ML042330027

Duke Energy Corporations Proposed Findings of Fact and Conclusions of Law Regarding Contention I

ML042330027
Person / Time
Site:	Catawba
Issue date:	08/06/2004
From:	Repka D Duke Energy Corp, Winston & Strawn, LLP
To:	Atomic Safety and Licensing Board Panel
SECY RAS
References
50-413-OLA, 50-414-OLA, ASLBP 03-815-03-OLA, RAS 8354
Download: ML042330027 (39)

Text

WA s g 36 L/

August 6, 2004 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD DOCKETED USNRC August 17,2004 (3:15PM)

OFFICE OF SECRETARY RULEMAKINGS AND ADJUDICATIONS STAFF In the Matter of:

DUKE ENERGY CORPORATION (Catawba Nuclear Station, Units 1 and 2)

)

)

)

)

)

)

)

)

Docket Nos.

50-413-OLA 50-414-OLA DUKE ENERGY CORPORATION'S PROPOSED FINDINGS OF FACT AND CONCLUSIONS OF LAW REGARDING CONTENTION I David A. Repka WINSTON & STRAWN, LLP 1400 L Street, NW Washington, D.C. 20005-3502 Timika Shafeek-Horton DUKE ENERGY CORPORATION P.O. Box 1006 Mail Code: EC1 1X-1128 Charlotte, N.C. 28201-1006 ATTORNEYS FOR DUKE ENERGY CORPORATION

-eFtp1,7te= Sec'/- 057

TABLE OF CONTENTS BACKGROUND....

2........................

A.

The License Amendment Request.............................

2 B.

The Evidentiary Hearing.............................

4 II. FINDINGS OF FACT.............................

6 A.

MOX Fuel Lead Assemblies..............................

6 B.

Duke's MOX Fuel LOCA Analysis..............................

.7 C.

Fuel Relocation -

LEU Fuel.............................

I I D.

Fuel Relocation -

MOX Fuel.............................

1 6 Filling Ratio.............................

17 M5Th! Cladding Ductility.............................

21 Fuel-Cladding Interaction.............................

23 MOX Fuel Relative Power at High Burnup.............................

25 E.

Assessment of Relocation Impacts.............................

26 Stacking the Conservatisms.............................

26 Running the Numbers.............................

28 F.

Credibility Determinations.............................

32 III.

CONCLUSIONS OF LAW..................

34 i

August 6, 2004 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of:

)

)

DUKE ENERGY CORPORATION

)

)

DocketNos.

50-413-OLA (Catawba Nuclear Station,

)

50-414-OLA Units l and 2)

)

)

DUKE ENERGY CORPORATION'S PROPOSED FINDINGS OF FACT AND CONCLUSIONS OF LAW REGARDING CONTENTION I In accordance with 10 C.F.R. § 2.754 and the scheduling order of the Atomic Safety and Licensing Board ("Licensing Board"),' Duke Energy Corporation ("Duke") submits in the form of a partial initial decision these proposed findings of fact and conclusions of law on Contention I in this proceeding. These proposed findings of fact and conclusions of law address and resolve all contested issues raised by Contention I.

Contention I was derived from certain proposed contentions submitted by the Blue Ridge Environmental Defense League ("BREDL"). It was re-framed and admitted by the Licensing Board by Memorandum and Order of March 5, 2004.2 An evidentiary hearing was conducted on Contention I before the Licensing Board on July 14 and 15, 2004, in Rockville, Maryland. These proposed findings of fact and conclusions of law are based on the evidentiary "Order (Regarding Proposed Redacted Memorandum and Order, and Proposed Schedule Changes)," May 25, 2004, at 2.

2 Duke Energy Corporation (Catawba Nuclear Station, Units I and 2), LBP-04-4, 59 NRC 129, 183 (2004).

I

record in this proceeding and, in total, support a determination that Duke has demonstrated, by a preponderance of the evidence, that with respect to the matters raised in Contention I there is reasonable assurance that the proposed license amendment will not endanger public health and safety.

The proposed findings of fact and conclusion of law are presented in sequentially numbered paragraphs.

The first section, "Background," describes the approval at issue, the contention addressed in this partial initial decision, and the evidentiary hearing on that contention.

The second section, "Findings of Fact," presents the specific findings of fact relevant to resolving Contention I. The final section, "Conclusions of Law," sets forth the Licensing Board's conclusions as necessary to resolve Contention I.

I.

BACKGROUND A.

The License Amendment Request

1.

This proceeding relates to the License Amendment Request ("LAR") filed by Duke on February 27, 2003. In the LAR, Duke requested NRC authorization to use four Mixed Oxide ("MOX") fuel lead assemblies at the Catawba Nuclear Station ("Catawba") or the McGuire Nuclear Station ("McGuire"). Duke Testimony, ¶¶ 5, 11. The LAR was subsequently amended to apply to Catawba only. Id., ¶ 11. Duke is currently targeting Catawba Unit 1, Cycle 16 (ClC16), with a Spring 2005 startup, for initial insertion of the MOX fuel lead assemblies. Id., ¶ 14.

2.

Duke's proposal to use four MOX fuel lead assemblies is part of an ongoing nuclear non-proliferation program of the United States and the Russian Federation. The goal of the program -

administered by the U.S. Department of Energy ("DOE") -

is to dispose of surplus plutonium from nuclear weapons by converting that material into MOX fuel and using the fuel in nuclear power reactors. Id., ¶ 12.

2

3.

The current proposal for four MOX fuel lead assemblies supports the potential future use of larger quantities of MOX fuel at either Catawba or McGuire. Any such future "batch" use would be subject to a separate NRC licensing action. Id.

4.

The use of lead assemblies in commercial reactors to demonstrate new fuel technologies is consistent with NRC Staff preference and guidance documents. Tr. 2699-2700 (Shoop). It is an approach that has been applied routinely to facilitate fuel design improvements.

Duke Testimony, ¶ 165; Duke Rebuttal Testimony, 1 77.

5.

The MOX fuel lead assemblies will be included in cores that will be predominately comprised of Low Enriched Uranium ("LEU") fuel assemblies. For ClC16, the LEU fuel will be predominately Westinghouse fuel of the Robust Fuel Assembly design. Duke Testimony, ¶ 15.

6.

The MOX fuel lead assemblies will be loaded in non-limiting locations in the core. That is, the Catawba cores containing MOX fuel lead assemblies will be designed to ensure that the MOX fuel assembly power is not the highest assembly power in the core. Id.,

¶ 137; Duke Rebuttal Testimony, ¶ 72.

7.

On April 5, 2004, the NRC Staff issued a Safety Evaluation ("SE), concluding that the health and safety of the public will not be endangered by operation with four MOX fuel lead assemblies at Catawba. Duke Testimony, ¶ 16; Exhibit 38.

8.

The Advisory Committee on Reactor Safeguards ("ACRS") Subcommittee on Reactor Fuels reviewed the LAR during a public meeting on April 21, 2004. Subsequently, the full ACRS reviewed the MOX fuel lead assembly LAR in a public meeting on May 6, 2004. In a letter dated May 7, 2004, the ACRS concluded that "under the restricted circumstances considered in both the Duke Power application and the NRC Staff's safety evaluation, the four 3

mixed oxide lead test assemblies in non-limiting core locations that do not contain control rods can be irradiated in either of the Catawba reactor cores with no undue risk to the public health and safety." Duke Testimony, ¶ 17.

B.

The Evidentiarv Hearing

9.

BREDL Contention I relates to Duke's analysis of a design basis Loss of Coolant Accident ("LOCA"). The contention asserts that:

The LAR is inadequate because Duke has failed to account for differences in MOX and LEU fuel behavior (both known differences and recent information on possible differences) and for the impact of such differences on LOCAs and on the [design basis accident ("DBA")] analysis for Catawba.3

10.

The issue in Contention I was derived from several BREDL proposed contentions, most of which related to the projected dose and public health consequences of design basis LOCAs and beyond-design-basis ("severe") accidents.

The Licensing Board subsequently clarified that Contention I does not include issues and analyses related to dose consequences. 4

11.

The LOCA analysis is the only design basis accident analysis at issue in Contention I. Duke Testimony, ¶ 31. Accordingly, the contention focuses on the methodology and acceptance criteria in NRC regulations in 10 C.F.R. § 50.46 and 10 C.F.R. Part 50, Appendix K. Id., at ¶ 8. As became clear by the time of the evidentiary hearing, the only issue in dispute concerns the need for evaluation, in the LOCA analysis, of the impact of fuel relocation. No 3

LBP-04-04, 59 NRC at 183.

4 "Order (Confirming Matters Addressed at April 6 Telephone Conference)," April 8, 2004 (at 2) ("With respect to Contention I, this contention encompasses those calculations involved in the determination of events up to and including LOCAs and DBAs, but does not include analyses related to releases either in containment or offsite."); see also Tr.

1726-36.

4

other aspect of the Duke MOX fuel design basis LOCA analysis was disputed.

Tr. 2457 (Lyman).

12.

Duke presented a panel of four qualified experts on Contention I. Duke's experts were Mr. Steven Nesbit, an Engineering Supervisor employed by Duke who is the MOX Fuel Project Manager; Mr. Robert Harvey, a Senior Engineer employed by Duke and responsible for LOCA analyses; Mr. Bert Dunn, an Advisory Engineer employed by AREVA Framatome ANP, Inc., responsible for LOCA analyses; and Dr. J. Kevin McCoy, an Advisory Engineer employed by AREVA Framatome ANP, Inc., with expertise in the fields of metallurgy and materials engineering.

Duke Testimony, ¶¶ 1-5; Duke Rebuttal Testimony, ¶¶ 1-4.

The extensive, relevant experience of all four Duke experts is demonstrated by their statements of professional qualifications. Exhibits 47-50.

13.

The Duke panel submitted direct testimony dated July 1, 2004, which was slightly revised at the evidentiary hearing. Tr. 2098-99 (Nesbit, McCoy). Duke submitted rebuttal testimony dated July 8, 2004. Two of the panel members also submitted supplemental rebuttal testimony, limited to addressing proposed Exhibit C (for identification), on July 20, 2004.

14.

The NRC Staff presented a panel of three qualified experts on Contention I. The Staff's experts were: Ms. Undine Shoop, a Reactor Systems Engineer in the Office of Nuclear Reactor Regulation ("NRR"); Dr. Ralph Landry, a Senior Reactor Engineer in NRR; and Dr.

Ralph Meyer, a Senior Technical Advisor for Core Performance and Fuel Behavior in the Office of Nuclear Regulatory Research.

The extensive, relevant experience of these three Staff witnesses is demonstrated by their statements of professional qualifications. Exhibit 37.

5

15.

The NRC Staff panel submitted direct testimony on July 1, 2004, which was replaced by a slightly revised version dated July 14, 2004. The Staff panel submitted rebuttal testimony on July 8, 2004.

16.

BREDL submitted limited direct and rebuttal testimony on July 1, 2004 and July 8, 2004, from Dr. Edwin Lyman. Dr. Lyman added to that testimony during the evidentiary hearing. See generally Tr. 2455-2613. As is clear from his curriculum vitae, Dr. Lyman has degrees in theoretical physics and. physics, and has been involved for a number of years in nuclear regulatory policy matters. Exhibit 25. However, Dr. Lyman does not have direct experience in LOCA analyses. Tr. 2455-57 (Lyman). His familiarity with these matters is limited to his review of the public literature. Tr. 2457 (Lyman). The weight of his testimony is, therefore, viewed accordingly in these proposed findings of fact.

II.

FINDINGS OF FACT A.

MOX Fuel Lead Assemblies

17.

The MOX fuel lead assemblies will be manufactured in France under the direction of AREVA. The MOX fuel assembly design is based on the AREVA Advanced Mark-BW fuel assembly, a fully qualified LEU fuel design. Duke Testimony, ¶ 13.

18.

Like the Advanced Mark-BW, the MOX fuel assemblies will use M5Th fuel rod cladding. M5Tm is an advanced cladding material that provides superior corrosion resistance, lower irradiation growth, and better ductility retention than Zircaloy-4. M5Tm has been reviewed and approved by the NRC and is currently being used in fuel operated in nine domestic reactors.

M5T"' is also being used in foreign nuclear power reactors, including MOX fuel applications.

Id., m¶ 48, 79; Staff Testimony, ¶ A.30.

19.

MOX fuel is very similar to LEU fuel. MOX fuel pellets are comprised of a small amount of plutonium oxide mixed with uranium oxide (typically depleted uranium in the range 6

of 0.2 - 0.3% 235U). MOX fuel used in European reactors utilizes plutonium that has been recovered from reprocessing LEU fuel. This type of plutonium contains at least 20% 24GPu and is classified as reactor grade ("RG"). The Catawba MOX fuel lead assemblies will utilize weapons grade ("WG") plutonium which is less than 7% 240Pu. Duke Testimony, 1 21.

20.

MOX fuel has been previously approved by the NRC for use in commercial reactors. Moreover, there is a substantial experience base with MOX fuel. More than thirty nuclear power reactors in Europe are currently using MOX fuel in quantities in excess of the four proposed MOX fuel lead assemblies. Id., ¶¶ 23-24. Experience indicates, for example, that MOX fuel failure rates are commensurate with those of LEU fuel. Id., ¶ 25; Tr. 2432-33; 2454 (Nesbit).

B.

Duke's MOX Fuel LOCA Analysis

21.

Duke's MOX fuel LOCA analyses are discussed in Section 3.7.1 of the LAR (Exhibit 1) and in Duke's November 3, 2003 response to an NRC Staff Request for Additional Information ("RAI") (Exhibit 2). The LOCA analyses address the requirements of 10 C.F.R.

§ 50.46 and 10 C.F.R. Part 50, Appendix K. Duke Testimony, ¶ 31.

22.

Under 10 C.F.R. § 50.46, a LOCA evaluation model must be used which either (a) realistically describes the behavior of the reactor system during a LOCA such that the uncertainty in the calculated results can be estimated and accounted for, or (b) conforms with the required and acceptable features of 10 C.F.R. 50, Appendix K. Whichever approach to the evaluation is followed, the results must meet the acceptance criteria stated in 10 C.F.R.

§ 50.46(b). Specifically the peak cladding temperature ("PCT") must not exceed 2200'F, the maximum local oxidation must not exceed 17%, the hydrogen generated must not exceed that which could be produced by oxidation of 1% of the total cladding, the core must remain in a 7

coolable geometry, and the core temperature must remain at an acceptable level for an extended period of time. NRC Staff Testimony, IT A.14; Duke Testimony, IT 32-33.

23.

The results of the LOCA analyses are used to define the "LOCA limits" (allowable core power peaking limits) which will be adopted to maintain plant operation consistent with the analyses that show compliance with the criteria of 10 C.F.R. § 50.46. Duke Testimony, ¶ 35.

24.

The MOX fuel lead assemblies were analyzed for a LOCA using the AREVA Appendix K methodology.

This is a conservative deterministic approach, in contrast to a realistic (best estimate) analysis. Id., m¶ 36, 43. The AREVA evaluation model was previously approved by the NRC for plants such as Catawba operating with LEU fuel. Id., ¶ 43.

25.

The AREVA evaluation model was adaptable for MOX fuel because LOCA response is primarily controlled by system phenomena.

However, specific MOX fuel characteristics were considered and addressed. A review of potential differences between LEU and MOX fuel that could affect a LOCA calculation was made, and, where necessary, changes were made to the AREVA evaluation techniques. Id., IT 43-47. No aspect of this adaptation of the AREVA evaluation model to MOX fuel was disputed. Tr. 2457 (Lyman).

26.

In the adaptation of the evaluation model for MOX fuel, it is also undisputed that in two key respects AREVA used a conservative approach by utilizing LEU fuel characteristics rather than MOX-specific characteristics. Specifically, the MOX assemblies were conservatively evaluated using neutron power-related characteristics (fuel neutronics coefficients) and decay heat characteristics appropriate for LEU assemblies. Duke Testimony, TI 46, 63-64.

27.

Because PCT is strongly influenced by decay heat, the use of the LEU decay heat model in the MOX fuel LOCA analysis has been estimated to be a conservatism of up to 750F on 8

PCT. Duke Testimony, 1 63. Dr. Lyman agreed that the decay heat model is a conservatism in the MOX fuel LOCA analysis and did not dispute Duke's numeric estimate of that conservatism.

Tr. 2458, 2510 (Lyman).

28.

Fuel cladding ballooning characteristics (i.e., "strain") and rupture effects are modeled in the AREVA evaluation model.

The modeling of fuel cladding in the LOCA evaluation model was specific to M5TM cladding characteristics. Duke Testimony, Im 51-53. In this regard, the deterministic evaluation model used unirradiated M5T cladding properties, maximizing the predicted strain. Id., ¶ 54.

29.

In contrast to one best estimate LOCA evaluation model, the AREVA Appendix K LOCA evaluation model does not specifically model fuel relocation effects. Rather, relocation is one potential non-conservatism in the model.

This is addressed by the use of specific conservatisms required to be built into a deterministic evaluation model, which collectively create substantial margins. Tr. 2374-76 (Harvey); Duke Testimony, m¶ 94-96; Duke Rebuttal Testimony, ¶ 14. In this respect, the AREVA Appendix K LOCA evaluation model is consistent with all approved Appendix K models. Tr. 2651 (Landry); Staff Rebuttal Testimony, I A.5.

30.

For the MOX fuel lead assemblies, LOCA calculations were performed for a variety of plant conditions.

The most limiting cases used to establish the LOCA limits (allowable peaking) for the cycles with MOX fuel assemblies were presented in the RAI response, Exhibit 2 (Table Q14-1). Duke Testimony, ¶¶ 55, 152; Tr. 2365 (Nesbit, Harvey).

The results of sample calculations presented in the LAR (Exhibit 1, Table 3-5) were not used for establishing LOCA limits. Tr. 2351 (Nesbit).

31.

The results of the relevant limiting case, and the corresponding regulatory acceptance criteria, were presented in Duke's testimony:

9

Peak Cladding Temperature (PCT) 2019.50F (22000F)

PCT at Ruptured Location 17500F (22000F)

Local Oxidation 5.2%

(17%)

Total Core Oxidation 0.4%

(1%)

(Hydrogen Generation)

Duke Testimony, ¶¶ 33, 55. The results are all within the NRC acceptance criteria.

32.

The maximum calculated cladding strain (ballooning) for this case is 51% with an associated blockage of 52% of the coolant channel surrounding the hot pin. This is well within the coolable geometry limit (specified by the AREVA evaluation model) of 90%. Id., ¶ 56; Duke Rebuttal Testimony, ¶ 69; Tr. 2440-42 (Dunn). The NRC Staff has previously reviewed the AREVA M5TM flow blockage model, and found that approach to be acceptable.

Duke Testimony, ¶ 87.

33.

Long-term coolability does not appear to be in dispute, given the revision made to Dr. Lyman's testimony, A.5. Tr. 2239 (Lyman). Nonetheless, the long-term coolability issue was satisfactorily addressed by Duke. Duke Testimony, ¶ 57; Duke Rebuttal Testimony, ¶ 19.

Compliance with the criterion in 10 C.F.R. § 50.46 related to total core oxidation (hydrogen generation), as shown in 1 31 above, also does not appear to be in dispute.

34.

Duke's MOX fuel LOCA analysis demonstrates that, with respect to PCT, the MOX fuel limiting case PCT (2019.50F from Exhibit 2, Table Q14-1) is not only within the regulatory acceptance limit; it also remains less than the limiting analysis of record for Catawba that is based on Westinghouse LEU fuel. Tr. 2372-73 (Harvey); Exhibit 2, at 30.

35.

The LOCA limits (i.e., the linear heat generation rate ("LHGR") or peaking values) for Catawba will specifically be established to ensure that Catawba operates at acceptable values for both the MOX fuel and the co-resident LEU fuel assemblies. Given the 10

core design, the MOX assemblies will not be the peak power assemblies (highest LHGR) in the core. Duke Testimony, IT 73, 137-140. From a LOCA analysis perspective, the MOX fuel assemblies therefore are not limiting. Tr. 2373 (Harvey).

36.

The MOX fuel large break LOCA analysis also included Catawba Unit 2 steam generator assumptions. The calculated PCT is therefore conservative relative to Catawba Unit 1.

Tr. 2367-68 (Nesbit).

37.

Duke also performed specific analyses which provide a direct comparison of MOX versus LEU fuel assemblies with the same peaking factor. These analyses show that there is only a small difference in the results (less than 400F in PCT) between the two fuel types, without taking any credit for beneficial characteristics of MOX fuel pellets versus LEU fuel pellets. Duke Testimony, ¶ 58; NRC Staff Testimony, I A.23. A calculated difference in PCT of this magnitude is inconsequential in a LOCA analysis. Duke Testimony, ¶ 58; Duke Rebuttal Testimony, ¶j 12.

38.

The NRC Staff concluded that the LOCA analysis supporting the use of MOX fuel lead assemblies -

as summarized in the LAR and the RAI responses (LAR "supplements")

was performed in an acceptable manner, is conservative, and demonstrates compliance with the acceptance criteria of 10 C.F.R. § 50.46. Staff Testimony, I A.20; Exhibit 38, Section 2.4.1.

C.

Fuel Relocation -

LEU Fuel

39.

Fuel relocation is a generic issue that is not unique to MOX fuel. The possible impact of fuel relocation on LOCA analyses for LEU fuel was recognized by the NRC Staff as Generic Issue ("GI") 92. Duke Testimony, 1 94. The GI was initially assigned a low priority and was subsequently dropped. Id. The generic issue was more recently acknowledged in NRC regulatory correspondence such as a memorandum from A.C. Thadani to S. Collins, "Research 11

Information Letter 0202, Revision of 10 C.F.R. § 50.46 and Appendix K" (June 20, 2002) (the "Thadani memorandum"). This document was admitted as Exhibit 27.

40.

Notwithstanding the GI and Exhibit 27, the NRC does not require modeling of fuel relocation under Appendix K. Tr. 2667 (Meyer). Instead, as noted above, Appendix K accounts for any nonconservatisms by compensating conservatisms. Staff Rebuttal Testimony,

¶A.5. Relative to a best estimate approach, Duke's expert has judged that an Appendix K methodology provides more than 6000F in margin for PCT. Duke Testimony, ¶ 62; Tr. 2383 (Nesbit).

41.

The progression of a LOCA is illustrated in Exhibit 6. As the fuel and cladding heat up in a LOCA the pellet-to-cladding gap is increased, reducing the heat conducted to the cladding; the cladding surface area increases, increasing the heat transfer to the coolant; and flow turbulence created by the cladding expansion improves the heat transfer to the coolant. The net effect on the cladding temperature is a short-term cooling of cladding at the location where the cladding expands to the point of rupture. Duke Testimony, ¶ 41. Because of the cooling effects, the ballooned or ruptured location is seldom the location of the calculated PCT.

Staff Testimony, m¶ A.42, A.47.

42.

The relocation issue relates to the possibility that during a LOCA fuel pellets will lose their integrity (break into small fragments) and fall to a lower portion of the fuel rod where the cladding has swelled. Duke Testimony, 1 90. The concern is that relocated fuel may generate too much power in a localized area, and thereby increase the cladding temperature at that location. Id., 1 92. The effect of relocation will depend upon the size of the cladding balloon and upon the filling fraction - that is, the percentage of the expansion space filled by relocated fuel.

12

43.

The most comprehensive experimental evidence available for evaluation of fuel relocation derives from the KfK tests performed at the FR2 reactor in Germany, illustrated in Exhibits 15 and 16. These tests clearly show (Exhibit 15) the increased cooling at the ruptured location in the case where there is no relocation. Duke Testimony, 1 149. They further show (Exhibit 16) in the case where relocation occurs, the cooling benefits are a near match for the detrimental effect of relocation on cladding temperature. Id., ¶ 150; Tr. 2404 (Harvey). In this case, the PCT still occurs at a non-ballooned location on the fuel pin; the relocation effect therefore remains bounded. See Exhibit 16.

44.

Contention I was originally derived from BREDL proposed Contention 10, which cited a presentation made to the NRC on October 23, 2003 by the French Institute de Radioprotection et de Sfiret6 Nucleaire ("IRSN"). IRSN is an agency in the French government that conducts nuclear research. Duke Testimony, ¶ 88. The October 2003 IRSN presentation is Exhibit 28. Exhibit 28 was offered as the basis for the assertion that relocation may introduce a significant uncertainty with respect to the LOCA analysis for the MOX fuel lead assemblies.

45.

Exhibit 28 specifically references certain tests from the VERCORS series conducted by IRSN and its predecessor organization. 5 The VERCORS tests involved irradiated MOX fuel, but are not relevant to a LOCA analysis because they were directed at severe accident consequences and were conducted at temperatures much higher than fuel temperatures experienced during a design basis LOCA. Duke Testimony, m¶99-108.

5 IRSN's predecessor was the Institute de Protection et de Sufret6 Nucleaire ("IPSN"). For convenience, hereafter IRSN and IPSN are not distinguished and are referred to collectively as IRSN.

13

46.

Separately, at a conference in Aix-en-Provence, France in 2001, IRSN presented a calculation of the possible effect of fuel relocation on cladding temperature at the ruptured location on the fuel rod. That presentation is included in Exhibit 4 and Exhibit 29.6

47.

The 2001 IRSN presentation in Exhibits 4/29 involves a calculation, not a test.

Tr. 2459, 2482 (Lyman). The IRSN calculation suggests fuel relocation could increase cladding temperature at the ruptured location (not necessarily the PCT) by slightly less than 320'F (313 0F). The increase in local oxidation was 7% at the ruptured location. The IRSN calculation was for a filling fraction of 0.7. Duke Testimony, ¶ 153.

48.

The 2001 IRSN calculation presented in Exhibits 4/29 is for LEU fuel, not MOX fuel. It does not include any estimate of an additional relocation effect on cladding temperature or oxidation due to LEU-MOX fuel differences. Tr. 2482 (Lyman).

49.

The 2001 IRSN calculation presented in Exhibits 4/29 takes no credit for heat transfer and cladding cooling benefits associated with swelling and rupture of the cladding (which are demonstrated in the FR2 tests, illustrated in Exhibits 15 and 16). Tr. 2459-60 (Lyman); Duke Rebuttal Testimony, 1 73; see Exhibit 4, "Discussion" following paper and presentation material (first page, "Answer by C. Grandjean").

50.

The IRSN calculation of possible impact on cladding temperature due to LEU fuel relocation is referenced by the NRC Staff in Exhibit 27 (the "Thadani memorandum"),, but is not specifically endorsed by the NRC Staff. Exhibit 27, Attachment 5, is simply a list of reported calculations and results potentially germane to Appendix K. On its face, the Thadani memorandum considers relocation only as one of several issues to be considered if 6

Exhibit 29 was offered by BREDL. Exhibit 4 was offered by Duke. Exhibit 4 includes the "Discussion" following the authors' paper and presentation. That "Discussion" was not included in the BREDL Exhibit 29.

14

the NRC chooses to revise its regulations for LOCA analyses by removing conservatisms to make the regulations risk-informed. Exhibit 27, Attachment 4, at "Introduction;" see also Duke Testimony, ¶ 60-62, 96; Duke Rebuttal Testimony, ¶ 14.

51.

Notwithstanding Exhibit 27, Duke's MOX fuel LOCA analysis remains an Appendix K analysis under the current regulations and includes all of the inherent Appendix K conservatisms. Duke Testimony, 1 96; Duke Rebuttal Testimony 1 14 (see also the discussions at ¶¶ 24, 40 above).

52.

In contrast to the calculation presented by IRSN in 2001 (Exhibits 4/29), another calculation was presented at the same conference in Aix-en-Provence, France by Electricite de France ("EDF") and Framatome ANP. This presentation is Exhibit 3, offered by Duke. Using less conservative assumptions than IRSN, EDF calculated an increase in cladding temperature due to relocation of 540F, resulting in the presenter's conclusion that relocation need not be taken into account in safety regulation. Duke Testimony, ¶ 146. This calculation again relates to LEU fuel only. Tr. 2485-86 (Lyman).

53.

In total, the calculations related to fuel relocation presented or reported in Exhibits 3, 27, and 4/29, do not establish a LEU/MOX fuel difference that would be material to Contention I. They establish a range of possible impacts of fuel relocation for LEUfiuel, where the impact depends upon the assumptions used in the analysis. In contrast, from the FR2 tests illustrated in Exhibits 15 and 16, the detrimental effect of fuel relocation on PCT for LEU fuel appears to be more than offset by cooling benefits.

54.

Given that the Duke MOX fuel LOCA analysis utilizes an approved Appendix K evaluation model, that Duke has specifically addressed applicable differences between MOX fuel pellets and LEU fuel pellets in a conservative manner in that analysis, that the NRC does not 15

require specific modeling of relocation effects in Appendix K models, and that the relevant acceptance criteria have been satisfied, Duke has satisfactorily demonstrated that the MOX fuel lead assemblies meet the design basis LOCA requirements. Duke's sensitivity analyses also confirm no significant difference between MOX and LEU fuel for large break LOCA performance (Exhibit 1, Table 3-5), and no further findings on Contention I are necessary.

Nonetheless, the following findings further address fuel relocation issues.

D.

Fuel Relocation -

MOX Fuel

55.

Contention I focuses on the concern of BREDL's witness, Dr. Lyman, that fuel relocation may have a greater impact on LOCA performance for MOX fuel assemblies with M5Th' cladding. Lyman Testimony, A.4. Dr. Lyman, however, has not quantified the effect that he believes would be uniquely attributable to MOX fuel. Tr. 2461 (Lyman). His argument is a qualitative one -

that there are features of MOX fuel assemblies that could exacerbate the relocation effect relative to the LEU fuel assemblies in the core. Id. BREDL's argument therefore is not one based on specific evidence; rather it is based on speculation and claimed uncertainty.

56.

As previously noted, the IRSN calculation presented in Aix-en-Provence, France in March 2001, documented in Exhibits 4/29, does not include any calculation of change in cladding temperature for MOX fuel relative to LEU fuel. Tr. 2482 (Lyman).

57.

IRSN presented an updated calculation of relocation effects at Argonne National Laboratory in May 2004. This presentation is included in Exhibit 5 (offered by Duke) and Exhibit 30 (offered by BREDL). This IRSN calculation suggests a difference in fuel relocation impact on PCT between LEU fuel and MOX fuel of only 180F. Duke Testimony, ¶ 155. This is 16

a negligible difference in a LOCA calculation and supports Duke's position that the fuel relocation issue is not uniquely or even primarily a MOX fuel issue. Id.

58.

In presenting this MOX fuel calculation, IRSN attributed the 18'F MOX fuel increment to the higher initial stored energy in MOX fuel. Exhibit 5, at 8. The MOX fuel lead assemblies, however, will have lower initial stored energy than the LEU fuel assemblies in the Catawba core, due to operation at lower peaking. They will also have lower decay heat, due to the characteristics of MOX fuel. Duke Testimony, ¶¶ 137-142 (see also the discussion at m¶ 90-95 below). Based on these factors, the MOX fuel assemblies should involve a benefit, not a penalty, relative to LEU fuel and the potential impact of fuel relocation. Duke Testimony, 1 156.

59.

In the absence of tests or calculations to support the concern that MOX fuel and M5Tm cladding will exacerbate relocation effects relative to LEU fuel, Dr. Lyman raised only certain "qualitative features" of MOX fuel and M5T' cladding that, in his view, could involve a difference. Tr. 2461 (Lyman). Each of these features is discussed below. The weight of the evidence is that these "qualitative features" do not involve significant differences between MOX and LEU fuel and do not create significant uncertainties for a LOCA analysis or for public safety.

Filling Ratio

60.

Dr. Lyman's first "qualitative feature" involves a speculative theory that MOX fuel pellets may experience greater fragmentation than LEU fuel pellets, and thus MOX fuel relocation will involve greater "fill fractions" with commensurately greater fuel relocation effects. Lyman Testimony, ¶ A. 1 1.

.61.

In the opinion of Duke's expert, Dr. McCoy -

a qualified materials engineer-the breakup of either LEU or MOX fuel during a LOCA would occur by the same process and 17

should produce nearly identical pellet fragmentation. He anticipates that there should be no significant difference between MOX fuel and LEU fuel. Duke Testimony, ¶ 121. A filling ratio for MOX fuel less than 0.7 (the fraction assumed in the most extreme IRSN relocation calculation) would be expected. Id.

62.

Dr. McCoy's testimony is supported by micrographs of an irradiated LEU fuel pellet (Exhibit 21) and an irradiated MOX fuel pellet (Exhibit 22). The micrographs illustrate that the cracking and the cracking pattern -

which would lead to fragmentation -

are similar for the LEU fuel pellet and the MOX fuel pellet. Id., ¶¶ 116-118.

Specifically, the crack locations in the MOX fuel pellet are not influenced by the plutonium rich agglomerates unique to MOXfuel. Id.,

118.

63.

Dr. Lyman speculates that fragmentation may be influenced by the formation of a "rim" microstructure in the fuel. Lyman Testimony, ¶ A.1 1. He bases this only on observations in Exhibit 29 that finer fuel fragmentation is associated with higher burnups (i.e., those exceeding 40-45 GWD/t). Id. His argument is that development of a "rim" region also occurs at higher burnups; therefore, the "rim" region and fragmentation must be related. However, no mechanistic explanation is given for a correlation between development of a "rim" microstructure and fuel fragmentation. Duke Rebuttal Testimony, ¶ 26.

64.

In fact, BREDL's own exhibit, Exhibit 31, at 432, states that the "fuel relocation process is not specific of high burnup fuel. It was also observed for [sic] fuel rod having a burnup as low as 48 MWd/t." This is a very low burnup consistent with only a few days in the reactor. Tr. 2463-64 (Lyman). This undercuts the proposed correlation of fuel fragmentation to development of "rim" regions in high burnup fuel rods.

18

65.

Dr. Lyman's theory is that MOX fuel pellets and LEU fuel pellets differ because in MOX fuel pellets "high-burnup rim-like regions" will emerge in the outer layers of the plutonium agglomerates distributed throughout the pellet. Lyman Testimony, I A. 11. In LEU fuel pellets, the "rim" microstructure appears first at the extreme periphery (i.e., the "rim") of the fuel pellet. Duke Rebuttal Testimony, 1 25. According to Dr. Lyman, and assuming a "rim" region/fuel fragmentation correlation, there could therefore be more fragmentation in MOX fuel.

Lyman Testimony, ¶ A. 1 1.

66.

Dr. McCoy testified that "rim" regions are tougher and more resistant to cracking.

Duke Rebuttal Testimony, 1 27. His expert opinion is that "rim" regions will not contribute to a loss of mechanical integrity or increase the susceptibility to relocation. Id. In fact, the extra toughness imparted by the formation of the "rim" microstructure supports the conclusion that plutonium agglomerates will not yield fine particles. Id., ¶ 31. Dr. Dana Powers of the ACRS has also stated that MOX fuel would have inherently a low fragmentation tendency. Id., ¶ 32.

67.

Dr. Lyman speculates that, because MOX fuel has a lower thermal conductivity and a higher radial temperature gradient than LEU fuel, it would experience greater fuel fragmentation due to thermal shock in a LOCA. Lyman Testimony, ¶ A.4. However, Exhibit 10 shows that the difference between MOX and LEU fuel in this regard is small and does not change significantly with burnup. Duke Testimony, 1 68. Dr. McCoy explained that there will be no significant differences in thermal gradients or thermal stresses between MOX fuel and LEU fuel, and the effect of these stresses on fragmentation will be similar in the two types of pellets. Duke Rebuttal Testimony, ¶ 34.

68.

The NRC Staff expert, Dr. Meyer, testified that the volume of "rim" material might be 25% greater in MOX fuel than in LEU fuel. Staff Testimony, 1 39; Tr. 2644 (Meyer).

19

But he further testified that the MOX material would be more "plastic" than the LEU material and, therefore, could form fewer particles. Tr. 2655 (Meyer).

69.

If fine particles are formed, a substantial portion of these particles would not be retained in the ruptured cladding balloon and therefore would not increase the filling fraction or affect LOCA performance. Staff Testimony, I A.40; Duke Rebuttal Testimony, ¶ 35. Dr. Meyer testified that this effect is illustrated by tests at Argonne National Laboratory documented in Exhibit 40. Fine fragments were not present in the fuel rods after the tests, apparently because many were blown out of the rod at the time of rod depressurization. Tr. 2624-28 (Meyer).

70.

Given anticipated cladding strains for Catawba and the available space within the fuel rod for relocation, filling ratios for MOX fuel are likely to be 0.5 to 0.6, and in any event are unlikely to meet or exceed the 0.7 assumed in the limiting IRSN calculations. Tr. 2639 (Meyer);

Tr. 2642-43 (Meyer).

71.

Dr. Lyman referenced the CABRI tests reported in Exhibit 51 to support his argument that fuel fragmentation may be greater in MOX fuel pellets. However, these tests involved fuel behavior under design basis "reactivity initiated accident" conditions.7 Exhibit 51, at 1. These test conditions are very different from a LOCA and the test results are not representative of LOCA fuel behavior or LOCA fuel fragmentation. Tr. 2672-75 (Meyer).

72.

In sum, no persuasive quantitative or qualitative evidence supports the BREDL assertion that there will be a significant difference in fragmentation and filling ratio between MOX and LEU fuel. In fact, the weight of the evidence supports a conclusion that there will be no significant difference.

Moreover, even if such a difference existed, the weight of expert 7

While "reactivity initiated accident" is the terminology used in the exhibit, this is also commonly referred to in the United States as a "reactivity insertion accident."

20

opinion is that any fine particles that might be created in MOX fuel rods will not impact LOCA performance through fuel relocation effects.

M5TM Cladding Ductility

73.

Dr. Lyman's second "qualitative feature" involves his speculative theory that fuel relocation effects will be more severe because of the properties of the M5Th fuel cladding.

Theoretically, if cladding strain (i.e., ballooning) is increased due to M5TM properties, there would be more room for relocation and therefore there could be more severe impacts on cladding temperature and oxidation. His specific concern is focused on the "greater retained ductility" of M5ThM as compared to Zircaloy-4 cladding. Lyman Testimony, ¶ A.13.

74.

This "qualitative feature" does not involve a MOX-LEU fuel difference.

Cladding ballooning is an M5T'I cladding issue, not a MOX fuel issue cognizable under Contention 1. Cladding ballooning is unaffected by the fuel pellets inside.

Staff Rebuttal Testimony, I A.9. M5T cladding has been reviewed by the NRC and approved for use in fuel assemblies in the United States. Duke Testimony, 1 48.

75.

The LOCA evaluation model for the MOX fuel lead assemblies incorporated M5'-specific properties in the evaluation of fuel cladding ballooning and rupture. Id., ¶1 48-54.

Consistent with Appendix K, the model used unirradiated cladding properties to maximize the predicted strain (strain decreases with irradiation of the fuel rod). Id., ¶ 54. The maximum cladding strain in the limiting case was 51 percent. As noted above, this is well within the coolable geometry limit. Id., ¶ 56.

76.

Relative to Zircaloy-4, M5Tm ductility does not decrease as much with irradiation as Zircaloy4 (that is, M5Th! cladding has greater retained ductility than Zircaloy-4). Id., ¶ 113.

Nevertheless, M5Tm is most ductile in the unirradiated state; the use of unirradiated M5Tm 21

properties in the LOCA analysis therefore maximizes the extent of ballooning considered. Duke Rebuttal Testimony, 1 58.

The NRC Staff concluded that ballooning size was adequately accounted for in the MOX fuel LOCA analysis. Staff Rebuttal Testimony, ¶ A.9.

77.

Dr. Lyman is concerned that the greater retained ductility of M5Tm relative to Zircaloy-4 would increase fuel relocation impacts of M5rm fuel rods relative to Zircaloy-4 fuel rods. Lyman Testimony, ¶ A.13. Again, however, the comparison to Zircaloy-4 rods is not a relevant comparison. It is not a MOX fuel-LEU fuel difference cognizable under Contention I.

78.

In any event, the difference between M5Tn and Zircaloy-4 is not significant.

M5Tm experiences less strain at rupture than Zircaloy-4 in the unirradiated state, but the two materials have approximately equal strain potential near the end of irradiation (that is, the ductility v. irradiation curves converge).

Duke Testimony, ¶ 113.

There is little expected difference in the consequences of fuel relocation due to cladding differences. Id., ¶ 114.

79.

A presentation by EDF in May 2004 at Argonne National Laboratory, introduced by the NRC Staff as Exhibit 41, provides data from rod burst tests using increasing temperatures that are typical of a LOCA. This "ramp" testing specifically shows that M5Th! cladding actually does not develop larger balloons than Zircaloy-4 under LOCA conditions. Staff Testimony, A.35.

80.

Dr. Lyman accepted that the EDF ramp tests in Exhibit 41 are more similar to the conditions experienced during a LOCA than are previous steady-state creep tests.

Lyman Testimony, ¶ A.14.

However, he questions whether the testing takes into account other inhomogeneities in Zircaloy-4 that could negatively impact ductility, potentially leading to smaller balloon sizes. Id. This is, however, speculation. Moreover, from operational and safety 22

perspectives, it would not be advisable to use inferior cladding to compensate for a speculative uncertainty. Duke Rebuttal Testimony, 1 59; see also Staff Rebuttal Testimony, ¶ A.9.

81.

In sum, no persuasive quantitative or qualitative evidence supports the BREDL assertion that there will be a significant difference in relocation impacts between MOX and LEU fuel due to M5S' cladding. The weight of the expert opinion is that there would be little difference in relocation effects between fuel rods with M5T' and Zircaloy-4 cladding.

Fuel-Cladding Interaction

82.

The next "qualitative feature" cited by Dr. Lyman and potentially suggesting differences between MOX and LEU fuel involves speculation on potential effects of fuel pellet-to-cladding interaction. The conjecture is that "tight fuel-clad bonding may delay the onset of fuel relocation." Lyman Testimony, ¶ A.12. Dr. Lyman cites a recent Nuclear Energy Agency

("NEA") report (Exhibit 34) as confirming that MOX fuel is more resistant to cladding failure due to pellet-to-cladding mechanical interaction ("PCMI") than LEU fuel. While acknowledging that this phenomenon "is not well-understood," he speculates that it "may imply that the pellet-clad bond is weaker for MOX fuel, in which case MOX fuel may have a greater propensity to earlier and more excessive fuel relocation than LEU fuel." Id.

83.

Exhibit 34, the March 2004 NEA report offered by BREDL, makes no suggestion that pellet-to-cladding chemical bonding is a possible explanation for differences that may exist between the PCMI performance of MOX and LEU fuels. Duke Rebuttal Testimony, m¶ 41-42.

84.

BREDL subsequently offered proposed Exhibit C (for identification) to support Dr. Lyman's theory. Proposed Exhibit C is merely a description of issues to be discussed at the workshop that was later the subject of Exhibit 34. Duke Supplemental Rebuttal Testimony, A¶ 6-

7. Proposed Exhibit C has no evidentiary value.

23

85.

Duke's expert, Dr. McCoy, testified that the processes that lead to any chemical bonding are the same for LEU and MOX fuels and the bond strengths are similar because of similarities in fuel chemistry and in operating conditions. He expects that pellet-to-cladding bonding will have similar effects (if any effect at all) on LOCA performance in MOX and LEU fuels. Duke Testimony, ¶ 123.

86.

Duke's experts also suggested that any fuel pellet-to-cladding bonding might be beneficial in controlling cladding ballooning and relocation. Id., TI 124-128. This unquantified beneficial effect is not credited in any MOX LOCA analysis. It is also not credited in the calculations of possible relocation effects performed by IRSN and EDF. Id., ¶ 128.

87.

The NRC Staff expert, Dr. Meyer, disputed Dr. Lyman's correlation of the additional resistance of MOX fuel to cladding failure by PCMI to fuel-to-cladding chemical interaction on bonding. He attributes the difference to the greater plasticity of MOX pellets.

Staff Rebuttal Testimony, I A.7. This characteristic has nothing to do with bonding between the pellets and the cladding. Id.

88.

To the extent any pellet-to-cladding bonding exists, the NRC Staff expert also provided his opinion, based on LOCA tests at Argonne National Laboratory, that there will be no effect of such bonding on the size of the balloon. The balloon is a function of the cladding, regardless of whether the fuel pellet is MOX or LEU. Tr. 2641-42 (Meyer).

89.

In sum, no persuasive quantitative or qualitative evidence supports the theory of a MOX-LEU difference with respect to pellet-to-cladding bonding that would affect the degree of fuel relocation. The weight of the expert testimony supports the conclusion that MOX fuel and LEU fuel will be similar in this regard, and that there will be no negative MOX effect on fuel relocation.

24

MOXFuel Relative Powver at High Burnup

90.

BREDL has also hypothesized, based on the October 2003 IRSN presentation to the NRC Staff (Exhibit 28), that fuel relocation in the MOX fuel assemblies may have a more significant impact than relocation in LEU fuel because the power in MOX fuel is higher at end-of-life than the power in LEU fuel. The theory is that, with fuel operating at higher power, the relocated fuel fragments would have a higher decay heat power, contributing to higher cladding temperatures. Duke Testimony, ¶¶ 129-130.

91.

'The IRSN statement in Exhibit 28 (slide 21) about higher end-of-life power of MOX fuel relative to LEU fuel is based on European MOX fuel experience and does not apply to the Catawba MOX fuel lead assemblies. The Catawba MOX fuel lead assemblies will operate at lower power at end-of-life than LEU fuel. Id., ¶ 131.

92.

The effect noted by IRSN (a tendency for higher end-of-life power for MOX fuel) is also a characteristic of European RG MOX fuel.

In contrast, with respect to neutronic characteristics, WG MOX fuel such as will be used at Catawba is much closer to LEU fuel than is RG MOX fuel. Id., ¶¶ 132-136 (see Exhibit 23).

93.

Dr. Lyman cited Exhibit 30, the May 2004 IRSN presentation at Argonne National Laboratory (slides 8-9), for the proposition that for MOX fuel there will be "higher initial energy" that would enhance fuel relocation impacts. Lyman Testimony, ¶ A.10. The Catawba lead assemblies, however, as noted previously, will be placed in non-limiting core locations. The MOX fuel lead assembly relative power will be lower than the maximum LEU assembly relative power at all times. Duke Testimony, T1 137-140.

94.

The MOX fuel decay heat also will be 3-5% lower than the decay heat from a LEU assembly with the same burnup and same operating power level. This is a beneficial 25

difference for MOX fuel not credited by IRSN. Id., T¶ 63, 141. As noted previously, as an analysis conservatism, the Duke MOX fuel LOCA analysis uses the LEU decay heat model. Id.,

63.

95.

In sum, there is no persuasive quantitative or qualitative evidence to support BREDL's assertion that the MOX fuel lead assemblies will have more significant relocation effects because of greater relative power. In fact, the evidence shows the opposite. The MOX fuel lead assemblies will operate at a lower relative power than co-resident LEU fuel, and MOX fuel has lower decay heat than LEU fuel. These factors will be beneficial for MOX fuel with respect to fuel relocation.

E.

Assessment of Relocation Impacts

96.

The finding of facts above provide an adequate basis on which to conclude that BREDL has not demonstrated significant differences between MOX and LEU fuel behavior (specifically with respect to the fuel relocation phenomenon) to cause significant impact on a LOCA or a LOCA analysis. Nonetheless, the parties have also addressed the issue of Contention I in terms of the (a) conservatisms built into the LOCA analysis and (b) the hypothetical impacts of relocation on the LOCA analysis results and acceptance criteria.

Each of these two approaches is discussed below.

Stacking the Conservatisms

97.

The conservatisms in the Appendix K LOCA model and in Duke's MOX fuel LOCA analysis more than compensate for any uncertainty created'by the relocation issue, particularly any relocation impacts attributable to differences between MOX fuel and LEU fuel (assuming for argument that such differences and impacts even existed). See generally Duke Testimony, ¶¶ 60-65, 158-161; Duke Rebuttal Testimony, ¶¶ 11, 75-76.

26

98.

There are two types of conservatisms: first there are those inherent in the Appendix K methodology; and second, there are those additional conservatisms incorporated specifically into Duke MOX fuel LOCA analyses. Duke Testimony, 1 60.

99.

Major conservatisms inherent in the approved Appendix K model are:

A decay heat approximately 10 percent higher than the 95 percent confidence level and 30 to 35 percent higher than the nominal expectation. (This is separate from the decay heat conservatism for MOX fuel discussed below.)

The use of the Baker-Just oxidation (metal-water reaction) correlation, which incorporates a reaction rate to 30 to 50 percent above the available data.

The use of the highest allowable local power level. Plants actually operate with peaking significantly lower than the limiting values defined to be acceptable by the LOCA analyses.

(This conservatism was quantified for representative Catawba operation cycles in the Duke Testimony, m 74, 76-77.)

The use of full double-ended break areas. The NRC is currently reviewing the use of substantially reduced break areas based on the risk of the large break occurring.

Leak before break analyses indicated that the probability of an instantaneous double-ended break is very low.

Limiting Emergency Core Cooling System ("ECCS") bypass assumptions. Nearly all of the ECCS water delivered to the reactor coolant system prior to the end of blowdown is assumed to be bypassed, and therefore unavailable for core cooling.

Id., ¶ 61.

As previously noted, the overall effect of conservatisms in the Appendix K methodology has been estimated to be 6000F margin relative to best estimate LOCA analysis results. Id., ¶ 62.

100.

Additional conservatisms in the MOX fuel lead assembly large break LOCA analysis include:

The use of the LEU decay heat model. MOX fuel falls 3 to 5 percent below LEU fuel decay heat during the time period of 27

importance. This has been estimated to provide a conservatism of up to 750F on PCT. Id., ¶ 63.

The use of LEU fuel neutronics coefficients. The magnitude of this conservatism was not calculated. Id., ¶ 64.

Given the core designs for the Catawba cores with the MOX fuel lead assemblies, the lead assemblies will not be the peak power assemblies in the core. Id., % 73, 137-140; Tr. 2373 (Harvey).

101.

BREDL points to one factor not included in the Appendix K LOCA evaluation model as a potential non-conservatism (fuel relocation).

In sum, however, any uncertainty regarding the impact of that issue -

even IRSN's calculated results for LEU fuel -

is clearly bounded by the known conservatisms in the Duke Appendix K evaluation approach that are included precisely for the purpose of addressing analysis uncertainties. Duke Testimony m¶ 158-160; Duke Rebuttal Testimony, 1 74; Staff Rebuttal Testimony, ¶ A.5. These conservatisms provide reasonable assurance that any uncertainty associated with fuel relocation does not pose an undue safety risk.

102.

Experimental data presented by Duke actually show that relocation effects for LEU fuel are negligible because of the offsetting cooling effects associated with cladding swelling and rupture. Exhibits 15 and 16. Moreover, there are no data in the record suggesting a MOX incremental impact on PCT. The NRC Staff has concluded that the uncertainties related to MOX fuel are "adequately understood." Staff Rebuttal Testimony ¶ A.12.

Running the Numbers 103.

Dr. Lyman has attempted to assess the importance of the relocation issue by adding a series of penalties to Duke's MOX fuel LOCA analysis results. Lyman Testimony, ¶ A. 16; Lyman Rebuttal Testimony, ¶ A.R-3. This approach is not technically accurate because it involves "cherry picking" a relocation effect from one model (IRSN's calculation, for example) and arbitrarily adding only that one effect to an Appendix K licensing calculation that 28

intentionally already includes compensating conservatisms. Tr. 2669-70 (Meyer). At best, this approach provides a crude estimate of a limiting impact of fuel relocation. Id.

104.

Given that the Appendix K evaluation model does not include relocation for LEU fuel, and includes other conservatisms discussed above, a more relevant analysis for MOX fuel assemblies -

and more precisely for measuring the difference between LEU and MOX fuel, which is the subject of Contention I -

would simply add a relocation impact based on the difference in impact on cladding temperature between MOX fuel and LEU fuel. As discussed above, no such difference between MOX fuel and LEU fuel has been demonstrated.

The difference appears to be zero. Tr. 2670 (Meyer).

105.

In any event, the correct baseline for Dr. Lyman's additive analysis is not 2018'F as suggested in his initial testimony (¶ A.16), nor is it the 1841'F suggested in his rebuttal testimony (¶ A.R-3). The appropriate baseline is the highest ruptured node temperature (because the relocation effect occurs at the ruptured location) from the Appendix K LOCA analysis that formed the basis for the MOX fuel lead assembly LOCA limits (Exhibit 2, Table Q14-1). This is 1750'F. Duke Testimony, ¶ 152, 154; Duke Rebuttal Testimony, 1 73.

106.

Dr. Lyman would add to this baseline ruptured node temperature the 313F LEU fuel relocation penalty calculated by IRSN in 2001 (Exhibits 4/29). Lyman Rebuttal Testimony,

¶ A.R-3. As previously noted, this IRSN estimate is not supported by any experiment. It makes no allowance for cooling benefits associated with cladding swelling and rupture. Duke Rebuttal Testimony, ¶ 73. It also assumes a conservative 0.7 filling fraction. Duke Testimony, ¶ 153.

Therefore, it was characterized by the Duke expert, Mr. Dunn, as a very "pessimistic" number.

Id., ¶ 152. The NRC Staff expert would also view it as an "upper bound." Tr. 2669 (Meyer).

29

107.

A 300+0F relocation penalty cannot be derived by comparing numbers from the FR2 experimental results in Exhibit 15 (no relocation case) and Exhibit 16 (relocation case). The two exhibits reflect two different tests at two different sets of conditions (e.g., different heatup rates or power) and therefore comparisons across the two tests are not technically valid. Tr.

2337-39 (Dunn, Nesbit). In fact, as noted previously, the exhibits actually show that detrimental relocation effects are a "near match" to beneficial heat transfer effects. Duke Testimony, ¶ 150.

108.

Dr. Lyman next would add to the baseline an arbitrary 200F penalty supposedly derived from Exhibit 16. Lyman Rebuttal Testimony ¶ A.R-3. The penalty, however, is not technically valid. It is based on the faulty assumption that the PCT (usually at a non-ruptured location) will always be 200F higher than the maximum cladding temperature at the ruptured location. The PCT at a non-ruptured location is independent of the ruptured node temperature.

Tr. 2340-41 (Nesbit).

109.

Dr. Lyman does not, in his calculation, include any MOX fuel analysis conservatisms, such as the conservatism of up to 750F on PCT associated with Duke's use of a LEU fuel decay heat model in the MOX fuel LOCA analysis. Dr. Lyman agreed that it would be appropriate to do so. Tr. 2510 (Lyman). The NRC Staff witness also agreed that, if it is appropriate to add a relocation penalty, it is equally appropriate to add this MOX fuel decay heat benefit. Tr. 2671 (Meyer).

110.

Dr. Lyman next argued that any "additional MOX effects" (such as a greater filling ratio) should be added. Lyman Rebuttal Testimony, ¶ A.R-3. However, no such number is offered (only the "qualitative features" discussed above). The NRC Staff expert expects the number to be zero. Tr. 2670 (Meyer). Moreover, it should be noted that the IRSN calculation 30

for LEU fuel relocation of 3130F already includes a relatively high filling ratio of 0.7. Lyman Testimony, I A. 16. Therefore, there is no basis for an additional MOX penalty.

111.

Adding 17500F and 313'F yields 20630F. This would be a maximum cladding temperature at the ruptured location that is greater than the PCT for the limiting MOX fuel LOCA analysis (2019.50F), but still well below the acceptance criterion (22000F).

Duke Testimony, ¶¶ 154, 55.

(Duke uses 3200F as the approximation of the IRSN relocation calculation in its testimony.) Taking credit for the undisputed decay heat model conservatism of up to 750F would reduce the total to below 20000F, which is bounded by the Duke MOX fuel LOCA analysis PCT (2019.57F). Although no margin is required, this result also provides approximately 2000F in margin to the acceptance criterion.

112.

The same approach can be applied to the oxidation acceptance criterion. Local oxidation in Duke's limiting LOCA analysis case is 5.2%. Duke Testimony, ¶ 55. However, the local oxidation on the ruptured fuel pin is 3.0%. Id., $ 154. IRSN calculated an increase in local oxidation of 7.0% at the ruptured location. Id., at ¶ 153. Even with this "pessimistic" prediction, a total estimated oxidation of 10.0% would result -

which is still less than the acceptance criterion of 17%. Id.,¶ 1 54.8 113.

Dr. Lyman on rebuttal hypothesized that there should be further conservatism included in the oxidation assessment because (a) maximum cladding oxidation typically occurs at the ruptured location and (b) an increase in cladding temperature at that location will increase the oxidation. Lyman Rebuttal Testimony, I A.R-3. However, even calculating oxidation at the ruptured location using the higher temperatures assumed at that location when a relocation 8

The NRC Staff expert, Dr. Meyer, also agreed with this assessment that oxidation would not exceed 17%. Tr. 2636-38 (Meyer). Although the Staff withdrew his reference to an independent calculation to verify his opinion, his opinion based on experience remains valid.

31

penalty is added, the Duke expert, Mr. Dunn, anticipated the local oxidation to be still less than 11.0%. Tr. 2451-53 (Dunn).

114.

In sum, this upper bound estimate approach provides further assurance that the deterministic criteria for PCT and local oxidation of 10 C.F.R. § 50.46 would be satisfied even if fuel relocation effects were considered for the MOX fuel assemblies. See also Staff Testimony,

¶ A.42.

F.

Credibility Determinations 115.

The witnesses presented by Duke and the NRC Staff demonstrated their directly relevant experience.

They were clearly well-versed in the relevant technical and regulatory issues and presented their case in a cogent, consistent, and articulate manner. The facts and expert opinions they presented are clearly entitled to great weight and deference.

116.

Dr. Lyman, on the other hand, demonstrated only limited experience related to LOCA analyses, materials science and engineering, and the other areas in which he presented views and arguments. He appeared to create technical advocacy positions by reviewing the technical literature and extracting discrete points with little regard for the original context of those points. Some specific examples follow.

117.

The initial proposed contention cited the VERCORS tests from the IRSN presentation to the NRC Staff from October 2003. However, as previously noted, those tests were at severe accident conditions and do not provide evidence of fuel relocation at LOCA conditions (see the discussion at ¶ 45 above).

118.

Dr. Lyman's initial testimony cited a baseline cladding temperature for adding MOX fuel penalties of 2018F. Lyman Testimony, ¶ A.16. Dr. Lyman did not recognize that 32

the relocation effect was at the ruptured location, which is not the site of the PCT in the LOCA analysis.

119.

Dr. Lyman in his rebuttal testimony erroneously drew correlations from Exhibits 15 and 16, without recognizing that the test conditions in the two exhibits were not directly comparable.

Moreover, as noted at 1 108 above, he extracted an invalid 20'F penalty from Exhibit 16. PCT is not quantitatively a function of maximum temperature at the ruptured node.

Tr. 2340-41 (Nesbit).

120.

Dr. Lyman attempted to draw conclusions regarding fuel fragmentation during a design basis LOCA by referencing Exhibit 51 (the CABRI tests). However, as noted at ¶ 71 above, these tests involved fuel behavior under design basis reactivity initiated accident conditions and are not representative of LOCA fuel behavior. Tr. 2672-75 (Meyer).

121.

Dr. Lyman in his testimony attempted to associate fuel relocation with the development of a high-burnup "rim" region as noted above. Lyman Testimony, I A.1 1. He cited observations of fuel relocation in LEU fuel with rod burnups exceeding "around 48 GWD/t." Id. However, as noted at 1 64 above, BREDL's own exhibit, Exhibit 31, states that the fuel relocation process is not specific to high burnup fuel, and has been observed for a "fuel rod having a burnup as low as 48 MWD/t." Dr. Lyman acknowledged this as a "mistake." Tr. 2463 (Lyman). Beyond that, however, because fuel with burnup of 48 MWD/t is practically fresh fuel, Exhibit 31 demonstrates that the supposed correlation of fuel fragmentation/relocation to development of "rim" regions in high burnup or MOX fuel is pure conjecture.

122.

In total, Dr. Lyman is presenting an "uncertainty" theory in Contention I in which the "uncertainty" is a moving target. As one "uncertainty" is addressed, a new one is raised based on reviewing literature that often is not applicable.

This seriously undermines Dr.

33

Lyman's technical credibility and the weight that can be assigned to his opinions. His qualitative speculation of differences between LEU and MOX fuel fails to support Contention I. The expert testimony of the Duke and NRC Staff witnesses provide reasonable assurance that the MOX fuel assemblies meet the design basis LOCA requirements.

III.

CONCLUSIONS OF LAW 123.

Based upon the complete evidentiary record and the findings of fact and credibility determinations set forth above, Contention I should be resolved on the merits in favor of the applicant, Duke (supported by the NRC Staff), and against the intervenor, BREDL.

124.

With respect to the issues raised in Contention I, Duke and the NRC Staff have specifically shown by a preponderance of the credible evidence that there are no significant differences between MOX and LEU fuel behavior that would impact fuel performance in a LOCA, the LOCA design basis accident analysis, or compliance with applicable NRC regulations.

125.

With respect to the issues raised in Contention I, Duke and the NRC Staff have specifically shown by a preponderance of the credible evidence that the proposed license amendment should be granted in accordance with 10 C.F.R. § 50.92(a).

Consistent with the general standards of 10 C.F.R. §§ 50.40(a) and 50.57(a)(3), there is reasonable assurance that operation of Catawba with four MOX fuel lead assemblies as described in the LAR will present no undue risk to public health and safety and will be consistent with NRC regulations.

126.

All issues, arguments, testimony, or exhibits presented by the parties but not addressed herein should be found to be without merit or unnecessary for issuance of a Partial Initial Decision on Contention I.

127.

In accordance with 10 C.F.R. §§ 2.760(a) and 2.764(a) the Licensing Board's Partial Initial Decision should be effective immediately and should constitute final action of the 34

NRC within forty (40) days of the date of issuance unless a petition for review is filed in accordance with 10 C.F.R. § 2.786(b) or the Commission directs otherwise. Any petition for review must be filed within fifteen (15) days after service of the Partial Initial Decision and shall conform with the requirements of 10 C.F.R. §§ 2.786(b)(2)-(3) and must be based on the grounds specified in 10 C.F.R. § 2.786(b)(4). In accordance with 10 C.F.R. § 2.786(b)(1), a petition for review is mandatory in order for a party to have exhausted its administrative remedies before seeking judicial review.

Respectfully submitted, David A. Repka WINSTON & STRAWN, LLP 1400 L Street, NW Washington, D.C. 20005-3502 Timika Shafeek-Horton DUKE ENERGY CORPORATION P.O. Box 1006 Mail Code: EC11X-1128 Charlotte, N.C. 28201-1006 ATTORNEYS FOR DUKE ENERGY CORPORATION Dated in Washington, District of Columbia This 6'h day of August 2004 35

UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of:

DUKE ENERGY CORPORATION (Catawba Nuclear Station, Units 1 and 2)

)

)

)

)

)

)

)

)

Docket Nos.

50-413-OLA 50-414-OLA CERTIFICATE OF SERVICE I hereby certify that copies of "DUKE ENERGY CORPORATION'S PROPOSED FINDINGS OF FACT AND CONCLUSIONS OF LAW REGARDING CONTENTION I" in the captioned proceeding have been served on the following by deposit in the United States mail, first class, this 6th day of August, 2004. Additional e-mail service, designated by *, has been made this same day, as shown below.

Ann Marshall Young, Chairman*

Administrative Judge Atomic Safety and Licensing Board U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 (e-mail: AMY~nrc.gov)

Thomas S. Elleman*

Administrative Judge 5207 Creedmoor Road, #101 Raleigh, NC 27612 (e-mail: ellemangeos.ncsu.edu)

Office of Commission Appellate Adjudication U.S. Nuclear Regulatory Commission Washington, DC 20555 Anthony J. Baratta*

Administrative Judge Atomic Safety and Licensing Board U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 (e-mail: AJB5@nrc.gov)

Office of the Secretary*

U.S. Nuclear Regulatory Commission Washington, DC 20555 Attn: Rulemakings and Adjudications Staff (original + two copies)

(e-mail: HEARINGDOCKET(nrc.gov)

Adjudicatory File Atomic Safety and Licensing Board Panel U.S. Nuclear Regulatory Commission Washington, DC 20555 1

Susan L. Uttal, Esq.*

Antonio Fernandez, Esq.*

Margaret J. Bupp*

Office of the General Counsel U.S. Nuclear Regulatory Commission Washington, DC 20555 (e-mail: slu(nrc.gov)

(e-mail: axf2(inrc.gov)

(e-mail: mjb5(nrc.gov)

Diane Curran*

Harmon, Curran, Spielberg &

Eisenberg, LLP 1726 M Street, N.W.

Suite 600 Washington, DC 20036 (e-mail: dcurran(harmoncurran.com)

David A. Repka Counsel for Duke Energy Corporation 2

DC:367244.1