ML20236P524
Text
y 5]
p u
i LNQ
' United States Department of the Interior i
- f$f.v 3l
' J/~?
GEOLOGICAL SURVEY t
' ' ' +. ' '
RESTON, VIRGINI A 22092 n <-
December 24, 1975 g-.,.
,r~..
_ '. l >.!0
?-
/
f2-24-OS'q
- a..
x,.: 4 y
w....... $
' Chief, Site l Analysis Branch
' O y ('.,
Mr. Nilliam P. Gammill
/-
' Division of Technical Review Office of Nuclear Reactor Regulation-U.S. Nuclear Regulatory Commission
+
Washington, D.C.
20555 1
t J
Dear Mr. Gammill:
Enclosed is a draft review of the Amendments 31, 32 and 34 of the FSAR for the Pacific Gas and Electric Company's Diablo. Canyon site, Units 1 and 2, San Luis Obispo County, California, Docket Nos.-50-275.
and 50-323. This review was prepared by Frank A.'McKeown, who reviewed the geology, and James F. Devine, who reviewed the seismology.
Mr. McKeown was assisted by Holly Wagner, David'McCullo'h, and ll c
Robert Yerkes; Mr. Devine was' assisted by' Robert Page. and Wayne l
Thatcher.
I We are transmitting this draf t to provide'a basis for discussions at l
]
l our planned meeting in Bethesda on December 30.-
q l
Sincerely yours,
<, '. Qg (.,(
Fred N. Houser Deputy Chief-l l
Office of Environmental Geology l
Enclosure g-
]
1
)
ul'
}"
}
.l
\\.
yy i
f
'w f 3 i, I o
a,.. i g
~
8708120405 870729
.PDR FOIA-PDR
- ggigNOR87-214
- ~ ~ n) /
J.
F. Devine (Scismology)
Diablo
. yon
?
FSAR Amendments 31, 32 end 34 1
q l
PACIFIC GAS AND ELECTRIC' COMPANY DIABLO CANYON SITE, UNITS 1 AND 2 SAN LUIS OBISPO COUFTY, CALIF 0PJ:IA AEC DOCKET NOS. 50-275 AND,50-323
~ Geology and Seismology This is a review of the geological and seismological information contained in Amendments 31, 32, and 34 of the Final Safety Analysis Report (FSAR) for the Diablo Canyon nuclear power plant site. Amendment 37, containing important discussion of the gr'ound response pertinent to seismicity, was received in early November and too late to be considered in this review.
The amendments were prepared by the' Pacific Gas and Elcetric Company (PG&E) in response to a request in a letter dated February 12, 1975, from the Nuclear Regulatory Commission (NRC) for certain additional information relevant to design basis earthquake i
issues, which have been the principal problems requiring additional i
carth sciences information and analyses.
To support assertions in the l
FSAR through Amendments 11, 19 and 20, five requests for information (referred to as questions in the Amendments) were made.
I l
2.17.
Provide additional discussion and arguments for determining the maximum carthquake that can be expected on faults of various ranks within the San Andreas system.
Relate the discussion to historic scismicity.
2.18.
Provide. additional docume.ntation, including scismic 2:cficction.
profiles, on the inters'cetion of the ilosgri fault.:one with the Transverse Range faults.
Include geologic maps southward of those provided in the FSAR showing the structural relation-ships of the Transverse faults and structures having a northwest trend.
2 l
2.19.
Provide additional documentation, including sc.ismic profiles,_
on the northern reaches of the !!osgri fault' :ene.
Include a 3
fuller development of your views on the structural relationship
]
of the llosgri fault to the San Simeon fault.
2.20.
Provide additional information on the loc ~ation of.the:1927 cVent, together with its probable mechanism. _ Discuss probable relationships of this event'to the geologic structure in the region.
2.21.
Provide your evaluation of the maximum credibic carthquake on the.llosgri' fault zone.
Assuming this event occurs along.the J
segment of the !!osgri fault' zone nearest the site, evaluate its' response spectrum at the site and compare it with the design response spectrum.
The response in the FSAR to the questions has provided considerable additional gedlogic and seismologic information and analyses.
llowever, unaribiguous answers to the questions have not been achieved.
Nbny uncertainties in tl}e data and interpretations still exist.
l Among the most important of these are:
- 1) the location l
f and mechanism of'the 1927 carthquake, 2) the exact ~ relation of the j
i 11osgri fault zone to faults in the Transverse Range system and the-L San 'imeon fault, 3) the continuity of some faults, 4)' the relative S
s' amounts of dip-slip and strike-slip movement on~ the llosgri fault' :one,.
5) the sense of displacement on parts of the'llosgri zone, 6) f identification and correlation of acoustical units, and 7) kinematic relations among different fault zones.
l In addition to these uncertainties, some information shown on the l
profiles is not shown on the maps and vice versa, and some prof 11'e i
data are not included that are important to evaluate the extension or l
character of some faults.
Because geologic maps developed from seismic I
3 reflection profiles are based upon much interpretation'that may differ i
acong several interpreters, it was necessary for the purposes of our review to make independent interpretations of the seismic profiles.
These independent interpretations are somewhat dif ferent than the j
interpretations presented in Amendments 31 and 32.
The major differences are briefly described in appropriate sections of this review.
l t
l Although some changes in, and additions to, geologic and scismologic details have been made in Amendments 31, 32, and 34 compared with i
previous data in the FSAR, no major changes can be made in our conclu-sions that were stated in the review of the FSAR, and Amendments 11,19, and 20, which was transmitted to the NRC from the Director j
of the United States Geological Survey by letter of January 28, 1975.
j
[
1 The pertinent statement in our previous conclusions was as follows:
" Earthquakes along the EBZ1 presumably would not be as large 'as expected on the San Andreas fault, however, from the information presently at hand we can find no evidence that would preclude the occurrence of an earthquake as large as events characteristic of subparallel strike slip faults, which bound basins, such as the Santa Maria, in the San Andreas system and which do not transect structural provinces." The size of an car ~thquake on faults tha't bound basins was not specified in this conclusion.
For reasons stated in s6bsequent 1As defined in the FSAR, EBZ refers to the East Boundary fault zone, which is the IIosgri fault zone.
P 9
4 parts of this review, however, the magnitude of the design basis earthquake for the Diablo Canyon nuclear reactor site should be in the range of 7.0 to 7.5 and located on the llosgri fault one.
This is based principally on the fact that the November 4, 1927, earthquake had a magnitude of 7.3 and that the best estimates of its location indicate that it could have occurred on the Hoagri fault.
Selected comments important to an evaluation of Amendments 31, l
32, and 34 are outlined below.
Amendment 31 NRC Question 2.18 On figures 8 and 9 relative displacement on the Hosgri fault between Point Buchon and Point Sal is shown to be down on the east.
On figure 10 relative displacement on the southern extension of the Hosgri fault south of Point Sal is down on the west, which is compatible with the argument that the Hosgri fault is the east boundary of a portion of the Santa Maria Basin.
Changes in direction of relative movement, i
however, are very suggestive of lateral displacement, which may have occurrred af ter development of the basin and bounding faults.
On page 9, reference is made to fig.11 as evidence that no scarp-forming scismic events have occurred on the southernmost part of the 11osgri fault since prior to the h'isconsinan stage of the Pleistocene.
It is true that no offset of the ocean floor is evident on fig. II, llowever, close inspection of fig. 11 shows offset of tha post-h'isconsinan unconformity when sighting niong it or placing a
5
(
straightedge along the mapped trace.
Also, faulting o.f tne post-Wisponsinan sediments cannot be precluded bec'ause a change in acoutical signature is evident' across an upward projection of the fault shown in figure 11.
The change in the acoustical signature of unit A2 I.
across the fault is quite clear and may be evidence of lateral movement on the fault.
{
l It is not cicar frca the profilec in figures 13a and 13b that the I
disturbed. zones in them that are inferred to represent the West Ilosgri fault are the same.
At least three additional faults can be inter-preted in the profile of figure 13b.
Also a disturbed zone appears
)
l to be between statioils 133 and 136 in the profile 'of line 13a.
(
Kelcz, Bartlett, and Polaris survey' lines criss-cross this area and additional evidence from them to support or n,egate the suggested l
correlatica of disturbed zones should be demonstrated.
j I
An independent interpretation of the scismic profiles in the l
~
offshore arca from about Point Sal to about five miles south of Point Arguello indicates that the Hosgri fault extends at least five miles south of Point Arguello and does not turn eastward as suggested in Amendment 31.
Although the Lomp7c fault' zone appears to have offset the sea floor, and may therefore be considered cap,abic-of movement again, its length of only about eight miles as inferred by the applicant appears to be incompatibic with a magnitude 7.3 earthquake.
An independent interpretation of the scismic profiles _ in the area of the Lompoc f ault differs from that of,the applicant in that it shows that the Lompot i
9 9
)
1 6
i fault zone is about twenty miles long; the longest singic fault in the zone is.about fifteen miles in length.
brthermore, the displace-nont'is dip slip or possibly oblique slip, rather than reverse slip as suggested by the appiteant.
12C Question 2.19 I
As noted in the previous section the sense of. displacement on
\\
~
the southern part of the Hosgri fault is up on' the west side, fi'gure i
1 (N), and therefore is not compatible with its being primarily related to basin development.
IIowever, an alternative interpretation suggests the displacement on the Hosgri fault in figure 1 to be do*,rn on the west.
Figure 1 (N) has three buried faults not shovn on Plate I.
This
^
leads to questions concerning the interpretation of so:te of the data in the report.
Another instance of faults shown in profile but not on a map is l
1 l
seen from comparison of fig. 4 (N) and Pinte I.
The correlation of faults between Lines 16 and 12 (figs. 3 (N) and 4 (N)) is questionable.
A profile along Line 14 would help.
Also', an interpretation of Line 10 should be included.
Although the straight coast line between Cambria and Point Estero suggests that the extension of the San Simeon fault is just offshore; data are lacking to prove this.
None of tiie data present'ed in 4
1 l
Amendment 31 preclude the San Simeon fault from intersecting the 11osgri fault offshore between Cambria and Point Estero.
The two j
l faults even as sh'own on Plate II (N) are less than 2.5 miles apart.
7 and could very well be tectonically coupled to each other by an en echelon or anastomosing series of faults which is characteristic of faults in the coast ranges.
Figs. 7a (N) and 7b (h') are very pu:: ling.
They show an inficction in the scafloor over the llosgri fault, and a drastic cfiange in the thickness and acoustical signature of unit A2, assuming A2' is correlative with A2.
In addition to vertical dis}ilacement, lateral displacement, which is not mentioned, could be interpreted from these profiles.
However, the basis for separating'A2' from A3 is not 1
apparent.
Similarly it is not apparent -why unit A, cast of the fault, is termintted.
It appears to continue to the cast' edge of these profiles.
On figure lla (N) the A2 unit east of the fault at station 119 is correla'ted with the Monterey formation (p. 8, NRC Question 2.19, l
j amend. 31), but the signature of the A2 unit west of this fault is completely different. This 11thologic change, as elsewhere, suggests lateral displacement.
NRC Question 2.20 On page 10 it is reasoned that both the Hosgri and h'est !!osgri faults can be climinated as sources of the'1927 carthquake because neither the sea floor nor the post-Wisconsinan unconformity are offset in the epicentral area of the earthquake.
This reasohing is not satis-I factory because typically surface rupturing of a fault is discontinuous, and of fset may not be detected if the displacement had a large lateral j
i component.
Furthermore, as stated on page 4 of this review, the base of post-Wisconsinan sediments is of fset, and a f ault in thn mediments
8 cannot be precluded in figure 11.
The evidence, therefore, to eliminate I
l the Hosgri fault as the source of the 1927 earthquake is inadequate.
As previously stated, the length of the Lompoc fault shown by the applicant appears to be incompatible with the magnitude of the 1927 carthquake.
Figure 1 shows that segments of the Hosgri fault zone, the Lompoc I
fault, rurisima fault, and Lion's Head fault occur within the error circle of Cawthrop and error ellipse of Engdahl for the 1927 earthquake.
However, all of the faults are outside of the area designated by Smith as the " inferred distribution of aftershock sequence of the 1927
{
earthquake." The 1927 carthquake, therefore, cannot be unequivocally located on any one of thece faults.
The Hosgri fault, however, is closer to the center oi the estimate of error than the other faults and, therefor'e, must be considered as a possible fault on which to locate the earthquake.
l l
Amendment 32 NRC Question 2.17 Although this section contains descriptions and explanations of the "---kinematics of structural behavior in the south-central California region-- " contemporary seismic activity is.not fully explained.
Also, we do not agree with some statements, given as f act.
For. example, on page 2 it is stated as f act that the 1927 M.7.3 earthquake occurred on the Lompoc fault.
This is not fact but a highly controversial assumption.
Item 2 on page 2 of this amendment indicates that the,Lompoc and San Andreas are the i
~
~
9 only faults in the southern Coast Ranges.that "ref1'ect substantial fl l
late Quaternary surface deformation." As defined on pago 3 of this amendment, " substantial" c1carly includes the San Simeon fault, which as stated on page 6 of this review may be coupled with the llosgri fault.
The attempt'to explain the large magnitude by using the logic that the -
Lompoc fault is in a transition zonc between the Coast Ranges and
~
h'estern Transverse Ranges applies to other faults in the zonc including-the southern part of the Hosgri fault.
Amendment 34 NRC Question 2.21 Tlie maximum credible earthquake of 6 1/4 - 6 1/2 on the Hosgri fault zone used in this section to derive peak site ground acceleration is unacceptable because as stated previously the 1927 car'.hquake with a magnitude of 7.3 cannot be precluded from having occurred on the
!!osgri fault.
Conclusions Although the FSAR includes a considerable amount of new informa-tion and analysis, the only change that can be made in the original conclusions transmitted to the NRC on January 28, 1975, is to be more specific in our estimate of the design basis earthquake.
This is based upon the following facts and judgments.
1.
The llosgri fault zone is more than 90 miles long and may even be tectonically coupled to the San Simeon fault as they are within 2.5'
4 10 miles of each other and both form parts of the eastern boundary of the Santa Maria basin.
2.
Marked changes in thickness and signature of acoustical units across the Hosgri f ault zone in several profiles indicates evidence of lateral slip. This was noted in our review of January 28, 1975, but such changes are even more abundant in the profiles of Amendment 31.
Right lateral movement is reported for the San Simeon fault.
These data suggest that displacements on the Ho,sgri fault are related to the highly active San Andreas plate-boundary system.
3.
The length of the Lompoc fault appears incompatible with the magnitude of the 1927 earthquake.
4.
The Hosgri fault is closer to the center of the esticates of crror of both Engdahl and Gawthrop than any other f ault.
It is there-fore a possible source of the 1927-earthquake.
5.
Equivocal evidence related to vertical displacement on the Hosgri fault in the epicentral area of the 1927 earthquake does not clininate it as a source.
Surface fupture is generally discontinuous, and if lateral slip c.: curred, it probably would not be detected.
Offset of the base of post-Visconsinan sedicents and probable faulting of them is evidence of post-Pleistocene covement.
For the above reasons and discussions gi.ven in the review, we conclude that the 1927 carthquak, could have occurred on the Hosgri f ault and that a similar earthquake with a magnitude in the ranse of 7.0 - 7.5 could occur in the future anywhere along the Hosgri f ault.
I
[
~
l' 11 6.
We repeat our opinion that, for sites within 10 km of the surf ace i
expression of a fault, the description of maximum carthquake ground motion by means of a singic acceleration value and a standard response spectrum may not be an appropriate representation of the ground motion for design purp'oses.
However, if a singic acceleration value is to be q
]
designated by extension of the existing strong motion data base, the i
0.5 g acceleration offered by. the applicant is inadequate for representing,
a magnitude 7.0 to 7.5 carthquake at a point on the Hosgri fault nearest
{
I the plant site when used in conjunction with a conventional analysis.
l The appropriate value should be developed in association with a response I
spectrum taking into account the proximity of the f ault, site conditions and the esticated larger earthquake than the magnitude 6 1/3 - 6 1/2 used l
by the applicant in these studies.
4 4
0
----- _ _ _. _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ _ _