ML18346A547

From kanterella
Jump to navigation Jump to search
NRC Presentation CCF Public Meeting 12122018 No Notes
ML18346A547
Person / Time
Issue date: 12/12/2018
From: Ching Ng
NRC/NRR/DRA/APOB
To:
Ng C
References
Download: ML18346A547 (16)


Text

Common Cause Failure (CCF)

NRC/External Stakeholders Public Meeting Dec 12, 2018

2 Outline of Issues to Discuss

  • SDP Basic Events Modeling
  • Recent and Notable Interactions with Industry
  • Industry Proposals
  • Stakeholders Comments
  • Going Forward

3 SDP Basic Events Modeling

  • NRC uses the alpha factors methodology:

- Documented in NUREG-2225, Basis for the Treatment of Potential Common-Cause Failure in the Significance Determination Process, Sept 2018, ML18274A198.

  • Failures attributable to the Performance Deficiency in the event or condition are modeled as failures.
  • Equipment and operator action successes-are not set to FALSE (probability of 0.0). Basic events are left at nominal failure probability.

4 Recent CCF-related SDPs

  • Plant Hatch EDG coupling degradation and failure in 2009.
  • Oyster Creek Electro-matic Relief Valve (EMRV) in 2014.
  • Dresden EMRV in 2014.
  • Oyster Creek EMRV failure in 2016.

5 Notable Interactions with Industry

  • Exelon letter to Bill Dean on January 2017 and NRC response (ML17066A245).

- Outlining several points including a second Oyster Creek EMRV failure in 2016.

- Concerns over SDP realism, CCF, HRA, and SRA-licensee engagement.

  • January 2018 - received Revision 1 of Industry white paper (ML18016A125)
  • May 2018 - Internal stakeholders call to discuss proposed response

6 Industry Proposals

  • Industry does not challenge the methodology used to estimate the base CCF probabilities.
  • NEI proposed:

(1) Perform sensitivities to determine if CCF is a major driver; (2) Based on the strength of CCF defense strategies, apply a qualitative factor (as oppose to 1x Conditional CCF values).

7 Question 1. Under what circumstances can Conditional CCF values be adjusted for CCF defense strategies?

Question 2. If Conditional CCF values can be adjusted, how much?

(Qualitative or Quantitative)

Fundamentally, it is an attempt to answer these 2 questions:

What Are We Trying to Address?

8 Question 1. Under what circumstances can Conditional CCF values be adjusted for CCF defense strategies?

Stakeholders perspectives:

- Never (the method of alpha factors is scientifically-sound)

- Sometimes (sensitivities calculations showing CCF is driving, only when CCF defense strategies are presented, etc.)

- During the detailed risk evaluation (DRE), or during the SERP meeting Stakeholders Comments

9 Question 2. If Conditional CCF values can be adjusted, how much? (Qualitative or Quantitative)

Stakeholders perspectives:

- During the detailed risk evaluation (quantitative)

- Current alpha factors have already accounted for known CCF defense

- Current alpha factors only account for average CCF defenses

- Current data is not up-to-date/sparse

- 1x nominal CCF? 0.5x Conditional CCF? What are the bases?

- Causal alpha factors?

- During SERP (qualitative) as part of the five Risk-Informed Decision making principles Stakeholders Comments (cont.)

10 Going Forward

  • NRR will put these elements for trial-use for 1 year

- Perform CCF sensitivity evaluation

- Allow licensees to demonstrate their CCF defense strategies warrant the adjustment of the conditional CCF values.

  • In parallel, NRR will implement a User Need Request (UNR) to request RES investigate the causal alpha factors method in more detail.
  • We will consider insights gained from the UNR and the trial use period in determining whether RASP Handbook update is needed.
  • Details - a memo from APOB Branch Chief to DRA Director

11

12 Backup Slides

13 Definitions of CCF

  • Two or more individual components fail or are degraded, including failures during demand, in-service testing, or deficiencies that would have resulted in a failure if a demand signal had been received.
  • Components fail within a selected period of time such that success of the PRA mission would be uncertain.
  • Component failures result from a single shared cause and coupling mechanism.
  • A component failure occurs within the established component boundary.

14 NRCs Approach to CCF

15 Adjusting CCF for EDG A Fail-To-Start

16 EDG Fail-To-Start CCF Data