e vrD 6 6 5 2 0 3 9 6 7 0 9 ar i u h AuW 2 9 6 1 4 4 4 3 6 8 9 lt g l o c) BM 1 1 2 2 2 2 2 2 2 2 2 cen bw t7 ( a8 mi m B9 s l ee 1 u3 a sr ~ l l3t so p8 o ac eo t 2 6 5 9 9 8 4 4 8 7 1 ut h .) 3 2 9 2 2 2 9 9 2 2 1 l 1. t 4e F g M. T i M 1 1 3 7 7 1 7 1 5 6 6 a0, 8h s .t e 1t t n eH( 5 8 6 2 2 7 0 5 1 2 0 7 eyr 2 no W 3 2 2 2 2 3 3 3 3 3 8ml a n ei pt / ew si Sc 1 ytd i e / var sd fj s 5ama yee oo e eih tir ohx; ili r r i yP e l t of cb u r bfb 8 4,4 4 4 8 8 2 2 8 7 nep pee sro amq ah mom 2 0 4 2 2 4 6 9 7 7 6 od ps asr mt u e 2 1 1 1 1 1 1 1 1 1 iuae cs mi N s uW s tl n ae S( A acsn e b inno g0 diit a8e a a r5 g 3 rlt9 o r A n ron1 tea o ioo1 shh pc1 tc e f i 2 4 5 6 8 9 0 2 4 5 7 f d s l oft 7 7 7 7 7 7 8 8 8 8 8 o n y0 esi o oa / / / / / / / / / / / il zud e e i 1 1 1 0 0 7 3 1 2 1 1 d bd il T tdd 3 3 1 3 1 2 0 1 1 0 0 nemn rpe ana / / / / / / / / / / / egea o r der 8 8 9 9 3 4 0 9 4 0 4 as r 0 0 0 0 0 0 1 0 0 1 0 mns).hso t ec I ona 2a ui ro 0t all fthUa bl c dt mu sl an nef le. yeEides uo au ( essa FN df t ea C c .ncr d 1 2 3 4 5 6 7 8 9 0 1 y s s e e pd l th 1 1 a s k gj eu nc et cocyoeto pa eraxrhah SB DGroPTl s ab cd l

.l A-14 l Table A.4 Comparison of Cumulative Gross Thermal Energy Production with Calculated Fuel Burnup from Start of Operatinns in 1970 to April 1,1987-(Millstone 1) Total Cumulative Total Burnup Genss Thermal Energy Spent Fuel in Batch (MWD x 10-3) Batch No. (MWD x 10-3)- 1 65.10 2 750.88 3 701.61 4 483.91 5 547.54 6 660.68 7 769.78 8 833.05 9 843.78 ) 10 915.25 11 917.21 12* 612.74 ' 13* 329.55 Total 8440.25 ~~8440.01 %urnup in fuel remaining in the core. 4 'I w- ., no e,aw

~ 3-A-15 Table A.5.' Comparison of Radioactive Inventories in Re8ctor Core' and Spent Fuel Basin (Millstone 1). The Assumed Refueling Scenario is Described in Section' A.3.4 no ctor soent ruel storace 8esine isotope Core sn /s7 7A /s? I on /s7 en iss (Redlooctivity, Curles) H3 4.95EM 1.38EM 1.37E+5 1.35E+5 1 J1 E+5 C 14 1.02E4 J4.12E4 4.12 E42 4.12 E4 4.12E42 Co 58 8.81 EM ' 2.2 9E44 1.26E*4 5.12 EM 8.54 E+2 Co 60 1.64E+5 3.72E+5 - 3.15E+5 3.04E+5 . 2.85EM Kr 85 5.35E*5 1.41 EM 1.39E+6 1.37E46 1 J3E+6 Rb 86 6.22 E+4 1.01 E +4 1.05E+3 3.44 E+1 3.84E-2 Sr 39 4.71 E+7 8.39E46. 3.63E+6 1.03EM - 8J3EM - Sr 90 4.25E46 1.42E+7 1.42E+7 1.41 E+7 1 J9E+7 Y 90 'd.37E+6 't.43E+7 1.42 E+7 1.41 E+7 1 J9E+7 Y 91 6.06E+7 1.18E+7 5.7M+6 1.98E+6 2.21E+5 Zr 95 8.70E+7 1.94E+7 1.00E+7 3.70EM 5.10E+5 le 95 8.91E+7 2.54E+7 1.70E +7 7J5EM 1.11 E,46 No 99 8.78E+7 - 1.49E+4 -3.12E 3 neg. neg. Tc 99m 7.69E+7 1.43E*4 3.01E-3 neg.b neg.b Ru 103 7.2 M+7 1.53E+7 5.21E+6 1.03E+6 4.07E*4 Ru 106 2.48E+7 1.72E+7 1.53E+7 1.29E +7 9.13E+6 Rh 106 2.63E+7 1.72E+7 1.57,E +7 1 J 9E+7 - 9.13E46 Sb 125 9.07E +5 1.19E46 1.14 E+6 1.07E,46 9.48E,+5 Sb 127 4.97E+6 8.21 E+3 1.39E-1 nog. neg. To 125m 1.93E+5 2.84EM 2.76E+5 2.61 E+5 2.31E+5 To127 4.92E*6 2.21 E+5 1.45E+5 8.06E*4 2.52E+4 To127m 6.61E+5 2.18EM - 1.48E+5 - 8.23E44 2.57E44 To129 1.49E+7 2.74EM - 7.79E M 1.17E44 l 2.6SE42 To129m 2J4E+6 4.21 E+5 1.20E+5 1.79E*4 4.12 E+2 To 132 6.72E+7 3.74E44 8.64E-2 neg.8 neg.b i129 1.75E40 7.15E+0 7.15E+0 7.15E+0 7.15E,40 l 1 31 4.74E+7 1 J2E+6' 6.3 5E+3 2 J8E,*0 neg. I 1 32 6.83E+7 3.85E*4. 8.90E*2 neg. neg.b Xe 133 9.7tE+7 7J9EM 2.30E+2 1 J1 E-3 neg.b Cs 134 6.10E+6 7.90E*6 7.47E+6 6.86E+6 5.80E+6 Cs 136 2.10E+6 2.05E+5 8.13E+3 6J6E+1 3.91E-3 Cs 137 5.84EM 2.02E+7 2.01 E+7 2.00E+7 1.97E+7 Be137m-5.53E+6 1.91 E+7 1.90E+7 1.89E+7 1.87E+7 Be 140 0.36E+7 5.19E*6 1.90E +5 1.30E+3 6.41 E-2 Le140 8.54 E+7 5.97E+6 2.19E+5 't.50E+3 7.37E-2 Co 141 7.94E+7 1 J2E+7 3.61E+6 5.07E+5 -t.03E+4 Ce 144 6.05E+7 2.64E+7 2.27E +7 1.8t E+7 1.16E+7 Pr 143 7.37E+7 5.44E46 2.41E+5 2.19E+3 1.90E-1 Pr 144 6.00E+7 2.6dE+7 2.27E +7 1.81 E+7 1.16E +7 141 147 3.16E +7 1.Ed+6 3.36E+4 1.0SE4 1.10E-3 Se 151 2.44E44 8 72E44 8.21 EM 8.19E+4 8.16E+i. Eu 154 4.61 EM 1,34E4 1.32 E*6 1.29E+6 1.25E+6 Eu 156 5.8% EM S J6E+5 5.10E44 7.76E*2 1.80E-1 Np 239 9.98E46 5.59E*4 2.88E+5 2.88E+3 2.88E+3 Pu 238 9.33E44 4.51E+5 4.53EM 4.54EM 4.54E+5 Pu 239 2.49EM 8.89EM 8.89E*4 8.89E*4 8.89EM Pu 240 3.14EM 1 J0E+5 1.30E+5 1.30EM 1.30EM Pu 241 7.19EM 2.29E+7 2.27E+7 2.25E+7 2.19E+7 Am 241 8.46E+3 2.88E+5~ 2.94E+5 3.03E+5 3J1EM On 242 2.09EM 1.45E*6 ' 1.12E46 7.60E+5 3.50E+5 Qe 244 6.72 E*4 2.27E45 2.25EM 2.23E+5 2.19Ee eSpent fuel pool inventory includes discherges from 11 refuelings cover-Ing the period from August 1972 through the projected refueling of April 1 987 b less then 10~3 Curles. neg. a 4

A-16 Table A.6 Comparison of Radioactive Inventories of Most Recently Discharged Fuel Batch (Batch 11) with Longer Aged Dis-charged Batches (Batches 1-10) (Millstone 1) Soont Fuel Batch 11 e Spent Fuel Batch 1-10b isotope 5/1/57 7/1/57 10/1/57 4/1/55 /l/57 7/1/57 60/1/57 4/4/BB I (Redloactivity, Curles) l H3 2.27EM 2.24EM 2.21 EM 2.15E+4 1.16E +5 1.15E+5 1.13 E+5 1.10Ee C 14 5.18E+1 5.1 SE+I 5.1 BE+1 5.18E+1 3.61 E+2 3.61 E*2 3.61 E+2 3.61 E+2 Co 58 2.28E4 1.26EM 5.10E+3 8.49E+2 1.18E +2 6.48E+1 2.63E+1 4.39E +0 Co 60 7.64E+4 7.48E 4 7.24 E+4 6.77EM 2.45E+5 2.40E+5 2.32 E+5 2.17EH \\ Kr 65 2.39E+5 2.36E e 2.32E+5 2.2 5E+5 1.17E+6 1.16E+6 1.14 E+6 1.10EM Rb 86 1.01 E+4 1.05E+3 3.44E+1 3.84E-2 neg.c neg. neg. neg. Sr 89 8.39E+6 3.63E+6 1.03E+6 8.33EM 5.39E+3 2.33E+3 6.60E+2 5.3 5E+1 Sr 90 1.93EM 1.92 E+6 1.91E+6 1.89E+6 1.23E+7 1.23E+7 1.22 E+7 1.21E+7 Y 90 1.93E+6 1.92E+6 1.9t E+6 1.89E+6 1.23E+7 1.23E+7 1.22 E+7 1.21 E+7 Y 91 1olSE+7 5.74E+6 1.93E+6 2.21 E+5 2.11 EM 1.02 E+4 3.44 E+3 3.94 E+2 Ir 95 1.94 E+7 1.00E+7 3.69E+6 5.09E+5 5.90E+4 3.05EM 1.12 E+4 1.55E+3 Nb 95 2.53E+7 1.69E+7 7.33E+6 1.11 E+6 1.31 E+5 6.76EM 2.49EM 3.44 E+3 Mo 99 1.49EM 3.12 E-3 neg. neg. neg. neg. neg. neg. Tc 99m 1.43E4 3.01 E-3 neg. neg. neg. neg. neg. neg. Ru 103 1.53E+7 5.21 E+6 1.03E+6 4.07EM 1.09E+5 3.73E+2 7.3 5E +1 2.91E+0 Ru '106 1.12E+7 9.98E 4 8.40E M 5.95E+6 5.98E+6 5.30E +6 4.48E46 3.18E46 Rh 106 1.12E+7 9.98E+6 8.40E 4 5.95E+6 5.98E+6 5.30E+6 4.48EM 3.18E+6 Sb 125 4.17Ee 4.00E e 3.76E+5 3.31 E+5 7.76E+5 7.44Ee 6.99E+5 6.16E +5 Sb 127 8.21 E+3 1.39E-1 neg. neg. neg. neg. neg. neg. To125m 9.42EM 9.39EM 9.04 E4 8.07EM 1.89Ee 1.82 Ee 1.70E6 1.50E+5 To127 2.14E+5 1.40E6 7.79E+4 2.43E+4 7.15E+3 4.85Ee 2.70E+3 8.44 E+2 To 127m 2.10E+5 1.43E+5 7.95E+4 2.48EM 7.30E+3 4.95E+3 2.76E+3 8.62 E+2 To129 2.74~6 7.79E+4 1.17E44 2.68E+2 3.85E+0 1.09E40 1.64 E-1 3.76E-3 Te 129m 4.2 5+5 1.20E6 f.79EM 4.12 E+2 5.9tE+0 1.68E+0 2.52 E-1 5.77E-3 To 132 3.74E4 8.64 E-2 neg. neg. neg. neg. neg. neg. ) s I129 8.84E-1 8.86E-1 8.86E-1 8.86E-1 6.26E+0 6.26E+0 6.26E40 6.26E40 l 1 31 1.22E+6 6.35E+3 2.28E*O neg. neg. neg. neg. neg. I 132 3.85E+4 8.90E-2 neg. neg. neg. neg. neg. neg. Xe 133 7.29E+5 2.30E+2 1.21E-3 neg. neg. neg.

nog, neg.

Cs 134 3.53E+6 3.34E+6 3.07EM 2.59E+6 4.37E+6 4.13E+6 3.80E46 3.21 E+6 Cs 136' 2.05E+5 8.13E+3 6.26E+1 3.91 E-3 neg. neg.

neg, neg.

Cs 137 2.83E+6 2.82 E+6 2.80E+6 2.77E+6 1.73E+7 1.73E+7 1.72 E+7 1.70E+7 Ba 137m 2.67E+6 2.66EM 2.65E+6 2.62E+6 1.64E+7 1.63E+7 1.63E+7 1.61E+7 8a 140 5.19E +6 1.90E+5 1.30E+3 6.41E-2 neg. neg. neg. neg. l.a 140 5.97E+6 2.19E +5 1.50E+3 7.37E-2 neg. neg. neg. neg. Co 141 1.32E+7 3.61 E+6 5.07E+5 1.03E+4 1.31 E42 3.57E +1 -- 5.02 E+0 1.01 E-1 Co 144 1.91 E+7 1.65E+7 1.32E+7 8.43E+6 7.23 E+6 6.23E+6 4.98E+6 3.19E+6 Pr 143 5.44E+6 2.41 EM 2.19E +3 1.90E-1 neg. neg. neg. neg. Pr 144 1.91 E+7 1.65E+7 1.32E+7 8.43E+6 7.23 E+6 6.23E+6 4.98E+6 3.19EM Nd 147 1.54 E+6 3.36EM 1.05E+2 1.10E-3 neg. neg. neg. neg. Sm 151 9.31 E +3 9.30E6 9.28E+3 9.2 5E+3 7.29E44 7.28EM 7.26EM 7.24E+4 Eu 154 2.89E+5 2.85E+5 2.79E+5 2.68E+5 1.05E+6 1.04EM 1.02 E+6 9.75E+5 Eu 156 8.38E +5 5.18EM 7.76E+2 1.83E-1 neg. neg. neg. neg. Np 239 5.36E+4 5.26E +2 5.2 6E+2 5.26E+2 2.35Ee 2.35E+3 2.35E+3 2.3 5E+3 Pu 238 6.73E+4 6.87E+4 7.02 EM 7.18EM 3.84E+5 3.84E+5 3.83Ed 3.82 E+5 Pu 239 9.2*E+3 9.28E+3 9.2 8E+3 9.28E+3 7.96EM 7.96EM 7.96EM 7.96E*4 Pu 240 1.55E+4 1.55E*4 1.55EM 1.55E+4 1.15E+5 1.1$E+5 1.1 SEe 1.1SE+5 Pu 241 3.73E+6 3.70E+6 3.65E+6 3.56EM 1.92 E +7 1.9%E+7 1.88E+7 1.84E+7 Am 241 6.01 Ee 7.00E+3 8.48E+3 1.14E+4 2.82 E+5 2.87E+5 2.95E+5 3.09E+5 On 242 1.31 E+6 I.0', E+6 6.86E+5 3.16E+5 1.39E+5 1 08E+5 7.33E+4 3.47EM On 244 5.88EM 5.84E+4 5.79EM 5.68E+4 1.68E +5 1.67E+5 1.65E+5 1.62 E+5 aFuel batch 11 is projected discharge during April 1987 b Fuel batches 1 -10 we5* Curl es.'5*d ***"**" ^"S"** 1 '72.nd Octob.c 1985. d*"* cneg. = less than 10*

A-17 Table A.7 Decay Heat Released from Spent Fuel Inventory for Various Discharged Fuel Batches (Millstone 1) Decay Heat Released by Batch Date, End of Batch Sizea May 1, 1987 July 1, 1987 April 1,1988 Irradiation (Metric Tonnes) (Kilowatts, Thermal) 1 08/31/72 5.13 1.8 1.8 1.8 2 08/31/74 38.13 22.0 21.9 21.5 3 09/11/75 26.40 21.8 21.7 21.2 4 09/30/76 22.73 15.2 15.1 14.8 5 03/10/78 22.73 18.4 18.3 17.7 6 04/27/79 27.13 23.5 23.3 22.4 h 7 10/03/80 30.79 30.3 29.9 28.2 8 09/11/82 35.19 41.5 40.3 35.9 9 04/12/84 31.53 67.4 63.6 50.9 10 10/01/85b 32.63 146.0 132.7 91.8 11 04/01/87b 30.61 909.0 537.7 210.5 Totalc 1-10 272.38 387.9 368.5 306.3 Totalc 1-11 302.99 1297.0 906.3 516.8 aSee Table A.3. b rojected dates. P cTotals may not equal sum of the entries due to rounding of decimals.

39 3 5 86 0 5 s 1 1 14127214291100005 6 e 21 2 - 9 l re ~ . = h es 47123622305733101 0 to 310056287,21335172 1 h a T 0 l 1 21002630192200001 5 y 1 21 1 1 9 n .l a e sw m e h s g ca 86 9664903270711 9 no t 70 - 4213780759162 8 m a, 9 Bh 38 - - 3657083300010 3 a c 1 11 1 9 l t d ea e u8 t F u h 37 5354630987225 9 b t c 29 - - 7693293482290 1 r i na 8 eE 54 - 5163023401020 4 t p 2 1 2 9 s Sf i o ) d s T up r A s ou e E 52 9573642399423 3 i i n )1 b H 89 - 8684065394220 1 rr m7 t au u Y 57 - 1776003501040 4 a VB e N A 2 - 2 9 e n C h rl o h E t oa c D y s f t t a ol a L 82 1580 2101423 7 c tTl 8 A 15 - - 8128 - - 3316210 4 e i a 6T d eeM l O 69 - 0587 - - 35110S0 4 Hh( e T 2 2 9 e t u h y n F F t ano O coi t f e t n T 21 459 5868443 9 o Dda e N 32 - - 742 - - 1316280 5 ni p5E e oed S C 60 - - - 388 - - 3511050 4 c t pa R 3 2 9 n er E a sDr P l n I ( a os b i nf 12 402 3668402 8 e t oo 68 - - - 272 - - 6637270 9 ui 4 h btd 61 - - - 289 - - 2411060 5 T i un 3 2 9 rbE ti nrr d ot e e Cnt 43 630 3349345 7 .t of 13 - - 743 - - 0317230 8 s %i eCA 3 d 69 - - - 188 - - 3711070 4 1l iee 2 2 9

0. s l gm cai 0e utT p

nn no oey 94 874 5540264 8 at ica 74 - - - 974 - - 1859290 0 ho. d rc 2 t ss aee 62 - - - 089 - - 2411070 7 ir RPD 3 2 9 s o sft eou l b 8 - ei A 83a 329 2464971 8 egr 47 - - 402 - - 2118150 0 t at e 1 at n 75 - - 090 - - 1221070 8 cno l b 3 3 9 i ec a dc T nrt i enpa e r. h s t p 6477444890112 s el r o 0 550333445334444 l hao t 90991111111222222 a st p o 9 t aom s r r h h s s a 'e r u u u u u m m o 0Ti I SYZNRCC8CPEPPPPAC T a g Fl .l t

A-19 Table A.9 Reactor and Fuel' Cycle Parameters for Ginna (Sources: Refs. 1-4) Assemblies in core: 121 Licensed thermal power: 1520 MWth (gross)a Thermal power corresponding to maximum dependable capacity: 1499 MWth (gross) Nominal initial metric tonnes of heave metal (IMTHM) per assembly: 0.375 MT Average refueling cycle interval (since initial commercial operation): 12.6 months Average number of assemblies per discharge: 1975-1980: 37 1981-1987: 24 Average IMTHM per discharge: 1975-1908: 15.3 MT 1981-1987: 9.0 Average number of fuel cycles per assembly: 1975-1980: 3.27 1981-1987: 5.04 Average period of irradiation (including down time):- 1976-1980: 3.3 years 1981-1987: 5.0 years Authorized storage pool capacity: 1016 aOn March 1,1972 the Atomic Energy Commission authorized an. increase in gross thermal power from 1300 to 1520 MW.

A-20 Table A.10 Summary of Operational Milestone for Ginna (Source: Ref. 4) ) Date of Initial Criticality: November 8, 1969 Date of First Electricity Generation: December 2, 1969 Date of Commercial Operation: July 1, 1970 Lifetime Cumulative Data: (January 1, 1968-March 31, 1986) Hours, Generator on Line: 107,134.3 hours Gross Thermal Energy: 149.26 x 106 MWh Capacity Factor (MDC net): 70.3% O 6 e St 0 0 'an 'I '-

l b t e e eh u l l vgFl e. b ii o un a t et o) 478111415346086 f e l aWnPT g i l e M 0832741719583051 ey a uspn( 124692478012568 h x v msSi 1111222222 t o a uo Crf tf g G o oo n n i s s t e u un es bn r n ve o o i iil et f s tl o 8 r a abo 0865234408225048 a0 d B l mP 236048268135802 w2 e ue 111222333344 d6 t e msn r0 n 9 usi a u a CA h0 o r c o d, c t nl a S aa o t e a a t7) ge b n y 8s 297196305176270 nm n as/y 363162090966993 ,i t i c y1 a 88416306284173 dy o G ea/d 544433322111 d v n DD5( aa n l e d i ch lu s sf o e ) uo c h pT l c) .uM 3593821184422303 ps s t7 gn/ 3934522588573367 e e a8 vrD 9600909408075536 l n.i B9 AuW 6608667567137025 ansl 1 BM 133332222333444 t osb l ( et om eo m re ut 4 gs F y1 s s ~ v6 sa t n t 8253855683686566 a4 e no h .) 7798777778888888 e n6 g M. T 7715733931591955 .h0n5 ei o i M pt Sc eH( 4831466541979999 7 s yt e e W 11 1111111 8el m fj /d e9o oo 1 ut7 s r /l a5 c m 6. s yP 5 r s ni n s ah e oix0o e mt r i cc t o i i mi e l )) l r gt l uW bfb 7789711068403544 nopna b S( mom 34 2344432222222 o o pi l m u e (( i pal u e N s t ac s 1 s anstl s A iinoa a 1 d it c A ada 6 ret ,2 1 e rrneN 0 l n i oorE 1 b o t caG a f i 1 235678901234567 f s wI s T oft 7777777788888888 o ydR i oa //////////////// ll rO 7318843987551801 deba y e i tdd 2130212021220230 numhe t ana //////////////// efe h i DE r 2423143234133234 sft c r 0010000000000000 mf so .a I ooa f ap r so t a fthe ac h cne sganm

d. d l

yiE oi d e e. ae tt ez uo d w ti FN d 5 e .cr 1234567890123456 yss4h eo th 1111111 ask3t ajh t ot c o c 1. t nc era aru et pa DG r0AdPA SB ab c de

4 A-22 Table A'.12 Comparison of Radioactive Inventories in Reactor Cora and Spent Fuel Basin (Ginna) Reector Soent Fuel Storaon Busin8 ~~ isatope Core 5/1/87 7n / B7 4/L/f8 (Redioectivity, Curles) H3 3.32[M 0.29EM 9.2 0E+4 8.82 E+4 C 14 6.42 E+1 2.64E+2 2.o4 E +2 2.64E+2 r Co 58 3.57E+5 5.93E4 3.2 6E+4 2.21 E+3 Co 60 3.20Ee 5.97E+5 5.84EM 5.29EM Kr 85 3.73E+5 9.84Ee 9.74EM 9.27E+5 Rb 86 6.53E e 7.22 E+3 7.48E+2 2.74E-2 Sr 89 3.55E+7 3.53E+6 1.53E+6 3.50E+4 Sr to 2.95E+6 1.02E+7 1.01 E+7 9.95E+6 Y 90 3.15Ee 1.02 E+7 1.01E+7 9.95E+6 Y 91 4.57E+7 5411 E '4 2.48E+6 9.54E+4 Zr 95 6.41 E+7 8.64E+6 4.46E+6 2.27E+5 m 95 6.34 E+7 1.12E+7 7.51 Ee 4.93Eg5 Mo 99 6.83E+7 7.03 E+3 1.48E-3 neg.b Tc 99m 5.89E+7 6.77E+3 1.42 E-3 neg. Ru 103 5.85E+7 7.86E+6 2.88E+6 2.09E+4 Ru 106 1.95E +7 1.09E+7 9.71E46 5.78E+6 Rh 106 2.15E+7 1.09E+7 9.71 E+6 5.78E+6 Sb 125 6.04Ee 7.11 E+5 6.82 E+5 5.65Eg5 Sb 127 4.12 E+6 4.33E+3 7.35E-2 neg. To125m 1.27E+5 1.70Ee 1.65E+5 1.37E+5 To127 4.05E+6 1.19E +5 7.79E M 1.36E+4 To127m 5.19E+5 1.17EM 7.95EM 1.38E+4 To129 1.21E+7 1.38E+5 3.93E+4 1.3 5E +2 To129m 1.80E+6 2.12 E+5 6.03EM 2.07E+2 To 132 5.33E+7 1.83E+4 4.23 E-2 neg.b i129 1.27E+0 5.32E+0 5.32 E+0 5.32Ep i 1 31 3.76E+7 6.00E+5 3.12 E+3 ug.b i i 132 5.42 E+7 1.89E+4 4.36E-2 neg.b Xe 133 7.64E+7 3.52 Ee 1.11 E+2 neg. Cs 134 5.82 E H5 6.3 5E+6 6.00E+6 4.66E+6 Cs136 1.87E +6 1.26E+5 4.99E +3 2.40E-3 Cs 137 4.21 E+6 1.48E+7 1.47E+7 1.44E+7 Be 137m 4.00E+6 1.40E+7 1.39E+7 1.37E+7 Be 140 6.55E+7 2.47E+6 9.07E+4 3.05E-2 Le 140 6.74E+7 2.85E+6 1.04E+5 3.51 E-2 Ce 1 41 6.28E+7 6.34E +6 1.72E+6 4.91 E+3 Ce 144 4.24 E+7 1.38E+7 1.19E+7 f,09E+6 Pr 143 5.71 E +7 2.54E+6 1.12 E +5 8.86E-2 Pr 144 4.27E +7 1.38E+7 1.19E+7 6.09E+6 Nd 147 2.48E+7 7.42 E+5 1.62EM neg.D se 151 1.42 E+4 5.14E+4 5.13E4 5.10EM Eu 154 4.09E+5 1.09E+6 1.07E+6 1.01 E+6 Eu 156 7.22 E*6 7.58E+5 4.68EM 1.66E-1 Np 239 7.81 E+8

• 3.02EM 3.2 6E+3 3.2 6E +3 Pu 238 1.01 E+5 4.46E+5 4.46E+5 4.46E+5 Pu 239 1.35E M 5.2 5E+4 5.25EM 5.25EM Pu 240 2.02 EM 8.60E+4 8.60EM 8.61 EM Pu 241 4.85E46 1.52 E+7 1.51 E+7 1.46E+7 ha 241 4.990+3 2.10E6 2.14E+5 2.32E+5 On 242 1.91E+6 9.33E+5 7.20E+5 2.2 5E+5

.Os 244 1.2 SEM 3.59E+5 3.56E+5 3.46E+5 s$ pent fuel pool Inventory includes dischargas fece 15 refuelings cover-Ing the period from Aprii 1983 through the projected refueling of April l967. b less then 10*3 Curles. neg. r

3 e A-23 Table A.13 Comparison of Radioactive Inventories of Most Recently Discharged Fuel Batch (Batch 16) with Longer Aged Dis-charged Batches (Batches 2-15) (Ginna) Spent Fuel Batch 16e Spect Fuel Batch 2-158 Isotope 5n is7 7n fs7 ion fs7 en f as sn fs7 7a fs7 i ofi f s7 en f as (Redloactivity Qarles) H3 9.89f +3 9.80E+3 9.66E+3 9.39E+3 8.29EM 8.22 E+4 8.10E44 7.88E+4 C 14 2.20E+t 2.20E44 2 J0E+t 2.20E+t 2.42 E4 2.42r42 2.42 E+2 2.42 E+2 Co 58 5.77E44 3.18E+4 1.29EM 2.15E+3 1.60Ee 8.78E4S 3.57E+2 5.94 E+t Co 60 9.92 E44 9.70E 4 9.39EM 8.79E+4 4.98E+5 4.87E+S 4.71 EM 4.41E+5 Kr 85 1.07E+5 1.0$Fe 1.04E +5 1.00E+5 8.78E+5 8.68Ee 8.54Ee 8.2 7E+5 Rb 86 7.22 E+3 7.48E4 2.45E44 2.73E-2 neg. neg. nog. neg. $r 89 3.50E46 1.52E46 /. 29E+5 3.48E 4 2 J9EM 1.04E44 2.93E 0 2.38E+2 $r 90 84S6E +S 8.53Ee 8.48E e 8.38E +5 9.32E4 9J8E46 9.23E4 9.12 * +6 Y 90 8.57Ee 8.53Ee 8.48E+5 8.38Ee 9.32E4 9.28E46 9.23E*6 - 9.12E+6 Y 91 5.04E+6 2.45E+6 8.23E +5 9.41 EM 6.86E44 3.33E44 1.12 EM 1.2 8E+3 Zr 95 8:47E4 4.37E+6 1.62 E+6 2.23E+5 1.64Ee 4.48EM 3.13E +4 5.60E44 Mb 95 1.09E+7 7.33E46 3.19E4 4.63E+5 3.68Ee 1.89Ee 6.96E*4 9.57E+3 leo 99 7.03 E +3 1.48E-3 neg.c neg. neg. neg. neg. neg. Tc 99m 6.77E +3 1.42E-3 neg. neg. neg. neg. neg. neg. Ru 103 7.86E+6 2.68E+6 SJ8Ee 2.09E*4 1.18E44 4.02 E+3 7.93E4 3.l asE41 'Ru 106 5.82E46 5.19E*6 4.37E46 3.09E4 5.06E46 4.51E+6 5.80E+6 2.69E+6 Rh 106 5.82E4 5.19E46 4.37Ee 3.09E+6 5.06E4 4.51 E4 3.80E+6 2.69E+6 Sb 125 1.84E+5 1.76E+5 1.65E+5 1.46E+5 SJ 8E*S 5.06E M 4.75E e 4.19E+5 sb 127 4.33Ee 7.35E-2 neg. neg. neg. neg. neg. neg. To125m 4.13EM 4.12 EM 3.97EM 3.55E*4 1.78E+5 1.24Ee 1.15E+S 1.02EM Te127 1.08E+5 7.05EM 3.93E+4 1 J3EM 1.09E44 7.42E6 4.14E+3 1.29E +4 To 127e 1.06EM 7.19E *4 4.01 EM 1.2 5E+4 1.12E44 7.58E+3 4.22E+3 1.32E +3 To129 1.38Ee 3.93E44 5.88E6 1.3 5E +2 6.98E44 1.98E41 2.97E*0 6.82E-2 To129m 2.12 E+5 6.03E44 9.04E+3 2.07E42' 1.07E42 3.05E41 4.57E+0 1.05E-1 To 132 1.83EM 4.23E-2 neg. neg. neg. neg. nog. neg. 1129 4.22 E-1 4.23E-1 4.23E-1 4 J3E-1 ' 4.89E40 4.89E40 4.89E40-4.89E40 1 131 6.00E+5 3.12E +3 1.12E40 neg. neg. neg. neg.

nog, i 1 32 1.89EM 4.36E-2 neg.

neg. neg. neg. neg. neg. Xe 133 3.52Ee 1.11 EC neg. neg. neg. neg. neg. neg. Cs 134 2.26E+6 2.13E+6 1.96E*6 1 66E+6 4.09E+6 3.87E+6 3.55E+6 3.01E+6 Cs 136 1.26Ee 4.99E +3 3.84E41 2.40E-3, neg. neg. neg. neg. Cs 137 1.34E+6 1.34E4 1.33E46 1.31E+6 1.34E+7 1.34E+7 1 J3E+7 1.31E+7 Se 137m 1.27E*6 1.26E*6 1.26E+6 1.24E46 1 J7E+7 1.26E+7 1 46E+7 1.24E+7 Be 140 2.47E46 9.07EM 6.19E4e 3.05E-2 neg. neg. neg. neg. Le 140 2.85E46 1.04E6 7.15E+2 3.51E-2 neg. neg. neg. neg. Ce 141 6.34E46 1.73E+6 2.43Ee 4.91E+3 2.5M+3 6.89Et -- 9.69E44 1.96E*0 Co 144 8.25E46 7.11 E+6 5.68E4 3.64E46 5.58E+6 4.81E+6 3.84E+6 2.46E*6 Pr 143 2.54E+6 1.12Ee 1.02E+3 8.86E-2 neg. neg. neg. neg. Pr 144 8.2 5E*6 7.11 E 4 5.68E+6 3.64E+6 S.58E46 4.81E+6 3.84E+6 2.46E*6 Nd 147 7.42Ee 1.62EM 5.08E44 neg. neg. neg. neg. neg. Se 151 3.47E+3 3.47E+3 3.46E +3 3.45Ee 4.79E44 4.79EM 4.78E+4 4.76EM Eu 154 1.67Ee 1.65Ee 1.61 Ee 1.55E+5 9.19E+5 9.06E +5 8.88E+5 8.53Ee Eu 156 7.58Ee 4.68EM 7.02 E*2 1.66E-1 neg. neg. neg. neg. - Np 239 2.74E44 4.59E4 4.59E42 4.59E4 2.80E+3 2.80E+3 2.80E+5 2.80E+3 Pu 238 4.87E44 4.95E+4 5.04E+4 5.15EM 3.97Ee 3.97EM 3.96E+5 3.95EM Pu 239 3.0$E+5 3.05E+3 3.05E+3 3.05E+3 4.95EM 4.95EM 4.95EM 4.95E44 Pu 243 6.01 E+3 6.01 E6 6.01 Ee < 6.02E+3 8.00E*4 8.00EM 8.00E44 8.00E+4 Pu 241 1.58E46 1.57E+6 1.55E*6 1.5IE46 1.37E+7 1.35E+7 1.34E+7 1.31 E+7 Am 241 2.05E e 2.47E+3 3.10Ee 4.33Ee 2.08Ee 2.12 E+5 2.17E+5 2 J8E+5 De 242 7.57E e 5.85E+5 3.96EM 1.82E+5 1.75Ee 1.36EM 9.21 EM 4.30E+4 Cm 244 8.06E44 8.00E*4 7.93E*4 7.78E44 2.78Ee 2.76EM 2.74E+5 2.68E+5 3 eFuel batch 16 la projected discharge during April 1987. b Fuel batches 2-15 we5*Curles."8'd *" a9eii t 972 and A9ri d'**h* 1986 Cneg. = less then 10*

A-24 s Table A.14 Decay Heat Released from Spent Fuel Inventory for Varf ous Discharged Fuel Batches (Ginna) Decay Heat Released by Batch Date End of . Batch Size + May 1,1987 July 1, 1987 April 1,1988 Irradiation (Metric Tonnes) (Kilowatts, TEermal) 2 04/13/72 18.772 8.5 8.5 8.4 I 3 12/31/73 3.195 2.9 2.9 2.8 4 03/08/75 11.583 14.3 14.2 13.9 5 01/28/76 14.778 18.1 18.0 17.6 6 04/14/77 16.375 20.5 20.4 19.8 7 03/23/78 16.376 15.8 15.7 15.1 8 0'i/J9/79 15.976 14.7 14.5 14.0 9 03/28/80 14.378 14.7 14.5 13.7 10 04/17/81 11.183 13.7 13.4 12.4 11 01/25/82 9.586 19.0 14.6 13.2 12 03/25/83 7.988 17.2 16.5 14.2 13 03/01/84 9.186 28.6 27.1 22.0 14 02/28/85 9.985 50.9 47.2 35.3 15 03/30/86 9.586 96.1 85.8 56.5 16 04/01/87b 9.586 437.2 260.4 107.7 Totalc 2-15 331.0 313.3 259.0 Totale.2-16 768.3 573.7 366.8 aSee Table A.11. b rojected dates. P CTotals may not equal sum of entries due to rounding of decimals.

A-25 6 i cs"-.= e w r.. ~ w a o g e -. m N ~ I '.'".*.a. d.*.".*.". N N.*. ".*.9 " *. % .. ~~ -.~.--oooo ~. N.4 (M 1 g...o~m...~@~~guo~ 7

• • • • • 195**

5"."N" 8 .-~ a em~,-e=~omeone ~ 6a .. M. ~.'g @e ~a M - N O. ' g i sus". n-w: .= : u m l,. n2 m u - - ~ m o s a c e. u 8 ".3 .a. .m....@MNM S.w N.

e. 6.

.A M P. p g o. m. ~.. n. ~. Mg....... .gg omo-o. g g I .~ - ..~~~ m . g... ..S....*...... m.: ~. - -. ".... m. c o o - o. ~ 5 m g. gg L tt y E. !. !.

• N S E.. m... m o. m.

5 d*g 19*5*1N*** g .g gown.o-onom g y - -g J W I 2.* .v. 5 .R. --m.~ -~m

m.. ~.. ~.... w m. o..

1..

.~gomm E g o-omo g 3 g - 7, .%.d 1 .=a.. O N. N.e.. e P. N . - ~ g o o m

e. m.

a. o.... N W

P..d e1 .o o - o

o..

e. .g...~... e gw g g g k,. = i w v O.. N M nn. ~N~NN m. ~.A. o~o~ M RE

• g

-g afIIdada.wddAdAadasA f k. y

3. -

m ~ m --me n o N . I. !. *

• N. *.".E.

~ m u. E.. S 4 g o o m .. N '.". N"** g ~ --o g; .g E ..m. ~8. mm.. - o

o..

~. g. l e. ~. S. M~g. . P. m g y i g. "2 83 m. ~. l i.. ~.. o..w -~ ou o. ..af iddaftidadsdida f fg 1 d has m. ~ m. - ~ g.. . o.e - o o.. g g* e g.. ~.. .~m. gs .M _g o' .= n o .she -. g.. .. ~... l. e.m a--.o.~. .... N f4;, N.. ..= -. g.. ~. .~ ~o~ g go ~ 9 ... m.. 3. ~. S. o.. ...m... m. m... ~.. ~

m. e.

~ ~ v. ~--o ~ = =. =. M N TE t -g ~ a.

o. 8.~A.....o..~n.

1

2..

mmm...mm. a .--.----~~~~~.g .O O >N Q a

~ I i 9 g i o g4 R _8 I 1 9 ) 4 1 fe 8 1 R I I # I o o e R cr 2 uo 8 S I 1 ( 9 \\' 4 1 7 o 8 9 1 6 I R o 1 I 7 9 y r 0 o o t 6 v 8 I s I 9 i g h 1 R E n i t I o a 1 re 5 p O o 8 'R ,j 7 I 1 o - 9 e 1 p n 1 o 4 t s l l I i I M R I , u {1 I 1 I 6 A i he 7 e I r 9 u 1 g i F 5 0 5 O 1 1 5I32~exW2W gt2xwIF o9xm t

e o APPENDIX B s f BROOKHAVEN NATIONAL LABORATORY MEMORANOUM DATE: August 27, 1986 To: W.T. Pratt f FEM : K.R. Perkin H. Conriell susJECT: Impact of Revised Reaction Rate Equation on the Likelihood of Zirconium Fires in c Drained Spent Fuel Pool (Task 5) The SNL investigation 1 of the potential for cladding oxidation. during loss of fual pool inventory accidents has been controversial due to many unique features of the. postulated "beyond design basis accident." The purpose of the BNL investigation (FIN A-3786) has been two-fold: 1. Provide an independent assessment of several important. areas of the. phenomenological-treatment of the SFUEL code.1 2. Provide an estimate of the likelihood and consequences of the postu-lated - accidents so that the risk can be compared. to the risk of severe reactor accidents evaluated in typical PRAs. The purpose of Task 5 of FIN A-3786 was to re-evaluate the oxidation rate equation used in the SFUEL code and to perform a sensitivity study to demon-strate the influence of the reaction rate on the results of the SFUEL analy-- sis. ThIoxidation rate equation is alto a key factor which af.fects the possi-ble propagation of Zircaloy fires to low power (i.e., older) spent fuel bun-dies. The uncertainty irt, propagation calculations with SFUEL is addressed in Task 3. A -letter report summarizing the results of Task 3 is in preparation and will be submitted to the NRC, Project Manager by September 10,'1986. Discussion After an extensive review of the zirconium /Zircaloy reaction rate data (Attachment 1) and a second review of some new German data (Attachment 2)', we have concluded that the reaction rate used by Benjamin et al.1 is representa-tive of the existing data. For the purposes of the sensitivity study, we have adopted the two parameter oxidation curve suggested; by Weeks. Weeks' two parameter curve is given by: 2 w /t = 3.09 x 10s exp(-56600/RT) (1) where: w is the oxygen consumption (mg/cm2) t is time (sec) .T.is the clad temperature (K) R is the gas constant (1.987 cal /K)

} 1 ~

0' c

Memo to T. Pratt from K. Perkins and H. Connell I August 22, 1986 Page 2 i-1 Weeks' equation'is equivalent to that suggested by Benjamin i except that'it i provides a Wooth transition to the-self-sustaining oxidation regime (above 800*C) and does not put undue emphasis on the threshold effect'of a shift in oxidation rate due to metallic phase change. ( We have varied the reaction rate by a factor of four based on the data ] scatter in the temperature range of 800 to 900*C-(where self-sustaining oxida-tion is initiated). Only a slight change '(s50*C) in the initiation tempera-ture occurs for this broad range of uncertainty in the oxidation rate. This translates into an uncertainty. of 125% in the critical decay; power. We be-lieve that this insensitivity to the oxidation rate equation basically con-i firms the SNL analysis 1 for. zirconium fire initiation in a. dry spent fuel j pool. As Benjamin et al.1' pointed out, the most sensicive. parameters' for clad q i fire initiation are the decay heat level and the fuel rack geometry (related j to natural circulation flow resistance). Thus, for BWRs with low power den-sity and relatively open fuel storage racks, the critical cooling time (to ensure that air cooling will keep the fuel rods below 800*C) is about'l to 5 i months. Whereas PWRs with higher power density ' and tighter storage racks j require 2 months to 2 years (the longer time is required for the new high den- ] sity storage racks). Note that even temperatures as low as 650*C can be expected to cause clad failure and release of some fission products if the temperatures are sustained 1 over a long period (several hours). However, below 800*C the energy from oxi-dation is insufficient to significantly increase the fuel rod temperature. Conclusions We conclude that the SNL code (SFUEL) and the clad oxidation rate equa-I tion used therein accurately represents the potential for self-sustaining oxi-I dation in a drained fuel pool. The largest uncertainty appears to be due to-1 uncertainties in natural convection flows in the transition flow regime.. Changes in the storage rack configuration result in large changes in the cal-culated flow rate and correspondingly large changes in the " critical ' power level" (above which self-sustaining oxidation is predicted to occur). Based on our review of the cladding oxidation rate model and the sensi-tivity study, we conclude that the conditional probability of self-sustaining - clad oxidation and resultant fission product release, given a loss of pool in-tegrity event, is about 10% to 40% for BWRs and 16% to 100% for PWRs, depend-ing on the storage rack configuration. J In terms of power level, our sensitivity studies indicate that the criti-I cal power level (abcve which self-sustaining oxidation will occur) varies from i about 50 kW/MTU (for cylindrical racks with large openings) to 6 kW/MTU for l the new high dersity PWR fuel storage racks. j j u l 23

1

• s

'a ~ Memo to T. Pratt from X. Perkins and H. Connell q August 22, 1986 1 Page 3 Recommendations We recommend that spent fuel not be stored in high density racks until it has been stored for 2 or more years in the old style cylindrical racks with adequate coolant openings (3 or more inch diameter. holes). We also reconnend that a test progra:n be initiated to confinn the capa-bility of natural air convection cooling capability for high density storage racks. Such tests could be performed with old low power spent fuel (2 to 4 kw/MTU) and minimal instrumentation (such as thennoccuples placed near the top of the fuel bundle). References 1. Benjamin, A. S., McCloskey, D. J., Powers, D. A., Dupree, S. A., Spent Fuel Heatup Following Loss of Water During Storage," NUREG/CR-06A9,' March 1979. 2. Pisano, N. A., Best, F., Benjamin, 'A. S., Stalker, K. T., "Tha Potential for Propagation of a Self-Sustaining Zirconium 0xidation Following Loss of Water in a Spent Fuel Storage Pool," Draft Report, January 1984 KRP/HC/csc Attachments 1 e A

• i

{ l x;

o.

g..

~.,. ' Attachment 1 il l w RROOKHAVEN ' NATIONAL LABORATORY %5 . MEMORANDUM l r d DATtt: . March 27, 1985- ~ To: K. Perkins 'FROM: L. J. Teutonico t i FINA-3786-E;udyofseyondDesignBasisAccidentsinSpent sueJECT: q Puel Fools } I2 Two Sandia reports e deal wich the question of rapid zirconium 'oxida-tien in a spent fuel pool following loss of water. Both the computer modeling and tha experimental simulation. as. described in these reports, suggested that in certain ' fuel racking configurations (a) a self-sustaining sirconium-str oxidation reaction can be initiated, and (b) th'is self-sustaining. reaction can propagate from one region of a pool'to another.' There are large uncertainties associated with the phenomenology of aircaloy oxidation and its propagation in spent fuel assenhlies. This preliminary report on Tasks' 3, 4,;5 of the' subjece FIN- (Uncertainties in oxidation Propagation, SFUELN Computer Code Validation, Impact cf Revised Reaction Rate Equation, respectively) addresses some of thes's uncertainties and their effects on the-initiation and propagati5n of a self-sustaining sircaloy-air oxidation reaction. s i 1) The propagat' on ratas of rapid sircaloy clad exidation. in air from the A hottest section of fthe pool (after a loss of water incident) to adjacent ,1 sections were estimated (in Ref. 2) under the conditions: that 'the spent fuel. ~ in the hottest crect! ion of the pool was generatihg 30 kw/MTU in a room, main-s tained as constant temperature. As pointed out by Han, this estimate should I be re-calculated g under inadequate room ' ventilation conditions, to simulate ..r t properly the. condit1'ons at many licensed facilities. Similarly, additional calculations should be performed in which the hot spent fuel decay power As b varied from 20 to 90 kw/MTU for both the adequate and inadequate room ventila-tion conditions. These studies would determine how sensitive the oxidation sprop'agation is to ; the decay power of the spent fuel scored adjacent to hot i s fuel, assuming the input oxidation rate data are known with sufficient accu-j racy. a

r s s Memorandum To: K. Perkins March 27, 1985 Page 2 2) The above assumes the zirealoy-air reaction race equation used in the l Sandia work is sufficiently accurate. There are a number of uncertainties associated with this equation. We discuss each of these uncertainties in turn. A. Experimental Data: A literature search "~I" has revealed that there is a great deal of data for zirconium oxidation; most of it, however, ia con-cerned with oxidation in steam or oxygen. The data for zirconium (zircaloy)- . cir oxidation presented in Refs. 1 and 2 appear to be, the best ava11atte. These are shown in Figure 1. The. authors (of the SNL reports) fit the da:a with three separate Arrhenius plots over the temperature range 500-1500*C; one break occurs at the a-S transformation temperature for zirconium, the other at the temperature at which the oxide undergoes a monoclinic-tetragonal trans-formation. (N.B. two of the sets of data are for zirconium, the other for s zircaloy-4). These assumptions are reasonable. It should be noted, however, that there is no a priori reason to expect that the data would be fit by an Arrhenius expression, particularly above the a-8 transformation temperature where a number of different pstocesses are occurring simultaneously (discussed further below); therefore the use of the Arrhenius expression should be viawed in this case only as a computational tool. It is difficult to assess the validity of the data employed. What are really required are new experiments to determine the oxidation race of zircaloy in air over the temperature range of interest, for both isothermal lnd non-isothermal conditions. B. Kinetics: The question was raised 2 as to whether the assumption of parabolic kinetics was valid. Data were presented (from Refs. 86 and 126) which show examples of linear as well as cubic kinetics. .g However, they all apply at temperatures below the a-$ transformation temperature. Since almost all rapid oxidation occurs above the a-8 transformation temperature, where the oxidation rate is controlled by one or more diffusion processes, ;the assump-tion.of parabolic kinetics appears to be reasonable.

v: i t i o Memorandum To: K. Perkins f,' March 27,1985 s Page 3 y i i C. Zirconium vs. Zircaloy: It is assumed in the Sandia work that the i oxidation rates of zirconium and zirealoy are essentially the same. work by Pavel and Campbell 36 Recent has shown that this is not the case. Oxidation in steam of both pure zirconium and zircaloy-4 was studied in the temperature range of rapid oxidation (1000*C-1500*C). It was found that at all tempera-tures the oxidation rate of zircaloy-4 was highsr than that of zirconium; the ratio of the { two rates is approximately 3 at 1000*C and decreases with increasing temperature to a value of approximately 1.5 at 1500*C (cf. Figure 2). The higher oxidation rate of Zirculoy-4 is attributed to increased oxygen diffusivity in the oxide phase; a lower activation energy was observed, implying that some mechanistic differences exist. Analogous results are ex-pected to apply for oxidation in air. D. Oxidation Model: The oxidation in steam of both zirconium and air-caloy-4 (in the temperature range 1000-1500*C) is a multi phase layer pro-cess.136 Not only is an oxide layer formed, but also (beneath it) a layer of oxygen-stabilized o phase (zirconium or zircaloy). The multi phase model is I only significant above the a-S transformation temperature (approximately 900*C),,but this is exactly where rapid oxidation occurs. The_ parabolic rate constants for oxide layer growth, a-layer growth, ned oxygen consumption were determined in Ref.136 from experimental data and coizputer modeling. ,j The rate i of cxygen consumption is significantly higher at all temperatures than the l rate of oxide formation for both' rirconium and zircaloy-4 For zirconium the ratio of oxygen consumption rate to oxidation rate is approximately 4 at 1000*C and increases with tvreasing temperature to a value of approxim 5.4 at 1500*C; for zircaloy-4 the corresponding values are approximately 3.0 ) r! and 4.5 at 1000*C and 1500*C, respe:tively (cf. Figure 2). Although these results were obtained for oxidation in steam, analogous results are again expected for oxidation in air. 4 E. Effect of Nitrogen: Before discussing the reaction of zirconium with air, let.us consider the reaction with nitrogen alone.148-151 The rate of l _______.______._a

B r u a Memorandum To: K. Perkins March 27, 1985 Page 4 reaction of nitrogen with zirconium is much less than the corresponding reac-tion rate with oxygen; weight gain data after one hour (800*C<T(1200*C) 151 indicate that zirconium reacts with nitrogen about 20 times slower than with oxygen. The overall process is very similar tu oxidation in view of the high solubility of nitrogen in zirconium, and involves a large amount of dissolu-tion along with film formation. In the case of nitriding in the a-region, a two phase diffusion process describes the behavior wheream S phase nitriding involves three phases (nitrogsn, like oxygen, stabilizes the a phase, 11eading to a wide range of a between the nitride and the 6-matrix). The reaction product is zirconium nitride (ZrN); the reaction is exothermic, releasing approximately 82 kcal/ mole. (The energy released in forming the oxide is approximately 262 kcal/ mole.) The thickness of the zirconium nitride layer I has been found " to be much smaller than that of the dissolution zone (in the temperature range 750*C-1000*C) which indicates that the race constant for l film formation is considerably smaller than the race constant for nitrogen dissolution. In fact, at 1000*C, 84% of the total nitrogen uptake was aue to dissolution in the metal. The role of nitrogen in the high temperature reaction of zircon'nm with air has been investigated. The reaction process is multiphase in nature. Adjacent to the 8 phase of the zirconium is a layer of a phase (stabilized by both oxygen and nitrogen) and a surface .cyer of Zr0. In general, a certain 2 amount of nitride (ZrN) is foYmed. For temperatures up to approximately 1050*C the nitride is found as a layer between the stabilized a phase and the o,xide layer; above 1050*C the nitride occurs as discrete particles dispersed in the oxide. m. It is doubtful whether any appreciable amount of nitride is formed in the l l problem currently being considered. At the lower temperatures (during heat up) the reaction rate is very slow. Once rapid oxidation 13, initiated (approximately 900*C) the self-sustaining reaction proceeds very quickly, and ~"

e up Memorandum To: K. Perkins March 27, 1985 Page 5 there may not be sufficient time for ZrN to be formed. Any nitride that does form, however, will contribute to the chemical energy release for the self-sustaining reaction. The reaction rate of zirconium is higher with air than with oxygen alone. The explanation advanced is that nitrogen dissolves in Zr0. By replacing 2 oxygen ions in the oxide structure, the higher valency nitrogen can increase the anion vacancy concentration, thus permitting a higher rate of diffusion of oxygen through the anion-deficient zirconia. In sum, there are a number of uncertainties associat.ed with the ziresloy-air reaction equation. These are particularly important above 900*C where rapid oxidation occurs. The most significant appear to be (1) the difference in the oxidation rates of zirconium and zircaloy, and (ii) the multiphase nz,ture of the oxidation process itself at these temperatures. The results given above in Section C and D (i.e. for zirconium vs. zircaloy-4, and oxygen consumption rate vs. oxidation rate, respectively) apply to or.idation in steam only. Analogous results are expected for oxidation in air, i.e. it is ex-pected t, hat the oxidation rate in zircaloy will be greater than that in zir-conium, and the race of oxygen consumption will be greater than the rate of oxide formation in both materials. The relative magnitude of these effects be deduced from the steam oxidation data. cannot What are required are new experiments and computer modeling (similar to those carried out U6 by Pavel and Campbe11 for oxidation in steam) for the high temperature reaction of zir-conium and zircaloy with air. In lieu of these, we suggest that ' additional calculations be performed for two other zirconium-air reaction correlations which will serve as bounds for those v presented in Figure 1. (a) The high temperature correlation for zirconium (above the phase change of Zr0 ) 2 should be multiplied by a factor mi to account for the higher reaction rate in zirca-loy. (b) The correlations above the a-8 transformation temperature should be divided by a factor m2 to account for the difference in oxygen consumption rate and rate of oxide formation. be considerod. Values of mi and m2 as large as five should I -- D

~ 9 9 d Memorandum To: K. Perkins Page 6 March 27, 1985 REFERENCES 1.

Benjamin, A.

S., McCloskey, D. J., Powers, D. A., Dupree, S. A., " Spent Fuel Heatup Following Loss of Water During Storage," NUREG/CR-0649, March 1979. 2. Pisano, N. A., Best, F., Benjamin, A. S., Stalker, K. T., "The Potential for Propagation of a Self-Sustaining Zirconium oxidation Following Loss of Water in a Spent Fuel Storage Pool," Draf t Report, January 1984. 3. Han, J. T., Memo to M. Silberbarg, May 21, 1984 4. Dougir.ss, D. L., "The Physical Metallurgy of Zirconium," Atom. Energy Rev.1,,4(1963) 71. 5.

Lustman, B., Kerze, F., " Metallurgy of Zirconium," McGraw-Hill, NY (1955),

6.

Miller, G.

L., " Zirconium," Butterworths Science Publications, London (1954). 7.

Douglass, D.

L., "The Macallurgy of Zirconium," International Atomic Energy Agency, Vienna,1971. 8. Pemsler, J. P., " Diffusion of Oxygen in Zirconium and its Relation to oxidation and Corrosion," J. Electrochem. Soc. 105 (1958) 315. 9. Mallect, M. W., Albrecht, W. M., Wilson, P. R., "The Diffusion of Oxygen in Alpha and Eeca Zircaloy-2 and Zircaloy-3 at Electrochem. Soc. 106 (1959) 181. High Temperatures, J. -~

10. Debuigne, J.,

Lehr, P.,

"Sur la Determination des Coefficients de Diffusion de L'Oxygene Dans la Systeme Zirconium-Oxygene," C. R. Acad. Sci. Paris 256 (1963) 1113.

11. Sainforth, G., Jacquesson, R'., Laurent, P., " Diffusion de L'Oxygene Dans le Zirconium," Proc. Colloque sur la Dif fusion a l' Etat Solide," North Ho11 cud Publishing Co., Amsterdam (1959) 79.

12. Davis, M.,

Montgomery, K. R., Standring, J., "Th for Oxygen in Alpha-Zirconium," J. Inst. Metal e Diffusion Coefficient 81 (1961) 172. r

13. Cabrers, N.,

Mott, N.

F., " Theory. of Oxidation of Metsis," Rep. Prog. Phys. 12,(1948) 163. 2

14. Campbell, W.

E., Thortas. U. S., "The Oxidation of Metals," Trans. Electrochem. Soc.,9,1,(1947) 623. w#"*8**'

d Memerandum To: K. Perkins Page 7 March 27, 1985 REFERENCES (CONT'D)

15. Uhlig, H.

H., " Initial' Oxidation Rate of Metals and the Logarithmic Equation," Acts Metall..,4, (1956) 541.

16. Draley, J. E., Bradhurst, Report, ANL-7000 (1964) 223.D. H., Van Drunen, C.

J., Argonne National Lab.

17. Dawson, J.

K., Baugh, Una C., White, J. F., " Observations on the Early Stages of Oxidation cf Zirconium and Zirealoy-2," Electrochen. Technol (1966) 137. .4

18. Cox, B., "The Oxidation and Corrosion of Zirconium and its Alloys," V.

J. Electrochem. Soc._105 (1961) 24. Mechanism of Oxide Film Crowth an

19. Cox, B.,

Nucl. Mater.,2_5, (1968) 310." Low Temperature (<30G'C) Oxidation of Zircaloy-s 5

20. Culbransen, E. A.,Andrev, K.

F., " Oxidation of Zirconium Alloys in Water Vapor Atmospheres Containing Trace Amounts of OxyEen at 375 and 575'C, Electrochem.' Technol. 4 (1966) 99.

22. Britton, C.

F., Wanklyn, J. N., " Inhibition by Boric Acid of the Oxidation of Zirconium in High-Pressure Steam." J. Nucl. Mater.,5, (1962) 326.

22. Britton, C.

F., Arthurs, J. V., Wanklyn, J. N., "Further Studies on the Steam," J. Nucl. Hacer. Inhib.ition by Boric Acid of the Oxidation of Zirconium in 51.5,(1965) 263. ~~

23. Demant, J.

T., Wanklyn, J. N., " Effects of Contamination on oxidation of Zirconium in Steam," Corronion 2,2, (1966) 60. 2

24. Cox, B., Reed, J. A.,

Report AERE-K-4459 (19634" Oxidation of a Zr 2, SNb Alloy in Steam and Air,"

25. Johnson, A.

B., Jr., " Effects of Nuclear Radiation on the Corrosion, Hydriding and oxide Properties of Six Zirconium Alloys," ASTM STP 458 (1969) 301. ,.g

26. Douglass, D.

L., " Corrosion Mechanism of Zirconium and Its Alloys-III. Solute Dist ribution Between Corrosion Films and Zirconium All.ov i Substrates," Corr. Sci. 5,(1965) 347. j

27. Kass, S., Crozier, J.

D., Shubert, F. L., "Ef fects of Silicont Nitrogen, and Oxygen on the Corrosion and Hydrogen Absorption of Zircaloy-2," Corrosion,2,,0,(1964) 350t.

. _ - _ - _ _ _ _ ~ _ - - o. er Memorandum To: K. Perkins Page 8 March 27, 1985 REFERENCES (CONT'D) l

28. Culbransen, E. A., Andrew, K.

F., " Oxidation of Zirconium Between 400 and 800*C," J. Metals (1957.) 394

29. Cox, B., " Rate-Controlling Processes During the Pre-transition Oxidation of Zirconium Alloys," J. Nucl. Mater. 13_,1,(1969) 48.

30. Flint, 0., Varley, J. H.

O., "On the Mechanism of oxide-Film Formation on Zirconium," J. Phys. Chem. Solids,6,,(1958) 213.

31. Cox, B.,

High Temperature Aqueous Environments," J. Nucl. Mater." Effects of Irr 28 (1968) 1.

32. Cox, B.,

Alcock, K.,

Derrick, F.

W., "The Oxidation and Corrosion of Zirconium and its Alloys,' J. Electrochem. Soc. 108 (1961) 129.

33. Cox, B., "The Effect Kinetics of Zirconium Oxidation," Reactivity of Solidsof Fission Fragments and Neut (DE BOL1, J. H.,

Ed.) Proc. Fourth international Symposium on the Reactivity of Solids, Amsterdam (1960) 425. )

34. Asher, R.

C.,

Davies, D.,

Hall, A.,

Kirstein, B.

A.,

Harriott, J.

W.,

White, P.

J.,." Effects of Radiation on the Oxidation and Hydrogen Absorption of Zirconium Alloys in Steam," Electrochem. Technol. 4 (1966) 231.

35. Rosa,. C.

J., J, Le'ss-Common Metals,1,6, (1968) 173." Oxidation of Zirconium, A Critical Review

36. Kofstad, P.,

"Righ Temperature oxidation of Metals," Wiley and Sons, New York (1966).

37. Belle, J., Mallett, M.

W., " Kinetics of the High Temperature oxidation of Zirconium, J. Electrochem. Soc. 101 (1954) 339.

38. Wallwork, G. R., Smeltzer, W. W., Rosa, C.

J., "The Par b lic oxidation Kinetics.of Alpha-Zirconium at 850*C," Acta Metall.,12 (1964) 409 ao ,2,

39. Bradhurst, D. H., Draley, J. E., Van Drunen, C.

J., "An Electrochemical y Model for the oxidation of Zirconium," J. Electrochem. Soc. _112 (1965) 1171.

40. Charles, R.

O.,

Barnarte, C., Culbransen, E.

A., " Prolonged Oxidation of Zirconium at 350 and 450*C," Trans. AIME 212 (1958) 101.

t

_Q__n..,--- .amw. era

d Memorandum To: K. Perkins Page 9 March 27, 1985 REFERENCES (CONT'D)

41. Porte, H. A., Schniziein, J. G., Vogel, R. C., Fischer, D. F., " Oxidation of Zirconium and Zirconium Allys," J. Electrochem. Soc.

107 (1960) 506.

42. Osthagen, K., Kofstad, P.,

" Oxidation of Zirconium and Zirconium-Oxygen Alloys at 800*C," J. Electrochem. Soc. 109 (1962).204.

43. Westerman, R.

E., "High-Temperature oxidation of Zirconium and Zircaloy-2 in Oxygen and Water Vapor," J. Electrochem. Soc. 111 (1964) 140.

44. Beranger, G.,

Lacombe, P.,

" Contribution a l' etude de la cinetique de l'oxydation du zirconium-a et de la diffusion de l'oxygene dans le metal sous-jacent a l'oxyde," J. Nucl. Mater. H (1965) 190. . 45. Hillner, E., "The High-Temperature oxidation Kinetics of Zirconium it; 0xygen," Electrochem. Technol.,4,(1966) 132. Drv

46. Cubicciotti, D.,

"The oxidation of Zirconium at High Temperatures," J. Am. Chem. Soc. ],2,(1950) 4138.

47. Hussey, R.

J., Smeltzer, W. W., "The Mechanism of Oxidation of Zirconium in thn Temperature Range of 400 to 850*C," J. Electrochem. Soc. I11 (1964) 1222.

48. Debuigne, J.,

Lehr, P.,

" Corrosion seche du zirconium non allie," Corrosion of Reactor Materials (Proc. Conf. Salzburg,1962) 2 IAEA, Vienna (1962,) 105.

49. De Celas, B. Beranger, G., Lacombe, P., "L'oxydation a haute temperature.

du zirconium et de ses alliages binaires a faible teneur en element d' addition," J. Nucl. Mater, g (1969) 1.

50. Korobkov, I. I., Ignatov, D. 'v., Yevstyukhin, A. I., Y of Zirconium and Certain Alloys of its Base,"

Meta 11urgiya i Meta 11ovedenie chistykh Metallov 1 (1959) 144.

51. Mackay, T.

L., " Oxidation of Single Crystal and Polycrystalline "O Zirconium," Trans. AIME 227 (1963) 1184

52. Saur, G., Laue, H.

J., Borchers, H., "Untersuchungen uber dio oxydation von Reactor-Grade-Zirkon," Metallurgic l_5,(1961) 9. 5

53. Madeyski, A.,

Poulton, D.

J.,

Smeltzer, W.

W., "Parabolie. Oxidation . Kinetics of Alpha-Zirconium," Acta Metall. E (1969) 579.

f l' o l ? l l ) Memorandum To: K. Perkins Page 10 March 27, 1985 REFERENCES (CONT'D) i i

54. Sense, K.

A., "On the oxidation of Zirconium," J. Electrochem. (1962) 377. Soc. 109 l

55. Phainikar, C. A., Baldwin, W. M., " Scaling of Zirconium in Air," Am. Soc.

Testing Mater. Proc. Sjl (1951) 1038.

56. Caribotti, D.

J.,

Green, H.

M.,

Baldwin, W.

M., Jr., Report AECU-3013 (1955). 't

57. Beranger, G.,

Lacombe, P.,

" Contribution a l' etude de la cinetique d'oxydation du zirconium," C. r. Acad. Sci., Paris 258 (1964) 6147.

58. Nomura, S., Akutsu, C., " Oxidation Kinetics and oxide Film Breakaway of Zirconium and its Alloys at High Temperatures," Electrochem. Technol. 4 (1966) 93.

59. Debuigne, J., Lehr.

P., "Sur l'oxydation du zirconium," C. r. Acad. Sci., Paris 254 (1962) 3710. ~ )

60. Kofstad, P., "Effect of Impurities on the Defects in oxides and their Relationship to oxidation of Metals," Corrosion 2]+ (1968) 379.

61. Fassel, W. M., Jr., "The High-Temperature Oxidation of Hetals," University of Utah Progresc Report for the Period June 1 to September 26 (1952)

NP-4246.

62. Takamu'ra, J.,

Sasaki, Y.,

" Study on the Oxidation --Resistance of Zirconium," Technical Reports of the Engineering Research Institute, Kyoto University, Vol. IX, No. 6, Report No. 58 (1959).

63. Misch, R.

D., Van Drunen, C., "The Oxidation of Zirconium Binary Alloys in 700*C 0xygen for Times up to.200 Days," Proceedings of the USAEC Symposium en Zirconium Alloy Development, CEAP-4089, Vol. II (1962). j

64. Perdereau, M., Bardo11e, J., " Etude cinetique de l'oxydation de zirconium aux temperatures elevens," C, r. Acad. Sci., Paris _256 (1963) 4665.

I

65. Ronnquist, A., Shiels, S., " Reaction of Zirconium with Oxygen, Nitrogen,

- Nr I and Air, a Review of Pubitshed Data," Forsvarets Forskiningsanstalt Institution M, Stockholm Report A 14 (1966).

66. Smeltzer, W. W., Haering, R. R., Kirkaldy, J. S.,

Short Circuit and Lattice D!,.' fusion of Oxygen," Acta Metail. j[ {1961) 880." i 1 d

h Memorandum To: K. Perkins Page 11 . March 27, 1985 REFERENCES (CONT'D)

67. Levitan, J.,

Draley, J. E., Van Drunen, C. J., " Low-Pressure Oxidation of Zirconium," J. Electrochem. Soc. 114 (1967) 1086.

68. Nierlich, J., Paidassi, J., "Sur la Cinatique de la Reaction du Zirconi avec l'oxygene sous -basces pressions aux temperatures elevees, C.

r. um Acad. Sci., Paris 267 (1968) 1429.

69. Paidassi, J.,

Nierlich, J.,

"Sur les phenomenes de germination et de croissance de la sircone, ainsi que de striation du sitconium, au cours de la reaction de ce metal avec l'exyEene sous basses pressions, aux temperatures elevees," C. r. Acad. Sci., Paris _267-(1968)'1085.

70. Beranger, G., Lacombe, P.,

"Phenomenes de germination et de croissance au Metallurgic, Presses Universitaires de France- (1966) 181.cou

71. Ranucci, L., Langeron, J.

P., "Oxydation du zirconium en phase 8 sous tre barses pressions, aspect superficial presente par les alliages s sirconium-oxygene," J. Nucl. Macer. 32,3,(1967) 79. I

72. Perdereau, M., Bardo11e, J., " Etude micrographique de l'oxydation manag

~ du zirconium," C. r. Acad. de Sci., Paris ee 257 (1963) 2477.

73. Pensler, J.

P., " Studies on the Oxygen Cradients in oxidizing Metals " J Electrochem. Soc. 113 (1966) 1242.

74. Hussey, R.

J., Smeltzer, W. W., "The Oxidation Kinetics of Zirconium in the Temperature Range 400-600'C," J. Electroches. Soc. JJf (1968) 564

75. Hagel, W.

C.,

Rosa, C.

J., " Oxygen Diffusion in Hypostoichiometric 2 Zirconium oxide in the Temperature Range of 875 to 1050'C, Trans 4 (1968) 1293. . AINE 242 ~

76. Rosa,

~ Zirconium Dioxide," J. Nucl. Mater. E (1968) 12.C. J., Hagel, V. C., ~0xygen-ion Diffusivity in Hypostoichiometric

77. Rosa, C.

J., "Evaluetton.of Oxidation Rates for J. Less-commons Metals 1,5 (1968) 35. Alpha-Zirconium," 5

78. Rosa, C.

J.,

Smeltzer, W.

W., "The Parabolic 0xidation Kinetics of Beta-Zirconium," Acta Metall.,1_3. (1965) 55.

79. Douglass, D.

L., " Corrosion Mechanism of Columbium, Alloys: 1. Diffusion Zirconium, and their Dioxide," Corrosion of Reactor Materials (Prnc. Conf. Salzburgo IAEA, Vienna (1962) 223. 1962) 2 ~~, )

o i . Memorandum To: K. Perkins Page 12 March 27, 1985 REFERENCES (CONT'D)

80. Douglass, D. L., Wagner, G., "The Oxidation of Oxygen-Deficient ~ Zirconia:

and its Relationship to the Oxidation of Zirconium," J. Electroches. Soc.' 113 (1966) 671. i

81. Smith, T., " Diffusion Coefficients and Anion Vacancy Concentrations for the Zirconium-Zirconiun Dioxide System," J. Electrochem. Soc. j)ll (1965)'

560.

82. Aronson, S.,

" Oxidation and Equilibrica in Nonstoichiometric Zirconium Dioxide Pcvder," J..Electrochen. Soc. 108 (1961) 312. i

83. Cox, B.,

Pensler, J. P., " Diffusion of Oxygen in Growing Zirconia Films," J. Nucl. Mater. 2,8,(1968) 73. 8

84. Madeyski, A., Smeltzer, W. W., " Oxygen Diffusion in Monoclinic Zirconia,"

Mater. Res. Bull.,3,(1968) 369.

85. Keneshea, F. J., Douglass, D. L., "The Diffusion of oxygen in Zirconia as a Function of Oxygen Pressure," 0xidation of Metals 3_ (1970) 1.

l

86. Hallett, M. ' W., Albrecht, W.

M., "High Temperature Oxidation of Two Zirconium-Tin Alloys," J. Electrochen. Soc. 102 (1955) 407.

87. Poulton, D.

J.,

Smeltzer, W.

W., "0xygen Diffusion in Monoclinic Zirconia," J. Electrochoes. Soc. 117 (1970) 378.

88. Douglass, D.

L., Van Landuyt, J., "The oxidation.of Zirconium: an l j Electron Microscopy Study of Zirconia Formed in Acta Mes 11. 1,3 (1965) 1069. the Thih-Film Region," i 3

89. Baranger, G.,

Duffaut, F.,

De Gelas, B.~, Lacombe, P., " Observations' en microscopic electronique de films minces de zircone obtenus par oxydation-du zirconium," J. Nucl. Hacer. J9, (1966) 290.

90. Missiroli, G.

F., " Electron Microscopic Observation of 0xidized Zirconium Films," J. Nucl. Hacer.,2_5,(1968) 93.

91. Douglass, D.

L.. Van Landuyt. J., " Transmission Electron Microscopy of - t Zirconium and Zirealoy-2 corrosion Films Formed in Air and Water," Proc. T; IXe Colloque de Metallurgic, Presses Universitaires de France (1966) 155.

92. De Gelas, B.,

Beranger, G.,

Lacombe, P.,-

" Etude par microscopic electronique en transmission de films minces d'oxyde obconus par oxydation d'alliages de. zirconium a teneur en element d' addition infericare a 4% en - poids," J. Nucl. Macer. 28,(1968) 185. h

o .{ Memorandum To: K. Perkins Page 13 March 27, 1985 i REFERENCES (CONT'D) 93. Bailey, J. E., "On the Oxidation of Thin Films of Zirconium," J. Nucl. Mater. 8 (1963) 259. 94

Bailey, J.

E., "The ' 0xidation of Thin Films of Zirconium," Electron Microscopy, 1964,; Vol. A.. Non-Biology,. (Titibach,.M., Ed.) Publ. House Czech. Sci., Prague (1964) 395. 95.

Denoux, M.', Trillat,.J. J., "Sur l'obtention de couches minces orientees de airconium et leur oxydation " C. r. Acad. Sci., Paris 258 (1964) 4683.

96. Denoux, M., "Sur l'oxydation de. couches minces orientees de zirconium," C. r. Acad. Sci.. Paris 260 (1965) 5003. 97.

Denoux, M.,

Trillat, J.

J., " Etude de l'oxydation de couches minces orientees de zirconium; relation-d'opitaxie oxyde-metal," Proc. IXe Colloque de Metallurgic, Presses Universitaires de France (1966) 173.. 98.

Vahldiak, F.

W., "Epitaxial Oxidation of Zirconium at Various Tear-peratures," J. Less-Common Metals, 12,(1967) 19. 99. Gulbransen, E. A., Andrew,- K. F., " Oxidation of Zircaloy-2 and -3A at 300 to 850*C," Trans. AIME 212 (1958) 281. 100. Greenbank, J. C., Harper, S., " Solute Distribution in oxidized Zirconium Alloys," Electrochem. Technol. 4 (1966) 142.- 101. Cox,. B., "0xidation 'and Corrosion of Zirconium and its Alloys," Corrosion 16 (1960) 380t. 102. Cox, B., "Causes of a Second Transition Point Occurring During Oxidation of Zirconium Alloys," Corrosion J,8,(1962) 33t. 103. Peneler, J. P., "The Kinetics and Mechanism of oxida Film Growth Zirconium," Electroches. Technol. 4 (1966) 128, on 104. Greenbank, J. C.,

Harper, S., "The Mechanism of Breakaway 0xidation in Zirconium Alloys," Electroches. Technol. 4 (1966) 88.

105. Douglass, D. L.', " Oxide Plasticity in the oxidation Mechanism of m. ~ Zirconium and its Alloys," Corrosion Sci. 5 (1965) 225.. 106. Ke: 4, L. 'H.,

Beranger, G.,

De'.Celas, B.,

Lacombe, P.,

. Ett$de ) .tierographique du processus de desquamation (" breakaway") l'oxydation du zirconium," J. Less-Common Metals au cours de 14 (1968) 18h l j m

3 Memorandum To: K. Perkins Page 14 March 27, 1985 REFERENCES (CONT'D) 107. Roy, C. Burgess, B., "A Study of the Stresses Generated in Zirconia Filc:s During the Oxidation of Zirconium Alloys," 0xidation of Metals 2 (1970). 235. 108. Cox, B., " Processes occurring-During the Breakdown of 0xide Films on Zirconium Alloys," J. Nucl. Macer. R (1969) 50. 109. Gulbransen, E. A., Andrew,. K. F., " Breakaway Oxidation of Zirconium-Tin Alloys," Corrosion H (1958) 32t. 110. Culbransen, E. A., Andrew, K. F., " Kinetics of the Reactions of Zirconium with 0, N2 and H," J. Metals,J,(1949) 515. 2 2 111. Culdner, W. C. and Wooten, L. A., " Reactions of Zirconium with Cases at Low Pressure," Trans. Electrochem. Soc., 93: 223-235 (1948'). 112. Cubicciotti, D., "The oxidation of Zirconium at High Temperatures," J. Am. Chem, Soc., 72: 4138 (1950). i 113. Hayes, E. T. and Roberson, A. H., "Some Effects of Heating Zi :enium in Air, Oxygen, and Nitrogen," Trans. Electrochem. Soc., 96: 142 (1949). 114. Hallett, M. W.,

Nelson, H.

R., and Papp, C. A., "The Reaction of Zirconium Metal Surfaces with Oxygen," Report BMI-727, 1952. 115. Thomas, D. E. and Chirigos, J., " Oxidation of Zirconium and Relation to Righ-Temperature Corrosion in Water," Report WAPD-98, Oct.15,1953. 116. Chirigos, J. and Thomas, D. E., "The Mechanism of Oxidation and Corrosion of Zirconium," Proc. AEC Metallurgy Conf. March 1952, p. 337 Report TID-5084; also Report WAPD-53. 117. Porte, H. A., Schniziein, 7J. G.,

Vogel, R.

C., and Fischer, D. F., I " Oxidation of Zirconium and Zirconium Alloys," Argonne National Laboratory, ANL-6046, September 1959. 118. Biederman, R. R., Ballinger, R. C., and Dobson, W. C., "A Study of Zircaloy 4-Steam Oxidation Reaction Kinetics," EPRI NP-225, September T 1976. 119. Biederman, R. R. B., Sisson, R. D.,. Jr., Jones, J. K., and Dobson, W. C "A Study of Zircaloy 4-Steam Oxidation Reaction Kinetics," EPRI NP-7'l4 April 1978. O j .}}