05000237/LER-2008-004

LER-2008-004, Non-Conservative Core Spray Flow Utilized in LOCA Analysis

Docket Number
Event date:	04-25-2008
Report date:	06-24-2008
Reporting criterion:	10 CFR 50.73(a)(2)(ii)(B), Unanalyzed Condition
2372008004R00 - NRC Website

Dresden Nuclear Power Station (DNPS) Unit 2 is a General Electric Company Boiling Water Reactor with a licensed maximum power level of 2957 megawatts thermal.

A. Plant Conditions Prior to Event:

Unit: 02� Event Date: 04-25-2008 �Reactor Mode: 1� Mode Name: Power Operation Power Level: 99 percent Reactor Coolant System Pressure: Approximately 1000 psig

B. Description of Event:

On April 25, 2008, the fuel vendor (Westinghouse) notified Exelon of a non-conservatism in the Large Break Loss of Coolant Analysis (LBLOCA) analysis for Dresden Unit 2. The non-conservatism involved the amount of Low Pressure Core Spray (LPCS) flow that actually reaches the reactor core during a LBLOCA event. Several leakage paths can exist between the reactor vessel core spray penetration and the core spray header nozzles that inject flow above the reactor core. This leakage reduces the flow from the core spray header nozzles. The LBLOCA results, including the Peak Cladding Temperature (PCT), are sensitive to the amount of core spray flow reaching the reactor core region. This identified error resulted in an increase of 80°F in the calculated PCT. The 10 CFR 50.46 PCT acceptance criterion of 2200°F was exceeded by 30°F.

In the original Westinghouse analysis for Quad Cities 1 & 2 and Dresden 2 & 3, a bounding modeling approach for the core spray flow, including leakage values, was developed for all four units.

However, the initial LOCA analysis results with this approach were overly restrictive and additional Maximum Average Planar Linear Heat Generation Rate (MAPLHGR) margin was required. The core spray flow modeling was reassessed, utilizing more detailed unit specific information, including LOCA design inputs from Exelon, in an attempt to reduce the conservatism in the bounding modeling approach. During this reanalysis, the Dresden Unit 2 non-conservative error was introduced.

During the acceptance review of the Westinghouse reanalysis, the Exelon subject matter experts (SMEs) accepted the core spray flow modeling justification provided by Westinghouse. The Exelon SMEs, while knowledgeable in transients and LOCA, do not have detailed knowledge of each fuel vendor's methodology and codes, and did not adequately challenge why the revised in-shroud core spray leakage values were bounding for Unit 2.

Westinghouse's application of the Exelon supplied core spray values was subsequently identified by Westinghouse engineers as incorrect (non-conservative) for Dresden Unit 2.

Upon discovery of the non-conservative error, Dresden Operations took immediate actions to apply a maximum average planar linear heat generation rate (MAPLHGR) administrative limit on Dresden Unit 2 to bring the plant configuration into compliance with 10CFR50.46 requirements.

On April 25, 2008 at 1829 hours0.0212 days <br />0.508 hours <br />0.00302 weeks <br />6.959345e-4 months <br />, an Event Notification to the NRC was made. On May 23, 2008, a 30-day followup report was submitted in accordance with the10 CFR 50.46(a)(3)(ii) requirements.

This Licensee Event Report is being submitted in accordance with 10 CFR 50.73(a)(2)(ii)(B): 'The nuclear power plant being in an unanalyzed condition that significantly degraded plant safety.

C. Cause of Event:

Based on the results of the cause evaluation performed by Westinghouse, the apparent cause of this error was that the author and verifier of the Westinghouse analysis had an incomplete understanding of the evaluation model, coupled with a false sense of security due both to the bounding system response being essentially the same as the original response and by the fact that the MAPLHGR margin recovery items had been developed by a panel of technical experts.

A contributing cause was that the Exelon SMEs did not adequately challenge why the revised in­ shroud core spray leakage values were bounding for Unit 2.

D. Safety Analysis:

The safety significance of the event is minimal. During the duration of the event, Dresden Nuclear Power Station continued to operate within the requirements of the Technical Specifications. The error was limited to the calculation for a postulated scenario. Therefore, the consequences of this event had minimal impact on the health and safety of the public and reactor safety.

E. Corrective Actions:

Corrective actions implemented or planned by Westinghouse include:

Correction of calculation notes, reports, etc. impacted by the non-conservative LPCS flows A quality stand-down with employees to provide lessons learned from the cause evaluation - Procedure revision and additional guidance regarding the direction of conservatism and shroud leakages Corrective actions planned by Exelon include:

Training to address identification of critical parameters, and methods to evaluate and minimize risks associated with these critical parameters - Reinforcement of importance of technical rigor and questioning attitude

F. Previous Occurrences:

A search was conducted of the Issue Report database for the last two years concerning any LER discussing the area of inadequate design review for plant impact. One previous LER event was found.

05000237/1998-002-00, Unit 2 Reactor Scram From A Main Turbine Trip Due To Inadequate Design Review Performed During Modification The root cause of the Unit 2 scram was found to be inadequate design review during the modification of Yarway level switch replacement. During development of the modification, the impact that the modification would have on plant maintenance activities was not identified by the design team (Design Engineering, Plant Engineering, Maintenance, etc.). The safety significance of the event was minimal since all plant systems operated as designed.

Corrective Action from this event was limited to emphasizing to engineers the necessity of thorough design reviews, detailed industry event database searches, and the need for detailed questioning of maintenance and operations personnel during design scope meetings. This event differs enough in cause that the previous corrective actions would not had prevented the 2008 Westinghouse LBLOCA analysis event.

An OPEX review was also conducted and no events were identified that resulted in further insight for this report.