ML19242D540

From kanterella
Jump to navigation Jump to search
Seismic Risk Analysis for Battelle Memorial Inst,Nuclear Research Facility,West Jefferson,Oh
ML19242D540
Person / Time
Site: 07000008
Issue date: 12/29/1978
From:
TERA CORP.
To:
Shared Package
ML19242D535 List:
References
REF-PROJ-M-3 NUDOCS 7908150393
Download: ML19242D540 (41)
v • d • e


Text

{{#Wiki_filter:. O SEISMIC RISK ANALYSIS FOR BATTF11 F MEMORIAL INSTITUTE NUCLEAR RESEARCH FACILITY WEST JEFFERSON, OHlO x=- e: : LAWRENCE LIVERMORE LABOR ATORY P.O. Box 808 Livermore, California 94550 Attention: Mr. Don Bernreuter Project Manager {g Teknekron Energy Resource Anc%+s TERA CORPORATION 2150 ShCMuck Avenue Bemeiev. Ccofcrnic 94E 415 845 5200 Decerrber 29,1978 i (., /: D{.$he 7908150 ' 72,. ( .,e

TABLE OF CNTENTS Sec tion Pece 1.0 !NTRODUC TIO N AND 5UMM AR Y............................... 1-1 2.0 SEISM IC RISK ME THODOLOG Y................................. 2-1 Theory........................................................ 2-3 3.0 GE O L O G Y AND HYDR O L O G Y.................................. 3-1 4.0 SE!5MOLOGY.................................................. 4-1 5.0 C A LCUL A TIONS AND RESUL TS................................ 5-1 Input......................................................... 5-1 Results....................................................... 5-9 R e sp cn s e 5 p ec tr u m............................................. 5-1I 6.0 BIBLIOGRAPHY............................................... 6-i C (, Q, lCJ i ERA CORPORAilON

14 INTRODUCTION Ato

SUMMARY

in this report, TERA Corporation presents the results of a detailed seismic ris,( analysis of the Battelle Memorial Institute's Nuclect Resecrch Fa:ility ct West Jef f erson, Ohic. This report is one port of a Icrger effort being directed by the U.S. Nuclect Reculatory Commissicn. The NRC's obiective in comm:ssioning the overcli report is to essess and improve, to the extent practicable, the coility of this facility to withstand adverse ncturcl phenomenc without loss of ccpccitv. This report focuses on ecrthquckes; the other nctural hczords, which c e c0 dressed in sepcrcte reports, cre severe wecther (strong winds and tornados) and floocs. The overall report will provide on essessment of the consequences of an occicent resulting from any of these nature! phenomenc. The essessment will express c quantitctive probcbilistic measu e of the potentici structurcl damcge and the relecse function. It will ciso provide a pruocbilistic estimate of the resulting dose of rcdiocctivity to the public. Inis stucy was performed uncer contreet to the Lawrence Livermore Lccorctorv (LLL). The study wcs directed by D. L. Bernreuter of the LLL Nuclect Test Engineering Division. At TER A, the study was managed by L. Wight. To ensure credible results, very sophisticated but well-occepted techniques were employed in the onclysis. The calculationc! method we used, which is bcsed on Cornell's work (1968), hos been previously applied to safety evoluotions of major projec ts. The historical seismic record was estcblished af ter o review of ovcilcble litero-ture, consultation with cperators of local seismic crrays and excminction of appropricte seismic data bases including the USGS, LASA, NOAA, USC and GS, and NElS bcses. Secouse of the aseismicity of the region around the :ite, on onelysis different from the conventionci closest opproach in a tectonic province wcs adopted. I-! s,8 h b TERA CCRPORADCN

Earthquakes es for from the site es 600 km were included, es wcs the possibilitv of earthquckes et the site. In addition, various uncertainties in the input wer? eclicitly consimred in the mclysis. For exemple, c!!owance wcs mode for both the uncertainty in predicting mcximum possible earthquckes in the region onc the effect of the dispersion of dato cbout the best fit ettenuation relation. The cttenuction relationship we coplied is, we fee!, the best avcilcble. it is derived from two of the most recent, advanced studies relcting ecrthqucke intensity reports and occeleration cnd is unique in that it incorporctes a thorough molys ' af the effects from the e 18o6 Charleston, South Ccrolim., dcrthquake; e !t is bcsed on o recent analysis of cimost 1500 world-wide strong motion records; cnd It is consistent with +5e newly cNoilcble strong motion e accelerotion dcto for the Ecstern-Centrel United States. Finally, cnd most important, the project hcs benefited from significant contr'- butions from, and finct review by Professor R. Herrmann, St. Louis University, a seismologist with pcriiculcr expertise in the local and regionc! seismology. The results of our risk cnolysis, which include o Scyesicn estimate of the un-cortointies, are presented in figure I-l expressed cs re' urn period cecelerotions. The best estimate curve indicater that the BMI facility will emerience 5% g every 200 years and 10% g every 900 years. The bounding curves roughly represent the o,ne stondcrd deviction confidence limits cbout our best estimete, reflecting the uncertainty in certain of the input. Detailed examinction of the results show that the accelerotions are very insensitive to the details of the source region geometries or the historical ecrthquake statistics in ecch region md that each of the source regions contributes almost equally to the cumulative risk at the site, if required for irructural analysis, 2,:celeration response spectro for the site can be constructed by scaling the mecn response spectrum for cllu/um in W ASH 1255 by these peak accele:ctiens. ~L[ g 1-2 4 TERA CORPCRAiiCN

9 a 10,000 7 I f i / i 4 /' ~C/ .c .; k ' ' /s' j l p @ (' I \\ f / c / '2

  • /

l.000 s / / / m. / / s f f f / / / / / / /'/ f' 5 / f ,/ = / ,/ / / =: / / r 100 -/ / ,' /,' ~ f / s ~ E //i / / / / // ~ // / t I/ /// ~ / // / / / / i t 10 / i I l l 1 l I i t i l l i l I l l I I i i I l l l I i I l 2 4 6 8 10 12 14 PEAK HORIZONTAL ACCELERATION (% g) FIGURE I-[ RETURN PERIODS FOP, SEISMIC ACCELERATION AT THE SMI WEST JEFFERSON FACILITY b+ I[j , Cj IERA CORDORATCN

a 2.0 SE!SMIC RISK METHODOLOGY A seismic risk analysis is only as credible cs the risk cnclysis methodology onc the input to it. This section presents the bcsis for our selection of c probcbc-listic Poisson model for the risk cssessment at the BMI facility. There cre generally two distinctly different cpproaches to seismic risk onclysis: probcbclistic cnd deterministic. Using the oeterministic cpprocch, the cnclyst judgmentally decides thct cn ecrthqvcke of a given magnitude or intensity occurs at c specific location. He then ettenuctes tYa ground motion from the ecrthquake source to the site end determines the effects of that qvcke. The problem in using this cpprocch is inct it is difficult to define the mergin of safety or the degree of conservctism in the resulting design percmeters. Anclysts are of ten csked to provide information on the "mcximum possible" or "most probcble" ecrthquakes for design purposes, but the deterministic approach does not ecsily provide those cnswers. A probcbclistic cpprocch, on the other hcnd, qucntifies the uncertainty in the number, size, and locction of possible future ecrthquakes cnd allows en analyst to present the trade-off between more costly designs or retrc'.its and the economic or social impoet of a fciture. Because the product of a probcbolistic cpproach is o measure of the seismic risk expressed in terms of return period, this trede-off con ecsily be qucntified. Although the probcbilistic cpproach requires significantly more effort than the deterministic cpprooch, it hcs the following advcntages: it qucntifies the risk in terms of retum peris #: e It rigorously incorporates the complete, m rice! seismic o record; it con incorporate the judgment ed exper mee of the e analyst; 2-1 ' ' ') EA CORPCRAilCN

4 it occounts for incomplete knowiecge regcrding the loco-e tion of faults; e it has the flexibility to cssess the risk ct the site in terms of spectral ccceleration, velocity, displacement, or ecrth-qucke intensity. The method is particularly appropricte for the BMI fccility for two recsons. First, cs will be shown below, the Columbus crea is very cseismic cnd it woulc therefore be very difficult, using conventionc! deterministic methods, to estcb-lish a design ecrthquake mognitude. Second, the seismicity of the ecstern Unitec Stctes is very diffuse and cennot be correlatec with surface faulting cs it con be in the westem United Stctes. The location of the cesign ecrthouckes in the ecstern United States is therefore pcrticulcrly uncertain. The strength of the probabclistic approcch is its ability to quantify these uncertainties. The credibility of the probcbilistic approach has been estcblished throv9 ' ce-toiled technical review of its cpplication to several important projects and crecs. Recent applicaticns include essessments of the seismic risk in Boston (Corne!!, 1974), the Sen Francisco Boy Area (Vogliente, 1973), the Puget Sound Arec (S tepp, 1971) and continentcl United Stctes (Algermissen and Perkins, 1976). Results of these studies have been applied to, among other crecs: Development of long-range ecrthqucke engineering re-e search goals; Plcnning decisions for urban development; e e Environmental hczcrds ossociated wiih the milling of uranium; and Design considerations for radioactive weste repositories. e This diversity of cpplicction dernonstrates the inherent flexibility of the risk assessment cpproach. h 2-2 g \\lG TERA con:crdTCN

THEORY The risk calculations can be fundamentally represented by the total probcbility theorem P [ A/m and r] f P [A] M ') f (r) dmdr I = R where P indicates probcbility. A is the event whose probability is sought, end M and R are continuous, independent rcndom vericbles which influence A. The probcbility that A will occur een be calculated by multiplying the conditionc! probcbility of A, given events m cnd r, times the probabilities of m cnd r, cnd integrating over cil possible values of m cnd r. In our assessment of the BMI facility, A will be taken cs maximum acceleration cnd therefore P [ A/m cnd r ] will be cerived from dcta relcting peck acceleration to epicentrol distence and ecrthqucke mognitude. Of ten known es ettenuation data, these cota are usucilv lognormcIly distributed cround c mecn relationship of the form (McGuire,l 77c). M 2 A = C; e (R + r,) The distribution on earthqucke mognitude, fM(m), cun re dily be derived from cn octual or postulated frequency relationship of the form log N = o-bM where N is tne number of earthquakes having magnitude grecter than M, cnd o and b cre constants chcrocteristic of the pcrticulcr source region under con-sideration. It follows (Cornell,1968) that f con be derived from the cumulative M distribution function, FM, which has the form, D' ,i $ 2-3 he TERA CCRPCRATICN

M = k (1-e-OM) F where k is a normalizing constcnt and S = bin l 0. The distribution on distence, f (r), depends on the geometry of the problem p under consideration. For simple geometries, the distributions con of ten be integrated analytically. Reclistic geometries, however, require numerical evcl-vation of the integrcl. A very versctile computer progrcm has been developed (McGuire,1 7eb) that incorporates the theory presented cbove with c numericci integrction scheme that allows for evcluction of very complex source-site geometries. The theory of seismic risk cssessment by this cpprocch is outlined below. First, the historical ecrthquake record cnd locci citenuation dotc cre combined with the experience of the analyst to produce the functioncl relctionships cp-plicable to the crea under consideration. The source regions cre divided into circular sectors and proportionc! seismicity is o!!ocated to each sector. The total expected number of events cousina mcximum occelerations at the site grecter then a pcrticulcr test cecelerction cre obtained by summing the events from each sector within ecch source region. The risk associated with this test accelerction is then eclculcted under the conventionci cssumption thct ecrth-quakes have c Poisson distribution in time. It then follows that the return period is simply the reciprocal of the risk. t er i; .f 2-4 TERA CCRoCRATICN

3.0 GEOLOGY AND HYDROLOGY The Battelle West Jefferson Nuclecr Science Area is located in M.cdison County, Ohio, about 15 miles west of downtown Columbus (Figure 3-1), in en crec where the geology is principcily giccici in origin. The following sections which cre bcsed principcily on BMI (1074), briefly summcrize the geology cnd hydrology. Glocial deposits at the surface of the Bettelle site were deocsited cs the Wis-consin ice sheet, the Icst of the four grect gicciers of the Pleistocene Age, ~ receded. Some subsurfcce glacici deocsits probcbly originated curing the first of perhcps two major advcnces of the Wisconsin ice sheet; some of the deep giccici deposits in the buried-volley system probcbly cre relcted to glccici stcges ecrfier then the Wisconsin sheet. Glacial deposits comprise two main types: (I) till, or meterict icid down directly cs the ice sheet receded cnc wested awcy, which occurs in this crec principolly cs ground moraine or till plain; cnd (2) outwesh, or send and gravel, depositec in stratified Icyers by meltwater. The till is an unstratified metrix of comperctively impermeable cicy containing rock f ragments, sand, and grevel. The upland crecs of the West Jef ferson Battelle site cre covered with till in depths vcrying from 60 feet to more than 200 feet. In some picces, sand and grovel outwosh deposits underlie the ground morcine or are interbedded with the till at shallow depths; however, the sand and grave: deposits in the crea cre thin and discontinuous and cre thinly covered with alluvium (river-Icid deposits) deposited by the stream during overflow periods. Beneath the giccial and alluviel deposits in the crec are several hundred feet of cimost horizontal beds of limestone, dolomite, and shcle, which comprise the bedrock of the creo. These rocks are of Devonian and Silurion oges. Their surface contours are approximately 750 to 800 feet mecn sec level (MSL). The bedrock surfoce in the crea is deeply cut by a buried-volley system cerved by j 3-1 TEiiA CCRDCRATiCN

/ i- /w \\\\ /' .,,,..C. N 5Y -

  • g/

\\ w- \\ _ h -- I s." / . ',2.. 's.. ~ \\, z'mQ -lw:.aec ; + l /

- ^

c,- N, n. ,g. s. - t .~ 6 s s M D \\ T'\\ M., /. / t.:n ~ < b % W, c=. ' v, d}[ N>~[ ' ( ,( / \\ / N /- . /m N, c.-/ / -7/,/ s.:h(. >.- ..fs 1 3 N f x . X s f \\;y a<a s w.,..A._, x-g x x, y -_ n w. y mk i. ;. J. t3. Y,'.. -.1 ?<. =g s t }i:N.j?- c., t =t - :tz::::: y \\ ,L \\% ; ~ e -( x,. - - i g y s \\ ,\\ N. N b { ( f se \\ N\\L MO ~ %u -z -- c il ~- - ~ -K, pNgYstrf h.~ l E. 3 a +: ; aEFFE P - ,a p .,/- 2 . s,\\ 3, 1 / .~p Iq x i

v y {-

./) p' i N " :'

?'

- t,/ y. h= s f 0 ~ 7' h_. & [' %x, / L /< s-. ,~ i \\,. \\ > N-m, q Q-q s-s v 4 h lr ar- - 5 /. ; ,n >~ [ L a/\\ 4 / \\ ~I / ~s g Ss??e m' Jire FIGURE 3-I ROAD MAP OF BATTELLE'S WEST JEFFERSON SITE VICINITY oh '/d TERA CORPOR ATON

streams that drained the crea before the period of Pleistocene glaciation. The distance frcm soil surface to bedrock surface on the site varies from a few feet in places along Darby Creek to more than 200 feet in the buried-valley crea nect the northwest corner of the property. Clay cnd fine scnd are the principcl deposits in the deeper part of the buried-volley ;ystem and are not a source of ground water. There are two ocuifers, or sources of water, in the site crec. The shallow ocuifer is, of course, the dense clcy till. The deep, or principo!, aquifer is the limestone bedrock underlying the til!. Ecrlier wells in the site crea ranged in depth fro n 10 to 40 feet, which placed them in the glacial deposits. Till is not very permectie end yields water slowly. The effective velocity of water moving through cicv under a hydrculic gradient of one percent is reported to be below 0.004 foot per de for water moving through silt, sand, and loess under the some gradient, the rate is cbout 0.052 to 0.065 foot per day. Wcter movement in the till ct the Bcttelle site is probcbly within the rcnge of the latter figures, since the hydrculic grcdient of the water table in the crec is only slightly grecter than one percent. The present wells et the Scitelle fccility lie below the surface of the bedrock. The north well is cbout 150 feet deep, the centrally loccted well in the Life Sciences crea is 161 feet deep, and the south well is 138 feet deep. Bedrock was encountered at approximately 103 feet below the surface in drilling these wells. A new gelogic feature of the site is the crtificici icke covering en cree of cbout 25 acres that was formed by domming Silver Creek stream south of and down gradient from the reactor site. The surface elevation of the icke is 891 feet MSL. The source of ground water in the site crea is local precipitation. Rechcrge to the shallow equifer takes picce relatively uniformly over the area. Contours of the water table, which cre cbout 40 feet below the surface, are o subdued replica of the surf ace topography. Ground water moves downslope at right angles to the contours md follows a path similcr to s;rface runoff. In this case, surface runoff moves downslope into the icke, thence through the controlled dcm on the site 3-2 i TEiiA CORDCRATICN I.b h

into Big Derby Creek. All ground water in the site crec and thct entering on the site is ciready nect its p! ace of dischcrge. Test borings corrried out in 1970 for cn addition to the Het Lcberctory recifirmed the geology described cbove. Only isolcted pockets of water were encountered during the boring cnd foundation-piling excavation operations. These pockets were readily pumped out cnd remained dry, which indiccted thct there is no interconnection of the pockets with the icke. Flood hydrology calculation indiccted a capacity of relecsing water thct was about three times the inflow rate mecsured during the Jcnucry 1959 flood-It can be concluded thct the icke hcs not cdversely offected the hycrologv of the crec. h b 'f l f [j 3-3 TERA CORPCRATiON

4.0 SEISMOLOGY While the detailed elements of the seismic risk assessment are discussed in Section 5.0, the historicci seismic record is of such significance thct it is dis-cussed sepcrately below. A complete evoluction of the; historical record is the keystone to the risk cssess-ment because of the important time end spctic! distribution information it contains. With regard to time, the record provides detciled historicc! ecrthqucke frequency informction thct con best be representec by the relationship, log N = o-bM. The spatic! distribution of ecrthquckes cround the site een of ten be used to delinecte seismic source regions within which ecrthquckes have common chcrocteristics. Unfortunately, ecrthquckes have been reliably reported only since the 1930s when c nctionwide ecrthqucke instrumentction progrcm wcs stcried. The pre-1930 record is a very vcluable supplement to the recent recorded dotc but cue to spcrse settlement and secttered intensity reports, much of these dato cannot be reliably used in developing ecrthqucke statistics. Our genero! coprocch is to use the recent recorded data to determine the statistics for magnitude six cnd less earthquakes and to include the entire historical record in determinction of statistics for Icrger ecrthquake mcgnitudes. We have collected and integrated the data from several seismic data bcses to ensure the most complete coverage. The primary source of data was from Nuttli (1978). Data sources consulted by Nuttli include Earthquake History of the United States (Coffman and von Hoke,1973), United States Earthouckes (U.S. Depcrtment of Commerce) for the yecrs 1926 through 1972, Prelimincrv Determination of Eoicenters (U.S. Geological Survey) for the yecrs 1972 through 1974, Earthouckes of the Stcbic interior, with Emphasis on the Midcontinent (Docekol,1970), A Contribution to the Seismic History of Missouri (Heinrich, 1941), Seismolocical Notes (Seismological Society of America) for the yecrs 1911 through 1975, Guerterly Seismolociccl Bulletins of Saint Louis Universitv i.lf 4-1 TERA CORPCRATION

(Stauder et al., 1974-1976) for the interval June 1974 through March 1976, un-published lists of earthquakes compiled by J. E. Zollweg of Scint Louis Univer-sity, c list of ecrthquakes compiled by M. M. Vermo cnd R. F. Blakely of Indienc University and the Prelimincry Scfety Anclysis Reports for proposed nuciecr power plant sites at Mcrble Hill (Jefferson County, Indienc), Colloway (Ccilowev County, Missouri), Koshkonong (Jefferson C oun ty, Wisconsin), Hcrtsville (Troudcle-Smith Counties, Tennessee), Perry (Lcke County, Ohio) cnd Sterling (Coyugc County, New York). For identification of possible seismic source regions outside the region investi-geted by Nuttli (centrol United States), we excmined the seismic dctc bcse mainicined by the U. S. Depcrtment of Commerce NOAA. The dcte for the site crec were checked and extended to 1977 by compcring them with other independent dctc. The Alexcndric Lcboratories of Teledyne Geotech mcintcins c data bcse whien consists of LASA, t ISCS, USC & GS cnd NEl5 dctc, onc these data provided the most complete check oM extension. The dctc through 1974 were checked ogcinst the USGS dotc, provided by D. Perkins. Besices these direct compcrisons, the availability of local unreported dctc was considered by checking with loccl experts: Professor G.A. Bollinger Virginio Polytechnic institute and State University Professor Shelton Alexander Pennsylvanic State University Professor Walter Pilant University of Pittsburgh The resulting integrated data bcse for the crea cround Columbus is plotted in Figure 4-l. Note that the data bcse is specified in terms of earthqucke mognitude; this follows from our emphcsis on the recent recorded dotc, which is in terms of mognitude. Nuttli's data bcse is already in terms of magnitude; when we require converting intensities to magnitudes in other data bcses, we use the Gutenberg-P,ichter relationship (1956)M = l 4-2 TER A CORPORAECN (,;f

"2 "'atmo_a.g,,,,, ,,o m er. + Mv.g

n

,, ' *- w. / ,,. m. f ..-/,.. f t,.., o / .. n 7 w...+. 9.y'

  • v*

~...)<, l.),'., I 'd.* ;~,J. ... p. 4, - x 3 -l-tf-7k,\\'J'/

  • sor.

-rd ) , '..f ~\\,}'h 's .\\{ ;

  • o

+. m-

  • ;Ne /

1'/ .t U

.x

. '... / . y </ g. g. ,L m}..,'Q. ,e ,'f.. + - - J,, cc r-y.,_..q w y...

n.. :.

7, ~.. i, % g..,,i g.,.c_. ~ N. 'o l. s V C. ,x n... + .,o..,.,, .).4 o....., .,o....s, _ Q,g,,U o6o.m,=69 30. %j 010. n.. i, 4 1

c. ;
e..

y w h s,I p /y 9o9 70 4 I-lu n <t 's - I

-p A--- A,m aO 26 Si iSMiriI Y li'l Vit 'It ilI Y On 2

ilMI SIII

M = 1.3 + 0.ei,. The validity of this relction for the ecstern United States has been confirmed by compcrisons between it and the more current seismic dctc bcse (Chinnery cnd Rodgers,1 73). There are severcl importent fectures of the historicci dctc that must be incluced in cny risk assessment. The most important is the obvious c!vstering of seismicity in several localities. The most cppcrent clustering is in the New Madrid crec, site of the famous 1811-I2 ecrthquckes. These ecrthquckes, which were the lcrgest ever experienced in the ecstern United States, resulted in intensities es high cs V (M.M.) in the Pittsburgh crec. The shocks were felt cs for awav cs Boston, Massachusetts, and the totcl felt crec wcs by fcr the lcrgest ever experienced on this continent. The crec wcs well known for its seismicity even before the 1811-12 earthcuckes, with historicci occounts going back even into Indicn legends. As ccn be seen from Figure 4.1, the seismicity cround New Mcdrid hcs been relatively contcined, thus suggesting c local tectonic origin. Another cree of s?gnificent seismicity is the region cround Anne, Ohic. This crec nos been subjected to several earthquckes thct produced moderate dcmage (Bradley cnc Bennett,1965) including those of: b = 5.3 June 18,1875 m b " 4'7 September 19,1884 m b = 5.3 September 30,1930 m b = 5.3 September 20, 1931 m b = 5.3 March 2,1937 m March 9,1937 n'b = 5.3 Other crecs of repetitive seismic octivity inc!vae the Fairport-Clevelcnd, Ohio crec, the Attica, New York crec, and the Anna, Ohio, crec. The activity in the 4-3 .en TERA CORPORAilON lk J

Fairport-C!eveland, Ohio, crec has bean minor. The largest ecrthquckes associated with this orea are of intensity V (M.M.). The Atticc, New York, crec emerienced en ecrthqucke of intensity Vill (M.M..) (August 12, 1929) cnd two earthqvckes of intensity VI (Mi.M.). Another importcnt, although very distent, source region exists along the St. Lawrence Seewey. This region, which is one of the two most octive regions in Ccoadc, has experienced several ecrthquakes-most notcbly the February 28, 1925, St. Lawr ence Ecrthe S,<e. M , 0, i h bb.k v + TERA CORoCRAECN

5.0 CALCULATIONS AND RESULTS In the previous sections, we have described the regional seismicity cround Pitts-burgh cnd have discussed the most cppropriate method of risk onc!ysis. In this section, we cpply these concepts to the BMI site. The detciled input to the calculational model is described below, followed by a presentction of the results. IN;"? As described in Section 2.0, Seismic Risk Methodology, the input to c probcbil-istic risk cssessment comprises ecrthqvcke frequency relctions, ettenuction functions and a specification of loccl source regions. Because risk essessment calculations are very sensitive to the particulcr composition of the input, we consulted with several eminent seismologists during the prepcration of input for the BMI facility anclysis. Major contributions in this effort were made first by Professor R. Herrmenn (St. Louis Univers:. <), and Professor S. Alexander (Pennsylvanic State University). Source Recions Af ter a thorough review of the historical seismicity (Figure 5-1) and geo-logic /geophysicci parameters such as gravity, mognetics, tectonics, cnd surfcce geology, it wcs egreed that the most appropriate source regions should be very similcr to those defined by Algermissen and Perkins (1976). Their definition of the source zones wcs based en the recsoncble assumption that future earthqucke occurrences will Mve the some general statistics cs historical earthquakes and that the historicm variation of ecrthqvcke statistics from region to region can be used to delimit general source regions. The finc! definition of the source region's boundaries was bcsed en the average sepcration distance for earthquakes of the Icrgest intensities. The representatien of the source regions synthesized c!! the available historical seismicity data and the state of knowledge of the relationship between geologic structure and historical seismicity. We depart frorn their definition of source regions only where it is necesscry to provide more resolution into the seismicity around the site, or to analyze the uncertainty in definition of source regions. U f-o r: U V J f_ 5-1 TERA CORDCRATICN

"2 98_, ~3 94 gy .y) gy, oc 04 4 + %>q t +.* ~ ~;~. av:,x 1 , /- e,. I I '. *'l v 4 - [ 3.. a i e. ' =+y,2.3 z ~' a T&.. .l 1 .-,,' i g. 4, . )'. l >.H:;m */.1 . C' > 4, e---- .= + ... ?, . m>,. - g* vii, k c/

-{ +

N>..,.. s... s...T 40_. .g t. 4 +, / p J, " 4 y ,a s s.c. m( 3- ,.:y, l,p/ * :,A,,4 u + t ,..33 pg v', c t ,r..,,. m-

s.

y.. e,.. - 41 ).7, :.

t. l. v

,_t ~ u, s.9 T _. - C ) Y u ,, ;.(. l s. n- /, l, W, N <,' l(. e / 's / x ~ f ..m..,, .'{ J.., v w-40.m. 49 .l - ~~ e '. O. m, 5 9 z O60 m,=69 O. ,f-O ro-m,.19 eg I 1) c_ $ _, v +.x 9,y t 1 _1 T <7y V h > J<;> \\ i V) \\ / () 0 ?O .= =.. =, <,, - U W s imni., y,y r,l ge,q gg gy y ggg vgg.!!11I Y ()! => 76 % O IlMI,f11 7

Figure 5-2 shows the cppropriate source regions for the BMI facility. The finc! determinction of the most copropricte set of regions evolved from sensitivity studies. The regions cre generally those contained in c 400-kilometer radius around the site. Our modificctices to the Algermissen end Perkins source regions cre cs follows. First, source zone 64 bcs been truncated cnd the seismicity c!!occted in proportion to the crea remcining. Second, source zone 61 (New Modrid) hcs beer extended up the Wcbcsh Valley to the north in accordance with current NRC tectonic interpretction. Finc!!y, we hcve divided source zone 62 into two pcrts in on ettempt to better segregate the Anne, Ohio, seismicity from the rest of the zone. The three citernctive segregations used in our anclyses cre presentec in Figure 5-3. The first model is bcsed on the work of Nuttli cnd Herrmer (personal communicction, August 1978) in their seismic bczerd anc!ysis of the centrcl United States. The other two segregations are, in our judgment, recsoncble c!ternatives to this. In the second ecse, the Anna seismicity is constrained to occur in the vicinity of the nistoriccl ecrthquake cctivity. The geophysical bcsis for this pcrticulcr zonction is that the Anna c ec is at a hinge of the two geologic structures, the Findley Arch cnd the Kcnkakee Arch, where there could be locc! stress concentrctions. The third ecse is built upon the historical oseismicity of the Columbus vicinity. Nuttli's catclog shows thct in the Icst 70 years no earthquckes have been reported within 60 kilometers of Columbus. Given the apparent geologic sicbility of the Columbus crec cs determined from, for example, basement contours, we judge that continued eseismicity of the crea is quite plcusible. Source Recion Seismicity Algermissen md Perkins (1976) calculated the rates at which ecrthquckes cccur in each of their source regions bcsed on the seismic dctc cvailable at that time (1974). These rates, which cre related to coefficients in the expression tog N = c - b I cre presented in Toble 5-1. r/ 1oiI F. !j Q Ot 5-2 TERA CORPCRAiiON

- - -,N/ \\ \\ t ) g \\ .N g 62 J' 67 (SEE *stTE FIG -~ ' 3) 1 66 l-64 - ~. ~ j tm [ 61, ~' ~ ~ - .3f. 3 N c' l \\ q ~ L --d ', ) t r p Ww\\ \\ , ~. ) W u > > - N..e Q P(% e / FIGURE 5-2 SEISMIC SOURCE REClONS USED IN THE ANALYSIS (Source Region Numbe-s From Aigermissen and Perkins,1976) c- /, R -) + ~.- TERA CORPORAilCN

f 'f f 63 63 ' 63 I /% / 62-EAS g _62-EAST 62-62-5 I )*S (MNA NA ^l,2.,'T \\,, %,.,., sir 1-

~.~

,j ASd!SMIC N \\ '~ 4'4 - .A e %ST REGION g-() (2) (3) FIGURE 5-3 THREE ANNA OHIO SEISMIC MODELS USED IN THE CALCULATIONS b ' 9 () R + TERA COGCRAIlCN

TABLE 5-1 ALGERMISSEN AND PERKINS (1976) EARTHQUAKE PARAMETERS Maximum No. of MM Mcximum Zone Historicci intensity V's cer No. Earthoucke 100 Yecrs b-vclue 6i X 84.5 .50 62 Vill 22.0 .50 63 Vill 22.1 .6L 64 Vill 54.4 .59 66 Vill 13.0 .59 67 Vil 7.8 .59 f J.,

4 [

5-3 ~ TERA CCRDORAilCN

These parameters define the incremental distribution of ecrthqucke magnitudes (thct is the number of ecrthquckes between I cnd I + dI) up to an hvpt esized maximum intensity. The Algermissen cnd Perkins (1976) mcximum mognituces corresponded to the lcrgest historically observed earthquckes. We updcted this dcto bcse to 1977 as described in Section 4.0 and our re-excmination of the seismicity in each region, except source region 6!, inciccted that there was no chcnge in the ecrthquake stctistics cnd therefore no bcsis for citering the Algermissen cnd Perkins statistical parameters. Source region 6 I has recently been carefully reviewed by Nuttli (1978), and his enc!ysis, which includes substantici microseismic dctc, indicctes thct the frequency of lcrge ecrthquckes is substantictly greater than cs reported by Algermissen and Perkins. We use Nuttli's values in our calculations. The mcximum mognitude ecrthqucke is a very uncertain parameter, particulcrly in the less seismic crecs that we cre considering here. Accorcingly, we judged that it wcs recsoncble and moderate to assume that each region wcs ccpcble of ecrthquakes of roughly one-hcif mognitude unit greater then the lcrgest magnitude ecrthqucke in the historicci record. The Gutenberg-Richter relctionship is here, ogcin, generally used to relate historicci intensity to mcgnitude. For calculationci simplicity, we c!se specify a lower cut-off mcgnitude for ecch region. Sensitivity studies showed that earthqvckes with magnitudes less then the lower cut-off do not affect the risk at the site. Finally, the calculational risk model requires that the dis-tribution of earthquakes be specified cs a complementary cumulative distribution (number of earthquakes with magnitudes greater than M) rather than incre-mentally. The results of this integration, up to the upper cut-off mcgnitude, are presented in Table 5-2 clong with the upper and lower cut-off magnitudes. Attenuation The ottenuation relationship was chosen for the credibility it has obtained from extensive review and evo'uction. None of the other ovcilable relationships (McGuire,19766 lists 25 published relations) hos been reviewed or scrutinized cs carefully cs the components of the one used in this analysis. h.- 5-4 iOQ e. O ' TERACDTTPORAilON

TABLE 5-2 LARCEST N=N 10-D I *,,-M ) MAXIMUM HISTORICAL o EARTHQUAKE EARTHQUAKE ZONE o '"o b MAGN!TJDE ( M t.', ) 61 10.3 4.0 0.75 8.0 X 62-Annc 0.14 4.0 0.92 6.5 Vill 62-Ecst 0.42 4.0 0.92 5.5 Vll 63 0.63 4.0 1.13 6.5 Vill 64 1.29 4.0 1.00 6.5 Vill 66 0.36 4.0 1.05 6.5 Vill 67 0.21 4.0 1.05 6.0 Vil c. u h 5-5 TERA CORDOhilON

The bcsic cpproach is to develop the functionc! forrr. of the relationship by synthesizing the results of several previous investigations. The specific relction-ship is then defined by a fit of the resulting functioncl form to the only ovcilcble strong motion dct: in Central /Ecstern United States. The functional form is developed from three sepercte regression eno!yses. The cnclyses resulted in best fits to the dcto for: Site intensity vs. distence and epicentrcl intensity e Peck accWration vs. site intensity, ecrthqucke magnitude and distance Ecrthquake mag titude vs. epicentrc! intensity e The first of these relationships is contained in Professor Bollinger's contribution to USGS Professional Paper 1028 on the Chcrieston, South Carolinc, 1886 ecrthcucke (Bollinger,1977). Bollinger's analysis of the 800 intensity obserse-tions from that earthqucke resulted in the development of a new intensity att en-uction relation that is similcr to other published relctions but hcs the odced credibility of being bcsed on the most complete set of Ecs+ Cocst dctc. Scilinger's use of the octuci intensity observctions rather then the isoseismcis permits the specification of fractile verictions to the fit. Bollinger's 50 percent fractile relationship is I = I, + 2.87 - 0.0052 a - 2.88 log a I = site MM intensity I, = epicentrol MM intensity a = epicentrol distcnce (km) Figure 5-4 compcres several of Bollinger's fractile relations with other recently published ottenuation functions. 5-6 .ih 'O TERA CCRDORATlCN

The return period cssociated with the specified occeleration is then the recipro-col cf the risk. It follows from the definition of return period that accelerations with a pcrticulcr return period have a 63 percent probability of being exceeded within the return period. Our estimate of the seismic risk represents the weighted results from 18 indivi-dual calculations. The five calculations represent six bcse cases cnd 12 pertur-bations of input parameters about inese bases. The perturbations are weighted by subjective estimates of their probcbility of occurrence to derive a weighted best estimate of the seismic hczcrd. The pcrameters that cre considered uncertcir, and which cre included in our estimate of the risk cre the intercept of the ottenuation r-lation expressed through the value of 7, cnd the value of the occeleration dispersion. The bcse ecses cre considered to consist of the following input-e The three sepcrate definitions of Annc, Ohio, sc,urce regions e Mcximum ecrthqucke = largest historicci plus one-hcif mognitude unit e Attenuation intercept, 7 = 0.90 e Acceleration dispersion, Jf.nA = 0.60 We chcrocterize the uncertainty in these dato by considering that the value 7 = 0.9 to be also 70 percent prcbable with perturbations of 7 = 1.0 and 7 = 0.80 to be respectively 15 percent probcble. We further weight the accelerotion dispersion of 0.60 at 70 percent with 15 percent weights respectively being cssigned to 0.50 and 0.70. The three Anna, Ohio, source region definitions cre weighted equally. The t.-st estimate in Figure 5-7 is the weighted summation of these 15 calcula-tions. The plus one stonderd deviation is derived from 50 percent-50 percent b h, l9l ItRA CORPCRATICN

X-N Ns N s N 's CENTR AL & E ASTERN L.5. \\ Howell & Schultz (l9751 gy N s N N N \\. N \\ Vill - s \\ \\ CENTR AL L.S. /- x C v;te & N et:1. (1975' sV \\ h Vll - NORTHEASTERN U.S. 'xs s Cornell & Me z (1974) \\ N N,N'\\ Ecliinge- (19- - K t vi O N e: \\\\ 90% ~R ACT!_E q V-g \\ 'N \\ 75% FR ACTILE s \\ \\ IV - 50 % \\ FRACTILE \\ \\ \\ i i i i i i i i i 10 20 30 40 50 100 200 300 500 1000 2000 EPICENTRAL DISTANCE (km) FIGURE 5-4 COMPARISON OF ATTENUATlON FUNCTIONS nc ,) 0) ~ TtRA CORPORADON

The acceleration relation that couples with Bo!!inger's relationship was de' rived from cnolysis on nearly 1500 world-wide accelerograms (USNRC,1977). Exten-sive statistical analysis resulted in the following correlction log A = 0. m l + 0.24 M - 0.68 log a + y H AH = peak horizontal acceleration M = earthqucke mcgnitude y = region-specific percmeter. Finally, we ogcin use the Gutenberg-P,ichter reictionship to relcte ecrthqvcke mognitude to epicentrol intensity, M = l.3 = 0.61,. Combinction of these three correlations results in log AH = 0.0 M + 0.0905 - I.08 log a - 0.0007 a + y The value of y wcs determined by fitting this relationship to the only avciloSie acceleration date from Eastern /Centrol United States (Herrmann et al.. ! ~ / and

5GS,1976). Note from Table 5-3 that three instruments did not trigger during the March 25, 1976, earthquake. The accelerations at these stations are assumed to be just under the trigger level for the instruments, I percent g. Although the recording sites for these dato are similcr to the BMI site, we allow for the possibility for some site amplification in these dato by considering three d
fferent fits to the data. We judge that the most appropriate fit to the dctc is for y = 0.9.

In consideration of the importance of this porometer, we also include the citernctive values of y = 0.8 and y = 1.0 in our analysis. These three reictionships are compared to the data in Figure 5-5. As Figure 5-5 suggests, it is very important to consider the mognitude of the dato dispersion about our meon attenuction relationship. Eoch of the component relationships that were synthesized into our ottenuation relationship were them-selves best fits to data with cssociated dispersion. ' Because the data set usec in / 5-7 TERA CORDORAIiON

TAB'.E S-3 LOCATlONS OF ONLY APPRODRIATE STRONG MOTION RECORDINGS I 6 13 75 N MADRID, MISSOURI 2 3 25 76 ARKABUTLA DAM, vilS$1SSIPPI L TOE 3 L CREST 4 R ABUT 5 TIPTONvlLLE, TENNESSEE 6 N MADRID, MISSOURI 7 WAPPAPELLO DAM CREST 8 TOE 9+ MEMPHIS, TENNE'SEE 10+ SARDIS DAM, MISSISSIPPI ll+ POPLAR BLUFF, MISSOURI

  • Instrt. ment, set at I percent g, did not trigger bh 7tRA CORPORADON

t DCC. B-5~ 4- \\ C#I# #C'"T "S U 2- '7 MACNn 2 4.ea 6/13/75 E rts:

  • 5 I OC -

E d / 5

  • 3 Ie L-,

2 6d 3; - ::,. va.*_i_ .2 7 "~ 3.22 .e-: C / 2 q e = a T

3
  • 10 -

o I e-

~

2 tv i e-16 x 2 4- =..e-0.1. = 2-I 6 4 6 6 100'. I 2 4 6 B 10 2 4 6 8 100 2 DISTANCE (um) Attenuction Function Usec in This Stuey

  • Deto (Tooie 5.3)

FIGURE 5-5 COMPARISON BETWEEN ATTENUATION FUNCTION AND DATA

i a

. _ + TER A CORCRATON

these individucl enclyses is diverse and not readily available, we choose to essess the dispersion for our attenuation relationship through consideration of other data sets and other attenuation analyses. The stctistical properties of peck occeleration cre usually chcracterized in terms of the notural logcrithm of acceleration and thus the dispersions are dispersions of in (AH). Typical stenacM devictions of tnis parameter rcnge from.51 (McGuire,1974) tc l.2 (Esteve,1970) with a median value close to the value of 0.707 determined by Donoven (1974). Since these essessments of the data dispersion are statistical overages over cll passible site conditions, trcvel paths, and tectonic settings, we judge thct the value of 0.60 is a reasoncole best estimate of the one standctd deviction dispersion m accelerction for our specific site. Beccuse the data bcse from which our attenuation relation was derived consists of predominently for-field accelerations, the relction is less vclid in the necr-field. We account for this by limiting the peck occelerations in the necr field. The de!cils of this nect field response are not important beccuse of the aseismicity of the Columbus crea cnd the distance from the site to significent sources. This complete attenuction relctionship is presented in Figure 5-6 for several mognitudes. RESULTS The results were obtained by computer calculations with a risk analysis code (McGuire,1976b) that is bcsed on the work of Cornell (1968). The basis for this opproach wcs summcrized in St.ction 2.0. As described in Section 2.0, the computer code calculates, for circular sectors within each source region at the site, the expected number of ecrthqvckes caus-ing (rcelerotions greater then a specified acceleration and this is done for ecch source region and the host region. The expected numbers are summed for each region, and the resulting risk cciculated from risk = 1.0 - exp(- totcl expected number). gh,/ "Q 5-9 TEiA CORDCRAilCN

I OX.< 8-o- 4- . F r-

OC -

6-e- Y 4 IOa G-n 6" s. 2-1 = 6 8 100 2 a 6 e 100: I 2 4 6 6 10 2 OtST ANO (mm) FIGURE 5-6 ATTENUATION RELATIONSHIP USED IN THE ANALYSIS ( 4 =1.0) / 5 e TERA CCRDORATION

10,000 i i i s i I I / i I I 6 p / i c j/ i 2 hc'Y r p l /- 6 i l ch l l /' C I 1,000 f / / / / / / / / / / / C / / ,s' c i .l ,s'- 5 / /. n / / / 5 / / -= 100 i i 1 5 / / / / s ~ O / / / 5 / / / / / / ~ i 2 // / 1 I / \\ / / i 1 l/, / \\ l i / 10 / f i i i i i i i i i l 2 l l i I l t I I i I 2 4 6 8 10 12 i4 PEAK HORIZONTAL ACCELERATION (% g) FIGURE 5-7 / 9 ',J r. ,/O RETURN PERIODS FOR SEISMIC ACCELERATION -v AT THE BMI WEST JEFFERSON FACILITY e ERA CORPORADON

weighting of the six more conservative runs and similcrly for the minus one standard deviction. RESPONSE SPECTRUM These results define the peck horizontal cecelerction et the fccility for various retum periods. We have also determined on cppropricte response spectrum for the site since some structures and equipment et the BMI facility have sufficiently low fundamentcl frequencies to experience spectrci emplification of the ground motion. The response spectrum for the site clectly ccanot be developed in associction with a specific ecrthqucke; our return period accelera-tions represent on integrated effect et the site from cn extraordincry vcriety c' earthquakes and the response spectrum must reflect this. Accordingly, we judge that the shcoe of the spectrum should be similar to the Newmcrk-Blume statisticcity-bcsed spectra from which Regulatory Guide 1.60 evolved. Because of cn cimest total lock of good dctc, the absolute level of spectral accelerations appropriate for design is very difficult to determine. For exemple, it is well known that attenuation in the Ecstern United States is much less rcpid than in the West (Alsup,1972). Since the bcsis for Regulatory Guide 1.60 is exclusively western dctc, one might crgue that the appropricts response spectral amplitudes should be in excess of the mean, perhaps the one standard deviction level, to occount for the lesser attenuction. Alternatively, given the objective of best estimate results with minimum conservatism, it could be crgued that the mecn response spectrum for alluvium in WASH 1255 is most oppropriate. There is, unfortunately, very little qucntitative basis for choosing between these citernatives. In cur finct consideration, we emphasize two points. First, the controversy surrounding the nature of attenvotion (Q ond its possible frequency cependency) cnd second, recent calculctions at Lowrence Livermore Lcboratory which show that the effects of strcight line cpproximations of statistical spectra make Regulatory Guide 1.60 slightly more conservative than a one standard deviction spectrum. Accordingly, it is our judgment that the mean response spectrum for alluvium presented in WASH 1255 is the most appropriate for.cn ysis of the BMI facility. U 7 l' 5-1! TERA CORPCRATCN

in summcry, we have combined the best cvoilable input data with the most credible tools of seismic risk cnolysis to determine the return period of occe!- erotion at the BMI fccility. The results, shown in Figure 5-7, occount for the dispersion of the data about the functional relationships used in the mocel. Further, the results are insensitive to verictions in the source zone geometries or seismic histories. Response spectrcl accelerctions can be determined by sec!ing the mecn response spectrum in WASH 1255 to the desired peck occe!erction. U 5-l2 TERA CORPCRATION

6.0 BIBLIOGRAPHY Algermissen, S.T.,1969. Seismic Risk Studies in the United States. 4th World Conference on Ecrthqucke Engineering, V. I, p.14. Algermissen, S.T.,1974. Seismic Risk Studies in the United Stctes. 5th World Conference on Ecrthquake Engineering. Algermissen, S.T. cnd Perkins, D.M.,1976. A Probcbilistic Estimate cf Mcximum Acceleration in Rock in the Contiguous United States: USGS Open File Report 76-416. Alstp, S.A., 1972. "Estimction of Upper Mcntle O Benecth tSe United States from P Amplitudes," PhD thesis, George Wcsnington University. n Bcttelle Memorict institute Revised FSAR for the Bcttelle Resecrch Reccier Docket No. 50-6 Merch 26,1974. Bollinger, G.A., 1977. "Reinterpretction of the Intensity Dctc for the 1886 Chcrieston, South Carolinc Earthquake." USGS Professional Pepe.- 1028, in Press. Bollinger, C. A., 1973. " Seismicity of the Southeastern United States." B5SA 63(5), p.1785. Bradley, E.A., and Bennett, T.J.,1965, "Ecrthqucke History of Ohio": Seismol. Soc. Americe Bull., v. 55, p. 745-752. Brczee, R.J., 1976. An Anclysis of Ecrthqucke Intensities with Respect to Attenuation, Magnitude end Rate of Recurrence. NOAA Tech. Mem. EDS N6SDC-2. Chimery, M.A. and Rodgers, D.A. 1973. "Ecrthqucke Statistics in Southern New Englcnd," Earthquake Notes, Vol. 44, pp 89-103. Cornell, C.A.,1968. " Engineering Seismic Risk Analysis," BSSA 58(5) p.1583. Cornell, C A., and Merz, H.A.,1974, A Seismic Risk Analysis of Boston: Jour. Structurcl Div., A.S.C.E.,110, no. ST 10, Proc. Paper I 1617, p. 2027-2043. Cornell, C.A. and Mertz, M.A.,1974. " Seismic Risk Analysis of Boston." ASCE Annual Mgt, Preprint #2260. Cincinnati, Ohio. Corne!!, C.A. and Vcnmcrke,1976. The Major Influences on Seismic Risk. 6th World Confere e on Earthqucke Engineering. Decekcl, J.,1970, Ecrthquakes of the sicble interior, with emphasis on the mideontinent (Ph.D. dissert.): Lincoln, Univ. Nebroskc, v. I,169 p., v. 2, 332 p. n, 1 ,s

g o i u

h 6-1 e TEiiA CORPORATION

Donovan, N.C.,1974 "A Statistical Evoluction of Strong Motion Dctc including the February 9,1971, San Ferncndo Earthquake," World Conf. Ecrthqucke Eng., 5th, Rome 1973, Proc., v. I, p. 1252-1261. Esteva, L.,1970. " Seismic Risk and Seismic Design Decisions," M Hansen, R.J., ed., Seismic Desien for Nuclear Power Plants; Cambridge, Mcssochusetts Inst. Tecnnology Press, p. 142-182. Evernden, J.F.,1970. " Study of Regional Seismicity and Associated Problems," BSS A g(2) p. 393. Fex, F.L.,1970. '$eismic Geology of w Ecstern Unitec States, Bulletin of tne Assoc. of Engin. Geologists vil (1) p. 21. Gupta, incra N., cnd Nuttli, Otto W.,1975, Spctic! Attenuation of intensities for Centrol United States Earthquake: submitted to Seisme!. Soc. Americ: Bull. Gutenberg, B. end Richter, C.F.,1956. Ecrthquake Magnitude, Intensity, Energy, end Acceleration, BSSA V.32, p.163. Herrmenn, R.S.,1974, Surface Wave Generation by Centro! United States Eartn-quakes (Ph.D. dissert.): St. Louis, Saint Louis University, 267 p.

Herrmcnn, R.B.,
Fischer, G.W.,

Zollweg, J.E., 1977. The June 13, 1975 Earthquake and its Relationship to the New Madrid Seismic Zone, BSSA Vol. 67, No. I,p.209.

Howell, B.F.,

1974. " Seismic Regionclization in North Americe Bcsed on Average Regicnci Seismic Hczerd index," BSSA g(5) p.1509. Howell, B.F., and Schultz, T.R.,1975. Attenuation of Modified Merecili intensity with Distance from the Epicenter: Seismo!. Soc. America Bull., V. 65, p. 651-665. Liu, S.C. and Fogel, L.W.,1972. Ecrthquake Environment for Physical Design: A Statistical Anclysis. The Bell System Technical Journal, Vol. 51, No. 9, p. 1957. McGuire, R.K.,1974. " Seismic Structural Response Risk Analysis, incorporating Peck Resoonse Regressions on Earthquake Magnitude and Distance," Mcsso-chusetts inst. Technology, Dept. Civil Eng., Research Rept. R74-51, 371 p. McGuire, R.K.1976c. "The Use of Intensity Data in Seismic Hczard Anclysis." 6th World Conference on Eorthquake Engineering. McGuire, R.K.,1976b. " FORTRAN Computer Program for Seismic Risk Anal-ysis:" U.S.G.S. Open File Report 76 - 67. McGuire, R.K.,1977c. "A Simple Model for Estimating Fourier Amplitude Spec-tro of Horizontal Ground Accelerotion." Submitted to BSSA. / oT U2 6-2 TERA CCRDORATiCN

McGuire, R.K., 19776. Effects of Uncertcinty in Seismicity on Estimates of Seismic Hczord 'or the East Coast of the Unitec States. BSSA Vol. 67 No. 3,p.627. Milne, W.G. and Davenport, A.C., 1969. " Distribution of Earthqucke Risk in Canada." BSS A p(2) p. 729. Necioglu, A., cnd Nuttli, O.W.,1974. "Some Ground Motien and Intensity Reic-tions for the Centrcl United Stctes:" Earthquake Engineering cnc Struc-tural Dyncmics, v. 3, p.I I l-l I 9. Nuttli, O.W., 1973. State of the Art for Assessing Ecrthqucke Mczercs in ine United States, U.S. Army WES Misc. Pcper S-73-1. Nuttli, O.W., 1973c. " Seismic Wave Attenuation and Mcgnitude Relctions for Eastern North Americe:" Jour. Geophys. Resecrch, v. 78, p. 876-885. Nuttli, O.W.,1974. " Magnitude-recurrence relation for centro! Mississippi Vcliey ecrthquakes": Seismol. Soc. Americo Bull., v. 44, p. I 189-1207. Nuttli, O.W.,1978. "Tne seismicity of the central United States": to capect in Nuclecr Geolocv published by the C.S.A. Nuttli, O.W., and Zollweg, J.E., 1974. "The fielction between Felt Are: cnc Mognitude for Centrcl United Stctes Ecrthquckes:" Seismol. Soc. America Bull., v. 64, p. 73-85. Oliveira, C.,1974. " Seismic Risk Anclysis," Ecrthqucke Engineering Resecrch Center Report EERC 74-l. Schnabel, P.B. and Seed, H.B.,1973. Accelerctions in Rock for Ecrthquckes in the Western United States, BSSA V. 63, p. 501. 6 Seismological Society of Americe, Seismologicc! notes: published in the Seismol. Soc. Americo Buil. since 1911. Stauder, W., and Nuttli, O.W.,1970. " Seismic Studies: South Centrc! lilinois Earthquake of November 9,1968:" Seismol. Soc. Americo Bull.

v. 60, p.

973-981. S touder, W., Best, J., Cheng, S.H., Fisher, G., Kramer, M., Morrissey, S.T., Schaefer, S., and Zollweg, J., 1974-1976. Southeast Missouri regional seismic network qucrterly bulletins Nos. I-7: St. Louis, Scint Louis University. Stepp, J.C., 1971. An Investigation of the Ecrthqucke Risk in the Puget Sound Area by Use of the Type l Distribution of Longest Extreme. Ph.D. Dis-sertation Penn. State Univ. - Geophysics. Stepp, J.C., 1974. " Analysis of Completeness of the Earthquake Scmple in the Puget Sound Area cnd its Effect on Stctistical Estimates of Earthqvcke Hazcrd." Proceeding of Conf. on Microzonoticm, Seattle. Oi 6-3 TERA CORPCRA7iCN

Str-et, R.L., Herrmann, R.S., and Nuttli, O.W.,1974. "Ecrthquake Mechanics in the Central United States": Science v.184, p.1285-1287. Trifunce, M.D. cnd Brody, A.C.,1975. "On the Corre!ction of Seismic Intensity Secles with the Pecks of Recorded Strong Ground Motion." BSS A g(l) p. 139. U.S. Depcrtment of Commerce, NOAA,1973. Ecrthqucke History of the United States. Edited by Cof' mon and Hake. U.S. Geological Survey, Prelimincry Determination of Epicenters, publishad annucily by the Nationcl Ecrtnqucke informction Service, Denver, Coloredo. U.S. Geological Survey,1976. Seismic Engineering Program Report, April-June 1976, Menlo Park, Calif. U.S. Nuclecr Reguictory Commission, 1977. The Correlation of Peck Ground Accelerction Amplitude with Seismic Intensity and Other Physical Pero-meters. NUREG-0143. Vogliente, V., 1973. Foreccsting the Risk inherent in Ecrthqucke Resistent Design. Ph.D. Dissertation, Civil Eng. Dept. Stanford University. s g$ _Uh g 6-4 TER A CORPOR ADON}}

Retrieved from "https://kanterella.com/w/index.php?title=ML19242D540&oldid=4893829"