ML17331B244

Seismic Hazard Sensitivity Analysis DC Cook Nuclear Plant Bridgman,Mi

ML17331B244
Person / Time
Site:	Cook
Issue date:	04/09/1992
From:	RIZZO, P.C.
To:
Shared Package
ML17331B245	List: ML17331B244
References
NUDOCS 9402250239
Download: ML17331B244 (58)

Text

REX'ORT SEISMIC &LZAIEDSENSITIVITYANALYSIS DONALDC. COOK NUCLEAR PLAZA'RIDGlVRQf, MICHIGAN PaoszcT No.89-654.30 Arm. 9, 1992 PAUL C. Ruzo Assoc'.TEs, INc.

300 Omoao Darvz MomoEvu.LE, PENNsvx.v~txA. 15146 PHoNE:

(412) 856-9700 TELEFAx: (412) 856-9749

'~n 9402250239 9402ii PDR ADOCK 05000315 P.

PDR

TABLEOF CONTlPl'IS PAGE LI

~ ~

ST OF TABLES.............................................................................

u LI

~ ~

ST OF FIGURESo ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ oo ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~

~ ~ ~ ~ ~ o 111

1.0 INTRODUCTION

~ ~ ~ ~ ~ ~ ~ ot ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ to ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 1

2.0 BACKGROUND

2 3.0 METHODOLOGYo ~ oo ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ oo

~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o 3 4.0 APPROACH TO SENSIIIVITYANALYSIS...................................

4 5.0 RESULTS AND DISCUSSIONS t ~ ~ ~ ~ ~ o ~ ~ oo ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ t ~ ~ t 6 5.1 SEISMIC SOURCE ZONES

~ ~ toto ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 6 5.2 5.3 5.4 EISMICHYPhRhMEIM

~ ~ ~ ~ ~ ~

~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o 7 S

ROUND MOTIONMODEL................................................

8 G

SEISMIC HhZhRD hT THE DRESDEN SITE...............................

9 6.0 SUhQ~Y AND CONCLUSIONS.............................................

10 REHHKNCES TABLES FIGURES

LIST OF TABLES SEISMIC E&ZARDMETHODOLOGY COMPARISON OF LLNL,EPRI, AND RIZZO ASSOCIATES CASES CONSIDERED IN SEISMIC E&ZARDSENSITIVITYANALYSIS COMPARISON OF MEANSEISMIC E&ZARDFROM LLNLSE-6 BEST ESTIMATEAND ALTERNATESOURCE ZONE MODELS COMPARISON OF MEANSEISMIC HAZARDFROM LLNLAND RIZZO SKSAHCIYYPARAMHIMISIR-6 SOURCE ZONE MODEL COMPARISON OF MEANSEISMIC HAZARDFROM LLNLAND RIZZO SHISAHHMY PARAMHIRRS AGGREGATED OVER ALLLLNL SEISMIC EXPERTS EFFECT OF GROUND MOTION MODELON SEISMIC HAZARD

LIST OF FIGUEQS THE BASIC STEPS OF A PROBABILISTICSEISMIC HAZARD ANALYSIS SEISMICITY OF THE EASTERN UNXIEDSTATES BEST ES'IIMATESEISMIC SOURCE MODELLLNLEXPERT 6 EFFECTS OF ALTERNATESOURCE ZONES ON MEANHAZARDCURVE

~ENLOCHP I~ PER LLNL SE-6 DOMINANTSOURCE ZONES EFFECT OF SEISMICITY P IIAMEIKKON 15IZ PHRCENTILE HAZARDCURVE EFFECT OF SEISMICITY IARAMEIZKON MEDIANHAZARD CURVE EFFECT OF SEISMICITY P RAMEIZKON 85TH PERCENTILE HAZARDCURVE EFFECT OF SEISMIClTY P l!AMEIKNON MEANHAZARD CURVE EFFECT OF GROUND MOTION MODELS ON 15TH PERCENTILE HAZARDCURVE 111

LIST OF FIGVKH (Continued)

TED<;

EFFECT OF GROUND MOTION MODELS ON MEDIANHAZARD CURVE EFFECT OF GROUND MOTION MODELS ON 85TH PERCERI ILE HAZARDCURVE EFFECT OF GROUND MOTION MODELS ON MEANHAZARDCURVE

, 15TH PERCEYI ILEHAZARDCURVE DRESDEN SITE 50TH PERCENTILE HAZARDCURVE DRESDEN SITE 85TH PERCENTILE HAZARDCURVE DRESDEN SITE

REPORT SEISMIC HAZARDSENSH'IVITYANALYSIS DONALDC. COOK NUCLF~ PLANT BRIDGMAN,MICHIGAN

1.0 INTRODUCTION

This report describes the analysis performed by Paul C. Rizzo Associates, Inc. (Rizzo Associates) to study the sensitivity of the seismic hazard at the Donald C. Cook Nuclear Plant (Cook Nuclear Plant) site to inputs used to calculate the hazard.

Rizzo Associates'eport dated April 1, 1991, presented the project methodology and documented the resulting seismic hazard curves for the site in support of American Electric Power Service Corporation's (AEPSC) Individual'Plant Examination for External Events (IPEEE) program.

The project methodology was presented and discussed with the U.S. Nuclear Regulatory Commission (USNRC) Staff on May 10, 1991.

As demonstrated to the Staff, Rizzo Associates approach is similar to that used by the Electric Power Research Institute (EPRI, 1988).

The results were generally consistent. with those published by the Lawrence Livermore National Laboratory (LLNL,Bernreuter et al., 1989).

Consistent with NV'REG 1407 (USNRC, 1990), AEPSC plans to use both LLNLand Rizzo Associates hazard results in the IPEEE, so that the resulting ranking in dominant failure sequences includes the desired robustness.

In light of its review of the EPRI results, the Staff strongly encouraged sensitivity analysis to demonstrate the stability of the seismic hazard curves calculated by Rizzo Associates (April 1991).

The study reported herein was performed in accordance with the scope of work for Program 2 in our proposal dated April23, 1991. It addresses the Staff s comments and evaluates the sensitivity of the seismic hazard at the site to the inputs used to represent the variability in the phenomena and uncertainty in modeling.

The results of this study support the conclusions documented in Rizzo Associates'arlier report.

The following sections of this report discuss the background, review the project methodology relative to the LLNLand the EPRI methodologies, describe the approach to the sensitivity analysis and present the results, conclusions and recommendations.

0

2.0 BACKGROUND

Prior to the Rizzo Associates hazard analysis, the seismic hazard at the Cook Nuclear Plant site was defined only by the LLNLstudy (Bernreuter, et al., 1989)., EPRI's

'rogram (EPRI, 1988), which developed hazard curves for some 50 Eastern U;S. sites (the Cook Nuclear Plant site was not included), indicated that the approach used in accounting for the modeling uncertainty could result in relatively significant differences in the calculated hazard.

The Commission's review (USNRC, 1988) concluded that the EPRI methodology was equally as acceptable as LLNL's.

The USNRC's guidelines (USNRC, 1990) recommend that LLNLas well as-EPRI seismic hazard curves (at sites where both hazard curves exist) be used in the IPEEE, to demonstrate that the ranking of the dominant failure sequences is consistent.

Where alternate hazard curves do not exist, the guidelines allow their development and subsequent use in the IPEEE. 'Accordingly, the Cook Seismic Hazard Program objectives are to establish alternate seismic hazard curves using the EPRI methodology as closely as possible.

Due to its proprietary nature, not all elements and inputs of the EPRI methodology were available for use in this analysis.

Consequently, the project methodology could not duplicate the EPRI study, but preserves in intent and philosophy the significant steps of EPRI's hazard evaluation.

The Rizzo Associates report (April 1991),

documents the project methodology and the resulting seismic hazard curves at the Cook site and demonstrates that the hazard curves generally exhibit similar trends to those reported by LL'NL. The report concludes that the LLNLhazard curves are more conservative and exhibit a larger dispersion in the results, than the Rizzo Associates hazard curves.

3.0 METHODOLOGY Figure 1 presents the major steps of a seismic hazard calculation.

In addition to the input parameter variability, each step is associated with modeling uncertainty.

Differences in the seismic hazard methodologies result from the manner in which these uncertainties are considered in the analysis.

Table 1 presents the major steps in the seismic hazard calculation and compares the manner in which LLNL's, EPRI's and Rizzo Associates'ethodologies differ in treating the uncertainties associated with these steps.

Table 1 illustrates that, with the exception of the seismic source zone models, other elements of the Rizzo Associates'alculation follow the EPRI methodology.

As stated in Rizzo Associates report (April, 1991), the Project Methodology uses a total of 16 source zone maps, each describing the best estimate of independent experts or expert teams.

Since each best estimate is given equal weight in the analysis, this method avoids problems associated with aggregating the results in accordance with subjective weights given to the various alternate source zone maps.

The Project Methodology developed magnitude recurrence relationships for the seismic source zones using a standardized approach similar to EPRI's.

As EPRI's earthquake catalog was unavailable for earthquakes ofM,, (5.0 or I, < VII, Rizzo Associates independently developed the remainder of the earthquake catalog foQowing the methodology and incompleteness data developed by EPRI.

However, unlike EPRI, seismicity was assumed to be uniformly distributed over entire seismic zones and was not discretized over smaller areas within the zone.

This avoided the potential for unrealistic cell-to-cell fluctuations in the seismicity parameters possible in the EPRI methodology (USNRC, 1988).

At the outset, it was intended to use EPRI's ground motion model to calculate the seismic hazard.

However, after examining the recently recorded data from the 1988 Saguenay earthquake, the model was refined to improve its predictive capability relative to the recorded data.

The project methodology uses simulation to generate the statistical information for the hazard.

This is equivalent to EPRI's logic tree approach but, in certain cases, improves the representation of the tail regions of the distribution functions.

4.0 APPROACH TO SENSITIVITYANALYSIS Several parameters affect the seismic hazard functions.

Among these, the seismic source zone models, the seismicity parameters defining the recurrence rates of earthquakes, and the ground motion model are the more predominant.

Rizzo Associates'eismic hazard methodology uses published data (e.g., LLNLand EPRI) and judgment to select the mean and variability of the parameters with particular reference to the Cook Nuclear Plant site.

As the comments in the USNRC's Safety Evaluation Report (SER, USNRC, 1988) suggest, the elimination of nonstandard hypotheses (e.g., seismic source zonation and seismicity parameters) and outliers in ground motion models are of particular concern to the Staff. The approach to the sensitivity analysis is directed to these concerns.

This study develops and documents the effects of the choice of ground motion model and the seismicity parameters on the median hazard curves for the Cook Nuclear Plant site.

Although truncation of ground motion beyond a certain acceleration is conceivable, not enough data are currently available to justify this action.

Hence, the sensitivity study reported here does not address truncation.

Additionally, hazard curves for the Dresden site were developed to allow a direct comparison with EPRI's results.

The Dresden site was chosen because of its proximity to the Cook Nuclear Plant site and because the structures at the Dresden plant are founded on rock. The use of a rock model allows for the comparison of the Rizzo Associates methodology with that ofEPRI and LLNLwithout the need for considering the effects of soil amplification which is a separate issue to the assessment of seismic hazard.

The models for the source zones, recurrence relationships and attenuation relationships used for the Cook Nuclear Plant site could be used unchanged to derive the hazard curves for the Dresden site.

Table 2 presents the various cases analyzed to document the sensitivity of the seismic hazard at the Cook Nuclear Plant site.

Case 1 is as reported in Rizzo Associates report (April 1991). For the sensitivity study reported here, this is considered as the base case which willbe challenged by the results of the other cases.

Case 2 evaluates the effects of the choice of seismicity parameters.

As EPRI s seismicity parameters were unavailable, Case 2 considers the 11 LLNLsource zone models and the Rizzo

Associates source zone and incorporates the seismicity parameters reported by LLNL.

Cam 3 and 4 evaluate the effects of the specific attenuation laws included in the gmund motion model. Finally, Case 5 calculates the hazard at the Dresden site and allows comparison with results reported by EPRI.

5.0 RESULTS ANDDISCUSSIONS This section discusses the effects of the different inputs on the calculated hazard functions.

Because the LLNLand EPRI studies are well publicized and reviewed, the following discussion is set out within the framework of these studies and addresses the specific points of departure from their methodologies.

5.1 Smmc Souacz Zoos As shown on Figure 2, historic seismicity alone suggests that the Cook Nuclear Plant site is located in one of the least seismic areas of the eastern U.S. In evaluating seismic hazard potential, seismic source zones are generally delineated on the basis ofboth the historic seismicity, as well as geologic and tectonic data such as surface expression of

, faults and subsurface characteristics developed from geophysical information. Due to the general lack of surface expression of faults in the eastern U.S., the identification of tectonic features here has to rely extensively on geophysical information. This process is fraught with uncertainty in interpreting the data, and the uncertainty is reflected in the weights assigned to source zones proposed by EPRI and LLNLexperts on the basis of tectonics.

The EPRI methodology implements the Features Matrix Methodology to incorporate the uncertainty in delineating source zones on the basis of tectonics.

This method deteanines the probability that a seismic source zone is active as the product of the probability that any hypothetical zone is active given the existence of a particular combination of geophysical characteristics and the probability of the existence of each characteristic.

Interestingly, and as noted in the Commission's SER (USNRC, 1988)

"...the Feature Matrix Methodology (FMM) data presented by the different earth science teams were biased toward recorded (historic) seismicity. Allother characteristics being the same, the probability ofactivity assigned to a feature was generally larger ifit had a favorable association with seismicity than ifit had a favorable association with another characteristic, such as a particular tectonic feature.

To illustrate the effect of source zone uncertainty on the calculated seismic hazard, consider for example, LLNLs Seismic Expert 6 (SE-6) input. Figure 3 presents his best estimate source zone map.

One of his alternate maps was to consider the entire

eastern US as one seismic zone.

This was perhaps prompted by the, fact that geophysical data reveals no definite direction for basing source zones on tectonics.

Clearly this represents an extreme. case among alternates proposed by other experts.

Although Expert 6 assigned a weight of only 10 percent to this alternate, when included in the hazard calculations, the alternate map dominates risk at higher values of peak ground acceleration (PGA). Qualitatively, the alternate penalizes sites in seismically quiet areas, particularly such as the Cook Nuclear Plant site, by allowing that very large earthquakes could occur at or near the site.

Figure 4 iOustrates the above effect in terms of the calculated seismic hazard for the Cook Nuclear Plant site using LLNLSE-6's best. estimate map and seismicity and his alternate map and seismicity.

Both seismic hazard curves in Figure 4 are based on Rizzo Associates ground motion model. For reference, Figure 4 also includes the hazard curve calculated on the basis of Rizzo Associates seismicity parameters assigned to SE-6's best estimate source zones.

Table 3 compares the mean exceedance probabiTities for specific values of Peak Ground Acceleration (PGA) calculated from Expert 6's best estimate and alternate source zone models.

These results illustrates the effect of SE-6's alternate map on the calculated hazard particularly at PGA's associated with lower'exceedance probabilities.

In our opinion, an expert s best estimate seismic hazard model reflects his or her best judgment and therefore can be included as a fullyvalid model in the hazard calculations regardless of its effect on the final result.

The expert's alternate model and the probability assigned to it, on the other hand, are representative only to the extent that the expert appreciates the effect of this alternate on the outcome, and the manner in which the outcome willbe used.

The probabilities assigned to alternate that models may be interpreted differently with respect to the different experts and are therefore subjective.

The above analysis illustrates that the alternate source zone models in the LL'NLanalysis could have a significant effect on the calculated site seismic hazard and its variability.

5.2 Szrmrnrv Pmarmrzas The Rizzo Associates methodology uses a standardized procedure for calculating the seismicity parameters describing the recurrence rates for earthquakes in each source zone.

As detailed in Rizzo Associates report (April 1991), the seismicity parameters

calculated for the source zones are based on a uniform earthquake list for the site region and includes EPRI's incompleteness data.

The list assigns a uniform magnitude to all events, excludes aftershocks and foreshocks and events less than magnitude 4.0.

For most of the nuclear power plant sites evaluated by LLNL,the zone containing the site contributes most significantly to the seismic hazard.

For example, LLNL'sSE-6 best estimate map places the Cook Nuclear Plant site in Zone 22 (Figure 3). The hazard analysis using SE-6's best estimate map identifies Zones 22, 17, 9, and the residual zone predominant contributors to the site hazard.

Figure 5 presents the LLNL and Rizzo Associates best estimate recurrence relations for these zones in comparison with recorded data.

This figure illustrates that the parameters defined by Expert 6 overestimate the seismicity of the predominant source zones.

As expected, Figure 4 illustrates that Expert 6's seismicity parameters result in a

'ignificantly higher seismic hazard.

The curves shown in Figure 4 are developed for SE-6's best estimate source zones and use Rizzo Associates ground motion model.

Table 4 compares the mean exceedance probabilities for specific PGA's calculated from Expert 6's and Rizzo Associates'eismicity parameters assigned to the expert's best estimate source zones.

These results represent the degree of conservatism that can be attributed fo an individual LLNLexpert's seismicity parameters.

The influence-of-the seismicity parameters aggregated over all-11 LLNLexperts'est estimate seismic source zone models is illustrated on Figures 6 through 9 which present the aggregated hazard curves.

Table 5 presents and compares the corresponding mean exceedance probabilities for specific PGAs.

The results indicate that the mean and median hazard corresponding to the LLNLexperts'eismicity parameters are somewhat higher than predicted using Rizzo Associates seismicity parameters.

Additionally, the fractiles also illustrate that the 15 percent variability in the seismicity parameters assumed in the Rizzo Associates analysis results in a smaller dispersion of the seismic hazard.

5.3 GROUND MOTIONMODEL As stated above, the Rizzo Associates analysis uses a ground motion model consisting of four attenuation relationships.

Three of these correspond to the attenuation relationships recommended by EPRI for Eastern North America (EPRI, 1988) and the fourth is SE-2A attenuation reported by LLNL(Bernreuter, et al., 1989).

The study reported herein evaluates the sensitivity of the site seismic hazard to the ground motion model by evaluating and comparing the seismic hazard resulting from three different ground motion models:

~

Rizzo Associates model used in the project methodology (Rizzo, 1991);

~

The model used in the LLNLstudy (Bernreuter, et al.,

1989); and

- ~

The model used in the EPRI study (EPRI, 1988).

Allof the above analyses use the Case 1 of source zone maps and seismicity parameters presented in Table 1. Figures 10 through 13 compare the calculated site seismic hazard using Rizzo Associates, EPRI, and LLNLground motion models aggregated over all seismic experts.

These figures illustrate that the ground motion model used in the project methodology is less conservative than" LL'NL'sand somewhat more conservative than EPRI's model.

Table 6 presents the difference in the mean exceedance probabilities for specific values of PGA. Table 6 indicates that the use ofLLNL ground motion model results in a seismic hazard at the SSE level of 0.2g PGA, which is a factor of two higher than obtained using Rizzo Associates ground motion model.

5.4.

Szmvac Km'rne Dazsomr sxxz In addition to the above sensitivity analyses, the study reported herein uses the project methodology to calculate the seismic hazard at the Dresden site so as to allow a direct comparison with EPRI's results for Dresden.

The Dresden site is near the Cook Nuclear Plant site; accordingly, the source zone models for the Cook Nuclear Plant site were used unchanged to derive the hazard at Dresden.

EPRI's hazard curves at Dresden correspond to rock outcrop as the Dresden plant structures are founded on rock.

Figures 14 through 16 compare respectively, the 15th, 50th, and 85th percentile seismic hazard curves for the Dresden site developed using the project methodology to those reported by EPRI and LLNL. These figures show that, relative to EPRI's hazard curves, the Project Methodology yields hazard curves that are similar to but more conservative than EPRI's.

The dispersion in the hazard, represented by the 15th and 85th percentiles curves, is similar to EPRI's and significantly less than LLNL's.

10 6.0 SGV94LRY AND CONCLUSIONS This report presents the analysis performed to study the sensitivity of the probabilistic seismic hazard at the Cook Nuclear Plant site to source zone modeling, choice of seismicity parameters, and the ground motion model. The sensitivity of the results was evaluated by calculating the seismic hazard using alternate source zone models, LLNL and Rizzo Associates seismicity parameters, and LLNL,Rizzo Associates and EPRI ground motion models.

Additionally, Rizzo Associates methodology and models were used to calculate the seismic hazard at the Dresden site to allow direct comparison to the results reported by EPRI for that site.

The Project Methodology results in site seisnuc hazard curves that fall between LLNL's and EPRI's.

Accordingly, the Methodology is considered to be appropriate for use in the IPEEE.

Respectfully submitted,

)J~~~cc s~

Nisluhmt R. Vaidya Principal - Structural Engineering Enri ue Bazan-Zuri Senior Staff Consultant

~~M WilliamJ.

ohnson Vice President-Technology NRV/EBZ/WJJ/smw

I

83WHUPtCKS Bernreuter, D.L., L.B. Savy, R.W. Mensing, and J.C. Chen, 1989, "Seismic Hazard Characterization of 69 Nuclear Plant Sites East of the Rocky Mountains," Lawrence Livermore National Laboratory, NU~REG/CR-5250.

EPRI, 1988, "Seismic Hazard Methodology for the Central and Easter United States,"

EPRI NP-4726, 12 Vol.

Rizzo Associates, April 1, 1991, Final Report Probabilistic Seismic Hazard Analysis, Donald C. Cook Nuclear Plant, Bridgman, Michigan."

U.S. Nuclear Regulatory Commission, 1991, "Procedural and Submittal Guidance for the Individual Plant Exanination ofExternal Events gPEEE) for Severe Accident Vulnerabilities,:

NUTMEG-1407.

U.S. Nuclear Regulatory Commission, 1988, "Safety Evaluation Review of the SOG/EPRI Topical Report, EPRI NP-4726."

01

I clean.doc

TABLEI SEISMIC HA2,ARDMmaooorOGV COMPARISON OF LLNL,EPRI ANDRIZZO ASSOCIATES EPRI MEZZO ASSOCIATES Seismic Source Zone Model Individual experts gave best estimate models+alternates.

Earth Science Teams'sed structured &

consensus

approach, gave best estimate models+ alternates.

EPRI's best estimate models supplemented by LLNL'sand Rizzo Associates best estimate models.

Seismicity Parameters Based on expert' data, indep. or correlated, continuous over source zone.

Std. catalog, includes incompleteness, discrete pairs, fully correlated.

Std. catalog includes incompleteness, independent and continuous over source zone.

Ground Motion Model 5 experts weighted 8 attention laws. Gave best estimate's and self weights.

EPRI (NP-6074) law with two alternates.

EPRI ground motion model + LLNL's SE2-a.

Site Effects Generic sites, correction used in simulation based on expert input.

Generic sites, evaluated separately for different values of PGA.

Uses EPRI's site amplification factors consistent with PGA's.

Procedure Simulation Logic Tree Simulation u454

C~ Comnzazo m Serac H~mn SENSHxvHY ANhLYSIS CASE 1

2 3

4 5

COOK COOK COOK COOK DRESDEN SEISMICITYMODEL PCRA LLNL PCRA PCRA PCRA GROUND MOTION MODEL PCRA PCRA EPRI LLNL

=PCRA

TABLE3 COMPARISON OF MANSEISMIC HAZARD FROMLLNLSEA BEsr EsromTE ANDAL1ZRNA'IESOURCE ZONE MODELS PGA (G's)

EXCEEDANCE PROBABILMESFOR SE-6 BEsr EsrIMATE SEA ALTERNATE 0.1 0.2 0.4 0.6 1.0 1.51E-03 3.12E-04 3.91E-05 8.66E-06 7.86E%7

2. 14E-03 8.57E-04 2.54E-04

1. 12E44 3.07E-05

COMPARISON OF MEANSEISMIC HAZARD FROMLLNLANDRIZZO SHSMICm'. PAKQdEIZRS SE-6 SOURCE ZONE MODEL PGA (G's)

SK-6 SErsMIcrrv Rxzzo SErsMIcrrv EXCEEDANCE PROBABKXHES FOR 0.1 0.2 0.4 0.6 1.0 1.51E-03 3.12E44 3.91E-05 8.66E-06 7.86E-07 1.97E44'.22E-05 9.53E-06 2.80E-06

2. 12E-07

TABLE5 COMPARISON OF MEANSEISMIC HAZAIU)

FROM LLNLANDRIZZO SEISMICHY PARAMEIXRS AGGazeAmn Ovza ALLLLNLSEIsMIC ExPERTS PGA (G's)

EXCEZDANCE PROBABILITIESFOR RXZZO SEISMICrrV 0.1 0.2 0,4 0.6 1.0 1.79E-04 3.56E-05 4.56E46 1.00E-06 8.06E48 6.77E-05 1.42E-05 1.84E46 4.24E-07 3.97E-08

TABLE6 EFFECX'F GROUND MOTIONMODEL ON SEISMIC HAZARD PGA (G's)

MEANEXCEEDANCE PROBABILITKSFOR RIZZO-GMM 0.1 0.2 0.4 0.6 1.0 5.36E45 1.05E-05 1.22E-06 2.62E-07 2.25E-08 2.45 E-04 3.52 E-05 2.87 E-06 5.10 E-07 3.76 E-08 6.77E-05 1.42E-05 1.84E-06 4.24E-07 3.97E-08 t64S4

I ~

I rl4aa.doc

~

~ I

~

~ >

I OO u w rD Cl E OZ b, Sope e) Geometry of homogenow source zo~ lseismicityjtectonus) mo (b) Megnitvde recurrence model (frequency ve size) cs tD cs O<o g Q

~ QC Ol us l

Vl ~~

I CO i

Z, tt cs CI 10 1

10 0.1 g 0.1 5 d) Seismic bezerd curve c, OSpeetiion s7 m

~ 5 mg ~5 log distence (c) Ground motion pred(et(on model (ettenwtion)

FIGURE 1

THE BASIC STEPS OF A PROBABILISTlC SEISMIC HAZARD ANALYSIS SE)SLIIC HA2ARO ANALYSIS OONAlO C. COOK NUClEAR PlANT PREPAREO FOR REFERENCE BERNREUTER ET A(

(1989)

WESTiNGHOUSE ELECTRIC CORP.

PITTSBURGH, PENNSYLVANIA Paul C Rheo Associates, Inc.

CONSULTANT

~ s ~'

~

~ ~ )

CV I

r laJm cr x oz r4 I

+

+

~

~

~

o

~'

~ J a

I, ~

~ ~

~ ~

J

~J

+

~

~

~ ~

~ ~

~

~

J

. ~

~ 4 p

~

~

~

4 ~'

e.".$j "r,.

,'S,a..

~

~

~

~

m

>. m

~ 0, o~

Ch I

g cv I

C7

~

hs

~

~

~

~

~

~

~

~

~

~,

~

~

~

+

~ 4..'a~ ~,'

~

~

~

~ ~

~

~

~

~

~

~ ~

+

, '4

~ +

~

JS

~I4

~

~

~

'le

~'.

0

~ ~ J ~

~

~ ~

,o

%J~

r

~

~,

~ ~

I

~rr m O

~

LESS THAN 3.3

~

3.3 TO 3,9 3.9 TO 4.5 4.5 TO 5.1 a

5.1 TO 5.7 5.7 TO 6.3 X

6.3 TO 6,9 6.9 TO 7.5

REFERENCE:

NAllONAL GEOOEllC DATA CENTER FIGURE 2 SEISMICITY OF THE EASTERN UNITED STATES SEISMIC HAZARD ANALYSIS DONALD C. COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTS8URGH, PFNNSYLVANIA Hul C Rzzo Associates. Inc.

COHSULTAHTS

IO

~

4l lc 4J CO QX 0 Z cn Y 33 1~CZ g) Gl O Cl us g OY Cc LJ f

( 31 22 rn I

cn ~~

I cD 27 17 10 15 25 12 200 SCALE 200 MILES

REFERENCE:

BERNREUTER ET AL. (1989)

FIGURE 3 BEST ESTIMATE SEISMIC SOURCE MODEL LLNL EXPERT 6

SEISMIC HAZARO ANALYSIS OONALD C. COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTSBURGH, PENNSYLVANIA Paul C. Rizzo Associates, Inc.

CONSULTANTS

10 10 a>

5R CI Cl aa oa oQ Ol I

CV I

CO 0-I-10 0

CL 4Jo 10

-z C5 OX 4J

-5 10 Dzz 10 10 0

200 400 600 I

800 1000 PGA (cm/sec

)

LEGEND:

LLNL SEISMIC EXPERT 6 ALTERNATE SOURCE ZONES LLNL SEISMIC EXPERT 6 BEST ESTIMATE SOURCE ZONES

--LLNI. SEISMIC EXPERT 6 B.E.

SOURCE ZONES WITH RIZZO ASSOCIATES SEISMICITY PARAMETERS FIGURE 4 EFFECTS OF ALTERNATE SOURCE ZONES ON MEAN HAZARD CURVE SENSIVI1Y STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTSBURGH, PENNSYLVANIA Foul C R!zzo Associates, Inc.

CONSULTANTS

~

> ~

SEISMICITY PARAMETERS LLNL SE-6, SOURCE ZONE 9 SEISMICITY PARAMETERS LLNL SE-6, SOURCE ZONE 17 1

G w OX C 2 oz V

~

I Q 10 LLI 10 O

02:

v ga aM 4J+ 10 Vl I

16 10 O

Qa 10 55 4

45 5

MAGNITUDE, m 55 6

10 3.5 4

4.5 5

MAGNITUDE, m 55 oY ()

~ Q.

o~C Cl rn I

V) ~

I.

lO

+ 10 16.

10 O

SEISMICITY PARAMETERS LLNL SE-6, SOURCE ZONE 22 lL

+ 10 16 10 O

02:

SEISMICITY PARAMETERS LLNL SE-6, COMPLEMENTARY ZONE a4 sty 10 5.5 4

M 5

55 6

MAGNITUDE, m 10 4

4.5 5

MAGNITUDE, m 5.5 6

LEGEND:

~

RECORDED DATA "RIZZO ASSOCIATES BEST ESTIMATE RECURRENCE

LLNL SEISMIC EXPERT 6 BEST ESTIMATE RECURRENCE FIGURE 5 SEISMICITY PARAMETERS FOR LLNL SE-6 DOMINANT SOURCE ZONES SENSIllVlTY STUDY OONALO C. COOK NUCLEAR PLANT PREPAREO FOR WESTINGHOUSE ELECTRIC CORP.

PITTS8URGH, PENNSYLVANIA Paul C. Rhzo Associates, Inc.

COHSUI TANTS

I C) 10 10 qP TE.

SEISMIC HAZARD Ci'RVES BASED OW ':.f'L SEISMiC SOURCE ZONE MODELS AhiD Rl ZZO ASSOCIAT"..S GROUND MOTION MODEL.

Ol

<n I

M CV I

l 4

gj 10 (0O lY Q.

LJ

-5 O

10 Z

O 4J

-d 10 Xz 10 L

LEGEND:

-d 10 0

200 400 600 PGA (cm/sec

)

800 1000 RIZZO ASSOCIATES SEISMICITY PARAME.ERS LLNL SEISMICITY PARAMETERS FIGURE 6

EFFECT OF SEISMICITY PARAMETERS ON 15th PERCENTILE HAZARD CURVE SENSITMTY STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITI'SBURGH, PENNSYLVANIA Hul C Rizzo Associates, Inc.

CONSULTANTS

l~

C'

lp I

LA I

La

~l I

I I

10 10 NOTE:

SEISMIC HAZARD CURVES BASED ON LLNL SEISMIC SOURCE ZONE MODELS AND RIZZO ASSOCIATES GROUND MOTlON MODEL CL a Ya

~Q

~ Q 0 Q 4)

I I

LO 10 QlQ n

Wo 10 Z

O LJJ 4JoX Ld e

10 D

Z l

~

L 10 LEGEND:

10 0

200 400 600 PGA (cm/sec

)

800 1000 RIZZO ASSOCIATES SEISMICITY PARAMETERS LLNL SEISMICITY PARAMETERS FIGURE 7 EFFECT OF SEISMICITY PARAMETERS ON MEDIAN HAZARD CURVE SENSILIIYY STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTSBURGH, PENNSYLVANIA RLL1 C Rtzxo Associates. Inc.

CONSULTANTS

<t 4i

03 I

LA C)

W W W ID 10 10 NO K:

SEiSMIC I".AZARD CURVES BASED ON LLNL Sc,ISMIC SOURCE ZONK MODELS AND RIZZO ASSOCIA.ES GROUND MOTION MODE'.

0-Cl Cl aa W W a 0 w o-

~ o o Q Ol I

Vl CV I

CO 0-4 10 QlO Q.

LJ 10 '

W LJJOX 4J d

10 ZZ l

L X

fm a

10

-d 10 0

200 400 600 800 1000 LEGEND:

PGA (cm/sec

)

RIZZO ASSOCIATES SE:SMICITY PARAMETERS LLNL SEISMICITY PARAME ERS FIGURE 8 EFFECT OF SEISMICITY PARAMETERS ON 85th PERCENTILE HAZARD CURVE SENSITIVITY STUOY OONALO C.

COOK NUCLEAR PLANT PREPAREO FOR V/ESTINGHOUSE ELECTRIC CORP.

PITl SBURGH, P ENNSYLVANIA Foul C Rizzo Associates, Inc.

c0 NSUt.Tpp4Ts

,(l

~

ig

~ g)

I IA I

Ol C3 I+

i 4J (D

dED O Z 10 10 NOTc:

SE;SMIC HAZARD CURVES BASED ON L'L SEISMIC SOURCE ZONE MODEi S AND RIZZO ASSOCIATES GROUND MOTION MODEL.

io c) d d o 0

~ 0 o o.

Oi rnill C4 I

iO 0-I 4

10 CQO 0'L 4Jo 10 z

I~

LJoX 4J

-s 10 zz ll L

mod-d 10

-7

-b

~ 0 200 400 600 800 1000 LEGEND:

PGA (cm/sec

)

RIZZO ASSOCIATES SEISMICITY PARAMEl'ERS LLNL SEISMICITY PARAME.ERS FIGURE 9 EFFECT OF SEISMICITY PARAMETERS ON MEAN HAZARD CURVE SENSITIVI7l'TUDY DONALD C. COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTS8URGH, PENNSYLVANIA Hul C Rizzo Associates.

Enc.

CONSULTANTS

I '~

~ lu c) oz

~

~

-2 0

-3 10

'VOTE::

Sc.ISMIC HAZ-'.<0 CURV.=S I3ASED ON RIZZO ASSOCIAT 3 Sc.ISMICITY PARAMETERS.

>. 0-to cl g 0

~ W g 0 uo.

V ~

0 Ql i.a Ul CV I

lO 4

10 O

CL Wo 10 Z

CL OX 4J

-6 10 z

zp-10

-a 10 0

200 400 600 800 1000 PGA ~cm/sec

)

LEGEND:

Rl

'0 ASSOCIATES GROUND MOTION MODEL LLNL GPOUND MOTION MODE'oRI GROUND. MOTION MODEL FIGURE 10 EFFECT OF

- GROUND MOTION MODELS ON 15th PERCENTILE HAZARD CURVE SEt'ISITIVITY STUDY DorIA:0 c. cooK NUGI.EAR PwrIT PREPARED FOR

'WESTINGHOUSE ELECTRIC CORP.

PITTS8URGH, PENNSYLVANIA Foul' Hizzo Associates. Ir.

coNSUI ANTS

I lA CO 4J ~

~E cR

-2 10

-3 10 NOTE:

SEISMIC HAZARD CURVES BASED ON RIZZO ASSOCIATES SEISMICITY PARAMETERS.

0-a<

~ 4J o<

w +

g o o o-Cl M

C4 I

I CO 4

m 10 O

O CL LJ

'Z

-s 10 Cl I~

QJ LaJ

-b 10 Z.

10

-e 10 0

200 400 600 PGA (cm/sec2) 800 1000 LEGEND:

RIZZO ASSOCIATES GROUND MOTION MODEL LLNL.GROUND MOTION MODEL li EPRI GROUND MOTION MODEL FIGURE 11 EFFECT OF GROUND MOTION MODELS ON MEDIAN HAZARD CURVE SKNSITMlY STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

P ITTSBURGH, P ENNSYLVANIA Foul C Hizzo Associates. Inc.

CON SUI.TAUNTS

s I

C)

II4 I

m V I

Ol I

Lil w cv I

CO

-2 10

-J 10 I

(g 10 COO CL 0

-s o

10 z

O 4JoX W

-d 10 zz' Y

NOTE:

SEISMIC HAZARD CURVES 8ASED ON RIZZO ASSOCIATES SEISMICITY P ARAMETERS.

10 10 0

200 400 600 PGA (cm/sec

)

800 1000 LEGEND:

RIZZO ASSOCIATES GROUND MOTlON MODEL LLNL GROUND MOTION MODEL EPRI GROUND MOTlON MODEL FIGURE 12 EFFECT OF GROUND MOTION MODELS ON 85th PERCENTILE HAZARD CURVE SENSITIVITY STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTS8URGH, PENNSYLVANIA Hul C Hizzo AssocIates, Inc.

CONSUL.TANTS

10

-s 10 NO.E:

SE!SMIC HAZARD CURVES BASED ON RIZZO ASSOCIATES SEISMICITY PARAMETERS.

4 g) 10 CQO Q.

4J

-5 o

10 z

CI 4JI~oK 4J d

10 O

z 10 10 0

200 400 600 800 1000 PGA (cm/sec

)

LEGEND'IZZO ASSOCIATES GROUND MOTION MODEL LLNL GROUND MOTION MOD-L EPRI GROUND MOTION MODEL FIGURE 13 EFFECT OF GROUND MOTION MODELS ON MEAN HAZARD CURVE SENSITMTY STUDY DONALD C.

COOK NUCLEAR PlANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITI SBURGH, PENNSYI VANIA Foul C Rtzzo Associates. Inc.

CONSULTANTS

i'E

<0 I

CO 10 10 a4/VV CJ t:- 10' QlOK CL 10 CL UJO 10 D

z 10 10 0

g 10 200 400 600 800 1000 PGA (cm/sec')

LEGEND:

RIZZO ASSOCIATES EPRI (1988)

BERNREUTER ET AI (1989)

FIGURE 14 15th PERCENTILE HAZARD CURVE DRESDEN SITE SENSITVITY STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTSBURGH, PENNSYLVANIA Foul C R'ufo Associates, Inc.

CONSULTANTS

10 hl ~

G w 02 V

I 48 II Z 10 10 QI Ql0 O

10 4lOZ O4l4lx 10 4l Kz 10 10 O

II 10 200 400 600 PGA (cm/soc2) 800 1000 LEGEND:

RIZZO ASSOCIATES EPRI (19SS)

BERNREUTER ET AL. (1989)

FlGURE 15 50 "

PERCENTILE HAZARD CURVE DRESDEN SITE SENSITMlY STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC COQP.

PITI SBURGH, PENNSYLVANIA Foul C HIxxo AaeocIates, Inc.

CONSULTANTS

I LA (0

10 10 10 (Q

Ql0 Q.

10 LJJoz C5 LLt o

10 10 10 10 200 400 600 I

800 1000 PGA (cm/sec~)

LEGEND:

RIZZO ASSOCIATES EP Rl (1 988)

BERNREUTER ET AL (1989)

FIGURE 16 85~h PERCENTILE HAZARD CURVE DRESDEN SITE SENSITIVITY STUDY DONALD C.

COOK NUCLEAR PLANT PREPARED FOR WESTINGHOUSE ELECTRIC CORP.

PITTS8URGH, PENNSYLVANIA PdQ1 C R2ZO ASSOCldt "S, L1C.

CONSULTANTS

K e

W 0