ML16253A414

Declaration of Mark Leyse to Support the Hearing Request & Petition for Leave to Intervene by Bellefonte Efficiency & Sustainability Team/Mothers Against Tennessee River Radiation Regarding Tennessee Valley Authority'S License Amendment Req

ML16253A414
Person / Time
Site:	Browns Ferry
Issue date:	09/06/2016
From:	Leyse M BEST/MATRR
To:	NRC/SECY
SECY RAS
References
EPU, RAS 51300
Download: ML16253A414 (66)

Text

September 6, 2016 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE SECRETARY OF THE COMMISSION In the Matter of:  :

TENNESSEE VALLEY AUTHORITY  :

(Browns Ferry Nuclear Plant Units 1, 2, and 3;:

Docket Nos. 50-259, 50-260, and 50-296;:

NRC-2016-0118)  :

DECLARATION OF MARK LEYSE TO SUPPORT THE HEARING REQUEST AND PETITION FOR LEAVE TO INTERVENE BY THE BELLEFONTE EFFICIENCY AND SUSTAINABILITY TEAM/ MOTHERS AGAINST TENNESSEE RIVER RADIATION REGARDING TENNESSEE VALLEY AUTHORITYS LICENSE AMENDMENT REQUEST FOR EXTENDED POWER UPRATES FOR BROWNS FERRY NUCLEAR PLANT UNITS 1, 2, AND 3 I. Introduction I, Mark Leyse, declare that the following statements are true and correct to the best of my knowledge. I am sui juris. I am over the age of 18 years old.

1. The Bellefonte Efficiency and Sustainability Team/ Mothers Against Tennessee River Radiation (BEST/MATRR) has contracted my services to supply technical analysis and comments in support of their hearing request and petition to intervene in the Tennessee Valley Authoritys (TVA) license amendment request (LAR) for extended power uprates (EPU) for Browns Ferry Nuclear Plant (BFN) Units 1, 2, and 3; NRC-2016-0118.

1

2. I studied nuclear engineering at the University of Wisconsin at Madison from 1979 to 1980. I have a Bachelor of Arts in Fine Arts from the University of California at Berkeley, completed in 1985.

3. I have worked as a nuclear safety consultant since 2010. I have worked for New England Coalition on Nuclear Pollution, Riverkeeper, and Natural Resources Defense Council. As a nuclear safety consultant, I have written 10 C.F.R. § 2.206 enforcement action petitions, a 10 C.F.R. § 2.802 petition for rulemaking, and reports.

4. For Natural Resources Defense Council, I wrote a 10 C.F.R. § 2.802 petition for rulemaking, PRM-50-103,1 requesting post-Fukushima Daiichi accident revisions to 10 C.F.R. § 50.44, Combustible Gas Control for Nuclear Power Reactors. I also wrote three reports:

1) Preventing Hydrogen Explosions In Severe Nuclear Accidents: Unresolved Safety Issues Involving Hydrogen Generation And Mitigation;2 2) Preventing Hydrogen Explosions at Indian Point Nuclear Plant: Fact versus Industry Spin;3 and 3) Post-Fukushima Hardened Vents with High-Capacity Filters for BWR Mark Is and Mark IIs.4

5. On March 15, 2007, I submitted a 10 C.F.R. § 2.802 petition for rulemaking, PRM-50-84,5 to the NRC. PRM-50-84 was summarized briefly in American Nuclear Societys Nuclear Newss 1

Mark Leyse, PRM-50-103, October 14, 2011, (ADAMS Accession No. ML11301A094).

2 Mark Leyse, Author, and Christopher Paine, Contributing Editor, Preventing Hydrogen Explosions In Severe Nuclear Accidents: Unresolved Safety Issues Involving Hydrogen Generation And Mitigation, NRDC Report, R:14-02-B, March 2014. (available at: https://www.nrdc.org/sites/default/files/hydrogen-generation-safety-report.pdf : last visited on 08/28/16) 3 Mark Leyse and Christopher Paine, Preventing Hydrogen Explosions at Indian Point Nuclear Plant: Fact versus Industry Spin, NRDC IB: 13-01-F, February 2013. (available at:

https://www.nrdc.org/sites/default/files/IndianPoint-hydrogen-explosions-IB.pdf : last visited on 08/28/16) 4 Mark Leyse, Post-Fukushima Hardened Vents with High-Capacity Filters for BWR Mark Is and Mark IIs, Report for NRDC, July 2012, (ADAMS Accession No. ML12254A865).

5 Mark Leyse, PRM-50-84, March 15, 2007 (ADAMS Accession No. ML070871368).

2

June 2007 issue6 and commented on and deemed a well-documented justification forrecommended changes to the [NRCs] regulations7 by the Union of Concerned Scientists (UCS).

6. PRM-50-84 requested that NRC make new regulations: 1) to require licensees to operate light water reactors under conditions that effectively limit the thickness of crud (corrosion products) and/or oxide layers on fuel cladding, in order to help ensure compliance with 10 C.F.R. § 50.46(b) emergency core cooling system (ECCS) acceptance criteria; and 2) to stipulate a maximum allowable percentage of hydrogen content in fuel cladding.

7. Additionally, PRM-50-84 requested that NRC amend Appendix K to Part 50ECCS Evaluation Models I(A)(1), The Initial Stored Energy in the Fuel, to require that the steady-state temperature distribution and stored energy in the fuel at the onset of a postulated loss-of-coolant accident (LOCA) be calculated by factoring in the role that the thermal resistance of crud and/or oxide layers on cladding plays in increasing the stored energy in the fuel. PRM 84 also requested that these same requirements apply to any NRC-approved best-estimate ECCS evaluation models used in lieu of Appendix K to Part 50 calculations. (Best-estimate ECCS evaluation models used in lieu of Appendix K to Part 50 calculations are described in NRC Regulatory Guide 1.157.)

8. In 2008, the NRC decided to consider the safety issues raised in PRM-50-84 in its rulemaking process.8 And in 2009, the NRC published Performance-Based Emergency Core Cooling System Acceptance Criteria, which gave advanced notice of a proposed rulemaking, 6

American Nuclear Society, Nuclear News, June 2007, p. 64.

7 David Lochbaum, Union of Concerned Scientists, Comments on Petition for Rulemaking Submitted by Mark Edward Leyse (Docket No. PRM-50-84), July 31, 2007, (ADAMS Accession No. ML072130342),

p. 2.

8 NRC, Mark Edward Leyse; Consideration of Petition in Rulemaking Process, Docket No. PRM-50-84; NRC-2007-0013, Federal Register, Vol. 73, No. 228, November 25, 2008, pp. 71564-71569.

3

addressing four objectives: the fourth being the issues raised in PRM-50-84.9 In 2012, the NRC Commissioners voted unanimously to approve a proposed rulemakingrevisions to Section 50.46(b), which will become Section 50.46(c)that is partly based on the safety issues I raised in PRM-50-84.10

9. With Rui Hu and Professor Mujid S. Kazimi of the Massachusetts Institute of Technology, I coauthored a paper, Considering the Thermal Resistance of Crud in LOCA Analysis, that was presented at the American Nuclear Societys 2009 Winter Meeting.11

10. On November 17, 2009, I submitted a 10 C.F.R. § 2.802 petition for rulemaking, PRM-50-93.12 PRM-50-93 requests that NRC make new regulations: 1) to require that the calculated maximum fuel element cladding temperature not exceed a limit based on data from multi-rod (assembly) severe fuel damage experiments; and 2) to stipulate minimum allowable core reflood rates, in the event of a LOCA.

11. Additionally, PRM-50-93 requests that NRC revise Appendix K to Part 50ECCS Evaluation Models I(A)(5), Required and Acceptable Features of the Evaluation Models, Sources of Heat during the LOCA, Metal-Water Reaction Rate, to require that the rates of energy release, hydrogen generation, and cladding oxidation from the metal-water reaction considered in ECCS evaluation calculations be based on data from multi-rod (assembly) severe fuel damage experiments. These same requirements also need to apply to any NRC-approved best-estimate ECCS evaluation models used in lieu of Appendix K to Part 50 calculations.

9 NRC, Performance-Based Emergency Core Cooling System Acceptance Criteria, NRC-2008-0332, Federal Register, Vol. 74, No. 155, August 13, 2009, pp. 40765-40776.

10 NRC, Commission Voting Record, Decision Item: SECY-12-0034, Proposed Rulemaking10 CFR 50.46(c): Emergency Core Cooling System Performance During Loss-of-Coolant Accidents (RIN 3150-AH42), January 7, 2013, (ADAMS Accession No. ML13008A368).

11 Rui Hu, Mujid S. Kazimi, Mark Leyse, Considering the Thermal Resistance of Crud in LOCA Analysis, American Nuclear Society, 2009 Winter Meeting, Washington, D.C., November 15-19, 2009.

12 Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No. ML093290250).

4

12. PRM-50-93 was discussed briefly in the American Nuclear Societys March 2010 issue of Nuclear News.13 PRM-50-93 was also commented on by UCS.

13. Regarding PRM-50-93, UCS states:

In our opinion, [PRM-50-93] addresses a genuine safety problem. We believe the NRC should embark on a rulemaking process based on this petition. We are confident that this process would culminate in revised regulationsperhaps not precisely the ones proposed [in PRM-50-93] but ones that would adequately resolve the issuesmeticulously identified [in PRM-50-93]that would better ensure safety in event of a loss of coolant accident.14

14. On October 27, 2010, the NRC published in the Federal Register that it had determined that a 10 C.F.R. § 2.206 petition, dated June 7, 2010, I wrote and submitted on behalf of New England Coalitionrequesting that the NRC order the licensee of Vermont Yankee Nuclear Power Station (VYNPS) to lower the licensing basis peak cladding temperature of VYNPSmeets the threshold sufficiency requirements for a petition for rulemaking under 10 C.F.R. § 2.802.15 The NRC docketed the 10 C.F.R. § 2.206 petition as a petition for rulemaking, PRM-50-95.16 PRM-50-95 was discussed briefly in the July 30, 2010 issue of Plattss Inside NRC.17

15. My expert opinions and comments in this declaration are based both on my professional experience and on my review of relevant aspects of TVAs license amendment request for EPUs for BFN Units 1, 2, and 3.

13 American Nuclear Society, Nuclear News, March 2010, p. 36.

14 David Lochbaum, Union of Concerned Scientists, Comments Submitted by the Union of Concerned Scientists on the Petition for Rulemaking Submitted by Mark Edward Leyse (Docket No. PRM-50-93),

April 27, 2010, (ADAMS Accession No. ML101180175), p. 1.

15 Federal Register, Vol. 75, No. 207, Notice of consolidation of petitions for rulemaking and re-opening of comment period, October 27, 2010, pp. 66007-66008.

16 Mark Leyse, PRM-50-95, June 7, 2010, (ADAMS Accession No. ML101610121).

17 Suzanne McElligott, Inside NRC, July 30, 2010.

5

II. Background The loss-of-coolant accident (LOCA) Analyses AREVA Conducted to Help Justify the Amendment Request for the EPUs for BFN Units 1, 2, and 3

16. The Federal Register notice that the NRC published on July 5, 2016 regarding the proposed EPUs for BFN Units 1, 2, and 3 states:

The Power Uprate Safety Analysis Report (PUSAR) summarizes the results of safety evaluations performed that justify uprating the licensed thermal power at BFN. The PUSAR uses GEH [General Electric-Hitachi]

GE14 fuel as the principal reference fuel type for the evaluation of the impact of EPU [extended power uprate]. However, the BFN units will utilize AREVA ATRIUM 10XM fuel, with some legacy ATRIUM 10 fuel, under EPU conditions. Therefore, the AREVA Fuel Uprate Safety Analysis Report (FUSAR) for Browns Ferry Units 1, 2, and 3 and fuel related reports are provided to supplement the PUSAR and address the impact of EPU conditions on the AREVA fuel in the BFN units. The AREVA analyses contained in the FUSAR have provided disposition of the critical characteristics of the GE14 fuel and have been shown to bound ATRIUM 10XM and ATRIUM 10 fuel.18

17. The AREVA LOCA analyses that were conducted to help justify the LAR for the EPUs for BFN Units 1, 2, and 3 are discussed in three AREVA reports: ANP-3377NP (regarding ATRIUM 10XM fuel), ANP-3378NP (regarding ATRIUM 10XM fuel), and ANP-3384NP (regarding ATRIUM 10 fuel). An important result of a LOCA analysis is the value that the maximum temperature the cladding of the fuel rods is predicted to reach: the peak cladding temperature (PCT). The LOCA analyses regarding the EPUs for BFN Units 1, 2, and 3 18 NRC, Applications and Amendments to Facility Operating Licenses and Combined Licenses Involving Proposed No Significant Hazards Considerations and Containing Sensitive Unclassified Non-Safeguards Information and Order Imposing Procedures for Access to Sensitive Unclassified Non-Safeguards Information, NRC-2016-0118, Federal Register, Vol. 81, No. 128, July 5, 2016, p. 43666.

6

discussed in ANP-3377NP, ANP-3378NP, and ANP-3384NP, predicted PCTs of 2030°F,19 2008°F,20 and 2086°F,21 respectively.

18. The overall predicted PCT is 2030°F for ATRIUM 10XM fuel, which is used at BFN (Units 1, 2, and 3).22 AREVAs analyses were performed for a [reactor] core composed entirely of ATRIUM 10XM fuel at beginning-of-life (BOL) conditions. Calculations assumed an initial core power of 102% of 3952 MWt, providing an analysis licensing basis power of 4031 MWt.

The 2.0% increase reflects the maximum uncertainty in monitoring reactor power, as per NRC requirements. 3952 MWt corresponds to 120% of the original licensed thermal power (OLTP) and is referred to as extended power uprate (EPU).23

19. And the overall predicted PCT is 2086°F for ATRIUM 10 fuel, which is used at BFN (Units 1, 2, and 3).24 Apparently, the plan for BFN is that all three reactors will primarily use ATRIUM 10XM fuel after the EPU is implemented. The plan is to maybe include some ATRIUM 10 fuel in a transition cycle along with ATRIUM 10XM fuel after the EPU is 19 AREVA, Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU), ANP-3377NP, Revision 3, Attachment 11 Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No:

ML15282A184), pp. 6.1, 6.3, 6.9, 8.6.

20 AREVA, Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM 10XM Fuel (EPU),ANP-3378NP, Revision 3, Attachment 13 Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No: ML15282A185), pp. 2.3, 5.1, 5.4, 6.1.

21 AREVA, Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU), ANP-3384NP, Revision 3, Attachment 15 Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No: ML15282A187), pp. 2.2, 5.1, 5.4, 6.1.

22 AREVA, Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU), ANP-3377NP, Revision 3, Attachment 11 Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No:

ML15282A184), p. 1.2.

23 Id., p. 1.1.

24 AREVA, Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU), ANP-3384NP, Revision 3, Attachment 15 Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No: ML15282A187), p. 1.1.

7

implemented. At EPU power, any ATRIUM-10 fuel would be in its third cycle of operation.25 The Computer Safety Model that AREVA Used to Conduct LOCA Analyses for the Amendment Request for the EPUs for BFN Units 1, 2, and 3

20. AREVA has stated that [t]he models and computer codes used by AREVA for LOCA analyses

[regarding the EPUs for BFN Units 1, 2, and 3] are collectively referred to as the EXEM BWR-2000 Evaluation Model. The EXEM BWR-2000 Evaluation Model has been approved for reactor licensing analyses by the NRC.26

21. The EXEM BWR-2000 Evaluation Model LOCA calculations for the EPUs for BFN Units 1, 2, and 3 were performed in conformance with 10 CFR 50 Appendix K requirements and satisfy the event acceptance criteria identified in 10 CFR 50.46.27 In regard to the zirconium-steam reaction that would occur in the event of a LOCA, 10 C.F.R. 50 Appendix K, I.A.5 requires that

[t]he rate of energy release, hydrogen generation, and cladding oxidation from the metal-water reaction shall be calculated using the Baker-Just [correlation].28 25 Id., p. 1.2.

26 AREVA, Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU), ANP-3377NP, Revision 3, Attachment 11 Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No:

ML15282A184), p. 1.1.

27 Id.

28 NRC, Appendix K to Part 50ECCS Evaluation Models, (This information is available at:

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appk.html : last visited on 09/02/16).

8

III. 10 C.F.R. 50 Appendix K I.A.5: Historical Background Experimental Data Demonstrates that 10 C.F.R. 50 Appendix K, I.A.5 Is Non-Conservative

22. 10 C.F.R. 50 Appendix K, I.A.5 is non-conservative. As discussed in this declaration, experimental data, along with appropriate citations, demonstrates that the Baker-Just correlation is inadequate for use in computer safety models like AREVAs EXEM BWR-2000 Evaluation Model.

23. The Baker-Just correlationused in Appendix K to Part 50 ECCS evaluation calculations dates back to 1962.29 The Baker-Just correlation is primarily based on data from Alexis Lemmon and W. A. Bostroms experiments,30 which were conducted in the 1950s.31 Bostroms experiments were conducted above the temperature range of design-basis accidents. Lemmon and Bostroms experiments were conducted with tiny inductively heated Zircaloy-2 specimens.32 (Lemmons specimens were Zircaloy-2 cylinders that were 2.0 inches long and 0.5 inches in diameter.33) There are radiative heat losses in experiments conducted with inductive heating, which affect a specimens oxidation kinetics.34 29 Louis Baker, Jr. and Louis C. Just, Studies of Metal-Water Reactions at High Temperatures: III.

Experimental and Theoretical Studies of the Zirconium-Water Reaction, ANL-6548, May 1962, (ADAMS Accession No: ML050550198).

30 G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003, p. 2.

31 W. A. Bostrom, The High Temperature Oxidation of Ziracloy in Water, WAPD-104, March 1954, (ADAMS Accession No: ML100900446) and Alexis W. Lemmon, Studies Relating to the Reaction Between Zirconium and Water at High Temperatures, Battelle Memorial Institute, BMI-1154, January 1957, (ADAMS Accession No: ML100570218).

32 V. F. Urbanic and T. R. Heidrick, High-Temperature Oxidation of Zircaloy-2 and Zircaloy-4 in Steam, Journal of Nuclear Materials 75, 1978, p. 252.

33 Alexis W. Lemmon, Studies Relating to the Reaction Between Zirconium and Water at High Temperatures, Battelle Memorial Institute, BMI-1154, January 1957, (ADAMS Accession No:

ML100570218), p. C-4.

34 G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003, pp. 4-5.

9

24. Regarding how radiative heat losses in inductive specimen heating experiments affect oxidation kinetics, a 2003 paper by G. Schanz states:

[Ocken] stated that [the] advantage [of experiments with inductive (Urbanic and Heidrick) and direct electrical heating (Biederman, et al.) of a specimen in a cool environment35] would be the temperature gradient from heated specimen to cool surrounding[s], leading to temperature gradients in the cladding wall in the same sense as in a reactor. In total disagreement with the argument of Ocken, the author of this paper stresses the advantage of a constant cladding wall temperature and thus of a better defined specimen temperature, as provided in furnace experiments! ... This argument was already used by Sawatzky, et al., who performed similar studies with inductive specimen heating as Urbanic and Heidrick. Sawatzky reached an important improvement of the specimen temperature homogeneity by only optimizing the geometry of the specimen and registered considerably increased reaction rates36 [emphasis added].

25. Radiative heat losses in an experiment conducted with inductive heating cause a specimens zirconium-steam reaction rates to decrease below what they would be if there were no radiative heat losses. The very experiments that the Baker-Just correlation is primarily based on would have had radiative heat losses that decreased zirconium-steam reaction rates. Lemmon and Bostroms experiments certainly did not replicate the oxidation kinetics that would occur in a nuclear reactors core, in the event of a LOCA. Yet the Baker-Just correlationrequired by Appendix K to Part 50 I.A.5is almost entirely based on the results of their experiments. This fact alone is evidence that the Baker-Just correlation is likely inadequate for use in computer safety models like AREVAs EXEM BWR-2000 Evaluation Model.

26. Results of larger scale experiments discussed in this declaration, along with appropriate citations, present far more conclusive evidence that the Baker-Just correlation is indeed inadequate for use in computer safety models like AREVAs EXEM BWR-2000 Evaluation Model. For example, on November 24, 2015, Aby Mohseni, Deputy Director of the NRCs 35 G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003, pp. 4-5.

36 Id.

10

Division of Policy and Rulemaking, disclosed to Leyse that an NRC (TRACE code) computer simulation (using the Baker-Just correlation) of a Westinghouse design-basis accident experiment (FLECHT Run 9573), under-predicted cladding and steam temperatures at the elevation of the hottest section of the tests fuel rod simulators.37 A computer safety model is supposed to over-predict temperatures in order to ensure an adequate margin of safety.

27. If a reactors power level is set too high after being qualified by LOCA analyses that do not ensure an adequate margin of safety, a real-life LOCA would lead to a beyond design-basis accident. In other words, if a reactors power level is set too high after being qualified by LOCA analyses that do not ensure an adequate margin of safety, in the event of a LOCA, the criteria set forth in 10 C.F.R. § 50.46(b) would be violated: 1) the PCT would exceed 2200°F;

2) the maximum cladding oxidation would locally exceed 0.17 times the total cladding thickness before oxidation; 3) the total amount of hydrogen generated from the chemical reaction of the cladding with steam would exceed 0.01 times the hypothetical amount that would be generated if all of the metal in the cladding cylinders surrounding the fuel, excluding the cladding surrounding the plenum volume, were to react; 4) the reactor core geometry would not remain amenable to cooling; 5) there would not be long-term cooling of the reactor core; the core temperature would not be maintained at an acceptably low value and decay heat would not be removed for the extended period of time required, as a consequence of the long-lived decay of fission products that remain in the core.

37 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

11

In 1971, in the Licensing Hearings for Indian Point Nuclear Plant Unit 2, Union of Concerned Scientists Alleged that the Baker-Just Correlation Is Inadequate for Use in Computer Safety Models that Simulate Loss-of-Coolant Accidents

28. The Indian Point Nuclear Plant Unit 2 licensing hearings were held because the Citizens Committee for the Protection of the Environment and other intervenors opposed the licensing of Indian Points Unit 2 reactor. The Union of Concerned Scientists (UCS) provided technical expertise for the Citizens Committee. UCS alleged that Westinghouse lacked a foundation for its claim that its emergency systems would prevent a meltdown in the event of a loss-of-coolant accident.38 Prior to the hearings, UCS had stated that until an independent third party reviewed and assured the performance of emergency systems they couldnt support the licensing and operation of any additional power reactors in the United States.39

29. In the Indian Point Unit 2 licensing hearings, UCS contended that results of the First Transient Experiment of a Zircaloy Fuel Rod Cluster (FRF-1) experiment, which was conducted at the Transient Reactor Test Facility (TREAT), a nuclear reactor in Idaho, indicated that the zirconium-steam reaction is more severe than industry claimed.40 According to scientists at Oak Ridge National Laboratory, as of 1971, the FRF-1 experiment was conducted under the most 38 Daniel F. Ford, Henry W. Kendall, James J. MacKenzie, A Critique of the New A. E. C. Design Criteria for Reactor Safety Systems, Union of Concerned Scientists, October 1971; I.A. Forbes, D.F. Ford, H.W.

Kendall, J.J. MacKenzie, Nuclear Reactor Safety: An Evaluation of New Evidence, Nuclear News, 14, No. 9, September 1971.

39 Henry W. Kendalls A Distant Light: Scientists and Public Policy has reprinted I.A. Forbes, D.F. Ford, H.W. Kendall, J.J. MacKenzie, Nuclear Reactor Safety: An Evaluation of New Evidence, Nuclear News, 14, No. 9, September 1971. For the quoted passage see Henry W. Kendall, A Distant Light: Scientists and Public Policy, (New York: Springer-Verlag, 2000), p. 35.

40 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), pp. 2297-2299.

12

realistic loss-of-coolant accident conditions of any experiment to date.41 UCS believed that results of the FRF-1 experiment indicated Con Edisons license application should be re-evaluated.42

30. It was reported that industrys computer safety model (using the Baker-Just correlation) vastly under-predicted the extent of the zirconium-steam reaction that occurred in the FRF-1 experiment,43 indicating the model was unfit for simulating the type of LOCAs that could occur at Indian Point.

31. In fact, data from the FRF-1 experiment indicates that computer safety models (using the Baker-Just correlation) under-predict the quantity of hydrogen produced by the Zircaloy-steam reaction. In the experiment, at fuel rod temperatures of about 1800°F, the Zircaloy-steam reaction generated 1.2 +/- 0.6 liters of hydrogen. In the Indian Point Unit 2 licensing hearing, Westinghouse, which had performed experimental simulations of loss-of-coolant accidents, and conducted computer simulations of such accidents, testified that their computer safety models (using the Baker-Just correlation) predicted that there would be no zirconium-steam reaction at 1800°Fthat no hydrogen would be produced in a LOCA if local temperatures of the fuel rods were to reach 1800°F.44 41 R. A. Lorenz, D. O. Hobson, G. W. Parker, Final Report on the First Fuel Rod Failure Transient Test of a Zircaloy-Clad Fuel Rod Cluster in TREAT, ORNL-4635, March 1971, p. 75.

42 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), pp. 2297-2298.

43 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 1, 1971, (ADAMS Accession No. ML100350644), pp. 2152, 2166-2167.

44 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 1, 1971, (ADAMS Accession No. ML100350644), pp. 2152-2153.

13

32. However, in the Indian Point Unit 2 licensing hearings, Dr. John Bernard Roll, a manager in Westinghouses Nuclear Fuel Division, testified that data hadnt been accurately recorded in the FRF-1 experiment. He claimed the test results didnt prove anything.45 Unfortunately, the AEC wasnt too eager to replicate the test with accurate data measurements in order to investigate whether its results were valid or not. It decided to kill funding for the TREAT reactors LOCA test program.46

33. In the Indian Point Unit 2 licensing hearings, Daniel F. Ford of UCS asked a number of questions about the Baker-Just correlation. Ford questioned whether or not the Baker-Just correlation was valid; that is, he questioned whether or not the Baker-Just correlation is adequate for use in computer safety models that simulate LOCAs. Regarding Fords questions, Dr. Roll stated: The line of questioning between Mr. Ford and myself really questioned the validity and applicability of the assumptions which Baker and Just made, and whether or not the validity of these assumptions in any way through a question or use of the equation [the Baker-Just correlation] in the analysis of the loss of coolant accident.47

34. By alleging that the Baker-Just correlation is inadequate for use in computer safety models that simulate LOCAs, UCS also alleged that 10 C.F.R. 50 Appendix K, I.A.5 is non-conservative. In regard to the zirconium-steam reaction that would occur in the event of a LOCA, 10 C.F.R. 50 Appendix K, I.A.5 requires that [t]he rate of energy release, hydrogen generation, and cladding 45 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), pp. 2297-2299.

46 W. B. Cottrell, ORNL Nuclear Safety Research and Development Program Bimonthly Report for March-April 1971, ORNL-TM-3411, July 1971, p. x.

47 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 16, 1971, (ADAMS Accession No. ML100350625), p. 3863.

14

oxidation from the metal-water reaction shall be calculated using the Baker-Just

[correlation].48 In 1971, in the Licensing Hearings for Indian Point Nuclear Plant Unit 2, Dr. John Bernard Roll, a Manager in Westinghouses Nuclear Fuel Division, Made False Statements Under Oath, Defending the Baker-Just Correlation

35. In the Indian Point Unit 2 licensing hearings, after Dr. Roll testified that data hadnt been accurately recorded in the FRF-1 experiment, he testified that Westinghouses FLECHT tests were superior to the FRF-1 experiment in terms of replicating how fuel rods would perform in an accident. He claimed that the FLECHT results reaffirmed the validity of the industrys computer safety model (using the Baker-Just correlation) for simulating the extent of the zirconium-steam reaction that would occur in the event of a LOCA.49 In fact, some of the FLECHT results did just the opposite.

36. In 1971, the year after Westinghouses FLECHT tests had been completed, employees of Westinghouse, including Dr. Roll, testified in the licensing hearings for Indian Point Unit 2, because the Unit 2 reactor is a Westinghouse design. The testimony of Dr. Roll served to counter charges that a reactor accident would be worse than industry claimed. Dr. Rolls job was to review, interpret, and model data from experiments that simulated LOCAs.50 When he was under oath, Dr. Roll made false statements, defending the Baker-Just correlations use in 48 NRC, Appendix K to Part 50ECCS Evaluation Models, (This information is available at:

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appk.html : last visited on 09/02/16).

49 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 1, 1971, (ADAMS Accession No. ML100350644), pp. 2297-2299.

50 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), p. 2297.

15

the industrys computer safety model for simulating the extent of the zirconium-steam reaction that would occur in the event of a LOCA.

37. After Dr. Roll was sworn in, Leonard Trosten, the attorney representing the Indian Point Unit 2 license applicant, Con Edison, asked him to respond to UCSs allegation that industrys computer safety model (using the Baker-Just correlation) vastly under-predicted the extent of the zirconium-steam reaction that occurred in the FRF-1 experiment,51 indicating the model was unfit for simulating the type of LOCAs that could occur at Indian Point.

38. After Dr. Roll testified that data hadnt been accurately recorded in the FRF-1 experiment and that the test results didnt prove anything,52 he stated:

Id like to add further that we [Westinghouse] have, as a part of our work, in particular under the FLECHT program, reviewed the extent of zirc-water reaction, under what we considered to be much more representative conditions [than those of the FRF-1 experiment], that is zircalloy clad fuel rods with our particular time and temperature histories and our particular coolant content, that is our particular water conditions, and I believe as reported in the documentation summarized in the FLECHT reports we find very good agreement with the Baker-just equation

[correlation], and so we believe in summary that the Oak Ridge report53 [on the FRF-1 experiment] presents a single data point to germaneness to our specific application must be questioned inasmuch as the data point was not, the test was not run to substantiate the Baker-Just equation [correlation] [emphasis added].

And secondly, in summary, the work that we have done under the FLECHT program and reported in the FLECHT reports we believe reaffirms our use of the Baker-Just equations in evaluating zirc-water reaction under our conditions of loss of coolant accident [emphasis added].54 51 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 1, 1971, (ADAMS Accession No. ML100350644), pp. 2152, 2166-2167.

52 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), pp. 2298-2299.

53 R. A. Lorenz, D. O. Hobson, G. W. Parker, Final Report on the First Fuel Rod Failure Transient Test of a Zircaloy-Clad Fuel Rod Cluster in TREAT, ORNL-4635, March 1971, p. 75.

54 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), p. 2299.

16

39. In the Indian Point Unit 2 licensing hearings, Dr. Roll failed to mention that in the FLECHT program, part of the FLECHT Run 9573 test bundle incurred thermal runaway, as a result of the heat generated by the zirconium-steam reaction.

40. As stated above in Section IV.A, on November 24, 2015, Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, disclosed to Leyse that an NRC (TRACE code) computer simulation (using the Baker-Just correlation) of FLECHT Run 9573, under-predicted cladding and steam temperatures at the elevation of the hottest section of the tests fuel rod simulators.55 More than four decades after the Indian Point Unit 2 licensing hearings, the truth has been revealed: Dr. Roll made false statements, defending the Baker-Just correlations use in the industrys computer safety model for simulating the extent of the zirconium-steam reaction that would occur in the event of a LOCA.

41. After Dr. Roll testified that the work Westinghouse had done under the FLECHT program and reported in the FLECHT reports...reaffirms our use of the Baker-Just equations [correlation] in evaluating zirc-water reaction under our conditions of loss of coolant accident,56 Daniel F.

Ford of UCS asked him a number of questions about FLECHT results and the Baker-Just correlation. Ford asked Dr. Roll to describe the techniques of FLECHT measurement of zircalloy-water reaction that were used in your FLECHT tests.57 55 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

56 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), p. 2299.

57 Id., p. 2300.

17

42. Dr. Roll answered Ford, explaining:

The measurement that we took in evaluating the result of our FLECHT test with regard to extent of zirc-water reaction were in fact metalographic cross-sections at various enlargements from which the experienced metalographers can infer [the]

nature of the phases in the cross-section. That is they can determine the portion of the original zircalloy which remains as original zircalloy. That portion which is oxygen saturated, that portion which is in fact converted to zirconium oxide. With these direct measurements at a number of cross-sections, one can then calculate explicitly the quantity of zirconium which has been converted to zirconium dioxide and the quantity of zirconium which is oxygen saturated from which you can then determine the total quantity of zirconium which has in fact reacted in some way with the oxygen.58

43. Dr. Roll further explained:

I believe the technique of looking at zirconium and zirconium oxide is in itself a primary source of data and need not be substantiated somewhere else. The question is, how do we know what is the extent of [the] zirconium and oxygen reaction. The answer is, you know this by looking at the quantity of zirconium which has been converted to zirconium oxide.59

44. Attempting to explain that the FLECHT program data affirmed that the Baker-Just correlation is adequate for use in computer safety models that simulate LOCAs, Dr. Roll concluded:

Let me refer to WCAP-7665 figures on page B-20, in particular and on B-23.60 I believe the answer to your question, does the prediction; i.e., Baker-Just, go over the top of the data? I think the answer is essentially yes, looking particularly at the figure on page B-20.61 (The figure on page B-20 of WCAP-7665 is copied below; it is Figure 1 of this declaration.)

58 Id., p. 2302.

59 Id., p. 2303.

60 See F. D. Kingsbury, J. F. Mellor, and A. P. Suda, Materials Evaluation, Appendix B of WCAP-7665. F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse Electric Corporation, WCAP-7665, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, April 1971, (ADAMS Accession No: ML070780083), Appendix B, pp. B-20, B-23.

61 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), p. 2303.

18

45. The document Dr. Roll referred to is Materials Evaluation, Appendix B of WCAP-7665.

Materials Evaluation explains that; In order to properly analyze FLECHT [tests conducted with zirconium test bundles], information regarding the amount of energy released during a test by the metal-water reaction was required. The purpose of the materials evaluation portion of the FLECHT program was to determine the extent of metal-water reaction in thetests [conducted with zirconium test bundles] and compare it with the predictions of an analytical model.62 Westinghouses analytical model used the Baker-Just correlation.

46. To conduct the materials evaluation, metallographic specimens were selected from the FLECHT programs zirconium test bundles.63 Then the thicknesses of the oxide layers of those specimens were compared to predicted (calculated) oxide layer thicknesses that were simulated (generated under the same temperature conditions that generated the real-life selected specimens). Westinghouses analytical model, using the Baker-Just correlation, predicted (calculated) the oxide layer thicknesses. Materials Evaluation, Appendix B of WCAP-7665 explains that [t]he calculated oxide thickness datawere obtained using the Baker and Just parabolic rate equation [Baker-Just correlation] and the detailed temperature-time output of the thermocouples located at the sections examined.64

47. Supporting Dr. Rolls conclusion, Materials Evaluation, Appendix B of WCAP-7665, concluded that [t]he Baker-Just parabolic rate equation [Baker-Just correlation] appears to provide a satisfactory basis for determining the overall extent of metal-water reaction.65 62 See F. D. Kingsbury, J. F. Mellor, and A. P. Suda, Materials Evaluation, Appendix B of WCAP-7665. F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse Electric Corporation, WCAP-7665, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, April 1971, (ADAMS Accession No: ML070780083), Appendix B, p. B-1.

63 Id., Appendix B, pp. B-2, B-3.

64 Id., Appendix B, p. B-19.

65 Id., Appendix B, p. B-24.

19

48. Also, supporting Dr. Rolls conclusion, Materials Evaluation, Appendix B of WCAP-7665, explains that, as shown in Figure B-12 (see Figure 1 of this declaration below), [i]t is evident that the calculated thicknesses are consistently high, with the error increasing with increasing oxide thickness. The calculated oxide thickness datawere obtained using the Baker and Just [correlation].66 66 Id., Appendix B, p. B-19.

20

Figure 1 (of this declaration). The figure on page B-20 of WCAP-7665 that Dr. Roll referred to in his testimony in the Indian Point Unit 2 licensing hearings 21

Cherry-Picking Experimental Data: The Main Problem with Dr. John Bernard Rolls Testimony

49. As stated above, in the Indian Point Unit 2 licensing hearings, Dr. Roll failed to mention that in the FLECHT program, part of the FLECHT Run 9573 test bundle incurred thermal runaway, as a result of the heat that was generated by the zirconium-steam reaction. Dr. Roll also failed to mention that in the materials evaluation of the FLECHT program, samples were not taken from the section of the FLECHT Run 9573 test bundle that incurred thermal runaway.67 In other words, there was cherry-picking of experimental data in the FLECHT program.

50. A section of the FLECHT Run 9573 test bundles zirconium cladding essentially caught on fire.

The cladding burned in steamthen, when cooled, shattered like overheated glass doused with cold water. (A photograph of the destroyed test bundle is depicted below in Figure 2.) In WCAP-7665, Westinghouse referred to the severely burnt, shattered section of the FLECHT Run 9573 test bundle as the severe damage zone and noted that the remainder of the [test]

bundle was in excellent condition.68 67 See F. D. Kingsbury, J. F. Mellor, and A. P. Suda, Materials Evaluation, Appendix B of WCAP-7665. F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse Electric Corporation, WCAP-7665, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, April 1971, (ADAMS Accession No: ML070780083), Appendix B, p. B-4.

68 F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, WCAP-7665, April 1971, (ADAMS Accession No. ML070780083),

p. 3.97.

22

Figure 2. The severe-damage zone of the FLECHT test bundle from Run 9573

51. Metal experts at Idaho Nuclear Corporation examined metallographic specimens that were selected from the FLECHT Run 9573 test bundle as well as three other zirconium test bundles from the FLECHT program. They wanted to determine the extent of the zirconium-steam reaction that had occurred at different locations of the test bundles. However, they did not examine any metallographic specimens from the FLECHT Run 9573 test bundles severe damage zone.69 Metallographic specimens were not taken from the severe damage zone. By way of an analogy what they did would be like trying to determine how severely trees burned in a forest fire by ignoring trees reduced to ash and only examining those that had been singed.

69 F. D. Kingsbury, J. F. Mellor, A. P. Suda, Materials Evaluation, Appendix B of Westinghouses PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, WCAP-7665, April 1971, (ADAMS Accession No. ML070780083), pp. B.1-B.25.

23

52. Westinghouse is likely the company responsible for the cherry-picking. They likely only sent Idaho Nuclear Corporation sections of the test bundle that were in decent shapeor metallographic specimens that were taken sections of the test bundle that were in decent shape.

But who knows what actually happened? The main thing is that there was cherry-picking. The identity of the culprit is less important.

53. Regardless of who cherry-picked the singed cladding samples, Westinghouse was enabled to downplay the extent of the zirconium-steam reaction. This is a serious problem. In a reactor accident, the reaction between zirconium and steam generates a lot of heat and leads to a meltdown. One thing is certain, in the Indian Point Unit 2 licensing hearings, Dr. Roll, a Westinghouse employee, made false statements, defending the Baker-Just correlation. His false statements were intended to support the claim that the Baker-Just correlation is adequate for use in computer safety models that simulate LOCAs.

Problems with the Metallurgical Data from the FLECHT Program

54. Four of the FLECHT tests were conducted with bundles of heater rods sheathed in zirconium alloy (Zircaloy) cladding. Those tests are FLECHT Runs 2443, 2544, 8874, and 9573.

FLECHT Runs 8874 and 9573

55. There are significant problems with Westinghouses examinations of the metallographic cross-sections that were taken from test rods from FLECHT Run 9573, because Westinghouse did not obtain metallurgical data from the locations of the rods from Run 9573 that incurred thermal runawaythe severe-damage zone. FLECHT run 8874 had also incurred thermal runaway.

And Westinghouse did not obtain metallurgical data from the locations of the rods from Run 24

8874 that incurred runaway oxidation.70 It is probable that the locations of the test bundles from Runs 8874 and 9573 that Westinghouse did examine were steam starved: the examined locations had limited oxidation because they were only exposed to a limited amount of steam.

56. It is reasonable to assume thatas in the CORA-2 experiment, in which local steam starvation conditions are postulated to have occurred71in FLECHT Runs 8874 and 9573, violent oxidation essentially consumed much of the available steam, so that time-limited and local steam starvation conditions, which cannot be detected in a post-test investigation, would have occurred.

57. Therefore, Westinghouses application of the Baker-Just zirconium-steam correlation (used in computer safety models) to the oxide layers on the test bundles from FLECHT Runs 8874 and 9573 were to locations that most likely were steam starved or partly steam starved (hydrogen produced by the zirconium-steam reaction would have also diluted the available steam).

Clearly, that is not a legitimate verification of the adequacy of the Baker-Just correlation for use in computer safety models.

58. Subsequently, the NRC applied the Baker-Just and Cathcart-Pawel correlations to the metallurgical data from the four FLECHT Zircaloy experiments:72 unfortunately, the NRC did not apply the Baker-Just and Cathcart-Pawel correlations to metallurgical data from the locations of FLECHT Runs 8874 and 9573 that incurred thermal runaway. Hence, NRCs 70 See F. D. Kingsbury, J. F. Mellor, and A. P. Suda, Materials Evaluation, Appendix B of WCAP-7665. F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse Electric Corporation, WCAP-7665, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, April 1971, (ADAMS Accession No: ML070780083), Appendix B, p. B-4.

71 S. Hagen, P. Hofmann, G. Schanz, L. Sepold, Interactions in Zircaloy/UO2 Fuel Rod Bundles with Inconel Spacers at Temperatures above 1200°C (Posttest Results of Severe Fuel Damage Experiments CORA-2 and CORA-3), Forschungszentrum Karlsruhe, KfK 4378, September 1990, p. 41.

72 NRC, Denial of Petition for Rulemaking (PRM-50-76), June 29, 2005, (ADAMS Accession No:

ML050250359), pp. 21-22.

25

analyses are not legitimate verifications of the adequacy of the Baker-Just and Cathcart-Pawel correlations for use in computer safety models.73 FLECHT Runs 2443 and 2544

59. There are also significant problems with Westinghouses examinations of the metallographic cross-sections that were taken from test rods from FLECHT Runs 2443 and 2544.

60. A Westinghouse report states that two of the FLECHT experimentsRuns 2443 and 2544 with Zircaloy test bundles had unintended internal gas pressure increases, at the middle sections of the bundles, which caused the Zircaloy cladding to balloon and move away from the heat source of the internally heated rods and from the location of the thermocouples.74 The actual temperatures of the Zircaloy cladding of the test bundles at the middle section were lower than the temperatures Westinghouse recorded. Therefore, the quantity of oxidation which occurred at the middle sections of the test bundles from FLECHT Runs 2443 and 2544, occurred at lower temperatures than Westinghouse claimed.

61. The thickness of each oxide layer would have been accurately measured; however, the examiners concluded that the thicknesses of the oxide layers from the middle sections of the test bundles from FLECHT runs 2443 and 2544 had been produced at higher temperatures than they were actually produced at. Hence, the metallurgical data was erroneously associated with cladding temperatures that were too high. Clearly, Westinghouses metallurgical data from FLECHT Runs 2443 and 2544 is not valid for performing a legitimate verification of the adequacy of the Baker-Just correlation for use in computer safety models.

73 Id.

74 F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse Electric Corporation, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, WCAP-7665, April 1971, (ADAMS Accession No: ML070780083), p. 3-95.

26

62. The NRCs subsequent analyseswhich used data from FLECHT Runs 2443 and 2544are also not legitimate verifications of the adequacy of the Baker-Just and Cathcart-Pawel correlations for use in computer safety models.75

63. (Interestingly, in Westinghouses comparison of eight metallurgical samples from run 2443, taken from two feet above and below the midplane location, all of the measured oxide thicknesses exceeded the predicted oxide thicknesses.76)

Problems with the Analysis of FLECHT Run 9573 Continue in the Post-Fukushima Era

64. 45 years after the Indian Point licensing hearings, the NRC does not seem concerned that industrys computer safety models still under-predict the extent of the zirconium-steam reaction. On November 17, 2009, Mark Leyse submitted a 10 C.F.R. § 2.802 petition for rulemaking, PRM-50-93,77 to the NRC that discusses the section of the FLECHT Run 9573 test bundle that incurred thermal runawaythe severe-damage zone.

65. As part of its technical analysis of PRM-50-93, the NRC did a computer simulation of what occurred in FLECHT Run 9573. They wanted to compare the results of their simulation to the data that Westinghouse reported on FLECHT Run 9573. However, there was a big problem with the NRCs simulation. They did not simulate the section of the test bundle that incurred 75 NRC, Denial of Petition for Rulemaking (PRM-50-76), June 29, 2005, (ADAMS Accession No:

ML050250359), pp. 21-22.

76 In all eight cases measured oxide thicknesses were less than 0.1 x 10-3 inches thick; however, all the predicted thicknesses were zero inches. See F. D. Kingsbury, J. F. Mellor, A. P. Suda, Westinghouse Electric Corporation, Appendix B, Materials Evaluation, of PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, p. B-9.

77 Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No. ML093290250).

27

thermal runawaythe severe-damage zone.78 (Or if they did simulate that section, they decided not to release their findings.)

66. By way of an analogy: what the NRC did would be like simulating a forest fire and omitting trees reduced to ash and only simulating those that had been singed. After doing such a bogus simulation one might try to argue that trees actually do not burn down in forest fires. The NRC basically did just that. They used the results of their simulation to argue that the Baker-Just correlation is adequate for use in computer safety models that simulate LOCAs.79

67. On January 31, 2013, Leyse gave a presentation to NRC Chairwoman Allison M. Macfarlane and the four NRC Commissioners. They invited Leyse to present his views on a panel addressing public participation in the NRCs rulemaking process.80 In his presentation, Leyse discussed the NRCs computer simulation of FLECHT Run 9573. He stated: You cannot do legitimate computer simulations of an experiment that incurred runaway oxidation by not actually modeling the section of the test bundle that incurred runaway oxidation. So, the staffssimulations were frankly a waste of money. Leyse offered to meet with the NRC staff members who were (and still are) reviewing PRM-50-93, to discuss it, try to sort things out, expedite things.81 78 NRC, Draft Interim Review of PRM-50-93/95 Issues Related to Conservatism of 2200 degrees F, Metal-Water Reaction Rate Correlations, and The Impression Left from [FLECHT] Run 9573 , October 16, 2012, (ADAMS Accession No. ML12265A277), pp. 7-9.

79 NRC, Draft Interim Review of PRM-50-93/95 Issues Related to Conservatism of 2200 degrees F, Metal-Water Reaction Rate Correlations, and The Impression Left from [FLECHT] Run 9573 , October 16, 2012, (ADAMS Accession No. ML12265A277), pp. 7-9.

80 NRC, Public Participation in NRC Regulatory Decision-Making, Transcript of Proceedings, January 31, 2013, (available at: http://www.nrc.gov/reading-rm/doc-collections/commission/tr/2013/20130131b.pdf ).

81 NRC, Public Participation in NRC Regulatory Decision-Making, Transcript of Proceedings, January 31, 2013, (available at: http://www.nrc.gov/reading-rm/doc-collections/commission/tr/2013/20130131b.pdf ),

pp. 55-56.

28

68. After everyone on the panel concluded their presentations, Chairwoman Macfarlane stated: Let me first note that I think Mr. Leyse demonstrated and has been and is continuing to be in the process of demonstrating that the public actually has a lot of valuable input. The public actually knows things that people at government agencies dont know and may not be aware of, and actually, the social science literature is ripe with this information as well, confirming this is true.82 Later on, Commissioner William Magwood assured Leyse that he and the other Commissioners would instruct their staff to follow up on his criticism of the NRCs computer simulation of FLECHT Run 9573.83

69. The NRC Commissioners seemed receptive to Leyses allegation that the computer simulation of FLECHT Run 9573 was inadequate. However, a couple of months after the meeting on public participation, the NRC staff released yet more of its technical analysis of PRM-50-93, including a statement that their simulation of FLECHT Run 9573 over-predicted the extent of the zirconium-steam reaction.84 The NRC staff simply reiterated their claim that the results of their simulation of FLECHT Run 9573 show that the Baker-Just correlation is adequate for use in computer safety models that simulate LOCAs.

70. In November 2015, after Leyse made a series of additional complaints, the NRC finally disclosed the results of a computer simulation of FLECHT Run 9573 that included the section 82 NRC, Public Participation in NRC Regulatory Decision-Making, Transcript of Proceedings, January 31, 2013, (available at: http://www.nrc.gov/reading-rm/doc-collections/commission/tr/2013/20130131b.pdf ),

pp. 65-66.

83 NRC, Public Participation in NRC Regulatory Decision-Making, Transcript of Proceedings, January 31, 2013, (available at: http://www.nrc.gov/reading-rm/doc-collections/commission/tr/2013/20130131b.pdf ), p.

83.

84 NRC, Draft Interim Review of PRM-50-93/95 Issues Related to Minimum Allowable Core Reflood Rate, March 8, 2013, (ADAMS Accession No. ML13067A261), p. 4.

29

of the test bundle that incurred thermal runawaythe severe-damage zone. And the simulation under-predicted temperatures Westinghouse had reported for that section.85 The NRCs Computer Simulation of FLECHT Run 9573 that Included the Section of the Test Bundle that Incurred Thermal Runawaythe Severe-Damage Zone

71. The FLECHT Run 9573 test bundle incurred thermal runaway around its seven foot elevation.

WCAP-7665 states: Post-test bundle inspection indicated a locally severe damage zone within approximately +/-8 inches of a Zircaloy grid at the 7 foot (ft) elevation. The heater rod failures were apparently caused by localized temperatures in excess of 2500°F. WCAP-7665 also states: During the test, heater element failures started at 18.2 seconds... At the time of the initial failures, midplane [at the 6 foot elevation] clad temperatures were in the range of 2200-2300°F. The only prior indication of excessive temperatures was provided by the 7 ft steam probe, which exceeded 2500°F at 16 seconds (2 seconds prior to start of heater element failure).86

72. The NRC conducted TRACE code computer simulations of FLECHT Run 9573 and found that TRACE under-predicted temperatures that were reported by Westinghouse at the 7 ft elevation of the test bundle. On November 24, 2015, Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, sent Leyse an e-mail regarding the NRCs TRACE computer simulation of FLECHT Run 9573. In his e-mail, Mr. Mohseni disclosed findings of 85 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

86 F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, WCAP-7665, April 1971, (ADAMS Accession No. ML070780083),

p. 3.97.

30

the completed simulation [for] the cladding and steam temperatures at the 7-ft elevation (at 18 seconds).87

73. According to Mr. Mohsenis e-mail, when the TRACE code used the Cathcart-Pawel and Baker-Just correlations, it predicted cladding temperatures of 1526 K (2287°F) and 1561 K (2350°F), respectively. And, when TRACE used the Cathcart-Pawel and Baker-Just correlations, it predicted steam temperatures of 1370 K (2006°F) and 1397 K (2055°F),

respectively. Those are predicted cladding and steam temperatures for the FLECHT Run 9573 test bundle at the 7-ft elevation, at 18 seconds.88

74. Westinghouse reported that at 18.2 seconds, heater rod failures occurred around the 7 foot elevation when cladding temperatures were in excess of 1644 K (2500°F). (Who knows how high the cladding temperatures actually were; they could have been hundreds of degrees Fahrenheit higher than 1644 K (2500°F).)

75. And Westinghouse reported that at 16.0 seconds, a steam probe at the 7 foot elevation recorded steam temperatures that exceeded 1644 K (2500°F). And a Westinghouse memorandum stated that after 12 seconds, the steam-probe thermocouple recorded an extremely rapid rate of temperature rise (over 300°F/sec).89 (Who knows how high the steam temperatures actually were at 18 seconds; they were likely hundreds of degrees Fahrenheit higher than 1644 K (2500°F).)

87 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

88 Id.

89 Robert H. Leyse, Westinghouse, Nuclear Energy Systems, Test Engineering, Memorandum RD-TE 616, FLECHT Monthly Report, December 14, 1970. This Memorandum is available at Appendix I of PRM-50-93. See Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No: ML093290250),

Appendix I.

31

76. Taking the time difference of 0.2 seconds (between 18 and 18.2 seconds) into account, when TRACE used the Cathcart-Pawel and Baker-Just correlations, it predicted cladding temperatures that were at least 200°F and 140°F lower, respectively, than the temperatures Westinghouse reported. That is non-conservative.

77. When TRACE used the Cathcart-Pawel and Baker-Just correlations, at 18 seconds it predicted steam temperatures that were about 500°F and 450°F lower, respectively, than the temperatures Westinghouse measured at 16 seconds. Westinghouse also reported that after 12 seconds, steam temperatures were increasing at a rate greater than 300°F/sec. So steam temperatures were even greater at 18 seconds than they were at 16 seconds. Hence, the TRACE predictions for steam temperatures are non-conservative.

78. The FLECHT Run 9573 results indicate that the currently used zirconium-steam reaction correlations, such as the Cathcart-Pawel and Baker-Just correlations, are inadequate for use in computer safety models like the NRCs TRACE code and AREVAs EXEM BWR-2000 Evaluation Model.

79. This is powerful evidence that the Baker-Just correlation is inadequate for use in computer safety models that simulate LOCAs. This also means that 10 C.F.R. 50 Appendix K, I.A.5 is non-conservative. In regard to the zirconium-steam reaction that would occur in the event of a LOCA, 10 C.F.R. 50 Appendix K, I.A.5 requires that [t]he rate of energy release, hydrogen generation, and cladding oxidation from the metal-water reaction shall be calculated using the Baker-Just [correlation].90 90 NRC, Appendix K to Part 50ECCS Evaluation Models, (This information is available at:

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appk.html : last visited on 09/02/16).

32

IV. Evidence Supporting Contention The Experiments Behind the Baker-Just Correlation

80. The Baker-Just correlationused in Appendix K to Part 50 ECCS evaluation calculations dates back to 1962.91 In order to develop the Baker-Just correlation, Louis Baker, Jr. and Louis C. Just partly relied on data from their own experiments. Their experiments were conducted at the melting temperature of zirconium, (in which fine [zirconium] wires were directly heated in water and the hydrogen evolution from the resulting molten droplets was measured to calculate the reaction rate).92 The melting temperature of zirconium approximately 3362°F (1850°C)is far greater than the temperature range of design-basis accidents, which have a maximum temperature of 2200°F (1204.4°C).

81. The Baker-Just correlation is primarily based on data from Alexis Lemmon and W. A.

Bostroms experiments,93 which were conducted in the 1950s.94 Bostroms experiments were conducted above the temperature range of design-basis accidents. Lemmon and Bostroms experiments were conducted with tiny inductively heated Zircaloy-2 specimens.95 There are 91 Louis Baker, Jr. and Louis C. Just, Studies of Metal-Water Reactions at High Temperatures: III.

Experimental and Theoretical Studies of the Zirconium-Water Reaction, ANL-6548, May 1962, (ADAMS Accession No: ML050550198).

92 G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003, p. 2.

93 G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003, p. 2.

94 W. A. Bostrom, The High Temperature Oxidation of Ziracloy in Water, WAPD-104, March 1954, (ADAMS Accession No: ML100900446) and Alexis W. Lemmon, Studies Relating to the Reaction Between Zirconium and Water at High Temperatures, Battelle Memorial Institute, BMI-1154, January 1957, (ADAMS Accession No: ML100570218).

95 V. F. Urbanic and T. R. Heidrick, High-Temperature Oxidation of Zircaloy-2 and Zircaloy-4 in Steam, Journal of Nuclear Materials 75, 1978, p. 252.

33

radiative heat losses in experiments conducted with inductive heating, which affect a specimens oxidation kinetics.96

82. Regarding the Bostrom and Lemmon experiments that were used to help develop the Baker-Just Correlation, a 1978 Journal of Nuclear Materials paper states:

Bostrom inductively heated specimens of Zircaloy-2 in water (with a steam bubble enveloping the specimen) under isothermal conditions and determined Kp in the temperature range 1300-1860°C by the hydrogen evolution method. Lemmon measured the rates of reaction between Zircaloy-2 and steam in the temperature range 1000-1700°C by inductively heating specimens in steam at 50 psia [pounds per square inch absolute] and measuring the rate of hydrogen evolution.97

83. Describing Lemmons experiments in more detail, Lemmons own 1957 report, Studies Relating to the Reaction Between Zirconium and Water at High Temperatures states:

The reaction between solid Zircaloy 2 and steam at 50 psia was measured over the temperature range 1000 to 1690°C. ... The Zircaloy 2 specimens were heated by electrical induction and reacted with flowing steam at a pressure of 50 psia. ... The [Zircaloy 2] specimen was supported on a thermocouple protection tube and enclosed inside a Vycor tube; it was inductively heated to the reaction temperature by power applied through the induction coil.98

84. Lemmons specimens were Zircaloy-2 cylinders that were 2.0 inches long and 0.5 inches in diameter.99 96 G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003, pp. 4-5.

97 V. F. Urbanic and T. R. Heidrick, High-Temperature Oxidation of Zircaloy-2 and Zircaloy-4 in Steam, Journal of Nuclear Materials 75, 1978, p. 252.

98 Alexis W. Lemmon, Studies Relating to the Reaction Between Zirconium and Water at High Temperatures, Battelle Memorial Institute, BMI-1154, January 1957, (ADAMS Accession No:

ML100570218), pp. C-1, C-2, C-3.

99 Id., p. C-4.

34

85. Regarding radiative heat losses experienced in Lemmons experiments, Lemmons own 1957 report states:

The passage of steam through the reactor [unit] greatly increased the heat losses from the samples; and a large increase in power to the induction coil was required. Sample temperatures dropped as much as 100 or 200°C below the desired temperature before the power adjustment was effective.

This sometimes took as long as [five] min.100

86. Regarding how radiative heat losses in inductive specimen heating experiments affect oxidation kinetics, a 2003 paper by G. Schanz states:

[Ocken] stated that [the] advantage [of experiments with inductive (Urbanic and Heidrick) and direct electrical heating (Biederman, et al.) of a specimen in a cool environment101] would be the temperature gradient from heated specimen to cool surrounding[s], leading to temperature gradients in the cladding wall in the same sense as in a reactor. In total disagreement with the argument of Ocken, the author of this paper stresses the advantage of a constant cladding wall temperature and thus of a better defined specimen temperature, as provided in furnace experiments! ...

This argument was already used by Sawatzky, et al., who performed similar studies with inductive specimen heating as Urbanic and Heidrick.

Sawatzky reached an important improvement of the specimen temperature homogeneity by only optimizing the geometry of the specimen and registered considerably increased reaction rates102 [emphasis added].

87. Radiative heat losses in an experiment conducted with inductive heating cause a specimens zirconium-steam reaction rates to decrease below what they would be if there were no radiative heat losses. The very experiments that the Baker-Just correlation is primarily based on would have had radiative heat losses that decreased zirconium-steam reaction rates. Lemmon and Bostroms experiments certainly did not replicate the oxidation kinetics that would occur in a nuclear reactors core, in the event of a LOCA. Yet the Baker-Just correlationrequired by Appendix K to Part 50 I.A.5is almost entirely based on the results of their experiments. This 100 Id., p. C-7.

101 G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003, pp. 4-5.

102 Id.

35

fact alone is evidence that the Baker-Just correlation is likely inadequate for use in computer safety models like AREVAs EXEM BWR-2000 Evaluation Model. Discussing larger scale experiments, this declaration presents far more conclusive evidence that the Baker-Just correlation is indeed inadequate for use in computer safety models.

Experiments in which Zirconium-Steam Reaction Rates Occurred that Exceed the Rates Predicted by Computer Safety Models

88. In this declaration, I provide information about experimental results that indicate the currently used zirconium-steam reaction correlations, such as the Baker-Just correlation, are inadequate for use in computer safety models like the NRCs TRACE code and AREVAs EXEM BWR-2000 Evaluation Model. When using the currently used zirconium-steam reaction correlations, computer safety models under-predict the zirconium-steam reaction rates that occurred in the experiments discussed in this declaration. Computer safety models are supposed to over-predict reaction rates in order to ensure an adequate margin of safety. The experimental results discussed in this declaration are evidence that the NRC and nuclear industrys computer safety models under-predict the zirconium-steam reaction rates that would occur in the event of a design-basis accident (LOCA), meaning that the amendment request for the EPUs for BFN Units 1, 2, and 3 should be denied.

Oxidation Models Are Unable to Predict the Fuel-Cladding Temperature Escalation that Commenced at Low Temperatures in the PHEBUS B9R-2 Test

89. The PHEBUS B9R test was conducted in a light water reactoras part of the PHEBUS severe fuel damage programwith an assembly of 21 uranium dioxide (UO2) fuel rods. The B9R test 36

was conducted in two parts: the B9R-1 test and the B9R-2 test.103 A 1996 European Commission report states that the B9R-2 test had an unexpected fuel-cladding temperature escalation in the mid-bundle region (see Figure 3 below); the highest temperature escalation rates were from 20°C/sec (36°F/sec) to 30°C/sec (54/°C/sec).104

90. Discussing PHEBUS B9R-2, the 1996 European Commission report states:

The B9R-2 test (second part of B9R) illustrates the oxidation in different cladding conditions representative of a pre-oxidized and fractured state.

This state results from a first oxidation phase (first part name B9R-1, of the B9R test) terminated by a rapid cooling-down phase. During B9R-2, an unexpected strong escalation of the oxidation of the remaining Zr occurred when the bundle flow injection was switched from helium to steam while the maximum clad temperature was equal to 1300 K [1027°C (1880°F)]. The current oxidation model was not able to predict the strong heat-up rate observed even taking into account the measured large clad deformation and the double-sided oxidation (final state of the cladding from macro-photographs).

No mechanistic model is currently available to account for enhanced oxidation of pre-oxidized and cracked cladding105 [emphasis added].

103 G. Hache, R. Gonzalez, B. Adroguer, Institute for Protection and Nuclear Safety, Status of ICARE Code Development and Assessment, in NRC Proceedings of the Twentieth Water Reactor Safety Information Meeting, NUREG/CP-0126, Vol. 2, 1992, (ADAMS Accession No: ML042230126), p. 311.

104 T.J. Haste et al., In-Vessel Core Degradation in LWR Severe Accidents, European Commission, Report EUR 16695 EN, 1996, p. 33.

105 Id., p. 126.

37

Figure 3. Local Cladding Temperature vs. Time in the PHEBUS B9R-2 Test106

91. Today, in 2016, oxidation models still cannot accurately predict the local fuel-cladding temperature escalation that commenced in PHEBUS B9R-2 in steam when local fuel-cladding temperatures were 1027°C (1880°F). The PHEBUS B9R-2 results indicate that the currently used zirconium-steam reaction correlations, such as the Baker-Just correlation, are inadequate for use in computer safety models like AREVAs EXEM BWR-2000 Evaluation Model.

92. The fact that PHEBUS B9R-2 was conducted with a pre-oxidized test bundle makes its results particularly applicable to high burnup fuel. High burnup fuel rods would also be pre-oxidized: when high burnup (and other) fuel rods are discharged from the reactor core and loaded into the spent fuel pool, the fuel cladding can have local zirconium dioxide (ZrO2) 106 G. Hache, R. Gonzalez, B. Adroguer, Institute for Protection and Nuclear Safety, Status of ICARE Code Development and Assessment, in NRC Proceedings of the Twentieth Water Reactor Safety Information Meeting, NUREG/CP-0126, Vol. 2, 1992, (ADAMS Accession No: ML042230126), p. 312.

38

oxide layers that are up to 100 m thick (or greater); there can also be local crud layers on top of the oxide layers, which can sometimes also be up to 100 m thick.

Low Temperature Oxidation Rates Are Under-Predicted for FLECHT Run 9573

93. Westinghouses PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report (hereinafter: WCAP-7665) states that, [t]he objective of the PWR FLECHTtest program was to obtain experimental reflooding heat transfer data under simulated loss-of-coolant accident conditions for use in evaluating the heat transfer capabilities of PWR emergency core cooling systems.107 The FLECHT tests were conducted with bundles of heater rods sheathed in zirconium alloy (Zircaloy) cladding. Runaway oxidation was not expected to occur in any of the tests; however, the FLECHT Run 9573 test bundle incurred runaway oxidation (see Figure 4 below).

107 F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse Electric Corporation, WCAP-7665, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, April 1971, (ADAMS Accession No: ML070780083), p. 1.1.

39

Figure 4. Section of the FLECHT Run 9573 Test Bundle that Incurred Runaway Oxidation

94. The FLECHT Run 9573 test bundle incurred runaway oxidation around its seven foot elevation.

WCAP-7665 states: Post-test bundle inspection indicated a locally severe damage zone within approximately +/-8 inches of a Zircaloy grid at the 7 ft elevation. The heater rod failures were apparently caused by localized temperatures in excess of 2500°F. WCAP-7665 also states:

During the test, heater element failures started at 18.2 seconds... At the time of the initial failures, midplane [at the 6 foot elevation] clad temperatures were in the range of 2200-2300°F.

The only prior indication of excessive temperatures was provided by the 7 ft steam probe, which exceeded 2500°F at 16 seconds (2 seconds prior to start of heater element failure).108

95. The NRC conducted TRACE code computer simulations of FLECHT Run 9573 and found that TRACE under-predicted temperatures that were reported by Westinghouse at the 7 ft elevation of the test bundle. On November 24, 2015, Aby Mohseni, Deputy Director of the NRCs 108 Id., p. 3.97.

40

Division of Policy and Rulemaking, sent Leyse an e-mail regarding the NRCs TRACE computer simulation of FLECHT Run 9573. In his e-mail, Mr. Mohseni disclosed findings of the completed simulation [for] the cladding and steam temperatures at the 7-ft elevation (at 18 seconds).109

96. TRACE under-predicted cladding and steam temperatures at the 7-foot elevation of the FLECHT Run 9573 test bundle. TRACE is supposed to over-predict temperatures in order to ensure an adequate margin of safety. The Baker-Just and Cathcart-Pawel zirconium-steam reaction correlations were used for the TRACE simulations. The TRACE simulations need to be considered as evidence that the NRC and nuclear industrys computer safety models under-predict the zirconium-steam reaction rates that would occur in the event of a design-basis accident (LOCA).

FLECHT Run 9573a Comparison between Computer Safety Model Predictions and the Results Westinghouse Reported

97. According to Mr. Mohsenis e-mail, when the TRACE code used the Cathcart-Pawel and Baker-Just correlations, it predicted cladding temperatures of 1526 K (2287°F) and 1561 K (2350°F), respectively. And, when TRACE used the Cathcart-Pawel and Baker-Just correlations, it predicted steam temperatures of 1370 K (2006°F) and 1397 K (2055°F),

respectively. Those are predicted cladding and steam temperatures for the FLECHT Run 9573 test bundle at the 7-ft elevation, at 18 seconds.110 109 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

110 Id.

41

98. Westinghouse reported that at 18.2 seconds, heater rod failures occurred around the 7 foot elevation when cladding temperatures were in excess of 1644 K (2500°F). (Who knows how high the cladding temperatures actually were; they could have been hundreds of degrees Fahrenheit higher than 1644 K (2500°F).)

99. And Westinghouse reported that at 16.0 seconds, a steam probe at the 7 foot elevation recorded steam temperatures that exceeded 1644 K (2500°F). And a Westinghouse memorandum stated that after 12 seconds, the steam-probe thermocouple recorded an extremely rapid rate of temperature rise (over 300°F/sec).111 (Who knows how high the steam temperatures actually were at 18 seconds; they were likely hundreds of degrees Fahrenheit higher than 1644 K (2500°F).)

100. Taking the time difference of 0.2 seconds (between 18 and 18.2 seconds) into account, when TRACE used the Cathcart-Pawel and Baker-Just correlations, it predicted cladding temperatures that were at least 200°F and 140°F lower, respectively, than the temperatures Westinghouse reported. That is non-conservative.

101. When TRACE used the Cathcart-Pawel and Baker-Just correlations, at 18 seconds it predicted steam temperatures that were about 500°F and 450°F lower, respectively, than the temperatures Westinghouse measured at 16 seconds. Westinghouse also reported that after 12 seconds, steam temperatures were increasing at a rate greater than 300°F/sec. So steam temperatures were even greater at 18 seconds than they were at 16 seconds. Hence, the TRACE predictions for steam temperatures are non-conservative.

111 Robert H. Leyse, Westinghouse, Nuclear Energy Systems, Test Engineering, Memorandum RD-TE 616, FLECHT Monthly Report, December 14, 1970. This Memorandum is available at Appendix I of PRM-50-93. See Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No: ML093290250),

Appendix I.

42

102. The FLECHT Run 9573 results indicate that the currently used zirconium-steam reaction correlations, such as the Cathcart-Pawel and Baker-Just correlations, are inadequate for use in computer safety models like the NRCs TRACE code and AREVAs EXEM BWR-2000 Evaluation Model.

Low Temperature Oxidation Rates Are Under-Predicted for the CORA-16 Experiment 103. When Oak Ridge National Laboratory (ORNL) investigators compared the results of the CORA-16 experimenta BWR severe fuel damage test, simulating a meltdown, conducted with a multi-rod zirconium alloy bundlewith the predictions of computer safety models, they found that the zirconium-steam reaction rates that occurred in the experiment were under-predicted. The investigators concluded that the application of the available Zircaloy oxidation kinetics models [zirconium-steam reaction correlations] causes the low-temperature

[1652-2192°F] oxidation to be underpredicted.112 104. It has been postulated that cladding strainballooningwas a factor in increasing the zirconium-steam reaction rates that occurred in CORA-16.113 However, it is unsubstantiated that cladding strain actually increased reaction rates.

105. To help explain how cladding strain could have been a factor in increasing the zirconium-steam reaction rates that occurred in CORA-16, the NRC has pointed out that an NRC report, NUREG/CR-4412,114 explain[s] that under certain conditions ballooning and deformation of 112 L. J. Ott, Oak Ridge National Laboratory, Report of Foreign Travel of L. J. Ott, Engineering Analysis Section, Engineering Technology Division, ORNL/FTR-3780, October 16, 1990, p. 3.

113 L. J. Ott, W. I. van Rij, In-Vessel PhenomenaCORA: BWR Core Melt Progression Phenomena Program, Oak Ridge National Laboratory, CONF-9105173-3-Extd.Abst., Presented at Cooperative Severe Accident Research Program, Semiannual Review Meeting, Bethesda, Maryland, May 6-10, 1991.

114 R. E. Williford, An Assessment of Safety Margins in Zircaloy Oxidation and Embrittlement Criteria for ECCS Acceptance, NUREG/CR-4412, April 1986, (ADAMS Accession No: ML083400371).

43

the cladding can increase the available surface area for oxidation, thus enhancing the apparent oxidation rate115 [emphasis not added].

106. Regarding this phenomenon, NUREG/CR-4412 states:

Depressurization of the primary coolant during a LB LOCA or [severe accident] will permit [fuel] cladding deformation (ballooning and possibly rupture) to occur because the fuel rod internal pressure may be greater than the external (coolant) pressure. In this case, oxidation and deformation can occur simultaneously. This in turn may result in an apparent enhancement of oxidation rates because: 1) ballooning increases the surface area of the cladding and permits more oxide to form per unit volume of Zircaloy and 2) the deformation may crack the oxide and provide increased accessibility of the oxygen to the metal. However deformation generally occurs before oxidation rates become significant; i.e., below 1000°C [1832°F]. Consequently, the lesser importance of this phenomenon has resulted in a relatively sparse database.116 107. NUREG/CR-4412 states that there is a relatively sparse database on the phenomenon of cladding strain enhancing zirconium-steam reaction rates.117 NUREG/CR-4412 also explains that it is possible to make a very crude estimate of the expected average enhancement of oxidation kinetics by deformation;118 the report provides a graph of the rather sparse119 data.

The graph indicates that the general trend is for cladding strain enhancements of zirconium-steam reaction rates to decrease as cladding temperatures increase.120 108. NUREG/CR-4412 has a brief description of the rather sparse data; in one case, two investigators (Furuta and Kawasaki), who heated specimens up to temperatures between 1292°F 115 NRC, Draft Interim Review of PRM-50-93/95 Issues Related to the CORA Tests, August 23, 2011, (ADAMS Accession No: ML112211930), p. 3.

116 R. E. Williford, An Assessment of Safety Margins in Zircaloy Oxidation and Embrittlement Criteria for ECCS Acceptance, NUREG/CR-4412, p. 27.

117 Id., pp. 27, 30.

118 Id., p. 30.

119 Id.

120 Id., p. 29.

44

and 1832°F, reported that [v]ery small enhancements [of reaction rates] occurred at about

[eight percent] strain at [1832°F].121 109. In fact, NUREG/CR-4412 states that only one pair of investigators (Bradhurst and Heuer) conducted tests that encompassed the temperature range1652°F to 2192°Fin which zirconium-steam reaction rates were under-predicted for CORA-16. Bradhurst and Heuer reported that [m]aximum enhancements occurred at slower strain rates. However, the overall weight gain or average oxide thickness in [the Zircaloy-2 specimens] was only minimally increased because of the localization effects of cracks in the oxide layer. 122 A second report states that Bradhurst and Heuerfound no direct influence [from cladding strain] on Zircaloy-2 oxidation outside of oxide cracks.123 (In CORA-16, in the temperature range from 1652°F to 2192°F, cladding strain would have occurred over a brief period of time, tens of seconds, because cladding temperatures were increasing rapidly.)

110. Clearly, it is unsubstantiated that the estimated cladding strain accurately accounts for why reaction rates for CORA-16 were under-predicted in the temperature range from 1652°F to 2192°F. First, there is a relatively sparse database on how cladding strain enhances reaction rates. Second, the little data that is available indicates that cladding strain may only slightly enhance reaction rates at cladding temperatures of 1832°F and greater.124 111. Furthermore, ORNL papers on the BWR CORA experiments do not report that any experiments were conducted in order to confirm if in fact cladding strain actually increased 121 Id., p. 30.

122 Id.

123 F. J. Erbacher, S. Leistikow, A Review of Zircaloy Fuel Cladding Behavior in a Loss-of-Coolant Accident, Kernforschungszentrum Karlsruhe, KfK 3973, September 1985, p. 6.

124 R. E. Williford, An Assessment of Safety Margins in Zircaloy Oxidation and Embrittlement Criteria for ECCS Acceptance, NUREG/CR-4412, p. 30.

45

zirconium-steam reaction rates and accounted for why reaction rates were under-predicted in the 1652°F to 2192°F temperature range for CORA-16.

112. There is also one phenomenon the NRC did not consider in its 2011 analysis of CORA-16:

[t]he swelling of the [fuel] claddingalters [the] pellet-to-cladding gap in a manner that provides less efficient energy transport from the fuel to the cladding,125 which would cause the local cladding temperature heatup rate to decrease as the cladding ballooned, moving away from the internal heat source of the fuel. The CORA experiments were internally electrically heated (with annular uranium dioxide pellets to replicate uranium dioxide fuel pellets), so in CORA-16, the ballooning of the cladding would have had a mitigating factor on the local cladding temperature heatup rate, which, in turn, would have had a mitigating factor on zirconium-steam reaction rates.

113. CORA-16 is an example of an experiment that had zirconium-steam reaction rates that were under-predicted in the low temperature range from 1652°F to 2192°F by computer safety models. The CORA-16 results indicate that the currently used zirconium-steam reaction correlations, such as the Baker-Just correlation, are inadequate for use in computer safety models like AREVAs EXEM BWR-2000 Evaluation Model.

125 Winston & Strawn LLP, Duke Energy Corporation, Catawba Nuclear Station Units 1 and 2, Enclosure, Testimony of Robert C. Harvey and Bert M. Dunn on Behalf of Duke Energy Corporation, MOX Fuel Lead Assembly Program, MOX Fuel Characteristics and Behavior, and Design Basis Accident (LOCA)

Analysis, July 1, 2004, (ADAMS Accession No: ML041950059), p. 43.

46

Computer Safety Models Fail to Accurately Predict the Onset of the Fuel-Cladding Temperature Escalation that Commenced in the LOFT LP-FP-2 Experiment (in the Design-Basis Accident Temperature Range) 114. In the LOFT LP-FP-2 experiment, there was a fuel-cladding temperature escalation that commenced when fuel-cladding temperatures were lower than the 2200°F PCT limit.

115. Computer safety models have failed to accurately predict the onset of the fuel-cladding temperature escalation that occurred in the LOFT LP-FP-2 experiment.

116. Regarding a fairly recent computer safety model (ASTEC V1.3 code) simulation of the LOFT LP-FP-2 experiment, a 2010 paper, Recent Advances in ASTEC Validation on Circuit Thermal-Hydraulic and Core Degradation states:

The onset of core uncovery and heat-up was very well reproduced by ASTEC (fig. 17), but the onset of temperature escalation in the upper part of the CFM [center fuel module] was delayed.126 117. In Recent Advances in ASTEC Validation on Circuit Thermal-Hydraulic and Core Degradation, in figure 17 (see Figure 3 of these comments), the graph of the cladding-temperature values in the ASTEC V1.3 simulation of the LOFT LP-FP-2 experiment depicts that the onset of the temperature escalation (at the 1.067 m elevation) commenced at a temperature greater than 1700 K (2600°F); figure 17 (see Figure 3 of this paragraph) also shows that in the experiment the actual onset of the temperature escalation (at the 1.067 m elevation) commenced at a temperature well below 1500 K (2240°F)definitely below 2200°F.127 Hence, the difference between the calculated and actual experimental value for the onset of the 126 G. Bandini et al., Recent Advances in ASTEC Validation on Circuit Thermal-Hydraulic and Core Degradation, Progress in Nuclear Energy, 52, 2010, p. 155.

127 Id.

47

temperature escalation (at the 1.067 m elevation) is greater than 200 K (360°F)a significant difference.

Figure 3. Onset of the Temperature Escalation that Occurred in the LOFT LP-FP-2 Experiment (at the 1.067 m Elevation)128 128 Id.

48

An Experiment for which the Quantity of Hydrogen Produced by the Zirconium-Steam Reaction at about 1800°F Is Under-Predicted by Computer Safety Models: The FRF-1 Experiment 118. The First Transient Experiment of a Zircaloy Fuel Rod Cluster (FRF-1) experiment conducted in the Transient Reactor Test Facility (TREAT) facilitywas not a large-scale experiment yet UCS and the authors of a report on the FRF-1 experiment129 claimed that, as of 1971, it simulated the most realistic loss-of-coolant accident conditions of any experiment to date.130 119. Data from the FRF-1 experiment indicates that computer safety models under predict the quantity of hydrogen produced by the Zircaloy-steam reaction. In the experiment, at fuel rod temperatures of about 1800°F, the Zircaloy-steam reaction generated 1.2 +/- 0.6 liters of hydrogen. In the Indian Point Unit 2 (IP-2) licensing hearing, Westinghouse, which had performed experimental simulations of loss-of-coolant accidents, and conducted computer simulations of such accidents, testified that their computer safety models predicted that there would be no zirconium-steam reaction at 1800°Fthat no hydrogen would be produced in a loss-of-coolant accident if local temperatures of the fuel rods were to reach 1800°F.131 120. In the IP-2 licensing hearing, Dr. Jack Roll of Westinghouse contended that data from the FRF-1 experiment was not reliable, because the measurement of the extent of [zirconium-steam] reaction was in fact by an inferred route, and there were no direct measurements taken, 129 R. A. Lorenz, D. O. Hobson, G. W. Parker, Final Report on the First Fuel Rod Failure Transient Test of a Zircaloy-Clad Fuel Rod Cluster in TREAT, ORNL-4635, March 1971.

130 Henry W. Kendall, A Distant Light: Scientists and Public Policy, Springer-Verlag, New York, 2000, p.

43.

131 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 1, 1971, (ADAMS Accession No. ML100350644), pp. 2152-2153.

49

that [t]here was a large uncertainty in the measurement of total hydrogen evolution during the experiment, and that there was an uncertainty in the temperatures of the fuel [rods] during the experiment.132 Westinghouse concluded that it is not possible to know if the data from the FRF-1 experiment actually demonstrated that the extent of the zirconium-steam reaction was higher (or much higher) than would be predicted by computer safety models.

121. Unfortunately, there was not a means to confirm if Westinghouses claims were correct or not, because the Atomic Energy Commission decided to discontinue funding for the TREAT facility loss-of-coolant accident experimental program.133 The FRF-1 experiment could not be replicated; its results could not be confirmed.

V. Contention 122. Based on BEST/MATRRs vested interest in the safe operation of BFN, BEST/MATRR members are personally affected and aggrieved by the EPUs that are proposed for all three of BFNs General Electric (GE) Mark I boiling water reactors (BWR). The defective, antiquated BWR Mark I design performed poorly in the Fukushima Daiichi accident. In the accident, three BWR Mark I reactors melted down, generating hundreds of kilograms of explosive hydrogen gas. Hydrogen then detonated at different times, destroying three reactor buildings, which released large quantities of harmful radioactive material into the environment.

123. BEST/MATRR believes that the amendment request for the EPUs for BFN Units 1, 2, and 3 must be denied. The proposed EPUs would increase BFNs current licensed steady-state 132 Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.: Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642), pp. 2297-2299.

133 W. B. Cottrell, ORNL Nuclear Safety Research and Development Program Bimonthly Report for March-April 1971, ORNL-TM-3411, July 1971, p. x.

50

reactor core power level for each unit from 3,458 megawatt thermal (MWt) to 3,952 MWt, constituting a thermal power level increase of approximately 14.3 percent for all three units.

The proposed EPUs would increase BFNs original licensed thermal power level of 3,293 MWt for each unit by approximately 20 percent for all three units.134 124. Petitioners requests for leave to intervene and a hearing are supported by this declaration.

BEST/MATRR alleges that non-conservative computer safety model analyses were performed in order to justify the EPUs for BFN Units 1, 2, and 3. As explained in this declaration, experimental data, along with appropriate citations, indicates that the EPU analyses under-predict the rates of the chemical reaction between zirconium and steam that would occur in the event of a LOCA. This means that the analyses under-predict the rates in which energy (heat) is released, hydrogen generated, and zirconium fuel-cladding oxidized by the zirconium-steam reaction.

125. AREVA has stated that its EXEM BWR-2000 Evaluation Models LOCA calculations for the EPUs for BFN Units 1, 2, and 3 were performed in conformance with 10 CFR 50 Appendix K requirements and satisfy the event acceptance criteria identified in 10 CFR 50.46.135 In regard to the zirconium-steam reaction that would occur in the event of a LOCA, 10 C.F.R. 50 Appendix K, I.A.5 requires that [t]he rate of energy release, hydrogen 134 NRC, Applications and Amendments to Facility Operating Licenses and Combined Licenses Involving Proposed No Significant Hazards Considerations and Containing Sensitive Unclassified Non-Safeguards Information and Order Imposing Procedures for Access to Sensitive Unclassified Non-Safeguards Information, NRC-2016-0118, Federal Register, Vol. 81, No. 128, July 5, 2016, p. 43666.

135 AREVA, Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU), ANP-3377NP, Revision 3, Attachment 11 Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No:

ML15282A184), p. 1.1.

51

generation, and cladding oxidation from the metal-water reaction shall be calculated using the Baker-Just [correlation].136 126. As discussed above, in Section IV (of BEST/MATRRs hearing request and petition to intervene regarding the LAR for the EPUs for BFN Units 1, 2, and 3), 10 C.F.R. 50 Appendix K, I.A.5 is non-conservative.

127. BEST/MATRRs members must not be subjected to the LAR for the EPUs for BFN Units 1, 2, and 3, because the proposed LAR is justified by non-conservative Appendix K computer safety model evaluations.

128. Contention: The EPUs for BFN Units 1, 2, and 3 must not be granted because the EXEM BWR-2000 Evaluation Models LOCA calculations for qualifying the EPUs for BFN Units 1, 2, and 3 are scientifically indefensible.

129. Contention: TVA has not scientifically demonstrated that at higher power levels (3,952 MWt) that in the event of a LOCA, at any of the BFN units, the PCT would not exceed the 10 C.F.R. § 50.46(b)(1) PCT limit of 2200°F.137 130. Contention: The health and safety of BEST/MATRRs members, as well as that of the general public, must not be threatened by scientifically indefensible EPUs for BFN Units 1, 2, and 3.

131. The health and safety of BEST/MATRRs members must not be threatened because the requirements of 10 C.F.R. 50 Appendix K, I.A.5 were defended by an industry professional, Dr.

136 NRC, Appendix K to Part 50ECCS Evaluation Models, (This information is available at:

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appk.html : last visited on 09/02/16).

137 NRC, § 50.46 Acceptance criteria for emergency core cooling systems for light-water nuclear power reactors, (This information is available at: http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0046.html : last visited on 09/04/16).

52

John Bernard Roll, a manager in Westinghouses Nuclear Fuel Division, who made false statements when he was under oath in the Indian Point Unit 2 licensing hearings.

132. The health and safety of BEST/MATRRs members must not be threatened because FLECHT program data was cherry-picked. In order to defend the Baker-Just correlation, Dr.

Roll discussed the cherry-picked FLECHT program data when he was under oath in the Indian Point Unit 2 licensing hearings. Besides the fact that FLECHT program data was cherry-picked, there were problems with the metallurgical data from the FLECHT program, as explained in this declaration.

133. The health and safety of BEST/MATRRs members must not be threatened because the NRC is considering an amendment request for EPUs for BFN Units 1, 2, and 3, which is dependent on Appendix K LOCA analyses, after the NRC disclosed that a computer simulation of FLECHT Run 9573, including the section of the test bundle that incurred thermal runaway, under-predicted temperatures Westinghouse had reported for that section.138 134. The health and safety of BEST/MATRRs members must not be threatened after the NRC has disclosed powerful evidence that the Baker-Just correlation is inadequate for use in computer safety models that simulate LOCAs, which means that 10 C.F.R. 50 Appendix K, I.A.5 is non-conservative.

135. The LAR for the EPUs for BFN Units 1, 2, and 3 must be denied.

136. AREVAs analyses (conducted to help justify the EPUs for BFN Units 1, 2, and 3) also under-predict the PCTs that would occur in the event of a LOCA for ATRIUM 10XM fuel and ATRIUM 10 fuel, respectively. If the EPUs for BFN Units 1, 2, and 3 were granted and power 138 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

53

levels of the BFN reactors were set too high, in the event of a LOCA at one of the BFN units, the PCT would exceed the 10 C.F.R. § 50.46(b)(1) PCT limit of 2200°F.139 And if the PCT were to exceed the 2200°F limit, the LOCA would (by definition) become a beyond design-basis accident. If one of the Browns Ferry reactors were to melt down, hundreds of kilograms of explosive hydrogen gas would be generated. It is likely that the hydrogen would then explode and destroy a reactor building, releasing large quantities of harmful radioactive material into the environment, as occurred in the Fukushima Daiichi accident.

137. It is unacceptable to subject BEST/MATRRs members to the dangers of granting EPUs for BFN Units 1, 2, and 3. AREVAs Appendix K LOCA calculations are supported by false statements that were made by a manager in Westinghouses Nuclear Fuel Division, Dr. Roll, when he was under oath in the Indian Point Unit 2 licensing hearings. AREVAs Appendix K LOCA calculations are supported by the cherry-picked FLECHT program data that Dr. Roll discussed when he was under oath in the Indian Point Unit 2 licensing hearings in order to defend the Baker-Just correlation.

138. It was over four decades ago that Dr. Roll, a manager in Westinghouses Nuclear Fuel Division, made false statements when he was under oath in the Indian Point Unit 2 licensing hearings. That was in an Atomic Energy Commission, the NRCs predecessor, licensing hearing. In a contemporary licensing hearing, if Dr. Roll were to make false statements and not disclose important experimental data when he was under oath, he would be in violation of 10 C.F.R. § 52.4, Deliberate misconduct.140 139 NRC, § 50.46 Acceptance criteria for emergency core cooling systems for light-water nuclear power reactors, (This information is available at: http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0046.html : last visited on 09/04/16).

140 10 C.F.R. § 52.4, Deliberate misconduct, (This information is available at: http://www.nrc.gov/reading-rm/doc-collections/cfr/part052/part052-0004.html: last visited on 09/05/16).

54

139. 10 C.F.R. § 52.4(b) states: Deliberate misconduct means an intentional act or omission that a person or entity knows: (i) Would cause a licensee or an applicant for a license, standard design certification, or standard design approval to be in violation of any rule, regulation, or order; or any term, condition, or limitation, of any license, standard design certification, or standard design approval.141 140. 10 C.F.R. § 50.46(b) is the regulation that would be violated, as a consequence of Dr.

Rolls false statements and failure to disclose important experimental data when he was under oath in the Indian Point Unit 2 licensing hearings.

141. It is unacceptable to subject BEST/MATRRs members to the consequences of Dr. Rolls violation of 10 C.F.R. § 52.4, Deliberate misconduct.

142. On November 17, 2009, Mark Leyse submitted a 10 C.F.R. § 2.802 petition for rulemaking, PRM-50-93,142 which addresses issues similar to those raised by BEST/MATRR in this Contention and in this declaration. However, the NRC is still reviewing PRM-50-93, more than six years after it was submitted. It is difficult to know how long the NRC will continue reviewing PRM-50-93. But there is ample evidence that the Browns Ferry EPU analyses under-predict the zirconium-steam reaction rates that would occur in the event of a LOCA. For example, as discussed in the this declaration and Section IV of this hearing request, on November 24, 2015, Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, disclosed to Leyse that an NRC (TRACE code) computer simulation (using the Baker-Just correlation) of a Westinghouse design-basis accident experiment (FLECHT Run 141 Id.

142 Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No. ML093290250).

55

9573), under-predicted cladding and steam temperatures at the elevation of the hottest section of the tests fuel rod simulators.143 143. A computer safety model is supposed to over-predict temperatures in order to ensure an adequate margin of safety. If a reactors power level is set too high after being qualified by LOCA analyses that do not ensure an adequate margin of safety, a real-life LOCA would lead to a beyond design-basis accident in violation of criteria set forth in 10 C.F.R. § 50.46(b).

144. BEST/MATRR and Leyse are not aware of any actions that the NRC has taken or of any information notices that the NRC has sent licensees, after finding that its TRACE computer safety model under-predicted cladding and steam temperatures for FLECHT Run 9573, at the elevation of the hottest section of the tests fuel rod simulators.144 145. The NRC has sent out information notices in other instances in which a computer safety models simulations indicated that NRC regulations could be violated. For example, the NRC sent out Information Notice No. 98-29: Predicted Increase in Fuel Rod Cladding Oxidation, after Westinghouse notified the NRC that one of its computer safety models may predict higher fuel temperatures and internal pressures at high burnup conditions. This, in turn, may lead to code [computer simulation] resultsthat do not meet the loss-of-coolant accident (LOCA) criterion in 10 CFR 50.46(b)(2)145 [emphasis added].

146. Furthermore, the TRACE computer safety model simulation that under-predicted cladding and steam temperatures for FLECHT Run 9573 demonstrates that the Baker-Just correlation is 143 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

144 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

145 NRC, Information Notice No. 98-29: Predicted Increase in Fuel Rod Cladding Oxidation, August 3, 1998, (ADAMS Accession No: ML003730714), p. 1.

56

inadequate for use in computer safety models that simulate LOCAs. As explained in this declaration, there is additional data from other experiments, along with appropriate citations, that also demonstrates that the Baker-Just correlation is inadequate for use in computer safety models that simulate LOCAs. That means that 10 C.F.R. 50 Appendix K, I.A.5 is non-conservative. In regard to the zirconium-steam reaction that would occur in the event of a LOCA, 10 C.F.R. 50 Appendix K, I.A.5 requires that [t]he rate of energy release, hydrogen generation, and cladding oxidation from the metal-water reaction shall be calculated using the Baker-Just [correlation].146 147. Nonetheless, the NRC is considering an LAR for EPUs for BFN Units 1, 2, and 3, which is dependent on non-conservative Appendix K LOCA analyses. By overlooking the deficiencies of computer safety models, the NRC undermines its own philosophy of defense-in-depth, which requires the application of conservative models.147 148. The health and safety of BEST/MATRRs members must not be threatened by the fact that the NRC has not concluded its review of PRM-50-93after more than six years. The fact alone that, on November 24, 2015, Aby Mohseni disclosed to Leyse that an NRC computer simulation (using the Baker-Just correlation) of FLECHT Run 9573, under-predicted cladding and steam temperatures at the elevation of the hottest section of the tests fuel rod simulators,148 is reason enough to deny the LAR for EPUs for BFN Units 1, 2, and 3.

146 NRC, Appendix K to Part 50ECCS Evaluation Models, (This information is available at:

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appk.html : last visited on 09/02/16).

147 Charles Miller et al., NRC, Recommendations for Enhancing Reactor Safety in the 21st Century: The Near-Term Task Force Review of Insights from the Fukushima Daiichi Accident, SECY-11-0093, July 12, 2011, (ADAMS Accession No: ML111861807), p. 3.

148 Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

57

149. Furthermore, the NRC must not grant the LAR for EPUs for BFN Units 1, 2, and 3, because AREVAs Appendix K LOCA calculations are supported by false statements that were made by a manager in Westinghouses Nuclear Fuel Division, Dr. Roll, when he was under oath in the Indian Point Unit 2 licensing hearings. AREVAs Appendix K LOCA calculations are supported by the cherry-picked FLECHT program data that Dr. Roll discussed when he was under oath in the Indian Point Unit 2 licensing hearings in order to defend the Baker-Just correlation. That was over four decades ago; however, in a contemporary licensing hearing, if Dr. Roll were to make false statements and not disclose important experimental data when he was under oath, he would be in violation of 10 C.F.R. § 52.4, Deliberate misconduct.149 150. The health and safety of BEST/MATRR members must not be compromised by Dr. Rolls violation of 10 C.F.R. § 52.4. And the health and safety of BEST/MATRR members must not be compromised by the application of the non-conservative Appendix K model that has been employed to help qualify the proposed EPUs for BFN Units 1, 2, and 3. AREVAs LOCA analyses regarding the EPUs for BFN Units 1, 2, and 3 predicted PCTs of 2030°F for ATRIUM 10XM fuel150 and 2086°F for ATRIUM 10 fuel;151 however, those PCTs were calculated with a non-conservative Appendix K model, which used the Baker-Just correlation.152 By definition, a 149 10 C.F.R. § 52.4, Deliberate misconduct, (This information is available at: http://www.nrc.gov/reading-rm/doc-collections/cfr/part052/part052-0004.html: last visited on 09/05/16).

150 AREVA, Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU), ANP-3377NP, Revision 3, Attachment 11 Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No:

ML15282A184), pp. 6.1, 6.3, 6.9, 8.6.

151 AREVA, Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU), ANP-3384NP, Revision 3, Attachment 15 Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No: ML15282A187), pp. 2.2, 5.1, 5.4, 6.1.

152 AREVA, Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU), ANP-3377NP, Revision 3, Attachment 11 Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No:

ML15282A184), p. 1.1.

58

non-conservative model does not ensure an adequate margin of safety. And, if a reactors power level is set too high after being qualified by LOCA analyses that do not ensure an adequate margin of safety, a real-life LOCA would have a PCT that exceeded the 10 C.F.R. § 50.46(b)(1)

PCT limit of 2200°F.153 151. If the PCT were to exceed 2200°F, it would be a beyond design-basis accident. If one of the Browns Ferry reactors were to melt down, hundreds of kilograms of explosive hydrogen gas would be generated. It is likely that the hydrogen would then explode and destroy a reactor building, releasing large quantities of harmful radioactive material into the environment, as occurred in the Fukushima Daiichi accident. Clearly, the health and safety of BEST/MATRR members must not be compromised by EPUs for BFN Units 1, 2, and 3.

The Amendment Request for the EPUs for BFN Units 1, 2, and 3 Must Be Denied 152. I allege that non-conservative computer safety model analyses were performed in order to justify the EPUs for BFN Units 1, 2, and 3. When using the Baker-Just correlation, computer safety models like AREVAs EXEM BWR-2000 Evaluation Model under-predict the zirconium-steam reaction rates that occurred in experiments discussed in this declaration.

Computer safety models are supposed to over-predict reaction rates in order to ensure an adequate margin of safety. The experimental results discussed in this declaration are evidence that AREVAs EXEM BWR-2000 Evaluation Model under-predicts the zirconium-steam reaction rates that would occur in the event of a LOCA, which means that the amendment request for the EPUs for BFN Units 1, 2, and 3 must be denied.

153 NRC, § 50.46 Acceptance criteria for emergency core cooling systems for light-water nuclear power reactors, (This information is available at: http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0046.html : last visited on 09/02/16).

59

Respectfully submitted on behalf of BEST/MATRR,

/s/

Mark Leyse P.O. Box 1314 New York, NY 10025 markleyse@gmail.com Dated: September 6, 2016 60

REFERENCES PERTAINING TO MARK LEYSES BACKGROUND Mark Leyse, PRM-50-103, October 14, 2011, (ADAMS Accession No. ML11301A094).

Mark Leyse, Author, and Christopher Paine, Contributing Editor, Preventing Hydrogen Explosions In Severe Nuclear Accidents: Unresolved Safety Issues Involving Hydrogen Generation And Mitigation, NRDC Report, R:14-02-B, March 2014. (available at:

https://www.nrdc.org/sites/default/files/hydrogen-generation-safety-report.pdf : last visited on 08/28/16)

Mark Leyse and Christopher Paine, Preventing Hydrogen Explosions at Indian Point Nuclear Plant: Fact versus Industry Spin, NRDC IB: 13-01-F, February 2013. (available at:

https://www.nrdc.org/sites/default/files/IndianPoint-hydrogen-explosions-IB.pdf : last visited on 08/28/16)

Mark Leyse, Post-Fukushima Hardened Vents with High-Capacity Filters for BWR Mark Is and Mark IIs, Report for NRDC, July 2012, (ADAMS Accession No. ML12254A865).

Mark Leyse, PRM-50-84, March 15, 2007 (ADAMS Accession No. ML070871368).

American Nuclear Society, Nuclear News, June 2007, p. 64.

David Lochbaum, Union of Concerned Scientists, Comments on Petition for Rulemaking Submitted by Mark Edward Leyse (Docket No. PRM-50-84), July 31, 2007, (ADAMS Accession No. ML072130342).

NRC, Mark Edward Leyse; Consideration of Petition in Rulemaking Process, Docket No. PRM-50-84; NRC-2007-0013, Federal Register, Vol. 73, No. 228, November 25, 2008, pp. 71564-71569.

NRC, Performance-Based Emergency Core Cooling System Acceptance Criteria, NRC-2008-0332, Federal Register, Vol. 74, No. 155, August 13, 2009, pp. 40765-40776.

NRC, Commission Voting Record, Decision Item: SECY-12-0034, Proposed Rulemaking 10 CFR 50.46(c): Emergency Core Cooling System Performance During Loss-of-Coolant Accidents (RIN 3150-AH42), January 7, 2013, (ADAMS Accession No. ML13008A368).

Rui Hu, Mujid S. Kazimi, Mark Leyse, Considering the Thermal Resistance of Crud in LOCA Analysis, American Nuclear Society, 2009 Winter Meeting, Washington, D.C., November 15-19, 2009.

Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No. ML093290250).

American Nuclear Society, Nuclear News, March 2010, p. 36.

61

David Lochbaum, Union of Concerned Scientists, Comments Submitted by the Union of Concerned Scientists on the Petition for Rulemaking Submitted by Mark Edward Leyse (Docket No. PRM-50-93), April 27, 2010, (ADAMS Accession No. ML101180175).

Federal Register, Vol. 75, No. 207, Notice of consolidation of petitions for rulemaking and re-opening of comment period, October 27, 2010, pp. 66007-66008.

Mark Leyse, PRM-50-95, June 7, 2010, (ADAMS Accession No. ML101610121).

Suzanne McElligott, Inside NRC, July 30, 2010.

REFERENCES NRC, Applications and Amendments to Facility Operating Licenses and Combined Licenses Involving Proposed No Significant Hazards Considerations and Containing Sensitive Unclassified Non-Safeguards Information and Order Imposing Procedures for Access to Sensitive Unclassified Non-Safeguards Information, NRC-2016-0118, Federal Register, Vol. 81, No. 128, July 5, 2016.

AREVA, Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU), ANP-3377NP, Revision 3, Attachment 11 Browns Ferry Units 1, 2, and 3 LOCA Break Spectrum Analysis for ATRIUM 10XM Fuel (EPU) (Non-Proprietary), August 2015, (ADAMS Accession No: ML15282A184).

AREVA, Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM 10XM Fuel (EPU),ANP-3378NP, Revision 3, Attachment 13 Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM 10XM Fuel (EPU) (Non-Proprietary),

August 2015, (ADAMS Accession No: ML15282A185).

AREVA, Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU), ANP-3384NP, Revision 3, Attachment 15 Browns Ferry Units 1, 2, and 3 LOCA-ECCS Analysis MAPLHGR Limits for ATRIUM-10 Fuel (EPU) (Non-Proprietary),

August 2015, (ADAMS Accession No: ML15282A187).

NRC, Appendix K to Part 50ECCS Evaluation Models, (This information is available at:

http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appk.html : last visited on 07/23/16).

Louis Baker, Jr. and Louis C. Just, Studies of Metal-Water Reactions at High Temperatures: III.

Experimental and Theoretical Studies of the Zirconium-Water Reaction, ANL-6548, May 1962, (ADAMS Accession No: ML050550198).

G. Schanz, Recommendations and Supporting Information on the Choice of Zirconium Oxidation Models in Severe Accident Codes, FZKA 6827, 2003.

62

W. A. Bostrom, The High Temperature Oxidation of Ziracloy in Water, WAPD-104, March 1954, (ADAMS Accession No: ML100900446) and Alexis W. Lemmon, Studies Relating to the Reaction Between Zirconium and Water at High Temperatures, Battelle Memorial Institute, BMI-1154, January 1957, (ADAMS Accession No: ML100570218).

V. F. Urbanic and T. R. Heidrick, High-Temperature Oxidation of Zircaloy-2 and Zircaloy-4 in Steam, Journal of Nuclear Materials 75, 1978.

Alexis W. Lemmon, Studies Relating to the Reaction Between Zirconium and Water at High Temperatures, Battelle Memorial Institute, BMI-1154, January 1957, (ADAMS Accession No:

ML100570218).

Aby Mohseni, Deputy Director of the NRCs Division of Policy and Rulemaking, e-mail to Mark Leyse, regarding the NRCs TRACE computer simulation of the FLECHT Run 9573 test bundle, November 24, 2015, (ADAMS Accession No: ML15341A160).

Daniel F. Ford, Henry W. Kendall, James J. MacKenzie, A Critique of the New A. E. C. Design Criteria for Reactor Safety Systems, Union of Concerned Scientists, October 1971; I.A. Forbes, D.F. Ford, H.W. Kendall, J.J. MacKenzie, Nuclear Reactor Safety: An Evaluation of New Evidence, Nuclear News, 14, No. 9, September 1971.

Henry W. Kendalls A Distant Light: Scientists and Public Policy has reprinted I.A. Forbes, D.F.

Ford, H.W. Kendall, J.J. MacKenzie, Nuclear Reactor Safety: An Evaluation of New Evidence, Nuclear News, 14, No. 9, September 1971. For the quoted passage see Henry W. Kendall, A Distant Light: Scientists and Public Policy, (New York: Springer-Verlag, 2000).

Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.:

Indian Point Station Unit No. 2, Docket No. 50-247, November 2, 1971, (ADAMS Accession No. ML100350642).

R. A. Lorenz, D. O. Hobson, G. W. Parker, Final Report on the First Fuel Rod Failure Transient Test of a Zircaloy-Clad Fuel Rod Cluster in TREAT, ORNL-4635, March 1971.

Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.:

Indian Point Station Unit No. 2, Docket No. 50-247, November 1, 1971, (ADAMS Accession No. ML100350644).

W. B. Cottrell, ORNL Nuclear Safety Research and Development Program Bimonthly Report for March-April 1971, ORNL-TM-3411, July 1971.

Atomic Energy Commission, In the Matter of: Consolidated Edison Company of New York, Inc.:

Indian Point Station Unit No. 2, Docket No. 50-247, November 16, 1971, (ADAMS Accession No. ML100350625).

F. D. Kingsbury, J. F. Mellor, and A. P. Suda, Materials Evaluation, Appendix B of WCAP-7665. F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse Electric Corporation, WCAP-7665, 63

PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, April 1971, (ADAMS Accession No: ML070780083), Appendix B.

F. F. Cadek, D. P. Dominicis, R. H. Leyse, Westinghouse, PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report, WCAP-7665, April 1971, (ADAMS Accession No. ML070780083).

S. Hagen, P. Hofmann, G. Schanz, L. Sepold, Interactions in Zircaloy/UO2 Fuel Rod Bundles with Inconel Spacers at Temperatures above 1200°C (Posttest Results of Severe Fuel Damage Experiments CORA-2 and CORA-3), Forschungszentrum Karlsruhe, KfK 4378, September 1990.

NRC, Denial of Petition for Rulemaking (PRM-50-76), June 29, 2005, (ADAMS Accession No:

ML050250359).

Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No. ML093290250).

NRC, Draft Interim Review of PRM-50-93/95 Issues Related to Conservatism of 2200 degrees F, Metal-Water Reaction Rate Correlations, and The Impression Left from [FLECHT] Run 9573 ,

October 16, 2012, (ADAMS Accession No. ML12265A277).

NRC, Public Participation in NRC Regulatory Decision-Making, Transcript of Proceedings, January 31, 2013, (available at: http://www.nrc.gov/reading-rm/doc-collections/commission/tr/2013/20130131b.pdf ).

NRC, Draft Interim Review of PRM-50-93/95 Issues Related to Minimum Allowable Core Reflood Rate, March 8, 2013, (ADAMS Accession No. ML13067A261).

Robert H. Leyse, Westinghouse, Nuclear Energy Systems, Test Engineering, Memorandum RD-TE-70-616, FLECHT Monthly Report, December 14, 1970. This Memorandum is available at Appendix I of PRM-50-93. See Mark Leyse, PRM-50-93, November 17, 2009, (ADAMS Accession No: ML093290250), Appendix I.

NRC, § 50.46 Acceptance criteria for emergency core cooling systems for light-water nuclear power reactors, (This information is available at: http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0046.html : last visited on 09/04/16).

G. Hache, R. Gonzalez, B. Adroguer, Institute for Protection and Nuclear Safety, Status of ICARE Code Development and Assessment, in NRC Proceedings of the Twentieth Water Reactor Safety Information Meeting, NUREG/CP-0126, Vol. 2, 1992, (ADAMS Accession No: ML042230126).

T.J. Haste et al., In-Vessel Core Degradation in LWR Severe Accidents, European Commission, Report EUR 16695 EN, 1996.

L. J. Ott, Oak Ridge National Laboratory, Report of Foreign Travel of L. J. Ott, Engineering Analysis Section, Engineering Technology Division, ORNL/FTR-3780, October 16, 1990.

64

L. J. Ott, W. I. van Rij, In-Vessel PhenomenaCORA: BWR Core Melt Progression Phenomena Program, Oak Ridge National Laboratory, CONF-9105173-3-Extd.Abst., Presented at Cooperative Severe Accident Research Program, Semiannual Review Meeting, Bethesda, Maryland, May 6-10, 1991.

R. E. Williford, An Assessment of Safety Margins in Zircaloy Oxidation and Embrittlement Criteria for ECCS Acceptance, NUREG/CR-4412, April 1986, (ADAMS Accession No:

ML083400371).

NRC, Draft Interim Review of PRM-50-93/95 Issues Related to the CORA Tests, August 23, 2011, (ADAMS Accession No: ML112211930).

F. J. Erbacher, S. Leistikow, A Review of Zircaloy Fuel Cladding Behavior in a Loss-of-Coolant Accident, Kernforschungszentrum Karlsruhe, KfK 3973, September 1985.

Winston & Strawn LLP, Duke Energy Corporation, Catawba Nuclear Station Units 1 and 2, Enclosure, Testimony of Robert C. Harvey and Bert M. Dunn on Behalf of Duke Energy Corporation, MOX Fuel Lead Assembly Program, MOX Fuel Characteristics and Behavior, and Design Basis Accident (LOCA) Analysis, July 1, 2004, (ADAMS Accession No: ML041950059).

G. Bandini et al., Recent Advances in ASTEC Validation on Circuit Thermal-Hydraulic and Core Degradation, Progress in Nuclear Energy, 52, 2010.

Henry W. Kendall, A Distant Light: Scientists and Public Policy, Springer-Verlag, New York, 2000.

NRC, Information Notice No. 98-29: Predicted Increase in Fuel Rod Cladding Oxidation, August 3, 1998, (ADAMS Accession No: ML003730714).

Charles Miller et al., NRC, Recommendations for Enhancing Reactor Safety in the 21st Century:

The Near-Term Task Force Review of Insights from the Fukushima Daiichi Accident, SECY 0093, July 12, 2011, (ADAMS Accession No: ML111861807).

10 C.F.R. § 52.4, Deliberate misconduct, (This information is available at:

http://www.nrc.gov/reading-rm/doc-collections/cfr/part052/part052-0004.html: last visited on 09/05/16).

65

UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE SECRETARY OF THE COMMISSION In the Matter of:  :

TENNESSEE VALLEY AUTHORITY  :

(Browns Ferry Nuclear Plant Units 1, 2, and 3;:

Docket Nos. 50-259, 50-260, and 50-296;:

NRC-2016-0118)  :

CERTIFICATE OF SERVICE I hereby certify that on September 9, 2016, I posted the foregoing DECLARATION OF MARK LEYSE TO SUPPORT THE HEARING REQUEST AND PETITION FOR LEAVE TO INTERVENE BY THE BELLEFONTE EFFICIENCY &

SUSTAINABILITY TEAM/ MOTHERS AGAINST TENNESSEE RIVER RADIATION REGARDING TENNESSEE VALLEY AUTHORITYS LICENSE AMENDMENT REQUEST FOR EXTENDED POWER UPRATES FOR BROWNS FERRY NUCLEAR PLANT UNITS 1, 2, AND 3. This has been filed on the NRCs Electronic Information Exchange system this 9th day of September, 2016. It is my understanding that as a result, the Commission, Atomic Safety and Licensing Board, and parties were served.

Respectfully submitted,

/s/

_ Garry Morgan BEST/MATRR PO Box 241 Scottsboro, AL 35768 Phone: 256-218-0124 E-mail: best@matrr.org September 9, 2016 37