ML20212M387
| ML20212M387 | |
| Person / Time | |
|---|---|
| Site: | 05000000 |
| Issue date: | 12/31/1977 |
| From: | Bonilla M, Mark R INTERIOR, DEPT. OF, GEOLOGICAL SURVEY |
| To: | |
| Shared Package | |
| ML20150F500 | List:
|
| References | |
| FOIA-86-391 77-614, NUDOCS 8608250470 | |
| Download: ML20212M387 (11) | |
Text
_ _ _.
p
. (.
Ull!TEDSTATE.b.
DEPAllTHCl:T Of Tilt IllTERIOR GEOLOGICAL SURVEY I.
I
!! EGRESS 10;; A :ALYSIS OF EAf TilQUAl'E li%illTUDE A :D SURFACE FAULT LEl'GT!! USil!G Tile 1970 DATA 0F !.DNILLA AliD CUCl:A!!All By R. K. Mark and M. G. Ponilla RheadW@l Prepared in coepera tion 1.'ith U. S. liuclear Regulatory Commission.
OPEli FILE REPORT 77-G14 This report is preliminary and lias not been edited or revit: ed for conformity trith Geclogical Survey standards and nomenclature.
llenlo l'arl:, California 1977 8608230470 860801 PDR FOIA PDR
~
HOUGH 86-391 3
o
(
(
f! EGRESS 10!! Ai!Al.Y5IS OF Eb"mQUA:'E S. lITUDE Af t0 SUP. FACE FAULT LEllGTil USitlG Tile 1970 DATA 0F C0flILLA Al D BUCllA!!Att By P.. I'. liark and H. G. Conilla Introduction.
The report of Conilla and Duchanan (1970) includes re-gressions of fault length on carthquake macn'itude that can be used to estimTte most probable length of surface rupture given earthquake acgni-tude. lhose regressions, hcwever, have sometimes been incorrcctly used to esticate :agnitude from' fault length, as pointed out by Mark (1977).
Using the data of Conilla and Buc!janan, this report gives regressions of cartinguahe magnitude on length of surface rupture tiiat can be correct);
used to estimate most probable nagnitude if the length of surface rupture is given.
It also gives the /egressions of length of rupture on cacaited that can be nred to estimate cost probable length of rupture given carth-quake magnitude.
In tabic 1 and figures 1-5 tha numbering and lcttering system used to designate fault geography and fault types is the same as in Conil1a and Buchanan (1970).
!! umbers 1-49 include surface ruptures that occurred in I.*0rth America and numbers 50-140 include ruptures culside of iorth America. The fault types are indicated by letters as follows:
A, narnal-slip faults; 0, reverse-slip faul ts; C, normal oblique-slip faults; 0, reverse oblique-slip faul ts; and E, strike-slip faults.
Use of the regree.sion lines.
The regrer.". ion of log length on magnittJe 9
~
1 Pow sto*f)I B -i t4-
(-
(
(Log L=a+bli) can be used to estimate the most probable rupture length given magnitude, and the regressica of magnii se (. log length (M=a 4b Log L) can be used to estimate the most probable magnitude given ru; ture length.
The estimtion or ' maximum magnitudes' for a given rupt:;re
~
length requires the use of one-sided confidence limits (Marl:, 1977).
P.crerences cited Bonilla, II. G., and Buchanan, J. H.,1970, Interim report on world wide historic surface faulting:
ll.S. Geol. Survey open-file rept.,
32 p.
14arl:,
P.. K.,1977, Applica tion of linear statistical models of carth ;ra.'.c magnitude versus faulL longth in estienting marim:a exp::tcble
- v. 4 p. 4 6 + - 4 6 c,, A ug t s t.
Geolot;y, 4n--in:es..
o en J
(
(~
Table i Regression analysis of magnitude - surface rupture length data from Bonilla and Duchanan (1970).
Log (L)a+b*M M=a i b, Log ( L)-
2 set n
r f
a b
s a
b s
1-49 20 0.372 10.64
-0.91 0.35 0.51 5.23 1.08 0."
50-140 33 0.217 8.57
-1.49 0.40 0.55 6.56 0.54 0.
1-140 53 0.257 17.62
-0.96 0.34 0.53 6.03 0.76 0.
A 14 0.175 2.55
-0.69 0.28 0.45 6.19 0.63 0.
D 7
0.003 0.01 not significant C
7 0.459 4.24
-2.81 0.61 0.38 6.08 0.75 0.
D 5
0.006 0.02 not significant E
20 0.48i 16.87
-1.08 0.39 0.52 4.96 1.24 0.'
A+C 21 0.279 7.37
-1.46 0.40 0.45 6.13 0.70 0.'
o B4D 12 0.033 0.34 not significant C4Dit 32 0.367 17.42
-1.24 0.40 0.55 5.62 0.93 0.'
EllB915W CiD 12 0.230 2.99
-2.79 0.59 0.57 6.62 0.39 0.
BiE 27 0.299 10.65
-0.71 0.32 0.56 5.71 0.94 0."
A+C4E 41 0.350 23.94
-1.20 0.39 0.49 5.56 0.99 0.7
~B+D4E 32 0.251 10.07
-0.81 0.32 0.60 5.98 0.78 0.E.
I!otes "n" is the number of cases.
2 "r " is the fraction of the variance expidined by the regression.
It ranger, fr-(no linear relationship) to 1 (perfect linear relationship).
"f" is a measure of statistical significance of the regression and is equal to r
((1-r2)(n_g)),
"L" is in kilometers.
"s" is the standard error of the estinite.
2 s
is equal to the rer.idual
.un of e.-
crrors about the regression line divided by the dogrees.of. freedom (i.e., n-2).
4 "a r *
(
~
=:-
-.(.
0.~
700
-t COO 500
/c 400
-g t
300 0
C' 200
.-o---
O r;
WORLDWlDE DATA e
U HU 100 7
O o 90 E
GO -~
g -----
70 b
l vf l...
co--
'O L
50 0 i O sa
-4cl
/
g, f
=
o 40 b, _.,
s 0
e v
o L'
(,
=
- 9. -
D t./
+
C-C fQ
-m
.e.
A.*
U g
O J
(6 O
L C
D 0
L-tO o
o O
9
$,/
I-
/
DI o
-g o
I f
o e
/
e 7
z U
G
/
-- /
J O
, /
e, t _
}uj D*
4
-. O -
-J N#
O
-5(*
^
3 A
X!
T
,O.,e/l 6-
~
2 w <'
- . /
\\*%
oq s.
o p
O V
\\
x.,
000
(
700 GOO 500 -
/
400 300 200
_. (:
G NORTH AMERICAN DATA e
U V
U 80 0 -
7 O 90 l
_lf-00 x
l 70 s.'
~
-l - - -
D CO
=
^- I/
4-,E O
< SD l' /
40-30
~-
O e /
y
=
f DH 0
D 20 C
U 3
U O<
L:.
C D
L3 l0 C
5 f"
.. /
~
g
=
._./
/
s 7
=
c
/
f O
7 U
J 5
-/
/
<e
- 0) >
O~.1 3
n-o
-p n'
d 9.-
y
.f J
g
.._ g,
<F
~-
yr.i P4 C
k$'
em
=
-(_
1 I__
.,00 700 I --
GOO 500 400 300 200 a
N OR MAL-SLIP FA ULT DATA tr W
W W
100
. ~
O 90 l
.a eo i
/
x l-l l
~
70
-I
/
l J
co
-~
=
n D
't l
l
/
l A
50 O
O
/
O
/
40 e,
/
W Le
.i o
a p
fL 3
...u..
m 20 j--oc
- -r- - -
W efe) o o
jN 4A
- *g C
0 M
LE l0
^
~ ~~
O v
N/
I c.
t o
.' I
/
0 7
/
l l_ _l
=
~~/-.--
\\
si l-- -
W G
/
r 5
/
)
to e,
t.0 -
O l-
/
a.
w
_,,o:.
g;'
t-o.
/
.Nf
.O f,l
=
s,.
+ - *.
-(
('
00")
.- 700
-./.
.[]..
COO 500 t
400
/
300 o
200 o
a NORMAL OBLIQUE-SLIP FAULT DATA e
W D
800 7
/
~
O 90 j_,
l s 80 1
I
/_/ -.I
/__ /_.i
=
70 l
b
/ /
oc
, //
2<
L;.
50
/'_.
2 o
Q 40 fI
~
30 c
3 h
C.
amers o
c 0
u 9.,
O b
C D
01 Lt.
10 0
9
/
0
/.~
H 0
7 l
'l
/
U C
5
/-- l
/
d 4_
6 p.
J 80 3
A O'
+
fo co 2
o-9
- y..j,
..o. - -
y y
.o
.o
.-. (._,
000
,703 600 500 400 b,i 500 200 O
a STRIKE-SLIP FAU LT DATA e
Ws d'
/
y sco o co 7 -._
-J
/ ]
00 l
~x ro l
~
I
/
/
I l
4//-
i-2 l
GO Dq f.'/--
( L.
50 of-o 5
3o A
4 0
c W 30
=
n' D
F-Q.
20 w
O
$e D
En La.
IO 3
o
/
/
I
/
/
j e
=
=
/
A w
c
-o J
l s
/
/
S 3
- 4.,
a..n n
o D'
V
&p-vw 2
.5
. +.
4 a
g 03 v
ge o
d\\
11.
s.
u
./.
{
===r se v k
-i
n.
., ' (,
~
W 10 Un;f ATTAClinE::T 6
i n
...,x.,,
r#.3'yd%
United States Department of the Interior
( Mei'!-.
5 ceor.ocicar, sultvsv Q~d.h. '>h
\\,.'$.
Denver Fedend Ce nier Denver. Colorado 0022:~
Liox 25046, Stop 966 8 8 p r*1 Y tit t 0 4 Tee.
April 13, 1978 Dr. Eli A. Silver Division of ::atural Sciences Applied Scicncen P.uilding University of California, Santa Cru::
Santa Cruz, Californi.a 95064
Dear Eli:
Your let.ter of rei ruary 27,197S, indicates your continued interest in the cvaluation of earth scicoce data and interpretatiaa of it that tay bear upon the safe operatien of the Diab.lo Canyca rauclear reactor.
Uc chare your conccrr. for the Dichla facility, as ucil au for other nuclear react or f aci'itier. f u the United St.ates.
You r s ta ttuen t, houever, that
"...The uh".)le quert ion o f naxieum potenti al ca rthqual.c and saf e shutdoten carthquake..." be reopened in not justified.
Th: fact is that the question can never closed. tiuch dehnte over varicus aspects of tha
(
f acility is. cont in ting, and t-till continue.
am-.
Uhen vc first received ycur letter, we considered renpending to your pg$))
.crecific argument.u.
This, heuever, would acrve no purpose.
- Cranted, ncv geologi.c infornation has been acquired and neu interpretations have been saade during : 5 laat 2 years.
The data, however, crc onJy good enough for pce:aiscibic interpretations.
Furthernore, almost no new informatien that veuld help us answer a much note it;ies: tant and diffi-cult question has becu acquired.
This question conce run hcu you deter-minc the ground not ion to be expect.cd at a particular nice for an assumed cartiiquai:n.
lhe deteroinarinn involveu not j u:.:t a f ault-length magnitule relatiormhip hut. other cource pa rar:cters, at tenuation functicon, and t hich ground r.: t ion parameter in ::ont :igni f J enat to chich parts of
'a facility. All of thene are no poor). kno :n and numhcr ; calculated from th n range so greatly that argumente over thether a mar,nitude-7, 7.5, or it carthquake should be the no-calic.I design catthquake become of secondary luportance.
A]no, geotor.ints' a rgunen:." ove r uh, t her 10, 50, 100, 200, etc., kilo:..eturn of offnet han occurred along a particular fault have very little relevance to e,round metien calculations, it is very iuportant to establish a mav:imum size earthquele b.n;ed upon geology and neinmology', but attenptu t o he very specf (!c about cir.c canact be made rigorounty.
Hence 15 0 ) '. recent paper in Ceclov.
0.
6, p.
231-2 3.'s ) ir an exano]c ci the quaiuf tr. of reeleci.tn anal re i: m l < n i..t : when U'
thev alter.ot to cr.t inar. nae.n i t u.l.; iren fis): 1 en. ' h.
.he
- i
- .e" of an g
ca r t.hytua' e,,
nien 1: a ses y s.oni i es e c. c.,.pte.n t 4. y, i. certainly one of
.(
- neve ral centrolliar. factuen e f. ;,rniual c:pt itto, en t.t p:s t d
.it.- a g.::r t,i e u la r.
.. i t e, tsu t ether fattorn ate v.,r-l',por r an t.
An ic? rpli p. r f '. : 1.i r l y.
's
'W e - u4-
=
l 1
s
(
C
.t
(
important to the liiablo Canyon site ir ther ground motion to be expec cil at or very close (<10 laa) to the source (rupturinp, Ientt) of
.n cart h-
-~
quake.
As you prebably know. no incanure:"ents of gi ound.. oliva have ever been unde within 10 km of the source of a large earthr:uake.
Au you a)no probably know, hi:;h accelerationc have been measured near coinc small carthquaken.
Ihr. ever, the peak :otionn for these carthquahcu, uhich are higher than any ccucidered f or the 1)iablo nite, did not caunc :tructural damage.
Critical f actorn for engineerin;; design are hou r.uch and in uhat f o r:n the encrg: is transnitted to the sit.e and into the facility.
Enginecen cont,ider that t he energy t.rann:al.t ted into the f acility is limi.ted by the ult.inate strent.h au-l transmisnibi.lity cha"actcrint.ics of the tochs.
Also it is postulated that t he eccheni:.e.s of f rac t ur e propagat ion preclude ever iatrr a<;in;, a: sents of energy arriving at any given loc.zli t -
f rom progrer.,circly ] <:nper t apt urc zonen.
lienc e, enginecrc believ there is an upper bound to the inpet functions which should be contddcred for design.
T,c c au n <.. the hirher magni::ude creund cotion fun <:rinun :necif ted in Ciretila r 672, i.nti used at the Diablo C.avon site. ev.ocd thenc limits _ they nuct. according!.y be subiected tn -nceiff.cd reducticas.
To repeat a state:.. cat nade at the beginning of thic larter, ve are concerned ever data that n.iy be reicvant to the cafe operatica of any nuclear rnactot.
As inplied al ove, hetic"ct, you and other
- ..ay be
(
rainintf t he uron. quant.ican.
A7.1 of the data and EncuN.dge recuired :.o cay that a reactor can be operr.t.ed eith :.cro rich durine itn lif e :131 never exist.
Uhat nuet eventually be decided is th acceptable icvel of rick.
Th l.: rationale chould he, and is graduall;. being, applied to
~
other potentially lcthal n n e.ade prcductn cuch an pollution, pesti-cides, automobiles (a herribic hillcr), trancportat ion of toxic che:-ical<.
dann, changes in ecological balance., etc.
!!ced I go on?
Uhether or not, or hou soon, estinaten of risk ufil bc incorporated fornally into reviews of maior facilitics or activities is not eithin our p Evie..
ito..ever, ue [o wdnt to e. ;ibanine tha: th.
kind of neu geologic infornarion you cit e in not likely to contrilute to any cub-stantial changen in anses::m.nts of the :at e operat ten of the Diablo Canyon nuclear re.netorn.
If and when noncone or nm. p.roup formally precent n ':cIl-docosaented neu inf ernation to the I:uclear gegula*.ory Comminnion, the UTCS uill prohahjy be requocted to tcvicu it.
Our rele in to tuvice intornation at the request of t he :'gG and to bring to their attention inf ornation that may be relev. int to a reactor uit.c.
We do net nahe judginent r. on the :;afety of a nacicar reactor.
~
Best regards,
/
4.-
,./i<.sJ.'
- l. A.thUrown h. h. )
s q
-g.