ML21029A319

Final Accident Sequence Precursor Analysis - Quad Cities Nuclear Power Station, Unit 2, Electromatic Relief Valve 3D Failed to Actuate Due to Out-of-Specification Plunger (LER 265-2020-002) - Precursor

ML21029A319
Person / Time
Site:	Quad Cities
Issue date:	02/18/2021
From:	Dale Yeilding NRC/RES/DRA/PRB
To:
Dale Yeilding
References
LER 265-2020-002
Download: ML21029A319 (6)

Text

ASP Program Analysis - Precursor Accident Sequence Precursor Program - Office of Nuclear Regulatory Research Quad Cities Nuclear Electromatic Relief Valve 3D Failed to Actuate Due to Out-of-Power Station, Unit 2 Specification Plunger LER: 265-2020-002 Event Date: March 30, 2020 CDP = 3x10-5 IR: 05000265/2020002 Plant Type: General Electric Boiling Water Reactor (BWR) 3 with a Mark 3 Containment Plant Operating Mode Mode 4 (0% Reactor Power)

(Reactor Power Level):

Analyst: Reviewer: Completion Date:

Dale Yeilding Christopher Hunter 2/18/2021 1 EXECUTIVE

SUMMARY

On March 30, 2020, with Unit 2 in shutdown for a refueling outage, electromatic relief valve 2-0203-3D of the automatic depressurization system (ADS) failed to actuate during performance of a technical specification (TS) surveillance test. All other Unit 2 ADS valves were successfully cycled open and closed. Licensee investigation revealed the cause of the failure was an out-of-specification plunger received from the vendor. The failed actuator for the ADS valve had been originally installed in March 2018 and the licensee determined that the ADS valve was inoperable for the entire fuel cycle.

This accident sequence precursor (ASP) analysis reveals that the most likely core damage sequences involve internal event scenarios that result in a loss of service water transient with failure of SRVs to close and failure of all high-pressure injection sources [reactor core isolation cooling (RCIC) and high-pressure coolant injection (HPCI)]. This accident sequence accounts for approximately 55 percent of the increase in the core damage probability (CDP) for the event. The mean CDP for this event is 3x10-5, which is dominated from internal event scenarios (approximately 97 percent of the total CDP). The risk from internal floods, seismic events, high winds, and tornados, contributes less than 3 percent to the total CDP for this event. Due to the lack of modeling of internal fires, the impact of internal fires is considered qualitatively.

No licensee performance deficiency (PD) associated with this event was identified and, therefore, an ASP analysis was performed since a Significance Determination Process (SDP) evaluation was not performed.

2 EVENT DETAILS 2.1 Event Description On March 30, 2020, with Unit 2 in shutdown for a refueling outage, electromatic relief valve 2-0203-3D of the ADS failed to actuate during performance of a TS surveillance test. All other Unit 2 ADS valves were successfully cycled open and closed. A Licensee investigation revealed the cause of the failure was an out-of-specification plunger received from the vendor.

1

LER 265-2020-002 The failed actuator for the ADS valve had been originally installed in March 2018 and the licensee determined that the ADS valve was inoperable for the entire fuel cycle.

The failed ADS valve actuator was dissembled and inspected to determine the cause of the failure. The plunger width was found to be different from the top to mid-length indicating the legs were splayed and this plunger width was outside the manufacturer's tolerance. The plunger width was not a parameter that was measured during vendor dedication until 2019, after the rebuild of the actuator in 2018. Additional information is provided in licensee event report (LER) 269-2020-002, "Electromatic Relief Valve 3D Did Not Actuate Due to Out of Specification Plunger, (ADAMS Accession No. ML20149K600).

2.2 Cause Licensee investigation revealed the cause of the failure was an out-of-specification plunger received from the vendor.

3 MODELING 3.1 Basis for ASP Analysis The ASP Program uses SDP results for degraded conditions when available (and applicable).

Region 3 did not identify a licensee PD associated with the ADS valve failure. NRC inspectors determined that the failure was not reasonably foreseeable and preventable by the licensee and, therefore, is not a performance deficiency. The LER was reviewed and closed in inspection report 05000265/2020002, Integrated Inspection Report 05000254/2020002 and 05000265/2020002, (ADAMS Accession No. ML20225A008). An independent ASP analysis was performed because no licensee PD was identified and the potential for risk significance of this event.

Additional LERs were reviewed to determine if unavailabilities existed concurrently with the ADS valve failure. One windowed event of a concurrent degraded condition was identified.

LER 265-20-001, "Loss of Both Divisions of Residual Heat Removal Low Pressure Coolant Injection Due to Swing Bus Failure to Transfer, (ADAMS Accession No. ML20140A093),

reported a motor control center (MCC) failed to transfer between buses during the performance of testing. This resulted in the licensee declaring both loops of low-pressure coolant injection (LPCI) inoperable. The failure of the auto-transfer would prevent the LPCI injection valves from opening. The apparent cause of the event was a manufacturing error in the LPCI swing bus time delay relay. Prior to installation of the relay in 2018, bench testing and calibration were successfully performed. An analysis assuming the failed MCC auto-transfer and ADS valve for a maximum of 1 year shows that the MCC auto-transfer has a negligible impact on the risk of ADS valve failure and, therefore, is not further considered in this ASP analysis.

3.2 Exposure Time The Licensee reported that the ADS valve was found to be inoperable for the entire operating cycle. Therefore, the exposure time was set to the maximum of one year in accordance with of Volume 1 of the RASP Handbook, "Risk Assessment of Operational Events (RASP) Handbook,"

Revision 2.02, (ADAMS Accession No. ML17348A149).

3.3 Analysis Type A condition analysis was performed using Revision 8.59 of the standardized plant analysis risk (SPAR) model for Quad Cities Nuclear Power Station (Unit 1) Test and Limited Use Model 2

LER 265-2020-002 (TLU) 1, revision dated September 24, 2020. Note that this Unit 1 model is being used for a Unit 0F 2 analysis. This SPAR model includes the following hazards:

• Internal events,

• Internal floods,

• Seismic,

• High winds, and

• Tornados 3.4 SPAR Model Modifications No SPAR model modifications were required for the analysis of this event.

3.5 Analysis Assumptions The following assumptions were required to reflect the plant status in modeling of this condition assessment:

• Basic event ADS-SRV-CC-ERV3D (electromatic relief valve 203-3D fails to open) was set to TRUE to represent the failure of the ADS valve to open.

4 ANALYSIS RESULTS 4.1 Results The mean CDP for this analysis is calculated to be 3x10-5. The ASP Program threshold is 10-6 for degraded conditions; therefore, this event is a precursor.

The parameter uncertainty results for this analysis is provided below:

Table 1. Parameter Uncertainty Results 5% Median Pt. Estimate Mean 95%

1x10-6 2x10-5 3x10-5 3x10-5 1x10-4 4.2 Dominant Hazards 2 1F The dominant hazard is internal events (CDP = 2.8x10-5), which contributes approximately 97 percent of the total CDP. The following table provides the contribution of all hazards that are included in the Quad Cities SPAR model.

1 If base SPAR model changes are needed to reflect the as-built, as-operated plant or analysis-specific modifications are needed, Idaho National Laboratory (INL) will create a TLU model that allows for the completion of analyses in a timely fashion. TLU models do not undergo the full complement of quality assurance checks that the models of record require before being posted for normal use.

2 The CDPs in Sections 4.2 and 4.3 are point estimates.

3

LER 265-2020-002 Table 2. Dominant Hazards Hazard CDP  %

Internal Events 2.8x10-5 97.3 Internal Flood 6.4x10-7 2.2 Seismic 1.6x10-7 < 1%

High Winds 2.3x10-8 Negligible Tornados 2.6x10-9 Negligible The lack of modeling of internal fires scenarios is a key uncertainty and is considered qualitatively in Section 4.4.

4.3 Dominant Sequences The dominant accident sequence is loss of service water 34 (CDP = 1.6x10-5), which contributes approximately 55 percent of the total CDP. The sequences that contribute at least 5 percent to the total CDP are provided in the following table. The event tree with the dominant sequence is shown graphically in Figure A-1 of Appendix A.

Table 3. Dominant Sequences Sequence CDP  % Description LOSWS 34 1.6x10-5 53.7 Loss of service water initiating event; both RCIC (random) and HPCI (service water dependent) fail; the safe shutdown system is unavailable due to the loss of service water; and common-cause failure (CCF) of ADV valves results in failure to depressurize the reactor resulting in core damage.

LOCHS 29 3.1x10-6 10.7 Loss of condenser heat sink initiating event; RCIC or HPCI are successful; suppression pool cooling fails; and CCF of ADV valves results in inability to depressurize the reactor resulting in core damage.

SLOCA 33 3.0x10-6 10.2 Small loss-of-coolant initiating event; HPCI fails; and CCF of ADV valves results in inability to depressurize the reactor resulting in core damage.

LOMFW 23 1.7x10-6 5.7 Loss of main feedwater initiating event; both RCIC and HPCI fail; suppression pool cooling fails; and CCF of ADV valves results in inability to depressurize the reactor resulting in core damage.

4.4 Key Uncertainties The key modeling uncertainty for this analysis is lack of internal fire scenario in the Quad Cities SPAR model. To address this uncertainty, the risk information provided by the licensee for various risk-informed applications (e.g., technical specification change). Quad Cities currently does not have a licensee PRA for internal fires and references its individual plant examination for external events (IPEEE) for a quantitative assessment of the risk associated with internal fires.

4

LER 265-2020-002 Fire scenario(s) that would significantly impact the risk of this event (i.e., a single ADS failure to open) require a fire to result in a partial loss of ADS function due to a failure of at least 1 of the 4 unaffected ADS valves. In addition, dominant fire scenarios that do not directly affect the ADS function but increase its importance (e.g., internal fires that affected RCIC and/or HPCI) could significantly impact the risk of this event. Fire scenarios that result in a complete loss of ADS function or spurious ADS valve actuation would not result in a risk increase for this event. A review of the Quad Cities IPEEE results did not reveal significant fire scenarios that would result in an additional loss of ADS valve redundancy (i.e., failure of the four unaffected ADS valves) without a complete loss of system function.

A quantitative estimate of the risk impact for internal fire scenarios in which ADS is a key mitigation feature could not be determined from available information within the IPEEE.

However, it is not expected that the impact from applicable internal fire scenarios would result in the overall CDP to increase to the significant precursor threshold of greater than or equal to 10-3. It is likely that the CDP, including internal fires, would remain in the 10-5 range.

5

LER 265-2020-002 Appendix A: Key Event Tree Figure A-1. Quad Cities Loss of Service Water Event Tree A-1