ML20136F318
| ML20136F318 | |
| Person / Time | |
|---|---|
| Site: | 05000000, Vogtle |
| Issue date: | 10/04/1984 |
| From: | Knight J Office of Nuclear Reactor Regulation |
| To: | Novak T Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML082840446 | List:
|
| References | |
| FOIA-84-663 NUDOCS 8410170075 | |
| Download: ML20136F318 (29) | |
Text
..l, 5
UNITED STATES
. P h gl NUCLEAR REGULATORY COMMISSION h ';o g c[p/. j
- Asa
- :oTcs, o. c. 23sss o.
\\..... J' CCT 0 4 'SS4 MEMORANDUM FOR:
Innma5 & NoviK, M stant g Pge g g for Licensing, UE FROM:
James P. Knight, Assistant Director for Components & Structures Engineering, DE
SUBJECT:
DRAFT SAFETY EVALUATION REPORT - GEOLOGY AND SEISMOLOGY -
V0GTLE, UNITS 1 AND 2 PLANT NAME:
Vogtle Electric Gen uating Plant, Units 1 and 2 DOCKET NUMBERS:
50-424/425 LICENSING STAGE:
OL Review PROJECT MANAGER:
Melanie Miller, LB4 Enclosed are the geology and se'ismology sections for the Vogtle Draft SER, SRP Sections 2.5.1, 2.5.2 and 2.5.3.
The report was prepared by Ina B. Alterman, Geologist and A. Ibrahim, Seismologist.
As stated in the SER, the staff concludes that there are no capable faults at the site or in the region around the site and that the SSE and OBE are adequate.
A six month geological and geophysical fault investigation, undertaken to address a possible nearby fault, the Millet Fault, postulated in the USGS Open-File Report 82-156, failed to produce any evidence for the existence of the fault.
The GSB staff reviewed the study and concurred with the applicant's conclusions.
A recent clarification of the USGS position with respect to the 1886 Charleston, S.C. earthquake has resulted in NRC-sponsored probabilistic and deterministic studies concerning seismicity in the eastern U.S.
At the conclusion of these studies we will be assessing the need for modified positions with respect to eastern U.S. sites.
At this time we see no need to modify the position taken for Vogtle during the CP review and the staff does not consider the issue an open item.
M q,\\
J. P. Knight, Assistant Director for Components & Structures Engineering, DE t
Enclosure cc:
R. Vollmer S. Brocoun J. Knight I. B. Alterman E. Adensam A. Ibrahim M. Miller D. Eisenhut L. Reiter l, ',,
l gmo[ DOM 4
M k c
V0GTLE ELECTRIC GENERATING PLANT, UNITS 1 AND 2 DRAFT SAFETY EVALUATION REPORT 2.5 Geology and Seismology For this SER, the staff has reviewed all available relevant geologic and seis-mologic information obtained since the issuance of the CP-SER and supplements to the CP-SER in 1974 (USNRC, 1974) in accordance with the SRP.
In the CP-SER the staff and their consultants, the U.S. Geological Survey, concluded that:
(1) Geologic and seismologic investigations and information provided by the applicant provide an adequate basis'for determining that no faults exist at, or in the immediate vicinity of, the plant site that could localize seismicity.
(2) Ground motion values of 0.20g and 0.12g for the Safe Shutdown Earthquake (SSE) and the operating basis earthquake (OSE), respectively, are ade-quately conservative.
Since the issuance of the CP-SER, the applicant.has performed further detailed
~
geologic and geophysi~ cal investigations of the site and site region. This includes geologic mapping of the excavation for the main power block area, and a fault investigation prompted by a United States Geological Survey open-file report postulating the existence of the Millet Fault seven miles south of the plant site. A staff review of this investigation is discussed in Section 2.5.3
.of this SER.
During the current review, the NRC staff identified the following issues for evaluation:
.f 09/26/34 2.5-1 VCGTLE SER INPUT SEC 2.5
.7
...e
- . t
-l i
(1) new geological and seismological information discovered sinca the CP
+
review (2) the postulated Millet Fault south of the Vogtle site (3) significance of clastic dikes and associated structures at the site and in the site region j
'l (4) the adequacy of the seismic design response spectrum Much of the new geologic and seismic data has been developed from research in the southeastern United States, particularly in the Charleston, South Carolina area. During past licensing decisions the NRC and AEC have held to the posi-
~ tion that the relatively high seismic activity within the Coastal Plain Pro-vince in the vicinity of Charleston, S.C., including the 1886 Modified Mer-calli Intensity (MMI) X earthquake, was, for licensing decisions, related to a unique tectonic structure there. ~Therefore, in the context of the tectonic p
province approach, an MMI X earthquake should not be assumed to occur anywhere else. This conclusion was based primarily on the persistent historical sefs-
]
micity that has characterized the meizoseismal zone of the 1886 Charleston, S.C.
It was also based on evidence, though not strong, of unique geologic structure.
Lacking definitive information, the NRC-AEC based its conclusions in part on advice from the U.S. Geological Survey.
In 1973, with AEC funding, the USGS began extensive geologic and seismic investigations'in the Charleston, S.C. region.
These studies are still under-way. As a result of these investigations, a great deal of information has been obtained, but the sou.rce mechanism of the seismicity still is not known.
L
_ Many working hypotheses have been developed based on the research data.
These hypotheses are described in the Virgil C. Summer Safety Evaluation Report (USNRC, 1981), and will not be discussed here, except to say that some of i
these theories postulate that an earthquake the size of the Charleston, S.C.
1 earthquake of 1886 could recur in other areas of the Piedmont, Atlantic Coast'al Plain, and continental shelf in addition to the epicentral area.
e-s
-09/25/8A 2.5-2 VCGTLE SER INPUT SEC 2.5
,, 3.'
. [*
Because of the wide range of opinions within the scientific community con-cerning the tectonic mechanism for the Charleston, S.C. seismicity, the USGS clarified its position regarding the localization of the seismicity in the vicinity of Charleston, S.C., including the 1886 MMI X earthquake (Novem-
~ber 18, 1982 letter from James F. Devine, USGS, to Robert E. Jackson, NRC).
The NRC staff has formulated an interim position concerning eastern seismicity in general and Charleston, S.C. seismicity in particular (see Appendix A to this SER and March 2, 1983 memorandum from R. Vollmer to H. Denton). As part of future research efforts described in that position, the NRC staff is addressing the uncertainties about eastern seismicity by probabilistic studies funded by NRC and conducted by Lawrence Livermore National Laboratory (LLNL).
At the conclusion of these studies, the NRC staff will assess the need for a modified position with respect to specific sites.
In the interim, considering the speculative nature of most of the eastern seismicity hypotheses, the low probability of large earthquakes in the eastern U.S. and present knowledge of the geology and seismology of the region, the NRC staff considers the Vogtle design basis appropriate. The staff does not consider this issue an open item.
After careful review of the new information as provided and evaluated by the applicant, the staff' concludes that there is no basis for altering its con-clusions stated in the CP-SER concerning the safety of the Vogtle site.
The staff has evaluated the FSAR and subsequent documents and infor:,: tion including excavation and trench mapping, and the fault investigation report,
" Studies of Postulated Millet Fault" (Georgia Power Co., 1983).
The staff has concluded that the applicant has (1) performed satisfactory site and regional geologic and geophysical investigations, (2) reviewed all available pertinent literature, and (3) provided the staff with all information necessary to evalu-ate, assess, and support the. applicant's conclusions concerning the safety of the Vogtle site from the geologic and seismologic standpoint.
In addition, the staff finds the applicant has satisfied the requirements of and is in compliance with applicable portions of the following:
09/26/84 2.5-3 V0GTLE SER INPUT SEC 2.5
I (1) Appendix A to 10 CFR 50 (2) Appendix A'to 10 CFR 100 (3) SRP Sections 2.5.1, 2.5.2, and 2.5.3 (4) RG 1.70 " Standard Format and Content of Safety Analysis Reports for Nuclear Power Plants," Rev. 2 (5) those portions of RG 1.132, " Site Investigations for Foundations of Nuclear Power Plants," applicable to the development of geologic and seismologic information relevant to the stratigraphy, lithology, geologic history, and structural geology of the site (6) RG 4.7, " General Site Suitability Criteria for Nuclear Power Stations"
'(7) RG 1.60, " Design Response Spectra for Seismic Design of Nuclear Power g'-'y Plants" In-the following sections, the staff reviews the geologic and seismologic information and bases for its conclusions.
2.5.1 Basic Geology The three fundamental geologic concerns addressed in this OL review, in order to confirm the geologic-safety of the site, were:
(1) the large body of new information rapidly accumulating.in the south-eastern U.S., in the Piedmont and Coastal Plain, partly because of NRC-f.unded research, that has resulted in greater knowledge of the sub-surface and modifications of interpretations of the tectonic history of eastern North America; (2) ~the problematic, and almost ubiquitous, occurrence of clastic dikes in the
- j..
upper Eocene and lower Miocene strata of Georgia, South Carolina and North Carolina, the origin of which had not been investigated in depth; and 09/26/84-2.5-4 VCGTLE SER INPUT SEC 2.5 r
i
.?
(3) the possibility of a fault of unknown age postulated by Open-File Report 82-156 of the U.S. Geological Survey to occur seven miles south of the Vogtle site.
The conclusions reached for which the following sections provides the back-ground and justification are:
(1) No new information detrimental to the safety of the Vogtle site has been uncovered.
(2) Although the origin of the clastic dikes is still not demonstrated with certainty, their apparent great age, in the hundreds of thousands to millions of years, assures tha't they are not a safety concern to the plant site.
(3) Geological and geophysical investigations confirm that no fault is present at or near the site that has offset any strata younger thar. 40 million D,
t years old.
2.5.1.1 Regional Geology The Vogtle Plant is located within the Atlantic Coastal Plain province about 25 miles southeast of Augusta, Ga., and on an upland surface about 100 feet above and adjacent to the Savannah River. At the plant site Coastal Plain sediments range in thickness from 800 to 1,000 feet, consist predominantly of sandstone, shale, limestone, claystone, and marl, and range in age from Upper Cretaceous (138 to 63 million years before the present (mybp)) to Holocene (10,000 yBP to the present).
2.5.1.1.1 Stratigraphy Except for' river alluvium and gravels of Quaternary age (2 mybp-Present), the youngest, most extensively exposed formation in the region is the Hawthorn (or Altamaha) Formation, a red and yellow, thick-bedded sandy clay, of Miocene age (25-5 mybp).
09/26/84 2.5-5 VCGTLE SER INPUT SEC 2.5
, N,.-
Underlying the Hawthorn Formation is the Upper Eocene (55 to 38 mybp) Barnwell Group, with a basal soluble limestone, and a variety of rapidly changing assorted sandy and sandy limestone formations and facies that change rapidly both laterally and vertically.
The oldest exposed rock unit, the Blue Buff marl of the Lisbon Formation, is a' calcareous, silty, grayish fossiliferous unit, distinctive in lithology and
. fauna, and'of middle Eocene age (50-40 mybp).
The marl is limited in exposure in the banks of the Savannah River which has begun to cut down through it.
Because of its clay content, its density and compaction, the marl is the bearing stratum upon which the applicant chose to build the main power block.
The Eocene formations rest unconformably on Paleocene (63-55 mybp) clays and black lignitic sands which are also distinctive. These overlie the late Cretaceous (96-63 mybp) Tuscaloosa Formation unconformably.
The Cretaceous sediments are the oldest o,f,,the late Mesozoic (240-60 ' mybp) marine trans-gression deposits that constitute the Coastal Plain covermass.
Adjacent to and northwest of the Atlantic Coastal Plain Province is the Pied-mont Province. The boundary between the two provinces is known as the FaM Line which is approximately 25 miles northwest of the site.
Rocks that char-acterize the Piedmont disappear beneath the Coastal Plain sediments at the Fall Line,. but ne structurally signthcant boundary exists.
2.5.1.1.2 Structure and rectonics In several places in the Piedmont, west and north of the Coastal Plain, Tri-assic down-faulted basins filled with distinctive Triassic sedimentary rocks and some Triassic-Jurassic basaltic igneous rocks are exposed.
Similar basins have been recognized beneath the Coastal Plain. One such basin, the Dunbarton
' Basin, trending northeastward, has been identified on the basis of aeromagnetic 4
anomalies and drilling for the Savannah River Plant (Marine and Siple, 1974).
The Vogtle plant overlies the basin, close to the northern boundary. As this basin is downfaulted, presumably into Piedmont rocks, it is assumed that the rocks below and surrounding the basis are Precambrian (older than 570 mybp) j.
and early Paleozoic (570-240 mybp) metamorphics.
10/02/84 2.5-5 V0GTLE SER INPUT SEC 2.5 E.
t
.f 4
-At the surface, the Coastal Plain strata overlying the Piedmont and Triassic
^
basins, is gently-dipping towards the southeast, and, on the large scale, relatively undeformed, reflecting the relative tectonic q'uiescence of a passive continental margin at great distances from a lithospheric plate boundary, where most seismic, volcanic, and tectonic activity occur, according to the plate tectonics paradigm.
I Recent deep seismic reflection profiling (Cook, et al,1979)' has identified a large scale detachrant surface under the Applalachians, from the Allegheny Plateau to at least the central part of the Southern Piedmont, indicating a large allo m.nonous mass above the continental basement.
There is, however, no cerninty that the detachment continues under the Coastal Plain (Iverson and Smithson,1983).- Although it has been suggested that this detachment surface may be localizing seismicity (Seeber and Armbruster, 1982), this has not been demonstrated. As this is also one of the hypotheses regarding the source of the Charleston 1886 earthquake, it is being addressed in the Interim Charleston Position of the NRC (s'ee Appendix A of the SER).
/^-
Geophysical evidence has suggested the possibility that the Piedmont and its l
extension under the Coastal Plain is constructed of several discrete, litho-logically and geophysically distinct masses (or terrains).
These are postulated to have coalesced by accretion during middle and late Paleozoic tectonic events as the pre-Atlantic Iapetus Ocean closed, resulting in the continental collision of North America with Africa (Williams and Hatcher,1982). Although attempts to correlate modern seismicity with these terrain boundaries have been made (Wheeler and Bollinger, 1984) the correlation is not convincing and, therefore, this hypothesis has not gained much favor. Moreover, the Vogtle site is not near the postulated terrain boundaries and thus is not affected by this information.
A recent U.S. Geological Survey report (Prowell, 1983) catalogues faults of Cretaceous and Cenozoic age (63 to 2 mybp), primarily in the Coastal Plain and the Piedmont. The faults of those ages closest to the Vogtle site are the l
Belair en echelon faults at least 25 miles north of the site, and an unnamed pair of small faults forming a ten-foot wide graben in a kaolin mine about 27 miles 10/02/84 2.5-7 VCGTLE SER INPUT SEC 2.5 i
k-
north of the site, in South Carolina.
These faults offset upper Cretaceous sediments.
No seismicity has every been associated with these faults. There are no other faults listed that are near the site, or within 50 miles.
It is significant to note that the Millet Fault, postulated by the same author in an earlier report (Faye and Prowell,1982), is not included in the more recent catalogue of documented Cretaceou's and younge'r faults.
2.5.1.1.3 Clastic Dikes Dike-like structures are common and widespread in the younger Tertiary sedi-mentary strata of the Miocene (24-5 million years before the present (mybp)).
They are found in the upper Coastal Plain from Georgia to North Carolina in two upper Miocene formations, the Barnwell Formation and the overlying Haw-thorn Formation.
The conclusions drawn by the applicant from a limited reference and full investigation required by the staff is that the dikes do not represent a safety issue because they appear tc be very old, between 10,000 to 100,000 yrs. The staff concurs with this assessment and suggests the possibility of the dikes having formed close to 20 mybp.
The following provides some of the background and information that were the bases for the above stated conclusions.
The origin of the structures, primarily " clastic" dikes, some with associated faults and folds, is still not understood.
Because of the lack of detailed information, they have been proposed, among other possible causes, to have had a seismic origin, possibly related to the Charleston 1886 earthquake (Seeber and Armbruster 1983), the result of subsidence, differential compaction, weathering and soil formation, or infilled extension cracks in soil.
The dikes have been studied by several authors (Heron, et al,1971; Zupan and Abbott,1975; Secor, 1979; McDowell and Houser, 1983).
None of these, however, was a long-term systematic study to attempt to relate all the features, nor to determine the age, extent, and geometry.
i 10/02/34 2.5-3 VC3TLE SER INPUT SEC 2.5
^ ~
At'the request of the staff, the applicant did more reference work and recon-naissance field observations to provide further information, ia an effort to determine if the dikes represent a safety-related concern.
During the course of their study, a large waste disposal trench was dug within the Vogtle exclusion area, two miles south of the plant, for burying construction-related wastes.
The trench, about 900 ft. long, 75 ft. wide and 35 ft. high, exposed several interesting features, including subsidence sags, faults, dikes and small diapiric structures.
Along with_other features seen in the vicinity and within 30 miles of the plant, the trench exposure provided more information on some aspects of dike genesis in the area.
The applicant logged the trench and submitted a detailed report and analysis. A staff geoscientist toured the area with the applicant's consultants and examined the trench prior to the mapping.
The report describes several lithic units exposed in the trench, interpreted
( N.
to be the upper sands of the Barnwell, the lower massive sandy clay of the Haw-thorn which is truncated by erosion and overlain unconformably by a layer of i
fine gray aeolian sands.
The strata are warped into several folds, with the downwarps accompanied commonly by normal faults and graben-like structures, indicative of differential subsidance of the strata into voids below.
Narrow, relatively planar dikes of clay emerge from the Barnwell sands, commonly along the faults and, often, unrelated to faults.
The dikes appear to flare out upward in the dense, compact sandy clay, into a myriad of branches or distributaries, irregular and non planar in shape, sometimes somewhat vertical but curving to an arch over downwarps. The gray clay dikes are distinguished only by their light color that stands out against the red or yellow host strata.
In some areas along the trench, blobs and clumps of the same material as the dikes can be seen near the contact with the overyling sands.
The applicant reports channeling and weathering of the middle thick sandy clay dikeJbearing unit to a soil below the overlying gray sand.
The dikes appear l,..
to be truncated at the contact with the aeolian sand:.
l I
09/26/34 2.5-9
'iCGTLE SER INPUT SEC 2.5 i
i
Elsewhere in the site area, dikes often are fairly planar, much thicker than in the trench, to about 3-4 in. as opposed to 1-3 in. in the trench.
The most striking feature of the thicker ~ dikes outside the trench is the grading from a thin. clay center outward to a yellow sandy border and a deep red rind or crust bounding the dike. The red crust is resistant to the characteristic weather-ing of the host strata, and thus stands out in relief within the sandy units in which they are found.
A few exposures along the local roads show narrow dike-like projections in coarse sand units in which there appears to be no distinction except in color between the dikes and host stratum. The applicant reports distinctions in the clay content of the " dikes" and the host stratum.
These observations have led the applicant to conclude in the report that the dikes are primarily a weathering phenomenon in which groundwater has made its way along pre-existing fractures, bearing and depositing transported clays or leaching out chemicals to alter the character of the fracture zones. The altered fractures then appear to have been intruded by clastic or clay material.
In addition, the surface weathering, soil development, channeling, and trunca-tion of dikes at the aeolean sand-dike bed contact suggest to the applicant a great length of time from the formation of the dikes and the weathering and erosion of the upper part of the unit.
The estimate based on weathering alone is 10,000 to 100,000 yBP.
Althougn the staff does not agree with all aspects of the applicant's report, particularly the mode of formation of the dikes, it concurs that they are very old, probably having formed early in the development of the strata in which they are found, which are between 24-20 mybp.
As the dikes intrude sandy strata in narrow channels and flare outward into very dense clay layers, which is the reverse of what would be expected if they were liquefaction structures, it is suggested that the strata must have been very loosely consolidated or still saturated in order for the dike material to penetrate what is now almost impermeable clay layers.
Such a condition would exist in the early stages of sedimentation almost 20-25 mybp.
09/25/84 2.5-10 VCGTLE SER INPUT SEC 2.5
~
2.e.,
j f
f Based therefore on curient knowledge, and considering the likely great age of the dikes, there is no evidence that these features represent a safety issue I
for the plant, whatever their origin.
s:
l w
2.5.1.2 Site Geology It is the applicant's, view that extensive core drilling and mapping of the main power block excavatisi Ms pe;Mded evidence that the bearing stratum is sound, m.
lacking faults, solutionteg'; or any other geologic feature that may represent a safety concern.-
Although small depressions are observable in.several localities in and around the plant site, drilling has? sh'own they are the result of solutioning of the Utley Limestone, a thin, fossiHferous, ' basal unit of the Barnwell Group.
Overlying strata, such as that seen in the trench, have subsided into solution cavities. The applicant reporti that no evidence for solutioning in the Lisbon clay-marl bearing stratum below the' Utley Limestone has been 'found in drilling f.:n or in the excavations.
(
y Some reservations that the staff has regarding characteristics of the bearing stratum are addressed in Section 2.5.4 of this SER.
^
.4 e
).,
I 4
- (b
,s h
/"
s h
L k-%
b
+
t, P,
s s
ut '
/
,\\
.(
2.5.2 Vibratory Ground Motion The conclusion reached during the construction permit (CP) review by the staff, the staff's consultant, the U. S. Geological Survey and the U. S.
ArmyCorpsofEngineers,wasthat0.20g(SSE)and0.12g(OBE) accelerations were adequate.
During the Operating License (OL) review, the NRC staff seismological review was based on geologic and seismologic information in the Vogtle PSAR and FSAR and other available literature. The review concentrated on the following:
f*"3 1)
A review of the seismicity of the region and examination of the association of earthquakes with geologic and tectonic features.
2)
A determination of the vibratory ground motion at the site from the maximum historical earthquake within the tectonic province, and from recurrence of the 1886 Charleston, S. C. earthquake.
3)
A comparison of the ground motion estimated in (2) above with the SSE proposed for the site.
Our review indicates that those conclusions reached during the CP review regarding the adequacy of the SSE and OBE at Vogtle are still appropriate.
2.5.2.1 Seismicity
,-s
i' S,,
Bulletins of the southeastern U. S. Seismic Network describe the seismicity 'ince 1977 in the vicinity of'the Vogtle Electric Generating s
Plant (VEGP). Before that, most of what is known about the site region seismicity was based mainly,on intensity data.
In general, the seismicity within 50 miles of the site is very low. The maximum historical event within that radius is intensity IV.
Within 20,0 miles of the site the odly earthquake of epicentral intensity greater than or equallto VII was the Union County, South Carolina earthquake of January 1,1913. x This earthquake, which occurred in the southern Piedmont had an epicentral intensity of VII (Stover, et sal 1984) i' and was not felt.at the site of the VEGP.
In addition,to this earthquake, jn, larger earthquakes at distances greater than 200 miles were examine'd.
The New Madrid, M0 eprthquake sequence of 1811-1812 occurreri in the New Madrid Seismic Zone ab ut 530 miles NW of the VEGP and O cluded a maximum epicentralintensityofXI(Stoveretal,1984). Nuttli (1973) indicated that this earthquake was felt in Georgia with a maximum intensity of VI.
The Giles County, VA earthquake of May 31, 1897 occurred about 280 miles I
north of the site in the Valley and Ridge province and had an epicentral intensity VIII (Bollinger, 1973).
Bollinger indicated that intensity III may have been felt at the site. Another earthquake was the New Brunswick earthquake of January 9,1982, which was about 1250 miles frcm the Vogtle site and had a magnitude of 5.75 and an epicentral intensity of VI. This earthquake occurred in the Piedmont-New England Tectonic province.
3
, (%. '
s
. _s An earthquake of significance to the VEGP is the Modified Mercalli intensity (MMI) X (Stover et al 1984) at Charleston-Summerville South Carolina. This is the largest historic event along the eastern seaboard of the U. S. and occurred in a concentration of seismic activity in the a
Atlantic Coastal Plain province about 78 miles from the site. The intensity of this earthquake at the site was VI. Other earthquakes in the Atlantic Coastal Plain are discussed ir. Section 2.5.2.2 2.5.2.2 Tectonic Provinces and Maximum Historical Earthquakes The Vogtle site is located in the Atlantic Coastal Plain province. This
' province extends from the Fall Line (the southern boundary of the Piedmont Province) 25 miles to the northwest, to the edge of the continental shelf to the east and southeast.
Other tectonic provinces within 200 miles of the, site include, the Valley and Ridge province. the Blue Ridge and the Piedmont.- Other than the Coastal Plain province, the above provinces and all other provinces outside the 200 miles radius are at sufficient 1
distance so as not to have any impact on the Vogtle seismic design.
In the Southern Appalachian area, the staff has for the purpose of licensing treated the southern Piedmont as a separate area within the assumed l
L Piedmont-New-England tectonic province (i.e. McGuire, summer, Catawba l-SER's). Based on the available information it was not possible to relate L
past earthquakes to geological structures in the southern Piedmont or the Coastal Plain province.
Except for the Charleston, South Carolina area where the high seismicity cannot be considered typical of the rest of the jf
. -s Atlantic Coastal Plain, this province is characterized by low to moderate seismicity. The largest reported earthquakes in the Atlantic Coastal Plain province are the Asbury Park, N.J. earthquake of 1927 with MMI VII, and the Wilmingtcn, Delaware earthquake of 1871 with MMI VII. Therefore, an event similar to the MMI VII mentioned above should be considered as:the maximum historical earthquake likely to effect the site.
Based on the estimated felt area, the Asbury Park earthquake and the Wilmington earthquake (Kafka, 1980) had an estimated magnitude less than 5.0.
Nuttli and Herrmann (1978) indicated that an appropriate equivalent magnitude to an epicentral intensity of VII is a magnitude 5.3.
It is the staff's conclusion that the maximum random earthquake in the Coastal Plain
,ggs province can conservatively be defined or having an estimated magnitude of e-a
' asi 5.3.
The August 31, 1886 Charleston, South Carolina earthquake -is listed with epicentral intensity X.
The center of the area of maximum intensity was located near Middleton, S.C.
The Charleston-Summerville, South Carolina region is presently under investigation.
Interpretations that have been published so far regarding the cause of the Charleton earthquake differ considerably as far as the possible mechanisms are concerned.
-The staff current position, as in the past (V. Summer Nuclear Station, SER) that in accordance with the tectonic province approach (Appendix A to 10-CFR 100), the effects of a recurrence of an event the size of the lfh f'
, i. -- s
. Charleston earthquake in the Charleston-Summerville area shall be postulated to assess its influence on the VEGP. Additional discussion of the Charleston earthquake is found in Section 2.5.1 and Appendix (A) of this report.
2.5.2.3 Safe Shutdown Earthquake At the CP stage (SER-CP Suppl. 1) the staff co'ncluded that the maximum site intensity will be no greater than VII and the SSE acceleration of 0.2g used for the Vogtle units to be adequately conservative.
The staff position regarding the VEGP site is that the following seismic issues should be considered for the SSE design.
g.s
(;, y
, ;~
1.
The maximum randcm event in the Coastal Plain tectonic province, an event of MMI VII equivalent to mb = 5.3 (Nuttli and Herrmann,1978) in the vicinity of the site.
2.
An event of the size of the 1886 Charleston earthquake MMI X occurring in the vicinity of the Charlesten Sumerville area about 78 miles from the site.
Based on the tectonic province approach the staff finds that the maximum random event in the Coastal Plain was of MMI VII. The resulting mean value of peak horizontal acceleration at the site was estimated to be 0.13g (Trifunac and Brady, (1975).
In recent Safety Evaluation Reports
.(for example, Hope Creek, Millstone, Limerick) the staff has indicated P.'
~
~
~
~b~
that site specific spectra obtained from appropriate suites of earthquake strong motion records are more in accord with the controlling earthquake size, frequency content of response spectrum and local site conditions than are standard Reg. Guide 1.60 spectra.
In this method the use of the peak acceleration and Reg. Guide 1.60 spectrum shapes are replaced by spectra obtained from earthquakes within half a magnitude of the safe
. shutdown earthquake recorded at a distance less than 25 km with geologic conditions similar to those at the site.
It is the staff position that spectra obtained by this method is a more realistic estimate of the seismic ground motion for Vogtle. The spectra should be based on an appropriate ensemble of records with mbig = 5.'3
.5 obtained at a soil site within.25 km of the source. The staff position has been that the 84th
- V.n percentile spectrum.is appropriate for describing ground motion to be used in evaluating the design spectra of nuclear power plants.
Previous staff reviews of site specific spectra for soil sites (Hope Creek, NUREG, Wolf Creek, NUREG-0881, and Palo Verde, NUREG-0857) indicate that the Reg. Guide 1.60 spectrum anchored at 0.2g is adequate for describing ground motion for a magnitude 5.3 event.
With respect to the Charleston earthquake of 1886 Nuttli et al (1979) estimated the magnitude (m ) to be 6.6.
The distance between the Vogtle b
site and the meisoseismal area of the 1886 earthquake is 78 miles. Using ths equations derived by Nuttli (1983) and Campbell (1981) the staff g,,
4 g
.~
~
\\
7-estimated the 84th percentile accelerations at VEGP due to the reoccurrence of such an event to be less than 0.29 Based on consideration of both the local magnitude 5.3 and the reoccurrence at an earthquake the size fo the 1886 earthquake in the Charleston area, the staff considers the Reg. Guide 1.60 response spectrum anchored at O'.2g used for the design of Vogtle to be acceptable.
(
I 2.5.2.4 Operating Basis Earthquake '
The applicant has proposed 0.12g for the acceleration level corresponding
. cags to the operating basis earthquake. This represents more than half of the
' (',q
'3/
SSE acceleration 0.20g, consistent with Appendix A to 10 CFR 100 which indicate that the OBE to be at least half of the SSE.
2.5.3 Surface Faulting 2.5.3.1 Postulated Millet Fault For the construction permit, the applicant's geological investigations concluded that there was no-surface faulting in the vicinity of the Vogtle site.
Prior to submission of the FSAR for the OL review, a document released by the U.S. Geological Survey, Open-File Report (OFR)82-156, (Faye and Prowell,
^
1982) postulated the existence of a fauit, the Millet Fault, seven miles south 10/0 /84 2.5-11 V0GTLE SER INPUT SEC 2.5 p,
I as*
3
of the site. Although the report was a water resources study of the hydrology and geology of the Coastal Plain in the vicinity of the Savannah River, fr. direct evidence from secondary sources suggested to the authors the possibility of a
-fault across the Savannah River.
The report further postulated a second fault, the Statesboro Fault, 32 miles south of the plant.
'As interpreted by Faye and Prowell, the Millett Fault trends northeastward across the Savannah River, is approximately 40 miles long, and has vertically offset the buried Tria'ssic/ Cretaceous contact +600 ft. on the southeast side of the fault.
The main evidence for the inferred fault came from a comparison of well cut-tings taken several years before the OFR study from two water wells, P5R and AL66..s interpretation of the lithic fragments suggested that Triassic rocks were present below -1100 ft in PSR, but above 600 ft in AL 66 four miles to A
the south.
Further, an examination of surface and groundwater flow records (j
over a period of forty years indicated some anomalous characteristics which Faye and Prowell interpreted as resulting from a subsurface barrier.
By extrapolation from the Belair Fault 35 miles to the north, the OFR inferred that an impermeable gouge zone above the postulated Millett Fault forced the south flowing groundwater from a lower aquifer to a higher one on the south side of the fault. As the trace of the postulated fault traverses a segment of the Savannah River where a straight stretch of the river changes to a more characteristic meandering flow pattern, the OFR considered this observation further support for the inferred fault.
Evidence for the trend and length of the postulated faults was by extrapola-tion from other post-Cretaceous Coastal Plain faults in the southeastern United States.
While no age of faulting was suggested, the OFR indicated that rocks at least throu'gh the Eocene epoch (55-8 mybp) were involved.
-09/25/34 2.5-12
'/CGTLE SER INouT SEC 2.5
, _ ~
2.5.3.2 Fault Investigation 2.5.3.2.1 Introduction At the request of the NRC staff, the applicant undertook a detailed investi-gative program, as the age and therefore capability of the inferred faults were undetermined.
In October,1982, Georgia Power Company submitted to the NRC a report of a fault specific investigation entitled, " Studies of Postulated Millett Fault" (Georgia Power Company, 1983). The postulated faults, the Millett and States-boro Faults, were the subjects of the utility's fault investigation, with the primary focus on the former which was closer to the plant.
Techniques Used in the Investigation:
SCS utilized a wide range of techniques to explore the surface and subsurface for both geologic and hydrologic information in order to locate and date the fault.
Included in the study was (1) field geologic mapping, (2) aerial and Landsat imagery for remote sensing analysis, (3) core drilling on both sides of the Savannah River and straddling the interval of the two wells described in the open file report, (4) petrographic, x-ray and heavy mineral analyses of core samples, (5) downhole geophysical studies including gamma, neutron and electric logging, (6) seismic reflection profiling, (7) regional geophysical a
and seismicity studies, (7) well water level monitoring, (8) groundwater modelling, and (9) analysis of surface water flow.
Staff Review:
Because of the wide range of techniques used, several NRC staff reviewers have contributed to this evaluation of the utility's report; a geologist covering the varied geologic investigations, a seismologist for the historic and pre-sent day seismicity of the area, a geophysicist for the seismic reflection prof'iling, and a hydrologist reviewing the surface and groundwater study.
In addition, the geologist and one of the geophysicists have visited the site
. m.s region with the NRC project manager and SCS staff to examine the cores drilled 09/26/84 2.5-13 VCGTLE SER INPUT SEC 2.5
for this study, the stratigraphy, and various aspects of the surface features
- ~,
in order to have first hand experience in evaluating the fault investigation report.
==
Conclusions:==
The conclusion arrived at by the applicant, based upon the results of the study, is that there is no evidence for a capable fault; and that if there is a fault, which could not be detected by any of the techniques used by the
~
applicant, it is older than 40 mybp. The staff agrees that this conclusion is consistent with the reported information and results of the various investi-
.gative' techniques.
2.5.3.2.2 Summary of Fault Investigation and Results A brief summary of these results, the applicant's views, and the bases for the applicant's and staff's conclusions follow.
2.5.3.2.2.1 Geologic Investigation:
The geologic investigation included (1) field mapping and remote sensing to identify evidence for surface faulting; (2) core drilling with petrographic, x-ray and heavy mineral. analysis of the strata in the cores in order to correlate layers from core to. core to determine any offset of the strata; (3) downhole geophysical logging, which identifies individual strata by characteristic signa-tures that are dependent upon the physical properties of the rock units, also with a view to correlating the strata from core to core; and (4) review of other geophysical studies.
1.
Field mapping and remote sensing techniques failed to uncover any evidence
.of surface faulting, or linear features indicative of surface or near surface rupture.
Staff review of some of the original imagery used for the study and field checks in the site area verified this conclusion.
2.
Twelve drill cores were taken along two parallel north-south lines, cross-ing the inferred trace of the Millett Fault, one in South Carolina close 09/26/34 2.5-14 VCGTLE SER INPUT SEC 2.5
-n~,,
to the two wells studied by Faye and Prowell, and one in Georgia. Along both lines, one core was taken from the location closest to the trace of the fault and the others equally spaced north and south of the fault.
Eight holes were cored in Georgia, about one mile apart north-south, and four in S.C.
Visual examination, petrographic and mineralogic analyses identified distinctive marker beds. One of these, the Blue Bluff marl, appeared consistently between 100 ft above sea level to 100 ft below, showing no change in elevation (other than that due to the gentle regional dip to the south) on either side of the postulated fault in Georgia and South Carolina. This indicates that-no vertical offset and therefore no fault of the type postulated in the USGS open file report is present down to strata at least 40 my old at the inferred location of the postulated fault.
Furthermore, Core VSC-4, on strike with, and 200 ft. from, well AL66 in which the USGS interpreted Triassic rocks at an elevation of -600 ft, was the deepest hole in the study. The core at -1000' was still in distinctive Cretaceous Tuscaloosa sands.
The staff examined this core and agrees that it has none of the characteristics of Triassic rocks and c]
looks much like other Cretaceous samples..This result is in direct
_ disagreement with the open-file report interpretation upon which the fault is postulated.
3.
Downhole geophysical logs, especially the gamma log, provided distinctive signatures that verified the petrographic identification of the strata, and in particular the Blue Bluff marl, confirming the continuity of the unit across the inferred location of the Millett Fault.
2.5.3.2.2.2 Seismic Reflection Study:
A nineteen mile acoustical seismic reflection survey was conducted in the Savannah River in the vicinity of the postulated Millett Fault. The survey used three different energy sources, Uniboom, 10 cubic inch air gun and 20 cubin inch air gun. to obtain high resolution and deep penetration of the sub-surface formations.
The water depth in which the survey was carried out ranges between 10-25 feet.
k 09/26/84 2.5-13 VCGTLE SER INPUT SEC 2.5
--a
-s r~.
e-
-.v-r,.
e y *
,.i The reflection survey identified staeral key horizons, A through I, at differ-ent depths ranging between +70 to -1150 feet.
Some of the horizons, such as reflector E, which correlates with the top of the limestone that is the lowest unit in the Barnwell group, of late Eocene age (40 MPBP), and reflector G, which represents the unconformable contact between the unnamed sands of the Middle Eocene Lisbon Fornation and the top of the Paleocene kaolinitic clay (60. mybp) are well correlated with adjac'ent core holes.
Some of the reflectors are well defined while others are weakly defined. Some of the horizons showed small features which may be indicative of past. channelization or buried karstic surfaces. The continuity of the reflectors above the Triassic / Cretaceous contact and the absence of noticeable displacement in the higher horizons above -500 feet elevation indicate the absence of faulting within the last
~
60 million years in the vicinity of the postulated Mil *lett Fault.
This con-clusion is in agreement with that of the applicant's report, that no capable fault has been identified by the seismic reflection data obtained for this study.
In addition, seismic reflection data obtained by the applicant from the Savannah
('N, River Plant investigation of the deeper horizons in the vicinity of both plants suggest the possibility of a small normal offset of the Triassic / Cretaceous contact of 50-100 ft in the vicinity of the postulated fault.
No evidence, however, for offset in younger horizons can be detected.
2.5.3.2.2.3 Seismology Study:
The available seismicity information includes (1) felt earthquakes, (2) recent instrumentally located events, and (3) data from the Savannah River Plant array, just across the Savannah River from the Vogtle site. The applicant concludes that historic seismicity reveals no evidence of active faulting in the a a.
msw.
This conclusion is consistent with the data.
The seismicity.raar the site has been scattered and low level-(maximum MM intensity [ No clustering of earth-quakes is occurring near the postulated Millett or"Statesboro Faults.
2.5.3.2.2.4 Hydrology Study:
Faye and Prowell used, as part of their case, several hydrologic arguments to support the existence of the postulated fault.
These' arguments were thoroughly 09/26/84 2.5-16 V0GTLE SER INPUT SEC 2.5 g
- -, - - ~
~--s p
7 r
,i
^
investigated in this study by the applicant.
The issues of different unit base flows in river reaches (generally above and below the postulated fault), water well levels across the postulated fault, and groundwater piezometric surface contours were addressed by their study. After carefully reviewing all the 4
contractor's hydrologic evaluations, the staff concurs in their conclusion that the hydrologic data provide "no basis to support or preclude the existence of a fault."
2.5.3.2.2.5
==
Conclusions:==
On the basis of (1) the continuity of strata across the inferred location of the postulated Millett Fault as determined by (a) drill cores, (b) downhole geophysical logging and (c) seismic reflection profiles, (2) the absence of Triassic rocks at levels above -1000 ft, as shown by drill core VSC-4 on the south side of the postulated fault, (3) the scattered and low level seismicity and (4) the hydrologic information which neither indicates nor disproves the presence of a fault, the applicant's report concludes that there is no evi-g'ETg dence for the existence of a capable fault'in the vicinity of the Vogtle plant.
\\./
The staff has carefully reviewed the report, has visited the site and examined the cores, the logs, and remote sensing imagery, and field checked the surface features. The reviewers consider the applicant's conclusion to be consistent with the data as reported, and conclude, therefore, that no capable fault as defined in Appendix A to 10 CFR Part 100 is present in the vicinity of the Vogtle plant, based on all presently available data.
Further support for this conclusion comes from Prowell's later (1983) report on Cretaceous and younger faults of eastern North America. The Millett and Statesboro Faults are not included as documented faults.
It is concluded, therefore, that no surface faulting capable of localizing earthquakes is present at the plant site or in the site vicinity.
09/26/34 2.5-17 VCGTLE SER INPUT SEC 2.5
4 a
REFERENCES 6
Bollinger, %
A., " Seismicity of the Southeastern United States," Bull. Seism.
4.
A Soc. Amer., v. 63, Oct. 1973.
Campbell, K. W. "A Ground Motion Model for the Central U.S. Based on Near-Source Acceleration Data," Prof. Conf. On Earthquakes and Earthquake Engineering, The
-Eastern U.S., v. 1, pp. 365-376, 1981.
Cook, F. A. et al.,1979, Thin-skinned tectonics in the crystalline Southern Appalachians, C0 CORP Seismic Profiling of the Blue Ridge and Piedmont, Geology,
- v. 7, pp. 563-567.
Faye, R. E. and Prowell, D. C., 19.82, Effects of Late Cretaceous and Cenozoic faulting on the geology and hydrology of the Coastal Plain near the Savannah River, Georgia and South Carolina:
U.S. Geological Survey Open-File Report 82-156, p. 73.
Georgia Power Company, Studies of Postulated Millett Fault, Vogtle Electric Generating Plant, October, 1982.
Heron, S. D., Jr., Judd, J. B., and ' Johnson, H. S., Jr. ; Clastic Dikes Associated b'
with Soil Horizons in the North and South Carolina Coastal Plain," Geological Society of America Bulletin, v. 82. pp. 1801-1810, 1971.
s Iverson, W. P. and Smithson, S.,
1983, Reprocessed C0 CORP Southern Appalachian reflection data:
Root zone to Coastal Plain, Geology, v. 11, p. 422-425.
Kafka, A.
L., " Earthquake Hazard Studies in New York State and Adjacent Areas,"
Lamont-Doherty Geological Observatory of Columbia University, Palisades, N.Y., 1980.
Marine, I. W. and Siple, G. E.,1974, Buried Triassic basin in the central Savannah River area, South Carolina and Georgia: Geological Society of America Bulletin, v. 85, no. 2, pp. 311-320.
McDowell, R. C., and Houser, B.B., Map Showing Distribution of Small-Scale Deformation Structures in a part of the Upper Coastal Plain of South Carolina and Adjacent Georgia," U.S. Geological Survey Miscellaneous Field Studies Map MF-1538, 1983.
Nuttli, 0. W., "The Mississippi Valley Earthquakes of 1811 and 1812.
Intensities, Ground Motion and Magnitude," Bull. Seism. Soc. Amer., v. 63, pp. 227-248," February 1973.
Nutt-li, O. W., R. Rodriguez, and R. B. Herrmann, " Strong Ground Motion Studies for South Carolina Earthquakes, NUREG/CR-3755, November 1983.
10/02/84 i
VCGTLE SER I"PUT REFERENCES
~
. +
Nutt11, O. W., G. A. Bollinger, and D. W. Griffith, "On the Relation Between
-Modified Mercalli Intensity and Body Wave Magnitude," Bull. Seism. Soc. Amer.,
- v. 69, 1979.
Nuttit, O. W., R. Herrmann, " State of the Art for Assessing Earthquake Hazards in the United States," U.S. Army Corp of Engineers, WES Report #12, 1978.
Prowell, D. C., 1983, Index of faults of Cretaceous and Cenozoic age in the
. eastern United States,.U.S. Geological Survey Map MF-1269.
1 Secor, D. T., " Geological Investigation at the Chem-Nuclear Waste Storage Site, Barnwell, South Carolina, February 24, 1980." Allied-General Nuclear Services.
Seeber, L. and Armburster, J. G.,1981, The 1886 Charleston, South Carolina earthquake and the Appalchian detachment, Journal of Geophysical Research,
- v. 86, no. 39, pp. 7874-7894.
Siple, G. E.,." Geology and ground water of the Savannah River Plant and 7
Vicinity, South Carolina," U.S. Geological Survey Water Supply Paper 1841, P. 57-60,'1967.
Stover, C. w., B. G. Reagor, and S. T. Algermissen, United States Earthquake Data File, U.S.G.S. Open-File Report 84-225, 1984.
Trifunac, M. D. and A. G. Brady, On the Correlation of Seismic Intensity Scales with-the Peaks of Recorded Strong Ground Motion," Bull. Seism. Soc. Amer.,
-s
.(
3
- v. 65, 1975.
U.S. Nuclear Regulatory Commission, Safety Evaluation Report, Alvin W. Vogtle Nuclear Plant Units 1, 2, 3, and 4, March 8, 1974; Supplement No. 1, May 1, 1974.
USNRC, " Safety Evaluation Repurt Virgil C. Summer Nuclear Plants," Docket No. 50-395, NUREG-0717, 1981.
USNRC, " Safety Evaluation Report Related to the Operation of Palo Verde Nuclear Generating Station, Unit No.
1," NUREG-0857, 1982.
USNRC, " Safety Evaluation Report, Hope Creek Generating Station, Units 1 and 2," Docket Nos. 50-354/355, NUREG-
, 1984.
USNRC, " Safety Evaluation Report Related to the Operation of Wolf Creek Generating Station, Unit No. 1," NUREG-0881, 1982.
Wheeler, R. L. and Bollinger, G.
A., 1984, Seismicity and suspect terranes
.in the southeastern United States Geology, v. 12, p. 323-326.
Williams, H. and Hatcher, R.
D., Jr.,1982 Suspect terranes and accretionary history of the Appalachian Orogen, Geology, v.10, p. 530-536.
Zupan, A. J. W., and Abb'ott, W. H., " Clastic Dikes" Evidence for Post-Eocene
(?) Tectonics in the Upper Coastal Plain of South Carolina," South Carolina i
State Development Board, Division of Geology, v. 19, pp. 14-23, 1975.
10/02/84 2
VCGTLE SER INPUT REFERENCES
- /
?
APPENDIX A Interim Position on Charleston Earthquake for Licensing Proceedings The NRR Staff position with respect to the Intensity X 1886 Charleston earth-quake has been that, in the context of the tectonic province approach used for Itcensing nuclear power plants, this earthquake should be restricted to the Charleston vicinity. This' position was based, in part, on information provided by the United States Geological Survey (USGS) in a letter dated December 30, 1980 from J. E. Devine to R. E. Jackson (see Summer Safety Evaluation Report).
The USGS has been reassessing its position and issued a clarification on November 18, 1982 in a letter from J. E. Devine to R. E. Jackson. As a result of this letter, a preliminary evaluation and outline for NRC action was forwarded to the Commission in a memorandum frem W. J. Dircks on November 19, 1982.
The USGS letter states that:
ps "Because the geologic and tectonic features of the Charlecton region are similar to those in other regions of the eastern seaboard, we conclude that although there is no recent or historical evidence that other regions have experienced strong earthquakes, the histor-ical record is not, of itself, sufficient grounds for ruling out the occurrence in th'se other regions of strong seismic ground motions e
similar to those experienced near Charleston in 1886. Although the probability of strong ground motion due to an earthquake in any given year at a particular location in the eastern seaboard may be very low, deterministic and probabilistic evaluations of the seismic hazard should be made for individual sites in the eastern seaboard to establish the seismic engineering parameters for critical facilities."
The-USGS clarification represents not so much a new understanding but rather a more explicit recognition of existing uncertainties with respect to the causative structure and mechanism of the 1886 Charleston earthquake. Many hypotheses have 09/26/5A 1
VGGTLE SER INPUT APP A
l" been proposed as to the locale in the eastern seaboard of future Charleston-size earthquakes.
Some of these could be very restrictive in location while others would allow this earthquake to recur over very large areas.
Presently none of f
these hypotheses are definitive and all contain a strong element of speculation.
We are addressing this uncertainty in both longer-term deterministic and shorter-term probabilistic programs. The deterministic studies, funded primarily by the i
Office of Research of the NRC should reduce the uncertainty by better identifying -
(1) the causal mechanism of the Charleston earthquake and (2) the potential for the occurrence of large earthquakes throughout the eastern seaboard.
The probabi-listic studies, primarily that being conducted for NRC by L'awrence Livermore National Laboratory (LLNL) will take into account existing uncertainties. They willhaveastheiraimtodeterminedifferences,ifany,bEtweentheprobabi-lities of seismic ground motion exceeding design levels in the eastern seaboard (i.e. as affected by the USGS clarified position on the Charleston earthquake) i and the probabilities of seismic ground motion exceeding design levels elsewhere in the central and eastern U.S.
Any plants where the probabilities of exceeding design level ground motions are significantly higher than those calculated for
]
other plants in the Central and Eastern U.S. will be identified and evaluated for possible further engineering analysis.
Given the speculative nature of the hypotheses with respect to the recurrence of large Charleston-type earthcuakes as a result of our limited scientific knowledge and the generalized l'ow probability associated with such events, we 4
do not see a need for any action for specific sites at this time.
It is our position, as it has been in the past, that facilities should be designed to withstand the recurrence of an earthquake the size of the 1886 earthquake in the vicinity of Charleston. At the conclusion of the shorter-term probabi-listic program and during the longer-term deterministic studies, we will be assessing the need for a modified position with respect to specific sites.
f' 09/26/84 2
VCGTLE SER INPUT APP A
.