RNP-RA/12-0030, Report of Changes to or Errors Discovered in an Acceptable Loss-of-Coolant Accident Evaluation Model Application for the Emergency Core Cooling System

Report of Changes to or Errors Discovered in an Acceptable Loss-of-Coolant Accident Evaluation Model Application for the Emergency Core Cooling System

ML12128A057
Person / Time
Site:	Robinson
Issue date:	04/12/2012
From:	Hightower R Progress Energy Carolinas
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
RNP-RA/12-0030
Download: ML12128A057 (6)

Text

Progrss E ergy10 CFR 50.46(a)(3)(ii)

Progress Energy Serial: RNP-RA/12-0030 APR 2 2012 United States Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555 H. B. ROBINSON STEAM ELECTRIC PLANT, UNIT NO. 2 DOCKET NO. 50-261/ RENEWED LICENSE NO. DPR-23 REPORT OF CHANGES TO OR ERRORS DISCOVERED IN AN ACCEPTABLE LOSS-OF-COOLANT ACCIDENT EVALUATION MODEL APPLICATION FOR THE EMERGENCY CORE COOLING SYSTEM Ladies and Gentlemen:

Carolina Power & Light Company (CP&L), now doing business as Progress Energy Carolinas, Inc.,

is submitting the enclosed report with discusses changes in the peak cladding temperature (PCT) for both the Large Break LOCA as a result of changes made to the fuel design for Cycle 28 and the Small Break LOCA analysis methodology for the H. B. Robinson Steam Electric Plant (HBRSEP),

Unit No. 2. The last report was submitted to the Nuclear Regulatory Commission by letter dated November 23, 2011 covering the changes through November 14, 2011. This report provides the changes covering the period of November 15, 2011 through March 16, 2012. HBRSEP, Unit No. 2 entered Mode 3 on March 16, 2012, thus entering the Mode in which the analysis is required and starting the 30 day reporting requirement.

This submittal satisfies the notification of a significant change [greater than 50'F change in calculated peak cladding temperature (PCT)] for both the Large Break LOCA (LBLOCA) and Small Break LOCA (SBLOCA), as required by 10 CFR 50.46(a)(3)(ii). The details of the changed conditions and impacts are contained in the Attachment I to this letter. The latest PCT estimates for the LBLOCA and SBLOCA are included in Attachment II. Since the changes were identified as part of a reanalysis, the schedule for providing a reanalysis or other actions to show compliance with 10 CFR 50.46 as discussed in 10 CFR 5046 (a)(3)(ii) are considered complete and no additional actions beyond reporting are required. This document contains no new Regulatory Commitments.

If you have any questions, please contact me at (843) 857-1329.

Sincerely, Richard Hightower Supervisor - Licensing/Regulatory Programs Progress Energy Carolinas, Inc.

Robinson Nuclear Plast 3581 West Entrance Road Hartsville, SC 29550 P

United States Nuclear Regulatory Commission Serial: RNP-RA/12-0030 Page 2 of 2 WRH/msc Attachments:

I.

Report of Changes/Errors in Loss-of-Coolant Accident Evaluation Models for the Emergency Core Cooling System II.

Peak Cladding Temperature Estimates c:

V. M. McCree, NRC, Region II A. Billoch-Colon, NRC Project Manager, NRR NRC Resident Inspector, HBRSEP

• t*".*.,t*"'*r¸* *'*'

United States Nuclear Regulatory Commission Attachment I to Serial: RNP-RA/12-0030 Page 1 of 3 H. B. ROBINSON STEAM ELECTRIC PLANT, UNIT NO. 2 REPORT OF CHANGES/ERRORS IN LOSS-OF-COOLANT ACCIDENT EVALUATION MODELS FOR THE EMERGENCY CORE COOLING SYSTEM This report provides an estimate of the effect on peak cladding temperature (PCT) of changes and error corrections in the Loss-of-Coolant Accident (LOCA) evaluation models (EMs) and EM applications for the Emergency Core Cooling System (ECCS) at the H. B. Robinson Steam Electric Plant (HBRSEP), Unit No. 2, covering the period of November 15, 2011 through March 16, 2012.

Large Break Loss-of-Coolant Accident (LBLOCA) Evaluation Model CHANGED CONDITION PCT IMPACT ('F) 3/13/12 - Change in analysis of record from AREVA EMF-3030(P)

+ 264 to AREVA ANP-2973(P). See discussion below for more detail.

2/10/12 - Sleicher-Rouse heat transfer correlation was programmed

+ 14 incorrectly into S-RELAP5 Cumulative Impact

+278 LBLOCA was reanalyzed to evaluate a fuel design change in Cycle 28 to fuel cladding (from Zircaloy-4 to M5) and to the lowest assembly grid (from bimetallic to HMP). The methodology remained AREVA's Realistic LBLOCA methodology, EMF-2103(P)(A). This replaces the Cycle 23 LBLOCA analysis as the analysis of record (AOR). Note that the Cycle 23 analysis was dispositioned against Cycle 27 core and plant parameters making it applicable as the Cycle 27 AOR. There was a significant increase in the 95/95 PCT result relative to the previous analysis:

+132 0F (2084°F in the Cycle 28 analysis versus 19520F in Cycle 23). The error adjusted Cycle 28 95/95 PCT result represents an increase of 278°F over the Cycle 23 95/95 PCT with error corrections through Cycle 27.

The discussion below examines differences in three areas and assesses the PCT impact of these differences in Cycle 28 relative to the Cycle 23 analysis. The discussion examines the effects of

1) differences in the analysis approach, 2) differences in design inputs, and 3) random parameter sampling on the analysis results.

The largest reason for the increase in PCT is the difference in analysis approach. Additional conservatisms were added to the Cycle 28 analysis which address items from the 2010 Calvert Cliffs transition to AREVA fuel (see NRC Accession Number ML110390224). The increase in PCT can be attributed, in part, to the use of once-burned rods and the assumptions regarding the availability of offsite power, which are Items 8 and 5 in the list below, respectively. Additional analysis approach changes are also summarized below.

United States Nuclear Regulatory Commission Attachment Ito Serial: RNP-RA/12-0030 Page 2 of 3

1. The assumed reactor core power for the Robinson Realistic Large Break Loss-Of-Coolant Accident (RLBLOCA) is 2346 MWt. This value represents the plant rated thermal power of 2339 MWt with a maximum power measurement uncertainty of 0.3 percent (7 MWt) added to the rated thermal power. The power was not sampled in the analysis; this results in an assumed power level that is either equal to or greater than if it had been sampled. A higher initial power level results in a zero to positive increase in PCT.

2. The RLBLOCA analysis was performed with S-RELAP5 which requires both the void fraction to be less than 0.95 and the clad temperature to be less than 900 'F before the rod is allowed to quench. This results in a zero to positive increase in the predicted PCT when compared to analyses not subject to these constraints.

3. The RLBLOCA analysis was performed with a version of S-RELAP5 that limits the contribution of the Forslund-Rohsenow model to no more than 15 percent of the total heat transfer at and above a void fraction of 0.9. This results in a minimal increase in the predicted PCT when compared to analyses not subject to these constraints.

4. The split breaks are now allowed to vary between a calculated minimum break size and double the area of the cold leg pipe, the same as guillotine breaks. This did not have an effect on PCT results.

5. In concurrence with the NRC's interpretation of GDC 35, a set of 59 cases was run with a Loss of Offsite Power (LOOP) assumption and a second set with a No-LOOP assumption.

The highest PCT result considering both sets of 59 cases is reported as the maximum PCT.

The previous analysis sampled the availability of off-site power. Since the highest PCT from both sets were used, this results in a zero to positive increase in the predicted PCT.

6. During recent RLBLOCA EM modeling studies, it was noted that changes in the treatment of cold leg condensation were warranted. Water entering the downcomer post-accumulator injection remained sufficiently subcooled to absorb downcomer wall heat release without significant boiling. To address this, saturated fluid conditions are provided at the downcomer entrance which conservatively reduces both the downcomer driving head and the core flooding rate. It is noted that test results indicate that fluid conditions entering the downcomer range from saturated to slightly subcooled. Hence, it is conservative to force an approximation of saturated conditions for fluid entering the downcomer. This results in a zero to positive increase in predicted PCT.

7. The fuel pellet thermal conductivity treatment has been altered to address concerns raised in NRC Information Notice 2009-23. The RLBLOCA analysis applies a burnup-dependent adjustment for code bias. The revised approach replaces the linear expression with a polynomial expression and is multiplicative instead of additive. The polynomial adjustment addresses burnup-dependent degradation of thermal conductivity. This results in a positive increase in PCT.

8. Once burned fuel is now modeled in the analysis in addition to the fresh fuel. This results in a zero to positive increase in PCT.

9. Decay heat uncertainty is not sampled. Rather a conservative approach is used to determine the decay heat produced in the reactor. This results in a zero to positive increase in PCT.

United States Nuclear Regulatory Commission Attachment Ito Serial: RNP-RA/12-0030 Page 3 of 3 Several plant and fuel design inputs also changed. The impacts of these changes on the PCT result were not individually assessed. A qualitative PCT assessment of some of the design input changes is shown below. The main design input changes are summarized as follows.

1. The fuel cladding changed from Zircaloy-4 to AREVA's M5 cladding.

2. The lowest spacer grid changed from a bimetallic design to High Mechanical Performance (HMP).

3. In addition to 0, 2, 4, and 8 w/o gadolinia-bearing fuel rod, rods with 6 w/o gadolinia were also analyzed.

4. Fuel pellet density increased from 95% to 96%.

5. The RHR pump curve decreased by 3% from the value assumed in EMF-3030.

6. The range of containment volume considered in the analysis changed from 1.95 - 2.23 to 1.96 - 2.02 million cubic feet.

7. Steam generator tube plugging decreased from 10% to 6%. Steam generators do not play a significant role in heat removal during a large break LOCA.

8. The assumed FQ decreased from 2.62 to 2.5.

The remainder of the variation is explained by random parameter sampling, which is a normal part of the Realistic methodology. The effects of random parameter sampling are examined using the nominal (50/50) PCT result in Cycle 28 to Cycle 23. Whereas the increase in the 95/95 PCT result from Cycle 23 to Cycle 28 is 132°F, comparing the 50/50 PCT results shows that the difference in the analysis results is not as significant: 87'F (1608'F in the Cycle 28 analysis versus 1521'F in Cycle 23). In accounting for the PCT differences between the two analyses, approximately 45°F (1 32'F-87°F) can be attributed to random parameter sampling.

In summary, the PCT change from Cycle 23 to Cycle 28 can be attributed to differences in changes in the both the design and the analysis approach. The changes result in a more conservative PCT than was obtained in the previous analysis of record, EMF-3030.

Small Break Loss-of-Coolant Accident (SBLOCA) Evaluation Model CHANGED CONDITION PCT IMPACT (fF) 3/13/12 - Change in analysis of record from AREVA EMF-2914(P)

- 123 to AREVA ANP-2972(P). See discussion below for more detail.

12/1/11 - The gamma smearing factor for the original analysis did not

+34 bound the Cycle 28 core design.

2/10/12 - Sleicher-Rouse heat transfer correlation was programmed

- 19 incorrectly into S-RELAP5 Cumulative Impact

- 108 The Robinson SBLOCA Methodology changed from XN-NF-82-49(P)(A) using ANF-RELAP to EMF-2328(P)(A) using S-RELAP5. The reanalysis was done to evaluate a fuel design change in Cycle 28 to fuel cladding (from Zircaloy-4 to M5) and to the lowest assembly grid (from bimetallic to HMP). The decrease in PCT is due to the change to a more advanced methodology.

United States Nuclear Regulatory Commission Attachment II to Serial: RNP-RA/12-0030 Page 1 of 1 H. B. ROBINSON STEAM ELECTRIC PLANT, UNIT NO. 2 PEAK CLADDING TEMPERATURE ESTIMATES The current peak cladding temperature (PCT) estimates associated with Loss-of-Coolant Accident (LOCA) Emergency Core Cooling System (ECCS) evaluation models are listed below. These estimates include the cumulative effects of significant and non-significant error corrections and evaluation model changes through March 16, 2012.

Event PCT (OF)

Large Break LOCA, ECCS Injection Mode 2098 Small Break LOCA, ECCS Injection Mode 1507