ML15156B331
| ML15156B331 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 06/09/2015 |
| From: | Arizona Public Service Co |
| To: | Office of Nuclear Reactor Regulation |
| DiFrancesco N, NRR/JLD, 415-1115 | |
| References | |
| Download: ML15156B331 (114) | |
Text
Palo Verde Nuclear Generating Station Seismic Hazard Re-evaluation NRC Public Meeting Rockville, Maryland June 9, 2015
Agenda
- Introduction/Overview
- Seismic Source Characterization SSHAC
- w/ Focus Area Questions
- Ground Motion Characterization SSHAC
- w/ Focus Area Questions
- Site Response
- w/ Focus Area Questions
- Probablistic Seismic Hazard Analysis
- Seismic Hazard Screening 2
3 Overview
- Safety is a core value for APS and Palo Verde Nuclear Generating Station
- New and extensive seismic hazard re-evaluation shows plant can safely withstand earthquakes
- Seismic re-evaluation was performed with independent experts in a transparent and open process
- Using new regulatory guidance, latest scientific methodologies and site-specific information, analysis demonstrates the plants design is earthquake safe
SSHAC - Objectives Update Seismic Source Characterization (SSC) and Ground Motion Characterization (GMC) models for use in Probabilistic Seismic Hazard Assessment (PSHA)
Site-specific SSC model developed following Senior Seismic Hazard Analysis Committee (SSHAC) Level 3 guidelines Regional GMC model for the southwestern U.S. (SWUS) developed by SWUS GMC SSHAC Level 3 project Communication between SSC and GMC projects ongoing during development of models SSHAC guidelines are summarized in NUREG/CR-6372 and NUREG-2117 Implemented PPRP data collection recommendations 4
- Quaternary Geologic Mapping;
- Spectral Analysis of Surface Waves;
- P-S Suspension Logging; - Broad Band Seismic Array Installation
Seismic Source Characterization SSHAC 5
Workshop 1: April 9-10, 2013
- Significant Issues and Data Needs
- Technical Integrator (TI) Team presentations, plus 14 Resource Expert presentations Workshop 2: September 24-25, 2013
- Alternative Interpretations
- TI Team presentations, plus 12 Proponent Expert presentations Workshop 3: April 23-24, 2014
- Preliminary Model and Hazard Feedback
- TI Team presentations, plus 2 Resource Expert presentations 6
SSC - Participatory Peer Review Panel (PPRP) Interactions 7
- Project Kickoff Meeting: January 21, 2013
- PPRP members, GMC Project Manager, and Project Sponsor
- Review Project Plan, SSHAC training
- Working Meetings
- PPRP members attended selected SSC TI Team working meetings throughout course of project
- PPRP Field Review of Geologic Mapping:
February 4-6, 2014
- Update on SSC Activities (post-Workshop 3):
June 18, 2014
Additional Interactions with SSC PPRP (contd) 8
- Final Briefing: July 10-11, 2014
- Update on SSC Activities: August 1, 2014
- Teleconference to Resolve Remaining Issues with SSC Report: February 19, 2015
- PPRP Closure Letter: February 26, 2015
Tectonic Setting of PVNGS 9
New Data, Models, and Methods for SSC 10
- New Data:
- Compiled earthquake catalog for study region (400 km radius)
- Quaternary geologic mapping of site vicinity (40 km radius), in collaboration with Arizona Geological Survey
- Site geophysical data collection for site response analysis:
borehole suspension logging and SASW
- New Models:
- Uniform California Earthquake Rupture Forecast, Version 3 (UCERF 3) - USGS, CGS, SCEC - Field et al. (2013)
- New Methods:
- Capture time-dependent behavior and uncertainties for high slip-rate strike-slip sources in CA (Equivalent Poisson Rates, or EPR)
- Virtual faults within the areal sources
SSC Overview 11
12 SSC: Areal Seismic Sources
13 SSC: All Fault Sources
14 SSC: Hazard-Significant Fault Sources
15 SSC Focus Area Question 1 NRC SSC Q1: Provide clarification on how the uncertainty derived from the difference in cumulative slip rates on Quaternary faults in Arizona in comparison to the geodetic slip rates in southern Arizona were evaluated and incorporated into the model. In addition, explain why faults that were originally part of the USGS National Seismic Hazard Mapping Project (NSHMP) and that are now part of the PVNGS SSC include alternative possibilities for geologic and geodetic slip rates while other PVNGS SSC faults that were not included in the NSHMP do not consider geodetic slip rates.
The differences in geologic and geodetic rates were evaluated in the following manner:
- Obtain E-W geodetic extension rates from Resource and Proponent Expert Corne Kreemer
- Compared to transect in central NV where fault slip rates and geodesy match (Koehler and Wesnousky, 2011)
- Evaluated the likelihood that faults are not expressed geomorphically (How many faults can hide in landscape? At what rates?)
- Evaluated how many faults required to produce geodetic extension rate
- Compared geodetic extension rates to strain rates implied by seismicity in source zones 16 SSC Focus Area Q1 Response
17 Transects of geodetically based east-west extension rates (in mm/yr) from Kreemer SSC Focus Area Q1 Response
18 Geodetic Extension Rates (in mm/yr) and % of extension that can be explained by fault slip rates SSC Focus Area Q1 Response
Could faults contribute to geodetic rate, but remain undetected in landscape?
19 Preservation and degradation of fault scarps in the desert landscape, modified after dePolo and Anderson (2000)
SSC Focus Area Q1 Response
How many normal faults required to produce geodetic extension rate?
20 Number of normal faults required to explain an east-west extension rate of 1.4 mm/yr (assumes various vertical separation rates and all faults are normal, strike north-south, and dip 50°)
SSC Focus Area Q1 Response
How many strike-slip faults required to produce geodetic extension rate?
21 Number of N40°W strike-slip faults required to explain an east-west extension rate of 1.4 mm/yr (assumes various dextral rates and all faults are strike-slip, strike N40°W, and dip 90°)
SSC Focus Area Q1 Response
Strain explained by source-zone recurrence rates Earthquakes in source zones also generate geodetic strain Calculation steps:
- Recurrence parameters in each source-zone cell Moment Rate in cell
- Moment Rate in cell, cell volume, etc. strain rate in cell
- Integration of strain rate along a transect deformation rate along transect 22 SSC Focus Area Q1 Response
Comparison to Kreemers Extension Rates 23 Black:
GPS extension rates derived by Kreemer Red:
Extension rates from earthquakes in areal source zones SSC Focus Area Q1 Response 1.4 mm/yr 1.5 mm/yr 0.8 mm/yr 0.1 mm/yr
Petersen et al. (2013) attempted to include geodetic rates in NSHMP SSC model Concluded not to use in off-fault or background zones Excluded block models (for faults), since they include no internal deformation and yield high rates Included two fault-based combined inversion models (Bird; Zeng and Shen), but gave them a low weight (0.2)
Low weight since data and models are relatively new, and are based on the limited understanding of why certain areas and models yield much higher short-term geodetic measurements when compared to longer-term geologic rates 24 SSC Focus Area Q1 Response
Summary of Geologic vs Geodetic Rate Evaluation Multiple assessments were made to understand if geologic data could allow, support, or even explain the relatively high geodetic extension rates The TI Team judged the geodetic rates were anomalously high compared to other technically defensible, long-term rate information for faults The distributed seismicity in the source zones generates deformation rates that are roughly consistent with the measured extension rates, and may help explain the difference between geological and geodetic slip rates on faults Unresolved questions regarding the applicability of using geodesy to model earthquake potential (temporal signals or perturbations, short-term vs long-term rates, seismogenic vs aseismic) led the TI Team to question the usefulness of directly applying the geodetic rates in the SSC To include a component of geodetic rates for fault sources, the geodetic rate applied to NSHMP fault sources by the USGS was incorporated directly into the PVNGS SSC 25 SSC Focus Area Q1 Response
PVNGS SSC model included geodetically-derived rates for only NSHMP fault sources This approach added a geodetic component of rate to the major faults that are common to both models Incorporates results of detailed study performed by USGS and other researchers for NSHMP TI Team did not include a geodetically-modeled rate on other faults in PVNGS SSC TI Team does not consider geodetic models to be as viable a representation of seismogenic strain accumulation as those models based on geologic slip-rate data 26 SSC Focus Area Q1 Response
27 NRC SSC Q2: With respect to the geologic mapping project, provide clarification regarding the stratigraphic correlation of Quaternary units used throughout the site area and site vicinity maps, please explain:
a) how the 16 Quaternary units on the site map resolve into five units on the site vicinity maps b) the rationale for singling out the river terrace unit (Qorh) as a stand-alone unit c) the apparent mismatch of units between the site area map and the site vicinity map along the north-south trending Qorh river terrace (along the eastern side of the site area map) with the Qi1 alluvial fan surface (to the north and the south of the Qorh unit on the site vicinity map)
SSC Focus Area Question 2
SSC Focus Area Q2
Response
28
29 NRC SSC Q2a: explain how the 16 Quaternary units on the site map resolve into five units on the site vicinity maps
Response
The site area map displays 16 Quaternary alluvial units and was mapped in more detail than the site vicinity maps The site vicinity maps display five Quaternary alluvial units, which represent a coarser grouping of the surficial deposits and was designed to place an emphasis on units that are approximately tens to hundreds of thousands of years and older that can then be interrogated for the presence of fault scarps and other signs of tectonic deformation The following table shows the general grouping of the site area map units within the generalized site vicinity map units SSC Focus Area Q2a Response
30 SSC Focus Area Q2a Response Site Vicinity Map Unit (40-km)
Site Area Map Units (8-km)
Qy - Holocene alluvium, undiff.
Qyc - Modern stream channel deposits Qy2 - Late Holocene alluvium Qy1 - Holocene alluvium Qyf - Fine-grained Holocene alluvium Qy - Holocene alluvium undiff.
Qye - Quaternary eolian deposits Qyp - Fine-grained deposits in playas Qi3 - Late Pleistocene alluvium Qi3 - Late Pleistocene alluvium Qi - Middle and late Pleistocene alluvium, undiff.
Qi2 - Middle to late Pleistocene alluvium Qi2 - Middle to late Pleistocene alluvium Qi - Middle and late Pleistocene alluvium, undiff.
Qi1 - Middle Pleistocene alluvium Qi1 - Middle Pleistocene alluvium Qorh Early to middle Pleistocene alluvium (of highest terraces along Hassayampa River)
Qi - Middle and late Pleistocene alluvium, undiff.
Qo - Early Pleistocene alluvium Qo - Early Pleistocene alluvium
31 NRC SSC Q2b: explain the rationale for singling out the river terrace unit (Qorh) as a stand-alone unit
Response
The Qorh unit delineates deposits associated with the highest terraces along the Hassayampa River. The Hassayampa River and additional Qorh deposits are located east of the site area map boundary Qorh is singled out to distinguish this specific type of deposit from other alluvial deposits of similar age (Qi1) that are not associated with the Hassayampa River Qorh appears on Pearthree et al.s (2006) geologic map of the Wintersburg quadrangle SSC Focus Area Q2b Response
32 NRC SSC Q2c: explain the apparent mismatch of units between the site area map and the site vicinity map along the north-south trending Qorh river terrace (along the eastern side of the site area map) with the Qi1 alluvial fan surface (to the north and the south of the Qorh unit on the site vicinity map)
Response
The 8-km site area map is based on Pearthree et al.s (2006) mapping of the Wintersburg quad, expanded to cover the entire site area The eastern portion of the 40-km site vicinity map is based on 1:100k mapping by the AZGS, modified by Phil Pearthree for this project. The western portion is based on original mapping The apparent mismatch between units is a function of comparing maps produced at two different scales with two different levels of detail Simplification and lumping of units in the site vicinity map Site vicinity map does not include small polygons and thin veneers SSC Focus Area Q2c Response
33 Portion of site area (8-km radius) map Portion of site vicinity(40-km radius) map 8-km radius Qorh Qi1 NE SE SSC Focus Area Q2c Response
NRC SSC Topic 3a: With respect to the areal source zones, provide the rationale for a.
choosing to model the spatial variation of the recurrence rate using variable, but continuous and relatively smooth seismicity rather than using uniform rates similar to the previously completed SSHAC Level 2 PVNGS SSC 34 SSC Focus Area Question 3a
Spatial pattern of earthquakes in SBR source zone is not consistent with the assumption of uniform seismicity 35 SSC Focus Area Q3a Response Map showing the PVNGS catalog (color coded by magnitude bin) and Seismotectonic sources
SSC TI Team adopted three alternative sets of magnitude weights, which yield different degrees of smoothness and are consistent with spatial distribution of the catalog data (representing the center, body, and range of technically defensible interpretations regarding degree of smoothing) 36 SSC Focus Area Q3a Response
NRC SSC Q3b: With respect to the areal source zones, provide the rationale for b.
not using a floor during the smoothing analysis of recurrence parameters as the host zones and other zones have cells with a rate of zero 37 SSC Focus Area Question 3b
- The cells with zero rate occur only in alternative realizations of the recurrence maps (as a way to represent uncertainty in rate), not in the mean recurrence maps
- The penalized likelihood approach introduces a natural floor by modeling the entire likelihood function of the rate 38 SSC Focus Area Q3b Response
Even when the earthquake count is zero, the likelihood function takes non-zero values (dark blue curve) 39 SSC Focus Area Q3b Response Natural Floor
Comparison to USGS, 2014 Recurrence rate (M>5) in host cell 40 SSC Focus Area Q3b Response
NRC SSC Q3c: With respect to the areal source zones, provide the rationale for c.
not using earthquakes with magnitudes lower than M4.67 to determine the recurrence parameters for the Eastern source zones 41 SSC Focus Area Question 3c
The SSC model used M3.33 and greater data for eastern source zones and M4.67 and greater for western sources (more abundant data) 42 SSC Focus Area Q3c Response
Typographical error in SSC Report Captions 43 Note: SSC Report is not a docketed report SSC Focus Area Q3c Response
Ground Motion Characterization SSHAC 44
SSHAC Workshops and Working Meetings (GMC)
Workshop 1: March 19-21, 2013 Critical Issues and Data Needs TI Team presentations, plus several Resource Expert presentations SSHAC Training Workshop 2: October 24-25, 2013 Proponent Models and Alternative Interpretations TI Team presentations, plus several Proponent and Resource Expert presentations Workshop 3: March 23-24, 2014 Preliminary GMC Model and Hazard Feedback TI Team presentations, with inputs from several Resource Experts 17 Formal Working Meetings 2 planning meetings and 1 extended meeting with Resource Expert presentations 45
Additional Interactions with GMC PPRP 46 Project Kickoff Meeting: August 27, 2012 TI Team, Project Technical Integrators, PPRP members, Hazard Analysts, Management and Project Sponsor attended in person Review Project Plan, discuss interface with SSC and Site Response Working Meetings PPRP members attended all GMC TI Team working meetings in person or via webinar throughout course of project PPRP Briefings (post-Workshop 3)
Pre-Briefing on May 14, 2014 and Final Briefing on July 17-18, 2014 PPRP members attended in person Review of Technical Report Several written and oral communications with PPRP members while resolving PPRPs comments and finalizing the Technical Report Endorsement letters received on February 24, 2015 and March 10, 2015
Reference rock ground-motion model (regional SWUS GMC SSHAC project)
- Median ground motion (two different source regions: Greater AZ and California/Mexico aka Regions 1, 2 & 3)
- Aleatory variability Interface with Site Response
- Accounting for: (1) differences in the rock conditions, and (2) site amplification Capture uncertainty in each part
- Epistemic uncertainties
- Avoid double-counting Ground-Motion Characterization Models 47
Data:
NGA-West2 strong-motion data set (PEER)
European strong-motion data set (RESORCE)
Taiwanese data (Lin et al., 2011)
Arizona Database (PEER)
Finite-fault simulations (SCEC)
New Data, Models and Methods for GMC 48 NGA-West2 RESORCE PEER-Arizona Lin et al (2011)
Finite-Fault Simulations Median for Greater AZ sources SS and NML events SS and NML events Kappa for Arizona rock site (Univ. Research task)
Earthquakes in Arizona Median for California/Mexico sources Earthquakes in California /Mexico (200-400 km)
Single-Station Sigma X
X X
Single path-to-region Sigma for California/Mexico sources Earthquakes in California /Mexico (200-400 km)
HW scaling X
Models:
Candidate median GMPEs: NGA-West2 and European GMPEs Adjustment for path effects Aleatory variability for residuals fatter than Gaussian distribution tail (Mixture model)
New Data, Models and Methods for GMC (Contd) 49 PVNGS - Greater Arizona Sources PVNGS - Distant California &
Mexico Sources Abrahamson et al (2014)
X X
Boore et al (2014)
X X
Campbell and Bozorgnia (2014)
X X
Chiou and Youngs (2014)
X X
Idriss (2014)
X Akkar et al (2014a, 2014b)
X Bindi et al (2014a, 2014b)
X
Methods:
Sammons map approach to develop weights to GMPEs, including comparison with empirical data Treatment of additional epistemic uncertainty Single-station sigma and single path-to-region sigma approaches New Data, Models and Methods for GMC (Contd) 50
Available Data:
10 earthquakes from California and Mexico recorded by 9 stations in central Arizona Systematic deviations from the average distance scaling for a given source-site pair Path term is a regional term rather than site specific (single-path-to-region)
Regionalization of GMC Model:
Greater AZ vs Regions 1, 2 & 3 51 Partially non-Ergodic GMC Path effects can be estimated, removing the path term from the aleatory variability in the traditional ergodic approach
Treatment of Path Adjustment 52 Mean Path Term Branches:
Upper and lower branches as mean path +/- 1.6 Combine results for Regions 2 and 3 Additional epistemic uncertainty included at T > 1 sec Scale the standard deviation at T > 2 sec by variance ratio of the Al Atik and Youngs (2014) additional epistemic variability with respect to its value at 2 sec PHI SP-R:
Evaluated at T = 0.5, 1 and 2 sec Constant for T < 0.5 sec and T > 2 sec
GMC Inputs to PVNGS Site Response 53 SWUS GMC host kappa value is 0.041 sec:
Estimated using the inverse random vibration theory (IRVT) approach (Al Atik et al., 2013) for a reference VS30 of 760 m/sec FAS HF slope for seven candidate GMPEs and nine short-distance scenarios
- PEER Report 2013/12 0
10 20 30 40 50 10
-4 10
-3 10
-2 10
-1 ASK14 - M7 - R20 - Vs760 Frequency (Hz)
FA (g-s)
Calc FAS k =0.043 sec SWUS GMC host profile (VS and density)
Reference VS profile for the host region is the Kamai et al. (2013*) profile for VS30 of 760 m/sec Reference density compatible with reference VS profile
GMC Inputs to PVNGS Site Response (Contd) 54 Target kappa value is 0.033 sec (ln = 0.5)
PEER Report 2014/09 VS profiles (2 generic profiles) for kappa estimates From 11 profiles (10 SASW and 1 existing PVNGS)
One profile for stiff soil sites (VS30670 m/s), and one profile for firm rock sites Extrapolated to source depth and merged into the host VS profile at depth with VS 3.5 km/s Chapter 4 130 m 130 m
NRC GMC Q1: Considering the limited bandwidth of the data recorded by the TA array, provide additional detail on the process used to estimate the target site kappa values and their uncertainties. In addition, describe in more detail how the site amplification at the recording stations was accounted for in the estimates of site kappa.
55 GMC Focus Area Question 1
Data Availability and Issues 12 earthquake in Arizona recorded by TA stations in central Arizona M1.2-3.4; R 10-300 km TA bandwidth limitations
- Sampling rate = 40 samples/sec
- Max freq = 16 Hz (80% of Nyquist) 56 GMC Focus Area Q1 Response
Combine the two methods to capture uncertainty AS and DS Methods 57 M > 3 M < 1.6 KAS computed between 9 and 16 Hz KDS computed between 4 and 16 Hz GMC Focus Area Q1 Response
AS and DS Methods: Results 58 GMC Focus Area Q1 Response Mean 0 = 0.033+/-0.014 sec Std. Deviation in 70 km 13 msec for AS method 6 msec for DS method Mean frequency-independent regional Q = 900+/-300 r estimates extrapolated to zero dist.
DS estimates larger than AS Large scatter in individual r Not explained by distance or site stiffness Possibly due to complex Q structure or shallow site resonance not accounted by average crustal amplification transfer function Results scatter is associated with differences between AS and DS approaches
BB Method 59 Fits entire spectrum Uses all the events and sites simultaneously The FAS is inverted for 0, stress parameter, and M Common crustal path damping parameter Q(f) and geometrical attenuation were used 1 starting model and 4 final models Q
mean 0(s) mean M Amplification Starting Model 200 0.68 0.040 1.97 rock/soil Model 1 200 0.68 0.033 2.14 rock/soil Model 2 152 0.72 0.034 2.17 rock/soil Model 3 1000 0.00 0.024 2.00 rock/soil Model 4 200 0.68 0.034 2.37 unity Best estimate So. Cal Q(f)
Freq. indep. Q No crustal ampl.
Modified from PEER Report 2014/09 GMC Focus Area Q1
Response
BB Method: Results 60 Mean 0 = 0.033 sec ln = 0.5 Corresponds to 0.020-0.054 sec Scatter evaluated by sensitivity analyses Results scatter is informed by sensitivity analyses Mean of individual inversion results for 14 sites from best estimate model (Model 1)
Variation in fixed input parameters Q(f)=Qo*f^n geo-spreading
stress drop crustal amplification 32% change in median by varying fixed parameters (taken independently)
Q0, eta and k are coupled due to limited bandwidth and distance range A ln = 0.5 in an appropriate value for epistemic uncertainty on median 0 estimate 60 GMC Focus Area Q1
Response
Uncertainty in Target Kappa Estimate 61 Consistent mean 0 estimates Consistent scatter:
AS/DS: 0.033+/-0.014 sec COV=0.014/0.033=0.43 ln=~0.41 BB: ln=0.5 (logarithmic range)
The BB Model 3s results (constant Q) are within the range of uncertainty from both approaches Consistent resulting Q (850-1050)
Within the scatter of independent estimates for the region GMC Focus Area Q1 Response
Treatment of Site Amplification 62
- Transfer functions were computed from the two reference extrapolated profiles They represent linear-elastic amplification, excluding any damping, from the source region (assumed to be at 10 km) to the surface at the site
- In all three approaches of kappa estimation (AS, DS, BB), the FAS are corrected for crustal amplification 1
elastic (Q) 3 5
16 Hz Stiff Soil Firm Rock 1 Hz GMC Focus Area Q1 Response
NRC GMC Q2: Provide additional detail regarding the evaluation of candidate GMPEs for PVNGS for distant California and Mexico sources. Specifically, describe in more detail the evaluation of Kishida et al. (2014) of the potential effects of the Q differences on the ground motions, which provides justification for the conclusion that although there are differences in Q between California and Arizona, these differences do not lead to a significant discrepancy in the average distance attenuation over the distance range of 200-400 km.
63 GMC Focus Area Question 2
Regional Q Evaluation:
Phillips et al. (2013) Model 64 Inversion for Q in individual bands along with moment and corner frequency source parameters Q(f) evaluated at 0.5 deg (50 km) grids Higher Q (lower attenuation) in Arizona (towards mid-west in general) as compared to California Effect more pronounced at higher frequency GMC Focus Area Q2 Response
Regional Q Evaluation: Phillips et al. (2013) Model (Contd) 65 Zoom in at regional level:
Compute Q(f)=Qo*f^n at grid points Q changes do not appear drastic Look at ray-paths from CA events to central Arizona Gridded values obtained from Phillips et al. (2013)
Q(f) ranges Q(f) < 100 150 < Q(f) < 225 100 < Q(f) < 150 225 < Q(f) < 340 340 < Q(f) < 500 500 < Q(f) < 750 750 < Q(f) < 1150 5 Hz 1 Hz GMC Focus Area Q2 Response
Regional Q Evaluation: Phillips et al. (2013) Model (Contd) 66 Path weighted attenuation (sum of 1/Q) in the 200-400 km range 14 NGA-West2 events with epicenters in CA recorded by stations in CA and AZ Q differences between CA and AZ are visible for short periods, but disappear at longer periods CA events contribute to PVNGS hazard at long periods (T > 0.5 - 1 sec)
Changes in Q between CA and AZ are negligible for hazard-significant frequencies From Kishida et al. (2015) - Chapter 5 GMC Focus Area Q2 Response
Adequacy of Attenuation with Distance Empirical check of evaluation:
6 events in Region 1, 4 events in Region 2 and 4 events in Region 3 PEER report compares distance scaling for four periods (0.2, 0.5, 1 and 2 sec)
Data corrected for event term and VS30 scaling (500 m/s)
Focus of hazard is for low frequencies from CA earthquakes The attenuation from California to central Arizona is not different from the attenuation within California at low frequencies 67 Chapter 5 GMC Focus Area Q2 Response
68 Attenuation with Distance: Region 1 (1 sec)
GMC Focus Area Q2 Response EQID 1005 EQID 1009 EQID 1047 EQID 1053 EQID 1182 EQID 118
69 Attenuation with Distance: Region 2 (1 sec)
GMC Focus Area Q2 Response EQID 1018 EQID 1020 EQID 1028 EQID 1067
70 Attenuation with Distance: Region 3 (1 sec)
GMC Focus Area Q2 Response EQID 280 EQID 1004 EQID 1017 EQID 1058
NRC GMC Q3: With regard to the candidate GMPEs for PVNGS for the Greater Arizona Sources, provide additional rationale for limiting the use of the Bindi et al. (2014) for magnitudes greater than M7.
71 GMC Focus Area Question 3
Problematic intermediate-frequency magnitude scaling (M>7):
Apparent in the 2011 version Resolved for PGA in the 2014 version, but not for 5 Hz Application limited at M7 for all periods Limitations in Bindi et al. (2014) GMPE 72 From TI Teams WS#3 presentation (Day 1) 5 Hz 100 Hz 1 Hz GMC Focus Area Q3 Response
NRC GMC Q4: Provide additional detail on the development of the common function form used to fit the candidate GMPEs. Specifically, please discuss how model parameters such as depth to Vs=1 km/s and 2.5 km/s (which are present in some of the candidate GMPEs) are accounted for in the final functional form.
73 GMC Focus Area Question 4
Two Alternative Distance Metrics:
RRUP-based common-form models (Model A)
RJB-based common-form models (Model B) > no explicit ZTOR scaling (a8=0)
Applicable Style-of-Faulting:
The common form is derived for SS and NML earthquakes only and a single REV factor is added at the end Greater AZ Median Model: Common-form Functional Form 74
(
)
(
) (
)
2 2
2 0
7 8
9 2
2 4
5 6
10 1
2 1
3
(
, )
( )
( )
( )
( )
( )
( )(
- 5) ln
( )
( )
( )
( )(
5.5) 5.5
( )(
6.5) 5.5 6.5
( )(
6.5) 6.5 A
RUP TOR RUP TOR NML RUP RV Ln SA M R Z
F T a T a T R a T Z a T F a T a T M
R a T a
T F a T a T M
for M a T M
for M
a T M
for M
=
+
+
+
+
+
+
+
GMC Focus Area Q4 Response
Fitting Process:
VS30 = 760m/s, M5 to M7.5, RX = -200 to 18 km (RX>0 for RJB=0km conditions), dip of = 50° (NML events)
Site parameters, Z1.0 and Z2.5 are set to their default values for VS30=760 m/s by NGA developers Basin depth is not a significant issue for soft-rock sites HW model added to RRUP-based common-form models No HW model added to RJB-based common-form models Common-form Models: Fitting 75 GMC Focus Area Q4 Response
Site Response 76
PVNGS Site Response
- Two elements
- Adjustment of motions from SWUS Reference Rock to PVNGS rock
- Site response proper (conversion of PVNGS rock motions to the surface, taking into account the effect of soils) 77
Site Geophysical Data Collection for Site Response Analysis 78
Vs Comparison (SASW, Suspension Logging, Downhole, and Cross-Hole) 79
Adjustment from SWUS Reference Conditions to PVNGS Rock (1)
Inputs:
- From SWUS Project: Reference (host) Vs and density profiles, host kappa; PVNGS (target) kappa, including uncertainty
- From PVNGS PSHA Project: PVNGS (target) Vs and density profiles from local and regional data (Deep Profile)
Approach (Vs-kappa adjustment)
- Assume linear behavior
- Use quarter-wavelength approach
- Consider uncertainty in PVNGS profiles and kappa
- Result: 9 adjustment factors (transfer functions in Fourier space*), with associated weights
- requires IRVT-adjustment-RVT for calculation of Spectral Acceleration 80
Adjustment from SWUS Reference Conditions to PVNGS Rock (2: Inputs)
- Reference kappa
- 0.041 s
- PVNGS kappa
- Median: 0.033 s
- ln[kappa]: 0.5 81 0
500 1000 1500 2000 2500 3000 0
1000 2000 3000 4000 Depth (m; from bottom of shallow profile)
Vs (m/s)
UB Profile Base Case Profile LB Profile Warren (1969)
SWUS Reference Profile
Adjustment from SWUS Reference Conditions to PVNGS Rock (3: Results) 82 0.1 1
10 0.1 1
10 100 Adjustment Factor (on Fourier Spectrum)
Frequency (Hz)
LB Profile, Median kappa (0.12)
LB Profile, UB kappa (0.09)
Median Profile, LB kappa (0.12)
Median Profile, Median kappa (0.16)
Median Profile, UB kappa (0.12)
UB Profile, LB kappa (0.09)
UB Profile, Median kappa (0.12)
UB Profile, UB kappa (0.09)
Site Response Inputs (shallow profile, range is representative of the 3 sites)
- Stratigraphy from UFSAR and PSAR
- Vs from suspension logging, SASW, and UFSAR
- Degradation Curves
- Sands: EPRI (1993) and Peninsular Range
- Vucetic and Dobry with a broad range of PI Approach:
- Profile randomization following SPID
- Standard SHAKE-style calculation using RVT 83
Site Response: Shallow Profile 84 Randomized Vs profiles for base case (BC) median profile
0 0.5 1
1.5 2
2.5 3
3.5 0.001 0.01 0.1 1
10 Median SAF Spectral acceleration, (g)
PGA 20 Hz 10 Hz 5 Hz 2.5 Hz 1 Hz 0.5 Hz Amplification Factors (ReferenceSurface; base case) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.001 0.01 0.1 1
10 Log-standard deviation SAF Spectral acceleration, (g)
PGA 20 Hz 10 Hz 5 Hz 2.5 Hz 1 Hz 0.5 Hz 85
NRC Site Response Q1a: Provide additional detail regarding the Vs-kappa adjustment factors. Specifically, provide the bases for a.
the host Vs profile 86 Site Response Focus Area Question 1a
87 SWUS GMC host profile (VS and density):
Reference VS profile for the host region is the Kamai et al.
(2013*) profile for VS30 of 760 m/sec Representative of NGA-West 2 rock sites Consistent with Host kappa value of 0.041 s Reference density compatible with reference VS profile
- PEER Report 2013/12 Site Response Focus Area Q1a Response
NRC Site Response Q1b: Provide additional detail regarding the Vs-kappa adjustment factors. Specifically, provide the bases for b.
the target deep Vs profile (including the use of a logarithmic standard deviation of 0.35 to develop the upper and lower profiles) 88 Site Response Focus Area Question 1b
Top portion (andesite, basalt, flow-breccia, and tuff)
Vs from suspension-log data (LCI-B-2 profile)
Thickness (& uncertainty) from UFSAR boreholes (consistent with Warren, 1969*)
- Geological Society of America Bulletin, 80(2), 257-282; used by Lockridge et al. (2012) for earthquake location 89 0
500 1000 1500 2000 2500 3000 0
1000 2000 3000 4000 Depth (m; from bottom of shallow profile)
Vs (m/s)
UB Profile Base Case Profile LB Profile Warren (1969)
SWUS Reference Profile Site Response Focus Area Q1b Response
Basement portion (granodiorite, metagranite)
Vs and thickness from Warren, 1969)
Used 10% uncertainty for thickness of top basement layer 90 0
500 1000 1500 2000 2500 3000 0
1000 2000 3000 4000 Depth (m; from bottom of shallow profile)
Vs (m/s)
UB Profile Base Case Profile LB Profile Warren (1969)
SWUS Reference Profile Site Response Focus Area Q1b Response
Rationale for 0.35 uncertainty in velocity:
Recommended by SPID (Section B.3.2) for sites where geophysical information such as very limited shear-wave velocity data exists 91 0
500 1000 1500 2000 2500 3000 0
1000 2000 3000 4000 Depth (m; from bottom of shallow profile)
Vs (m/s)
UB Profile Base Case Profile LB Profile Warren (1969)
SWUS Reference Profile Site Response Focus Area Q1b Response
NRC Site Response Q1c: Provide additional detail regarding the Vs-kappa adjustment factors. Specifically, provide the bases for c.
the target kappa value used for the kappa adjustments and whether the input FAS were corrected to the site kappa of 0.033 sec or a lower baserock kappa value 92 Site Response Focus Area Question 1c
Basis for target kappa value of 0.033 sec was provided in response to GMC Topic 1 Input (Reference Rock) FAS were corrected to 0.033 sec (+uncertainty), which corresponds to Arizona rock 93 Site Response Focus Area Q1c Response
NRC Site Response Q1d: Provide additional detail regarding the Vs-kappa adjustment factors. Specifically, provide the bases for d.
use of a logarithmic standard deviation of 0.5 to determine the upper and lower site kappa values 94 Site Response Focus Area Question 1d
Basis for logarithmic standard deviation was provided in response to GMC Topic 1 Applied as follows:
95 Kappa (sec)
Weight 0.033*exp[-1.28*0.5]=0.0174 0.3 0.033 0.4 0.033*exp[+1.28*0.5]=0.0626 0.3 Site Response Focus Area Q1d Response
NRC Site Response Q1e: Provide additional detail regarding the Vs-kappa adjustment factors. Specifically, provide the bases for e.
the scenario events (magnitudes and distances) used to develop the input spectra for the Vs-kappa adjustment factors 96 Site Response Focus Area Question 1e
Vs-kappa adjustment was applied to each of the ground motion spectra used in the site-response calculations Spectral shapes are based on the 10-4, 10-5, and 10-6, High-and Low-frequency Controlling Earthquakes, calculated using the approach in RG 1.208 97 Site Response Focus Area Q1e Response
Controlling Earthquakes 98 Site Response Focus Area Q1e Response
NRC Site Response Q1f: Provide additional detail regarding the Vs-kappa adjustment factors. Specifically, provide the bases for f.
not including the Vs-kappa adjustment factors as additional epistemic uncertainty on the median GMMs instead of capturing this variability as part of the variability in the site amplification functions 99 Site Response Focus Area Question 1f
- Because NUREG/CR-6728 Approach 3 was used to convolve the rock hazard and site response, decision to include uncertainty in Vs-kappa adjustment factors as an uncertainty in the amplification functions has no effect on the mean hazard (no effect on the GMRS)
- This decision was made for the sake of computational efficiency in hazard calculations 100 Site Response Focus Area Q1f Response
Probabilistic Seismic Hazard Analysis Results 101
Hazard Sensitivities (rock) 102
Hazard Sensitivities (rock; 2) 103
Hazard Results on Soil 104
Hazard Results on Soil (2) 105 0.01 0.1 1.
10.
0.1 1
10 100 Spectral acceleration, g Spectral frequency, Hz Mean Horizontal Soil UHRS and GMRS 10- UHRS 10- UHRS GMRS 10- UHRS
106 PSHA Summary and Observations
- Approach and Scope
- SSHAC Level 3 SSC and GMC Models
- Site response used newly acquired site-specific SASW and suspension-logger data (as well as data from UFSAR), and followed Appendix B of SPID Robust characterization of seismic hazard at PVNGS and its uncertainty
- Most important contributors to hazard
- Area source zones (mainly SBR and East) dominant for 1 and 10 Hz and for 10-4, 10-5, 10-6
- Faults (mainly San Andreas, Cerro Prieto, and San Jacinto) important only at 1 Hz for 10-4
107 PSHA Summary and Observations Largest contributors to epistemic uncertainty in hazard
- GMC
- Common-form models for Greater AZ
- NGA-West 2 models for CA-MX faults (low frequencies only)
- Path term for CA-MX faults (low frequencies only)
- Additional epistemic uncertainty for CA-MX faults (low frequencies only)
- Total sigma for CA-MX faults (low frequencies only)
- Site Response
- Vs profile (BE, UR, LR) (low frequencies only)
- Degradation Curves (high frequencies only)
- SSC
- Mmax of SBR source (low frequencies only)
Seismic Hazard Screening 108
109 PVNGS Current Licensing Basis
- UFSAR Section 2.5 Geology, Seismology, and Geotechnical Engineering
- 10 CFR Part 100, Appendix A, site characterization Safe Shutdown Earthquake (SSE) - 0.20g peak ground acceleration (PGA) value (UFSAR Figure 2.5-94)
- UFSAR Section 3.7 Seismic Design
- Seismic Category I Structures, Systems and Components (SSCs) designed to spectral response curve anchored at 0.25g PGA (UFSAR Figure 3.7-1)
110 Screening Evaluation
- In accordance with the 10 CFR 50.54(f) letter:
- Re-evaluated Seismic Hazard Ground Motion Response Spectrum (GMRS) was developed for purpose of screening for additional evaluations
- PVNGS screening evaluation performed based on comparison of Ground Motion Response Spectrum (GMRS) with Design Spectral Response Curve
111 Screening Evaluation PVNGS Seismic Category I SSCs have code-allowable seismic capacities above re-evaluated GMRS
112 Screening Evaluation
- Risk Evaluation Screening (1 to 10 Hz, and >10 Hz)
- Design Spectral Response Curve exceeds GMRS response curve
- Not required to perform a Seismic Risk Evaluation
- Not required to perform a Spent Fuel Pool Evaluation
- Not required to perform the High Frequency Confirmation Interim Actions
- PVNGS meets criteria in Screening, Prioritization, and Implementation Details (EPRI, 2013) for screening out
- Interim actions are not required
- Conclusion
- Further action is not required for NTTF Recommendation 2.1 Seismic Review
113 Questions
114 Backup Slide