ML20035A704

Summary of 930309 Meeting W/Llnl Re Probabilistic Seismic Hazard Estimates for Eastern Us.List of Attendees & Meeting Agenda & Handouts Encl

ML20035A704
Person / Time
Issue date:	03/11/1993
From:	Sobel P Office of Nuclear Reactor Regulation
To:	Bagchi G Office of Nuclear Reactor Regulation
References
NUDOCS 9303290179
Download: ML20035A704 (67)

Text

p

&%( Rlg t

i

[v-c,..

o UldiTED STATES

! " 3.,.q j

NUCLEAR REGULATORY COMMISSION D

WASHINGTON. D. C. 20555 y

g a

/

gg 11 12

~*

NEMORANDUM FOR:

Goutam Bagchi, Chief Structural and Geosciences 5 ranch Division of Engineering Technology THROUGH:

Robert Rothman, Chief f/

Geosciences Section NL Structural and Geosciences Branch Division of Engineering Technology FROM:

Phyllis Sobel, Geophysicist Geosciences Section Structural and Geosciences Branch Division of Engineering Technology

SUMMARY

OF A PUBLIC MEETING TO DISCUSS RECENT LAWRENCE LIVERMORE NATIONAL LABORATORY'S (LLNL) PROBABILISTIC SEISMIC i

HAZARD ESTIMATES FOR THE EASTERN UNITED STATES (EUS)

A public meeting was held on March 9,1993 to discuss revised probabilistic seismic hazard estimates for the nuclear power plant sites in the EUS (east of the Rocky Mountains). A list of attendees is attached as Enclosure 1.

The agenda and handouts are attached as Enclosure 2.

Phyllis Sobel, NRC, opened the meeting by discussing the purpose for the revised calculations.

In the last few years LLNL has improved their approach by applying the methodology at Department of Energy (D0E) sites.

Based on LLNL work done for DOE in the last few years, NRC felt that limited reelici ation of the LLNL experts was needed to refine the estimates of uncert

'v in seismicity parameters and ground motion models.

The prelir results presented at the March 9 meeting are these improved and updatt

.les.

After undergoing NRC review, the new results could be used i

in future actions.

)

m Jean Savy, LLNL, described the revised hazard calculations.

In probabilistic seismic hazard analyses, the experts define maps of source zones and then describe the seismicity of each zone in terms of the rate of earthquakes versus magnitude for each source zone.

In the 1980's the LLNL experts were elicited for a and b values (the intercept on the y-axis and the slope of the rate of earthquakes versus magnitude). Then the experts were elicited for the uncertainties associated with these a and b values.

This procedure led to I

higher uncertainties at the larger magnitudes.

260006 jgy -3 [

rr o

.~

~

c en 7'-

f

..;p

.t

. w:

}

%(

' ~ ' ' *

/r

~

(.

ji

'l(t

.}

9303290179 930311

~

. w:.

.::w LAIL p {-g p

(..

t g-In 1992, seismicity recurrence rates (including uncertainty) were elicited by frequency of occurrence at specific magnitudes. All the seismicity experts were more comfortable with this approach. The result of the elicitation was that the uncertainties associated with the seismicity recurrence curves were reduced at all magnitudes. The 1992 elicitation was limited to the seismicity recurrence rates and upper magnitude values and zonation was not re-elicited.

l For the 1980's LLNL study, each ejround motion expert provided several ground I

motion medels which estimated ground motion as a function of magnitude and i

distance.

In the 1992 elicitation of ground motion experts, emphasis was placed on updated ground motion estimates based on the current state of knowledge and uncertainty estimates. The ground motion experts were asked to provide peak ground acceleration and spectral acceleration estimates for specific magnitudes and distances. Then LLNL aggregated the ground motion inouts into a composito nrnund mntinn mndel.

This composite ground motion model is used in the hazard calculations.

The revised hazard estimates at the EUS plant sites are lower than the 1989 estimates, principally due to the revised recverence curves.

The revised LLNL calculations were still higher than Electric Power Research Institute (EPRI) hazard calculations for EUS sites. Additional feedback with the LLNL ground motion experts is planned for this spring and the final estimates will be produced later this spring. At that time the NRC will make the results publicly available.

i l

gngtosistM*M l

l Phyllis Sobel, Geophysicist Geosciences Section Structural and Geosciences Branch Division of Engineering Technology l

l

Enclosures:

As stated cc:

J. Richardson B.D. Liaw L. Ruth A. Murphy, RES L. Abramson, RES l

DISTRIBUTION:

Central File PDR l

ECGB R/F l

GBagchi RRothman PSobel ECGB:DE

/ -

DE EC

DE

~

PSobel(/ -

othman GBagchi 3/1/93 3/p/93 3/gp/93

F ^ld da$ M 5

[

ATTENDEES PUBLIC MEETING EASTERN U.S. SEISMIC HAZARD MEETING (Revision of Livermore seismic hazard estimates for Eastern U.S. sites)

March 9, 1993 Holiday Inn Crowne Plaza, Rockville NAME AFFILIATION l1 5 Soh NTc / ver rl D \\s/

Q y

lf %

O,b "'

,An kt f 4 L e e Q wu x.m s

,/ \\[A, W h a t.

Il 4)Tl-)rd n+1 N 6c /N Cl?-

Oake Ilcd lu, 13 M j Phal b/rt)6 Hl./)M!

SQE M

• bu M S C-s JSk is o.

Li A-w N il t / Notin l O E beu

9. Moore G o h C ~,, _ 6 es. 6 d.A. duke FP g L 0

c2a Wd4u W

sku,,

G., at alm Pc NcE kt?(./

42o Gs /ruEUAbe /nsnt Jean 5p-t/y

LLNL, m 'h'2 [ c Icr &

RK.H,xnneAnn G)<cc,,a;lElees:j'cwe,q sus Giere - Ko cn ug.a. /oze 2 6 M em u te,,

orc-ec c do e (~

>nf Nn $2 rief+a

%c

$ni iTW kn w wer</ 6Pisoid

'bu w %I SY8xeY h f~i 7 Lusw we

U.

ATTENDEES PUBLIC MEETING EASTERN U.S. SEISMIC HAZARD MEETING (Revision of Livermore seismic hazard estimates for Eastern U.S. sites)

March 9, 1993 Holiday Inn Crowne Plaza, Rockville NAME AFFILIATION

$ n UTA M H A b t (+I

(). i W 2 C' JU A'l E E u FARucWL AJ U K A E C

[%

Or &

YlL'lk'kf E

&nn e,

St-PrA G Oe c.'tu h<kk 4

t D a < Q. kocJeen N O M A R C Ql n %ww hos a a %vn a P&+s l'D iI?4 CJ

///vms Corio Lh eu ik - A i, be c McS r$o6 LO h o rcrc,o '

50cs CIaouan EtEc h Ces

_ ' cl. 2Ec-RvER M. T. 4//{. O ketse n% ice cs Dc5 ) DP-6 L I Wisu

\\/Nov >

Pwt c.Roo Assocairs /tby IT ANNnM Acl O Coto.

E h\\ &nsmdc Ah' i+,. :-, Ts &

u cu<- w & >H w.a u

~

0 0

k nC/cSw( M c

EASTERN U.S. SEISMIC HAZARD MEETING March 9,1993 9:00 am - 3:00 pm Holiday Inn Crowne Plaza, Rockville AGENDA 9:00 Introduction P.Sobel 9:15 Procedure for elicitation of seismicity J. Savy input parameters, summary of elicitation 9:45 Procedure for elicitation of ground motion, summary of elicitation 10:15 Break 10:30 Sensitivity studies l

- Effcc cf scismicity update

- Effect of ground motion update

- Combined effect i

11:30 EPRI vs 1989 LLNL vs 1993 LLNL results P. Sobel 12:00 Lunch break 1:00 Discussion All 2:30 Summary P.Sobel t

i

,. l t

?

4 l

I

(

h i

I I

r I

i LLNL EASTERN U.S.

SEISMIC HAZARD PROGRAM i

I I

4 I

I l

l e

i I

f h

1 i

m.....

..m n

I'n I LL13 DOBEL MARCH 9,1993 i

J

)

e d

I

[

i n

J I

d t

f l

I e

I i

?

P

i l

BACKGROUND I

j LLNL developed probabilistic seismic hazard o

estimates at all operating plant sites east of the j

Rocky Mountains for the NRC (NUREG/CR-l I

5250).

I i

i

'(

EPRI developed seismic hazard estimates at 57 of o

the Eastern U.S. sites.

j o

The differences between the LLNL and EPRI seisritic liarard estituates were addressed in l

NUREG/CR-4885 (1987) and will be studied in the l

NRR/ DOE /EPRI resolution program.

l

+

i o

Based on LLNL work done for DOE in the last j

few years, NRC had LLNL conduct a limited.

l reelicitation of the seismicity and ground motion i

experts to refine estimates of uncertainty.

i l

)

i

j a

1992 ELICITATION OF SEISMICITY EXPERTS 2

1 l

In NUREG/CR-5250, higher uncertainties i

i o

j associated with seismic recurrence rates at j

j moderate and large magnitudes due to 1980's l

elicitation process (elicitation of a and b values l

l (intercept on the y-axis and slope) and their j

uncertainties).

2 i

i 4

In 1992, to improve the modeling of o

l uncertainty, seism.icity recurrence rates j

j (including uncertainty) were elicited by-l frequency of occurrence at specific magnitudes.

4 t.

l o

Seismicity experts more comfortable with 1992 j

j approach.

i

~

o Uncertainty in seismicity recurrence curves i

reduced.

l i

-i o

Zonation was not re-elicited.

l 4

i i

l

j l

l ELICITATION OF GROUND MOTION EXPERTS i

i i

Updated ground motion estimates based on o

current state of knowledge.

o Peak ground acceleration and spectral acceleration estimates were made for specific magnitudes and distances.

o Greater emphasis piaced on uncertainty estimates.

4 i

o LLNL aggregated experts' ground motion inputs into a composite ground motion model.

4 i

1

_m.

. =..

m m-i SCHEDULE l

i May 1992 Workshop on expert elicitation.

l I

Nov 1992 Seismicity experts elicitation l

complete.

Dec 1992 Ground motion experts elicitation complete.

i Sensitivity studies.

Preliminary hazard estimates.

l l

l Mar 1993 Public meeting to present results.

l t

i April 1993 Additional feedback with ground J

motion experts.

l June 1993 Final spectral hazard-estimates.

i

+

5 t

i Eastern U. S. Seismic Hazard Meeting March 9,1993 1

9:00 a.m. - 3:00 p.m.

Holiday Inn Crown Plaza, Rockville l

Updating the 1989 (NUREG-5250) i Seismic Hazard Results Jean B. Savy Lawrence Livermore National Laboratory P. O. Box 808, L-196 Livermore, California 94550

~

Project Background f

O In January 1989, LLNL published the results of a multi-year of efforts in estimating the Seismic Hazard at the Nuclear Plant Site East of the Rockies (NUREG/CR-5250,8 volumes)

O The funding agency was NRC/NRR L

1

}

i i

i l

l x

EG93-205JB3vg/3/93

,2

r 1

The Goals of the EUS study were two fold:

d

~l i

O Develop a good central estimate (median hazard)

O Characterize the total uncertainty in the estimate of the hazard t

4 r

4 L

EG93-205JB3vg/3/93 - 3

There are four steps in our probabilistic seismic hazard analysis Zonation Rscurrence

' ' ' ~

Log N a

l Fault 1 -,

300-0

'X 500 km A Site Area B

-1

'x,

-2

=

300-500 km 4

5 6

7 8 Magnitude

=

Attenuation Seismic hazard curve calculation PGA" H(y)"

(9) _\\b 10-1 b-Attenuation a

f curve for 10-2

/

magnitude M 10-3 d

10 i

I I

m y

R Distance 0.1 g 0.5g 1.0 g y

~

Shaded area = P(PGA > a lM,R)

L EG93-205JBSvg/3/93,- 4

_..,...s

......... ~..

.. _, _. -, _... _, _ _..., -.. - - -. - ~..... _.. _... - _..

. ~. -

' We use a Monte Carlo simulation to propagate uncertainties h

! Wz, g

311p W # kh zonation A.. lh m

$fSS H (y) -

'~

kh N bitNbkbk Simulation i 7

i D

w, N'

o

(

s Attenuation ik w, (g; e [{

...-.=:...-

'N o

N

=-

x N

W k

" -'

~

e 1

h

...?

17 w, H(y)m

[sigHsfi6%g i

um--

A.

w,,

Recurrence

-- Mean g,

Median

~Y l

l l

L EG93-205JUSvg/3/93 - 5

Expert Elicitation A.

Workshop on Expert Elicitation LLNL, May 27,1992.

Panel George Apostolakis R. Keeney R. Clemen P. Morris S. Kaplan R. Winkler Topics and panel Recommendations 1.

Elicitation of Ground Motion Information Information to be elicited: experts' judgment about level of ground motion at a site given an earthquake of magnitude, m, a distance, r, l

from the site i

i EG93-205JBSvg/3/93

,6

~ -. _ _

Expert Elicitation cont'd h

2.

Aggregation of Elicited Information Preferred aggregation: aggregate seismicity and ground motion information separately prior to combining to assess seismic hazard Recommendations Aggregate ground motion probability distributions a.

Analytical methods using weights acceptable No weights "best" Equal weighting is one useful aggregation Should recognize dependencies among experts i

b.

Sensitivity of hazard results to aggregation methods should be investigated and reported J

L EG93-205JBSvg/3/93 - 7

..___... _ _ _,. _,.. _. -. _.,.. ~.. _. _...

Expert Elicitation cont'd Recomrnendations Elicit each expert's probability distribution of the level of ground a.

Inotion

- probability is personal

- allow for some " vagueness" or uncertainly in specifying probability b.

Elicit probability distribution directly, e.g., fractiles, rather than indirectly in terms of ground motion models and distributional parameters Elicitation based on personal interviews with possible feedback and c.

re-clicitation d.

Documentation of resources and reasoning used in formulating distributions is important i

4 i

EG93-205JBSvg/3/93 - 8

- - - - _ _... _ -.. _,.. _ _ _ _. ~,,.. _ - -. -. _ _ _. _ _ _, _ -

- -. ~ -.

m..-

IV Data Diagnostic Tools (DDT)

Two Purposes 1.

Check input data for S-and G-experts for:

sanity consistency (with empirical data and with physics principles) gross errors in transcription logics reasoning validity of assumptions and supporting data for selected models 2.

Provide the experts with unbiased information to help them formulate or refine their opinion analysis of available data uncertainties estimates graphical displays of available data in a variety of different formats comparisons of experts input to data and/or to other experts opinions L

EG93-205JBSvg/3/93 - 9

_, _. ~. _

I e

l r

i Seismicity Data Diagnostic Tools 1

F_) Data Diagnostic Tools for checking on expert's seismicity estimates include:

• Comparison of upper magnitude cutof f, Mu, with the largest recorded magnitude I

within a zone

• Comparison of expected frequency estimates with the historical data within a zone

- :284,84,34 years t

l

• Expressing the estimated frequencies in terms of estimated return periods

• Comparison of 4

estimated frequency curve mathematically estimated expected frequencies based on the historical data and a model for incompleteness corrected" historical data I

a e

EG93-205 TBSvg/3/93 - 10

Seismicity Data Diagnostic Tools cont'd h

i El Comparison of the experts' uncertainty bounds for expected frequencies with uncertainty bounds for mathematical estimates of expected frequencies based on the variability of the histc>rical data, i.e., the historical data represents a sample

• based on resampling (bootstrap) methods

• assumes either occurrence of earthquakes within a zone is Poisson occurrence of earthquakes throughout the EUS is a stationary process over time

• Comparison of regional carthquake occurrence rates based on expert's seismicity estimates historical data, allowing for variability in

- number of earthquakes that ht re occurred

- recorded magnitudes

-incompleteness of the historical data EG93-205JBSvg/3/93 - 11

. = - -

< a and b uncertainties can lead to unrealistically large variability on the occurrence estimates at large magnitudes Log N JL Log N = a - b m s

N NN b/

i 8m 3

magnitude D

I EG93-205JBSvg/3/93 - 12 L.

i c

0 m

vg wNW CD v CD ID m -

s V @ CD G

V c

- g cD v cp N

owo UCMG 0

N N

c o

N

~

oo g

. CD CD CD r

IA 1

i oo 0

- CD l

- In CD m--

3 w

d, f

O@-@

e y

N D CD G in CD N NOG G

Un{

C I

" - b.

m

'O GmG l

t 0

m 0 D.D b.D.D IJ t

c _O - 3

.D - 3 l

c i

o i

b f

I m

l 3-~ W l

0 E-O r

a l

l O

L G

O o

.a g

g D

O o.

U g

N e

x c

O E

b O

o i

N O

E L

0 y

o t

L I

D

'+

.h iO

/

c O

O l

g L

a o

OmD t

l V

1 0

Ox o

T D

1 c

j s

n G

• \\

o5 u

m i

E I

1 l

O r

i V

O W

[

f G

Q Q

t i

I e

9 3

Q 0

0 0

o 0

~

i

?

s f

3 i

i

.n

+

f

[

i 4

b 4

r 4

a

~

i i

L

~.

I t

.I i

Table 2A j

Summary ofInterviews Disinbuuons Max Mag Zonation i

j Expert for M u Update Update l

Bollinger Quadratic Some upward (~ +.2)

OK, but possible (bell-shaped)

N.M 7.5 -+ 7.35 minor changes l

CZ 5.5 -+ 5.75 4

4 Holt Triangular No Changes No changes j

l Johnston f3 Revised upward (Paleo studies)

No changes

-i tor N.M.,

Removed some alternatives j

(7.4 --> 8.3), Charl. 5.8 -+ 7.0) i Beta

.i Kafka Revised all, small changes Change map weights

{

)

]

Diamond Lawson Beta Most B.E retained, some Removed smalle.r zones in bounds reduced.

Arkansas.

I l

Long Triangular OK Major change Giles County 6.5 OK, but would change Charl.

Beta preferred

-+ 7.0 (Z8) f Others mostly unchanged i

4 q

i i

Pommeroy Triangular No changes No changes Stepp Beta and diamond Few BE re/' iced. Charl. raised No changes j

i

+

t 6.9 -> 7.2 2

Some ranges reduced Toks6z Triangular No changes in B.E, mostly No changes reduced ranges

?

3 l

4 4

s 2

3 l

4 i

1

-n

[

s i

~

i Table 2B Summarv of Interviews cont'd Frequency i

i Expen Estimates impact - Comments i

Bolhnger Preserved BE Median estimates mostly unchanged revised ranges Mean slightly lower comfonable with new process Ilolt Preserved B.E, greatly reduced Reduced mean bounds unchanged median j

e Johnston Preserved B.E, reduced bounds Slight reduction of uncenainty from

)

zonation; large reduction from seismicity ranges i

/

i Kafka Mostly preserved B.E, greatly Large reduction in mean

.l

~

reduced bounds Slight reduction in median

j i

I Lawson Mostly preserved B.E, greatly Drastic reduction of mean values i

reduced bounds No change in median j

)

Long

[ Mostly no enange m u.t piist.

Reduced mean estimate i

values)

Median unchanged Greatly reduced bounds i

Pommeroy Retain B.E, Retain median, drastically reduced mean Bounds, factor of 3-5 i

i Stepp Revised, giving high weight to Little changes in median f

previous B.E Great reduction of uncenainties, mean i

j Keep b value, reduce bounds (factor of 2 to 5) i Toks6z Mostly preserved initial B.E. t.

j Little change in median, greatly reduced

(

factor of 2 to 3 for bounds mean l

i j

l l

i l

l l

l i

(

-., _ -.. ~

l Plot GM Symbol Model No.

i 1

RV-1 A 2

RV-5 A (X2) 3 G16-A3 4

SE-IA 5

SE-2A 4

10 i

+

3 3

3 jh i

2 se M%N

~

U N

\\x h

s\\..

8 q

3 e

2 lo' N

\\

s N:

\\ \\

1 0

N 4

10 n n 5 en om, a

n, e> om, n n 5 <> om.

" o.-

"o

~

o o

l I

DISTANCE-FM P

Figure 3.4 Best estimate PGA models listed in Table 3.5 plotted for magnitudes of 5 and 7.

Rock base Case.

r t

_49-l

t i

EUS Ground Motion Model Panel Members

]

(G-Panel) i 1989 Studv - NUREG.5250 e

1 Dr. David M. Boote (1)

Dr. Kenneth Campbell i

I Professor Otto W. Nuttli (1) (2)(3) f Professor NafiToksoz 14 Professor Mihailo Trifunac (I) j Dr. John Anderson j

Dr. John Dwyer i

Notes:

Provided the final sets of ground motion models used in 1989

)

study 3

(1)

Participatcd as a mcmber of the SEP EUS Ground Motion Panel l

t 4

i (2)

Also member of the Seismicity Panel l

j (3)

Left the Panelin June 1986 l

i 1

5

.]

i 4

i l

k i

i i

l

)

1 2

4 i

w 4

e J

~

l t

l t

l Plot GM l

i, Symbol Model No.

l 1

1-RV-2A 2

RV-5A (X3) 3 SE-1A (X4)

l 4

4 SE-2A (X4)-

5 SE-1 A (XI) l 6

Comb-IA.

4 10

.I l.;

lo 9

N k

-ww 2

s'

+

io

~

SR x

T I

'1 1

4 0

10 f

4 i

i i

l i

o

\\

10

, a, w-, '

, a, w -, '

I u n u

n u n, a, wm,

• o "o

"o o

l i

DISTANCE-KM

/

i i

4 l

Figure 3.5 Remaining PGA models listed in Table 3.5 plotted for l

magnitudes of 5 and 7.

Rock base case.

i l

,=

W Eastern U. S. Update Ground Motion Modeling List of Ground Motion Experts t

2 l

K. Aki USC

=

1 W

G. Bollinger - M. Chapman VPI D. Boore USGS-K. C=npbel!

EQE J. Fletcher USGS l

J R. Hemnann St. Louis University M. Trifunac USC a

i r

j N

L I

.l i

f J

3 f

5 9

i W

9 W

b I

1 n

L

en 1992 Ground Motion Eligitation M

'l l

I I

I I

I

'Is I

i 1

I T

s l

I x

s

)

I Ni M

l s

J N

C l

l l

s 1

o c

s j

l I

,9 1

l NI s

2 i

i O

I IN 1

N 2

t i

I I

l s

s s c

t c

I I

I c

l l

s(

' 85%

h N

l I

I i

i I

s 1

I I

O I

I I

-N 50 %

'15%

I I

R, R,

R, Distance R,

R, R,

Distance Set of multiple attenuation models Composite model J

V EG93-205JBSvg/3/93 - 13

e

~

J 3

1 3

9 M

/

3

/gv S

BJ e

5 c

0 M

n 2

3 a

9 t

G s

E iD N

l I

, i R

n sw coiy mC5uO o

i t

a t

i e

c c

M n

il a

E t

s i

n D

o i

to M

dn I

iI I

I iI i

N u

o r

G 9

89 1

~e@a $rSa tac 2O g

yj

^ ",.

S.

a n

v 15, 50. and 85.

-a percentiles

=

orIthmetIc means 8n horord curves using S & G experts (g

7

.1

'j g.

2 d mean gqqty" m

3*

b

.01 gg

.4 o-

.p 9

E

.001 o

N e

e m

2e o

~

o M

1e-4 N

5 iL

% 86

{

o n ie-5 x

n M-n Mfeddyp 5

0 m

e 1e-6 e

.o E

M o.

1e-7 0

500 1000 m

acceleration cm/ secs *2 g

pilgrim EUS update, 1992 Selsmicity, 1989 g.m. methodology, 6

SHEET 4 1989 9.m.model, M>=5

I 53 E8

=g 15.,

50. and 85.

-a percentiles

=

arithmetic means 8n harard curves using 5 & G experts

', 3

.1 9.n<

t 0

y'E 18 yg h

.01 o.

I, o

E

.001 e

e i

3 o

m 1"_4 N

E N

e N

O 7

T m ie-5 I

n M

n 5

.oe 1e-6 no

.o 0-N L

0-2

~

le-7 O

500 1000 N

acceleration cm/ secs *2-g pilgrim EUS update, benchmark PGA Ci SHtti 4 M >= 5, 1989 seismicity and 9.m.

4 *

.,. -...,, _ -. - - -.. _ ~ _ -.. _ -. -......., _ -. _. _.. _ _. _... -........ ~.

b i

Ji $

u.s g

percentiles 15.,

50. and 85.

ga

=

orithmetIc means an hazard curves using 5 & G experts

',,3 1-

$n t

c m

5" h

.01 gg o-v 0

0 h'

.001

\\

a o

\\

T En B

ie-4 3N

\\

E o

m ie-5 T

nX n

n

.o 5

e 1e-6 n

.o o

E M

o-le-7 3

0 500 1000 N

acceleration em/sec**2 I

braidwood 5

EUS update, 1992 Selsmicity, 1989 g.m. methodology, E

SHEET 4 1989 g.m.model, M>=5

381Da nj en? 8 gf0 aT.a2* 5 N5 xn E m0 *.N2 $ m g I N 7

I 1ngb 5 5. ga'

,S.

u 0

0

'0 s

• 1 t

re p

5 x

t 8

e G

d n

2 a

5 m

0 g

c 9

5 n

e s

d i

s

/

n u

0m Aa 5s 0e G

1n s 5

Py a e n

t v

o ki N

e rc mr i

=

u t

ai sc c a

mm r

hs eI d

e ci It d ne ier l

es tm o e ob nh r c

et e Q

c 9

o

,8 h

a eI w e9 rr t1 ea da p

d

\\

p5 i

au

=

s 0

1 1

1 4

5 6

7 rS>

U 0

0 b EM 0

e e

e e

l l

1 l

tem bO eEovo0N*

m o.eD.0La t

4 T

E E

H S

1 r

i

+

t

(

i e

_ ts!

O N

M u

U m

\\

E U

1 O

~

G 0

W U

u rJ l

f w

Ct' CL 4

tJ

/

c 4

3 C

c O

E J

i m

a

>I I

I i

i 1

VJd d

V LO (D

DO G

G I

i oz d

G D

D 0

b oJ i

3J D

l i

C

{

O C a

L U

.D E acuepaaoxa jo qosd l

4 d

i

)

1 a

k t

N J

f 1

O O

O

.-e N

N N

U bm

\\

E U

U g

7 b

O o

~

O W

b If)

U U

O m

b i

M

)

Z b

[

I i

0

/

1 3

l m

D L

3 1 I

I I

i e

C V C*

w y

p g

O O

s a

o O

o o

o~

3 CD

-H w

w U

C C C L D DE G30Pp3G3xa jo agoad t

L 1

5 34

}

+

1 i

i

I braldwood i

oId 89 ccay -C l

medIen & 85'/. pIotted NEW-92--N vs

.1 i

.01 e

U E

.001 C

m 00 0x le-4 o

C

+

o le-5

.O O

L Q.

le-6 1e-7 O

500 MO accel-cm/sec**2

BRAIDHOOD meadlan & 85% plotted

.1

.01 ou 5

.001 tg i

L 0

L u

g le L s

le-5 1

o e

.oo L

le-6 c

1e-7 O

500 1DO accel-cm/ secs *2 6

0

.,-..,-+-e

.r.

.,4

- - -.. ~ -

---..r.

+

, - ~ -, -

• g o

V M

i

_ O 3

0 t

I t

f N

H U

O i

M l

\\

E U

U I

I al O

~

s C

e 0

l t

to u

u o

C

~u.

M Z

i 19 i

a

/

/

6 o

M U

L 3 ;

i OC-

~

v LD W

N (5)

O

^

O G

I I

I I

C G

O O

O O

oN em

+

+

+

e M

C o

U EE acuepaaoxa 30 qoad l

~.

P t

I O

_ (53 (5)

U I

m O

• -o N

m CD N

o o

U i

m i

o

\\

E m

u 1

0 Z

0 0

I in uu i

I 0

N l

2 LJ Z

.x B a

e, 65

/

/

Ev 6

6 3o D-C O.

w

\\*

DW l i

I I

I I

C (D g

o

~

V 1A t.0 N

G G

I I

i i

ed G

0 0

0 0

g

~

a w

w w

EV O

E acuepaaoxa 30 qo sd i

r MILLSTONE housner meadlan & 85'/. plotted

.1

.01 eo L

g

.001 -L o

L N

N le-4 E'

e

.\\'

le-5 f

~N o.

le-6 1

i 1e-7g 500 1000 acce1-cm/sec:*2 c--

6 6

0 0

8 0

1 2*

ces

/

m c

0 l

0 e

8 5

c ca I

RP

[

nae m

L S

L V -

0 eL1 1

1 4

5 6

7 0

0 nN 0

e e

e e

oL t L l

l l

l s

n l

a l

e i

mm e U'a meo u g O jO f

I

f e

w

,8 ED2 nog <n o g%.rmO oaiaNe" E T n5)f5no N1? 0 q 5 I w.

y d$? go, 79. * ~ zy a.

-w v

v i

h d

t 5

' 0 8

' 0 0

v 1'

y s

c t

re w

p w

x 5

8 e

m G

d 2

n s

m a

5 a

c g

0 g

e 5

n s

d i

/

n s

0m Aa 0c sG u

m e

5s 5

Py 1n s ni t

-y e

o ki r

a rrc e v

.i ai

= m r t

r u

a mm sc c r

ahs w

e cl y

ehne ei d

l 1tie r e

es v

n a c

b tr n

,9 nh z c

8 a

a et h o e9 cl t1 rr ra ea ad

+

p p5 eu

=

O hS>

6 r

1 1

1 4

s 6

7 U

0 0

s EM r

0 e

e e

e i

t 1

1 w

r, e Um bQ ~ eEmmeDND o mj

• o= 0 e o.0 L 0 4

r T

EEH S

m w

v e

v e

I C _.

..s

-.o S8

?SD perTthmetic 15.,

50. and 85.

-2 entiles

=

ar means On hazard curves using 5 & G experts

', 3 6"

.1

2. <n

^

t E

?$

zB b

.01 (g

o-a?

p e

0 E

.001 o

e e

t 1

o o

0 1

W ie-4

\\

N o

o T

m ie-5

+'

85th n

x n

n

.o 5

m ie-6 n

.o o

o L

N 0-1 I

1e-7 0

500 1000 N

acceleration em/sec

• 2 y

shearon harris EUS update, benchmark PGA 2

SHEET. 4 M > = 5, 1989 selsmicity and g m.

- ~

rm

$8

~) w t

I-a i,

1. a x

D percentiles 15.,

50. and 85.

-2

=

orithmetIc means On using 5 & G experts

'S bazard curves zm i

.1-S.n<

t 0

P'$

m zo b

.01 yg o-J. %

~

'F 0

0 E

.001 a

e e

m

.i o

a 0

2 W

1e-4 O

e

+

o n ie-5 T

+'

n 85th x

nean n

n

.o

'S o

le-6 m

.o o

o L

N l

b l

1e-7

'*i" i

0 500 1000 N

acceleration cm/ secs *2 g

t i

shearon harris i

EUS update, 1992 Seismicity, 1989 g.m. methodology, d

1989 g.m.model, M)=5 SHEET 4

l l

v

N-m

?Eo 15.,

50. and 85.

g3 l' u i ot per gen t i 1 es

=

orithmetic means

,a g hazard curves using 5 & G experts zg 1-S.n<

^

t b

m zs tn b

.01 28 a.

. 4 m

o E

.001 o

e T

r

\\

-a o

0 3"

N 1e-4 N

0

.o.

N te.

O T

m ie-5 k

mean 85th E

g o

le-6 m

o

.o 0

N Lo.

median A

le-7 3

0 500 1000 N

acceleration cm/sec**2 5

shearon harris N

EUS update,- 1989 Seismicity, PGA A

SHEET 4 1992 Experts' Attenuation Models

_ _ _ _ _.__. -, - _ -. ~, _ _ _ - _ _ _ _ _ _.. _. _.... _ -...... _. - -

.-.. - = =.

...I r

~

(

v t

-wn

-O

'; w -t 15.,

50. and 85.

ga pergentiles arithmetic means

.a n r

O l

hazard curves using 5 & 3 experts a

,m

.O 1-

2. n<

to 5

0

+

m 25 t

.01 Rn

x. o O

1 o-y a

v wr 0

9 E

.001 a

o o

\\

2 0

m.

0 3

le-4 y

N 5

0 N

o r

m ie-5 T

+'

n Xn

-n n

.O 85th 3

e 1e-6 m

.o o

o L

N 0-S mitan i

le-7 4

0 500 1000 N

acceleration cm/sec882 shearon harris s

EUS update, 1989 SelsmicIty, PGA sHECT 4

1992 Composite g.m.

accross experts

-.. /

(#

?8 2o I?tm< /]q

• 5 pergentiles 15.,

50. and 85.

ga

=

arithmetic means ano hazard curves using 5 L G experts

,g 1-9.n<

t b

m

?*

b

.01 gg 8

bf 0

0 E

.001 e

e T

u O

CD l

o

\\

W 1e-4

\\

N o

\\

B.

4-N l

o l

m ie-5 T

i n

l

~

x n

j n

5 1e-6 8se

.o ocan a

O

.L N:

o-S i

mitan a

le-7 4

0 500 1000 N

acceleration cm/sec8*2 g

shearon harris j

EUS update, 1992'Selsmicity, PGA 5

SHEET 4 1992 Experts' Attenuation Models

-, ~ _ _

- -. ~,. -..., _ -. -...

. - -. ~.

. _ -, ~.. _

G J

wn

-o

{&_- ' \\ );,

?8o 15.,

50. and 85.

-2 percenti1es

=

arithmetic means En hazard curves using 5 & 3 experts

,3

.1-S.n<

^

tc 5

o m

20 m

b

.01 gg a.

-- r-0 0

E

.001 G

c

?

m D

Q 0

\\

j N

1e-4 N

\\

0

\\

E.

w m

o T

m i e-5

+'

n 1-xn E

~

.o c

le-6 85th M

e

.o "N

o L

N o-S te-7

'" di "

0 500 1000 N

acceleration em/sec**2 q

shearon harris i

1 EUS update, 1992 Seismicity, PGA A

SHEET 4 1992 Composite g.m.

accross. experts L

= - - -.,

w w e--

w e-

-.-w.y aw w e e' e-

-v v v - w - vr,w,*e--vsw-

-.w-.r-+e--,-----,----e*,,.,

.c 3----,-.

-v m -w er e,-

e

-o-r-me->w-.x---,--

-+w,ee,

-m*w sur-,

-,-w,wwe,--sv--- +,,,,

+w.,,--.%,--

e. - -,,,

.,er-m-.i..--w-aeyen-an

- m i

.c e

.m mm

-O 2 n mn a

10000. year return period constant percentile spectra:

$n 15.,

50. and 85.

"., m3 pereentiles

=

o 1000r

2. m<

m

.m.-

'z a i

m

+

~

nn XO e a i

w a u

100 r p

b n

g o

\\

nuan o

85th e

E

/

h m

so o

N 10 F13dia3 -

~

0 3

[

?

w b

~

x n

?,

i r n

C n

e g

l 4

Ns oa 1

i I

I t t I t I i

R 1

1 it I i t

i iI t i t t e

10 N

.01 1

1 per1od ( s ec-)

g s hear o n. harris EUS update, benchmark PGV 2

SHEET.17 M > = 5, 1989 seismicity'and 9.m..

l

~

I l

6

-l. u.. t i

2m

^

$8

• Eo 1

2 10000. year return period constant percentile spectra:

$n 15., 50. and E5.

'E percentiles

=

2 to 10000r 9<

i g

• .~

2m

(

h8

{

e 1000 7 mm

~p l/

m. 4.,,

e e

s*

L.

N e

100 3

5 o

w o,

e.

~

n 5

10 E

g o

N A=

......i

......i 1

.1 1

10 100 m

frequency (hz) p-shearon harris B-EUS update, benchmark PSV E

sstti 17 M >= 5, 1989 seismicity and g.m.

_v wn

-o 30 e n D2 10000. year return period constant percentile spectra:

Eno percentiles 15.,

50, and 85.

a

=

,m O

10000 7

a. n<

n we 2o

\\

An e

XO 1

- a a

u t g

e 1000 r U

~F 1

n 85th e

+

c

~

0

-- n Ns s

edim ra

/

s

+

100r o

u f

.w W

U

~

'O d

i

-m n

i X

10 r n

n C

wn e

N s

N i

,,,,i o

,.,,,.i

,,,,,,i 1

e 4

.1 1

10-100 N

frequency (hz) y shearon harris N

EUS update, 1992 Seismicity, 1989 g.m. methodology, u

SHEET'17 1989 9.m.modei, M>=5

. -. ~._........ -_. _. __.. _.. _ _. _ _... _.. _ _.. - -. _..... _... _ _... _ _. _,... _ _.,.. - -..., _...... _ _..... _ _. _... _,.... _ _, _ _,,........ _..., _ _.. _ _ _.. _ _.. _. _, _ _., _. _,...,,

. _ =...-

.s l

de t.n n

~O 20 n

D i.

1 10000. gear return period c o n r,t a n t percentile spectre:

8n 15.,

50. and 85.

g f

percentiles

=

10000r

- 2. <

n i

,,n a-zo

\\

g wm Nn 6'

X. O2 ti 3

7 ~

BM

)

1000 _-

U n

o e

N O

/

__ miim 1,

e se t

N o

100 r 0

{

i

/

i-O g

O N

n 5

10 r n

1 C

n o

i N

NN

,....I

.,,,,1

,,1 m

1 e

.1 1

10 100 m

frequency (hz) g shearon harris i

EUS update, 1989 Seismicity, PSV Po SHECT 17 1992 Composite g.m.

aCCrOSS experts vr.-m.4-.

.-.-- -*.~-*, -., --

e..w..~~----.~.u..-.,---.-,v.-..-v<-v

.-.e

-v w

,,.. ~ -- -

--,----~~v.----+---,-

w e,.1

-,++---ew,--+,vv.-,-

v----,www.v..---w,--,.---,~---,~-w

+w.

.y t,n M

• • O soV i

3 2

10000, year return period constant percentile spectra:

yg l

15.,

50. and 85.

zg percent 11es

=

o 10000r

2. n<

n ms

~

zo o

\\

~

nn XO

{'

e 1000 r

~F M

U C

g man 85th y

3 y

mdIan s

l

. a0 L

N o

100 r

/

0 b

U W

0

/

b

~

N nX 10 r n

E n

C N

.sN

,,,,,1

,,,,,,i

,,,,,i

?

1 e

.1 1

10 100 N

frequency (hr) shearon harr.is

?x EUS update, 1992 Seismicity, PSV o

sHtti 17 1992 Compostte.g.m. accross. experts O

b

. _ -. _... _.,. _, _... _.. -.. _.. _. _.. _ _ _ ~ - _,.. - _ - _ _.

. -.. ~..,.. _ _ - -.,.., _.. -. _, - -.. _ - -. -., -,.,, _ - _ - _ _.. _,. _.. _ _.... _...... ~,, - -,. -..

4 1

i l

EUS Update Study 1993 t

Conclusions j

1.

Seismicity models in the EUS 1989 Study were revisited by re-eliciting the S-s expens' opinions.

l Central tendencies were generally preserved Uncertainties in occurrence rates became smaller 2.

Ground motion modeling was updated. Uncertainty modeling will be finalized by June 93.

j Present uncertainty is similar to 1989 study.

4-ne< rence in PSV was larper at hieher cercentiles 85%, and mean.

Central tendency slightly lower for both PGA and PSV.

1 3.

Combined effect of Seismicity and g.m. update i

i Generally mean PGA reduced by factor of 8 to 100 Generally medians PGA only slightly changed (factor 1 to 5) j For PSV the overall tendency was to decrease the Uniform Hazard Spectra with some reduction in overall uncertainty (15-85th percentiles). Large l

1 reduction in means.

4 5

!i i

i 1

[

4

)

s 4

e l

i i

l I

1 J

i i

l 1

4

?

1 i

i i

COMPARISONS.BETWEEN 1989 LLNL, I

1992 L.LNL ANP EPRI 1

SEISMIC HAZARD ESTIMATES i

i I

t i

PIIYLLIS SOBEL 1

MARCH 9,1993 i

i t

i

't l

i i

.~. -

,e

SHEARON HARRIS COVPAR SON OF MEANS PROB OF EXCEED /YR 1.0E+00 1.0E-01 1.0E-02 e ~"#

s s

~+

1.0E-03

+.

h

'+.

1.0E-04

~+ +..

x s

\\

1.0E-05 s

j

,h-1.0E-06

's s s

1.0E-07 0

100 200 300 400 500 600 700 800 9001000 ACCELERATION - CM/SEC**2

- - EPRI ~+- LLNL - 89

• LLNL - 92

~

~

PILGRIM COVPATSO\\ OF VIEANS 9

PROB OF EXCEED /YR 1.0E+00 1.0E-01

+

1.0E-02 i

+-

\\

A

\\

~

• p

~-

a._

< nc no a..

T.

~

1.0E-04

- - - +

s

's s

1.0E-05

's

~

~

s N

t 1.0E-06 1.0E-07 O

100 200 300 400 500 600 700 800 9001000 1

ACCELERATION - CM/SEC**2

- - EPRI

+- LLNL - 89

• LLNL - 92 1

4 -

g g g

I b

1 4 '

a h

+

I f

I O

O O

t

7...........

3 C')

t O

E o

...............n....._

O

?-

N e

1 k

v i

.O o

E i

g o

3 u

j

......d I

t O

....e 8

9

.e

..e

.e

......... e..

e g

..e g

ev=

m, hh.

o

%imsP 1

m I

I i

o o

o o

oO O

o o

o o

LO o

10 N

5-=

v-l h

CRYSTAL RIVER - 10,000 YR RETURN PERIOD COMPARISON OF 85TH PERCENTILES SA (cm/s/s) 2000 1500 1000

....g

... v...... : s..

1

.t 500 y,....,...

f s

.... g~u' p' ' ',

x.

c 0

0.1 1

10

- 100 1000 Frequency- (Hz)

LLNL - 89 + LLNL - 92

• EPRI

• 5 o

ce d e er ese new 1.ev=

g--

e.-

-y e s-

• re.%wt==i-w i6-.6

,, - + =.mawe

-cs-->-

+-M-.e toi++

e e

-'-=s=

TMwi s

e-E w-

'-=**ee-w eiemad.mm e+

9e m

+e

• -es.w--

.te is-ee-rrs'er-epr-

• 'e'a J-eh-2T-e-a-,

e-+e.--eWw+mm-%-ew%e-ay.,,,%-w..,e,ie.e-ys

BYRON - 10 000 YR RETURN PERIOD s

COMPARISON OF 85TH PERCENTILES SA (cm/s/s) 2000 1500 1000 500

-4 _.

,s-

/

.,..s,.

.....1

..... r][

s

....el

..l

..*.l 0

0.1 1

10 100 1000 Frequency (Hz)

" LLNL - 89 + LLNL - 92

• EPRI i

SHEARON HARRIS - 10,000 YR RETURN PERIOD COMPARISON OF 85TH PERCENTILE!S SA (cm/s/s) 2000 1500

, ~..-

1000 500

.._t_.....

1 1..

1 }-

.I.

w.

.m

.n.

.s.

.....z

.....f

...1

.....l

...t

\\

0.1 1

10 100 1000 1

I i

Frequency (Hz)

LLNL - 89 + LLNL - 92

• EPRI

~"~

__-~..____.-_--..._-_-.,..._.._---.-_s.

.... -.. ~..

-_.,-..-...~,,..,_..._,___._.-...-,.-,.,._.._.-_.__!

VOGTLE - 10,000 YR RETURN PERIOD COMPARISON OF 85TH PERCENTILES SA (cm/s/s) 2000

~.'s*r~.'.*.%.-

.~.%s.

c '. '..

1 s.'1..

• .*..t s1 s, t

~'**.

1000 t

• * *

• i *.*.*,. t I

T

• * * * ' **=.....

... s f..

500

=

. * = * *..

...y

.N N.

g.

....t,

<Ts.

..., s 0.1 1

10 100 1000 Frequency (Hz)

LLNL - 89 + LLNL - 92

• EPRI

~'~

_____...m,-

.w

.,--...--,-.......--....w,,

.---.-,------.....--,..,----,,,-.+.-,--we,..-i~,~~.---~m.m.

.~,~

m,+,..

v,

,--.-.m...-.

PILGRIM - 10,000 YR RETURN PERIOD COMPARISON OF 85TH PERCENTILES SA (cm/s/s) 2000 1500 z.

1000

-a.*

.* ;..i..

i..

... ~ ".

f....

s'f...

,./

3

. <s.

.1 500

.... W<+s

....I

....I

....I

....I 0

0.1 1

10 100 1000 Frequency (Hz)

-"~ LLNL - 89 + LLNL - 92

• EPRI

.,-,---,,._,.,...._,mm,._.-__.-,__,.,.--..m.-

-____,.-__.m.__.m..

SEABROOK - 10,000 YR RETURN PERIOD COMPARISON OF 85TH PERCENTILES SA (cm/s/s) 2000 1500

.c '.. -

1000

....j;........

,, /..

m.

_.s...

500

._ g.

s.

..,s l

.e e

a

..e 0.1 1

10 100 1000 Frequency (Hz)

~~ LLNL - 89 + LLNL - 92

• EPRI

- -