2nd Workshop on Earthquake Engineering for Nuclear ...indico.ictp.it/event/a04181/session/27/contribution/16/material/0/... · IAEA/ICTP Workshop on Earthquake Engineering for Nuclear

Iliill

International Centre for Theoretical Physics

United NationsEducational, Scientificand Cultural Organization

International AtomicEnergy Agency

H4.SMR/1645-14

"2nd Workshop on Earthquake Engineering for NuclearFacilities: Uncertainties in Seismic Hazard"

14 - 25 February 2005

Site Specific Data - Treatment of Uncertaintiesaccording to SSHAC Level 4 Methodology

S. Tinic-Kizildogan

Civil Eng. SIAEarthquake Eng. & Structural Dynamics

Switzerland

Strada Costiera I I, 34014 Trieste, Italy - Tel. +39 040 2240 I I I; Fax +39 040 224 163 - [email protected], www.ictp.it

IAEA/ICTP Workshop onEarthquake Engineering for Nuclear Facilities - Uncertainties in Seismic Hazard

Assessment

"Site specific data- Treatment of uncertaintiesaccording to SSHAC Level 4 Methodology"

Trieste, Italy, 14-25 February 2005

Unit 15 - Sener Tinic-Kizildogan

Content of the Presentation

• Introduction• SSHAC Level 4 Methodology• PEGASOS-Project (2000 - 2004)

- Methodology- Site Response Characterisation Methodology- Site Response Characterisation Process

• Geological and Geotechnical Data Assessment• Supporting Computations• Expert Models• Lessons Learned and Conclusions• Summary of the Presentation• References and Glossary

IAEA/ICTP Workshop - February 2005

Introduction

First PSHA's for Swiss NPP's are performed between 1984-1996PSA's have shown that earthquakes are a significantcontributor to the estimated core damage frequencyHSK required a state-of-the-art hazard evaluation andissued methodological guidelines which closely resemblethe Study 4 Level methodology in NUREG/CR-6372(SSHAC, 1997)Swiss NPP's planned the PEGASOS- Project based on theSSHAC recommendations.The first phase of the PEGASOS is performed in 2000 -2004.The second phase will be completed by the end of 2006.

SSHAC Level 4 Methodology

The differences between the LLNL (1989) and EPRI (1989)hazard results led to establish the SSHAC to providemethodological guidelines on the PSHA.SSHAC concluded that the main reasons for thedifferences were procedural rather than technical.

SSHAC Report (NUREG/CR-6372) provides technicaladvice and procedural guidance at 4 different "levels ofcomplexity" of a PSHA- Study.


Table 3-1 Degrees of PSHA Issues and Levels of Study

ISSUE DEGREE

A

Non-controversial; and/orinsignificant to hazard

B

Significant uncertainty anddiversity; controversial; andcomplex

C

Highly contentious; significantto hazard; and highly complex

DECISION FACTORS

•Regulatory concern

•Resources available

•Public perception

STUDY LEVEL

1

TI evaluates/weights models based onliterature review and experience; estimatescommunity distribution

2

TI interacts with proponents & resourceexperts to identify issues and interpretations;estimates community distribution

3

TI brings together proponents & resourceexperts for debate and interaction; TI focusesdebate and evaluates alternative interpretations;estimates community distribution

4

TFI organizes panel of experts to interpret andevaluate; focuses discussions; avoidsinappropriate bshavici on part of evaiuaEors;draws picture of evaiualors' estimate of shecommu»ily's composise dissibutioD; hasuStisnaie responsibility for project

NUREG/CR-6372

SSHAC Level 4 Methodology

Important Characteristics of SSHAC Level 4 PSHA

• Sampling the state of knowledge of the informed scientific communityby expert elicitation and the using the aggregated views

• Distinguishing between

- Epistemic uncertainty - knowledge-based (e.g., lack-of-knowledge) orinformational uncertainties (e.g. associated with a limited data)Epistemic uncertainty results from imperfect knowledge aboutearthquakes and their effects and can be reduced with advances inknowledge and collection of additional data.

- Aleatory variability (Randomness)- probabilistic variability that resultsfrom natural physical process (the size, location and time of the nextearthquake and the resulting ground motion)Aleatory variability is irreducable without the inclusion of additionalpredictive parameters.


PEGASOS-Project (2000 - 2004)

PEGASOS is the Probabilistic Seismic Hazard Assessment for fourSwiss Nuclear Power Plant Sites according to SSHAC Level 4Methodology- Worldwide second project after Yucca Mountains- Worldwide first project for existing NPP's- Worldwide first project with site-specific soil response evaluation fully

integrated to the soil hazard evaluation- Added values to the "state of the art" of seismic hazard anayses regarding

to max. ground motions and some other aspectsDeliverables of the Project:- Seismic hazard curves at three depths for peak (100 Hz) and spectral

accelerations for probabilities of exceedance between 10~1 and 107 withthe epistemic uncertainties

- Uniform hazard specta for the 10"3 to 10"7 annual probabilities of occurancewith associated uncertainties

- Deaggregation of hazard results with respect to magnitude and distance

PEGASOS-Project Methodology

Rock Hazard Site Specific Soil Hazard

SP1: Source Characterisation SP2: Ground Motion Characterisation

4 Expert Groups per 3 Experts = 12 Experts

SP3: Site Response Charact.

5 Experts

Hazard CurvesThe epistemic uncertaintyabout various inputs isorganized and displayedby means of logic trees.

Dr. C. Sprecher, NAGRJA

4 Experts

Earthquakeand Sftuunie

vP" ^ ^;v• * %

- -^<iK,

£__

CattingSs5arC«S

f

PeeunencB Cohres \Mat* Mar^ i t»^£ ,

I3 5 1 % -

sisgnttud?- ^

/re-

§ :

I :

*;

\

i Mccel

\ ^

• \ 1

; • : • • • : •

llllli:!lllliiiyiiilSsplilBfiSi a

iiiiliililiJlli: iBsjSJaSgSSISSMlES

ifflHSSSHlS: M KM

Site specificHazard Spectra

e-g. P - « H /Yeisr


PEGASOS - Site Response Characterisation

• Objective: To quantify the surficialgeology effects on thecharacteristics of the earthquakeground motion

• Results: A set or several sets ofscale factors given as function offrequency at the surface,foundation level of the reactorbuilding and the safety-relevantbuildings

Site specificTransfer Function

•g"toa<to1s

Site specifichazard Spectra

ag. P -10- 4 /Yea-

. - • • -

'J; * - 50%

Frequency

Frequency

PEGASOS - Site Response CharacterisationMethodology

Computation of Amplification Factors

/ / . ' • / . • • • ; • ; . ' • • • , : / / .

Numerical Simulations

AccelerationResponse Spectra B

Frequency [f]

'/•;;:: Acceleration .-'.-.Response Spectra A

Damping; .

Frequency tq

Amplification Factor[Sets*]

Frequency [f]


PEGASOS - Site Response Evaluation Process

Generation of Geological and Geotechnical ModelsWorkshop 1: Key Issues and Data NeedsAdditional Site InvestigationsElicitation Meeting on Properties of Soil ProfilesSupporting ComputationsWorkshop 2: Evaluation of ModelsElicitation Interviews, Development of Initial Logic treesWorkshop 3: Initial Feedback on Experts' EstimatesModel RevisionsModel ImplementationWorkshop 4: Feedback on Expert's EstimatesExpert Model RevisionsExpert Meeting on Maximum Ground MotionDevelopment and Review of Final Expert Models

ExpertsProject Management, TFI

Uncertainties in Site Response Characterisation

Epistemic uncertainty is due to- Imperfect knowledge of material properties

(uncertainties in the shear wave velocities, G-modulusand damping curves, unknown variation of the materialproperties over the site)

- Inappropiate or too simple modeling (1-D or 2-D insteadof 3-D modelling, simplifed material modelling,equivalent linear instead of true non-linear)

Aleatory variability is due to- S-PV Waves- 2D-effects


Geological and Geotechnical Data Assessment

• Collecting all available and relevant geological and geotechnical data• Digitalizing all available data, generating a document database• Evaluation of the quality of all basis geotechnical data• Compilation of all available geological and geotechnical data

- Integration of data in the GIS- Generate geological models- Store the data in the Project Database

• Submit data information to experts prior to Workshop 1• Discussion at the Workshop

- Is the data quality acceptable for site-specific evaluations ?- Is additional data needed ?- Can uncertainty be reduced through additional investigations?- Are resources available in the framework of project deadlines and budget

constraints?

Geol. and Geotechn. Data Asessment ( cont.)

Evaluation of the Quality of Former GeotechnicalInvestigations at the NPP-Sites

• Evaluation criteria- Accepted standard procedure at the time of data

gathering

- Acceptance of these procedures today- Experience of the organization performing the

investigations and interpretation results- Properties in expected, reasonable range for these

geological formations


Geol. and Geotechn. Data Asessment ( cont.)Evaluation of the Quality of Former Geotechnical Investigations atthe NPP-Sites (cont.)

• Field tests- Geophysical borehole tests- Hydrostatic test- Plate test- Standard penetration test- Load-settlement tests- Surface refraction tests- Up-hole tests- Cross-hole tests- Down-hole tests

Evaluation of the Quality of Former GeotechnicalInvestigations at the NPP-Sites (cont.)

• Laboratory tests- Classification- Specific weight- Grain size distribution- Direct shear test- Compression test- Triaxial test- Resonant columns tests- "Stress controlled dynamic tests"- "Strain controlled dynamic tests"- Cylic triaxial tests



Compilation of all available Site Geology Data

•Geological information from all the boreholesdrilled on the four NPP sites

•All available well data in a radius of some twokilometers around each site.

• Additional regional geological data (maps, deepboreholes, seismic reflection data, digitalelevation models, etc.)

Location map


J

Sunmsiy Site 1 Vs Profile

Formation

UnconsolidatedRocks

WeatheredOpalinuston

Opalinuston

Lias & Keuper

Muschelkalk

Depth BelowSurface (m)

From

0

9

-11

74

224

To

9

-11

74

224

End ofModel

Vs Range

Min

150

850

1300

1750

Average

380

400?

1000

1S20

2200

Max

500

1200

2700

3100

PROSEIS, Switzerland

Jr ' S *"

r i

i:J§L

530

1 B30

1| S > 990

[ 1440

""'*"} *•^^Ks ^ - —

Site 4 VsProfile

Formation

UnconsolidatedRocks

Aquitanian

Chattien

Depth BelowSurface (m)

From

0

9

580

To

9

580

End ofModel

Vs Range

Min

240

800

Average

400

1500

Max

470

2100

PROSEIS, Switzerland

IAEA/ICTP Workshop - February 2005 10

Geol. and Geotechn. Data Asessment ( cont)Geological and Geotechnical Database

PROJECTDATA DB & DATAFLOW

SPl Source Characteriasatton

EarthquakeCatalogue DB

CAT *

CnmpumfonsDSUP #•

# • QC Certificate

SP2 Ground Motion Characterisation SP3 Srte Response Characienasatson

Wave Form DSWAF •

Fxp«rtsnientsEXA

ExpertModels OB

EXM

M input1.tKiO; D8

* HiO •

Geol. and Geotechn. Data Assessment ( cont.)

Site Geology Database

• Maps of the bedrock geometry using a GISsoftware package.

• Contour maps of depth to additional deepergeological units

• Selected vertical geological profiles• Additional vertical profiles computed along any

expert-defined cross section.Bedrock Map: NPP 2 x 2 km 1:5,000Subsurface Map: Top Opalinuston 1:5,000Subsurface Map: Top Lias 1:5,000Subsurface Map: Top Muschelkalk 1:5,000Geological Section: NPP_Centre 1:5,000Geological Section: NPP-C 1:5,000


<3-

t

&

t

t-

5-

S

. • • • • • ' •

i /;: /V

LJL..1 - - — - —

;

PROSEiS,

•

• • ' , '

Switzerland

.<

-

1

' - - . ' - • "

" • < • • :

, , :

' " ' . '

• • ' ' -

r ^ *

1"" i

— - g;


Site Geotechnical Database

Using the same boreholes as used for the collection of the geologicaldata, several categories of SP3 relevant rock physical properties havebeen gathered and integrated in the Geotechnical Database. Thephysical properties collected were:

•P wave velocity•S wave velocity•Grain size distribution (grain size and sediment type)•Atterberg consistency limits•Density•Water content•Porosity•Elastic properties•Non linear material properties (shear modulus and material damping asa function of shear strain)



• Geological, geotechnical data together with related qualityassessments were presented at WS-1 to the experts

• Experts required additional investigations:- Ambient vibration measurements to elaborate possible

Vs Profiles based on Dr. Donat Fah's inversiontechnique

- Additional ambient vibration measurements (HA/ dataprocessing to determine the fundamental frequenciesaccording to Nakamura-Method)

- Spectral Analysis of Surface Wave (SASW) to check theolder crosshole data at one site

Geological and Geotech. Data Asessment ( cont.)

Measurement of S-Wave Velocity Profiles, Donat Fan, Johann Wossner, SED

Baznau: KKBIrb; KKB2:g: KKB3:r, KKB4:m

J.--,

VSJ

• • • &

.'„'.*.^

Figure 1. Sites of the ambient vibration measurements at NPPsite Beznau, and derived H/V ratios.


Geological and Geotech. Data Asessment ( cont.)

Conclusions:The structure can not beapproximated by a one soil-layer - over-bedrock structure.Since other velocitymeasurements are available atthe site the combined VSprofile (in cyan) is proposed.

2.5

„ 2

11.5>

1

0.5

Beznau 1:b;2:g; r:5;8:m; bl

-

ack: BeznauJap2; cyan: Combined

• ; . • • . •

I 1

I

0.1 0.2 0.3 0.4 0.5 0:6 0.7Depth |km]

Geological and Geotech. Data Asessment ( cont.)Measurement of S-Wave Velocity Profiles, Donat Fan, Johann Wossner, SED

2.5

2

0.5 -

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Depth |km]

Structural models obtained from the different inversions.

Conclusions:The structure can be approximated by a one soil-layer — over-bedrock structure.Inversions from HA/ ratios in such case are considered to be reliable.


Geological and Geotech. Data Asessment { cont)

. .:J_Jii£:__

Geological and Geotech. Data Asessment { cont)Gosgen 2: SASW inwersion with 3 to S Layers

I

3-Layer Model (RMS = 35.2)

5-tayer Model (RMS - 4.0)

4-Layer Mode! (RMS - 3.2)

S-Layer Model (R«S » 3.7)


Geolog. and Geotech. Data Assessment ( cont.)

Results of additional geological Investigations were presented at theExpert Meeting on "Validation of Soil Profiles"

NPP Site 1

Cross-hole, downhole, 1980

Microtremor. Oyo-Corp. 2000

Microtremor. Fan2001

NPP Site 2

Refraction Seismic,1967

Cross-hole anddown-hole, 1972

Microtremor Fah,?r.31

SASWRi.OT-<incG'C'D'..1':. >CC

NPP Site 3

Cross-hole, 1975

Micotremor Fah,2001

Nakamura,^i-t.onance.2001

NPP Site 4

Cross-hole and up-hole1980,1981

Nakamura,Resonance,2001

In the Framework of PEGASOS-Project


Experts discussed potential value of additional S-Wave reflectionmeasurements. According to different contacts it was not possible to executesuch measurements in the given time and budget frame of the PEGASOS-Project. It was also questionable if these measurements can give a clearstatement for the typical nuclear plant environment because of the typicalgeological conditions and the availability of the adequate shear wave source onthe market at that time.

Based on all compiled information experts decided on the control values of theshear wave velocities for each site and requested to introduce az1/4dependencywithin the uncohesive loose soil and to check the transfer functions for weakmotion and adjust the velocities so that the fundamental frequencies obtainedfrom HA/ ambient vibration measurements are matched.

Further experts specified soil profile randomization constraints.



Soil Profile Randomization Constraints- A bound on the uncertainty of 30% of the median velocity is imposed. This represents ± 2 a.- G/Gmax (y) and D(y) are randomized about the base case values. A log normal distribution is

assumed with aln of 0.35 at a cylic shear strain of 3x10-2. An upper and lower bound truncation of2 a is used

- An inverse correlation between G/Gmax and D(y) curves is assumed. In developing therandomized equivalent linear properties, full correlation is assumed between the G/Gmax (y) andD(y) curves

- The fundamental natural frequencies must be within the site specific frequencies (e.g. for Beznau1.8 Hz and 3.0 Hz)No randomization will be undertaken for the bedrock

- Site specific constraints about layer thicknesses

Geol. and Geotechn. Data Assessment (cont.)

Shear Wave Velocity Profiles at Site 1

L ;

fo = 2.5Hz '



Dynamic Material Properties for 1-D Site SimulationsG/Gmax(y)

1993. 75 KN/m2

Geolog. andDynamic

: 0,35-

0,3-

'' 8 0,25*i e

i | W-: &: 0.15-

| 0,1*

: 0,05- •

Geotech. DataMaterial Properties

D(y)

il" "CTC ~* Oi "~Z

. _ .

- — -

Assessment (cont.)for 1-D Site Simulations

~/jy

t /

^ . : —

0 - " " • ! • • :

0.0001 0,00t 0,01 0,1 1 10 !: Sicar Strain!^


Vs profile for Site 2, elicitated on the Meeting, February 4, 2002

• 1 * \ ; :

30 .

50

100 *—

*A&-~ --*—•>

• : . :

•

i : fo

' ! i

= 4.7 Hz ' ' 1

• • t •

j : j ! •

4. Geological and Geotechical Data (cont)Vs-Profiles for Site 3, elicitated on the Meeting, February 4, 2002


4. Geological and Geotechical Data (cont)VS-Profile for Site 4, elicitated on the Meeting, February 4, 2002

500 1000

-^-LL

. . .

'fo

\

1500

= '.2

1,1

Hz

2000

• - "

Supporting Computations

1-D Equivalent-Linear Site Response Computations- Time History Methode (SHAKE) each case with 15 TH's- Random Vibration Theory Method (RVT) with and without soil

profile randomization (Silva, 1972)

1-D True Non-Linear Site Effect Computations (Modaressi.GDS,Geodeco using SUMDES and DYNAFLOW)

1-D Site Effect Computations for Oblique Wave Incidence (Bard, Fah)

2-D Site Effect Computations- Aki-Larner Technique (Bard)- 2D SH computations (Fah)


Supporting Computations (cont.)

Overview of the supporting computations for thehorizontal component of ground motion

Site

NPP Site 1

NPP Site 2

NPP Site 3

NPP Site 4

TOTAL

RVT with soilrandomization

numberof cases

90

60

120

60

330

numberofruns*

4'590

3'060

6'120

3'060

16'830

Shake

numberof runs

225

225

270

180

900

1-D true non-linear

number ofruns

30

35

30

0

95

2-D

2-methods

SH-PSV

yes

yes

yes

* 51 runs per case


1- D Site Response Simulations according to the RVT-method (Silva)

NPP

Site 1

Site 2

Site 3

Site 4

Number ofVelocity SoilProfiles

B1B2B3

G1

L1L2

M1

Number ofDynamicMaterialProperties

M1

M1M2

M1M2

M1M2

Number ofMagnitudes(Mw)

3 (5, 6, 7)

3 (5,6,7

3 (5,6,7)

3(5,6,7)

Number of groundmotion levels

10 (0.05 to 1.5 g)

10 (0.05 to 1.5 g)

10 (0.05 to 1.5 g)

10 (0.05 to 1.5 g)



Visualisation and Comparison of RVT and SHAKEResults- Effect of different Time Histories (SHAKE runs)

- Effect of Ground Motion Level (RVT runs)- Comparison of RVT and SHAKE runs- Effect of Input Level (RVT runs)- Effect of Soil Randomization- Effect of Magnitude (RVT runs)- Maximum Strains

Supporting Computions (cont.)

Effect of Different Time Histories

Figure 21. Amplification Spectra I SIM, Profile 1, Material 1- Surfaces Layer (PGA = 0.4 g)

I 2-1

Motbn 1

Motion 9

Motion 2

Motion 10

Motion 3

Motion 11

Frcqu

Motion 4

Motion 12

• ney(Hz)

Motions

Motion 13

Motion 6

Motion 14

Motbn 7

Motion 15

Motbn 8

«•«••••• Geometric Mean



• Effect of Different Time Histories

Figure 10. Acceleration Response Spectra Site, Profile 1, Material 1 -Surface Layer (Outcropping Motion PGA = 0.4

S)

- M O T

- M O T

MOT

•MOT

- M O T

-MOT.

- M O T

- M O T

- M O T

MOT

MOT.

MOT

•MOT

-MOT

-MOT

- M e d s



site - Profile 1, Material Set 1, PGA = 0.1 g

Surface

10"

Frequency (Hz)

lo2

time hrsL 1- lime hist 2- lime hraL 3- lime hist. 4- time hist 5- lime hist. 6

_ tima hist. 8- time hist. 9 j.. trma hist. 10 "... time hist. 11,. time hist. 12.. time hist. 13... time hist. 14_ time hist 1E

median - o ^ -

10" 10'

Frequency (Hz)




- Profile 1, Material Set 1, PGA= 0.4g

Surface

10° 101

Frequency (Hz)

time histtime hist.time hist

— tim« hlattime histtime hist.time hist.

— time histtime hist

....... time hist

"___ time hist„ time hist

time hist IS

10" 10"Frequency (Hz)


•Effect of Ground Motion Levels

site - Profile 1, Material Set 1, M =6, SurfaceW

Soil Randomization

Am

plifi

catio

n

01C

Af

/

I t:::/-,• -f\ •'"'"

—^,—

*~"—<~.-~.•*-—^_-

-~ -L ^ r

• PGA = 0.05 g |: : — PGA = 0.10 g |

—. PGA = 0.20 g |i PGA = 0.30 g |

PGA = 0 . 4 0 g |:: •—- PGA = 0.50 g j: : PGA - 0.75 g |:: PGA = 1.00 g |

PGA = 1.25 g |:: PGA = 1.50 g |

*-r

10° 101 102

Frequency (Hz)


Effect of Input Level (RVT runs)

"reo

q=

010"

site, Mw= 7, Surface, f= 3.0 Hz

— GP1M1GP1M2

.\..

10u 10'PGA(g)

GP1M1: Median Amplification Factors Soil Profile 1, Material 1

GP1M2: Median Amplification Factors Soil Profile 1, Material 2

Supporting Computions (cont.)•Effect of Input Level (RVT runs)

c4o

•5.

<2

—LP1M2 :

LP2M1 :LP2 \A2 :

site, MN -1

>

D m), f=4.C)Hz

10-1PGA(g)

io° 10'

LP1M1: Median Amplification Factors Soil Profile 1, Material 1LP2M1: Median Amplification Factors Soil Profile 2, Material 1LP1M2: Median Amplification Factors Soil Profile 1, Material 2


Expert Models

• Logic Trees for- Median Amplification of Horizontal Ground Motion- Median Amplification of Vertical Ground Motion- Aleatory Variability of Horizontal Ground Motion- Aleatory Variability of Vertical Ground Motion

• Estimation of Maximum Ground Motions

Expert Models (cont.)* Example for Median Amplification of Horizontal

Ground Motion

SP3. Median Site Amplification for Horizontal Ground Motion

Vel_Profiles Nonlin_Prop Approach Evt_Subsets Mag_Depen DepenJType

Rvt

Profile./ 0.33

/ Profile,A. 0.34

\ Profile.

1

3

05Sm_&_Dr_l

03

05

No

Yes

1.0Dependencel

63Dependence2

05

No

S "\ Yes

05

* 1.0Dependencel

/ 05^ " \ Dependence2

05


Expert Models (cont.)

Weights for Alternate Velocity Profiles

Profile

Site 1, Profile 1

Site 1, Profile 2

Site 1 Profile 3

Site 3, Profilei

Site 3, Profile 2

Expert 1

0.5

0.2

0.3

0.8

0.2

Expert 2

0.4

0.2

0.4

0.7

0.3

Expert 3

0.45

0.35

0.20

0.7

0.3

Expert 4

0.5

0.3

0.2

0.7

0.3

Weights for non-linear material properties

Model

Ishibashi&Zhang

Hardin&Drenevich

Expert 1

0.65

0.35

Expert 2

0.5

0.5

Expert 3

0.30

0.70

Expert 4

0.3

0.70

Experts' Models (cont.)• Epistemic uncertainties of site amplification function

- Surface- Magnitude 6- PGA 0.10gon rock

Sin ampificalior (fractllas): , Gurfscs, ME, PGAO.


Experts1 Models (cont.)Contribution of the SP3 logic tree branches to the median amplificationat 5 Hz for Magnitude 6 und PGA rock = 0.23g at one site

Experts j

Shear-wave velocity profile :Nonjnear im^|a|prope ii1ies.Site..effects simulation method |

P-SV effects !

J = M : \ i

•: • ! [ : ' i

2D affects ! T i ' • \ • \ .

Site effects simulation method IP-SV effects }

2D effects

Shear-wave velocity profileN.on^nea.r.rr^erjalprppe.rties

_..__Ste^^s.simu,iafion..rneth.od__.

2D effects

Shear-wave velocity profileNon-linear material propertiesSite effects simulation method

2D effects

] • n i '

1 ;.:j §t± I ; •..; :....

1 j • | ; :

": i ": I ': rirrr•• j 'i ; I

* i "'~r' ' ! 1 :at • ; i :•% : ! ! i 1

5 HzMagnitude 6PGA.oa = 0.23g

Legend

Weights• »0.0-0.1. 0.1-0.2• 0.2-0.3O 0.3 - 0.4O 0.4-0.5• 0.5-1.0

1.0 1.5 2.0

Spectral acceleration on soil Ig]

Lessons Learned and Conclusions

The uncertainties on the soil shear wave velocity profiles and dynamic soilproperties are major contributors to the uncertainty in the soil hazard.These uncertainties can be generally reduced by additional geological andgeotechnical site investigations. The cost/benefit of such investigations shouldbe examined carefully for existing plant sites.The uncertainty caused by non-linear material properties increases for highground motion levels.This uncertainty in the non-linear material properties can be reduced byadditional geotechnical investigations to determine non-linear soil properties.

To analyse the site specific seismological data (real earthquake records) may bea simple and an effective approach to specify more reliable shear wave velocityprofiles and to reduce the uncertainties.To increase the reliability of the site specific data will lead to estimate themaximum ground motion precisely.The aleatory variability is considered in the expert models of mean groundmotion as separate branches for P-SV and 2-D effects, these effects areincluded also in the logic trees for the aleatory variability. The attenuationmodels for rock motion contains these effects too. This means aleatory effectsare included three times in the entire process of soil hazard evaluation. In theend this leads to unrealistic high soil hazard results.


Comparison of Site Area Investigations

IAEA Safety Standard Series No. NS-G-3.3Evaluation of Seismic Hazards for NPP's3.17 Site area studies should include the entire area coveredby the plant, which is typically one square kilometer

3.18 The database should be developed from detailedgeological, geophysical and geotechnical studiescomplemented by in situ and laboratory testing

3.19 The following investigations of the site area performed:Geological and geotechnicalHydrogeological investigationsInvestigations of site effects

3.20 All the data required for assessing the dynamic soil-structure interaction should be acquired in the course of theseinvestigations

3.21 The data are typically presented on maps at a scale of1:500 and with appropriate cross-sections:

PEGASOS

Fulfilled, geological and geotech. Datais compiled for an area of 4. km2

fulfilled

The existing data fulfill therequirements.Hydrogeological data existsAmbient vibration measurementsExisting real earthquake records atsome sites were presented at WS1.However the experts did not considerthese data to validate Vs profiles

fulfilled

PEAGASOS Data are on the GIS,might be presented on maps at therequired scale of 1:500

Site-Specific Seismological Data

Comparison of the acceleration response spectra on the rock (weathered)

Y-Direction (North)

SierentzN, ML=4.4, h=15 km,D=54km Wuto. Y, ML=3.9,h=25 km,D=21 km Saint-Dife, M=5.5, D=159 km Frick, H=4.1. H=21 km, D= 6 km" I

70

*" 60

(0

E 50oc

— 40co

• ^ 30

icce

0 1d z

\*>

i

1

Frequenzen in Hz5% Damping

\

\

V

oA

10

-JI

H100


Site Specific Seismological Data

NPP Beznau, Comparison of Acceleration Response Spectra ofEarhquake on the 05.12.2004 and the 28.06.2004 Sensor F1 (Surface) in X-

Direction (Ost-17°)

0.09 -

0.08 -

O) 0 0 7 "

~ 0.06 -

| 0.05a« 0.04 -u

< 0.03 -

0.02 -

0.01 -

0.1

/

I V

r

p1\

28.06.2004, MLM.O, D=6 km05.12.2004, ML-5.1, D>60 km

\'V\/'A

\

nSJ\ V

1 10Frequency

100

Site Specific Seismological Data

NPP Beznau, Comparison of Acceleration Response Spectra ofEarhquake on the 05.12.2004 and the 28.06.2004 Sensor F1 (Surface) in Y-

Direction (Nord -17°)

0.09 -

0.08

0.07 -Dl

- 0.06 -c

1 ° 0 5 '• | 0.04u< 0.03 -

0.02 -

0.01

0 1

/

/

j

/

\

IJ

Jk

J

1Frequency

\28.06.2004, ML=4.0, D=6 km05.12.2004, ML=5.1, D=60 km

\\

\

V

s*

\V.

10 100


Site-Specific Seismologicai Data

NPP Beznau, Comparison of Acceleration Response Spectra ofEarhquake on the 05.12.2004 and the 28.06.2004 Sensor F1 (Surface) in

Vertical Direction

0.09 -

0.08 -

0.07 -O)

- 0.06 -c• | 0.05 -

» 0.04 -oo

< 0.03 -

0.02 -

0.01 -

0 1

/

J

11f

J

\

Frequency

28.06.2004, ML=4.0, D=6 km "05.12.2004, ML=5.1, D=60 km

>

10 00

Summary

SSHAC-Level 4 Methodology is fully implemented in Site-Response Characterisation.PEGASOS geological and geotechnical site investigations fulfill the related IAEA Safety Standard NoNS-G-3.3 with except for site specific seismologicai data.Geological and geotechnical database contain all available technical information and serve a basis forfurther investigations.The results of some additional site investigations are ambiguous. They do not help to reduce theuncertainties, on the contrary they increase uncertainties.The experts considered the uncertainty on the geological and geotechnical data using different shear-wave velocity profiles and multiple set of non-linear material properties.Each expert used different weights for each Vs-profile and non-linear material property set in his logictree model.The RVT-Method with Soil Profile Randomization is an effective method for large amount of siteresponse simulations. The chosen constraints for Soil Profile Randomization avoided physicallyunrealistic soil profiles.The large amount of epistemic uncertainty in soil hazard is caused by the uncertainty in the sitespecific geological and geotechnical data. This uncertainty can be reduced by appropriate site specificinvestigations.The treatment of aleatory variability is an unresolved issue in the probabilistic soil hazard evaluation.The solution fora realistic and physically more reliable site specific soil hazard evaluation will be achallenge for the involved technical community in the near future. The interaction betweengeoscientists, seismologists and engineers will lead a big contribution to solve this complex issue.


Join to the

A joint CVCT. trf ;r;c

* European $e$£moio*;ic&! Commission

c

References and Glossary

NUREG/CR-6372, Recommendations for Probabilistic Seismic Hazard Analysis: Guidance onUncertainty and Use of Experts, Senior Seismic Hazard Committee, Main Report, April 1997

PEGASOS- A Comprehensive Probabilistic Seismic Hazard Asessment for Nuclear Power Plants inSwitzerland, Abrahamson N. et al, 12th European Conference on Earthquake Engineering

IAEA, Safety Standard Series No. NS-G-3.3, 2002, Evaluation of Seismic Hazard for Nuclear PowerPlants

Fah, D., Kind F., and D. Giardini, 2001, A theoretical investigation of average HA/ ratios, GeophysicalJournal Int., 145 535-549

Silva, 1976, W.J. Body Waves in a Layered Anelastic Solid. Bull. Seismic Ass. Am. 66 (5) 1539 -1554

Jongmans D., Demanet D.,1993, The importance of surface waves in vibration study and the use ofRayTeigh waves for estimating the dynamic charesteristics of soils, Engineering Geology 105 -113Pecker Alain, 2004, An estimate of maximum ground surface motion for non.zero velocity, C.R:Mecanique 332 725- 730, www.sciendirect.com

Pecker Alain, accept, for publ. 2004, Maximum Ground Surface Motion in Probablistic Seismic HazardAnalyses, Journal of Earthquake Engineering


References and Glossary (cont.)

PEGASOS:_Probabilistische_Erdbeben-Gefahrdungs-Analyse fur KKW-StandOrte in derSchweiz

SSHAC: Senior Seismic Hazard Committee

SED: Schweizerischer Erdbeben-Dienst (Swiss Seismological Service)

SASW: Seismic Analysis fo Surface Waves

RVT: Random Vibration Theory (Silva, 1972)

SHAKE: A Computer Program for Conducting Equivalent Linear SeismicResponse Analyses of Horizontally Layered Soil Deposits , Schnabel,Lysmer&Seed, 1972


Documents

2nd Workshop on Earthquake Engineering for Nuclear ...indico.ictp.it/event/a04181/session/27/contribution/16/material/0/... · IAEA/ICTP Workshop on Earthquake Engineering for Nuclear