ML17193A323
| ML17193A323 | |
| Person / Time | |
|---|---|
| Site: | Vogtle  |
| Issue date: | 06/27/2017 |
| From: | Ed Miller Special Projects and Process Branch |
| To: | |
| Miller E, 415-2481 | |
| Shared Package | |
| ML15316A419 | List: |
| References | |
| NCP-2017-008 | |
| Download: ML17193A323 (6) | |
Text
Page 1 of 8 NCP-2017-008 Section C
==
Introduction:==
As the Non-Concurrence Process (NCP) approver for NCP-2017-008, I have read and considered the submission from the non-concurring staff. I continue to acknowledge and applaud the efforts of these staff members to raise these concerns, and the efforts of other staff to actively participate in the resolution of the issues identified. Given the length of the review time for the subject document, I determined it was appropriate to document some of the history of the Vogtle license amendment request (LAR) which is the subject of this NCP. The safety evaluation (SE) for this LAR was originally routed for concurrence in the summer of 2015.
During that routing, a non-concurrence was tendered by EEOB 1 staff (NCP-2015-009, ADAMS Accession No. ML15322A197) and supported by the EEOB branch chief. Following disposition of the NCP and, in part, based on some of the issues raised by the non-concurring staff, the NRC staff developed a significant number of additional questions regarding this LAR.
Exploration of these issues resulted in the licensee removing some of the actions originally included, providing additional justification for selected technical specification (TS) actions, and introducing additional constraints to implement the risk-informed TS completion time (RICT) program.
Although the arguments tendered in the original Section C of NCP-2015-009 for the validity of a RICT program remain valid, it is appropriate to document the changes that have resulted and the positive results of the NRC staff working collaboratively through difficult and highly technical issues. Among others, some of the more notable changes were:
- Additional justification was provided to demonstrate how defense-in-depth is maintained while in a RICT;
- Additional constraints were added for addressing the possibility of common-cause failure for both emergent failures and planned maintenance while in a RICT;
- Some actions were identified as not being appropriate for applying a RICT;
- Additional constraints were applied to TS loss of function (TS LOF) situations;
Additionally, it is important to note the generic impact as well. Although the Vogtle LAR did not directly reference TSTF-505, Revision 1 (TSTF-505), which provides model TS changes to adopt RICTs in accordance with NEI 06-09, Revision 0-A, the application was considered a pilot for implementing a RICT program. The concerns identified during the NRCs review of the Vogtle LAR also caused the NRC staff to suspend approval of TSTF-505 (ADAMS Accession No. ML16281A021), Revision pending resolution of the subject issues. That effort is currently ongoing.
This summary of the history was also included as an addendum to NCP-2015-009.
Summary of Issues:
This summary was provided to the non-concurring individuals and their comments on the summary have been incorporated. The non-concurrence has 10 objections to the proposed license amendment as listed below:
1 At the time, the branch designation was EEEB. Due to a reorganization, the branch designation was changed to EEOB.
Page 2 of 8 NCP-2017-008 Section C
- 1. The proposed amendment is risk-based and would modify the TSs in a manner that doesnt meet the current licensing basis of Vogtle, safety margins, defense-in-depth, and NRC regulatory requirements. Deterministic basis for TS sections 3.8.1 specified in SE section 3.2.1 through 3.2.3 state that RMAs are adequate to maintain safety margins and defense-in depth and therefore meets current regulations. The evaluations provided in this section and the conclusions are incorrect. Therefore, it must be corrected or removed from the SE.
- 2. The TS Loss of Function (TS LOF) action statements represent an unanalyzed condition. Two supporting examples of aberrant system behavior are postulated. The first derives from operating experience from the Forsmark plant in Sweden. The second by postulating removal of a battery charger from service and experiencing voltage spike.
For LOOP/LOCA scenarios, the ESF buses feeding the vital ac buses would not have any power for at least 20 seconds until the onsite power system breakers for ESF system are closed.
- 3. PRA Functional is a new term and not recognized in Commission policy, regulations, guidance documents, NRC endorsed standards or technical specifications. Further, utilizing a backstop as long as 30 days represents new Commission policy. There is no acceptable deterministic basis provided for 30 days backstop CT in the SE.
- 4. Completion Times that allow continued operation of the plant past the deterministically established times do not preserve safety and are not in accordance with 10 CFR 50.36 or RG 1.93.
- 5. The LAR renders the front-stop meaningless because the backstop of 30 days is so much greater than most front-stops.
- 6. Given that the NRC has suspended approval of TSTF-505 and there are substantial similarities between the Vogtle LAR based on NEI 06-09 and the TSTF, it is not appropriate to issue the LAR until resolution of issues associated with TSTF-505 is resolved. The NRC must withdraw approval of Topical Report NEI 06-09, Risk-Informed Technical Specifications Initiative 4b, Risk-Managed Technical Specifications (RMTS)
Guidelines, since the Vogtle would be operated outside the design bases requirements.
- 7. Due to the potential for large-scale consequences for onsite and internal events that can hinder recovery efforts, the common assumption of a 24-hour mission time for electrical systems supporting ECCS for PRA evaluations is unjustified. As an example, provide the basis for 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> mission for EDGs given that a LOOP event maybe a result of extremely severe weather, or the onsite vital AC systems (Batteries, inverters, UPS etc.)
given that such equipment is required for all modes of operation.
- 8. The justification referenced from TSTF-426 shouldnt be used because electrical system TSs were specifically excluded from the TSTF based on inadequate technical and regulatory basis.
- 9. Since the Risk-Informed Technical Specifications Initiative 4b methodology is risk-based and not addressing deterministic requirements specified in 50.46, 10 CFR 50 Appendix A, GDC 17, UFSAR Chapters 6, 8, and 15 design bases, the NRC must require the Vogtle to either request exemptions from NRC requirements and changes to the licensing basis to approve this LAR or Section 3.8 of the TS should not be included within the scope of this LAR similar to I&C systems (note: I&C systems are excluded).
- 10. Since the licensee has not corrected the electric power system design vulnerability identified in Bulletin 2012-01 at Vogtle Units 1 and 2, the plant operation with the approval of this TS amendment along with the Open Phase Condition deficiency will put the plant in an unsafe condition. Therefore, NRC should not approve CT extensions for TS Section 3.8, Electric Power Systems.
Page 3 of 8 NCP-2017-008 Section C Response to Issues:
Issue 1
Description:
The non-concurring individuals assert that the proposed amendment is risk-based and would modify the TSs in a manner that doesnt meet the current licensing basis of Vogtle, safety margins, defense-in-depth, and NRC regulatory requirements. Specifically, they indicate that the deterministic basis for TS sections 3.8.1 specified in SE section 3.2.1 through 3.2.3 state that RMAs are adequate to maintain safety margins and defense-in depth and therefore meets current regulations, which they believe is incorrect and, therefore, must be corrected or removed from the SE.
Issue 1 Response: This issue was originally raised in NCP-2015-009 (See core objection 1) and the justification for this issue is substantially similar to that provided previously. In addition, as discussed in the introduction, through the ongoing dialogue, additional changes were made to the Vogtle SE. These changes include new constraints TS 5.5.22.f.1 and TS 5.5.22.f.2 (discussed further under Issue 2), which generally limit the plant to operating in previously evaluated and acceptable configurations providing further confidence that maintain safety margins and defense-in depth are satisfactorily retained. Therefore, the previous determination that the proposed LAR would not adversely impact the manner in which the TSs meet applicable regulatory requirements and maintain defense in depth remains valid.
Issue 2
Description:
The non-concurring individuals asserted that the TS Loss of Function (TS LOF) action statements represent an unanalyzed condition. Two supporting examples of aberrant system behavior are postulated. The first derives from operating experience from the Forsmark plant in Sweden. The second by postulating removal of a battery charger from service and experiencing voltage spike. For LOOP/LOCA scenarios, the ESF buses feeding the vital ac buses would not have any power for at least 20 seconds until the onsite power system breakers for ESF system are closed.
Issue 2 Response: The TS LOF action statements proposed for addition to the TSs contain a note that applies a number of additional restrictions listed in Section 5.5.22 of the TSs. Of particular importance with respect to this issue is TS 5.5.22.f.1 and TS 5.5.22.f.2. The first restriction is that any SSCs credited in the determination of PRA functionality must be the same SSCs relied upon to perform the specified TS safety function. As such, the pedigree of these SSCs to accomplish the function has already been evaluated and determined to be sufficient.
By the second restriction, the current state of the train, though TS inoperable (i.e., the associated TS LCO is not met), must be capable of meeting all design basis success criteria parameters for all design-basis accident scenarios. Thus, the plant would have one train capable of responding to a design basis accident and remain in an analyzed condition.
Therefore, the SE provides adequate basis that the TS LOF actions do not represent an unanalyzed condition.
In the Vogtle application, for TS 3.7.14, the licensee proposed a specific use of alternative SSCs to accomplish room cooling for an engineered safety features room. Specifically, the licensee proposed to open doors to maintain room temperatures, which would not be in accordance with the limitations of TS 5.5.22.f.1. For this instance, the NRC staff specifically evaluated the proposed actions and usage of equipment for the stated purpose. This evaluation concluded that the actions and usage of SSCs was acceptable for the stated purpose and an exception to limitation was added to the TS action. However, since the NRC staff did review a specific
Page 4 of 8 NCP-2017-008 Section C evaluation of the condition in the context of a license amendment request to incorporate it into the plant licensing basis, it would represent an analyzed condition, if approved.
The first supporting example provided by the non-concurring individuals referenced an actual event that occurred at the Forsmark plant in Sweden where voltage spikes resulted in the lockup or tripping of safety systems. While this operating experience showed potential design vulnerabilities, the subject amendment does not propose to change the design of the onsite electrical systems at Vogtle. Instead, it proposes a change to how they are operated.
Specifically, how long the plant may remain in a degraded, but analyzed condition. Any such design issues that the non-concurring individuals believe exist should be addressed through NRC processes for addressing such issues.
The second example provided postulated a situation where a battery was without any chargers for an extended period of time. I agree with the non-concurring individuals that, as described, the scenario would result in degrading battery voltage over time and would eventually reach a point where spurious operation of systems could occur. This configuration would not be allowed by the restrictions of TS 5.5.22.f.2, which requires at least one train to be capable of meeting all design basis success criteria parameters for all design-basis accident scenarios. For the licensee to meet this criteria, at least one charger would need to be able to maintain battery voltage and the sequence of events described in the example would not be possible.
A similar consideration is applicable to the LOOP/LOCA scenario described. Since the proposed actions restrict plant operation to a condition where at least one train is capable of meeting all design basis success criteria parameters for all design basis accident scenarios, plant response to a LOOP/LOCA would be bounded by the current analysis.
Issue 3
Description:
The non-concurring individuals asserted that PRA Functional is a new term and not recognized in Commission policy, regulations, guidance documents, NRC endorsed standards or technical specifications. Further, they asserted that utilizing a backstop as long as 30 days represents new Commission policy. Lastly, they assert that there is no acceptable deterministic basis provided for a 30-day backstop CT in the SE Issue 3 Response: The development of the RICT program has been over the course of many years since the initial policy statement in an environment of significant Commission interest and awareness of risk-informed activities. The selection of South Texas Project as a pilot application for a RICT program, and the associated review of and final issuance of the amendment in 2007 (which included the concept of PRA Functional and a 30-day backstop, ADAMS Accession No. ML071780186) have been included in periodic status SECY papers to the Commission. These updates included increasing specificity and detail as it was available from the pilot application. The Commission has not, at any point, expressed that the NRC staff is proceeding in a manner that is contrary to their intent. Further, the express term PRA Functional does exist in current NRC guidance. For example, and as part of establishing the framework for risk-informed TSs, the concept and usage of PRA Functional is currently described in IMC-0326, Operability Determinations & Functionality Assessments for Conditions Adverse to Quality or Safety. Therefore, issuance of the proposed amendment or usage of the term, and associated concept of PRA Functional does not appear to be contrary to Commission policy.
Page 5 of 8 NCP-2017-008 Section C As indicated in the NRC staff SE, the 30-day backstop assures that the TS equipment is not out of service for extended periods, maintaining the intent of equipment allowed outage times in general as opposed to permanent design changes, and is a reasonable upper limit to permit repairs and restoration of equipment to an operable status. This determination is aligned with previous NRC staff application of engineering judgement in determining equipment allowed outage times using all information and tools available at the time.
Issue 4
Description:
The non-concurring individuals asserted that Completion Times that allow continued operation of the plant past the previous deterministically-established times do not preserve safety and are not in accordance with 10 CFR 50.36 or RG 1.93.
Issue 4 Response: This issue was originally raised in NCP-2015-009 (See core objection 3) and the justification for this issue is substantially similar to that provided previously. Therefore, the previous determination that the proposed LAR would not adversely impact the manner in which the TSs meet applicable regulatory requirements and maintain defense-in-depth remains valid.
It is important to note that the current NCP indicates that Sections 3.2.1 through 3.2.3 of the current SE states that RMAs are adequate to maintain safety margins and defense in depth.
This is inaccurate as these sections do not contain such an assertion. The SE does not conclude that the RMAs alone are adequate to maintain safety margins and defense in depth.
Instead, as indicated in the SE, they are one part of a multi-faceted evaluation that included consideration of the plant design and potential configurations allowed by the proposed TSs. In addition, as discussed in the introduction, through the ongoing dialogue, additional changes were made to the Vogtle SE. These changes include new constraints TS 5.5.22.f.1 and TS 5.5.22.f.2 (discussed further under Issue 2), which generally limit the licensee to operating the plant in previously evaluated and acceptable configurations. This provides further confidence that maintain safety margins and defense-in depth are satisfactorily retained. The conclusion reached in the SE is that the RMAs, along with other considerations, provide additional confidence that the intent of the plants design basis is maintained.
Issue 5
Description:
The non-concurring individuals asserted that the LAR renders the front-stop meaningless because the backstop of 30 days is so much greater than most front-stops.
Issue 5 Response: The amendment would implement a program whereby the licensee has the option to exercise a RICT. When exercising a RICT, there are a number of attendant programmatic requirements including evaluations to be performed, actions to be taken, and cumulative monitoring of the change in risk from the RICT. In many instances, the licensee may not wish to exercise the RICT program. For this situation, the completion time would be governed by the existing front-stop. In such circumstances, the front-stop remains a real and meaningful limit within the TSs. In either case, safety of the plant is adequately maintained and managed.
Issue 6
Description:
The non-concurring individuals asserted that because that the NRC has suspended approval of TSTF-505 and there are substantial similarities between the Vogtle LAR based on NEI 06-09 and the TSTF, it is not appropriate to issue the LAR until resolution of issues associated with TSTF-505. They also asserted that NRC must withdraw approval of Topical Report NEI 06-09, Risk-Informed Technical Specifications Initiative 4b, Risk-Managed Technical Specifications (RMTS) Guidelines, since the program would allow operation outside of design bases requirements.
Page 6 of 8 NCP-2017-008 Section C Issue 6 Response: The suspension of TSTF-505 was an action taken by the NRC based upon the issues raised during the review of the Vogtle 4b LAR. As the non-concurring individuals indicate, although the Vogtle LAR is not implementing TSTF-505 directly, it is the pilot application for implementing a RICT program and will inform TSTF-505 as a model change to the standard TS to implement a RICT program. The letter that suspended TSTF-505 (ADAMS Accession No. ML16281A021) indicated that the NRC would continue its review of in-house applications as plant-specific licensing actions, as long as the issues raised in the letter were resolved in those reviews. The licensee for Vogtle has resolved these issues to the staffs satisfaction as indicated by the current EEOB concurrence on the SE. Therefore, the fact that TSTF-505 remains suspended does not preclude issuance of the Vogtle LAR.
Issue 7
Description:
The non-concurring individuals asserted that due to the potential for large-scale consequences for onsite and internal events that can hinder recovery efforts, the common PRA evaluation assumption of a 24-hour mission time for electrical systems supporting ECCS is unjustified. The non-concurring individuals provided a specific example requesting the basis for a 24-hour mission for EDGs given that a LOOP event may result from extremely severe weather and would not be recovered within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
Issue 7 Response: In the supporting information provided for this issue, the non-concurring individuals indicate that restoration of offsite power after a significant event may take significantly longer than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. This issue appears to conflate the 24-hour mission time used in PRA evaluations with the concept of mission time used in the design of the plant (e.g.,
factoring assumptions of restoration of offsite power into the design of EDG fuel oil tank capacity). Although sharing a similar name, the PRA mission time is used to delineate the time required to be analyzed in the risk analysis when determining whether a particular sequence results in success or failure (i.e., core damage). The PRA standard specifies a minimum of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for this mission time, but it can be longer (e.g. 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />) if stable plant conditions have not been achieved at 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, such as in the case of external hazards, e.g. tornadoes, external flooding etc. If the analysis shows that the plant reaches a safe and stable condition at 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, there is limited benefit to continuing the analysis. If a stable condition has not been reached in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, the PRA either assumes a core damage for that sequence or must identify and incorporate an extended analysis time until safe and stable plant conditions are achieved for the accident sequence. Therefore, the nominal use of a 24-hour mission time in PRA evaluations does not equate to the length of time a piece of equipment needs to operate to mitigate an accident. Further, the approach represents an efficient method of determining whether particular sequence leads to core damage or not.
Issue 8
Description:
The non-concurring individuals asserted that the justification referenced from TSTF-426 shouldnt be used because electrical system TSs were specifically excluded from this TSTF based on inadequate technical and regulatory basis.
Issue 8 Response: In the SE for the Vogtle LAR, the NRC staff does reference that a 24-hour backstop upon loss of operability is consistent with the maximum acceptable CT in TSTF-426, Revise or Add Actions to Preclude Entry into LCO 3.0.3 - RITSTF [Risk-Informed TSTF]
Initiatives 6b & 6c. TSTF-426 supports extending the time before shutdown upon loss of TS operability from 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> up to a maximum of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> based on an in-depth evaluation on the impact of the loss of operability on the design basis and the development of additional, pre-identified defense-in-depth RMA type measures included in the TSs. While the NRC staff may have previously found an inadequate basis for application of TSTF-426 to electrical TSs,
Page 7 of 8 NCP-2017-008 Section C there are significant differences between the limitations of TSTF-426 and the Vogtle LAR. Most notably, TSTF-426 did not have the requirement that the licensee must make a determination that the design basis success criteria can still be met. This is a significant limitation within the RICT program and, combined with the requirement to develop RMAs, serves to assure that there is sufficient remaining capability and only a small, acceptable, increase in risk for the limited time the CT can be extended with all trains TS inoperable. From this evaluation, the NRC staff SE concludes that the licensees use of a 24-hour backstop for TS inoperable but PRA Functional equipment provides adequate defense-in-depth when applied together with the conditions that 1) the same SSCs relied on in the LCO be available and 2) the design-basis function success parameters can still be achieved.
Issue 9
Description:
The non-concurring individuals asserted that since the Risk-Informed Technical Specifications Initiative 4b methodology is risk-based and not addressing deterministic requirements specified in 50.46, 10 CFR 50 Appendix A, GDC 17, UFSAR Chapters 6, 8, and 15 design bases, the NRC must require the Vogtle to either request exemptions from NRC requirements and changes to the licensing basis to approve this LAR or Section 3.8 of the TS should not be included within the scope of this LAR similar to I&C systems, further asserting that I&C systems are excluded from the RICT program Issue 9 Response: This issue was originally raised in NCP-2015-009 (See core objection 1) and the justification for this issue is substantially similar to that provided previously. In addition, as discussed in the introduction, through the ongoing dialogue, additional changes were made to the Vogtle SE. These changes include new constraints TS 5.5.22.f.1 and TS 5.5.22.f.2 (discussed further under Issue 2), which generally limit the plant to operating in previously evaluated and acceptable configurations providing further confidence that maintain safety margins and defense-in depth are retained satisfactorily. Therefore, the previous determination that the proposed LAR would not adversely impact the manner in which the TSs meet applicable regulatory requirements and maintain defense-in-depth remains valid. With regard to I&C TSs, these TS are not categorically excluded from the RICT program or TSTF-505 and the licensee did not propose their inclusion in this LAR.
Issue 10
Description:
The non-concurring individuals asserted that since the licensee has not corrected the electric power system design vulnerability identified in Bulletin 2012-01 at Vogtle Units 1 and 2, the plant operation with the approval of this TS amendment along with the Open Phase Condition deficiency will put the plant in an unsafe condition. Therefore, NRC should not approve CT extensions for TS Section 3.8, Electric Power Systems.
Issue 10 Response: As discussed in addressing the Forsmark operating experience cited in issue #2 above, the subject amendment does not propose to change the design of the onsite electrical systems at Vogtle. Instead, it proposes a change to how they are operated.
Specifically, how long the plant may remain in a degraded, but analyzed condition. Evaluation of licensee responses to the bulletin and any subsequent NRC regulatory actions, if any, should be through the normal processes for implementing those actions. Therefore, this topic should not be addressed as part of this LAR.
Page 8 of 8 NCP-2017-008 Section C
==
Conclusion:==
As discussed in detail above, I do not find the issues raised by the non-concurring individuals to be inimical to the issuance of the LAR for Vogtle. Additionally, I find that RICT program does provide an appropriate means of determining a Completion Time for the selected TS Required Actions. Therefore, I am directing the NRC staff to continue routing the SE through concurrence along with providing this non-concurrence to all those who have previously concurred for their consideration.