ML031110315

Outline of RES Activities to Be Undertaken in Support of NRR Significance Determination Process Davis Besse Head Wastage Issue

ML031110315
Person / Time
Site:	Davis Besse
Issue date:	06/05/2002
From:	Matthew Kirk Office of Nuclear Regulatory Research
To:
References
FOIA/PA-2003-0018
Download: ML031110315 (3)

Text

Outline of RES Activities to be Undertaken in Support of NRR Significance Determination Process:

Davis Besse Head Wastage Issue 3P June 2002 - Mark Kirk (mtknrc~aov) - Draft for Comment 2MEMEMM- "

Fig. 1: Photographs of wastage area after removal from Davis Besse head. Photographs taken from P June 2002 presentation made by FENOC to the A CR5.

Short Term Focus: Estimate the probability of rupture of the un-backed cladding at pressures between operating pressure (2165 psi) and 2500 psi.

Tasks Task 1. Measurements Measure the wastage cavity and cladding thickness / surface characteristics. Plug containing wastage cavity is currently at the B&W hot cells in Lynchburg, VA. Point of contact is Mark McLaughlin, mmclauahlint5firstenernycorp.com, 419-321-8295, FENOC field activities team leader.

Task 2. Failure Criteria Cladding failure criteria to be established by calibrating to 1972 burst disc experiments Task3. Finite Element Analyses Do bounding analysis to determine if there is any merit to performing a probabilistic computation.

In this analysis use the o thinnest cladding based on Fig. 14 from licensee root cause document o Largest geometry of exposed area based on Fig. 13 from licensee root cause document and based on Fig. 1 on the previous page.

o Lowest clad yield strength If "failure" (failure defined based on calibration to 1972 burst disc experiments) is not achieved below 2500 PSI then STOP!!!!!.

• If "failure" is achieved in the bounding analysis, then we need to do a probabilistic analysis, as detailed below:

• 5 to 10 different geometries o 1 (or perhaps 2) descriptions of the wastage geometry >3

o About 5 different uniform thicknesses of cladding

???? different yield strengths Task 4. Monte Carlo Analysis Monte Carlo analysis to consider the following as variables (or not).

• Failure Criteria: Not a variable. As established in Task 2.

• Wastage Geometry: First cut based on current model (Fig. 13 of licensee root cause report),

represented as either a single exposed area geometry or as two exposed area geometries that bound measurements). If two exposed areas are used Monte Carlo analysis to sample from them with equal frequency. Second cut based on measurements to be made at Lynchburg on or before 4th July 2002 (contact: Mark McLaughlin, mmclauchlin(firsteneroycorp.com, FENOC field activities team leader).

• Clad Thickness: First cut is a statistical distribution of thicknesses based on UT measurements taken from Fig. 14 of licensee root cause report. Second cut based on measurements to be made at Lynchburg on or before 4th July 2002 (contact: Mark McLaughlin, mmclaughlin(firsteneraycorp.com, FENOC field activities team leader). Cladding to be represented as a constant thickness, no undulations

• Strenqth: Gaussian distribution of yield strengths based on ?????. Hold hardening exponent fixed.

• Cracks: None modeled.

Output is a fragility curve of failure probability vs. pressure.

Task 5. Report ORNL to supply letter report to NRC RES.

Long Term Focus: Estimate the size of a wastage cavity that would result in rupture of the un-backed cladding at operating pressure (2165 psi)

Tasks Task 1. Corrosion Growth Rate Model

• Need to establish if we will use uniform or biased expansion model, or both? Need to write down the physical basis for both models, and why we pick one over the other

• Need to establish corrosion growth rate, with uncertainty.

Task 2. Finite ElementAnalyses

• Multiple wastage geometries

• Multiple (uniform) cladding thicknesses

• Multiple yield strengths Task 3. Monte Carlo Analysis Monte Carlo analysis to consider the following as variables (or not).

• Failure Criteria: Not a variable. As established in Task 2 under"short term focus".

• Wastage Geometry: Initial geometry based on measurements to be made at Lynchburg on or before 4th July 2002 (contact: Mark McLaughlin, mmclaughlinUfirstenergycorp.com FENOC field activities team leader). Larger geometries based on growth rate models established in Task 1.

• Clad Thickness: Based on measurements to be made at Lynchburg on or before 4 th July 2002 (contact: Mark McLaughlin, mmclaughlin~firstenergycorp.com, FENOC field activities team leader). Cladding to be represented as a constant thickness, no undulations

• Strenath: Gaussian distribution of yield strengths based on ?????. Hold hardening exponent fixed.

• Cracks: None modeled.

Output is a fragility curve of failure probability vs. cladding thickness, parametric in wastage area.

Task 4. Report ORNL to supply letter report to NRC RES.