ML20009C006

Advises That Util 810623 Info Re Buried Pipe Analysis Apparant Shear Wave Velocity of 2,500 Ft Is Compatible w/state-of-the-art.Newmark & Hall Article Seismic Design Criteria for Pipelines & Facilities Encl

ML20009C006
Person / Time
Site:	Fermi
Issue date:	07/17/1981
From:	Colbert W DETROIT EDISON CO.
To:	Kintner L Office of Nuclear Reactor Regulation
References
EF2-54096, NUDOCS 8107200155
Download: ML20009C006 (10)

Text

..

! Detroit 2000 Second Avenue b fi$$$3T T 2/ u s22e July 17, 1981 EF2-54096 i

f Mr.

L.

L. Kintner Division of Project Management Office of Nuclear Reactor Regulation U.

S. Nuclear Regulatory Commission Washington, D. C.

20553

Reference:

Enrico Fermi Atomic Power Plant Unit 2 NRC Docket No. 50-341

Subject:

Buried Pipe Analysis Shear Wave Velocity In our letter of June 23, 1981 EF2-53866, we stated that at Fermi 2, an apparent shear wave velocity of 2500 ft/sec. has been used.

The ' state of the art' design methodology in this area is based on the work of Drs. Newmark and Hall (1) carried over some period of years in connection with design studies for the Trans-Alaska pipe-line, the Canadian Arctic gasline, the Schio pipeline from Long Beach, California to Midland, Texas, as well as other special facilities.

The shear wave velocities to be used in design as suggested in ref. (1) are:

4000 ft/sec. for rock or permafrost, 3500 ft/sec. for massive gravel deposits, 3000 ft/sec. for sand and competent soils, and slightly lesser values for silt and clay deposits.

At Fermi 2, the buried pipes were installed in compacted rockfill.

As per ref. (1) a shear wave velocity of 3000+ ft/sec. could have been applied.

For the backfill compaction densities as placed in the field, the use of design shear wave velocity of 2500 ft/sec.

at Fermi 2 is compatible with the state of the art.

Sincerely,

/

M99 l William F.

Colbert Technical Director Enrico Fermi 2 WFC:dah g

9 I

Attachment ~

s l

8107200155 810717

~~

PDR ADOCK 05000341 A

PDR

g HOVEMEiEfl 1978

,TC1' f

Hist y y clo,

JOURNAL OF THE TECHNICAL COUNCILS I

OF ASCE I

SEISMIC DESIGN CRITERIA FOlt PIPEUNES AND FACILITIES *

.a By William J.11:11,8 F. ASCE and Nathan M. Newmark,' lion. M. ASCE (Rcylmd by the Technical Council on I.lfeline Earthquake Enrir:cering) i I

! Senenc DislGM Pmtosoeur The design criteria and reconunendations described herein late into account l

the seismic motions and seismic generated forces that have a rearcuable dqree of probability of occurrence along the route of a pipelinc. The ba';is for the l

selection of these criteria and recommendations involves consicler:stion of the acceptahl risk of caccedng the desi.n levels for the pipeline system and various f

classes of asmciated stiuctures, equipment and facilitics. For the mast critical N

classes, where failure, defined as exceeding the allowable recommntled levels, b

wr ul.1 have a bearing on life and safety of the population on ni ht adversely

/

alfcct the environment, or where for economic reasons interruption cf the service 3h l[

provitteil by the pipeline is not tolerabic, the margins of safety imili:it in these wT jl criteria are often greater than those riow used in the scismic deWr.n of major i}

R buildings in highly scismic regmns of the United States. For the least crincal t

classes, the martms of safety are at least as great ns those provid-d by current 'b S j

building codes such as the Uniform fluild ng Code or the Structural lingineers A,sociation of California (SF.AOC) Code (15). The procedures outlined will h

result in a design having appropriate factors of safety against seismic disturbances l

when combined with tl.e othe applicable operating and environme'u.tl conditioris, ll in accordance with principles developed for use in the de'.i n of nuclear 7

1 3

reactor power plants, the design criteria genefally encompass two levels of l

carthquake.harant. The lower level is that associated with a aturn l criod for i

'l the design carthquake of approx 50 yr-Ibo yr and is designated herein as the

{

7 j

Note. -Discussion open untit Apsil I,1979. To extend the closing date one mc.hih, D a written request must he fihd with the tiditor of technical Publications. ASCE. This V.

CN raper is ratt of the copyrir.l ted Journal of ibe Techmcal Councils of ASCli, Proceedings ol the American Sotiety of Civill:nr.inecis, Vol. lGt Ho.'ICI, November. lu18. Manuscript c ',

was submitted for scview for pessible public.ition on March 9,1978.

j

' Prof. of Civ. linstg., Univ. of Illinois. Usbane, Ill.

i

' Prof. of Civ.1 ngrg., asi.l in the Centet for Advanced Study, limeritus, Univ, of

's Illinois, Usbar.a. Ill.

91 P

~

93 SEISMIC DESIGN CRITERIA M

NOVEMnt.rt 1970 TCt In assessing the importance of the accelerations and velocities for which n.

" Design Probable Tarthquake." The higher Icvel is that associated with a I..n the design is to be made, the maximum ground accclerations, in th:mselves, return period, of the order of about 20') yr-Mi yr or more, and is deurn et are of Icss significance than the accumulated effects of the larger number of as the "D-sign.'.tasinmm Earthipid e." IJnder some situationsii may hc e spedient somewhat smaller accelerations that contribute to the prmcipal structural or g

to use only one such Icycl, g.:ncrally the luter.

element response. In general the significant effects of an earthquake are measu t

Conceptually one might consider the first carthquake as one through which more directly by the ntaximum groimd velocity than by the maximtym grotmd the pipeline should be able to operate and i ontinue operation after its occurrence acceleration. A single spike of hir.h accelciation may have much less significance on tesponse than would he computed by straightforward applicatmns of imc(

whiIc the larger carthquake should not produce damage that has not hcci l

anticipated m the design of the pipeline, structures or facilitics. Ilowever, to clastic analvds for dynamic systems.

j.

do this in a systematic way usually would involve an unreasonable degree of in the design of any system to resist seismic excitation, there are a number deugn effort and morcover often may be based on inaccurate or insufficient or parametern and design considerations that must be taken into acenunt. Am scisnac data for the region. When this is the case the relationship between these are the nngnitude of the carthquate for which the design is to be made, y

l the intensitics of the two design carth<pi.4 es normany has been taken arbitrarily

's i:overning the distance of the facihty from the fm;u, os f auh,,'

• p w"1 em1, rock

,}

as a factor of two, f nttenuation of motions with distance from the focus or epicenter.

The carthquake intensity by itself has limited significance in terms of design or permafrost conditions ns wellas the generni ccologic conditions m the vicim to resnt seismic motions. Of equal importai.cc are the structural parameters and the parameters governing the response of the facility or structure itself.

govermng response, such as stress or strain and deflection, that the designer hfost, if not all, of these parameters are subject to considerable uncertamty

l inten h to u,e for the particular carthquake harard selected. Normally these in their value, Uccause so many of the parameters involved have probabilistic cnteria are selected to make the Design Maximum liarthquake govern the design, (rather than deterministic) distribut;ons, it is not proper to take cach of th Furthermore the criteria are such that, in the event of the smaller earthquake, !

with a high degree of conservatism because the resulting comb, ed degree o m

I the pipeline and facilitics, if properly designed in accordance with the recommen.

conservatism would then be untcasonabic. At the same time it dations, will generally he abic to continue operation, to have an assured margin of safety in the combined design corditions. Thus, Structures and aboveground piping, which acts like a structure, respond to a choice must be made as to the parameters that will be taken with large margins an carthquake in a way associared with their dynamic parameters. Iluried piping of safety and those that will be taken with more reasonable values closer to responds to carthquake motions by moving with the ground in such a way the mean or expected values.

,1 as to have neady the same curvature a.id nearly the same longitudinal strain The relation between magnitude of energy release in an carthquake and the I

as the ground. 'I hus, the design entcaia for these Iwo conditions must be dif ferent.

maximum ground mo' tion is complex. 'ihere are some reasons fer infernng that I

in general, for aboveground clements or structures, the design is carried out the maximum accelerations are, for exampic, nearly the same for all magnitudes by use of the concept of the design or response spectrum. The response spectrum of relatively shallow cartb nake for points near the focus or epicer:cr. Ilowever, conrept, however, does not apply to buried pipe. In the latter case the carthquake 3

for larger magnitudes, the valuet do not drop off so rapidly with distance from motions produce strains m the pipe that can be computed reasonably well from the epicenter, and the duration of v'..aking is longer. Consequently, the sta the canhquake ground motions and the estimated wave propagation velocity h

arismg from the carthquake. 'these dificiences are explained in more detail l

mean or expected values of grotmd rr..ons show a relationslup moreasmg wit q

magnitude, ahhough noi in a linear manner.

later m tlus paper.

i Mgn Sdunic MohnMn selecnng ec cadquuc MM M m b W.

the Cencral concept has been used that the carthquake magnitude nicc Susme Om.s Monous u.n nesponst be at least as large as those that have occurred in the past, and these carthquake Attual Versus Effectisc Entthquake Motions.- Although peak values of ground are generally considered to have equal probabilitics of occuning at any pomt I

motion may be assigned to the various magnitudes of earthquakes, especially within regions of similar or closely related I,cologic character. In particular, m the vicmity of the si.. face expression of a f ault or at the cricenter, these the estimites of motion considered are appropriate for competent. materials motions are in general considerably greater than smaller motions that occur at or near the ground surface, including rock and permafrost, or competent many more times in an earthquake. Design carthquake response spectra are consolidated sediments at or near the surface. The values sclected nic ricarly independent of the' properties of competent surface mate,ials. It is consid

}

hased on "cfrectise" values of the acceleration, velocity and displacement, that the predominaht part of strong carthquake f round motion, fencrated by which occur sever il times during the canhquake, rather than isolated p ak values of instrmnental reading. T he effective harards scletted for determining d-sign a near shallaw carthquake cncrgy release, is representcJ by surface waves.

spectra may be as little as one. half the expcoted isolated peak instrument readints In general, these are propagated in a manner consistent with the properties for near carthquakes, ranging up to the latter valucs for distant carthquakes.

of the material at a depth considerably beneath the surface and are not'affected ficsirn response spectra descrmined from these parametres can take int.'

to a large extent by the surface properties themselves. The design values of j

account the varmus energy absorption met hanisms, both in ihr Fround and motion are based on the assumption that the same values are applicable m m the cicment, including radialmn of energy into the ground f rom the ec p.n f. 's a particular zone for all competent soils.

,(

W " "'

a 94 NOVEMBER 1978 iCl SEtSMIC DESIGN CRITERIA 95 l

An example of design seismic motions are those given in Table I f the characteristics of an initial clastic followed by a perfectly phstic relation iip, scismic zones for the Trans-Alasta piieline (11). These zones are charauer ic.Iihen the ductility factor must be defined in the fashion given in Re s.

or l

by the magnitude of earthquake considered as the Design Maximum Eanhqu 4 14 by use of an equivalent clastoplastic relation drav'n to make the energy For each r.f the zones evo sets of effective ground motion vnfues are hsted~

The first set, entitled " Ground Motion," ir. chides those values that ma dastoplask cum aff the stability of slopes or the liquefactien of cohesionless materiah, and ne anmunt d inelas dehne, n h a gm m *M Wl?

also the values which should be used to infer the. strains in unde ground suffcring undue danwge abo aUccts b repnse, m m M & a m

'i The second set of values, entitled, " Structural Design," lists those valu it and the corresponding defo mations and dcHections. The allowat Ic va oes are to be used for the design of structures or other facilities. These values of dustility depcnd on the material of which the structure is ina e an on take into account impl;citly the size of the structure, soil-structure interaction IIS '""""cr or e nstruct on, principally the way in which jomts are ma.

and structural response and are Ecnerally less than those used for definin

""*PC**I^*

"IC *"""'#"

I soil instabilities. Obvioiuly the actual values are transient values at variab!

techniqucs and attention to details possess hirh ductility. Under ccr u e rcum-i

}

times, and only the peak cife.: live design values are listed. The design motions 5'""CCS Y ##"

"E I

gisen are for the horizontal direction and may occur with equal probability to undergo kcal Mng. N h remns k h% W-1 f

. in either of two orthogonal horizontal directimns more or less simultaneously' design must be verified to detesmine that. the snaterials themsel es an t se f fabrication procc>scs, and especially for scinforced concrete tye t etai s o TAllLfi 1.-Design Seismic Motione construction, are conisolled in such a way that the value of ducidity used can actually he achieved while maintaininE the requisite margin of sie :ngtt;; it is

_ _. _ _. _ = = _. -

Ground Motion recommended that the structuse as a whole, including details, bc inade capable Structural Desig" of developing a ductility factor of at least 1.5 times that used m the design Accesoro.

Voloc-Accelora-Veloc-spectrum, possi'.de ductility levels under ordinary conditions are covered m tion of ity, in tion of ity, in Ref.14. Where the permissilde level of stiuctural response does not mvolve gr avity, inches Displace-

gravity, inches Displace.

yicMing at all, then the ductility factor used W Iimited to a value of unity.

ar. a per-por mont, in as a per-por m ont,in Magnitude ce De second mchos centa00 second inches Hnponse and Dolgn Spects2.-The response spectrum (9,10,13,1 l) is a plot the masimum transient response to dynamic motion of a simple dynamic (I

N. --

N system having viscous damping. An clastic response spectrum has peaks and 3.5 and 8 60 29 22 33 16 12 Wim but in general has a roughly trapeioidal shape, similar to the upper 45 22 16 22 11 s

l t of Fig. I. Spectral amplification factors for horizontal motion,in the clastic

]

~-

5 5

4 tange, for damping values of 2%, 3%, 5%, and 7% critical, taken from Ref.

5.5 6 (75 percentile values) are given in 'Iable 2.

To draw de clastic response spectrum for any Design Maximum liarthquake It is reconunended that the design motions to be used in the vertical direction a

f be taken as two-thirds of the value in the horiznntal direction.

motion for a structure, one takes the,alues of ground motion for any one of the zones from Tablo I, using the "structusal desirn" values, and apphes

{

The maxim :m ground motion values.givcn in Tahic I, as coscred earlier the appropriate amplification factors f rom Talde 2 for the particular percentage hesem, may be considerably less than the isolated peak values of motion (as of dampmg to the asceleration, velocity, and displacement, respectively. 'Ihc measured by inst:uments) that concspond to the magnitudes of earthquakes tn:d to these various zones.

d.unping values in Tables 2 and 3 nie intended to represent primarily material l

that mi,;ht be assi and structural damping. One obtains in this way a roughly trapezoidal form I:faele Spotral Amplificalmn.-The clastic response of a simple dynamic of response spcetrum similar to the curves in Fig.1. The intenections of the system subjected to motion of its suppo[rt is affected to a large extent by the upper two knees of the clastie responsc apcctrum are determined by the amplified g

damping in the structure. This damping is usually capressed in terms of the motion lines. The two Inwer knees, at the higher frequencies, are taken as percentage of it c " critical value" of damping. Values of damping for particular 1

8 hz and 33 hz, respectively The value of the spectral acceleration at 33 L strue tures or sisuctural types are covered in Refs.10 and 14 The importance and beyond is taken as the maximum ground acceleration for the clastic sesponse of dimping is indicated by the large ef fect of damping on the clastic spcoral ampliin asion.

spect ra.

Spectra also may be drawn for the operating earthquake for any ione, w hcie The ductihty factor of a structtire or element is defined as the value of the ground motion values are taken as half of those that cmrespond to the deformation or strain x., which the strutture or element can s assain before larger carthquake. In general, the amplification values, because of the different fatture relative to that value x, for which it departs appreciably from clastic comhuons it is dehned precisely on!y for an "cIntopintic" relatmn. Ilow cs er-values of damping that might be used for the lower intensity caithquake, will whese the Imd deform.nion or sorceirain cune is one w hich does not hne not be the same as for the lasger carthquake.

}

To determine the design specira for acceleration (or scismic coefficient) for s

$7 NOVEMBER 1p78 IC' ICI SEISMIC DESIGN CRITERIA clastic the inciastic case one takes the appropriate value of dnctility r higher than 33 hz, the design acceleration level is the same as the i

3 for the scismic design class (defined latcr herein) anci divid acceleration (10,13). Typical design spectra for the three scismic desirn classes elastic displacement and velocity bounds by the value orductilit are shown in Fig. l.

j a

The valocs of the controlling cla. tic acceleration bound I v r, are divided From the procedure described, it is clear that the intensity of casthquake rnotion as defined by the applicable responsc spectrum, must be considered by the quantity (2 - 1)"', in which is We deih o

quencies in the light of the way in which that carthquake motion is used in design.

T ABLE 2 -.$pectsel Ampilfication Factors, Horizontal. Elastic Range in other words, one would prescribe a lower value of acceleration ho used

~

with a procedure that involves the use of working stresses than a procedure 8

ll.at involves yield point (or limit) strengths. One cannot compare the carthquake accclerations prescribed by vatious codes without takinginto account the design Damping. per Amplifica' ion factor

~

critesia used in the codes. T he Uniform iluilding Code of the United States, cent critical Acceleration

veloci, 4

which generally is based on the SilAOC Code, has up to the present time UI (2)

(3) j used wo: Ling stress desif,n criteria, and the seismic coefficients described i 2

3.,

y7 2.2 the SliAOC Code are consistent with those values. One would have to increas 3

2.9 N

I the scismic cocificients in the codc to arrive at values comparable with those 5

2.5 2[9 1

I j

2.2 develor * ' * "in, which tre to be used at yictd levels.

g u-CAssgricAyloN roR Sosuc UtstGN l

TABLE 3.-Emampfe of De nping and Ducti'ity Levels for Various Design Classes I

5 cnd Earthquakes llecause of the importance of the amount of deformation or stress that can be permitted in buildings of various types subjected to carthquakes, guidance is necessary in arriving aI an appropriate means of selecting the dc>irn require-

" Pe Ductility Earthquake Class cent critical factor

'. ments. For this purpose, a scismic classification system, encompawing three l

p; classea, is recommended f r use.

~

Class I includes those items of equipment (including instruments) perf orming '

Design pmbable g

2 1.5 g

3 2

l vital functions that must remain nearly clastic, or any items for which the l

allowable probabihty of cxceedmg design levels must be extremely low. Obvious-l gg 3

ly, items that are essential for the safe operation of the pipeline or any facihty Design maximum I

3 f, where damage tv the particular unit would cause extensise loss of n

5 life or majos ensiionmental damage, would be in Class I. Other items might ill 7

5 be included in Class I if failure of such items would entad large cost., in repair i

or replacement, or lengthy sh;.tdown of the pipeline.

\\ O Class 11 includes aboveground piping or buildings and equipment that can n

x s

deform inciastically to a moderate extent without loss of funuion. lhis class h

~'

],,

4[!

e also includes any items for which the allowable psobability of excee ling design

• /M limits can be somewhat larger than in Class I. Ilowever, piping which might

,I 4-pj g

Ng ir A

9 failin a brittle mode, or whose falium might tend to propagate over considerable s

i d ['*""'/ y A

' n)

- f

's, fy%

distances, causing extensive damage or powihty, danger to life in populated I

}

regions, or both, perhaps should be put in Class I or in a classification intermediate between Classes I and II.

j

[

[g

' \\(

Class Ili includes, in general, buildings or equip, ment that can be permitted

_."..n g

to deform significantly, or any items that are not essential for safety; it includes p\\

/ g, K "

x I those iteins for which the allowable probability of exceeding design hmits can k

I/ % 'g^ j. N s

be moderately higli. Ilowever, buildings that contain Class I or Class !! items fN

~ - '

' /

% f ed {

/ \\

N,

\\/

cf, anJ which might damage or put out of action those items if the Claw 111 buildings Sk'ould dc[orm excessively, should be moved to a higher class, peihaps interme-x

~

t r,..

,,w, diate between Classes Il and 111.

" """b "U "" "

II *

"I

'I "E

IIG.1.-Elastic Responso Sportrum and Design Spectre the design spectra for the vatious seismic design classes is given i:1 'fabic 3.

~ ^

.-u..._.

99 SEISMIC DESIGN CntT ERIA N

NOVEMDER 1978 TCl 11 is. 2(f)),

fail re by a properly designed and supported aboveground pipelina These give results that are consi., tent with the class definitions above and even though one or two supports may lose contact with the pipe. For undergrou the criterion that the Design Maximum Earthquake, with its higher int nsi s oukt m general give snore stringent require'ments than the Design probab pipe such motions might cause severe local distortion and wriniling but not necessardy collapse or rupture.

One of the most important response parameters is the level of permissible gesponse,1:or some structures the response must remain in the clastic range.

Of sroNsE OF Pirrunts eso Sinuciunes Thisis normally not the case with any clement of a pipeline, for large i efouaaticas t

e generally can be permitted to occur provided rupture does not tale place w The response of a structure or clernent is dependent on its strength, damping a consequent hazard to the public or to the environment. Ilowever, et is important c iaracteristics, and the stress-strain or load-deformatie.: relationship for the to recognise the fact that the level of response must be selected in a manner structure or element considered. The response also is affected to a large degree consistent with the selection of the eartfuguake itself, in order to scach a y soil-structure interaction m those instances where the structure is supported consistently reasonable margin of safety,it is surgested that the levtl of response g

in or m soil, and not on rock. In particular, the following assumptions normally permiwh!c in the pipeline under estreme conditions, t.c., for the maximum marnitude of earthquake and the maximum intensity of motum, involve a f

flelowground or buried pipe, as depicted in l'igs. 2(a) and 2(b) is considered considerahic degree of deformation but short of rupture of the pir'c.

to nyove with the ground m such a way as to have nearly the same longitudinal V'"Id3'". amt Wrinkling as 17 unction of Allunable Pipe Defur.. '*iun.-Above-stram as the ground. These requirements impose both compressive and tensile M d M8 ground piping and sunctures contain% sesMng clesnem cm"l orces in the pipe as wcll as lateral bending. Of course, this assumption is normally would fall in Clau 11 or Class ill depending on their importance g

and influence on sarciy. ihe meinoas or anaiysis and the ricuen speciia used i

lOl Tof are the same as for strustuscs having similar or resated paperoes, flowever, in piping one must take account of the stress condde.ations (inctuling long term 91 corrosion potential) at joints or connections or points of suppoit to nisure that, the ductility levels consistent with the design classification can be met. These Q

- - /M \\

g(ncrally can be met by piping in Class 11, but in Class lit local strengthenin p

ICI (d) of the pipe may be required if advantage is desired to be tal.cn of the lower

[

design acceleration values for that class.

Stresses in the' materials of either pipe or other structur I clements, for

/

4 h -Q

=

{

carthqua.e and primary stresses s.umbined, should be hnute I to minimum l

specified yield strent.th vahics in general. Ilowever, for the Design Maximum

~

""~

I' "I

til Earthquake, the values mirbt be increased to the average attual yictd point.

i I.arger values of defosmation might be perniissible in extremely duct le structures FIG. 2.-Piping Configuratione depending upon the nature of the pipeline, contents handled. environmenta concern and safety required but the limiting values applicable for Class !!!

7 valid only so long as the material surrounding the pipe remains relatively intact.

should not be execeded. Development of applicable stress and deformation criteria n other words, the assumption applies only if the material surrounding the to accommodate the scismic criteria outlined, as well as applicahic code provision pipe does not liquefy or the material surrounding the pipe is not grossly distusbed (l.2) lor exarnple, no,mally entails considerable effort by the designer and ofte Under thpefaction conditions, the pipe is no longer supported directly by the involves review by cognizant governmental re6ulatory agencies.

material and the possibility of further large deformations must be considered, 1.ocal wrinkling theoretically (17) may begin,at compressive strains g,iven by I or abo.cground pmelines [irig. 2(j)] the motion of the ground is imparted to the pipe through the piers on supports under the pipe. The deformation of the following expression these supports must be considered. as must also tipping or tilling. Of course, i

-(I) t hquefatnon of the foundatica material under the support can mean lou cf e = 0.6 -

{

abt ' ptwet.

R t en-nh rain.n alsd must be gn en. beth for busied and abos cse..un.1 rigwt ne.

m which r is the wall thickness and R is the pipe radius, both in the same to ibe sclun e m. nons ainmg p,.m t.mits < reums the g q.cl ne V,,

41... :

unist Actual pipe normally will begin to wrinkle at strains one-third to one-quarter her u. " it dat is emtr.is...cser.i(,ct m es..ur e.<,,t.c,,

I that. $nlue. Strains of the order of 4 to 6 times as great. can be sustaincJ l '. t r h.i i':c.- r.< r.1 r t r.r.e....a,

....,,y..,,.,,,,,, o. r....

Ileuose of the shift

,,06.m dyr of nig at du compacuion w rinkle.

t-

. :< t, ~ a..,,.

' t she neusta! a mis away Isom the compressive side w hen w rinktmg occm s.

4 8,,

s, i,n.,t,,tenns will bc tonsulerably le5s than the compteuis e str ams under i

s-re,<,.

i

....s

i..

101 SEISMIC DESIGN CiuiERtA fCf j!

la general, this alternativs is slightly conserv:tive for most cases and 100 NOVEMBEfl 1978 TC, ll these conditions. If the stresses causing wrinkling arista in part from therrnal

.idequate since its degree of conservatism is relatively sma.

l effects or other secondary sources, which is usually the case, the likelihood Grasily Loads -The effects of gravity loads, when structures detorm la by a considerable amonnt, can be of impoitance, in accordance w j

of failure as reduced even more.

recommendations of most entant codes, the effcces of gravny loads are

.sion Cantam ANo Pnocaounts ron Srnucrunts Aho Aeovranoumo Pinums added directly to the primery vnd canhquake effects. In gracral, in comp the auual deflection the effect of travity loads, one must take into accousa Dnign Considerstluus.--For the design classifications used, C! ass ill is and not that corresponding to the reduccd seismic coefficient. In other w considered as falling under the provisions of extant codes fot ordinary buildings if one designs for one.fifth of the actual acceleration, as one does wtie Thus, the concept is implicit in the recommendations made herein that Class scismic Class 111, Ibc actual total lateral denections of the structure 111 items should not have design levels lower than those for the applicable by muhiplying the clastically computed deflections for the design a 4

codes, such as Itef.15. Normally it would be c=pected that major structures Un9mmetrical Structures and Torsion.-Consideration should be g by five, and aboveground piping will be placed in Class 11, except under ciicumstances where buildings, piping, or equipment can be permitted to deform a great deal, effecis of torsion on unsymmetricalstructures,and even on symmc nical ytru e

l ding where torsion may arise accidentally, because el various reawns, inc u or are not essential for safety, or will not seriously damage any c!cments or lack of homogeneity of the structures, or the phased wave motions d items that arc essential fer safety.

f After selecting (Le classification, the design spectra can he drawn b, use in carthquakes. The accidental eccentricitics of the horizontal forces of structural design motions of the type given in Tahic 1, and the applification, l by current codes require that 5% of the width of the structure m th and damping and ductility values like those of Tables 2 and 3. The design of the carthquake motion considered be used as an accidental eccentric coefficients are then determined. One iay choose to use the sunple methods stresses arising from the actual eccentricity shoulJ be combincd with tho i

of analysis picscribed in the various huildirn codes. In the event that a dynamic arising from the accidental cccentricity in all cases. The effect of eccen analysis is made it is secommended i;iat h response spechum technique be is to produce a grester suess on one side of the structure than on the l

used. The various mcies of response are computed, and then for each madal and the outcr walls and columns willin generalbe subjected to larrer def f requency the spectrum amphiication faciers are read from a plot similar to and forces aban would be the case if the struuure were cu that of i:ig. l.The various modal valurs for ttress deformation. or other res,onse Oscrturning and Moment und Shear laishibution.-In rencral when mo uniformly.

at a particular point are then combined for the various modes 15y taking the l

analysis techniques are not used,in a complex structure or in one havin square root of the sums of the squarcs of the individual modal responses.

degrees-of.fsecdom, it is necessary to have a method of defining

'Ihe dynamic analy>is procedure should generally be used 4 + complex or design forces at each mass point of the structure in order to be able to co r-simplified unusual structures, but it is quite adequate in many cases ir h

the shears and moments to be used for design throughout the structure. T code procedures with the appropriate scismic coef ficicner.ets.w

' irom the method dcocribed in the SliAOC Code (15)is preferable for this purpose.

response spectra constructed by the pmcedure descrit in.

i f

Attention is called to the fact that the design spect

.. l'ig. I for Class I, Class II, and Class 111 can he used only to obtain accclcration levels or Dssion Cantna Ano Pnoctounts von Bumto Perunt scismic coefficients but not deflecti ms or def ormations. In order to obtain dirlacements or dellections, one must multiply the design spectra by the For buried piping, the pipe in general will deform with the a,round, an j

ap,4epriate value of ductility factor, such as that given in Table 3. In general, i

strain in the ground will be transmitted to the piping without attenuation this will lead to displacements that are equal to or greater than the clastic faults intercept the pipeline, and the motion is greater than that which ca spectral displacements in all cases. For frequencies higher than about 2 hz, be absorbed by change in cross sections of the pipe itself, the structura the total displacements are slightly to considerably greater than the corresponding must be made to prevent the pipe from ruptpring because of the fault clasuc displacements, but for lower frequencies /they are precisely the same.

These considerations are covered in the fotfowing.

Combining llortionf al and Vertical Scismic Motions.-For those parts of In general, carttuguake motions on buried pipe produce essentially structures or components that are af fected by motions in various directions, rather than primary, ef fects since the strains o. deformations are fixed in r

1

- the ret respon'.. n t y A computed by cither one of two methods. 'Ihe first and the sire of the pipe or thickness of the wall, or quality of material, d method involves computing the responses in a particular direction at a particular not affect the strains appreciably. 'the implications of yielding and wrink point for cach of the directions independently and then taking the squarc root ;

in the pipe caused by ground motions, as covered earlier, may have a of '

. c'il nares of the resulting responses as the combined response.

different from the cifcci of secondary stiesses on aboveground structures.

Alternatively, one can use the procedure of taking 100% of the motion in one Stanins in lluried Pipe.-liccause a buried pipe conforms to the strains a direction, combinul with 407 of the motions in the other two mthogonal defo,mations in the medhun in which it is placed, both longitudinal strain a j

directions, then adding the absolute values of the resulting responses to obtain curvature are induced in a buried pipeline. Actually, because of slip between t

• {

the matimmn combined value in a member or at a point in a particular direction.

j j

• - -N '- - - A-aa w..:. a.z.:...;,

e 103 SEISMIC DESIGN CRITERIA NOVEMBER :D8 tci ICs OcndIh3 r02 e with x, the shearing d'istortion y in the element is y== (F/c) ces the pip and the medium, and local deformations between the two, incia.fm some si cht ova!!ing of th* pipe, the d,:fo>mations of the p pc...sy be e hti ma im m value is given by r

m than that of the medium. It a generally not dJsitable to consider a reductio 9

f rom the stram in the medium for the above reason however. One can mak

'~

inferrences about the relative motier.s betwycn nearby pois.ts in a pipe as outlined

- b

'~

here (3). For enmple, consider two points at a distance b apart, and consider a displacement p at point I and p plus an increment at peint 2, in the same r!

directien as at point 1. If a wave is pro'pagated from point I towards point

. (10)

A with a displacement of the form given by c,,, = v, n.

j For either Eq. 7 or 10, slippage of the soil against the cicment may reduce

. (2) -

p = f(x -- ct) -

the force transmitted to it from that corresponding to the strems determined m which c is the velocity of this particular wave propagation and iis the time' i from the equations.

v' n.clocity, then the various derivatives of the displacement p with respect to x and t are In applying the preceding expressions the valus of wave prepar given by the following relations (4):

f c, to be used in arriving at she pipe strain or curvatore is the cifcctisc veloc applicahic to the type of motion and medium being considered. In the case j

8 Jp Jp of shear wave cifects, which is typical, the effective value normally should j;

, (3)

I

{ "/' (* - C'); p =/~(x-ct) not be taken as the vahse at the surface nos the value at great depth in underl i

de a'p strata, but instead as the value representative of the actual motion rf the medium

......,,.... (4) surrounding the point of interest; in general the propagation takes place in a

{ " -C/'(2 - C'); p = c'f*(x - ct) manner represented in Section 5 of Ref. 8. lixamples of effective selocity for f ;,

From the first of Eqs. 3 and 4 one derives Ibc following result:

sheating type propagatitm in diff,: rent media that might be representative unde spe<:ial circumstances aic 4,000 fps for rock or permafrost,3,500 fps for m giavel deposits,3.000 fps for sand and competent soils, and slightly lesse ap g g

........... (5 )

for silt and clay deposits. Obviously, significant relative motians can occur at c at at ground medium Oanssuon zones (e g., rock to soil) and these situation mu stnd s milaily, from the second of Eqs. 3 and 4 one obtains i -

te cancfully considered in design.

/

d'p I d'e On a gross basis considening relative settlement a strain in the pipe of the 4.ft order of 0.004 has been a common operatins limit of deformatimi; for P

.......... (6) a,a y 3,3 j

a diam pipe, for example, this cmreslumds to a radius of curvatme of the pipe f

In the care where p is in the direction of x, then the strain e is obtained of the order of 500 n due to relative settlement. This would correspond to i

{

from Eq. 5, and the maximum strain at point I is therefore about a 1-in. settlement over a lent - af 20 ft, or a 2-ft satiement over a

{

length of '00 ft.

c A reast 'able critcrion for permisdble deformation to avoid supture appears v'

.,,,.,,,, (7)

I e~=1 -

to be of the order of 1% to 2% strain in modern steel pipe at any section, c mputed on a nominal basis, or appsoximately twice as much at points of i

In the case where p is perpendicular i,o t, either horbontally or vertically, Juess concentration, such as near welds or abrupt cross-sectional changes, takin i

the max _ mum curvature et point I is obtained from Eq. 6, and is as follows:

into account the local strain or stress concentr3tions. To reach a str i

order of 0.01 (corresponding to a ductility factor of moout 5) wouhl correspond a~

-g to a radius 0.4 as large as thosejust cited which would give relative displacements Curvature =

of the order of about 2.5 times those computed.

c Hu ed Wpe.-Herause a hun. d p4e is sub. W essentiah to e

ec in which a,, as the manimum acccleraH.i at point I. The strains corresponding lontdudinal strain that is fised in arnount, it can he considured as having a d ng to sin h cifccts are commonly quite small-

'Vhen the disp!acements m the region considered are associated with horitimial I "

"E f. *. I. ""

"E E#"#

shearing displacements occurring without longitudinal or extensional stram. ther-

1. 1. P the displacernent p is perpendicular to the wave front. For thi* case she.c i*

'#' #5P#

atiso an estensianal deformation of an embedded element such as a t :w' "

g letumalmet Q ures m uQ quoted h abaW ten

• h or tunnel. but the relations toscaning it are slightly different it. m I,

e i swime thu al.e snatesi.gl quahtees as measured by esanunen temperature ro t t e e-

• s, the uw where p is numal to x, but the elemers wn.nicac.:

104 NOVEMBER 1978 t r. i TCl SEISMIC DESIGN CRITERIA 105 or in appropriate other ways (e.g., fracture mechanics concepts) will insure estinistes of the maximum permanent displacement after the excitation has against brittle or ductile fracture at these levels of strain. The permissibic stum stopped in general, the transient displacement docs not eseced about twice levels would have to t e reduced in the event such assurance is not possible, the maximum permanent displacement, I ut may be considerably less, especially Although brittle tractures are not likely to occur under compressive strains when the permanent displacements are large.

t in the pipe, wrinkling or buckling can occur. In general, cu:'ent piping codes in studics of slope stability, the value of N is determined by taking it as require that the strali, in piping that is constrained be limited te somcwhat generally equal to that constant acceleration of 41 ' earthquake applied as a constant dynamic force which gives a factor of safety of one for the slope.

less than the minimum specified >ictd strength. This is probably desirable from l

the point of view of operating limits since buckling may produce difficuhics If one-half the maximum ground motion acce!cration is taken as N, or if the in operation, but insofar as allowable maxima are concerned, it does not appear j slope has a factor of safety of one with 1/2 A statically applied, one can determine from liq.12 the maximum downslope displacemc it, and normally i

necessary to limit the compressise strain even for buried piping to values less this will be a few inches in magnitude.

{

than about t% to 2% strain. This, howe er, should be the simi of the strains p

)

at a peint in a given direction from all sources, including thermal, pressure, j

It should be pointed out that most embankments and copetially most carth and scismic deformation. Particular care needs to be exercised at bends, either l and rock fdl dams are designed for a considerably smaller factor of safety

+

}

side bends, over bends, or sag bends, however, to avoid bucklirg or compressive than that which would correspond to a value of dynamic intor of safety of l

i or tensile failures that arise from the combined longitudinal stress and moment.

1.0 as previously defined. Values of displacements of several feet are quite i

In some cases where special provision must be made for deformation of nosdht in well designed dams and dikes without failure ou urring, and have belowaround p.iping, the mounting of the piping in tunnels with special supports, been experienced in practice. In any event, if the ratio of N/.I is greater than j

as shcwn in Fig. 2(c) may be woithy of consideration; normally however this about 0.5, one does not or linarily consider that failure of an embankment will techn que is quite exp;nsive.

. occur. Where the values are less than this, then the embankment design must i

be considered more carefully and in detail. The same comments apply to dikes Seccin Geoncamen Desics Psovisions I

and similar structures, induding gravel pads or berms.

it is estimated that slips of one foot or less for a slide of normal dimension Stability and D namic Mosements of Slopes.-The movement of slopes and I

would cause plastic deformation but not cracking of standaid dimension metal 3

embanLments under scismic conditions is covered in Ref. 3. I or e ;ual resistaaces buried pipe; pipe made of blittle material normally could not uithstand such to shdm, g m both directions the maximum motion is given by the relation:

m mion.

V' [I - N\\

j Where slopes are unstahic and on the verge of sliding.tatically, the use

.... (11)

I of liq.12 would indicate that large displacements would occm Such rcE ons I

i 2gN A

should be avoided or, as an alternative, contingency plans shouhl be descloped.

I in which Vis the maximum ground velocity in the carthquake, g is the acceleration Obviously where exceptional cases are encountered special engineering consider.

I of grasity, N is a measure of the dynamic resistance to sliding (as determined ations may I e necessary, from the constant horizontal or inclined force which, if apt. lied after several I.lquefaction potential.-.One of the most serious consequences of an carthquake cycles of shakir.g and consequent loss of strength, wouhl produce sliding with is the effect of changing the properties of inundated samh or cohesionless 7

a factor of safety of unity), and A is the maximum ground motion acceleration materia's so that they become " quick" or develop a liquefied condition. One in the earthquake considered. The most serious ca<e is that in which slip takes method of dealing with this problem is covered in Ref.16. Ilowever, it must piace in only one direction, corresponding to an unsymmenical resistance to be noted that several peaks of high acceleration have little influence compared motion. The maximum displacement under carthquake loading, as computed with that of long. sustained motions, even of a less intense naturc; the latter can have a serious effect on the liquefaction potential. For this reason, it is from the extreme values for a number of calculations from different carthquakes, is approximated by the relation:

expected that high intensitics of only 2 or 3 spiken of acceleration can be discounted and the liquefaction cifects can be computed for longer durations of shaking with a corresponding lower acceleration level not less tlian the de'ign ground s

. (12) 1-motion value such as those presented in l'able 1. This procedure also tales 2xN \\

A/N account of the high damping values expected.

't.he vahic's given by I.q.12 becomes highly o'verconservative when N/A is l{ffect of I'ar : Mutions.-The tiansient disphccments at some distance away less than about 0.15. In this case,1,q.12 should be replaced by an upper

,g gg Isund of:

si do wt emqd tone WiMi@amn A ph 6 V' for, estimating fault motions is presented in Ref.12. Ilowever, at or adjacent (13) to a fault, large relative motions can occur. In rock, these relative motious 2M take place over fairly.hort distanca, and therefo e for buried pipe provisions Eq s. 12 and 13 are based on rigid. plastic resistance to sliding, and give good must be made to cushion the pipe against the abrupt changes in displacement.

SEISMIC DESIGN CRITEftlA 107 10G NOVEMflEll 1978 TC1

.This can be done most readily by arranging for the excavation of the,,,

Engineering and Design, Vol. 20, No. 2,1972, pp. 303-322.

f in which the pipe is placed to have relatively shallow (45* or less) side st*,feb

6. Newmark, N. M., J Study of Vertical and florizontal EarthecoAr Spectra. USA EC Irport WASil.l235, Consuhing Engiucering Services, Superintenent of Documents,

,i with a h.mited depth (..ot more than 3 fi-5 ft) of gravel cover over the pipe U.S. Govt. Printing Office, Wehington, D.C., Apr.,1973.

3

[ Fig. 2(bll, so that the pipe will b lifted up and out rather than crushed and

7. Newmark, N. ht., "Scivnic Design Caiteria for Structures and Facihtie.. *l,m Alaska I

cracked if a fault occurs transvetsc to the pipe. With nearly ventical slopes Pipeline System," f roceedings U.S. National Con /trence on farthqua#r En8m" ring.

f l Fig. 2(a)l, the rnotiou would tend to cut or constrict the pipe in a way that

" Esuhquake Enginecting Research Institute,1975, pp.91-103.

I

3. Newmask N hl., et al "hfethods for Determmmg Site Characteristio, Proceedings j

would cause danger of over:, training an I possibly failure.1-ault mot. ions m soil r,wM Cor(creib eri Alictmnation for Safer Construction Research and I,

ate not nearly so serious as they are in rock because they do not occur so afpfsca,jun, p3p. UNESCO.Univenity Washington-ASCE.Acadmy of hicchanics, absuptly. In soil they woulJ be expected to correspond to more gradual Seattle, Wash., October. November,1972. Vol. I, pp. 113-129.

displacements in which the local pressures ruinst the pipe would cause some

9. Newmark, N. ht., and 11 11, W. J., "Scismic Design Criteria for Nuclest Reactor Facihues," Trourdhss hunh World Con /trencr on fartAquaAc Engwrmg. Santiago, defo mation r

• the pipe but not the Lind of crushing or damage that would Clule. Vol. II, B-4,1969, pp. 37-50.

be caused by laultmg m rock; fu some cases it may bc desirable to place the

10. Newmark, N. hl., stal llall, W. J., "ProccJurcs and Criteria for Earthquake Resistant pipe on the surface [ Fig. 2(c)] on in a hcrm on the surface Fig. 2(d).

Desir,n," liuilding tractice: for Disaster Alitigation National Bureau of Standards, Vertical fault motions are not generally serious for metal pipe since the pipe lluilding Science Sctics M Sept.,1972, pp. 2u9-2.16.

norma ly has the capability to resist n 4ft or 5.ft vcrtical displacement without II, Nt+ m A N. lLI.. and lMI. W. J.. "Sristnic De.i n 3 c(tra for fran, A f.Mn Pipthne,"

t f I

undue difficulty, coriesponding toloss of one or Iwo of the aboveground suppo "

'[8' # I'A #*'## C#'N""" "" E"'#h "" A' # "# ""'#"3' V"L i' i9I4' EP' i

ifit can become free it can accommodate sigmficant motions ove: large distancca.

12. Newmark, N. hf., and Itall, W. J., "papeline Design to Resist Large 17ault Displace-Whcre fault motions might occur, it is important that the depth of cover ment," Proceedings U.I National con /<rence on for Aquale Engin<cimg. Earthquake over the pipe be limit-d, as explained "bove, and that anchors or bcnds of Engineering Research Institute.1915, pp. 416-425.

any sort not be placed within a distarice of at least 200 ft cither side of the

13. Newmask, N. ht., and IMi, W. J., " Vibration of Structures Induced by Ground hiotion." in S;.mA and hbra:Un llandbooA. C. hf. Ilarns and C. E. Crede, eds.,

l cxpected fault area. It is also des.iral.lc that the pipeline intersect the fault 2nd Ed., Chapter 29, hfcGraw Ilill llook Co.,1976.

at nearly r85 t angles, or in any case not make an angle less than about 45*

14. Newmark, N. ht., mm. Rosenblueth, E, f undamentals of EarthquAc Englatering, h

I with the fault trace. If the geometry and direction of relative slip are known Prentice.Ilatt, Inc., Enclewood Chtfs, N J.,1971.

I it may be desirable to incline the pipe axis to the fault slightly'to produce

15. "Itec"mmended Lc'al l'o'cc Req.utements and Commentary," Strmtural Engtncers either knsion or compression in the pipe depending upoa the design criteria

"'f(" "[ 'ff*" *["" *[{

.hg,,f;o*n of So.I Liquefaction Potential

^

I 3

g and to accommodate such items as the type of pipe mate,ial, and over bends During Easthquakes," Report No. EERC 73 23, Earthquake Engineering Research or sag hends. Also, long. term creep ef f ects near Inhs may need consideration Center, University of Cahfornia Ilerkeley, Cahf.,1975.

in design.

17. Wdson, W. hi and Newmask, N. bl., "1he Strength of *Ihin Cylindrical Shells as I

Columns," University of Illinois Emperiment Station Bulletin No. 255, U bana, Itl.,

lI Acunowtroovtur Feb.1933.

11' i

3 The criteria and procedures described have been developed by the writers j

over some period of yrans in connection with design studies for the Trans-Alaska pipeline (7,11), the Canadian Arctic gashne, the Schio pipeline from Long Ileach, I

Calif. to Midla nd, Tex., as well as for ot her special facilitics. The recommendations made herein are not to be construed as representing an official position pertaining i

j to any of the projects identified.

j f

Arrtuoix 1.-Rrraneness I. ASAIE Guidefor Gas Transmission and Distrobution riping Sprems, (including Federal e

3 S *fety Standards), American Society of hicchanical Engineers, New York, N.Y.,

g gg 3

2. t.iquid Petroleum Transportation l'iping Systems, A NSI BJf.41977, Amencan National Sundards *nstitule, New York, N.,Y.,1971.

. u. M., "Prects of liaithquakes on Dams aml Enhankneents," Geotechni-

3. N-wmw que. Lon tun, England. Vol. XV, No. 2, June,1965, pp.139-159.

4. Newm.uk, tv..*.t., "Problen s in Wave Propication in Soil and Rock," hacerdmA' 8

Intonatumal Symposium on it' ave Propanarson and Dynamir Fr.>perties of Eo'th Alate rials. Umversity of New hiexico Press, Albuqncique, N.ht.,1968, pp. 7-26

5. Newmaik, N. hl., "liutthquake I esponse Analysis of Reac tor Structures," A us t" I

,