ML20128D287

Forwards Util Forwarding Addl Info Resulting from 830708 Meeting at Structural Mechanics Assoc,Inc Re Severe Accident Risk Assessment

ML20128D287
Person / Time
Site:	05000000, Limerick
Issue date:	07/29/1983
From:	Martin R NRC
To:	Nilesh Chokshi, Ibrahim A, Reiter L NRC
Shared Package
ML19292B772	List: ML19292B771 ML19292F169 ML20127A269 ML20127A345 ML20127A390 ML20127A443 ML20127A464 ML20127A814 ML20127A820 ML20127A923 ML20127A943 ML20127A959 ML20127B042 ML20127B062 ML20127B403 ML20127B689 ML20127B713 ML20127B794 ML20127B884 ML20128B972 ML20128C022 ML20128C044 ML20128C071 ML20128C089 ML20128C105 ML20128C120 ML20128C138 ML20128C168 ML20128C203 ML20128C270 ML20128C286 ML20128C294 ML20128C313 ML20128C327 ML20128C337 ML20128C366 ML20128C403 ML20128C503 ML20128C842 ML20128C867 ML20128C938 ML20128C976 ML20128D257 ML20128D287 ML20128D427 ML20128D470 ML20128D479 ML20128D514 ML20128D662 ML20128D680 ... further results
References
FOIA-84-624 NUDOCS 8505280552
Download: ML20128D287 (1)

Text

- - -

Yo: L iz e,+er A Ebal.*~

T zek)

N CLorsh 3

Peau.ny j2. u p

Le s E

cht/. d fr>>: A lh a.rt:~

nTW y D ar

>% l 4 r nll o n.

it slowlo'

.coi ~ h d.th. u d sy piac k

g 52 g 2 841015 SHOLLYB4-624 PDR n.-.

..n.,, - - -,-- <,

6 s.

PHILAD ELPHIA ELECTRIC COM PANY

\\

9 2301 MARKET STREE7 P.O. BOX 8699 PHILADELPHIA. PA.19101 33w A RD S. B A W ER. J R-

' lll.*:::.".'....

12 51341-4000 SWGENEJ SRADLEY

...Au..uA....

July 27, 1983

.woocaw4.ca>6Lews E C. MIRK M A bb k.

, M M AMgm CORMELL PAhlk A W SM S ACM so. Amo s cuccan.s.

Docket Nos. 50-352 1 o.. a. u,66s a. 4 a.

entMt A. M.N EMM A S M S3 Mr. A. Schwencer, Chief Division of Licensing U. S. Nuclear Regulatory Commission Washington, D. C.

20555 Subj ec t :

Limerick Generating Station, Units 1 & 2 Raquest for Additional Information har Mr. Schwencer:

I Inc. on the Severe Accident Risk Assessment,As a result of the July 8,1983 m additional informa tion was provided to the NRC's consultant (Jack Benjamin and Associates, Inc.).

It consisted of tie following enclosed information:

1)

Average Magnitudes and Distances for Seismic Ha ard 2)

Coordinates for Piedmont Zone 3)

Structural Mechanics Associates calculations of Earthquake Component Factors 4)

" Comments on Effective Ground Acceleration Estimates" by R. P. Kennedy 5)

Letter frorn R. Campbell (SMA) to J. Reed (JB&A) da ted July 11, 1983.

Very truly your Enclosures 3d {2 Copy to:

See Attached Service List C, v,. CM)W"

9%

cc:

Judge Lawrence Brenner (w/ enclosure)

Judge Richard F. Cole (w/ enclosure)

Judge Peter A. Morris (w/ enclosure)

Troy B. Conner, Jr., Esq.

(w/ enclosure)

Ann P. Hodgdon (w/o enclosure)

Mr. Frank R. Romano (w/o enclosure)

Mr. Robert L. Anthony (w/o onclosure)

Mr. Marvin I. Lewis (w/o enclosure)

Judith A. Dorsey, Esq.

(w/ enclosure)

Charles W. Elliott, Esq.

(w/ enclosure)

Jacqueline I. Ruttenberg (w/o enclosure)

Thomas Y. Au, Esq.

(w/o enclosure)

Mr. Thomas Gerusky (w/o enclosure)

Director, Pennsylvania Emergency (w/o enclosure)

Management Agency Mr. Steven P. Hershey (w/o enclosure)

Angus Love, Esq.

(w/o enclosure)

Mr. Joseph H. White, III (w/o enclosure)

David Wersan, Esq.

(w/o enclosure)

Robert J. Sugarman, Esq.

(w/o enclosure)

Martha W. Bush, Esq.

(w/o enclosure)

Spence W. Perry, Esq.

(w/o enclosure)

Atomic Safety & Licensing Appeal Board (w/o enclosure)

Atomic Safety & Licensing Board Panel (w/o enclosure)

Docket & Service Section (w/o enclosure)

Robert K. Weatherwax (w/ enclosure) l

.e l

l l

r---

ENCLOSURE 1 LIMERICK SEISMIC HAZARD STUDY AVERACE MAGNITUDES AND DISTANCES Feak Acceleration, g Seismogenic

.J.

Zone ab. man 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.10 5.8 m

5.2 5.3 5.3 5.4 5.4 5.4 Fledeont

-=

R 26.

19.

16.

15.

14.

13.

6.3 a

5.4 5.5 5.6 5.7 5.8 5.8 5.8 5.9 5.9 R

31.

24.

20.

18.

17.

16.

15.

14.

14 5.0**

5.1**

5.0 a

Tectonic Zones R

25.

21.

5.5 a

4.9 4.9 4.9 4.9 Crustal R

40, 24 17.

14.

Blocks 6.0 a

5.3 5.3 5.3 5.3 5.3 5.2 5.2 R

51.

42.

36.

32.

29.

27.

26.

6.8 m

5.7 5.9 5.9 6.0 6.1 6.2 6.2 6.2 6.3 6.3 Decollement R

55.

40.

32.

28.

25, 22.

21.

19.

18.

17.

a le everage body-wave magnitude, R is everase distance in km.

The average magnitude has cointributions f rom an adjacent sone with "b, max = 6.5.

1 e

e

ENCLOSURE 2 LIMERICK SEISMIC HAZARD STLTY COORDINATES FOR PIEDMONT ZONE-Long. @)

Lat.(N)

Long. N)

Lat.(N)

B2.00 ~

36.60 77.30 36.50 80.80 36.80 77.50 34.90 79.40 37.40 77.40 37.70 78.70 38.20 77.40 38.30 77.60 39.50 77.30 39.,00 77.60 39.80 74.40 39.50 76 40 40.30 76 30 39.50 75.60 40.50 75.30 39.80 74.00 41.50 73.60 41.00 73.30 43.00 73.10 41.10 73.20 44.60 72.00 41.20 72.40 46.10 49.60 41.20 70.30 47.30' 66.80 45.10 69.60 47.90 67.80 47.10 S

L

I CF f:b No Ym.s STRUCTURAL Pase DArgtoiI4iBf.

mECHAnlCS couunnis BY

""""""" ASSOCIATES CHKO.3Y Daft I

I "N

a cene. c....

J ENCLOSURE 3

~

.... :.fh.plei.. 5 on.y ortent Facto es.

C Ear l

TA<

~wo?if; &Mri.AchtaL pluo ve S c d daechon.

A lsoA~6e... % t :_.

-fo.r_._..:...

. r e y o n o $....{. e c. ] e g...._. g

. cz s@6u-ND,. ~ana. ~ :.eMme.~t-FS A12 fu..

aze.

....z..

SecWm i. S. 7 2.r6:

and.-

......u..

._.L i_...

ccm he 4.eck.

3ke.

k a n: : raponae

.n.e.pm med.; _.-

s.

cc,oI:,i;, f " ~ 4.:.HG,.' &e.

k r,z o,a K r [ ~o a a.'i: q f4

'4 a.*/ ~

m s.

.,%a or:M~ak. :L:LaMzo,,xd r

ois e. 2. & _.: 4 c d 2.

.. _ r...

cu td. Jbli. ;i97 e). -

a. :l g / c cc h r s,

,,oe. ( n e.o n m ie

~

^

r Cois.gctl....

..c.

. 1 T h&

l@erick;...d.

rg; iou.g O

CMen.o n:..-- e-nlL. - Mo&---.l2.-

yeo.

- - - - z -- >- h;".

l a.

n oe'~

T.'

P.'~.~~

,...).

L Ak (C7 OM Y M

c n..Ac

' oric~c cdo~ ~ckd.'r4ph l Cf W,hf5.che-cla ~

chet.ccuceds$ics:

of - - 17 c ep.aftvrw cmeder - (_s i lsred k v --

7E.L Ecc 'Fac+ ors ctnd f

VNu6 O

l v

Descripl&n F

l.

Ecc E cc !..

l B

' 3.D G e r re reti

'I.04 0.12-1

.I 2D Horizor,fd GenercJ o.e2 a.'os 2

Horizonlal. No conf l*y 0 9 3 -

3 2D o 06 4

i H o ez +- Vert. Gen erd 124 0 06 l

0.03 5

1 Hor-z +

verb, No ceup i9 1.16 0

G Urd ctxia.I Response onl I.00 0.

y Reference.

Ne w m cu.u., a. M., ct.nd

.Ha.II, d.T., 'De velopme n t of

~

of uleclect nu clear power P "^Is "

I C6teria.

for se.csm c re v:ew tJUQE G { CR - 0099, M a.q.

1 78 9

I'"'

OF s:b No f

gg STRUCTURAL PAGE my cAva-1 I

mECHROICS couusnis

""""""" RSSO.C..I,ATES CHKO.eV DATE-1 I

wm a e..

O Cm 1

3D G e n e r a.I

.eesponse X

Y 2

Vecfor g

g g

O'3 D I..

. X*...

f J..

j tool, Cov.f thg Y

0. co7 0.67 0.67 gq I 67 1.47 t 67 2.89 Meckan EQ.

I I

I 100/, Co q ling, E.

_4..

...._4

!ao*/.,40f.>40f.,:

Eq: l'O40 o.40 0 40

_ 3.*

^:

l EQ_

o.27_

0.27 O. 2 7 t

ea,

r. a,

i. s,

2. h s. - - - - -- -

O

.Duign

.Y I

ga.

Ab Coqol'hg ' ~

• o.67 1

c. 67 12o MedNa,w x

1 EQ ao. coapling

'Eq o,40 EQ O.27 1

o.27 a 40

f. r 1 0

Best

1. Q 1 08 i

=

Meclicln I.1 I p yg, F

t.04 Ecc Wo v.S h r D00L4 0 2.89_

f.00 s

~

MC CliC\\ ri 2.69

3

' TiTLa STRUCTURAL PAos cr

-tb &

sY oars I

mECHAnlCS couvents

==""" ASSOCI ATES CHNO.sY DATE I

I WM a casee. c..

O e.

x a p v s cs.

c w a s,- -- ca ceau a.

'j Cowp />d Ex&anz e r

~

V8 dor

..........-.._.4 h

• R

/O o /.

In. /b Ra. -

so

~ 1 7

y EEQ-1 1

j t

..._..y

. g,g i..

l.

2.67' 247 2 (7

( 0 ) 2 f*7 g.

.y.

r-1.

I~.

I

i.,,

5_......

__.-,............._L....

.......L_..._--...y..

.- I... _ -.

l'

- itoo/;

.0M g.

' Xg g

phw

.' a:

.o o

.:p.

. ;2..

eq I

f O

.._ ;e..

._._;_..._..._..7r..yEQ

... O O-O

[3.

r-

'... ".,I '

~~~ j -

'l I.

.1

. t.

,a

..)

O (9. 67)

t 4

n

=

0 1 Co

f. '.

M.

g

^

gy3 I

P 4

to o'l.' Inf oot.-

l Y

k ea L

4, f

i I

r Y

eo 0.67 i

i t

0.67

7. SG a..

S e.

Q I00-l. %c o}

y y

l Pl,w a

O

'EG YEQ f

O

4 Tiv'LE g

STRUCTURRL PAGE CF J:b No mECHRnlCS couuamis EY DATE I

f

" RSS,O,,C,,I,RTES CHKO.EY DATE-1 I

wm

.O b(

/f )

c.o74 G

0 !b

• 0 O'd 0 12 -

A.vf.

j'..-

Cm a

z a s en. HoeizoMTA L R E S POM S E OU L 9' X

3 Ve c/ov-R R

.Deez.qs N

loof Cov.ffd X

1 I

1 414 gq_

b j

yecl. Law ygq t

/ Co 4,llcl^...

E 04 04 Ioo.

Q yao y,.

4a.,

gq 14

1. 4 1 98 i

i then-d-

y j

1 00 gg No. coLpLy Y

I VIL cl L ct.g EQ l

No kg oi,3 gq 04 c

I o.4 108 Woii[

De*9^

1.414.

o.y14 z

htedian I.96 O&>t DMp

t. oo -

c.928 Hection 1.08 O

o,714 + 0.9 28 o.s2 AvO F

scc 2

5 PA88 CF

-1:b peo Ta7ts g STRUCTURAL BY DATE I

I mECHRniCS couwenis RS,SOCIRTES cHuo. sy oAva f

I m

Cowpled.

Ex bre rnes y

Z vecitre g

a In Ph u

• e a 1

I 2/3 2

%s of Phw X

1

~

ga Z*

O o

E T

,fi 1

L 2 fi/c ) =

0. f15 h _,

s 0

f y

z, Veum-In PNace-v ea 1

2'9 L

i rz 1

I Od o} Pb%

I

~

EQ

~

a z%

E 1

O I

h:

Am l!

0.058 6

O.115

• 0.05 8 o.09 A

B Y

I ECC n

g O

' E8 6

T8 or Job No g STRUCTURRL rass av Dars i

I mECHR01(S COMMENTS

""""""" RSSOCIRTES CHKO.SY DATE-I I

WM a cene.

c....

O Cm 3 ~

2o Horizonted Response. Alo

% p l' hF E.fe r

!c.

Cu D..

vea;$;

^ ^ ite&dc '

'. co

~

f Meclian Vec hr t.0 8 G e.sf once.

fachrr i

'b. o e = 0 93 A br,o! fe worst Wche :

6 f_. '..

p.2 M S//......

o,o6 2...

g cm 4>

2o one Hore'zon/er/.

Plux Wrhcd Gewal %

Q

. )t.

.y ye ck,-

g g

'X f

I-

DQn.

Sq

?

o. roh o.67 lbo*/.

Conf 'h3 Yea

f. 6 7 1.67 2%

'N" Ygq 1

Cotrylm$

YEQ 0.27 0.27 100'l.

I 80

/.2 7

f. 27 l

D4gm Yeo f

d MO'Ch*f'dg y

o.67 l

gg 1.20 t

a. 6 7 Media ~

v 1

ea 1

Alo cov-pliAg U

Ye a 0.z7 l

I.O 4 p

G I

O 27 l

Besl :

I'30

t. 3 2

f. 60 Avg f

1,24 Ecc Worsf s 1.20 1,16 o->

T'T'8 7

CF J:b No-

~gg STRUCTURRL PAos eV oats i

I mECHR0lCS couusnis

""""""" RSS,O,C,,I,ATES

~

CHMD.BY DATE I

I wem O

Conpl.ed

& f re>rt e -

Vec &

y*

y*

100 -/.

ln. Ph m Y

1 1

eQ 8

Y O.67 0.47 gg 167

/ 67 2 35 foo /.

0+t of fbu Y

1 1

ga

~

Y c

'o gq

!. 4 14 y

y t

1

...... l3 - N [

' #lt 4 I4 G

=

0085 Gxbewte.1 l.

% cokpb.d 0

.I.

V Ya g

loo J (h '(fnx y,g

)

i Y

0.67 Ea 120 y

o g, 100 '/-

OkCofPbkx V;q y

o

~

Y O

O ca 1

o 1

L i l'20l1]lc,

=

0.03t o. o e s + 0. 03 /

M g

,,,3 I

a O

8

~

sob N f

TiTLa STRUCTURAL p..es CF sY oats i

I mECHROICS couusnis

"""""" ASSOCI ATES CHKD.SV DATE I

I wm a c.ne.

c....

e O

i Hor zon{cd

+ mlical,

th coup ag Caae s c

Qefe.c la Cu 4

F

'l

=

1. t co R

t04 For eeire,rtes ja =

0.03

. i..

e O

i i

t l

I I

e 4

l l

O

,___..._r____.._

+-

E arthqua.ke Compone,tt Cebth o.b5 ygygg F49'h9 d Lb.rwterne 9

OF Job Na-g STRUCTURAL PAGE

~

MkR oar,12 f f iaf i

i mECHA0KS couvents sv

- ASSOCIATES O *""

DESIGM PR O CE DURE SRSS Co M stM A'rlo M OF R.ES PO M SES Caos 1

so s en ceal cu V: o.6,H

~~

~

Vec W D

100 }, Co4lrrrg Yga O.47 a 6, 447 Sqgg 4 y

4 1

. co m pm d.

ega

~

I 56

/. 5 6 15G 2 71 M ecli %

y Yg Z

R e

. f***l (0 4 l'^$..

v, g

YGa o.17 0.27 0.27

^'

EQ 0-40 0 40 0.40 Q

... j, $7

. q, g y.

g,_ gy 2, g g_

Y I

l 60.

Mo cog i yea 0 67 Z ga, g

O.69 1

1.56

~

Medid UC Cowp(g No YEQ O.27 E

o.40 gg l

o.22 o 40 t.t I D M ^ b4 eel, u =

'2 '!/g y 0.94 J

ae.c

,. o,,

• AO' O

M e d.i %

I. t(

O.94

& I. 4 0 f

0.vp o}

s l 17

=

Ea Corn pon eat Comb factheu SR ss F'coce du re to TITLE STRUCTURAL paes or fob No-RkR f2 rri Sf mECHAnlCS couments i

sY Dave

""""""" ASSOCI ATES CHKO.sY DATE-I I

WM a cene. cw Cornpuke f

freyn.

2x bre m es

/%u rn.c e 3 p.grr- '5;lre>ra-see Pye 3.

In phu ve c fo,., (73)(a.s,y

~

h of P M vedor =

r3 04 6 (2.67) o.ta p

a n -pa

/ 56

/npk u Vecivv p-L (! s') -

o. o,4 o

^

h = (arg of 0.16 o.074 )

o. 1 1. ' ' ~

^ ^ ^ ^ ~ ' * * " ~ ~

e h

O

~

.. n.

EQ Corne. C<nnhn.

SRss Proce du re 11 787 W STRUCTURAL caos or fob No-MkR-oavat2 e tt i B' mECHRniCS couvents sv

""""" AS.S.O.Cl.ATES cano.ay ears I

I wnummuminum Cctoe a 2.

H oriz on t a.L Aeopo w s O k,

G e ntr-n k-Co m Y

E Ve ch:rr-a g

DeQn Xect 1

I loo [ Cougedy ---

%q ~~

" - f - -- -

I G.

B.

A Y

4 i

EA l

. 100'I.. Cowf la'tg. -..

ag.--

o,4 o.s 2.

l. 4 I,4 D%^

X f

eq cogloy.-.

.. --. - : 3... - -.-

^

.No 1

i MLdik Xgg f

.NQ Coht

-.E O,4...

gg.

f-o, 4 1.06 i

tr/orsb k

?z 10 W

I* ll I4',

t, y M_e ua n, _

Meda.hn

,, g g 2

/

1 + t. 31,

p. tS f

ECC 2

h 3 p valu u

%pn pos.1 cole aw,n eo coa A

=

Page s.

ca.

Co u pwl.

I,, ph u veem = a m g,

od o} ph m vedv =

4 (o

WkkpM ln ph.u vedor s V5 g, o.ose aseypAu

yecs, i

g..

o.13.+o.058 e

0.09.

O 2

m ECC Stess' F'rocedure 12 787L' oF Job No g STRUCTURAL PAos av oAve /2 :If 8f mECHAnlCS couusare

""""" ASSOCI ATES CHuo. sY DATE-I I

WM a cene. c..

b Cm 3

2D ho r r'z hio CosP !!'9 pp to c-2

!. 31 f = 0'O6 f

ECC Ccroe 4 On.e.

H+V Ge n e reu'.

x y

vechrr-g y

96 X

1

(

cq D^$

Y O.61 0.67 I ' 'O ga d

i20

/. n o X

-. pg g m

i Coqbd..

Y gg O. 2. 7

0. 2.7 f.a, y I. 2 7 I 79 6 Y

1 ga U^Coufld Ygg o.67 i

1.20 I

o,67 Nech ~v V

t yg

• W Ygq o.1.,

1. • O30 f

o.27 worsf D M.9 n l 7o g_g4y Medice n *

7 9 6 B n/

P"La n I. c o 1, p a g Meck cen

1. o.1 4 Y

0 9 4 7 + 1.15 8

f. o s,

p Ecc a

Compde p W exlre m a Cara cG r-4 3 l2 'u ^2 a. (P 7 )

Co*PM

/4 ghee vedoc 236

& 2 3 y,,,,,,,

I Ont0}

P'"

hclor-1 I d'a lln conpled.

p,

L L?* ', s o n.

in pA u ve ao,-

i. aos Osa o} phaos 1

A.

o.095 + 0 03' n ns i

Ecc 5RSS Procedure 73

~

g g STRUCTURAL paea oF Job No-av oars 12 If: Bf R

mECHAnlCS couusmis

""""""" ASSOCI ATES CHKO.SV DATE-I I

WM a cene. c..

veektal

- Alo c p/,9 Cm s One Horizo,M

+

ou kg

/0 Ch 4-flnCohfM v

l 15 8 g

1.16 f

a Ecc f:

0 03 c~ c un mag eg, _

V

[.O i

.C.

T*

a ecc O

!. O e

.---c.---

,--r--,,-w

---

,wg.------r-----,,..,---%.

,-,y- - - - - - - - - - -_

4 l

i ENCLOSURE 4 COMMENTS ON EFFECTIVE GROUND ACCELERATION ESTIPMTES by R. P. Kennedy February, 1981 1.

INTRODUCTION Structural Mechanics Associates (SMA) has estimated fragility curves defining the frequencies of seismic induced failures for structures and components in nuclear generating stations. These frequency of failure estimates are defined as a function of the effective peak ground acceler-ation level (EPA). The EPA is the ground acceleration level at which a broad frequency content structural response spectrum should be anchored for the purposes of predicting structural damage. For predicting structure and component damage at sites, SMA has used the median broad frequency content structural response spectrun for appropriate ground media as defined in Reference 1.

SMA has assumed 3 to 5 near-peak response excursions approaching the levels defined by this structural response spectrum anchored to the EPA.

This approach is most applicable when dealing with longer duration ground motions which contain a broad range of frequency contents such as the Taf t recording from the 1952 Kern County, California earthquake (local magnitude Mt = 7.2, range from causative f ault = 40 lan) or the highway test laboratory recording from the 1949 Olynpla, Washington earthquake (Mt = 7.0, range = 29 km). For such earthquakes, the EPA to which a broad frequency content response spectrum is anchored and the instrumental peak acceleration (IPA) should be essentially the same.

Such records result in 3 to 5 structural response excursions approaching the levels defined by the structural response spectrum.

1

l However, the frequency of strong ground motion in areas of low seismicity will be mostly due to lower magnitude earthquakes (ML $5.7) and shorter ranges (less than 20 km). Ground motion ~s from such earth-quakes have characteristics like those recorded at the Gilroy Array from the 1979 Coyote Lake, California earthquake (Mt = 5.7, range = 7 km) or the Helendy Ranch Barn record from the 1972 Bear Valley, California earth-quake (Mt=4.7, range =6km). These records have narrow frequency content, and within the majority of the frequency range of interest (2 to 10 Hz) their structural response spectra is seriously overpredicted by the Reference 1 broad frequency content median spectra when this spectrum is anchored at the IPA. Secondly, only a single cycle of strong structural response occurs from these records because of their limited du;ation and energy content. Thus, for these records the IPA cannot be used as a basis for predicting the level of structural response which is approached 3 to 5 times during an earthquake.

Since structural damage predominately depends upon multiple cycles of strong response, the IPA cannot serve as a good indicator of structural damage for these earthquakes.

In these cases, damage is better described by an EPA which is much less than the IPA. This EPA represents the ground acceleration for an equivalent long duration record with broad frequency content which causes 3 to 5 response cycles and results in the same expected damage level as the actual record.

2.

EFFECTIVE PEAK VERSUS INTRLNENTAL PEAK AND SUSTAINED PEAK ACCELERATIONS SMA is currently engaged with Woodward-Clyde Consultants in a research propam sponsored by the Nuclear Regulatory Comission (EC) to define effective ground motion parameters useful in predicting structural damage. Tuis section briefly sumarizes some of the tentative findings to date.

For the purpose of predicting clastic response of structures in the amplified acceleration frequency range (2 to 10 Hz), median broad frequency content response spectra such as those from Reference 1 are more accurately anchored to an EPA defined by:

2

. - =

{

[

A 1.25

• A3F

.(1)

=

E rather than the IPA. The quahtity A3p represents the third-highest acceleration peak of a filtered acceleration time-history record. The filter should be chosen to pass all frequency content below about 8 Hz and filter-out all frequency content above about 9 Hz. The quantity A3F corresponds closely with what Nutt11 (Reference 2) has defined as sustained peak acceleration.

Therefore, Equation 1 together with Nutt11's definition of sus-tained acceleration can be used to define an EPA (A ) to be used to E

estimate elastic response of a structure within the 2 to 10 Hz frequency range. However, elastic response is not a good measure of damage.

Based upon current work, two ground motion time histories with the same spectral acceleration values at the structure's natural frequency can lead to vastly different nonlinear response or damage for the same structure model. For instance, one can compare (Reference 3) the maximum nonlinear response (damage) from the Melendy Ranch Barn record (Magnitude 4.7) with that computed from the Taf t record (Magnitude 7.2).

It is found that the Melendy Ranch Barn record must be scaled to produce spectral accelerations between 1 and 2 g's at the structural natural frequency to produce the "same level of damage as a 0.59 spectral acceleration from the Taf t record.

Thus, for Melendy Ranch, the spectral acceleration must be two to four times as great as for Taf t to produce the same level of structural damage.

Similar conclusions are reached for the Coyote Lake records (Magnitude 5.7) versus the Olympia record (Magnitude 7.0) or Taft. Thus, for

^

obtaining in EPA corresponding to a given level of structural damage, Equation (1) should be modified, as follows:

1.25 AD F

• Ap (2) 3 The f actor F must be established as a function of ground motion character-Istics for a constant level of structural damage. For magnitudes greater than about 7.0 and ranges greater than about 40 km. F can be taken as 3

e F

unity and Equation (1) can be used to predict the EPA corresponding to structural damage. However, with magnitudes less than'.about 5 and ranges less than about 20 km. F should have a value greater than 2 for predicting structural damage. As a consequence, the EPA (A ) should range from D

less than 0.6 to 1.25 times the sustained ground acceleration (A37) depending upon the earthquake magnitude and hypocentral range with the lower factor being appropriate to low magnitudes (less than about 5.0) and short hypocentral ranges (less thali 20 km). The appropriate ratio of A /A3F is strongly influenced by the duration of strong ground motion.

0 The Sm fragility curves for nuclear generating stations are best anchored to the damage EPA defined by Equation (2). For sites where strong motion generally results from low magnitude earthquakes I would judge that the 90% confidence bounds on F are:

1.0 1 F 1 3.0 (3)

Because of the tentative nature of the research conducted to date and the controversy of the subject of EPA versus IPA, I would recomend that F be conservatively selected for use in Equation (2) and that the EPA be defined by:

AD at A p (4) 3 with Nutt11's sustained peak acceleration (Response 2) being used to define A3p, B

3.

UPPER BOUND CUTOFF ON EFFECTIVE PEAK ACCELERATION The EPA is being used as a measure of damage to structures with a fundamental natural frequency in the 2 to 10 Hz frequency range. The Modified Mercalli Intensity (Im) is also a measure of damage. Although I, is a subjective scale, it probably correlates best to damage of N==

4

r conventional structures which generally have natural frequencies in the 0.3 to 3 Hz range. Because both EPA and I, are measures of damage capability of ground motion, these two quantities should be closely cor-related with each other. Thus, one should be able to estabitsh upper bounds on the EPA irrespective of frequency of exceedance if upper bounds exist on intensity.

Table 1 describes the earthquake effects (damage) corresponding to each of the I scale levels. These damage descriptions can be used m

to define upper bounds on the EPA corresponding to a given I, level.

Masonry A construction corresponds to earthquake resistant masonry struc-tures designed to the Uniform Building Code in California (Zone 4).

Masonry B construction is reinforced and represents well-engineered masonry structures in UBC Zones 0 or 1.

Masenry C construction represents well-constructed unreinforced masonry structures. The SMA methodology used to develop the fragility curves for structures and components at sites with low seismicity will predict very substantial damage and/or at least partial collapse of 50% of these masonry structures for 3 to 5 cycles of the following effective peak ground acceleration (EPA) levels:

Masonry Type 50% Damage EPA Levels (g's)

C 0.25 - 0.3 8

0.4 - 0.5 A

0.6 - 0.8 Thus, very serious damage to a large number of Masonry A, B, and C structures would be predicted by the SMA methodology to correspond to EPA levels of less than 0.8, 0.5, and 0.3 g's, respectively. The SMA methodology for predicting damage levels has been benchmarked against observed damage in past earthquakes in which substantial damage was observed for sustained ground motions corresponding to these levels.

5

r Based upon the damage descriptions in Table 1, serious damage to at l' east some Masonry A, B, and C construction correspond to I, levels X, IX, and VIII, respectively. Comparing the EPA levels defined above for each of these levels of damage, one would estimate that I, of X would correspond to an EPA of 0.6 to 0.8 g's or less, I, of IX corresponds to an EPA of 0.4 to 0.5 g's or less, and I, of VIII corresponds to an EPA of 0.25 to 0.3 g's or less based upon the described damage to masonry construction. Even considering unce'rt'ainty in the correlation between the two descriptors of damage (I, and EPA), an upper band must exist on EPA for a given I, level. Otherwise, the two indicators of damage would be contradictory.

In my judgement, an upper bound on EPA can be estimated by assigning the EPA ground motion levels defined above to an intensity value one level lower than that for which a given type of masonry construction damage is considered appropriate. Thus:

Intensity, I, UpperBoundEPA(g's)

IX 0.8 VIII 0.5 VII 0.3 VI 0.2 In my judgement, the EPA values given in this table represent conservative upper bounds for the corresponding intensity levels. These EPA levels would result in the prediction of substantia 11y more damage than that from which the intensity level is defined.

If upper bound intensity levels, I

, are defined for a site, then the EPA levels should also be limited to being below the upper bound levels defined above when the SMA fragility curves are used to predict structure and component damage. Unless these limits to EPA are applied, one would predict substantially more damage than could possibly corres-pond to a given upper bound intensity level. Even with these Ifmits, I judge that the level of predicted damage would correspond to at least one intensity level higher than the upper bound intensity level.

6

r

~

e REFERENCES 1.

"A Study of Vertical and Horizontal Earthquake Spectra", WASH 1255, Nathan M. Newmark Consulting Engineering Services, prepared for USAEC, April,1973.

2.

Nuttli, O. W.,

"The Relation of Sustained Maximum Ground Acceleration and Velocity to Earthquake Intensity and Magnitude", Paper S-73-1 published in State-of-the-Art for Assessing Earthquake Hazards in the United States, U. S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi,1979.

3.

Kennedy, R. P., Short, S.

A., Newmark, N. M., "The Response of A Nuclear Power Plant to Near-Field Moderate Magnitude Earthquakes",

Paper K 8/1, to be presented at Sixth International Conference on Structural Mechanics in Reactor Technology, Paris, France, August 17-21, 1981.

M 7

1-w:

s g g STRUCTURAL ENCLOSURE 5 mECHRnlCS

- RSSOCIATES SMA 14302.01 wm c...

c....

5160 Bach Street, Newport Beach, Cahf. 9266o (714) 833-7552 w

July 11, 1983 Dr. John Reed J. R. Benjamin and Associates, Inc.

Mountain Bay Plaza 444 Castro Street, Suite 501 Mountain View, California 94041

Dear John:

During the Limerick PRA review meeting on July 8,1983, you raised some questions which I could not answer off the top of my head. The following responses should clear up the issues.

The first question was regarding an apparent inconsistency between the response factors for piping and for valves. These response factors should be the same as valve response is predicated upon piping response.

Initially the piping response factors were developed on the assumptioli that 1% damping had been used in the original design analysis when in fact 0.5% was specified. The valve response factors were calculated using the correct damping of 0.5%. The text in the original report draft erroneously tried to explain the difference. A review of the work uncovered the error in the piping response factors and the cal-culations were changed. Unfortunately the text in the report was not properly changed. The response factors for piping on page 5-59 are low and should be the same as those for valves on page 5-60.

The descriptions on page 5-60 of how the valve response factor differs from the piping response factor should be deleted as there is no difference.

You raised another question regarding the deletion of SRV loading from the generic derivation of piping and pipe support fragilities. The SRV loading was purposely lef t out of the generic derivation to examine the more conservative case.

If the majority of the loading is seismic, the resulting capacity factor is lower than if the majority of the loading is from other sources. For instance:

=

c "n

1. SRV F

s "SSE

r 1

,a e' s l

Dr. John Reed J. R. Benjamin and Associates, Inc.

' July 11, 1983 Page two Q

where strength factor relative to the SSE F

=

s failure stress

=

oc stress from normal pressure and dead weight load o

=

n stress from SRV load

=

oSRV stress from SSE load

=

OSSE is a greaEer+por" tion of code allowable results inSRV Assuming the sum of ( OSSE + 0 the case where OSSE the lowest valve of Fs. Likewise for supports, we derive the capacity factor for supports which carry only seismic loading as they are the governing cases.

A third question was raised regarding variability in testing due to use of spectral test methods. The attached pages from NUREG-CR1706 describe the problem and attempt to estimate the uncertainty.

I hope this explanation will clear up these issues. If you have any further questions please don't hesitate to call.

I will be away for several weeks, but both Ravi Ravindra and Greg Hardy should be available to answer any further questions.

Very truly yours, STRUCTURAL MEC NICS ASSOCIATES, INC.

/

Robert D. Campbell Project Manager RDC:mw cc:

H. Hansell, P.E.Co.

=.

l 4.4.5 Variability Due to the Use of Soectral Testina Methods In Chapter 3 the use of spectral testing methods,versus a power spectral density approach was discussed. Smallwood in Reference 23, stated that it is important that a spectrum not be considered as a suffi-cient specification. His study of a large variety of synthesis methods concluded that a different method may produce significantly different results. SWRI, in Reference 22, performed an independent evaluation of test methods and concluded that with current spectr a methods, insuff-i cient excitation of equipment modes might occur. Neither reference attenpted to quantify the magnitude of the variability that might he

- experienced.

The problem stems from the fact that a synthesized time-history l

contains peaks and valleys in acceleration or force and that if a signif-icant valley in the time-history input corresponds to a predominant fre-quency of the equipment being tested, the response may be significantly less than the mathematically anticipated and required response.

A much better approach, as recomended in Reference 23, is to synthesize a time-history that corresponds to a power spectral density j(

which envelopes the RRS rather than make the direct step from the RRS to

[

the synthesized time-history. This approach tends to smooth out the li input time-history, resulting in less chance for an equipment mode to b

coincide with a significant valley. Most testing laboratories do not, however, have this capability.

Quantification of the variability that might result from use of spectral methods could be accomplished by analytical studies that com-pared responses to synthesized time-histories keyed to power spectral densities as opposed to those synthesized directly to match the RRS.

I l

4 1

l 4-38

1 The variability due to the use of spectral test methods is expected to be smaller than that due to instrument and coittrol error and acceleration time, history variation and is considered to be prinitpally a modelfing uncertainty in that most of the variability could be eliminated by using more sophisticated test equipment. Without conducting a study to quantify the variability, we would estimate.the i 26 range of response to be within i 255 equating to a S # "II # *I* 8 ~ 0.M W Su ~ 0.10.

C R

0 t

I i

\\-

3 1.

t,

20. Smith, P. D., S. Bumpus and 0. R. Maslenikov, "LLL/ DOR Seismic Conservatism Program Part VI, Response to Three Input Cogonents".

UCID-17959 (draft report). Lawrence Liverinore Laboratory, Ljvermore, California April, 1979.

21. USm c Regulatory Guide 1.60, " Design Response Spectra for Seismic Design of Nuclear Power Plants".

22. Kana. D. D., and R. W. LeBlanc, "An Evaluation of Seismic QualificationTestsforNuclearPowerPlantsEquipment"(draft) prepared for the USEC by SWRI, Contract No. AT (49-24)-0372.

23.

" Seminar on Understanding Digital Control and Analysis in Vibration Test Systems", sponsored by Goddard Space Flight Center, Jet Propulsion Laboratory and The Shock and Vibration Information Center, held at Goddard Space Flight Center on 17-18 June,1975 and at the JPL on 22-23 July, 1975.

1 em 4 45

-