ML19225D324

Attachment 2 - Callaway Energy Center Seismic Probabilistic Risk Assessment in Response to 10 CFR 50.54(1) Letter with Regard to NTTF 2.1: Seismic

ML19225D324
Person / Time
Site:	Callaway
Issue date:	08/12/2019
From:	Ameren Missouri
To:	Document Control Desk, Office of Nuclear Reactor Regulation
Shared Package
ML19225D321	List: ML19225D323 ML19225D324 ULNRC-06524, Response to March 12, 2012, Information Request, Seismic Probabilistic Risk Assessment for Recommendation 2.1
References
ULNRC-06524
Download: ML19225D324 (92)

Text

Attachment 2 to ULNRC-06524 Callaway Energy Center Seismic Probabilistic Risk Assessment in Response to 10 CFR 50.54(1) Letter with Regard to NTTF 2.1: Seismic

. .*._ r  : ..

L: ,

__t:__, . ... .

. (. 4

t,,* _., . - ..

L:..

• ::bi Page 1 of 92 to ULNRC-06524 Table of Contents 1 .0 Purpose and Objective 6 2.0 Information Provided in this Report 7 3 .0 CEC Seismic Hazard and Plant Response 11 3 1

. Seismic Hazard Analysis 11 3 .2 Seismic Hazard Analysis Methodology 11 3.3 Seismic Hazard Comparisons and Insights 20 3.3.1 Comparison ofHazard Curves from NTTF 2.1 Seismic Hazard Submittal and PRA Site Response Analysis 21 3 .3 .2 Probabilistic Seismic Hazard Analysis Technical Adequacy 21 3 .3 .3 Uncertainties in the Seismic Hazard Results from Input Parameters and Models 22 3.4 Horizontal and Vertical Response Spectra 22 3 .4.1 Derivation of Vertical Response Spectra 23 3 .4.2 Ground Motion Response Spectra at Elevation $40 FT 24 3.4.3 Foundation Input Response Spectra at Elevation 829 FT 26 3 .4.4 Foundation Input Response Spectra at Elevation 808.5 ft 28 3 .4.5 Response Spectra for the Alternate Emergency Power System 29 4.0 Determination of Seismic Fragilities for the S-PRA 32

4. 1 Seismic Equipment List 32

4. 1 1

. SEL Development 32

4. 1 .2 Relay Evaluation/Spurious Breaker Trip Evaluation 34 4.2 Walkdown Approach 34 4.2.1 Significant Walkdown Results and Insights 35 4.2.2 Seismic equipment List and Seismic Walkdowns Technical Adequacy 35 4.3 Dynamic Analysis of Structures 35 4.3.1 Fixed-baseAnalyses 35 4.3 .2 Soil Stmcture Interaction (551) Analyses 36 4.3.3 Structure Response Models 36 4.3.4 Seismic Structure Response Analysis Technical Adequacy 37 4.4 SSC Fragility Analysis 37 4.4. 1 SSC Screening Approach 39 Page 2 of 92 to ULNRC-06524 4.4.2 S$C Fragility Analysis Methodology 40 4.4.3 SSC Fragility Analysis Results and Insights 41 4.4.4 $SC Fragility Analysis Technical Adequacy 41 5.0 Plant Seismic Logic Model 42 5.1 Development ofthe $-PRA Plant Seismic Logic Model 42 5 .1 .1 General Approach 42 5 .1 .2 Initiating Events and Accident Sequences 42 5 .1 .3 Modeling of Correlated Components 43 5 .1 .4 Modeling of Human Actions 43 5.1 .5 Seismic LERF Model 44 5.2 5-PRA Plant Seismic Logic Model Technical Adequacy 45 5.3 Seismic Risk Quantification 45 5.3.1 5-PRA Quantification Methodology 45 5.3.2 5-PRA Model and Quantification Assumptions 46 5 .4 SCDF Results 49 5.5 SLERF Results 57 5.6 5-PRA Quantification Uncertainty Analysis 62 5.7 5-PRA Quantification Sensitivity Analysis 64 5 .7. 1 Truncation Limits for Model Convergence 64 5.7.2 ACUBE Post-Processing 68 5.7.3 HazardlntervalStudy 69 5.7.4 Non-Safety Component Fragility Sensitivity 69 5.7.5 Mission Time Sensitivity 69 5 .7.6 On-site FLEX Equipment Sensitivity 69 5.7.7 Model Sensitivity to Seismic HRA 70 5.7.8 Model Sensitivity to Seismic HRA Bin Definition 70 5.7.9 Model Sensitivity to Ex-Control Room Actions 70 5 .7. 10 Cumulative Sensitivity 70 5.7.11 Plant Modification Sensitivity 70 5.7.12 Summary of Sensitivity Study Results 70 5.8 5-PRA Quantification Technical Adequacy 71 Page 3 of 92 to ULNRC-06524 6.0 Conclusions 72 7.0 References 73 8.0 Acronyms 76 Appendix A - Summary of S-PRA Peer Review and Assessment ofPRA Technical Adequacy 78 Page 4 of 92 to ULNRC-06524 Executive Summary In response to 10 CFR 50.54(f) letter issued by the NRC on March 12, 2012, a seismic probabilistic risk assessment ($-PRA) was performed for Callaway Energy Center. The $-PRA effort included performing a probabilistic seismic hazard analysis (PSHA) to develop seismic hazard and response spectra at the plant using state-of-the-art seismic source models and attenuation equations; seismic response analysis of structures, fragility analysis of structures, systems and components (S$Cs); developing a logic model and performing risk quantification. The S-PRA effort underwent a final peer review by a team of experts. The comments ofthe reviewers were addressed and incorporated into the S-PRA as applicable.

The S-PRA identified risk-significant sequences and S$Cs with their risk ranking and showed that the point estimate seismic core damage frequency ($CDF) is 4.86E-06 per year, and the seismic large early release frequency (SLERF) is 3.41E-06 per year.

Sensitivity studies were performed to identify critical assumptions, test the sensitivity to quantification parameters and the seismic hazard, and identify potential areas to consider for further reducing seismic risk.

These sensitivity studies demonstrated that the model results were robust to the modeling and assumptions used.

Two plant actions are committed to: (1) modification ofthe AEPS transformer anchorage and (2) increasing the gap between the sprinkler head and fire protection insulation to improve fragility estimates based on the risk insights from the 5-PRA.

Page 5 of 92 to ULNRC-06524 1.0 Purpose and Objective Following the accident at the Fukushima Dai-ichi nuclear power plant resulting from the March 1 1, 2011, Great Thoku Earthquake and subsequent tsunami, the Nuclear Regulatory Commission (NRC) established the Near-Term Task Force (NTTf) tasked with conducting a systematic and methodical review of NRC processes and regulations and to determine if the agency should make additional improvements to its regulatory system. The NTTF developed a set of recommendations intended to clarify and strengthen the regulatory framework for protection against natural phenomena. Subsequently, the NRC issued a 50.54(f) letter on March 12, 2012 [1], requesting information to assure that these recommendations are addressed by all U.S. nuclear power plants. The 50.54(f) letter requests that licensees and holders of construction permits under 10 CFR Part 50 re-evaluate the seismic hazards at their sites against present-day NRC requirements and guidance.

A comparison between the re-evaluated seismic hazard and the design basis for Callaway Energy Center (CEC) has been performed, in accordance with the guidance in EPRI 1 025287, Screening Prioritization and Implementation Details (SPJD) For the Resolution of Fukushima Near Term Task Force Recommendation 2.1 : Seismic [101, and previously submitted to NRC [3]. That comparison concluded that the ground motion response spectrum (GMRS), which was developed based on the re-evaluated seismic hazard, exceeds the design basis seismic response spectrum in the 1 to 10 Hz range, and a seismic risk assessment is required. A seismic PRA (5-PRA) has been developed to perform the seismic risk assessment for CEC in response to the 50.54(f) letter, specifically item (8) in Enclosure 1 of the 50.54(f) letter.

This report describes the 5-PRA for Callaway Energy Center (CEC), and provides the information requested in item (8)(B) ofEnclosure 1 ofthe 50.54(f) letter [1J. The 5-PRA model has been peer reviewed (as described in Appendix A) and found to be of appropriate scope and technical capability for use in assessing the seismic risk for CEC, identifying which structures, systems, and components (SSCs) are important to seismic risk.

This report provides a summary regarding the 5-PRA as outlined in Section 2.0.

The level ofdetail provided in the report is intended to enable the NRC to understand the inputs and methods used, the evaluations performed and the decisions made because of the insights gained from the CEC 5-PRA.

Page 6 of 92

Attachment 2 to ULNRC-06524 2.0 Information Provided in this Report The following information is requested in the 50.54(f) letter [1], Enclosure 1, Requested Information Section, paragraph (8)(B), for plants performing a 5-PRA:

1 . The list of the significant contributors to seismic core damage frequency (SCDf) for each seismic acceleration bin, including importance measures (e.g., Risk Achievement Worth, Fussell-Vesely and Bimbaum)

2. A summary of the methodologies used to estimate the SCDF and seismic large early release frequency (SLERF), including the following:

i. Methodologies used to quantify the seismic fragilities of SSCs, together with key assumptions.

ii. SSc fragility values with reference to the method of seismic qualification, the dominant failure mode(s) and the source of information.

iii. Seismic fragility parameters.

iv. Important findings from plant walkdowns and any corrective actions taken.

v. Process used in the seismic plant response analysis and quantification, including the specific adaptations made in the internal events PRA to produce the 5-PRA model and their motivation.

vi. Assumptions about containment performance.

3 . Description of the process used to ensure that the 5-PRA is technically adequate, including the dates and findings of any peer reviews.

4. Identified plant-specific vulnerabilities and actions that are planned or taken.

Table 2-1 provides a cross-reference between the 50.54(f) reporting items noted above and the location in this report where the corresponding information is discussed.

The content ofthis report is organized as follows:

. Section 3 .0 provides information related to the CEC seismic hazard analysis.

. Section 4.0 provides information related to the determination of seismic fragilities for CEC SSCs included in the seismic plant response.

. Section 5 .0 provides information regarding the plant seismic response model (Seismic accident sequence model) and the quantification of results.

. Section 6.0 summarizes the results and conclusions ofthe 5-PRA, including identified plant seismic issues and actions taken or planned.

. Section 7.0 provides references.

. Section 8.0 provides a list of acronyms used.

. Appendix A provides an assessment of 5-PRA technical adequacy for response to NTTF 2.1 Seismic 50.54(f) Letter [1], including a summary ofthe CEC 5-PRA peer review.

The SPID defines the principal parts ofan 5-PRA, and the CEC 5-PRA has been developed and documented in accordance with the SPD. The main elements of the 5-PRA performed for CEC in response to the 50.54(f) Seismic letter correspond to those described in Section 6. 1 1 of the SPDJ, i.e.:

. Seismic hazard analysis Page 7 of 92 to ULN RC-06524

. Seismic structure response and SSC fragility analysis

. Systems/accident sequence (seismic plant response) analysis

. Risk quantification Table 2-2 provides a cross-reference between the reporting items noted in Section 6.8 of the SPID, other than those already listed in Table 2-1 , and provides the location in this report where the corresponding information is discussed.

Page 8 of 92 to ULNRC-06524 Table 2-1 : Cross-Reference for 50.54(f) Enclosure 1 S-PRA Reporting 50.54(1) Letter Reporting Item Description Location in this Report List of the significant contributors to $CDf for each 1 . . . . Section 5.0 seismic acceleration bin, including importance measures.

Summary of the methodologies 2 used to estimate SCDF and Sections 3.0, 4.0 and 5.0 SLERF.

Methodologies used to quantify 2(I) the seismic fragilities of SSCs, Section 4.0 together with key assumptions.

SSc fragility values with reference to the method of 2(u) seismic qualification, the Table 5-3 dominant failure mode(s), and the source of information.

Table 5-3 provides fragilities for

. . . . . . . the top risk significant SSCs 2(ui) Seismic fragility parameters.

based on standard importance measures such as F-V or RRW.

Important findings from plant 2(iv) walkdowns and any corrective Section 4.2 actions taken.

Process used in the seismic plant response analysis and quantification, including the 2(v) specific adaptations made in the Sections 5 1 and 5.3 internal events PRA model to produce the 5-PRA model and their motivation.

. Assumptions about containment 2(vi) Sections 4.3 and 5.5 performance.

Description of the process used to ensure that the 5-PRA is 3 technically adequate, including Appendix A the dates and findings ofany peer reviews.

Identified plant-specific 4 vulnerabilities and actions that Section 6.0 are planned or taken.

Page 9 of 92 to ULN RC-06524 Table 2-2 : Cross-Reference for Additional SPID Section 6.8 S-PRA Reporting spIn Section 6.8 Item (1) Description Location in this Report A report should be submitted to the NRC Entirety of the submittal addresses this.

summarizing the S-PRA inputs, methods and results.

The level of detail needed in the submittal should Entirety ofthe submittal addresses this. This report be sufficient to enable NRC to understand and identifies key methods of analysis and referenced determine the validity of all input data and codes and standards.

calculation models used The level of detail needed in the submittal should Entirety of the submittal addresses this. Results be sufficient to assess the sensitivity of the results sensitivities are discussed in the following to all key aspects ofthe analysis sections:

5.7 (S-PRA model sensitivities) 4.4 fragility screening (sensitivity)

The level of detail needed in the submittal should Entirety of the submittal report addresses this.

be sufficient to make necessary regulatory decisions as a part ofNTTf Phase 2 activities.

It is not necessary to submit all of the S-PRA Entire report addresses this. This report documentation for such an NRC review. Relevant summarizes important information from the S documentation should be cited in the submittal, and PRA, with detailed information in lower tier be available for NRC review in easily retrievable documentation.

form.

Documentation criteria for a $-PRA are identified This is an expectation relative to documentation of throughout the ASME/ANS Standard [4]. Utilities the S-PRA that the utility retains to support are expected to retain that documentation application of the S-PRA to risk-informed plant consistent with the Standard. decision-making.

Note (1): The items listed here do not include those designated in SPID Section 6.8 as guidance.

Page 10 of 92

Attachment 2 to ULNRC-06524 3.0 CEC Seismic Hazard and Plant Response This section provides summary site information and pertinent features including location and site characterization. The subsections provide brief summaries of the site hazard and plant response characterization.

The Callaway Energy Center (CEC), Unit 1 plant is a single Westinghouse 4-ioop pressurized water reactor located in central Missouri approximately 10 miles southeast of Fulton and 25 miles east-northeast of Jefferson City. The CEC Unit 1 plant used the Standardized Nuclear Unit Power Plant Systems standard design, including seismic design. The regional and site (local) geology is described in additional detail in the Probabilistic Seismic Hazard Analysis (PSHA) completed to support the CEC Unit 1 S-PRA [2]. The CEC Unit 1 site ground surface is at elevation (EL) 840 feet (fi), which represents the control point for development ofthe Ground Motion Response Spectra (GMRS).

The geologic column underlying the CEC Unit 1 site consists of 30.5 ft of engineered backfill overlying a thick sequence of sedimentary rocks. The foundation material and foundation elevation for the CEC Unit 1 plant structures are described in Table 3-1 The geotechnical profiles developed to support the PSHA are developed using the extensive borehole and geophysical data gathered at the CEC Unit 1 site, and at the nearby site investigated to support a second unit at Callaway [2].

Table 3-1 : Category 1 Structures and Geotechnical Foundation Material Geotechnical Foundation Applicable Category 1 Structure Material Elevation (ft)

Reactor Building, Diesel Generator Building, Engineered Backfill 829 Ultimate Heat Sink Cooling Tower Auxiliary/Control Building, Emergency Service Sdjt Rock 808.5 Water Pumphouse Alternate Emergency Power System (located Accretion-Gley and 7,5OOft NW ofnuclear island) 832 Glacial Till Soils 3.1 Seismic Hazard Analysis This section discusses the seismic hazard methodology, presents the final seismic hazard results used in the S-PRA and discusses assumptions and important sources of uncertainty.

The seismic hazard analysis determines the annual frequency of exceedance for selected ground motion parameters. The analysis involves use of earthquake source models, ground motion models, characterization ofsite response (e.g., soil and sedimentary rock column), and accounts for the uncertainties and randomness of these parameters to arrive at the site seismic hazard. The initial set of information regarding the CEC Unit 1 site hazard was provided to the Nuclear Regulatory Commission (NRC) in the seismic hazard information submitted in response to the NTTf 2. 1 Seismic information request [3]. As further discussed below, an updated PSHA and site response analysis (SRA) were performed for the CEC Unit 1 site [2].

3.2 Seismic Hazard Analysis Methodology The following method was used to perform the seismic hazard analysis.

The hard-rock PSHA is based on the Central and Eastern United States Seismic Source Characterization (CEUS-SSC) Project [4], and the Electric Power Research Institute (EPRI) Ground Motion Model (GMM)

Review Project [5]. Both the CEUS-SSC and the EPRI GMM Update Projects were executed according to the latest seismic hazard guidelines as published in [6] and [7]. The CEUS-SSC Project was a Senior Seismic Hazard Analysis Committee (SSHAC) Level 3 project consistent with the guidance from [6] and

[7]. The EPRI GMM Update Project was a SSHAC Level 2 update of a previous SSHAC Level 3 GMM.

Page 11 of92

Attachment 2 to ULNRC-06524 Both projects considered available data, models and methods proposed by the larger technical community that were relevant to the hazard analysis. Both projects represent the center, body and range of technically defensible interpretations informed by the assessment of existing data, models and methods.

The uncertainties in source geometry, recurrence parameters, maximum magnitude and ground motion prediction equations are propagated in the hazard estimates through the logic tree formalism [2]. A review of published references and updated earthquake catalogs since the CEU$-SSC and EPRI GMM models were published found that those models continue to be appropriate for P$HA [21. The CEUS-$SC Project included the development and use ofa comprehensive earthquake catalog through 2008. An assessment of seismicity since 2008 determined that recurrence rates and maximum magnitudes from the CEUS-SSC model continue to be appropriate [2]. An evaluation of induced seismicity determined that the seismic hazard at the CEC Unit 1 site is not impacted by suh events [2]. All credible seismic sources that may contribute significantly to the frequency of exceedance of vibratory ground motion at the CEC Unit 1 site are taken into account. The PSHA accounts for: (1) CEUS-$SC distributed seismicity source zones out to a distance of at least 640 kilometers (km) and (2) Repeated Large Magnitude Earthquake seismic sources within or near 1,000 km [2].

The effect of geologic deposits and geotechnical properties on ground motions at the CEC Unit 1 site are accounted for by performing a site-specific SRA. The site profiles used for the SRA were based on an updated review of available geologic and geophysical data characterized at the CEC Unit 1 site and the nearby site investigated for a potential second unit. Given these two sets of data, the CEC Unit 1 site is considered well characterized. The site profiles used for the $RA represent an update to the site profiles used for the NTTF 2. 1 submittal [2] and [3] In the NTTF 2. 1 submittal [3] the upper 3O feet of soil were modeled as glacial and post-glacial deposits. However, these materials were removed under the Nuclear Island at the CEC Unit 1 site and replaced by engineered backfill.

Thus, the Nuclear Island at the CEC Unit 1 site has 30.5 ft of engineered backfill and approximately 2,400 ft of sedimentary rocks over hard rock. Relatively flat-lying strata underlie the CEC Unit 1 site; therefore, a one-dimensional site response analysis was determined to be appropriate. Guidance for performing $RA is provided by [8], [9] and [10]; Reference [1 0] is referred to as the Screening, Prioritization and

• Implementation Details (SPID). The general approach outlined in the SPDJ is to define the uncertainty and variability in key site response input parameters as part of deriving the site amplification factor (AF) and

.. -.--- its variability. Appendix B ofthe SPUJ provides direction for the assessment ofboth epistemic and aleatory .

unceiainties in key site response input parameters [10].

The SRA uses a logic tree approach to represent epistemic uncertainty in the shear-wave velocity (VS) profile, the geologic layer dynamic properties, the ground motion inputs and the upper crustal damping associated with the parameter kappa [2]. Three (3) base-case VS profiles were used for the SRA. The VS values, layer elevations and depths are shown in Table 3-2 and the VS profiles are shown on Figure 3-1 for the upper 350 fi. The SPA included two sets of non-linear dynamic properties modeled two ways accounting for non-linear and linear behavior [2] For each base-case profile and dynamic property model, kappa values were quantified [2] to ensure that the range of kappa is consistent with the guidance from the SPID [10].

Page 12 of 92

Attachment 2 to ULN RC-06524 Table 3-2: Base-Case Median V Profiles Used for the CEC Unit 1 Site Response Analysis Profile P1 Profile P2 Profile P3 Top of Layer Depth To Depth To Depth To Elevation Vs Vs Vs 13 I3 (ft) (fps) (fps)1 (fps) 840 1350 0 1100 0 1350 0 809.5 2337 30.5 2337 30.5 2337 30.5 781.7 4052 58.3 4052 58.3 4052 58.3 777.6 6045 62.4 6045 62.4 6045 62.4 767.2 3714 72.8 3714 72.8 3714 72.8 747.6 8358 92.4 8358 92.4 8358 92.4 707.8 6950 132.2 6950 132.2 6950 132.2 675 8319 165.0 7234 165.0 HardRock 165.0 HardRock 2168.5 HardRock 2168.5 Notes: (1) fps is feet per second.

$heit Wave Veocity(ft/sec) 0 1 OX c 4 smo £000 1 O 9 IXc C

so

-I Jt

J i

4. .,

L° Figure 3-1: Base-Case Median Vs Profiles Down to a Depth of350 ft Used for the CEC Unit 1 Site Response Analysis Page 13 of 92 to ULNRC-06524 Consistent with guidance from the SPDJ [10] the SRA is based on a random vibration theory approach using input ground motions that span a wide range of input amplitudes. The input ground motions used for the SPA were based on scaling three (3) deaggregation earthquakes (represented by magnitude and distance) defined at each of 1 1 mean annual frequencies of exceedance, for each of the seven spectral frequencies assessed as part of the hard rock PSHA. Consistent with the $PID [1 0], epistemic uncertainty in source spectra was modeled as either a single-corner or double-corner spectral shape. Appropriate weights are applied to each branch of the SPA logic tree.

Because the various structures at the CEC Unit 1 site are embedded to different depths, it is important that the SPA provide ground motions that are appropriate for use in S$I analysis and account for the potential influence ofsoil confinement above the foundation elevations. Guidance provided by [9] and [1 1] account for these conditions.

The SRA for shallow-embedment structures (EL 829 fi) was based on the full soil column (EL 840 fi),

where the full set of strain-iterated properties is retained for each of the layers modeled. The soil column is then truncated at the foundation elevation (829 fi), and the SRA is repeated, using the strain-iterated properties from the full column; no further strain iteration is permitted. The AFs from the truncated column are used to derive the foundation Input Response Spectra (FIRS) at EL 829 fi. In the PSHA report [2]

these are designated as FIRS-i In essence, the shallow-embedment structures are analyzed as surface structures.

The $RA for deeper embedded structures (EL 808.5 fi) is also based on the full soil column (EL 840 11),

where soil column outcrop motions at the foundation elevation (808.5 fi) are used to derive the FIRS. In the PSHA report [2] these are designated as FWS-2. In essence, the deeper embedded structures are analyzed as embedded structures.

The results of the SRA consist of AFs that describe the amplification (or de-amplification) of reference rock motion as a function of spectral frequency and input reference rock ground motion amplitude. The AFs are represented in terms of a mean amplification value and an associated standard deviation for each spectral frequency and input rock ground-motion amplitude. For subsequent use in deriving hazard curves the AF values are retained at each end branch of the SRA logic tree. Consistent with the SPID [9], a minimum AF of 0.5 was employed when combining the SRA results with the hard-rock reference hazard results.

Figure 3-2 shows the AF results at EL 840 ft considering the overall weighted AF given the weights assigned to each branch ofthe SRA logic tree. Figure 3-3 shows similar results at EL 829 ft and Figure 3-4 shows the results at EL 808.5 ft. Several AF sensitivity comparisons are shown in the PSHA report [2]

including the impact of each base-case profile, each of the two ground motion point-source models, and each of base-case sets of dynamic property curves. In general, the AF is most sensitive to the base-case profile, with profile P3 being different than P1 and P2 because the overall profile thickness is significantly less for P3 compared to either P1 or P2. A SRA sensitivity assessment was completed to address a peer review Fact & Observation in which additional Vs epistemic uncertainty was added to profiles P2 and P3 between depths of 30.5 and 1 65 ft. This sensitivity assessment demonstrated that the AF distribution is not impacted by adding in additional median Vs epistemic uncertainty over this depth range.

The site response results are combined with the hard-rock PSHA results to obtain hazard curves at elevations of interest that are consistent with the annual frequencies of exceedance of the hard-rock hazard curves [2]. The results are combined using Approach 3 of NUREG/CR-6728 [17], consistent with the methodology described in the SPID [10]. The combination of the hard-rock hazard results with the amplification factors is performed on a hazard curve by hazard curve basis for each individual seismic source and median ground motion model branch. Given thecomplexity ofthe logic tree used for the SRA, a review ofthe AFs was completed and ultimately two groups ofAFs were selected to derive hazard curves at the surface. A hazard curve fractile sensitivity study was performed to assess the impact of the AF grouping process, in which a portion ofthe epistemic uncertainty is transferred to aleatory uncertainty. The Page 14 of 92 to ULN RC-06524 sensitivity study showed that the AF grouping approach has essentially no impact on the mean hazard or on any of the hazard fractiles above the mean for all levels of ground motion. Also, there is no impact on fractiles below the mean hazard at ground motions less than approximately ig [2J.

t

. . -= .

Page 15 of 92

Attachment 2 to ULNRC-06524 rM

--- M Sv I 0061 M

1SA on

,q%,.(yfH7 f#.q*ytHJ 10

- .L%

_t_:::::i 3061 K

___p%

• S.86 Figure 3-2: CEC Unit 1 Site Total Amplification Factors EL $40 ft (Note: Quantities in the upper-right-hand corner represent the hard-rock input 100 Hz SA in gs.)

Page 16 of 92 to ULNRC-06524 0.06 E_M 3

[::-

q.yH?]

4 M( 0.32 F.q.yfHJ M-.*

4 c\ 0.53 M 1

M -

4 1.38

.: M.

,1\(

• 3.06 Vq;.y (1*]

3;

+dj ,7\

5.84 Figure 3-3: CEC Unit 1 Site Total Amplification Factors EL 829 ft (Note: Quantities in the upper-right-hand corner represent the hard-rock input 100 Hz SA in gs.)

Page 17 of 92 to ULNRC-06524 i: 0.06 TM 0.11 I

F.q.:y(H

[: j 019

[: 032 I

090 H Sd / \

2.15 1 I F(.q.(YtH fHJ L-: i 306

432 I

I Fq**ytHi M

584 z

Figure 3-4: CEC Unit 1 Site Total Amplification Factors EL $08.5 ft (Note: Quantities in the upper-right-hand corner represent the hard-rock input 100 Hz SA in gs.)

Page 18 of 92 to ULNRC-06524 The 1 00 Hz spectral acceleration (assumed to be peak ground acceleration) is the ground motion parameter used for risk quantification in the Seismic PRA. The 100 Hz spectral acceleration hazard curves (mean and fractiles) for EL $40 ft. are displayed in Table 3-3 and on Figure 3-5 The mean hazard curves for each of the seven spectral frequencies associated with the ground motion model are displayed on Figure 3-6.

Table 3-3: 100-HZ SA Mean and Fractile Hazard Curves for the CEC Unit 1 Site at Ground Surface EL 840 ft Spectral Annual Frequency of Exceedance Acceleration [gJ MEAN 5TH 15TH 50TH 85TH 95TH 0.01 2.60E-02 9.33E-03 . l.70E-02 2.53E-02 3.40E-02 5.05E-02 0.02 9.43E-03 3.40E-03 4.96E-03 8.l4E-03 1.23E-02 2.52E-02 0.03 5.80E-03 1 66E-03 2.60E-03 4.76E-03 7.87E-03 1 .7 1E-02 0.04 4.15E-03 9.33E-04 l.58E-03 3.23E-03 6.l2E-03 l.23E-02 0.05 3 1 1E-03

. 5.77E-04 9.6 1E-04 2.3 1E-03 4.77E-03 9.47E-03 0.06 2.43E-03 3 .69E-04 6.50E-04 1 .69E-03 3 .92E-03 7.80E-03 0.07 1.95E-03 2.60E-04 4.52E-04 1.24E-03 3.23E-03 6.73E-03 0.0$ l.59E-03 1.73E-04 3.36E-04 9.25E-04 2.70E-03 5.81E-03 0.09 l.32E-03 1.27E-04 2.54E-04 7.20E-04 2.27E-03 4.95E-03 0.1 1.11E-03 9.06E-05 1.85E-04 5.68E-04 1.94E-03 4.35E-03

0. 1 5 5 .43E-04 3 .02E-05 6.27E-05 2. 1 6E-04 8 .50E-04 2.45E-03 0.2 3 .08E-04 1 .43E-05 3 .0 1 E-05 1 .08E-04 4.3 8E-04 1 .47E-03 0.25 1.92E-04 7.90E-06 1.78E-05 6.45E-05 2.58E-04 9.06E-04 0.3 1 .28E-04 4.96E-06 1 1 3E-05

. 4. 14E-05 1 .54E-04 6. 1 2E-04 0.35 8.84E-05 3.55E-06 7.62E-06 3.O1E-05 1.03E-04 4.22E-04 0.4 6.32E-05 2.48E-06 5.43E-06 2.14E-05 7.1OE-05 2.94E-04 0.5 3.49E-05 1.42E-06 3.12E-06 1.23E-05 4.09E-05 1.53E-04 0.6 2.09E-05 7.99E-07 1.96E-06 7.46E-06 2.60E-05 8.53E-05 0.7 1.33E-05 5.05E-07 1.31E-06 4.8$E-06 l.78E-05 5.33E-05 0.8 8.85E-06 3.48E-07 8.41E-07 3.49E-06 1.25E-05 3.67E-05 0.9 6. 12E-06 2.36E-07 5.9 1 E-07 2.40E-06 8.66E-06 2.44E-05 1 4.36E-06 1.62E-07 4.20E-07 1.76E-06 6.54E-06 1.73E-05 2 3.63E-07 7.20E-09 2.46E-08 1.32E-07 5.64E-07 1.49E-06 3 6.16E-08 6.14E-10 2.57E-09 1.77E-0$ 8.98E-08 2.64E-07 5 5.68E-09 1.59E-11 8.21E-11 9.33E-10 7.32E-09 2.63E-08 6 2.64E-09 4.42E-12 2.61E-11 3.53E-l0 3.27E-09 1.30E-08 7 1 .4 1E-09 1 .46E- 12 8.86E- 12 1 .46E- 1 0 1 .63E-09 6.8 1 E-09 8 8.24E-10 5.44E-13 3.90E-l2 6.66E-11 8.59E-10 4.l3E-09 9 5.13E-10 2.29E-13 1.69E-12 3.46E-ll 5.O1E-l0 2.55E-09 10 3.36E-10 9.54E-14 7.84E-13 1.77E-11 3.12E-lO l.77E-09 Page 19 of 92 to ULN RC-06524 1E-O2 Mean S

5th tE-03 15th 0

0 ,.*.

50th

+BSth 1EO4 U

• 95th 03 0

1E-O5 C

0 I 1O7 I.E-OS 001 0.1 10 Acceteraton Figure 3-5: 100-HZ SA Mean and Fractile Hazard Curves for the CEC Unit 1 Site at Ground Surface EL $40 ft 1.0+00 1.E-O1 03F1z =1.0Hz

• 2.5Hz 1.E-02

- 1E-03

- 1.E-04 1.E-05 1.E-06 1.E-07 i.E-OS 0.01 0.1 1 10 Acceleration [g]

Figure 3-6: Mean ilazard Curves for the CEC Unit 1 Site at Ground Surface EL $40 ft 3.3 Seismic Hazard Comparisons and Insights This section compares the surface control point hazard curves developed to support the NTTf 2. 1 Seismic Hazard submittal [3] to that developed to support the 5-PRA [2], and provides a briefsummary ofthe PSHA technical adequacy and key uncertainties.

Page 20 of 92

Attachment 2 to ULN RC-06524 3.3.1 Comparison ofHazard Curves from NTTF 2.1 Seismic Hazard Submittal and PRA Site Response Analysis As noted above, the site profiles used for the $RA were based on an updated review ofthe available geologic and geophysical data characterized at the CEC Unit 1 site. In the NTTF 2. 1 submittal [2] the upper 3O ft of soil was modeled as glacial and post glacial deposits. However, these materials were removed at the CEC Unit 1 site and replaced by engineered backfill which represents more competent material (increased stiffness and density). Because the engineered backfill is stiffer than the natural soils, the general trend was to result in decreases in the median AFs (lower impedance relative to the natural soils) which result in a decrease in the surface control point hazard curves.

Figure 3-7 compares the mean surface control point seismic hazard curves for spectral frequencies of 100.0, 10.0 and 1 .0 Hz from the NTTF 2. 1 Seismic Hazard submittal [3], assessed by the NRC [121, with the results from the Seismic PRA [2]. The relative decrease between the NTTF 2. 1 submittal and the Seismic PRA is the result oflower impedance between the upper 30.5 ft of geologic material when compared to the underlying thick sequence of sedimentary rock. The median VS values for the engineered backfill, based on results from test pads of the backfill, are larger than for the natural soils that were removed at the site; median VS values of 1, 1 00 to 1 ,350 feet per second (fps) for the engineered backfill [2] compared to median VS values of 400 to 625 fps for the upper 3 .9 ft and 836 to 1307 fs for the remaining 26.2 fi of natural soils [3].

LE+OO

. SxfaceMei1Kz5PR5

-- -5ifacMen10ftz5ffiA 1.E-O1 SLafacMean1O0f65PRA Nfl1:L1Mealf%

1E-02 -NWL1Mean1OHz

-NTtL1MeWW1z 11-03 11-04 11-05 I

11-06 16-07 0.01 0.1 10 Acceleration [gJ Figure 3-7: Mean Hazard Curves for the CEC NIT 1 Site at Ground Surface EL $40 ft 3.3.2 Probabilistic Seismic Hazard Analysis Technical Adequacy The CEC S-PRA seismic hazard methodology and analysis was subject to an independent peer review against the pertinent requirements in the ASME/ANS PRA Standard [13]. The 5-PRA was peer reviewed relative to Capability Category II for the full set of requirements in the Standard. After completion of the subsequent independent assessment, the full set of supporting requirements was met. The seismic hazard analysis was determined to be acceptable for use in the $-PRA. The peer review assessment, and subsequent disposition of peer review Facts and Observations (F&O) through an independent assessment, is further described in Appendix A and References [14] and [15].

Page2l of92 to ULN RC-06524 3.3.3 Uncertainties in the Seismic Hazard Results from Input Parameters and Models The PSHA results [2] were reviewed to identify and understand the sources of uncertainties and related assumptions that are important. Hazard assessment and sensitivity studies document the relative contribution to the total hazard by seismic source, the deaggregation of hazard by magnitude and distance, the impact on AF distribution for the $RA input uncertainties and the sensitivity to adding in additional epistemic uncertainty to the median Vs for the SRA site profiles.

The P$HA [21 includes an assessment ofthe hazard sensitivity to epistemic uncertainty in particular PSHA input variables (i.e., GMPE, seismicity of distributed sources, maximum magnitude of distributed sources, etc.), which is measured by the variance in the hazard contributed solely from epistemic uncertainty in a specific input variable, normalized by the variance in the total hazard. The results ofthis process are shown on Figure 3-8 which displays the variance deaggregation for the spectral frequency of 100 Hz (PGA) at the surface control point (EL 840 fi). Consistent with experience from several other P$HAs, the dominant contributor to the total variance is the epistemic uncertainty in the ground motion model. As the mean annual frequency of exceedance gets lower, the epistemic uncertainty in both maximum magnitude and the three magnitude-range cases used for deriving recurrence rates become more significant.

80%

I 00 Hz Hazard MAFE=1 11E-3 I 70% sMAFE= I 280-4 MAFE= I 33E5 sMAFE = 436E-6 60%

1. MAPE = 3.637 MAPE = 5.68E-9 50%

C 0

40%

C U

30%

20%

10%

0%

GM Cluster and Distrtbuted 3 Cases tar 2 Reaiza6ons Site Oistibuted 0stributed Mean 4od4 Source Mniax Distributed tot Distributed Amphtico%on Sc4Jtce Soutce Model Seismic6t Sesmcity Stanch Seismogenic Depth Figure 3-2: Variance Deaggregation ofthe CEC Unit 1 Site PSHA Logic Tree Inputs for the Spectral Frequency of 100 HZ 3.4 Horizontal and Vertical Response Spectra This section provides the horizontal and vertical GMRS at the surface control point (EL 840 ft), the horizontal and vertical FIRS at foundation elevations 829 ft and 808.5 ft, and the horizontal response spectra at the location ofthe Alternate Emergency Power System (7,5OOft NW of CEC Nuclear Island).

Page 22 of 92 to ULNRC-06524 3.4.1 Derivation of Vertical Response Spectra The vertical response spectra were developed based on the corresponding horizontal response spectra, by scaling with appropriate vertical-to-horizontal (V/H) ratios. The derivation of V/H ratios follows the guidance developed by [16] and [1 7], in which generic V/H ratios are developed that can be used at nuclear power plants in the CEU$. The generic V/H ratios are adjusted to account for the CEC Unit 1 site conditions and the site-specific horizontal GMRS and FLRS. Two models were used. The first V/H model is based on interpolation between the V/H ratios for Western United States and CEUS rock site conditions from [17].

The second set of V/H values is based on empirical ground motion models. Consistent with the guidance from [16], the empirical relations from [18] and [19] were considered appropriate. The two models were used to derive a mean V/H ratio to establish a mean vertical GMRS and FIRS. The final shape and values of the recommended V/H ratios for the three elevations are consistent with the mean V/H models given by

[1 6]. The mean V/H ratios are shown in Table 3-4.

Table 3-4: Recommended Mean V/H Ratios V/H Ratio V/H Ratio V/H Ratio V/H Ratio V/H Ratio V/H Ratio Frequency Frequency GMRS FIRS-i FIRS-2 GMRS FIRS-i FIRS-2 (Hz) z)

EL $40 ft EL 829 ft EL 808.5 ft EL 840 ft EL 829 ft EL 808.5 ft 0.10 0.694 0.697 0.697 3.56 0.694 0.697 0.697 0.13 0.694 0.697 0.697 4.52 0.694 0.697 0.697 0.16 0.694 0.697 0.697 5.00 0.694 0.697 0.697 0.20 0.694 0.697 0.697 5.74 0.694 0.697 0.697 0.26 0.694 0.697 0.697 7.28 0.694 0.697 0.697 0.33 0.694 0.697 0.697 9.24 0.694 0.697 0.697 0.42 0.694 0.697 0.697 10.00 0.694 0.697 0.697 0.50 0.694 0.697 0.697 11.72 0.694 0.697 0.697 0.53 0.694 0.697 0.697 14.87 0.694 0.697 0.697 0.67 0.694 0.697 0.697 18.87 0.694 0.697 0.697 0.85 0.694 0.697 0.697 23.95 0.735 0.729 0.733 1.00 0.694 0.697 0.697 25.00 0.747 0.741 0.744 1.08 0.694 0.697 0.697 30.39 0.807 0.799 0.798 1.37 0.694 0.697 0.697 38.57 0.870 0.860 0.853 1.74 0.694 0.697 0.697 48.94 0.899 0.889 0.880 2.21 0.694 0.697 0.697 62.10 0.895 0.887 0.883 2.50 0.694 0.697 0.697 78.80 0.858 0.852 0.855 2.81 0.694 0.697 0.697 100.00 0.806 0.804 0.813 Page 23 of 92 to ULNRC-06524 3.4.2 Ground Motion Response Spectra at Elevation $40 FT The horizontal and vertical GMR$ at EL 840 fi are displayed on Figure 3-9 with the spectral acceleration values shown in Table 3-5.

12 Horizonta1 GMRS I .0 Vertical GMRS C

0 0.8 a;

a; 0 0.6 0

0 a;

0.4 0

0%

0.0 0.1 1 10 100 Frequency (Hz)

Figure 3-9: CEC Unit 1 Site Vertical and Horizontal GMRS at EL 840 ft

- .-.-==

Page 24 of 92 to ULNRC-06524 Table 3-5: CEC Unit 1 Site Vertical and Horizontal GMRS at EL $40 ft Horizontal Vertical Horizontal Vertical frequency frequency GMRS GMRS GMRS GMRS

[UzJ . . [HzJ .

(SA 111 g) (SA 111 g) (SA in g) (SA in g) 0.10 0.0058 0.0040 5.00 0.8716 0.6045 0.13 0.0086 0.0060 5.74 1.0303 0.7146 0.16 0.0126 0.0087 6.12 1.0716 0.7433 0.20 0.0182 0.0127 6.51 1.0843 0.7521 0.26 0.0265 0.0183 6.89 1.0759 0.7462 0.33 0.0385 0.0267 7.28 1.0661 0.7395 0.42 0.0563 0.0391 9.24 0.8770 0.6083 0.50 0.0754 0.0523 10.00 0.8236 0.5712 0.53 0.0782 0.0542 11.72 0.7774 0.5392 0.67 0.0901 0.0625 14.87 0.7992 0.5543 0.85 0.1042 0.0723 18.87 0.8246 0.5719 1.00 0.1115 0.0774 23.95 0.8007 0.5884 1.08 0.1192 0.0827 25.00 0.7947 0.5940 1.37 0.1402 0.0972 30.39 0.7766 0.6269 1.74 0.1623 0.1125 38.57 0.7081 0.6157 2.21 0.2173 0.1507 48.94 0.6368 0.5723 2.50 0.2530 0.1755 62.10 0.5449 0.4875 2.81 0.2912 0.2020 78.80 0.4365 0.3743 3.56 0.4305 0.2986 100.00 0.3899 0.3144 4.52 0.7234 0.5017 Page25 of 92 to ULN RC-06524 3.4.3 foundation Input Response Spectra at Elevation 829 fT The horizontal and vertical FIRS-i at EL 829 fi are displayed on Figure 3-10 with the spectral acceleration values shown in Table 3-6.

I .4 I .2 to 0

, 08 G) 0 0

0.2 0.0 01 1 10 100 Frequency (Hz) figure 3-10: CEC Unit Site Vertical and Horizontal fIRS-i at EL $29 ft Page 26 of 92 to ULNRC-06524 Table 3-6: CEC Unit 1 Site Vertical and Horizontal FIRS-i at EL $29 ft Horizontal Vertical firs- Horizontal Vertical firs-Frequency frequency firs-i i firs-i i

[Hz] . . [Hzl *

(SA in g) (SA in g) (SA in g) (SA in g) 0.10 0.0058 0.0040 3.56 0.3164 0.2205 0.13 0.0085 0.0059 4.52 0.4589 0.3198 0.16 0.0125 0.0087 5.00 0.5482 0.3820 0.20 .

0.0181 0.0126 5.74 ..

0.7219 0.5030 0.26 0.0262 0.0183 7.28 1.0889 0.7588 0.33 0.0381 0.0266 9.24 1.2163 0.8476 0.42 0.0557 0.0388 10.00 1.1806 0.8227 0.50 0.0745 0.0519 11.72 1.0526 0.7335 0.53 0.0772 0.0538 14.87 0.9278 0.6465 0.67 0.0889 0.0620 18.87 0.9181 0.6398 0.85 0.1026 0.0715 23.95 0.8565 0.6242 1.00 0.1095 0.0763 25.00 0.8396 0.6221 1.08 0.1167 0.0813 30.39 0.7993 0.6390 1.37 0.1358 0.0947 38.57 0.7466 0.6420 1.74 0.1543 0.1075 48.94 0.6547 0.5818 2.21 0.1977 0.1378 62.10 0.5498 0.4874 2.50 0.2246 0.1565 78.80 0.4428 0.3774 2.81 0.2501 0.1743 100.00 0.3969 0.3190 Page 27 of 92 to ULN RC-06524 3.4.4 Foundation Input Response Spectra at Elevation 808.5 ft The horizontal and vertical FIRS-2 at EL 808.5 fi are displayed on Figure 3-1 1 with the spectral acceleration values shown in Table 3-7.

0.9 0.8 0.7 a)

C 0 0.6 I-0 0 0.5 0

0 0.4 0

0 0 0.3 a

0 02 0.1 0.0 0.1 1 10 100 Frequency (Hz)

Figure 3-11: CEC Unit Site 1 Vertical and Horizontal FIRS-2 at EL 808.5 ft .

Page 28 of 92

Attachment 2 to ULNRC-06524 Table 3-7: CEC Unit 1 Site Vertical and Horizontal FIRS-i at EL 808.5 ft Horizontal Vertical Horizontal Vertical Frequency Frequency FIRS-2 FIRS..2 FIRS.2 FIRS2

[Hz] jHzJ (SA in g) (SA fl g) (SA in g) (SA Ifl g) 0.10 0.0058 0.0040 3.56 0.3262 0.2273 0.13 0.0085 0.0059 4.52 0.4834 0.3369 0.16 0.0124 0.0087 5.00 0.5601 0.3903 0.20 0.0181 0.0126 5.74 0.6620 0.4613 0.26 0.0262 0.0183

• 7.28 0.7192 0.5012 0.33 0.0321 0.0266 9.24 0.6390 0.4453 0.42 0.0557 0.0388 10.00 0.6356 0.4429 0.50 0.0745 0.0519 11.72 0.6763 0.4713 0.53 0.0772 0.0538 14.87 0.7794 0.5431 0.67 0.0890 0.0620 18.87 0.8077 0.5628 0.85 0.1027 0.0716 23.95 0.7390 0.5419 1.00 0.1094 0.0763 25.00 0.7293 0.5429 1.08 0.1167 0.0813 30.39 0.7050 0.5625 1.37 0.1359 0.0947 38.57 0.6384 0.5444 1.74 0.1549 0.1079 48.94 0.5460 0.4807 2.21 0.1989 0.1386 62.10 0.4582 0.4044 2.50 0.2264 0.1578 78.80 0.3628 0.3103 2.81 0.2524 0.1759 100.00 0.3185 0.2590 3.4.5 Response Spectra for the Alternate Emergency Power System A horizontal GMR$ was developed for the AEPS location at the CEC site. The AEP$ is located about 7,500 ft northwest of the CEC Unit 1 Nuclear Island, and the AEPS foundation slab is at EL 832 ft, about 8 ft lower than the surface control point (EL 840 fi). The AEP$ foundation overlies natural soils.

The AEPS GMRS was developed by deriving scale factors that represent adjustments needed to derive the AEPS GMRS relative to the surface control point (Nuclear Island) GMRS. These adjustments were derived by: (1) reviewing available geologic and geoteclmical information, (2) developing a model for the subsurface below the AEP$ location and (3) completing simplified site response which is used to derive the relative AF differences between the AEPS location and the Nuclear Island. Because the relative adjustments were based on simplified site response, the resulting AEPS GMRS was smoothed to account for site response input model uncertainties which were explicitly modeled when deriving the GMR$ for the Nuclear Island surface control point. The AEPS smoothed horizontal GMRS is shown on Figure 3-8 with the spectral acceleration values shown in Table 3-8. For comparison, Table 3-8 lists the spectral acceleration values for the surface control point (EL 840 ft) and the unsmoothed and smoothed GMRS at the AEPS location (EL 832 ft).

Page 29 of 92 to ULNRC-06524 AEPS Horizontal GMRSJ SOS C

0 cc 0

0 0

0

< 0.4 0

0 a 0.3 .

0 0.2

+-

0,1 00 .,

01 1 10 100 Frequency (Hz)

Figure 3-12: Horizontal GMRS for the AEPS Location at the CEC Unit 1 Site Page 30 of 92 to ULNRC-06524 Table 3-8: Horizontal GMRS at the CEC Unit 1 Site and at the AEPS Location Ground Motion Response Spectra (g)

Frequency Nuclear Island AEPS AEPS (Hz) Surface Control Point EL 832 ft EL 832 ft EL 840 ft Unsmoothed Smoothed 0.1000 0.0058 0.0063 0.0063 0.1269 0.0086 0.0092 0.0092 0.1610 0.0126 0.0133 0.0133 0.2043 0.0182 0.0190 0.0190 0.2593

• 0.0265 0.0274 0.0274 0.3290 0.0385 0.0397 0.0397 0.4175 0.0563 0.0581 0.0581 0.5000 0.0754 0.0782 0.0782 0.5298 0.0782 0.0813 0.0813 0.6723 0.0901 0.0951 0.0951 0.8532 0.1042 0.1128 0.1128 1.0000 0.1115 0.1240 0.1240 1.0826 0.1192 0.1348 0.1348 1.3738 0.1402 0.1710 0.1710 1.7433 0.1623 0.2266 0.2266 2.2122 0.2173 0.3864 0.3864 2.5000 0.2530 0.5407 0.5407 2.8072 0.2912 0.7615 0.7615 3.5622 0.4305 1.0515 0.9200 4.5204 0.7234 0.6886 0.9200 5.0000 0.8716 0.5870 0.8800 5.7362 . 1.0303 0.4512 0.8500 6.1219 - 1.0716 0.4187 0.8500 6 5076 1 0843 0 4286 0 8500 6.8933  ;=z 1.0759 0.4617 0.8500 7.2790 1.0661 0.5162 0.8500 z+/-:

9.2367 0.8770 0.8798 0.8500 0.8236 0.8877 0.8500 11.7210 0.7774 0.7712 0.7712 14.8735 0.7992 0.7101 0.7101 18.8739 0.8246 0.7040 0.6800 23.9503 0.8007 0.5652 0.6500 25.0000 0.7947 0.5243 0.6400 30.3920 0.7766 0.5272 0.6000 38.5662 0.7081 0.5604 0.5604 48.9390 0.6368 0.5363 0.5363 62.1017 0.5449 0.4749 0.4749 78.8046 0.4365 0.3892 0.3892 100.0000 0.3899 0.3529 0.3529 Page3l of92 to ULN RC-06524 4.0 Determination of Seismic Fragilities for the $-PRA This section provides a summary of the process for identifying and developing fragilities for S$C that participate in the plant response to a seismic event for the Callaway Energy Center $-PRA. The subsections provide brief summaries of these elements.

4.1 Seismic Equipment List A seismic equipment list ($EL) was developed that includes those $$C whose seismic-induced failure could either give rise to an initiating event or degrade capability to mitigate a seismically-induced initiating event.

The $EL was developed for the end states of core damage and large early release. The methodology used to, develop the SEL is generally consistent with the guidance provided in the EPRI S-PRA Implementation Guide [27].

4.1.1 SEL Development The $EL includes all plant components and structures whose seismic-induced failure could either give rise to an initiating event or degrade capability to mitigate a seismic-induced initiating event. A preliminary SEL is developed based on seismic-relevant portions of the IE PRA model. This preliminarily SEL is supplemented by a series of reviews intended to identify seismically risk-significant components not modeled by the IE PRA.

The steps followed in developing the full $EL are as follows:

. The internal event logic model was reviewed to identify all physical components associated with the modeled basic events. Equipment that is captured through rule-of-the-box considerations, e.g., equipment contained on a skid or in a cabinet, that can be subsumed into the major skid equipment or into the cabinet, were identified. For such equipment, the seismic fragilities for the containing equipment consider all the equipment in the box. The rule-of-the-box components were included in the SEL and tracked by identification of the parent item. The resulting SEL applicable for walkdowns includes approximately 730 items not counting rule-of-the-box components.

. The Callaway IPEEE was included in the analysis to capture additional information associated with components on the $EL such as seismic classification and component locations.

. Seismically-induced equipment failures potentially resulting in an initiating event were added to the SEL.

. The main structures retained for evaluation in the 5-PRA have been included in the SEL. The structures associated with the SEL equipment are the following:

0 Auxiliary Building (AB) 0 Control Building (CB) 0 Diesel Generator Building (DGB) 0 Essential Service Water System Pumphouse (ESWS) 0 Ultimate Heat Sink (U}{S) 0 Reactor Building (RB) 0 Hardened Condensate Storage Tank (HCST) 0 Alternate Emergency Power System (AEPS)

. Instrumentation is not always explicitly modeled in the IE PRA due to the high level of redundancy.

This screening criterion is potentially not applicable to a 5-PRA because of the possibility of a seismic event inducing a common mode failure (i.e., full correlation of seismic failure) of similar equipment that would circumvent such redundancy. For this reason the more detailed Page 32 of 92 to ULNRC-06524 instrumentation review performed for the Callaway F-PRA was used to supplement the SEL developed in the first bullet.

Internal events modeling often screens flow diversion paths based on the size of the potential diversion path. The internal events analysis uses a single failure as one ofthe criterion for screening.

This screening criterion is potentially not applicable to a 5-PRA because of the possibility of a seismic event inducing a common mode failure (i.e., full correlation of seismic failure) of similar equipment that would result in multiple diversion paths being impacted. Spurious opening of passive valves is not considered a realistic seismic-induced failure mode; thus the primary concern is the spurious opening of power operated valves associated with relay chatter. This is in essence the same concern of spurious actuation due to hot shorts in the F-PRA. The Callaway F-PRA was therefore reviewed for additional equipment that was included in the Multiple Spurious Operations (MSO) evaluation. Equipment identified in this analysis was added to the SEL.

Containment penetrations are screened from explicit modeling in the IE PRA based on a number of criteria, included the size of the penetration. In case of a seismic event, it could be envisioned that multiple small bore lines could fail to be isolated due to a common mode, seismic-induced failure. For this reason, the containment penetrations were reviewed to assess if the screening criterion was still applicable in a seismic scenario. Spurious opening of passive valves is not considered a realistic seismic-induced failure mode; thus the main focus was on power operated valves that could be spuriously opened by seismic-induced relay chatter. Additional equipment was added to the SEL if the screening criteria were not applicable.

ISLOCA pathways are screened from explicit modeling in the IE PRA based on a number of criteria, including the size of the interfacing lines. This screening criterion is potentially not applicable to a S-PRA because of the possibility of a seismic event inducing a common mode failure (i.e., full correlation of seismic failure) of similar equipment that would result in multiple valves opening.

Spurious opening ofpassive valves is not considered a realistic seismic-induced failure mode; thus the main focus was on valves that could be spuriously opened by seismic-induced relay chatter.

These valves have been reviewed to evaluate whether any additional equipment needed to be added to the baseline SEL.

Distributed systems are not normally modeled in the IE PRA because oftheir low failure probability but are vulnerable to seismic-induced failure. Distributed systems have been added to the SEL in a representative fashion.

0 Inclusion of piping associated with fluid systems credited in the 5-PRA.

0 Inclusion of cable trays associated with electrical systems credited in the 5-PRA.

0 Inclusion of ducts associated with HVAC systems credited in the 5-PRA.

A number of checks have been run to ensure that individual assumptions or modeling simplifications in the internal events model would not result in incorrectly leaving out equipment from the analysis.

All the input generated by the previous steps has been assembled into one summary table representing the Callaway 5-PRA SEL The internal fire and internal flooding PRAs were used to gain insights on risk significant flood and fire. These fires and flood sources are not included in the SEL but are provided to the fragility team so that they can be investigated for additional seismic related vulnerabilities.

A qualitative comparison with previously completed Seismic Equipment Lists.

Page 33 of 92 to ULN RC-06524 4.1.2 Relay Evaluation/Spurious Breaker Trip Evaluation During a seismic event, vibratory ground motion can cause relays to chatter. The chattering of relays potentially can result in spurious signals to equipment. Most relay chatter is either acceptable (does not impact the associated equipment), is self-correcting, or can be recovered by operator action. An extensive relay chatter evaluation was performed for the CEC S-PRA, in accordance with EPRI 1025287, Screening, Prioritization and Implementation Details (SPID) [28] Section 6.4.2 and ASME/ANS PRA Standard [29]

Section 5-2.2. Fragility analysis was performed for relays with the potential to impact SEL component functions. Those relays identified as having a significant impact on core damage frequency (CDF) and large early release frequency (LERF) were functionally screened. The evaluation resulted in most relay chatter scenarios screened from further evaluation based on no significant impact to component function.

Table 4-1 lists relays with significant cOntributors to risk, along with their function and disposition in the 5-PRA with appropriate seismic fragility or operator action.

An evaluation of spurious trips of breakers was also performed for low and medium voltage switchgear.

The functionality of breakers was analyzed through evaluation of the site-specific seismic qualification testing performed for design basis and verification ofEPRI NP-5223-SL [30] generic equipment ruggedness spectra (GERS) caveats. The major types ofbreakers at the plant are air breakers (medium voltage), draw-out type breakers (low voltage), and molded case circuit breakers. Molded case circuit breakers inherently have high seismic capacity. The switchgear fragility which house breakers were evaluated through EPRI NP-6041 -SL [3 1 ] conservative, deterministic failure margin (CDFM) criteria. The seismic capacities used in this evaluation were based on the site-specific qualification testing and EPRI GERS in EPRI NP-5223-SL [30]. This evaluation addressed fragility for high frequency sensitive components as discussed in Section 6.4.2 ofthe SPD [28].

Table 4-1 : Summary of Disposition of Risk Significant Relays Relay Function Disposition Relay_O. 1 8DG Can cause DG circuit breaker Fragility analysis performed and closure, affecting bus incorporated explicitly into the 5-PRA operation and leading to loss model.

ofAC power.

RelayO.33 Linked to various systems and Fragility analysis performed and components incorporated explicitly into the 5-PRA model.

4.2 Walkdown Approach This section provides a summary of the methodology and scope of the seismic walkdowns performed for the 5-PRA. Walkdowns were performed by personnel with appropriate qualifications as defined in EPRI NP-6041-SL [3 1] Section 2 and the requirements in the ASME/ANS PRA Standard [29] Section 5-2.2.

Each seismic review team (SRT) utilized for the 5-PRA included seismic engineers with extensive experience in fragility assessment and seismic walkdown training. Walkdowns of those SSC included on the seismic equipment walkdown list were performed to assess the as-installed condition ofthese SSC for use in determining their seismic capacity (e.g. anchorage and lateral support), and performing initial screening, to identify potential 1111 spatial interactions and look for potential seismic-induced fire/flood interactions. The fragility walkdowns were performed in accordance with the criteria provided in EPRI NP-6041-SL [3 1]. The information obtained was used to provide input to the fragility analysis and 5-PRA modeling (e.g., regarding correlation and rule-of-the-box considerations).

Page 34 of 92

Attachment 2 to ULNRC-06524 The seismic fragility walkdowns were conducted on all accessible SEL equipment including equipment inside the RB. The fragility walkdowns included the evaluation of seismic interactions, including the effects of seismic-induced fires and flooding.

In addition to evaluating individual components and associated systems on the SEL, the walkdown reviewed the fire protection system. The fire protection piping was found to be well supported and not susceptible to anchorage failures.

A concern for seismic-induced fires is from flammable gases and liquids. Thus, the walkdown included these sources and their proximity to components on the SEL. Potential fires due to hydrogen piping in SEL buildings and transformers in SEL buildings are examples of scenarios that were evaluated. The potential for fire initiation due to seismic failure ofhigh-voltage non-safety electrical cabinets was also investigated.

The potential for seismically-induced flooding was also evaluated. During the walkdowns, potential spray and flooding scenarios from piping systems and SEL components were reviewed. Flood sources, including the fire-protection system, were evaluated.

Pre-action fire protection systems were also considered and were investigated for seismic-induced flooding due to potential tripping of ionized particle smoke detectors for the case of nearby block walls failing, thereby producing dust in the air and initiating false alarming.

4.2.1 Significant Walkdown Results and Insights Consistent with the guidance from EPRI NP 6041 -SL [3 1], no significant findings were noted during the CEC seismic walkdowns. Components on the SEL were evaluated for seismic anchorage and interaction effects in accordance with SPID [28] guidance and ASME/ANS PRA Standard [29] requirements.

The walkdowns also assessed the effects of component degradation, such as corrosion and concrete cracking, for consideration in the development of SEL fragilities. In addition, walkdowns were performed on operator pathways, and seismic-induced fire and flooding scenarios were assessed, and potential internal flood scenarios were incorporated into the CEC S-PRA model. The walkdown observations were used in developing the SSC fragilities for the $-PRA.

4.2.2 Seismic equipment List and Seismic Walkdowns Technical Adequacy The CEC S-PRA SEL development and walkdowns were subjected to an independent peer review against the pertinent requirements (i.e., the relevant $FR and SPR requirements) in the PRA Standard [32], also followed up by an F&O closure review.

The peer review assessment, and subsequent disposition ofpeer review findings, is described in Appendix A, and establishes that the CEC S-PRA SEL and seismic walkdowns are suitable for this $-PRA application.

4.3 Dynamic Analysis of Structures This section summarizes the dynamic analyses ofstmctures that contain systems and components important to achieving a safe shutdown. The methodologies used to develop fixed-base and soil structure interaction

( $51) models, and to perform SSI analyses are discussed.

4.3.1 Fixed-base Analyses All major CEC structures are founded on soil. Therefore, fixed-base analyses were not applicable. It is noted however that intermediate fixed-base modal analyses were performed to check the modeling fidelity when Lumped Mass Stick Models (LMSM) were recreated from design models, as well as for validation purposes.

Page 35 of 92 to ULNRC-06524 4.3.2 Soil Structure Interaction (551) Analyses All major structures are founded on or embedded in soil and therefore an S$I analysis was required for a realistic estimate ofresponse. In addition to the structures, an $SI analysis was performed for the Refueling Water Storage Tank (RWST).

The ACS SASSI [33] program was used to perform the 551 analyses for the 3-D finite element model of the AB/CB, while the EKSSI program [34] was used to perform the 551 analyses on the LMSM ofthe RB, the DGB, the ESWS, the liFTS, and RWST. frequency-dependent impedance functions for layered media were used in all analyses.

Soil layers and properties are primarily based on the profiles developed in Probabilistic Seismic Hazard Analysis Seismic Probabilistic Risk Assessment Project, Callaway Energy Center, Unit 1 [35]. The 551 analyses used these soil layers, along with the Ground Motion Response Spectrum (GMRS) seismic hazard and the corresponding hazard-consistent strain-compatible properties provided in [35], to (a) generate time histories and soil column data for the Foundation Input Response Spectra FIRS-i and FWS-2 compatible to the GMRS, and (b) convert the generated time histories to in-structure response spectra (ISRS) for use as seismic input to the building models.

Modeling uncertainty was addressed by developing best-estimate (BE), upper bound (UB), and lower bound (LB) soil stiffness models. For all buildings, the general approach followed was that the soil stiffness uncertainty dominates all other uncertainties. An 551 response analysis was performed for each ofthe BE, UB and LB cases for the AB/CB, RB, DGB, ESWS, and UHS.

Ground motion input for the 551 analyses was based on the site-specific GMRS which is anchored to a O.39g horizontal peak ground acceleration (PGA). A single set of artificial time histories (one time history per direction) was generated based on the criteria presented in ASCE 43-05 [36] Section 2.4 and NUREG 0800 Section 3 .7. 1 [37] Acceptance Criterion 1B, Option 1 Approach 2. Additional criteria used, include checks for statistical independence, strong-motion duration, power spectral density and Arias intensity, to ensure that the resulting time histories are suitable without any deficiencies of power across the frequency range ofinterest. The adequacy ofthe set ofartificial time histories was further verified by comparing their ISRS to the average ISRS from 551 analyses using 5 sets of time histories generated from real earthquake seeds and spectrally matched to the site-specific GMRS.

A series of sensitivity studies were performed to check the validity of stiffness variability and cracking assumptions.

The generated ISRS have a 84% non-exceedance probability and are suitable for CDFM analysis per the criteria of EPRI NP-604 1 -SL [3 1] Section 2 and EPRI 1 0 1 9200 [3 8] Appendix A. ISRS with amplified narrow frequency content were clipped for comparison to broad-banded test response spectra, typical of most nuclear power plant components. The guidance in EPRI TR-103959 [39] Section 3 was followed for the peak clipping process.

4.3.3 Structure Response Models The ANSYS [40] finite element program was used to develop the AB/CB 3-D finite element fixed-base model, while the GTSTRUBL sofiware [41] was used to develop the fixed-base LMSM of the remaining building structures (RB, DGB, ESWS, and UHS). The modeling effort was guided by these primary goals:

. Models are to provide the capability to estimate realistic seismic demand on in-scope SSC.

. Models are to satisfy review criteria of EPRI 1025287, Screening, Prioritization and Implementation Details [28], Section 6.3.1.

The Auxiliary Building and the Control Building share a common foundation; therefore, they were modeled and analyzed together. All other buildings are on independent foundations; therefore, they were modeled and analyzed as independent structures.

Page 36 of 92 to ULNRC-06524 A detailed 3-D finite element fixed-base model was developed for the AB/CB to capture its irregular geometry with respect to lateral load paths, as well as the global lateral load path, the vertical flexibility of floor slabs and horizontal flexibility of floor diaphragms, which would be difficult to realistically model with a LMSM. The finite element model was developed from plant drawings and related documents. Each floor of the building above the foundation is constructed of reinforced concrete slabs supported by reinforced concrete walls and columns. The slabs were modeled using shell elements, whereas steel beams and concrete columns were modeled using beam elements. Furthermore, the AB/CB contain a significant amount of active SEL electrical and mechanical equipment at multiple elevations including batteries, switchgear, and transformers. The mass of these equipment and their distribution on its floor was accordingly accounted for in the finite element model.

fixed-base LM$M for the RB, DGB, ESWS and TiNS were developed from plant design documents.

Specifically, design-basis fixed-base LMSM were obtained for these structures and enhanced as needed to more realistically assess the demands on equipment. Enhancements included adding elements to represent the load path through concrete fill above the foundation, adding outrigger nodes to automatically capture torsional and rocking effects in time histories, and enhancing torsional response capabilities. Enhanced LMSM met the modeling criteria of SPID [28] Section 6.3 .1 and were considered suitable for 5-PRA response analysis. Criteria were verified by a review ofdrawings, review ofplant design basis calculations, and by supplemental calculations.

Concrete cracking was considered for both the 3-D finite element model and the LMSM, per the guidance of ASCE 4-1 6 [42]. To determine cracking, the design shear stresses on each floor were checked against the allowable code limits. In cases when the stresses were determined to exceed the code the stiffness of the affected elements was reduced accordingly.

The guidance of ASCE 4-16 [42] Section 3.2 was applied to properly select structural damping for the buildings. The applied damping values are intended to have a conservative bias to meet the intent of EPRI NP-6041-SL [31] Table 2-5 for estimation of 84% non-exceedance probability in-structure response.

Buildings are primarily reinforced concrete shear-wall structures and applied damping is typically in the range of 4% to 7%. for 551 analysis, the effective foundation damping is implicitly accounted for through the frequency-dependent impedance functions used.

A simplified LMSM was also developed for the HCST to capture the seismic response using 551 analysis.

Effective masses for the sloshing and impulsive modes were per EPRI NP-604l -SL [3 1] Appendix H. A 5% structural damping value was applied per EPRI NP-604l -SL [31].

4.3.4 Seismic Structure Response Analysis Technical Adequacy The CEC 5-PRA Seismic Structure Response and Soil Structure Interaction Analysis were subjected to an independent peer review against the pertinent requirements in the PRA Standard [32].

The peer review assessment, and subsequent disposition of peer review findings, is described in Appendix A, and establishes that the CEC 5-PRA Seismic Structure Response and Soil Structure Interaction Analysis are suitable for this 5-PRA application.

4.4 SSC Fragility Analysis The primary goal of the fragility analysis was to identify and analyze critical SSC. A critical SSC item is one that ranks low in terms of seismic ruggedness and is also important to plant safety. A screening process was followed to select SSC for analysis. The process included review of the plant seismic design bases, performance of seismic walkdowns, and application of industry practice related to seismic margin and 5-PRA studies.

The screening evaluation was performed for the 1 .2g spectral acceleration level, which corresponds to the 2 screening lane of EPRI NP-6041 -SL [3 1] (refer to EPRI NP-6041 -SL [3 1], Table 2-3 and 2-4). The seismic capacity inferred by this value is:

Page 37 of 92 to ULNRC-06524 Sc 1 .2g high confidence of low probability of failure (HCLPF) seismic capacity, as spectral acceleration The 1 .2g value associated with screening is applicable to the 5% spectral acceleration at the ground across a broad frequency range (from about 2 to 8 Hertz per EPRI NP-6041-SL [31] Page 2-44). Spectral accelerations above this range are less damaging and are less important to both structures and equipment.

A HCLPF value for screened-out equipment and structures was determined by comparing the spectral peak of the GMRS to the above screening capacity. A combined logarithmic standard deviation c (for randomness and uncertainty) was assigned per the SPID [28] guidance.

The guidelines of EPRI NP-604 1 -SL [3 1 J Table 2-3 were applied for screening of civil structures. The guidelines of EPRI NP-6041-SL [3 1] Table 2-4 were applied for screening of equipment and subsystems.

After initial screening, an item was either screened-in (a fragility was performed) or screened-out (a surrogate value was given).

Analysis was performed for all screened-in items. In addition, analysis was performed for some screened-out equipment to increase the 5-PRA model capability. The resulting sample of SSC is broad and supportive of a robust 5-PRA model. Some screened-out items required verification of issues that could not be resolved by design review or walkdown. A set of action items was created to track items designated for fragility analysis and to resolve outstanding screening issues. EPRI reports TR-10l9200 [38], TR 1 03959 [39] and NP-6041 -SL [3 11 were used as the basis for calculation of seismic fragility parameters.

The lognormal model was used.

Screened-out items were those judged to have a relatively high seismic fragility. Fragility parameters for this category of SSC were addressed by a surrogate element(s) similar to that described in EPRI report TR 1 0 1 9200 [3 8] As an option to increase 5-PRA model capability, seismic fragility parameters applicable to an individual screened-out equipment item were also provided. With this option, the reliance on surrogate elements may be reduced.

Table 4-2: summarizes key attributes of the fragility analysis. More details are provided in following sections and in supporting project documents.

Table 4-2 : Key Attributes of CEC Fragility Analysis Attribute Description Seismic equipmentlist The SEL was provided by the plant response model team to the fragility analysis team. The SEL was the basis for the scope of $SC addressed by the plant fragility

_z- analysis.

The above process ensured that the fragility analysis addressed all SSC explicitly or implicitly credited in the plant response model.

Screening process The screening process of EPRI NP-6041-SL [31] was applied. This process is widely used in the nuclear power industry for 5-PRA and similar beyond-design-basis studies.

Screening level The project earthquake screening level was the 1 .2g spectral acceleration level identified in EPRI NP-6041-SL [10] Section 2. This screening level was found to be sufficient for evaluation of structural failure modes of buildings and passive equipment.

Page 38 of 92

Attachment 2 to ULNRC-06524 Table 4-2 : Key Attributes of CEC Fragility Analysis Attribute Description Walkdown scope and Comprehensive seismic walkdowns were performed and documented as part of the procedures EPRI NP-604 1 -SL [3 1 ] screening process. All equipment items credited in the plant response model were included in the walkdown scope. Other than a small number of inaccessible items, all equipment was inspected by the SRI.

For the walkdowns, emphasis was placed on inspection of as-built anchorage and lateral support, investigation for seismic interactions (including seismic-fire and seismic-flood) and checks for seismic vulnerabilities documented in the earthquake experience database. The walkdowns also helped the fragility analysis team become familiar with the plant construction.

Earthquake The applied earthquake was based on site specific GMRS that were produced by the characterization recent probabilistic seismic hazard analysis [35].

Building seismic response New ISRS were produced for the 5-PRA project using the site-specific earthquake analysis motion. New building models were created for this purpose including detailed 3-D finite element models for critical buildings.

Treatment of screened-out Potential failure of screened-out equipment was addressed through development of equipment surrogate elements. Fragility data were derived from the screening levels and assigned to the surrogate elements.

Fragility analysis methods The methods in EPRI reports NP-6041-SL [3 1], TR-1019200 [38] and TR-l03959

[ 39] were used for calculation of seismic fragility parameters. The level of analysis effort applied for an SSC item was tied to the best understanding of its importance to the plant seismic response, and feedback from the plant response model was used to sharpen the focus on the most important items. fragility calculations were performed in stages to make use of the plant response feedback. Seismic analyses were initially performed using the CDfM method of EPRI NP-6041-SL [3 1].

Refined CDFM analysis were then performed for dominant contributors to risk and if they remained dominant to risk, another refinement using separation of variables (SOV) method was performed for realistic variabilities.

Documentation Supporting documentation was created to detail the scope, methods and results of fragility analysis, to verify quality, allow revisions and upgrades, and support regulatory and peer review.

4.4.1 SSC Screening Approach 4.4.1.1 Overview The methods in EPRI reports NP-6041-SL [31], TR-1019200 [38] and TR-103959 [39] were used for calculation of seismic fragility parameters. The seismic fragility of an SSC item is defined as the conditional probability of its failure at a given value of acceleration. The following parameters define the seismic fragility for any specific S$C item:

FGAmed =

median capacity, stated as peak ground acceleration fir =

logarithmic standard deviation, randomness flu =

logarithmic standard deviation, uncertainty

=

SRSS(j3r, f3u) logarithmic standard deviation, combined Page 39 of 92 to ULNRC-06524 The above parameters are tied to a lognormal probability distribution. FGAmed represents the best estimate ofthe seismic capacity (50% probability offailure). The /3 parameters address the variability ofthe estimate.

Parameter fir (randomness) accounts for sources of variability that cannot be reduced by more detailed studies or more data.

In general, a fragility is associated with the failure to perform an assigned function. Non-performance may be a consequence ofstructural failure, ofmalfunction ofan electro-mechanical component, or ofsome other type ofphysical change. The governing failure mode is identified in the final listing offragility parameters.

Parameter FGAmed corresponds to earthquake severity and is defined in terms of horizontal PGA at the control point. The fragility analysis accounts for propagation ofearthquake ground motion to the SSC item location and the resulting dynamic response of the item and its supporting structure.

4.4.1.2 Initial Analysis Seismic analyses were initially performed using the CDFM method of EPRI NP-6041-$L [3 lJ. Each analysis produced a HCLPF capacity for the SSC item. Nominally, the HCLPF is the capacity at which there is 95% confidence of less than 5% probability of failure. Fragility parameters were then produced using the scaling approach ofEPRI TR-1019200 [3$] Section 3.4. This is equivalent to the Hybrid Method discussed in EPRI 1025287 [28] Section 6.4. 1 The median capacity was estimated from the HCLPf value using the following equation:

FGAmed = (PGAc)

• e2325t FGAc = HCLPf capacity, stated as peak ground acceleration To produce the initial FGAmed for SSC, the HCLPF value was calculated and the corresponding fic value is estimated based on 5-PRA experience. Per EPRI 1 025287 [28] Table 6-2, fic = 0.45 was typically applied for an item subject to in-structure demand andflc = 0.35 was typically applied for an item subject to demand at the ground level. In some cases, an item-specific fi?c was applied. Unless noted otherwise, the randomness component of/fr was set to 0.24 per EPRI 1025287 [28] Table 6-2.

Application ofthe CDFM to screened-in SSC as a first step provides these benefits:

1 . The method and criteria are straightforward and not overly reliant on analyst judgment.

2. Sorting the critical SSC by risk-significance is more dependable when seismic fragilities are based on a common method.

3. It is practical to develop item-specific fragility parameters for a large population or equipment (especially when existing plant seismic calculation methods are similar to CDFM).

The CDFM analysis criteria followed are based on EPRI NP-6041-SL [3 1] Table 2-5 but include the recommended updates of TR-l 0 1 9200 [3 8] Appendix A.

4.4.2 SSC Fragility Analysis Methodology The HCLPF values from the initial analysis were supplied to the plant response model team for preliminary analysis of seismic core damage frequency (SCDF) and large early release frequency (SLERF). Based on that initial data, dominant contributors were identified, and more rigorous methods of analysis were applied for this subset of screened-in SSC. Under this approach, an existing CDFM analysis was used as a reference and a more refined CDFM analysis was performed. If the item remained significant to risk, the SOV approach was used as a more refined analysis method per EPRI TR-l03959 [39] and more realistic variabilities were determined. In general, the refined analysis is expected to produce a more accurate median capacity estimate and more accurate log standard deviations.

Page 40 of 92 to ULNRC-06524 4.4.3 S$C Fragility Analysis Results and Insights Section 5 reports the fragilities of the risk important contributors to the SCDf and SLERF. Detailed

( separation ofvariables) calculations have been performed for selected highest risk-significant SSC, as well as for other selected components.

4.4.4 SSC Fragility Analysis Technical Adequacy The CEC 5-PRA SSC fragility Analysis was subjected to an independent peer review against the pertinent requirements in the PRA Standard [32].

The peer review assessment, and subsequent disposition of peer review findings, is described in Appendix A, and establishes that the CEC 5-PRA SSC Fragility Analysis is suitable for this 5-PRA application.

• * , , tJ -

Page4l of92

Attachment 2 to ULNRC-06524 5.0 Plant Seismic Logic Model This section summarizes adaptation of the CEC internal events at-power PRA model to create the $-PRA plant response (logic) model.

The seismic plant logic model includes combinations of structural, equipment, and human failures that could give rise to significant core damage and large early release sequences. Quantification of this model yields total SCDF and $LERF, including contribution ofboth seismic-induced and non-seismically induced unavailabilities, and the identification of important seismic risk contributors. The quantification process also includes an evaluation of uncertainty, which provides perspective on how modeling and parametric sources of uncertainty affect the S-PRA insights.

5.1 Development of the $-PRA Plant Seismic Logic Model The CEC seismic logic model was developed from the internal events at-power PRA model ofrecord. The internal events model was adapted in accordance with the EPRI Seismic PRA Implementation Guide [27]

and ASME/ANS PRA Standard [32] requirements. This process included adding seismic fragility events to the logic model, eliminating portions of the logic model irrelevant to seismic risk, and adjusting the human reliability analysis to account for response during and following an earthquake. The final seismic logic model is a large fault tree with single top events for SCDF and SLERF. The following summarize each ofthe major changes made to the internal events logic model during 5-PRA Model development.

5.1.1 General Approach Seismic-induced initiating events that could give rise to significant accident sequences were first identified, including seismic-unique initiating events such as building failure. Next, SSCs whose seismic failure could cause an initiating event, or degrade plant response to an initiating event, were identified and consolidated into an SEL. Initially conservative fragility groups and lognormal fragility parameter estimates were identified for each SEL item.

The seismic hazard was discretized into ten (1 0) intervals, each with a representative ground motion level and occurrence frequency. Fragility events representing seismic failure of individual components, or groups ofcomponents were developed for each ground motion interval. These fragility events were inserted into the fault tree using internal events basic events as targets. Human failure events relevant to seismic sequences were quantified using a screening process accounting for earthquake impact on performance shaping factors, and the resulting seismic human error probabilities were incorporated into the S-PRA quantification process. The resulting 5-PRA model is capable of quantifying at-power seismic-induced CDF and LERF, including the contributions of both seismic-induced and non-seismic hardware failures.

5.1.2 Initiating Events and Accident Sequences The initial step of the 5-PRA was to systematically identify earthquake-caused initiating events that have the potential to give rise to significant accident sequences. In the initiating event identification process, a hierarchy was developed to ensure earthquakes exceeding an OBE are modeled, and that their frequency is apportioned to an appropriate induced initiating event. The CEC S-PRA includes the following seismic-induced initiating events:

. Direct to Core Damage and Large Early Release

. Large LOCA

. Intermediate LOCA

. SmailLOCA I Very Small LOCA

. Main Streamline Break Outside of Containment Page 42 of 92

Attachment 2 to ULNRC-06524

. Loss of all Component Cooling Water

. Loss of Vital DC Bus

. Spurious Safety Inj ection Signal Due to Relay Chatter

. LOOP and SBO

. Loss of All Service Water The direct to core damage and large early release initiating events include seismic-unique failures such as building collapse. The potential for seismic-induced very small LOCA is modeled following non-LOCA initiating events. The initiating event and mitigating systems impact of seismic-induced fires and floods is also included quantitatively in the 5-PRA.

While the CEC 5-PRA uses a largely unmodified version of the internal events accident sequence and system modeling, some changes were required to reflect potentially risk-significant seismic sequences that were not already included in the base internal events model. The significant modifications are listed below:

. Added logic to reflect seismic failures that lead directly to core damage and large early release

. Added logic to reflect seismic-induced very small loss of coolant accident following non-LOCA initiating events.

. Disabled credit for recovery of offsite power.

. Updated the mutually exclusive logic to prevent non-sensical failure combinations

. Modification ofthe recovery rule file to apply the seismic human reliability analysis

. Used FRANX to create and insert logic reflecting seismic-induced initiating events and mitigating equipment failures.

5.1.3 Modeling of Correlated Components Fully correlated components were assigned to correlated component groups so that all components in the group fail at the same time with the same probability based on the seismic magnitude for each hazard bin.

The model assumes fully correlated response of same or very similar equipment in the same structure, elevation, and orientation. Correlated component groups were developed consistent with the above mentioned criteria and based on insights from component walkdowns.

5.1.4 Modeling of Human Actions The CEC Seismic HRA consists ofthe following tasks:

. Operator action identification and definition

. Feasibility assessment

. Screening quantification

. Detailed quantification

. Model integration Operator actions to be modeled by the 5-PRA were identified and defined using the guidance of EPRI 3002008093 [43] Chapter 3, consistent with supporting requirement SPR-Dl of the PRA Standard [32J.

. All operator actions modeled by the IF PRA are identified.

. Any operator actions with a screening human error probability (HEP) applied from the internal events HRA (i.e., not detailed analysis provided) are not credited in the 5-PRA.

Page43 of 92

Attachment 2 to ULN RC-06524

. Pre-initiator HFEs are independent of the initiating event, and therefore there is no seismic impact to these actions.

. Following initial quantification, a review is performed to identify (if applicable) seismically risk-relevant operator actions not already modeled in the IE PRA. Examples include recovery actions for mitigation equipment impacted by seismic-induced relay chatter.

The feasibility of each identified operator action is assessed using the guidance of EPRI 3002008093 [43]

Section 4.2. The purpose of the feasibility assessment is to determine if successful completion of each operator action is even possible in the event of an earthquake. All actions determined to be infeasible must either not be modeled by the S-PRA or have their associated HEPs set to 1 .0. The feasibility assessment

-. considers seismic impact on the following performance shaping factors: time, manpower, cues, procedures and training, accessible location and environmental factors, and equipment accessibility, availability, and operability.

All operator actions carried forward from the IE PRA have been determined feasible, in the context of internal initiating events. The conclusion was that the internal events feasibility assessment of each action remains valid in the context of seismic events, and any exceptions are addressed by the screening quantification process.

In the course ofthe HRA analysis, a number ofHRA events were identified and refined. Significant HRA events are identified during the review of the quantification results. Depending on the risk significant of the actions, an iterative process is used to refine and update the HRA events and to assess their impact on quantification results. The following events went through a detailed HRA refinement:

. $H2-EF-XHE-FO-ESWREC - OPS FAILS TO MANUALLY START AND ALIGN E$W SYSTEM BEFORE RX TRIP

. SH2-EF-XHE-FO-MANESW - OPERATOR FAILS TO MANUALLY START AN]J ALIGN ESW SYSTEM

. SH2-FB-XHE-FO-FANDB - OPERATOR FAILS TO ESTABLISH FEED AND BLEED

. SH2-OP-COG-CCW - OPERATORS FAIL TO DIAGNOSE LOSS OF CCW

. SH2-OP-XHE-FO-AEPS1 - OPERATOR FAILS TO ALIGN AEPS

. S112-OP-XHE-FO-DEPRES - OPERATOR FAILS TO COOLDOWN ANIJ DEPRESSURIZER RCS

. SH2-OP-XHE-FO-NSAFP - OPERATOR FAILS TO SUPPLY NSAFP WITHIN 45 MNS

. S112-OP-XHE-FO-RCPTRP - OPERATORS FAIL TO TRIP RCP FROM CONTROL ROOM Additionally, based on risk insights from the $-PRA quantification process, a seismic specific operator action (NE-XHE-FO-EDG-RLYSET) was defined in accordance with supporting requirement SPR-D2 of the PRA Standard [32] to address recovery from the effects of seismic-induced relay chatter for relay group Relay_O. 1 8DG. Feasibility of this ex-control room action has not yet been assessed via a dedicated walkdown. A sensitivity analysis was performed to capture this impact on model results.

5.1.5 Seismic LERF Model The transition from Level 1 sequences to Level 2 sequences developed for the IE PRA is considered valid for the 5-PRA. A specific initiator evolving into a release scenario is generated by a seismic event does not change the phenomenology associated with containment failure and individual components associated with containment isolation function. The components associated with containment failure are captured in the SEL and are included within the scope of the fragility analysis and the subsequent PRA modeling.

Operator actions involved in the Level 2 analysis are similarly captured in the human reliability analysis.

Page 44 of 92

Attachment 2 to ULNRC-06524 Currently, all containment penetrations are treated as correlated and are assigned to one fragility group (SF RB-PEN). Failure of containment penetrations is modeled separately from Reactor Building collapse via the fragility associated with Reactor Building equipment hatches.

5.2 S-PPA Plant Seismic Logic Model Technical Adequacy The CEC S-PRA seismic plant response methodology and analysis were subjected to an independent peer review against the pertinent requirements in the ASME/ANS PRA standard [13J.

The 5-PRA was peer reviewed relative to Capability Category II for the full set of requirements in the Standard. Afier completion of the subsequent independent assessment, the full set of supporting requirements was met with the exception of SPR-B2 and SPR-E6 which were not reviewed by the independent assessment team. The Seismic plant response analysis was determined to be acceptable for use in the 5-PRA. The Peer review assessment, and subsequent disposition of peer review Facts and Observations (F&O) through an independent assessment, is further described in Appendix A and References [1 4] and [15].

5.3 Seismic Risk Quantification 5-PRA quantification involves assembling the results ofthe seismic hazard analysis, fragility analysis, and the seismic accident sequence model to estimate the frequencies of core damage and large early release.

The risk quantification considers both seismic failures and non-seismic failures, and the applicable operator actions. This section describes the S-PRA quantification methodology and important modeling assumptions.

5.3.1 S-PRA Quantification Methodology The Callaway 5-PRA is a large fault tree with separate top events for SCDF and SLERF. The FRANX software is used to create an integrated one-top PRA model. The fault tree is quantified using FTREX through the PRAQuant graphical user interface (GUI), results are post-processed using ACUBE, and seismic human error probabilities are applied via cutset post-processing with QRecover. The approach used by the FRANX tool is a scenario-based approach which divides the seismic hazard curve into discrete seismic magnitude intervals. PRAQuant is used to quantify each scenario (i.e., %GO1, %G02, and finishing with the final hazard interval selected). Each seismic interval is assigned a representative ground motion magnitude and corresponding occurrence frequency. The seismic interval frequencies become the frequency (of occurrence) for the seismic initiators, and the representative magnitude at each interval is used as input to the fragility calculations for the fragility events modeled in the 5-PRA.

Seismic HRA dependencies for CDF and LERF were assessed at each seismic HRA bin and the impact of dependent combinations is explicitly included in the S-PRA model quantification and results.

The 5-PRA quantification generates a wealth of data, which generally require post-processing to extract meaningful insights. At a minimum, the Callaway 5-PRA results are presented in the following forms:

. Total seismic CDF and LERF

. Fractional CDF and LERF contributions of each ground motion interval

. CCDP and CLERP for each ground motion interval

. Occurrence frequency for each ground motion interval

. Ground motion at which the estimated likelihood of core damage and release are 1 .0 (plant-level fragility)

. Documented cutset review, including sampling of non-significant cutsets

. Uncertainty analysis for CDF and LERF Page 45 of 92 to ULNRC-06524

. Dedicated sensitivity studies to quantitatively evaluate key assumptions and sources of model uncertainty 5.3.2 $-PRA Model and Quantification Assumptions Significant assumptions and sources ofuncertainty for the CEC S-PRA are summarized as follows:

Hazard Uncertainty The hazard intervals used for quantification are tailored for the current risk profile as an iterative approach was used to determine the number and width of intervals which would result in an increasing seismic CCDP/CLERP percent by hazard interval (up to the PAF) and would also result in a seismic percent CDF/LERF by hazard interval distribution skewed around the plant HCLPF value. A study on hazard intervals was completed and demonstrates convergence based on varying the hazard intervals. Additionally, the epistemic uncertainties identified through the SSHAC process in the hazard are effectively covered via the P$HA in the distribution that describes the entire family of hazard curves. The seismic hazard

( exceedance frequency over a range of credible PGA) includes both mean and fractile curves. When FRANX discretizes the hazard and injects seismic initiators into the .rr database, it assigns distribution parameters to characterize the uncertainty of each hazard interval. While the hazard uncertainty is significant, it is quantitatively included in the uncertainty quantification.

Fragilities No major assumptions or sources of model uncertainty were identified during the fragility analyses that merit a sensitivity study in the final risk quantification. Similar to the hazard, SSC fragility estimates are also subject to significant uncertainty but are quantitatively included in the uncertainty evaluation as each fragility is represented as a lognormal distribution using its median capacity (Am), uncertainty variable (Beta U, which is the uncertainty (state-of-knowledge) of what the true median capacity is), randomness variable (Beta R, which represents how the failure probability of SSC changes based on different seismic intensities) and composite uncertainty variable (Beta C).

Model Development Significant assumptions and source of model uncertainty identified during the development of the 5-PRA model are characterized for their impact on the 5-PRA results below:

1 . Correlation The CEC 5-PRA is performed under the generic assumption of full correlatiOn of seismically-induced failures of similar components. This is a recognized conservative assumption that is commonly used in the developed of 5-PRAs and is consistent with industry practice. While research activities are performed in the industry to investigate alternative modeling of potential correlations between seismic failures, these are judged to not be ready yet to be implemented in the CEC 5-PRA. Correlation grouping was initially determined based on equipment type and building location. For dominant contributors, correlation grouping was iteratively refined where appropriate based on additional considerations consistent with SFR-A2 (e.g., building elevation, anchorage similarity, orientation, etc). For significant fragility groups (FV > 2% for CDF or LERF), as identified by the fragility ranking sensitivities, the basis for the correlation assumptions is reviewed.

2. LLOCA Seismically-induced failure of the pressurizer supports are assumed to lead to unrestrained motion of the pressurizer and subsequent failure of the pressurizer surge line. No other concurrent failures resulting from the unrestrained motion of the pressurizer are assumed.

This is a recognized conservative assumption used for modeling purposes as it is understood that the motion of the pressurizer can be impeded by other structures in the containment and that pipe flexibility can probably accommodate a significant motion from the pressurizer before generating a full break ofthe line. This modeling assumption is expected to artificially increase the importance of seismically induced LLOCA. To reduce the modeling uncertainty associated with this modeling assumption a more refined assessment of the actual impact of the pressurizer motion once the Page 46 of 92 to ULNRC-06524 supports are failed could be performed, although to deterministically or probabilistically assess the impact of the unrestrained motion of the pressurizer is beyond current practice in the S-PRA.

Additionally, further assessment is not warranted at this time based on LLOCA due to pressurizer failure not being a significant contributor to the model results.

3 . MLOCA The generic MLOCA fragility estimates available in NUREG/CR-4840 [2 1 ] are considered applicable to the CEC $-PRA. These estimates are recognized as dated and generic (i.e., not plant or design specific). A plant-specific fragility analysis for pipe break is an alternative approach which is not performed for the CEC S-PRA on the basis that seismically-induced Intermediate LOCAs are not expected to be significant or lead contributors to the CEC seismic risk profile. Note that the EPRI S-PRA implementation guide provides an alternative source for generic MLOCA fragility parameters (with a higher median capacity) which may be a potential sorirce of margin. Additionally, further assessment is not warranted at this time based on MLOCA not being a significant contributor to the model results.

4. VSLOCA RC$ leakage within the very small LOCA range is conservatively assumed to occur for all seismic events above the OBE, regardless of the presence of the CVC$ system. The intent ofthis modeling is based on ASME/AN$-RA-Sb-201 3 [2J supporting requirement SPR-B8 which imposes this approach to ensure that some form of RC$ makeup is demanded for all seismic sequences. Note that the Code Case [13] essentially subsumes this requirement into SPR-Al The .

approach is conservative, especially at low ground motion levels, given that much ofthe earthquake operating experience did not involve loss of coolant.

5 . LOCA Location The same LOCA split factions used in the internal events PRA were assumed to be also applicable to the S-PRA. Such split fractions assign equal probability ofa LOCA to happen to any RCS loop. A plant-specific investigation to assess whether any of the RCS loops has an appreciably different seismic capacity is an alternative approach to validate this assumption but it was not performed for the CEC S-PRA as this goes beyond current practice in 5-PRA and is judged not to be a significant limitation in the analysis. Also, note that limiting a seismically-induced LOCA on a specific RCS loop is in some extent inconsistent with the generic item #1 above on full correlation between seismic-induced failures. This is a recognized inconsistency that is driven by the practical limitations in developing detailed fragility estimates for each individual RCS pipe section.

6. Building Failure A seismic event that is strong enough to induce catastrophic failure of a Seismic Category I building is also assumed to generate a loss of offsite power event that is not recoverable within the 5-PRA mission time. This is considered a realistic assumption.

7. Building Failure Catastrophic events such as major structural collapses are assumed to lead directly to a core damage scenario and a large early release. This assumption is considered conservative because ofthe relatively conservative approach normally used to generate the fragility of major structural collapses. There are no sensitives that are explicitly designed to monitor the epistemic uncertainty associated with this assumption but the importance of the scenario leading directly to core damage has been monitored during the development of the CEC 5-PRA and the fragilities associated with these scenarios have been refined to a point where it is considered cost effective (i.e., the effort invested in the fragility analysis for these scenarios matches the relative importance ofthe events).

8 . Rule-of-the-box A number of basic events in the CAFTA internal events model file refer to MSWs, MFWs, MFRVs, and MFR bypass valve actuators which are expected to be rule-of-the-box with the MSWs, MFIVs, MFRVs and MFR bypass valves themselves. A dedicated entry in the SEL is therefore not provided for these subcomponents.

Page 47 of 92 to ULN RC-06524

9. Rule-of-the-box A number of basic events in the CAFTA file refer to limit switches associated with MOVs. The limit switches do not normally have a dedicated tag number and they are expected to be rule-of-the-box within the MOVs themselves. A dedicated entry in the $EL is therefore not provided for these subcomponents.

10. Mission Time The CEC S-PRA uses the same overall mission time used by the IE PRA (i.e., 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />). This is a critical assumption that allows the use of the same set of event trees originally developed for the IE PRA. This is based on the inherent assumption that offsite power is recovered within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> from the event. The CEC $-PRA does not credit the LOOP recovery that is credited in the IE PRA, but the overall mission time of the TE PRA inherently assumes that LOOP is fully recovered within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. While this is a reasonable assumption for seismic events of lower magnitude (considering that event for the Mineral 5.8 Earthquake, which was relatively severe, offsite power was recovered at the station in the evening of the day of the event, thus roughly 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> afier the event), it may be less realistic for a seismic event of very high magnitude (at Fukushima, where maj or infrastructure damage was also caused by the tsunami, offsite power was restored 1 1 days after the event). One can therefore conclude that the epistemic uncertainty associated with this assumption is higher for seismic-induced events that are high-consequence events (e.g., Large LOCA) and is of less importance for lower magnitude and lower consequence events. Maintaining the same overall mission time is a practical approach that is consistently used in the current practice and an alternative approach of re-evaluating the mission time for each individual sequence and component/system may require extensive logic change and generation of new success criteria thermal-hydraulic simulations. Extending the mission time beyond 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> would also introduce significant departure from the as-operated condition of the plant because the longer the time from event and the higher the possibility that damaged equipment may be repaired or that non-proceduralized activities may be performed. Repairs and non-proceduralized are not normally included in the PRA. This would therefore introduce significant uncertainties in the fidelity and realism of the seismic risk profile. The assumption that the same mission time is applicable, is supplemented by extending the definition of the SEL; for example including equipment needed for refueling ofthe DG (i.e., the $-PRA does not solely rely on the DG day tank).

The CEC FLEX Integrated Plan describes the site program for coping with earthquakes requiring mitigation beyond 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. While, conservatively, the internal events and S-PRAs do not credit FLEX implementation, the existence of the systematic and regulated FLEX process increases confidence that the plant can mitigate longer-term accident conditions, and this is especially true for the low to mid-level ground motion events for which the $-PRA does not already consider to have resulted in core damage and release.

1 1 Seismic Initiating Event Tree Development of the seismic initiating event tree ($IET) involves a ranking of seismic initiating events from greatest to least in terms of potential risk significance, with the purpose of ensuring each ground motion level is assigned to the most challenging initiating event that could be credibly induced by that ground motion level. This ranking involves judgment and is a source of epistemic uncertainty. For example, while large LOCA occurs earlier in the SIET than small LOCA, it is not immediately clear that seismic-induced large LOCA is more challenging that a small LOCA in terms of risk significance. The actual risk significant of these induced initiating events depends on the mitigating systems they demand and the fragilities for those systems. One consequence of the SIET initiating events being disordered is that a larger fraction of the seismic frequency may be inappropriately apportioned to a plant impact of lesser consequence. This source of epistemic uncertainty may be assessed via sensitivity studies using the 5-PRA model.

12. Accident Sequence Models, Success Criteria, and System Models The CEC internal events model includes internal events initiator basic events in the accident sequence modeling and also in some cases within system models to address conditional system dependencies and conditional Page48 of 92

Attachment 2 to ULNRC-06524 dependencies for I&C signals. Seismic induced initiating events are included into the model based on their corresponding internal event initiating event. This introduces a source ofuncertainty since the internal event logic may be limited in addressing the uniqueness of a seismic induced event that could satisfy multiple internal events at the same time (e.g., Loss of offsite power and very small LOCA at the same time).

13 HRA The selection of breaking points for the seismic HRA plant damage bins are lefi to interpretation andjudgement ofthe analyst. The definition ofPlant Damage Bin SF13 describes the boundary between Plant Damage Bins SH3 and $H4 as a source of uncertainty. As it is observed that the majority of the fragilities with median capacity between O.4g and O.6g are actually concentrated in the O.5g to O.6g range, a sensitivity can be performed in the quantification notebook to evaluate the epistemic uncertainty associated with the boundary between SH3 and $H4 pushing the threshold as low as O.5g PGA. It is assumed that the long time frame branches in the EPRI seismic HRA decision tree can be used for the seismic version of operator action OP-XHE-FO ECLRS2. This is based on the fact that this is an action that takes place more than 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> after the period of strong shaking and also because of the fact that the internal events HRA indicates a potential longer time frame for recovery, which is currently not credited in the internal events model.

This represents a partial refinement of this operator action (i.e., as opposed to manipulating the HRA calculator file).

14. Level 1 and 2 Linkage The interface between Level 1 and Level 2 is assumed applicable to the S-PRA without changes. This assumption is supported by the fact that there are no late releases with containment failure timing close to the 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> timing used to define LERF in the internal events Level 2 analysis. It is nevertheless recognized that, because the potential LERF re categorization timing is essentially dependent on offsite conditions it has inherent uncertainties.

15 Spatial Interaction The spatial interaction between the block walls at elevation 2016 of the Control Building are not explicitly modeled as impacting the NK equipment in that area. This is based on the significant difference between the functional HCLPF ofthe modeled components and the HCLPF of the block walls, which is similar in turn to the direct to CD and direct to LERF fragility associated with soil failure. It is therefore judged that no significant completeness uncertainties are added by this model approach while on the other hand the model is maintained less heavy. .

Quantification -

The 5-PRA quantification method itself is a recognized limitation. The CEC 5-PRA model discretizes the hazard into 10 intervals, calculates SSC failure probabilities at each interval, and CCDP/CLERP at each interval using the plant response model. In addition, the minimal cutset upper bound approximation used by FTREX to calculate CCDP/CLERP from cutsets can be over-conservative where the conditional probabilities exceed 0. 1 as is the case for the dominant sequences. To minimize, but not eliminate, this conservatism, ACUBE is applied to all CDF and LERF cutsets at each hazard interval. Seismic HRA dependency is explicitly assessed and modeled for CDF and LERF.

5.4 SCDF Results This section presents the base SCDF results, a list of SSCs that are significant contributors, including risk importance measures, and a discussion of significant cutsets The total point estimate SCDF is 4.86E-06 per reactor year (note that ACUBE was used to post-process the CAFTA cutsets, and all cutsets are post-processed through ACUBE). Table 5-1 provides the PGA, earthquake occurrence frequency, truncation limit, ACUBE CCDP, ACUBE CDF, and contribution for.

each ground motion interval. Note that the reported SCDF includes a 0.90 plant availability factor. The dominant PGA range is between 0.4 and 0.5g.

Page 49 of 92 to ULNRC-06524 Table 5-1 : Contribution to SCDF by Acceleration Level PGA Frequency . ACUBE ACUBE Interval Truncation °° CUF (g) (/yr) CCDP CDF

%GO1: O.lg to O.2g 0.14 8.02E-04 1.OOE-09 1.30E-04 1.05E-07 0.22%

%G02: 0.2g to 0.3g 0.24 1.80E-04 5.OOE-09 8.12E-03 1.46E-06 3.0%

%G03: 0.3g to 0.4g 0.35 6.48E-05 5.OOE-09 9.64E-02 6.25E-06 12.9%

%G04: 0.4g to 0.42 1.67E-05 5.OOE-09 3.1 1E-Ol 5.20E-06 10.7%

0.45g

%G05: 0.45g to 0.47 1.16E-05 1.OOE-08 5.28E-01 6.13E-06 12.6%

0.5g

%G06: 0.5gto 0.52 8.13E-06 5.OOE-08 7.78E-01 6.33E-06 13.0%

0.55g

%G07: 0.55gto 0.57 5.87E-06 2.OOE-07 8.12E-01 4.76E-06 9.8%

0.6g

%G08: 0.6gto 0.7g 0.65 7.60E-06 5.OOE-07 8.60E-01 6.54E-06 13.5%

%G09: 0.7g to 0.8g 0.75 4.45E-06 5.OOE-07 8.81E-01 3.92E-06 8.1%

%G10: >0.8 0.88 8.85E-06 2.OOE-06 8.94E-01 7.92E-06 16.3%

Total 4.$6E-05 Table 5-2 identifies the top 10 SCDF cutset groups. Note that the CEC seismic cutset reviews are performed by first identifying the unique cutsets (independent ofground motion interval). Then a combined frequency

( based on the MCUB calculation) for each unique cutset is totaled across all ground motion intervals, and the unique cutsets are ordered by decreasing combined frequency. This approach allows better prioritizing the review of significant cutsets. Note that post-processing of cutsets with ACUBE does not output new cutsets but only updated frequencies. Therefore, the cutset review is performed on the CAFTA cutset file.

Page 50 of 92 to ULNRC-06524 Table 5-2: Summary of Top 10 CUF Cutsets ID Cutset Description FL-DGA-fTS FL-DGB-FTS FL-SHLD-S CDF Sequence T(1S)507P: A seismic-FL-S-Ti I FL-Ti -508 FL-Ti SAE-S 1 0 FL- induced loss of offsite power occurs.

Ti 5-507 PAF I RELAYO. i 8DG Both diesel generators (NEOi and RELAY_O. 1 8DG-CFF I SF-IE-Ti I SF-IF-Ti - NEO2) fail to start due to relay failures CFF COMBINATION_15 NE-XHE-FO- for automatic start of the EDGs EDG-RLYSET I OP-XHE-FO-AEPS1 combined with the operator failing to start the diesel generators following a seismic event. No AC power is available to NB buses due to the operators failing to align AEPS. The operator fails to manually operate the TDAFP following battery depletion resulting in core damage.

2 FL-DGA-FTR FL-DGB-FTR FL-ESWA- CDF Sequence T(i S)507P: A seismic-FAILS FL-ESWB-FAILS FL-SHLD-S induced loss of offsite power occurs.

FL-S-Ti FL-Ti -508 FL-Ti SAE-SiO FL- Both diesel generators (NEOi and T1S-507 I PAF SF-IE-T1 SF-IE-Ti-CFF NEO2) fail due to the loss of service SF-NGXC-i I SF-NGXC-i-CFF I water cooling which is the result of failure of motor control centers NGO5E OP-XHE-FO-AEPS i OP-XHE-FO-and NGO6E. No AC power is available AEPSi REC to NB buses due to the operators failing to align AEPS. The operator fails to manually operate the TDAFP following battery depletion resulting in core damage.

3 FL-DGA-FTR FL-DGB-FTR FL-SHLD-S CDF Sequence T(i S)507P: A seismic-FL-S-Ti I FL-Ti-508 I FL-TiSAE-SiO I FL- induced loss of offsite power occurs.

Ti 5-507 PAF SF-IE-T1 SF-IE-Ti -CFF I Both diesel generators (NEOi and SF-XNGX- i SF-XNGX- i -CFF NEO2) fail due to the loss of room cooling and/or low fuel flow which is OP-XHE-FO-AEPS1 I OP-XHE-FO-the result of seismic-induced failures of AEPS 1 REC transformers XNGO3 and XNGO4. No AC Power is available to NB buses due to the operators failing to align AEPS.

The operator fails to manually operate the TDAFP following battery depletion resulting in core damage.

Page5i of92 to ULNRC-06524 Table 5-2: Summary of Top 10 CUF Cutsets ID Cutset Description 4 FL-DGA--FTR FL-DGB-FTR FL-ESWA- CDF Sequence T(1 $)$07P: A seismic-FAILS FL-ESWB-FAILS FL-SHLD-S induced loss of offsite power occurs.

FL-S-Ti I FL-Ti -508 FL-T 1 SAE-S 1 0 I FL- Both diesel generators (NEO 1 and Ti$-507 I PAP SF-IE-Ti SF-IE-Ti-CFF NEO2) fail due to the loss of service SF-NGXC-2 SF-NGXC-2-CFF water cooling which is the result of OP-XHE-FO-AEPS i OP-XHE-FO- failure of motor control center NGO1A AEPS 1_REC and NGO2A. No AC power is available to NB buses due to the operators failing to align AEPS. The operator fails to manually operate the TDAFP following battery depletion which results in core damage.

5 FL-NOFP$ FL-S3-508 FL-S-$3 I FL-S-Ti I CDF Sequence S(3)508P: A seismic-PAF SF-EEGOiX I SF-EEGO1X-CFF SF-induced loss ofoffsite power occurs and IE-S3 SF-IE-53-CFF SF-IF-Ti SF-IF-Ti postulated to cause a seismic-induced CFF OP-XHE-FO-LTLOCA OP-XHE-FO- VSLOCA. Both CCPs fail due to insufficient room cooling which is a LTLOCA_REC result of seismic-induced failure of room cooler fans SGL12A and SGLi2B. Both SIPs fail due to insufficient room cooling which is a result of seismic-induced failure of room cooler fans SGLO9A and SGLO9B. The operator fails to perform long term cooldown and depressurization which results in core damage.

6 FL-$3-S08 FL-S-S3 FL-S-Ti I CDF Sequence S(3)508P: A seismic-PAF SF-IE-S3 SF-IE-S3-CFF SF-IE-Ti induced loss of offsite power occurs and is postulated to cause a seismic-induced SF-IE-T1-CFF SF-SGLXXX SF-VSLOCA. Both CCPs and SIPs fail due SGLXXX-CFF OP-XHE-FO-LTLOCA OP-XHE-FO-LTLOCA_REC to seismic-induced loss of CCW heat exchangers EEGO 1 A and EEGO 1 B.

The operator fails to perform long term cooldown and depressurization which results in core damage.

Page 52 of 92 to ULNRC-06524 Table 5-2: Summary of iop 10 CUF Cutsets ID Cutset Description 7 FL-DGA-FTR FL-DGB-fTR FL-ESWB- CDF Sequence T(1 S)SO7P: A seismic-FAILS FL-SHLD-S FL-S-Ti FL-Ti-508 I induced loss of offsite power occurs.

FL-Ti SAE-SlO FL-Ti 5-507 I PAP Both diesel generators (NEO1 and RELAYO.27CB I RELAYO.27CB-CFF I NEO2) fail due to the loss of room SF-IE-Ti I SF-IE-Ti-CFF I SF-NGO2 SF- cooling and/or low fuel flow which is a NGO2-CFF OP-XIIE-FO-AEPS 1 OP-XHE- result of seismic-induced failure of FO-AEPS1REC common relays that control transformers XNGO3 andXNGO4. No AC power is available to NB buses due to the operators failing to align AEPS.

The operator fails to manually operate the TDAFP following battery depletion which results in core damage.

8 FL-DGA-FTR FL-DGB-FTR I FL-SHLD-S I CDF Sequence T(l S)507P: A seismic-FL-S-Ti FL-Ti-S08 FL-T1SAE-SiO FL- induced loss of offsite power occurs.

Ti 5-507 I PAF I RELAYO.23 Both diesel generators (NEO 1 and RELAYO.23-CFF RELAYO.27CB I NEO2) fail due to the loss of room RELAYO.27CB-CFF I SF-IE-Ti SF-IF-Ti - cooling andlor low fuel flow which is a CFF OP-XHE-FO-AEPS 1 I OP-XHE-FO- result of seismic-induced failure of AEPSiREC independent relay failures that control transformers XNGO3 and XNGO4. No AC power is available to NB buses due to the operators failing to align AEPS.

The operator fails to manually operate the TDAFP following battery depletion which results in core damage.

9 FL-DGA-FTR FL-DGB-FTR FL-ESWB- CDF Sequence T(i S)SO7P: A seismic-FAILS FL-SHLD-S I FL-S-Tl FL-Ti-S08 induced loss of offsite power occurs.

FL-T1SAE-SlO FL-T1S-S07 PAF Diesel generator NEO1 fails due to low RELAYO.27AB I RELAYO.27AB-CFF fuel flow which is a result of seismic-SF-IF-Ti SF-IE-Ti-CFF SF-NGO2 SF- induced failure of fuel transfer pump NGO2-CFF OP-XHE-FO-AEPSi OP-XHE- PJEO1A. Diesel generator NEO2 fails FO-AEPSiREC due to the loss of service water cooling which is a result of seismic-induced failure of Breaker NGO2O1 No AC.

power is available to NB buses due to the operators failing to align AEPS.

The operator fails to manually operate TDAFP following battery depletion which results in core damage.

Page 53 of 92 to ULN RC-06524 Table 5-2: Summary of Top 10 CDF Cutsets ID Cutset Description 10 FL-DGA-FTR FL-DGB-FTR FL-ESWA- CDF Sequence T(1S)SO7P: A seismic-FAILS FL-SHLD-S FL-S-Ti FL-T1-508 I induced loss of offsite power occurs.

FL-T1SAE-S1O FL-T1S-507 PAF Both diesel generators NEO1 and NEO2 RELAYO.23 RELAYO.23-CFF Sf-IE-T1 fail due to the seismic-induced failure of SF-IE-T1-CfF SF-NGO1 Sf-NGO1-CFF 480V load center and failure of fuel OP-XHE-FO-AEPS1 OP-XHE-FO- transfer pump PJEO1B which is a result AEPS1_REC of seismic-induced relay failure. No AC power is available to the NB buses due to the operators failing to align AEPS which is a result of operators failing to align AEPS. The operators fail to operate TDAFP which results in core damage.

Notes:

1 . Events with a suffix -CFF are correlation factors (set to 1 .0) and are added by FRANX.

Table 5-3 summarizes the fragilities with a CDF Fussell-Vesely importance of greater than 2%. The Fussell-Vesely for each fragility was approximated by assuming the component was rugged then recalculating the failure probability at each ground motion interval, imposing those failure probabilities onto the cutsets, recalculating the CDf by ACUBE, and finally calculating the percent reductions of total CDF that the improved capacity affords.

Offsite power (Sf-IE-Tl) has the greatest importance, which is due to its low assumed capacity, its assumed non-recoverability, and the significant plant impact that losing offsite power creates. Seismic-induced very small LOCA (SF-IE-53) because of its low capacity and that it is assumed to occur simultaneously with any seismic event. Seismic-induced failure ofthe main control board (SF-RLOXX) is significant due to its low capacity and its conservatively assumed impact on operator actions following a seismic event. Note that non-seismic failures generally do not contribute significantly to SCDF.

Page 54 of 92 to ULNRC-06524 Table 5-3: SCDF Importance Measures Ranked by Fussell-Vesely Fragility Description F-V Am (g) I r Failure Mode Method SF-IE-T1 Seismic-Induced Loss of Offsite Power 0.53 0.3 1 .OE-07 0.55 Yard-Centered Generic Loss of Offsite (Conservative Power Estimate)

SF-IE-S3 Seismic-Induced Very Small Break 0.06 1 0.5 1 0.4 1 .OE-05 Structural failure SPRAIG Section LOCA ofsmall impulse 5.4.4.2 Option 3-lines leading to HCLPF based on RCS pressure Safe Shutdown boundary breach Earthquake Sf-RLOXX Seismic-Induced Failure of the Main 0.054 0.63 0.32 0.24 Functional CDFM (Surrogate Control Board Element)

Relay_O. 1 8DG Relay Fragility Group 0.044 0.72 0.27 0.58 Chatter SOV SF-IF-SW Seismic-Induced Failure of Service Water 0.039 0.24 0.38 0.24 Loss ofNon- Generic Nuclear Safety (Conservative Equipment Estimate)

SF-NNOX Seismic-Induced Failure of 120 VAC 0.023 0.914 0.38 0.24 Functional CDFM Based on Seismic Test Data SF-FR-YDXFR Seismic Rupture ofthe Yard Transformer 0.023 0.24 0.38 0.24 Loss ofNon- Generic Nuclear Safety (Conservative Equipment Estimate)

SF-SB1O2X Seismic Failure ofW Cabinet for reactor 0.02 0.91 0.38 0.24 functional CDfM Based on trip switchgear train A/B Seismic Test Data Relay_O.33 Relay Fragility Group 0.02 0.84 0.32 0.24 Chatter CDfM Page 55 of 92 to ULNRC-06524 Table 5-4 identifies the relative contribution of each initiator to total CDF for those initiators comprising the top 95% of CDF, or individually contributing greater than 1% to total CDF. Note that this list doesnt account for consequential initiating event contributors.

Table 5-4: Relative Contribution of Each Initiator to Total SCUF Initiator Contribution Loss ofOffsite Power 39.4%

Very Small LOCA 34.7%

Loss ofSeal Cooling 12.3%

ATWS 8.8%

Direct to Core Damage 3.3%

Loss of Service Water 1%

Page 56 of 92

Attachment 2 to ULNRC-06524 5.5 SLERF Results This section presents the base SLERF results, a list of S$Cs that are significant contributors, including risk importance measures, and a discussion of significant cutsets.

The total point estimate SLERF is 3.41E-06 per reactor year (note that ACUBE was used to post-process the CAFTA cutsets, and all cutsets are post-processed through ACUBE). Table 5-5 provides the PGA, earthquake occurrence frequency, truncation limit, ACUBE CLERP, ACUBE LERF, and contribution for each ground motion interval. Note that the reported SLERF includes a 0.90 plant availability factor. The largest individual contribution to SLERF is from interval %G06, which represents ground motion between 1.Ogto 1.2g.

Table 5-5: Contribution to SLERF by Acceleration Level PGA Frequency . ACUBE ACUBE Interval Truncation °° -

(g) (/yr) CLERP LERF

%GO1 : O.lg to O.2g 0.14 8.02E-04 1 .OOE-1 1 1 .4E-06 1 .09E-09 0.0%

%G02: 0.2gto 0.5g 0.32 2.73E-04 1.OOE-09 1.OE-04 2.73E-08 0.8%

%G03: 0.5g to 0.6g 0.55 1 .40E-05 1 .OOE-09 1 .1E-02 1 .48E-07 4.3%

%G04: 0.6gto 0.8g 0.69 1.21E-05 2.OOE-09 3.3E-02 4.OOE-07 11.7%

%G05: 0.8g to ig 0.89 4.49E-06 5.OOE-09 1.3E-01 5.76E-07 16.9%

%G06: lg to 1.2g 1.10 1.89E-06 1.20E-08 3.1E-01 5.92E-07 17.4%

%G07: 1.2gto 1.4g 1.30 1.O1E-06 4.OOE-08 5.4E-0l 5.45E-07 16.0%

%G08: 1.4g to 1.6g 1.50 5.68E-07 8.OOE-08 6.7E-0l 3.80E-07 11.1%

%G09: 1 .6g to 2g 1 .79 5.29E-07 2.OOE-07 8.OE-01 4.23E-07 12.4%

%G10: >2g 2.20 3 .63E-07 2.OOE-07 8.8E-Ol 3.20E-07 9.4%

Total 3.41E-06 Table 5-6 identifies the top 10 $LERF cutset groups. Note that the CEC seismic cutset reviews are performed by first identifying the unique cutsets (independent ofground motion interval). Then a combined frequency (based on the MCUB calculation) for each unique cutset is totaled across all ground motion intervals, and the unique cutsets are ordered by decreasing combined frequency. This approach allows better prioritizing the review of significant cutsets. Note that post-processing of cutsets with ACUBE does not output new cutsets but only updated frequencies. Therefore, the cutset review is performed on the CAFTA cutset file.

Table 5-6: Summary of Top 10 SLERF Cutsets ID Cutset Description 1 PAF SF-SOIL SF-SOIL-CFF Seismic-induced soil failure occurs and leads directly to LERF.

2 PAF SF-NSSG I SF-NSSG-CFF Seismic-induced failure of the steam generators occurs and leads directly to LERF.

Page 57 of 92 to ULNRC-06524 Table 5-6: Summary of Top 10 SLEPT Cutsets ID Cutset Description 3 FL-DEP FL-S3-$08 FL-$-$3 FL-$- LERF Sequence CEll $25P and CDF Ti EARLY NOBYPASS PAF SF- Sequence FL-53-508: A seismic-induced IE-53 SF-IE-53-CFF SF-IE-Ti SF- loss of offsite power occurs and is IE-T1-CFF SF-RB-PEN SF-RB-PEN- postulated to cause a seismic-induced CFF SF-SGLXXX SF-SGLXXX-CFF VSLOCA. The CCPs and SIPs fail due to seismic-induced failure of room coolers SGL12A, SGL12B, and SGLO9A and SGLO9B. The operators fail to cooldown and depressurize. Containment Isolation fails due to seismic-induced failure of the reactor building penetrations which results in LERF.

4 FL-DEP I FL-NOFPS FL-53-508 FL- LERF Sequence CET1 525P and CDF 5-53 FL-S-Ti EARLY NOBYPASS Sequence FL-53-S08: A seismic-induced PAF SF-EEGOiX SF-EEGO1X-CFF loss of offsite power occurs and is SF-IE-53 SF-IE-53-CFF SF-IE-Ti postulated to cause a seismic-induced SF-IE-T1 -CFF SF-RB-PEN SF-RB- VSLOCA. The CCPs and SIPs fail due to PEN-CFF seismic-induced failure of the CCW heat exchangers. The operators fail to cooldown and depressurize. Containment Isolation fails due to seismic-induced failure of the reactor building penetrations which results in LERF.

5 FL-DGA-FTR FL-DGB-FTR FL- LERF Sequence CET2S2OP and CDF ESWA-FAILS FL-SHLD-S FL-S-Ti Sequence FL-Ti 5-507: A seismic-induced FL-Ti-S08 FL-T1SAE-SiO FL-T1S- loss of offsite power occurs. Both diesel 507 EARLY NOBYPASS PAF generators (NEO1 and NEO2) fail due to the SFLIETi SF-IE-Ti-CFF SF-NGO1 seismi-induced failure of48OV load center SF-NGO 1 -CFF SF-RB-PEN SF-RB- NGO 1 The operators fail to manually align PEN-CFF OP-XHE-FO-AEPS1 OP- AEPS. The operators fail to manually XHE-FO-AEPS 1REC operate TDAFP after battery depletion.

Containment Isolation fails due to seismic-induced failure of the reactor building penetrations which results in LERF.

6 FL-DGA-FTR FL-DGB-FTR FL- LERF Sequence CET2S2OP and CDF SHLD-S FL-S-Ti I FL-Ti-SOS FL- Sequence FL-TiS-507: A seismic-induced T1SAE-SlO FL-T1S-S07 loss of offsite power occurs. Both diesel generators (NEO1 and NEO2) fail due to EARLY NO BYPASS PAF seisimc-induced relay failure. The RELAY O.27AB RELAY O.27AB-operators fail to manually align AEPS. The CFF SF-IE-T1 SF-IE-Ti-CFF SF-operators fail to manually operate TDAFP RB-PEN SF-RB-PEN-CFF after battery depletion. Contamment OP-XHE-FO-AEPS1 OP-XHE-FO- Isolation fails due to seismic-induced AEPS1REC failure of the reactor building penetrations which results in LERF.

Page 58 of 92 to ULNRC-06524 Table 5-6: Summary of Top 10 SLERF Cutsets ID Cutset Description 7 FL-DGA-FTR FL-DGB-FTR FL- LERF Sequence CET2S2OP and CDF SHLD-$ FL-S-Ti FL-Ti-508 I FL- Sequence FL-Ti 5-507: A seismic-induced TiSAE-SiO I FL-TiS-507 loss of offsite power occurs. Both diesel generators (NEOi and NEO2) fail due to EARLY I NOBYPASS I PAF I seismic-induced relay failure. The RELAYO.27CB RELAY O.27CB-operators fail to manually align AEPS. The CFF I SF-IE-Ti I SF-IE-Ti-CFF I SF-operators fail to manually operate TDAFP RB-PEN SF-RB-PEN-CFF OP-XHE-after battery depletion. Containment FO-AEPS1 OP-XHE-FO-AEPS1REC Isolation fails due to seismic-induced failure of the reactor building penetrations which results in LERF.

8 FL-S-Ti FL-Ti-503 I EARLY LERF Sequence CET1S25P and CDF FLEXAFWFAIL I NOBYPASS PAF Sequence FL-Ti-S03: A seismic-induced RELAYO.27AB RELAYO.27AB- loss of offsite lower occurs. AFW fail due CFF SF-IF-Ti SF-IE-Ti-CFF SF- to FLEX AFW failure, operators failing to RB-PEN SF-RB-PEN-CFF control SG level after a complex event, and seismic-induced relay failure. The CCPs AL-XHE-FO-SBOSGL AL-XHE-FO-and SIPs fail due to seismic-induced relay SBOSGLREC failure. Containment Isolation fails due to seismic-induced failure of the reactor building penetrations which results in LERF.

9 FL-S-Ti FL-Tl-S03 LERF Sequence CET1S25P and CDF Sequence FL-Ti-503: A seismic-induced EARLY FLEXAFWFAIL loss ofoffsite power occurs. AFW fails due NOBYPASS PAP RELAYO.27CB I to FLEX AFW failure, operators failing to RELAYO.27CB-CFF SF-IF-Ti SF-IE-Ti-CFF SF-RB-PEN SF-RB-PEN- control SG level after a complex event, and seismic-induced relay failure. The CCPs, CFF JAL-XHE-FO-SBOSGL AL-and SIPs fail due to seismic-induced relay XHE-FO-SBOSGLREC failure. Containment Isolation fails due to seismic-induced failure of the reactor building penetrations which results in LERF.

10 FL-S-Ti FL-Tl-S03 EARLY LERF Sequence CET1S25P and CDF FLEXAFWFAIL NO_BYPASS PAF Sequence FL-Ti-S03: A seismic-induced SF-IF-Ti SF-IE-Tl-CFF SF-NGOi loss ofoffsite power occurs. AFW fails due SF-NGO i -CFF I SF-RB-PEN I SF-RB- to FLEX AFW failure, operators failing to PEN-CFF AL-XHE-FO-SBOSGL I AL- control SG level after a complex event, and XHE-FO-SBOSGLREC seismic-induced relay failure. The CCPs and SIPs fail due to seismic-induced failure of 480V load center NGOi. Containment Isolation fails due to seismic-induced failure of the reactor building penetrations which results in LERF.

Page 59 of 92 to ULN RC-06524 Table 5-6: Summary of Top 10 SLERF Cutsets ID Cutset Description Notes:

1 . Events with a suffix -CfF are correlation factors (set to 1 .0) and are added by FRANX.

Table 5-7 summarizes the fragilities with a LERF Fussell-Vesely importance of greater than 2%. The Fussell-Vesely for each fragility was approximated by multiplying the median capacity by a factor of 1 ,000,000, then recalculating the failure probability at each ground motion interval, imposing those failure probabilities onto the cutsets, recalculating the LERF by ACUBE, and finally calculating the percent reduction of total LERF that the improved capacity affords.

Fragilities most important to LERF are those representing SSCs whose failure affects containment isolation and those failures that lead directly to LERF. Fragilities such as Seismic-induced soil failure (SF-SOIL) and seismic-induced failure of the steam generator support (SF-NSSF) lead directly to LERF. Whereas seismic-induced failure of the reactor building penetrations (SF-RB-PEN) result in contaimnent isolation failure. Note that random (non-seismic) failures generally do not contribute significantly to the seismic LERF.

Page 60 of 92 to ULNRC-06524 Table 5-7: $LERF Importance Measures Ranked by Fussell-Vesely Fragility Description F-V Am (g) r Failure Mode Method SF-SOIL Seismic-Induced Soil Failure 19.5% 1.67 0.26 0.24 Soil Bearing Capacity CDFM SF-NSSG Seismic-Induced Failure ofthe Steam 19.4% 1.86 0.42 0.09 Structural (SG SOV Generator Supports Column)

SF-RB-PEN Seismic-Induced Failure of the Reactor 14.6% 1 .78 0.26 0.24 Shear Failure CDFM Building Penetrations SF-IF-Ti Seismic-Induced Loss ofOffsite 9.3% 0.3 1.OE-07 0.55 Yard-Centered Loss of Generic Power Offsite Power (Conservative Estimate)

SF-F-AB-2047-F Seismic-induced Flooding from Fire 8.8% 1 .44 0.32 0.24 Boundary Pressure CDFM Protection Piping in the Auxiliary Failure Building 2047 level SF-CC Seismic-Induced Failure ofthe 3.0% 2.21 0.26 0.24 Impact with SC-I CDFM Communications Corridor Structures SF-SBOXX-2 Seismic-Induced Failure ofRP Cabinet 2.6% 0.74 0.38 0.24 Functional CDFM Based on Seismic Test Data SF-NNOX Seismic-Induced Failure of 120 VAC 2.2% 0.914 0.38 0.24 Functional CDFM Based on Distribution Panels NNO 1 NNO2, Seismic Test Data NNO3 and NNO4 SF-IE-S3 Seismic-induced very small break 2.2% 0.5 1 0.4 1.OE-05 Structural failure of SPRMG Section LOCA small impulse lines 5.4.4.2 Option 3 leading to RCS HCLPF based on pressure boundary Safe Shutdown breach Earthquake SF-SBOXX-l Seismic-Induced Failure ofRP Cabinet 2. 1% 0.88 0.38 0.24 Functional CDFM Based on SBO29AJB/C/D, SBO32A/B/C/D Seismic Test Data RelayO. 1 8DG Relay Fragility Group 2.0% 0.72 0.27 0.58 Chatter SOV Page6l of92 to ULNRC-06524 5.6 S-PRA Quantification Uncertainty Analysis Parametric uncertainty in the S-PRA results originates from seismic hazard curve uncertainty, the $SC fragility uncertainties, and basic event failure parameter uncertainties from the internal events PRA.

Parametric uncertainty quantification was performed using the UNCERT code in conjunction with the FRANX sampling equation method. Table 5-8 documents the SCDF and SLERF uncertainty quantification results, using Latin Hypercube with 10,000 samples and ACUBE applied to all cutsets.

Table 5-8: Uncertainty Quantification Results

. Point . Standard Metric . Mean 5th Median 95th Estimate .

Deviation SCDF 1.37E-05 4.55E-05 2.55E-06 2.03E-05 1.65E-04 9.27E-05 SLERF 6.80E-07 3.32E-06 1.18E-07 1.24E-06 1.26E-05 8.05E-06 The error factor (defined as [95th/Median] or [Medianl5t9) for $CDF and SLERF is approximately eight (8) and ten (10) respectively, which is generally high compared to the internal events PRA. This is caused by relatively high fragility uncertainty, indicated by parameters fi and fir, as well as hazard uncertainty.

The relatively high error factor (as compared to the internal events PRA) resulting from the seismic uncertainty quantification is expected and consistent with industry operating experience.

Figure 5-1 and Figure 5-2 depict the probability density function and cumulative distribution function for

$CDF and $LERF. Note that the median was skewed to the right ofthe mode for seismic CDF and LERE.

This is not atypical for Seismic PRAs and is likely due to the high CDF contribution associated with higher g levels (i.e., hazard interval %GlO). This shows the impact of the larger uncertainty for the hazard at higher acceleration levels given that the CCDP (0.9) is relatively insensitive to uncertainty.

Page 62 of 92 to ULN RC-06524 UI I

/

I 1E-? 1E 1 -3 1,

7\

\

I /

1EC3 1Ea 1E-Ol Figure 5-1: Seismic CDF Uncertainty (10,000 sample LII and ACUBE on All Cutsets)

The resulting histogram cumulative distribution function indicates that there is approximately a 90%

likelihood that seismic CDf is less than 1E-04/yr and a 30% likelihood that seismic CDF is less than 1E-05/yr.

ff zJz EzZZ__

I :E-z___  :

, E7 *

• SE3 1E3 Fl G1--- - -_

E4 E 1 (I .. 1 Figure 5-2: Seismic LERF Uncertainty (10,000 sample LII and ACUBE on All Cutsets)

Page 63 of 92

Attachment 2 to ULN RC-06524 The resulting histogram distribution function indicates that there is approximately a 90% likelihood that seismic LERF is less than 1E-05/yr and nearly a 50% likelihood that seismic LERF is less than 1E-06/yr.

Model uncertainty is introduced when assumptions are made by the S-PRA model and inputs to represent plant response, when there may be alternative approaches to particular aspects of the modeling, or when there is no consensus approach for a particular issue. For the CEC S-PRA, the important model uncertainties are addressed through the sensitivity studies to determine the potential impact on $CDF or SLERF.

Completeness uncertainty relates to potential risk contributors that are not in the model. The scope of the CEC S-PRA is for at-power operation, and does not include risk contributors from low-power shutdown operation, or for spent fuel pooi risk. In addition, there may be potential issues related to factors that are not included, such as the impact of aging on equipment reliability and fragility. Other potential issues include impacts of plant organizational performance on risk, and unknown omitted phenomena and failure mechanisms. By their nature, the impacts on risk of these types of uncertainties are not known.

5.7 S-PRA Quantification Sensitivity Analysis The CEC S-PRA includes quantitative sensitivity studies for the following elements to assess model stability and sources of epistemic uncertainty:

. Truncation Limits for Model Convergence

. ACUBE Post-Processing

. Hazard Interval Study

. Non-Safety Component Fragility Sensitivity

. Mission Time Sensitivity

. On-Site FLEX Equipment Sensitivity

. Model Sensitivity to Seismic HRA .

. Model Sensitivity to Seismic HRA Bin Definitions ..

. Model Sensitivity to Ex-Control Room Actions

. Cumulative Sensitivity

. Plant Modification Sensitivity 5.7.1 Truncation Limits for Model Convergence In the framework of seismic PRA, truncation is only one ofthe aspects used to assess model stability, others being considerations on the extension ofthe hazards integration and size and number ofthe integration bins

( discussed in other sections). Truncation is more relevant at lower g levels (where additional cutsets can increase CCDP) and it becomes less and less relevant at higher g levels, with CCDP approaching 1 .0. At the top level, if CCDP is 1 .0, truncation is irrelevant.

QU-B3 of ASME/ANS RA-Sb-201 3 [201 specifically discusses assessing truncation against the overall model and provides examples of truncation limits for internal events. In traditional truncation studies performed for Internal Events PRA, cutsets are truncated until successive reductions in the truncation limit by one decade result in a total plant CDF and LERF which converge to a value less than 5% different from the previous quantification. For seismic PRAs, demonstrating convergence using the traditional definition of less than 5% change from the previous quantification can be challenging due to more complex models Page 64 of 92

Attachment 2 to ULNRC-06524 and computing and memory limitations and is unnecessary to prove model stability, based on the discussion above.

The truncation study is performed across each ground motion using a range oftruncation limits as opposed to a single truncation limit for all ground motions. The percent change shown per decade is calculated as the delta result for each hazard interval contribution in comparison to total CDF/LERF. ACUBE was applied to all cutsets for each ground motion interval. It can be estimated that the sum of each hazard interval contribution would represent the total change in CDF/LERF. Summing the values in Table 5-9 and Table 5-10 results in an overall change in CDF and LERF of -4 5% and l9% respectively.

The truncation study documented in Table 5-9 and Table 5-lOis performed across each ground motion using a range of truncation limits as opposed to a single truncation limit for all ground motions. The percent change shown per decade is calculated as the delta result for each hazard interval contribution in comparison to total CDF/LERF. ACUBE was applied to all cutsets for each ground motion interval. It can be estimated that the sum of each hazard interval contribution would represent the total change in CDf/LERF. Summing the values in Table 5-9 and Table 5-10 results in an overall change in CDF and LERF of 15% and 19%

respectively.

At low g levels, where the hazard interval bins contribute less than 5% to total CDF and LERF, decreasing the truncation limit further for these bins would not impact the overall risk insights from the 5-PRA model, even though additional insights at low g level may be gained through generation of additional cutsets (although it can be noted that at lower g level the seismic failures have lower probability and the insights tend to converge to the insights ofthe internal events PRA). At high g levels, where the CCDP/CLERP of the hazard interval bins is approaching 1 .0 (the PAF at 0.9 limits the CCDP to .9), no additional insights can be gained by decreasing truncation since regardless of the number of cutsets generated the numerical solution will not be impacted. Based on the assessment of each hazard interval as compared to total CDF/LERF it is shown that for each decrease in truncation, a significant impact to the total results from any one hazard interval does not occur (i.e., the percent change against total ACUBE CDF/LERF is less than 5% for each hazard interval). Due to computing and memory limitations, quantification below these truncation limits cannot be achieved at this point; however it is qualitatively assessed that the overall risk insights would not change, even iftruncation limits could be lowered for a given hazard interval bin. Based on the above considerations, it is judged that the model is sufficiently converged, consistent with the intention of thestandard.

=-

Table 5-9: Model Convergence on Truncation Value CUF -

Percent

. Truncation Change Scenario ACUBE CUF Delta . Notes Level Against Total ACUBE CDF

%GO1 l.OOE-08 5.93E-08 - -

l.OOE-09 1.05E-07 4.52E-08 0.09% (1)

%G02 5.OOE-08 1.03E-06 - -

5 .OOE-09 1 .46E-06 4.35E-07 0.90% (1)

%G03 5.OOE-08 4.55E-06 - -

5.OOE-09 6.25E-06 l.70E-06 3.49% (1)

%G04 5.OOE-08 3.82E-06 - -

Page 65 of 92 to ULNRC-06524 Table 5-9: Model Convergence on Truncation Value CUF -

Percent

. Truncation Change Scenario ACUBE CUF Delta . Notes Level Against Total ACUEE CDF 5.OOE-09 5.20E-06 1.37E-06 2.82% (1)

%G05 1 .OOE-07 4.77E-06 - -

1 .OOE-08 6 1 3E-06

. 1 .36E-06 2.79% (1)

%G06 5.OOE-07 5.36E-06 - -

5 .OOE-08 6.33E-06 9.70E-07 2.00% (1)

%G07 2.OOE-07 4.76E-06 - - (1)

CCDP=O.9 528E-06 5.19E-07 1.07% (2)

%G08 5.OOE-07 654E-06 - - (1)

CCDP=O.9 6.84E-06 3.04E-07 0.62% (2)

%G09 5.OOE-07 3.92E-06 - - (1)

CCDP=O.9 4.0 1 E-06 8 .56E-08 0. 1 8% (2)

%G1O 2.OOE-06 7.92E-06 - - (1)

CCDP=O.9 7.97E-06 4.96E-08 0.63% (2)

Notes

( 1) Due to computing and memory limitations, quantification below this truncation level was not achievable.

( 2) Due to computing and memory limitations, quantification below the previous truncation level for this interval was not achievable; therefore, the delta for this interval was calculated against the ACUBE CDF assuming a CCDP=O.9 (PAF).

Page 66 of 92

Attachment 2 to ULNRC-06524 Table 5-10: Model Convergence on Truncation Value LERF Percent

. Change

. Truncation ACUBE Scenario Delta Against Total Notes Level LERF ACUBE LERF

%GO1 1.OOE-1O 7.98E-1O - -

1.OOE-11 1.09E-09 2.95E-1O 0.01% (1)

%G02 1 .OOE-09 2.73E-08 - - (2) 1.OOE-1O 5.1OE-08 2.37E-08 0.70% (1)

%G03 1 .OOE-08 8.09E-08 - -

1 .OOE-09 1 .48E-07 6.74E-08 1 .97% (1)

%G04 2.OOE-08 2.97E-07 - -

2.OOE-09 4.OOE-07 1.03E-07 3.02% (1)

%G05 5.OOE-08 5.09E-07 - -

5 .OOE-09 5 .76E-07 6.63E-08 1 .94% (1)

%G06 1.20E-07 4.75E-07 - -

1 .20E-08 5 .92E-07 1 1 7E-07

. 3 .44% (1)

%G07 4.OOE-08 5.45E-07 - - (1)

CLERP=O.67 6.77E-07 1.32E-07 3.86% (3)

%G08 6.OOE-08 3.8OE-07 - - (1)

CLERP=O.8 4.54E-07 7.43E-08 2.18% (4)

%G09 2.OOE-07 4.23E-07 - - (1)

CLERP=O.9 4.76E-07 5 34E-08 1 .56% (5)

%G1O 2.OOE-07 3.20E-07 - - (1)

CLERP=O.9 3 .27E-07 6. 8 1 E-09 0.20% (5)

Page 67 of 92 to ULNRC-06524 Table 5-10: Model Convergence on Truncation Value LERF -

Percent

. Change

. Truncation ACUEE Scenario Delta AGainst Total Notes Level LERF ACUBE LERF Notes:

( 1) Due to computing and memory limitations, quantification below this truncation level was not achievable.

(2) Due to the quantification run time for interval %g02 at 1.OOE-lO, the l.OOE-09 run was chosen for the base case truncation level. This was judged to be acceptable due to the less than 1 % change in total seismic LERF between 1.OOE-09 and 1.OOE-lO.

(3) Due to computing and memory limitations, quantification below the previous truncation level for this interval was not achievable; therefore, the delta for interval %G07 was calculated against the ACUBE CDF assuming a CCDP =0.67 (the CCDP of interval %GO8).

(4) Due to computing and memory limitations, quantification below the previous truncation level for this interval was not achievable; therefore, the delta for interval %GO8 was calculated against the ACUBELERF assuming a CLERF =0.8 (the CLERP ofinterval %G09).

(5) Due to computing and memory limitation, quantification below the previous truncation level for this interval was not achievable; therefore, the delta for this interval was calculated against the ACUBE LERF assuming a CLERF = 0.9 (PAP).

5.7.2 ACUBE Post-Processing Due to the rapidly increasing memory requirements of ACUBE with larger numbers of cutsets it is not always possible to evaluate an entire cutset file with ACUBE. This typically results in an iterative process to determine the number of cutsets to run based on computing capability.

An evaluation was performed to illustrate the impact of ACUBE post-processing on the S-PRA results for CDF and LERF. ACUBE post-processing is performed on each ground motion interval scenario, consistent with the FRANX scenario proces-First, the evaluation was performed considering each individual PGA interval bin, and post-processed all of its cutsets with ACUBE. Following this evaluation, all cutsets for the grouped cutset files were processed with ACUBE. A summary is shown in Table 5-1 1 and in Figure 5-3.

Table 5-11: ACUEE Post Processing Evaluation for Seismic CDF and LERF Percent CUF Percent LERF ACUEE Cutsets Total CUF Total LERF Change Change 10 6.41E-05 - l.09E-05 -

500 5.80E-05 -9.5% l.08E-05 -1.5%

1000 5.22E-05 -10.0% 7.93E-06 -26.3%

All 4.86E-05 -6.9% 3.41E-06 -57.0%

Page 68 of 92 to ULNRC-06524 J . -

i__* - * -

3_-;*- -

- (T;_ *

-.- . -: flj: .. -. . , ,

- .I I

IMIJE-Cio I I ioI ACUBE Cu(stt Iüt-Pcesug

1. stii:: CDT I S2tL:c: ] [RI Figure 5-3: ACUBE Post-Processing Evaluation for CUF and LERF 5.7.3 Hazard Interval Study The CEC $-PRA discrefizes the seismic hazard for both CDF and LERF. For CDF, the hazard is discretized, starting at O.lg, into 10 intervals with varying widths (.lg for %GOlto %G03 and %G08 to %G1O, and 0.05 g for %G04 to %G07). for LERF, the hazard is discretized, starting at 0. ig, into 10 intervals, with varying widths (0.lg for , and %G03, 0.3g for %G02, and 0.2g for %G04 to %G10). Convolution ofthe mean hazard with the point estimate plant level fragilities for core damage and large early release confirmed that the CDF and LERF estimates are converged at the selected 10-interval hazard discretization.

5.7.4 Non-Safety Component Fragility Sensitivity Non-safety components basic events were assigned a generic fragility value and were assumed to be fully correlated. This treatment is generally regarded to be conservative. However, to ensure that the impact of crediting non-safety equipment with a generic fragility is not a significant contributor to seismic risk, a sensitivity analysis was performed by setting the associated N$CI fragility to true. The sensitivity case resulted in a 0.6% increase in CDf and a 0% change in LERF.

5.7.5 Mission lime Sensitivity A mission time of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> was used in the CEC S-PRA model consistent with the internal events model.

For this sensitivity study, all CEC basic events associated with a mission time were altered from the standard 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> mission time to a 48 hour5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> mission time. The sensitivity case resulted in a 1 .2% increase in CDF and a 0% change in LERF.

5.7.6 On-site FLEX Equipment Sensitivity For this sensitivity the potential impact of crediting FLEX was investigated. FLEX equipment has the potential to be incorporated into a plant given a severe accident on site and with some assumed conditions during the extreme event. The S-PRA model currently includes minimal credit for this equipment in the baseline model (flags set to 0.99). For this sensitivity the flags were modified and set to an optimistic value of 0.1 The sensitivity case resulted in a 4.8% decrease in CDF and a 0.8% decrease in LERF.

Page 69 of 92 to ULNRC-06524 5.7.7 Model Sensitivity to Seismic HRA Detailed seismic HRA was performed to target significant post-initiator actions during the model refinement process; however, many ofthe post-initiator actions are still assessed via the conservative EPRI screening method. Two evaluate the S-PRA models overall reliance on operator actions and highlight the benefit of detailed seismic HRA, two sensitivity studies were run: (1) setting all seismic bin HEPs to a value of one, and (2) setting all screening seismic bin HEPs to their nominal values, with the exception of bin $114. The sensitivity case resulted in a 5.8% decrease in CDF and a 1 .4% decrease in LERF. Note that since all HEPs for the $H4 bin are already set to one in the base case, the impact of these sensitivities is limited to CDF intervals %GO1 -%G07 and LERF intervals %GO1-%G03.

5.7.8 Model Sensitivity to Seismic HRA Bin Definition An inherent uncertainty associated with S-PRA development relates to the binning for the HEP seismic hazard groups to the number of seismic hazard bins modeled in the S-PRA. This sensitivity addresses three sensitivity cases to address the uncertainty regarding the cut-off definitions for the seismic HRA bins: (1) shifting cut-off breaking points down by applying the $112 cut-off (O.5g) for $111, applying the $113 cut-off (O.6g) for $112 and defining new cut-off of O.8g for S113, (2) shifting cut-off breaking points up by removing credit for $H1 applying the $111 cut-off (O.2g) for $112, and applying the $112 cut-off (O.5g) for SH3, and (3) extending cut-off breaking point for $113 by extending $H3 cut-off to 0.8g. The sensitivity case resulted in a 6.0% decrease in CDF and 1 .3% decrease in LERF for Case 1 a 171 .4% increase in CDF for CDF and a 5.9% increase for LERF for case 2, and a 0.1% reduction in both CDF and LERF for Case 3 Note that binning of the HEPs does have a noticeable impact on the results but the bin definitions used in the base case were defined to be realistic based on the relevant component fragilities used to defined plant damage.

5.7.9 Model Sensitivity to Ex-Control Room Actions For this sensitivity study all ex-control room actions are set to TRUE. This sensitivity cases resulted in a 7.6% increase in CDF and a 1 .0% increase in LERF.

5.7.10 Cumulative Sensitivity For this sensitivity study, the median capacity for fragility groups SF-IE-$3, SF-RLOXX, and $F-NNOX were doubled to assess the cumulative impact on the PRA model. Additionally, flags events (FLEXACTODCFATL and FLEXAFWFML) for FLEX equi5ment were modified to an optimistic value of 0.1 This sensitivity case resulted in a 1 8 .93% decrease in CDF. The model results were then passed through the fragility ranking tool to ensure that no masking of risk-significant contributors was occurring.

Review of the results show no masking of risk-significant contributors.

5.7.1 1 Plant Modification Sensitivity Currently the CEC S-PRA model credits the impact of two plant modifications which have not occurred.

For this sensitivity the pre-plant modification median capacities for fragilities SF-XPBO5-AM (O.28g) and

$F-F-CB-1974-$-AM (O.59g) were evaluated. This sensitivity case resulted in a 16.5% increase in CDF.

5.7.12 Summary of Sensitivity Study Results Table 5-12 summarizes the quantitative sensitivity study results discussed in preceding sections.

Table 5-12: Summary of Quantitative Sensitivity Study Results Study ACDF ALERF Non-$afety Component Fragility +0.6% 0%

Mission Time +1 .2% 0%

On-site FLEX Equipment -4.8% -0.8%

$eismic HRA -5.8% -1.4%

Page 70 of 92 to ULNRC-06524 Table 5-12: Summary of Quantitative Sensitivity Study Results Study ACDf ALERF Seismic HRA Bin Definition -6.0% -1.3%

(Case A)

Seismic HRA Bin Definition 1 71 .4% 5.9%

(Case B)

Seismic HRA Bin Definition -0.1% -0.1%

(Case C)

Ex-Control Room Actions +7.6% +1.0%

Cumulative Sensitivity -1 8.93% Not Evaluated Plant Modification +16.5% Not Evaluated 5.8 $-PRA Quantification Technical Adequacy The CEC S-PRA quantification methodology and analysis were subjected to an independent peer review against the pertinent requirements in the ASME/ANS PRA Standard [13]. The 5-PRA was peer reviewed relative to Capability Category II for the full set of requirements in the Standard. After completion of the subsequent independent assessment, the full set of supporting requirements was met with the exception of SPR-B2 and SPR-E6. The seismic plant response quantification analysis was determined to be acceptable to use in the 5-PRA. The peer review facts and Observations (F&O) through an independent assessment, is further described in Appendix A and References [14] and [15].

Page 71 of 92 to ULN RC-06524 6.0 Conclusions A S-PRA has been performed for Callaway Energy Center in accordance with applicable requirements of the ASME/ANS PRA Standard [13]. The S-PRA shows that the point estimate seismic CDF is 4.86E-05/yr and the seismic LERF is 3 .41E-06/yr. Uncertainty, importance, and sensitivity analyses were performed.

Sensitivity studies were performed to identify critical assumptions, test the sensitivity to quantification parameters and the seismic hazard, and identify potential areas to consider for the reduction of seismic risk.

The model was used to identify potential plant modifications that would benefit the 5-PRA risk metric results. Two plant modifications were specifically committed to based upon the results of the 5-PRA, the AEPS transformer anchorage modification (SF-XPBO5), and a plant modification to double the gap between the sprinkler head and the insulation (SF-F-CB-l 974-5). These plant modifications are credited in the base model S-PRA results. These sensitivity studies demonstrated that the model results were robust to the modeling and assumptions used.

Table 6-1 : Regulatory Commitments and Completion Schedule Modification Completion Date AEPS Transformer Anchorage Modification Prior to plant startup from Refueling Outage 24 Sprinkler Head Clearance Prior to plant startup from Refueling Outage 24 Page 72 of 92

Attachment 2 to ULN RC-06524 7.0 References

[1] USNRC (E Leeds and M Johnson) Letter to All Power Reactor Licensees et al., Request for Information Piwztant to Title 1 0 of the Code of federal Regulations 50. 54(f) Regarding Recommendations 2.1, 2.3, and 9.3 of the Near-Term Task Force Review oflnsights from the Fukushima Dai-Ichi Acciclent March 1 2, 20 1 2. NRC ADAMS ML12053A340.

[2] RIZZO International Inc., 20 1 9, Probabilistic Seismic Hazard Analysis, Seismic Probabilistic Risk Assessment Project, Callaway Energy Center, Unit 1 Project No: 1 l-4695B, Revision 2, February 20, 2019.

[3] Ameren Missouri, 2014, Ameren Missouri Seismic Hazard and Screening (CEUS Sites) Response to NRC Request for Information Pursuant to 10 CfR 50.54(f) Regarding Recommendation 2. 1 of the Near-Term Task Force Review of Insights from the Fukushima Dai-Ichi Accident, Docket Number 50-483, March 28, 2014.

[4] Nuclear Regulatory Commission, 2012, Central and Eastern United States Seismic Source Characterization for Nuclear Facilities, NUREG-21 15, U.S. Nuclear Regulatory Commission, Washington, D. C., February 2012.

[5] Electric Power Research Institute, 2013, EPRI (2004, 2006) Ground-Motion Model (GMM)

Review Proj ect, Report 300200071 7, Electric Power Research Institute, Palo Alto, CA, June 2013.

[6] Senior Seismic Hazard Analysis Committee, 1997, Recommendations for Probabilistic Seismic Hazard Analysis: Guidance on Uncertainty and Use of Experts, NUREG/CR-6372, Washington DC, 1997.

[7] Nuclear Regulatory Commission, 2012, Practical Implementation Guidelines for SSHAC Level 3 and 4 Hazard Studies, Revision 1 NUREG-2l 1 7, U.S. Nuclear Regulatory Commission, Washington, D. C., April 2012.

[8] Nuclear Regulatory Commission 2007, A Performance-Based Approach to Define the Site-Specific Earthquake Ground Motion, Regulatory Guide 1 .208, U.S. Nuclear Regulatory Commission, Washington, D.C., March 2007.

[9] Nuclear Regulatory Commission, 2010, Interim Staff Guidance on Ensuring Hazard-Consistent Seismic Input for Site Response and Soil Structure Interaction Analysis, Interim Staff Guidance DC/COL-ISG-0l7, U.S. Nuclear Regulatory Commission, Washington, D. C., March 2010.

[1 0] Electric Power Research Institute, 201 3 Seismic Evaluation Guidance: Screening, Prioritization and Implementation Details (SPID) for the Resolution of Fukushima Near-Term Task Force Recommendation 2. 1 : Seismic, Electric Power Research Institute, Palo Alto, CA: Report 1025287, 2013.

[1 1] Nuclear Energy Institute, 2009, Consistent Site-Response/Soil-Structure Interaction Analysis and Evaluation, White Paper, ADAMS Accession No. ML091680715, Nuclear Energy Institute, June 12, 2009.

[12] Nuclear Regulatory Commission, 2015, Callaway Plant, Unit 1 StaffAssessment of Information Provided Pursuant to Title 10 ofthe Code offederal Regulations, Part 50, Section 50.54(f), Seismic Hazard Reevaluation Relating to Recommendation 2.1 of the Near-Term Task Force Review of Insights from the Fukushima Dai-Ichi Accident (TAC No. MF3739, Letter to Ameren Missouri, April 21, 2015.

[1 31 ASME/ANS RA-S-Case 1 20 1 7 Case for ASME/ANS RA-Sb-20 1 3 Standard for a Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications, The American Society of Mechanical Engineers, New York, November 22, 2017.

Page 73 of92 to ULN RC-06524

[14] Westinghouse, 201 8, Peer Review of the Callaway Seismic Probabilistic Risk Assessment, Report PWROG-1 8044-P, Revision 0-A, Risk Management Committee PA-RMSC-1476, July 2018.

[1 5] Westinghouse, 2019, Peer Review Closure Report, Report PWROG-19011, Revision 0, Risk Management Committee

[1 6J Electric Power Research Institute, 2015, High Frequency Program, Application Guidance for Functional Confirmation and Fragility Evaluation, Electric Power Research Institute, Palo Alto, CA: Report 3002004396, July 2015.

[17] McGuire, R. K, et al., 2001, Technical Basis for Revision of Regulatory Guidance on Design Ground Motions: Hazard- and Risk-Consistent Ground Motion Spectra Guidelines, NUREG/CR 6728, U.S. Nuclear Regulatory Commission, October 2001.

[1 8J Campbell, K. W., and Bozorgnia, Y., 2003, Updated Near Source Ground-Motion (Attenuation)

Relations for the Horizontal and Vertical Components of Peak Ground Acceleration and Acceleration Response Spectra, Bulletin ofthe Seismological Society ofAmerica, Vol. 93, No. 1, pp. 314-331.

[19] Gulerce, Z., and Abrahamson, N. A., 201 1, Site-Specific Design Spectra for Vertical Ground Motion, Earthquake Spectra, Volume 27, No. 4, pp. 1 023-1047, 201 1.

[201 ASME/ANS RA-Sb-20l3, Addenda to ASME/ANS RA-S-2008 Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications, 2013.

[21] NUREG/CR-4840, Procedures for the External Event Core Damage Frequency Analyses for NUREG-l 150, USNRC, November 1990.

[22] PRA-SPRA-002, Revision 0-E, Callaway Energy Center Seismic Probabilistic Risk Assessment Quantification Notebook.

[23] RG 1 .200, Revision 2, An Approach for Determining the Technical Adequacy of Probabilistic Risk Assessment Results for Risk-Informed Activities, USNRC, March 2009.

[24] NET 12-13, External Hazards PRA Peer Review Process Guidance, August 2012.

[25] SeismicEvalitation Guidance: Screening, Frioritization andlmplementation Details (SFID)for the Resolution ofFukushima Near-Term Task force Recommendation 2. 1: Seismic EPRT, Palo Alto, CA: 2013. 1025287.

[26] NUREG-l 855, Revision 1, Guidance on the Treatment ofUncertainties Associated with PRAs in Risk-Informed Decision making, March 2017.

[27] EPRI 30020007091, Seismic PRA Implementation Guide, Electric Power Research Institute, Palo Alto, CA, December 2013.

[28] EPRI Report 1 025287, Screening, Prioritization and Implementation Details (SPDJ) for the Resolution of Fukushima Near-Term Task Force Recommendation 2. 1 : Seismic, February 2013.

[29] ASME/ANS RA-S-2008, Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications, including Addenda B, 2013 American Society of Mechanical Engineers, New York, September 30, 2013.

[30] EPRT Report NP-5223-SL, Revision 1, Generic Seismic Ruggedness ofPower Plant Equipment, August 1991.

[31] EPRI, A Methodology for Assessment ofNuclear Plant Seismic Margin, NP-6041-SL, Revision 1, August 1991.

Page 74 of 92 to ULNRC-06524

[32J A$ME/ANS RA-S CASE 1 Case for ASME/ANS RA-Sb-20 1 3 Standard for Level 1/Large Early Release frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications, American Society of Mechanical Engineers, New York, November 22, 2017.

[33] ACS SASSI, NQA Version 2.3.0 Including Options A and FS An Advanced Computational Software for 3D Dynamic Analysis Including Soil-Structure Interaction User Manuals Revision 7 September 26, 2012.

[34] E. Kausel, EKSSI v3.1, A Program for the Dynamic Analysis of Structures Including Soil-Structure Interaction Effects.

[35] Rizzo Associates Document 1 1-4695A, Revision 0, Probabilistic Seismic Hazard Analysis Seismic Probabilistic Risk Assessment Projt, Callaway Energy Center, Unit 1 .

[36] ASCE 43-05, Seismic Design Criteria for Structures, Systems and Components in Nuclear Facilities.

[37] USNRC NUREG-0800, Standard Review Plan, Section 3.7.1, Revision 4, Seismic Design Parameters, December 2014.

[38J EPRI, Seismic Fragility Applications Guide Update, TR-1019200, December 2009.

[39] EPRI, Methodology for Developing Seismic Fragilities, TR-103959, June 1994.

[401 ANSYS, Inc., ANSYS Mechanical Version 15.0.

[41] GTSTRUDL, Version 32.

[421 ASCE 4-1 6, Seismic Analysis of Safety-Related Nuclear Structures and Commentary.

[43] EPRI, An Approach to Human Reliability Analysis for External Events with a Focus on Seismic, TR-3 002000709.

[44] NRC Letter, U.S. Nuclear Regulatory Commission Acceptance on Nuclear Energy Institute Appendix X to Guidance 05-04, 07-1, and 12-13, Close-Out of facts & Observations (F&Os),

dated May 3, 2017 (ADAMS Accession NO. ML17079A427).

[45] PRA-SPRA-00l Revision 0-F, Callaway Energy Center Seismic Probabilistic Risk Assessment Modeling Notebook. *.--

Page 75 of 92 to ULNRC-06524 8.0 Acronyms ACB Auxiliary and Control Buildings ADAMS Agencywide Documents Access and Management System AEPS Alternate Electric Power System AFW Auxiliary Feedwater ANS American Nuclear Society ASME American Society of Mechanical Engineers ATWS Anticipated Transient without SCRAM BE best-estimate CAFTA Computer Aided Fault Tree Analysis CCDP Conditional Core Damage Probability CCW Component Cooling Water CDF Core Damage Frequency CDFM Conservative Deterministic Failure Margin CEC Callaway Energy Center CEUS Central and Eastern United States CLERF Conditional Large Early Release Frequency CLERP Conditional Large Early Release Probability CVCS Chemical and Volume Control Systems EDG Emergency Diesel Generator EL Elevation EPRI Electric Power Research Institute FIRS Foundation Input Response Spectra FLEX Diverse and Flexible Mitigation Capability F&Os Facts & Observations F-PRA Fire Probabilistic Risk Assessment fsp feet per section FTR Fail to Run FTS Fail to Start F-V Fussell-Vesely GERS generic equipment ruggedness spectra GMC Ground Motion Characterization GMM Ground Motion Model GMRS Ground Motion Response Spectra GUI General User Interface HCLPF High Confidence of Low Probability of Failure HEP Human Error Probability HFE Human Failure Event HLR High Level Requirement HRA Human Reliability Analysis IE Internal Events IEPRA Internal Events Probabilistic Risk Assessment ISRS Instructure Response Spectra JCNRM Joint Committee on Nuclear Risk Management LB lower bound LERF Large Early Release Frequency LLOCA Large Loss of Coolant Accident LMSM Page 76 of 92 to ULNRC-06524 LOCA Loss of Coolant Accident LOOP Loss ofOffsite Power LOSP Loss ofOffsite Power MLOCA Medium Loss of Coolant Accident MSO Multiple Spurious Operations NEI Nuclear Energy Institute NRC Nuclear Regulatory Commission NSCI Non-Seismic Class I NTTF Near-Term Task Force OBE Operating Basis Earthquake PAF Plant Availability Factor PGA Peak Ground Acceleration PRA Probabilistic Risk Assessment PSHA Probabilistic Seismic Hazard Analysis PWR Pressurized Water Reactor RCS Reactor Coolant System RRW Risk Reduction Worth RWST Refueling Water Storage Tank SBO Station Blackout SCDF Seismic Core Damage Frequency SEL Seismic Equipment List SIET Seismic Initiating Event Tree SLERF Seismic Large Early Release Frequency SOV Separation of Variables SPID Screening, Prioritization and Implementation Details 5-PRA Seismic Probabilistic Risk Assessment SRA Site Response Analysis SSC Seismic Source Characterization (used in Seismic Hazard discussion)

SHHAC Senior Seismic Hazard Analysis Committee SSC Structures, Systems and Components.

551 Soil Structure Interaction TDAFP Turbine-Driven Auxiliary Feedwater Pump -

UHRS Uniform Hazard Response Spectra 3i .

UHS Ultimate Heat Sink -1 r r rfi1:r UP upper bound Jz; 7Id. ,. C-;:  :.:..:

1, V/H vertical-to-horizontal ,-

VSLOCA Very Small Loss of Coolant Accident .

• )! ; *, *

,I

) ;_ .... t3:

i) 3!; .

j 1]*-

Page 77 of 92 to ULNRC-06524 Appendix A - Summary of $-PRA Peer Review and Assessment of PRA Technical Adequacy This Appendix provides a summary of the peer review of the CEC S-PRA, the peer review f&O closure reviews, and provides the bases for why the $-PRA is technically adequate for the response to the NRCs request under 50.54(f) [1].

The CEC $-PRA was subjected to an independent peer review against the pertinent requirements of the ASME/ANS PRA Standard [13] as detailed in Section A.l The information presented in this Appendix establishes that the S-PRA has been peer reviewed by a team with adequate credentials to perform the assessment, establishes that the peer review process followed meets the intent of the peer review characteristics and attributes in Table 16 of RG 1 .200, Revision 2 [231.

A.1 Overview of the Peer Review The peer review assessment [1 4], and subsequent disposition of peer review findings [1 5] are summarized in this Appendix. The scope of the reviews encompassed the set of technical elements and supporting requirements (SRs) for the SHA (seismic hazard), SFR (seismic fragilities), and SPR (seismic PRA modeling) elements for seismic CDF and LERF.

The CEC 5-PRA peer review occurred during the week of June 18, 2018. The CEC 5-PRA F&O closure review occurred during the week of March 1 1 201 9.

A.2 Summary of the Peer Review Process The June 201 8 peer review [14] was performed against the requirements in the PRA Standard Code Case for Part 5 [13], using the peer review process defined in NEI 12-1 3 [24]. for supporting requirements in the Code Case that referred back to requirements in Part 2, Addendum B of the PRA Standard [21] was utilized.

The NET 12-13 S-PRA peer review process [24] involves an examination by each reviewer oftheir assigned PRA technical elements against the requirements in the Standard [20] and [1 3] to ensure the robustness of the model relative to all of the requirements.

Tmplementing the review involves a combination ofa broad scope examination ofthe PRA elements within the scope of the review and a deeper examination of portions of the PRA elements based on what is found during the initial review. The supporting requirements (SRs) provide a structure which, in combination with the peer reviewers PRA experience, provides the basis for examining the various PRA technical elements. Tf a reviewer identifies a question or discrepancy, that leads to additional investigation until the issue is resolve or an f&O is written describing the issue and its potential impacts, and suggesting possible resolution.

for each technical element, i.e., SHA, SfR, and SPR, a team oftwo (at least two) were assigned, one having lead responsibility for that area. for each SR reviewed, the responsible reviewers reached consensus regarding which of the Capability Categories defined in the Standard that the PRA meets for that SR, and the assignment of the Capability Category for each SR was ultimately based on the consensus of the full review team. The Standard also specifies high level requirements (HLR). Consistent with the guidance in the Standard, capability categories were not assigned to the HLRs, but a qualitative assessment of the applicable HLRs in the context of the PRA technical element summary was made based on the associated SR Capability Categories.

As part of the review teams assessment of capability categories, F&Os are prepared. There are three (3) types of f&Os defined in NET 12-1 3 [24]: findings which identify issues that must be addressed in order for an SR (or multiple SRs) to meet Capability Category II; Suggestions which identify issues that the reviews have noted as potentially important but not requiring resolution to meet the SRs; and Best Practices Page 78 of 92

Attachment 2 to ULNRC-06524 which reflect the reviewers opinion that a particular aspect ofthe review exceeds normal industry practice.

The focus in this Appendix is on Findings and their disposition relative to this submittal.

A.3 Peer Review Team Qualifications The June 2018 CEC $-PRA peer review was led by Mr. Kenneth Kiper ofWestinghouse Electric Company.

Team members included: Dr. Martin McCann ofJack Benjamin & Associates, Dr. Glenn Rix of Geosyntec Consultants, Inc., Stephen J. Eder of facility Risk Consultants, Inc., Dr. Ram Srinivasan, independent consultant, Colter D. Somerville of Southern Nuclear Operating Company, Rick Summit ofRLS Consulting, and Deepak Rao of Entergy Nuclear. The lead and reviewer qualifications were reviewed by Ameren Missouri and have been confirmed to be consistent with requirements in the ASME/ANS PRA Standard

[20] and [13] and the guidelines ofNEI-12-13 [24J. The members ofthe peer review team were independent of the CEC 5-PRA. They were not involved in performing or directly supervising work on any PRA Element evaluated in the overall CEC 5-PRA.

Mr. Kenneth Kiper, the team lead, has over 35 years of experience at Westinghouse and, previously at Seabrook Station, in the nuclear safety area generally and PRA specifically for both existing and new nuclear power plants. He has led a number of peer reviews, including reviews of internal events PRAs, internal flood PRAs, fire PRAs, high wind PRAs, and several 5-PRAs Dr. Martin McCann was the lead for the review of the Seismic Hazard Analysis (SHA) technical element.

Dr. McCann has over 35 years of experience in engineering seismology including site response analysis and specification of ground motion. Dr. McCann has served as SHA lead reviewer for a number of recent 5-PRAs.

Dr. McCann was assisted by Dr. Glenn Rix. Dr. Rix has nearly 30 years of experience in seismic hazard evaluation, geotechnical earthquake engineering, and performance-based and risk-based analyses. Dr. Rix joined Geosyntec in 2013 after a distinguished 24 year career as a Professor in the School of Civil and Environmental Engineering at the Georgia Institute of Technology specializing in geotechnical and earthquake engineering. Dr. Rix has participated in a number of S-PRA peer reviews.

Mr. Stephen J. Eder led the Seismic Fragility Analysis (SFR) review. Mr. Eder is an industry leader in the seismic fragility analysis of systems and components at nuclear power facilities with over 35 years of experience. He has led the seismic probabilistic risk analysis fragility peer reviews for Beaver Valley, Davis Bessie, Fermi, Perry, and Vogtle nuclear power plants. He has provided overall technical direction for walkdowns and seismic fragility analyses for a number of U.S nuclear power plants.

Mr. Eder was assisted by Dr. Ram Srinivasan and Mr. Colter D. Somerville. Dr. Srinivasan has over 45 years of experience in the nuclear industry, principally in the design, analysis (static and dynamic, including seismic), and construction ofnuclear power plant structures. He is actively involved in the Post-Fukushima Seismic Assessments (NRC NTTF 2. 1 and 2.3) and is a member of the NEI Seismic Task Force and the ASME/ANS JCNRM, Part 5 Working Group (Seismic and other External Hazards PRA). He has participated on several previous 5-PRA peer reviews, either as reviewer or utility consultant.

Mr. Colter D. Somerville is the seismic technical support engineer for the Southern Nuclear Operating Company (SNC) nuclear power plants, including VogUe, Hatch, and Fancy. He has participated in several peer reviews defending the seismic fragility work for the SNC plants. He was a working observer at two 5-PRA peer reviews in 2017, Indian Point 2 and Diablo Canyon.

Mr. Rick Summitt was the lead for the review of the Seismic System Response Analysis (SPR) technical element. Mr. Summitt has 3 8 years of experience in risk analysis and has used both deterministic and probabilistic techniques to manage and supply lead technical guidance in the assessment of safety and reliability concerns for both nuclear and chemical projects.

Mr. Summitt was assisted by Mr. Deepak Rao. Mr. Rao is a nuclear engineer and mechanical engineer with over 30 years experience working in the nuclear power industry. He has supported a number of Page 79 of 92 to ULN RC-06524 industry PRA peer reviews, including internal events, Fire PRA, Seismic PRA and High Winds PRA in the last several years. He has participated in a number ofpeer reviews, including internal events PRAs and 5-PRAs.

Dr. Mohamed Talaat from Simpson, Gumpertz & Heger, Inc. and Mr. JC Patel from Wolf Creek Nuclear Operating Company served as working observers for the SFR and SPR technical elements, respectively.

My observations and findings that Dr. Talaat or Mr. Patel generated were given to the peer review team for their review and ownership. As such, Dr. Talaat and Mr. Patel assisted with the review but were not formal members of the peer review team.

A.4 Summary of the Peer Review Conclusions The review teams assessment of the SPRA elements is excerpted from the peer review report [14] as follows. Where the review team identified issues, these are captured in peer review findings, for which the dispositions are summarized in the next section of this Appendix.

A.4.1 Seismic Hazard Analysis The Standard requires the seismic hazard input to the 5-PRA be determined on the basis of a site-specific probabilistic seismic hazard analysis (PSHA). The site-specific PSHA must be based on current geological, seismological, and geophysical data; local site topography; and surficial geologic and geotechnical site properties. The Callaway PSHA:

1 . Used the existing regional Seismic Source Characterization (SSC) model for the Central and Eastern United States (CEUS);

2. Used the regional CEUS ground motion characterization (GMC) models; 3 . Evaluated the effects of local site conditions on the ground motions that would be experienced by plant structures, systems and components; and

4. Considered the potential for other seismic hazards, including ground failures that might occur due to soil liquefaction, slope failures, fault displacement.

The existing regional-scale SSC model for the CEUS largely meets the intent of the Standard for sites in the CEUS. At the same time the Standard is clear that the SSC model must be based on current information.

The review team understands that some work was done to compile an up-to-date earth science database (available since the CEUS SSC model was developed) that could impact the estimate ofthe seismic hazard at the site. These impacts could be modifications to existing seismic source models or the addition of new sources. The work that was done in terms of compiling new information and evaluating its impact does not fully meet the intent of the Standard. Specifically, the documentation needs to be enhanced to explicitly summarize the information reviewed and the technical basis for determining no impacts on the PSHA input models used. For this reason, SHA-C4 which relates to the evaluation ofthis new information was Not Met.

The PSHA requirements associated with incorporating the effects oflocal site conditions on ground motions are defined in $HA-E. The effect ofsite conditions were modeled by means ofamplification factors derived from probabilistic site response analyses that incorporate site-specific information on surficial geologic deposits, and site geotechnical properties. Epistemic uncertainty and aleatory variability in shear wave velocity, layer thickness, and nonlinear properties are considered and propagated separately in the site response analyses.

The Standard requires that a screening analysis be performed to assess whether in addition to vibratory ground motion, other seismic hazards, such as fault displacement, landslide, soil liquefaction, or soil settlement, need to be included in the S-PRA. As part of the PSHA, a systematic evaluation was carried out for other hazards.

Page 80 of 92

Attachment 2 to ULN RC-06524

$HA-J defines the requirements for documentation of the PSHA. These requirements set a high bar with regard to the documentation that should be provided. It must support PRA applications, peer review, and future updates. Overall, the documentation for the PSHA is generally complete and meets the intent of the Standard; therefore the Supporting requirements for SHA-J are met. The documentation of the Callaway PSHA is provided in a group ofdocuments, including the PSHA contractor self-assessment. However, there are elements of the PSHA documentation that could be better integrated and enhanced. Several of the findings are associated with PSHA documentation issues and, if these are adequately addressed, will result in PSHA documentation that meets the intent ofthe Standard.

A.4.2 Seismic Fragility Analysis The SFR assessment covered the three principal elements of the fragility analysis; namely, site-specific seismic response analysis, plant walkdown, and fragility analysis calculations. A summary of the three elements are briefly summarized below.

Site-specific seismic response analysis ofthe various buildings housing the SEL items was performed using the ground motions corresponding to the Uniform Hazard Response Spectra (UHRS) shape provided in the plant Probabilistic Seismic Hazard Analysis (PSHA) reports. The Reference Level Earthquake (RLE) utilized corresponds to the GMRS, which is anchored to O.39g peak ground acceleration (PGA). Selection of this level was based on the interim results of the PRA quantifications and corresponds to the expected failure levels of most of the top contributors to seismic risk. The appropriateness of this ground motion level was reviewed relative to the capacities oftop contributors to CDF/LERF. The review showed that the selected O.39g PGA is appropriate for top contributors to CDF. However, the failure level of the top contributors to LERF occurs for ground motion levels significantly exceeding O.4g PGA.

The seismic input to the building response analysis is based on a single time history set (3 components) matched to the GMRS Structural properties (uncracked concrete) and strain dependent soil properties compatible with the GMRS conditions were used in the building models. Out-of-plane cracking of floor slabs was not considered. The development ofbuilding models relies heavily in the modification of existing lumped-mass stick models. These modifications are documented in the respective calculations. Vertical floor slab flexibility was neglected.

For the Auxiliary and Control Buildings (ACB), a new 3D finite element building model was developed.

Structural stiffness variation was not addressed. Dynamic analyses included soil-structure interaction (551) analysis of structures. 551 analysis for the ACB was performed using the software ACS-SASSI (3D finite element model) while for the other buildings (lumped-mass stick models) EKSSI was used. EKSSI uses impedance functions to represent the soil mass and stiffness properties. The 551 analyses considered three soil profiles; best-estimate, lower-bound, and upper-bound; and a median-centered structural model.

Instmcture Response Spectra (ISRS) were generated at various locations in the different buildings for the fragility analysis ofthe components. Some ofthe ISRS results did not appear to be realistic.

A walkdown of the Callaway Nuclear Plant was performed as part of this Peer Review. The walkdown review focused on the dominant risk contributors in seismic CDF and LERF, but also included some additional example SSCs to confirm findings or observations made by the Seismic Review Team. The peer review walkdown included the Auxiliary Building, Control Building, Diesel Generator Building, Essential Service Water System (ESWS) Pumphouse, ultimate heat sink (UHS) Cooling Tower, Turbine Building, and the Yard.

The peer review team observed the plant to be spacious, well laid out, and free of congestion. Equipment was observed to be well anchored, and piping and pipe supports appeared to be rugged. The level of seismic housekeeping in the plant was observed to be very good. It was a general observation that the robustness of the plant was not reflected in the current seismic capacities determined by the fragility analysis.

The 5-PRA seismic walkdowns by the seismic review team (SRT) were performed in four sessions and reviewed attributes for functional capacity, anchorage, and spatial interactions for each component. During Page 81 of 92 to ULN RC-06524 the walkdown, seismic vulnerabilities were identified and documented by the SRI. However, it became apparent to the peer review team that no follow-up walkdowns had been conducted to verify and ensure that the governing failure modes used in the fragility calculations were realistic and plant specific.

The 5-PRA seismic walkdown addressed seismic-induced fire and flood. The seismic-induced fire review focused on the hydrogen piping system, which was found to be rugged. However, the potential for fire initiation due to seismic failure of high-voltage non-safety electrical cabinets was not investigated. The seismic-induced flood review focused on the fire protection system and identified several vulnerabilities.

However, pre-action fire protection systems were assumed to be dry and were not investigated for potential seismic vulnerabilities as a potential flood source.

Fragility parameters were calculated for all the SEL items credited in the plant S-PRA Model. No capacity-based screening was performed. Rather, surrogate fragility parameters were developed and used for the componentsjudged to be ofhigher seismic capacity. The failure modes considered in the building structure and retention pond fragility evaluations included structural integrity, soil bearing capacity, and seismic interaction considerations. The failure modes considered by the fragility evaluations included anchorage, functional, and seismic interaction considerations.

The fragility calculations for the top contributors to SCDF were reviewed and found to be from an initial set of conservative deterministic failure margin (CDFM) hybrid fragilities with conservative assumptions.

These fragility calculations contain various significant conservatisms and thus are not realistic. At the time ofthe peer review, it became apparent that more rigorous methods had not yet been employed to determine realistic median capacities and associated uncertainty parameters for the top contributors to SCDF and SLERF. The peer review noted that there was a plan in place to perform this additional analysis:

. Refinement of fragility data for dominant contributors is to be performed. It is assumed fragility data will be updated as needed after this refinement.

. More rigorous methods for calculation for fragility parameters may be applied to a small subset of SSC after initial S-PRA results are available and the dominant contributors to risk are identified.

Methods will be per EPRI TR-103959 including numerical simulations and/or the separation of variables approach.

The peer review team was able to perform the peer review using the documentation received from the proj ect team. Aspects of the review were difficult and required considerable research as well as question and answer sessions.

In summary, the fragility analysis generally meets the applicable requirements of the ASME/ANS RA-S Standard CODE CASE 1 [1 3]. However, the work is not complete for developing realistic fragilities for top contributors to SCDF and $LERF.

A.4.3 Seismic Plant Response The Callaway SPR model integrates the site-specific hazard, fragilities and system-analysis and accident sequence aspects. The Callaway SPR model appropriately modified the full Power Internal Event model to include seismically-induced initiating events including industry experience from EPRI guidance documents, seismically-induced initiating events caused by secondary hazards (internal and external floods),

seismically-induced SSC failures and non-seismically induced unavailabilities and human actions.

The assessment of potential seismic initiating events considered a wide range of potential consequences defined by the internal events analysis, postulated external initiating events and industry guidance documents. The disposition of some initiating events as being encompassed by loss of offsite power may be non-conservative for lower accelerations where the generic capacity of non-safety equipment is lower than that assumed for LOSP.

The model accurately maps defined fragilities to the appropriate components in the PRA model and expands the internal events to address seismic impacts. However, the fragilities as mapped may overstate the Page 82 of 92

Attachment 2 to ULNRC-06524 correlation among relays and may result in links with non-correlated equipment causing broad impacts on components and systems. The impact of this conservatism is not fully resolved and may be addressed by future refinements. The conditional flooding and fire events postulated for seismic events are also mapped appropriately, but in some cases the mapping may not reflect the conditions of the seismic event. As an example, some fire piping faults are screened based on preaction design but that may not be appropriate for seismic events where spurious actuation of detectors is possible.

The $-PRA includes operator actions from the IEPRA, with the HEPs evaluated based on seismic specific challenges. The S-PRA model does not include any seismic-specific operator actions. The addition of operator actions to recover the impacts of relay chatter should be considered if the actions are feasible and impact risk significant relay chatter.

The model assembly incorporates the hazard, fragility and $SCs in a manner supporting quantification. The quantification process provides an approach for integrating this information in order to develop both CDF and LERF results and to propagate uncertainties. The uncertainty and sensitivity studies provide additional insights into the CDf and LERF characteristics as well as model realism and areas of the study that are sensitive to model assumptions.

A.5 Revision of S-PRA Model and Documentation Following the peer review, the $-PRA model and documentation were updated to address all f&Os from the 201 8 $-PRA Peer Review [14] except for Finding 25-1 9 (linked to $PR-B2) and Finding 25-12 ($PR E6). In addition, CEC generated closure documentation for each ofthese F&Os. Subsequently, the updated S-PRA model and documentation were subjected to an independent assessment in March 201 9 of the F&O closure.

A.6 Finding Closure by Independent Assessment and Focused Scope Peer Review An independent assessment of CEC s resolution of open 5-PRA F&Os was performed in March 201 9 and is documented in PWROG-l 901 1-P [l5J. The process used for the independent assessment is outlined in Section X.l .3 (Close Out F&Os by Independent Assessment) ofAppendix X to NET 12-13 [24], which has been accepted by the NRC [44], with two conditions:

1 . Use ofNew Methods: A PRA method is new if it has not been reviewed by the NRC staff. There are two way new methods are considered accepted by the NRC staff: (1) they have been explicitly accepted by the NRC (i.e., they have been reviewed, and the acceptance has been documented in a safety evaluation, frequently-asked-questions, or other publicly available organizational endorsement), or (2) they have been implicitly accepted by the NRC (i.e., there has been no documented denial) in multiple risk-informed licensing applications. The NRCs treatment of a new PRA method for closure of F&Os is described in memorandum U.S. Nuclear Regulatory Commission Staff Expectations for an Industry Facts and Observations Independent Assessment Process, dated May 1 2017 (ADAMS Accession No. ML1 7l2lA27l).

2. Use ofAppendix X in its entirety: In order for the NRC to consider the F&Os closed so that they need not be provided in submissions of future risk-informed licensing applications, the licensee should adhere to the guidance in Appendix X in its entirety. Following the Appendix X guidance will reinforce the NRC staff s confidence the F&O closure process and potentially obviate the need for a more in-depth review.

The result ofthis independent assessment was intended to support future CEC licensee amendment request submittals, other regulatory interactions, risk-informed applications, and risk-informed decision-making.

Finding resolution reviewed and determined to have been adequately addressed through this independent assessment are considered closed and no longer relevant to the current PRA model, and thus need to be carried forward nor discussed in such future activities.

Page 83 of 92

Attachment 2 to ULNRC-06524 A.6.1 Selection of Independent Assessment Team Members The independent review team consisted of Mr. Kenneth Kiper, Dr. Marty McCann, Mr. Glenn Rix, Mr.

Steve Eder, Dr. Ram Srinivasan, Mr. Rick Summit and Mr. Deepak Rao. Note that all the reviewers were members of the team that performed the 201 8 5-PRA peer review [14]. The independent assessment team qualifications are discussed in Section A.3.

A.6.2 Host Utility Preparation CEC provided the complete and relevant review material to the independent assessment team in advance to allow the reviewers to prepare and conduct a more efficient technical review. As input to the review, CEC provided the following documentation:

. Exact wording of each original F&O within scope of the independent assessment

. A summary description of how each F&O was dispositioned

. CEC self-assessment of whether the F&O closure involved an upgrade or a maintenance activity, based on the definition ofupgrade vs. maintenance documented in the PRA Standard [321.

. Documents that were revised to resolve the F&Os A.6.3 Offsite Review All material generated in supported to the F&O closure activities performed by CEC were provided to the independent assessment team two weeks before the onsite review and consensus session. The review team started the review and familiarization of the documentation.

A.6.4 Onsite Review and Consensus During the onsite review and consensus session, the team achieved the following for each reviewed F&O:

. Consensus on the status of the F&O (i.e., CLOSED, OPEN, or PARTIALLY CLOSED). This conclusion was reached through a review of the original basis and description of the F&O and on the techuical work and documentation provided by CEC to resolve the issued identified in the F&O.

. Consensus on whether the activities performed to close the F&O are to be considered maintenance or upgrade, per the appropriate definition ofthe PRA Standard.

If the F&O was associated with an SR that was originally judged as Not Met or Met at Capability Category I, upon confirming closure of the associated F&Os, the SR has been re-assessed to reach consensus on whether the intent of the SR is now Met at Capability Category II or higher.

A.6.5 Treatment of New Methods All of the changes to the CEC 5-PRA were classified as either PRA maintenance or PRA upgrade by the independent assessment team. Therefore, no new methods were identified during the independent assessment.

A.6.6 Use of Remote Reviewers All team members participated on-site except for the two SHA reviewers (Dr. Marty McCann and Mr.

Glenn Rix). This remote participation was supported with web and teleconference to the onsite review team. The SHA reviewers were needed only for a limited number ofF&O reviews, and thus, these reviewers participated remotely.

A.6.7 Status of Findings at End of Independent Assessment The following bullets summarize the independent assessment conclusions:

Page 84 of 92

Attachment 2 to ULNRC-06524

. SHA: 5 f&Os were closed, 0 remain open. Two SRs (SHA-C4 and SHA-Hi) were reassessed and determined to be Met at Capability Category II. No new issues were identified for element SHA.

. SFR: 27 F&Os were closed, 0 remain open. SFR-D3, SFR-E3, SFR-E5 and SFR-Fl were reassessed and determined to be Met at capability category II. No new issues were identified for element SfR.

. SPR: 32 F&Os were closed, 2 remain open but were not in-scope for the IAT during this review.

SPR-D2 and SPR-E5 were reassessed and determined to be Met at Capability Category II. No new issues were identified for element SPR.

A.6.$ final Independent Assessment Report A final report was provided at the end of the independent assessment, which documented the review and its conclusions [1 5]. This report includes the following information:

. Descriptions ofthe F&O independent assessment process.

. Description of the scope ofthe independent assessment.

. A summary of the review teams decisions for each finding within the scope of the review, along with the rationale for determination of adequacy of inadequacy for closure of each finding in relation to the affected portions of the associated SR.

. For each finding, assessment of whether the resolution was determined to be a PRA upgrade, maintenance update, or other, and the basis for that determination.

The final report included each ofthe independent assessment team members resumes and summary of their experience as it applies to qualification guidelines of NET guidance documents and the ASME/ANS PRA Standard.

This report will be retained by CEC in accordance with maintenance of their peer review and PRA recordkeeping practices, and it is available for review and audit.

A.6.9 Summary of Independent Assessment Team Conclusions of the F&Os reviewed, the independent assessment team concurred that all can be considered closed (with exception to the two f&Os that were not reviewed). As a result ofthe closure ofthe associated f&Os, SRs originallyjudged as Not met are now judged to be Met.

The independent assessment team recognized a significant amount of work invested in the resolution of the F&Os from the original peer review, including the generation ofnew fragility, additional sensitivity studies, improved documentation, and an additional walkdown for risk-significant seismic-induced flood and fire sources. The independent assessment team concluded that, as a result ofthe closure ofthe associated f&Os, the CEC 5-PRA more realistically reflects the current seismic risk at the site.

A.6.1O Compliance of Independent Assessment with NRC Conditions As indicated in Section A.6, the NRCs acceptance of the F&O closure process described in Appendix X to NET 12-13 [24], as documented in [44], includes two conditions. The independent assessment of the finding closure for the CEC 5-PRA satisfies the conditions as follows:

1 . Use of New Methods: As indicated in Section A.6.5, new methods were not employed in the resolution of findings associated with the CEC Unit 1 5-PRA. Therefore, this condition does not apply.

2. Use ofAppendix X in its Entirety: The finding closure process encompasses all ofthe elements of Appendix X Page 85 of 92

Attachment 2 to ULNRC-06524 Therefore, the application of the Appendix X process to the closure of the findings identified during the CEC S-PRA peer review is in conformance with the NRCs requirements.

A.7 Summary of Technical Adequacy of the $-PRA for the 50.54(1) Response The set of supporting requirements from the ASME/ANS PRA Standard that are identified in Tables 6-4 through 6-6 of the SPIB [25] define the technical attributes of a PRA model required for a $-PRA used to respond to the 50.54(1) letter. The conclusions ofthe peer review discussed above and summarized in this submittal demonstrates that the CEC $-PRA model meets the expectations for PRA scope and technical adequacy as presented in RG 1 .200, Revision 2 [23J as clarified in the SPID [25]. The main body of this report provides a description ofthe S-PRA methodology, including:

. Summary of the seismi hazard analysis (Section 3 .0)

. Summary of the structures and fragilities analysis (Section 4.0)

. Summary ofthe seismic walkdowns performed (Section 4.2)

. Summary of the internal events at power PRA model on which the S-PRA is based, for CDF and LERF (Section 5.1)

. Summary of adaptations made in the internal events PRA model to produce the seismic PRA model and bases for the adaptations (Section 5.1).

Detailed archival information for the 5-PRA consistent with the listing in Section 4.1 of RG 1 .200 Rev. 2

[23] is available if required to facilitate the NRC staffs review ofthis submittal.

The CEC 5-PRA reflects the as-built and as-operated plant as of the cutoff date for the 5-PRA, 3/8/2019.

The peer review observations and conclusions noted in A.4, the F&O closure review discussion in Section A.6, demonstrate that the CEC S-PRA is technically adequate in all aspects for this submittal. Subsequent to the CEC peer review, the peer review findings have been appropriately dispositioned, and the S-PRA model has been updated to reflect these dispositions and further refine several fragility values. The results presented in this submittal reflect the updated model as of March 2019.

A.8 Summary of $-PRA Capability Relative to SPID Tables 6-4 through 6-6 The PWR Owners Group performed a full scope peer review of the CEC internal events PRA and internal flooding that forms the basis of the 5-PRA to determine compliance with ASME/ANS PRA Standard [20]

and [13]and Regulatory Guide 1 .200 [23] in April 2019. This review documented findings for all supporting requirements (SRs) which failed to meet at least Capability Category II.

The PWR Owners Group peer review of the CEC S-PRA was conducted in June 201 8 An independent.

assessment of 5-PRA F&O closure was conducted in March 2019. The results of these peer reviews are discussed above, including resolution of $Rs not assessed by the peer review as meeting Capability Category II, and resolution of peer review findings pertinent to this submittal. The peer review teams expressed the opinion that the CEC 5-PRA model is of good quality and integrates the seismic hazard, the seismic fragilities, and the systems-analysis aspects appropriately to quantify core damage frequency and large early release frequency. The general conclusion of the peer reviews was that the CEC 5-PRA is judged to be suitable for risk-informed applications.

. Table A-l provides a summary ofthe disposition of SRsjudged by the peer reviews to be not met, or not meeting Capability Category II.

. Table A-2 provides a summary ofthe ofpotentially important sources ofuncertainty in the CEC 5-PRA Page 86 of 92 to ULNRC-06524

. Table A-3 provides a listing of plant modifications that have not yet been incorporated into the S PRA.

. Table A-4 provides an assessment ofthe expected impact on the results ofthe CEC $-PRA of those SRs and peer review Findings that have not been fully addressed.

Table A-i: Disposition of SRs Assessed as Not Met or Not Met at Ca ability Category II Assessed Capability Associated Open Impact on S-PRA SR#

Category Finding F&Os Results SPR-B2 Not Met 25-19 See Table A-4 SPR-E6 Met at CCI 25-12 See Table A-4 A.9 Identification of Key Assumptions and Uncertainties Relevant to the S-PRA Results The ASME/ANS PRA Standard [13] includes a number of requirements related to identification and evaluation ofthe impact of assumptions and sources ofuncertainty on the PRA results. NUREG-l 855 [26]

and EPRI 1016737 provide guidance on assessment ofuncertainty for applications of a PRA. As described in NUREG-1 855, sources of uncertainty include parametric uncertainties, modeling uncertainties and completeness (or scope and level of detail) uncertainties.

. Parametric uncertainty was addressed as part ofthe CEC 5-PRA model quantification (See Section 5 .7 of this submittal)

. Modeling uncertainties are considered in both the base internal events PRA and the 5-PRA.

Assumptions are made during the PRA development as a way to address a particular modeling uncertainty because there is not a single definitive approach. Plant-specific assumptions made for each ofthe CEC 5-PRA technical elements are noted in the 5-PRA documentation that was subject to peer review, and a summary of important modeling assumptions is included in Section 5.3.2.

I Completeness uncertainty addresses scope and level of detail. Uncertainties associated with scope and level of detail are documented in the PRA but are only considered for their impact on a specific application. No specific issues of PRA completeness were identified in the S-PRA peer review.

A summary ofpotentially important sources ofuncertainty in the CEC S-PRA is listed in Table A-2.

Page 87 of 92 to ULN RC-06524 Table A-2: Potentially Important Sources of Uncertainty in CEC $-PRA PRA Element Summary of Treatment of Potential Impact on S-PRA Sources of Uncertainty per Peer Results Review Seismic Hazard The CEC $-PRA peer review The seismic hazard reasonably team noted that both the aleatoiy reflects sources of uncertainty.

and epistemic uncertainties have been addressed in characterizing the seismic sources. In addition, uncertainties in each step of the hazard analysis were propagated and displayed in the final quantification of hazard estimates for the CEC site.

Seismic Fragilities The CEC $-PRA peer review Several of the sensitivity studies team noted that the seismic described in Section 5.7 of this response analysis of buildings report evaluate the impact of was supplemented by a number changes to fragilities on the 5-of sensitivity studies used to PRA results as one means of assess important modelling assessing the impact of fragilities assumptions. These sensitivity uncertainties on the $-PRA studies included difference in results.

response between single and multiple time histories; effect of spatial incoherency of ground motion in building response; and control building in-structure sensitivity study. These studies served to reinforce engineering judgments and assumptions implemented in the fragility analysis of $SCs.

Seismic PRA Model The CEC S-PRA peer review A characterization of the mean team noted that parametric SCDF and $LERF is provided in uncertainty analyses and Section 5.6 ofthis report. Several sensitivity studies are performed sources of model uncertainty are and risk importances results are discussed in Section 5.7 along provided. with sensitives performed to evaluate the impact of possible changes to address these.

Page 88 of 92 to ULNRC-06524 A.1O Identification of Plant Changes Not Reflected in the S-PRA The CEC $-PRA reflects the plant as ofthe cutoffdate for the S-PRA, which was 3/8/201 9. Table A-3 lists significant plant changes subsequent to this date and provides a qualitative assessment of the likely impact of those changes on the S-PRA results and insights.

Table A-3 : Summary of Significant Plant Changes Since S-PRA Cutoff Date Description of Plant Change Impact on S-PRA Results MP 16-0021 : MDV53 and MDV55 Replacement Impacts configuration risk model only, potential This modification replaces two existing 345kV to no longer require the security diesel for circuit breakers and four vertical disconnect switchyard breaker support. Minimal impact.

switches in the Callaway Switchyard. In addition, 4 vertical disconnect switches: MDV71A, MDV71B, MDV75A, and MDV8 lA are replaced with Ameren standard Southern States type RDA 1 horizontal disconnect switches.

MP 17-0028 : Switchyard Breakers Replacement Impacts configuration risk model only, potential MDV45, MDV5 1 and MDV85 The following

, to no longer require the security diesel for changes are part of the modification: switchyard breaker support. Minimal impact.

. The ITE Switchyard breakers were at the end of their service lives and were replaced. The existing breakers utilize a two-pressure system with pneumatic

,mechanism to trip and close the break contacts. The new breakers are puffer .

type breakers that have a spring operated mechanism to trip and close the break . .

contacts I

. Relays and cabling associated with I breakers MDV41 MDV43, MDV45,

, . . S MDV5 1 and MDV85 are at the end of their useful lives and will be replaced and * . - . ==-=

indicators associated with the breakers -______

updated.

S S .

Existing 345kV vertical disconnect switch MDV55B is replace with Ameren standard Southern States RDA-l type double-end break disconnect switches.

MP 09-0048: Cathodic Protection System FPRA Documentation Only Upgrade This modification modifies the various components of the CPS : Rectifiers, Deepwell Anode Beds, and Shallow Anode Beds.

Page 89 of 92 to ULN RC-06524 Table A-3 : Summary of Significant Plant Changes Since S-PRA Cutoff Date Description of Plant Change Impact on S-PRA Results MP 16-0041 : Provide Power to New Maintenance FPRA Documentation Only Storage Building This modification provides a new 300 Loop power feed and transformer to provide available power source to the new Maintenance Storage Building, the Steam Generator Storage Facility and the Quality Control Radiographic Bunker by installing new power feed from 289PG30401 to new switch 289PG3240 1 new transformer XPG324, and 480V power panel PPPG324. The existing feed from Construction Quality control Building and the Construction Fab Shop can no longer be used.

Switchyard Physical and Cyber Security fPRA Documentation Only Modification 10/12/2 - This mod added more fencing, lighting, cameras and card readers to the switchyard.

FLEX Portable Equipment Mods Waiting on Industry resolution of data and HRA issues. Sensitivities assessed in S-PRA.

MP 08-0027: Main Turbine Controls Replacement FPRA Only. Minimal Impact.

This modification replaced the existing GE Mark II turbine control system, the turbine over speed protection system, and the Low Load Valve control system. The existing control systems were analog and were replaced with digital technology to eliminate single-point vulnerabilities, address the limited support available for the Mark II system, address obsolescence, improve reliability, and reduce operator burden.

MP 16-0024: SGKO5 Supplemental Cooling Internal Flooding and Fire. Minimal Impact.

System The modification includes the addition of two supplementary cooling systems (one to provide cooling from $GKO5A and one to provide cooling from $GKO5B) comprised of normally open fire dampers, circulating fans, and cross-train ductwork with isolation dampers. This modification also includes modifications to the 2032 elevation slab and concrete masonry unit walls ofthe Class 1E equipment rooms along with associated structural steel support for wall penetrations, fans, dampers, and ductwork.

Page 90 of 92 to ULNRC-06524 Table A-3: Summary of Significant Plant Changes Since S-PRA Cutoff Date Description of Plant Change Impact on $-PRA Results MP 16-0039: Upgrade EHC Room HVAC The FPRA only. Minimal impact.

Main Turbine Controls Upgrade project MP 08-0027 will install new digital controls in the EHC room increased the heat load in the room. The new equipment requires that all room penetrations must have fire-rated seals. Two new 5-ton HVAC units were installed and existing windows units SGE13 and $GE2O were removed and fire dampers were installed in the new ducts.

MP 17-0006: ESW Water Hammer Mitigation All PRAs Documentation Only. Internal Flooding Modification This modification adds pressurized small pipe additions, minimal impact.

accumulators at the CACs, CRACs, and Component Cooling Water (CCW) heat exchangers that inject non-condensable gas at low pressure to reduce the pressure response present at those components.

MP 1 8-0042: ESW Water Hammer Modification All PRAs Documentation Only. Internal Flooding CRAC Tie Ins - - Install safety-related system small pipe additions, minimal impact.

branch locations during $GKO4A!B Technical Specification Outages to facilitate the future Control Room Air Conditioning (CRAC) air accumulator injection points to mitigate the water hammer event on the supply and return piping of the A and B CRACs, SGKO4A!B.

17-0020, 17-0021, and 17-0037 Chemical FPRA Documentation Only Addition Pad and skid MP 17-0005 : Instrument Tunnel Sump Pump FPRA minimal impact.

This modification replaces the instrument sump pump 7-5/16 impeller with a 8-3/8 impeller for PLFO7A. A 7-7/8 impeller is acceptable for PLFO7B.

MP 14-0032: Install CCW Vent Valves Modification completion delayed. Internal Installation of 7 vents on the CCW B train and Flooding only. Minimal impact.

12 vents on the CCW A train, and modification ofpipe EG-196-HBC-4 to address voiding concerns.

MP 10-0010: Modify Cooling Tower Basin Level FPRA Only minimal impact.

Instrumentation Installation of new level indication system for the Callaway Cooling Tower Basin and uninstall the obsolete capacitance style indication.

Page9l of92 to ULN RC-06524 Table A-4: Summary of Open Finding F&Os and Disposition Status SR F&O Description Basis Suggested Resolution Disposition

$PR-B2 25-1 9 The draft documentation provided to the A significant number of f&OS Provide official The CEC internal events peer team included a summary of the status remain open and there is no documentation of all PRA model will be peer of Findings against the IF-IF PRA and clear basis for considering the open F&Os (with regard reviewed and F&Os FPRA, either Open, Closed, or Pending potential resolution to be not to the S-PRA model) and dispositioned. Upon Closed (But considered open for S-PRA). In relevant to the S-PRA. provide ajustification for resolution of F&Os, the addition, the summary indicated whether any open F&Os that are updated internal events the resolution to the F&O had been judged to be not model will be pulled in as integrated into the S-PRA. Additional relevant to the $-PRA. the base model for the $-

information was provided regarding that Alternatively, close the PRA.

integration, indicating that some ofthe open F&Os and assess the F&Os were documentation only and thus, impact of the changes on would have no impact on the S-PRA. Other the S-PRA model, open F&Os were listed as no anticipated whether they are impact on the S-PRA model. However, maintenance or upgrade this SR is based on relevance to the S-PRA, changes. For any not whether there would be a significant identified as upgrade, a impact. Ofthe 181 F&Os, 89 were listed as focused peer review Open or Pending Closed (Considered open would be required to for the $-PRA). While some of these are close the F&O.

documentation only F&Os, the majority of the open F&Os are more than documentation.

$PR-E6 25-12 Although many elements of the internal The current status of the CEC Elevate the internal The CEC internal events events LERF assessment is currently at least internal events LERF analysis events assessment to PRA model will be peer category II, the specific identified SRs are identified that several of the meet Cat II and validate reviewed and F&Os met at category I. Therefore the S-PRA is required SRs were met at Cat I. that the improvement is dispositioned. Upon assessed at Cat I. Since the S-PRA utilizes the appropriate for the S- resolution of F&Os, the internal event PRA model these PRA. updated internal events

$Rs carry forward and restrict model will be pulled in as the SR to Cat I. the base model for the 5-PRA.

Page 92 of 92