ML19345D582
| ML19345D582 | |
| Person / Time | |
|---|---|
| Site: | Summer  |
| Issue date: | 12/12/1980 |
| From: | Nichols T SOUTH CAROLINA ELECTRIC & GAS CO. |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8012160123 | |
| Download: ML19345D582 (44) | |
Text
. R: H SOUTH CAROUNA ELECTRIC a GAS COMPANY CotunsiA, Scuin CAnounA aseia
- m CEC Ifi /M 12 33 T.C.Neewots Jm.
u.
2 - o.
December 12, 1980
%.e iiCIIERVICES
~"
- GCH Mr. Harold R. Denton, Director Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Con
- ::iission Washington, D. C. 20555
Subject:
Virgil C. Summer Nuclear Station Docket No. 50/395 Seismology Questions
Dear Mr. Denton:
South Carolina Electric and Gas Company, acting for itself and as agent for South Carolina Public Service Authority, herewith forwards forty-five (45) copies of revised pages to responses to 361. series questions previously transmitted to the Staff by letters dated 9/5/80 and 10/8/80. These responses are being submitted by letter in order to expedite the Staff review and will subsequently he incorporated in the FSAR by amendment.
Very truly yours,
/
T. C. Nichols, Jr.
4 RBW:TCN:rh cc:
V. C. Summer G. H. Fischer T. C. Nichols, Jr.
E. H. Crews, Jr.
O. W. Dixon, Jr.
O. S. Bradham D. A. Nauman R. B. Clary W. A. Williams, Jr.
B. A. Bursey J. L. Skolds J. B. Knotts W. F. Kane A. J. Murphy R. E. Jackson D. Cash P. Sobel NPCF/Whitaker File s012160 M b
VIRGIL C. SUMMER NUCLEAR STATION - FINAL FSAR REVISIONS TO RESPONSES TO 351. SERIES QUESTIONS (DECEMBER 1980)
REMOVE INSERT Pages 361.14-1 Pages 361.14-1 361.14-2 361.14-2 361.14-3 361.14-3 361.14-4 361.14-4 361.15-1 361.15-1 361.15-2 361.15-2 361.16-1 361.16-1 361.17.1-1 361.17.1-1 361.17.1-2 361.17.1-2 361.17.2-3 361.17.2-3 Figures 361.17.2-1g (p. 361.17.2-10)
Figures 361.17.2-19 (p. 361.17.2-10) 361.17.2-lh (p. 361.17.2-11) 361.17.2-lh (p. 361.17.2-11) 361.17.2-2g (p. 361.17.2-18) 361.17.2-29 (p. 361.17.2-18)
Pages 361.17.3-1 Pages 361.17.3-1 361.17.4-2 361.17.4-2 361.17.4-3 361.17.4-3 361.17.4-3a 361.17.4-8 361.17.4-8 361.17.4-8a 361.17.4-8a 361.17.4-8b 361.17.4-11 361.17.4-11 361.17.4-12b 361.17.4-12b 361.17.4-13 361.17.4-13, Figures 361.17.4-20a 361.17.4-20b 361.17.4-21a 361.17.4-21b 361.17.4-22a 361.17.4-22b 361.17.4-23a 361.17.4-23b Pages 361.18-4 Pages 361.18-4 361.18-6 361.18-6 Figures 361.18-2 (p. 361.18-12)
Figures 361.18-2 (p. 361.18-12)
Pages 361.19-2 Pages 361.19-2 361.19-4 361.19-4 361.19-4a 361.19-5 361.19-5 361.20-3 361.20-3 361.21-2 361.21-2 361.21-2a 361.21-4 361.21-4 Figures 361.21-2 (p. 361.21-6)
Figures 361.21-2 (p. 361.21-6)
361.14 With respect to FSAR Question 361.7-2, the microzonation of seismicity associated with Monticello Reservoir comprises three major zones (Teledyne Geotech Technical Report 79-8, pages 11-17):
(a) single zone near the north end of the reservoir; (b) an east-west zone, containing possibly four subzones, across the central part of the reservoir; and (c).
a zone near the end of the reservoir consisting possibly of 1
two subzones. Produce composite fault plane solutions of l
each zone (or subzone, where possible).
Include first motions l
derived from MEQ-800 portable seismographs and the six-station USGS network (Talwani, Induced Seismicity and Earthquake Prediction Studies in South Carolina,1979e, page 16).
RESPONSE
Before the installation of the six-station USGS network, hypocentral locations were obtained routinely using the four-station SCE&G network together with data from JSC and MEQ-800 portable seismograph (as and when deployed). These data were also used to obtain composite fault plane solutions of events sub-divided according to their locations and depths. After the installation of the USGS network and its incorporation with the SCE&G network, data from 10 channels were recorded on magnetic tape. This fonns the Monticello seismo-graphic network.
l To compare the hypocentral locations of events recorded on the Monticello network (tape recording) with those on the SCE&G network (helicorder recording),
about 100 events recorded on both were selected. The epicentral locations obtained from tne two recordings agreed well (latitudes within
+0.3 (s500m) l 361.14-1
i for 78% events, longitudes within 10.3' (s450n) for 72% events). However only 53% of the depths agreed within 1,0.6 km.
As the angle of incidence (used in the fault plane solutions) critically depends on the hypocentral depths, for detailed construction of fault plane solutions, only events recorded on the Monticello network were used.
RESULTS The various clusters discussed below had been well defined by the end of 1978, since most of the spreading occurred 6 months after impoundment (1977),
and 90% of the epicentral area was covered in the first year. Therefore, the July-December 1978 data, covering all observed clusters, comprise an excellent representative sample of the total observed seismicity. The 1979 data have been reduced, incorporated into the total site seismicity, and are discussed in detail in Appendices II and IV of " Evaluation of Site-Specific Sei:micity at the Virgil C. Summer Nuclear Station," as formally submitted on the docket in December 1980. However, the basic conclusions presented here are unaffected by the 1979 data sample.
From the USGS facilities in Golden, Colorado, visual playbacks were obtained 4
for the data recorded on the 10-station Monticello network, Using thess high-I quality data, it was possible to divide the seismicity into five clusters (Figure 361.14-1). At taast one composite fault plane solution was obtained for each cluster. These are summarized in Table 361.14-1. One solution each was obtained for Clusters I and V, two for Cluster III, three for Cluster II t
i
[
361.14-2 l
n.
and four for Cluster IV. Except for solutions 3 and 4 for Cluster IV, all composite fault plane solutions are for events with depths ranging from 0 to I km. All fault plane solutions indicate that thrust faulting is the predominant mechanism. Some events, especially the deeper events (1-2 km) (Cluster IV, Solutions 3 and 4) exhibit a component of strike-slip motion.
Two predominant orientations of the nodal planes, N-S and fM-SE were noted.
The P axes, as would be expected for thrust faults, are predominantly close to horizontal (Table 361.14-1). However their azimuth is not consistent.
The slip vectors were obtained for both nodal planes for each of the fault plane solutions. These are summarized in Table 361.14-2. A persual of Tables 361.14-1 and 361.14-2 indicated that the orientations of the P axes and of the slip vectors were related to the rock units the events were associated with. So they were separated on the basis of their geologic association (Tables 361.14-3). The geology of Cluster IV is uncertain, and hence the rock unit has been labelled " mixed."
In Group A (Table 361.14-3), i.e., when the epicenters are located on the l
country rock (granofel, CBGN) or on the intrusive rock (granodforite), the P axes (azimuth measured clockwise from north) are remarkedly consistent, 0
and lie between 80 and 92. Two possible slip vectors are indicated, one I
0 0
0 90 -1200 and the other 225 -265, both indicating a northerly set striking striking fault plane. Other data were examined to decide between the two.
l 361.14-3 1
i l
i j
The events comprising Group B (Table 361.14-3) are in shallow migmatite and l
probably reflect local orientation of fractures. The spatial association of different fault plane orientations is summarized in Figure 361.14-2.
0 0
In Group C (Table 361.14-3), consistent P axes directions (53 -66 ) (in-dicating a NW-SE fault plane) were obtained, but the calculated slip vectors are not consistent. Two solutions included in Group C are for deeper events.
t j
SUMMARY
OF RESULTS j
1.
The results obtained from an analysis of data recorded on the Monticello i
network for the period July-December 1978, are representative of the total observed seismicity, i
I 2.
Several fault plane solutions were obtained for the different clusters.
All of them indicate thrust faulting as the mechanism.
l 3.
Although there was some variation in the orientation of the nodal planes, two predominant sets of fault plane solutions were obtained. The nodal planes are oriented predominantly N-S and NW-SE.
4.
The P (compressional) axis is usually taken to represent the direction of i
maximum stress. Here two sets of consistent P axes were obtained. As there should be only one direction of maximum REGIONAL TECTONIC stress in so small an area, the conclusion can be drawn is that one or both inferred directions of maximum regional tectonic stress are erroneous. This observation has three possible implications:
361.14-4
,......m
,. ~...,,.., _...,. _... _ -,...
,-_.--,,.___.-m-..,,-..
,,----,r.-.-,,,.m...-..-_..,-,--.,,v-.,
~.m,
361.15 Discuss the relationship between the stress field determined from the fault plane solutions and (a) local (<150 km) structural / tectonic geology and (b) the stress field determined in the two USGS deep wells located west and southwest of the i
reservoir (Talwani, Induced Seismicity and Earthquake Prediction Studies in South Carolina,1979e, page 21).
RESPONSE
As discussed in response to Question 361.14, the inferred direction of maximum horizontal stress (a)) at hypocentral depths is NE-SW. The local ( 150 km) structural / tectonic geologic framework has a pronounced NE-SW orientation.
The NE-SW geologic grain would suggest chat, at the time of fonnation of those rocks, the direction of maximum horizontal stress would be NW-SE. Thus, the stress field inferred from fault plane solutions is representative of the present day stress field, whereas that inferred from tha local geology probably would represent a paleostress field.
In the two USGS wells no stress orientations are available from the hydrofracture data (Zoback, pres. com). Other features of these measurements are addressed in response to Question 361.21, and are discussed in Appendix III of " Evaluation of Site-Specific Seismicity at the Virgil C. Sumer Nuclear Station," as fonnally submitted on the docket in December 1980.
Regionally, orientation of the principal stresses has been obtained from fault plane solutions at Lake Jocassee (about 170 km to NW of Monticello Reservoir). Stresses have also been measured at Bad Creek (about 10 km NW of Jocassee Dam) by hydrofracture (Haimson, 1975) and by overcoring (Schaffer et a_1., 1979). These results are summarized in Table 361.15-1.
361.15-1
Table 361.15-1 Average Principal Stress Values Depth Below ogMin oHN*
Location Surface bars Direction bars Direction Method 0
0 Bad Creek 236 159+25 N30 W 228+55 N60 E Hydrofracture Bad Creek 181 184 N320W 293 N57 E Overcoring 0
Lake Jocassee
< 2 km NW NE Fault Plane Solutions Monticello
< 2 km NW NE Fault Plane Reservoir Solutions Thus, the stress directions inferred from fault plane solutions at Monticello Reservoir are in agreement with those in the vicinity of Lake Jocassee which were inferred from fault plane solutions, and with those obtai.1ed from over-coring and hydrofracture measurements in the Bad Creek well. All these data indicate that ej (the largest principal stress)is horizontal and oriented NE-SW, c2 is also horizontal and oriented NW-SE, and a3 (the least principal stress) is vertical. A more detailed analyses of composite fault plane solutions, in situ stress measurements, and tectonic and structural geology is contained in Appendixes III and IV of " Evaluation of Site-Specific Seismicity at the Virgil C. Summer Nuclear Station," as formally submitted on the docket in December 1980.
REFERENCES Haimson, B.D., 1975, Hydrofracturing stress measurements, Bad Creek Pump Storage project, Report for Duke Power Company,19 p.
Schaeffer, MF., Steffens, R.E., and Hatcher, R.D., 1979, In Situ stress and its relationship to joint formation in the Toxaway gneiss, Northwestern South Carolina, Southeastern Geology, M, 129-143 p.
361.15-2
361.16 On the basis of the more recent seismicity reports by Teledyne Geotech (1978a,b,c,d; 1979a,b,c) and Talwani (1979a,b,c) for Monticello Reservoir, update the discussions of the spatial and temporal distribution of hypocenters and their relationship to the local (<150 km) strtctural/ tectonic geology.
RESPONSE
The monthly epicentral locations have been provided in the various quarterly seismicity reports and are not presented here. Details of epicentral migration are discussed in response to Question 361.17.2, and in Appendix II of "Evalu-ation of Site-Specific Seismicity at the Virgil C. Summer Nuclear Station,"
as formally submitted on the docket in December 1980.
The pattern of seismicity noted in the earlier reports appears to be continuing.
The seismicity appears to be related to the local (<10 km radius) geology rather than to any regional feature. The seismicity still appears to be related to existing features as inferred from an analysis of fault plane solutions. Most of the seismicity is still located in 3 broad bands (Figures 361.16-1 and 361.16-2).
After the initial spure of seismicity following impoundment, there has been a marked decrease in the seismicity. The decreased level of seismicity was interrupted by discrete swarm episodes, the most prominent occurring in October 1979. The bulk of the larger (ML " 2.0) events occurred in 3 swarms, i.e., 26 events in January-February 1978, 9 in October-November 1978 and 19 in December '79, 54 or 77% occurred in 3 swarms. After the initial seismicity associated with the impoundment, the cause of the other swarms is not very clear. However, there has been no noticeable change in the magnitude of the largest events in any of the swarms (<M 2.8).
L 361.16-1
' 361.17 Re: The maximum earthquake potential under Monticello Reservoir during the lifetime of the Virgil C. Summer Nuclear Station.
(1) In response to FSAR Question 361.7, the applicant presents arguments that earthquakes associated with Monticello Reservoir cannot be very large because of the shallow focal depths, 0.5 km.
Two problems exist with that argument:
(a) the assumption is made that the vertical extent of the fault plane equals the focal depth, and (b) calculated focal depths are as much as 4 km (Talwani,1979c,d). Justify assumption (a) in view of focal mechanisms calculated by Talwant that indicate nodal plane dips between 300 and 600 (Talwani et al.,1980), or propose a different assumption that can be justified. Determine whether new evidence exists that the focal depths are really different from those published (Talwani,1979 and Talwani et al.,1980).
Using the justified assumption and the most recent estimates of focal depth, evaluate the maximum earthquake potential under Monticello Reservoir. Present and justify the limitations or uncertainties of that estimate.
RESPONSE
The argument that earthquakes associated with Monticello Reservoir cannot be l
l very large because of shallow focal depths (0.5 km) cannot be justified in l
l light of recent data. However, a case can be made that some of the depths calculated from helicorder data (SCE&G network) are in error.
t In th'. period July-December 1978, 173 - events were recorded on analog tapes (Monticello network). Of these,170 or 98% were shallower than 2 km (68% with l
361.17.1-1
a depth < 1 km). When these depths are compared with those obtained from helicorders (SCE&G network), only 53% agree to within 29 6 km, whereas for 47% of the events, the depth differences range between fp.6 and f.2.6 km.
These observations indicate that although the SCE&G network give adequate epicentral locations, the depth estimates are unreliable.
Discussion of maximum earthquake potential under Monticello Reservoir is contained in response to Question 361.18, and in Appendices VII, VIII and IX of " Evaluation of Site-Specific Seismicity at the Virgil C. Summer Nuclear Station," as formally submitted on the docket in December 1980.
1 361.17.1-2 l
The hypocentral depths obtained from the SCE&G network are not accurate enough to document any changes. However, qualitatively, there does not appear to have been a marked increase with time of the hypocentral depths.
Qualitatively, we can see from Figures 361.17-2-la through lh, that the epicentral growth rate was fastest in the period December 1977-Varch 1978.
During that period the epicenters covereo approximately 70% of all the epicentral area occupied (through 12/79). During the next nine months, (4/78 - 12/78) a further increase of about 20% of the total area occurred.
In 1979, the growth rate was much less (as was the seismicity). A detailed review of site seis-micity is contained in Appendix II of " Evaluation of Site-Specific Seismicity at the Virgil C. Sumer Nuclear Station," as formally submitted on the docket in December 1980.
In summary, after rapid epicentral growth associated with the initial impoundment, further growth has been extremely limited. A possible explanation is that initially the seismicity was associated with the heavily jointed regions--predominantly in the migmatite units. Since then seismicity has spread up to the less permeable country rocks. The country rock is associated with fewer and possibly tighter joints (Secor, pers. com.).
Continued spreading would indicate the presence of permeable jointed rocks, and would not (in the absence of through-going faults) suggest an increase in the maximum earthquake potential.
361.17.2-3
- - ~., - - -,.,.
, -. - - - ~. -. -., -.. -. -,
-,p---,,.-
.-.,p.
,,~- - -..
...---,m.--
-w-
-- +
, 3
m
- 0NTICELLO
- 79
- 0NTICELLC 9/f4-M/77
- f PI i2 Ji.
p.
n:..
2'
.i s. '
l_,
n.
2i. '
- 22. ' p.'
- p. ' i..'
J, a
L.
- 3
!C
. Jg J
ka i..
EJ r
a I2 r
u
()
a..
J l
2 1
s,.
u, sc T
J j
\\ \\./.M I
-c v
of a,
a y1u 9
V
.q a..
s
~..
a s
I i
)
'yge o g,.
1
<g
-~
m,.
J f.
s.
~.
I u.00 1
u,:o'
- ONTICE LO 10/1-7/ 79
- 0NTICELLC fq/8-M/77
"
- 22.
21.
23..
MS.
J T..
J4..
I_
" $2.
- 21. '
$3. '
J J 7. '
- s.'
l _
Qs. '
J ls H.
J p<
a
>t e
a' bJ l
h' p.
4-(
s g
p..
tc J
-g N
r.
\\
t 4a
'^
/ aa
/ l
\\
g I>,l.
C
{-
l
(
-s
_.l ov, 3
V C
%/
/
i j
b 7
~
W" a
.I l
i v
s\\tN t
F" e-J a
L-J
> 'i y
_ /w [ /,g
,L-
-l
' s _/
.g g 4
a...
I..
_l i
l r
m,.
I..
i i
l a
p n.ca.
i u.ca 4
i I
1 FI GURE 361.17.2-1 g l
l 361.17.2-10
"0NTICELLO 10/fS-21/17 MONTICELLO 10/22-3f/77 fI 22. ' ai. '
go.
- ps.
2 7. '
- s.
fI ta.
ai. '
as.
- Ms. ' s t. ' p s.
t J
L.
J a4 wy E
J 5
n.
.a,.
f g
s a.
W U
h a.
1 a.
l l' s 0
O f
9
/\\
~
r I
T
?
I'u U
tr h /
}
',\\
m.
1
)
j s
1
'rS 4 0 J
uM 0
q 1
q q
]
(*.
-4 L..
4 '.
34.00' I
34.30
=0NTICELLO f1/19 MONTICELLO 12/79 p t,, _.,, _,,
p, e t,,; n; p_.
,_.,, _. p; j
ig L.
J
,f e
<,K
% lt m
==
J b ({
^*
,1 f
(
L' '
1 4
1
=
-4
,~ev, y
J w
e i~x
~
a{
ll ~
o a
I g
g p
a.
I
)
)
a.
1 V '73 W 0 s
M0
]
'N
~
j a
J J
m.
m.
Iq '.
s....
o e.
~
34.03 34.00, FIGURE 361.17.2-l h 361.17.2-11
MONTICELLO fO/I3-21/79 M '>,2.0 t
3f 22.
El.
20.
J9.
)8.
17.
JS.
1 I
- s.s i
23.0 22.s 22.0
._2 t 1 2i.0 J.s O
n e
20.0 L
4 9.5 39.O N
/
/
1s.s f !
(#
,_16.0
)
27.5 17.0 1s,s J 8.0 l
0 5
(KM) 34.00 i
i l
i l
i l
I FIGURE 361.17.2-2 q 3'1.17.2-10
361.17 (3) Prior to the most recent technical report (Taleni,1979d) t nearly all the seismicity under Monticello Reservoir has occurred in previously jointed or fractured rock. At a recent meeting with the applicant and his consultants (2/21/80 at the plant site), Talwani indicated that earthquakes were beginning to occur in the nore competent plutonic rock, similar to the rock which is under the plant. Does this indicate that the reservoir is inducing more energetic earthquakes (breaking fresh or unfractured rock) 1.e., earthquakes with greater stress drops?
Explain consequences with respect to maximum earthquake potential.
RESPONSE
The pre =ence of seismicity in the more competent plutonic rocks does not suggest that the reservoir is inducing more energetic earthquakes.
It merely indicates that seismicity in these rocks occurred after that in the more fractured rocks, because these rocks are associated with fewer fracturet.
Due to lower permeability, these rocks would not be associated with seismicity, when more permeable rocks are available for fluid flow. However, after seismicity (and presumably fluid flow) in the heavily fractured rocks has i
stopped, seismicity will be observed in the less permeable plutonic rocks.
The seismicity is still associated with joints and fractures (as indicated by an analysis of fault plane solutions). This is also supported by calculated fracture energies (see Appendix VII of " Evaluation of Site-Specific Seismicity of the Virgil C. Sunner Nuclear Station," as fonnally submitted on the docket in December,1980) which suggest that events are occurring along existing frictional surfaces, since the energies are too low to break new rock. Thus, the presence of seismicity in the plutonic. rocks does not increase the maxi-mum potential earthquake.
361.17.3-1
1 I
A.
PEAK ACCELERATION ESTIMATES To estimate amolitudes of peak acceleration, the Brune (1970, 1971) model of seismic sources is used to determine the root-mean-square acceleration a and duration Td of direct shear waves at the site of ms interest, and a simple application of random vibration theory is used to estimate the peak acceleration a from a and T. This method has been shown to be p
rms d
j appropriate by comparison to recorded ground motions in California (McGuire,1980; McGuire and Hanks,1980; Hanks and McGuire,1980) and is applied te the pro-ject site with certain modifications as discussed below.
)
The model used here to estimate characteristics of motion begins with a description of the Fourier amplitude spectrum of displacement u, at 4
source-to-site distance R, caused by shear waves in the far field. The initial description given here is for ground motion in all directions at a site
]
located in a uniform, isotropic full space. A correction factor to account for free surface amplification, vectorial partitioning of energy into instrumental l'
components, and radiation pattern will be introduced below so that predictions can be compared with observations.
The salient characteristics of the spectrum u are shown in Figure 361.17.4-la. The Icng period level n is given by (Haskell,1964):
o Mo Go
- 4 p R33 Rg W
I where M is seismic mcment, R9, is the radiation pattern of the shear excita-o tion, o is density, and 3 is shear wave velocity.
In Figure 361.17.4-la, the I
L corner frequency f is inversely proportional to source radius r and can be a
l estimated with the relation (Brune,1970):
l o = 2.343 (2) l f
2r 361.17.4-2
1 I
1 High frequency (f>>fo) spectral amplitudes fall off as f', y = 2, and this is an important feature of the assumed model. Hanks (1979) has argued from an observational basis that the Y=2 model is the one generally applicable to crustal earthquakes.
Fletcher (1980) has observed values of y greater than 2 for aftershocks of the Oroville earthquake but only for high corner i
frequencies probably associated with the rise time function of the rupture rather than with the rupture dimension.
The Fourier amplitude spectrum of acceleration ~ can be obtained from u by multiplying by (2nf)2, leading to the typical spectrum shown in Figure 361.17.4-lb. Amplitudes for all frequencies are decreased by the anelastic attenuation factor i
I ka = exp (-rfR/QS)
(3) where Q is the specific attenuation. Anelastic attenuation for typical values of Q and S, and for distances of interest, is only important for frequencies >>f o
4 The earthquake stress drop ao is related to M and r in the Brune g
(1970, 1971) model by:
ao =
(4) 16r3 Relations (1) Through (4), together with Figure 361.17.4-1, constitute the model used to estimate shear wave Fourier amplitude spectra of accelera-tion.
361.17.4-3
-.-..e..-
,--,,.**---**ev'
---**=*e
+ -
- r+
e-=--
et=,e--+---
r vw&
- -= w
v
=
I
)
)
To estimate Fourier amplitudes of single components of motion at the ground surface, we account for the free surface effect (a factor of 2).
vectorial partitioning of energy into two components of equal amplitude 1/[2,and the root-mean-square (rms) of the double-couple radiation pattern (0.6), or a combined correction factor of 0.85 The division of energy into j
two equal components of motion is considered appropriate for the low nagn~itude, near-source ground motions of interest here, in spite of recent strong motion data obtained in California which show high-frequency vertical motions to have i
I 361.17.4-3a
~ ~ ~ ~ ~ ~ " ~ ' ~ ~ ~ ~ ~ ~ ~ ~ ~ " ~
i I
However, they do form a preliminary basis for estimating ground motions at the small source-to-site distance requested by NRC.
Estimates of peak acceleration for the six combinations of magnitude I
and distance requested by NRC are shown in Table 361.17.4-2, along with the para-meters used to obtain these estimates. For one case, where R = 5.8km for a Mt
= 5.3 event, the significant distance of 5.8 km was arrived at using a horizontal distance of 3.0 km and a hypocentral depth of 5.0 km.
The 3.0 km horizontal distance was chosen as representing the closest " cluster" of events (Cluster IV, Figure 361.14-1) of the five reported " clusters" surrounding the site, cor-responding to a " mixed", more fractured, rock type unit in comparison to the less fractured granodiorite bedrock at the plant site. The focal depth of 5.0 km was chosen based on physical constraints such as:
- 1) stress drop; 2) lithostatic I
pressures; and 3) recent near-field observations of peak accelerations which sug-gest that, because of a lack of attenuation of strong ground motion in the range of 0 to 5 kr from the fault trace, the energy-causing strong ground motion ef-fectively is generated at depths of 5 km rather than at the surface. For several cases (indicated by parenthese) the distances involved are less than several source diameters (R < 4r) so that the estimates are actually made for near-field conditions where the theory is not strictly applicable.
In these cases the esti-mates are conservative.
Comparison of Vertical and Horizontal Peak Accelerations The October 15, 1979, Imperial Valley earthquake has indicated vertical peak accelerations in the near-field which are comparable to those for horizontal motion. This has raised the question of whether vertical motions for design in the near-field of earthquakes should be equal to horizontal design i
l 361.17.4-8
.-~.
i J
f motions. To answer this question, peak acceleration data recorded in the near-field of a variety of earthquakes has been compiled. A comparison of vertical and horizontal peak accelerations is presented here.
Figures 361.17.4-20a and 361.17.4-20b show, respectively, peak vertical accelerations and peak horizontal accelerations, plotted versus magnitude. These data were recorded at distances Rf of 10 to 20 km, the distance used is the distance to fault rupture if this is known, or focal distance otherwise. At magnitude 6.5 the vertical data seem to be as large as the horizontal, due largely to the Imperial Valley earthquake. At smaller magnitudes, vertical data are consistently lower than the horizontal, by a factor of about 0.5 at M L
= 5.0, Note that for many of the records shown in Figures 361.17.4-20a and i
361.17.4-20b the two horizontal component peak accelerations were available, but the vertical was not.
Figures 361.17.4-21a and 361.17.4-21b show a similar comparison for f
5 j[ Rf < 10 km. While fewer data are available at this distance, they indicate the same conclusions as discussed earlier for 10 j[ Rf < 20 km.
Figures 361.17.4-22a and 361.17.4-22b indicate similar conclusions for data in the range 0 j[ Rf < 5.0 km, although the earthquake records available at these close distances are few.
Another way to compare these data is to plot them versus distance. Figures 361.17.4-23a and 361.17.4-23b show vertical and horizontal data for 5 j[ ML < 6.
With the exception of one record from the Coyote Lake earthquake obtained at Rf
= 3 km, these data indicate that vertical peak accelerations are less than hori-zontal peak accelerations for ML in the range 5.0 to 6.0, for all distances in the near-field.
361.17.4-8a i
. -,.,.. _ _ -,...,,.., _.,, _, -,,... -... _ _,.,..,,,.,, -,. ~.
i 1
4 i
i From these comparisons we conclude that vertical peak accelerations are generally less than horizontal peak accelerations in the near-field, for magni-tudes less than 6.
A-
'?ropriate ratio of vertical-to-horizontal peaks is 0.5 for magnitudes less than about 5.5.
B.
RESPONSE SPECTRA ESTIMATES Response spectra were estimated based on a procedure developed by New-J l
mark and Hall (1969) and later modified by Mohraz (1976) and Johnson and
~'
Traubenik (1978) to incorporate site conditions, and magnitude and distance respectively.
A general description of the procedure used by Johnson and Traubenik (1978) to obtain ground motion ratios and amplification factors is as follows:
1.
Ground motion ratios, v/a and ad/v, for various percentiles 2
using a normal and a log-normal distribution were calculated.
2.
The selected response spectra were normalized to obtain amplifica-tion factors. The procedure is to obtain, at each frequency point, the ratio of the spectral response to the maximum ground motion (i.e. amplication factor) for acceleration, velocity, and displacement in the corresponding frequency ranges. Averaged i
g 361.17.4-8b 1
P l
R = 10 km (Figure 361.17.4-9). Similar comments regarding conservatism apply to Figures 361.17.4-7 and 361.17.4-S, because a for these events are conser-p vatively estimated as noted in Table 361.17.4-2.
1 Figure 361.17.4-10 is presented to show the comparison of a Magnitude i
i 4.0 earthquake at R =3 km, and a Magnitude 5.3 earthquake at R =5.8km to the design OBE and SSE Spectra (where R is the significant distance from the focus of energy release to the plant site).
The former event, a M =4.0 earthquake, represents the maximum 3
L induced earthquake by Monticello Reservoir and has been discussed in detail in response to Question 361.18 and in the report " Evaluation of Site-Specific Seismicity at the Virgil C. Summer Nuclear Station," as formally submitted on the docket in December,1980. The latter event, a M =5.3 L
earthquake, has been suggested by the NRC to be representative of the mag-i nitude associated with a Modified Mercalli Intensity VII event (the SSE j
for the V. C. Summer Plant).
Even though the spectrum due to the M =4.0 and R=3.0 km event L
exceeds the original OBE spectrum in certain frequency regions as shown in Figure 361.17.4-10, there is no adverse effect on the structure and equipment.
This is due to the fact that a 2% structural damping value was used in the original design and Regulatory Guide 1.61 allows 2% damping for prestressed concrete structures and 4% damping for reinforced concrete structures.
In the evaluation of the effect due to the M =4.0 spectrum, 2% and 4%
L composite damping values were used. The time history method was used to generate the structural responses and floor response spectra. Four horizontal i
components of two M = 4 aftershocks of the 1975 Oroville, California L
361.17.4-11
,--v-
-,=,-r----,..r.
p.
_.,,,-,.,~.-m
,yyw<w,,-
,e-m,,
,m
-ey-+e...._.,.-w
-.n.-.e-y.-w--em~-,
]
occurrence of such an event is very low, based on the results of probability
{
studies presented in Section 316.19, and in comparison to similar studies at other sites. Moreover, there is no evidence to suggest that the original design basis for the subject facility should be changed. Therefore, the applicant concludes that the seismic design for the V. C. Sumer plant is adequate.
This analysis has treated two earthquakes, one of M =4.0 and L
a the other of M =5.3, as a hypothetical case. The report "Evalation of Site-g 1
Specific Seismicity at the Virgil C. Summer Nuclear Station," formally submitted on the docket in December,1980, contains a detailed acccunt of the selection of a magnitude (M ) 4.0 earthquake as representing the largest earthquake which t
can be induced by Monticello Reservoir. For conservatism, a magnitude 4.5 earthquake was assumed to occur near the plant and the effects on the plant seismic design were analyzed. This analysis is reported in Appendix X of the aforementioned report.
361.17.4-12b i
1
RE."ERENCES Brune, J.N. (1970), " Tectonic stress and the spectra of seismic shear waves from earthquakes",
J_. Geophys. Res. 75, 1997-5009.
Brune, J.N. (1971). Correction, J_. Geophys. Res. 76, 5002.
Fletcher, J. B. (1980), " Spectra from High-Dynamic Range Digital Recordings of Oroville, California, Aftershocks and Their Source Parameters", Bull.
Seis. Soc. Am., 70, 3, June, pp. 735-755.
Hanks, T.C. (1979), "b-Values and y-n Source Models:
Implications for Tectonic Stress Variations along Crustal Fault Zones and the Estimatien of High Fre-quency Strong Ground Motion",
J_. Geophys. Res.
Hanks, T.C. and McGuire, R.K. (1980), "The Character of High Frequency Strong Ground Motion," (in preparation).
Johnson, J. A. and Traubenik, N.L. (1978), YMagnitude-Dependent Near-Source Ground Motion spectra", Proc. A.S.C.E. Conf., Earthquake Engineering and Soil Dynamics, Pasadena, June, p. 530-539.
i McGuire, R. K. (1980), " Geophysical Estimates of Seismic Shear Wave Motion",
Proc., 7th World Conf. on Earthquake Eng., Istanbul.
McGuire, R.K., and Hanks, T.C. (1980) "RMS-Accelerations and Spectral Amplitudes of Strong Ground Motion During the San Fernando, California Earthquake,"
Bull. Seis. Soc. Am., 70, 5, Oct.
Mohraz, B. (1976), "AStudy of Earthquake Response Spectra For Different Geological Conditions, " Bull. Seis. Soc. Am., v.66, No. 3, June.
, Newmark, N.M., and Hall, W.J., " Seismic Design Criteria for Nuclear Reactor Facilities," Proc., 4th World Conf. on Earthquake Eng., Santiago, vol. 2
- p. B5-1-85-12.
Thatcher, Wayne, and Hanks, T.C. (1973), " Source Parameters of Southern California Earthauakes", Journal of Geophysical Research 78, 8547-8576.
Vanmarche, E.L., and Lai, S.P., " Strong Motion Duration of Earthquakes", M.I.T.
Dept. of Civil Eng., Report R77-16, 32p.
McGuire, R.K. (1980), " Geophysical Estimates at Seismic Shear Wave Motion, Proc, 7th World Cont. on Eartnquake Engineering, Istanbul, September Tsai, N. C.,"Transfonnation of Time Areas of Accelerograms", Proceedings of ASCE, Engineering Mechanics Division, Vol. 95, No. EM3, June 1969.
361,17.4 13
I 2000
}
KtY:
C IMPERIAL VALLEY (1979)
O coy 0TE LArt (1979)
]000-6 SANTA BARBARA (1978) 900-o san rtanAnoo (1971) 800-
+ CAL!r0RNIA 700-x wtsitanu.s.(ExctPrcA.)
- J"'A" 600-500-400-N U
x 300-Kuw in E
p 200-u
~
l z
O
+
H F-c 100-E 90-
+
g
)
80-
+
W 70-
+a e-u 60-
+
u 50-c
++
y 40-a
+
30-
+,
+
+
+
+
20-
+
+
+
10,
2.00 3'.00 4'.00 S'.00 6'.00 7'.00 8'.00 l
LOCAL MAGNITUDE Figure 361.17.4-20a.
Vertical Peak Accelerations Versus Magnitude for 10 5 Rf < 20 km.
2000 KEY:
O IMPERIAL VALLEY (1979)
O C0Y0TE LAKE (1975) 6 5ANTA BARBARA (1978) 2000-o san FERNAN00 (1971) 900-
+ CAL!rcRNIA 800-a WESTERNU.S.(EICEPTCA.)
700-X JAPAN 600-4 ONERS
+
500-
++
b 400-
+
+
b y
300-
,+
- +
^
g t
o W
A
+
l p
200-
+
8 iv;,+
+
+
+
y f
y
+
++
+
H A
H
+
c 100-
+
+4 4+.
W 5
90-
- +
+
.+
80-
++ +
+
d 70-
+
g
+
8 so-
+
++
+++
+
+
u 50-
+
c
+
+ ++
+
s 40-4
+
+ ++
30-
+
++
++
+
20-
+
+
+
10 2.00 3.00 4.00 5.00 6.00 7:00 8.00 LOCAL MAGNITUDE i
Figure 361.17.4-20b.
Horizontal Peak Accelerations Versus Magnitude for 10 2 Rf < 20 km.
i 2000 try:
C IMP [ RIAL VALL[y (1979)
O cov0TE LAKE (1979) o SANTA BA uA u (197s) 2000-900-O SAN RRNANN (1971) 800
+ cAtt m NrA 700-wisitRN U.s. (ExctPr cA.)
=
X JAPAN 600-e 9 gy, 500-400-m N
y 300-m W
La t.np 200-
+
8 8
e 2
o I
H
++
g C
200-90-s 80-0 x
d 70-M 60-a 50-
+
c E
40-2
+
30-X 20-102.00 3'.00 4'.00 5'.00 6'.00 7'.00 8.00 LOCAL MAGNITUDE Figure 316.17.4-21a. Vertical Peak Accelerations Versus Magnitude for 5 mRf <10 km.
2000=
KEY:
O IMPERIAL VALLEY (1979)
O C0Y0TE LAKE (1979) 6 $ANTA BARBARA (1978) 2000-Q SAN FERNAM (1971) 900-
+ CAttr0RNrA 800-
= WEsTERNu.s.(ExcEPTcA.)
700-X JAPAN 600-
$ ONER$
500-
+
+
400-a g
a
+
y U
M 300-M i
M
- 5 e
E v's N
B p
200-
+
e e
S
$+
e Z
+
o e
H
+
w
+
+
+
c 100-e a:
90-
+
80-x
+
d 70-
+
+
u 60-
+
u 53-
+
c 40-20-x x
20-10 2.00 3.00 4.00 5.00 6.00 7.00 8.00 LOCAL MAGN]TUDE Figure 361.17.4-21b. Horizontal Feak Accelerations Versus Magnitude for 5 1 Rf.10 km.
. ~. - -...
2000 m
1000-900-m 800-700-m U
600-U 500-400-N g
300-
+
o LLJ
+
v)g 200-8 m
U 2O H
I--c 200-90-a 80-d 70-u 60-x a
50-i c
+
40-uy:
O IMPERIAL VALLEY (1979) 30-O C0YOTE LAKE (1979) 6 SANTA BAasARA (1978) o SAN FERNANDO (1971)
+ CALIFORNIA 20-
= WESTERN u.s. (ExcEPT cA.)
M JAPAN
$ OThERs 10,
2.00 3.00 4.00 5.00 S.00 7.00 8.00 LOCAL MAGNITUDE Figure 361.17.4-22a.
Vertical Peak Accelerations Versus Magnitude, 0 2 Rf < 5.
2000 ee 1000-900-800-m 700-O 600-X m
500-E m
400-O e
N
+0 B
+
M 300-x O
5 8
M O
8 g
200-2 2o
+
H c
200-90-
_3 80-d 70-60-g 50-C 40-ggy O IMPERIAL VALLEY (1979) 30-o C0YOTE LAKE (1979)
O SANTA BARBARA (1978)
C SAN FERNANDO (1971) 20-
+ CALIFORNIA WESTERN U.S. (EXCEPT CA.)
N M JAPAN
$ OTHERS 10 2.00 3.00 4.00 5.00 6.00 7'.00 8.00 LOCAL MAGNITUDE Figure 361.17.4-22b. Horizontal Peak Accelerations Versus Magnitude, O s Rf 45.
2000 KEY:
O IMPERIAL VALLEY (1979)
O C0Y0TE LAKE (1979) 1000-6 5AN'A BARBARA (1978) 900-o SAN FERNANDO (1971) 800-
+ CALIFORNIA I WESTERNU.s.(EXCEPTCA.)
700-N 'AN 600-
.500-400-N y
300-
+
U La vig 200-
+
0 8
o O
2 a
o
+
+
+
[
c 200-90-
++
_J 80-e d
70-a 60-c a
50-c
+
+
40-
+
+
30-
+
20-
+
2 3
i $ $ 1'O 20 DISTANCE (KM)
Figure 361.17.4-23a.
Vertical Peak Acceleration Versus Distance for 5.0 sML
<6.0.
2000 1000-900-800-700-600-500-400 D
+
b 0
+
+a
+
+
M 300-A
+'
e o
+
O m
.A y
200-a e
8
+
Z
+
0 0
H C
100-h F-90-
__s 80-
+
+
+
d 70-60-
+
a 50-
+*
+
C
[
40-KEY:
C IMPERIAL VALLEY (1979)
O C0YOTE LAKE (1979) n 6 SANTA BARBARA (1978)
Q SAN FERNANDO (1971) 20-
+ CALIFORNIA N WESTERN U.S. (EXCEPT CA,)
M JAPAN
$ OTHERS 1
2 0
y 5 $ 1'O 20 DISTANCE (KM)
Figure 361.17.4-23b.
Horizontal Peak Accelerations Versus Distance for 5.0 2 M, 46.0.
I i
1.27 and 1.21, respectively. Both the Lake Jocassee data and the Monticello Reservoir data were then used to obtain recurrence rates. The object of using the Lake Jocassee data was to see if the calculations yielded meaningful results which could be compared with the observed seismicity there, i
RECURRENCE RATES If T is the mean return period of an earthquake exceeding a magnitude M, then L
for two years, N, the number of earthquakes per one year exceeding that magnitude T = 2/N. For Lake Jocassee, where 4 years data have been used, t = 4/N. For Monticello Reservoir, the return periods of 1,10, 40 and 50 years correspond to N = 2, 0.2, 0.05 and 0.04, respectively. In the calcula-tions a b*-slope value of 2.67 (b-value of 1.34) is used. (It yields more conservative results than by using that obtained by Utsu's formula.) The magnitudes (M ) corresponding to these return periods are given in Table l
g 361.18-2 where they are compared with results from Lake Jocassee.
The largest earthquake recorded at Lake Jocassee (in about 7 years since its impoundment) is of the same order as predicted for a 40-50 year return period.
Thus, it appears that the estimates obtained by using b-value extrapolations yield usable order of magnitude values. For Monticello Reservoir, this method suggests that the largest earthquake with a return period of 10-50 years is Mt = 4.0 - 4.5.
l A note of caution is in order. In about 3 years of continuous monitoring, no event with ML > 3.0 has been observed.
It is entirely likely that the 1
[
maximum size of the earthquake is more closely related to the local geology,-
fractures and stress conditions rather than to the conclusions of a statistical exercise.
(
361.18-4 s
,n
-~ ~. ~.,,, -,. ~,,,,...,,,, -. -, -,
,-.n,
been suggested (Talwani,1980, Personal Com.) that it may be associated with i
the King's Mountair belt - Charlotte belt contact.
If it is assumed, however, that the Union County event was a random event, and could occur anywhere in the Piedmont (as has been done in Section 2.5.2.9 of the FSAR), then the largest event that could occur near the Monticello Reservoir is MM Intensit)
VII.
From the nature and extent of the isoseismals of the Union County event, its depth has to be at least -5 km.
SUMMARY
From the empirical analysis which is based on all existing data, the largest induced earthquake is estimated to be 'ML 4.0 with a depth of at least 3 km.
The largest tectonic earthquake is estimated at MM Intensity VII (or 'ML 5.3 for eastern U.S.) with a depth of at least 5 km. The return period for such an event is 4500 years, in keeping with expected annual probabilities of exceedance for an event such as the SSE (Discussed in detail in response to Question 361.19).
Supplemental discussion of induced seismicity is contained in Appendices VIII and IX of " Evaluation of Site-Specific Seismicity at the Virgil C. Sumer Nuclear Station," as formally sub-mitted on the docket in December 1980.
REFERENCE
- Aki, K., (1965) Maximum likelihood estimate of b in the formula log N =
a + bm and its confidence limits. Bull. Earthq. Res. Inst. Vol. 43,
- p. 237-239.
Gupta, H. K., and B. K. Rastogi, (1976), Dams and Earthquakes, Elsevier, Amsterdam, 229 pp.
Gutenberg, B. and C. F. Richter (1944), Frequency of Earthquakes in California, Bull. Seis. Soc. Am., 34, 185-188.
i Severy, N. I., G. A. Bollinger, and H. W. Bohannon, Jr., (1975), A Seismic Comparison of Lake Anna and other Piedmont Reservoirs in the Eastern U.S.A., 1st Internat. Symposium on Induced Seismicity, Banff, Alberta, Canada, Sept. 15-19, 1975.
- Utsu, T., (1965), A method for determining the value of b in a formula log N = a - b M showing the magnitude-freauency relation for earthquakes. Geophys. Bull. Hokkaido Univ. 13,,99-103 (In Japanese with English Summary).
361.18-6 s
~ - - -,,,
- ~-
.-e--,n
--.,,,,-..,r,.
--w,~.--,
,a,,_ee,--
..,,--..,,,,--,rr.,
l 4
10 ll o
o 10'-
MONTICELLO O
/.67 2
2 10 _
e e
JOCASSEE E
j W
2.54 aMbLg 1-o (JOCASSE)
\\ \\
0.1 -
k
.05
.04 -
DURATION ISEC) 10 100 1000 10,000 I
I dl I
- 01 I
I I
I I
I I
I l
l l
l 0
1 2
3 4
5 ML FI GU RE 361.18-2 361.18-12
Plain, Piedmont, Blue Ridge, Valley and Ridge, and Appalachian Plateau zones.
For each zone, the historical occurrence of earthquakes was determined using the Bollinger catalog. Rates of occurrence of earthquakes with fi.M. intensity greater than V were determined by correcting each intensity level for incomplete reporting of events. A Richter b-value of 0.5 was asstned because this has been widely reported in the literature for eastern North America; the data used in this study indicate a b-value of approximately 0.5, although some statistical variations from zone to zone occur. Maximum possible intensities for each zone were selected to be equal to the largest historical earthquake plus one unit; sensitivity to this selection is discussed below.
The M.M. intensity at a site Is was estimated from the following equation:
Is = 3.08 + Ie - 1.34 1n a a>10 km Is " le a110 km where le is epicentral M.fi. intensity and a is epicentral distance to the site.
Uncertainty in this estimate was modeled by a Gaussian probability distribution with a standard deviation of 1.19 intensity units, except that the distribution was truncated so that site intensities could not exceed epicentral intensities.
The above equation is very close to that reported by Bollinger for the 1886 Charleston earthquake; the Gaussian distribution is a typical one used to represent uncertainty in ground motion estimates.
The first two rows of Table 361.19-1 indicate annual probabilities of exceedance (and corresponding return periods) for various intensity levels, using the models just described. The annual probability of exceedance associated with intensity VII is.22x10-3, implying a return period of 4500 years.
If the assumption on maximum earthquake size is modified to say that the maximum possible event j
361.19-2
the Bollinger catalog is primarily in the form of intensities rather than magni-tudes. The maximum value of Mb was estimated by determining the largest historical earthquake in each zone in terms of le, converting this to Mb by the above equation, and raising this value of Mb by 0.25 magnitude units.
l Thus, for instance, a largest historical intensity of VII would imply a largest l
possible magnitude Mb of 5.5, wnich in turn corresponds to intensity VII-VIII.
A depth of 5 km was used to represent the source of energy during tnese events.
A Richter b-value of 0.9 was used for these calculations, because this is thougnt to best represent the distriution of Mb values for tectonic earthquakes in the area. This value provides more conservative probabilities than a value of 1.0, which would be implied by the previously referenced b-value for intensities and the above relation between Mb and Io.
For this analysis, ground motion was characterized by peak horizontal accelera-tion; this was estimated using equations derived by Nuttli (1979) for the central U.S., which are thought to be appropriate for the southeast U.S.
The theory of Nuttli was modified to estimate the peak acceleration from a sustained accelera-tion, and to estimate tne average of two horizcatal motions rather than the larger. A log normal probability distribution was used to represent uncertainty in the estimate of peak acceleration.
For the FSAR sources, with parameter values and attenuation model as discussed above, Table 361.19-2 gives annual probabilities and return periods associated with various levels of acceleration. A peak acceleration of 0.15 g has an estimated return period of approximately 3,400 years, by this analysis.
k For purposes of comparison, we have also estimated annual probabilities of
)
exceedance from induced earthquakes. These were assumed to occur at 1 km depth, 361.19-4 3
. ~ - -
_ _ ~. - _ _. _,
and to have magnitudes (M ) in the range 1.6 to 3.6 which corresponds to b
local magnitudes in the range 2.0 to 4.0 using the Nuttli relation:
M - 1.023 Mb + 0.3 t
The rate of occurrence of induced seismicity in this magnitude range was estimated to be 35 events per year (Talwani, personal communication,1980),
and the Richter b-value was assumed to be 1.34 based on Talwani's analysis of observed events.
The Nuttli attenuation equation was used to estimate accelerations for these events, as it was for tectonic earthquakes. By comparison with the recorded earthquake of August 27, 1978 (M = 2.7, R + 0.8 km), Nuttli's theory esti-t mates 0.12 g for this event, which is in excellent agreement with the (digitized) peak horizontal acceleration values of 0.11 g and 0.13 g.
Table 361.19-2 indicates the calculated annual probabilities associated with induced seismicity. These probabilities are approximately 100 times as large as those for tectonic events, primarily due to the large number of them (35 per year), their shallow depth, and their proximity to the facility.
It is also a property of small magnitude events close to the site that they can generate larp peak accelerations, but may produce no damage because of short duration and lack of energy at lower frequencies.
A supplemental probabilistic analysis of reservoir-induced seismicity, using alternate input parameters, is contained in Appendix IX of " Evaluation of Site-Specific Seismicity at the Virgil C. Summer Nuclear Station," as for-mally submitted on the docket in December,1980.
361.19-5
m (f) In over two years of observations, no increase in hypocentral depth beyond s2 km has been observed based on depth computations using magnetic tape da ta. As described in response to Question 361.17.4, a M s5.3 earth-L quake would require at least a 5 km depth.
(2) Observations at Lake Keowee (Talwani et al,1979) su; est the migration of seismicity from one joint system to another. Multiple events have been observed at Lake Jocassee (4/21/76). However, the magnitude of no event has exceeded (M ) 2.1.
Subsurface data with detailed descriptions of joint L
patterns are not available and thus it is not possible to " estimate the possibility of such an event".
For a detailed discussion of the magnitude and orientation of available stresses at Monticello Reservoir, reference is made to Appendices III, IV, and VII of " Evaluation of Site-Specific Seismicity at the Virgil C.
Summer Nuclear Station," as formally submitted on the docket in December, 1980, and the response to Question 361.21, following.
REFERENCES Talwani, Pradeep, D. Stevenson, D. Amick, and J. Chiang, (1979) An Earthquake Swarm at Lake Keowee, South Carolina, Bull. of the Seism. Soc. of Am., Vol.
t 69, No. 3, pp. 825-841.
- Talwani, P., B.K. Rastogi and D. Stevenson, (1980) Induced Seismicity and Earthquake Prediction Studies in South Carolina, Tenth Technical Report, USGS Contract No. 14-08-0001-17670, 212 pp.
361.20-3 l
,-,--,-,+-r.-,.,
n
- ~,
...er-
,,---.n-
.v-,-.,-,
,n.
substantially greater than the vertical stress. This stress configuration is appropriate for thrust faulting. The data in Figure 361.21-2(a) indicate that the tendency for this thrust-type faulting is limited to the upper 300 m or so, as the measurement at 486 m shows that the vertical stress is no longer the least principal stress (a requirement for thrust faulting).
Figure 361.21-2(b) shows the magnitude of in situ stresses at Monticello 2.
At shallow depths both horizontal stresses are greater than the vertical stress. However, at depths from 200 m to 320 m, both horizontal stresses are approximately equal to the vertical stress.
The observation at a depth of 400 m could possibly be reopening of pre-existing fractures. The hydrofracture at 508 m depth entered a pre-existing fracture. Thus, the stress value lies between the maximum and minimum stress value. No data were obtained below a depth of 650 m.
From Figures 361.21-2(a) and (b) we note the following:
The maximum stress difference in a vertical plane (o nor, max Cver) displays a remarkable trend. With depth the stress difference diminishes.
It becomes equal to zero at a depth of approximately 300 m in Monticello #2 well.
Similar conditions are achieved in Monticello #1 well at a depth of approxi-mately 1100 m.
The depth dependent decay in the stress difference must indicate similar decay in the distortional strain energy stored in the bed-rock. A depth interval for which the distortional strain energy is very small defines a stress " barrier." This barrier cannot be penetrated by a dislocation plane which has been initiated in a region of high deviatoric stress (e.g., above or below the stress barrier). Hence, the barrier is 361.21-2 i
an important factor influencing the size of a potential seismic source as well as the size of a potential stress drop.
It should be noted that the stress " barrier" occurs at different depths in both hydrofracture wells.
Hence, it is anticipated that the plane representing the stress carrier is a " warped" surface with highly variable depth of occurrence. The occur-rence of reservoir-induced earthquakes must be confined to the rock mass j
above the barrier surface.
361.21-?a
The basic assumptions in estimating tne magnitude is the applicability of the Brune model and the Thatcherand Hanks relationship to smaller magnitudes.
If all these assumptions are satisfied, Mt can be estimated from equation (1) and (2) for assumed values of r, the source radius. For source radii of 10 and 100 m, the calculated potential magnitudes are 0.9 and 2.9, respectively.
A maximum stress drop of 100 bars was assumed in the response to Question 361.17.4.
This maximum stress drop has been revised to 25 bars according to additional analyses as reported in Appendices III, VII and X of the report
Evaluation of Site-Specific Seismicity at the Virgil C. Sumer Nuclear Station," as formally submitted on the docket in December,1980.
RE FERENCES Brune, J.N. (1970) Tectonic stress and the spectra of seismic shear waves from earthquakes, Journal of Geophysical Research, 75_, 1997-2009.
Brune, J.N. (1971) Correction, JGR, 76, 5002 Thatcher, W., and T. C. Hanks (1973) Source Parameters of Southern California earthquakes, JGR, 78,, 8547-8576.
Zoback, M.D., et al., (1977) Preliminary stress measurements in central California using hydraulic fracturing technique, PAGE0PH,115,135-152.
361.21-4
3-j sg
\\t?
I sf o -=
\\\\.t
~
a =
L a
h U =m I
== :"
o o a.
-a
{
I I
t I
-2 tr.
I t
f f
f.
f I
f f
f f
f
~
=
a a
n a
a a
a d
SH 313M 0111 ' Hid30 E-I 5" u m
ST w
\\t?
o 2 2 a=
0 Q g_
=
+
- '/
&,@' \\k
~
T
=
~
e5
~
I
=
g
-=
=
=
o m 2
__g
,/
t t
t t
f f
I a
N w
se se a
a a
a a
a c
SH313M0111
- Hid30 Fl GU RE 361.21-2 361.21-6
-.