ML19323C294
| ML19323C294 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 04/18/1980 |
| From: | WOODWARD-CLYDE CONSULTANTS, INC. |
| To: | |
| Shared Package | |
| ML13329A099 | List: |
| References | |
| TASK-03-06, TASK-3-6, TASK-RR NUDOCS 8005150294 | |
| Download: ML19323C294 (30) | |
Text
.
O s oos i s o m f
Development of Instrumental Response Spectra with Equal Probability of Exceedance for SONGS Unit 1 PREPARED FOR SOUTHERN CALIFORNIA EDISON P.O. BOX 800 ROSEMEAD, CALIFORNIA 91770 April 18,1980 KA Woodward Clyde Consultants w I
---,-,,_.-.,.,a-,
-.n
- - - -,,, ~. - -. -
o TABLE OF CONTENTS Page 1.0
SUMMARY
AND FINDINGS 1-1 1.1 Summary 1-1
1.2 Findings
1-1 2.O METHODOLOGY 2-1
~
2.1 Identification of Seismicity Sources 2-1 2.2 Characterization of Activity of 2-1 Seismicity Sources 2.3 Characterization of Attenuation of 2-2 Ground Motion 2.4 Specification of Earthquake Magnitude 2-2 Versus Rupture Length Relationship 2.5 Calculation of Probabilities of 2-3 Exceedance
3.0 ASSESSMENT
OF INPUT DATA 3-1 3.1 Geometry of Sources 3-1 3.2 Activity of Sources 3-1 3.3 Attenuation Relationships 3-2 3.4 Magnitude-Rupture Length 3-3 Relationship 4.0 RESULTS 4-1 4.1 Probability of Exceedance 4-1 Relationships for Ground Motion Parameters 4.2 Development of Equal Probability 4-1 Instrumental Response Spectra 4.3 Contributio'ns of Source, Magnitude and 4-2 Distance to Probability of Exceedance 9
^
o
'o t
5.0 SENSITIVITY ' OF THE PROBABILITY OF 5-1 4
l EXCEEDANCE TO INPUT DATA VARIATIONS
]
5.1 Introduction 5-1 5.2 Results of Sensitivity Studies 5-1 l
6.0 REFERENCES
6-1 I
i i
3 i
1 t
F 1
i I
4 5
l i
I
^
l i
i i
l
1.0
SUMMARY
AND FINDINGS 1.1 Summary The objective of this study was to provide basic input data to be used in evaluating the probabilities of exceeding the de-sign spectrum for San Onofre Nuclear Generating Station (SONGS) Unit 1.
Specifically, this study provides instrumen-tal response spectra with equal probabilities of exceedance for use in judging the probabilities of exceedance of the SONGS Unit 1 design spectrum.
The approach to evaluating probabilities of exceeding various levels of a ground motion parameter is shown schematically in Figure 1-1 and described in Section 2.
The basic input requires the following information:
e Identification of seismicity sources and characteriza-tion of activity of each source; e Specification of magnitude-rupture length relation-ships; e Selection of attenuation relationships for peak ground acceleration and spectral ordinates at selected structural periods.
The assessment of these input items is discussed in Section 3.
The results of the probability of exceedance calculations are described in Section 4.
The analyses results presented herein are a product of a care-fully carried out evaluation of existing data.
Best estimate and in some cases conservative input parameters with appro-1-1
proriate probability distributions were considered in the initial analysis as described to the NRC staff in Washington on 20 March 1980 (results shown in Sections 2 through 4).
Subsequent to that meeting and in response to NRC concerns expressed during the meeting, sensitivity analyses were carried out on various input parameters.
The results of these studies are discussed in Section 5.
All cited references are listed in Section 6.
1.2 Findings
The results of the probability analyses as described in Sec-tions 2 through 4 are summarized in Figure 1-2 in the form of equal probability spectra.
Specifically, instrumental response spectra with equal probabilities of exceedance corresponding to peak accelerations of 4/10, 1/2-and 2/3-g are presented. The corresponding probabilities of exceedance 10-3, 6 x 10-4 and 1 x 10-4 These results are are 2 x consistent with those presented in the 20 March 1980 NRC meeting.
After concideration of the results of sensitivity studies on various input parameters (Section 5),
it is concluded that the results presented' in Figure 1-2 reflect a reasonable but conservative estimate of equal probability of exceedance spectra for the SONGS site consistent with the regional historic seismicity data as well as the geologic constraints on prehistoric seismicity.
1-2
T t
INPUTS _
ANALYSIS RESULTS Source Seismicity Model Location Recurrence Magnitude Range Obtain probability distribution of spectral ordinates by combining over probabilities Calculate probabilities of exceeding specified Attenuation Model of occurrence of different magnitude earthquakes, distance response spectral ordinates to fault, and spectral acceleration Site Conditions for a given distance Transmission Path Conditions Magnitude and Distance Exposure Evaluation Criteria Period of interest Probability of Exceedexe Figure 1-1 -Schematic Representation of Approach 4
e G
FREQU ENCY (c os) soo so no to is e
7 s
a e
i.s a
or o.s o.s o.a o.is o.e 30 0 d.
A- [-. ;
[.vl-r-+--+ 7.
j (v ya, e.*..*.37--+-
co
.---M /v.h j a l '7
--.Q.. v-e,,A;a/ A- ;.'
,.j
.a., -q so
- - -. -. eo n 1 o,,. - -
..u.,x
.-:,5 4L-r.
x 5
- < \\p m q r g _;.. gu, s ',/.h. 1. / ' / ' N \\{ g \\ y,ff g /
r
. + }./ f / V A / N < s.y'
--f V-
/ 5 vs1-y ys s
s so f.. -
w-
-4,#-
. - < : N v /!-
W /.,4 n,.,p,on.oii,,, v.,',r,' / u4 i ', '.V/=, / /. N y -.tY
,. so
- w..
u-.~X. l:'. --.1.+V.,4.Y/.X /mm'x v l 1.,1/2
/
,.b/>f V'V s1 'V/W*,",0_"4
?/b.9,W/>{ 'n ~)WX ANf%'/ A ' ann /YI'/// /\\W.+ +N A
X 'd
.o r
.o N> -, A&--). y\\ fX A.-
/
- 2.,n a so
~&y v
--,- +
-,,/'*,,P}[{'l)['O&'~&!/Nl7 e
so
- _ h'/'/.J9d. - -
e^)/t A / '/ / Y V ' ' '
~. i /',' /NN /l J '
,,2:.
t //.*y ' T~'
Lt 6
? ! X ~! 1~$ $ /)f/ D 7 k $ ^x10-4 4 4
f-l3
' /D'V)'
f
~7
?t >
I/ I,i,S'$$/d>,XI'k*" N I
.N_!-t. %-. N'A v-~
a
/
/
/
^
-mgv y
-~
e
,-.->,- m.. _,c.,(_N.1,-xSm.A/ g.,/. ?
y' -
k e
r3 o
..q/.c.,O g / 4
/ ',. -r-, - y.--,<--+u-s-.--
?
v
' J, %U. :'r /-. ~ - -
b
'., J~.N ~ Q 4 - N '? !-
..'l'
2 : '. ', J4*
' /' '
/ i i
/
M. I e
e
~
'~-
-/./ / ^ e\\..N [A A f4Q _,, /
P
- g#4 5, y y, '/ y y y y.4._/g,/ v'
- 4
,,,, s 0
.a~<. -_
N N /76 4
.t,y v'/
/ / y 3' s p
,3 -
s,r' s
y s.
x s j.,/.
-/14w.x -% <w t.h VV. '
3
.: J. /.3. < lA X.. -/ {..' 4 N V,,Nv' /.n/. / '<
r.-
's ' '
a
-l,.f ',( '/l\\ ' '.-}..l /.9-{/i,d.x
~
f --
sp
's ',A \\ h '/< Y, ~/M / L. /, '
's.
'< I N/
(m.'s' V
-y-~ ~- -,'
y @M
-o-
>'f, y ! l,'*{,Y Y
h.:l
, haxb f/ /{ / Nf,
,k
-- s,-.,.n : A 4- {. e -.$*b! / f,)h>?,Y e-s t
.,. /..y Ny y-L -.
N- 'N, /y -,'n/ /q/
/.'6,
/
..t*r s
s
- -~-r 2~vt h:
G V
,ryJ.s :
.c
?
/
rp'//
.-- r- *r.-+--M -- -
o
.- y
/
?'-~
W, ' < /g M '/X VQ>.h,, f/1^ /'k N i V A,.,/ ',.t N, /,h'/.f..l'Ii.-
- j
- / M._. N. '/\\;
.4' s
-1
+
y-y. :.--y
> a -v i
,N'l*[. A '/- 2,N',b,;'V.
~ -
s
...-,-r
^ ;/+ v -.- ; -
/ A /._ \\ v 7 M/ }' '[y'/ '.(
.,}'s / /
y_. a /. 'W e/,'
l N
/
-j m - p,r s. 3
' %v,.4,.,< ya--2 fy r.--
s.
. p. _.-fu A, N
A,s
, s s
f 7
x (O I \\
(.,
s
? ' ',L,,,.
/-
e
.....,.s.,
4.,.
s
'. ' /p ~ 0.,A'V* L2--.-4
. 7,.9 s, 3. L./
~,-v-~
e
- /,,'f ;.f.. -
L-a-,t.D,4 L}-.
.,4 /
4 -,1- %-. -.
j-_.4 /.n fl-% ' Y',4%
a
,4. A tJQ Yt
$. '. -../ A t' L <? f I t n'-.,L.L 2 ' ' ' ' .tn. 4
/z..,r'..(D W'R/ : ** (. f-'g' Y Lh'! )b'.RW '. 'X :-
')
- '. 'f..a ',c'.-,L '.,: ) ' <.' '
s
' h/.--.
s //;
// 'V V i p')
/
s u /i>y4 ',a
'1)< v ;<L'< A D,
s ifX' N 4
/4
- /,k:/}s[. Q..N 4
' $a'[,( # }\\ !'\\( / /1 '@)W' v
4 sa/,../[4,[
J)-,
/
N c.-
n..
,,-) -. /.
- 5 l4, f -. U. ;. *,'"M"9 "
a.
e s
....N N./ N /
l
/
D 0 02
.,,....,-r s
?
N /,,
.' f a / \\/+ :/ ','(jn'-.,..~.,V,,,,y'
- - - p' ' v'. / \\. e,.* x/.-e,2
. a.N. t./ 's
/.X s
/
". N s
k N,&e& (r.f,../ '
m
~./Ni\\ \\
[/'j
/_ \\
,e k,$.;.-,fe/ / j i p Ny'/.~-N'f'<.gr[4;^4 ',
r
.' 9 1 y
-y a
.. u
,.-.s,
-.--m.
7.., 2
' 4h., k. J/ '(f,. b N.d'b'.n.INI b, n.,
N
[c
\\M
-f
,. -.>.c,,'y. ~.
N y' c% f
/..
.,. % c _ -.,%,--
-g N/
N -
/
/-
y:
N s.;,
s 1
, k
/ '$,
' :N, '
' j'. 's
\\
/
~
./
.oi
.on
.c s
.o.
.os
.co.e
.s
.s
.e e
a s
e e so PERIOD (secs.)
i Figure 12 - Instrumental Response Spectra Associated with Various Annual Probability of Exceedance Levels i
l
4 2.0 METHODOLOGY This section provides a description of the methodology that was used to calculate probabilities of exceeding different levels of ground motion transmitted to the site from future earthquakes.
The basic components of this methodology are described below.
2.1 Identification of Seismicity Sources The sources which can generate earthquakes within the study region we r r.
identified and the geometry of these sources was delineated.
2.2 Characterization of Activity of Seismicity Sources A Poisson model was assumed for the occurrence of earthquakes.
The parameter of this model is the average number of earthquakes of different magnitude per unit time.
This parameter is assessed f rom the Gutenberg-Richter relationship expressed as log N(M) = a - bM (2-1) i which N(M) = number of earthquakes greater than or equal to magnitude M in unit time a and b = empirical constants.
The constants a and b were estimated from the analysis of historic seismicity and geologic data.
For engineering purposes, bounds on earthquake magnitude, M,
may be used.
The lower bound is the minimum magnitude of engineering significance (for
- example, a
magnitude 4).
The upper bound represents the magnitude of a maximum credible earthquake tht.t each fault is capable of generating.
Equation 2-1 was modified to account for the lower and upper bounds on earthquake magnitude.
2-1
L 6
2.3 Characterization of Attenuation of Ground Motion The attenuation of ground motion was assumed to follow a
relationship of the type shown below:
D bN 2
(R + C) 3c (2-2)
S=b e
y in which F = ground motion parameter of interest (peak acceleration and spectral velocity for several periods up to 1 sec)
M = earthquake magnitude R = source-to-site distance b,b2,b3 = coefficients y
KM 2
C = magnitude-dependent constant = K1e (2-3)
C = random error term 2.4 Specification of Earthquake Magnitude Versus Rupture Length Relationship The approach used in the present analysis incorporates a line-source model which includes the length of rupture during the earthquake in a
manner similar to, that suggested by Der Kiureghian and Ang., 1977.
The release of earthquake energy was assumed to be along a line rupture of the fault.
Closest distance from the site to the fault rupture was used in calculating the attenuation of earthquake ground motion.
7ne following form of relationship was used for the rupture length, L
for an earthquake of magnitude M:
R L
= exp (Cy + C M), fo r M _<_My (2-4)
R 2
= exp (C3 + C M), for M > M (2-5) 4 1
The constants' C,
C, C
and C 2
3 4 were estimated froi. historic 1
data.
2-2
c>
v 2.5 Calculation of Probabilities of Exceedance saa 'ean number of events in which a given level of ground motion parameter is exceeded at the site during the period of interest was calculated.
Contributione of different magnitude earthquakes occurring on various sources a.d at different distances from the site were included in this calculation.
The mean number of events was then used to calculate the probability that at least one such event would occur at the site during the specified period of interest.
The computer program SEEP developed by Woodward-Clyde Consultants (Kulkarni et al.,
1979) was used to perform the probability calculations.
The program displays the overall probability of exceeding a specified level of a ground motion parameter.
In
- addition, contributions to the overall probability from the different sources, earthquake magnitudes, and distances are also displayed.
2-3
5 6
3.0 ASSESSMENT
OF INPUT DATA 3.1 G_eometry of Sources Table 3-1 lists fault lengths, closest distance to each fault from the site of SONGS Unit 1, and the maximum magnitude assessed for each fault in the study region.
The San Andreas, San
- Jacinto, Whittier-Elsinore, Palos Verdes
- faults, and the hypothesized OZD were selected for this analysis because the maximum earthquakes on these faults are expected to produce significant ground motions at the SONGS site.
Fault lengths were obtained principally from Jennings (1975),
with more recent maps consulted as applicable.
The closest point from the site to each fault was obtained by measuring a line from the site to the fault and perpendicular to the trend of the fault.
In order to measure the endpoints of the faults, the faults were approximated by straight lines, with no bends.
This approximation is generally close to the actual fault geometry.
The approximate lines were structured such that the distance to the part of the fault closest to the site remained unchanged.
3.2 Activity of Sources The magnitude-frequency relationship for each fault was determined from its moment rate using the method of Anderson (1979).
In this
- method, all slip is assumed to occur seismically.
Then the total moment rate (i.e.
the product of slip rate, facit length, fault width, and shear modulus) is partitioned among events of different magnitude assuming the Gutenberg-Richter magnitude-frequency relationship, which must be truncated at some maximum magnitude.
The magnitude scale used with the method of Anderson (1979) is the moment magnitude scale (Hanks and Kanamori, 1979).
These magnitude values do not differ significantly from M except for g
3-1
s s
magnitudes greater than 7.5.
A moment magnitude of 8.0 was used to represent the maximum magnitude on the San Andreas fault.
This value is compatible with the moment magnitude of 7.9 for the 1857 Fort Tejon earthquake obtained by Hanks and Kanamori (1979).
Using the slip rate, maximum magnitude, fault length values, and 11 en assumed fault width of 15 km, shear modulus of 3x10 dyne /cm2, and
'b' value of 0.85, the 'a' value of each fault was calculated.
Table 3-2 shows the inputs used in the probability analysis.
The central segment of the San Andreas fault (north of Wrightwood) is judged to be more active than the southern segment.
However, for the analysis, an average 'a' value was assumed for the entire fault.
This is conservative, since the more active central segment is farther away from the site.
The selected 'a' value and the
'b' value for the combined faults and the corresponding number of large earthquakes calculated from Richter relationship are compared with other the Gutenberg estimates for Southern California in Table 3-3.
It can be seen that the expected number of large earthquakes calculated using j
the selected 'a' and
'b' values is only exceeded significantly by the number obtained from Anderson (1979) for an equivalent set of faults.
Therefore, the selected recurrence parameters are con-servative compared to those obtained from seismicity data.
It should be noted that the estimates from siesmicity data generally j
pertain to much larger areas than that of the present study.
3.3 Attenuation Relationships The form of the attenuation relationship given in Equation 2-2 was used for the present study.
The parameters in this relationship were estimated for tectonic environment, transmission path characteristics, and local site conditions appropriate to SONGS.
Expressions were derived for attenuation 3-2 r
4 of peak ground acceleration and spectral velocities for several selected periods.
Figure 3-1 shows the attenuation relationships for peak ground acceleration.
Figure 3-2 illustrates the attenuation relationships for spectral velocity at a period of 04 seconds.
An upper bound on each ground motion parameter was specified through a
truncated distribution as described in Kulkarni et al.,
1979.
The upper bound values of the ground motion parameters used in this truncation accommoh te an uncertainty band of greater than 3a about the mean.
3.4 Magnitude-Rupture Length Relationship The calculation of the closest distance to site requires the extent of fault rupture length for a
given magnitude earthquake.
For magnitudes greater than approximately 6-1/4, the dimension of the rupture surface was established based,on the relationship for strike-slip faults given in Slemmons (1977).
For magnitudes less than 6-1/4 the relationship given in Patwardhan and others (1975), which was derived on the basis of correlation between earthquake magnitude and length of aftershock zone, was selected.
3-3
6 TABLE 3-1 FAULT PARAMETERS USED IN SEISMIC EXPOSURE ANALYSIS Distance from SONGS Estimated Fault or Zone Length Site to Closest Maximum of Deformation (km)
Point on Fault (km)
Magnitude San Andreas 540 92 8-1/4 San Jacinto 260 69 7-1/2 i
Whittier-Elsinore 230 37 7
Ilypothesized OZD 200 8
6-1/2 Palos Verdes 100 18 6-1/2 4
e~
,w
t a
TABLE 3-2 DATA ON SEISMIC ACTIVITY OF SOURCES USED IN SEISMIC EXPOSURE ANALYSIS Average Number of Slope
'b' for the Earthquakes Greater Gutenberg-Richter Fault or Maximum than M
=4 in One Year Relationship s
Zone of Deformation Earthquake on the Entire Fault loginN(M) = a - bM Central and Southern San Andreas 8-1/4 15.04 0.85 San Jacinto 7-1/2 3.39 0.85 Whittier-Elsinore 7
2.04 0.85 Hypothesized OZD 6-1/2 0.82 0.85 Palos Verdes 6-1/2 0.41 0.85
r TABLE 3-3 COMPARISON OF DIFFERENT ESTIMATES OF a AND b VALUES FOR SOUTilERN CALIFORNIA Normalized Time w.r.t.
this study Source Region Interval a
b N(M>6)
N(fl>7 )
1.
This study S.
California (Pleistocene 4.74 0.85
( 0. 44/y r)
(0.062/yr)
(specific faults) slip rate) 1 1
2.
Anderson S.
California (Pleistocene) 4.99 0.86 1.56 1.52 (1979)
Californja 1934 - 1943 4.77 0.85 1.09 1.07 3.
Richter S.
(1958)
(300,000 km )
4.
Ryall et al.
Southern California before 1932 2.53 0.55 0.39 0.78 2
(1966)
(60,000 km )
5.
Ryall et al.
Southern California 1932 - 1961 4.30 0.79 0.83 0.96 2
(1966)
(60,000 km )
6.
Hileman et at.
S.
California 1932 - 1971 5.36 0.98 0.69 0.52 2
(1973)
(238,600 km )
7.
Hileman et al.
Los Angeles Area 1932 - 1971 4.33 0.93 0.63 0.62 (1973)
Imperial Valley 4.27 0.85 Parkfield 3.64 0.80
i _.
x 0.3
\\
'N N
N N
N Nt.
N N
N NN>,N++
\\
\\
\\
N
' r N
N N
'N 5
\\
\\
\\y x:
g
\\
T+
N N
\\ x \\.
1
\\s N
N 8 0.03
\\
1
\\
\\\\
N-e N
y\\
s
\\
\\
0.01
\\
\\7
\\
\\
Ni
\\
\\
\\.:
N 0.003
\\
Relationships shown are for the Median Values
\\
Standard Deviation of f n a = 0.38 p
0.001 3
10 30 100 300 Closest Distance (km) i i
Figure 31 - Attenuation Relationships for Peak Ground Acceleration
4 300
.,,,g g
,,,,g 100 N
\\
\\
\\
\\h N
N N
30
\\
\\
\\
\\*
N, \\
{
~
\\
\\
\\>
N N
N
\\~
\\
\\
N N
$ 10
\\
\\
\\
i N
Y+
\\
N E
x#
x
\\
s S
f
\\
\\
\\_
3
\\
\\,
N
\\
q x
\\
\\_
x
\\
\\
\\
\\
\\
\\-
\\
\\:
Relationships shown are for the Median Values Standard Deviation of f n S = 0.38 y
3
l l
'l 0.3 3
10 30 100 300 Closest Distance (km)
Figure 3 Attenuation Relationships for Response Spectral Velocity (Period = 0.4 seconds, Damping = 0.02) l
4.0 RESULTS This section describes briefly the main results of the probabilistic evaluation of seismic exposure.
4.1 Probability of Exceedance Relationships for bround Motion Parameters The mean number of events per year,1, in which a given level of a ground motion paramecer is exceeded at the site was obtained direc tly from the output of the computer program.
The annual probability of exceedance, p, was then calculated from g
g = 1 - e-A.
(4-1; p
The annual probabilities of exceeding different levels of peak ground acceleration are shown in Figure 4-1.
Figures 4-2 through 4-5 show the annual probabilities of exceeding different levels of spectral velocity at selected periods.
4.2 Development of Equal Probability Instrumental Response Spectra Instrumental response spectra associated with desired annual probability of exceedance levels were obtained from relationships shown in Figures 4-1 through 4-5 for various periods.
The spectral values corresponding to annual probability of exceedance levels of 2 x 10-3, 6 x 10-4 and 1 x 10-4 are presented below.
S (cm/sec) corresponding to y
Period _
Ps = 2 x 10-3 P
= 6 x 10-4 Ps = 1 x 10-4 3
1 0
ZPA = 0. 4g ZPA = 0.5g ZPA = 2/3g 0.1 10.8 13.5 17.6 0.2 30 38 50 0.4 59 76 103 1.0 75 100 140 l
{
l 4-1 I
The instrumental response spectra corresponding to these three annual probability of exceedance levels are shown in Figure 1-3.
4.3 Contribution of Source, Magnitude and Distance to Probability of Exceedance The annual probability of exceedance values presented above include all of the contributions from the different magnitude earthquakes which could occur along the lengths of the sources examined in this study.
Each source, distance, and earthquake magnitude combination contributed a different amount to t?e total exposure.
As can be observed from the table below, the total probability of exceedance was dominated by the contribution from the Hypothesized OZD.
Relative Contribution from the Hypothesized OZD to the Total Period Probability of Exceedance 0-0.04 sec.
94 - 99 percent 0.1 93 - 99 0.2 92 - 98 0.4 84 - 96 1.0 50 - 60 Earthquakes with magnitudes 5-1/2 to 6-1/2 rupturing to distances within 8 to 12 kilometers of the site provided the vast majority of these contributions from the Hypothesized OZD.
For each other source, most of its contributions similarly came from the larger magnitude earthquakes with rupture at the closest distance from the site.
4-2 l
<a
+
l 1
1 l
10-1 1
e i
4 i
i i
\\
e 10-2 c
u 1
w o
b g
iino o.
l
%=
i C
d 10-3 2
cn q
oif n
~
o' 10-4 I
I I
I o
0.2 0.4 0.6 0.8 1.0 Peak Ground Acceleration (g)
Figure 4 Annual Probabilities of Exceeding Different Levels 4
of Peak Ground Acceleration 1
i l
i l
l t
n.
10-1 i
i i
i 6
i
=
g 3n-2 7,
g w
o b
g f
.o g
n.
~n 3
h10-3 Period = 0.1 sec Damping = 0.02 4
I I
I 10 O
4 8
12 16 20 Spectral Velocity (cm/sec)
Figure 4 2 - Annual Probabilities of Exceeding Different Levels of Spectral Velocity (Period = 0.1 sec, Damping = 0.02) i
e x
l 10-1 i
i i
- 8 10-2 m2
~
g w
C t
R E
^m E
5 10-3
~
Period = 0.2 sec Damping = 0.02 10-4 I
I I
I O
10 23 30 40 50 Spectral Velocity (cm/sec) i i
I i
i Figure 4 3 - Annual Probabilities of Exceeding Different Levels of Spectral Velocity (Period = 0.2 sec, Damping = 0.02) i 1
1 1
I
~
.a.
10-1 i
j 4
i g
i e8 10-2 n2 oy w
-o b
g S
i et 3
e 2 10-3 Period = 0.4 sec
~
Damping = 0.02 10-4 I
l O
20 40 60 80 100 Spectral Velocity (cm/sec)
Figure 4 Annual Probabilities of Exceeding Different Levels of Spectral Velocity (Period = 0.4 sec, Damping = 0.02)
o b.
10-1 i
g i
g i
i i
$10-2 I.
g w
O
?
g S
9 c.
~n E
$ 10-3
~
Period = 1 see Damping = 0.02 104 I
I I
f I
O 30 60 90 120 150 Spectral Velocity (crn/sec)
Figure 4 5 - Annual Probabilities of Exceeding Different Levels of Spectral Velocity (Period = 1 sec, Damping = 0.02)
o b
9 5.0 SENSITIVITY OF THE PROBABILITY OF EXCEEDANCE TO INPUT DATA VARIATION '
5.1 Introduction The expected values of annual probability of exceedance presented above were based on the best estimates of the input data from the assessment described in Section 3.
To examine the sensitivity of those results to variations of the input data, the probabilities of exceeding various levels of instru-mental peak ground accelerations have also been evaluated for extreme ranges of the various input parameters.
In these j
sensitivity analyses, variations have been limited to the data pertaining to the Hypothesized OZD rather than the entire study region because, as discussed in Section 4.3, the Hypothesized OZD dominates the total exposure of the site.
5.2 Results of Sensitivity Analyses The results of the analyses for an instrumental peak accelera-tion of 0.5 g, which are typical for other.lavals of instru-mental peak acceleration, are summarized below:
)
e for slip rate variations within the 0.30 to 0.68 mm/
i year range assigned to the Hypothesized OZD (value in analysis documented in Section 4 was 0.5 mm/ year),
j the annual probability of exceedance ranges between 4 x 10-4 and 8 x 10-4;
\\
e for values of the parmeter
'b' (of the Gutenberg-Richter rela'tionship) between 0.7 and 1.0, the annual probability of exceedance varies from 3 x 10-4 to 8 x 10-4; 5-1 l
i
B g
a for a maximum magnitude earthquake of M= 7 on the Hypothesized OZD s
together with the maximum slip rate for the Hypothesized OZD, the annual exceedance was calculated probability of calculated for M to be within 10% of that s 6-1/2; e using dispersion values varying with magnitude from 0.62 for magnitudes 4 to 4-1/2 to 0 38 for magnitudes 6 to 6-1/2 with the mean attenuation sented in Section 3, has minor effect (apprrelationships pre-percent increase) oximately 20 on the annual probability of exceed-ance for the magnitude range 5-1/2 discussed in Section 4.3, provides theto 6-1/2 which, as the Hypothesized OZD contribution; vast majority of The use of this type of dispersion relationship, decreasing with i.e.
increasing magnitude, is consistent with the observed trend of the empirical d t
and Bornstein (1978) provide dispersion values as a ata.
Donovan function of acceleration level accelerations greater (varying from 0.3 for ation of 0.05 g).
than 0.3 to 0.48 for an acceler-
- Also, the use of dispersion values (larger than 0.38) with ship selected for SONGS would bethe mean attenuation relation-j general, published too conservative.
In dispersions relationships with higher reported exhibit a correspondingly lower acceleration and cover a wider range in sit mean peak than the e conditions attenuation relationship used in the present study (Idriss, 1978 contains a summary of widely used attenuation relationships).-
As a result, the calcu-lated probabilities of exceedance using a acceleration level would be decreased providi lower mean or no effect on the calculated probabilities ng little 5-2 L.
p
.x.
Considering the observations above, the annual probability of exceedance for an instrumental peak ground acceleration of 0.5 g was found to be in the range of 3 x 10-4 to 9 x 10-4 for reasonable variations of the input data.
Because this represents a small variation in probability from the calcu-lated 6 x 10-4 and because the variations evaluated above are considered extreme, the calculated equal probability spectra shown on figure 1-3 are considered reasonable and conservative for the SONGS site.
5-3
EE.
Na a 6.0 REFER E!CES
- Anderson, J.G.,
1979.
Estimating the seismicity from geological structure for seismic risk studies.
BSSA 69 135-158.
Der Kiureghian, A.,
and Ang, A.H.S.,
1977, A fault-rupture model for seismic risk analysis.
Bull. Seism. Soc.
Am.,
Vol. 67, No.
4, pp. 1173-1194, August.
- Donovan, N.C.,
and Bornstein, A.E.,
- 1978, Uncertainties in seismic risk procedures:
Journal of the Geotechnical Engineering Division, American Society of Civil Engineers, v.
- 104,
- n. GT7, July, p.
859-887.
- Hanks, T.
and H.
Kan amori, 1979.
A moment magnitude scale.
JGR 84 2348-2350.
- Hileman, J.A.,
C.R.
Allen and J.M.
Nordquiest 1973.
Seismicity of the southern California region,, 1 January 1932 to 31 December 1972.
Seismological Laboratory, California Institute of Technology, 481 p.
- Idriss, I.M.,
- 1978, Characteristics of earthquake ground motions:
State-of-the-art paper presented at American Society of Civi] Engineers Specialty Conference on Earthquake Engineering and Soil Dynamics,
- Pasadena, California, June 19-21.
- Jennings, C.W.
1975.
Fault map of California:
California Division of Mines and Geology, California Geological Data Map Series, Map no.
1, scale 1:250,000.
Kulkarni, Ram B.,
K.
Sadigh, and I.M.
Idriss 1979.
Probabilistic evaluation of seismic exposure.
2nd U.S.
National Conference on Earthquake Engineering, Stanford University, August 22-24.
j Patwardhan, A.S.,
D.
Tocher, and E.E.
Savage 1975. Relationship between earthquake magnitude and length rupture surface based on aftershock zones.
Program for 70th Annual Meeting of the Seismological Society of America.
- Richter, C.F.
1958.
Elementary Seismology.
W.H.
Freeman, San Francisco and London, 768 p.
- Ryall, A.,
D.B.
Slemmons and L.D.
Gedney 1966.
Seismicity, tectonism, and surface faulting in the western United States during historic time.
BSSA 56 1105-1135.
6-1
- Slemmons, D.B.
1977.
State-of-the-art for assessing earthquake hazards in the United States; Part 6:
Faults and earthquake magnitude with appendix on geomorphic features of active fault zones.
U.S.
Army Engineer Waterways Experiment Station, Vicksburg, Contract No. D\\CW 39-76-C-0009, 129 p.,
plus 3 7 p. Appendix.
l l
l l
i 6-2 I
. l