Upload
haliem
View
212
Download
0
Embed Size (px)
Citation preview
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.
IAEA/ICTP Workshop - February 2005
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.
IAEA/ICTP Workshop - February 2005
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
IAEA/ICTP Workshop - February 2005
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]
IAEA/ICTP Workshop - February 2005
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
IAEA/ICTP Workshop - February 2005
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
IAEA/ICTP Workshop - February 2005
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
IAEA/ICTP Workshop - February 2005
Geol. and Geotechn. Data Asessment ( cont.)
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
IAEA/ICTP Workshop - February 2005
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
IAEA/ICTP Workshop - February 2005 11
<3-
t
&
t
t-
5-
S
. • • • • • ' •
i /;: /V
LJL..1 - - — - —
;
PROSEiS,
•
• • ' , '
Switzerland
.<
-
1
' - - . ' - • "
" • < • • :
, , :
' " ' . '
• • ' ' -
r ^ *
1"" i
— - g;
Geol. and Geotechn. Data Asessment ( cont.)
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)
IAEA/ICTP Workshop - February 2005 12
Geol. and Geotechn. Data Asessment ( cont.)
• 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.
IAEA/ICTP Workshop - February 2005 13
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.
IAEA/ICTP Workshop - February 2005 14
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)
IAEA/ICTP Workshop - February 2005 15
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
Geolog. and Geotech. Data Assessment ( cont.)
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.
IAEA/ICTP Workshop - February 2005 16
Geolog. and Geotech. Data Assessment ( cont.)
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 '
IAEA/ICTP Workshop - February 2005 17
Geolog. and Geotech. Data Assessment ( cont.)
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!^
IAEA/ICTP Workshop - February 2005 18
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
IAEA/ICTP Workshop - February 2005 19
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)
IAEA/ICTP Workshop - February 2005 20
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
Supporting Computations (cont.)
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)
IAEA/ICTP Workshop - February 2005 21
Supporting Computations (cont.)
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
IAEA/ICTP Workshop - February 2005 22
Supporting Computions (cont.)
• 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
Supporting Computions (cont.)
Effect of Different Time Histories
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)
IAEA/ICTP Workshop - February 2005 23
Supporting Computions (cont.)
Effect of Different Time Histories
- 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)
Supporting Computions (cont.)
•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)
IAEA/ICTP Workshop - February 2005 24
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
IAEA/ICTP Workshop - February 2005 25
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
IAEA/ICTP Workshop - February 2005 26
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.
IAEA/ICTP Workshop - February 2005 27
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.
IAEA/ICTP Workshop - February 2005 28
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
IAEA/ICTP Workshop - February 2005 29
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
IAEA/ICTP Workshop - February 2005 30
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.
IAEA/ICTP Workshop - February 2005 31
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
IAEA/ICTP Workshop - February 2005 32
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
IAEA/ICTP Workshop - February 2005 33