NRC-24-0027, Partial Response to Request for Additional Information for License Amendment Request Regarding Risk Informed Approach to Performance ECCS Strainer Performance
| ML24115A095 | |
| Person / Time | |
|---|---|
| Site: | Fermi  |
| Issue date: | 04/24/2024 |
| From: | Domingos C DTE Electric Company |
| To: | Office of Nuclear Reactor Regulation, Document Control Desk |
| References | |
| NRC-24-0027 | |
| Download: ML24115A095 (1) | |
Text
Christopher P. Domingos Site Vice President DTE Electric Company 6400 N. Dixie Highway, Newport, MI 48166 Tel: 734.586.5025 Fax: 734.586.5295 Email: christopher.domingos@dteenergy.com 10 CFR 50.90 April 24, 2024 NRC-24-0027 U.S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, DC 20555-0001 Fermi 2 Power Plant NRC Docket No. 50-341 NRC License No. NPF-43
Subject:
Partial Response to Request for Additional Information for License Amendment Request Regarding Risk Informed Approach to Performance ECCS Strainer Performance In Reference 1, DTE Electric Company (DTE) submitted a License Amendment Request (LAR) to for a Risk Informed Approach to ECCS Strainer Performance. In Reference 2, an email from Mr. Surinder Arora to Mr. Eric Frank dated March 20, 2023, the NRC sent DTE a Request for Additional Information (RAI) regarding the LAR. A partial response to the RAI is provided in. Mark-ups of LAR NRC-23-0020 Attachments are provided in Enclosure 2. The remaining responses to the RAI will be submitted by May 31, 2024.
No new commitments are being made in this submittal.
Should you have any questions or require additional information, please contact Mr. Eric Frank at (734) 586-4772.
References:
- 1) DTE Letter NRC-23-0020, License Amendment Request for a Risk Informed Approach to ECCS Strainer Performance, dated June 13, 2023 (ML23164A232)
- 2) NRC E-mail Capture, Fermi 2 - Request for Additional Information for License Amendment Request Regarding Risk-Informed ECCS Strainer Performance Evaluation (L-2023-LLA-0092)", dated March 20, 2024 (ML24080A391)
USNRC NRC-24-0027 Page 2 I declare under penalty of pe1jury that the foregoing is true and correct.
Enclosure:
Executed on Ap *1 24, 2024 C41 F-t~-~
Christopher P. Domingos Site Vice President
- 1. Response to Request for Additional Information
- 2. LAR NRC-23-0020 Attachment Mark-ups cc: NRC Project Manager NRC Resident Office
- Regional Administrator, Region III Michigan Department of Environment, Great Lakes, and Energy to NRC-24-0027 Fermi 2 NRC Docket No. 50-341 Operating License No. NPF-43 Response to Request for Additional Information to NRC-24-0027 Page 1 On June 13, 2023, DTE Electric Company (DTE, the licensee) requested, in accordance with the provisions of Title 10 Code of Federal Regulations (10 CFR) 50.90, an amendment to modify the Fermi 2 Updated Final Safety Analysis Report (UFSAR) to describe the methodology used to address the impact of potential debris sources from a postulated high-energy line break on the Emergency Core Cooling System (ECCS) suppression pool strainer performance. The proposed amendment would revise the licensing basis as described in the Fermi 2 UFSAR to allow the use of a risk-informed methodology to address potential debris sources beyond those currently evaluated in a deterministic methodology. The licensees submittal also requested exemptions to 10 CFR 50.46 and associated general design criteria.
10 CFR 50.46 requires that the ECCS be capable of performing its safety function considering the most challenging loss-of-coolant accidents (LOCAs). To assure that long-term core cooling (LTCC) is maintained, the NRC has concluded that the effects of debris on the ECCS must be considered as part of the analysis. The NRC has reviewed the information provided by the licensee and determined that the additional information requested below is required for the staff to make a regulatory decision regarding the request.
The RAIs included in this transmittal were discussed with the licensee during a regulatory audit.
The NRC staff and licensee have discussed these issues during the audit and the licensee is familiar with the information requested via these RAIs. The licensee understands that the NRC requires information in response to these questions to complete its review of the license amendment request.
The NRC staff did not include RAIs that it determined could be affected by the licensees current effort to quantify the potential for additional Min-K to be generated or RAIs that could be affected by the related change in Min-K source term. Additional RAIs, if required, will be sent after the NRC reviews supplemental information submitted by the licensee. If the licensee believes that its additional work will affect the responses to some of the questions in this request, the responses to such questions may be submitted after the work regarding the Min-K source term is complete.
(A) Technical Specification Branch (STSB) RAIs:
STSB-RAI-2 Update the FSAR markup to describe the risk-informed analysis methodology and how it relates to the design basis deterministic requirements for Fermi. The description should clearly explain the design basis for the strainers including the debris limits and how they affect operability. The markup should clearly describe what the risk analysis presented in the LAR is used for, what debris is covered, and how it supplements the deterministic design requirements. The acceptance criteria for the risk analysis should be included. For example, the risk results must remain in Region III of RG 1.174. If the risk analysis will be used to evaluate debris source discoveries that may exceed these design basis values, the markup should outline how such an analysis would be performed and specify the acceptance criteria that will be used. Alternately, if it is intended that the risk analysis be used only to evaluate the present condition, that should be stated. A one-time to NRC-24-0027 Page 2 use of the analysis would not obviate the need to monitor the plant to assure that the risk results remain valid even after changes that may occur in the plant.
DTE Response:
In response to this RAI, Fermi herein revises Fermis LAR Attachment 1-2, UFSAR Page Markups, with the following additions to the mark-up of the discussion of Regulatory Guide 1.82 in the Fermi UFSAR.
KEY to Mark-up Below:
- Original discussion of Regulatory Guide 1.82 in the Fermi UFSAR is retained, shown in plain typeface.
- Original mark-up to discussion of Regulatory Guide 1.82 in the Fermi UFSAR as part of the Fermi LAR Attachment 1-2, is retained, shown in italics typeface.
- Additional mark-up to discussion of Regulatory Guide 1.82 in the Fermi UFSAR made in response to this RAI is shown in underlined-italics typeface.
- Revised wording is shown by a combination of crossed out and underlined-italics typeface.
The complete mark-up is shown below:
Consistent with Section D, the DTEs response to NRC Bulletin 96-03 committed to replace the original Residual Heat Removal (RHR) and Core Spray (CS) suction strainers with new, larger passive strainers designed to meet the sizing criteria of Revision 2 of this regulatory guide. The new strainers, which were designed and installed in RF06, are of the General Electric (GE) optimized stacked-disk [OSD]
design. Whereas the original design sizing was predicated on the deterministic assumption of 50% plugging, the new OSD strainers were designed under the commitment to satisfy the mechanistic design methodology described in Revision 2 of the Regulatory Guide. In their closure of the Fermi response to Bulletin 96-03, the NRC expressed their understanding that the design of the Fermi OSD strainers was performed in accordance with the method provided in NEDO-32686, BWROG Utility Resolution Guidance. The NRC SER that approved the URGs did not accept its proposed analytical methodology for calculating debris head loss and instead stipulated that the calculation of debris head loss would need to be based on vendor supplied analytical correlations developed from tested performance. This requirement is satisfied by utilizing the debris head loss methodology in the NRC-approved GE Licensing Topical Report NEDO-32721P-A, except as modified to correct elements of the method affected by errors identified in GE Safety Communication 08-02, (hereafter referred to as NEDO-32721P-A, with the understanding that NEDO-32721P-A includes the corrections in GE Safety Communication 08-02).
DTE has amended its ECCS suction strainer licensing basis consistent with the U.S.
Nuclear Regulatory Commissions Policy Statement on probabilistic risk assessment (PRA). The methodology employed is identified as a risk over to NRC-24-0027 Page 3 deterministic approach. The basic principle of the risk over deterministic analysis is that operational risks posed by accident scenarios, debris sources, and assumptions considered during original ECCS strainer design are addressed by the deterministic design basis (shown to be acceptable using NEDO-32721P-A).
Operational risks posed by newly identified failure modes and debris sources, arising from component identification tags/labels and from Min-K insulation applied in containment penetrations, are addressed by risks over or beyond the deterministic design basis and are quantified based on the extent to which they impact incremental changes to core damage frequency (CDF) and large early release frequency (LERF).
The risk analysis is described in detail in Attachment 3 of License Amendment Request for DTE Fermi Risk Informed ECCS Evaluation (SERCO-REP-DTE-22609-02 R1). At a high level, transported post-LOCA debris quantities from postulated pipe breaks at each Class 1 weld location were computed and compared to deterministic limits based on NEDO-32721P-A. In addition, fiber debris thickness accumulated on active strainers during each break scenario was compared to an assumed 1/8th-inch strainer failure criterion. CDF and LERF contributions for each break where deterministic limits or 1/8th-inch debris thickness were exceeded were summed and the total CDF and LERF from all break cases were compared to the risk region definitions in US NRC Reg Guide 1.174. The acceptance criterion of the analysis pertaining to newly identified failure modes and debris sources, arising from component identification tags/labels and from Min-K insulation applied in containment penetrations was to remain within Region III of RG 1.174. Acceptance criteria for debris defined in Fermi Calculation DC-5979 Rev-B remain based on NEDO-32721P-A. Consequently, the conditions under which strainers are considered operable are when (1) the quantities of debris other than those attributable to component identification tags/labels and from Min-K insulation applied in containment penetrations are within limits defined in NEDO-32721P-A, and (2) the total CDF and LERF from all break cases, including the effects of component identification tags/labels and from Min-K insulation applied in containment penetrations are within Region III of RG 1.174. These two conditions comprise the acceptance criteria that would need to be satisfied if new potential debris is discovered.
The risk analysis and the amended ECCS suction strainer licensing basis apply only to use of a risk-informed methodology to address potential debris quantities arising from component identification tags/labels and from Min-K insulation applied in containment penetrations exceeding those currently evaluated in the deterministic licensing methodology. The risk analysis in the amendment applies to the present condition, i.e., the amendment is not intended to allow accommodation of additional debris in the future. The risk analysis does not apply to the design debris loads, (defined in Fermi Calculation DC-5979 Rev-B) that are successfully addressed using the methodology defined in NEDO-32721P-A.
to NRC-24-0027 Page 4 The scope of the amendment does not apply to any existing configuration controls, containment cleanliness programs, or containment inventory monitoring practices.
These are not being modified or relaxed by the amendment. Existing controls include equipment labeling procedures and engineering review of all changes to containment insulation and other potential debris sources. Fermi does not anticipate any new procedural changes arising from the LAR. Fermi intends to observe all existing material controls with emphasis on labels, Min-K insulation, and fiber insulation.
DTE has demonstrated that the risk significance of potential additional debris loads posed by tags/labels and Min-K in containment penetrations is very-low as defined by Region III of RG 1.174 in terms of incremental change to core damage frequency (CDF) and incremental changes to large early release frequency (LERF).
Regulatory Guide 1.174 permits consideration of accident scenario frequency in combination with accident scenario magnitude provided that the change in risk risk-informed analysis (1) meets current regulation, or is accompanied by requested exemptions; (2) is consistent with the defense-in-depth philosophy; (3) maintains adequate safety margin; (4) is small and consistent with the NRC Safety Goal policy; and (5) is monitored by performance measurement strategies that are discussed in Attachment 3 to the Fermi LAR.
Key aspects of the analysis supporting the amendment that cannot be changed without NRC prior approval are: (1) baseline risks must remain in RG 1.174 Risk Region 3 as computed using the approved assumption of 100 ft2 of total obstruction caused by overlapped miscellaneous debris (tags/labels) transported to the clean strainer area of all active strainers; (2) the definition of strainer success/failure accepted by NRC (i.e., 1/8th inch of fiber on any single strainer causing core damage); and (3) observance of all existing material controls with emphasis on labels, Min-K insulation, and fiber insulation.
STSB-RAI-3 Update the FSAR markup to specify the key aspects, or methods, of the risk-informed analysis that cannot be changed without prior NRC approval.
DTE Response:
Please see Fermis response to STSB-RAI-2 in which key aspects of the risk-informed analysis that cannot be changed without prior NRC approval are discussed in a revised UFSAR mark-up.
STSB-RAI-4 Update the FSAR markup to define the acceptance criteria for the effects of debris on strainer performance. Define the conditions under which the strainers are considered operable. List the design basis requirements. Provide a clear definition of the deterministic limits beyond which a risk analysis is required. Define the debris that is evaluated by the risk-informed analysis and to NRC-24-0027 Page 5 any other limitations on its use. Describe how any discovered new potential debris must be addressed.
DTE Response:
Please see Fermis response to STSB-RAI-2 in which acceptance criteria, necessary conditions for strainer operability, design basis requirements, definition of the debris that is evaluated by the risk-informed analysis, and requirements that apply in the event of discovery of new potential debris are discussed in a revised UFSAR mark-up.
STSB-RAI-5 The FSAR markup, Insert 2, should be revised from change in risk to the risk-informed analysis that must meet the 5 Key Principles of RG 1.174.
DTE Response:
The revised UFSAR mark-up provided in response to STSB-RAI-2 includes the correction from change in risk to risk-informed analysis that must meet 5 Key Principles of RG 1.174.
STSB-RAI-7 Identify the material types of the non-RMI insulation shown as blue spheres in figures 2-5 and 2-6 of Attachment 3 (Serco Calculation). Clarify whether this material is accounted for in the existing deterministic debris loads or if this material was newly discovered. Identify the material of the vertical red lines in the torus in figure 2-5 or confirm that they are not debris sources.
DTE Response:
In Figures 2-5 and 2-6 of Attachment 3 (Serco Calculation SERCO-REP-DTE-22609-02 R1),
blue spheres denote isolated Nukon applications on equipment hangers, valve bodies, and whip restraints. Spherical diameter is scaled to represent appropriate volume at each location. Much smaller red spheres denote isolated Min-K applications, generally on whip restraints, but it is difficult to capture a CAD image containing both Nukon and Min-K in a meaningful graphic.
Long Min-K insulation sleeves residing in containment penetrations are also colored red.
Min-K present in containment penetrations is not accounted for in existing deterministic debris loads. All other insulation materials illustrated in Figures 2-5 and 2-6 are included in existing deterministic debris loads.
Vertical structures in the torus (colored red) represent vent lines that are not insulated and do not represent debris sources. No materials or structures in the torus affect debris generation calculations discussed in the LAR.
STSB-RAI-8 Describe how it is determined that closed check valves that are credited for system isolation for debris generation are holding pressure. Describe how the assumption in the analysis that the MS drain valves are shut and verified shut is implemented operationally.
to NRC-24-0027 Page 6 DTE Response:
If a closed check valve were not holding system pressure as intended, normal full power operation could not proceed. In accordance with Fermi specific operations procedures and engineering programs, system check valve function is confirmed during power ascent and monitored during power operation for leakage.
Similarly, full power operation could not proceed if Main Steam (MS) drain valves were not successfully closed and holding pressure. In accordance with Fermi specific operations procedures and engineering programs, MS drain line isolation is confirmed during power ascent and remain closed in support of power operations STSB-RAI-9 Describe the current design basis for area occluded by miscellaneous debris, and the updated design basis amount for the miscellaneous debris identified in the containment. In the response, provide the limit associated with strainer qualification (likely 6 ft2 of sacrificial circumscribed area), the amount of labels discovered in the containment that may transport to the strainer, and any updated deterministic limit (increase to the 6 ft2 of sacrificial circumscribed area) if applicable. Clearly state how any increase above the strainer design value is evaluated by deterministic and/or risk informed methods. Describe how any updated design limit above the design value of 6 ft2 of sacrificial circumscribed area is defined and controlled in the plant.
Provide the basis for using 100 ft2 of sacrificial area in the risk informed analysis and how this value relates to the area of labels that will be the plant design limit following implementation of the LAR. Use consistent terms in the response. For example, use either sacrificial strainer area or total label area in descriptions (overlapped or not overlapped debris area). If necessary, describe how these values are assumed to be correlated to each other and the circumscribed area design limit. It may add clarity if this information is provided in tabular format. Provide the circumscribed and total strainer areas used to determine the strainer surface area assumed to be blocked by miscellaneous debris in the strainer calculations.
DTE Response:
General Information Comparisons of debris area made in the LAR are often presented in terms of non-overlapped debris found in containment, because this is the material condition found during walk downs and event investigations. Many sub-questions of this RAI focus on sacrificial area caused by overlapped debris on the strainers. Both perspectives are useful, but careful attention must be given to the basis of each comparative statement. To improve clarity, each quantitative description will be appended with a label of (overlapped) to indicate the condition of miscellaneous debris on a strainer, or (non-overlapped) to indicate the condition of miscellaneous debris as found in containment.
to NRC-24-0027 Page 7 Consistent with guidance that allows a 50% overlap between individual (non-overlapped) debris pieces resulting in a 25% reduction in total obstructed area (overlapped) at the strainer, the relationship between the two conditions is defined as: (overlapped) = 0.75 x (non-overlapped).
The amount of miscellaneous debris found in containment that may transport to all active strainers is 87 ft2 (non-overlapped).
Table STSB-RAI-9.1 summarizes several pertinent miscellaneous debris area quantities, expressed in both overlapped and non-overlapped units.
Table STSB-RAI-9.1 Miscellaneous-Debris Summary.
Debris Area Overlapped Non-Overlapped Design basis total obstruction
6 0.75
= 8 RHR strainer design basis 6 x 0.643 = 3.86 8 x 0.643 = 5.14 CS strainer design basis 6 x 0.357 = 2.14 8 x 0.357 = 2.86 Newly identified misc debris 87 x 0.75 = 65.25
Baseline total obstruction
100 0.75
= 133.33 Total design basis + New debris 6 + 65.25
= 71.25 71.25 0.75
= 95 Baseline > (Total DB + New debris) 100 > 71.25 133.33 > 95 Miscellaneous debris margin 100 71.25
= 28.75 133.33 95
= 38.33 The total face area of each GE optimized stacked disk strainer = 387.42 ft2 (Fermi, Impact of Additional Containment Penetration Min-K on ECCS Suction Strainer Design, EFA-E11 004, Rev. A, p.7 of 48). Minor variations may exist between strainers because of differences in mounting brackets and structural braces. The total face area of each strainer as installed is smaller than the manufactured face area by approximately 4 ft2. Face Area refers to the perforated plate surface that supports active flow.
The circumscribed area of each GE optimized stacked disk strainer = 48.77 ft2. (Fermi, Impact of Additional Containment Penetration Min-K on ECCS Suction Strainer Design, EFA-E11-16-004, Rev. A, p.7 of 48). Minor variations may exist between strainers because of differences in mounting brackets and structural braces. Circumscribed Area refers to the approximately cylindrical shape that would result from wrapping an entire strainer in plastic, i.e., the three-dimensional convex outer perimeter.
to NRC-24-0027 Page 8 Debris Load Factors Strainer head loss qualification calculations are documented in Reference GEH Nuclear Energy, Parametric Strainer Head Loss Evaluation Fermi Unit 2, 0000-0155-4058, Rev. 0, PROPRIETARY (December 2012), where it is determined that the maximum total allowable blockage of circumscribed strainer area by miscellaneous debris is approximately 6 ft2 (overlapped). This value of obstructed strainer area was determined by iteration to be consistent with design-basis debris loads and with hydraulic Net Positive Suction Head (NPSH) calculations for both the RHR and CS systems. The strainer qualification calculation divides total obstructed area (overlapped) between one RHR strainer operating at 11,000 gpm and one CS strainer operating at 6,100 gpm according to their relative flow rates to determine debris load factors of 11,000 (11,000 + 6,100)
= 0.643 for the RHR strainer and 6,100 (11,000 + 6,100)
= 0.357 for the CS strainer.
Assumed baseline flow rates of 15,540 gpm for operating RHR strainers and 8,200 gpm for the operating CS strainer are consistent with high-flow, high-temperature hydraulic calculations with pump overspeed, but the highest pump flow rate determined in the hydraulic resistance calculation for each system is assigned to all operating pumps in the system for the purpose of increasing the debris accumulation rate and minimizing the time to strainer failure. The assumed baseline flow rates, which are higher than the strainer qualification flow rates, are not expected to occur in practice, but are analyzed for completeness in the RHR and CS hydraulic calculations. The risk-informed LAR does not change pump performance requirements or qualification calculations in any way.
When strainer qualification load factors are applied to the total design-basis obstruction of circumscribed area by miscellaneous debris (overlapped), it is determined that each operating RHR strainer can accommodate 6 2 x 0.643 = 3.86 2 of miscellaneous debris (overlapped) on a circumscribed debris bed, and each operating CS strainer can accommodate 6 2 x 0.357
= 2.14 2 of miscellaneous debris (overlapped) on a circumscribed debris bed. Because RHR system NPSH requirements are more stringent and all strainers are the same size, the miscellaneous debris load allowed for RHR strainers (both overlapped and non-overlapped) is the more limiting factor.
Strainer qualification calculations also include 21.6 ft3 of fiber (when the design basis limit for Min-K is also counted as fiber). At the design basis fiber load, the theoretical maximum debris bed thickness on the RHR strainer would be (21.6 x 0.643) 387.42 x 12 = 0.43
. This result suggests that the larger anticipated debris loads would be capable of filling the approximately 1-inch gaps between strainer disks and/or bridging between the plate edges to form a circumscribed debris bed configuration. The baseline assumed failure thickness of 1/8th-in. on all active surfaces would not be capable of completely filling interstitial gaps between plates, so the design-basis obstruction of 6 ft2 (overlapped) subtracted from fully circumscribed strainers does not provide a sensitive risk-discrimination metric, because all high-fiber debris scenarios that could form a circumscribed condition are already assigned to ECCS failure by assumption.
to NRC-24-0027 Page 9 Head Loss Scaling Argument GE strainer qualification calculations subtract the miscellaneous debris area obstruction (overlapped) from the circumscribed area to determine an effective approach velocity that yields total strainer head loss consistent with NPSH requirements. Total strainer head loss equals the sum of clean strainer head loss and debris-induced head loss. For the RHR strainers, the effective circumscribed approach velocity is:
=
=
11,000 (48.77 2 3.86 2)448.8
= 0.55 Inside the debris bed, water velocity would deaccelerate to a strainer face velocity of only
=
11,000 (387.42 2) 448.8
= 0.0633 which controls clean strainer head loss and is approximately 8.7 times slower than the velocity experienced at the outer face of a circumscribed debris bed.
More recent industry-wide strainer performance assessments, consistent with guidance, generally subtract miscellaneous debris area from the clean face area of all active strainers in proportion to their initial flow rates. The baseline assumption for total sacrificial debris area of 100 ft2 (overlapped) defines the total obstruction (overlapped) across all active strainers. Applying the design-basis load factor, one active RHR strainer would receive 100 x 0.643 = 64.3 2 (overlapped). This amount of obstruction on the clean strainer would induce a face approach velocity of
=
11,000 (387.42 2 64.3 2)448.8
= 0.076 which is (0.076 0.0633) 0.0633
x 100 = 20% faster than would be experienced by a clean strainer. As expected, clean-strainer head loss would increase somewhat under the baseline debris obstruction.
However, a 1/8th-in. thin bed would also experience nearly the same approach velocity as the obstructed strainer face, and this is (0.55
) (0.076
)
= 7 times slower than the qualification calculations allow for the circumscribed debris bed. Because both clean strainer head loss and debris-bed head loss are proportional to fluid velocity (or velocity squared), the reduction in debris-induced head loss through a smaller amount of fiber, with the baseline miscellaneous debris applied to one RHR strainer, would be much greater than the increase in clean strainer head loss caused by the baseline miscellaneous debris obstruction.
Thus, the combination of a small baseline failure load (1/8th-in. fiber thickness) and a reasonably bounding baseline sacrificial area (100 ft2 (overlapped)) assures that successful break scenarios having less than 1/8th-in. of fiber on all strainers also have total RHR-strainer head loss less than the design basis maximum. Failed break scenarios that exceed the 1/8th-in. debris-thickness to NRC-24-0027 Page 10 criterion contribute their assigned frequency to total risk associated with the baseline 100 ft2 (overlapped) limit.
Application of the proportional miscellaneous debris obstruction to one RHR strainer ensures that the conclusions of this scaling argument hold for suppression pool cooling configurations where only one RHR and one CS strainer are operating, and that the conclusions bound all other strainer configurations where more than two strainers collect debris.
Plant Implementation The baseline total strainer obstruction of 100 ft2 (overlapped), which equates to 100 2 0.75
=
133.33 2 (non-overlapped) of debris in containment, was chosen to bound the 87 ft2 (non-overlapped) of recently identified miscellaneous debris that may transport to all active strainers plus the total design-basis allowance. Note that the newly identified material represents an additional total strainer blockage potential of 87 2 x 0.75 = 65.25 2 (overlapped) that is not included in the strainer design basis. The last two rows of Table STSB-RAI-9.1 confirm that the baseline assumption bounds the sum of the total design basis and the newly identified debris in both the overlapped and non-overlapped configurations, and allows some margin for unexpected emergent issues.
Use of 100 ft2 (overlapped) of total obstructed strainer area as a baseline risk-evaluation assumption does not change the deterministic design-basis limit of 6 ft2 (overlapped) of total obstructed strainer area. No update to design basis strainer obstruction is needed to account for miscellaneous debris newly identified in containment. Fermi will continue to enforce metal tag labeling requirements, containment cleanliness procedures, and manage hydraulic calculations to meet the existing 6 ft2 (overlapped) deterministic qualification limit that is based on one RHR strainer and one CS operating in a configuration representing the suppression pool cooling mode.
The purpose of analyzing the 100 ft2 (overlapped) baseline total obstruction area is to show that the risk of ECCS failure posed by a conservatively large, assumed obstruction (enough to account for the design basis and the newly identified material) is acceptable by RG1.174 Risk Region definitions. Quantified risk associated with 100 ft2 (overlapped) of total obstruction area represents risk over, or greater than, the deterministic (RoverD) 6 ft2 (overlapped) limit on total obstruction area. While the deterministic 6 ft2 (overlapped) design basis will not change following LAR acceptance, the 100 ft2 (overlapped) of total obstruction area establishes a new licensing basis limit that Fermi cannot exceed, in either the overlapped or non-overlapped configuration, without informing the NRC. Existing containment configuration controls and margin embedded in the baseline obstruction limit ensure that the 100 ft2 (overlapped) licensing basis will not be exceeded.
STSB-RAI-10 Miscellaneous debris (tags and labels) that may cover greater area than the original design assumption can affect the areal density of non-fibrous debris types (other than Nukon or LDFG) that may collect on the strainer. Describe how the increase in the areal density of debris is evaluated. Discuss whether the change in the areal density of the other debris types is bounded to NRC-24-0027 Page 11 by the design basis analysis. Discuss how any increased bed thickness or increased velocity though the debris bed is considered in the evaluation.
DTE Response:
The 1/8th-in baseline debris-thickness failure criterion is selected to preclude formation of a contiguous filter medium that can lead to collection of non-fibrous debris types and formation of a high-density thin-bed. A principal objective of the LAR is to assign all such debris combinations to ECCS failure, without any need to quantify head loss. Because of this approach, investigations of increased areal density do not affect the risk tabulation.
It is, however, important to understand if risk is sensitive to the baseline value of obstructed strainer area. A forthcoming LAR supplement, prepared to answer RAIs related to parameter sensitivity, reports very little change in risk compared to the baseline when assumed strainer area obstruction is changed from the baseline value of 100 ft2 to 65 ft2 and 120 ft2 (overlapped). The primary effect that obstructed strainer area introduces to the risk quantification is controlling how much debris is required to reach the defined failure thickness. Recall that Min-K volume, traditionally a non-fibrous debris type, is counted as fiber in the comparison of debris bed thickness to the failure criterion. This conservative approximation was made to simplify treatment of the fiber binder present in the largely particulate micro-porous insulation material and ensure that Min-K quantity directly contributes to identified ECCS failures.
The following discussion is provided to quantify the increase in debris areal density caused by the increase in assumed strainer blockage by miscellaneous debris and compare the increased areal density to the design basis analysis.
The baseline assumption of 100 ft2 (overlapped) of total strainer face area blockage increases the areal density {2
} of any given debris load compared to the areal density of the design basis.
For the RHR strainer using the design basis load factor (see response to STSB-RAI-9), the ratio of the generic areal density under baseline obstruction
to the generic areal density under design basis obstruction
is
=
(387.42 2 100 2 x 0.643)
(387.42 2 6 2 x 0.643)
= 383.56 323.12 = 1.19 where is the mass density of debris type and is the volume of debris type. The sum of debris masses is the same in the numerator and denominator of the ratio, so the result depends only on the ratio of the adjusted strainer areas. The baseline obstructed area increases the areal density of any given debris load by approximately 20%.
The response to STSB-RAI-9 explains that at the design basis fiber load, the theoretical maximum debris bed thickness on the RHR strainer would be (21.6 x 0.643) 387.42 x 12 =
0.43. thick. The risk-informed methodology implemented in the LAR imposes strainer failure at a debris thickness of 1 8
= 0.125., which is 3.4 times thinner and effectively has an areal density that is 3.4 times lower than that of the design basis for any given debris combination. The reduction in areal density imposed by the 1/8th-in. baseline failure criterion is much greater than the increase in areal density caused by the baseline 100 ft2 (overlapped) miscellaneous debris obstruction.
to NRC-24-0027 Page 12 The combination of a minimal debris load failure criterion (1/8th-in) and a reasonably bounding strainer obstruction limit (100 ft2 (overlapped)) keeps risk evaluation areal densities much lower than the design basis areal density.
Velocity changes through the debris bed that are induced by the baseline strainer obstruction are explained in the response to STSB-RAI-9. For the limiting RHR strainer, at the baseline failure thickness of 1/8th-in., velocity approaching the debris is lower by a factor of 7 compared to the design-basis circumscribed debris configuration.
Debris loads in the design-basis analysis of the limiting RHR strainer have a much greater approach velocity, thickness, and areal density than debris loads present at the 1/8th-in. baseline failure thickness, despite the larger strainer obstruction (100 ft2 (overlapped)) assumed in the baseline risk evaluation. Greater approach velocity, thickness, and areal density all imply greater head loss, indicating that the effect of larger strainer area obstruction on non-fibrous (and fibrous) debris types is bounded by the design basis analysis.
STSB-RAI-13 In the description of the torus cooling mode on page 39 of 94 of the Serco calculation (PDF pg.
111) it is stated that the flow rate is 10,000 gpm rated. Explain what rated means in this case.
What flow rates for torus cooling mode are considered in design basis analyses? Provide a discussion that demonstrates that the design basis conditions are adequately evaluated by the risk-informed analysis.
DTE Response:
The term rated used to describe 10,000 gpm of RHR flow in the original strainer sizing calculation is somewhat non-specific and often used interchangeably with nominal, meaning only that the pumps can easily supply the stated flow, which is described in round figures as a target value descriptive of the operational mode. Operators have flexibility and discretion to set exact flow rates as determined by accident conditions.
The response to STSB-RAI-9 explains that the most recent strainer qualification calculation divides debris between one RHR strainer operating at 11,000 gpm and one CS strainer operating at 6,100 gpm in suppression pool cooling mode. The strainer qualification calculation establishes the design basis for use of 6 ft2 (overlapped) as the total amount of flow area obstructed by miscellaneous debris, and it places a slightly greater emphasis on flow through the RHR system, which has more stringent NPSH limitations than the CS system. These revised flow rates determine debris load factors of 11,000((11,000+6,100) ) =0.643 for the RHR strainer and 6,100/((11,000+6,100) ) =0.357 for the CS strainer. Debris load factors describe the proportion of debris accumulated by each strainer. Load factors can be time dependent if pump flow rates change.
A forthcoming LAR supplement, prepared to respond to RAIs related to parameter sensitivities, provides a discussion that demonstrates that the design basis conditions are adequately evaluated by the risk-informed analysis. The supplement compares baseline single-division Low Pressure to NRC-24-0027 Page 13 Coolant Injection (LPCI) flow to risk quantified for several alternate pump flow configurations, including suppression pool cooling.
(B) Vessels and Internals Branch (NVIB) RAIs:
DTE General Response to NVIB RAIs No adjustments (up or down) were made to NUREG-1829 weld break frequencies to account for inspection results or industry standard mitigation practices. Fermi has no adverse findings from weld inspections. NUREG-1829 includes consideration of all degradation mechanisms highlighted in NVIB questions and assumes adherence to industry best practice weld management initiatives, which Fermi implements. Uncertainties in weld break frequency are addressed by quantification of risk assuming Arithmetic Mean aggregation of the NUREG-1829 expert panel elicitation. It is not typical for risk-informed applications to adjust individual weld break frequencies, though some studies such as the BWROG Fleet Vulnerability Study have applied "bottom up" weighting factors by plant system, which are also provided in NUREG-1829. Bottom-up, system-specific weighting factors are not applied for risk quantification in the Fermi LAR.
NVIB-RAI-1 The Serco calculation, page 25, (PDF pg. 97) states that 1103 welds were selected as the possible LOCA locations in the risk-informed application of CASA Grande. The licensee further stated that The review further identified a subset of 924 welds that are considered active LOCA locations given the at-power plant configurationA subset of only 921 welds represents potentially active LOCA locationsThe plant configuration provides for auto-isolation of 37 welds that reduce the number of welds to 887 unique locations The licensee further reduced the LOCA location to 884 welds.
Discuss how 1103 welds were reduced to 924 welds although the licensee did state that the 924 welds are considered active LOCA locations. Explain why 179 welds (1103-924) were eliminated from consideration. (2) Clarify whether 884 welds or 921 welds are included in the risk analysis. (3) Discuss whether every one of the 884 welds has been ultrasonically examined at least once. (4) Clarify whether 884, 924, 921, or 1103 welds are considered in the scope of the license amendment request (i.e., in-scope welds).
DTE Response:
(1) The purpose of identifying active LOCA locations on welds is to locate the center of damage zones that can generate debris if the weld ruptures. All welds in the initial comprehensive set are evaluated for debris-generation potential. Welds not carried forward for analytic quantification are deemed to have negligible contribution to the risk of ECCS failure caused by debris sources.
Candidate break locations started with ISI welds identified on Piping and Instrumentation Diagram (P&ID) isometric drawings. Of the candidate locations, welds were further evaluated based on location relative to valves and on system configuration with the purpose to NRC-24-0027 Page 14 of confirming whether failure could result in ECCS suction strainer clogging. The down selection process is documented in several locations of Att-3 (Appendix D of LAR Att-3, pdf p114 of 202). Appendix D includes the full list of Fermi welds evaluated and considers valve positions during normal operation.
Some welds in the initial set are outside of containment or downstream of normally closed isolation valves and were removed from further quantification. High Pressure Coolant Injection (HPCI) and Reactor Core Isolation Cooling (RCIC) steam line hanger support welds were also removed from the risk quantification.
(2) All welds appearing on ISI isometric drawings are considered for debris generation and ECCS failure potential. Of the welds selected for risk quantification, 37 welds are protected by automatic isolation valves (isolable) and 887 are analyzed as nonisolable (including 3 MS Drain Line Isolation welds that could logically be screened from further consideration, but remain in the quantification. (See Att-3, pdf p.25 of 202).
(3) Every one of the 884 welds have not been ultrasonically examined at least once. In accordance with ASME Section XI, ASME Code Case N-716-1, and BWRVIP-75-A, ultrasonic examinations are performed on a sampling basis.
(4) See response to Question 1 above. For the purpose of the following responses, we define in-scope welds to include 37 that are protected by automatic isolation valves (isolable) and 887 analyzed as nonisolable (including 3 MS Drain Line Isolation welds that could logically be screened from further consideration, but remain in the risk quantification). The total number of in-scope welds (924) is never considered as a single combined population for the purpose of risk quantification.
NVIB-RAI-2 Section 6.5 of the Serco Calculation, page 83 (PDF pg.155), states that All Class I welds at Fermi are either Category A, welds with no known cracks that are made from materials that are considered resistant to IGSCC, or Category B, welds made from material that is considered susceptible to IGSCC but have been mitigated by stress improvement prior to two cycles of operation.
Discuss whether Category A welds have recorded indications from the ultrasonic examinations performed during inservice inspection (ISI) intervals even though the report states the welds have no known cracks (i.e., difference between an indication and a crack). (2) Discuss any degradation occurred in the Category B welds after mitigation. If degradation was identified, discuss the corrective actions. (3) Discuss whether Class 2 welds in the drywell/containment are to NRC-24-0027 Page 15 considered as in-scope welds (i.e., are they part of the risk analysis). If not, clarify why the Class 2 welds in the containment are not included in the risk analysis.
DTE Response:
(1) No Class 1 piping welds at Fermi have indications unacceptable to ASME Section XI.
(2) No degradation of Intergranular Stress Corrosion Cracking (IGSCC) Category B welds has been identified after mitigation.
(3) Class-2 welds are not considered as in-scope welds in the Fermi LAR as locations for a LOCA, because they are generally on low-energy systems having minimal debris generation potential and negligible risk contribution via ECCS strainer obstruction.
NVIB-RAI-3 Section 6.5 of the Serco Calculation, page 83 (PDF pg. 155), states that Pressure retaining welds are inspected in accordance with ASME Code Case N-716-1 Alternative Classification and Examination Requirements. This Code Case prioritizes inspection of risk significant welds and welds potentially susceptible to a degradation mechanism, such as Intergranular Stress Corrosion Cracking (IGSCC) or thermal fatigue Ten percent of Class I welds are examined over a ten-year interval...
Discuss whether the in-scope welds have been examined per the ASME Code,Section XI. (2)
Discuss whether the same ten percent of the Class 1 welds are examined per Code Case N-716-1 during every 10-year ISI interval or different population of welds are examined in different 10-year ISI intervals. (3) Discuss the percent of in-scope welds that have not been inspected and will not be inspected to the end of the license renewal period. (4) For the in-scope welds that will never be inspected, discuss whether the higher probability of failure value was used for these welds than for the inspected welds in the risk analysis.
DTE Response:
(1)
Consistent with ASME Code,Section XI requirements, 100% of in-scope welds have not been examined. Fermi inspections have found no indications adverse to ASME weld management standards.
(2)
The current ISI 4th Interval is the first full Interval implementing Code Case N-716-1.
Pending any programmatic changes, the same ten percent of Class 1 welds will be examined per Code Case N-716-1 during future 10-year ISI intervals.
(3)
Unless programmatic weld inspection requirements change, or adverse indications are found in the inspected weld population, the same 90% of Class 1 welds will not be inspected to the end of the license renewal process. Changes to programmatic requirements, made in response to changes in ASME standards or industry best practice initiatives, or adverse indications found during periodic inspections may result in additional inspections of to NRC-24-0027 Page 16 previously uninspected Class 1 welds.
(4) NUREG-1829 weld failure frequencies were not increased nor decreased to account for past or future inspection results. NUREG-1829 weld failure frequencies applied in the LAR are consistent with values used in the Fermi PRA, and they assume compliance with industry weld inspection, maintenance, and mitigation requirements.
NVIB-RAI-4 Discuss any in-scope welds that are fabricated with nickel-based Alloy 82/182. (2) Discuss whether these inscope welds have been mitigated to reduce their susceptibility to stress corrosion cracking. (3) Discuss whether any in-scope nickel-based Alloy 82/182 welds that have not been mitigated. (4) For those unmitigated in-scope welds, discuss how they are being inspected to monitor their structural integrity. (5) Discuss whether the failure probability value in the risk analysis is increased to account for the unmitigated nickel-based Alloy 82/182 welds.
DTE Response:
(1 - 3) Welds within the scope of GL 88-01 have been mitigated to reduce their susceptibility to stress corrosion cracking.
(4)
Class 1 welds that are not within the scope of GL 88-01 or N-716-1 are subject to VT-2 inspection during the Reactor Pressure Vessel Pressure Test.
(5)
NUREG-1829 weld failure frequencies were not increased nor decreased to account for degradation mechanisms and industry standard mitigation practices that are already considered in the NUREG failure frequencies NVIB-RAI-5 Discuss the number of austenitic stainless steel welds that are susceptible to intergranular stress corrosion cracking (IGSCC) and are included in the risk analysis. (2) Discuss whether these in-scope austenitic stainless steel welds have been mitigated to reduce their susceptibility to IGSCC (excluding the improvement in primary system water chemistry). (3) For those in-scope unmitigated austenitic stainless steel welds, discuss the corrective actions (excluding water chemistry improvements). (4) Discuss whether the failure probability value in the risk analysis is increased to account for the unmitigated austenitic stainless steel welds that are susceptible to IGSCC.
DTE Response:
(1) 38 BWRVIP-75A Category A welds and 106 BWRVIP-75A Category B welds are included in the risk analysis.
(2) These welds have been mitigated to reduce their susceptibility to IGSCC.
(3) Class 1 austenitic stainless steel welds outside the scope of BWRVIP-75A have not been mitigated and no corrective actions have been required.
to NRC-24-0027 Page 17 (4) NUREG-1829 weld failure frequencies were not increased nor decreased to account for degradation mechanisms and industry standard mitigation practices that are already considered in the NUREG failure frequencies.
NVIB-RAI-6 BWR owners have implemented inspection guidance of BWRVIP-75-A, BWR Vessel and Internals Project Technical Basis for Revisions to Generic, Letter 88-01 Inspection Schedules.
Discuss whether inspection guidance of BWRVIP-75-A has been implemented for the inservice inspection of in-scope welds at Fermi. (2) If affirmative, discuss how inspections of in-scope welds are carried out per the BWRVIP-75-A. (3) If the topic report is not implemented, discuss the reason.
DTE Response:
(1) Fermi currently implements BWRVIP-75-A for inservice inspection of in-scope welds.
(2) Ten percent of Category A welds and 25% of Category B welds are ultrasonically examined every interval as part of the ISI Program. Examinations are performed in accordance with ASME Section XI Appendix VIII and the EPRI Performance Demonstration Initiative.
(3) Not applicable.
NVIB-RAI-7 Discuss whether failure of flanges and bolts of a piping system, nozzle penetrations to the reactor vessel such as control rod drive mechanisms and standby liquid control system are considered as potential LOCA locations and included in the risk analysis.
DTE Response:
Mechanical couplings are not included as potential LOCA locations, because:
(1) The spatial coverage of welds is generally considered sufficient to capture potential debris generation that might occur near failed couplings, especially in a relatively small BWR drywell.
(2) NUREG-1829 specifically excludes mechanical couplings, so assigning NUREG-1829 break frequency to couplings would artificially reduce or dilute the intended weld break frequency.
(3) Mechanical couplings having multiple physical connection points are not suspected to allow large catastrophic pressure release the same as for Double Ended Guillotine Break (DEGB) of a weld.
(4) Multiple bolt failures on a flange needed for an equivalent size large break significantly reduce the estimated event frequency compared to weld failure of the same size.
to NRC-24-0027 Page 18 Nozzle penetrations such as control rod drive mechanisms and the standby liquid control system are also not included as potential LOCA locations, for similar reasons as not considering mechanical couplings. In addition, nozzle connections having proper structural integrity are assumed to be more mechanically robust than other Reactor Coolant Pressure Boundary welds, leading to negligible comparative catastrophic break frequency and a longer expected service life for inspection, identification, and mitigation of any evolving degradation mechanisms.
NVIB-RAI-8 The Serco Calculation, page 33 (PDF pg. 105), Table 2-1 shows the failure probability for feedwater is 1.57E-4. Explain why this failure probability is lower than other piping in the table such as main steam and RWCU.
DTE Response:
Table 2-1 cites on-demand failure probabilities for isolation valves. The entry for Feedwater also notes that the valve type is a check valve (CHK). Mechanical simplicity leads to higher reliability (lower failure probability) for check valves than for Air Operated and Motor Operated valves used on Main Steam and Reactor Water Clean Up (RWCU) systems.
(C) Probabilistic Risk Assessment Licensing Branch B (APLB) RAI APLB-RAI-1 In Attachment 3 (Serco calculation) of the license amendment request (LAR) on page 51 of 94 (PDF pg. 123), the double ended guillotine break (DEGB) model is represented as the baseline model. The NRC staff noted that partial breaks were not considered in the sensitivity analysis of the change in Core Damage Frequency (CDF). In accordance with NEI 04-07, Pressurized Water Reactor Sump Performance Evaluation Methodology, Revision 0, December 2004, (ADAMS Accession No. ML050550156), to assure defense in depth is maintained, mitigative capability must be maintained through all break sizes up through DEGB. Also, LARs in response to GL 2004-02, Potential Impact of Debris Blockage on Emergency Recirculation During Design Basis Accidents at Pressurized-Water Reactors, for PWRs have traditionally computed the CDF considering both the continuum break model (where pipes are assumed to exhibit partial breaks) and the DEBG-only model in sensitivity analyses. NUREG-0800, Standard Review Plan, Section 15.6.5, Loss-Of-Coolant Accidents Resulting from Spectrum of Postulated Piping Breaks Within the Reactor Coolant Pressure Boundary, (ADAMS Accession No. ML070550016), discusses the staff performing an evaluation on whether the entire break spectrum (break size and location) has been addressed and demonstrating compliance with requirements of 10 CFR 50.46. 10 CFR 50.46 requires that ECCS cooling performance must be calculated in accordance with an acceptable evaluation model and must be calculated for a number of postulated loss-of coolant accidents of different sizes, locations, and other properties sufficient to provide assurance that the most severe postulated loss-of-coolant accidents are calculated. Understanding that partial hemispherical breaks occurrence affects the CDF, the to NRC-24-0027 Page 19 NRC staff is requesting that the licensee provide justification for considering only DEGB in its analysis.
DTE Response:
For Fermi, and in general for all BWR, analysis of DEGB spherical Zone of Influence (ZOI) is more consistent with the current strainer licensing basis. However, if the NRC agrees that hemispherical partial breaks defined in NEI-04-07 [Ref. 1] (a PWR guidance document) are useful and appropriate for the LAR, Fermi can generate standard CASA Grande output for the continuum break model and report risk for partial breaks as a sensitivity case.
Fermi chose to use DEGB ZOI as the baseline risk evaluation case because: (1) of consistency of the DEGB methodology with the existing strainer licensing basis; (2) to avoid the gap in present guidance that does not provide for a smooth transition between large hemispherical breaks and a full DEGB, which introduces the non-intuitive statistical effect that DEGB carry a vanishingly small probability in the continuum break model and implies that continuum-model risk cannot be fully interpreted without a corresponding DEGB analysis, and (3) because subject matter expert experience with PWR risk-informed analyses suggests that DEGB-model risks bound continuum-model risks, thus facilitating a more efficient analysis and review process for the Fermi LAR.
The current licensing basis for the Fermi GE stacked disk strainers was developed under BWR Utility Resolution Guidance (URG) guidelines [Ref. 2] that predates the use of partial hemispherical breaks discussed in PWR guidelines developed for NEI-04-07 [Ref. 2], but is considered to be fully compliant with NUREG-800 and all requirements for calculating ECCS cooling performance for a number of postulated LOCA of different sizes, locations, and other properties. Current automated DEGB analyses of all weld locations in containment as postulated break locations are substantially more comprehensive than the manual methods formerly used to compute debris sources and verify, for PWR and BWR strainer designs, that the most severe postulated loss-of-coolant accidents are calculated.
Hemispherical breaks can never generate more debris than a DEGB at the same location because hemispherical breaks have the same radius of ZOI for every debris type, but only half of the volume of the corresponding spherical DEGB. However, the methodology used for continuum-break frequency conservation does introduce smaller breaks on large pipes at more locations that can refine the transition between fiber debris volumes that pass assumed strainer failure criteria and those that fail, which is useful for examining the exceedance function of cumulative risk for abrupt transitions (cliffs) near the RG-1.174 Risk Region boundaries. Figure APLB-RAI-1.1, prepared for a forthcoming LAR supplement to answer RAIs related to sensitivity parameters, illustrates the risk exceedance function (blue line) for all DEGB non-isolable breaks, defined with respect to the baseline 1/8th-in. fiber thickness failure criterion. The DEGB risk exceedance function is relatively smooth in the region near 1.0E-06/yr, which forms the boundary between RG1.174 Risk Region II and Risk Region III.
The baseline strainer flow history lies in a group of alternate strainer flow Cases and gives a risk contribution from non-isolable breaks of approximately 8.0E-7/yr. (The LAR to NRC-24-0027 Page 20 supplement includes penetration Min-K as a debris source and reports associated risk of exceeding strainer design-basis Min-K limits.)
Figure APLB-RAI-1.1. DEGB risk-exceedance function for 1/8th-in. fiber thickness failure criterion.
Higher break-size sampling resolution provided by the continuum break model spreads NUREG-1829 break frequency available in any defined break size interval across all welds that are larger than or equal in diameter to the upper bound of the given size interval.
However, frequency conservation requires that the fractional frequencies carried by hemispherical breaks postulated within the size interval are inversely proportionally to the number of breaks in the interval (i.e., smaller). Frequency conservation for the DEGB break model works similarly, except the number of break size intervals is determined solely by the number of pipe sizes rather than by a statistical sampling strategy; but again, NUREG-1829 break frequency available in any break size interval defined between pipe sizes is shared across all welds larger than or equal to the upper bound of each interval. Because pipe sizes are necessarily included as interval boundaries in the continuum model, the DEGB methodology can be viewed as a coarse discretization of continuum break sizes where any break occurring in a size interval is simulated as a single spherical DEGB ZOI rather than as a set of smaller hemispherical ZOI. Conservation requires that the sum of frequencies for multiple small intervals, and multiple breaks per interval, between two pipe sizes (continuum model) can never exceed the sum of frequencies over larger intervals between the same two pipe sizes (DEGB model) when the two approaches share the coarse interval boundaries in common.
Fiber Thickness Failure Spectrum 10*3 ~-----~-----~-----~----~
><' 10*4
- \\I 5-
~
i', 10-5 C
Q)
O'
~
u..
~ 10*6. """""""""""""""""""""""""""""""""""""'" :*''''''''''''''''''''
C rn
~
Q)
Q) u il:i 10*7 10*8 '----~--'----~--'------='----"---......,
10*2 10*1 10° 101 Suppression Pool Total Fiber Volume (tt3) to NRC-24-0027 Page 21 Cumulative break frequency within any size interval at a single weld, whether summed over multiple assumed hemispherical breaks or represented by a single DEGB, is only counted as an ECCS failure if the amount of debris generated exceeds the strainer performance metric of interest. In a hypothetical example where fiber insulation totally fills containment, it is possible that all hemispherical breaks in a size interval do not exceed the strainer limit and zero risk is recorded, but the spherical ZOI, being twice as large, does exceed the strainer limit and the full frequency of all hemispherical breaks postulated at that weld is recorded as risk. The true heterogeneous nature of problematic insulation in the Fermi containment makes it even more likely that a portion of hemispherical breaks postulated by the continuum model at one weld location will pass, but the spherical ZOI will fail and contribute a bounding failure frequency (i.e., risk).
Higher break-size sampling resolution is generally more useful for generating a smooth risk distribution when applied to large dry PWR containments having sparse, or highly heterogenous insulation applications, but in the smaller Fermi BWR drywell, spherical DEGB breaks tend to homogenize spatial variations between isolated, spot insulation locations, because they consume a larger percentage of the total containment volume than hemispherical breaks placed at the same locations, and generate correspondingly smooth debris spectra. For perspective, the smallest DEGB capable of generating the 9 ft3 of fiber needed to reach a 1/8th-inch strainer failure criterion in the baseline flow condition is approximately 18 inches in diameter.
References:
- 1. Nuclear Energy Institute, Pressurized Water Reactor Sump Performance Evaluation Methodology, NEI-04-07, Rev., December 2004.
- 2. BWROG, "Utility Resolution Guide for ECCS Suction Strainer Blockage - Volume 1 & 2, NEDO-32686-A," October 1998.
to NRC-24-0027 Fermi 2 NRC Docket No. 50-341 Operating License No. NPF-43 LAR NRC-23-0020 Attachment Mark-ups
FERMI 2 UFSAR A-42 REV 24 11/22 Topical Report NEDO-32721P-A, except as modified to correct elements of the method affected by errors identified in GE Safety Communication 08-02.
A.1.83 REGULATORY GUIDE 1.83 (July 1975, Revision 1), INSERVICE INSPECTION OF PRESSURIZED WATER REACTOR STEAM GENERATOR TUBES This guide is not applicable to Fermi 2, which is a BWR.
A.1.84 REGULATORY GUIDE 1.84 (September 1983, Revision 21), DESIGN AND FABRICATION CODE CASE ACCEPTABILITY--ASME SECTION III, DIVISION 1 The Fermi 2 plant is in compliance with Regulatory Guide 1.84.
To ensure integrity of the RCPB commensurate with its important safety function, Fermi 2 has applied the code cases of the ASME Boiler and Pressure Vessel Code Section III, to design, fabrication, erection, and testing of Class 1 components within the limitations set forth in 10 CFR 50, Section 50.55(a).
For specific identification of the code cases used, refer to Table 5.2-3.
A.1.85 REGULATORY GUIDE 1.85 (September 1983, Revision 21), MATERIALS CODE CASE ACCEPTABILITY--ASME SECTION III, DIVISION 1 To ensure integrity of the RCPB commensurate with its important safety function, Fermi 2 has applied the code cases of the ASME Boiler and Pressure Vessel Code Section III, to design, fabrication, erection, and testing of Class 1 components within the limitations set forth in 10 CFR 50.55(a) and Regulatory Guide 1.85. Thus the Fermi 2 RCPB is in compliance with the positions of this guide.
For specific identification of the code cases used, refer to Table 5.2-3.
A.1.86 REGULATORY GUIDE 1.86 (June 1974), TERMINATION OF OPERATING LICENSES FOR NUCLEAR REACTORS This guide is not presently applicable to Fermi 2. At the time of decommissioning and dismantlement of the Fermi 2 plant, Edison intends to follow procedures in compliance with this guide.
A.1.87 REGULATORY GUIDE 1.87 (June 1975, Revision 1), GUIDANCE FOR CONSTRUCTION OF CLASS 1 COMPONENTS IN ELEVATED TEMPERATURE REACTORS (SUPPLEMENT TO ASME SECTION III CODE CASES 1592, 1593, 1594, 1595, and 1596)
This guide is not applicable to the Fermi 2 BWR.
, (hereafter referred to as "NEDO-32721P-A", with the understanding that "NEDO-32721P-A" includes the corrections in GE Safety Communication 08-02).
delete period Add Insert 2 from the following page as a continuation of A.1.82 to NRC-24-0027 Page 1 to NRC-24-0027 Page 2 INSERT 2 DTE has amended its ECCS suction strainer licensing basis consistent with the U.S. Nuclear Regulatory Commissions Policy Statement on probabilistic risk assessment (PRA). The methodology employed is identified as a risk over deterministic approach. The basic principle of the risk over deterministic analysis is that operational risks posed by accident scenarios, debris sources, and assumptions considered during original ECCS strainer design are addressed by the deterministic design basis (shown to be acceptable using NEDO-32721P-A). Operational risks posed by newly identified failure modes and debris sources, arising from component identification tags/labels and from Min-K insulation applied in containment penetrations, are addressed by risks over or beyond the deterministic design basis and are quantified based on the extent to which they impact incremental changes to core damage frequency (CDF) and large early release frequency (LERF).
The risk analysis is described in detail in Attachment 3 of License Amendment Request for DTE Fermi Risk Informed ECCS Evaluation (SERCO-REP-DTE-22609-02 R1). At a high level, transported post-LOCA debris quantities from postulated pipe breaks at each Class 1 weld location were computed and compared to deterministic limits based on NEDO-32721P-A. In addition, fiber debris thickness accumulated on active strainers during each break scenario was compared to an assumed 1/8th-inch strainer failure criterion. CDF and LERF contributions for each break where deterministic limits or 1/8th-inch debris thickness were exceeded were summed and the total CDF and LERF from all break cases were compared to the risk region definitions in US NRC Reg Guide 1.174. The acceptance criterion of the analysis pertaining to newly identified failure modes and debris sources, arising from component identification tags/
labels and from Min-K insulation applied in containment penetrations was to remain within Region III of RG 1.174. Acceptance criteria for debris defined in Fermi Calculation DC-5979 Rev-B remain based on NEDO-32721P-A. Consequently, the conditions under which strainers are considered operable are when (1) the quantities of debris other than those attributable to component identification tags/labels and from Min-K insulation applied in containment penetrations are within limits defined in NEDO-32721P-A, and (2) the total CDF and LERF from all break cases, including the effects of component identification tags/labels and from Min-K insulation applied in containment penetrations are within Region III of RG 1.174. These two conditions comprise the acceptance criteria that would need to be satisfied if new potential debris is discovered.
The risk analysis and the amended ECCS suction strainer licensing basis apply only to use of a risk-informed methodology to address potential debris quantities arising from component identification tags/labels and from Min-K insulation applied in containment penetrations exceeding those currently evaluated in the deterministic licensing methodology. The risk analysis in the amendment applies to the present condition, i.e., the amendment is not intended to NRC-24-0027 Page 3 to allow accommodation of additional debris in the future. The risk analysis does not apply to the design debris loads, (defined in Fermi Calculation DC-5979 Rev-B) that are successfully addressed using the methodology defined in NEDO-32721P-A.
The scope of the amendment does not apply to any existing configuration controls, containment cleanliness programs, or containment inventory monitoring practices. These are not being modified or relaxed by the amendment. Existing controls include equipment labeling procedures and engineering review of all changes to containment insulation and other potential debris sources. Fermi does not anticipate any new procedural changes arising from the LAR. Fermi intends to observe all existing material controls with emphasis on labels, Min-K insulation, and fiber insulation.
DTE has demonstrated that the risk significance of potential additional debris loads posed by tags/labels and Min-K in containment penetrations is very-low as defined by Region III of RG 1.174 in terms of incremental change to core damage frequency (CDF) and incremental changes to large early release frequency (LERF). Regulatory Guide 1.174 permits consideration of accident scenario frequency in combination with accident scenario magnitude provided that the risk-informed analysis (1) meets current regulation, or is accompanied by requested exemptions; (2) is consistent with the defense-in-depth philosophy; (3) maintains adequate safety margin; (4) is small and consistent with the NRC Safety Goal policy; and (5) is monitored by performance measurement strategies that are discussed in Attachment 3 to the Fermi LAR.
Key aspects of the analysis supporting the amendment that cannot be changed without NRC prior approval are: (1) baseline risks must remain in RG 1.174 Risk Region 3 as computed using the approved assumption of 100 ft2 of total obstruction caused by overlapped miscellaneous debris (tags/labels) transported to the clean strainer area of all active strainers; (2) the definition of strainer success/failure accepted by NRC (i.e., 1/8th inch of fiber on any single strainer causing core damage); and (3) observance of all existing material controls with emphasis on labels, Min-K insulation, and fiber insulation.