ML19344D904
| ML19344D904 | |
| Person / Time | |
|---|---|
| Issue date: | 07/31/1980 |
| From: | Justus P Office of Nuclear Reactor Regulation |
| To: | Barrows A CALIFORNIA, UNIV. OF, LOS ANGELES, CA |
| References | |
| NUDOCS 8008260309 | |
| Download: ML19344D904 (3) | |
Text
-
/
-- MW
. J. u W '.
i,U...
JUL 311980 f
i Mr. Allen G. Barrows Dept. of Earth and Space Sciences i
University of California (UCLA)
Los Angeles, CA 90024 i
f
Dear Mr. Barrows:
l Thanks for your willingness to discuss the San Fernando fault zone data that you published with Messrs Kahle, Weber, and Saul (p. 127-135, U.S.D.C.,
r
- v. 3, 1973). The following are the questions that I spoke of on the phone i
Tuesday.
I h
As far as we know, your report and map of surface breaks resulting from the San Fernando earthquake of February 9,1971 renders the San Fernando fault zone the best docunentedrmverse-oblique slip fault in the world. As you know NRC has compared aspects of the Verona Fault Zone, Vallecitos Valley, CA to the San Fernando and, by analogy, concluded that one-reter of reverse-oblique net slip on the Verona is an appropriate conservative structural 3
i design criterion.
I We have analyzed aspects of the surface offset data displayed on your rap
[
' data which might aid us in interpreting the analyses. Our prelimirra.y and would like to discuss with you and your co-authors aspects of the raw analysis of the vertical slip measurements shows the mean to be 39 cm, the standard deviation to be 31 cm, and 97% of the measurecents to be one meter; the =casure.ents range from 2 to 160 cm and show a positvely skewed " normal" distribution (see curve E).
r In considering the error bars.for, or limits of, these statistics we need to have your evaluation of the accuracy of the field measurements and a sense of l
their representativeness. In particular:
(a) to within how many cm or within what % actual value is a field measurement repeatable? (b) does a field measure-ment (as shown on your cap) represent the maxinum offset at that point or the j
average offset for that particular site? We would appreciate it if you would discuss these limits to interpretation and any others that you think may be j
important.
I i
- ma>
= = =
- l un>
3o 03 o 0 w9
- w.. ee.a=====, aa.=two e** cei s s t
, NBC PosM 3ts (Ms) NRCM 0240
-7
- =
+>
i JW.31iS80 2-I Also, please cosnider this more theoretical, seni-rhetorical, question, what muld you expect the frequency distribution of surface offsets of a reverse-oblique fault of San Fernando characteristics to be? At the I
accent, we feel that it ought to range fron zero (no-offset crack) to sore,
maximua value, with a preponderance of small offsets and fewer large off-l sets, such as shown on (A):
t i!
I I
i i
Such a distribution implies that the fault surface exposed is doubly teminated and has a crest, such as shown on (b):
i i
I l
I; rather than a plateau scarp (C) which would yield distribution (D):
I i
In fact, based on our extraction of measurements froc ycur nap, the distribution curve has this fom (E):
I i~
l 1
l l
l i
the fault surfaces, as predicted from the actual distribution, ought to
~
l appear like this (F), devoid of zero-offset cracks:
i I
l
- = = >
..n.. w I
n>
Nac roax m o.m uncx ono
- ...---r---.."...='-..
,v
, 3 1 1980 Also, please consider this more theoretical, semi-rhetorical, question, what.would you expect the frequency distribution of surface offsets of a reverse-oblique fault of San Fernando characteristics to be? At the moment, we feel that it ought to range from zero (no-offset crack) to some maximum value, with a preponderance of small offsets and fewer large off-sets, such as shown on (A):
'"]
ey$2b sokw o[Y ];kr,l,ba,(
(A)
(upy be *<.rcsN scarIs "
o a 14.iftet x
- -4 a
Such a distribution implies that the fault surface exposed is doubly terminated and has a crest, such as shown on (3):
F'
.,, ////< ll!" '
F rather than a plateau scarp (C) which would yield distribution (D):
enbebb sadw. Es V
(C)
- s !/j Il'"
(nl,,-l h.\\,3,a E<
j 7j (n)
-gm, r-
/ /t a-. 1 1.mer In fact, based on our extraction of measurements from your map, the distribution curve has this form (E):
abaA kNn0!rf, Abt35"'
W "i.... 4 s.st a..,
the fault surfaces, as predicted frcm the actual distribution, ought to appearlikethis(F),devoidofzero-offsetcracks:
l m
4 m
p n
c e
i i
DISTRIBUTION:
jut.311980 3-CENTRAL FILE GSB ROG If we assuce that the expected distribution (A) is. correct, then to convert fron (E) actual distribution to (A) expected distribution.~any measurements frcn zero to thirty c:n cust be added to (E), or alternatively, (E) is not like (A) because i
many scall or no-offset data points were onitted. We guess that nurerous small i
scarps or zero-offset cracks were present but were either indistinguishalle from the background irregularities or of a questionable origin, such as due to t
lurch, slur.p, slide or differential compaction, and therefore, in any case, i
not included in the fault-offset data-set. Could you coment on these i
assumptions?
Other aspects of this study that we night want to discuss later on are the interoretption of strike slip and horizontal dip slip (your' transverse dis-placenent) data, variation of offset within and between fault seg ents, and 8
the establishment of guidelines for expedient field reasure ent of surf ace
{
offsets designed to yleid accurate parametric offset-codels of fault never.cnts for predictive or planning (structural design or ha:ard zoning) purposes.
We'd greatly appreciate hearing from you and Messrs. Kahle, Weber and Saul.
I Sincerely, i
i Philip 5. Justus, Geologist Geology Section i
Geosciences Branch Division of Engineering t
bcc:
R. Jackson R. McMullen S. Wastler 6
1 i
t I
i I
..D 1h..
. = = *
..RJustus:sbg..............
w.*
.. 2/31/.S0..
unc was m nm.uncu a2.o
- "..........,-~.-~.n
" =
-