Text

Forwards Draft Beyond DBA in Spent Fuel Pools (Generic Issue 82). Formal Nureg/Cr Will Be Completed After NRC Review
	ML17347B682
Person / Time
Site:	Vermont Yankee File:NorthStar Vermont Yankee icon.png
Issue date:	02/05/1987
From:	Perkins K; BROOKHAVEN NATIONAL LABORATORY
To:	Throm E; Office of Nuclear Reactor Regulation
Shared Package
ML17347B681	List: ML17347B680; ML17347B682;
References
	CON-FIN-A-3786, REF-GTECI-082, REF-GTECI-NI, TASK-082, TASK-82, TASK-OR NUDOCS 8704060324
	Download: ML17347B682 (188)
	v • d • e

GLIB Deportment of Nuclear Energy BROOKHAVEN NATIONALLA8OPATORY ASSOCIATED UNIVERSITIES, INC.

Upton, Long Island. New York 1 "973 (516) 282'147 FTS 666'ebruary 5,

1987

'r.

Edward Throm Division of Safety Review and Oversight U.S. Nuclear Regulatory Commission Phillips Bldg., Hail Stop 244 7920 Norfolk Avenue Bethesda-,

HD 20814 RE:

FIN A-378S

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 wganizations within BNL.

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, CC A. Benjamin, SNL H. Connell W.T. Pratt H. Reich V.L. Sailor T. Teichman A. Tingle J.

Weeks R.A. Bari (w/o encl.)

W.Y. Kato ( "

)

K.R. Perkins, Group Leader Containment 5 Systems Integration Group

A

~

y I

0

NUREG/CR-BNL-NUREG-BEYOND DESIGN-BASIS ACCIDENTS IN SPENT FUEL POOLS (GENERIC ISSUE 82)

V.L. Sailor, K.R. Perkins, and H. Connell Containment E Systems Integration Group and J.

Meeks Materials Safety Application Group Department of Nuclear Energy Brookhaven National Laboratory

Upton, New York 11973 (DRAFT)

January 1987 Prepared for Division of Safety Review and Oversight Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Mashington, DC 20555 Contract No. DE-AC02-76CH00016 FIN A-3786

l

~ ~

s

~ ~

ABSTRACT This investigation has provided an integrated assessment of the risk of beyond design basis accidents in spent fuel pools for two surrogate plants (a

PMR and BMR).

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 consequences were found to be about 12 person-rem/Ry and 130 person-rem/Ry for the BMR and PMR

plants, respectively.

These. estimated risk re'suits 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 is 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.

g ~

\\

~ ~

~

1 TABLE OF CONTENTS ABSTRACT.................

LIST OF TABLES...........

LIST OF FIGURES..........

ACKNOWLEDGEMENTS.. ~... ~..

SUMMARY

.oooooooo. ~ oo. ~

os

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Paae ill 1x xii i xv S-1 1 ~

INTRODUCTION ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~

o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

1.1 1.2 1.3 1.4 1.5 1.6 1.7 Previous Investigations.................,".....

Related Events............................""

Risk Potential.............................-...

Discussion of Spent Fuel Storage Pool Designs Selection of Surrogate Cases for More Detailed Report Content................................

References of Section 1.......................

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

and Features......

Studieso

~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~

1>>1 1-3 1-4 1-4 1<<5 1-6 1-6 2 ~

ACCIDENT INITIATING EVENTS AND PROBABILITY ESTIMATES.................

2-1 2.1 Loss of Water Circulating Capability............................

2-1 2.2 Structural Failure of Pool......................................

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............

2.2.1.2 Seismic Hazard Estimates for the Millstone and Ginna Sites............................

2.2.1.3 Seismic Fragility of Pool Structures.......

2.2.1.4 Seismically-Induced Failure Probabilities..

~ ~ ~ ~ ~

~. ~ ~ ~ ~

~ ~ ~ ~ ~

2-8 2-9 2-10

~ ~ ~ ~ ~ 2 4 2.2.2 Structural Failures of Pool Due to Missiles.........

2.3 Partial Draindown of Pool Due to Refueling Cavity Seal Fai lures

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

2.4 Pool Structural Failure Due to Heavy Load Drop.............

2.5 Summary of Accident Probabilities..........................

2.6 References for, Section 2...............".""."""."".

~ ~ ~ ~ ~

2-12 2-16 2-18 2-19

~ ~ ~ ~ ~ 2~12 3.

EVALUATION OF FUEL CLADDING FAILURE""......."".""".".".." 3-1 3.1 Summary of SFUEL Results........................................

3-1 3.1.1 Summary Model Description...""."."."

3.1.2.

Clad Fire Initiation Results..........."

3.1.3 Clad Fire Propagation........'. """""

~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

3-1 3-2 3~3 3.2 Validation of the SFUEL Computer Code...........

3.3 Conclusions Regarding SFUEL Analyses............

3.4 References for Section 3........................

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

3-10 3-12 3-13

g ~

~

~ ~

~Pa e

4.

CONSEQUENCE EVALUATION............,......,,...........,...,....,,....

4.1 Radionuclide Inventories.....................,.....,.......,....

4.2 Release Estimates'.............................,........,.....,..

4-1 4-1 4-1 4.2.1 Estimated Releases for Self-Sustaining Cladding Oxida-tion Cases (Cases 1 and 2)............""'""""".".

4.2.2 Estimated Release for Low-Temperature Cladding Failure (Cases 3 and 4) ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

4-2 4-4 4.3 Off-Site Radiological Consequences....;......."."..""""."4-5 4.3.1 Scenarios for Consequence Calculations..........."......

4-5 4.3.2 Consequence Results......................................

4-5 4.4 References for Section 4................................ "... ". 4-6

~

RISK PROFILEo ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

5 5.1 Failure Frequency Estimates.....................................

5-1 5.1.1 Spent Fuel Pool Failure Probability......................

5-1 5.1.2 Spent Fuel Failure Likelihood............................

5-2 5.2 Conclusions Regarding Risk......................................

5.3 References for Section 5...............".... "..""""...""

6.

CONSIDERATION OF RISK REDUCTION MEASURES... ~ o o" e. o o""". "."""

5-2 5-2 6-1 6 1 Risk Prevention

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~

6.2 Accident Mitigation.........................."......"".......

6.3 Conclusions Regarding Preventive and Mitigative Measures........

6.4 'eferences for Section 6.............."."" "."""".""."

6-1 6-2 6-3 6-3 APPENDIX A - RADIOACTIVE INVENTORIES~ ~ ~ ~ ~ ~. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~. ~.. ~ ~. ~ ~. ~ ~. ~ ~.....

A-1 A.1 INTRODUCTION..~.....~....... ~ o o.o. ~ o.. o.. ~ ~.o. o o ~..o. ~ o o.""o" A.2 SIMULATION OF OPERATING HISTORIES ~ ~ ~. ~ ~. ~ ~ ~ ~ ~ ~. ~ ~. ~ ~"~ ~~....".

~ ~ ~ ~

A-1 A-1 A.2.1 A.2.2 A.2 3 Thermal Energy Production vs Time........................

Fuel Burnup Calculations.................................

Calculation of Radioactive Inventories...................

~ ~ ~ ~

A-1 A-1 A-3 A.3 DATA FOR MILLSTONE 1

~ """"""o ~"~. ~." ~

~ ~"".."""."".A-4 A.3,1 A.3 2 A 3.3 A.3.4 A.3.5 Reactor and Fuel Cycle Parameters........................

History of Operations....'................................

BMR Fuel Assembly Model Used in ORIGEN2 Calculations.....

Calculated Radioactive Inventories......................,

Decay Heat................".."."."."................

~ ~ ~ ~

~ 0 ~ ~

~ ~ ~ ~

A<<4 A-4 A-5 A-6 A-6

~Pa e

A.4.1 A.4.2 A.4.3 A.4.4 A.4.5 Reactor and Fuel Cycle Parameters............................

History of OPerations................."""-"".-..........

PMR Fuel Assembly Model Used in ORIGEN2 Calculations.........

Calculated Radioactive Inventories............ """"...""

Decay Heat...................................................

A.4 DATA FOR GINNA."..........,...,...,....,.............,...,...,,....

J A-7 A-7 A-7 A-8 A-8 A-9 A.5 REFERENCES FOR APPENDIX A. ~.. ~. ~.................

~ ~.................

A-9 APPENDIX 8 - IMPACT OF REVISED REACTION ON THE LIKELIHOOD OF ZIRCONIUM FIRES IN A DRAINED SPENT FUEL POOL...... ~....... ~.....~.....

8-1 8 e I INTRODUCTION~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Bo2 DISCUSSIONS ~ ~ ~ ~ o ~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

8' CONCLUSIONS ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Be4 RECOMMENDATIONS ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~

8.5 REFERENCES

FOR APPENDIX 8'........

~.......

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

8-1 8-1 8-3 8-3 8-4

~

~ ~

Table LIST OF TABLES

~Pa e

S.1 2.1 2.2 243 2'

2.5 Estimated Risk for the Two Surrogate Spent Fuel Pools from the Two Dominant Contrfbutors......................................;..

S-6 Data on Spent Fuel Basins (as of December 31, 1984)...............

1-8 Typfcal Spent Fuel Pool Dimensions and Mater Inventories..........

2-24 Decay Heat as a Function of Time Since Last Refueling (Data f,rom Appendix A).........................;........................

2-24 Examples of Thermal-Hydraulic Transient Parameters, Assuming Complete Loss of Pool Coolant Circulation.........................

2-24 Fragility Parameters Assumed in This Study for Spent Fuel Storage Pools.....................................................

2-25 Meightfng Facto>s Assigned to the Various Hazard and Fragility Curves for the Millstone Case.....................................

2-25 2.6 Events in Which'nflated Seals Have Failed........................

2-26 2.7 2.8 3'

3.2 Estimated Distribution of Human Error fn Heavy Crane Operations...

2-27 Assumptions Used fn Calculating the Hazard of Catastrophic Struc-tural Damage to Pool Resulting from the Drop of a Shipping Cask...

2-28 Summary of Estimated Probabilities for Beyond Design Basis Acci-dents fn Spent Fuel Pools Due to Complete Loss of Water Inventory. 2-29 Summary of Critical Conditions Necessary to Initiate Self-Sustainfng Oxidation..............................................

3-14 Summary of Radial Oxidation Propagation Results for a High Densftiy PMR Spent Fuel Rack with a 10 Inch Diameter Inlet and Perfect Ventilation...............................................

3-15 3.3 Summary of Radial Oxidation Propagation Results for a Cylin-drical PMR Spent Fuel Rack with a 3 Inch Diameter Hole and Perfect Ventilation........................."" "."""..""..3-16 3.4 3.5 3 ~ 6 Summary of Radial Oxidation Propagation Results for a Cylin-drical PMR Spent Fuel Rack with a 1.5 Inch Diameter Hole and Per feet Ventilatfon......................""."""-.""""".. 3-16 Summary of Radial Oxidation Propagation Results for Various PMR Spent Fuel Racks with No Ventilation.............,..

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

3 17 Comparison of SNL Small Scale Oxidation Tests to Calculations with CLAD ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 3~18

Table 4.1 4.2 Comparison of Radioactive Inventories of Equilibrium Core with Spent Fuel Assemblies for Selected Isotopes (Millstone 1).........

Estimated Radtonuclide Release Fraction During a Spent Fuel Pool Accident Resulting in Comple'te Destruction of Cladding (Cases 1 and 2) ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~...

~Pa e

e 4-8 4-9 4.3 Estimated Releases of Radionuclides for Case 1 in Which a Zirconium Fire Propagates Throughout the Entire Pool Inventory (Morst Case)

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~. ~.. ~ ~ ~ ~ 4-10 4.5 4.6 4.7 5.1 A.1 A.2 A.3 A.4 A.5 A.6 A.7 A.8 A.9 A.10 Estimated Releases of Radionuclides for Case 2 in Which Only the Last Discharged Fuel Batch Suffers a Zirconium Fire...............

Estimated Releases of Radionuclides for Cases 3 and 4 in Which Low-Temperature Cladding Failures Occur.................."....".

Comparison of Radioactive Inventories of Equilibrium Core with Spent Fuel Assemblies for Selected Isotopes (Ginna)...............

CRAC2 Results for Various Releases-Corresponding to Postulated Spent Fuel Pool Accidents with Total Loss of Pool Mater...........

Estimated Risk for the Two Su> rogate Spent Fuel Pools from the Two Dominant Contributors.........................................

Reactor and Fuel Cycle Parameters for Millstone 1.................

Summary of Operational Milestones for Millstone 1.................

Su+nary of Spent Fuel Batches in Millstone 1 Storage Basin (Mith Projections to 1987)................................,.....,.

Comparison of Cumulative Gross Thermal Energy Production with Calculated Fuel Burnup from Start of Operations in 1970 to April 1, 1987 (Millstone 1)............................,..........

Comparison of Radioactive Inventories of Reactor Core and

Spent, Fuel Basin (Millstone 1)......................................,...

Comparison of Radioactive Inventories of Most Recently Dis-charged Fuel Batch (Batch 11) with Longer Aged Discharged Batches (Batches 1-10) (Millstone 1).........,...........,........

Decay Heat Released from Spent Fuel Inventory for Various Dis-charged Fuel Batches (Millstone 1)............,........,........,.

-= Radionuclide Contributions to Decay Heat for Various Spent Fuel Batches (Millstone 1)........,..................................

Reactor and Fuel Cycle Parameters for Ginna.......................

Summary of Operational Milestone for Ginna........................

4-11 4-12 4-13 4-14 5-4 A-11 A-12 A-13 A-14 A-15 A-16 A-17 A-18 A-19 A-20

Table A.ll A.12 Paoe Summary of Spent Fuel= Batches in Ginna Storage 'Basin (With Projections to 1987)...............;..............................

A-21 Comparison of Radioactive Inventories in Reactor Cor'e and Spent Fuel Basin (Gfnna).........'.......................................

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-charged Fuel Batches (Ginna)..............'........................

A-24 A.15 Radionuclide Contributions to Decay Heat for Various Spent Fuel Batches (Ginna)...................................................

A-25

/ 'i

~

~ ~

x$ i

LIST OF FIGURES F1oure Paoe 2.1 Seismic Hazard Curve for the Millstone Site.........'..............

2-30 2.2 The 15, 50 and 85 Percentile Hazard Curves for the Millstone ite. ~ ~ eee

~ e. ~.e

~ e ~.eee.e'...e...........e...........e.e..e

".ee..e S

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 Seismic Hazard Curve for Ginna.................................;..

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 I).......................;..............;..............

2-36 2.8 Cross Section of a Typical Pneumatic Seal.........................

2-37 2.9 2.10 3.1 Cross Section of Inflated Pneumatic Seal Seated in the Reactor Vessel Flange and Inner Surface of Cavity Wall..................;.

2-38.

Uninflated Pneumatic Seal with Steel Hold-down Ring...............

2-39 Comparison of CLAD to SNL data for Test 4....................,....

3-19 A.1 Millstone 1:

Operating history 1976-1984.................,...,.

A 26

xiv

xv ACKNOWLEOGEMENTS This work was performed for the Reactor Safety

.I'ssues Branch of the Division of Safety Review and Oversight, NRR/NRC.

The NRC Managers for the

'rogram were Mr. E. Throm and Or. H. Wohl who prov'ided 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 A8H) 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.

Or. 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.

Hr.

A.

Aronson implemented the ORIGEN2 code and provided the calculations for spent fuel pool fission product inventories for the actual discharge histories.

The authors are especially grateful to Hs. S. Flippen for her excellent.

typing of this report and for cheerfully accepting the numerous additions and revisions to this manuscript.

xvi

S-1 BEYOND DESIGN-BASIS ACCIDENTS IN SPENT FUEL POOLS (GENERIC ISSUE 82)

SUMMARY

S. 1 INTRODUCTION Generic Safety Issue 82, "Beyond Design Basis Accidents in Spent Fuel Pools,"

was assigned MEDIUM priority in Novembe'r 1983.~

In its prioritiza-tfon, the NRC staff took account of two factors that had not been considered in earlier risk assessments:z 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, A

2.

A theoreti cal model."

suggested the possibi1 ity of catastrophi c Zircaloy fire, propagating from assembly to assembly in the event of complete drainage of water from the pool.

S.l.1 Previous Investi ations The Reactor Safety Study (which did not take account of the two factors above) concluded that the risks associated with spent fuel storage were ex-tremelyy 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.a." inves-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.

Calculations with SFUEL indicated that, for some storage configurations and decay times, the Zircaloy cladding could reach temperatures at which the

S-2 exothermic. oxidation would become self-sustaining with resultant destruction 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 become involved.

Cladding fires of this type could occur at, temperatures well below the melting point of the UO> fuel.

The cladding,ignition point is about 900'C compared to the fuel melting point of 2880'C.

S.1.2 Related Events There is no case on record of a significant loss of water inventory from a domestic, commercial spent fuel storage pool.

However, one recent incident occur ed 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 open, or when transfer of a spent 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 Nore recently a

pneumatic seal failure in the Hatch spent fuel basin which released appr oximately. 141, 000 gallons of water resulted in a drop in water level in the pool of about five feet.~

S.1.3 Re ort Ob ectfve-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 ripks are defined in terms of

S-3 the probabilities of various initiating events that might compromise the structural integrity of the pool or its cool'ing capability, the probability of a system failure, given an initiating event, fuel failure mechanisms, given a system failure, potential radionuclide releases, 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.l.4 S ent Fuel Stora e Pool Desi ns The configurations of spent fuel storage pools vary from plant to plant.

In BWR's, the pools are located 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 above the top of the pressure 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 located 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 wei rs separate different sections of the tr ansfer and stor-age systems.

Nore details concerning various configurations are given in Section 2.3 and Table 1.1.

S.1.5 Selection of Surro ate Cases for Nore Detailed Studies Two "older vintage" plants were'selected to serve as BWR and PWR surro-

, gates for more detailed 'studies.

The choices, Hillstone 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

gl I ~

S-4 operating histories

. of the two surrogate plants were modeled to obtain a

realistic radioactive inventory in the various spent fuel batches.

S.2 ACCIDENT INITIATING EVENTS AND PROBABILITY ESTIHATES Accident initiating events that have been considered include pool heatup due to loss of cooling water circulation capability, structural failure of pool due to seismic events or missiles, partial draindown of pool due to pneumatic seal failure, and structural failure of pool due to a heavy load drop.

Estimates of the likelihood for each of these initiators are provided in

.Section

2. It is concluded that the dominant initiators are structural fail-ures resulting from a

seismic events

(-2xl0- /Ry) and heavy load drops

(-3xlO-s/Ry).

Uncertainties in the probability estimates are quite large, being at least an order of magnitude in either direction.

In the case of seismic events, the seismic hazard and structural fragilities both contribute to the uncertainty range.

For heavy load drops, human error probabilities and I

structural damage potentials are the primary sources of uncertainties.

S.3 EVALUATION OF FUEL CLADDING FAILURE The SFUEL computer code developed at Sandia National Laboratories (SNL) by Benjamin et al.,s 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 range of decay he'at 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, SFUELIM."

The BNL evaluations of SFUELIM have led to the conclusions that the modi-fied code gives a reasonable estimate of the potential for propagation of a

cladding fire from high power to low power spent fuel and that the code pro-vides a valuable tool for assessing the likelihood of a catastrophic fire for a variety of spent fuel configurations iri the event that the pool is drained.

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 Radioactive Inventor)es The radioactive inventor'ies contained in the spent fuel pools (as of April 1987) for Millstone 1

and Gfnna were calculated using the ORIGEN2 com-puter

code, based on the operating 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.

5.4.2 Release Estimates Fractional releases for various groups of radionuclides were estimated based on the physical parameters characterizing the SFUEL fai lure scenario.

Thus, four source terms were estimated corresponding to the four accident scenarfos.

S.4.3 Off-Site Radf olo ical Consequences

/

Off-site'radiological consequences were calculated using the CRAC2 com-puter code.s Because

'of. several features fn the health physics modeling in the CRAC2 code, the population dose results appear to be of limited value.

The most meaningful measure of the accident severity appears to be the inter-diction area (contamfnated land area) which in the worst cases was about two orders of magnitude greater than for core-melt accident.

No "prompt fatali-ties" 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 obtafn the risks which are summarized.fn Table S.l.

As noted

above, the population dose results are of limited value because they are driven by decontamination levels assigned within the CRAC2 code.

Thus the

S-6 land interdiction area is included in Table 5.1 as a

more meaningful repre-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 hazard.

Table S.1 Estimated Risk for the Two Surrogate Spent Fuel Pools

~

from the Two Dominant Contributors Accident Initiator Seismic induced PWR pool failure Seismic induced BWR pool failure Cask drop* induced PWR pool failure Cask drop* induced

. BWR pool failure Spent Fuel Pool Fire Probability/Ry 1.6x10 s

1.8x10-~

3.1x10-s 2.5x10-s Health Risk (Man-rem/Ry) 37 71 Interdiction Risk (Sq. e./Ry) 8.4x10 4

7.6x10-s

~001 1.1x10

<<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 CRAG 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.

~

e S-7 S.6 CONSIDERATION OF MEASURE WHICH MIGHT REDUCE CONSEQUENCES A number of potential preventive and 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.

S.7 References for Summar 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 (SAND77-1371),

May 1979.

4.

N.A. Pisano, F. Best, A.S. Benjamin and K.T. Stalker, "The Potential for Propagation of a Self-Sustaining Zirconium Oxidation Following Loss of Mater 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.)

5.

IE Bulletin No. 84-03:

"Refueling Cavity Water Seal," U.S. Nuclear Regu-latory Commission, Office of Inspection and Enforcement, August 24, 1984.

0 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 8.

A.G. Croff, "ORIGEN2:

A Versatile Computer Code for Calculating the Nuclide Composition and Characterlstlcs of Nuclear Materials,h Nuclear

~33 l, 2 1. 62, pp. 333-332, 3

p 2

l333.

9.

L.T. Ritchie, J.D. Johnson and R.H. Blond, Calculations of Reactor Acci-dent Conse uences Version 2, CRAC2:

Com uter Code User's Guide, prepared by Sandia National t.aboratories for the U.S. Nuclear Regulatory Commis-

sion, NUREG/CR-2326 (SAND81-1994), February 1983.

1-1 1.

INTROOUCTION Generic Safety Issue 82, "Beyond Oesign Basis Accidents in Spent Fuel Pools,"

was assigned NEOIUtl priority'n November 1983.~

In its prioritiza-tion, the NRC staff took account of two factors that had not been considered in earlier risk assessments:z 1.

Spent fuel is currently being stored rather than shipped for repro-cessing or repository disposal, resulttng in much larger inventories of spent assemblies in reactor fuel basins than had previously been anticipated;

and, 2.

A theoretical model suggested the possibility of catastrophic Zi rca-loy fire, propagating from assembly to assembly in the event of com-plete drainage of water from the pool.

1.1 Previous Investi ations The Reactor Safety Study~

(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, 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 assemblies, and The storage structures were designed to accommodate the forces and missiles generated by violent storms.

Probabilities of pool failures due to external events (earthquakes, mi s-siles) or heavy load drops were estimated to be in the range of 10- /year.

1-2 Radioactive release estimates were based on melting of 1/3 of a"core for var-ious decay

periods, with and without filtration of the building atmosphere (see Ref. 2, Table I 5-2).

Subsequent to the Reactor Safety Study, A.S. Benjamin et al.

1nvestigat-ed the heatup of spent fuel following drainage of the pool.

A computer code,

SFUEL, was developed to analyze, thermal-hydraulic phenomena occur ring when storage racks and spent assembl1es become exposed to afr.

The computer model takes into account decay time, fuel assembly

design, storage racks

design, packing dens1ty, room ventilation and other variables that affect the heatup of the fuel.

Calculations with SFUEL indicated that, for some storage configurations

.and decay times, the Zircaloy cladding could reach temperatures at wh1ch the exothermic oxidation would become self-sustaining with resultant destruction of the cladding and fission product release.

The possibility of propagation to adjacent assemblies

{f.e., the cladding would catch fire and burn at a hot enough temperature to heat neighboring fuel assemblies to the fgnitfon point) was also identif1ed.

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 U02 fuel.

The cladding ignition point is about.

900'C compared to the fuel melting point of 2880'C.

Uncertainties fn the SFUEL calculations were primarily attributed to un-certaint1es in the zirconium oxidation rates.

Further work was done to refine the SFUEL computer model and to compare calculated results w1th experimental data.4 These more recent results have generally confirmed the ear lfer concepts of a Zircaloy fire which, given the right conditions, will propagate to ne1ghborfng assemblies.

However, compar1-sons to out-of-pile heat-up data have not shown good agreement with the code.

The authors noted that more work fn several areas was needed to define more prec1sely the conditions and configurations which allow or prevent propaga-tion.

I 4

1-3 Several studies have been conducted on al ternati ve 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.s 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 pools 1.2 Related Events.

There is no case on record of a significant loss of water inventory from a domestic, commercial spent fuel 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 open, 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.~

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.~

Also, had the transfer of spent fuel been fn 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 unique 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

4 wide.

About 40 operating (or soon to.operate) reactors use inflatable seals.

However,'ecause of design differences, the Haddam Neck failure does not appear to be directly applicable to the other plants.

It is noted that BNR plants have permanent steel bellows seals to fill the gap between the reactor flange and the cavity bearing plate.

This issue is discussed more fully fn Section 2.3.

1.3 Risk Potential The risk potentials of "Beyond Design Basis Accidents in Spent Fuel Pools" are defined in terms of the probabilities of various initiating events that might compromise the structural integrity of the pool or its cooling capability, the probability of a system failure, given an initiating event, fuel failure mechanisms, given a system failure, potential radionuclide releases, 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.

1.4 Discussion of S ent Fuel Stora e Pool Desi ns and Features The general design criteria for spent fuel storage facilities are stated in Appendix A of 10 CFR 50,s and are discussed more fully in Regulatory Guide

$ 3 lo The pool structures, spent fuel racks and overhead cranes must be design-ed to Seismic Category I standards.

It is required that the systems be de-signed (1} with capability to permit appropriate periodic inspection and test-ing of components important to safety, (2) with suitable shielding for radia-tion protection, (3) with appropriate containment, confinement, and filtering

'systems, (4) with a residual heat removal capability having reliability and testability that reflects the importance to safety of decay heat and other

1-5 residual heat

removal, and (5)'o prevent significant reduction in fuel stor-age coolant inventory under accident conditions.~

The configurations of spent fuel storage pools vary from plant to plant.

Table 1.1 lists 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 fs 38 feet.)

Dur-ing refueling, the cavity above the top of the pressure vessel is flooded to the same elevation as the storage

pool, so that fuel assemblies can be trans-ferredd 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 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 container 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 systems.

More details concerning various configurations are given in Section 2.3.

1.5 Selection of Surro ate 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.

1-6 It should be noted that both surrogate plants have relatively large in-ventories of spent fuel ass'emblies in their spent fuel basins.

Accident initiating events and their probabilities are covered in Section 2.

Fuel cladding failure scenarios based on the SFUELIW Computer Code are evaluated fn Section 3.

'ncluded 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 terms 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

l.

"A Prforftfzatfon of Generic Safety Issues," Division of Safety Technolo-gy, Office of Nuclear Reactor Regulation, U.S. Nuclear Regulatory Coamfs-

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 Commfss4m, 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 Mater During Storage,"

prepared for the U.S.

Nuclear Regulatory Comm)ssion by Sandia Laboratories, NUREG/CR-0649 (SAND77-1371),

May 1979.

4.

N.A. Pfsano, F. Best, A.S. Benjamin and.K.T. Stalker, "The Potential for Propagation of a Self-Sustafnfng Zirconium Oxidation Following Loss of Mater 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.)

1-7 5.

D.D. Orvis, C.

Johnson, and R.

Jones, "Review of Proposed Dry-Storage Concepts Usino Probabilistic Risk Assessment,"

prepared for the Electric Power Research Institute by the NUS Corporation, EPRI NP-3365, February 1984.

6.

IE Bulletin No. 84-03:

"Refueling Cavity Water Seal," U.S. Nuclear Regu-latory Commission, Office of Inspection and Enforcement, August 24, 1984.

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.

9.

Code of Federal Regulations, Title 10, Part 50, "Domestic Licensing of Production and Utilization Facilities, Appendix A, 'General Design Cri-teria for Nuclear Power. Plants,'eneral 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.

Table 1.1 SMR'ss DATA DN SPENT FUEL STORAGE BASINS.

included are spent fuel storage inventories as of Oeceaber

1981, fractions of core in storage, comparisons with thc 'reference case'f radionuclide inventory, locations of spent fuel basins, and seisatc design bases of pools.

Plant Theraal Noaber of Spent Fuel Power Fuel Asscablics Stored lnventorya tHMt)

~

in Corea (No. of Assemblies)

Stored inventory Fractions of Core Radioactivity Relative to Reference Case tper cent)

Storage Pool tocationd Selsiil c Design Basise Sig Rock Point'ro>>ns Ferry-1 Sro>>ns Ferry-2

~

Drowns Ferry-3 Bruns>>ick-1 Sruns>>lck-2 Cooper Dresden-1 Dresden-2 Dresden-3 Ouane Arnold Fitzpatrick Grand Gulf-1 Hatch-1 240 3293 3293 3293 2136 2136 2381 700 2527 2527 1658 2136 3833

. 2136 81 761 761 I

761 560 560 548 164 721 721 368 560 N/A 560 172 1068 889 1768 f

1056 9

921 985 221 h

2011 576 816 140 2.05 1.40 1.16 2.31 1.89 1.65 1,80 0.48 h

2.78 1.57 1.16 0.00 0.25 1.9 16 ~ 1 38.2 76.1 16.0 40.2 42.9 3.36 h

70.3 26.0 35.6 0.0 6.1 AS, grd RB ~ ele RS, ele RB, ele RB, elc RS ~ ele RB ~ el e RS, ele RB, elc RS, cle RB, ele H/A RB, ele DOE*0.059 DBE<0.209 OBEYED.209 OBEYED.209 OBEYED.169

.OBE*O.l6g OBEYED.29 OBEYED;209 OBEYED.29 OBb0.29 OBEYED.129 OBE 0.159 OBEYED.159

Table 1.) (Cont'd)

Plant Theraal Neaber of Pomr Fuel Assembl les (HMt) ln Corea Spent Fuel Stored lnventorya (Ko. of Asseablles)

Stored inventory Fractions'f Core Radioactlvlty Relative to Reference Case (per cent)

Sel sole Storage Pqol Design Location4 Baslse Hatch-2 Ilumboldt Bay LaCrosse LaSalle-1 LaSalle-2 Lfeerfck-1 Hl1 1 stone-1 Hontfcello-Nlne Hlle Point-1 Oyster Creek Peach Bottea-2 Peach Bottoa-3 P Ilgrlo-1 guad Cltles-1 guad Citfes-2 2436 220 165 3323 3323 3293 2011 1670 1850 1930 3293 3293 1998 2511 2511 560 172 72 N/A N/A N/A 580 484 532 s

560 764 764 58d 724 724 1284 251 207 0:

1346 1137 1244 1375 1361 1212 1128 1730 412 2.29 1.46 2.88 I

0.00 t

0.00 0.00 2.32 2.35 I

2.34 2.46 1 78 1.59 1.94 2.39 0.57 55.8 3.2 4.8 0.0 0.0 0.0 46$ 7 39.2 43+3.

47+5 58.6 52.4 38.8 60.0 14.3 RB, ele N/A AB, grd RB ~ ele RB, ele RB, ele RB, ele RB, ele RB, ele RB, ele RB, eke RB. ele RB, ele RB, ele RB, ele OBE$ 0.15g OBE$ 0.509 DOE*0.12g SSE$ 0.20g SSE$ 0.20g SSE$ 0 139 DBE$0.17g DBE$0.12g OBE$0.llg.

DGE$0.22g OBE$ 0.12g

~

OBE$0.12g OBE$ 0 159 OBE$0.24g OBE$ 0.24g

Table 1.1 (Cont'd)

Plant Thermal Number of Spent Fuel Power Fuel Assemblies Stored inventory (Nt) in Corea (Ko. of Assemblies)

Stored inventory Fractions of Core Radioactivity Relative to Reference Case (per cent)

Seismic Storage Pqol Design

,Locationu Gasise Susquehanna-1 Susquehanna-2

~

Vermont Yankee Mash. Hucl.-2 3293 3293 1593 3323 764 N4 368 H/A 1174 0 I 0.00 O.OO 3.19 0.00 Oo0 O.D 50.8 0.0 RB, ele RB ~ ele RB, ele H/A SSEID.lg SSE*D.ig OBEYED.149 SSE~0.32g Footnotes a)

Source:

U. S. Kuclear Regulatory Cocaission, Licensed 0 eratin

Reactors, KUREG-0020, Vol. 9, Ko. 1, January 1985.

b)

(Stored Assemblies)/(Assemblies in Core).

c)

"Reference source Tera'sslies a thermal po~er of 3000 Nt. stored inventory fran ten annual discharges, last discharge six months ago,

'otal Inventory 1150 assemblies.

Source term relative to "Reference Source Term'as not been corrected for age of fuel In storage.

d)

Location:

RB ~ reactor building, AB ~ auxiliary building, grd ~ pool at grade level, ele poo) at high elevation In building.')

Seismic design basis as a function of the gravitational acceleration (g):

OBE ~ design basis earthquake, or equivalent as used for older vintage plants; SSE a safe shutdown earthquake as defined in 10 CFR 100, App. A; Entry shown is the horizontal component.

I f)

Brunswick-1 has in storage 160 PMR + 656 BMR a'ssemblies, equivalent to 1056 SMR'assemblies.

g)

Brunswick-2 has in storage 144 PMR t 564 BMR assemblies, equivalent to 924 SMR assemblies.

h)

Dresden Units 2 and 3 have two pools in one structure.

The data cited are total of the two.

I)

N/A data not available.

Table 1.1 (Cont'd)

PMR'st OATA OH SPEHT FUEL STORAGE BASINS.

Included are spent fuel storage inventories as of Oeceaber

1984, fractions of core in storage, comparisons with the "reference case" of radionuclide inventory, locations of spent fuel basins, and seismic design bases of pools.

Plant Thermal Number of Spent Fuel Power fuel Assemblies Stored lnventorya Stored inventory (KMt) in Corea (Ho. of Asseablies)

Fractions of Core Radioactivity

~ Relative to Reference Case (per cent)

Storage Pool Locationd Seismic Oesign Basise Arkansas-1 Arkansas-2 Beaver Valley-1 Byron-1 Callaway-1 Calvert Cliffs-1 Calvert Cliffs-2 Catawba-1 Cook-1 Cook-2 Crystal River-3 Oavis Besse-1 2660

'57 r

N/A H/A 3411 2700 2700 H/A 3250 3411 2544 2772 H/A 217 217 H/A

'93 193 177 177 2568 177 2815

'77 388 168 104 H/A 9

868 H/A 9

553 171 199 2.19 0.95 0.66 0.00 H/A 9

'4.00

'/A 9

2.87 0.97 1.12 56.3 I

'6.7 I

17.6 0.0 N/A 9

108.0

'/A 9

93.1 24.6 31+2 AB, grd AB, grd FB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd OBE*O.Zg OBE 0.2g SSE 0.125g SSE~0.29 SSE*0.29 OBEi0.15g OBEi0.159 SSE~0.129 SSE~0.209 SSE~0.209 SSEi0.109 OBEi0.159 Oiablo Canyon-1 3338 Farley-1 2652 N/A 157 N/A 114 H/A 0.73 H/A 19.3 AB, grd AB, grd DOE~0.4g SSEi0.109

Table 1.1 (Cont'd)

Plant Thermal Numsber of Spent Fuel Pouer Fuel Assesblles Stored lnventorya (HMt)

In Core>

(No. of Asseablles)

Stored inventory Fractions of Coreb Radloactlvlty

, Relative to Reference Case (per cent)

Storage Pool locatfond Selsmlc Design Basis" Farley-2 Fort Calhoun Glnna Haddaa Neck Indian Point-1 lndlan Point-2 Indian Point-3 Kewaunee Halne Tankee HcGu Ire-1 HcGulre-2 Hlllstone-2 North Anna-1 Horth Anna-2 Oconee-1 2652 1500 1520 1825 2758 3025 1650 2630

34) 1 3411 2700 2775 2775 2568 157 133 121 157 h

0 193 193 121 217

~ 193 N/A 217 157 157 177 62 305 340 545 160 332

'40 268 577 91 N/A 376 9

220 g

1037 0.39 2.29 2.81 3.47 1.72 Oo73'.21 2.66 0.47 H/A 1.73 9

1.40 9

5.86, 10.5 34.4 42.7 63 4 47.4 21.9 36.5 69.9 16.1 H/A 46.8 9

38.9 9

150.5 AB; grd AB, grd

, AB, grd AB, grd AB, grd AB, grd AB. grd AB~ grd AB, grd AB. grd AB, grd

'AB, grd AB, grd AB, grd AB~ grd SSE~O, IOg

'BEs 0.179 ODE*0.209 I

OBEYED.17g DBEi0.10g OBEYED.159 DBE~0.159 DBE~0.129 DBEiO.I'99 SSE~O.I59 SSEa0.159 DBE~0.17g 55E~0.129 SSE80-129 OBEYED.IOg

Table 1.1 (Cont'd)

Plant Theraal Hmber of'pent Fuel Po~er Fuel Assemblies Stored Inventorya (Nlt)

~

in Corea (Ho. of Assemblies)

Stored Inventory Fractions of Core Radioactivity Relative to Reference Casec (per cent)

~

Storage P~ol I.neat ion Sei saic Design Oasise Oconee-2 Oconee-3 Palisades 2568 2568 2530 Rancho Seco-1 Robinson-2 Salea-I Sales-2 San Onofre-I San Onofre-2 San Onofre-3 2772 2300 3338 3411 1347 3410 3390 Palo Verde-I H/A Point Beach-1 1518 Point Beach-2

~

1518 Prairie Island-1 1650 Prairie Island-2 1650 177 177 204 K/A 121 121 121 121 177 157 193 193 157 217 217 218 480 H/A 9

524 9

601 260 152 296 265 94 i) 217 1.23 2.35 H/A 9

4.33 9

4.97 1.47 0.97 1.53 1.37 0.60 1.00 0.00 31+6 59.5 H/A 9

65.7 9

82o0 40.7 22 3 51.2 46.8 8.1 34.1 0.0 AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB. grd AB, grd AS'rd OBEYED. Ig OBEYED. Ig OBEYED.209 SSE~ 0. 20g DBE~O.IBg OBEYED.IBg SSE~O.I29 SSE$ 0.129 SSEr0.259 OBEYED.209 DBEi0.209 OBEYED.209 OBEYED.509 SSE~0.679 SSE~0.679

Table 1.1 (Cont'd)

Plant Therwal s

Number of Spent Fuel Pouer Fuel Assemblies Stored Inventorya (HMt)

In Corea (Ho. of Assenblles) r Stored Inventory Fractions of Core Radl oactl vlty Relative to Reference Casec (per cent)

Sel sslc Storage Pqol Des Ign Locatlono Basise Sqquoyah-1 Sequoyah-2 St. Lucle-1 St Lucle-2 Sunvner-1 5urry-1 Surry-2 Three Hlle Island-1 Three Hlle Island-2 Trojan Turkey Point-3 Turkey Point-4 Materford-3 Yankee Rove 3411 193 3411 193 2700 217 2775 2441 2441 2535 157 157 157 177 177 3411 2200 2200 H/A 600 193 157 157 H/A 76 2560

'/A 65 130 352 H/A 52 9

608 208 312 445 430 H/A 250 0.34 0.67

).62 H/A 0.33 9

3.87 1.18 0.00 1.62 2.83 2.74 H/A 3.29 11.5 23.0 43.8 H/A 9.2 9

94+5 29.8 0.0 55.1 62.4 60.3 H/A 19.7 AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd AB, grd SSEi0.189 SSE*O.IBg OBEYED.IOg SSE*0.10g SSE 0.15g SSE~O.I59 SSE~0.159 DBE~O. 129 SSE 0.129 OBEYED.259 OBEYED.15g DBE 0.159 SSE*0.10g Hone

Table 1.1 (Cont'd)

Plant Thermal Iiueber of

~

Spent Fuel Po~er Fuel Assemblies Stored Inventorya

'tored Inventory (Igft) ln Care (iio. of Assemblies)

Fractions of Core Radioactivity

'elative to Reference Casec (per cent)

Sei smi c Storage Pqol Design I.ocatlona Oasisc Lion-1 Lion-2 3250 3250 193 193 9

863 9

4.47 9

145+3 AB, grd AB, grd SSENO.ILg SSE~D.I79 Footnotes for Table 2 a)

Source:

U. S. IIuclear Regulatory Caaxfssfon, Licensed Operating Reactors, NREG-0020, Vol. 9, No. 1, January 1985.

b)

(Stored Assemblies)/(Assemblies ln Core).

I c)

'Reference Source Tera" assumes a thermal power of 3000 Nt. stored Inventory fram ten annual discharges, last discharge six months ago, total inventory 700 assemblies.

Source ters relative to "Reference Source Term" has not been corrected far age of fuel ln storage.

d) location:

RS i reactor building, AS auxiliary building, FS fuel building,- g ~ pool at grade level ~ e

~ pool at high elevation In bul Iding.

e)

Seismic design basis as a fraction of the gravitational acceleration (g):

OSE design basis earthquake.

or equivalent as used for older vintage plants; SSE

~ safe shutdown earthquake as defined In 10 CFR 100, App. A.

Entry sho~n ls the horizontal component.

f)

II/A data not available.

Spent fuel basin shared by two units.

Entries shown are totals.

h)

Indian Point-1 Is permanently shutdown.

'I l

I)

IIII-2 Is Indefinitely shutdown.

))

Diablo Canyon originally used the 'Double Design Earthquake."

ODE accclcratlon' DBE. ~ Later, mare elaborate ana!ysls was done to postulate an earthquake of 0.5g associated with the llasgri Fault.

~ I 0

2-1 2.

ACCIDENT INITIATING EVENTS AND PROBABILITY ESTIMATES 2.1 Loss of Water Ci rculatin Ca abi lit The spent fuel basfns 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 fn Table 2.2 (data extrapolated to the 1987 scheduled refuelfngs).

The cooling systems provided for spent fuel pools typically have a capacity fn the range of 15 to 20x10 Btu/hr (4.4 to 5.9xlO 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.

(Note:

~

In a situation where the stored inventory was

small, an equilibrium temperature, below the boiling point, would 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 capabi lfty 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 cfrculatfon until exposure of fuel to afr fs quite long.

Even in the most pessfmfstfc case cfted fn Table 2.3 (Docket No. 50-247),

the water level fn the pool would drop only about, 6 inches per hour.

Thus, there appears to be considerable time available to restore normal cooling or to implement one of several alternative backup options for cooling.

For licensing purposes, ft has been accepted that the time interval for restoring cooling-manually from available water sources is adequate without requi ring active (automatic) 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.z 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 value assumed in MASH-1400). s This combined with the conditional probabili-ties of failure/non-availability of the second "train" yielded a combined fre-quency of a pool heatup event of 3.7x10-z/Ry.

(This estimate appears to be somewhat conservative since no "pool heatup events" are on record after

-10~

r reactor year s 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 1.4xl0-6/Ry for an accident initiated by loss of spent fuel pool cooling.

Originally, the spent fuel pool at the Gfnna plant had only one installed cooling train with a "skid-mounted" backup pump and heat exchanger.

However, a

second cooling'rain 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-~/Ry, i.e., about a factor of 3 smaller than for the BWR case analyzed in Ref. 2.

2.2. Structural Failure of Pool 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 f'n 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.

2~3 2.2.1 Structural Failure of Pool Resultin from Seismic Events Procedures and'onventions for a detailed probabili stic ri sk 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, e.g.,

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 I), 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 inconsistent with the hazard uncertainty."

The specific applicability to spent fuel pools of Reed's assertion is discussed in Section

'.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.~

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).

The methodology has been further refined in a

subsequent

study, "EUS Seismic Hazard Characterization Project" SHC) ~ 9,10 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 elicitation 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 funct'ions 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 manner.

Provision is made at various stages for peer review of the methods and input opinion, feedback to the experts and critical evaluation of the re-sults.

M 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 Hotion 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 b4 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.

In particular, numerical tabulat1ons of the sets of hazard curves (such as those shown 1n F<gs.

2.1 and 2.2) and their dertvat1ves, (dh/dal.~

lf~

for each reactor s1te would be helpful.

Also, 1t appears that the 1'ocal sit~

geol ogy needs more rigorous consideration in the deri vati on of the hazard curves (see below).

Members of the Peer Review Panel have suggested several ways in which the methodology could be reffried (see Ref.

10, Section 7

and Appendices D.1-D.4).

Many of these suggestions were implemented in'he final feedback process and were included fn the final results reported fn 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. Q.l).

Comments from NRC licensees and their consultants indicated objections to appl1cation of the results to specific

s1tes, noting that the s1te specific correction factors in the ground motion models were too simplified to ade-quately take local geological factors into account (Ref.

10, App. D.6).

Also, the criticism was made that the results were not adequately tested agains recent historical. records.

In order to illustrate the hazard

curves, their range of uncertainties

'nd comparison with other

studies, a series of figures taken from Ref.

IO for the Millstone site 1s reproduced 1n Figs. 2.1-2.4.

Figure 2.1 1s the hazard -curve obtained from combining the "best esti-mate" results for.all experts in the SHC study (including the seismic and the ground motfons panels).

The curve plots frequency of exceedance per year vs.

peak ground acceleration.

Ffgure 2.2 fllustr'ates the uncertainties fn the hazard curve (15, 50, and 85 percentfles) derived from the spread fn expert opinion and the self-confidence factors in the input parameters.

It can be seen that the spread between the 15 and 85 percentfles 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 SEP.

or the SHC

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'he SEP study by one of the experts 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-paring 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 recomnendations 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) 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 fn the equations used for conversions from intensity to mag-nitude and in the values for the rate of earthquake recurrence (Ref.

10, pg.

8-1 et seq.).

As would be expected the 280 year historical hazard curve (de-noted by "H") falls below the SHC data because it does not 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.

The speculat)on 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." "~

These result-from the liquefaction and venting to the surface 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.

The latest

{prior to 1886) occurred about 1100 years ago.

Sand blows from prehistoric earthquakes have been un-earthed recently in the region e'xtending from near

Savannah, GA as far north as Myrtle

Beach, SC.~~

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.

Recently Thorson et al. reported the existence of apparent sand blow craters in eastern

-Connecticut.~~

These craters were recently examined by a

USGS field team and assessed as not being of the same nature as those observed in South Carolina.~a 2.2.1.2 Seismic Hazard Estimates for the Millstone and Ginna Sites The "best estimate" and the median, 15 and 85 percentile seismic hazard curves developed by the SHC project for the Millstone site are shown in Figs.

2.1 and 2.2.

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 appr oxi-mately 1 g (980.7 cm/secz),

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; however, the SEP project included data for Ginna.

Unfortunately, the format of the SEP results, which were directed primarily at obtaining site specific spectra, cannot readily 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'000 year return times, tabulated for the two plants in the SEP study.a The hazard'urve resulting from this synthesis is shown in Fig. 2.5.

Because of the

'igher upper magnitude cutoff at the Ginna site, as perceived by experts, (Millstone:

MMI ~ 8.0 vs. Ginna:

MMI ~ 8.2),

we have assumed the upper cut-off PGA of the hazard curve to be 1.25g.

Although this is recognized to be a

somewhat pessimistic assumption, it serves the useful purpose of illustrating the sensitivity of the cal'culated seismic risk to the upper cutoff of the haz-ard curve.

2.2.1.3 Seismic Fra ilit of Pool Structures Fragility curves specifically for spent fuel pools have never been devel-oped.

It is necessary therefore, to rely on fragility assessments for other structures which appear to be of similar construction to spent fuel storage pools.

It must be recognized that this procedure introduces an additional element'f uncertainty in the final risk estimates 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

BMR plant is aggravated by the fact that the pool structure extends typically from the 60 to the 100 foot elevations above grade with the resultant amplification of the seismic bending stresses relative to the lower el'evations of the structure.

For the present

analyses, two, somewhat diverse sets of fragility esti-

mates, have been used:

I) the fragility curve developed by R.P.

Kennedy et al'.

for the Oyster Creek reactor building; and 2) the fragility of the Zion plant auxiliary building shear walls (north-south ground motion). ~4

2-10 In each'case, the fragility curve fs "defined by the following equation:

F{a) e f(xn a/A/SR3, (2.1) where F(a) is the probability of structural failure given a peak ground accel-V

eration, PGA

~ a.

e(

) is the normal distribution function, A is the median fragility level (i.e., the acceleration at which th'ere is a

50% probability of failure) and BR is the logarithmic standard deviation expressing the random-V ness in the value of A.

A third parameter, Bu, is used to express the un-certainty in the median value and is used to generate upper and lower confi-V dence limits.

For example, it can be shown that the substitution for A in Eq.

2.1 of A'

e

~u and A'

u generate respectively the 84 and 16 percentile curves.

Thus, a set of fragility curves can be generated from three parameters, A,

gR and-8.

The data used for generating the "Kennedy" and the "Zion" 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 i'n the U.S. in the mid 1960's.

The Kennedy fragility curve is sho~n 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 Seismicall

-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 I

(~a)i F(a)$

da.

o (2 2) where Pi g

is the failure probability obtained from the convolution of hazard curve i with fragility curve j,

~dH/dalai is the derivative of the hazard curve i (i.e., the annual frequency of occurrence of peak ground accel-

eration, a,

and F(a)>

is failure probability at acceleration, a,

for

2-11 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 carried out for each pair of curves with weighting factors assigned to each of the curves in each set..

The resultant collection of Pi j gives a probability distribution which expresses the uncertainties fn 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 (~i or ~j) 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, Pf ~ $0)iQj Pi j / )ldi (2.3) where

)~i

~

1,

$uj

~

1 and )~i,j

~

$~i~j

~

1.

The weighting factors assigned by BNL for the Millstone case are given in Table 2.5.

As can be seen from the table, 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 percentile 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 1.0g, the mean probability of failure, Pf, derived from the 24 sets of Pi,j, using the weighting factors listed in Table 2.5 and Equation 2.3, was Vf ~ 2.2x10

/year (Millstone).

In the case of Ginna, only a single hazard curve (Fig. 2.5) was

used, 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 "Zion" fragility curves are more appropriate, than the "Kennedy" curves.

2-12 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 1.25g, the mean probability resulting from the convolution of the single hazard curve with the six weighted fragility curves was Vf 1.6x10

/year (Ginna)

The difference between the estimates for Millstone and

Ginna, 2.2xl0-vs. 1.6x10-, should 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 Nissiles 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.

for an average U.S. site derives a

probability of <1x10- /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

/year (Ref. 25, pg. 4-58).

The damage caused by Missiles generated by turbine failure depends on the orientation of the turbine axis relative to the structure, as well as the fre-quency of turbine failure.

An analysis by Bush yields a probability of

-4x10

/year for spent fuel pool damage from a turbine failure missile.

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 Refuelin Cavit Seal Failures On'ugust 21, 1984,.the Haddam Neck Plant experienced a failure of the refueling cavity water seal, while preparing for refueling.

The water level

2-13 i n the refueling cavity dropped by about 23 feet to the top of the reactor

~

vessel flange within 20 minutes a loss of approximately 200,000 gallons, or a leak rate of about 10,000 gal-lons per minute.~7 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.

A]though the seal failure did not result in an accident or in the release of radioactivity, the inci'dent 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.

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.

t BWR's have a permanently installed stainless steel bellows to seal the

gap, and are, thus, not subject to failure of the Haddam Neck type.

Hany PWR's seal the gap with gaskets held down by a bolted flat steel ring.

Such systems have experienced difficulties in achieviag tight seal be-cause of surface irregularities and small vertical and concentric offsets in the two flanges.

Consequently, many pl ants have converted to inf1atabl e (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 summer 1985 some 45 units used pneumatic seals in the refueling cavity.

a 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).

According to the re-sponses of the licensees, even if a pneumatic seal should deflate, the leakage would be expected to be small 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'ead to extrude through the gap.

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'nowledge 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 -1xl0-~/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-~/Ry.

's is obvious from Table 2.6, a seal failure does not necessarily result fn the rapid loss of water inventory from spent fuel transit or storage loca-tions.

The limited experience indicates that the most probable time for a

refueling cavity seal to fail is shortly after installation, while the cavity volume is being filled with water.

According to the analyses supplied by licensees in response to IE Bulletin No. 84-03, the failure of a pneumatic refueling cavity seal in most PMR plants would not result in massive leaks because of the relatively narrow gap to be sealed and the geometric shape of

'he seal.

Also, leaks from seal failures in transfer 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

0

2-15 serious loss of pool water inventory resulting from a pneumatic seal failure to be in the range of "1xlO- /Ry.

Even a large loss of water inventory from the spent. fuel pool does not necessarily result 1n uncovering and subsequent failure of fuel.

Host spent fuel bas1ns are constructed with weirs below the transfer gates which preclude complete drainage of the pool, even in the event of a catastrophic Haddam'eck 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:

L1censee responses to IE Bulletin 84-03 did not always provide information about the elevations of we1rs and tops of stored assemblies.)

In the event of a draindown of the pool to near the top of the fuel as-

sembl1es, there would still be time (1/2 to 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months ) 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."

A pool heatup event similar to the partial draindown scenario described above was considered by the NRC staff in Ref.

2.

A conditional probability for failure to restore adequate makeup water was taken to be Sx10-z, based purely on judgement.

Because of higheJ radiat1on levels in the partial drain-down scenario, it is est1mated that the probability of fa1lure to. restore adequate makeup water to be somewhat larger, i.e.,

-1x10

~.

. Given all of the above, the probability of a pneumat1c seal failure'which.

results 1n exposure to air of stored spent fuel with resulting clad failure is estimated to be of the order of P

lxlO-6/Ry.

2-16 2.4 Pool Structural Failure Due to Heav Load Dro MASH-1400 considered the probability of structural damage to the pool due to the dropping of a fuel transfer cask (Ref. 3, pg. 1-97).

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 failure (for a single unit plant) was 4.5xl0-7/Ry.

The above frequency was based on a crane 'failure probability of 3x10 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 damage to the pool wall if a crane failure oc-cur red.

Human error was not considered.

Since spent fuel is not currently being shipped, this hazard does not ex-ist at the present time.

However, 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.

have reexamined the cask drop probability and have used the following probabilities:

Mechanical failure of crane 3x10-s/operating hour Electrical control failure of crane

~ 3x10-~/operating hour Human error

~ 6x10-"/lift.

As can be seen, human error dominates the Or vis estimates for probability of a cask drop.

The Orvis datum for human error was based on a study by Garrick et al.s 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 fn the range of 10 s to 10 4 per operation appears to be consistent with data listed in the NRC handbook on human 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.

2-17 Obviously, not all human failures associated with the lifting and moving of a spent fuel shipping cask would result in structural damage to the pool.

The section of the pool where, the cask is set down has an impact pad,to absorb the impulse of a

dropped cask.

Accidents in unloading the cask'rom or reloading on the transport vehicle would not involve the pool.

Only horizontal movements of the cask above a structurally critical sec-tion of the pool would pose the threat of structural damage.

As noted

above, WASH-1400 assumed that the sensitive section is the vertical wall at the pool edge.

It was implicitly assumed that all load drops on the pool edge would result in structural failure.

This assumpt1on appears to be too simplistic and consequently too conservative for the following reasons:

~

many "load drops" would be partially attenuated by crane mechanisms which limit descent

rates, and reduce impact energy, 0

~

in case of some "off-center" hits, the full potential impact energy would not be absorbed by the pool edge (cask tilted, one end strikes floor first), and

~

account should be taken of exterior cask 'fitt1ngs (e.g.,

cooling vanes) which absorb some 1mpact energy.

No rigorous structural analyses have been performed to s<dpe the range of damage to a pool.edge from a cask drop.

In the absence of such analyses, it has been necessary to estimate. the conditional probability of catastrophic structural damage g1ven a cask -drop 1n the vicin1ty of the pool edge.

It is estimated that the conditional probability is less than 100% and greater than 1%.

A conditional probability of 10% has been arbitrarily selected for the hazard calculat1on and 10'nd 1% used for defining the range of uncertain-ties.

Since human error, rather than mechanical or electrical failure, appears to dominate the hazard arising from shipping cask movements, the various steps 1n the crane operation have been identified in Table 2.7, which also lists the types of human error associated with each step.

The distribution of failure

2-18 frequency in the. various steps has been estfmated 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 operat1ons of this type.)

It will be noted that most steps in.the crane operation do not jeopardize the structural integrity of the pool.

Only fn steps Sa and. 5b {see Table 2.7) could the cask strike the pool edge.

An accident of the type Tfsted in 5a (horizontal movement with cask not high enough to clear the pool edge) would probably not cause serious damage because of the:limited kinet1c energy of the cask associated with the slow veloc1ty of horizontal crane movements.

Thus, only step 5b fn Table 2.7 fs considered fn the hazard calculation.

For purposes of calculat1ng the cask drop hazard, i.e., the probability of catastrophic structural damage to the pool result1ng from a cask dropp1ng on the pool

edge, the assumptions listed fn 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 fa1lure due to cask drop on pool edge caused by mechanical or el ectrfcal failure of crane

~ 3.5x10-7/Ry.

Probabfl1ty of structural failure due to cask drop on pool edge caused by human err or

~ 3.lxlO-s/Ry.

If the failure rates summarized fn Table 2.8 are assumed to be statis-tfcally fndependent, then the uncertainty in the overall failure r ate is domi-nated by the uncertainty 1n the probab1lity of pool failure.

Thus the overall uncertainty is about a factor of ten 1n either direction.

2.5 Swear of Accident Probabilities The probability estimates made fn Sections 2.1-2.4 are summarized in Table 2.9.

These include only those accidents that result in the complete loss of pool water fnventory.

It will be seen that shipping cask drop result-ing from human error and seismic induced failures dominate in the hazards.

As

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. Coomercial Nuclear Power Plants,"

U.S.

Nuclear Regulatory Commission, NUREG-75/014

{MASH-1400), October 1975, App. I, Section 5.

4.

Rochester Gas and Electric Corporation, Oocket No. 50-244, "Design Crite-

ria, Ginna Station, Spent Fuel Cooling System,"

EMR

1594, Revision 1,

October 10, 1979; U.S. Nuclear Regulatory Commission, Safet Evaluation b

the Office of Nuclear Reactor Re ulation Su ortin Amendment No.

65 to Provisional 0 eratin License No.

DPR-18 Rochester Gas and Electric Cor oration 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 Safet Anal ses Procedures Guide, Yol. 2, pr e-pared for the U.S. 'Nuclear Regulatory Comnission by Brookhaven National Laboratory, NUREG/CR-2815

{BNL-NUREG-51559) Yol. 2, Rev. 1, August 1985.

6.

John M. Reed, "Seismic Probabilistic Risk Assessment for Critical Facili-ties," preprint of 'paper presented at the 1985 ASCE Spring Convention, Denver Colorado, April 29 - Hay 3, 1985.

~

2-20 7.

II.L~ 8, K.G!

ill, O.II.

N <<1i.Il., ~AA h

Seismic Zonation for Sitin Nuclear Electric Power Generatin Facilities in the Eastern United States, prepared for the U.S.

Nuclear Regulatory Comnission by Rondout Associates, Inc., NUREG/CR-1577, May 1981.

8.

U.S. Nuclear Regulatory Commission, NUREG/CR-1582, Seismic Hazard Anal-

sis, 5 Volumes:

Volume 1,

Overview and Executive Summar O.L.

Bernreuter and C.

Minichino, April 1983.

Volume 2, A Methodolo for the Eastern U.S.,

Lawrence Livermore National Laboratory/

TERA Corporation, August 1980.

Volume 3, Solicitation of Ex ert 0 inion, Lawrence Livermore National Laboratory/

TERA Corporation, August 1980.

Volume 4, A lication of Methodolo Results and Sensitivit

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.M. Mensing and D.H.

Chung-,

Seismic Hazard Characterization of the Eastern United States:

Methodolo and Interim Results for Ten Sites, prepared for the U.S. Nuclear Regulatory Conmis-sion by Lawrence Livermore National Laboratory, NUREG/CR-3756 (UCRL-53527), April 1984.

10.

D.L. Bernreuter, J.B.

Savy, R.B. Reusing, 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 National Labora-tory, UCID-20421, April 1985.

11.

This matter is under consideration but has not yet received NRC sponsor-ship (private communication, O.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 Velocit in Rock in the Carti uous United States, U.S. Geological

Survey, open file report 821033 (1982).

13.

E.g.,

see R.A. Kerr, "Charleston 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.

Gohn, 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 PE 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.B.

Jacobson, "Evidence for Three Moderate to Large Prehistoric Holocene Earthquakes Near Charleston, SC," Proceedin s of the Third U.S.

National Conference on Earth uakes En ineerin Au ust 24-28,

1986, Charleston, 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 (7) Earth-quakes Induced Sand Blows in South Carolina,"

Proceedin s of the Third

.U.S. National Conference on Earth uake En ineerin Au ust 24-28,

1986, Char leston SC, pp.

197-208 (published by the Earthquake Engineering Re-search Institute).

18.

S.F. Obermeier, private coomunication, Sept.

26, 1986.

19.

R.M.

Thorson, W.S.

Playton and L.

Seeber, "Geological Evidence for a

Lgp tlt l

E ttt k

l Egg,'~gt,kt,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. Cornell, R.D.

Campbell, S.

Kaplan and H.F.

Perla, "Probabilistic Seismic Safety Study of an'Existing Nuclear Power Plant,"

Nuclear En ineerin and Desi n

59, (1980) 315-338.

24.

L.E. Cover, M.P. Bohn, R.D. Campbell and D.A. Wesley, Handbook of Nuclear Power Plant Seismic Fra ilities, prepared for the U.S. Nuclear Regulatory Comnission by Lawrence Livermore National Laboratory, NUREG/CR-3558 (UCRL-53455), June 1985.

25.

D.D. Orvis, C. Johnson and R. Jones, Review of Pro osed Dr -Stora e Con-ce ts Usin Probabilistic Risk Assessment, prepared by NUS Corporation for the Elect ic Power Research Institute, EPRI NP-3365, Feb 1984.

26.

S.H.

Bush, Probability of Damage to Nuclear Components Due to Turbine Failures," Nuclear Safet

, 14, No. 3, May-June, 1973.

27.

U.S. Nuclear Regulatory Comnission, Office of Inspection-and Enforcement, IE Bulletin No. 84-03:

"Refueling Cavity Water Seal,"

August 24, 1984.

(This was subsequently followed up by more detailed instructions for evaluation:

"Inspection Requirements for IE Bulletin 84-03, "Refueling Cavity Water Seals,"

Temporary Instruction 2515/66, Ins ection 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 mi'd-1985 a few licensees had not yet filed responses.

29.

Response

of Baltimore Gas and Electric to IE Bulletin No. 84-03.

~

2-23 30.

B.J.

Garrick et al.,

"The Effect of 'Human Error and Static Component Failure on Engineered Safety System Reliability," Holmes 5 Narver, Inc.,

HN-194, November 1967.

31.

A.D. Swain and H.E. Guttmann, Handbook of Human Reliabilit Anal sis with Em hasis on Nuclear Power Plant A lications, prepared for the U.S.

Nu-clear Regulatory Commission by Sandia National Laboratories, NUREG/CR-1278 (SAND80-0200), August 1983.

2>>24 Table 2.1 Typical Spent Fuel Pool Dimensions and Mater Inventories Length/Midth/Depth (feet) 40/26/39 a 43/22. 25/40. 25b aBWR, Vermont Yankee.

bPMR, Ginna.

Pool Volumes.

(cubic feet) 4.1x104 3.4x10" Nominal Mater Inventory (cubic feet) 3.5xl04 3.3x104 Table 2.2 Decay Heat as a Function of Time Since Last Refueling (Data from Appendix A)

Plant Deca Heat Load (10 Btu/hour) eca

>me Since ast hutdown for Refue in days 9

days years year Millstone-2 Ginna 4.43 2.62 3'0 1.96 2.38 1.59 1 ~ 76 1.25 Table 2.3 Examples of Thermal-Hydraulic Transient Parameters, Assuming Complete Loss of Pool Coolant Circulation Docket No.a Rate of Time of Bof 1-Off Temp. Increase Boilingb Rate

('F/hr)

(hours)

(gpm)

(ft3/hr) 50-325 50-250 50-271 50-247 50-344 5.0 9 7

<3 13.0

<6.3 13' 9.3

>20 4 8

>11 28 262 N.A.

14 131 57 534 34 318 aSee Ref. l.

bHours after complete loss of cooling capability.

2-25 Table 2.4 Fragility Parameters Assumed in Thi s Study for Spent Fuel Storage Pools Structure A

(g)

'R g'ef.

Oyster Creek Reactor Building 0.75 0.37 0.38 24 Zion Auxiliary Building Shear Walls (N-S motion)b 1.1 0.12 0.20 25 aDesignated as the "Kennedy" fragility curves in the text.

bDesignated as the "Zion" fragility curves in the text.

Table 2.5 Weighting Factors Assigned to the Various Hazard and Fragility Curves for the Mill-stone Case Sei smi c Hazard Curves:

MILLSTONE GINNA "Best Estimate" 15% Confidence Curve Median Curve 85$ Confidence Curve "i

0.50 0'0 0.30 0.10

>100 1.00a 1.00

~Flit C

"Kennedy", Median

, 16+

, 84$

"Zion", Median

, 16$

84$

a"Synthesized" Curve.

~4l

'.45 0.15 0.15 0.15 0.05 0.05

$a)g

> 1.00

+ld 0.15 0.05

'.05 0 45 0.15 0.15 T.00

2-26

~ 0

~ ~

Table 2.6 Events in Which Inflated Seals Have Failed Date Plant Seal Location Cause Total Leakage 9/72 Pt.

Beach Transfer Gate Failure of air supply 11,689 gal.

10/76 Brunswick 2 Inner Pool Gate Air leak in seal plus compressor power supply failure (Pool level dropped 5")

8/84 Haddam Neck Cavity Seal 1

10/84 San Onofre 2~

Gate Seal 11/84 San Onofre 2

Cavity Seal~

5/81 Arkansas Nuclear Transfer Gate One-2 Maintenance error, air supply shutoff Design weakness, seal shifted Air compressor power failure Manufacturing defect, seal rupture 1000 gal/min 200,000 gal.

in 20 min.

20,000 gal.

12/86 Hatch Pool-Canal Flexible Joint Valve to compressed air supply closed 141,000 gal.

~No spent fuel was in the storage pool.

~Failure occurred during installation and leak testing.

2~27 Table 2.7 Estimated Distribution of Human Error in Heavy Crane Operations.

These Estimates, Hade by BNL Staff, are Based on Engineering Judgement and are Not Supported by Actuarial Data.

Operational Step 1.

Install rigging 2.

Positioning of crane over load, apply tension 3.

Lift load 4.

Star t horizontal tr avel 5b. Horizontal travel 6.

Lower load Possible Human Errors Wrong slings (e,g., hoist rigging not qualified for task)

Improper installation (shackle, pins, etc.)

Crane hook not over center of gravity (load upset as tension applied)

Control error (wrong hoisting speed unintentional reversal of direction)

Control error (move wrong direction, liftor lower instead of move)

Control error (unintentional, rever sal of motion, overshoot stopping point)

Load not high enough to clear obstacles Control error or delayed rigging failure resulting in load drop Control error (wrong direction, descant too fast)

Estimated Fraction of Total'rror Frequency (Per Cent) 10 10 15 10 10 10 7.

Positioning of crane over re-ceiving cr adle and set-down load Inaccurate positioning cradle capsizes during set-down Set down too rapid 20 10 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 oper ator.

2-28 Table 2.8 Assumptions Used in Calculating the Hazard of Catastrophic Structural Damage to Pool Resul ting from the Drop of a Shi pping Cask Item Number of fuel shipments (eventual rate to reduce accumulated inventory) per week Number of passes over pool cage per shipment Fraction of horizontal movement when cask is above pool edge Total operational time in each

movement, minutes per lift Time over pool edge per lift, seconds per lift Mechanical failure rate of crane, per operating hour Electrical failure rate of crane, per operating hour Total accident rate from human error, failures per lift Fraction of human error cask drop accidents occurring during horizontal motion of crane, fraction of total Conditional probability of structural failure of pool given a cask drop at pool edge loca-tion, failures per drop Assumed Value 2

2a 0.25 10 10 3xl0- e 3x10-e 6x10 "

0.01 0.1 Uncertainty Range 0.1 to 0.5 8to 30 5 to 20 10-'o 10-s 10-s to 10-s 10 4 to 10

~

5xl0 s to 5x10-z 10 z to 1.0 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.

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 Accident Estimated Probability/Ry Hillstone Ginna Loss of Pool Cooling Capability Seismic Structural Failure of Pool Structural Failure from Tornado Missiles Structural Failure from Aircrash Structural Failure from Turbine Missile Loss of Pool Mater Due to Pneumatic Seal Failure Structural Failure from Cask Drop~

1.4x10 s

2.2xl0 s

<1x10-8

<1x10-"

4xlO"7 3.1x10"s 5.7x10-7 1.6xlO-s

<1xl0-8

<1x10-"

1x10- s 3.1xlO-s

~After 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

/Ry t

~Typical PMRs map have a failure frequency due to turbine missiles on the order of 4x10 but Ginna's pool is shielded from the turbine.

2-30 Io IO MILLSTONE "8EST ESTIMATE CD ag Io CDX 4J b.

~o IO O

Cf" Io D

IO IO 0 200 400 600 800 PEAK GROUND ACCELERATION (cm/sec

)

IOOO Figure 2.1.

Seismic Hazard Curve for the Millstone Site.

The curve shown is the mean of the hazard curves generated 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).

2-31 IO IO MILLSTQNE l5, 50, 85 PERCENTILES td C3 o

m IO hl h.O 43

~

IO Rz IO l5 50 85 ICY 200 400 600 800 PEAK GROUND ACCELERATION (cm/sec

)'OOO 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 uncertainti es in the best choice of ground mo-tion models (Source:

Ref. 10, pg. 5-45).

2-32

~

~ o

~

I~

10 10

<<3 10 10 10

<<7 10 cO t

ACCELERAT ION ~SKC~~2

~MlLLSTONE 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);

2-33

-1 10 10 10 10 10 10

<<7 10 cv n

0 4

e h

0 -"

o 4J ACCELERAT ICN ~SEC~~Z Figure 2.4 Compar ison 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).

10-'10~

hl R

Cl 10 0X hJ h.O u

10-+

10-5 G1NNA 10 7 200 400 600 800 1000 1200 PEAK GROUND ACCKLERATlON (cm/sec~)

Figure 2.5 Seismic Hazard Curve for Gfnna.

This curve was Synthesized from the SHC "Best Estimate" curve for Hi11stone: (see Figure 2.1),

and pGA ratios for Hillstone and Ginna given in the SEp studies.

2-35 I.O FRAGILITY CURVE o

0.8 zhlD 0.6 U.

hl CLD 0.4 l6%

MEDIAN 84%

0.2 0

0.2 0.4 0.6 0.8 I.O PEAK GROUND ACCELERATION (g) l.2 Figure 2.6 Fragil ity Curves for the Oyster Creek Reactor

'uilding.

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.

2-36 0.3 MEAN=2.15 x 10 yr.

I-lD UJ Cl w 0.2 CQ cK ClO CL O.I 10 9 10 8 10 7 10 6 10 5 10 +

ANNUAL FAILURE FREQUENCY Figure 2.7 Probability Density for Seismically-Induced Fail-ure as a Function of Annual Failure Frequency.

The histogram was obtained from 24 convolutions of four hazard curves with six fragility curves and includes the weighting factors assigned to each curve.

V

~

2<<37 Figure 2.8 Cross Section of a Typical Pneumatic Seal (Source:

submission by Sequoyah Nuclear Plant, Oocket No.

50-327, 10/26/84 in response to IK Bulletin 84-03.

2-38 Idd+99 Jt AEACTM VESSEL AAr4$C (Sbf gdd<49$ ~JJ SIlPPgd T RIHd rSW lid Zvff) 5JIAAPP d'PSC5 AAHP ROuVPCP dr HII5 7'%AtHT CW44af 7P dEAL

~

~ i+

~

~ ~

JVJF/ACd'lACHIHFD dV HIIS 7'd ACCPNMOPATX'eddRAVOCUL (~wi~ur 4Ar Pr 5~ ddTWuH AAH48 Atvu CAVlrT WAL+8Q~PCIRr RiH4)

INSET/CATIPH (Sa fdSSCeCel)

~L PV HPSITICW O IHrWTZP 4 0 ~ IQPS CAVITT'lLL(4bH9$I) hIACQHCP, RFCPWPIrzOHN kvP dACHSaCr'P WITH OFVCCW'S TYM 'A PLASTIC 5TCEL, EIVHICH HA5 C.Edd THAH 4 M LFACHAdLF CHCPRIPFf J TV ~A; NPT.V CV58 777 NWiPIf

'4775I ACllRT SAC SIH7SACS Figure 2.9 Cross Section of Inflated Pneumatic Seal Seated on the Reactor Vessel Flange and Inner Surface of Cavity Wall (Source:

submission by Sequoyah Nu-clear

Plant, Oocket No.

50-327, l0/26/84 in re-sponse to IE Bulletin 84-03.

~

~.

2-39 Sf.

AtOCTA VIAL FCAIIPIOd PiS Cd4 iC W'F4TASLC DOL

~VI~ PLce~

W ~

4~% +w~w~'

P C~l ~T STATION JjY~PQ(~~

~

S I~\\I~

~ IIPII

~ PP I~IIIltlNCL P ~ I W D

IW

~PWWI Fi gure 2.10 Uninf1 ated pneumatic Seal with Steel Hol d-down Ring (Source:

submission by Indian Point Station, Unit 2, Oocket No. 50-247, 11/30/84 in response to IE Bulletin 84-03).

P

~

e e

~

y

~

3-1 3.

EVALUATION OF FUEL CLADDING FAILURE e

/

Two previous studies

~

have analyzed the thermal -hydraul ic 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.

~

3.1 Summar of SFUEL Results The SFUEL code was developed by Benjamin et al.

to analyze the behavior of spent fuel assemblies after an accident has drained the pool.

The results reported in Reference I indicated a wide r ange of decay power levels for which self-sustaining oxidation of the cladding would be predicted.

Several limita-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 Sugar Model Descri tion The SFUEL code was developed at SNL and is described in Reference I.

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.

3~2 3)

Temperature variations across the fuel rods are neglected.

4)

The afr flow patterns are one-dimensional.

5)

The spaces between adjacent basket walls are assumed to be closed to afr flow.

After the water is drained from the pool the fuel rods heat up until the buoyancy driven afr flow is sufficient to prev'ent 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 rises rapidly.

Reference 2 modified the SFUEL code to increase calculatfonal stability and assess propagation of Zircaloy "fires" from high power to low power assem-blies.

The SFUEL code was also modified by Pf sano et al.2 to elim1nate un-realistically high temperatures*

by non-mechanfst1cally removing each node as ft.reaches the melting point of Zi rcaloy dioxide (2740'C or 4963'F).

In the present 1nvestigation, the oxidation cutoff has been reduced to 1900'C (The melting point of Zircaloy 3450'F).

3.1.2 Clad Fire Initiation Results An extens1ve review of the cladding oxidation models used in

SFUEL,

~

fs given in Appendix B:

1 1.

The likelihood of clad fire inftiat1on is not very sensitive to the oxidation equation.

2.

The oxidatfon equation used in SFUEL is a

reasonable representation of the data.

d d

Pllltlyt 1tg 1

t 1dd'g p

tures as high as 3800'C were predicted.>

~

3~3 3.

The like1ihood of clad fire initiation is most sensitive to the decay heat lev'el and the storage rack configuration (which controls the ex-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 inlet orifice (3 inch.diameter) the natural convection cooling in air fs predicted to be adequate to prevent self-sustaining oxidation (cladding "fires") after 10 days of decay for BMR assemblies and 50 days for PWR assemblies.

However for the new high density fuel racks, natural convective flows are so restrict-ed that even after cooling for a year there is potential for self-sustaining oxidation.

As pointed out by Benjamin et al.~ 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 area 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 PMRs with low density storage

racks, and 1.0 for PMRs with high density storage racks.

3.1.3 Clad Fire Pro a ation The SNL investigations

~

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,,the phenom-enology of propagation is not well understood and there is considerable

3-4 uncertainty in these estimates.

Benjamin et al.,~

Johnsen4 have pointed out a

number of limitations ses.~

~

In order to put the present results in mentioning the most important limitations:

Pisano et al.,

Han and f n the previ ous anal y-perspective it is worth 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'o 2200'C'ut 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'elieved 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 SFUEL code had not yet been validated successfully against fuel 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 vy3 idated.

Specif-ically, the axial heat up (without oxidation) and the temperature at which self sustaining 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 term exposure at low temperatures (when oxide layers may flake off and expose fresh Zircaloy).

However, Appendix 8

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 melt1ng 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 SFUEL code (SFUEL1Wz) has sufficient flexibility to estimate the im-portance of oxidation propagation.

5.

Johnson" criticized the clad failure, criterion used 1'n the SNL ana-lyses.

~

He noted that the clad failure could occur at tempera-tures as low as 650'C if the thermal loading is 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 1s very likely and the potential for cladding "fires" is high due'o the effects of asymmet-ric heating (from adjacent high power bundles).

Propagation of cladding "fires" by particulate (i.e., spallat1on) or zir-con1um vapor transport has been investigated and eliminated in an approximate separate effects study by Pisano et al.z However,. propagation due to the heat flux (radiation and convection) from adjacent bundles fs pr.edicted 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.

8oth the power of the 1nitiating bundle and the power of the adjacent bundles have been varied as well as the ventilating conditions of the spent fuel bu1ld1ng.

It should be emphas1zed that SFUEL does not address the question of Zir-caloy oxidat1on propagation after clad melting and relocation.

For recently discharged fuel (less than 90 days),

or for severely restricted air flow (e.g high density PMR spent fuel racks) the oxidation reaction is predicted

3-6 to be. very vigorous and failure of both the fuel rods and the fuel rod racks fs expected.

Thus a large fraction of the fuel rods would be expected to fall to the bottom of the pool-forming a large debris bed.

If water fs not present in the bottom of the pool, the debris bed will remain hot and will tend to heat the adjacent assemblies from below.

The investigation of debris bed for-mation fs beyond the scope of the present study, but ft appears to be an addi-tional mechanism for oxidation propagation.

Rese1ts As pointed out fn Section 3.1.1 self-sustaining oxidation initiation is not very sensitive to the oxidation rate equation but ft fs dependent upon the calculated air flow (r elated to flow resistance) and the power level.

BMR spent fuel with its low power density and open flow 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 only a slight chance of propagation to older low power fuel bundles.

However, PWR spent fuel racks typically have a higher power density and more flow restrfction, thus self-sustaining oxidation can be initiated in fuel

'hat has been discharged for one year or more.

Two fuel building ventilation condftfons have been investigated as de-scribed below but ft must be recognized that. both of these assumptions corre-spond to very idealized condftfons that are unlikely to be duplicated in an actual accident.

Rather these idealized conditions are provided to demon-stratee the fmportance of the various assumptions.

For a beyond design basis seismic event, that catastrophically fails the pool, it seems likely the fail-ure of the fuel building may also occur.

However, Benjamin et al.~ have shown that a very large hole (at least 77 ft2) must be opened fn order to approxi-mate the perfect ventilation case.

3~7 Per feet 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 systems) or by a large opening (greater than 77 ftz) 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 1n Table 3.1 for three typical fuel rack configurat1ons.

Note that these are borderline" cond1tions 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 1mply that fuel rod failure is not predicted for power levels below these conditions.

The power level must be reduced substantially (about 20'L) to en-sure that the pred1cted clad temperature is below 650'C (the minimum tempera-ture at which clad failure and fission product release is likely to occur).

For power and flow conditions that are only slightly below the "critical" conditions ft should be obvious that the heat flux from a

much higher power adjacent bundle would have the potent1al to push the "non-critical" fuel over the self-susta1ning 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 oxidat1on in low power fuel bundles that have been cooled for one or more years.

In this context two limitations of the SFUEL1M~ code should be noted:

1.

The fuel stor age racks are assumed to be 1mmediately adjacent so that no air flow between racks is allowed.

(The numerical approach used to calculate the heat transfer 1s numerically unstable if flow is al lowed).

0 2.

All fuel storage racks are assumed to be identical so that the ques-t1on of propagat1on from high power cylindrical racks to low power h1gh density racks cannot be addressed.

3-8

~ I

~ ~

The first limitation probably represents 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, ft appears

that, under the idealized'erfect ventilation conditions, the provision of an air space of 6 to 12 inches around the periphery of re-cently discharged fuel would minimize the likelihood of oxidation propagation to low power spent fuel assanblies.

(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.

The propagation results with perfect ventilation are summarized in Table 3.2 for the high density rack configuration descr'fbed in Reference 2.

Note that the lowest power (11.0 kM/N'U) 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-dle that is about 2 years old (6.0 kM/HTU).

For a fuel assembly that has been discharged for about 10 days (90 kM/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

kM/HTU 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 suamarized 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 kM/MTU (corresponding to about

'0 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/MTU (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.

Inade uate Ventilation As previously mentioned the case of perfect ventilation implies a very high ventilation rate that is not normally possible.

Benjamin et al.~ 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.

Benjamin~ found that the heat-up of the building in-creased the likelihood of self-sustaining oxidation (i.e., decreased the decay power level necessary to initiate self-sustaining oxidation).

This section is intended to address the question of whether limited ventilatian also increases the likelihood of propagation to low power bundles.

Table 3.5 provides a

summary of propagation runs under inadequate venti-l.ation conditions.

For the analyses the high power assemblies are modeled to represent approximately 1/3 of the core for 1000 HWe plant and the fuel build-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

3-10 p

I ~

recently discharged assemblies uses up the oxygen supply before the lower power assemblies can be heated to the point of self-sustaining oxidation.

In view of th'e potential for fuel building.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 app'roximation for the potential for pr opagation.'.

Conclusions Re ardin Pro a ation Based on the previous results we have concluded that the modified SFUEL gode (SFUEL1Mz) 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 spent fuel that has been stored as'long as 2 years.

The investigation of the effect of insufficient ventilation in the fuel building indicated that oxygen depletion is a competing factor with heating of 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.

These results are in general agreement with the earlier SNL studies.

~

3.2 Validation of the SFUEL Com uter Code The SNL investigations

~

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

~

'undles would oxidize once oxidation had been initiated.

However, an attempt>

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) but

.the SFUEL code over-estimated the temperature transient after.air was intro-duced.

~

'I

~

3-11 The objective of this section is to use the revised~ oxidation rate equa-tion 1n SFUEL to analyze the SNL'mall 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.

The 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 (>850'C) the rods were heated with a very low flow rate of helium before air was 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 1s 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 oxidat1on.

Since the free convection and radiation calculations in SFUEL

~

were inappropriate to the test. configuration, Pisano et al.2 created a stripped down version called CLADz which used a matrix inversion routine to calculate radiation losses.

After several preliminary attempts to analyze the helium portion of 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 hi.gher heat capacity and conductivity than air it appears to contribute to establish-1ng the init1al cond1tions.

In order to provide an adequate simulation of the initial steady-state portion of the test we made two modificat1ons to the CLAD code:

Include helium properties with a

switch to air properties at the'tart of the transient.

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'xidation correlation,

- 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'orrelation has been reduced by a factor of four (note that this corresponds approximately to the data scatter).

Results The fevised 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 4ests.

Note that CLAD still overpredicts the peak temperature for the low flow rate test (4 and

5) but gives goad 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 oxygen must diffuse to the clad surface.

CLAD includes an oxygen depletion calculation but assumes that all the oxygen in each volume is imnediately available at the surface.

3.3 Conclusions Re ardin SFUEL Anal ses 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 sel f-susta1ning clad oxidat1on for a

variety of spent fuel configurations assuming that the pool has been drained.

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 calculat1on which has not been validated.

When oxidation is terminated at the Zircaloy melting temperature (assum-ing that the molten Zi rcaloy is relocated),

oxidation propagation only occursi:-

for spent fuel bundles which are already approaching the "critical" conditions-for self-sustaining oxidation (see Table 3.1).

However, this finding does not mean that oxidation propagation is unlikely.

On the contrary, for some high density storage configurations the "crit1cal" cond1tions are approached for 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 "crit1cal" conditions for initiat1on of oxidation and what the uncertainty is for a given spent fuel configuration.

The crit1cal cond1tions are summar-ized 1n 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 convect1on flow rate.

It is recommended that these free convec-t1on flow calculations be verified against large scale data.

Preferably the data would be obta1ned from spent fuel assemblies in typical storage racks (both high and low density).,

3.4 References for Section 3

l.

A.S. Benjamin, Q.J. McCloskey, O.A. Powers, S.A. Dupree, "Spent Fuel Heat--

up Following Loss of 'Mater Qur1ng Storage,"

NUREG/CR-0649, March 1979.

2 ~

N A.

Pi sano, F.

Best, A S.

Benjamin, K T. Stalker, "The Potential for Propagation of a Self-Sustaining Zirconium Oxidation Following Loss of Water 1n a Spent Fuel Storage Pool," Oraft Report, January 1984.

3.

J.T. Han, Memo to M.'Silberberg,

USNRC, May 21, 1984.

3-14

~

Table 3.1 Summary of Critical Conditions Necessary to Initiate Self-Sustaining Oxidation Spent Fuel Rack Configuration High Density PWR (6 assemblies per rack)

Inlet Orifice Hinimum Approx. Critical(1)

Diameter Decay Power Decay Time (inches)

(kW/MTU)

(days) 700 High Density PWR (6 assemblies per rack)

Cylindrical PWR Cylindrical PWR Cylindrical PWR Cylindrical BWR Cylindrical BWR 10 1.5 1'

3.0 90 45 14 70 360 10 50(2) 250(2) 180

<10 Critical cooling time is the shutdown time necessary to reach a decay power level below the minimum decay power for self-sustaining oxidation.

The cooling time to prevent cladding failure is at least 20% longer.

Note that these critical cooling times are somewhat lower than that found by Benjamin et al.~ since the orifice loss coefficient was modified at BNL

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 High Power Level Adjacent Power Level (kW/NTU)

(kW/mu)

Approximate Decay Time (days)

Propagation 11.0 19.2 90 90 5.9 5.9 5.9 4.0 365 365 365.

730 Yes Yes Yes No

Table 3.3;,.

Summary of Radial Oxidation Propagation Results for a Cylindrical PWR Spent Fuel Rack wi'th' 3 Inch Diameter Hole and Perfect Ventilation Approximate High Power Level Adjacent Power Level Decay Time (kW/NTU)

(kW/NTU)

(days)

Propagation 90 90 11.0 19 365 180 No Yes*

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

Table 3.4 Summary of Radial Oxidation Propagation Results for a Cylindrical PWR Spent Fuel Rack with a 1.5 Inch Di ame-ter Hole and Perfect Ventflation Approximate High Power Level Adjacent Power Level Decay Time (kW/NTU)

(kW/NTU)

(days)

Propagation

?

90 90 90 15 15 11.0 5.9 3.0 11.0 5.9 365 730 1100 365 730 Yes Yes No Yes No

3-17 Table 3.5 Su+nary of Radial Oxidation Propagation Results for Various PHR Spent Fuel Racks with No Ventilation Spent Fuel Rack-High Power Level Adjacent Power Level (kM/MTU)

(kV/ta'u)-

Propagation Cylindrical with 1.5 inch hole Cylindrical with 1.5 inch hole Cylindrical with 3 inch hole Cylindrical with 3 inch hole High Density with 10 inch hole 90 90 90 19.2 90 5.9 3'

5.9 11.0 4.0 Yes No (0~ depletion).

,No Yes No (0> depletion)

3-18 1

Table 3.6 Comparison of SNL Small Scale Oxidation Tests to Calculations with CLAO Air Flow Rate Test (1pm) 12 28.3

56. 6

Thermocouple failure.

Peak Tem eratures ata A

1570 1900 1400 1850 1960 1660

>2000*

2100 1800

1600 1500 p O-O-1400

~

1300 l~

1200 I<

n.

,~ 1i00 1000 900 c LAD Dale.

0 A top of center pin 0 nlcl-helgbt of center pin rnid-heiglit of inside of liner ooa 0

2 4

6 O

10 12 14 16 10 20 22 24 26 20 30 Vlme from introduction of Oxygen (minutes)

Figure 3. 1 Comparison of CLAD to SNL data for Test 4.

I

~

4-1 4.'ONSEQUENCE EVALUATION A

PWR and a

BWR reactor were selected for risk evaluation based on a pre-liminary screening

- of perceived vulnerability and the spent fuel pool 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.

,II 4 ~ 1 Radionucl ide Inventori es The radionuclide inventories for both the PWR and BWR pools were calcu-lated using the ORIGEN2 Computer Code~ for the actual operating and discharge histories for Ginna and Millstone 1.

The ORIGEN2 program in use at BNL was

'erified by comparison with results obtained at ORNL for identical 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 (Ginna).

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.$ ).

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

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 150 days for a BMR and 30 to 250 days for a PMR).

This leads to a conditional probability for a Zircaloy fire of.28 for a

BMR and

.40 for a

PMR.

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 Zircaloy fire is increased to a virtual certain-ty e 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 yea~s.

Subsequent propagation

- to low power bundles by thermal radiation is highly unlikely, 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 Releases for Self-Sustainin Claddin Oxidation Cases (Cases 1 and 2

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 Nfllstone 1 the probability is about 20>>.

If self-sustaining 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 (25), 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

a

~

4-3 data (tellurium), location of the isotopes (whether in cladding as activation products or in fuel pellets as "fission products),

and melting/boiling points of the element or o'xides of the element.

Comments on the estimates listed in Table 4.2 follow:'ellurium:

The releases shown assume the lower limit of Ref.

4 based on the tellurium release model recently proposed by Lorenz, et al.

The low release value assumes that a fraction of the Zircaloy 5

cladding relocates (melts and flows downward) before oxidation is complete.'lkali Earths:

Because of the high boiling points of the oxides of Sr and Ba, it is estimated that only a very small fraction (2x10-

)

of these elements of fission product origen in the fuel pellets es-cape.

It is estimated that 100K of the activation product Sr-89 and Y-91 contained in the Zircaloy cladding are released as aerosols.

Transition El'ements:

It is estimated that 100>> of the transition element activation products contained in the cladding are levitated as aerosols of the oxides (smoke).

Note that the small release frac-tion of Zr-95 (0.01) takes into account the large inventory of fis-

.sion product Zr-95 trapped in the fuel pellets.

It is assumed that only 10% of the activation products-Hn the assembly hardware escapes (see Table 4.2, Fe-55, Co-58, Co-60 and.Y-91).

The Co-60 fraction is corrected for its small content in the cladding.

~Antimon: 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 lathanides and actinides is estimated because of their chemical sta-bility, low vapor pressures and ceramic characteristics.

4 4

~ ~

~

~ v Case 1:

Case 1, the "worst" case, assumes an, accident.that. results.in a Zircaioy fire that propagates throughout the entire spent fuel in-ventory in the pool, and that. the accident occurs 30 days after the reactor was shut down for discharge of the last fuel batch.

The esti-mated releases of radfonuclfdes* are listed in Table 4.3.

These were obtained by combtnfng.the "30-day" Inventory given fn Column 3 of Table 4.1 with the release fractions listed in Table 4.2.

Case 2

Case 2 assumes an accident that resu'Its in a 21rcaloy fire that involves only the last fuel batch to be discharged, and that the accident occurs 90 days after the reactor was shut down for fuel dis-charge.

The estimated releases of radfonuclfdes are listed in Table 4.4.

These were obtained by combining the inventory in the last fuel batch (data tabulated in Table Am6 of Appendix A) with the release fractions in Table 4.2.

4.2.2 Estfmated Release for Low-Tem erature Claddin Failure (Cases 3 and 4

For a less severe accident in which fuel is exposed to af r but does not

'each temperatures at which a Zircaloy fire ignites, it is assumed that the

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 s1nce the last d1scharge 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-tion of the fuel assemblies

~ the decay time for the last discharged fuel batch fs 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 fn Table 4.5.

~

s~

r 4-5 4.3 Off-Site Radiolo ical Conse uences 4.3.1 Scenarios for. Conse uences 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 site surrounded by a constant population density of 106 persons per square mile;

~

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

and

~

the population in affected zones was relocated after 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

The radiological effects were calculated out to' distance of 50 to 500 miles.

r CRAC2 calculations were made for a

range of possible releases as de-scribed in Section 4.2.

The consequences are summarized in Table 4.7.

4.3.2 Conse uence Results There are several unusual characteristics of a

spent fuel accident that cause somewhat surprising results in the radiation exposure calculations.

Specifically, the radiation exposure is insensitive to fairly large variations in the estimated release.

This is due principally to the health physics assumptions within CRAG.

For the long lived isotopes (predominantly cesium),-

the exposure is due mainly to exposure after the area is decontaminated and people return to their homes.

The CRAG code assumes that decontamination will limit the exposure of each person to 25 rem.

Thus, for this type of release the long term whole body dose is limited by the population in the affected sectors.

(about 0;8 million people in the 16 sectors for a 50 mile radius) to about 3x10~ person-rem (only 3 of the 16 sectors are downwind).

4-6 3 ~

~

The extreme cases (1A; immediately after refueling and 1B and 1C; with the total fuel pool inventory involved) result in much higher releases but no significant change in population dose.'

more sensitive indication of the consequences for. a spent fuel accident is the interdiction area (the area with such a high level of radiation that it is assumed that it cannot ever be decontaminated).

As indicated in Table 5

the worst spent fuel accident is calculated to result in an interdiction area of 224 sq. miles.

This is about two orders'f magnitude higher than the interdiction area computed for reactor core melt accidents (about 1 to 10 mi2).

4.4 References for Section 4

1.

BNL Memorandum, V.L. Sailor and K.R. Perkins to W.T. Pratt, "Study of Be-yond Design Basis Accidents in Spent Fuel Pools,"

May 8 1985.

2.

A.G. Croff, "ORIGEN2:

A Versatile Computer Code for Calculating the Nu-=

clide Composltlons and Characterlstlcs of Nuclear Naterlals,"

Nuclear

~TR I, 2 1. 32, pp. 335-352, 5

p l333.

3.

Internal Memorandum, Brookhaven National Laboratory, from V.L. Sailor to R.A. Bark, "Comparison of BNL ORIGEN2 Calculations with ORNL,"

May 27, 1986 ~

4.

Memorandum of J.T.

Han to M. Silberberg, "Response to a

NRR Request to Review SNL Studies Regarding Spent Fuel Heatup and Burning Following Loss of Water in Storage Pool," U.S. Nuclear Regulatory Commission, (May 21, 1984) ~

5.

R.A. Lorenz, E.C.

Beahm and R.P Wichner, "Review of Telliurium Release Rt I

llIRP lpl lld 3

fd tt dff.'~ft f

the International Meetin on Li ht Water Reactor Severe Accident Evalua-tion, August 28-September 1,

1983, pg. 4.4-1, Anerican Nuclear Society Order

700085, ISBN 0-89448-1112-6.

~

4-7 6.

L.T. Ritchie, J.D.

Johnson and R.H. Blond, Calculations of Reactor Acci-dent Cons'eauences Version 2, CRAC2:

-Com uter. Code, User's

Guide, pre-pared by Sandia National Laboratories for the U.S.

Nuclear Regulatory Commission, NUREG/CR-2326 (SAN081-1994), February 1983.

4-S Table 4 ~ 1 Comparison of Radioactive Inventories of Equi 1 i-,

brium Core with Spent Fuel Assemblies for Select-ed Isotopes (Millstone 1)

Isotooe Equillbrlwa Core Spent Fuel d s woold (time atter Inst d Ischn e) d S

edr (Total Red lcect IvIty, Curl es)

Go 58 Co 60 Kr 85 Rb 86 Sr 89 Sr 90 Y SO Y 91 Ir 95 Hb 95 le 99 Tc 99m Ru 103 Ru 106 Rh 106 Sb 125 Sb 127 Te 125m Te 127 Te 127m Te 129 Te 129m Te 132 I 129 I 131 I 132 Xe 133 Cs 134 Cs 136 Cs 137 Ba 137m Ba 140 Ln 140 Ce 141 Ce 144 Pr 143 Pr 144 H5 147 Sm

'I 51 Eu 154 Eu 156 Np 239 Pu 238 Pu 239 Pu 240 Pu 241 ha 241 Oe 242 Oa 244 8+81 Et4 1,64E+5 585E+5 642E~

4 71E+7 445E&

4Q7Et6 5~06E+7 8.70E+7 8.91 E+7 8,78E+7 7o69E+7 743E+7 2+48 E+7 2 o63E+7 9.07K+5 4e97E&

1 o93E+5 4.92&6 6+61 E+5 f.49E+7 24.4&6 5.72E+7 1.75&0 io74E+7 6i83E+7 9.72E+7 5ol OEt6 2+1 OEK 5.84EW 5.53E+5 846E+7 8.54E+7 1 94E+7 5,05E+7 7 87E+7 6oOSE+7 3,1 6E+7 2 ~44E+4 4,61 E+5 5+61 Et6 9.98KB S 43E+4 2a49E+4 3 14E+4 7.1 SEA 8.86E+3 2.09&6 6e72&4 249Am 3.72E+5 1.41 EK

'I ~OI E+4 849E+6

'I o42&7

'I o43E+7 I 18E+7 1.94E+7 2 o54E+7 1,49Et4 1.43E+l I ~53E+7 1 o72E+7 1.72E+7 1.1 SEA 841 E+3 2o84E+5 2+21 E+5 2 18E+5 2.74E+5 441 E+5 3 o74&4 7'1 5E40 I 42E+6 3.85EM 7 USE+5 7.90&6 2.05E+5 2 o02E+7 I +91 E+7 5.1 SE+5 5.97&6 I Q2E+7 2.64E+7 5,44Et6 2.64E+7 I.54E+6 842E+4 I Q4&6 845E+5 5e59E44 4,$ E+5 So89&4 I QOE+5 2 USE+7 2 88E+5 1.45&6 2 47E+5 I 46EA 3 15E+5 I OS&4

\\ A5E+3

3+63&5 I i42E-:7 1 o42E+7 5.15E&

I oOOE+7 1 o70E+7 3o12E 3 3,01E 3 541 &4 I

53E+7 I o53E+7 1,1 4&6 1 o39E I 2 o76E+5 I o45E+5 I o48E+5 1 79E+4 1 DOE+5 8,6lE 2 7.1 5EK 645E+3 8,90E 2 2 ALOE+2 7,47EK 8.1 3E+3 2+5 E+7 1+SOE+7 1.90E+5 2.1 9E+5 3,61 &6 247E+7 2+41 E+5 2Q7E+7 346EW 841 &4 I Q2E&

5,1 OE+4 2.88E+3 4 53E+5 S,SSEt4 I DOE+5 2Q7E+7 2o94E+5

'I o12E&

245E+5 8,54E+2 2,85E+5 I 43&6 3o84E 2 8D3E+4 I QSE+7 1 QSE+7 241 E+5 5'E+5 1 o1 I E+6 neg.b neg.b 4,07E+4 9.1 3E+5 9.1 3EH 9,48E+5 neg,b 2.31 E+5 2.52EW 2.57EW 2,68E+2 4 12E+2 7.1 5EK ngg neg,b eg b

5.80EH 3o9lE 3 1 e97E+7 1,87E+7 6o41 E-2 737E 2

1,03E+4 I il6E+7 1iSOE I 1 ol 6E+7 1,10E 3

8 16E+4 I 45EH I oSOE I 2 +SSE+3 4e54E+5 8 CASSE~

1 DOE+5 2 19E+7 341 E+5 3o50E+5 2+1 9E+5 Spent fuel pool Inventory Includes discharges tram ll refuel lngs goveg lng the period from August 1972 through the pro)ected refu>>I Ing of Aprl I 1987 ~

neg, e less than I 0 3 Curles.

~

g

~

4-9 Table 4.2 Estimated Radionuclide Release Fraction During a Spent Fuel Pool Accident Resulting in.Complete Destruction of Cladding (Cases 1 and 2)

Release Fractiona Chemical Family Element or Isotope Va ue Used Uncertainty Range Noble gases Halogens Alkali Metals Chalcogens Alkali Earths Transition Elements Miscellaneous Lanthanides Transuranics Kr, Xe I-129, I-131 Cs, (Ba-137m)

Rb Te, (I-132)

Sr, (Y-90),

Ba (in fuel)

Sr, Y-91 (in clad)

Co-58 (assembly hardware)

Co-60 (assembly hardware)>

Y-91 (assembly hardware)

Nb-95, Zr-.95 (in fuel)

Nb-95, Zr-95 (in clad)

Mo-99 Ru-106 Sb-125 La, Ce, Pr, Nd, Sm, Eu Np, Pu, An, Cm F 00 1.00 1.00 0.02 2xl0 s

1.00 0.10 0.12 0.10.

0.01 1.00 1x106 2x10 s

1.00 1 x 10-'

x 10-6 0.5-1.0 0.5-1.0 10 "-10" 2 0.5-1.0 0.1-1.0 0.1-1.0 0.'1-1.0 10-' '.5-1.0 10-e 10-s 10-6-10-4 0.5-1.0 10-8 10" s 10-8 10-s aRelease fractions of several daughter isotopes are determined by their precursors, e.g.,

Y-90 by Sr-90, Tc-99m by Mo-99, Rh-106 by Ru-106, I-132 by Te-132, Ba-137m by Cs-137, and La-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.

4-10 Table 4.3 Estimated Releases of Radionuclides for Case 1

in Which a Zircaloy Fire Propagates Throughou" the Entire Pool Inventory (Morst Case)

Isotone Co 58 Co 60 Kr 85 Rb 86 Sr 89 Sr 90 Y 90 Y 91 Zr 95 Hb 95 Ho 99 Tc 99m Ru 103 Ru 106 Rh 106 Sb 125 Sb 127 Te 125m Te 127 Te 127m Te 129 Te 129m Te 132 I 129 I 131 I 'l32 Xe 133 Cs 134 Cs 136 Cs 137 Ba 137m Ba 140 La 140 Ce 141 Ce 144 Pr 143 Pr 144 Nd 147 Sm 151 Eu 154 Eu 156

+ 239 Pu 238 Pu 239 Pu 240 Pu 241 ka 241 Os 242 Oa 244 b

5 2>>74K+3 4,46EW 1>>41 EK 1>>OI E~

1>>68E44 2>>84K+I 2>>84EW 1>>l BE46 1.63K&

2>>1 3EW 1,49E-2 1>>43E 2 3.06E+2 3>>44K+2 3.44E+2 1.1 9EK 8>>21 E+3 5>>68K+3 4,42K+3 4,36E+3 5,48E+3 8,42E+3 7,48E+2 7.1 5EK 1 42K+5 7>>70E+2 749K+5 7>>90K&

2,05E+5 2.02E+7 1.91 E+7 1>>04&4 1.1 9K+4 1 42K%

2>>64EW 5,44EK 2>>64EW 1.54&0 842K 2 1 >>34&0 8a6E-I 5>>59E 2 4,51E 1

S>>89E 2 140K 1

249EW 248K 1

1>>45EK 247K 1

Time after Lnst Olschnrne b

S (Rnd loactlv Ity>> Cur I es) 1 ~ 51 E+3 3>>78K' 49E46 1>>05K+5 7>>26E+3 2>>84E>>4 i 2>>84&4 5>>75E+5 8>>39E+5 1>>42E46 neg,a neg.a 1,04E+2 3>>06E+2 3>>06E+2 1.1 4E&

1>>39E 1

5>>52K+3 2>>90K+3 2>>96E+3 1.56E+3 2.40E+3 1,72E 3 7>>1 5E&

645E+3 1>>78E 3 2 DOE+2 7>>47M 8>>1 3E+3 2,01 E+7 1>>90E+7 3,80 K+2 4>>38E+2 3.61 &0 2>>27EW 2,41E 1

2>>27&(

3,36E-2 8>>21E 2 1>>32&0 5,10E 2 2>>BBE 3 4>>53E-I S>>89E 2

'I QOE 1

2>>27&i 2>>94E 1

1>>12&0 245'nc 1>>02E+2 3>>42E44 143K&

3>>84E 2 1>>67E+2 2.78EW 2.78&4 2>>21 Et4 446K%

947Et4 neg,a nag.a 8,14E 1

1,83E+2 1>>83K+2 9.48K+5 neg a

4,62E+3 5.04E+2 5,1 4K+2 5.36EK 844K&

neg,a 7>>1 5EK nag,a neg,a neg,a 5.80EK 3,91E 3

1.97E+7 1.87E+7 nag neg,a 1,03E 2 1,1 6EW neg 4 -"

1 16EH neg,a 8,16E 2 1 D5EK nag,a 2>>BBE 3 4>>54E 1

'>>89E 2 I DOE 1

2>>l 9&(

34IE 1

3>>50E-1 2>>1 9E-1 aneg e

leSS than 10 Curlea

4-11 Table 4.4 Estimated Releases of Radionuclfdes for Case 2

in Which Only the Last Oischarged Fuel Batch Suffers a Zi rcaloy Fire I sotooe Co 5S Co 60 Kr 85 Rb 86 Sr S9 Sr 90 Y 90 Y 91 Zr 95 tO 95 Ho 99 Tc 99m Ru 'l03 Ru 106 Rh 106 Sb 125 Sb 127 Te 125m Te 127 Te 127m Te 129 Te 129m Te 132 I 129 I 131 1 132 Xe 133 Cs 134 Cs 136 Cs 137 Sa 137m Sa 140 La 140 Ce 141 Ce 144 Pr 143 Pr 144 Hl 147 Sm 151 Eu 154 Eu 156 Q 239 Pu 23S Pu 239 Pu 240 Pu 241 he 241 Oa 242 Qa 241 oa 5

2 QSEt3 9I1 7E+3 2,39Et5 1 01Et4 1 o79Et4 3I84Et3 3o86Et3 2+66Et4 1,62Et6 2.11Et6 1 o49E-2 1 o43E 2 3 o06Et2 244Et2 2 44Et2 4,1 7E+5 8J.l E+3 1,88E+3 4 QSEt3 440Et3 5,48E+3 8~42Et3 7,4&E+2 8,84E 1

1 42Et6 7o70Et2 749Et5 3.53Et6 2 o05Et5 2 83Et6 2,67E+6 1,04E+5 1.1 9Eti 1.32Etl 1,91 Etl 5o44Eto 1.91 Etl 1.54EtO 931 E 3 2,89E-1 Si37E-1 536E 2 6,BE-2 948E 3 1+55E 2 3.73Eto 6+01E 3 1 41 Eto 5o&SE 2 Time aiteI Last 015cnarae a

5 (Red loactiv lty, Cur les) 1 46Et3 S,68E+3 2 46Et5 1 o05Et3 7e75Et3 3o&2Et3 3,84E+3 1 QOEt4 847Et5 1 +41 Et6 neg o nag.a 1,04E+2 1,99E+2 1 o99Et2 1,00E+5 1,39E-l 1,88E+3 2,80E+3 2 o86E+3 1 o56Et3 2 i40E+3 1,73E-3 8,86E 1

6.35E+3 1,78E-3 2 i30Et2 3,34Et6 Sol 3E+3 2.81 Et6 2,66Et6 3,80E+3 4,38E+2 3.61 Eto 1,65Etl 2 ilE-1 1.65E+1 3,36E-2 930E-3 2,85E-l 5+82 E-2 neg a

6o87E 2 948E 3 1,55E 2 3.70Eto 7oOOE 3 1 +01 Eto 5i84E 2 ear 8,49Etl Sol 2E+3 245Et5 3o84E-2 1 i78Et2 3o78E+3 3 o7&Et3 4+99Et2 445Et4 944Et4 neg a

neg,a 8,1 4E-1 1 ol 9E+2 1.1 9E+2 3+31 Et5 neg,a 1,61 E+3 4,86E+2 4,96E+2 5.36EtO 844Eto neg a

8,86E 1

neg.a neg,a neg.

2.59Et6 3 F91 E-3 2.77Et6 2 +62 Et6 1 QSE-3 neg a

1,03E 2 So43EtO neg.a -"

8,43EtO neg,a 9 45E-3 2,69E-l neg.a neg.a 7,18E 2 948E 3

1,55E 2 3 '6Eto 1 o14E 2 3o16E 1

5+68E 2

aneg,

~ less than 10

Curles,

4-12 Table 4.5 Estimated Releases of Radionucl,ides for Cases 3

and 4 in Which Low-Temperature Cladding Failures Occur Isoto e

Case 3a Case 4b

(

a soactsvsty, urges Kr 85 I 129 I 131 I 132 Xe 133 6 '5E+5 3.58E+0 neg.c neg.c neg.c 2.39E+5 8.84E-1 1.22E+6 7.70E+2 7.29E+5 aCase 3

2 ~

3 ~

assumes:

last fuel discharge has decayed for 1 year.

no Zircalog fire occurs.

50% of the fuel rods develop leaks.

100~ release of noble gases and halogens from leaking fuel rods.

bcase 4

2 ~

3 ~

assumes:

last fuel batch discharged has decayed for 30 days.

no Xircaloy fire occurs.

100~ of fuel rods in last discharge devlop leaks.

100" release of noble gases and halogens from leaking fuel rods.

cneg.

~ less than 10 3 Curies.

~

I

~

+

~

4-13 Table 4.6 Comparison of Radioactive Inventories of Equili-brium Core with Spent Fuel Assemblies for Select-ed Isotopes (Ginna),

Isotone EquiIlbrlum Core Spent Fuel Poola a s (time atter last d Ischarae) ear avs Co 58 Co 60 Kr 85 Rb 86 Sr 89 Sr 90 Y 90 Y 91 Zr 95 Hb 95 lio 99 Tc 99m Ru 103 RQ 106 Rh 106 Sb 125 Sb 127 Te 125m Te 127 Te 127m Te 129 Te 129m Te 132 I 129 I 131 I 132 Xe 133 Cs 134 Cs 136 Cs 137 Ba 137m Ba 140 La 140 Ce 141 Ce 144 Pr 143 Pr 144 Hd 147 Sm 151 Eu 1 54 Eu 1 56 IIp 239 Pu 238 Pu 239 Pu 240 Pu 241 Aa 241 Oa 242 Oa 244 3.57E+5 340K+5 3+73K+5 6e53E+5 3.55E+7 2.95K&

3+i 5&6 i+57K+7 6+41 E+7 644K+7 6+83K+7 5,89E+7 5 o85E+7 I

95E+7 2 el 5E+7 6,04E+5 4o12&6 1 47K+5 4,05K+5 5ol 9E+5 I gl E+7 1 oBOEt6 5 +33K+7 1 47EtO 3.76K+7 5+42K+7 7,64E+7 5.82E+6 1,87&6 4 o21 Et6 4,00Et6 6~55K+7 6,74E+7 648K+7 4 44K+7 5,71 E+7 4 47K+7 2,48E+7 1 o42&4 4.09E+5 742K&

7,81 Et8 1 oOI K+5 1 D5&4 2 o02EM 4o85Et5 4,99E+3 1.91 E&

I 45K+5 (Total Rad loactlv Ityi 5.93E+4 5.97E+5 9+84K+5 742K+3 3.53EW 1 s02E+7 1.02K+7 5.11EW 8.64EW 1 o12E+7 7.03E+3 6 77E+3 7.86K&

I +09K+7 I +09K+7 7,1 1 E+5 4.33E+3 1,70E+5 1,1 9E+5 1.1 7E+5 1.38E+5 2ol2E+5 1.83K~

5.32&0 6,00E+5 1.89EW 3.52E+5 6.35E+6 1 46K+5 1

48E+7 1,40E+7 2.47&6 2.85EW 6D4&6 1 o38E+7 2 ~ 54K%6 1.38E+7 7,42E+5 5.1 4EW 1 o09EW 7e58E+5 3+02K+4 4,46E+5 545&4 So60&4 1 o52E+7 2 II OE+5 943K+5 3.59E+5 Cur les) 346&4 5,84E+5 9 o74E+5 7,48E+2 1 o53E&

1 01E+7 1,01 K+7 2 ~48Et6 4 46K%6 7+51 K+5

'1,48E 3 I i42E 3 2+88K&

9+71 Et6 9.71 E&

6+82K+5 735K-2 I I65E+5 7o79EW 7+95EM 3o93E+4 6o03&4 443K 2 542K&

3,1 2E+3 4 36K-2 1,1 1 E+2 6,00Et6 4,99E+3 1 <47K+7 1 o39E+7 9.07EW I o04E+5'

+72&6 1 oi 9E+7

'I +I 2K+5 1,1 9K+7 1.62EM 5+i 3EM 1,07Et6 4,68K+i 346K+3 4,46E+5 545&4 S i60E44 1 o5l E+7 2+1 4E+5 740K+5 3o56E+5 241 E+3 549K+5 947K+5 2 o74E-2 3~50EW 9~95K+5 9.95K+6 9 o54E44 247K+5 i+93K+5 b

neg.

neg ~

2.09K+4 5.78EH 5.78EH 5+65K+5 neg b

1 D7E+5 1 46K+4 1 QBEW 1,35E+2 2 07K+2 neg.b 542EW neg, b neg,b neg,b 4.66E+6 2,40E-3 1 ~44K+7 1.37K+7 3o05E 2 3+51 E-2 4 91E+3 6.09EH 8,86E 2 6+09K+6 neg 5.1 OE~

1 OIEt6 1 o66E I 306K+3 4 o46E+5 545EW 8~61 Et4 1,46K+7 2D2E+5 2 45E+5 3+46K+5 espent tuel pool Inventory includes discharges trae 15 retuel lngs cover lng the period from hprll 1983 through the proJected refueling of hprll 1987

'neg, a less than 10

Curles,

4-14

~

~ ~

Table 4.7 CRAC2 Results for Various Releases Corresponding to Postulated Spent Fuel Pool Accidents with" Total Loss of Pool Water Case Description 1A.

Total inventory 30 days after discharge 50 mile radial zone 1B.

Total inventory

. 90 days after discharge 50 mile radial zone 1C.* Total inventory 30days after discharge 500 mile radial zone 2.

Last fuel discharge 90 days after discharge 50 mile radial zone Whole Body Dose

~

(Man-rem) 2.6x10e 2.6x10 7.1x10~

2.3x10 Interdiction Area (sq. miles)

'24 215

'24

Note that the consequence calculations in NUREG-1150 are based on a 50 mile radial zone.

Case 1C is given as a sensitivity result.

II E%

~

g 5-1 5 ~

R!SK PROFILE de The likelihood and consequences of various spent fuel pool accidents has been estimated in the previous sections.

The risk is sumnarized in Table 5.1.

As previously mentioned, the exposure results are'ied to the health physics assumptions regarding decontamination and maximum allowable exposure.

Thus the 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'rder of magnitude in either'irection and is due princi-pally to uncertainty in the fragility of the pools and uncertainty fn the seismic hazard.

Note that the risk results are calculated for two surrogate plants and may not be applicable to generic pool types.

5.1 Failure Fre uenc Estimates 5.1.1 S ent Fuel Pool Failure Probabilit The likelihood of the various postulated spent fuel pool accidents was developed in Section 2 and summarized in 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 these potential accidents are plant specific and their frequency will vary widely from plant to plant.

In parti'cular, BWR's do not have pneumatic'seals so their failure frequency is zero.

5-2 5.1.2 S ent Fuel Failure Likelihood Previous investigations

~

of spent fuel behavior after a loss of pool integrity accident -focused'n the conditions necessary to initiate cladding "fires" after a

spent fuel pool has drained.

The present project has reevaluated these conditions using the SFUEL code~

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 I.O while for a BMR with low density storage racks the proba-bility of a clad fire was found to be 0.08.

5.2 Conclusions Re ardin Risk The overall risk due to beyond design basis acridents in spent fuel pools for the PMR surrogate plant fs about 130 person-rem/Ry and about 12 person-rem/Ry for the BMR surrogate.

These estimates are comparable to present esti-matess 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 CRAG 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.

0 The uncertainty in risk in terms of person-rem/Ry is driven principally by the uncertainty fn the likelihood of complete draining of the spent fuel pool which fs estimated to be at least an order of magnitude in either di rec-tion.

5.3 References for Section 5

1.

A.S.

Benjamin, D.J.

McCloskey, D.A.

Powers, S.A.

Dupree, "Spent Fuel Heat-up Following Loss of Mater During Storage,"

NUREG/CR-0649, triarch 1979.

0

5-3 2.

N.A.

Pi sano, F.

Best, A.S.

Benjamin, K.T. Stalker, "The Potential for Propagation of a

Self-Sust'aining 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).

5 4

~

Table 5.1 Estimated Risk for the Two Surrogate Spent Fuel Pools from the Two Oominant Contributors Accident Initiator Seismic induced PWR pool failure Seismic induced BMR pool failure Cask drop* induced PMR pool failure Cask drop* induced BMR pool failure Spent Fuel Pool Fire Probability/Ry 1.6x10-s 1.8xlO-e 3.lxlO-s 2.5xlO-e Health Risk (Han-rem/Ry) 37 71 6

Interdiction Ri sk (Sq. Ni./Ry) 8.4xlO "

7.6xl0 s

~001 l.lx10 4

After removal of accumulated inventory resumes.

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

6-1 6.

CONSIDERATION OF RISK REDUCTION MEASURES Due to diversity in the nature of initiating events 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 performed 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 Inventor

-. Most of the consequences of a

release of radioactivity from a catastropic 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. 'ne obvious measure for risk reduction is to transfer part of the inventory to alternative storage locations (e.g.,

see Ref. 1).

2.

Air Circulation - The one universal prevention measure 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 even old speot 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 then the likeli-hood of self-sustaining oxidation would be reduced by a factor of 5 or more compared to the high density storage configuration.

6-2 3.

Additional Coolin S stems - Although loss-of-pool cooling appears to be risk s1gn1ficant, an add1tional cooling system 1s, unlikely to be cost beneficial (unless the cooling system was substantially more unreliable than the two,surrogate sy'stems).

An additional cooling system would not affect the risk from pool failure events (seismic or cask drop acc1dents).

Thus the net risk reduction would be minimal unless loss-of-cooling were the dominant event.

4.

Im roved Procedures and E ui ent - The likelihood of cask drop acci-dents can be reduced by 1mprov1ng procedures, administrative controls and/or 1nstalling more reliable equ1pment.

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 specif1c basis.

It would appear to be useful to 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 response to the loading from a dropped cask.

6.2 Accident Miti ation 1.

Post-Accident S ra

- Water spray 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 cont1nue 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 building.

For those plants without sprays available, 1t seems.unlikely that the.ex-pense of a new safety grade spray system could be Justified consider-ing the large uncertainty 1n the r1sk.

'Temporary fire hoses were suggested by BenJamin et al.,~ but the radiation levels would make such ad hoc measures extremely 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 oxidation.

(The steam/D rcaloy reaction is also highly exothermic.)

~

'-3 2.

Filterino - For those pi ants with a

standb'y gas treatment system available, operation of the system has the potential to substantially'educe 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 Re ardin Preventive and Hiti ative 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 (a 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-

'ore 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, CAPRI NP-3365, February 1984.

2.

A.S. Benjamin,,D.J.

HcCloskey, D.A. Powers, S.A. Dupree, "Spent Fuel Heat-up Following Loss of Water During Storage,"

NUREG/CR-0649, March 1979.

~1 I

~l II ~

0

~

s ~

'I ~

A-1 APPENDIX A RADIOACTIVE INVENTORIES A.l INTRODUCTION Two older-vintage plants, a

BWR and a

PWR, were selected to serve as sur-rogates for estimat1ng the risks associated with "Beyond Design Basis Acci-dents in Spent Fuel Pools."

The purpose of thiC 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 1n the fuel assemblies stored in the spent fuel basins.

The surrogate plants were Millstone-1 (BWR) and Ginna (PWR).

A.2 SIMULATION OF OPERATING HISTORIES A.2 ~ 1 Thermal Ener Product1on 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 schedul ed

~ 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 1n Fig. A.l.}

~ A.2.2 Fu'el Bur nu 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 t

Northwest Laboratories (PNL) for the U.S. Department of Energy.

It should be

A-.2 noted that the inventory of spent fuel assemblies stored in the spent fuel basins at various points in time'isted 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 percent from the burnups calculated by the methods described in the following paragraphs.

These discrepancies do not have significant effects'n 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.,

(HWth/MT)i (MWthD)i/Di(MT) where for operating cycle, i, MWth/MT is the average specific power per met-r ic tonne of initial heavy

metal, (MWthD)i is the total thermal ener gy produced in Di days of the cycle, i, and MTcore 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, (MNt,MD/NT)) f (NN~M/MT))0) where is the su+nation over the several operating cycles, i, that the fuel

'was in the reactor.

(As noted

below, ORIGEN2 also calculates the specific burnup which provides a check on internal consistency 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

A-3 total thermal energy production over the preceding history of the plant (e.g.,

see Table A.4).

A.2.3 Calculation of Radioactive Inventories The average rad1onuclide content in each metric tonne of discharged fuel was calculated using the ORIGEN2 Computer Code.<

The code treats the reactor core as a homogeneous body operating at an average specific power.

Account is, taken of radionuclide decay dur ing and following irradiation, decay

chains, and success1ve neutron captures.

The BNL version of ORIGEN2 was benchmarked against the version in use at

, Oak Ridge Nat1onal Laboratory by calculating an identical

case, which yielded 1dent1cal 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 calculat1ons were made to select an appropr1ate set of t1me steps for which the sens1tivity was negligible.

(Shor ter time steps give higher precision results, but at the expense of increased computer time.

The crite-r1on adopted was that the time-step sensit1vity be less than 0.1~ in the cal-culated concentration of several key nuclides.)

In a mature operating nuclear power plant fuel management strateg1es are complicated (e.g.,

see Ref.

8).

Host fuel assemblies remain 1n the core for several operating cycles and are often shifted fn location during refueling so as to optimize burnup.

Also, U-235 enr1chment is var1ed.

ORIGEN2 as used at BNL did not take account of such detail, nor of the ax1al and radial distribu-t1on of the. power density.

Thus, the radioactivity calculated for a part1cu-lar assembly would not correspond exactly to an actual assembly.

Neverthe-less, the total calculated radioactivity in a d1scharged batch should be iden-tical to total fn 'a real batch (in so far as the precision 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.

II

A-4 As used at BNL, the input for each irradiation cycle is the average spe-cific power and the length of the cycl'e.

ORIGEN2 calculates the total average burnup of each fuel batch over the irradiation cycles during which ft was in the core.

This calculated burriup was cross checked against "hand" calcula-tions for each batch',

the "hand" calculations bei ng 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 fn the fuel cladding and assembly hard-ware.

For each plant, BMR and PMR, only a single fuel and assembly composi-tion was modeled which fs 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, and thermal power produc-tion of each radionuclide.

These are given at specified decay times for acti-vation products (in cladding, hard~are and trace elements in the fuel pel-lets), fission products and actinfdes.

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 C cle Parameters Table A.1 summarizes several of the major reactor characteristics and fuel cycle parameters for Millstone 1.

A.3.2 Hfstor of 0 erations 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'lant

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 averaging about 57 days.

A mor e detailed narrative of the plant operating hi story from 1970 through 1981 appears in Ref.,10, Appendix 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.

Since mid-1981, the plant has operated with nearly 100% unit service factor except for scheduled refueling outages.4 There have been 10 refueling campaigns since beginning of commercial 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 occurred at some-what irregular intervals, being dictated by unscheduled forced outages.

- Since

1981, refueling has been scheduled for approximately 18 month intervals, oc-curring in April or September.

During the lifetime of the plant the average fuel burnup has generally increased from about 20,000 MWD/HT in 1972 to about 28,000 MMD/MT at present.

A.3.3 BMR Fuel Assembl Model Used in ORIGEN2 Calculations A nominal BMR fuel element has been

modeled, based

'on data presented in Ref. 9.

This is an 8x8 element assembly of 2.75$ U-235 enrichment, containing 1.5873 kg of gadolinium burnable poison per metric tonne of uranium.

The fuel cladding is Zircaloy-2.

Other alloys present in the fuel assembly hardware include Zircaloy-4, Inconel X-750, SS302 and SS304.

The alloy contents of the assembly hardware are included with weighting factors to take account of the axial variation 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'or the presence of "crud" composed of Fe, Co, and Ni on the outer surfaces of the cladding and assembly hardware.

~

g.

A-6 A.3.4 Calculated Radioactive Inventor ies The calculated inventories of selected radi'onuclides* are listed in Table A.5 for the reactor core 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, 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, Ba-137m, Eu-154, Pu-239, Pu-240, Pu-241, Am-241, and Cm-244.

Table A.6 gives a comparison of the radionuclide 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.

3.3.5 ~gil Table A.7 suamarizes the decay thermal production in the 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 sumnar izes 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.

~fh

'I I

dl 115 3

d I

I I

I 1dl gp tial for biological concern, thermal

power, and total curies of activity.

c A-7 A.4 DATA FOR GINNA A.4.1 Reactor and Fuel C cle Parameters Table A.9 summarizes several of the major reactor characteristics and fuel cycle parameters. for Gi.nna.

A.4.2 Histor of 0 erations Several milestones in the operation of Ginna are sunmarized in Table A.10.

A narrative of the operating history from 1969 through 1979 can be found in Ref. 12, Appendix 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-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.ll, but repeating the ORIGEN2 calculations did not seem worthwhile since only minor changes in the spent fuel radioactive i n-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.~3 At the time of the April 1972 refueling, cladding distortions due to fuel densification was discovered and 61 assemblies were replaced (Ref.

12, pg.

F-56).

Thus, the entry in Table A.11 for the second discharge is incorrect.

The total burnup not accounted for fn the ORIGEN2 calculations amounts to 4..2% of the total thermal energy production from 1969 through April 1, 1987.

t The missing 4.2% burnup is for fuel discharged on or before April ],972.

A-8 A.4.3 PWR Fuel Assembl Hodel Used in ORIGEN2 Calculations A nominal PMR fuel element has been modeled based on data presented in Ref. 9.

This is a

17x17 element assembly (264 fuel elements 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 resu'its in lower neutron flux which results in lower 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.

'.4;4 Calculated Radioactive Inventories The calculated inventories of selected radionuclides* are listed in Table 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 April 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, Ba-137m, Eu-154, Pu-239, Pu-240, Pu-241, Pm-241, and Cm-244.

f dl lid h

d I

t H

I 1digp tial for biological concern,-thermal power and total curies of activity.

A-9 Table A.13 gives a comparison of the radionuclide inventories in the last fuel batch to be'discharged with the summation of the inventories contained in batches 2-15 discharged. between 1976 and 1986.

2.5.5 ~hh Table A.14 summarizes 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 summarizes 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 fs dominated by Cs-134, Rh-106 and Pr-144.

The actinides are relatively small contributors.

A.5 REFERENCES FOR APPENDIX A 1.

Nucleonics

Week, a weekly newsletter published by McGraw-Hill Publishing.

Co.,

New York, NY.

2.

~II@

Af, dhllhd hl hly hy h

5 I

Af y If

Center, Oak Ridge National Laboratory.

3.

Nuclear En ineerfn International, published monthly by IPC Business Press, Ltd., Sussex, England.

4.

U.S.

Nuclear Regulatory Comnission, Licensed 0 eratin

Reactors, NUREG-0020, Vols. 1-10, published monthly.

5.

U,S.

Department of Energy, Richland Operations

Office, Program

Office, Commercial Spent Fuel Management, (private communication from P.A. Craig, Director, June 11, 1985).

A-10 s ~

I 6.

A.G.

Cr off, "ORIGEN2:

A Versatile Computer Code for Gal cul ating the Nuclide Compositions and Characteristics of Nuclear Materials,"

Nuclea.

~ih 1

. i i. 62. ii. 335-352; i i i iiii.

7.

Internal Memorandum, Brookhaven National Laboratory from V.L. Sailor to R.A. Sari, "Comparison of BNL ORIGEN2 Calculations with

ORNL, May 27, 1986.

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

~Soclet, Vol. 32, pg. 706, June 1979.

9.

A.G. Croff, M.A. Bjerke, G.W. Morrison, and L.M. Petrie, Revised Uranium-Plutonium C cle PWR arid BWR Models for the ORIGEN Computer

Code, Oak Ridge National Laboratory, ORNL/TM-6051, September 1978.

10.

U.S.

Nuclear Regulatory Comnission, Inte rated Plant Safet Assessment S stematic Evaluation Pro ram, Millstone Nuclear Power Station, Unit 1, NUREG-0824, February 1983.

11'.S.

Nuclear Regulatory Commission, Nuclear Power Plant 0 eratin Ex eri-.

ence, 1974-1975, NUREG-0227, April 1977, pg. 65.

12.

U.S.

Nuclear Regulatory Commission, Inte rated Plant Safet Assessment S stematic Evaluation Pro ram,,

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 Storage Capacity," submitted to NRC April 2, 1984, Docket No.50-244.

Table A.l Reactor and Fuel Cycle Parameters for Millstone 1

(Sources:

Refs.

1-4)

Assemblies in core:

580 Licensed thermal power:

2011 MWth (gross)

Thermal power corresponding to maximum dependable capacity:

2006 '

MMth (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 interval (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 C

Average period of irradiation (including downtime):

about 72 months Authorized Storage Pool Capacity (as of 1985):

2184 assemblies

Table A.2 Summary 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 Une:

100,307.9 hours1.041667e-4 days 0.0025 hours 1.488095e-5 weeks 3.4245e-6 months Gross Thermal Energy:

184.83 x 10'Wh Capacity Factor (MDC net):

67.4"

Table A.3 Sugary of Spent Fuel Batches in Hillstone 1 Storage Basin (With Projections to 1987)

Spent Fuel Batch No.

Date of Number End of'f Irradiation Assemblies Weight H.H.

(HT)

Avgas Burnup (HWO/HT)

Decaya Days to 5/1/87 (days)

Cumulative Assemblies in Pool=

Cumulative" Gross Weight of Spent Fuel in Pool (HT) 1 2

3 5

6 7

8 9

10c 08/31/72 08/31/74 09/11/75 09/30/76 03/10/78 04/27/79 10/03/80 09/11/82 04/12/84 10/01/85 04/01/87 28 204 144 124 124 148 168 192 172 178 167 5.132 38.126 26.395 22.729 22.729 27.128 30.794 35.194 31.528 32.627 30.611 12686 19695 26581 21290 24090 24354 24998 23670 26763 28052 29963 5356 4626 4250 3865 3339 2926 2394 1693 1114 577 30 28 236 380 504 628 776 944 1136 1308 1486 1653d 8.95 75.47 121.52 161.18 200.83 248.16 301.89 363.29 418.30 475.22 528.63 aDecay days from end of irradiation to 5/1/87.

bGross fuel tonnage in pool includes heavy metal plus cladding and hardware but not including fuel racks.

Each assembly contains approximately'0.1833 metric tonnes of heavy metal, 0.0246 tonnes nf oxygen (in UO>) and 0.1119 tonnes ofihardware, totaling 0.3198 metric tonnes gross.

cProjected data.

dThe present authorized storage capacity is 2184 assemblies.

After the 04/01/87 refueling, the accumu-lated assemblies plus the 580 assemblies in the core would exceed the authorized storage capacity should a full core discharge be required.

Table A.4 Comparison of Cumulative Gross Thermal. Energy Production."

with Calculated Fuel Burnup from Start of Operations in 1970 to April 1, 1987 (Millstone 1)

Total Cumulative Gross Thermal Energy (MWD x 10 s)

N Total 8440.25 Spent Fuel Batch No.

~

9 10 12*

13*

Total Burnup in Batch (HWD x 10 a) 65.10 750.88 701.61 483.91 547.54 660.68 769.78 833 F 05 843.78 915.25 917.21 612.74 329.55 8440.01

Burnup in fuel remaining in the core.

A-15 Table A.5 Comparison of Radioactive Inventories in Reactor Core and Spent Fuel Basin (Millstone 1).

The Assumed Refueling Scenario is Described in Section'A.3.4 Isotone Reactor Core Soent Fuel (Rad Ioactl vIty, Stol aoe Baslno

~7'urles)

H3 014 Co 5S Co 60 Kr 85 Rb 86 Sr 89 Sr 90 Y 90 Y 91 Zr 95 Nb 95 llo 99 Tc 99m Ru 103 Ru 106 Rh 106 Sb 125 Sb 127 Te 125m Te 127 Te 127m Te 129 Te 129m Te 132 I 129 I 131 I 132 Xe 133 Cs 134 Cs 136 Cs 137 Ba 137m Ba 140 La 140 Ce 141 Ce 144 Pr 143 Pr 144 Hd 'l47 Sm 151 Eu 154 Eu 156 Hp 239 Pu 238 Pu 239 Pu 240 Pu 241 fia 241 Cm 242 Cm 244 4o95E+4 1.02E+2 S+SI Et4 1 o64Et5 5+35Et5 642Et4 4.7l E+7 4 45Et6 4.37Et6 6,06E+7 BoVOEtV 8,91 E+7 8.78E+7 7,69E+7 743K+7 2,48E+7 2 e63E+7 9oOVE+5 4.97E+6 1.93K+5 4.92E+6 6,61 K+5 I +49K+7 2 44E+6 6,72E+7 1.75Et0 4,74E+1 6,83Et7 9oVZE+7 6,1 OE+6 2,1 OEt6 5.84Et6 5.53Et6 8+36K+7 S,54E+7 7.94K+7 6+05K+7 7 47K+7 6+08K+7 3.1 6E+1 2,44Et4 4o6I E+5 5+61 Et6

'o98EtS 9e33Et4 2 o49Et4 3e1 4Et4 7.1 9Et6 So86E+3 2~09Et6 6eVZEti I QSEt5 4el2Kt2 249Et4 3.72E+5 1.41 Et6 1,01 Et4 8.39Et6 1,42E+7 1.43K+7 1 ~I BE+7 1.94E+7 2,54E+7 1,49E+4 1.43Eti 1,53E+7 1.72E+7 1.72E+7 1.1 9Et6 SQI E+3 2 o84Et5 2QI E+5 2 ol BE+5 2,74E+5 441 E+5 3,74Et4 7.1 5Et0 I 42Et6 3.85Et4 749Et5 7,90Et6 2,05E+5 2,02K+1 1.91 E+7 5.19Et6 5,97Et6 1,32E+1 2,64E+7 5,44Et6 2o64E+7 1 ~54Et6 SMEt4 1 44Et6 846Et5 5+59Et4 4+51 E+5 8,89Et4 1 QOEt5 249K+7 2 BBEt5 1.45Et6 247Kt5 1.37Et5 i,l2 E+2 I 46Et4 3.1 5E+5 I 49Et6 1 o05Et3 3.63Et6 I +42K+7 1 +42K+7 5.75Et6 1 oOOE+7 1 eVOE+7 3.1 2E-3 3iOI E-3 541 Et6 I +53K+7

~ I 53E+7 1.1 4E+6 I i39E I 2,76E+5 1,45K+5 1.48E+5 7.79K+4 I 40K+5 Bo64E-2 7.1 5Eto 6+3 5E+3 8,90E-2 2 30E+2 7,47E+6 8,1 3E+3 2,01 E+7 1 o90EtV 1.90E+5 Zol 9K+5 3.61 E+6 2 47K+7 2,41 E+5 2 QVEt7 3.36Et4 SQI E+4 1.32Et6 5'Et4 2.88K+3 4.53K+5 8,89Et4 1 DOEt5 247K+7 2o94Et5

'I el2Et6 245K+5 1,35E+5 4,12K+2 5.12E+3 3+04K+5 1 QVKt6 3.44Etl I o03Et6 1 +41 E+7 1 o4I K+7 I o98Et6 3.70Et6 7.35Et6 nege neg ~

I e03Et6 I 49EtV 1 49K+7 I +OVEt6 neg o 2,61 E+5 8,06Eti 8 43Et4 1.1 VEt4 I o79Et4 neg,b Vol 5EtO 2 QBEt0 neg e IQIE 3 6+86Et6 646EtI 2+OOK+7 I o89EtV 1 OOEt3 1 o50Et3 5.07E+5 1,$ E+7 2+1 9E+3 I FBI E+7 1.05K+2 S.l 9Eti

'I 49Kt6 1,16E+2 2 BBE+3 4.54Et5 So89Et4 1 e30Et5 245K+7 3o03Et5 V+60K+5 243Et5 I Dt E+5 4 ol 2Kt2 So54E+2 2~85Et5 1.33E+6 3i84E-2 843Et4 1 D9E+7 1 +39K+7 2DI E+5 5,1 OK+5 I 11Et6 neg,b neg.b 4.07Eti 9.1 3K+6 geI 3Et6 9,48K+5 neg,b 2+31 E+5 2.52Et4 2.51Et4 2 o68Et2 4+i 2E+2 neg,b neg.b negi neg e 5.80Et6 3,9'I E-3 I

97E+7 I,BVE+7 6o41 E-2 1 e37E-2

-t.03K+4 1 ol 6E+7 I o90E-I 1,1 6E+7 I oI OE-3 8 el 6Et4 I 45Et6 1,80E-I 2oBSEt3 4 54E+5 S.SQEti 1DOEt5 2'E+1 341 E+5 3.50K+5 2.1 9K+5 aspent fuel pool InventorY Includes discharges tran ll refuel lngs cover-ing the period trom hugust 1972 through the pro)ected refueling ot April

1987, bneg, e less than 10 3 Curios,

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)

Isotone Sooht Fuel Batch 11 a (Rod looctlv Itye Curios) ant Fuel Batch 1 10b~'AH H3 C 14 Co 58 Co 60 Kr 85 Rb 86 Sr 89 Sr 90 Y 90 Y 91 Zr 95 Nb 95 Mo 99 Tc 99III Ru 103 Ru 106 Rh 106 Sb 125

"'b 127 Te 125e To 127 Te 127m Te 129 Te 129'e'132 I 129 I 'l31 I 132 Xe 133 Cs 134 Cj 136 Cj 137 Ba 137la Bo 140 I.o 140 Ce 141 Ce 144 Pr 143 Pr 144 Hd 147 Sel 151 Eu 154 Eu 156 Hp 239 Pu 238 Pu 239 Pu 240 Pu 241 ha 241 Ce 242 aa 244 2 e27E+I 5.1 BEW 2e28&4 Te64E t4'49K+5 1.01 &4 SQQE&

1.93&5 1.93&6

'I el BE+7 1 e94K+7 2e53E+7 1 e49E44 1,43&4 1 e53E+7 1 el2E+7 1,1 2E+7 4.1 TE+5 841 E+3 9o42 &4 2 el 4E+5 2,1 OK+5 2o7lf+5 ldl E+5 3e74E+l 8,84E 1

I MKW ~

3e85ft4 7e29E+5 3.53&5 2 05E+5 2.83K&

2.67&6 5.1 9EW 5.97EK 1.32E+7 1.91 K+7 5.44' 91 E+7 1,54EW 9.31 E+3 2 89E+5 BASE+0 5e36EW 5.73 BC 948K+3 1 e55Et4 3.73&6 6 01K+3 1M&6 5.88'e24&4 5ol SEW 146EW.

Te48E44 2D6E+5 1 o05E+3 3.63&6.

1 e92EK 1.92&6 5.74&6 1 oOOE+7 1 e69E+7 3,12E 3 3o01 E 3 5e21 E46 9e98E+6 9e98E+6 loOOE+5 1,39E 1

949EM l,lOE+5 1.43E+5 7.79EW 1 DOE+5 Bo64E 2 Be86E 1

6e35E+3 Be90E 2 2,30E+2 3o34f46 8,1 3E+3.

2.82K&

2o66B6-1.90E+5 2.1 9E+5 3.61 E&.

1,65E+7 2,41 E+5 1

65E+7 3D6EM 9o30E+3 2,S5E+5 5el Ski 546K%

6o87&4-948E+3 1 e55EW 3.70&6 7eOOE+3 1

01 ft5 5e84EW 2 e21 E+4 5.1 BEW 5el OE+3 Te24ft4 2e32f+5 3e44E tl 1 e03&6 1e91 &6 1 e91 E+6 1.93&5 3e69&6 7e33&6 nag

nag, 1 e03f46 Se40E46 Se40E&

3.76E+5 flag ~

9,04K%4 To79&4 7e95&4 1 ol 7&4 1 79EM nag Bo86E 1

2QS&0 neg ~

1 e21E-3 3.07EW 6e26EW 2 oBOEt6 2o65E&

1 e30E+3 1 o50E+3 5.07K+5 I o32E+7 2 el 9E+3 1,32E+7 1,05E+2 9e28E+3 2e79E+5 7.76E+2 546K+2 7e02E 44 948 E+3 I e55EM 3.65&6 SelSE+3 6,86E+5 5e79&4 2,1 5&4 5.1 SEW Se49E+2 6e77E+4 2o25E+5 3e84K 2 Bo33&4 I e89E45 1.89K&

2o21 E+5 5o09E+5 1.11 &6

neg, nag, 4 o07Et4 5.95E&

5e95&6 3o31 Et5

nog, Se07&4 2 o43Et4 2e48EM 2e68E+2 4el 2E+2 neg ~

Be86E-1 nag

neg, hog ~

2.59&6 3o91E 3 2.77&6 2.62K&

6o41E 2 7o37E 2 1 e03EW 8.43K+6 I e90E-1 8o43E+6 1 elOE 3 9o25E+3 2e68E+5 I e83E 1

5o26E+2 Tel Bfti 9e28E+3 1 e55EW 3.56K&

1 el 4E+4 3el 6E+5 5,68E t4 1 ol 6E+5 3e61 E+2 1 el BE+2 2.45E+5 I el 7&4 heg 5o39E+3 1 e23E+7 1 e23E+7 2el 1 &4

'ogOEW

'I e31 E+5 neg

nag, 1 e09E+3 5e98&6 5.98&6 Te76E+5 nag ~

lo89E+5 7el 5K+3 7o30E+3 3o85E40 5.91 &0 neg 6e26E40 nog ~

neg ~

nag ~

ie37EK

neg, 1 e73E+7 I e64E+7 nag ~

neg ~

I e31 E+2 7e23E+6 neg 7e23EK

nag, Te29Eti 1 e05&6 neg 2e35E+3 3e84E+5 7 e96E+4 1 el 5E+5 1 e92E+7 2e82E+5 1 e39E+5 I e68E+5 1 ol 5E+5 3.61 E+2 6e48EW 2e40E+5 1 el 6E&

heg o 2o33E+3 1 o23E+7 1 e23K+7 1 e02E44 3e05E44 6.76EA hog e hag ~

3e73E+2 5e30&6 5.30E&

7e44E+5 flag ~

1 e82E+5 4 e85E+3 4 95E+3 1 e09E tO.

le68E40 heg ~

5e26&0 hag e heg ~

heg ~

4,1 3&6 heg o 1 e73E+7 1 o63E+7 heg e neg e 3e57bl "

643&6 nag e 6e23&6 heg ~

To28&4 1.04EW flog ~

2e35E+3 3.84E+5 7e96E44 1 el 5E+5 1,91 E+7 2.87E+5 1,08E+5 1.67E+5 1 el 3E+5 3o61 K+2 2 63E+1 2e32E+5 1 el 4&6 neg ~

6,60E+2 I 42K+7 I o22E+7 3e44E+3 I el2E+4 2o49EM flag ~

nag, To35K+l 4e48&6 4 o48K+6 6o99K+5 nag.

1 e70K+5 2eTOE+3 2e76E+3 1,64K 1

2o52K 1

neg e 6e26E40 flag ~

flag ~

nag.

3oSOEK nag ~

1 e72E+7 1 e63E+7 neg o neg.

5o02E&

4o98&6 nag 4o98EW nog o To26EW 1,02 &6 ne9 ~

2e35E+3 3o83K+5 7.96&4 1 el 5K+5 1 eSBE+7 2e95E+5 Te33&4 1 e65E+5 1 ~ 'I OE+5 3,61 E+2 4o39E40 2 el TE+5 1 el OE46 hag ~

5.35EW 1 e21 K+7 1 e21 E+7 3o94E+2 1.55E+3 3.44K+3

nag, nag, 2.91 E&

3.1 8K+6 3ol SE&

6el 6K+5 nag ~

1 e50E+5 8,44K+2 8,62E+2 3,76E 3

5e77E-3 nag.

6e26EK nag ~

nag ~

3e21 E+6 nag ~

1,70K+7 1.61 E+7 ha9 ~

nag, 1,01 E-1 3.1 9E e6 hag o 3ol 9EW5

nag, 7e24E+4 9.75E+5 nog o 2o35E+3 3e82E+5 7.96K+4 I el 5E+5 1.84K+7 3e09E+5 3.47K+4 1 o62E+5 afuol batch 11 Is pro)ected discharge durl'ng April 1987 ~

Fuel botches 1 10 re~a discharged between august 1972 ond October 1985, cneg

~ loss than 10 Curios

Table A.7

'Decay Heat Released from Spent Fuel Inventory for Various Discharged Fuel Batches (Millstone 1)

Deca Heat Released.b

.Batch O,E"d f 8

El ~,

9 Irradiati on>,

(Petri c Tonnes)

(Ki owatts, herma 1

2 3

5 6

7 8

9 10 11 08/31/72 08/31/74 09/11/75 09/30/76 03/10/78 04/27/79 10/03/80 09/11/82 04/12/84 10/01/85b 04/01/87b

5. 13 38.13 26.40 22.73

22. 73

27. 13

30. 79 35.19 31.53 32.63 30.61 1.8 22.0 21.8 15.2 18.4 23.5 30.3 41.5 67' 146.0 909.0 1.8, 21.9 21.7

~ 15.1

18. 3 23.3 29.9 40.3 63.6 132.7 537.7 1.8 21.5 21.2 14.8 17.7 22.4

28. 2 35.9 50.9 91.8 210.5 Total c 1-10 272.38 302.99 387. 9 1297.0 368. 5 906.3 t

Total c 1-11 a See Table A.3.

bprojected dates.

cTotals may not equal sum of the entries due to rounding of decimals.

306.3 516.8

Table A.8 Radionuclide Contributions to Decay Heat for Various Spent Fuel Batches.

The Percentage Contributions Depend on the Total Burnup of Each Batch, as well as Decay Time After End of Irradiation (Millstone ))

Isotope S ent Fuel Batch t/umber 5

6 7

RC N

F TA DECAY HEAT 0-Sr 90 Y 90 Zr 95 Nb 95 Rh 106 Cs 134 Cs 137 Ba 137m Ce 144 Pr 144 Eu 154 Pu 238 PLI 239 Pu 240 Pu 241 Am 241 Cm 242 7.48 35.73 a

0.43 9.02 30.29 1.22 2.14 2.16 1.84 0.19 7.57 0.01 6.79 32.44 6.14 29.33 2.15 4.85 1.54 1.90 0.22 7.96 0.04 3.03 7.33 1.14 1.79 0.23 7.34 0.05 0.98 1.76 8.77 8.43 29.44

28. 30 6.61 31.82 2.24 8.70 29.22

~ ~

2.63 4.66 1.36 1.78 0.24 6.70 0.02

~

6.32 30.21 WW&

3.74 8.45 28.29 3.15 5.38 1.16 1.68 0.24 5.84 0.03 6.18 29.52 0.81 5.15 8.28 27.80 3.32 5.31 1.10 1.61 0.24

. 5.12 0.03 5.85 27.92

~0 OW 1.89 7.65 7.87 26.43 0.06 0.64 3.52 5.33 0.99 1.49 0.24 4.22 0.03 5.23 24.97 5.75 1).63 6.95 23.34 0.26 2.93 3.30 4.49 0.88 1.27 0.22 2.92 0.05 3.78 18.06

~V sw 13.49

)6.26 5.16 17.34 0.79 8.80 3.03 3.72 0.57 0.90 0,17 1;61 0.21 2.34 11.17 0.01 0.02 22.53

)6.66 3@22 10.82 1 ~ 73 19.20 2.15 2.37 0.33 0.53 0.11

0. 70 1.21 1.39 4.96 1.13 2.33 27.)0 12.53 1.45 4.88 2.66 29.42 1.12 1.13 0.14 0.23 0.05 0.18

5. 52 Totalsb 98 08 97 08 94.87 95.98 94.59 94.47 94.13.

94.19 93.89 95.10 96.22 aDashes indicate less than 0.0)$.

bTotal percentage of.isotopes listed.

The balance of the decay heat is distributed among many other less important contributors.

Table A.9 Reactor and Fuel Cycle Parameters for Ginna (Sources:

Refs.

1-4)

Assemblies in core:

121 Licensed thermal power:

1520 HWth (gross)a Thermal power corresponding to maximum dependable capacity:

1499 HWth (gross)

Nominal initial metric tonnes of heave metal (IMTHH) per assembly:

0.375 HT Average refueling cycle interval (since initial commercial operation):

12.6 months Average number of assemblies per discharge:

1975-1980:

37 1981-1987:

24 Average IHTHM per discharge:

1975-1908:

15.3 HT 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 year s.

Authorized storage pool capacity:

1016 a0n March 1, 1972 the Atomic Energy Commission authorized an increase in gross thermal power from 1300 to 1520 HW.

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, Gener ator on Line:

107,134;3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months Gross Thermal Energy:

149.26 x 10 MWh Capacity Factor (MDC net):

70.3%

Table A.11 Summary of Spent Fuel Batches in Ginna Storage Basin (With Projections to 1987)

Spent Fuel Batch No.

1 2

3 5

6 7

8 9

10 11 12 13 14 15 16d Date of End of Irradiation 02/27/71 04/13/72 12/31/73 03/08/75 01/28/76 04/14/77 03/23/78 02/09/79 03/28/80

~

04/17/81 01/25/82 03/25/83 03/01/84 02/28/85 03/30/86 04/01/87 Number of Assemblies (37)

(47)c 8

29 37 41 41 40 36 28 24 20 23 25 24 24 Weight Avg.

H.H.

'urnup

{HT)

(HWD/HT) 14.778 6933 18.772 16695 3.195 30039 11.583 38043 14.778 36958 16.375 36022 16.375 27921 15.976 25451 14.378 26088 11 '83 27884 9.586 31054 7.988 33772 9.186 37532 9.985 40533 9.586 42360 9.586 45673 Decaya Dayl to 5/1/87

{days)

WW 5832 4869 4437 4111 3669 3326 3003 2590 2205 1891 1467 1156 792 397 30 Cumulative Assemblies in Pool 0

28 36 65 102 143 184 224 260 288 312 332 355 380 404..

428e Cumulativeb Gross Weight of Spent.

Fuel in Pool (HT) 0 18.4

23. 7 42.8 67.1 94.1 121.1 147.4 171.1 189.5 205. 3 218.4 223.6 250.0 265.8 281.6 aDecay days from end of irradiation to 5/1/87.

bGross weight of fuel stored in pool includes heavy metal plus cladding and hardware but not the fuel racks.

Each assembly contains approximately 0.4614 tonnes of heavy metal, 0.0620 tonnes of oxygen, 0.1345 tonnes of hardware, totaling 0.6579 tonnes gross.

cAt the time of the ORIGEN2 calculations some 56 assemblies could not be accounted for using available data.

dProjected data.

ehuthorized capacity is 1016 assemhlies.

A-22 Table A'.12 Comparison of Radioactive Core and Spent Fuel Basin Inventories in (Ginna)

Reactor Isotone Reactor Core Spent Fuel Stornoe Bnslna (Red lonctlvity, Curl es) 441488 H3 C14 Co 58 Co 60 Kr 85 Rb 86 Sr 89 Sr 90 Y 90 Y 91 Zr 95 Hb 95 Xo 99 Tc 99m RU 103 Ru 106 Rh 106 Sb 125 Sb 127 Te 125m Te 127 Te 127m Te 129 Te 129m Te 132 I 129 I 131 I 132 Xe 133 Cs 134 Cs 136 Cs 137 Bn 137m Bn 140 Ln 140 C

141 Ce 144 Pr 143 Pr 144 Hd 147 Sm 151 Eu 154 Eu 156 Np 239 Pu 23S Pu 239 Pu 240 Pu 241 Am 241 Os 242

.Os 244 3Q2EW 6+42K+I 3.57K+5 340K+5 3,73E+5 6.53K+5 3o55E+7 2.95EK 3 15EW 4 57E+7 6o41 E+7 6o34E+7 6 83E+7 5,89E+7 5+85K+7 1+95K+7 2 15E+7 6 04E+5 4.1 2K+6 1,27E+5 4.05EW 5,1 9E+5 I DI E+7 I oSOE+6 5.33K+7 1Q7&0 3+76K+7 5,42E+7 7+64K t7 5.82E %

1.87EW 441 E+6 4 IOOE46 6+55K+7 6.74K+7 648K+7 4 44K+7 5.71 E+7 447K+7 2 i48E+7 1 +42&4 4o09E+5 742K+6 7.81 EW 1 iOI E+5 145EM 2+02K+i 4+85E46 4.99E+3 1.91 &6 I 45K+5 949EM 2 64E+2 5.93&4 5,97E+5 9,84E+5 742K+3 3.53EK 1 02ee7 1.02E+7 5.11K&

8 64&6 1.1 2K+7 7a03E+3 6.77E+3 7,86K+6 1,09E+7 1 +09K+7 7,1 I E+5 4+33K+3 1.70E+5 1,1 9E+5 I oI 7K+5 1

38E+5 2 12E+5 1.83K+4 5.32K+0 6 OOE+5 1,89E44 3+52K+5 6.35EW I 46K+5 1,48K+7 1,40E+7 2.47K+6 2.85EK 6.34K&

1 +38K+7 2 ~ 54E46 1 +38K+7 7,42E+5 5,1 4E44 I +09&6 7'8K+5 3.02&4 4o46E+5 545EM 8 060K&

52K+7 2.1 OK+5 943K+5 3+59K+5 940EM 2 +64K+2 3.26K+4 5+84K+5 9.74E+5 7.48K+2 1 e53EK 1

01 E+7 1 aol E+7 2+48K+6 4 46E+6 7+51 E+6

'I o48E-3 I o42E-3 2,88E@6 9e7I EK 9.71 E+6 6,82E+5 7o35E-2 1,65E+5 7.79EM 7o95E44 3 o93E+4 6a03Et4 443K 2 5.32&0 3.1 2E+3 4o36E-2 1,11 K+2 6 +OOKED 4 +99K+3 1.47E+7 1 +39K+7 9,07EW 1,04E+5 '

72E+6 1.1 9E+7 1.1 2E+5

\\ ol 9K+7 1,62EW 5.1 3EW 1.07K+6 4.68EW 346K+3 4,46E+5 545EW So60E+4 1 o5I E+7 2ol 4E+5 740K+5 3.56E+5 8,82K+4 2,64 E+2 241 E+3 549K+5 947K+5 2 e74E-2 3.50EW 9.95K+5 9,95K+6 9o54EW 247K+5 4.93E+5 neg ~ b neg o 2,09K+4 5.78EK 5+78K+6 5e65E+5 neg.b 1 o37E+5 I,36K+4 1.38EW I +35K+2 2,07E+2 nag,b 5.32EK nag.b neq,b neg,b 4.66E+6 2 o40E-3 1.44E+7 1,37K+7 3,05E-2 3 ~ 51 E-2 4.91 E+3 6.,09K+6 8 o86E-2 6,09K+6 neg.b 5.1 OEW 1,01 EH I o66E 1

346K+3 4,46E+5 525K~

8,61 E+4 1,46K+7 2,32E+5 2.25E+5 3,46K+5 aspent fuel pool Inventory Includes discharges from 15 refuellngs cover-ing the period from Aprl I 1 983 through the proJected refueling of April 1987

'neg, a less than 10

Curles,

A-23 Table A.13 Comparison of Radioactive Inventories of Most Recently Discharged Fuel'atch (Batch

16) with Longer Aged Dis-charged'Batches (Batches 2-15)

(Ginna)

I SOtOO0 Spent Fuel Batch 160 Spent Fuel Batch 2 15 I

~T7Fo (Rad 1cacti v 1ty, Cur les)

H 3 C 14 Co 58 Co 60 Kr 85 Rb 86 Sr 89 Sr 90 Y 90 Y 91 Zr 95 Hb 95 Ho 99 Tc 99m Ru 103 Ru 106 Rh 106 Sb 125 Sb 127 To 125m Te 127 To 127m To 129 Te 129m Te 132 1 129 I 131 I 132 Xo 133 Cs 134 Cs 136 Cs 137 Ba 137m Ba 140 la 140 C

141 Ce 144 Pr 143 Pr 144 Hd 147 Sm 151 Eu 154 Eu 156 Mp 239 Pu 238 Pu 239 Pu 240 Pu 241 ha 241 Cm 242 Cm 244 9o89E+3 2o20E tl 5.7794 9.92' o07E 45 742 E+3 3.50&6 So56K+5 So57E+5 5o04E t5 Be47EK 1 e09E+7 7o03E+3 6,77E+3 7o86E46 5,82&6 5,82 Kt6 lo84E+5 4.33K+3 4,1 3EH 1

OBE+5 1

06E+5 1.38K+5 2.1 2E+5 1,83&4 ie22E-1 6,00E+5 1.89&4 3o52E+5 2o26E t6 1 o26E+5 1'o34EW 1 o27&6 2.47K%

2o85&6 6.34K%

Bo25E&

2o54E t6 8o25E46 7o42E+5 3,47E+3 1.67E+5 7e58&5 2o74E t4 4e87&4 3e05E+3 6.01 K+3 1 ~ 58E46 2.05E+3 7,57E+5 Se06&4 9oBO K+3 2e20ER 3ol BE44

~

9 70&4 1,05E+5 7o48E+2 1.52K&

8.53E+5 8,53E+5 2 o45&6 4.37&6 7,33ft6 1 o48E 3 1,42E 3 2 o68EK 5.1 9&6 5.1 9E&

1 o76E+5 7e35E-2 4,1 2ft4 7,05&4 7.1 9EW 3o93EW 6.03EW 4o23E-2 4o23E-1 3.1 2E+3 4,36E-2 1

11E+2'.13&6 4.99K+3 1.34&6 146K%

9o07ft4 1,04E+5.

1.73E+6 7,11 EK l,l2E+5 7.11K%

1.62K~,

3.47E+3 1 e65E @5 4.68&4 4.59M 4,95ft4 3.05Ee 6.01 E+3 1.57&6 2.47&5 5.85K+5 8.OOKED 9,66E+3 240&(

1 o29EW 9o39EW 1 o04E+5 2.45&i 4o29E+5 8o48E+5 So48E+5 8o23E+5 1.62K&

3ol 9&6 nag,c nag o 5o28E+5 4,37E't6 4 o37E+5 1 o65E+5 hag o 3o97EW 3.93EW 4

01 E+4 5oBBE+3 9o04E+3

nag, 4e23E-1 l,l2K%0 neg ~

nag, 1 o96E46 3.84K<

I o33E&

1 o26E46 6ol 9K%2 7el 3K+2 2 o43E+5 5.68&6 1 o02E+3 5.68K&

5.08EW 3o46E+3 1.61 E+5 7.02K%

4,59&2 5.04&4 3o05E+3 6ool E+3 1 55EK 3ol OE+3 3.96K%

7e93E t4 9.39E+3 240&(

2.1 5E+3 S.79EW 1

OOE+5 2e73E-2 3.48K&

BOSE+5 So38E+5 9.41 EA 243K+5 4.83K+5 neg o flag o 2.09K+5 3.09EK 3,09K+6 46E+5 heg o 3.55&4 1 D3E+4 1 o25Ea 1,35E+2 2,07K+2 neg.

4o23E-1 nag neg.

hag o 1.66&6 2 40E-3 1.31 &6 1 44&6 3o05E-2 3o51 E 2 4.91 E+3 3.64K&

So86E-2 3.64K&

neg.

3.45E+3 1.55K+5 lo66E 1

4.59K+2 5.1 3EM 3.05E+3 6.02E+3 1.51 E&

443K+3 1

82E+5 7o78E t4 Bo29EM 2 oi2E+2 1,60K+3 4o98E+5 So78Et5 nag.,'e39E44 9o32EW 9.32&6 6o86Kt4 1 o64E+5 3,68E+5 nag ~

hag ~

1 ol BE+4 5,06K+5 5.06&6 5o28E+5 nag ~

1,78E+5 1.09EH 1 ~1 2K%4 6.98EW 1 o07E+2 hag ~

4o89EK

nag, neg.

hag ~

4 09K&

hag e 1 o34E+7 1 o27E+7

nog, hag o 2.53E+3 5,58&6 nag ~

5o58E t6 neg ~

4,79EM 9.1 9E+5 neg ~

2.80E+3 3e97E+5 4e95EQ S.OOKED 1 37K+7 2oOBEt5 1.75K+5 2o78E+5 So22&4 2o42E+2 8o78E+2 4e87E+5 Bo68E+5 hag o 1.04K&

948K%

9o28E+5 3 ~33K%

8,48K+0 1 o89E+5 nag o hag o 4 o02E+3 io51 P6

.4 ~ 51 Et6 5o06E+5 nag.

1 o24E+5 7o42E+3 7,58E+3 1 o98EW 3.05EW hag o 4e89E tO nag heg o hag o 3e87E t6 hag ~

1 o34E+7 1 46K+7 nag.

flog ~

6,89E+2 -<<

io81 Et6 neg.

4,81 Et6 nag.

4,79EH 9.06E+5 heg o 2oSOE+3 3.97K+5 4.95EM S.OOKED 1 o35E+7 2.12E+5 1 46K+5 2.76E+5 8,1 0&4 2oi2E+2 3.57K+2 4,71 E+5 8,54E+5 hag o 2,93E+3 943&6 9o23E+S 1 12EH 3.1 3K~

6.96EH

nag, nag, 7,93E+2 3.80EH 3.8OEW 4.75E+5 nag o 1,1 6E+5 4 el 4E+3 482K+3 2.97EK 4.57EK nag ~

4o89E40

nag, hag ~

nag, 3.55K+5

nag, 1,33E+7 1 46K+7

nag, hag ~

9e69E4 3.84K&

nag, 3e84E+5 flag ~

4.78&4 SoBBE+5 hag ~

2oSOE+3 3 e96E+5 4e95E44 8oOOK+l 1 44K+7 2.1 7E<

941 EW 2,74E+5 7oBBE+4 2,42K+2 5.94K+I 4,41 E+5 8,27E+5 flag o 2o38E+2 9.1 2K+5 9.1 2K&

1 USE+3 5.60K+4 9 ~ 57K+3

nag, nag.

3,1 4K+i 2.69K+5 2 o69E+5 4.1 9E+5 flag o 1.02EW 1 o29EM 1.32K+3 6,82E 2 1 o05E 1

flag o 4.89&0 nag o flag e hag o 3.01 Kt6 nag ~

1,31 E+7 1 44K+7

nag, hag ~

1.96EK 2.46EH hag o 2,46K+5 naQ ~

ie76ft4 8,53E+5 hog e 2oBOE+3 3.95K+5 4o95ft4 B,00K+4 1 o31 E+7 2o28K+5 4.30EW 2e68E+5 aFuel batch 16 ls pro)ected discharge during hprll 1987, bFuol batches 2-1 5 we~0 discharged botlleen hprll 1972 and Apr ll 1986,

cnog,

~ less than 10 Curles.

A-24 Table A.14 Decay Heat Released from Spent Fuel Inventory for Various Discharged Fuel Batches (Ginna)

Date, End of Batch Sizea Irradiationa (Netric Tonnes)

Deca Heat Released b

Batch

~Nl, ri 1,

9 (Ki owatts, erma 2

04/13/72 3 '2/31/73 4

03/08/75 5

01/28/76 6

04/14/77 7

03/23/78 8

02/09/79 9

03/28/80 10 04/17/81 11 01/25/82 12 03/25/83 13 03/01/84

'14 02/28/85 15 03/30/86 16 04/Ol/87b 18.772 3.195 11.583 14.778 16.375 16.375 15.976 14.378 11.183 9.586 7.988 9.186 9.985 9.586 9.586 8.5 2.9 14.3 18.1 20.5 15.8 14.7

14. 7 13.7 15.0 17.2 28.6 50.9 96.1 437.2 8.5 2.9 14.2 18.0 20.4 15.7 14.5 14.5 1'3.4 14.6 16.5 27.1 47.2 85.8 260.4 8.4 2.8 13.9 17.6 19.8 15.1

14. 0 13.7 12.4

13. 2 14.2 22.0

35. 3

56. 5 107.7 Total c 2-15 Totalc.2-16 331.0 768.3 313.3 573.7 259.0 366.8 aSee Table A.11.

bProjected dates.

cTotals may not equal sum of entries due to rounding of decimals.

Table A.15 Radionuclide Contributions to Decay Heat for Yarious Spent Fuel Batches.

The Percentage Contributions Depend on the Total Surnup of Each Batch, as ue))

as Decay Time After End of irradiation (Ginna) isotope 5 ent Fuel Batch Number Sr 90 T 90 Ir 95 Hb 95 Rh 106 Cs l34 Cs l37 Ba )37m Ce )44 Pr )44 Eu 154 Pu 238 Pu 239 PU 240 Pu 241 ha 241 Cm 242 Cm244 7.32 34.95a.

0,33 8.89 29.85 1.33 2.90 1.84

) ~ BO 0 17 7.88 0.01 0.25 6.26 29.89 0.02 1.12 8.42 28.26 2 73 6.98 1.07 1.69 0.21 7,68 0.03 2.74':

5.56 26.57 0,05 1.77 7,93 26.64 3.37 9.46 0.81 1.49 0.20 6.60 0.03 8.45 5.61 26.82 0.08 2.26 7.94 26.67 0.01 3.48 9 12 0.82 1.49

'.20 6.20 O.D3 5.94 5.61 26 79 0.19 3.29 7,89 26 48 0.03 3.66 8.67 0.80 1.45 0.21 5.56 0.03 5.63 6.23 29.77 D.41 4.01 8.23 27.65 0.01 0.09 3.21 5.77 1.01 1.53 D.22 5.19 0.02 2+18 6.32 30.20 0.76 4.95 8.19 27.5D 0.02 0.23 3.05 1.76 1,05 1 ~ 47 0.22 4.67 0.02 1.52 6.07 29.01 HW 1.46

6. 72 7,89 26.51 0.05 0.52 3.20 4.71 0.97 1.38 0.22 4.07 0.02 1.64 5.61 26.79 2.86 9 38 7,38 24 79 O.l 1 1.21 3.41 4.83 0.84 1.27 0.2) 3.31 0.02 2.03 5.01 23.95 4.65 12.21 6.75 22.68 0.19 2.13 3.64 5.18 0.69 1.) 3 0.20 2.69 0.03 2,87 4.23 20.19 8.19 15.67 5.81 19.51 0.39 4.3) 3.6D 5.06 0.54 0.94 0.18 1.97 0.07 3.47 3.13

)6.37

')2.24 18.46 1.83 16.22 0 63 6.99 3.4) 4.82 0.41 0 74 0.15 1.39..

0.21 4.25 2.51 12.D1 18.16 20.18 3.63 12;18 1.07 1).84 2.88 3.95 0.28 0.53 0.))

0.82 0.70 4,35 1.60 7.64 0.04 0.08 24.96 19.07 2.33 7.84 1.78 19.75 2.02 2.58 0.)7 0.32

, 0,08 0.37 2 20 3.32 0.90 4.31

.1.0) 2.08 27.54 l5.65 1.35 4.52 2.24 24.8D 1.29 1.58 0.09 0.)7 0.04 0.13 6.22 2.52 Totalsb 97.52 97,10 96,93 96.67

. 96.29 I

95.53 94 93 94 44 94.05 g4.00 94.13 94.55 95.20 96.)5 96.44 aDashes indicate less than 0.0)S.

blotal percentage of isotopes listed.

The balance of the decay heat is distributed among many other less important contributors.

I

COO x

0)

U)O (9

~+

C9 0

hJ l.5 I.O 0.5 4

R@4 R@5 R4'6 R47 R48 R49 l976 l978 I980 l984

~

APPENDIX B BROOKHAVEN NATlONAL LABORATORY MEMORANDUM DATE:

TO:

August 27, 1986 M.T. Pratt K.R. Perkin a

H. Con ell Impact of Revised Reaction Rate Equation on the Likelihood of Zirconium Fires in a Drained Spent Fuel Pool (Task 5) e The SNL investigation

~ of the-potential for cladding oxidation during loss of fuel 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.

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.

The oxidation rate equation is also a key factor which affects the possi-ble propagation of Zircaloy fires to low power (i.e., older) spent fuel bun-dles.

The uncertainty in 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 0

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,

~ is representa-tive of the existing data.

For the purposes of the sensitivity study, we have adopted the two parameter oxidation curve suggested by Meeks.

weeks'wo parameter curve is given by:

w /t ~ 3.09 x 10 exp(-56600/RT)

(1) where:

w is the oxygen consumption (mg/cm~)

t is time (sec)

T is the clad temperature (K)

R is the gas constant (1.987 cal/K)

Nemo to T, Pratt from K. Perkins and H. Connell.

August 22, 1986 Page 2

Weeks'quation is equivalent to that suggested by Benjamin ~ except that it provides a

smooth 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

(+50'C) in the initiation tempera-ture occurs for this broad range of uncertainty in the oxidation rate.

This translates into an uncertainty of Q5% in the critical decay power.

Me be-lieve that this insensitivity to the oxidation rate equation basically con-firms the SNL analysis

~ for zirconium fire initiation in a dry spent fuel pool.

As Benjamin et al.~ pointed out, the most sensitive parameters for clad fire initiation are the decay heat level and the fuel rack geometry (related 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 1 to 5

months.

Whereas PWRs with higher power density and tighter. storage racks 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 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 Me conclude that the SNL code (SFUEL) and the clad oxidation rate equa-tion used therein accurately represents the potential for self-sustaining oxi-dation in a drained fuel pool.

The.largest uncertainty appears to be due to uncertainties in natural convection flo~s 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'X to 40'X for BWRs and 16% to 100% for pWRs, depend-ing on the storage rack configuration.

In terms of power level, our sensitivity studies indicate that the criti-cal power level (above which self-sustaining oxidation will occur) varies from about 50 kM/NTU (for cylindrical racks with large openings) to 6

kW/NTU for the new high density PWR fuel storage racks.

Memo to T. Pratt from K. Perkins and H. gonne))

August 22, 1986 Page 3

Recommendations Me recommend that spent fuel not be stored in high density racks until it has been stor ed for 2 or more years in the old style cylindrical racks with adequate coolant openings f3 or more inch diameter holes).

Me also recomnend that. a test program be initiated to confirm the capa-bility of natural air convection cooling capabil:ity 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 thermocouples placed near the to of the fuel bundle).

e op References 1.

Benjamin, A. S.,

McCloskey, D. J.,

Powers, D. A., Dupree, S. A., Spent Fuel Heatup Following Loss of Mater During Storage,"

NUREG/CR-0649, March 1979

Pisano, N. A., Best, F., Benjamin, A. S., Stalker, K. T., "The Potential for Propagation of a Self-Sustaining Zirconium Oxidation Following Loss of Mater in a Spent Fuel Storage Pool," Draft Report, January 1984.

'RP/HC/csc Attachments

0

e 4

Attachment, 1

BROOKHAVEN NATIONAL LABORATORY MEMORANDUM DATE:

FRQM:

March 27, 1985

K. Perkins Li Ji Teutonico

SUBJECT:

FIN h-3786 - Study of Beyond Design Basis hccfdents fn Spent Fuel Pools Twa Sandia reports i deal with the question of rapfd zirconium oxida-I 2 tion fn a spent fuel pool following loss of wateri Both the computer modeling and thc experimental simulation, as described in these reports, suggested that in certain fuel racking configurations (a) a self-sustaining zirconium-afr oxidation reaction can bc initiated, and (b) this self-sustaining reaction can propagate from one regian of a pool to another'here aro large uncertainties associated with the phenomenology of zfrcaloy oxidatfon and its propagation fn spent fuel assemblies This preliminary report an Tasks 3,

4, 5 of the subject FIN (Uncertainities in Oxidation Propagatian, SFUELIM Computer Code Validation, Impact of Revised Reaction Rate Equation, xespectively) addresses some of these uncertainftfes and their effects on the initiation and propagation of a self-sustafning zfrcaloy-air oxidation reaction.

1)

The propagatfon rates af rapfd zircoloy clad oxidation in air from the hottest sectfon of the pool (after a lass of water incident) to adjacent sectfons were estimated (fn Ref. 2) under the conditions that the spent fuel in the hottest section of the pool was generating 30 kw/MTU in a room main-tained at constant temperature.

hs painted out by Han, this cstfmate should be rc-calculated under inadequate room ventilation concLLtfons g

ta simulate properly the conditfans at many licensed facflftfes.

Sfmilax'ly, additional calculatfons should be performed in which the hot spent fuel decay power gs varied from 20 to 90 kwl>ITU for both the adequate and inadequate room ventilat-ionn conditions These studios would determine how sensitive the oxidation t

propagation fs to the decay power of the spent fuel stored ad)acent to hot fuel, assuming the input axidation rate data are known with sufficient accu-racv o

P C

Memorandum To:

K, Perkins Page 2

March 27, 1985 2)

The above assumees the zircaloy-air reaction rate equation used in the Sandia work is sufficien iciently accurate.

There are a

number of uncertainties associated with this. equation.

Qe discuss each of these uncertainties in turne h,

Experimental Data:

A Literature search 4

14~ has revealed that there is a great deaL of data for zirconium oxidation; most of it, however is can-t cerned with oxidation in steam or oxygen.

The data for zirconium (zircaloy)-

~ air oxidation presented in Refs. I and 2

appear to be the best available.

- These are shown in Figure 1.

.The authors (of the SNL reports) fit the data with three separate Arrhenius plots over the temperature range 500"1500'C; one break occurs at the e-B transformation temperature for zirconium, the other at the Cemperature at which the oxide undergoes a monoclinic-tetragonal trans-formation (N BE two of 'the sets of data are for zirconium, the other for zircaloy-4) ~

These assumptions are reasonabLe, It should be noted, however t Chat there is, no a priori reason to expect that the data wouLd be fit by an Arrhenius expression, particularly above the a-B transformation temperature where a number of different processes are occurring simultaneously (discussed further below); therefore the use of the Arrhenius expression should be viewed 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 Che oxidation rate of zircaloy in air over the temper t e

emperature range of interest, for both isothermal and non-isothermal conditions Bi Kinetics:

The question was raised as to whether the assumption of parabolic 'kinetics was valid.

Data were presented (from R f 86 rom e s.

and 126) which show examples of linear as well as cubic kinetics.

However, they all apply at temperatures below the c-B transformation temperature Since almost all rapid oxidation occurs above the e-B transformation temperature, where the oxidation rate is controlled by one or more diffusion processes,

@he assump-tion.of parabolic kinetics appears to be reasonable.

~ l ll t

Memorandum To:

K Perkins Page 3

March 27, 1985 C.

Zirconium vs. Zircalo :y.

It is assumed in the Sandia vork that the oxidation rates of zirconium and zircaloy are essential1y the same Recent vork by Pavel and Campbell has shovn th t thi I a

s s not the case.

Oxidation in steam af both pure zirconium and zircaloy-4 vas studied in the temperature range of rapid oxidation (1000 C-1500 C)

I f

t vas found that at all tempera-tures the oxidation rate of zircalay-4 vas higher than that of zirconium the ratio of the two rates is approximately 3 at 1000 C:and decreases vith increasing temperature to a

value of appraximateiy 1..5 at 1500'C (cf Pigure 2) ~

The higher oxidation rate of Zircaloy-.4 is attributed to increased oxygen diffusivity in the axide phase; a lover activation energy vas observed ct vat on energy vas observed, implying that same mechanistic differences exist.

Analogous results are ex-pected ta apply for oxidation in air.

~ D~

Oxidation Model:

The axidation in steam of both zirconium and zir-e caloy"4 (in the temperature range 1000-1500'C) is a multi-phase layer pro-cess.

Not only Is an'oxide layer formed, but also (beneath it) a layer of oxygen-stabilized c-phase (zirconium or zircaloy).

The multi-phase model is only significant above the a-8 transformation temper t

(

emperature (approximately 900 C), but this is exactly vhere rapid oxidation occurs.

The parabolic rate constants for oxide layer grovth a-layer grovth and.oxy t

oxygen'onsumption vere determined in Ref 136 from experimental data and computer d li mo e

ng.

The rate, of oxygen consumption is significantly higher at all temperat emperatures than the rate of oxide'ormation for both zirconium and zircaloy-4 F

y-.

or zirconium the ratia of oxygen consumption rate to oxidation rate is appraximat l 4

appra mately 4 at 1000'C and increases vith increasing temperature to a v l a ue o

approximately 5.4 at 1500'C; for zircaloy-4 the corresponding values are approxi l

3 ~

e approx mately 3.0 and 4.5 at 1000'C and 1500'C, respectively (cf Figure

2) ~

Although these results vere obtained for oxidation in steam analog na ogous results are again expected for oxidation in air.

S.

Effect of Nitrogen:

Before'iscussing the reacts ac on of zirconium vith t

air, let us consider the reaction vith nitrogen alone The rate of

~

Memorandum To:

K. Perkins Page 4

Harch 27, 1985 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 to..oxidation in view oi the high solubility of nitrogen in zirconium, and involves a large amount of dissolu-tion along with film formationi In the case of nitriding in the a"region, a

two phase diffusion process describes the behavior whereas 8-phase nitriding involves three phases (nitrogen, like oxygen, stabilizes the a-phase, Leading to a wide range of a between the nitride and the S~trix) ~

The reaction product is zirconium nitride (ZrN);

the reaction is exothermic, releas ing approximately 82 kcal/mole.

(The energy reLeased in forming the oxide is approximately 262 kcal/mole

)

The thickness of the zirconium nitride layer has been found to be much smaller than that of the dissolution zone (in the 149 temperature range 750'C-1000'C) which indicates that the rate constant for film formation is considerably smaller than the rate constant for nitrogen dissolution In fact, at 1000'C, 84K of the total nitrogen uptake was cue to dissolution in the metal The role of nitrogen in the high temperature reaction of zirconium with air has been investigated

~

The reaction process is multiphase in nature.

1S1 Adjacent to the 8-phase of the zirconium is a layer of n-phase (stabilized bY both oxygen and nitrogen) and a surface layer of Zr02 ~

In general, a certain amount of nitride (ZrN) is formed.

For temperatures up to approximately 1050'C the nitride is found as a layer between the stabilized a-phase and the oxide layer; above 1050'C the nitride occurs as discrete particles dispersed in the oxide It is doubtful whether any appreciable amount of nitride is formed in )he problem currently being considered At the lower temperatures (during heat up) the reaction rate is very slow.

Once rapid oxidation i> initSated (approximately 900'C) the self-sustaining reaction proceeds very quickly, and

Memorandum To':

K. Perkins Page 5

March 27, 1985 there may not be sufficient time for ZrN to be formed Any nitride that does form hovever vill. contribute to the chemical energy release for the self-sustaining reaction.=

t h,

P The reaction rate of zirconium is higher vith air than vith oxygen alone.

The explanatfon advanced. fs that nitrogen dissoLves in K Q B

1 n

r 2 y zep acing oxygen fons fn 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 associated vith the zircaLoy-air reaction equation.

These are particularly important above 900'C vhere rapid oxidation occurs' The most significant appeaz to be (i) the difference in 'he oxidation rates of zirconium and zircaloy, and (ii) the multiphase nature of the oxidation process itself at these temperatures.

The results given above in Section C and D (i.e for zirconium vs'ircaloy-4, and oxygen consumption zate vs oxidation rate, respectively) apply to oxidation in steam only.

Analogous results aze expected for,oxidation. in air, i.e. it is pected that the oxidationrate in zircaloy vill be greater than that in zir-

conium, and the rate of oxygen consumption vill be greater than the rate of oxide formation in both materials The relative magnitude of th ff t ese effects cannot be deduced from the steam oxidation data.

Mhat are re a e requ re are nev experiments and computer modeLing (similar to those carried out b

P e

out y Pavel and CampbeLLI36 for oxidation in steam) for the high t'perature reaction of zi~

conium end zircaloy vith air.

In lieu of these ve'ugge t th dd

'uggest t at additional calculatfons be performed for tvo other zirconium-air rereact on correlations vhich vill serve as bounds for those presented in'igure 1

( )

h h

gure

~

(a)

The high temperature correlation for zirconium (above the phase change of Zr02) should be muLtiplied by a factor mg to account for the higher reaction rate in zirca-Lo.

(b) y.

( ) The correlations above the cL-S transformation temperature should be divided by a factor m2 to account for the difference in oxygen consumption rate and rate of oxide formation.

Values of mg and m2 as large as five should be considered.

Memorandum To:

K. Perkins Page 6

March 27, 1985 REFERENCES Ben)amin, A

S ~, McCloskey, D ~ J.,

Powers, D ~ A, Dupree, S ~ A., "Spent Fuel Heatup Following Loss of Water During Storage

" NUREG/CR-0649

'fa h

1979 '

i rc 2 ~

3 ~

Pisano, N ~ Ai, Best, F ~, Ben)amin, 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," Draft Report, January 1984.

Han, J T', Memo to M Silberberg, May 21, 1984.

4 ~

5 ~

Douglass, DE L,

"The Physical Metallurgy of Zirconium," Atom.

Energy Rev 1 4(1963) 71 '

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

6 ~

Miller, G

L, "Zirconium," Butterworths Science Publications, London (1954) ~

7 ~

Douglass, D

Li, "The Metallurgy of Zirconium," International Atomic Energy Agency, Uienna, 1971 8o 9

Pemsler, J

P ~, "Diffusion of Oxygen in Zirconium and its Relation to Oxidation and Corrosion," Ji Electrochem Soc.

105 (1958) 315.

Mallett, M. W., Albrecht, Wo Mo

Wilson, P. R.,'The Diffusion of Oxygen in Alpha and Beta Zircaloy-2 and Zircaloy-3 at High Temperatures, J.

Electrochem.

Soc.

10& (1959).

181 10

'ie 12

Debuigne, J.,

Lehr, P ~,

"Sur la Determination des Coefficients d

t ens e

Diffusion de L Oxygene Dans le Systeme Zirconium-Oxygene,"

C.

Scio Paris 256 (1963) 1113

'ainforth, C, Jacquesson, Ri, Laurent, P., "Diffusion de L'Oxygene Dans le Zirconium,'" proc.

Colloque sur la Diffusion a 1'Etat Solide," North Holland Publishing Co., Amsterdam (1959) 79.

Davis, M., Montgomery, Ko Re, Standring, J ~,

"The Diffusion Coefficient for Oxygen in Alpha-Zirconium," J ~ Inst. Metal 89 (1961) 172

'3

Cabrera, N, Mott, N. F.,

"Theory

. of Oxidation of Metals,"

Re.

pro Phys 12 (1948) 163',

ep.

roy.

14, Campbell, W~

E.,

Thomas, U.

B.,

"The Oxidation of Metals,"

Trans.

EIectrochem.

Soc.

91 (1947) 623 ~

Memorandum To:

K Perkins Page 7

REFERENCES (CONT'D)

March 27, 1985

15. Uhlig, H.

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

'16 Draley, J. E., Bradhurst, D. H., Van Drunen, C. J.,

Argonne National Lab.

Report, ANL-7000 (1964) 223

'7

~ Dawson, J

K, Baugh, Una C., White, J. Fi, "Observations on the Early Stages of Oxidation of Zirconium and Zircaloy-2," Electrochem.

Technol.

4 (1966) 137

'8.

Cox, B.,

"The Oxidation and Corrosion of Zirconium and its Allovs," V.

Mechanism of Oxide Film Growth and Breakdown on Zirconfum and Zircaloy-2,"

J~ Electrochem.

Soc.

108 (1961) 24 ~

19'ox, B.,

"Low Temperature

(<300'C) Oxidation of Zircaloy-2 in Water," J.

Nucl ~ Mater.

25 (1968) 310, 20 Culbransen, E.

A,Andrew, K. F.,

"Oxidation of Zirconium Allovs in Water Vapor Atmospheres Containing Trace Amounts of Oxygen at 375 and 575'C, Electrochem.

Techno' (1966) 99

'ritton, C

F ~, Wanklyn, J. N., "Inhibition by Boric Acid of the Oxidation of Zirconium in High-Pressure Steam," J ~ Nucl Mater.

5 (1962) 326

'2

~ Britton, C ~ F,, Arthurs, J V., Wanklyn, J, N., "Further Studies on the Inhibition by Boric Acid of the Oxidation of Zirconium in High-Pressure Steaai," Ji Nucl. Mater.

15 (1965) 263 ~

23. Demant, J., T., Wanklyn, J. N., "Effects of Contamination on Oxidation of Zirconium in Steam," Corrosion 22 (1966) 60.

24 ~ Cox, B~, Reed, JA,, "Oxfdatfon of a Zr-2, 5Nb Alloy fn Report AERE-R-4459 (1963).

25 ~ Johnson, A. B., Jr.,

"Effects of Nuclear Radiation on the Corrosf on Hydrfding and Oxide Properties of Six Zirconium Alloys," ASTM STP 458 (1969) 301.

26 'ouglass, D. L.,

Corrosion Mechanism of Zirconium and Its Alloys-III.

Solute Distributfon Between Corrosion Films and Zirconium Allpy Substrates,"

Corr. Sci.

5 (1965) 347.

27'ass, S., Crozier, J.

D., Shubert, F. L., "Effects of Silicon~ Nitrogen, and Oxygen on the Corrosion and Hydrogen Absorption of Zircaloy-2,"

Corrosion 20 (1964) 350t.

Memorandum To:

K>> Perkins Page 8

REFERENCES (CONT'D)

March 27, 1985

28. Gulbransen, E

A, Andrev, K, F ~, "Oxidation of Zirconium Between 400 and 800'C," J ~ Metals (1957) 394 '

~t 29'ox, B ~,

Rate-Controlling Processes During 'the Pre-transition Oxidation of Zirconium Alloys," J NucL Mater'1 (1969) 48

'0>>

Flint 0

1

~

~ g Varley, J H>> 0, On the Mechanism of Oxide-Film Format ion on Zirconium," J ~ Phys.

Chem. Solids 6 (1958) 213.

31.

Cox, B., "Effects of Irradiation on the Oxidation of Zirconium AlLoys in High Temperature Aqueous Environments," J Nucl Mater.

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 of Fission Fragments and Neutron Irradiation on the Kinetics of Zirconium Oxidation," Reactivity of Solids (DE BOER, J H.,

Ed.)

Proc.

Fourth International Symposium on the Reactivity of Solids, Amsterdam (1960) 425 '

Asher, R.

C.,

Davies, D., Hall, A., Kirstein, BE A>>, Marriott, J ~

W.,

White, P.

J,,

"Effects of Radiation on the Oxidation and Hydrogen Absorption of Zirconium Alloys in Steam,"

Electrochem.

Technol.

4 (1966) 231

'5

'osa,.

C. J

, "Oxidation of Zirconium, A Critical Reviev of Literature,"

J>> Le'ss-Common Metals 16 (1968) 173

'6.

Kofstad, P ~,

"High 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, Wallvork, G

R

~ Smeltzer, W. W., Rosa, C

J.,

"The Parabolic Oxidation

~

Kinetics of Alpha-Zirconium at 850'C," Acta Metall.

12 (1964) 409 ~

39 ~ Bradhurst, D. H, Draley, J E.>

Van Drunen, C. J.,

"An ELectrochemicaL Model for the Oxidation of Zirconium," J.

Electrochem.

Soc.

112 (1965) 1171

40. Charles, R. G., Barnartt, S ~, Gulbransen, E.

A,

"Prolonged Oxidation of Zirconium at 350 and 450'C," Trans.

AINE 212 (1958) 101.

Memorandum To:

K. Perkins Page 9

REFERENCES (CONT'D)

March 27, 1985 41 42 ~

Porte, H

A., Schnizlein, J.

G ~, Vogel, R. C., Pischer, D. F., "Oxidation o

Zirconium and Zirconium Allys," J. Electrochem.

Soc.

107 (1960) 506.

Osthagen, K., Kofstad, P, "Oxidation of Zirconium and Zirconium"Oxygen Alloys at 800 C," J Electrochem.

Soc 109 (1962) 204

'3

'4

'5

'esterman, R~

E ~, "High-Temperature Oxidation of Zirconium and Zircaloy"2 in Oxygen and Water Vapor," J Electrochem.

Sac~

111 (1964) 140.

Beranger, G.,

Lacambe, P, "Contribution a

1'etude de la cinetique de

'oxydation du zirconium-a et de la diffusion de 1'oxygene dans le metal sous-)acent a 1'oxyde," J. Nucl. Mater.

16 (1965) 190.

H liner, E.,

"The High-Temperature Oxidation Kinetics of Zirconium in Dry Hilln r Oxygen," Electrochemi Technol 4 (1966) 132 ~

46 'ubicciotti, D.,

"The Oxidation of Zirconium at High Temperatures,"

J.

Am. Chem Soc 72 (1950) 4138 ~

Hussey, R, J Smeltzer, W. W.,

"The Mechanism of Oxidation of Zirconium in the Temperature Range of 400 to 850'C," J Electrochem.

Soc.

111 (1964) 1222 48

'9'ebui e

J gn,

J.,

Lehr, P.,

Corrosion seche du zirconium non allie,"

N Carrasion of Reactor Materials (Proc. Conf. Salzburg, 1962) 2 IAEA Vienna (1962) 105.

~

De Gelas, 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 29 (1969) 1 ~

50, Korobkov, I, I~, Ignatov, D. V., Yevstyukhin, A I~,

Yemel Yapov, V.

S ~

"Electronagraphic and Kinetic Investigations of the Process of 0 id of Zirconium and Certain Alloys of its Base,"

Metallurgiya x

ation Metallovedenie chistykh Metallov I (1959) 144

'1

'ackay, T.

L~,

"Oxidation of Single Crystal and Polycrystalline Zirconium," Trans.

AIME 227 (1963) 1184 ~

52+ Saur, G.,

Laue, H. J.,

Borchers, H.,

"Untersuchungen uber die Oxydation von Reactor-Grade-Zirkon,"

Metallurgic 15 (1961) 9 ~

0 53'adeyski, h.,

Poulton, DE J.,

S>eltzerr W

W s

"Parabolic.

Oxidation

.Kinetics af Alpha-Zirconium," Acta Metall.

17 (1969) 579 ~

Memorandum To:

K Pezkfns Page 10 REFERENCES (CONT'Dj March 27, 1985 54'ense K

A "On (1962) 377'

~

~ f the Oxidatfon of Zirconium " J, Electroch m

S 10 e

~

oc 55m Pbaalnikar Ce ho~, Baldwin, W

M.,

Scaling of Zirconium fn Air, Am. Soc Testing Mater Proc Sl (1951) 1038 ~

56'arihotti, DE J~

Green H.~ Mi, Baldwin, W

M., Jri, Report AECU-3013 57 'eranger, G>>

Lacombe, P.,

"Contribution a

1'etude d

1 i i d'o dati e

a c net que xy ation du zirconium,"

C~ r Aced. Sci., Paris 258 (1964) 6147.

.58 ~ Nomura

~

omura, S ~, Akutsu, C, Oxidation Kinetics and Oxide Film Breakaway of Zirconium and its Alloys at High Tempezatures

'El h

(1966) 93 ~

ectx'oc em.

Technol.

4 59 ~ Debuigne, J, Lehr, Pi, "Sur 1'oxydation du zirconium," C. r. Acado Scil, Paris 254 (1962) 3710 ~

60 ~ Kofstad, P.,

"Effect of Impurities on the Defects in Oxides and their Relationship to Oxidation of Metals," Corrosion 24 (1968) 379.

61 'assel W

of Utah Pxo r s

~

sel, W. M, Jr.,

The High-Temperature Oxidation of tf 1

" U eta s, niversizy NP-4246.

a xogress Report foz the Period June 1

to Septemb 2&

(1952) er 62 'akamura, J.,

Sasakf, Y.,

"Study an the Oxid ti

-R Zfr

~,

a on esfstance of rconfum," Technical Reports of'he Engineering Research Institute, K ota Univex'sity, Uol. IX, No 6, Report No 58 (1959).

63'fsch 700',

R. D., Van Drunen, C.,

The Oxidation of Zirconium Bf 00 C Oxygen for Times up to.200 Days," Proceedings of the USAEC Symposfum nary A oys fn on Zirconium Alloy Development, GEAP-4089, Vol. II (1962) ~

aux 64 ~ Perdereau, Mi, Bardolle, J"Etude cfnetique de 1'oxyd f d

o at on e zfrconfum ux temperatures elevees,"

Ci ri Acadia Sci., Paris 256 (1963) 4665 ~

and 65~ Ronnquist, h., Shiels, S ~, "Reactian of Zirconium with 0

Air, a

Review of Publfshed Data,"

Forsvarets Farskf nfngsans tal

~, "

xygen, Nitrogen, Institution M, Stackholm Report A 14 (196&).

rs n ngsanstalt 666 ~ Smeltzez',

Vo W, Haerfng',

R ~

Kirkaldy J.

S "0 id Shoz't Ci c

~

~ e x

ation of Metals by ort Circuit and Lattice Diffusion of Oxygen,- Acta Metall.

9 il9&lj

~

j 1 j 880.

Memorandum To:

K. Perkins Page 11 REFERENCES, (CONT'D)

March 27, 1985 67 ~

68 Levitan J

Dr e

~,

raley, J. E, Van Drunen, C

J.,

"LoM-Pressure Oxidation of Zirconium," J. Electrochem Soc 114 (1967) 1086.

Nier1fch J

avec l'o ene J j Paidassf, J~,

Sur la Cfnetique de Ia Reaction du Zi u

rcony.um xygene sous

-basses pressions aux

.temperatures

elevees, C. r.

head Sci

, Paris 267 (1968) 1429

'9 70 7lo

Paidassf, J.,

Nferlich, J, "Sur les phenomenes de germfnati d

issance de la zircone, ainsf que de striation du zirconium, au cours de la reaction de ce metal avec l'oxygene sous basses pressions, aux temperatures elevees,"

C. r Acad Sci., Paris 267

(%968) 1085 Beranger,. G Lacombe, P.,

"Phenomenes de germination et de croissance au cours de l'oxydation superficielle du zfrconium," Proc.

IXe Colloque de Metallurgic, Presses Universitaires de France (1966) 181.

Renucci, L., Langeron, J. P., "Oxydation du zirconium en phase B sous tres basses pressions, aspect superficiel presente par les alLiages zirconfu~xygene," J. Nucl. Mater.

23 (1967) 79 ~

72i Perdereau, Mi, Bardolle, J

, "Etude micrographique de l'oxydation menagee du,zirconium,"

C r~ Acad de Sci, Paris 25? (1963) 2477.

730 76m 75 76'emsler, J

P., "Studies on the Oxygen Gradients Xn Oxidizing Metals," J, Electrochem Soc 113 (1966) 1242 ~

Hussey, R J,

Smeltzer, W. W.,

"The Oxidation Kinetics of Zirconium the Temperature Range 400-600'C,"

J~ Electrochem. S~ ill (1968) 564,

Hagel, W

Ci,

Rosa, C

J,

"Oxygen Diffusion in Hypostoichiometric Zirconium Oxide in the Temperature Range of 875 to 1050'C T

AD (1968) 1293

'osa, Ci J.,

Hagel, W

C.,

"Oxygen-fon Diffusivity in Hypostoichfometric Zirconium Dioxide," J. Nucl ~ Mater.

27 (1968) 12.

77i Rosa, C~

J.,

"Evaluation of Oxidation Ji Less-Commons Metals 15 (1968) 35 ~

Rates for Alpha-Zirconium "

?8. Rosa, C

J.,

Smeltzer, W.

W.,

"The Parabolic Oxidation Kinetics of Beta-Zirconium," Acta Metall 13 (1965) 55.

79 Dougl'ass, D. L, "Corrosion Mechanism of Columbium, Zirconium, and Alloys:

I.

Diffusion of Oxygen in CoLumbium Pentoxide anc." Zfrconfum Dioxide," Corrosion. of Reactor Materials (Proc Conf. Salzburg, 1962) 2 XAEA, Vienna (1962) 223.

~

azurine 162 2,

Memorandum To:

K. Perkins Page 12 REFERENCES (CONT'D)

March 27, 1985 80 Dou l

. and it

Douglass, D. L, Wagner, G,

The Oxidation of Oxygen-Deficient. Zi s Relationship to the Oxidation of Zirconium," J ~ Electzochem Soc'.

en rcon a

113 (1966) 67le 81 Smith, T, "Diffusion Coefficients and Ani'on Vacancy Concentrations for the Zirconium"Zizconium Dioxide System,"

J Electrochem Soc.

112 (1965) 560m 82 Aronson f S ~ j "Oxidation and Equ 11ibrium in Nons toichiome tric Zirconium Dioxide Powder," J Electrochemi Soc 108 (1961) 312.

83'ox, B ~, Pemsler, J.

P ~, "Diffusion of Oxygen in Growing Zirconia Films,"

J. Nucl. Mater.

28 (1968) 73.

84 'adeyski, A., Smeltzer, W. W.,

"Oxygen Diffusion in Monoclinic Zirconia,"

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

'85 'eneshea, F

J, Douglass, DE L, "The Diffusion of Oxygen in Zirconia as a Function of Oxygen Pressure,"

Oxidation of Metals 3 (1970) 1.

86'allett, M.

Wi, Albrecht, W

M,

"High Temperature Oxidation of Two Zirconium-Tin Alloys," J. Electrochem Soc.

102 (1955) 407.

87'oulton, D. J,

Smeltzer, W.

W.,

"Oxygen Diffusion in Monoclinic Zirconia," J. Electrochoem.

Soc.

117 (1970) 378.

88 'ouglass, Do Lo, Van Landuyt, Ji, "The Oxidation of Zirconium:

an Elect'ron Microscopy Study of Zizconia Formed in the Thin"Film Region,"

Acta Metall.

13 (1965) 1069 ~

89 ~

90

'1 92 ~

Beranger,

. G.,

Duffaut, F, De Gelas, B ~,

Lacombe, P.,

"Observations en microscopic electronique de films minces de zircone obtenus par oxydation du zirconium," J. Nucl. Mater 19 (1966) 290.

Missiroli, G~ F., "Electron Microscopic Observation of Oxidized Zirconium Films," J. Nucl. Mater.

25 (1968) 93

'ouglassy D

L,

Van Landuyt, J,

"Transmission Electron Microscopy of Zirconium and Zizcaloy-2 corrosion Films Formed in Air and Water," proc.

IXe Colloque 'de Metallurgic, Presses Universitaires de France (1966) 155.

De

Gelas, B ~,

Beranger, G,

Lacombe, P ~,

"Etude par microscopic electzonique en transmission de films minces d'oxyde obtonus par oxydation d'alliages'e zirconium a teneur en element d'addition infericore a

4X en pofds," J. Nucl. Mater.

28 (1968) 185.

Memorandum To:

K. Perkins Page 13 REFERENCES (CONT'D)

March 27, 1985 93.

Bailey, J E,

"On the Oxidation of Thin Films of Zirconium," J. Nucl.

Mater 8 (1963) 259

'4

'ailey, J.

Ei, "The OxIdation of Thin Films of Zirconium," Electron Microscopy, 1964; Vol A, Non-Biology, (Titlbach, M, Ed.)

Publ.

H6use Czech. Sci, Prague (1964) 395.

95 ~

Denoux, M., Trfilat, J. J., "Sur L'obtention de couches mfnces orientees de zirconium et leur oxydatfon,"

C r. head. Sci., Paris 258 (1964) 4683.

96 ~

D

~

Denoux, M.,

Sur 1 oxydation de couches mfnces orientees de zirconium,"

f C. r. Acad. Sci., Paris 260 (1965) 5003.

97 ~

Denoux, M, Trillat, J~ Ji, "Etude de l'oxydation de couches mfnces orfen tees de zirconium; relation d 'epi taxie oxyde-metal,"

Proc.

IXe Colloque de Metallurgic, Presses Universitaires de France (1966) 173.

98

Vahldfek, Fi Q,

"Epitaxial Oxidation of Zirconium at Various Tear peratures," J.

Less-Common Metals, 12 (1967) 19.

e 99.

Culbransen, E ~ Ai, Andrev, K. F., "Oxidation of Zircaloy-2 and -3A at 300 to 850'C," Trans h&E 212 (1958) 281 100. Greenbank, J. C., Harper, S ~, "Solute Distribution fn Oxidized Zirconium Alloys," Electrochem Technol.

4 (1966) 142.

101 ~ Cox,. B., "Oxidation and Corrosion of Zirconium and its Alloys," Corrosfon 16 (1960) 380t 102'ox, B ~, "Causes of a Second. Transition Point Occurring During Oxfdation of Zirconium Alloys," Corrosion 18 (1962) 33t.

103 'emsler, J

P.,

"The Kinetics and Mechanism of Oxide Film Growth Zirconium," Electrochem

. Technoli 4 (1966) 128

'04

~ Greenbanky Jo Ci ~ Harper, S ~,

"The Mechanism of Breakaway Oxfdatfon Zirconium Alloys," Electrochem Technol.

4 (1966) 88 ~

105 'ouglass, Do Le, "Oxide Plasticity in the Oxidation Mechanism of Zirconium and its Alloys," Corrosion Scf.

5 (1965) 225.

106.

Ke: s, L.

H..

Beranger, G.,

De

Ge1as, B-,

Lacombe, P,,

micrographique du processus de desquamation

("breakavav")

au l'oxydatfon du zirconium," J.

Less-Common Metals 14 (1968)

0 Memorandum To:

K. Perkins Page 14 REFERENCES (CONT'D)

March 27, 1983 107'o Ci B

~

y, C. Burgess, B ~,

A Study of the Stresses Generated in Zirconia Films

~I During the Oxidation of Zirconium Alloys," Oxidation of Metals 2 (1970) 108 '

1 8 ~ Cox, B,

Processes Occurring. During the Breakdown of Oxide Films on

~<<

Zirconium Alloys," J Nucli Mater 29 (1969) 30.

109 ~ Gulbransen, E

Ai~ Andrew, K

F ~,

"Breakaway Oxidation of Zirconium-Tin Alloys," Corrosion 14 (1958) 32t.

110 ~

1

~ Gulbransen, E ~ A., Andrew, K F, Kinetics of the Reactions of Zirconium with 02, N2 and H2," J. Metals 1 (1949) 515.

Ill~ Guldner, W

G.

and Wooten, Li Ai, "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,"

Am. Chem Soc.,

72: 4138 (1950}

113. Hayes, E

T. and Roberson, Ae Ho, "Some Effects of Heating Zirconium in Air, Oxygen, and Nitrogen," Trans ~ Electrochem, Soc., 96:

142 (1949).

114 'all.ett, 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 High-.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'orte, H

A.,

Schnislein, J

G.,

Vogel, R. C, and

Fischer, D

F.,

"Oxidation of Zi.rconium and Zirconium Alloys,"

Argonne National Laboratory, ANL-6046, September 1959

'18

~ Biederman, R

R.,

Ballinger, R.

G.,

and

Dobson, Q.

G. >

"A Study of Zircaloy 4-Steam Oxidation Reaction Kinetics,"

EPRZ NP-225 1976 ~

eptember 119 'iederman, Ri R B, Sisson, R

D., Jr.> Jones.

J. K., and Dobson, W.

C "A Study of Zircaloy 4-Steam Oxidation Reaction Kinetics,"

EpRZ Np 734 April. 1978 ~

~

W

Memorandum To:

K. Perkins Page l5 REFERENCES (CONT'D)

March 27, 1985 120 ~

121

'harlot, Li A and

Johnson, A

B, Jr "Reaction of Zircaloy-2 with Water Vapor at 900 to 1200'C," on BNWL 120, July 15, 1965

Bostrom, W. A., "The High Temperature Oxidation of Zircaloy, in Water,"

WAPD-$04 ~

Westinghouse Electric Corporation 'arch 1954

'22

'emsler, J P, "Studies of the Oxygen Gradients in Oxidizing Metals, III Kinetics of the Oxidation of Zirconium at High Temperatures,"

J ~

Electrochem.

Soc., vol 112, pp ~ 447-484 '965

'23

'24

'25' 126 ~

127

'28

~

129

'30

~

Wallwork, G.

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 '965 White, J H,

reported in AEC Fuels and Materials Development

Program, Progress Report No 67, GAP-67, General'lectric Co,,

151, 1967

'eis trow, S, et al.,

"S tudy on High Temperature S team Oxidation of Zircaloy-4 Cladding

Tubes, Nuclear Safety Pro)ec t Second Semiannual

Report, 1975, KfK-2262, Karlsruhe,

233, 1976

'endall, L. F~, "Reaction Kinetics of Zirconium and Zircaloy-2 in Dry Air at Elevated Temperatures,"

Hanford Atomic Products Operation, Washington, Contract No W-31-109"Eng.-52, September 1955

'rquhart, A.

W and Vermilyea, D ~ A"Characterization of Zircaloy Oxidation Films,"

Zirconium in Nuclear Applications, ASTM STP

551, hmetican Society for Testing and Materials,

1974, pp. 463-478 ~

Sabol, Gi P.,

McDonald, S ~

G and hirey, G.

P ~, "Microstructure of the Oxide Films Formed on Zirconium-Based Alloys," Zirconium in Nuclear Applications, ASTM STP 551,. hmerican Society for Testing and Materials,

1974, pp 435-448'illner, E, "Corrosion of Zirconium-Base Alloys-An Overview," Zirconium in the Nuclear Industry, hSTM STP 633>

Ao Lo Lowe, Jr and G ~

W

Parry, Eds American Society for Testing and Materials, 1977, pp. 211-235

'rbanic, V. F.,

"Oxidation of Zirconium Alloys in Steam at 1ppp 1850'C,"

Zirconium in the Nuclear Industry, ASTM STP

633, Jr.

and C

W. Parry, Eds.,

hmerican Society for Testing and Materials,

1977, pp.

168-181

Memorandum To:

K. Pezkfns Page 16 March 27, 1985 REFERENCES (CONT'D) 131 ~

132m

Dobson, Q

C.,

Biederman, R

R.,

and Baliinger, R. C, "Zircaloy-4 Oxidation in Steam Under Transient Oxidizing Conditions,"

Zirconium in the Nuclear Industry, ASTM STP

633, h.

L.

Lowe, Jr.

and C.

W. Parry, Edsi, American Society foz Testing and Materials,

1977, pp 150-167

Sawatzky, A.,

Ledoux, G.

h,

and

Jones, S ~,

"Oxidation of Zirconium During a High-Temperature Transient," Zirconium in the Nuclear Industry, ASTM STP 633, A. Li Lowe, Jr.

and G ~ W. Parry, Eds.,

American Society for Testing and Matetials, 1977, pp. 134-149.

133

'34

~

Pawel, R.

E,

Perkins, R.

h,

McKee, R.

A.,

Cathcart, J.

V,

Yurek, C. J.,

and Druschel, R. E., "Diffusion of Oxygen in Beta-Zircaloy and the High Temperature Zircaloy-Steam Reaction,"

Zirconium in the Nuclear Industry, ASTM STP 633, A. L Lowe, Jr.

and CD W. Parry, Eds.,

American Society for Testing and Materials,

1977, pp. 119-133.

Hofmann, PE and Politis, C, "Chemical Interaction Between 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.

. 135

Ocken, H.,

Biederman, R

R.,

Harm,

~

CD R.,

and Westerman, 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 ~

137

'awel, R

E and Campbell, J

J,

"A Comparison of the High Temperature Oxidation Behavior of Zircaloy-4 and Pure Zirconium," Zirconium in the Nuclear Industry (Fifth Conference),

ASTM STP 754, D.

G. Fzanklin, Ed.,

American Society for Testing and Materials',

1982, pp. 370-389.

Hofmann, P ~,

Kezwin-Peck, D~,

and Nikolopoulos, P.,

"Physical and Chemical Phenomena Associated with the Dissolution of Solid UOg by Molten Zircaloy-4," Zizconium in the Nuclear Industry (Sixth Conference),

ASTM STP 824, D

C Franklin and Ri BE Adamson,'ds American Society. for Testing and Materials, 1984'pi 810 834

'38

'hung, H,

M. and

Thomas, Gi R, "High-Temperature Oxi.dation of Zizcaloy in Hydrogen-Steam Mixtures," Zfrconfum in the Nuclear Industry (Sfxth Conference),

ASTM STP

824, D

G~

Franklin and R

B.

Adamson, Eds American Society for Testing and Materials,

1984, pp. 793-809.

139 'eistikow, 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),

AST.'I STP

824, DE G ~ Franklin and R

Bo Adamson>

Ed' Amerfcan Society for Testing and Materials,

1984, pp. 763-779.

kg+

hg Memorandum To:

K. Perkins Page 17 March. 2T, 1985 140 ~

141 ~

REFERENCES (CONT'D)

Yurek, G J.,

Cathcart,.J>>

V., and Pawel, R>> E, "Microstructures of the Scales Formed on Zircaloy-4 in Steam at Elevated Temperatures,"

Oxidation of Metals 10, 255 (1976) ~

Pawel, R.

E>>,

Cathcart, J>>

U>>,

and McKee-,'. h.,

"The Kinetics of Oxidation of Zircaloy-4 in Steam at High Temperatures,"

J Electrochem.

Soc

126, 1105 (1979);

142 ~ Pawel.,

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. Mater. 82, 129 (1979).

143 ~

144 ~

145 ~

146 ~

14? ~

Pawel, R

E.,

"Oxygen Diffusion in the Oxide and. Alpha Zhases During Reaction of Zircaloy"4 with Steam from 10QQ'C to 1500'C,"

J.

Electrochem.

Soc.

126, ill'I (1979) ~

Pawel, R. E., Cathcart, J. V., and McKee, R. A., ",Anomalous, Oxide Growth During Transient-Temperature Oxidation of Zircaloy-4,"

Oxidation of Metals 14, 1 (1980) ~

I

Pawel, R>>

E and Campbell, J. J.,

"The Observation of,Effects of Finite Specimen Geometry on the Oxidation Kinetics of Zircaloy-4,"

J.

Electrochem.

Soc.

127, 2188 ()980).

1

Pawel, R

E.

and Campbell',

J ~ J.,

"The Oxidation of Pure Zirconium in Steam from 1000'o 1416'C," J. Electrochem. Soc'28, 1999 (1981).

Pawel, R

E and

Campbell, J ~ J ~,

"The Effect of Structural Changes in the Oxide on the Oxidation Kinetics of Zirconium in High Temperature Corrosion," NACE-6, 1983, p.

162 ~

148 'ravnieks, A>>,

"The Kinetics of the Zirconium-Nitrogen Reaction at High Temperatures,"

J. Am. Chem..Soc.

72 (1950) 3568 ~

149'osa, C. J>>, Smeltxez, M. W., "The Nitriding Kinetics of Zirconium in the Temperature Range ?50-1000'C," Electrochem Technol.

4 (1966) 149.

156

Mallett, M. M., Baroody, E ~ M., Nelson, H

R., and

Papp, C,

Surface Reaction of Nitrogen with Beta Zirconium and the Diffusion of Nitrogen in the Metal," Report BMI-709 (rev.), Dec.

12, 1951.

151 'vans, ED B.,

Tsangarakis, N ~,

Probst, H. B.,

and Garibotti, N.

"Critical Role oi Nitrogen During High Temperature Scaling of Zirconium,"

in High Temperature Gas"Metal Reactions in Mixed Environments; TifS-A1ME,

. 1972, p'>> 248>>'

~pa qo

+6

6. 20 x 10 exp (-29077/R11

-2 0

I 0

0

5. 76 x 10 exp (-52990IRT)

MONO-TETRAGONAL PHASE CHANGE OF Zr02 a -g PHASE CHANGE OF Zr -"02 SOLID S OLUTI ON S

1. 15 x 10 exp (-2734/RT) 3

-12 o

LEI ST I KON (1975) 0 NH l TE (1967)

HAYES AND ROBERSON (1945) 6 7

8 9

eve.'~~'4

~"'4 10 IT PK) 10 D

- 12 13 FIGURE 1 CQRRELAT IONS FOR Z IRCON I UN OXIDATION Al A I R (FROM REF U

4 Zy

TENCl l600 <500 <400

<500

<200

. i<00 lOOO IO

<Oec

'I IO C 4 AI,PHA \\.AYER,

,cm /s So 2

t

<0 7 O

2 I0~7 LI O

0 1r IO'XIOATIONOF Z< l

)

ANO Zc ~1 l~~ ~1 IN STE44l Sg c< Ox<OE I.AYER.,cmt/a C OXYGEN CONSVIIEO, t, lg/Cmt) t/a Sr

<0'OC 5

e IQC

<0 <o 5.Q 5 5 6 0 6.5

?.0

?.5 6.0

'I000/T

< ~

~

FIGURE 2 PARABOLIC RATE CONSTANTS FOR OXIDE LAYER GROWTH J

+-LAYER GROWTH~

AND OXYGEN CONSUMPTION FOR THE REACTION OF ZIRCONIUM (SOLID LINES)

AND ZIRCALOY-g (DASHED LINES) WITH STEAM<

THE RATE CONSTANTS FOR OXYGEN CONSUMED (WEIGHT GAIN)

WERE DETEPMINED FRP<~

MODELING ANALYSES (FROM REF,

$36),

Attachment 2

SROOKHAVEN NATlONAL LASORATORY MEMORANDUM DATE:

TO:

FROM:

January 27, 1986 Kenneth Perkins John Weeks

SUBJECT:

Parabolic Rate Constants for Oxidation of Zircaloy-4 in Dry hir On Harch 27,

1985, Lou Teutonico (1) provided you with an assessment of

. the knowledge of oxidation kinetics of sircaloy-4 in air and steam based on the. literature available at that time, as part of our overall assessment of the Sandia report',

NUREG/CR/0649, under our NRC contract FIN h-3786.

Among

~other things, Lou showed significant differneces in the oxidation kinetics of zirconium metal and sircaloy-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

\\

rates ar'e expected, except for the 1967 data by White on unal3oyed zirconium (3) which show significan'tly higher rates than would be expected by extrap-olation of results obtained at lower temperatures.

%hen 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.

hfter Teutonico left Brookhaven, I attempted to reevaluate the work he had done in the context of comparing the Pawel and Campbell data with the Sandia curve. It is immediately apparent that the Pawel and Campbell l

parabolic rate constants are considerably lower than the curve used by

4 diag Figure 1

shows this comparison i Subsequently, while at the Xnteznational Conference on Environmental Degradation of Nuclear Hat'erials in September, X discussed the subject further vith Dr. Hee Chung of-Azgonne National Laboratory and Dr. Friedrich Garzarolli of KWU, through whom X

requested the unpublished German data.

Chung pointed out that, vhile the rate controlling step fn the high tempezature oxidation af zirconium ar zircaloys

'L fs the diffusion of axygen Chrough the axfde and/or through the solid solution af oxygen in zircaloy thaC underlfes it fn both steam and air oxidatfan, there is a significant decrease fn Che axidation rate observed in a steam environment due to an effect of the hydrogen produced during this oxidation on these diffusian constants.

He pointed out that, vhile this effect has been observed by several vorkers, it fs not sufficiently quantified to permit us to use hfgh temperature steam data (such as some of his own, Chose of Prater and Courtright at PNL (4) and those of Pavel and Campbell at Oak Ridge) to estfmate axidatfon rates under our fuel pool accident scenario.

This leaves us, therefare, vith only the White data in the high temperature range.

Garzarolli advised me that most of the German data vere generated by Siegfried Leistikov at Karlsruhei Following the conference, I vzote both to Garzarolli and Lefstfkow, and from both sources received copies of Leistikov's more recent data Xn partfcular, Leistikov sent me, not only unpublished curves fn afr and steam for oxfdation kfnetics of zfrcaloy-4, but several fnternal reports, fn German, that contain the results of a fev short-term J

experiments above 1100'C hppendix X ta this memorandum gives Che cover letter from Leistikov and his recent unpublished data.

You will note from the letter that data at Cemperatures above 1100'C may be available in a year or

so.

I also received the report KFK 2587 dated March 19, )978, which shows some high temperature oxidation rate measurements on'ircaloy-4 in air,oxygen, and steamy I'e included this figure as Appendix 'II'he few data available above 1100 C show that the oxidation rates of zircaloy-4 are'uch greater in

\\air,than they are in either. oxygen or steam, Leistikow's new data show roughly a parabolic corrosion rate behavior (slope of 1/2 on the log log plot) 4 for the first 30-60 minutes in both air and steam They also show that the difference between the air and the steam rates increases with temperature After 30-60 minutes, how'ever, the rate at all but the highest temperature increased dramatically, especially in air.

This may be Rue either to difficulty in controlling the temperature of the highly exothermic zirconium/air oxidation, or to some "breakaway" type phenomenon in the surface oxides exposing the bare metal underneath.

Leistikow drew his curves to suggest a new leveling off, at least at 950 and 1000'C after long times (t >

~

90 min.) ~

At 1ower temperatures, 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 P'arabolic growth, which is more typical of diffusion controlled phenomenan such as are believed to occur at high temperatures.

The new Qerman data show a slope somewhere between 1/3 and 1/2 for the first 30 minutes or so.

In an r

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 curv'es 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 rates at 950,

1000, and 1100'C, on several of his curves in the first 60 minutes or so obtained in steam, and an the short-term data in Appendix II at 1160'C.

These rate constants are also shovn in figure l. It is apparent, thcrefare, that the German steam data and those of Pavel and Campbell for sircalay-4 in steam consistent in the temperture range in which they overlap.

The ncw German air data are consistant with some of their own work (at short exposure times) published some years earlier (5).

From the new German data, I suggest the rate equation:

t 3 ~ 09 x 10 exp (-.'

56,000 RT vhere W is in mg 02 reacted per square cm, t is in seconds, and T is in 'K.

The instantaneous

rate, dw at time t and temperature T

dt 3.09 x 10 exp (-

')

dw 1

8 569000 dt 2w RT The Sandia curve shovs an abrupt increase in oxidation

,vhich they attribute to the mono-tetragonal phase change of is given by (2) rate at 10 /T 7,

ZrOg.

As can be seen in figure 1, the Pavel and Campbell data do not show such an abrupt change at this temperature;

hovevcr, they vere obtained in steam.

The recent results of Prater and Courtright (4) (vhich vere presented at, the 1985 Symposium on Zirconium) shaw that for reactions in steam they find a similar 5ump at temperatures as high as 1500'C (1/T is 5.5 x 10 ") ~

This may be due to effects of the, hydrogen produced by steam reaction on the oxide structure on the sircaloy.

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 1 which is in wt. of Og reacted.

Since a considerable amount of the oxygen that reacts either from air" or steam exists in high concentration solid solutions in the xircaloy, 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 layers I would anticipate the fice energy of formation per gram atom of oxygen reacted be approximately the same for the zirconium oxygen solid solution as for ZrO~ at these high temperatures.

I have included Prater and Courtright's figure as Appendix III Conclusions Based on the information available to date it appears impossible to e

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 sircaloy-4 over those for unalloyed Zr observed by Pawel and Campbell in steam may also exist in air.

For temperatures from 800-1150'C, I think the new German data fit in well with what was prev'iously observed, and suggest using equation 1 given above'owever, 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 ).

cct,QeYo Kyoto Q.T.'ratt V.L Sailor L J Teutonico 0

1 L.J ~ Teutonico, Memo to K Perkins regarding FIN A-3786, March 27, 1985.

2 ~

R E ~ Pawel and J.J

~ Campbell, Zirconium in the Nuclear Industry, ASTH STP

754, pg

370, 1982

'i JoHo White, GEMP-67~ pg. 151, 1967 '

~

J.Ti 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 '

~

ST Leistikow et al., KFK-2262, pg. 233, 1976 '

l727 T

oG ll56 777 838

'27 656 560 l

Sandia Curve 0

2

~

CV

<<t-,

(Pawel et al}

Zr-4 in steam Zr in steam Sandia Curve

-lo Leistikow - l985 o Zr-4 in Air, t<30min

+ Zr-4inAir, t>60min o Zr-4in Steam, t<30min o Zr-4in Air, t< I5min (l978)

I 56600

=5.09xl08 exp ~R t

Zr-4 in Air,t<30 min 8

9 IO"/T, 'K lo l2 Figure 1

+Cat

v

~

o Appendix I liernforschungszentrum Karlsruhe Gcscaschatl mit tteschtanatct Hattuny ttatnt~nuncstannum ttarlstuna GmcH Poslfacn 56cO tt 'r50tt ttgrtatuna t Dr. John Weeks Materials Technology Division Brookhaven National Laboratory It>'tfut fttr M;tturt:tl-itttd Fcsfkorpcrforschttng II

~

tu p

~u ~ i

~ ~

U ton, Lon Island, N.Y. 11973 USA Oatunc 21 10.1985 - Th.

Bc~I: Dr. S. Leistikow Tctefctc 072irI e2-2915 Ihrc NIHung.

Dear Dr. Weeks,

Dr Qazzarolli was right in telling you about ouz Xircaloy-4 oxidation experi-ments in air which in fact are not yet published.

Sorry that we did not work above 1100 C which in fact could be done next year.

So Z send our curves 0

Zizcaloy-4 in a.iz 800 - 1100 C

Zircaloy-4 in steam and air 800-1100 C

o and add same other, more general publications we wrote on Zircaloy-4 behavior under fSR accident conditions.

g was aware of the problem tnormal operating/accident conditions) when I applied for presentation of my paper to the Monterey Conference.

Only your first positive reaction gave me hope.

A participation after the rejection of the paper was then excluded because we are contzibuting to LWR problems only under the safety aspect

- which in fact have their own series of conferences.

Zn case you have further questions don't hesitate to ask me.

We dispose in case of the air oxidation experiments about a lot of other informations.

Very sincerely,

-~.-r Gg-J,.(t=~,~

10 h,m/A fmg/drrP]

o~

~r

<o:

0 6

5 Luft ==

~

~I j

4 ~

rl/I P

0

~

p ~~

Dampf -"::

r'0 I

~j

.I

/

a a

jp A

~'4 J'

oO'r'r' 0

P 4

~

p</

os

~yt re4'7 r~

1100'CP

~4

~ 0 cL rr'

.r' r

~t

~

~r'or

~+

990/ >

~OO 6

5 4

900'~C

~

0,'t 0/5 10 15 coo'c 05 30 1,5 2

90 g,

5

)hi 300 fminl tsotherme Zircato -4/Oxidation VPl Alar'rh rior Mnec:on~>>nnhmnn I Ohhp 1<hh'f

>/A ldm>)

if.

~

I

~

.8.

6 11NPC~~

$NPC

~ ~

~

8 01 4

6 810 Ncaloy-I /Loft-xidaton 0,5 1

1,5 6

8 1

3 4

5 lh) 2 6

8 10(mril

0 l000 Gt>>IchlstunohIIIt leg/onI> )

TODD ZIRCALOY4 -HOCHTEMPERATUR-QXlOATlON OURCH VdASSERDAMPF, SAUERSTOFF UHQ LUFT 5 ABHANGIG-XEIT VOM REAXTlONSTEHPERATUR UNO ZElT 5-15 mfn, 900-130D 4C,1bar 0

0 I

5 min

.lfIfII'V.a."::

0

~

I 6twChtSZunchmt imp/dml) a000 lf00 0 0 0

ll00 l4Cl 2000 o lutt

~ Soutrstcll o

ftosstrdompf f000 10 min 0 O 6tyIIChfStunahmt tmy/dml 1 SOD Q)0 1100 QDD lo()

0

~g 35min 8 0

~

0>>.

SD I

Temperature

( +

Q SQQ 1OQ SOQ

. gOQ

~ PML-Pr'esent Study

ORNL-Cathcart, et. al.

AECL-Urbanic and Heidrick ANL-Baker and Just

'g iO ~

O 0

K

~ee

'iO 3 U

co-4 3.6 4.0 4.6 5.0 6.6 6.0 6.6 7,6

'i 0,000/Temperature

{K)

FfgMre 5

ML17347B682

Contents

Text

Dear Ed:

SUMMARY

8.5 REFERENCES

SUMMARY

Response

SUBJECT:

SUBJECT:

Dear Dr. Weeks,

Navigation menu

ML17347B682

Text

Dear Ed:

SUMMARY

8.5 REFERENCES

SUMMARY

Response

SUBJECT:

SUBJECT:

Dear Dr. Weeks,

Navigation menu

Search