ML20002B106
| ML20002B106 | |
| Person / Time | |
|---|---|
| Site: | Yankee Rowe |
| Issue date: | 12/05/1980 |
| From: | BOSTON EDISON CO. |
| To: | |
| Shared Package | |
| ML20002B092 | List: |
| References | |
| NUDOCS 8012090443 | |
| Download: ML20002B106 (27) | |
Text
__.
..g
- M,y )..: :
Enclosure B-
, u.
3 1
P
~
INTERIM SEISMIC' DESIGN BASIS FOR YA'IEE NUCLFAR POWER STATION ROWE, MASSACHUSETTS December 5, 1980 s
I By Yankee Atomic Electric-Company 25 Research' Drive e
Westborough, MA 01581-
-801N S OYf)
s-
- TABLE OF CONTENTS
?B&e 111
. LIST 0F. FIGURES................................................-
' l '. 0 '
-INTRODUCTION...............'..............................-
1 2.0
-BACKGROUND..........~..............'.......................
3 i'
3.0 SITE-SPECIFIC RESPONSE SPECTRUM.s........................-
5 I
4 '. 0 PROB ABILISTIC ANALYSIS OF SEISMIC IIAZARD................. -.
10 1
4
- 5. 0' SITE' AMPLIFICATION.......................................
16
6.0 REFERENCES
'18' 4
4 4
1 A
4 1
J i
1 11' 7
3 '
r
\\
l t
,3_
4,
.. f s
' Th
.1 '
- 3. ST OF FIGURES' Figure
- Title'
~
-1..
Interim' Seiomic Design Basis Spectrum 2.
. Historical Earthquakes and. Tectonic Provinces -
3.' :
1 Site' Specific Spectra 4.
' Calculated. Peak Groudd' Motion' Felt at.Rowe
~
.5.
Source Areas-46.
. Uniform Hazard. Spectra f
4 5
O
(
x 4
I
.i I
1 l-
-l s
v 4
~
5 E
d I
1 iii' I
t 1
i 4
a 4 4,.-_,- _ -..
4
,...a..
c u
1.0' INTRODUCTION AND
SUMMARY
Topic II-4 of the NRC's Systematic Evaluation Program.is'a-review of the geology'and seismology of the Yankee Rowe Site. The NRC,Lin its letter of-January 15, 1979 to Yankee Atomic Electric Company, stated the necessity for Yankee to assess'the siismic design criteria for the Yankee Rowe Site.
The NRC noted that,_ " Evaluation of various site-specific response spectra-methodologies may demonstrate a more realistic approach in determining seismic input".
Yankee has undertaken an extensive program to determine the Site-Dependent Response Spectra for the Rowe Site.-
This paper summarizes the results of 'lankee's program te develop Site-Specific Seismic Response Spectra leading to selection of a co0servative Interim Seismic Design basis for the Yankce plant.
The following conclusions and summary statements apply to Yarkee's seismic t
response spectra:
1.
The Interim Seismic Design Basis Spectrum (see Figure 1) was developed in accordance with the methodology of Appendix A, 10CFR100.
2.
The Interim Design Basis Spectrum has a median exceedance probability of less than 10-3 per year.
3.
The proposed Interim Design Basis is considered adequately conservative for an interim evaluation of the Rowe plant'because:
l a.
The' historical seismicity in the Rowe region is characterized by very infrequent earthquakes of low intensity.
i l
1
b.
The proposed Interim Seismic Design Basis Spectrum falls above the mean of the real site-specific data set for all periods and is above the 84th percentile of che real data set for periods above 0.9 seconds (see Figure 3),
The proposed spectrum (developed from real data) has been scaled c.
up to a peak ground accelegation of.lg and has been modified for conservatism to maintain a constant peak velocity for periods of 0.4 seconds and above (see Figure 1).
d.
The peak ground acceleration (Figure 4) of.lg is mare than three
' times greater than the calculated historical maximum peak acceleration for over 250 years of record in the Rowe area.
The proposed spectrum adequately incorporates site amplification.
e.
f.
The proposed Yankee respouse spectrum has a return period on the order of 1,500 years. Over the two year interim period, the chance of not observing this event is 99.8%.
Conversely, the chance of observing the 1,000 year event over the next two years is 0.2%,
which is less than that implicitly accepted by the NRC in recent licensing decisions.
2
4 2.0 BACKCROUND The Yankee Rowe facility is located in-the town of, Rowe, Franklin _ County,.
Massachusetts, _ on the east side of the Deerfield River, three quarters of a mile ' south of' the Vermont-Massachusetts border.. Most of the land in the
- immediate vicinity of.the' site is heavily forested, including parts of state and national forests; The only community with a population of more than i
2,500 within 16 km is North Adams, Massachusetts, located about 14 km to the west-southwest;1 Pittsfield, Massachusetts, which lies 34 km to the southwest, is the closest city or town with a population of greater than t
25,000.
Geologically, the site is located in the Green Mountain Section of the New England Physiographic Provinca (Reference 1).
Tecton. ally, the site is located on the east flank of the Berkshire - Green Mountain anticlinorium in the central axial zone of the Western New England Foldbelt (Reference 1). _The historical-seismicity in this tectonic province is of very low frequency and intensity. Figure 2 shows Yankee Rowe in relation to historical earthquake epicenters demonstrating the low historical seismicity of the Yankee Rowe site region. For the province, the largest earthquake not associated with geologic structure is an intensity V.
The largest earthquake in the province is an intensity VI, m =4.8 event, which occurred b
It is in Woburn, Quebec, approximately -338 km north-northeast of Rowe.
believed that this earthquake is associated with the mafic intrusive-fault complex near E.gantic, Quebec.
Peak horizontal ground motion at the Rowe site, based upon historic seismic activity (more_than 250-years), has been determined using the conservative-3
9
.g relationship of Nuttli and Herrmann_(1978). Figure 4 is'a plot of the
. calculated. horizontal ground accelerations that would have been felt at the Rowe-Site. As 'shown, Lthe maximum historical' acceleration was calculated to be.029g._:The earthquake that. generated this computed maximuct acceleration is.the 1755 Cape Ann ~ event located 214 km from the Rowe site.
4 d
a 4-1
.I l
4 Y
4 i
1, 1
+
~.--_.-..
{}i 3.0 SITE SPECIFIC RESPONSE SPECTRUM This section ' describes the ' evelopment and characteristics of the Site-d Specific Seismic Response Spectrum for the Yankee' Nuclear Power Plant.
It is emphasized that the Interim Design Basis Spectrum is more severe than this Site-Specific Spectrum.
The Site-Specific Response Spectrum for the Ysnkee Site was developed using
. actual recorded accelerogram data'from earthquakes which' approximate the magnitude, hypocentral distance, and the site foundation conditions of the controlling earthq'uake. This approach, which is in accordance with 10CFR100, Appendix A, is " site-specific" and avoids many of the problems inherent with determining ground motion by scaling event size or epicentral distance.
From t cent ceismological studies, it is known that the shape of the response spectrum is dependent on earthquake size, recording distance and the asscciated frequency-dependent attenuation effects. The proper selection of accelerograms which are used to construct the response spectra avoids the errors resulting from scaling.
3.1 Geology and Seismology j
The seismic evaluation for Yankee Rowe prepared by Weston Geophysical was
~
submitted to the NRC on February 23, 1979(1). This report entitled, " Geology and Seismology, Yankee Rowe Nuclear Power Plant", considered the seismicity within the site tectonic province, as well as adjacent tectonic provinces 4
I and structures to reach the conclusion, "that an Intensity VI(MM) is an appropriately conservative estimate of the Safe Shutdown Earthquake". The Safe Shutdown Earthquake is best characterized by a magnitude of 4.5 in i
the. report' entitled " Eastern United States Tectonic Structuras and Provinces 1
5
-l
- i Sign'ificant to_the Selection of a Safe Shutdown-Earthquakc". prepared by
~
fWeston Geophysical for the SEP Owner's Group and submitted to the NRC on Oct'ober: 16 -1979(2),
The occurrence of larger earthquakes at greater epicentral distances was
~
-also considered in the' seismicity evaluation of the Rowe site. -The magnitude of_ maximum historical activity located in the adjacent provinces (Ossipee, 1940; Cape Ann,~ 1755) are estimated to be near 5.5.
On the basis-of the historical activity, the' ground motion potential is conservatively estimated to be 4.8 in the site province and-5.7 in the White hountain series. This is a conservative representation of the maximum historical events migrated throughout their respective provinces and is consistent with 10CFR100 Appendix A.
3.2 Site-Dependent Response Spectra The report, " Site-Dependent Response Spectra - Yankee Rowe" prepared by Weston Geophysical and Dr. Erik Vanmarcke,(3) was transmitted to the NRC on February 29, 1980. That report describes the methods and results of a program to develop site-dependent response spectra.
The following are the ranges of magnitudes and distances for the two
~
categories of events that were conservatively defined to represent the ground motion potential.
Epicentral 2
Province Magnitude Range Distance Range (mb)
(km)
Site.
4.5-5.4 2-33 Distant 5.4-6.0 50-85 6
a.-
~
e v
g
e
~
s 1 Geological; investigations of the' site locale show that.the Yankee plant is situate'd~on glacial sediments.in the lower elevation of a broad-bedrock
~
valley with the bedrock surface beneathithe site dipping at 300 to 500 'to
~
Seismic refraction surveys and Ltest' borings have identified
.the' southeast.:
the glacial deposits as very dense till.' ' In situ velocity measurements,(4) transmitted to the NRC on April 5, 1979,-have determined that the 70 to 140 feet of glacial till beneath the. site has a compressional wave velocity of 6,700 to 7,000 f t/sec' and a shear rave velocity of 1,700 to 2,200. f t/sec (equivalent to many sedimentary bedrock types).
Using the seismicity and site conditions defined above, the worldwide strong motion data base was searched for accelerograms produced by earthquakes.
within the appropriate magnitude and-epicentral distance ranges and recorded at sites whose geologic setting and/or foundation conditions (based on shear wave velocity data, if available) resemble those at Rowe. The selected accelerograms were corrected for instrument response in accordance with state-of-the-art procedures, Response spectra were then computed for each available component.
j The data selected to represent the earthquake in the site province included 62 horizontal components (17 earthquakes). The earthquake magnitudes ranged f
from 4.5 to 5.4 (mean magnitude.4.8), and the epicentral distances ranged from 1.7 to 33.2 kilometers (mean epicentral distance 15.0 km). The selected accelerogram data set included 22 recordings from California (12 from 1975 Oroville earthquake sequence) cr.' 40 recordings from the 1976 Friuli, Italy earthquake sequence.
.For. the distant earthquake eight (8) components (three earthquakes) met 7
5 r--,
t the_ seismicity-distance criteria. These accelerograms, all from Ca lifornia earthquakes, have a magnitude : range of. 5.4_ to 6.0 ' (mean magnitude 5.7) and '
an epicentral distance range-of 53 to 85 kk.(mean epicentral distance 69-km).- The shorter epicentral. distances in the selected data. set account
~
for the slower attenuation for eastern United States earthquakes.
The seismic design response spectrum recommended for the Rowe site, was
~
obtained by enveloping the mean spectra derived for the site province and for distance sources. :The spectrum developed for a seismic event within the site province, governs at periods less than.5 seconds while the spectrum developed for a distant seismic event outside the site province governs at periods. greater than.5 seconds. The mean and 84th-percentile envelope of these spectra are shown in Figure 3.
3.3 Conclusion and Proposed Spectrum The aforementioned studies indicate that.a site-dependent response spectrum with peak ground acceleration of 0.06g to 0.07g is conservative for the Rowe site. This peak ground acceleration is twice as large as the maximum cale 21.ted historical value based on more than 250_ years of record.
To arrive at an Interim Design Basis Spectrum, the Site-Specific Response Spectrum has been scaled up'to a peak ground acceleration of 0.lg and its shape has been modified to maintain a constant peak velocity above a period of'O.4 seconds. This is consistent with Appendix A of 10CFR100 which assumes a peak ground acceleration of 0.lg.
Figure 1 shows Yankec's Interim Design Basis Response Spectrum.
It should be noted, that the Interim Design Basis Spectrum has a peak ground 8
.C acceleration (0.lg) more than three times higher than the maximum calculated
. historical value based on over 250 years of record. Furthermore,-the proposed spectrum is well above the mean of representative real data at
~
~
.all points and exceeds.the median. plus one ' standard deviation (84th s-percentile) offrepresentative-real data.at.-periods above 0.9 seconds.
4 1
a i
i I['
J.
i t '
4 I
1 '
r l
9
4.0 PROBABILISTIC ANALYSIS OF SEISMIC HAZARD Yankee-has performed' a' probabilistic analysis to estimate the peak ground' acceleration and uniform hazard spectra for dif ferent return' periods. at the Yankee Rowe site.
The probabilistic approach utilized is basically that developed by Cornell (5),
and further developed by McGuire(6). This method allows for the' utilization of geologic, seismolegic, geophysical and historic information.
It also accounts for uncertainty in the attenuation models.
Input parameters to the Cornell /McGuire program are:
1.
Source area geometry (provinces) 2.
Seismicity of each source area 3.
Upper ragnitude cutof f (upper bound) for each source area 4.
Attenuation model 5.
Measure of' uncertainty (c).
The source area geometries (tectonic provinces and structures), Figure 2, were determined by Weston Geophysical Corporation in their August 1979 report (2),_
Figure 5 is the actual source area geometry used in this analysis. These areas were defined by:
1.
Geologic interpretation 2.
Regional ~and local' geophysical data; i.e.,
gravity, aeromagnetic, seismic 3.
Historical' and recent seismic activity.
'10
8
.For t e. purpose of this' analysis, the seismicity and upper magnitude cutoff h
- for each source area were determined by Weston Geophysical. 'Mean annual recurrence rates were ' computed - by Weston, using only segments of ' the earthquake catalog. assumed. to be complete on the basis of known population distribution in~ time. ' Table 1 presents the input seismicity parameters including recurrence rates and. upper bounds'for all 13 source areas.
TABLE 1 SOURCE AREA SEISMICITY Province or
, Source Area Lower Bound Upper Bound Beta Rate /yr.
1 3.50 6.00 2.35'
.741
-2 3.50 5.30 2.60
.121 3
3.50 5.30 2.53
.156 4
3.50-5.50-
-2.56
.610 i
5 3.50 5.30 2.30
.377 6
3.50 5.50 2.30
.291 1-7 3.50 6.50 2.12
.724 8
3.50 7.50 1.50'
.624
~
9 3.50 6.00 1.80
.098
-10 3.50
~5.30
- 2.30
.131 11.
3.50 6.50 1.98
.460 12 3.50 7.50 2.19 1.758 13 3.50 6.50 2.07 1.062
- s.
s
'% }
4 l
11
~
\\
t 4
- +
A The at enuation model used in the analysis was developed by.Weston Geophysical, and is appropriate for the area, since it'is based on the observed intensity attenuation for 4 well documented Northeastern U.S.
earthquakes. - Peak ground acceleration and velocities were associated to the predicted site intensities using tae equation of McGuire, 1977.
~
The uncer'tainty in'the atianuation model is defined by in a or in vg = 0.70.
g
- This is approximately equal to a factor of 2.
The distribution associated with the uncertainty is truncated at three standard deviations.
The Cornell /McGuire program output is an estimate of the. peak ground acceleration (PGA) and velocity (PGV) at various probabilities. To determine the uniform hazard spectra, this output is used as input to NUREG/CR-0098.
For this analysis, the median uniform hazard spectra are presented for the 4
10-3 and 10-4 PGA '""Y at 5% damping.
Figure 6 is a.. plot of-the 10-3 and 10-4 uniform hazard spectra and the proposed Yankee Interim Design Basis Spectrum.
The results of this analysis indicated that the Yankee spectrum is between the 10-3 and 10-4 annual exceedance probability.
These.results are comparable to the 1,000-10,000 year return periods which have seen implicitly. accepted by the NRC in recent licensing decisions.
In particular, this return period must be-viewed with reference to the relationship betweenlit and the " risk" of its occurrence during the operating life of the facility. " Risk" is defined as the chance of some undesirable event. For new facilities, the operating life is usually about 40 years.
12
~
-The'following equations will~show the relationship between. return periods and'" risk",.
The return period T is defined by T =.1/p, where p is the annual probability of'some slue x being equalled or exceeded.
Using the fundamental rules for the calculation of: probabilities, one can derive expressionsIfor the probability of various ccmbinations of events 1.
The; probability that x.will occur next-year is, by-definition P = 1/T (1) 2.
The probability;that x will not occur next year is q =.1~
p = 1 - 1/T (2) 3.
The probability that x will occur at least once in the next j. years is the sum of. the probabilities of its occurrence in the 1st, 2nd, to jth years.
j = p + pq + pq2 +,,, + pq -1 j
P
= l_
qd = l - (17-1)3 (3).
T Equation (3) is commonly referred to as the risk equation and is written R = 1 - ( 1)3 (4)
T where R is the risk of seeing the event associated with the return period T over the.timeframe J.
13-
'd m(e e
For comparison purposes, it is'more meaningful to specif/ some risk-that must not.be exceeded-duringlthe design life and then calculate,the required
~
return perfod of th'e design event., Table 2 gives gives-the values'of the
~
appropriate rotarn period.T. corresponding to a number ' of values.'of - the -risk R and design periodj.
i.
TABLE 2 Risk Associated with' Return Period and Operating Life Operating Life (j)
. RISK (R) 2 5
10 20 30 40
= z RETURN PERIODS T 4
50 127
'250 500 T9 1000 2
100 250 500-1000.
1500 2000
.4 500 1250 2500 5000 7500 10000
.2 1000 2500_
5000.
10000 15000 20000 1
4 f
l t
14
. e'-
To illustrate the'use of this Table, suppose a' nuclear facility is being
~
se.'smically designed with a projected. operating life of 40 years, and that:
a 10-3 seismic response. spectrum is.specified. This equates to a 4% chance The table of experiencing the design event. during the 40 year period.
illustrates that a plant with an operating life of 10 years, requires onlyL a 250 year return; period response. spectrum to maintain an equal risk of i
4%. Likewise, over a 2 year interim period, a 50 year return period response spectrum would have an equivalent risk of 4%.
Based upon this,-it is apparent that event return period by itself is not a reliable measure of the hazard, and can be misleading if it is not viewed in the context of the risk associated with its occurrence during the operating life of the facility.
With respect to Yankee Rowe, the proposed response spectrum has a return period on the order of 1,000 years. Over the next 2 years, the chance of 1
not observing this event.is 99.8%.
Conversely, the chance of~ observing the event is only 0.2% which is less than the risk implicitly accepted by the NRC in recent licensing decisions.
t 15 4
.5 p.
_~
5.0 SITE AMPLIFICATION l
The issue of site seismic ' amplification has been addressed in two ways.
First,' site amplification was incorporated directly into 'the development of the site specific spectrum because this spectrum was developed fron real earthquake data which 'was recorded at sites with geologic / geophysical characteristics simIllar to those at Yankee Rowe.
L Second, a theoretical study of the potential amplification of ground motion from local bedrock topography and basin effects was undertaken by Drs.
Vanmarcke, Kausel, Aki and' Bouchon, using the Aki-Larner-Bouchon method.
A variety of _ wave types, P, SH, SV, and varying angles of incidence, were input to a two-dimensional model of the alluvial basin of the Deerfield' River geometry at the plant site' for the frequency range of 0 to 8 Hz, and i
amplifications of sinusoidal inputs have been evaluated.
Results indicate the absence of any large amplification of ground motion in the range of 0 - 2 Hz.
Amplification factors are restricted to values below 1.25 in this range.
This model does predict higher amplification above 2 Hz.
However, material
- damping in the soil column was neglected in this basin-ef fect study. When a realistic amount of. soil column danping is incorporated, basin effects vill decrease and amplification at higher frequencies is greatly reduced.
Based on these two studies, it is concluded that:
16 l
F i
6 Site amplification is adequately incorporated into the development o'
of the Yankee spectra.
There is no significant site amplifi:ation at the longer periods o
of the spectrum.
i 17
V
6.0 REFERENCES
1)
Weston Geophysical Corporation, " Geology and Seismology, Yankee Rove Nuclear Power Plant", January 1979.
2)
Weston Geophysical Corporation, " Eastern United States Tectonic
. Structures and Provinces Significant to the Selection of a Safe Shutdcwn
}
Earthquake", August 1979.
3)
-Weston' Geophysical Corporation,1980, " Site-Dependent Response Spectra, -
l Yankee Rowe", February 1980.
4)
Weston Geophysical Corporation, "In-Situ Velocity Measurements Yankee Nuclear Power Station - Rowe, Massachusetts", March 30, 1979.
5)
- Cornell, C.A., " Engineering Seismic Risk Analysis", Bull. tin of the Seismological Society of America, Vol. 58, No. 5, October 1968, pp.
1583-1606..
6)
- McGuire, R.K., " Fortran Computer Program for Seismic Risk Analysis",
United States Department of the Interior Geological Survey, 1976.
I i
18 1
J
~
./,
o%
' +p t.,
0, e,.
/
10' O
E
/\\
bR sio -
/
o i V
/
N/
N 5
10
-V
'N/
N/
N
\\
N 10 2 10-'
loo 30' PERIOD (SEC)
. INTERIM SEISMIC DESIGN BASIS SPECTRUM (5% DAMPING)
Figure 1 9
Mk 31k$m HISTORICAL EARTHQUAKES
- stew
%.A>hl.
,,r c
"..y g
p.
./
lp,f' e
a' I
5' '
[
k.
UPLIFT h
3 e
s *' g,r t
. [' ~
d'$' 9 Q d
pgg%
YANKEE
'" p
- A#
N APPAyAC8\\b 5
9u C
SOUTHEASTERN W
F R09 t N NEW ENGLAND
\\
PLATFORM 1
hfLA
?
/
4 4
. Md *
$ b, i
j*
Figure 2
. j/
/
0, 50 1Co V LLs f
o s;
is e.
v f
g
O 9
10'E/
h/
\\/
\\ r
/g eain oESiGn ens N
/
\\
~~%
\\
/
/
g
/
\\
W g
h
/
/
\\
845
=,e
.) /
h
/
63%
/
~
/
/
/
10
~
N l
~
79.,
PERIOD (SEC)
SITE SPECIFIC SPECTRA (5% DAMPING)
Figure 3
CALCULATED PEAK HISTORICAL GROUND MOTION FELT AT YANKEE ROWE(NUTTLI-HERRMANN ATTN) l l
320-300-280-260-240-LLL/ TERA 220-200-180 Z 160-su 14 0-YANKEE f
g INTERIM SPECTRUM O 120-100-s y
80-Z 60-40 =
M AX. HISTORICAL M
AT TENUATE D 20 2
A, 4r it u
.1'4
.1 S
.1 6
.17 18
.1'9
.20
. O'1
.02
.03
.O' 4
.O'S
.O'6 07
.O'8 O9
.1 0
.11
.1 '2
.1 3 ts ACC E LE R ATIO N (9)
s u
.N -
SOURCE AREAS g
y
\\
W
- y. _.. _ m N
\\,
7 Ex
/
(
\\
h p
/
N:
1
'/
/
i' 3
/
b$
6
/
BACKGROUND ROWE s
~ ' -
s.
't v
.o--
10-e
~
v
/
-. - ~
9
}
x
~
g,,,
/
./
V
/
5 4
f
~
13
,/
-s
'N.
f'"
,;- *g
~
^~ ~
W?u
,,,rg 9[sj e,[g
,N
/f[W 1 (VH;Iis Ing;
)
i!
2 u,
h
1
'9 o%
/ +,
sq 0,
q.
4, j
p 10'-
/
j----
- go_4 9
~
/
/
/
C
/
s-
/
10~3
~
/
/
,
- G r
/
5 4
w f
~*
R
/
/
8;o _
//
/
o
\\.
y
/
_ \\
/
//
INTERIM DESIGN BASIS \\
~
,/
l s'
/
j
/
/
I
- /
/
t
/
/
lO'l
\\
/.
I
~
l 10**
go-10 10 5
PERIOD (SEC)
INTERIM DESIGN BASIS SPECTRUM'AND
'NUREG/CR 0098 MEDIAN UNIFORM H4ZARD CURVES.
(5% DAMPING)
Figure 6 o'
.