ML19134A260

From kanterella
Jump to navigation Jump to search
30-Extremely Low Probability of Rupture (Xlpr)
ML19134A260
Person / Time
Issue date: 05/20/2019
From: Matthew Homiack
NRC/RES/DE/CIB
To:
Shared Package
ML19136A264 List:
References
Download: ML19134A260 (5)


Text

Extremely Low Probability of Rupture (xLPR)

Matthew Homiack Materials Engineer RES/DE/CIB

=

Background===

Purpose:

Determine whether the deterministic approach for evaluating leak-before-break, as outlined in the Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants:

LWR Edition (NUREG-0800), Section 3.6.3, continues to ensure that the probability of piping system rupture is extremely low when considering the presence of primary water stress-corrosion cracking (PWSCC).

Collaborative effort between NRC/RES and the Electric Power Research Institute, Inc.

(EPRI) culminating in xLPR Versions 1 and 2

xLPR Version 2 Key Developmental Documentation

  • Supports NRR review

- Completed xLPR Version 2 Code

- User Manual

- User Training

- External Review Board Reports

- Primary Water Stress-Corrosion Cracking Initiation Model Parameter Development, Con"rmatory Analyses, and Validaon

- Veri"caon and Validaon Tesng Results

- Sources and Treatment of Uncertainty

- Summary NUREG

Maintenance and Distribution

  • NRC/RES and EPRI completed Addendum to their Memorandum of Understanding on Cooperative Nuclear Safety Research to provide for code maintenance, support, and distribution
  • Maintenance

- Conduct under rigorous quality assurance program as in code development

  • Distribution

- Domestic: Finalizing arrangements with third-party distributor/maintainer

- International: Parameters being developed

Version 2 Code Leak-Before-Break Applications Sensivity Studies Focus: Identify which models and inputs contribute most to uncertainty and spread in the output NRC/RES approach developed and exercised through study of V.C. Summer and Tsuruga Findings: Top 3 parameters (all are inputs to PWSCC initiation models)

Proportionality constant Proportionality constant multiplier Welding residual stress at the inside diameter NRC/RES approach compared to EPRI-sponsored approaches The following methods were considered adequate and provided similar results:

Structural Integrity Associates: Most Probably Failure Point Direction Cosines, Degree of Separation Dominion Engineering: Machine learning: Gradient Boosting Decision Trees, Random Forest Decision Trees, Linear Support Vector Machines Engineering Mechanics Corporation of Columbus: Linear or Rank Stepwise Regression, Recursive Partitioning, Multivariate Adaptive Regression Splines Piping Systems Analyses Focus:

Generate system-level leak-before break results (i.e., consider multiple welds)

Exercise proposed acceptance criteria Assess effects of enhanced inspection and mitigation techniques NRC/RES results to be compared with independently generated results sponsored by EPRI