ML19257A904

Site-Specific Response Spectra.

ML19257A904
Person / Time
Site:	Oyster Creek
Issue date:	12/31/1979
From:	JOHN A. BLUME & ASSOCIATES, ENGINEERS
To:
Shared Package
ML19257A902	List: ML19257A901 ML19257A904
References
TASK-03-06, TASK-3-6, TASK-RR NUDOCS 8001090371
Download: ML19257A904 (17)

Text

SITE-SPECIFIC RESPONSE SPECTRA FOR THE OYSTER CREEK NUCLEAR POWER PLANT prepared for Jersey Centrel Power and Light Company Morristown, New Jersey December 1979 s URS/ John A. Blume & Associates, Engineers 130 Jessie Street (at New Montgomery)

San Francisco, California 94105 1707 296 800109 r~  !

CONTENTS Page INTRODUCTION ........................................................ I

SUMMARY

............................................................. 3 METHODOLOGY ......................................................... 4 Response Spectra ................................................. 4 Random Vibration Method .......................................... 4 Input Parameter for Determining Peak Ground Acceleration . . . . . . . . . 7 Site-Specific Design Spectra ..................................... 10 REFERENCES .......................................................... 14 TABLES 1 Peak Ground Acceleration for Various m Duration of S t rong Mot ion . . . . . . . .............................

. . . .bLg , Di s tance , and 9 2 UHS Data Set, Oyster Creek ....................................... 11 FIGURES 1 Oys ter Creek Mean Horizontal Spectra for 5% Dampi ng . . . . . . . . . . . . . . 12 2 Oyster Creek Mean Horizontal Spectra for 0, 2%, 5%,

and 10% Damping .................................................. 13 8

1707 297

- ii -

INTRODUCTION This report describes work performed by URS/ John A. Blume & Associates, Engineers (URS/Blume), in developing a site-specific response spectrum for the Oyster Creek Nuclear Power Plant. The report is prepared for the Jersey Central Power & Light Company of Morristown, New Jersey. The spec-trum is intended for use in engineering design review and analysis of exist-Ing facilities at the Oyster Creek plant.

At this time, the use of site-specific response spectra is encouraged by the L.5. Nuclear Regulatory Commission (NRC) staff. This is because of improved analytical methods and the increased number of strong motion earthquake records available for use in analysis.

The methodology used in the past to establish seismic criteria for most nuclear power plants in the eastern United States utilized the generic re-sponse spectra presented in NRC Regulatory Guide 1.60. This spectrum is anchored at a zero-period peak ground acceleration based upon the Modified Mercalli intensity (MMI) of the Safe Shutdown Earthquake (SSE). The appro-priate level of peak ground acceleration for an earthquake of a given inten-sity is determined from published studies of strong motion earthquake data that relate intensity to peak ground acceleration.

The problems associated with this method of developing response spectra, which have been widely discussed, are generally recognized to be significant.

Some of the major problems cited are the highly subjective nature of earth-quake intensity ratings, the large scattering of data in the intensity-acceleration relationship, and the conservatism of the NRC Regulatory Guide 1.60 spectra, which are based on ground motion records of earthquakes of varying magnitudes.and epicentral distances in different geologic environ-ments.

These problems are alleviated to a large extent by the technique for develop-Ing a site-specific response spectrum that is used in this study. This tech-nique is based on selection of a suite of strong motion earthquake records that characterize both the SSE (i .e. , magnitude and epicentral distance) and the geologic conditions of the power plant site.

.i_ 1707 298

This study begins with determination of the magnitude appropriate to the SSE /

maximum MMI of Vil developed in accordance with regulatory procedures prior to design of the Oyster Creek Nuclear Power Plant. Since this is the same intensity that would be arrived at following the current regulations of 10 CFR 100, Appendix A, the study does not review this development.

The following sections of this report explain the methodology utilized in de-veloping the site-specific response spectra for the Oyster Creek plant; they describe the calculations, provide the input parameters utilized, including the suite of strong motion earthquake records, and present results of the calculations and the site-specific response spectra.

1707 299

SUMMARY

In the mean value site-specific response spectrum developed for the Oyster Creek Nuclear Power Plant site, shown on Figure 1, the zero period peak ground acceleration is 0.079 The basic input parameters used to develop this spectrum are as follows. The TSE is an event of Modified Mercalli intensity VII immediately adjacent to the site. The corresponding magnitude (mbQ) is 5.3. This site is underlain by moderately deep alluvium. The characteristic epicentral distance is 5 km.

The earthquake mechanism is considered to be characterized by normal faulting.

1707 500 METHODOLOGY A computer program based on a new way of predicting the maximum response of single-degree-of-freedom systems to which the response spectra shape can be anchored was implemented to predict ground motion spectra at specific sites in the eas tern Uni ted States.

The methodology employed is an application of the random vibration theory (Vanmarcke, 1976) with the data needed for the solution based on empirical work (Street and Turcotte, 1977) and engineering Judgment. The anchor point of the response spectrum at period T = 1 sec and the maximum expected ground acceleration for rock is obtained for a specified spectral moment (g),a given epicentral distance, a strong motion duration, and a description of the spectrum within the period range of interest. The desired spectral shape can also be scaled with the corresponding transfer function so that the soil char-acteristics of the local site are included.

Response Spectra The maximum response is of practical value in the analysis and design of elas-tic structures. For a simple one-degree-of-freedom mathematical model of a structure with a natural period T and a damping ratio S, a plot of the peak response to a given excitation as a function of the period is known as the response spectrum, it is useful to present the displacement, velocity, and acceleration spectra in one plot, as in the tripartite response spectra. Here the ideal character-istics of a double-degree-of-freedom oscillator are exploited so that relative displacement (S g ) and pseudo-velocity (S p = UnDS ) are both presented.

Random Vibration Method Because of our limited knowledge of the physical process that produces seis-mic motion and the almost total absence of strong motion data for the eastern United States, a stochastic description of the earthquake process becomes in-evitable. Such a description is very useful in predicting the occurrence of events because of the availability of the analytical tools develc'oed in the theory of random processes. f 4_

When the theory of random process is applied to a time history ( (t)), the whole ensemble.of possible time histories which might have occurred as a re-sult of the phenomena under consideration, not just one time history, are described. An individual time history belonging to the ensemble is called a realizatlon.

A major simpilfication is normally used, in that the random process is assumed to be invariant with time in its statistical description. In this case, the stochastic or random process is called a stationary stochastic pro-cess, and we can write the autocorrelation function (E[X(ti )X(t2 )]), which is the expected value of the product of the value of :(t) at two different times ti and t2 , as a function of the difference T = t2-t1 : E(X(t )X(t2 ) = R (T)l.

i where R (T) is the autocorrelation function of the stationary stochastic pro-cess X(t) .

The spectral function, S (u), can be defined as the Fourier pair of R ( ),

so that 3_(t)

=

S_(u)e cd> (1) and S_(c) = dt (2) h .. R (t)e" Both the autocorrelation and the spectral functions are a complete represen-tation of the stationary stochastic process :(t)

Empirically derived relationships that relate the observed vertical displace-ment spectrum 0,(u) with the source spectral function S*(u) at a fixed dis-tance are available. Among them, the one found in Street and Turcotte (1977) is applicable for the eastern United States. It is:

3 knob gr (r/rg) 0,(u), when r s r S* (u) =

(3) 43083 p (pfp )1/2 Op (u) , when r > r O where p = soil density, taken as 2.5 gm/cm-3; s = damping, taken as 3.5 km/sec rg = a fixed distance parameter = 100 km; S*(u) = source spectrum at r, distance; and Dy (u) = observed vertical displacement spectrum.

, 1707 302

From the same reference it can be observed that the frequencies around T =

1 sec, (a = 2n), are well behaved. Also, the following relation between ob-served m and S*(T = 1 sec) has been derived:

3 (17.5 + ra"g)

S* (T = 1 sec) = 10 -

(4)

It is easy to obtain fran Equation 3 the site specific vertical displacement spectral level Op(T = 1 sec) for any specified distance P. This site vertical acceleration amplitude spectrum A,(T = 1 sec) is readily calculated by:

A ,(T = 1 sec) =

0,(T = 1 sec) - (22)2 (5)

The development up to this point has been for vertical motion (used in obtain-ing ng ;g). This can be transformed to horizontal acceleration amplitude by multiplying by the ratio of horizontal to vertical acceleration. In this case a mean value of 2.4 was obtained for this ratio f rom a study of 70 strong motion records of eastern United States earthquakes.

The power spectral density function (one-sided) of the ground motion C,(w) can be estimated by smoothing the Fourier acceleration spectrum, G,(T = 1 sec) =

h lA r(T = 1 sec)l2 (6) where s denotes strong-motion duration. As the assumption of stationarity implies a uniform energy distribution in time, the duration s must be adjusted so that an equivalent duration, called e ,g and the averaged G,(w) give the total energy content of the earthquake. Here the definition given in Vanmarcke and Lai (1977) will be used to that effect.

We need the predicted response spectrum ordinate at period T = 1 see to anchor the selected shape of the normalized response spectra. A nonstationary random vibration analysis following that of Vanmarcke and Lai (1977) is done to accomplish this task: Given C,(w) of a single-degree-of-f reedom system and the strong motion (equivalent) duration s g, the pseudo-velocity response spec-tra Sy(T,6) is predicted with the general form Sy(T,6) = ao p(sg) (7) 1707 303

in which cy(ag ) = time-dependent standard deviation of the pseudo-velocity response, evaluated at t = s,; ap is a peak factor function of the probabil-I ty of nonexceedance, p.

The variance 2c (t) is obtained by integrating the spectral density function over all frequencies, c2 (t) = C(u ,t)6 (8)

An approximate solution for moderate natural frequencies, including T = 1 sec (o = 2n) and for relatively large damoing values (such as 8 = 0.05) is, for T=1 sec:

C,(T = 1 sec) 1/2 o (s0) =

8S (9)

The peak factor is the product of an analysis of the maximum. Exact solu-tions for ap do not exist, but an approximate solution that has been exten-sively checked is, for T = 1 sec:

a p

= [2 In (-2Sg /In p)]1/2 (j o)

From Equations 7, 9, and 10, we have C,(T = 1 sec) In (-2sg/Zn p) 1/2 SU (T = 1 sec, S) =

(11) 48 Another value that can be approximated f rom these data is the peak ground acceleration, a . This is done empirically; from a set of strong motion data the ratio a#

k =

(12)

S p(T = 1 sec, 0.1) was estimated. For a in cm/sec2 and s p in cm/sec the mean value was k = 10.46.

Input Parameter for Determining Peak Ground Acceleration in order to determine the peak ground acceleration, certain parameters are needed as input. The primary input is the maximum magnitude expected at the ti 7 n. , 7

Ps-4

site. The magnitude scale used is m bLg devel ped by Nuttil (1973) and is most appropriate for the East Coast. Since the maximum historic event asso-clated with the Oyster Creek site is known in t..rms of epicentral intensity (MMI VII), a relation is needed between epicentral intensity and m bLg. Using relations developed by Nuttii and Zollwig (1974) and Street and Turcotte (1977) relating intensity to mb (body wave magnitude) and mb to mbLg (* 9"I-tude determined from the higher mode L wave), the following relation between m nd intensity was derived:

bLg "bLg

=

.485 I, + 1.92 (13)

Using the maximum MMI determined for the Oyster Creek site of VII one obtains an m bLg v lue f 5.3 The next parameter needed is the distance from the source to the site. A distance of 5 km was chosen to approximate the condition in which the earth-quake occurs under the site. This distance would represent an average focal depth for the site and eliminates the problem of having the acceleration achieve unrealistically high values as the source-site distance approaches zero.

The duration of strong shaking was estimated to be approximately three to five seconds (Bolt, 1973). These values are also consistent with estimates of duration measured for the strong motion records used to develop the site response spectra. The last parameter to input is the damping value, which for characterization purposes was put at 5%.

Using the parameters of m = 5.3, distance = 5 km, duration = 3.0 sec, and b4 damping = 0.05, a calculated ground acceleration of 0.072g was obtained for the site. This value of acceleration was then used to anchor the site spe-cific design spectrum derived for Oyster Creek. In order to see what a slight variation of the parameters would do to the ground acceleration, a preliminary sensitivity study was performed, the results of which are pre-sented in Table 1.

1707 305 TABLE 1 PEA: 3ROUND ACCELERATION (g) FOR VARIOUS *bLg, DISTANCE, AND DURATIOfJ OF STRONG MOT 10N l

Distance (km) 5 10 15 l

Duration (sec) 3.0 5.0 3.0 5.0 3.0 5.0 mean .0459 .0399 .0239 .020g .015g .0139 5.1 mean

+ .0579 .0489 .0299 .024g .0199 .016g la mean 0729 .0629 .036g .0319 .0249 .021g

• bLg 5.3 _ _ _ _ _ _ _______ _ _ _ _ _ _ _ __ __ _ _ _ _ _ _ _ _____ _ _ - _ _ _

mean

+ .0919 .075g .0459 .0389 .030g .0259 la mean .1149 .098g .0579 .0499 .0399 .0339 5.5 - - - - - ----- - - - - - - - - - - - - ----

mean

+ .1449 .1199 .0729 .060g .048g .040g lo 1707 306 Site Specific Design Spectra A totu: of 34 strong motion records were used to develop site specific response spectra. Twenty records were from the western United States and fourteen from the 1976 Friuli, Italy, earthquake. The accelerograms were chosen to satisfy certain criteria. The records were all recorded at soil sites and at distances of less than 54 km. The earthquakes ranged in magni-tude from 4.7 to 6.1. The peak accelerations varied from .0279 to .308g, with a mean peak acceleration of .115g and standard deviation of .073 (see Table 2). The spectra were all normalized to the zero period acceleration and then statistically combined (assuming lognormal distribution) to obtain a mean response spectrum. The spectrum was then anchored to the peak ground acceleration value determined by the above method. It is shown in Figure 1.

The spectra for 0, 2, 5, and 10% damping are shown together in Figure 2.

1707 $07 TABLE 2 UHS DATA SET - OYSTER CREEK Epicentral 1.D. Earthquake Location Datt/ Time Accelerometer Station Mg disttnce Component & pga (g)

P.n (kr)

An10 San Jose, Ca. 9-4-55 Bank of America, 5.8 10 N31W: .302 San Jose N59E: .308 40!3 Sar. Francisco, Ca. 3-22-57 Southern Pacific Bldg., 5.3 17 N45E: .04?

San Francisco N45h: .046 A018 Central California 4-S-61 Hollister City Hall 5.6 21 $01h: .065 N89h: .179 B223 Long Beach, Ca. 10-2-33 Hollywood Storage Bldet. , 5.4 35 N90h: .027 basement, los Angeles N .033 8031 Wheeler Ridge, Ca. 1-12-54 Taft Lincoln School Tunnel 5.9 54 N21E: . 0t 5 569E: .068 U305 Central California 4-25-5: Hollister Public Library 5.3 36 50th: .053 N89h: .050 V316 Torrance-Gardena, Ca. 11-14-41 Long Beach Public Utilities 5.4 6 N: .040 Bldg., Los A.ngeles E: .055 V329 Souihern California 3-18-57 Pt. Hueneme Research Lab. 4.7 5 S: .367 W: .089 V336 Lytle Creek, Ca. 9-12-70 Hall of Records, 5.4 30 N: .116 San Bernardino E: .059 h339 '

Southern California EJison, 5.4 34 S: .0e2 Colton E: .059 FC059 Friuli, Italy 5 11-76/ Forgaria-Corvino 5.3 10 NS: .190 2:4402 Eh: .308 FC131 9 11-76/ Forgaria-Cornino 5.5 16 NS: .095 163112 Eh: .115 TAR 133 9-11-76/ Tarcento 5.5 B NS: .204 163112 Eh: .105 FC138 '

9-11-76/ Forgaria-Cornino 5.9 15 NS: .133 163500 EW: .235 B143 '

9 11-76/ Bula $,9 14 gs: ,;33 163500 th: .306 FC152 9-15-76/ Forgaria-Cornano 6,1 10 ss: ;e3 031519 Eh: .218 8156 9-15-76/ Bula 6 6.1 NS: II0 031519 Eh: .096

-n- 1707 308

0 2

8 1

'6 1 G N

I P

M A

D 4  %

[

3 5 R

O F

A R

2 T 1

C E

P S

) L c A T

0 e N i.s O 1 (

T Z I

R O

H N

8 A

i. E

' 0 M K

E E

R C

6 R 0 E T

S Y

O 1

4

' c, E R

U G

I F

~ '

'O 7 6 5 4 3 2 1 0 0 0 0 O 0 0 i

^ow

,N' .

uOW

_ ON

0 2

.8 1

G N

I

.6 P 1

M A

D 0

1 i41 5

2 0

R i2 O F

1 A

R T

) C E

c P e

i0 s S 1 (

L T A T

N O

Z I

8 R O

i0 H N

A E

M K

i6 E 0 E R

C R

E T

S i4 Y 0 O 2

E

_ R

- U 2 G i0 I F

7 6 5 4 3 2 1 0 0 0 0 0 0 0 n

^cv a i 3i . .

C N '+ . ..

.DN

REFERENCES Bolt, B. A., 1973, " Duration of Strong Ground Motion," Fifth World Conference on Earthquake Engincering, Rome.

Nuttli, O. W., 1973, " Seismic Wave Attenuation and Magnitude Relations for Eastern North America," J. Coophys. Research, V. 78, pp. 876-885 Nuttli, O. W. , and Zollweg, J. E.,1974, "The Relation Between Fel t Area and Magnitude for Central United States Earthquakes," Seismological Society of America, Bullctin, V. 64, No. 1, pp. 73-85 Street, R. L., and Turcotte, F. T., 1977, "A Study of Northeastern North American Spectral Moments, Magnitudes, and Intensities," Seismological Society of America, Bulletin, V. 67, No. 3, pp 599-614.

Vanmarcke, E. H., 1976, " Structural Response to Earthquakes," Chapter 8 in Seismic Rick and Engineering Dccisions, C. Lommitz and E. Rosenblueth, Eds., Elsevier Publishing Company, Amsterdam - Oxford, - New York.

Vanmarcke, E. H. , and Lai, P. ,1977, Strong Nation Duratkn of Earthquakco, M.I.T. Department of Civil Engineers Research Report.

1707 MI