ML20008F129

to App X, Effects of Near-Field Earthquake Ground Motion on Structure & Equipment Design, from Rept Entitled Supplemental Seismologic Investigation.

ML20008F129
Person / Time
Site:	Summer
Issue date:	03/31/1981
From:	SOUTH CAROLINA ELECTRIC & GAS CO.
To:
Shared Package
ML20008F123	List: ML20008F122 ML20008F129
References
NUDOCS 8103120320
Download: ML20008F129 (6)

Text

,

O APPENDIX X EFFECTS OF NEAR-FIELD EARTIIQUAKE GROUND MOTION ON STRUCTURE AND EQUIPIENT DESIGN 8103120 3 M Re s on

, t9Sl

EFFECTS OF NEAR-FIELD EARTHQUAKE CROUND MOTION ON STRUCTURE AND EQUIPMENT DESIGN The extensive body of site-specific observational evidence pre-sented elsewhere in this report shows that in this highly heterogeneous envi ronment there is a_ limit of approximately one kilometer to the maximum fault dimension for any single -earthquake occurrence. This corresponds to a limiting magnitude of ML = 4.0. The calculations of ground motion presented in this Appendix were carried out both for this limiting magnitude event and for a larger assumed magnitude of ML =.

4.5 which provides a conservative assessment of ground motion effects cnt the -Virgil C. Summer "uclear Station.

The ML = 4.0, stress drop of 25 bars, and R = 2.0 km case gives a zero period acceleration (ZPA) value of 0.14 g, which is less than the Virgil C. Summer SSE ZPA value of 0.15g' for structures on rock.

Therefore, this maximum induced seismic event has no ef fect on struc-

. tures or equipment.

The ML = 4.5, stress drop of 25 bars, and R = 2.0 km case gives .

a ZPA value 'of 0.22 g, which is . higher than ' the Virgil C. Summer SSE 4 ZPA values $f 0.15- g for structures on rock and is lower th'an the SSE

-l  ;ZPA value of 0.25 g for structures on soil. In the original seismic analysis, a very conservative 2 percent- damping value was used. NRC Regulatory Guide 1.61 allows a 5 percent damping value for' prestressed

~

concrete and 7 percent damping value for reinforced concrete structures in:the SSE; analysis. Thus,.the Virgil-C. Summer (0.15 g) SSE spectrum at 2: percent damping is compared with the ML = 4.5 (mean value plus' one standard deviation) event' spectra at 5 percent-and 7 percent i

damping in Figure-1.. As shown in'the comparison, the.ML = 4.5 event

exceeds the Virgil C. Summer (0.15 g) SSE spectrum only in the i

. frequency region higher-than about 9 Hertz..-The dominant frequencies

of all Seismic Category 1 structures are lower than 9 Hertz except for

,.. the-interior concrete structures of the-Reactor Building.- However,

, :since.the original seismic analysis used.the1 artificial time history.

. (ATH),as? input and the the ATH's

response spectrum'always exceeds or 1

een 3 _ _ _ m

e .

j' 4' l equals the Virgil C. Su=cer SSE spectrum, additional conservatism of

. ,- the ATH method can be used to justify the original seismic design of f

the interior concrete structures and the equipment contained therein.

-To remove the conservatism of the ATH method, the Oroville accelerograms were ustd in statistical studies. Four horizontal

. components of two ML = 4.0 aftershocks of the 1975 Oroville, California earthquake, recorded at rock sites, were modified to match the 5 percent damping target spectrum in the mean a's shown in Figures 2 to 5. Four Oroville af tershocks were extended into 36 components by adjusting the time increment, which achieves the effect of shifting the frequency-content of the accelerogram (Tsai, 1969). The original Oroville aftershock accelerograms have time increments of 0.005 second.

Each component was extended into 9 components by using time increments 4 of 0'.0038, 0.0041, 0.0044, 0.0047, 0.005, 0.0053, 0.0056, 0.0059, and e

q 0.0062 second. The 36 accelerograms were used as input to the Reactor N Building seismic analysis. - The mean values -of the 2 percent floor

, response spectra were compared with the original Virgil C. Summer floor

. response spectra used in design (Figures 6'to 20). As shown in the comparison, the Virgil C. Summer spectra envelop the ML = 4.5 sean v

7- value spectra in_the resonance peak region and almost all other

.n frequency regions. Thus, it is concluded that the original Virgil C.

]{

Summer seismic- design is not exceeded by this ML = 4.5 event.

Some NSSS equipment are designed to floor response. spectra generated at 5 percent structural damping, for loading combinations containing SSE'and LOCA. This set of. Virgil C. Summer floor response

-spectra.is also~ compared with the ML = 4.5 mean value floor response

~

' spectra (Figures - 21i to 35). As shown in the comparison, the original Virgil C.' Summer _ seismic design is not' exceeded - by the ML 4.5 event

~ and continues to be valid'and'a'dequate; W As shown in Figures 6 through 35, the Virgil C. Summer SSE floor response. spectra envelop the M.

g =4.5 event: floor response spectra in the 1 a' '

~

resonance peak' region and in almost all other frequency regions.. The

~

A 4

4

%) . ~

X-2i Revision 1~

,7 March,:1981 x - +

b- ,

only exception is in the frequency region of 20 to 30 Hz, where the Virgil C. Summer response spectra are exceeded by a very small amount. For piping systems and other multiple degree of freedom systems, the frequencies of dominant modes are always lower than 20 Hz. Higher modes have smaller participation factors. Thus, the combined stresses of multiple modes due to the Virgil C. Summer response spectra will always be higher than those due to the Mg=4.5 event spectra. There are a few relatively rigid systems and equipment with frequencies of fundamental modes in the region of 20 to 30 Hz.

These systems and equipment all have high moments of inertia and large section sizes in ord'er to reach high frequencies, and are originally over-designed. 1 Therefore, the slight exceedance of the Virgil C. Summer floor response spectra in the frequency region of 20 to 30 Hz will not cause any overstress problems.

To demonstrate the additional margins available in systems design, the seismic stress, design-stress, and allowable stress are shown in Table 1 for the emergency feedwater and residual heat removal systems. The margins between the required input and the actual input values of equipment seismic qualification are also shown in Table 1. As shown in the table, ample margins are availabic in the original design to accommodate the reservoir induced seismicity.

Revision 1

~

X - 2a March, 1981-

..--a -

.~

Y , . '.

4 TABLE 1 V. C. SUMMER NUCLEAR PLANT SEISMIC BUILT-IN DESIGN MARGINS DESIGN STRESSES ALLOWABLE STRESSES COMPONENT OR REQUIRED SEISMIC STR'ESS OR ACTUAL INPUT G VALUES INPUT G VALUES EMERGENCY FEE 0 WATER 19,000 PSI 12,000 PSI 36,000 PSI l PIPING (T0 DATE)

,000 PSI 11,6@ N M,M N LA (T0 DATE)

TURBINE DRIVEN .36G/.36G/.21G

.5G/.5G/.4G EFW TURBINE TEST 1 TURBINE DRIVEN EFW .36G/.36G/.21G .48G/.48G/0.4G PUMP APPURTENANCES TEST RHR PUMP & MOTOR NL S SAFETY INJECTION .29G/.24G/.19G

CHARGING PUMP ANALYSIS 3.0G/3.0G/2.0G Revision 1 March, 1981

REFERENCES Tsai, N.C., 1969. Transforraation of Time Axes of Accelerograms. ASCE Proceedings, Engineering Mechanics Division, vol. 95, no. EM3.

I I

x-3 j

i e - - - - - , - - , . , . - - . . - , -