LIC-02-0054, Annual Report for 2001 Loss of Coolant Accident (Loca)/Emergency Core Cooling System (ECCS) Models Pursuant to 10CFR50.46

Annual Report for 2001 Loss of Coolant Accident (Loca)/Emergency Core Cooling System (ECCS) Models Pursuant to 10CFR50.46

ML021350236
Person / Time
Site:	Fort Calhoun
Issue date:	04/30/2002
From:	Ridenoure R Omaha Public Power District
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
LIC-02-0054
Download: ML021350236 (7)

Text

Omaha Public Power District 444 South 16th Street Mall Omaha NE 68102-2247 April 30, 2002 LIC-02-0054 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, D.C. 20555

References:

1. Docket 50-285

2. XN-NF-82-49(P)(A), Supplement 1, "Exxon Nuclear Company Evaluation Model Revised EXEM PWR Small Break Model", Revision 1, December 1994

3. EMF-2087(P)(A), "SEM/PWR-98: ECCS Evaluation Model for PWR LBLOCA Applications", Revision 0, June 1999

4. Letter from OPPD (R.T. Ridenoure) to NRC (Document Control Desk),

"Report of Significant Change/Error in the Large Break Loss of Coolant (LOCA)/Emergency Core Cooling (ECCS) Models and Evaluations Pursuant to 10 CFR 50.46", dated March 22, 2002 (LIC-02-0035)

SUBJECT:

Annual Report for 2001 Loss of Coolant Accident (LOCA)/Emergency Core Cooling System (ECCS) Models Pursuant to 10 CFR 50.46 In accordance with 10 CFR 50.46(a)(3)(ii), the Omaha Public Power District (OPPD) is submitting the annual 10 CFR 50.46 summary report for 2001. This summary report updates all identified changes or errors in the LOCA/ECCS codes and methods used by Framatome ANP to model Fort Calhoun Station Unit No. 1 (FCS). References 2 and 3, respectively, describe the Small Break (SB) and Large Break (LB) LOCA analysis methodology used by Framatome ANP for modeling plants, such as FCS.

OPPD has received the Framatome ANP 10 CFR 50.46 Annual Notification Report for 2001 for the SB and LB LOCA Analyses that are subject to the reporting requirements of 10 CFR 50.46.

For 2001 there were two SB LOCA Analysis Peak Clad Temperature (PCT) 10 CFR 50.46 Model Assessment errors of 0 'F and -1 'F. These errors are described in Attachment 1, for Errors Discovered during RODEX2 Verification and Validation (V&V) and an Error in the TOODEE2 Clad Thermal Expansion. Attachment 2 provides the 2001 SB PCT Margin Utilization Summary for Fort Calhoun Station. As a result of the -1 TF total of errors, the SB LOCA PCT changed from the baseline value (reported in the FCS Updated Final Safety Analysis Report) of 1865 TF to 1864 TF.

Employment with Equal Opportunity 4171

U. S. Nuclear Regulatory Commission LIC-02-0054 Page 2 In Reference 4 OPPD reported that the sum of the absolute values of errors/changes in the LB LOCA analysis of record (as contained in the FCS Updated Final Safety Analysis Report) exceeded 50 °F per 10 CFR 50.46(a)(3)(i). Subsequent to the Reference 4 Significant Error Report, there were two additional LBLOCA Analysis PCT 10 CFR 50.46 Model Assessment errors of +2 °F and +1 OF. These errors are described in Attachment 3, respectively, for an Incorrect Pump Junction Area used in RELAP4 LBLOCA Blowdown Analysis and an Incorrect Fuel Density Input to RSD2D2 LBLOCA Radiation Calculation. Attachment 4 provides the 2001 Large Break PCT Margin Utilization Summary for Fort Calhoun Station. As a result of these two errors the Large Break LOCA PCT changed from the Reference 4 value of 1956 OF to 1959 °F. The Large Break LOCA has been reanalyzed for the conditions associated with Cycle 21 operation, currently scheduled to begin in June 2002. The new baseline PCT value will be 1956 °F upon Cycle 21 startup.

In summary, the FCS PCT values for SB and LB LOCAs remain less than the 10 CFR 50.46(b)(1) acceptance criterion of 2200 OF.

Please contact me if you have any questions.

Sincerely, nMoDivis ager Nuc ar Ope ations c: E. W. Merschoff, NRC Regional administrator, Region IV A. B. Wang, NRC Project Manager W. C. Walker, NRC Senior Resident Inspector Winston & Strawn

LIC-02-0054 Page 1 Attachment 1 10 CFR 50.46 Small Break Model Assessment Errors Discovered during RODEX2 V & V The RODEX2 code system is used by Framatome ANP to determine the initial fuel stored energy and gap conditions for the initialization of the system blowdown and hot rod response calculations for their LOCA analysis.

Framatome ANP, in their SB LOCA Rodex2 (i.e., RDX2LSE and RODEX2-2A) model, actually used a smaller fraction of dish volume than the NRC approved 75% of the dish volume to accommodate gaseous swelling. This difference dates back to the mid-1980's to a mistake in a Rodex2 equation that caused the code to use an incorrect dish volume when calculating the allocation for swelling accommodation. Taking 75% of too small of a volume resulted in 57.4%

of the actual dish volume being available to accommodate swelling, with NRD (the number of radial nodes used to model the central dished part of the pellet) equal to 8. The error is larger for values of NRD less than 8.

The mistake was corrected in the predecessor to Rodex2-2A, and the swelling accommodation allocation was reduced from 75% to 57.4% in order to maintain agreement with the benchmark data, since the model had already been tuned to that data. However, the individuals making the change failed to change the value in the document.

The RDX2LSE model equation used to allocate the dish volume for swelling accommodation has a mistake that causes a large sensitivity to NRD. As NRD is reduced, the underestimation in the dish volume increases, and the volume allocated decreases. This could affect predicted temperature and gap conductance at moderate-to-high burnups. This error has been traced back to the oldest version currently available, and it appears to come from the original Rodex2 version from which RDX2LSE was branched. It was corrected in RODEX2-2A.

Framatome ANP created a development version of RDX2LSE for evaluation of the impact. In this version, the equation used to allocate dish volume has been corrected to not underpredict volume for lower values of NRD.

The nature of this error leads to a bounding 0 'F change in the calculated PCT for Fort Calhoun Station.

LIC-02-0054 Page 2 Attachment 1 (Cont'd) 10 CFR 50.46 Small Break Model Assessment Error in TOODEE2 Clad Thermal Expansion Framatome ANP uses TOODEE2 code in their SB LOCA analysis to model the behavior of the hot rod during the entire event.

TOODE2 uses a correlation for determining clad thermal expansion with three ranges; alpha phase, beta phase, and a transition region between alpha and beta phases. The formulation for the beta phase was programmed with an incorrect constant coefficient of 2.9E-6 instead of 3.2E-6.

This introduced a discontinuity between the transition region and the beta region and caused all clad thermal expansion calculations at temperatures above 1773°F to be underpredicted.

Framatome ANP reported that the code probably came to them containing the error in the 1970's. The error was recently discovered during the implementation of RELAP5 and was discovered while searching for other codes that calculate clad thermal expansion.

A development version of TOODEE2 was created which corrects the error and sample SBLOCA calculations were performed. Also, it was verified that the RELAP4 and RELAX codes implement the correlation correctly.

The nature of this error leads to a bounding -1 TF change in the calculated PCT for FCS.

LIC-02-0054 Page 1 Attachment 2 Fort Calhoun Station Small Break LOCA Peak Clad Temperature Margin Utilization Summary LICENSING BASIS Clad Temp (°F)

Analysis of Record 1865 MARGIN ALLOCATIONS (APCT)

A. Prior Permanent ECCS Model Assessments 0 B. 2001 10 CFR 50.46 Model Assessments (Permanent Assessments of PCT Margin)

1. Errors discovered during RODEX2 V & V 0

2. Error in TOODEE2 Clad Thermal Expansion -1 LICENSING BASIS PCT + MARGIN ALLOCATIONS 1864

LIC-02-0054 Page 1 Attachment 3 10 CFR 50.46 Large Break Model Assessments Incorrect Pump Junction Area used in RELAP4 Blowdown Analysis Framatome ANP document, XN-75-41 (A) Volume II ("Exxon Nuclear Company WREM-based Generic PWR Evaluation Model") Appendix provides an input listing for the LBLOCA example problem which shows both reactor coolant pump junctions incorrectly modeled with one-half of the cold leg pipe cross sectional area. There is no justification provided in the documentation for this non-physical modeling approach. This modeling approach was not used in the Siemens Power Corporation Non-LOCA, SBLOCA, and main steam line break ANF-RELAP models nor in the best estimate S-RELAP5 LOCA models being submitted to NRC for review and approval.

Further, this modeling was not used in the NRC-sponsored nuclear safety research programs such as Semiscale and LOFT experimental analyses. Therefore, modeling of the pump junction areas in the RELAP4 blowdown as one-half of the pipe area is different from normal practice and may produce misleading and erroneous break spectrum analysis results.

It was determined that the Cycle 20 analyses supporting Fort Calhoun Station used a smaller area for the pipe junctions. The analysis was corrected by using the full pipe areas. The LBLOCA guideline is also being changed to require the use of the full connecting pipe area.

The nature of this error leads to a bounding +2 'F change in the calculated PCT for FCS.

Incorrect Fuel Density Input to RAD2D2 Radiation Calculation RAD2D2 code is used to calculate the radiation heat sink temperature distribution at beginning of refill (EOBY time), and radiation heat transfer parameters needed for the Framatome ANP radiation heat transfer model incorporated in the TOODEE2 code. The Framatome ANP LBLOCA guidelines specify that the fuel density input to the RAD2D2 code should be the initial fraction of nominal fuel theoretical density.

The Fort Calhoun LBLOCA analysis for Cycle 20 calculation notebook incorrectly used fractional density of 0.94 as input to RAD2D2, rather than using the nominal fuel initial fraction of theoretical density of 0.9535 as specified by the Framatone ANP guideline. The fuel density was set to 0.94 based on an ambiguous description of the fuel density input in the RAD2D2 code users manual. The input description was interpreted as meaning that, for the volumetric heat capacity option being used, the fuel density should be set equal to 0.94 for input to RAD2D2.

Current licensing calculations for Cycle 20 were reviewed. It was discovered that FCS is the only affected plant. A sensitivity calculation was performed for the limiting PCT case with the correct fractional fuel density input to RAD2D2.

The nature of this error leads to a bounding +1 'F change in the calculated PCT for FCS.

LIC-02-0054 Page 1 Attachment 4 Fort Calhoun Station Large Break LOCA Peak Clad Temperature Margin Utilization Summary LICENSING BASIS Clad Temp (°F)

Analysis of Record 1905 MARGIN ALLOCATIONS (APCT)

A. Prior Permanent ECCS Model Assessments* +51 B. 2001 10 CFR 50.46 Model Assessments (Permanent Assessments of PCT Margin)

1. Incorrect Pump Junction Area in RELAP4 Blowdown +2

2. Incorrect Fuel Density Input to RAD2D2 Radiation +1 LICENSING BASIS PCT + MARGIN ALLOCATIONS 1959

• per Reference 4 of LIC-02-0054