ML19281A098
| ML19281A098 | |
| Person / Time | |
|---|---|
| Site: | Vallecitos File:GEH Hitachi icon.png |
| Issue date: | 02/28/1979 |
| From: | GENERAL ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML19281A099 | List: |
| References | |
| NUDOCS 7903060332 | |
| Download: ML19281A098 (22) | |
Text
.. _ _.. _ _ _ _ _
e 1sy a
-t -
i, EVALUATION OF SEISMIC HAZARD J
AT THE GENERAL ELECTRIC TEST REACTOR SITE, 3
$=1 m
sii ma BY RICilARD H. JAHNS 4
~
I REPORT TO GENERAL ELECTRIC COMPANY
[
\\
-s: - -
' FEBRUARY 1979
('
39030603%
EVALUATION OF SEISMIC EAZAED AT THE GENERAL ELECTRIC TEST REACTOR SITE, ALAMEDA COUNTY, CALIFORNIA by Richard H. Jahns Report to General Electric Company February 19'/9 C0 HOPED COPY
CONTENTS Pa ge Introduction 1
Background sketch 4
Vallecitos landslide complez 6
Verona fault 11 Conclusions 16,
References cited 18 e
4
INTRODUCTION The General Electric Test Reactor (GETR) Site occupies a part of the Vallecitos Nuclear Center at the northerly edge of Vallecitos Valley in Alameda County, California.
It lies about 3t miles south-southeast of Pleasanton, and is bordered on the northeast by the steep southwesterly front of the Vallecitos Hills.
Farther northeast, beycnd these hills, is Livermore Valley.
Both the site area and the adjacent Vallecitos Hills are underlain by the Livermore Gravels and other nonmarine sedimentary deposits of late Cenczoic age.
Terranes of older rocks are exposed in nearby areas at distances of a mile or more from the site.
In a general sense, the subject area
~
lies within an elongate structural block that is bounded by the northwest-trending Calaveras fault zone, 1.9 miles (3 1 km) southwest of the site, and the Livermcre fault zone, 4.3 miles (7 0 km) northeast of the site.
The Greenville fault zone, also with a northwesterly trend, lies farther ncrtheast.
The Calaveras is an active fault zone that can be re-garded as capable of generating a major earthquake.
Of more immediate concern in t'.1 context of seismic hazard at the GETR, however, is the question of whether an active fault may be present in ground nearer to the site.
On this score,
the existence of a Verona fault has been postulated in the Vallecitos Hills - Vallecitos Valley area by some geclogists, and it has been further suggested that such a fault could well be active or at least potentially active.
This raises serious questions concerning possible surface faulting and strong seismic ground motion in the site area.
Even though investigators who have inferred the presence of a Verona fault have not agreed as to its probable location, and even though viewC cencerning its geometry and sense of movement have shif.ed during the past two years, the existence of such a fault near the GETR has remained a possibility that cannot be firmly dismissed without definitive evidence of an un-broken geologic section.
Toward resolving the question of geologically young faulting in the site area, extensive and detailed geotech-nical investigations were made on behalf of General El'ectric Company by Earth Sciences Associates during the period 1977-1979 (ESA,1978a-d,1979).
Corollary attention to site geol-ogy has been given by other consultants of General Electric Company, and by representatives of the California Division of Minea and Geology, the Nuclear Regulatory Commission, and the U. S. Geological Survey.
Marked differences in inter-
.pretations and opinions among these various investigators and observers have emerged from a two-year series of reports, memoranda, conferences, and lively field discussions.
Most
~
of the expressed disagreement has been focused upon two questions:
Does a Verona fault exist in or near the GETR site area?
Do observed discontinuities in geologic materials of the site area represent past activities of faulting or of large-scale mass wasting (landsliding)?
The purpose of this report is to provide an independent review of data and conclusions deriving from geotechnical in-vestigations thus far completed, with emphasis on the two foregoing questions.
Particular attention is given to the most recent studies made by Earth Sciences Associates in 1978 and 1979, and to comments and questions offered by staff of the Nuclear Regulatory Commission in' 1978.
The treatment is more analytical than descriptive, and no attempt is made here to repeat either the details of arEuments advanced by others or the specific descriptions and documentation that already are in the record (e.g., ESA,1978a,1979).
I have carefully reviewed all written materials furnished me as per-tinent to the evaluation of seismic hazard at the subject site, and I also have reviewed related parts of the earlier published record.
An aggregate of eight days has been devoted to indepen-dent examination of exploratory trench exposures, study of geologic relationships in and near the site area, and field revi 'w of rela tionships at selected localities farther from the site.
Participation in five field inspections with other geologists involved in the overall study has been of consid-
4_
erable benefit in providing exposure to contrastinF inter-pretations and points of view.
BACKGROUND SKE'ICH Predecessors of a Verona fault were suggested more than a half-century ago.
As noted and described by ESA (1978a,
- p. IV-1 et seq.), a fault was shown on early published maps to extend southeastward from points near Pleasanton to a connection with the Williams fault about 2 miles east of the site area.
The trace of this inferred fault was variously associated with the southwesterly front of tne Vallecitos Hills, mainly on the basis of broad topographic considera-tions.
The name Verona was first assigned by Hall (1958) to a west-northwest-trending fault inferred by him to lie within the Vallecitos Hills.
Existence and location of the fault were based upon geomorphic expression, observed structural features in the Livermore Gravels, linear distribution of springs and ponds, and evidence suggesting subsurface strati-graphic offsets that would require a reversal of movements in post-Miocene time.
A new location for the fault was later postulated by Herd (1977), who correlated its trace with the southwesterly base of the Vallecitos Hills.
He cited geomorphic expression, along with a discontinuous line of springs and seeps along
~
the base of an " arcuate line of truncated spurs" as evidence for this location.
Thus he applied the same kinds of geo-morphic and hydrologic evidence as did Hall, but to an in-ferred fault trace significantly different in position and trend.
The evidence has been discussed by ESA (1976a,1979),
and I agree frcm independent observation that the existing relationships of topography, springs, and seeps in both areas are not in good accord with linear alignments, faceted spurs, and other features that could be expected alonF the trace of a geologically young fault.
No'r is it at all clear that tne stratigraphic relationships described by Hall require a Verona fault for their explanation.
Hall evidently thought of the Verona fault as a near-vertical feature, and Herd's mapping and description of a fault farther southwest in the site area suggest that he in-ferred the break to be steeply dipping.
Views of a possible Verona fault subsequently have changed, however.
It was pointed out, late in 1977, that some degree of geometric fit between trace and topography might be achieved through pos-tulation of a northeastward-dipping thrust fault daylighting along the base of the Vallecitos Hills.
Shortly thereafter, two exploratory trenches in the site area revealed rupture features in the Livermore Gravels and some overlying mater-ials that could be interpreted as expressions of either thrust faulting or large-scale landsliding.
~
An extensive landslide complex, considerably mcdified by erosion, was identified in 1977 along the front of the Vallecitos Hills (ESA,1978a), mainly on the basis of geo-morphic relationships and scattered observations of bedding attitudes in the Livermore Gravels.
Arguments to explain obcerved geologic features in the GETR area in terms of mass-wasting processes rather than faulting were subsequently referred to as "largely assertive" by some critics, a mildly pejorative characterization that could be equally well ap-plied to arguments that had been advanced for inferring the presence of a geologically young fault.
The differences in view actually have represented contrasting interpreta tions from a data base that has been far from complete.
The ad-ditional exploration and study discussed in following sections of this report was cimed at improvinF this data base.
VALLECITOS LANDSLIDE COMPLEX Both the hillside northeast of the GETR site and much of the site ground itself can be interpreted as parts of a very large ancient landslide complex.
The domain of gravi-tational failure extends over much of the southwesterly front of the Vallecitos Hills, and it evidently includes a belt of much lower rolling ground that flanks the base of the hill-slope.
Inferred geometric relationships of the slide com-
plex, as based in part or, topographic relationships and in part on the results of test drilling and exploratory trench-ing, have been indicated by ESA (1979).
In my judgment, the landslide model is probably correct for the areas in question, the landslide complex represents several episodes of ancient movement, and all major parts of the complex have become stabilized in pre-Holocene time.
This view is based upon the following relationships:
1.
Most of the landslide domain is grossly re-flected by present topography, with amphi-theater-like expressions on higher ground, bench-like features in intermediate areas, and relatively steep-faced snout-like ex-pressions along the base of the main hill-slope.
Lower and more distal parts of the complex also have topographic expression, albeit in much more subdued form.
Overall relationships are clearly recognizable on remote imagery.
2.
Oversteepened pre-slide topography is readily inferred for the southwesterly slope of the Vallecitos Hills, and the Livermore Gravels section involved in the ground failure in-cludes abundant silty and clayey materials with low shear strength.
Such materials are well exposed in various parts of the B, F, and G-series trenches, for ex-ample (ESA,1979).
3 The slide complex must be very old, as in-dicated by deep erosional dissection.
Moreover, the present headscarp features do not represent the positions of the or-iginal pull-away margins of the slide units, but instead have migrated north-eastward and have been heightened through post-slide erosion.
No rubble-filled pull-away features were exposed in the F-and G-series trenches, which instead re-vealed several steeply inclined slip sur-faces that could well represent the rela-tively tight roots of original pull-away features.
4.
As indicated by ESA (1979), these steeply in-clined slip surfaces are ancient features that do not disturb Holocene soil horizons.
They antedate modern surficial slide masses that have fresh topographic expression and disrupt the youngest soil horizons.
7 A major distal part of the slide complex can be approximately outlined by means of ex-posures in trenches T-1, B-1, and B-3 (ESA, 1978a, 1979).
Well-defined shear surfaces, as revealed in these trenches, in general dip northeastward at low angles and could well represent the simple or compound sole of a major part of the landslide complex.
6.
The daylight line of wha'. probably is a deeper-seated slide sole can be traced by means of exposures in the B-2-series trenches (ESA, 1979).
An apparently continuous shear sur-face, which does not correspond in position with a photolineament to the southwest, dips northeastward at low angles.
Like the shear surface exposed in trench B-1, it evi-dently reflects predominantly reverse dip-slip movement.
7.
The exposures in trenches T-1, B-1, B-3, H (not checked by me), and the B-2 series permit reconstruction of displacements along the two principal shear trends in distal parts of the landslide complex, as outlined by ESA.
Dating of soil strati-graphic units (Shlemon, Appendices A and B in ESA, 1979) that have been displaced
~
indicates a maximum of 3 feet of slip
during the past 10,000 to 20,000 years, 3 to 9 feet prior to that but since 70,000 to 125,000 years ago, and at least tens of feet post-dating the Livermore Gravels.
8.
Such multiple movements are by no means un-common in large landslide ccmplexes, in-cluding several that I have studied in the Palos Verdes Hills and the San Francisco Bay negion, California.
The t tire assemblage of data indicates that the Vallecitos complex is very old, perhaps dating back several hundreds of thousands of years, and that it has been reactivated on a minor scale dur-ing pluvial stages of Pleistocene time.
Such correlation between slide movement and times of wet climate has been established elsewhere (e.g., Ehlig and Ehlert, 1978; Stout, 1969), and it seems highly unlikely that the Vallecitos complex, with its tem-poral pattern of activity, could have been active in post-Pleistocene time.
O VERONA FAULT Most questions raised by staff members of the Nuclear Regulatory Commission and the U. S. Geological Survey con-cerning the claimed existence or non-exi~ tence of a Verona s
fault have been dealt with in the latest geologic report by Earth Sciences Associates (ESA,1979).
The treatment ranges from a well-reasoned discussion, with appropriate documenta-tion, of regional geology and structural evolution to specific descriptions of recently exposed features that bear on the presence or absence of a significant fault.
A careful review of the record to date suggests to me that existence of a Verona fault has not been established, but that such existence cannot be firmly ruled out.
I regard the following factors as most important among those militat-ing against the existence of such a fault:
1.
Difficulties in explaining the southwesterly slope of the Vallecitos Hills as a result of uplif t along a northeasterly-dippinE thrust fault.
The required offsets in Quaternary time should be accompanied by surface expressions that do not appear to be present.
P.
As pointed out by ESA (1979), the widespread occurrence of a shallowly buried paleosol in the site area indicates a long period cf landscape stability.
Such paleosol accumu-lations could be expected to become buried beneath younger alluvium derived from a rising (and advancing) fault scarp.
3 Occurrence of a major northeast-dipping thrust fault is not compatible with the geologic section beneath the nearby Livermore Valley (ESA, 1979).
4.
Exposures in trench E and seismic results along reflection line 1 (ESA,1979) appear to pre-clude a northwesterly extension of a Verena fault from the GETR area.
5 Available evidence indicates that it is all but impossible to extend a Verona fault south-eastward to a connection with inferred faults trending northeast (ESA,1979).
No one has yet seen an undeniable Verona fault, and much of the evidence cited for its existence and location either can be dismissed on the basis of detailed observation (e.g., aligned springs and seeps) or can be reasonably ex-plained in other ways.
Ncnetheless, the absence of such a fault has not been proved, especially in and near the GETR site area where it has not been practicable to trench all ground where a fault conceivably could be present.
Given this dilemma, let us then assume that a Verona thrust fault is indeed present in the site area.
If it is buried beneath parts of the Vallecitos landslide complex, it must be an ancient feature, with youngest movements at least 70,000 to 125,000 years ago.
On the cther hand, it could be assumed that the fault is represented by either or both of the distinctive shear surfaces exposed in trenches T-1, B-1, B-3, H, and the B-2 series.
If these shear surfaces were thus reFarded as tectonic in origin, the respective off-sets of dated soil-stratigraphic units would suggest approx-ima tely 0.05 millimeter per year as an average slip rate (strain relief) for either of the surfaces during the past 70,000 years (ESA,1979).
These surfaces thus would represent respec tive slip rates at least two orders of magnitude lesser than reported creep rates along the Hayward and Calaveras faults, as pointed out by ESA (1979).
Further comparisons are provided in the table on the following paFe.
Here it should be indicated that the San Jacinto and Whittier-Elsinore-Agua Caliente-Laguna Salada fault zones are characterized by strike-slip movements, whereas the other fault zones and systems are characterized by a decinance of reverse dip-slip movements and hence are ccre closely comparable with the shear surfaces in the GETR area.
Evidently these shear surfaces could not have been highly active tectonic fea tures during late Quaternary time.
If mul-Fault Averare f.vera re General or Length Earthouake Slio Reference pault (km)
Recurrence Rate Period 3.P.
System Interval (mm/yr)
(1000s (yrs) of years)
San Jacinto 440 400*
3 5,000 Whittier-Elsincre-Agua Cal-260 2,000*
0.8 6,000 1 ente-Lafuna Salada+
White Wolf +
53 2,000*
0.4 7,000 Sierra Madre +
90 300*
8 30 Sierra Madre #!
90 1,700 0.7 10 Raymond#
20 5,000 0.13 86 Lakeviev 25 2,000 0.7 500 GETR site ++
(either 8
20,000 0.05 70 shear)
- For M 7 event.
I 13 km segment extending eastward from Big Tujunga Canyon.
+ Lamar et al.,1973
- Allen et al.,197E.
++ ESA, 1979 e
tiple slip events are assumed to contribute about one meter of fault movement apiece, the corresponding recurrence inter-val has an average value of about 20,000 years between suc-cessive events.
This is at least an order of magnitude lonEer than estimated average recurrence intervals noted on the preceding page for active faults in California.
If more than one meter of movement is assumed for any event, the average recurrence interval becomes even greater.
One can only agree with the ESA (1979) conclusion that a maximum of one meter of net slip is a highly conservative assessment of the amount of offset that might occur on a thrust fault at the GETR site during any one event.
Further, it can be pointed out that such inferred maximum offset would daylight in ground not occupied by the reactor.
Indeed, no offset from tectonic causes is to be expected in the foundation area of the reactor, as no breaks in the Livermore Gravels can be projected into that area from the nearby exploratory trenches.
Under the most conservative assumptions of a real Verona fault that is expressed by the shear surfaces discussed in the foregoing paragraphs, and that has a length of 8 km, the fault cannot be reckoned as a major tectonic feature.
It must be considerably shorter than faults that have been judged cap-able cf generating M 6.5 earthquakes (e.g., Ray =ond and Sierra Madre as cc= parable thrust faults), and its activities in late Quaternary time cannot have been nearly as great.
In my judgment, assignment of 5.5 as a maximum value for magnitude of an earthquake generated along a Verona thrust fault re-presents an extremely conservative assumption.
CONCLUSIONS An extensive body of information now at hand justifies, to the best of my judgment, the following conclusions con-cerning seismic hazard at the General Electric Test Reactor Site:
1.
The hillside northeast of the site, together with a flanking strip of lower ground, is best interpreted as a domain of ancient landsliding.
The large landslide complex represents several episodes of movement, it has been considerably modified by erosion, and it was stabilized in pre-Holocene time.
2.
Existence of a Verona fault is very doubtful but cannot be firmly denied by the sum of evidence at hand.
The evidence does in-dicate that no Verona fault can exist as a major tectonic feature.
3.
If a Verona fault is assumed to exist, its youngest movements must have occurred at least 70,000 to 125,000 years ago if it is buried beneath parts of the Vallecitos landslide complex.
If it is not so buried and ins'tead is assumed to be represented by distinctive shear surfaces exposed in trenches T-1, B-1, B-3, H,
and the B-2 series, the average slip rate along these shear surfaces imposes constraints on amount of slip per event as linked with average recurrence interval between suc-cessive events.
4.
A maximum of one meter is a highly conserva-tive assessment of net slip that might oc-
~
cur along a Verona thrust fault at the GETR site during any one event.
No offset from tectonic causes is to be expected in the foundation area of the reactor, as no breaks in the Livermore Gravels can be projected into that area from the nearby exploratory trenches.
5.
A maximum of M 5.5 for an earthquake generated along a Verona thrust fault is an extremely conservative assumption.
REFERENCES CITED Allen, C. R., Crook, R., Kamb, B., Payne, C. M., and Proctor, R. J.,1978, Evidence for faulting recurrence in the Raymond and Sierra Madre fault zones: Amer.
Geophys. Union Trans., Tectonophysics Abstract T-226,
- p. 1210.
Earth Sciences Associates,1973a, Geologic investigation, General Electric Test Reactor Site, Vallecitos, Califor-nia: Report prepared for General Electric Ccmpany, Pleasanton, California (Feb. 1978 ).
Earth Scienc es Associates,1978b, Geologic evaluation of GETR structural design criteria: Report prepared for General Electric Company, Pleasanton, California (March 1978).
Earth Sciences Associa ces,1978c, Geologic investigation, General Electric Test Reactor Site, Vallecitos, Califor-nia: Addendum report prepared fcr General Electric Com-pany, Pleasanton, California ( April 1978).
Earth Sciences Associates,1978d, Landslide stability, General Electric Test Reactor Site, Vallecitos, Califor-nia: Report prepared for General Electric Company, Pleasanton, California (July 1976).
Earth Sciences Associates,1979, Geologic investigation, Phase II, General Electric Test Reactor Site, Vallecitos, California: Report prepared for General Electric Ccmpany, Pleasanton, California (Feb.1979).
Ehlig, P.
L., and Ehlert, K. W., 197E, Engineering geology of a Pleistocene landslide in Palos Verdes, in Lamar, D. L.
(editor), Geologic guide and engineering geology case histories, Los Angeles metropolitan area: Assoc. of Engineer-ing Geologists, First Annual' California Section Conference.
Los Angeles, p. 159-166.
Hall, C. A., Jr.,1958, Geology and paleontology of the Pleasanton area, Alameda and central Centra Costa Counties, California : University of California Publs. in Geol.
Sciences, v. 3k, p. 1 -90.
Herd, D. G.,1977, Geologic map of the Las Positas, Green-ville, and Verena faults, eastern Alateda County $9, 25 p.
Cali-fornia: U. S. Geol. Survey Open File Report 77-6
Lamar, D. L., Merrifield, P. M., and Proctor, R. J.,1973, Earthquake recurrence intervals on major faults in south-ern California, in Moran, D.
E.,
Slesson,'J.
E.,
- Stone, R.
O.,
and Yelverten, C. A.
(editors), Geology, seismology, and environmental ireact: Assoc. Engineering Geologists Special Publication, Los Angeles, University Publishers,
- p. 267-276.
Stout, M. L.,1969, Radiocarbon da ting of landslides in southern California and engineering geology implications, in Schumm, S.
A., and Bradley, W. C.
(editors), United States contributions to Quaternary research: Geol. Soc.
America Special Paper 123, p. 167-179 Stout, M. L.,1977, Radiocarbon dating of landslides in southern California: California Geolcgy, v. 30, p.99-105 e
.