ML20209C959
Text
,
5t Q ry g
4 Q
Q I
IRAY 211984 f
f HEMORANDUM FOR: James P. Knight, Assistant Director for Components & Structures Engineering Division of Engineering FROM:
Robert E. Jackson, Chief Geosciences Branch Division of Engineering
SUBJECT:
PRELIMINARY
SUMMARY
A ND EVALUATION OF ARTICLE CONTAINING NEW INFORMATION OR INTERPRETATIONS OF FAULTS IN THE NEAR OFFSHORE OF CENTRAL C0ASTAL CALIFORNIA (INCLUDING THE HOSGRI FAULT NEAR DIABLO CANYON)
On March 22, 1984, representatives of Pacific Gas and Electric Company (PG&E) informed the Geosciences Branch that they had received a preprint of an article entitled, " Post-Miocene Compressional Tectonics Along the Central California Margin", by J. K. Crouch and others, of Nekton, Incorporated. This paper was presented at the annual meeting of the Pacific Section of the American Association of Petroleum Geologists in
[
April,1984 It is published in " Tectonics and Sedimentation along the California Margin," J. K. Crouch and S. B. Bachman, Editors, Pacific Section of the Society of Economic Paleontologists and Mineralogists, 1984 We have completed a preliminary review of the article, had discussions with the principal author of the paper, PG&E and their consultants and brief discussions with the U. S. Geological Survey. We have attached our preliminary assessment of the potential effects that the interpretations contained in the paper could have on the seismic l
licensing aspects of Diablo Canyon.
The primary potential effect stems from the observation that the faults offshore may be more compressional in nature (one side of the fault moving up and over the other side) than previously understood, which was dominantly strike-slip (opposite sides of the fault moving past each other). The observations in the paper indicate that the Hosgri fault could bend over at depth and project beneath the site, possibly in closer proximity than the 5.8 kilometers now used for the Hosgri earthquake distance.
Based on our preliminary review we see no immediate basis-for modifying our previous conclusions regarding the seismic design bases at Diablo ge4Y Canyon. We base this conclusion on the following observations:
b @ M W h _/ e40521
, 4F ADOCK 05000275 omce >
~4G
.............-~~..
.~.. ~. - ~. ~ ~ ~ ~
~ ~ ~ ~ ~ ~ ~ ~
l*****I
~~~"" g g g;Q'j"""
. ~ ~.. - ~ ~ ~ ~ ~
~~~~..-~~
~~~~~~~"~~
~ ~ " ~ ~
l *^"
'"M ~~ V'N
................................... ~.
.... ~.... ~. - -
- ~ ~ ~ ~ ~ ~ -
~~~~~~~~~
5c ronu so no,somacu oado
- "*""*".*"7 OFFICIAL RECORD COPY 4., g3 4 u. _ m..
_. ~,.... -
. - * -- - - - --.. i*" '[ _ [ C,. E j i _ 1,ZT1
~
9
')
)
1.
The paper has just become available and, although in our judgement it appears to be a quality publication, the theories and basic thesis contained in it need to be reviewed in the total context of our evolving knowledge of California tectonics.
In the geologic and seismic area, many new hypotheses evolve and are often modified substantially with time.
It would, therefore, be premature at this time to assume all information contained and inferences drawn as established fact or fully accepted theory.
2.
During licensing activities for Diablo Canyon, this compressional aspect of faulting was included to varying degrees in the specification and validation of the ground motion, and although new ground motion estimates using information in the paper may be higher than previously assumed, the observations could also lead to lower estimates of both magnitude and ground motion.
3.
It is currently assumed that a magnitude 7.5 event can occur 5.8 kilometers from the site with a resulting.759 ground motion. This facility, therefore, already considers a near field event in its design.
4.
All onshore faults in imediate proximity to the site have been shown to be non-capable and this new information would not change that conclusion.
Finally, we have recommended and the Comission made it a condition of the Low Power License that a validiation of the seismic design bases he provided by July 1, 1988. This condition has been required to assure that all new information is evaluated by the most modern techniques.
Therefore, all new infomation, such as that contained in this paper, would be incorporated into that analysis.
In view of this license condition, we recommend that licensing action proceed and the new information be factored into the future validiation effort.
OT1 1aal Signe4 kg' 8
R. E. Jackson +'
Robert E. Jackson, Chief Geosciences Branch Division of Engineering
Attachment:
DISTRIBUTION:
As stated DCD 016 GSB RDG
/')/)
/)sn 7
..%B..........QE.;9.s.@ k$/...
.0.E[. f...
omer >
GsB....
.. Rilpthman....Spum..... I Da t
.RE ckeL-..
suRum=> RMcMu 1 s.1
........... ~......
omy..... 5 AS.84......... 5/!.WB4......... 5NB.4............ c&.......
.. 5h)84~~~
~ ~. - - -.. -
NRC FORM 310 (1HO) NRCM 0240 0FFiClAL RECORD COPY usom mi-au-ow
I PRELIMINARY
SUMMARY
AND EVALUATION OF ARTICLE CONTAINING NEW INFORMATION OR INTERPRETATIONS OF COASTAL CENTRAL CALIFORNIA INCLUDING THE H0SGRI FAULT SYSTEM NEAR DIABLO CANYON An article entitled, " Post-Miocene Compressional Tectonics Along the Central California Margin", by James Crouch, Steven B. Bachman, and John T. Shay was made available to the'NRC staff on March 22, 1984; by the Pacific Gas & Electric Company (PG&E) utility. The major thesis of this paper is that most of the' fault systems west of the San Andreas Fault in Central California, including the Hosgri Fault system, may not consist of vertical planar' fault surfaces in which the movement is strike-slip (i.e., the major component of movement is parallel and horizontal along the fault trace), but may have fault surfaces which curve and flatten out at depth toward the northeast. The authors suggest that all these faults, including the Hosgri fault system, are predominantly thrust faults which have resulted from movement in which the overlying blocks 3.i to the northeast have slid up and over the underlying blocks to the southwest. This interpretation could indicate that the Hosgri fault system, which passes on the seafloor about 5.8 kilometers (km) to the west of the site, may curve and flatten at depth, and may be extrapolated to pass beneath the plant site at distances less than 5.8 km. The seismic reflection lines also indicate to the authors that no movement has occurred on the Hosgri fault since late Pliocene time, over 2 million years ago.
However, continued earthquake activity and fault plane solution determinations for the region, which are generally compatible with the thrust fault interpretation, indicate that t-earthquakes can probably still be associated with these faults.
The new f M -39/
A-n&
... _._-_ - _.._ __ _ __ - _ _. m _
)
2-information that suggests this apparent different interpretation of the Hosgri fault system consists of several high resolution seismic reflection profiles which have been processed by state-of-the-art techniques to migrate subsurface structures to their correct geographic
. locations. These seismic reflection profiles were taken offshore from a few miles south of Point Buchon, near Diablo Ca'nyon south to about Capitan in the Santa Barbara channel. Although interpretation of only a few lines were shown in the preprint, hundreds of similar proprietary.
lines exist within the oil industry, and reportedly show similar thrust faults. Based on discussions with one of the authors, additional proprietary information, including seismic reflection lines and well data, also influenced their scientific determinations..
[
With regard to their thesis of compressional tectonics onshore west of the San Andreas fault, the authors cite numerous scientific publications which when considered in their entirety, suggest to them that tectonic movement to the southwest on northeast dipping thrust faults is the major mode of deformation in that region.
Evidence that the authors cite includes geometry of the folds, parallelism of the faults and folds, fault plane solutions of recent earthquakes, and plate tectonic motions reported by previous investigators.
The paper appears to be well thought out, written and documented.
)
i There are several possible implications of this hypothes.is on the seismic design of Diablo Canyon.
- 1. The Hosgri fault system, which was.used as the controlling structure in determining the magnitude of the controlling earthquake.in the seismic reanalysis of Diablo Canyon may approach the plant closer than the 5.8 km stated in the SER.
- 2. Faults in the vicinity of the site may be splays of the Hosgri fault system if it is interpreted to pass und r the site.
- 3. The ground motion resulting from the controlling earthquake on a thrust fault under the site may be different than that used in the SER. Also, as the geometry of' the fault may be different (gently 5
dipping) than that assumed by the applicant (vertical), the magnitude of the controlling earthquake calculated from a thrust fault may differ from the magnitude 7.5 that was assumed by the U.
S. Geological Survey and NRC staff.
i l
The interpretation of the seismic reflection profiles suggests that the Hosgri fault system begins curving toward the shore at depths of about 2 1/2 km. Data further to the Northeast (nearer Diablo Canyon) is not presented in the article. If the fault is assumed to have a horizontal attitude as it is extrapolated to the northeast of the seismic reflection lines, based upon this most conservative estimate, it could pass under Diablo Canyon at a depth of about 21/2 kms.
If the interpretations and extrapolations are indeed correct, the faulting I
)
)
would most likely pass beneath the site at a depth considerably greater than about 2 1/2 kms. Observations of well-studied overthrust belts elsewhere suggest that. thrust faults continue increasing in depth and eventually flatten out along a common fault referred to as the sole fault at the base of the~ system of thrust faults, which is usually much deeper than 2 1/2 kms. The model postulated by the authors shows the sole thrust to be 10-20 kms deep. There is, however, no definitive data in the article, at this time to determine if there is a sole thrust or to determine its depth beneath,the site.
During site validation investigations of the Diablo Canyon site in the middle to late 1960's, several faults were found. These faults were C
evaluated by mapping and age dating as they were exposed along the seacliff adjacent to the site, in a network of trenches dug across the site, and in t plant excavations during construction.
. *:g Most of the faults range in offset from a few inches to several feet, are discontinuous and disappear into folds in the rock.
There are several larger faults in the seaciiff area with displacements on the order of tens of feet.
1 The ages of last movement along these faults were bracketed between 15 million years, which is the approximate age of the rocks that are cut by L-l l
k
1
' the faults, and 80,000/120,000 years before present which is the age of the marine terrace on which the plant is located.
None of the faults offset the terrace or the deposits overlying it.
USGS geologists examined the faults exposed on the seacliff and in excavations at the site, and they and the AEC concluded. that the faults were not capable within the meaning of Appendix A 10 CFR 100.
The new information (Crouch and others,1984) is not significant wit!h respect to the capability of the' faults at the site because the faults do not displace the marine terraces and are, therefore, old and not capable.
If the thesis of the article is correct and the site faults 1
are related to that kind of regional motion, vertical offset in the terraces would be more obvious than if they were related solely to strike slip faulting which has been assumed to date. However, in the rugged terrane that lies between Pt. San Luis and Pt. Buchon it may be more difficult to identify all thrust faults that do not intersect the exposed seacliffs.
The interpretations and inferences in the paper also have a potential effect on magnitude and ground motion determinations. The assumption of reverse or thrust faulting on the Hosgri fault could result in different estimates of ground motion at the site than if it was assumed to be a strike-slip fault.
It is generally accepted today that reverse or thrust fault earthquakes of equivalent magnitude yield higher ground motions than those from strike-slip earthquakes.
In deriving and later i
verifying the adequacy of the seismic design basis in previous licensing
)
actions, no clear distinction was made as to which type of fault is assumed.
In the most recent Appeal Board hearing, however, most of the reasoning presented assumed strike-slip faulting.
In assessing the level of ground motion from a reverse fault, several key input parameters would need to be known including:
1
- 1. The correct estimation of the earthquake magnitude. The compressional regime may require assuming an earth, quake of a magnitu'de either higher, lower or the same as the 7.5 currently used. This may not be much different since one of the prime bases for choosing magnitude 7.5 was the occurrence of the 1927 magnitude 7.3 Lompoc earthquake which is believed by many to have occurred on 9
a reverse fault.
- 2. The ratio of strike-slip to reverse or thrust motion.
- 3. The distance from the fault to the site. This is dependent on the extent to which the Hosgri flattens out as it proceeds eastward.
If it steepens rapidly the fault could occur at a depth much greater than 2 1/2 km beneath the plant.
.A The estimation of near-field ground motion from a large earthquake is a difficult task frought with a large amount of uncertainty. A good deal of extrapolation and expert judgement is still required to make estimates at the magnitudes and distances needed for Diablo Canyon.
The range of results assuming different fault types and distances undoubtedly will exhibit extensive overlap.
For example, using a
i
)
relationship by Campbell (1983) which does take these factors into account indicates that the 0.75g peak free field ground acceleration already assumed for the Diablo Canyon site would be a median estimate for the " worst case" (a reverse or thrust' fault at about 2.5 km) magnitude 7.5 earthquake and a somewhat greater than mean plus one sigma estimate for the magnitude 7.5 earthquake on a strike-slip fault at 5.8 km distant from the site. These free field accelerations do not take into account reductions for buildings and/or embedment effects, spectral amplifications or the significance of the thrust motion assumption on the vertical acceleration determination. Other investigators and/or techniques could possibly yield higher or lower estimates.
It is apparent, however, that the existing design basis will accommodate large near field earthquakes of different types at different distances.
Uncertainty exists, and will continue to exist, however, as to defining the precise level of conservatism for each different scenario.
g Reference Crouch, James K., Steven B. Bachman, and John T. Shay, " Post-Miocene Compressional Tectonics Along the Central California Margin," in Crouch, J.K., and Bachman, S.B., Eds., 1984, Tectonics and Sedimentation Along the California Margin: Pacific Section S.E.P.M., Vol. 38, p. 37-54.
- 4. -
l l
FO'D9 &-39t A~'3Y E
- - - - -