ML20093B616
Text
-
c-g
,g.e Me LA@ -SE1SMic iMpeT NAy e..tLs E
AndL -
o v
a um a ony cow.un.ex T
.j
- t.a w:1.. <.t.. u c. :nsus x
AUG 0 7 bed i
MEMORANDUM FOR-
. Robert E. Jackson, Chief Geosciences Branch, DE
-THRU:
Leon Reiter, Leader Seismology Section Geosciences Branch, DE FROM:
Jeff_Kimball, Seismologist Seismology Section Geosciences Branch, DE
SUBJECT:
SEISM 0 LOGICAL INPUT PARAMETERS IN RELATION TO GEOSCIENCES AND GE0 TECHNICAL REVIEW 0F MIDLAND 1 & 2 On July 2, 1980 several staff members of the Geosciences and Hydrologic and Geotechnical Engineering Branches met to discuss various methods that could be used to generate a response spectra for the Midland site.
These methods included use of the Standard Revicw Plan (Reg. 1.60),
real time histories, and' Newmark-Hall (NUREG/CR-0098) using peak values for acceleration, velocity and displacement. Formerly the Michigan Easin (the region within which the s.ite is located) was considered to ba a tectonic province. Currently the staff position is that.the Central S uble Region cannot be subdivided into separate tectonic provinces and tnat the controlling earthquake is similar to those occurring in Anna, Chio (March 1937). Although a detailed probability study has not been a tempted, the applicant has suggested (response to NRC question Q361.7) that the Michigan Basin has less seismicity compared to the entire Central Stable Region.
The Anna Chio earthquake had an estimated body wave magnitude (eblg) of aout 5.3 and an intensity of about VII-VIII. Both magnitude and intensity ware used to generate different response spectra.
Results from the above cethods indicate that the S4th percentile response spectrum calculated from real time histories (fra SSSP M-5.3 +.5 soil sites) is similar to Reg.
Guide 1.60 anchored at.139 at requencies or mterest. When using the
-Standard Review Plan and the intensity of the controlling earthquake (VII-VIII),
the response spectra (Reg. 1.60) is anchored at.19g using the trend of the reans from Trifunac and Brady (1975). The applicant originally anchored the response spectra at.129 and used a Housner spectrum over the entire frecaency range. With this in mind (characterizing the earthquake in teras of magnitude and (or) intensity), I suggested that Reg. Cuide 1.60 anchored at.139 or appropriate real time histories could be used as response spectra for the site, f-
- p --
y.
l go om 567f,
l- - - - -
_p 1
AUG 07 One topic that was discussed following the presentation was possible problems due to soil amplification at the Midland site. This problem was briefly discussed in a memo from S. L. Wastler to you on March 17, 1980.
Because of the wave velocity contrasts in the natui al soils beneath the plant and the man-made fill supporting and surrounding plant structures, there may be amplification of the vibratory ground motion.
Same of the factors that need to be considered in specifying ground motion at the site include the following:
some buildings are founded on natural glacial till, some foundations are set on both the natural glacial till and plant fill, some are entirely on fill, other because of poor fill, are to be underpinned to till. The plant fill original design specifications ware not achieved (e.g., a shear velocity of 1350 ft/sec was adopted for the plant fill, but has now been reduced to 500 ft/sec). Also the effective acceleration value suggested (o.13 ) is that expected at the top of the 9
uppermost till unit but, because of specific plant fill properties and site soil conditions, may not be the peak acceleration at the top of the fill (plant grade). A computation done by Joe Kane (Geotechnical/ Engineering Section) using site soil properties Qg'lintf occurs at a frequency of erty containment indicates that amplification through the signiric about 3-4Hz. A soil response anal is is needed to know if the soil and fill conditions would produce anomalous accelerations within the fill.
Taking all the above factors into account the following ground notion specifications alternatives are possible:
bdt de.c 1.
Require the applicant to use suggested acceleration.(such as Reg.
'I t 5
- O b t
2.4 Require the applicant to use the S4th percentile spectra f yrombekhW locse -m,ds Sh set of real time histories as input at the top of the till. They h po.4-9 then would have to calculate soil-structure interaction, including k4E <
possible soil amplification for structures founded in the fill. This y
method.could involve the use of the computer program, SHAKE, by the applicant, which, because of apparent program. limitations, has some-u times calculated unreasonable acceleration levels at the around M w} amplii7* g ation for'%
powerprojects.gdtostu surface when us so1 tjagr nuclear J w$ a g r @yrde %{or w@di. 7q Q % 3 1"
,h;gh t w yt c
RequiretheapplicanttogthrregtimedataforM=5.3+.5,R425km 9
as input to the SHAKE program at a rock at rock sites. Use this dat p
$49, outcrop of the modeled soil column, then calculate thg o g atio ofn* #N bg the motion thru the glacial till and engineering fill.
ce e T
histories at the various foundation levels could then be calculated h**Ch taking into account the depth and dynamic properties of the till and fill below the various foundations. Again the program limitations of SHAKE must be looked at because of the great depth to the rock at this site, u _.
T L - -
2 L.
- TJ.:
jo a f..,' f.
q
.! AUG 0 71980 i
p 4.
Require the applicant to gather real-time data for M=5.3 +.5, h
.R4 25 km at specific soil sites assuming that this data silit
' ggg already represents soil and fill amplifications conditions at this particular plant. Assume the acceleration history occurs
- M%wM at the top of the fill and use the SHAKE program to 'deconvolve and calculate. accelerations and seismic motion at the various foundation levels. Again the SHAKE computational limitations apply.
b These alternatives were discussed on. July 22,.1980 by.R. E. Jackson.
L. Reiter, J. Kimball, T. Cardone. L. Heller and J. Kane of the Geosciences Branch and Geotechnical Engineering Branch.. L. Heller initiated discussion by saying he thought method number 2 would be the most appropriate because of the complex foundations to be utilized for this plant and because the spectra to be developed represents the motion of naturally occurring soils.
Others present agreed with this and the discussion centered on a statement of staff position. This position would take into account the fact that the applicants limiting earthquake differs from that currently accepted by the staff and that soil amplification problems should be addressed.
The major elements of the position would take the following form.
1.
The controlling earthquake is assigned a body wave magnitude of 5.3.
It should be noted here that this magnitude is.also suggested by Nuttli (1978) as the maximum when using Residual events (these left over after Ann'a, Wabash etc. are removed) for the Central United States.
4 2.
The applicant should use a collection of real time histories for M=5.3 +
.5, R<,25 km and soil sites. This collection could come from TERA Corporation (Seismic Hazard Analysis:
submitted to Lawrence Livermore Laboratory 1979) but it would be suggested that the applicant update this data set.
3.
The84thpercentile(meanplusonestandarddeviation)responsespectra should be used as input at the top of the uppermo,st coh'esive. till unit.
Above the till is a thin but variable sand layer and plant fill. The applicant could then calculate the motion that would occur at the various structural foundations as a part of the soil-structure interaction analysis i
which should adequately address soil amplification.
4 The NRC staff is aware of the limitations of SHAKE and SSI computational programs and recognize that unreasonable response acceTeration values are a possible outcome of the applicants analysis using a broad-band spectra 4
at the. top of the till. These results however, are needed for use in a qualitative sense to address the question of soil amplification and the seismic response of the plant fill.
4>
4
(
__.,..__,,,...__._m m
....~.,.,i_..i.
.]
_m
- e..
)
-s t..p i.UG g7 g39 Once the branch position:has.been forwarded, GSB, HGEB, and SEB should
- neet to discuss the current position in relation to what the applicant has designed for.
?
Jeff Kimball, Seismologist
-Seismology Section Geosciences Branch, DE
-cc:
R. Vollmer J. Knight.
- L. Reiter-R. Rothman
- T. Cardone J. Kane-L.. Heller F. Rinaldi
- J. Kimball F. Schauer R. McMullen H. Levin L. Heller D. Hood M+4.6t aN-N-4a er a
y f,
in'*.
~.
.....t....w
_ t.>..s.: ew
_;_.e ti e i e i.
i,.33 f
./<jO y
,;h. at : -
r-
' [Q f-
%(s d
I.$o 4
.g e
i
'N
- v Nb{%
/Cf_ c go',%:
'c' X
c
-s
',10f
-g 5
p s,..
K 6
3,tv@3
/je
,c
\\/'%
./
- c-s's
'$ n?,'b
(
,.y.
/-
~
,N a
0
/
/
b k[. :
(n,M"\\
[
/
i
.f
<ji
,y/
l
/
.ili'W t
gN,., *f
.j
(
/j
, -)' l'
/
. l' ?g
/
((c5. b.
)
/'
t.19.-
- V.k
- $'y*
GW'
,j s':
/
G l.c.
f b'
{
n x
c f
.f9
/
s-
,.,,./,.i n -- -
- . t,
,_,.. y _,.,,.,,.,_ _ _
- 3...,.
i i
i.
i i.-
9I 0.1 1.0 10.0 V 2.RICO-SEC.
4 e
e
- =--,m._
m w c
--u ya
-,-4 y
w,
-l - [f iuteo MIDLPrhih
-F5Pr9. No\\ 7 Sus by, eeroe. yrnnt naeut s 3 C.A 4 uhd 3
631-1-ModshiccbTAFT bne otsfor) % pmdoce afrIcN Whsam en*Wed & "m cgec ca
- Apyw4o kan pfpAy#5'spp rupeor ycra-eu1 Moo Y used ismpcl mss wiedel (Eq.M-9 kTabi< b i-3 {te cBE,46%t
~
M.3 3
F%R
- 73. 2.s -38 a 8 l% anal se cudockedAd,Vol &
cuk resses scil o.m ab hsc 3 % m&J LTnwe.hJ c< Mev~ph v
ne U$th k dbVvwint pmt OcceIercuew
- ca SSE eve.,E ck Js upo o aanship Aetno ts aced oc a. bc p
ow 9.cc oJ musin n spdev-
- 4. Set com&ho,s ed w 'ta L ate c, eems acceum w
w t. u rt.
's cowsidered +c ccn servccm. uh# ht44v:
ver t> c,oas'sde( an,gk.bcsson l
C
,...[_,.
. ~_ $... -_,2.. I li -.l.
....,.___.21,--.-...-
'~~...,_.-.,__.__SI_-
'