GO2-17-101, Response to Request for Additional Information (RAI) Related to One-Time 7 Day Extension of Completion Time for TS 3.5.1.A, 3.6.1.5.A, and 3.6.2.3.A
| ML17192A455 | |
| Person / Time | |
|---|---|
| Site: | Columbia  |
| Issue date: | 07/11/2017 |
| From: | Javorik A Energy Northwest |
| To: | Document Control Desk, Office of Nuclear Reactor Regulation |
| References | |
| GO2-17-101 | |
| Download: ML17192A455 (39) | |
Text
Alex L Javorik Columbia Generating Station P.O. Box 968, PE23 Richland, WA 99352-0968 Ph. 509.377. 2354 l F. 509.377. 8555 aljavorik @energy-northwest.com July 11, 2017 GO2-17-101 10 CFR 50.90 U.S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, DC 20555-0001
Subject:
COLUMBIA GENERATING STATION, DOCKET NO. 50-397 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION (RAI)
RELATED TO ONE-TIME 7 DAY EXTENSION OF COMPLETION TIME FOR TS 3.5.1.A, 3.6.1.5.A, AND 3.6.2.3.A
References:
- 1. Letter,GO2-16-119, A. L. Javorik (Energy Northwest) to NRC, "License Amendment Request for One-Time 7 Day Extension of Completion Time for TS Condition 3.5.1.A, 3.6.1.5.A, and 3.6.2.3.A," dated November 8, 2016 (ADAMS Accession Number ML16313A573)
- 2. Email, John Klos (NRC) to Lisa Williams (Energy Northwest),
"Columbia LAR extend RHR outage time 7 to 14 days, MF8794, formal release of RAIs," dated May 31, 2017 (ADAMS Accession Number ML17151A224)
Dear Sir or Madam:
By Reference 1, Energy Northwest submtted a License Amedment Request to extend the completion time by 7 days for Technical Specification (TS) actions 3.5.1.A, 3.6.1.5.A, and 3.6.2.3.A on a one-time basis. This one-time extension will be used to support preventive maintenance (PM) under Engineering Change (EC) 14635 which replaces the Residual Heat Removal (RHR) Division A subsystems pump and motor.
By Reference 2 the Nuclear Regulatory Commission (NRC) requested additional information related to the Energy Northwest submittal. Enclosure 1 to this letter contains the requested information for APLA RAI 01 through 05. Enclosure 2 to this letter contains the requested information for the remaining APL RAIs.
No new commitments are being made by this letter or the enclosure. If there are any questions or if additional information is needed, please contact Ms. L. L. Williams, Licensing Supervisor, at 509-377-8148.
The enclosed information does not alter the No Significant Hazards Consideration determination in the original submittal.
G02-17-101 Page 2of2 I declare under penalty of perjury that the foregoing is true and correct.
Executed this Respectfully, LI ~ay of :f"'/j , 2017.
Enclosures:
As stated cc: NRC RIV Regional Administrator CD Sonoda - BPA/1399 (email)
NRC NRR Project Manager WA Horin - Winston & Strawn NRC Senior Resident lnspector/988C
GO2-17-101 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION APL RAI 01 through APLRAI 05
COR A divisio n of F.~ E N E RC 0 N Response to RAls for a One-Time RHR A AOT Extension Columbia Generating Station Revision 0 Preparodb) : ~ @1/;1I Re\'ie\\cd b) : E~
I labib Shtaih. Energ) Northwest R~vie\\edh): -
~~~> /,,h1A~
~---~-~ --....-
""fO"'/Ff Blake Sm ith, Energ) North\\ est Page 1 I
Response to RAIs for a OneTime RHR A AOT Extension By letter dated November 8, 2016 (Agencywide Documents Access and Management System (ADAMS)
Accession No. ML16313A573), Energy Northwest (the licensee) submitted a License Amendment Request (LAR) for Columbia Generating Station, which proposed a onetime extension of increasing the Completion Time (CT) currently specified in Technical Specification (TS) conditions 3.5.1.A, 3.6.1.5.A, and 3.6.2.3.A from 7 days to 14 days for restoring Residual Heat Removal (RHR) Train A. The extended CT will allow sufficient time to complete the preventive maintenance to install a new pump and motor in the RHR Train A subsystem, which is planned for December 2017.
The Probabilistic Risk Assessment Licensing Branch (APLA) has reviewed the LAR and has identified areas where additional information is needed to complete its review. Responses to Request for Additional Information (RAI) 01 through 05 are provided below.
REFERENCES
- 1. Columbia Generating Station Unit 2 PRA Peer Review Report Using ASME PRA Standard Requirements, BWR Owners Group, 2009.
- 3. CGS PostFire Safe Shutdown Analysis NE028519, Rev. 11, 2016.
- 4. CGS System Impacts on PostFire Safe Shutdown, NE029435, Rev. 4, 2015.
- 5. CGS FPRA Quantification and Results, FPSA1RE0001, Rev. 2, 2006.
- 8. ASME/ANS RASa-2009, Addenda to ASME/ANS RAS-2008 Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications.
- 9. Guidance for Post Fire Safe Shutdown Circuit Analysis, NEI 0001, Rev. 4, 2016.
- 10. RG 1.200, An Approach for Determining the Technical Adequacy Probabilistic Risk Assessment Results for RiskInformed Activities, Rev. 2, 2009.
- 11. Fire PRA Methodology for Nuclear Power Facilities, NUREG/CR6850, 2005.
- 12. Practical Guidance on the Use of Probabilistic Risk Assessment in RiskInformed Applications with a Focus on the Treatment of Uncertainty. EPRI, Palo Alto, CA: 2012. 1026511.
Page 2
APLA RAI 01 The license amendment request (LAR) for Columbia Generating Station (CGS), dated November 8, 2016, states that the proposed change to the Technical Specification (TS) completion time has been evaluated using the riskinformed processes described in Regulatory Guide (RG) 1.177, An Approach for Plant Specific, RiskInformed Decisionmaking: Technical Specifications, Revision 1. Based on Section 2.3.1 of RG 1.177, the technical adequacy of the probabilistic risk assessment (PRA) must be compatible with the safety implications of the Technical Specification change being requested and the role that the PRA plays in justifying that change. RG 1.177 endorses the guidance provided in RG 1.200, "An Approach for Determining the Technical Adequacy of Probabilistic Risk Assessment Results for RiskInformed Activities, Revision 2, on PRA technical adequacy. RG 1.200 describes a peer review process utilizing American Society of Mechanical Engineers/American Nuclear Society (ASME/ANS) PRA standard RASa 2009, "Addenda to ASME/ANS RAS2008 Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications," as one acceptable approach for determining the technical adequacy of the PRA once acceptable consensus approaches or models have been established for evaluations that could influence the regulatory decision.
Based on Section 5.1 of LAR Attachment 5, the CGS internal events probabilistic risk assessment (IEPRA) underwent a peer review against ASME/ANS RASa2009, as clarified/qualified by Revision 2 of RG 1.200. The peer review was conducted using the industry peer review process guidelines in NEI 0504, Revision 2, Process for Performing Internal Events PRA Peer Reviews Using the ASME/ANS PRA Standard. Section 4.6 of NEI 0504 states:
It should be noted that even in cases where an SR [supporting requirement] has been assessed to meet CC [Capability Category] II or III, the review team may document an F&O [facts and observations]
finding. Such findings are typically for nonsystematic discrepancies that the PRA peer review team judges require correction.
LAR Attachment 6 provides and dispositions findinglevel F&Os associated with SRs that were assessed by this peer review as not meeting CC II. However, no information appears to have been provided regarding findinglevel F&Os associated with SRs that were otherwise met at CC II. Owing to their potential impact to the risk results of the proposed onetime completion time (CT) extension, provide all remaining findinglevel F&Os from the 2009 peer review and associated dispositions. Additionally, clarify whether any changes made to the IEPRA subsequent to the peer review constitute a PRA upgrade as defined by ASME/ANS RASa2009. For those changes that constitute a PRA upgrade, if any, indicate whether a focusedscope peer review(s) has been performed for those upgrades. As applicable, provide the findinglevel F&Os from the peer review(s) and explain how the F&Os were dispositioned for this application. If a focusedscope peer review(s) was not performed for these upgrades, then provide a qualitative or quantitative evaluation (e.g., sensitivity or bounding analysis) of its effect on the risk assessment results.
Response
Table RAI011 provides all remaining findinglevel F&Os from the 2009 peer review and associated dispositions. All changes made to the IEPRA subsequent to the peer review have been reviewed. None of these changes constitute a PRA upgrade as defined by ASME/ANS RA Sa2009.
Page 3
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 115 SupR: TIA initiating event does not appear to accurately credit Two possible choices: a) Resolved: YES ASA5 both nonsafety and safety instrument air. The initiating Model the TIA initiating event The peer review finding was resolved. In the updated event frequency appears to be based on the loss of non to impact systems for only PRA model, the TIA initiating event is modeled to safety air, but would not be applicable to failure of both loss of nonsafety air, or b) impact systems for only the loss of nonsafety air, and safety and nonsafety air. See Page 46 of the QU notebook. Modify the initiating event the potential loss of safetyrelated air is now modeled Initial Cutsets appear to fail depressurization, which appears frequency based on a loss of as part of the TIA accident sequence development.
to be from an assumed safety air failure. both safety and nonsafety Depressurization is included in the updated TIA event air, which would be a much tree, as safetyrelated air can support this function.
lower probability.
F&O 128 SupR: RCIC Gate GRCI11A22 (loss of RCIC Pump Room Cooling) Correct the logic under Resolved: YES SYA11 appears to be developed incorrectly. The logic for nonSBO GRCI11A22. Additionally, The peer review finding was resolved. The need for appears to assume a loss of HVAC is required to trip the modify the preinitiator HEP gland seal cooling in the event of a loss of RCIC pump pump on a miscalibration event. However, if the mis to represent miscalibration of room cooling was added to the pump cooling calibration event is bad enough, the trip may occur once the a single instrument under dependency modeling at gate GRCI11A22.
room heats up after start, given HVAC is working. Also in either A or B train gates Miscalibration of RCIC leak detection temperature RCIC under gate GRCI241, a single temperature monitor (GRHI241 and GRIC432). sensors could cause a spurious RCIC isolation, and results in failing closed the RCIC valves 8 and 63. However, Evaluate the potential impacts miscalibration basic events are now modeled in the the preinitiator HEP is for common mode failure of mis of RCIC gland seal leak / appropriate fault tree locations for the failing closed calibration of A and B sensors. In discussion with CGS staff, failure and revise the of RCIC valves 8 and 63. The modeling of it was judged warranted to evaluate and include the modeling as appropriate. temperature sensor miscalibration is not applicable to potential for RCIC gland seal leak and/or failure in this logic, the RCIC pump cooling dependency logic, and was as appropriate. In setting the event RCICHUMNRMTEMP3LL therefore removed from gate GRCI11A22.
to 1.0, the CDF goes up 33%. Miscalibration will be significantly increased when modeled as an independent miscalibration of a single instrument (versus a CCF event). A RCIC gland seal leak / failure could potentially impact the RCIC operability and room cooling requirements.
Page 4
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 129 SupR: Truncation Sensitivity is documented in the QU notebook Perform sensitivity studies to Resolved: YES QUB3 Section 3.3. Table 32 documents a 4% difference in CDF validate the LERF cutoff value, The peer review finding was resolved. Section 6.6 of between 1E011 and 1E12. However, LERF does not with additional runs the Level 2 Notebook was updated to enhance the document that use of 5E12 results in less than 5% increase documented below the LERF truncation sensitivity discussion to include in comparison to a decade drop in the cutoff value. The text present cutoff value of 5E12. additional truncation sensitivity runs. Extending the in 6.6 of the LERF notebook says that dropping the cutoff to truncation to 5E13/yr (an additional decade) results 5E13 is judged to 5% reduction (from 5E12). Since there is in an approximate 5% increase (Table 6.61 shows a a 3% drop from 1E12 to 5E12, and an 11% increase from 6.3% increase) in the LERF risk metric. This is 5E11 to 5E12, the use of 5E12 may not meet the consistent with the guidelines in SR QUB3 of the requirement, depending on the increase at 1E12 or 5E13 ASME/ANS PRA Standard.
(depending on which is calculated.
For the CGS Rev. 7.2 PRA model of record, the LERF Requirement of the standard to verify LERF cutoff is truncation is 2E12/yr. Extending the truncation to adequate. 2E13/yr results in a 4% LERF increase, which is consistent with the guidelines in SR QUB3 of the (This F&O originated from SR QUB3) ASME/ANS PRA Standard.
F&O 212 SupR: In the AC system notebook, PSA2SNAC R3, for the 230 kV Add the consequential LOOP Resolved: YES SYB5 BPA line, the unavailability was calculated to be 2.74E4 event (as a single event) into The peer review finding was resolved.
[EACTRLASHEW3D1, Ref. 90]. For the 115 kV BPA line, the the CGS models. Consequential LOOP modeling, as a single basic event, unavailability was calculated to be 5.48E4 [EACTRL has been added to the CGS PRA model.
BENTNW3D1, Ref. 90]. These events were originally credited as consequential LOOP when the CGS staff was questioned for the treatment of consequential LOOP as discussed in NUREG/CR6890 V1 Section 6.3. A review of the AC system model showed that EACTRLASHEW3D1 & EACTRL BENTNW3D1 should be more appropriately labeled as switchyard failures, which are actually independent from each other and from the postulated turbine trip.
(This F&O originated from SR SYB5)
Page 5
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 226 SupR: While checking the frequency calculation for the dominant Update the walkdown sheets Resolved: YES IFEVA5 flooding initiator IEFLDC502TSWU, the following errors and the calculations The peer review finding was resolved. The following have been identified: associated with flooding resolutions have been made:
- a. The formula in the cell for frequency calculation, initiating event associated D2*'Mean Pipe Failure Rates'!$D$110*'Mean Pipe Failure with the dominant flooding a. The formula in the spreadsheet cell for flood Rates'!$D$148, reflects the SW failure rate and the impact initiator IEFLDC502TSWU. frequency calculation, D2*'Mean Pipe Failure factor. This is incorrect since the flooding scenario is Check other calculations to Rates'!$D$110*'Mean Pipe Failure Rates'!$D$148, associated with the TSW piping break. identify similar issues if exist reflects the standby service water (SW) failure rate
- b. Discussion with CGS PSA staff indicated that the value in (e.g., sampling of calculations and the impact factor. This was noted to be incorrect PSA2FL0003 is incorrect and use the incorrect integrity associated with other flooding since the flooding scenario is associated with the management data. initiators). plant service water (TSW) piping break. This error has
- c. The walkdown sheet in PSA3FL0001 for area C502 has been corrected.
been determined to be incomplete, which missed the 100', b. The integrity management data used for IEFLD 3" TSW pipe listed in PSA2FL0001 Table 4. C502TSWU in the Initiating Events Frequency Development Report, PSA2FL0003, was incorrect.
Per discussion with CGS PSA staff and further review of the The equation in the flood frequency calculation calculations, the identified issues have been determined to spreadsheet and the value in PSA2FL0003 have been be not widespread. The identified flooding scenario is corrected.
dominant. Therefore, the identified issue could be risk c. The walkdown sheet in Walkdown Summary significant. Report, PSA3FL0001, for area C502 was determined to be incomplete, which missed the 100', 3" TSW pipe listed in PSA2FL0001, Table 4. Energy Northwest verified this TSW pipe length in the field and added it to the walkdown sheet for this zone.
- d. Further, all formulas in the flood initiating event spreadsheet were reviewed for correctness. A total of five frequencies were found to be incorrect and were revised. Only one of these initiators increased in frequency.
Page 6
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 23 SupR: It seems that the unavailability of Component Population Perform the Bayesian update Resolved: YES DAC9 and Demand Data for RCIC was used to exclude the for RCIC fail to start. It is also The peer review finding was resolved. Per the Data performance of Bayesian updates. The RCIC pump failure recommended to perform Analysis Notebook, PSA2DA0002, the RCIC would be generally a significant risk contributor. However, Bayesian update for RCIC fail operational history doesnt meet the Bayesian update its Bayesian update was not performed although it is a to run based on general data fit test. There is too little operational experience candidate for Bayesian update based on Risk Significance. estimate on RCIC run hours to perform a Bayesian update. Therefore, the RCIC Further investigation show that Bayesian update for RCIC fail based on the plant general data remained unchanged. The Data Analysis to start data can be performed. practices. Notebook was updated to reflect these insights.
This is an isolated issue, which then do not impact the overall capability category.
(This F&O originated from SR DAC9)
F&O 317 SupR: Sources of Level 2 / LERF model uncertainty are not Add an uncertainty analysis Resolved: YES LEF3 addressed in the L2 Notebook. Instead, the Level 2 notebook and evaluation similar to the The peer review finding was resolved. The Level 2 refers to the QU notebook. There are four prominent Level I QU notebook in the L2 uncertainty analysis was added to the PRA sources of Level 2/LERF uncertainty listed in Table 52 of notebook. Quantification (QU) Notebook and to Appendix K of PSA2QU0001 and only two potential sources of LERF the Level 2 Notebook.
model uncertainty in Table 53.
Given the wide range of phenomenological issues dealt with in Level 2 and the various assumptions made in the analysis, it seems likely that there would be additional plant specific sources of model uncertainty that should be discussed.
Understanding of the impacts of the various sources of uncertainty in the Level 2 model is important. However, it is believed that this is a documentation issue.
(This F&O originated from SR LEF3)
Page 7
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 39 SupR: EDG modeling: the fuel oil supply is modeled but only for Add to the convolution model Resolved: YES SYA6 the 6hr average mission time, no accounting for possibility for LLOP and EDG running for The peer review finding was resolved. Failure of of needing fuel for longer than 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> (i.e., makeup to the 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, the likelihood that makeup from the fuel oil makeup from the individual day tank not modeled). after the EDG day tank is fuel oil storage tanks to the respective EDG fuel oil empty, it would need to be day tank was added to the PRA model.
EDG is significant in the model. refilled.
(This F&O originated from SR SYA6)
F&O 43 SupR: The Initiating Events Notebook notes that no additional Include a precursor review in Resolved: YES IED2 initiating events have been identified as a result of a review the IE notebook. The peer review finding was resolved. The of precursor events. The Initiating Events report did not documentation for precursor review was improved.
elaborate on how the precursor event review was conducted, that is what searches were completed and what if any precursor events were precluded along with a basis.
The discussion about precursors occurred in the IE notebook section dealing with operator interviews. Appendix G documents these interviews, however there is no discussion relative to precursors. This is primarily a level of documentation issue and does not imply that an adequate review was not performed.
F&O 57 SupR: The Columbia PSA configuration control document, SYS4 Add LERF requirements to the Resolved: YES MUB2 34, Rev. 1, does not include discussion or criteria for LERF change and update process. The peer review finding was resolved. SYS434 impacts. Changes in LERF can impact risk applications. This Section 4.2.2 requires addressing significant LERF is a clarification to Section 3.7 to include LERF effects. impacts. Section 6.10 of SYS434 provides LERF requirements to the change and update process.
Page 8
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 61 SupR: No evidence was found that codes and models were used Add a discussion to the Resolved: YES LED1 outside their known limits of applicability, but no clear Deterministic NB identifying The peer review finding was resolved. Section 1.4 to documentation was found identifying and confirming that MAAP operational limits, Volume 1 of the Deterministic Calculations Notebook codes and models were used within known limits of accident sequences that documents MAAP operational limits, accident applicability. Per the selfassessment, there was a previous challenge MAAP limitations, sequences that challenge MAAP limitations, and F&O for CGS for this issue. and cautions in the use of cautions in the use of MAAP results.
MAAP results in that regard.
All computer codes / analyses have limitations regarding Add a discussion in For room heatup calculations, the operational limits of applicability and modeling capabilities. The PRA appropriate PRA limitations of GOTHIC are discussed (Numerical documentation should be enhanced to address such limits documentation to address Application, Inc., Vital Island Analysis of Loss of HVAC related to thermal hydraulic analysis, containment structural similarly for room heat up as Initiating Event) and the conclusion is made that response, and room cooling that can be impact the calculations and containment the analysis is well within the demonstrated application in the PRA. response, as appropriate. capabilities of the software.
(This F&O originated from SR SCB4)
Page 9
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 614 SupR: For loss of decay heat removal sequences and consideration Reevaluate the containment Resolved: YES LEB2 ECCS injection post containment failure, the Level 1 model failure location analyses and The peer review finding was resolved. The likelihood bases the likelihood of containment failure in the wetwell as either make the Level 1 and for wetwell (WW) failure was revised to be consistent 0.33 (BE CFFAILSINJECT) based on the older containment Level 2 modeling consistent in for the Level 1 and Level 2 PRAs. The existing structural evaluation (ME029177). This WW failure this regard. structural analysis (ME029177) is the appropriate location is assumed to result in an adverse environment that and applicable analysis. The Level 2 PRA had fails ECCS injection. The Level 2 models the likelihood of inappropriately credited the biological shield wall to containment failure in the wetwell as 0.04 based on a reduce the likelihood for WW failure.
reappraisal of the older analysis (accounting for impacts from the biological shield wall), as documented in Appendix F of the L2 NB (PSA2L20001). The L2 model uses this WW failure probability only to differentiate release magnitudes.
Based on discussion with CGS staff, the CGS staff estimated that CDF would decrease by 29% (from 1.15E5/yr to 7.8E 6/yr) if the WW failure likelihood was decreased to 0.04 in the Level 1 model. Level 2 LERF results would not be impacted by this change since the sequences under consideration are not early releases, however, other Level 2 results (e.g., late releases) would be impacted based on a decrease of CDF. Discussions determined cutset dependencies between the Level 1 and Level 2 are not impacted since the Level 1 BE in question is not utilized in the Level 2. Investigation of this issue demonstrates that the reanalysis / assumptions related to the impact of the biological shield wall on containment failure for over pressurization scenarios is significant. It is noted, however, that there is also reasonable uncertainty in general associated with the potential for loss of ECCS injection due to a failure low in containment. That is, injection may continue to operate post wetwell failure.
The likelihood of WW failure is modeled differently in the L1 and L2 logic, based on use of different containment failure analyses. CDF may be impacted by 29% (too conservative).
(This F&O originated from SR ASA5)
Page 10
Table RAI011 Disposition of Findings from the 2009 Peer Review for Supporting Requirements Graded as Met Finding Observations Recommendations Resolution F&O 63 SupR: Select IE frequencies updated using Bayesian process (IE NB Correct the IE Bayesian Resolved: YES IEC4 App B). Use of lognormal distribution is consistent with calculations. The peer review finding was resolved. The turbine industry practice. However, two of two of the updates that trip (TT), loss of feedwater (TF) and loss of main were checked do not appear correct. For TT, the very small condenser (TC) Bayesian updates were corrected.
error factor (1.1) would be expected to constrain the Constraints on update parameters were removed to posterior mean to be very close to the prior mean, and the produce the correct updated values.
posterior mean was not as close to the prior mean as expected. Upon review by CGS Staff, it was identified that the spreadsheet used to Bayesian update the initiating events contained certain calculational constraints that were not appropriate for the IE Bayesian update and therefore resulted in incorrect results for these cases.
An attempt to repeat the IE Bayesian update results for Turbine Trip (TT) and loss of FW using different Bayesian update software were unsuccessful, and CGS staff subsequently identified the cause of the incorrect result in the spreadsheet.
(This F&O originated from SR IEC4)
Page 11
APLA RAI 02 The CGS LAR states that the proposed change to the TS completion time has been evaluated using the riskinformed processes described in RG 1.177, Revision 1. Based on Section 2.3.1 of RG 1.177, the technical adequacy of the PRA must be compatible with the safety implications of the Technical Specification change being requested and the role that the PRA plays in justifying that change. RG 1.177 endorses the guidance provided in RG 1.200, Revision 2, on PRA technical adequacy. RG 1.200 describes a peer review process utilizing ASME/ANS RASa2009 as one acceptable approach for determining the technical adequacy of the PRA once acceptable consensus approaches or models have been established for evaluations that could influence the regulatory decision.
In Section 5.2 of LAR Attachment 5, the licensee explains that the fire PRA (FPRA) has not undergone a peer review against ASME/ANS RASa2009, as clarified/qualified by Revision 2 of RG 1.200.
Additionally, the licensee further clarifies that if such a review were to be performed, the FPRA would not meet all supporting requirements of ASME/ANS RASa2009 at a level of CC II, which the NRC staff, in general, anticipates is adequate for the majority of applications. Lastly, although some of the FPRAs methods were updated in 2006 to make use of guidance in NUREG/CR6850, the methodologies of the FPRA appear to still be largely based on the Individual Plant Examination of External Events (IPEEE) model, and there have been significant changes to FPRA methodology since the IPEEE model was issued in 2001.
Identify any gaps between the CGS FPRA model and the "Internal Fire Technical Elements" required by Revision 2 of RG 1.200 that are relevant to this submittal. In doing so, explain why these gaps do not have significant impact on the risk assessment results and the development of compensatory actions used to support this application. This may include discussion of relevant conservatisms in the CGS FPRA model and additional sources of defenseindepth, as well as, the risk significance of each to the application.
Response
CGS FPRA Gaps to Revision 2 of RG 1.200 Relevant to this Submittal Three gaps between the CGS FPRA model and the "Internal Fire Technical Elements" required by Revision 2 of RG 1.200 that are relevant to this submittal have been identified. These gaps, however, do not have significant impact on the risk assessment results and the development of compensatory actions used to support this onetime application, based on significant CGS defenseindepth:
- 1. The CGS FPRA models some of the fireinduced multiple spurious operation scenarios applicable to CGS. However, several multiple spurious operations scenarios applicable to CGS are not modeled by the FPRA. To examine this gap, the generic BWR MSO scenarios applicable to CGS were reviewed, as discussed in Table RAI021. This evaluation concludes that this gap does not have a significant impact on the risk assessment results and the development of compensatory actions used to support this application based on CGS defenseindepth.
Page 12
- 2. The CGS FPRA selects the cables relevant to components credited by the FPRA. The electrical cable location development includes the routing of cables in cable trays as well as cable terminations. However, the location development does not include the portions of FPRArelated cables where these cables pass through electrical conduits.
Cables routed in conduit through a physical analysis unit (PAU) or zone of influence could be damaged by a fire but are not included in the FPRA.
This gap is assessed to not have significant impact on the risk assessment results and the development of compensatory actions used to support this application. Due to the CGS defenseindepth, the probability of a fire in a single PAU impacting the redundant and diverse systems that provide core cooling and decay heat removal is very low. For fire areas where Division 2 (RHR B) provides postfire safe shutdown, the division is free from fire effects and is redundant to RHR A. For fire areas where Division 1 (RHR A) provides postfire safe shutdown (that is, fire areas where Division 2 is not protected from fire damage), this gap does not significantly impact the decision based on CGS defenseindepth as discussed in Table RAI022.
- 3. The CGS FPRA does not model multicompartment (MCA) fire scenarios. This gap is assessed to not have significant impact on the risk assessment results and the development of compensatory actions used to support this application. For MCA scenarios, the probability is very low that a single fire would disable all ECCS trains due to defenseindepth. The total percentage contributions from MCA scenarios to FPRA CDF and LERF tend to be small. MCA scenarios occur from the fire scenarios that produce a hot gas layer, which are a portion of potential scenarios for a PAU, and are produced only if the PAU fails to contain the hot gas layer, due for example to the relatively low probability failure of a penetration barrier. For CGS, it is expected that the contribution to CDF and LERF from MCA scenarios is small, and therefore this gap is not expected to have significant impact on the decision.
There are two MCA scenarios relevant to the RHR A AOT application. One is a hot gas layer occurring in the Division 1 switchgear or electrical equipment rooms that propagates to a Division 2 switchgear or electrical equipment room. These rooms are on a single elevation of the radwaste building. A multicompartment scenario that affects both divisions would be a thirdorder scenario, as a corridor separates the two electrical divisions. This is a very low frequency scenario. A second multicompartment scenario involves a fire in the RHR A room that propagates to the RHR B room. Only fires that produce a hot gas layer and pass through a failed barrier penetration will produce the scenario.
DefenseinDepth There are two trains of low pressure coolant injection (RHR B and RHR C) that are redundant to RHR A and provide core cooling. There are two high pressure injection sources, high pressure core spray (HPCS) and reactor core isolation cooling (RCIC), as well as an additional low pressure injection source, low pressure core spray (LPCS), that are redundant to and diverse from RHR A.
For decay heat removal, there is one train of suppression pool cooling (RHR B) that is redundant to RHR A, as well as containment venting and the main condenser for some fires. Based on Page 13
defenseindepth, the probability is very low that a fire occurring in the plant will fail core cooling or decay heat removal during the period that RHR A is out of service.
Additional Sources of DefenseinDepth FLEX provides additional defenseindepth and is not credited by the FPRA. Five prominent features of the FLEX strategy that reduce internal fire risk at CGS are the following:
- 1. The FLEX strategy makes available either of two FLEX diesel generators, which are capable of recharging plant batteries, if the onsite diesel generators are unavailable.
- 2. The FLEX strategy allows for manual operation of RCIC if dc battery power depletes before an ac power source can be aligned to provide battery charging.
- 3. The FLEX strategy includes a hardened containment vent system, which provides additional decay heat removal redundancy to suppression pool cooling. Hardened containment vent can be implemented from the control room, or locally in the absence of electrical power or control, if necessary.
- 4. The FLEX strategy provides an additional external injection source, which can be connected to any of the three RHR injection lines.
- 5. If RCIC suction is aligned to the suppression pool, then RCIC can continue operation for suppression pool temperatures up to 240 degrees Fahrenheit, which allows for extended operation of RCIC with containment venting, well beyond the 24hour mission time.
Risk Margin The RHR A AOT risk assessment results indicate significant margin to the regulatory thresholds of 1E6 ICCDP and 1E7 ICLERP without compensatory measures and 1E5 ICCDP and 1E6 ICLERP with compensatory measures. The significant margin supports the qualitative assessment of gaps and accommodates additional risk due to the nonRG 1.200 fire PRA for this onetime submittal. Further, CGS will institute additional compensatory measures during the 14day extended outage time for RHRA to provide additional safety margin. These measures will be hourly, roving, fire tours to reduce the likelihood of damaging fires. These fire tours will cover risksignificant fire areas. Roving fire tours provide detection and suppression (prevention) benefits, as well as confirmation of the adherence to transient combustible restrictions. Other compensatory measures will include equipment protection, reduction and management of transient combustibles, and procedural fire risk management actions.
Conclusion Gaps between the CGS FPRA model and the "Internal Fire Technical Elements" required by Revision 2 of RG 1.200 that are relevant to this submittal do not have significant impact on the risk assessment results and the development of compensatory actions used to support this one time RHR A AOT extension, based on three factors:
- 1. CGS plant systems provide significant defenseindepth to provide key mitigating functions,
- 2. FLEX provides further defenseindepth, which is not credited by the FPRA, and Page 14
- 3. CGS will institute compensatory measures during the extended outage time for RHRA to provide additional safety margin.
Page 15
Table RAI021 Evaluation of the Gap Involving Multiple Spurious Operations MSO Generic BWR MSO Scenarios DefenseinDepth Features Scenario with the Potential to Occur at Impact CGS Not modeled by the CGS FPRA:
2a (head vent LOCA),
CGS defenseindepth provides mitigation to address MSO scenario related to loss of reactor 2b (MSIVs fail to close or coolant pressure boundary. Given a loss of reactor coolant pressure boundary, there are two spuriously open),
trains of low pressure coolant injection (RHRB and RHRC), that are redundant to RHR train A 2c (main steam line drain to provide core cooling; there is one high pressure injection source high pressure core spray shutoffs spuriously open),
(HPCS), as well as an additional low pressure injection source, low pressure core spray (LPCS),
Loss of 2d (recirculation pump seal that are redundant to and diverse from RHR A. Based on divisional separation between reactor LOCA),
Divisions 1, 2 and 3, loss of all ECCS trains due to fire in a single PAU or fire area is not credible coolant 2e (SCRAM discharge volume (for Division 2 postfire safe shutdown fire areas, RHR B is free from fire damage; for Division 1 pressure leak),
postfire safe shutdown fire areas, there is significant defenseindepth, as discussed in Table boundary 2f (shutdown cooling RAI022).
isolation valves spuriously open),
For multicompartment fire scenarios, the probability is very low that a single fire would disable 2h (reverse flow through all ECCS trains. Therefore, this gap does not have significant impact on the risk assessment RHR),
results and the development of compensatory actions used to support this application.
N8 (main steam drain valves spuriously open during alternate shutdown cooling)
Page 16
Table RAI021 Evaluation of the Gap Involving Multiple Spurious Operations MSO Generic BWR MSO Scenarios DefenseinDepth Features Scenario with the Potential to Occur at Impact CGS Not modeled by the CGS FPRA:
2l (RHR minimum flow For core cooling, there are two trains of low pressure coolant injection, that are redundant to RHR train A spurious closure), (RHRB and RHRC); there are two high pressure injection sources, RCIC and HPCS, as well as an additional 2p (diversion of RHR to low pressure injection source, LPCS, which are redundant to and diverse from RHR A.
standby service water connection), For the RHR A AOT extension, reactor core cooling MSO scenarios that pose the most potential risk 2u (spurious HPCS impact involve PAUs for which safe shutdown Division 1 (RHR A) is protected from fire damage. For these fire scenarios, if sufficient fire damage occurs to impact Division 2 (RHR B and RHR C), then three trains of operation),
core cooling would be unavailable (RHR A is out of service and RHR B / RHR C are impacted by fire). The 2x (spurious opening of HPCS probability is very low that other sources of core cooling, HPCS, RCIC, and LPCS would all be impacted.
test return valves), The fire areas for which safe shutdown Division 1 is protected from fire damage consist of the following:
2y (spurious opening of RCIC DG3, DG5, M21, M27, M9, R21, RC19, RC1, RC9, RC6, RC7, RC8, RC12, RC13, R4, R6, R7, R18, test flow to CST valves), SW2, and TG12. Based on an analysis of credit by exclusion documented in Table RAI022, all trains of Reactor core 2aa (RCIC diversion through core cooling would not be impacted by fire. The most challenging fire scenarios from an MSO cooling minimum flow line), perspective would be those occurring in TG12. For this area, spurious, runon operation of HPCS caused 2ab (RCIC diversion of by hot short could cause RPV overfill and lead to the loss of RCIC due to water filling the steam lines.
suppression pool suction However, in this scenario, HPCS is successfully providing core cooling and the pump is not impacted by through minimum flow line), the overfill. Furthermore, DG 1 is free from fire damage for fires in TG12, and the cables that support LPCS operation are routed in the radwaste building and the reactor building. Therefore, the cables that 2ai (uncontrolled feedwater support LPCS cannot reasonably exist in the TG12 corridor, and LPCS would not be impacted by TG12 injection), fires, based on credit by exclusion.
2New1 (loss of RCIC high level trip), Based on divisional separation between Divisions 1, 2 and 3, loss of all ECCS trains and RCIC due to fire in 2New4 (drain CST to a single PAU or fire area is not credible. For multicompartment fire scenarios, the probability is very low suppression pool), that a single fire would disable all ECCS trains. Therefore, this gap does not have significant impact on 2New5 (recirculation the risk assessment results and the development of compensatory actions used to support this pumps fail to trip), application.
N6 (pumping CST to suppression pool)
Page 17
Table RAI021 Evaluation of the Gap Involving Multiple Spurious Operations MSO Generic BWR MSO Scenarios DefenseinDepth Features Scenario with the Potential to Occur at Impact CGS Not modeled by the CGS For decay heat removal, suppression pool cooling provided by RHR B and containment venting FPRA: provide redundancy to RHR A. For fire areas in which RHR B is the protected safe shutdown 4a (loss of suppression pool train, one train of suppression pool cooling is free from fire damage. For fire areas in which Decay heat cooling due to RHR valve RHR A is the protected fire safe shutdown train, the redundant RHR B train of suppression pool removal interlock interactions), cooling could be impacted by multiple spurious operation. However, nonhardened 4c (spurious opening of RHR containment vent via the standby gas treatment system and hardened containment vent heat exchanger bypass provide redundancy, and both of these decay heat removal systems can be manually operated valves) locally.
Page 18
Table RAI021 Evaluation of the Gap Involving Multiple Spurious Operations MSO Generic BWR MSO Scenarios DefenseinDepth Features Scenario with the Potential to Occur at Impact CGS For core cooling, there are two trains of low pressure coolant injection (RHRB and RHRC), that are Not modeled by the CGS redundant to RHR train A that provide core cooling; there are two high pressure injection sources, RCIC FPRA: and HPCS, as well as an additional low pressure injection source, LPCS, that are redundant to and diverse 5a (additional components from RHR A.
loaded onto diesel For the RHR A AOT extension, support system MSO scenarios that pose the most potential risk impact generator),
involve PAUs for which Division 1 is protected for safe shutdown. Fire in these areas could potentially 5c (SSW pump operation impact Division 2. For these scenarios, three of the six main core cooling system trains would potentially without flow path), be impacted RHR B, RHR C and RHR A (out of service). However, the probability is very low for the 5f (nonsynchronous remaining three core cooling system trains to be unavailable. Division 1 is the protected train from fire in paralleling of EDG with these areas, and Division 1 electrical distribution and service water remain free from fire damage, which Support offsite power), can support LPCS and RCIC operation. Furthermore, HPCS provides additional redundancy and diversity, systems 5g (nonsynchronous as the HPCS switchgear and electrical distribution are in a separate building (diesel generator building) paralleling of opposite from Division 1 and Division 2 switchgear and electrical distribution (radwaste building), and the HPCS divisions), service water cooling pump is located with the Division 1 service water cooling pump (service water 5h (nonsynchronous pump house A), a building separate from the Division 2 service water cooling pump (service water pump house B).
paralleling of offsite power sources), For decay heat removal, suppression pool cooling provided by RHR B and containment venting provide 5i (EDG operation without redundancy to RHR A. For fire areas in which RHR B is the protected safe shutdown train, one train of cooling water), suppression pool cooling is protected from fire damage. For fire areas in which RHR A is the protected 5l (loss of HVAC due to fire safe shutdown train, the redundant RHR B train of suppression pool cooling could be impacted by spurious operations) multiple spurious operation. However, nonhardened containment vent and hardened containment vent provide redundancy, and both of these decay heat removal systems can be manually operated locally.
Page 19
Table RAI021 Evaluation of the Gap Involving Multiple Spurious Operations MSO Generic BWR MSO Scenarios DefenseinDepth Features Scenario with the Potential to Occur at Impact CGS Modeled:
2i.1 (RHR flow diversion),
2i.2 (RHR flow diversion),
2j (RHR flow diversion),
2k (RHR flow diversion),
2m (RHR flow diversion),
2New9 (spurious closure of RCIC steam line return to Modeled by suppression pool),
the CGS 3a (potential opening of all n/a - The CGS FPRA models these scenarios explicitly FPRA SRVs),
3b (multiple SRV opening),
3c (spurious automatic depressurization),
4b (diversion of suppression pool cooling to drywell spray),
4d (spurious closure of RHR suction valves)
Page 20
Table RAI022 Evaluation of the Gap Involving Electrical Conduits for Fire Areas where Division 1 (RHR A) Provides PostFire Safe Shutdown FPRA Physical Analysis Unit Description DefenseinDepth Features (Fire Safe Shutdown Fire Area)
Core Cooling The three CGS onsite diesel generators are located in the diesel building in divisionallyseparate fire areas. Fire area DG3 houses the Division 2 diesel generator, DG 2. Cables that support HPCS (Division 3), as well as RCIC and LPCS (Division 1) cannot reasonably exist in the DG 2 room.
Based on credit by exclusion1, fires in this room do not impact HPCS, RCIC and LPCS.
DG3 (DG3) DG 441455 DG 2 Room Decay Heat Removal Hardened containment vent is not impacted by fires in the DG 2 room.
Hardened containment vent cables pass from the radwaste building into the turbine generator corridor (fire area TG12) and into the laundry storage area of the DG building, and do not pass into the DG2 room.
Local manual nonhardened containment vent does not rely on electrical power and therefore is not impacted.
Core Cooling Like fire area DG3, cables that support HPCS, RCIC and LPCS cannot reasonably exist in the DG 2 oil pump room. Based on credit by DG5 (DG5) DG 441 DG 2 Oil Pump Room exclusion, fires in this room do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Like fire area DG3, hardened containment vent and local manual non hardened local containment vent are not impacted.
1 Credit by exclusion relies on reasonable assurance that a cable is not located in a physical analysis unit, considering physical layout of plant equipment, as indicated in the note to supporting requirement CSA11 of the 2009 ASME / ANS PRA Standard.
Page 21
Table RAI022 Evaluation of the Gap Involving Electrical Conduits for Fire Areas where Division 1 (RHR A) Provides PostFire Safe Shutdown FPRA Physical Analysis Unit Description DefenseinDepth Features (Fire Safe Shutdown Fire Area)
Core Cooling This room contains instrument rack EIRP021, which contains Division 2 instruments required for safe shutdown (instrumentation for control room and cable spreading room fires). HPCS, RCIC and LPCS do not utilize instruments on EIRP021. Cables that support HPCS, RCIC and LPCS Reactor Building 501 Instrument M21 (M21) cannot reasonably exist in the Reactor Building 501 Instrument Rack Rack Room (H22/P021)
Room. Based on credit by exclusion, fires in this room do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Core Cooling This room contains instrument rack EIRP027, which contains Division 2 instruments required for safe shutdown. HPCS, RCIC and LPCS do not utilize instruments on EIRP027. Cables that support HPCS, RCIC and Reactor Building 522 Instrument LPCS cannot reasonably exist in the Reactor Building 522 Instrument Rack M27 (M27)
Rack Room (H22/P027) Room. Based on credit by exclusion, fires in this room do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Core Cooling This room contains instrument rack EIRP009, which contains Division 2 instruments important for safe shutdown. The functioning of RHR B and Reactor Building 471 Instrument RHR C is not impacted by fires in this area, as redundant channels of M9 (M9)
Rack Room (H22/P009) instrumentation are available.
Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Page 22
Table RAI022 Evaluation of the Gap Involving Electrical Conduits for Fire Areas where Division 1 (RHR A) Provides PostFire Safe Shutdown FPRA Physical Analysis Unit Description DefenseinDepth Features (Fire Safe Shutdown Fire Area)
Core Cooling The south valve room contains RHRV42B, which supports RHR B. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in the Reactor Reactor Building 522 South Building 522 South Valve Room. Based on credit by exclusion, fires in this R21 (R21)
Valve Room room do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Core Cooling Division 2 cables are routed in this radwaste building corridor. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in the Radwaste Building 467 Corridor. Based on credit by exclusion, fires in this RC19 (RC19) RW 467 Corridor C205 corridor do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Core Cooling Division 2 cables pass through this radwaste building elevation to travel from the vital island Division 2 electrical rooms to the cable spreading room. Cables that support HPCS, RCIC and LPCS cannot reasonably exist RC1a (RC1) Radwaste Bldg 437 in Radwaste Building 437. Based on credit by exclusion, fires on this elevation do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Page 23
Table RAI022 Evaluation of the Gap Involving Electrical Conduits for Fire Areas where Division 1 (RHR A) Provides PostFire Safe Shutdown FPRA Physical Analysis Unit Description DefenseinDepth Features (Fire Safe Shutdown Fire Area)
Core Cooling HPCS is not included on the remote shutdown panel. Cables that support HPCS cannot reasonably exist in this room. Based on credit by exclusion, RW 467 Remote Shutdown RC9 (RC9) fires in this room do not impact HPCS.
Room Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Core Cooling This room contains the Division 2 battery. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in this room. Based on credit by RC6 (RC6) RW 467 Battery Room 2 exclusion, fires in this room do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Core Cooling This room contains Division 2 electrical equipment. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in this room. Based on RW 467 Division 2 Electrical RC7 (RC7) credit by exclusion, fires in this room do not impact HPCS, RCIC and LPCS.
Equipment Room Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Core Cooling This room contains Division 2 switchgear. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in this room. Based on credit by RW 467 Division 2 Switchgear RC8 (RC8) exclusion, fires in this room do not impact HPCS, RCIC and LPCS.
Room Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Page 24
Table RAI022 Evaluation of the Gap Involving Electrical Conduits for Fire Areas where Division 1 (RHR A) Provides PostFire Safe Shutdown FPRA Physical Analysis Unit Description DefenseinDepth Features (Fire Safe Shutdown Fire Area)
Core Cooling This room contains Division 2 HVAC equipment. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in this room. Based on RC12 (RC12) RW 525 B A/C Room credit by exclusion, fires in this room do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Core Cooling This room contains the control room emergency chillers and cables for Division 2 HVAC equipment. Cables that support HPCS, RCIC and LPCS RW 525 Emergency Chiller cannot reasonably exist in this room. Based on credit by exclusion, fires RC13 (RC13)
Room in this room do not impact HPCS, RCIC and LPCS.
Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Core Cooling This room contains the RHR B pump. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in this room. Based on credit by exclusion, Reactor Building 422 RHRB R4 (R4) fires in this room do not impact HPCS, RCIC and LPCS.
Room Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Core Cooling This room contains the RCIC pump. Cables that support HPCS cannot reasonably exist in this room. Based on credit by exclusion, fires in this Reactor Building 422 RCIC Pump R6 (R6) room do not impact HPCS.
Room Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Page 25
Table RAI022 Evaluation of the Gap Involving Electrical Conduits for Fire Areas where Division 1 (RHR A) Provides PostFire Safe Shutdown FPRA Physical Analysis Unit Description DefenseinDepth Features (Fire Safe Shutdown Fire Area)
Core Cooling This room contains the RHRC pump room. Cables that support HPCS cannot reasonably exist in this room. Based on credit by exclusion, fires Reactor Building 422 RHRC R7 (R7) in this room do not impact HPCS.
Pump Room Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Core Cooling This room contains Division 2 MCCs. Cables that support HPCS and LPCS cannot reasonably exist in this room. Based on credit by exclusion, fires Reactor Building 522 Division 2 R18 (R18) in this room do not impact HPCS and LPCS.
MCC Room Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Core Cooling This room contains the Division 2 standby service water pump 1B. Cables that support HPCS, RCIC and LPCS cannot reasonably exist in this room.
Standby Service Water Pump Based on credit by exclusion, fires in this room do not impact HPCS, RCIC SW2 (SW2)
House 1B and LPCS.
Decay Heat Removal Hardened containment vent and nonhardened containment vent are not impacted, based on credit by exclusion.
Page 26
Table RAI022 Evaluation of the Gap Involving Electrical Conduits for Fire Areas where Division 1 (RHR A) Provides PostFire Safe Shutdown FPRA Physical Analysis Unit Description DefenseinDepth Features (Fire Safe Shutdown Fire Area)
Core Cooling Division 3 cables are routed in this corridor and fire impacts to these cables, which support HPCS, are modeled by the FPRA. Although some Division 3 cables are routed in conduit in this corridor, these cables produce no different FPRA results. The gap related to conduit routing produces no different FPRA result for Division 3.
Division 1 cables routed in this corridor are associated with DG 1 and are wrapped in Darmatt. The FPRA accounts for the protection of DG 1 cables in this PAU.
TG12 (TG12) 441 Turbine Generator Corridor Cables related to RCIC and LPCS pass from the radwaste building to the reactor building, and do not travel in the turbine generator corridor.
Cables that support RCIC and LPCS cannot reasonably exist in this corridor. Based on credit by exclusion, fires in this room do not impact RCIC and LPCS.
Decay Heat Removal Local manual hardened containment vent and nonhardened containment vent do not rely on electrical power and are not impacted.
Local hardened containment vent would be performed well after any fire occurring in the corridor, and thus the operator action travel pathway is not impeded.
Page 27
APLA RAI 03 The CGS LAR states that the proposed change to the TS completion time has been evaluated using the riskinformed processes described in RG 1.177, Revision 1. Based on Section 2.3.1 of RG 1.177, the technical adequacy of the PRA must be compatible with the safety implications of the Technical Specification change being requested and the role that the PRA plays in justifying that change. RG 1.177 endorses the guidance provided in RG 1.200, Revision 2, on PRA technical adequacy. RG 1.200 describes a peer review process utilizing ASME/ANS RASa2009 as one acceptable approach for determining the technical adequacy of the PRA once acceptable consensus approaches or models have been established for evaluations that could influence the regulatory decision.
In Section 5.3 of LAR Attachment 5, the licensee explains that the seismic PRA (SPRA) has not undergone a peer review against the ASME/ANS PRA standard RASa2009, as clarified/qualified by Revision 2 of RG 1.200. Additionally, while the licensee indicates that seismic hazards are a low risk contributor, the risk results presented in Table 2 of LAR Attachment 1 demonstrate that the incremental conditional core damage probability (ICCDP) contribution from seismic events (approximately 23%) is not much different than that from other hazards (i.e., internal events and fire). Moreover, considering that RHR pumps are stated as having high seismic capacity, it is not clear to the NRC staff that RHR Train A subsystem unavailability is insignificant to seismic risk on the basis of seismic correlation between RHR pumps alone and independent of other structures, subsystems and components, which may have a lesser seismic capacity and may not be correlated across trains. Lastly, the SPRA, which is based upon the IPEEE model, does not appear to address updated seismic hazard information and the significant changes to SPRA methodologies since the IPEEE model was issued in 2001.
Identify any gaps between the CGS SPRA model and the "Seismic Events Technical Elements" required by Revision 2 of RG 1.200 that are relevant to this submittal. In doing so, explain why these gaps do not have significant impact on the risk assessment results and the development of compensatory actions used to support this application. This may include discussion of relevant conservatisms in the CGS SPRA model and additional sources of defenseindepth, as well as, the risk significance of each to the application.
Response
CGS SPRA Gaps to Revision 2 of RG 1.200 Relevant to this Submittal Three gaps between the CGS SPRA model and the "Seismic Events Technical Elements" required by Revision 2 of RG 1.200 that are relevant to this submittal have been identified. These gaps, however, do not have significant impact on the risk assessment results and the development of compensatory actions used to support this onetime application, based on the following:
- 1. As discussed by the RG 1.177 guidance, technical specification completion time changes are relatively insensitive to uncertainties. This is because the uncertainties associated with CT changes tend to similarly affect the base case (before the change) and the AOT case (with the change in place). That is, the risks result from similar causes in both cases.
- 2. CGS plant systems provide defenseindepth.
Page 28
- 3. Conservatisms exist in the CGS SPRA model. Correcting the HPCS suction logic conservatism would significantly reduce the contribution of the seismic hazard to the risk result.
The gaps between the CGS SPRA model and the "Seismic Events Technical Elements" required by Revision 2 of RG 1.200 that are relevant to this submittal are listed in Table RAI031, with the basis to indicate that the gap does not significantly impact the riskinformed decision.
Additional Sources of DefenseinDepth FLEX provides additional defenseindepth and is not credited by the SPRA. Crediting FLEX would provide additional margin by:
- 1. Modeling two FLEX diesel generators, which are capable of recharging plant batteries,
- 2. Continued operation of RCIC after battery depletion,
- 3. Continued operation of RCIC for suppression pool temperatures up to 240 degrees F,
- 4. An additional external injection source, which can be connected to any of the three RHR injection lines, and
- 5. Use of hardened containment vent for decay heat removal.
Modeling FLEX in the SPRA would provide additional safety margin when RHR A is out of service.
Risk Margin The RHR A AOT risk assessment results indicate significant margin to the regulatory thresholds of 1E6 ICCDP and 1E7 ICLERP without compensatory measures and 1E5 ICCDP and 1E6 ICLERP with compensatory measures. The significant margin supports the qualitative assessment of gaps and accommodates additional risk due to the nonRG 1.200 seismic PRA for this onetime submittal.
Conclusion Gaps between the CGS SPRA model and the "Seismic Events Technical Elements" required by Revision 2 of RG 1.200 that are relevant to this submittal do not have significant impact on the risk assessment results and the development of compensatory actions used to support this one time application, based on several factors:
- 1. Technical specification completion time changes are relatively insensitive to uncertainties. This is because the uncertainties associated with CT changes tend to similarly affect the base case and the AOT case.
- 2. CGS plant systems provide defenseindepth.
- 3. Conservatisms exist in the CGS SPRA model. Correcting the HPCS suction logic conservatism would significantly reduce the risk impact of the AOT application relative to the acceptance guidelines.
Page 29
Table RAI031 Gaps between the CGS SPRA Model and Revision 2 of RG 1.200 Relevant to this Submittal Gap Description Impact to the RiskInformed Decision SPRA01 Conservative modeling of the HPCS For the RHR A AOT LAR, a modeling completeness issue was identified suction logic. in Table 42, Page 38 regarding conservative modeling of the HPCS suction logic. The signal for automatic transfer on low condensate storage tank (CST) level to the suppression pool is not credited. The valve transfers are modeled, but the fault tree logic requires operators to perform the suction transfer. For operators to manually perform the suction swap from the CST to the suppression pool, CST level indication is required, which relies on Division 2 electrical power. Since the CST is not credited by the SPRA, failure of the CSTs transfer to the suppression pool will fail HPCS suction. Therefore, seismicallyinduced unavailability of Division 2 electrical power fails both Division 2 ECCS pumps (RHRB and RHRC) as well as well as HPCS.
This modeling incompleteness overestimates the seismic risk significance of RHR A. For higher frequency, lower magnitude earthquakes in which the seismic failures of plant safety components (correlated and uncorrelated) are of lower likelihood, the unavailability of Division 2 electrical power (for example, the Division 2 diesel generator fails to start or run) causes a loss of Division 2 systems as well as Division 3 systems (HPCS). A sensitivity calculation was performed in which this modeling incompleteness was corrected in the SPRA model. Once this incompleteness is resolved, the incremental conditional core damage probability (ICCDP) contribution from seismic events is insignificant relative to the acceptance threshold.
Page 30
Table RAI031 Gaps between the CGS SPRA Model and Revision 2 of RG 1.200 Relevant to this Submittal Gap Description Impact to the RiskInformed Decision SPRA02 The existing component chatter The SPRA includes and models the following component chatter evaluation utilized seismic margins scenarios that are significant to the SPRA: a) unavailability of diesels assessment to exclude chatter due to component chatter; and b) correlated unavailability of scenarios, an approach that does not switchgear due to component chatter. It is expected that modeling of meet the industry consensus additional contact chatter scenarios would affect the base case and approach and the ASME / ANS the AOT case similarly. Further, HPCS, RCIC and LPCS each have Standard Capability Category II actuation and control logic designs that are diverse from each other requirements. as well as from RHR, which provides defenseindepth from seismic correlation of component chatter scenarios. Similarly, containment venting provides diversity from suppression pool cooling, and significant time is available to recover from component chatter impacts. This gap therefore produces no significant impact on the risk assessment results and the development of compensatory actions used to support this application.
SPRA03 The CGS SPRA does not utilize This gap produces no significant impact on the risk assessment results updated seismic hazard information. and the development of compensatory actions used to support this application. The updated seismic hazard information would affect the base case and the AOT case similarly. The seismic correlation across RHR trains, for the pumps as well as for structures, systems and components that support RHR, vary similarly to changes in seismic hazard for the base case and AOT case.
Page 31
APLA RAI 04 The CGS LAR states that the proposed change to the TS completion time has been evaluated using the riskinformed processes described in RG 1.177, Revision 1. RG 1.177 states riskinformed analyses of TS changes can be affected by numerous uncertainties regarding the assumptions made during the PRA models development and application. The LAR neither identifies nor assesses uncertainties associated with the fire and seismic PRAs, and Section 4 of LAR Attachment 5 clarifies that no formal assessment of uncertainties has been performed. Provide, consistent with the guidance in RG 1.177, assurance that uncertainties associated with the fire and seismic PRAs are not sufficient to change the conclusions of the LAR.
Response
As discussed by the RG 1.177 guidance, technical specification completion time changes are relatively insensitive to uncertainties. This is because the uncertainties associated with CT changes tend to similarly affect the base case (before the change) and the AOT case (with the change in place). That is, the risks result from similar causes in both cases.
For the CGS RHR A AOT completion time LAR, no deviations from these expectations are found.
The relevant gaps between Revision 2 of RG 1.200 and the CGS FPRA and SPRA are sources of completeness uncertainty and these uncertainties are evaluated in the responses to RAIs 02 and 03 and found to not change the conclusions of the LAR.
Page 32
APLA RAI 05 The CGS LAR states that the proposed change to the TS completion time has been evaluated using the riskinformed processes described in RG 1.177, Revision 1. Section 2.3 of RG 1.177 states compensatory actions that can mitigate any corresponding increase in risk should be identified and evaluated.
Due to analyses of risksignificant combinations documented in LAR Attachment 5, the licensee committed, in Section 3.6 of LAR Attachment 1 and LAR Attachment 4, to implement flood and fire watch tours as a means of early detection and risk reduction. However, no information by which to evaluate the effectiveness of these compensatory measures (e.g., frequency of tours) appears to have been provided. Clarify how the committed flood and fire watch tours offer appropriate restrictions to reduce risk from the corresponding risksignificant scenarios.
Response
The committed flood and fire watch tours will be performed on an hourly, roving basis during the 14day RHR A completion time.
The committed flood tours offer appropriate early detection and isolation benefits to reduce risk from corresponding risksignificant scenarios.
The committed fire tours offer appropriate detection and suppression (prevention) benefits, as well as confirmation of the adherence to transient combustible restrictions, prohibition of hot work, and confirmation of area detection, suppression, and fire door availability. These fire tours will reduce risk from the corresponding risksignificant fire scenarios.
Page 33
GO2-17-101 Page 1 of 3 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION APL RAI 06 and 07
GO2-17-101 Page 2 of 3 NRC REQUEST APLA RAI 06:
The CGS LAR states that the proposed change to the TS completion time has been evaluated using the risk-informed processes described in RG 1.177, Revision 1. As indicated in Section 2.3.7.2 of RG 1.177, the licensee should ensure that the Configuration Risk Management Program (CRMP) contains a number of key components, including the treatment of external hazards and Level 2 issues, either qualitatively or quantitatively, or both. However, while Section 3.7 of LAR Attachment 1 indicates that the CRMP makes use of quantitative insights provided by the internal events PRA, plant risk levels appear to only address core damage risk. Additionally, while the CRMP addresses fire hazards qualitatively, it does not appear to address other hazards, particularly external hazards such as those posed by seismic events.
Clarify how the CRMP proposed in the LAR treats external hazards and Level 2 issues consistent with guidance in RG 1.177.
ENERGY NORTHWEST RESPONSE TO APLA RAI 06:
The Columbia Generating Station (Columbia) CRMP uses qualitative and quantitative insights to address Level 2 issues. Qualitative insights are provided by the Primary Containment Control and Secondary Containment Control defense-in-depth safety functions. Quantitative insights are provided by the internal events Probabilistic Risk Assessment (PRA) model. The Paragon program calculates core damage frequency (CDF) and large early release frequency (LERF) impacts.
The Columbia CRMP uses qualitative insights to address external hazards. The following is from NUMARC 93-01, Revision 4A, Section 11.3.4.2, Qualitative Considerations, Item 7:
"External event considerations involve the potential impacts of weather or other external conditions relative to the proposed maintenance evolution. For the purposes of the assessment, weather, external flooding, and other external impacts need to be considered if such conditions are imminent or have a high probability of occurring during the planned out-of-service duration. An example where these considerations are appropriate would be the long-term removal of exterior doors, hazard barriers, or floor plugs."
Normally potential impacts of external events are not imminent or have a high probability of occurring during the planned out of service duration. At Columbia, if an external condition is imminent or has a high probability of occurring, then a high risk evolution (HRE) is included in the risk assessment to address the external hazard. An HRE will typically increase the Plant Risk Level by one level. Qualitative insights are used to address external hazards.
GO2-17-101 Page 3 of 3 NRC REQUEST APLA RAI 07:
The CGS LAR states that the proposed change to the TS completion time has been evaluated using the risk-informed processes described in RG 1.177, Revision 1.
Section 3.2 of RG 1.177 states [i]f the licensee concludes that the performance or condition of TS equipment affected by a TS change does not meet established performance criteria [as part of its Maintenance Rule program], appropriate corrective action should be taken, in accordance with the Maintenance Rule. Clarify if the additional out-of-service time for the RHR Train A subsystem is expected to result in exceeding the current established Maintenance Rule performance criteria for the RHR Train A subsystem. If so, describe the corrective action(s) that will be taken.
ENERGY NORTHWEST RESPONSE TO APLA RAI 07:
It is anticipated that the current unavailability performance criteria for Residual Heat Removal (RHR) system A will be exceeded when performing the pump and motor replacement. The performance criteria for unavailability associated with the Maintenance Rule program does not include infrequently performed maintenance activities such as pump and motor replacements. However, when a 12 year preventative maintenance task to rebuild a pump or motor or predictive maintenance tasks indicate that replacement or repair is required, the System Engineering Instruction, SYS-4-22, Maintenance Rule Program, allows the performance criteria to be reviewed and temporarily increased to account for the infrequently performed maintenance. The existing performance criteria for RHR system A will be reviewed and action taken per SYS-4-22 to temporarily revise the criteria to reflect the estimated unavailability that will be required to replace the pump and motor. The revised performance criteria is approved per SYS-4-22 either by the System Engineering Manager and/or the Maintenance Rule Expert Panel. If the newly established performance criterion is exceeded, then a Maintenance Rule (a)(1) Evaluation will be completed to determine if the system should be place in Maintenance Rule (a)(1) status.
If the system is placed in (a)(1) status, then a Performance Improvement Plan with corrective actions and goal(s) will be developed.