ML23076A044

Pb U2 Scram Barrier Presentation KT Session

ML23076A044
Person / Time
Site:	Peach Bottom
Issue date:	03/16/2023
From:	Frank Arner, Scott Rutenkroger Exelon Generation Co
To:	Kenneth Kolaczyk NRC/NRR/DRO/IRAB
References
Download: ML23076A044 (1)

Text

DISCLAIMER This training session is being recorded for future use in the NRCs knowledge management program. As such, please note that the recorded contents of the session, including any questions posted by audience members, will be preserved in accordance with the NRCs record management program and, moreover, subject to FOIA disclosure. Please refrain from including any sensitive information (i.e., SUNSI) in any questions that you may ask.

• Scott Rutenkroger, Peach Bottom Senior Resident Inspector PRESENTERS

• Frank Arner, Senior Reactor Analyst

PEACH BOTTOM UNIT 2 MAY 16 SCRAM 19 The Unexpected 40 Plans are Made for a Reason, Time Pressure, Purpose Commitment 19 Plan commitments, Just in Time Training, Walkthroughs, 20 Procedures Planning and Setup 19 19 Training and Crew Dynamics 19 Two Heads Partial Procedure vs Temporary Change and 10 50 Lineup Awareness, Plans and Purpose, Training 80 Stop Work Criteria, Peer Checks, Human Integrate Risk Management Standards Performance Tools Questions Turnover and Briefings 19 19 and Answers Quality of Turnover, Review of Actions, Communication 30 70 Ask, Listen, Understand, Reply, Communication Must and Purpose Flow Both Ways 19 When It Really Matters Best Laid Plans 60

4KV EMERGENCY POWER BUS ALIGNMENT NORMAL LINEUP LINEUP DURING 2EA OUTAGE LINEUP WITH 2SU LINKS REMOVED NORMAL RPS POWER SUPPLY ALIGNMENT

ALTERNATE RPS POWER SUPPLY ALIGNMENT

OFFSITE TRANSIENT A RPS TRIPPED

ERROR RESULTS IN BOTH U2 RPS BUSSES TRIPPED

BARRIER ANALYSIS PLANNING AND SETUP COMMITMENT TURNOVER AND BRIEFINGS

BARRIER ANALYSIS THE UNEXPECTED TRAINING AND CREW DYNAMICS

WHEN IT REALLY MATTERS BARRIER ANALYSIS QUESTIONS AND ANSWERS TWO HEADS 3 TAKEAWAYS FROM THIS EVENT

• If you think something is not a credible potential scenario, this one is a reminder that anything can happen. There were at least 5 actions that happened which typically would not occur (This shows the powerful effect of group think or dependency within each human task which can drastically increase all failure rates in a sequence of events)

• Resident Inspectors continue to be the eyes and ears for the agency - never underestimate the importance of each and every observation you have during post event response

• Initiating Events such as MSIV closure events can result in an increase in risk (impulse) which is equivalent to TWICE the entire baseline internal event risk expected over the course of the entire year - (With no other failures of mitigating equipment for event)

HUMAN ERRORS PRIOR TO EVENT

• EO asked the CRS if the partial procedure still needed to carried, CRS stated they would look into it and never got back to the EO. 1E-1 1 in 10 failure?

• 4th RO (tasked to reset 2A RPS) pulled a clean copy of procedure. 1 in 2 0.5?

• EO had the partial procedure and asked 4th RO if it should be used instead of clean copy. 4th RO told the EO to use the clean copy. Recovery Chance 1E-1?

• EO spoke with CRS and was directed to operate A components only, no B components. CRS told EO to N/A step 4.3 but gave no direction for steps 4.4 to 4.9 which is for lining up the alternate feed. NOT CLEAR, STOP? 1E-1

• EO called MCR to verify alarm was in per step 4.4 and 4th RO mistakenly said alarm was in, when it was not.

1 in 100 failure, simple verification alarm 1E-2?

• EO noted the output breakers for alternate feed were closed and step 4.5 would trip them. EO asked 4th RO if they should continue with step 4.5 and was told to continue. CLEARLY NORMALLY RESULTS IN STOP 1 in 10 failure 1E-1

• Incredible that perhaps ANY 1 of the above may have stopped this Event - above exceeds 1 in a million chance ...Clearly shows how dependency among crew thinking or among group think takes place. Easy to have said if asked to evaluate the above- not credible

POST EVENT MCR OBSERVATIONS

• Noted that SRVs seemed to be cycling quite a bit. As it turned out, the residents provided the SRA with computer information showing there were up to 33 SRV cycles within the 1st hour post event. An SRV failing open can then be estimated considering the high number of cycles and core damage cutsetsused to keep the current value in the model which increased the risk outcome due to higher probability of a stuck open SRV

• Residents were wondering 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> into the event why the opening of the MSIVs was not being pursued. 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> in, it was apparent there were issues which were causing procedure limits to be exceeded (i.e. 50 psid limit for re-opening of inboard MSIVs) Thus Reactor Feedpumps, high pressure source and Turbine Bypass Valves (Heat sink) could not be used.

• While Operators were following standard EOPs, there were steps to consider re-opening MSIVs but this did not occur until around 9 hours1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br /> later -Thus the SRA could NOT credit early MSIV recovery based on the challenges that were observed -Licensee denied challenges existed which did not comport w observations (Credit would have lowered risk) 26

WHY WAS RESIDENT INSPECTOR PRESENCE POST EVENT IN MCR CRITICAL IN RISK EVALUATION?

MSIV CLOSURE EVENT RISK

• PB Model of record indicated a CCDP for MSIV closure of 1E-5

• PB claimed their PRA MOR for this event would not be accurate and would overestimate the risk - stated SPAR model with proposed major revision decreasing operator failure rates by up to 25x to 125x should be used (So in essence they wanted incredible Human Error Probability reductions even after all of the PRE Event errors that occurred

• With a MSIV closure event frequency changing from 1/25/yr to 1.0/yr this drives the risk even with no other failures of other mitigating equipment such as HPCI or RCIC, remember we are not calculating what happenedBUT what COULD happen, i.e. MSIVs closed, x prob. Of HPCI failing, Prob. Of RCIC failing and failure to depressurize leading to core damage for example.

So If the PB internal event risk is 3E-6/yr including internal flooding, this ICDP risk increase was 6E-6 or basically equivalent to 2 times the risk of operating the plant over the entire year..Initiating Events create a spike in CDP and have a large impact to the Plant risk profile!

28

QUESTIONS The Office of Nuclear Reactor Regulation (NRR), Division of Reactor Oversight (DRO), will conduct an Inspector Town Hall OPERATING meeting on Thursday, March 9, 2023, from 2:00 pm - 3:30 pm.

The purpose of the meeting is to update inspectors on important developments impacting the Reactor Oversight REACTOR Process (ROP) and their inspection activities.

INSPECTOR Topics for the meeting will include the following:

• Overview of Regulatory Issues Summary 2022-02 Operational Leakage - Jay Collins, Senior Materials TOWN HALL Engineer, NRR/DNRL

• Inspection sample changes and guidance for calendar year 2023 - Manuel Crespo, Operations Engineer, NRR/DRO

• Resident Inspector Recruitment and Retention Status Update - Zack Hollcraft, Senior Operations Engineer, NRR/DRO