ML051190665

From kanterella
Jump to navigation Jump to search
Final Precursor Analysis - Perry
ML051190665
Person / Time
Site: Perry  FirstEnergy icon.png
Issue date: 09/01/2003
From: Eliezer Goldfeiz
NRC/RES/DRAA/OERAB
To:
References
LER 2003-004-01, LER 2004-004-01
Download: ML051190665 (10)
v • d • e


Text

1 Final Precursor Analysis Accident Sequence Precursor Program --- Office of Nuclear Regulatory Research Perry Nuclear Power Plant ESW A Pump Failure to Run Due to Shaft Failure and Inadequate Repairs Led to a Second Failure Event Dates: 09/01/2003 05/21/2004 LERs: 440/03-004-01 440/04-001-01 CDP = 1x10-6 April 13, 2005 Operating Condition Summary Description. On September 1, 2003, the plant was in Mode 1 at 100% of rated thermal power.

The Emergency Service Water (ESW) A pump at Perry Nuclear Power Plant (PNPP) was started. The ESW A pump failed to run after 42 minutes, resulting in loss of flow to its loads as documented in Licensee Event Report (LER) 440-2003-004-01 (Reference 1). The control room staff observed all ESW A pump flow indications for Residual Heat Removal A, Emergency Core Cooling A, and Division 1 Emergency Diesel Generator A at zero gallons per minute. The ESW A pump motor temperature began to rise. Operators then declared the ESW A pump to be inoperable. No motor protective trips occurred during the pump run failure event. Region III issued an inspection report on October 30, 2003 (Reference 2). The Office of Enforcement issued a final significance determination process finding letter on January 28, 2004, on the same event (Reference 3).

On May 21, 2004, the ESW A pump failed again. This second failure event, as documented in LER 440-2004-001-01 (Reference 4), was caused by inadequate repair of the first pump failure.

Region III issued an inspection report on July 2, 2004 (Reference 5).

Cause. The second ESP A pump failure was caused again by a failure of the pump shaft coupling sleeve. The shaft coupling sleeve failure was the result of improper coupling reassembly by plant maintenance personnel following the September 2003 failure. Inspection findings documented that intergranular stress corrosion cracking was the failure mechanism for shaft coupling.

The cause of the pump shaft sleeve failure was later identified to be due to licensees failure to follow vendor-specified reassembly instructions for the ESW pumps after maintenance events.

Instead of following vendor-specified reassembly instructions, the licensee relied on knowledge-based skills of their facility maintenance personnel for reassembly of the ESW pumps after maintenance events. This contributed to the first failure of the pump. Additionally, the pump coupling was not designed for sufficient stress margin to failure and this contributed to the second failure of the pump. Not following vendor-specified instructions to fix a safety-related pump and not providing adequate stress safety margin for the pump coupling were found to be major factors to the licensees performance which resulted in two non-compliance findings in a one-year period.

LER 440/03-004-01 2

Condition duration. On August 14, 2003, 1610 hours0.0186 days <br />0.447 hours <br />0.00266 weeks <br />6.12605e-4 months <br />, the ESW A pump was started and ran successfully until August 23, 2003. On September 1, 2003, the ESW A pump was restarted.

The pump ran for 42 minutes and then failed due to pump shaft coupling sleeve failure. The ESW A pump was declared inoperable on September 1, 2003, at 1717 hours0.0199 days <br />0.477 hours <br />0.00284 weeks <br />6.533185e-4 months <br />. The pump was repaired and declared operable on September 5, 2003, at 1855 hours0.0215 days <br />0.515 hours <br />0.00307 weeks <br />7.058275e-4 months <br />. The reactor was operating during the entire period between August 23 at 0607 hours0.00703 days <br />0.169 hours <br />0.001 weeks <br />2.309635e-4 months <br /> and September 5, 2003, at 0655 hours0.00758 days <br />0.182 hours <br />0.00108 weeks <br />2.492275e-4 months <br />. So, the operating condition involving the inoperable ESW A pump existed for 13.46 days (323 hours0.00374 days <br />0.0897 hours <br />5.340608e-4 weeks <br />1.229015e-4 months <br />).

On May 21, 2004, 0148 hours0.00171 days <br />0.0411 hours <br />2.44709e-4 weeks <br />5.6314e-5 months <br />, the ESW A pump was started for surveillance testing. The pump ran for about 2 minutes and then failed because the uppermost split ring coupling broke in half.

Region III documented in their inspection finding on this second failure event that "the primary cause for this failure was related to the cross-cutting issue of problem identification and resolution in that the licensee neither understood nor corrected the design deficiencies associated with the coupling" when it failed the first time on September 1, 2003. The pump was declared inoperable on May 21, 2004, at 0152 hours0.00176 days <br />0.0422 hours <br />2.513228e-4 weeks <br />5.7836e-5 months <br />. Following the failure, the licensee shut down the plant to replace the pump. The pump was repaired and declared to be operable on May 29, 2004, at 0513 hours0.00594 days <br />0.143 hours <br />8.482143e-4 weeks <br />1.951965e-4 months <br />. Region III inspectors found that the second failure event was due to the same cause as the first failure event (coupling installation error).

The ESW A pump ran for over 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> during maintenance on the control complex chillers from April 24, 2004, until 1732 hours0.02 days <br />0.481 hours <br />0.00286 weeks <br />6.59026e-4 months <br /> on May 13, 2004. Between May 13 and the second failure event on May 21, 2004, the ESW A pump ran for an additional 0.46 days (11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />) intermittently. The pump never ran for a 24-hour period continuously. So, it is questionable whether the ESW A pump would have been able to operate for a 24-hour mission time on demand during this period. From about 0435 hours0.00503 days <br />0.121 hours <br />7.19246e-4 weeks <br />1.655175e-4 months <br /> on May 13, 2004, until shutdown cooling was initiated at 0307 hours0.00355 days <br />0.0853 hours <br />5.076058e-4 weeks <br />1.168135e-4 months <br /> on May 23, 2004, a period of 9.96 days (239 hours0.00277 days <br />0.0664 hours <br />3.95172e-4 weeks <br />9.09395e-5 months <br />), it is assumed that the ESW A pump would not have been able to run for its 24-hour mission time on demand.

Therefore, the second operating condition involving the inoperable ESW A pump existed for a net period of 9.5 days (9.96 days - 0.46 days) or 228 hours0.00264 days <br />0.0633 hours <br />3.769841e-4 weeks <br />8.6754e-5 months <br />.

Since the same pump was inoperable due to run-failures on two separate occasions within a one-year period, the total inoperability period for the ESW A pump was found to be 22.96 days (551 hours0.00638 days <br />0.153 hours <br />9.11045e-4 weeks <br />2.096555e-4 months <br />).

Related event. None.

Recovery opportunity. Given a transient or loss of offsite power event, the ESW A pump shaft failure would have been considered to be non-recoverable in a timely manner.

Other related conditions or events during the condition period. A review of Region III-issued green SDP findings (finding with less than 1E-6 delta CDF values) for the same condition period was conducted in identifying potential overlapping operating conditions. It was found that none of the green SDP findings was applicable for evaluation of combined overlapping conditions for the same condition period.

LER 440/03-004-01 1 Since this condition did not involve an actual initiating event, the parameter of interest is the measure of the incremental change between the conditional probability for the period in which the condition existed and the nominal probability for the same period but with the condition nonexistent and plant equipment available. This incremental change or importance is determined by subtracting the CDP from the CCDP. This measure is used to assess the risk significance of hardware unavailabilities especially for those operating conditions where the nominal CDP is high with respect to the incremental change of the conditional probability caused by the hardware unavailability.

3 Analysis Results Importance1 The risk significance of the potentially inoperable ESW A pump due to shaft reassembly problem for a condition duration of 22.96 days (551 hours0.00638 days <br />0.153 hours <br />9.11045e-4 weeks <br />2.096555e-4 months <br />) was determined by subtracting the nominal core damage probability (point estimate) from the conditional core damage probability (point estimate):

Conditional core damage probability (CCDP) =

1.6E-6 Nominal core damage probability (CDP) =

4.5E-7 Importance (CDP = CCDP - CDP) =

1.2E-6 The estimated importance (CCDP-CDP) for the operating condition was 1.2E-6.

A uncertainty analysis was conducted for the operating condition. The mean estimates for CCDP, CDP, and importance were 1.629E-6, 4.558E-7, and 1.173E-6 respectively.

Dominant sequence Loss of condenser heat removal event followed by successful reactor scram, successful reclosure of safety relief valves, successful Feed Water system, failure of suppression pool cooling, failure of the Containment Spray system, failure of PCS recovery, and failure of containment venting.

Sequence LOCHS-07; Importance was estimated to be 3.5E-7. The events and important component failures in this sequence were as follows:

- Loss of condenser heat removal event

- successful reactor scram,

- successful reclosure of safety relief valves,

- successful Feed Water system,

- failure of suppression pool cooling,

- failure of the Containment Spray system,

- failure of PCS recovery, and

- failure of containment venting

- Onset of potential core damage Success-failure path for dominant sequence LOCHS-07 is shown Figure 1.

LER 440/03-004-01 4

Results tables Table 1 provides the conditional probabilities for 2 dominant sequences.

Table 2a provides the event tree sequence logic for the dominant sequences listed in Table 1.

Table 2b provides the definitions of fault trees used in event tree logic listed in Table 2a.

Table 3 provides the conditional (CCDP) cut sets for 2 dominant sequences.

Table 4 provides the definitions and probabilities for added basic events and condition-affected basis events.

Modeling Assumptions Assessment summary Assessment type - This event was modeled as an at-power condition assessment with the ESW A pump run failure for a 23-day period (551 hours0.00638 days <br />0.153 hours <br />9.11045e-4 weeks <br />2.096555e-4 months <br />).

Condition modeling and related assumptions -

1. Given a demand for the ESW A pump to run, it would have failed due to shaft failure (operating condition). The ESW A pump run failure would not have been recovered since the shaft failure could not have been recovered in a timely manner.

Model use - The Revision 3.11 Standardized Plant Analysis Risk (SPAR) model for Perry Nuclear Power Plant (Reference 6) was used for this condition assessment.

Model update to Revision 3.11 SPAR model -

CFAILED (CONTAINMENT FAILURE CAUSES LOSS OF ALL INJECTION) - In the baseline plant model, this event was judged to have a probability of 0.5. No references to a physical analysis (e.g., ultimate pressure capacity analysis) and/or structure-thermal hydraulic calculations to support the 0.5 probability assignment were documented in the Perry plant model documentation by INEEL staff as part of SPAR model update project.

After CVS (containment venting fails) failure event occurs, no more coolant injection could be provided to vessel. Saturated pool water may not cool the core adequately.

MARK III containment may not be in intact once containment (drywell) failure pressure reaches beyond the design pressure due to venting failure. A core damage event would be onset if CVS failure would occur. This CVS failure-based core damage finding is consistent with the plant models for other BWR-6/MARK III plants (e.g., Clinton, River Bend Station, Grand Gulf). So, basic event CFAILED was set to TRUE in the baseline plant model.

LER 440/03-004-01 5

Basic event probability changes -

Table 4 provides the basic events that were modified to reflect the operating condition being analyzed. The bases for these changes are as follows:

Given a demand (Transient or a LOOP event), the ESW A pump might have failed to run after its successful start. Operators could have been forced to use the ESW B pump for EDG B cooling and other decay heat removal cooling through the ESW B pump and its cooling loop.

SSW-MDP-FR-1A - This was set to TRUE to reflect the operating condition (the ESW A pump failed to run due to its shaft failure).

Uncertainty analysis and range for total importance due to operating condition -

The parameter estimates and the uncertainties regarding the numerical estimates of the parameters used in the model (parameter uncertainty) are calculated. These data and uncertainty distributions are then propagated through the modified version of the Revision 3.11 SPAR model for PNPP (Reference 6) to produce statistical uncertainty estimates.

Uncertainty analysis of the operating condition along with parameters was performed using the SAPHIRE code (Version 7.22). Default distribution types for applicable initiating events (e.g. loss of offsite power, transients) and basic events for components were documented in the Revision 3.11 SPAR model for PNPP. These uncertainty estimates and uncertainty estimates for condition-affected basic events were used in estimating mean condition-CDP values and mean condition-CCDP values. Other statistical values such as point estimates, 5% estimates, and 95% estimates were also calculated for CDP and CCDP analysis cases. Estimated statistical values for the operating condition are shown in Table 5.

LER 440/03-004-01 6

References 1.

FirstEnergy Nuclear Operating Company, LER 440-2003-004-01, "Emergency Service Water Pump Upper Shaft Coupling Sleeve Failure," dated January 29, 2004.

2.

USNRC, Region III, "Perry Nuclear Power Plant, NRC Integrated Inspection Report 50-440/2003-006," dated October 30, 2003. (ADAMS Accession No. ML033040217) 3.

USNRC Office of Enforcement, "Final Significance Determination for a White Finding (NRC Inspection Report 50-440/2004-005) (Perry Nuclear Power Plant) - EA-03-197,"

dated January 28, 2004. (ADAMS Accession No. ML040280577) 4.

FirstEnergy Nuclear Operating Company, LER 440-2004-001-01, "Emergency Service Water Pump Upper Shaft Coupling Sleeve Failure," dated September 18, 2004.

5.

USNRC, Region III, "Perry Nuclear Power Plant, NRC Special Inspection Report 50-440/2004-011," dated July 2, 2004. (ADAMS Accession No. ML041900080) 6.

Robert Buel, et al., "Standardized Plant Analysis Risk (SPAR) Model for Perry Nuclear Power Plant (Version 3.11)," by Idaho National Engineering and Environmental Laboratory, December 2004.

LER 440/03-004-01 7

Table 1. Conditional probabilities (point values) for dominant sequences Event tree name Sequence no.

Conditional core damage probability (CCDP)

Core damage probability (CDP)

Importance (CCDP - CDP)2 LOCHS 07 6.8E-7 3.4E-7 3.5E-7 LOOP 34-09 3.2E-7 1.5E-8 3.0E-7 Total (all sequences)1 1.6E-6 4.5E-7 1.2E-6 Notes:

1.

Total CCDP and CDP includes all sequences (including those not shown in this table).

2.

Importance is calculated using the total CCDP and total CDP from all sequences of all applicable event trees. Sequence level importance measures are not additive.

Table 2a. Event tree sequence logic for dominant sequences Event tree name Sequence No.

Logic

(/ denotes success; see Table 2b for top event names)

LOCHS 07

/RPS * /SRV * /MFW

  • PCSR
  • CVS

/RPS

  • HCS * /RCI * /DEP * /VA01
1. / indicates that top event is a success event in the event tree logic

LER 440/03-004-01 8

Table 3a. CCDP cut sets for LOCHS Sequence 07 CCDP Percent contribution Minimal cut sets1 Event Tree: LOCHS, Sequence 07 3.208E-07 46.91 PCS-XHE-XL-LTTRAN

  • RHR-XHE-XM-ERROR
  • CVS-XHE-XM-VENT1 8.806E-08 12.88 PCS-XHE-XL-LTTRAN
  • RHR-MDP-TM-TRNB
  • CVS-XHE-XM-VENT 6.800E-07 Total2 Table 3b. CCDP cut sets for LOOP Sequence 34-09 CCDP Percent contribution Minimal cut sets1 Event Tree: LOOP, Sequence 34-09 1.384E-07 42.56 SSW-MDP-CF-RUN
  • EPS-XHE-XL-NR07H
  • OEP-XHE-XL-NR07H 3.774E-08 11/94 EPS-DGN-FR-DGB
  • EPS-DGN-FR-DGC
  • EPS-XHE-XL-NR07H
  • OEP-XHE-XL-NR07H 3.200E-07 Total2 Notes:

1.

See Table 4 for definitions and probabilities for the basic events.

2.

Total CCDP includes all cut sets (including those not shown in this table).

Table 4. Definitions and probabilities for added basic events and condition-affected basis events Basic Event Name Description Probability Modified PCS-XHE-XL-LTTRAN OPERATOR FAILS TO RECOVER THE MAIN CONDENSER 1.000E+0 NO RHR-XHE-XM-ERROR OPERATOR FAILS TO START/CONTROL RHR 5.000E-4 NO CVS-XHE-XM-VENT1 OPERATOR FAILS TO VENT CONTAINMENT (DEP EVT) 5.100E-2 NO RHR-MDP-TM-TRNB RHR TRAIN B IS UNAVAILABLE BECAUSE OF MAINTENANCE 7.000E-3 NO CVS-XHE-XM-VENT OPERATOR FAILS TO VENT CONTAINMENT 1.000E-3 NO SSW-MDP-CF-RUN ESW PUMPS FAIL FROM COMMON CAUSE TO RUN 8.230E-7 NO EPS-XHE-XL-NR07H OPERATOR FAILS TO RECOVER EDG IN 7 HOURS 2.970E-1 NO OEP-XHE-XL-NR07H OPERATOR FAILS TO RECOVER OFFSITE POWER IN 7 HRS 1.365E-1 NO EPS-DGN-FR-DGB DIESEL GENERATOR B FAILS TO RUN 2.117E-2 NO EPS-DGN-FR-DGC DIESEL GENERATOR C FAILS TO RUN 2.117E-2 NO CFAILED CONTAINMENT FAILED TRUE NO (Note 1)

SSW-MDP-FR-PUMPA SSW PUMP A FAILS TO RUN TRUE YES (Note 2)

Notes:

1. Basic event probability is changed in the baseline plant model. Bases for change is documented in Basic event probability changes section of this report.
2. Basic event probability is changed to reflect the operating condition.

LER 440/03-004-01 9

Table 5. Uncertainty estimates for the operating condition Plant: Perry Nuclear Power Plant IR ID: 50-440/2003-006, 50/440/2004-011 SDP: EA-03-197 LER ID : 440-2003-004-01, 440-2004-001-01 Analysis type = Monte Carlo Samples = 10000; Seeds = 97453 Initiating event (IE)

IE ID Point estimate Mean estimate 5% estimate 50% estimate 95% estimate All internal initiating events CCDP for 1 year 2.550E-05 2.590E-05 2.448E-06 1.556E-05 8.340E-05 CDP for 1 year 7.031E-06 7.247E-06 2.069E-07 2.098E-06 2.981E-05 CCDP for 551 hours0.00638 days <br />0.153 hours <br />9.11045e-4 weeks <br />2.096555e-4 months <br /> 1.604E-06 1.629E-06 1.540E-07 9.787E-07 5.246E-06 CDP for 551 hours0.00638 days <br />0.153 hours <br />9.11045e-4 weeks <br />2.096555e-4 months <br /> 4.422E-07 4.558E-07 1.301E-08 1.320E-07 1.875E-06 Importance for 551 hours0.00638 days <br />0.153 hours <br />9.11045e-4 weeks <br />2.096555e-4 months <br /> 1.162E-06 1.173E-06 1.410E-07 8.468E-07 3.371E-06

LER 440/03-004-01 10 LI LATE INJE CTION CVS CONTAINMENT VENTING PCSR POW ER CONVERSION SY STEM RECOVE RY CSS CONTA INMENT SPRAY S PC S UPPRESSION P OOL COOLING VA ALTERNA TE LOW PRESS INJECTION LPI LOW PRESSURE INJE CTION DEP MANUAL REA CTOR DEP RES S HCS HPCS SPC SUP PRES SION POOL COOLING (EARLY)

RCI RCIC MFW FEEDW ATER S RV S RV'S CLOSE RP S RE ACTOR PROTECTION SYSTEM IE-LOCHS SS OF CONDENSE R HEA T S INK END-STATE 1

OK 2

OK 3

OK 4

OK 5

CD 6

OK 7

CD 8

OK 9

OK 10 OK 11 OK 12 CD 13 OK 14 CD 15 OK 16 OK 17 OK 18 CD 19 OK 20 CD 21 OK 22 OK 23 OK 24 CD 25 OK 26 CD 27 CD 28 CD 29 OK 30 OK 31 OK 32 OK 33 CD 34 OK 35 CD 36 OK 37 OK 38 OK 39 OK 40 CD 41 OK 42 CD 43 OK 44 OK 45 OK 46 OK 47 CD 48 OK 49 CD 50 CD 51 CD 52 T

1S ORV 53 T

2S ORVS 54 T

ATWS P 1 P 2 LI00 LI00 LI00 LI01 LI01 LI01 LI00 LI01 LI00 LI00 LI00 LI01 LI01 LI01 Figure 1. Transient Event Tree Showing Sequence LOCHS 07

Retrieved from "https://kanterella.com/w/index.php?title=ML051190665&oldid=5384250"