ML15208A460

Slides of Regulatory Conference for July 8, 2015

ML15208A460
Person / Time
Site:	Pilgrim
Issue date:	07/08/2015
From:	Entergy Corp, Entergy Nuclear Generation Co
To:	NRC Region 1
Marjorie McLaughlin
References
Download: ML15208A460 (45)

Text

PILGRIM NUCLEAR POWER STATION REGULATORY CONFERENCE REGULATORY CONFERENCE Safety Relief Valve SRV-3A July 8, 2015

Entergy Representatives Entergy Representatives John Ventosa COO Northeast Fleet John Dent Site Vice President John Dent Site Vice President David Noyes Director, Regulatory & Performance Improvement John Macdonald Sr. Manager, Operations g

p David Mannai Sr. Manager, Fleet Regulatory Assurance Thomas White Manager, Design Engineering Everett (Chip) Perkins Manager, Regulatory Assurance Clem Littleton Senior Lead Engineer Patrick Doody Senior Lead Engineer Patrick Doody Senior Lead Engineer Dr. Donald Dube PRA Consultant, ERIN Engineering Dr. M. S. Kalsi President, Kalsi Engineering Zachary Leutwyler Senior Specialist, Kalsi Engineering 2

OPENING REMARKS OPENING REMARKS John Dent Site Vice President Pilgrim Nuclear Power Station Pilgrim Nuclear Power Station 3

Agenda Agenda 4

INTRODUCTION INTRODUCTION Dave Noyes y

Director, Regulatory & Performance Improvement Pilgrim Nuclear Power Station Pilgrim Nuclear Power Station 5

Introduction Introduction

• We are extracting all of the learnings from this event g

g o Detailed root cause evaluation is underway; identified gaps in:

Equipment performance monitoring Operability evaluation and corrective action rigor Operability evaluation and corrective action rigor Post trip review performance o We are sharing those learnings with the fleet I

t h

b d

i dd i

• Improvements have been made in addressing conditions in the Corrective Action Program (CAP) since 2013 o Continued improvements in CAP implementation ongoing 6

Pilgrim performance is different in 2015 versus 2013

3-Stage Target Rock Safety Relief Valve g

g y

(SRV) 7 4 Safety Relief Valves

Full Range of SRV Operation Full Range of SRV Operation Risk-Significant Region 8

Pilgrim Approach and Methodology Pilgrim Approach and Methodology

• Best available (including new) information is derived

(

g

)

using:

o NRCs bounding analysis Plant specific PRA model o Plant-specific PRA model o Actual in-plant experience in light of physical inspection of SRVs o Results from component inspections at National Technical Systems (NTS) (f l W l L b )

(NTS) (formerly Wyle Labs) (New Information) o Analytical Work (New Information)

• Result after considering all plant-specific risk inputs is very low safety significance 9

PILGRIM SRV PERFORMANCE HISTORY PILGRIM SRV PERFORMANCE HISTORY John Macdonald Senior Manager, Operations Pilgrim Nuclear Power Station g

10

Key Events Key Events May 2011 3-Stage Target Rock SRVs installed during RFO 18 o

New model SRVs included all known corrective measures for main stem/piston loosening (identified in IN 2003-01 and GE SIL 646)

(identified in IN 2003-01 and GE SIL 646)

Feb 8, 2013 Cooldown following loss of offsite power and scram - winter storm Nemo o

Attempted use of SRV-3A to reduce reactor pressure on three occasions; response from tailpipe acoustic monitor was not as expected (114 psig, 101 psig, and 98 psig)

May 2013 SRV-3A Pilot Assembly replaced due to Intergranular Stress Corrosion Cracking of Bellows identified in 10 CFR Part 21 during RFO 19; SRV-3A cycled Aug 22, 2013 SRV-3A manually opened twice (923 psig and 823 psig) and reclosed during plant shutdown Oct 14, 2013 SRV-3A manually opened once (1080 psig) and reclosed during plant shutdown Jan 27, 2015 Cooldown following partial loss of offsite power - winter storm Juno o

Two attempts to use SRV-3C to reduce reactor pressure (220 psig and 260 psig)

F b 2015 SRV 3A d SRV 3C l

d i

t t

t f i t t

J f

d t

Feb 2015 SRV-3A and SRV-3C replaced prior to restart from winter storm Juno forced outage Mar 16, 2015 Target Rock issues 10 CFR Part 21 interim report based on Pilgrim SRV performance May 2015 3-Stage SRVs replaced with 2-Stage model during RFO 20 11 11

Limited Range of Known Impact Limited Range of Known Impact

• Extent of Condition Risk-Significant Region 12

Key Events and Timeline Key Events and Timeline

==

Conclusions:==

• At all times maintained multiple methods of reactor pressure control

• At all times had at least 2 SRVs available for pressure relief 2 SRV (SRV 3B & 3D) d d l d

li bl lti l

• 2 SRVs (SRV-3B & -3D) opened and closed reliably on multiple demands when called upon across the entire pressure range

• All SRVs consistently opened at high pressure

• SRV-3A opened 4 times at high pressure after failing to open at low pressure during winter storm Nemo - 2013 S

• All SRVs consistently reclosed under all conditions

• SRV-3B & -3D responded as expected more than 100 times Pressure control via SRVs available at all times across 13 13 Pressure control via SRVs available at all times across the entire pressure range

DELTA CORE DAMAGE FREQUENCY DIFFERENCES DIFFERENCES Dave Noyes y

Director, Regulatory & Performance Improvement Pilgrim Nuclear Power Station Pilgrim Nuclear Power Station 14

What is different?

What is different?

• NRC Staff conclusions driving the low-to-moderate safety significance o Common Cause Factor (CCF) consistently applies to failure-to-open (FTO) for all SRV (

d 1 i 6 FTO b bilit )

SRVs (assumed 1 in 6 FTO probability) o Failure-to-close (FTC) is credible (10X FTC probability) o SRV failure probabilities increase across the full reactor pressure vessel (RPV) pressure spectrum (10X FTO factor)

Pil i

l i

d i i th l

f t i

ifi

• Pilgrim conclusions driving the very low safety significance o SRV-3B & -3D performance affirmed by inspection and analysis supports a lower CCF rate o Based on valve performance and maintenance practices, use of a nominal FTC b bilit i i t probability is appropriate o All SRVs exhibited functionality at high RPV pressure o Based on operating history, a 2X nominal FTO probability is conservative and is based on Bayesian methodology Pil i

ill t th i

ti d t ti lt d th i

t thi

• Pilgrim will present the inspection and testing results and the impact this has on the risk assessment New information provides plant-specific best information 15

SRV INSPECTION / ANALYSIS RESULTS SRV INSPECTION / ANALYSIS RESULTS Thomas White Manager, Design Engineering Patrick Doody Senior Lead Engineer Senior Lead Engineer Pilgrim Nuclear Power Station 16

SRV 3A & 3C SRV-3A & -3C

• SRV-3A (2013) & SRV-3C (2015) failed-to-open on

(

)

(

)

p demand at relatively low reactor pressure

• Both valves removed from service in February 2015 and l

d replaced

• Comparison of valves:

o Similar conditions found on both SRV-3A & SRV-3C o Similar conditions found on both SRV 3A & SRV 3C o Physical inspection of SRV-3B (SN8) & SRV-3D (SN7) revealed only minor degradation (wear) that would not have impacted valve operation operation SRV-3A & -3C physical conditions distinctly SRV 3A & 3C physical conditions distinctly different from SRV-3B & -3D 17

3 Stage Target Rock SRV Testing 3-Stage Target Rock SRV Testing

• Testing of all 4 SRVs at NTS o All valves were removed during RFO-20 (April 2015) o SRV-3A & -3C with approximately 60 days in-service o SRV-3B & -3D with approximately 2-years in-service o As-found main stage test at 50 psig instead of safety mode setpoint o Successful lifts without high pressure pre-conditioning

• Valves were disassembled and inspected o Dimensional checks on individual parts o Observed by Entergy, Target Rock, Lucius Pitkin (LPI), and NRC

• Inspection and test results support Pilgrim position on reliable SRV p

pp g

p performance for the 4 SRVs, particularly SRV-3B & -3D SRV-3B & -3D maintained full functionality 18

Target Rock Part 21 Target Rock Part 21

• New 3-Stage SRV Model 0867F has unique combination g

q of internal features (used in Pilgrim 2011 - 2015) o Largest available throat diameter o Largest total steam drive flow to piston g

p o Highest opening speed o Cause Analysis

Disc/piston experiences largest test stand impact force resulting in p

p g

p g

piston becoming de-torqued and de-shouldered

Vibration and/or pressure pulsations cause detrimental wear on some steam supply systems Design changes to be determined by analysis and testing o Design changes to be determined by analysis and testing 10 CFR Part 21 analysis is consistent with Pilgrims 10 CFR Part 21 analysis is consistent with Pilgrim s analysis and experience 19

Main Stage Internal Parts Main Stage Internal Parts Example of Ring/Guide Wear 20

SRV Analysis Model SRV Analysis Model

• Calculates the Piston

• Calculates the Piston Delta Pressure (DP) required to climb up the ramp ramp

• Required Piston DP increases as angle increases increases

• Required Piston DP increases as friction ffi i t i Groove Depth coefficient increases R

i d Pi t DP i d

d t

l d f i ti Ramp Angle 21 21 Required Piston DP is dependent on ramp angle and friction coefficient

SRV Assessment of Stroke SRV-3B - New information

• Inspection revealed no measureable wear and guide OPEN p

g damage

• Target Rock declared guide reusable SRV-3D - New information SRV 3D New information

• Inspection revealed shallow wear groove and ramp angle

• Shallow ramp easily climbed by piston ring

.0025 depth 15°

~10° SRV-3A and -3C

• Both did not open at low pressures (98 to 260 psig), but both opened at higher system pressures 56°

.020 depth pressures New Information for SRV-3B & -3D

• Both opened at NTS at 50 psig; meeting the valve design specification

• Internal inspections explain why the valves operated per the design 22 22 SRV-3B and -3D operated at all pressures - SRV-3A and -3C had reduced operating range due to ramp angle

• Internal inspections explain why the valves operated per the design

Inspection Results p

Condition Affected Part Demand Cycles 50+

50+

8 (2 PWTs) 3 (1 PWT)

Grooves Worn in Guide NO Minor wear from piston contact Deep Ring Grooves Deep Ring Grooves Max Depth 0.0025 inches Max Depth 0.020 inches Max Depth 0.018 inches Max Ramp Angle Not Applicable

~10°

~39°

~56° Gap from Piston to 3-4 mils 3-4 mils Large Gap Large Gap Gap from Piston to Disc Shoulder 3 4 mils 3 4 mils Large Gap 13-16 mils Large Gap Not Measured Piston OK Shoulder not Wear on OD Threads Galled Wear on OD and Threads Galled Rings Normal Wear Normal Wear Fretting Wear Fretting Wear Rings Normal Wear Normal Wear Fretting Wear Fretting Wear Disc / Stem Shoulder not Disc length short Shoulder not Disc length short Threads Galled Disc machined off Pitch damage Threads Galled Severe thread pitch damage 23 SRV-3A & -3C physical conditions distinctly different from SRV-3B & -3D

SRV 3B Guide SRV-3B Guide No Guide Damage g

No Grooves Guide is re-useable Ring shadows on Guide surface are d b h t

caused by heatup SRV-3B opening and closing unaffected - No wear 24

SRV 3D Guide SRV-3D Guide Magnified Impression of Localized Wear in of Localized Wear in Guide on SRV-3D Max Localized Depth 0.0025 Inches Minor Wear Minor Wear Location 10° OPEN SRV-3D opening and closing unaffected by very minor wear 25

SRV 3A Guide SRV-3A Guide Magnified Impression from top and bottom ring grooves located at bottom ring grooves located at 1200 position Max Depth 0.020 inches 8 times deeper than SRV-3D Sh ll l

th Shallower ramp angle than SRV-3C 0.290 inches wide

~39° 0.240 inches wide Top Groove Bottom Groove OPEN OPEN SRV-3B & -3D comparatively little or no Guide wear 26

SRV 3C Guide SRV-3C Guide Magnified Impression from Bottom Ring Groove located at Top Groove 0 003 inches Bottom Ring Groove located at 0530 position Max Depth 0.018 inches 7 times deeper than SRV-3D St l

th SRV 3A 0.003 inches Steeper ramp angle than SRV-3A 0.215 inches wide OPEN

~56° B tt G

OPEN Bottom Groove SRV-3B & -3D comparatively little or no Guide wear 27

Kalsi Engineering Results Kalsi Engineering Results

• Valve modeling to assess opening/closing performance (using as-found conditions)

• Force balance analysis takes into account:

o Differential pressures o Coefficient of friction between internal parts o Spring force o Parts geometry Threads Stem and piston shoulders Effective ramp angle of grooves Thi d P t R i

P f

d b MPR d LPI

• Third-Party Review Performed by MPR and LPI Comprehensive modelling of all critical parameters Comprehensive modelling of all critical parameters affecting SRV operation 28

Kalsi Engineering Conclusions Kalsi Engineering Conclusions Opening Analysis p

g y

• SRV-3A & -3C minimum opening pressure estimated by analysis

• Minimum opening pressures:

SRV/Opening Pressure Effective Ramp Angle

• 3A: Est >200 psig

~39°

• 3A: Est. >200 psig

~39°

• 3C: Est. >300-400 psig

~56°

• 3B: Actual <50 psig None

• 3D: Actual <50 psig

~10° SRV-3A & -3C - Reduced ability at low pressure SRV 3A & 3C Reduced ability at low pressure SRV-3B & -3D - No loss of capability 29

Kalsi Engineering Conclusions Kalsi Engineering Conclusions Closing Analysis g

y

• Based on design, approximately two times more closing force available than opening force

• Ability to close was not compromised Ability to close was not compromised

• As-found conditions identified in Hatch and Browns Ferry 2-stage SRVs that failed to close in-service were evaluated (Ref IN 2003 01 GE SIL 646 Browns Ferry Licensee Event (Ref. IN 2003-01, GE SIL 646, Browns Ferry Licensee Event Report, Hatch and Browns Ferry Root Cause Analysis)

• Pilgrim piston tilting limited by existing thread and shoulder clearances on all SRVs clearances on all SRVs Cl i

bilit t

i d f l

Closing capability not compromised for any valve 30

New Information Summary/Conclusions New Information - Summary/Conclusions

• Additional Inspections, Testing, and Modelling/Analyses Support Minimal Loss of SRV Functional Capability Minimal Loss of SRV Functional Capability o CCF

Significant physical differences between SRV-3A/-3C and SRV-3B/-3D

This explains the performance differences at low pressure SRV 3A & 3C functioned at high pressure

SRV-3A & -3C functioned at high pressure

Differences should be reflected in the CCF o FTC:

New data demonstrates failure mode baseline probability is unaffected

All SRVs demonstrated closing capability at all pressures o FTO:

Testing and inspections of SRV-3B & -3D demonstrates failure mode baseline probability is unaffected p

y

Both SRVs demonstrated opening capability at all pressures New information provides best available information New information provides best available information for risk assessments 31

PILGRIM PERSPECTIVE ON NRC STAFF BOUNDING ANALYSIS STAFF BOUNDING ANALYSIS Dave Noyes y

Director, Regulatory & Performance Improvement Pilgrim Nuclear Power Station Pilgrim Nuclear Power Station 32

PRA Key Assumptions Comparison and y

p p

Reconciliation - MLOCA Run Title PNPS (CDF)

NRC (CDF)

Factor Difference Comment Delta Core Damage Frequency (CDF) 1.83E-08 1.34E-06

~ 100 Result: 2 orders of magnitude difference MLOCA Base Case 2.08E-07 1.19E-07

~ 1 Close agreement Similar Success criteria of 1 out of 4 SRV required to open MLOCA Conditional 2.27E-07 1.46E-06

~ 10 Large difference in CCF NRC CCF = 1.74E-01 PNPS CCF = 1.39E-03 SPAR model kept CCF group of 4 valves PNPS separated CCF group into 2 valves (SRV-3B & -3D)

Major difference is how CCF is treated 33 Major difference is how CCF is treated

Pilgrim Perspective MLOCA Pilgrim Perspective - MLOCA Run Title PNPS (CDF)

NRC (CDF)

Factor Comment

(

)

(

)

Difference Delta CDF 1.83E-08 1.34E-06

~ 100 Large difference in CCF NRC CCF = 1.74E-01 PNPS CCF = 1.39E-03 Delta CDF (New) 1.30E-07

~ 10 PNPS CCF = 1.0E-02

• Not credible to postulate a 1-in-6 chance that all SRVs will fail to Not credible to postulate a 1-in-6 chance that all SRV s will fail to open o Valves SRV-3B & -3D opened and closed more than 100 times o Disassembly of valves SRV-3B & -3D revealed minimal physical degradation which would not have prevented operation of valves over entire pressure range CCF i l

ti ith t

which would not have prevented operation of valves over entire pressure range o Given the actual performance of SRV-3B & -3D (operated ~ 100 times), inspection results, and analysis we have assigned a conditional CCF value of 1.0E-02 34 CCF is overly conservative without consideration of plant-specific information

CCF Sensitivity MLOCA CCF Sensitivity - MLOCA

• SPAR CCF is 2 orders of magnitude greater than Pilgrims o NRC treats all 4 SRVs as a CCF group of 4; Application of SPAR CCF o NRC treats all 4 SRVs as a CCF group of 4; Application of SPAR CCF generates a 1.74E-01 value which assumes loss of function of 2 or more valves (upper bounding condition) o Initial Pilgrim PRA analysis treated SRV-3B & -3D as a CCF group of 2 b

d t

l f

i ti d

l i

thi t d based on actual performance, inspection, and analysis; this generated a 1.39E-03 value (lower bounding condition) o Best-available information compels consideration of actual performance o Given actual performance of SRV-3B & -3D (operated ~ 100 times),

p

( p

),

inspection results and analysis we assigned conditional CCF value of 1.0E-02 o A CCF value of 1.0E-02 is conservative considering the performance of SRV-3B & -3D combined with functionality of SRV-3A & -3C at high SRV-3B & -3D combined with functionality of SRV-3A & -3C at high pressure SRV-3A & -3C degraded performance only at pressures below risk-significant range o This yields a Pilgrim Delta CDF for all MLOCA events of i

t l 1 30E 07 approximately 1.30E-07 35

PRA Key Assumptions Comparison and y

p p

Reconciliation - Non-MLOCA Run Title Pilgrim NRC Factor Comment Run Title Pilgrim (CDF)

NRC (CDF)

Factor Difference Comment Delta CDF 3.60E-07 (2X FTO) 2.20E-06 (10X FTO &

10X FTC)

~ 10 2X FTO most appropriate treatment No evidence supporting a 10X FTC Non MLOCA Base 2.89E-06 2.08E-05

~ 10 SPAR generally more conservative Conditional (2X FTO) 3.25E-06 Conservative application of 2X FTO in accordance with data contained in NUREG/CR 7037 and the methodology for (2X FTO) contained in NUREG/CR-7037 and the methodology for Bayesian updating in NEI 99-02 Conditional (FTO 10X &

FTC 10X)

See Comment 2.30E-05 Overly conservative - no observed failure mechanism that would tend to increase FTC rate; not realistic to use FTC with 10X multiplier FTC 10X) 10X multiplier 36 2X FTO most appropriate and any FTC increase is not credible

Pilgrim Perspective Non MLOCA Pilgrim Perspective - Non-MLOCA Run Title Pilgrim (CDF)

NRC (CDF)

Factor Difference Comment Delta CDF 3.60E-07 (2X FTO) 2.20E-06 (10X FTO &

10X FTC)

~ 10 2X FTO most appropriate treatment No evidence supporting a 10X FTC

• Overly conservative to assume a 10X increase in probability of FTO o Applying a 10X FTO to all SRVs is overly conservative (Note that the resultant FTO probability of 2.77E-02 would predict ~ 3 instances of SRV FTO during event when 100 demands were experienced) o Conservative application of 2X FTO in accordance with data contained in NUREG/CR-7037 and the o Conservative application of 2X FTO in accordance with data contained in NUREG/CR 7037 and the methodology for Bayesian updating in NEI 99-02

• Not credible to postulate a 10X increase in probability of FTC o RASP Handbook: The expected higher failure probability estimate can be assessed using engineering judgment through expert elicitation; In some cases the estimate may be derived through engineering judgment through expert elicitation; In some cases, the estimate may be derived through prior operating experience of the component o No industry history of FTC for valves modified and maintained in accordance with IN and GE SIL o Kalsi Engineering review of disassembled SRVs affirmed no projected FTC above nominal probability 37 Differences due to overly conservative FTO/FTC assumptions

Delta CDF Comparison Delta CDF Comparison Run Title Pilgrim (CDF)

NRC (CDF)

Factor Comment Run Title Pilgrim (CDF)

NRC (CDF)

Factor Difference Comment Total Delta 4.90E-07 3.54E-06

~10 Differences in application of best-available plant specific available plant-specific information Delta CDF MLOCA 1.30E-07 1.34E-06

~10 CCF application difference MLOCA difference Delta CDF Non MLOCA 3.60E-07 2.20E-06

~10 FTO/FTC application 38 Differences due to overly conservative assumptions

OPERATIONAL RISK MITIGATING FACTORS FACTORS John Macdonald Senior Manager, Operations Pilgrim Nuclear Power Station 39

Operational Risk Mitigating Factors Operational Risk Mitigating Factors

• EOP-01, RPV Control, provides options for high pressure injection into RPV (High Pressure Coolant Injection injection into RPV (High Pressure Coolant Injection

[HPCI], Reactor Core Isolation Cooling [RCIC],

Feedwater, Control Rod Drive, and Standby Liquid Control)

Control)

• EOP-01, RPV Control, provides alternate depressurization systems: HPCI, RCIC, Main Steam line d

i R

W Cl U

i l

d d

drains, Reactor Water Clean-Up in letdown mode

• Our procedures and training direct operators to initiate emergency depressurization when RPV level approaches g

y p

pp top of active fuel versus at the minimum steam cooling reactor water level

• Availability of diverse shutdown methods demonstrated Availability of diverse shutdown methods demonstrated 40

CONCLUSIONS CONCLUSIONS Dave Noyes y

Director, Regulatory & Performance Improvement Pilgrim Nuclear Power Station Pilgrim Nuclear Power Station 41

Conclusions Conclusions

• Public health and safety always protected, with margin

• Our initial conclusions on risk were based on empirical

• Our initial conclusions on risk were based on empirical data regarding performance of the SRVs in the plant o We performed testing, detailed inspections, and expert analysis o These validated the experience of the SRVs in the plant

• Updated plant-specific PRA evaluation provides best available determination of actual risk, based on:

o SRV-3B & -3D were distinctly different from SRV-3A & -3C T

SRV (SRV 3B & 3D) l k d o Two SRVs (SRV-3B & -3D) always worked o Narrow range of scenarios of concern with operation for SRV-3A & -3C Functional at high pressure o Increased FTC probability is not a credible concern for Pilgrim o Increased FTC probability is not a credible concern for Pilgrim Twice as much closing force Appropriate to remove the 10X factor o 2X FTO is appropriate R

lt Ri k i f

l f t i

ifi

• Result: Risk is of very low safety significance 42

Appendix M Analysis Summary Appendix M Analysis - Summary Number Subject Pilgrim Position 4.2.1.1 Defense in depth Other mitigating strategies remained available; alternative pressure control and high pressure injection 4 2 1 2 Reduction in safety margin Very small; SRV 3B & 3D remained operable 4.2.1.2 Reduction in safety margin quantified Very small; SRV-3B & -3D remained operable at all times; Increased FTC probability of any valves not credible; 2X FTO is appropriate 4.2.1.3 Extent of PD impact on other equipment None 4.2.1.4 Degree of degradation 2 valves FTO at low pressure below the risk-significant region significant region 4.2.1.5 Period of fault exposure time 12 months 4.2.1.6 Likelihood of success of Other mitigating strategies remained available; 43 licensees recovery actions g

g g

alternative pressure control and high pressure injection

CLOSING COMMENTS CLOSING COMMENTS John Dent Site Vice President Pilgrim Nuclear Power Station Pilgrim Nuclear Power Station 44

QUESTIONS AND DISCUSSION QUESTIONS AND DISCUSSION 45