ML13115A265

Submittal of 10 CFR 50.46 - 30-Day Special Report

ML13115A265
Person / Time
Site:	Sequoyah
Issue date:	04/23/2013
From:	James Shea Tennessee Valley Authority
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
Download: ML13115A265 (8)

Text

Tennessee Valley Authority, 1101 Market Street, Chattanooga, Tennessee 37402 April 23, 2013 10 CFR 50.46 ATTN: Document Control Desk U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001 Sequoyah Nuclear Plant, Unit 1 Facility Operating License No. DPR-77 NRC Docket No. 50-327

Subject:

10 CFR 50.46 Day Special Report

Reference:

TVA Letter to NRC, "10 CFR 50.46 Annual Report for Sequoyah Nuclear Plant, Units 1 and 2," dated November 20, 2012 [ML12333A241]

The purpose of this letter is to provide a 30-day report of changes to the calculated peak cladding temperature (PCT) for the Sequoyah Nuclear Plant (SQN) Unit 1 Emergency Core Cooling System (ECCS) evaluation models. This submittal satisfies the reporting requirements for a significant change or error in accordance with Title 10 of the Code of Federal Regulations (10 CFR) 50.46, "Acceptance Criteria for Emergency Core Cooling Systems for Light-Water Nuclear Power Reactors," paragraph (a)(3)(ii). The enclosed report provides a summary of the changes to the calculated PCT for the limiting ECCS analyses.

The PCT changes identified for SQN Unit 1 in the referenced annual report submitted on November 20, 2012, when expressed as the cumulative sums of the absolute magnitudes, exceed the 50 degrees Fahrenheit (OF) threshold for a significant change or error as defined in 10 CFR 50.46(a)(3)(i). Accordingly, any subsequently discovered change or error is considered significant for the purposes of reporting until such time as a re-analysis of the ECCS evaluation model is completed. By letter dated March 28, 2013, AREVA NP, Inc.

notified the Tennessee Valley Authority (TVA) an error affecting the Cathcart-Pawel uncertainty correlation used in the metal-water reaction calculation for Realistic Large Break Loss of Coolant Accident (RLBLOCA) applications. As indicated in the enclosed report, there is a 0°F effect on the updated (net) licensing basis PCT for the SQN Unit 1 Large Break LOCA (LBLOCA) analysis of record (AOR). Note that the identified error is not applicable to the SQN Unit 1 Small Break LOCA (SBLOCA) AOR.

Printed on recycled paper

U.S. Nuclear Regulatory Commission Page 2 April 23, 2013 Because there were no changes to the calculated PCTs, the updated (net) licensing basis LBLOCA PCT remains at 1785 0 F and the updated (net) licensing basis SBLOCA PCT remains at 1403 0 F, which are below the 2200°F limit. However, the cumulative sum of the absolute magnitudes of the LBLOCA PCT changes for SQN Unit 1 continues to be above the 50°F threshold defined for a significant change or error, thus necessitating this 30-day report. The due date for compliance with the 30-day reporting requirement is April 27, 2013.

This date falls on a Saturday, so the official due date for reporting is the next business day, which is April 29, 2013.

In addition, when the cumulative sum of the absolute magnitudes of the resulting PCT changes exceeds 50 0 F, 10 CFR 50.46(a)(3)(ii) requires the licensee to include with the 30-day report a proposed schedule for providing a re-analysis or taking other action as may be needed to show compliance with the 10 CFR 50.46 requirements. As presented in this report, compliance with the 10 CFR 50.46 requirements is demonstrated by the calculated LBLOCA and SBLOCA PCTs for SQN Unit 1 remaining below the 2200'F limit. Therefore, TVA has concluded that no proposed schedule for providing a re-analysis or other action is required.

There are no regulatory commitments in this letter. Please direct questions concerning this report to Clyde Mackaman at (423) 751-2834.

if Ily, YResp Shea icPresident, Nuclear Licensing

Enclosure:

10 CFR 50.46 30-Day Report of Peak Cladding Temperature Changes Enclosure cc (Enclosure):

NRC Regional Administrator - Region II NRC Senior Resident Inspector - Sequoyah Nuclear Plant

ENCLOSURE TENNESSEE VALLEY AUTHORITY (TVA)

SEQUOYAH NUCLEAR PLANT (SQN)

UNIT 1 10 CFR 50.46 30-DAY REPORT OF PEAK CLADDING TEMPERATURE CHANGES By letter dated March 28, 2013, AREVA NP, Inc. (AREVA NP) notified TVA of an error in the Cathcart-Pawel uncertainty correlation used in the Realistic Large Break Loss of Coolant Accident (RLBLOCA) applications. This error affected the calculation used to model the metal-water interaction during a loss of coolant accident (LOCA).

The RLBLOCA analysis of record (AOR) for SQN Unit 1 was issued in 2008 and determined the peak cladding temperature (PCT) to be 1809 degrees Fahrenheit (OF). The SQN Unit 1 plant specific analysis is detailed in Topical Report No. ANP-2695P, Revision 0, "Sequoyah Unit 1 Realistic Large Break Loss-of-Coolant Accident Analysis." This report was submitted to the Nuclear Regulatory Commission (NRC) as part of SQN Unit 1 Technical Specification Change TS-08-01. The plant-specific application analysis was found to be acceptable as discussed in the NRC Safety Evaluation Report, dated September 24, 2008.

The AOR for SQN Unit 1 for the Small Break LOCA (SBLOCA) was issued in 2008 and determined the PCT to be 1403 0 F. The SQN Unit 1 plant specific analysis is detailed in AREVA NP Calculation 51-5006843-001, "SQN SBLOCA Accident Analysis Profile."

In accordance with the reporting requirements of 10 CFR 50.46(a)(3)(ii), the following table details the cumulative changes affecting the Emergency Core Cooling System (ECCS) model calculated PCT for SQN Unit 1, and includes the PCT change associated with the Cathcart-Pawel uncertainty correlation error. The new AOR PCT change can be summarized as:

• The Large Break LOCA (LBLOCA) PCT for SQN Unit 1 decreased 24 degrees Fahrenheit (OF), and results in a current net licensing basis PCT of 1785 0 F. There was an estimated 0°F change in the SQN Unit 1 AOR PCT value as a result of this error.

• The SBLOCA PCT for SQN Units 1 remains unchanged from the AOR, with a net licensing basis PCT of 1403 0 F.

• The cumulative sum of the LBLOCA PCT changes for SQN Unit 1, expressed as absolute magnitude, is 106°F. This exceeds the 50°F threshold for a 30-day report in accordance with 10 CFR 50.46(a)(3)(ii).

E-1 of 6

TABLE Summary of SQN Unit 1 PCT changes Report Description LBLOCA Change in SBLOCA Change in Notes Year PCT (°F) LBLOCA PCT (°F) SBLOCA PCT °F) PCT (°F) 2008 AOR PCT (previous) 1809 1403 2008 Cold leg condensation under-predicted following cold leg 0 0 1 accumulator injection 2009 Thermal radiation heat transfer under- +5 0 2 predicted +5_2 2009 Reactor kinetics model coding error and heat conduction -30 0 3 algorithm logic error 2009 Fuel pellet thermal conductivity 0 0 4 degradation 2010 Liquid entrainment under-predicted in the 0 steam generator tubes 2011 High head ECCS injection delay time +24 0 6 increase 2011 Upper plenum modeling inhibits vapor flow into the top 0 0 7 of the hot bundle 2011 S-RELAP5 Sleicher-Rouse heat transfer -35 0 8 correlation equation error 2012 Non-conservative liquid fallback into 0 0 9 surrounding six (6) assemblies 2013 Cathcart-Pawel uncertainty correlation 0 0 10 error in RLBLOCA applications E-2 of 6

TABLE (Continued)

Summary of SQN Unit 1 PCT Changes Report Description LBLOCA Change in SBLOCA Change in Notes Year PCT (°F) LBLOCA PCT (°F) SBLOCA PCT (°F) PCT ( F)

Updated Licensing Net PCT 1785 1403 AOR PCT + Z A PCT Cumulative sum of PCT changes 106 0 ZIA PCTI Notes:

1), Cold leg condensation under-predicted following cold leg accumulator injection The LBLOCA analysis methodology under-predicts condensation in the Reactor Coolant System (RCS) cold leg after the accumulators empty. Because of this, the ECCS water entering the downcomer is sufficiently subcooled that it absorbs the downcomer wall heat without significant boiling. The lack of boiling in the downcomer leads to a higher water level in the core during reflood and a lower PCT. The condensation in the cold legs was increased by using very large multipliers on the interphase heat transfer in the cold legs. This resulted in saturated water entering the downcomer after cold leg accumulator injection. Sensitivity studies were performed using this revised model and no change in LBLOCA PCT was predicted.

Boiling in the downcomer is only significant in LBLOCAs, so the SBLOCA PCT was not affected.

2) Thermal radiation heat transfer under-predicted The LBLOCA analysis methodology accounts for thermal radiation heat transfer between the fuel and the reactor coolant. The model contains a correlation for determining water vapor emissivity and this correlation contained errors that resulted in the wrong values for water vapor emissivity being determined by the correlation. The water vapor emissivity correlation was corrected, and based on sensitivity studies, a bounding increase in LBLOCA PCT was determined to be 5°F. The thermal radiation heat transfer model used in the LBLOCA methodology is not used in the SBLOCA methodology.

3) Reactor kinetics model coding error and heat conduction algorithm logic error The LBLOCA analysis methodology accounts for the change in core power during the LBLOCA using a reactor point kinetics model. The point kinetics model in the version of RELAP5 used for the LBLOCA analysis was found to contain errors in the numerical solution algorithm indices and convergence criteria, which could affect the determination of transient core power. The version of RELAP5 used for the SBLOCA analyses did not contain these algorithm errors.

The LBLOCA analysis methodology accounts for conduction heat transfer in the fuel rods in its determination of PCT. The transient conduction heat transfer solution algorithm used in the methodology contained a logic error that assigned the heat capacity of the right boundary mesh point (N) from the next to last mesh point (N-2) instead of the adjacent mesh point (N-i). The E-3 of 6

SBLOCA methodology did not use the same transient conduction heat transfer algorithm as the LBLOCA methodology.

The point kinetics and heat conduction code errors were corrected in the version of RELAP5 used for the LBLOCA analysis, and sensitivity studies were performed using the updated computer program. These sensitivity studies established a 30°F reduction in LBLOCA PCT.

4) Fuel pellet thermal conductivity degradation In the LBLOCA methodology, the initial fuel pellet temperature distribution was being determined using a model that under-predicts the degradation in fuel pellet thermal conductivity at high core burn-ups. This resulted in lower initial fuel pellet temperatures being predicted and reduced initial fuel pellet stored energy. A proportional adjustment to the fuel temperature calculation was determined by comparing the fuel pellet temperature predications from the model to fuel pellet temperature data. The higher initial fuel pellet temperatures were assessed for their effect on PCT based on the time that PCT occurs. For SQN, Unit 1, PCT occurs during blowdown (less than 100 seconds), which resulted in no change in PCT from the higher initial fuel pellet temperatures.

In the SBLOCA methodology, the initial fuel pellet temperature distribution was also being determined using a model that under-predicts the degradation in fuel pellet thermal conductivity at high core burn-ups. This resulted in lower initial fuel pellet temperatures being predicted and reduced initial fuel pellet stored energy. However, for SBLOCAs, the time of PCT is much later than for LBLOCAs, so the initial stored energy has already been transferred to the reactor coolant. As a result, the initial fuel pellet temperature distribution did not affect SBLOCA PCT.

5) Liquid entrainment under-predicted in the steam generator tubes The LBLOCA methodology uses a bias on interphase friction at the steam generator tube sheet entrance to establish the magnitude of liquid entrainment in the steam generator tubes. The amount of liquid entrainment was found to be under-predicted due to a low value being specified for the interphase friction multiplier in the evaluation model.

During the reflood phase of a LBLOCA, some of the water droplets entrained in the flow from the core region are vaporized in the steam generator tubes due to heat transfer from the hot secondary side of the steam generator. The vaporization of the water in the steam generator tubes increases the pressure difference between the break and the core, typically called "steam binding," which affects the rate of core flooding and consequently PCT.

The interphase friction multiplier was increased as described in AREVA calculation E-2353-N90-59, "Evaluation of Interfacial Drag between Phases for UPTF and FLECHT-SEASET Tests." Sensitivity studies with the higher value for the interphase friction multiplier determined a bounding 12°F increase in LBLOCA PCT.

In a SBLOCA event, depressurization of the RCS is much slower and break flows are less. The only time the interphase friction occurs in the hot legs and steam generator inlet plenums tube regions is during the reflux condensation period when the flow at the tube inlet is counter-current. The flow at the tube inlet is primarily controlled by counter-current limitations.

Therefore, interphase friction and the amount of water retained in the steam generator tubes does not affect SBLOCA PCT results, and the modeling of this phenomenon is not part of the SBLOCA methodology.

E-4 of 6

6) High head ECCS injection delay time increase The ECCS flow to the RCS cold legs from the high head ECCS pumps is modeled in the LBLOCA and SBLOCA analyses as starting after the injection isolation valves are fully open.

The motor operator on these injection isolation valves has been modified, resulting in an increase in their opening time. The SBLOCA analysis uses a very long delay time for high head ECCS pump injection, so the longer opening time on the injection isolation valves did not result in an increase in the assumed delay time. However, the LBLOCA analysis required an increase in the high head ECCS pump injection time delay. The longer delay time for the high head ECCS pump injection reduces the initial volume of water available for core cooling, and resulted in a 24 0 F increase in LBLOCA PCT.

7) Upper plenum modeling inhibits vapor flow into the top of the hot bundle During the reflood phase of a LBLOCA, steam velocities in the hotter fuel assembly flow channels could prevent water above the core from draining back and quenching the hotter fuel pins. To ensure top-down quench does not occur, the LBLOCA modeling of the upper plenum was revised to not allow the core to be quenched from ECCS flow that enters from above the core. The nodalization of the upper plenum was revised and a high (reverse) loss coefficient was used to prevent water from flowing back into the core from the upper plenum. Subsequent analyses determined that there was no change in LBLOCA PCT.

In a SBLOCA, the flows between the upper plenum and the core have less of an effect on core cooling, as the core remains substantially covered, so much less steam is produced during the quenching of the core. That is, steam flow in the hot channel would not be high enough to prevent water in the upper plenum from draining back to the core. In addition, for a SBLOCA, the quenching of the core always occurs from the bottom of the core upwards, so SBLOCA PCT was unaffected.

8) S-RELAP5 Sleicher-Rouse heat transfer correlation equation error Sleicher-Rouse is one of the correlations used in the S-RELAP computer code for predicting convective heat transfer between the fuel and coolant single-phase vapor. This correlation is applicable to both the LBLOCA and SBLOCA analyses performed with S-RELAP5. During a review of the behavior of the Sleicher-Rouse correlation relative to other single-phase vapor heat transfer correlations, an error was discovered in the form of the correlation used in the S-RELAP5 implementation. The difference is related to the form of the equation for calculating the exponent of the temperature ratio correction term. The S-RELAP form of the Sleicher-Rouse heat transfer correlation has been updated to:

n = -[log 1O(Tw/Tg)]11 4 + 0.3 This correction resulted in a 35 0 F decrease in LBLOCA PCT for SQN Units 1. There was no change in SBLOCA PCT.

E-5 of 6

9) Non-conservative liquid fallback into surrounding six (6) assemblies During the reflood phase of a LBLOCA, steam velocities in the central core region could prevent water above the core from draining back and quenching the fuel pins. Therefore, the LBLOCA evaluation model should not allow the fuel pins to be quenched from water that enters the fuel assemblies from above the core. Preventing water from flowing back into the core from the upper plenum is accomplished by applying a high reverse form loss coefficient (FLC) to the flow paths between the hot assembly and central core with the upper plenum.

The evaluation model was found not to include the above described high reverse FLCs in the central core region flow paths. However, a subsequent review of the LBLOCA cases in the AOR determined that no water entered the fuel assemblies in the central core region from above the core, so there was no increase in PCT resulting from this error.

There was no effect on the SBLOCA PCT for the same reasons as stated in Note 7 (above).

10) Cathcart-Pawel uncertainty correlation error in RLBLOCA applications For RLBLOCA analyses, the rate-dependent correlation developed by Cathcart-Pawel is used to model the metal-water reaction during a LOCA. The rate constants for the Cathcart-Pawel equation are determined experimentally and the data are subjected to a statistical analysis to determine the relevant uncertainty parameters for the derived correlation. The RLBLOCA analysis uses a log-normal function for the uncertainty multiplier applied to the rate constant.

The formula and standard deviation were found to be incorrect.

Analysis of the error confirmed that the effect on previous RLBLOCA analyses was negligible.

There is no change to the SQN Unit 1 LBLOCA PCT value from this error. This error did not apply to the SBLOCA analysis and, therefore, had no effect on the SBLOCA PCT value.

E-6 of 6