ML20066B823

Forwards Results of Evaluation of Structural Adequacy of Hydraulic Control Unit Racks for Seismic Loads,Per SEP Topic III-6, Seismic Design Consideration. Control Rod Drive Hydraulic Control Unit Acceptable for Design Seismic Loads

ML20066B823
Person / Time
Site:	Oyster Creek
Issue date:	10/29/1982
From:	Fiedler P GENERAL PUBLIC UTILITIES CORP.
To:	Crutchfield D Office of Nuclear Reactor Regulation
References
TASK-03-06, TASK-3-6, TASK-RR NUDOCS 8211090202
Download: ML20066B823 (13)

Text

O s l o* ,

GPU Nuclear y g7 P.O. Box 388 Forked River, New Jersey 08731 609-693-6000 Writer's Direct Dial Number:

October 29, 1982 Mr. Dennis M. Crutchfield, Chief Operating Reactors Branch #5 Division of Licansing U.S. Nuclear Regulatory Commission Washington, D.C. 20555

Dear Mr. Crutchfield:

Subj ect: Oyster Creek Nuclear Generating Station Docket No. 50-219 Systematic Evaluation Program (SEP)

Topic III-6, Seismic Design Consideration Attached are the results of our evaluation of the structural adequacy of the Oyster Creek CRD hydraulic control unit racks for the seismic loads resulting from the site specific spectra.

The NRC evaluation conducted previously indicated that the structural integrity of the CRD hydraulic control units is still an open issue due to lack of design information. By letter dated November 24, 1981, GPU transmitted to the NRC our analysis entitled " Evaluation of CRD Hydraulic Control Units" which shows that the units are structurally adequate for SSE loads. Subsequently, in April 1982, the NRC requested additional information regarding ef fects of axial-bending stress interaction and whether the resulting stresses meet ASME Code, Service Level D allowables.

The attached analysis dated May 7,1982, responds to the NRC questions and demonstrates that axial-bending stress interaction ef fects are negligible and that Service Level D limits are met. Accordingly, all outstanding questions on the CRD hydraulic control units are considered resolved.

Very truly yours, A)

• QV)A _^}' '

Peter B. Fiedler Vice President and Director Oyster Creek

..Q PBF:lse Attachments

~

[ F211090202 821029 l _P PDR ADOCK 05000219 PDR GPU Nuclear is a part of the General Public Utikties System

d), , .

. cc: Mr. Ronald C. Haynes, Administrator Region I U.S. Nuclear Regulatory Commission 631 Park Avenue King of Prussia, PA 19406 NRC Resident Inspector Oyster Creek Nuclear Generating Station Forked River, NJ 08731 l

l l

I 1

4

-F

, .~ , - , - -m-_, , --- y-,,-,,m, , . _ , # .. .. - - , -.-., ,,w, ..,

h) 5 May 7, 1982 EVALUATION OF CRD HYDRAULIC CONTROL UNITS OYSTER CREEK NUCLEAR GENERATING STATION

Purpose:

To re-evaluate the structural adequacy of the Oyster Creek CRD hydraulic control unit racks for the seismic loads resulting from the site specific spectra (SSS) .

/

Background:

The seismic adequacy of the CRD control unit racks was previously evaluated in MPR analysis " Evaluation of CRD Hydraulic Control Units" dated September 15, 1981, Refer-ence 1. This analysis was provided to NRC representatives (Mr. T. Cheng and Dr. J. Stevenson) on October 19, 1981.

The results of this analysis indicated that for floor response spectra developed by MPR for the Oyster Creek SSS, the maximum seismic loads result in a bending stress in the tubular support frame of 30,120 psi. This stress is less l than the effective elastic stress of 34,250 psi at the pipe l l limit moment and was therefore considered to be acceptable.

In subsequent communications received from the NRC (Messrs.

Cheng and Stevenson) on April 1 and 19, 1982, the NRC l

l requested that the previous analyses be re-evaluated to l

determine if the " stress in limiting support elements are

' ' [

within ASME Service Condition D stress limits for supports

. {

1 i

i n . ,

wh3n consid3 ring d2cdwaight, exicl-bending intsraction

~

effects and the effects of element curvature". The analyses which follow respond to this request.

l

/

O l

! 1

Evnlention:

1. Seismic Stresses and Moments The evaluation of the Oyster Creek CRD control unit support racks is based on a generic finite element analysis performed by General Electric (GE) in Reference 2. The racks analyzed in Reference 2 have been confirmed by inspection to be iden'tical to those installed at Oyster Creek (see Reference 3) .

/

The significant results of the GE analyses are pre-sented below for the limiting support element - the 1-1/2" Schedule 40 curved pipe elements at the base of the support racks.

Nat'l Peak Bending Axial Earthquake Freq. Accel. Stress Stress Direction (HZ) (gs) (psi) (psi)

Worst Horizontal 2.27 11.5 335,300 3,755 Vertical 23.8 5.3 13,745 660 l

The above elastic stresses calculated separately for norizontal and vertical earthquake components are l

ratio'd to the peak acceleration valve's expected for i

l the Oyster Creek SSE and combined by the square root of the sum of the squares (SRSS) of the vertical and two l

l >

l l

horizontal components. The expected peak accelerations

)

s for the CRD control units are obtained from the ampli-fled floor response spectra curves presented in Reference 4 for the 0.165g SSS, 23' elevation and 7%

damping - the same inputs used in the previous analyses of Reference 1. The peak accelerations and the resulting bending and axial stresses are given below.

I Nat'l Peak Bending Axial Earthquake Freq. Accel. Stress Stress Total Direction (HZ) (gs) (psi) (psi) (psi)

Worst Hori-zontal 2.27 0.72 20,993 235 21,228 vertical 23.8 0.16 2,531 122 2,653 Total SRSS Bending and Axial Stress = 30,140 psi.

This value is within 1% of the maximum bending stress reported in Reference 1 and confirms that deadweight and axial stress effects are negligible. The net moment corresponding to the above stress is 30,140 psi x the section modulus for 1-1/2 Schedule 40 pipe, or 9,826 in-lbs. ,

The fact that the structural element in question is a curved beam can be accounted for by the application of l

l l

1 1

a stress intensification factor applicable to the maximum stre'ss on the concave side of the elbow. This factor is given in Reference 5, Table VII, as 3.I. Factor = 1 + 1.05 I 1 +1 bhI L R-C R-where d = pipe diameter = 1.900" c = pipe radius = 0.950 R = elbow radius A 4.5" I = moment of inertia = 0.31 in 4 For these values, S.I. Factor = 1.096 This indicates that the curvature effects are about 10%

at the worst location on the pipe element. The maximum peak stress is therefore:

Maximum combined peak stress = 1.096 x 30,140 psi, or l 33,033 psi. As indicated above, this stress is essentially all due to bending.

2. A11owables The allowables suggested by the NRC are the Service Level D allowables for component supports given in Subsection NF of the ASME Code,Section III. This subsection provides two alternative acceptance criteria:

i t

R

a. Elastic Analysis - NF-3231.1 permits .

ap' plication of F-1370.of Appendix F of Section III, which in turn refers to Appendix XVil-2000. Specifically, F-1370 states that for Level D loads the allowables given in XVII-2000 for normal loads may be increased by a factor of 1.2Sy/p , where Sy = yield stress = 25000 psi Ft = tensile allowable = 0.60 Sy, but not to exceed 0.7 Su/p , where Su = 45000 psi (from Appendix I of Section III for 24000 psi yield strength carbon steel).  ;

The allowable bending stress is then

1.2 Sy x 0.66 Sy, or 1.32 Sy.

It For the CRD racks, this allowable is 1.32 (25000) = 33000 psi l . This value is greater than the average bend-ing stress of 30,140 psi and.only slightly l

less than the local peak stress of 33,033 psi calculated above for the Oyster Creek racks.

i i Within the normal accuracy for such calcula-l l

b

. . tiens, tha rccks are consid3 red to cost thic

~

codterion.

b. Inelastic Analysis - A limit analysis in accordance with XVII-4000 is also permitted as an alternative to the elastic analysis.

This method requires that the lower bound l l,,.

• i t collapse moment, Mp, be no less tha 1.1 x the applied moment, where Mp = Sy x z x and zx = the plastic section modulus.

For 1-1/2 Schedule 40 pipe, the plastic section modulus is z K = 1.372 x Section Modulus s

= 1.372 x 0.326 in3

= 0.447 in 3.

Then, the lower bound collapse moment is:

Mp = 0.447 x 25000 = 11,180 in-lbs.

This value is greater than 1.1 x the applied moment, which is:

1.1 x 9826 in-lbs, or 10,809 in-lbs.

Thus the alternative, limit analysis criteria l for Service Level D loads are met.

1 l

l l

1

c. Buckling Infaddition to the above, the Code requires that the critical buckling load for beam sec-tions be at least 1.5 x the applied load.

From Reference 5, Table XVI, the critical buckling moment for a thin tube in bending is Mc 1 0.72 E rt 2, 1-v 2 where f E = 28 x 10 6 p,i v = 0.3 r = tube radius - 0.95 inch t = tube wall thickness = 0.145 inch.

Then, Mc > 440,000 in-lbs.

This moment is substantially in excess of 1.5 times the applied moment of 9826 in-lbs.

Therefore buckling is not a problem.

l

) . - - - - - - _ __ -_ _

f '

Summary of Re ulte REsults of analy$'es presented herein show that the stresses and moments in the limiting elements of the Oyster Creek CRD control unit support frames essentially meet the Service Level D allowables of the ASME Code for component supports for the 0.1659 site specific spectra and the interim floor response spectra given in Reference 4.

In addition, new floor response spectra have recently been I

generated by URS/Blume and Associates using a conventional time-history analysis method for the 0.165g site spectra.

These spectra are presented in Reference 6. These floor response spectra for the 23' elevation of the Oyster Creek reactor building have lower accelerations at the fundamental frequencies of the CRD control unit racks than those calcu-lated in Reference 4 and used in the above analyses. Spe-cifically, the comparable peak horizontal acceleration (at 2.27 Hz) is reduced from 0.72g to 0.589; the peak vertical acceleration (at 23.8 Hz) is reduced from 0.16g to 0.149 The net result of these reductions is to reduce the calcu-lated stresses and moments by about 19%. This reduces the highest calculated peak stress in the limiting element to approximately 26,760 psi, which is well within the clastic allowable of 33,000 psi. Similarly, 1.1 x the applied moment is reduced from 10,809 in-lbs to 8755 in-lbs which is well below the lower bound limit moment of 11,180 in-lbs.

l

BIScd Cn tha rG-cnolyoss prGsGnted haroin cnd tha Edditional margin resulting drom the use of the latest plant specific floor response spectra, the Oyster Creek CRD hydraulic control units are considered acceptable for the design -

seismic loads. ,

/

e l

~

l l

7

.c , Roforencns

1. MPR Analysis'" Evaluation of CRD Hydraulic Control Units" dated September 15, 1981. .

2. General Electric Report' 383HA853, " Hydraulic Control Unit, Seismic Analysis of, dated November,;1972.

3. MPR letter to GPUN dated November 30; 1981.

4. MPR-6h'1,"InterimFloor.ResponseSpectrafortheOyster Creek Reactor Building" dated October 15, 1981.

5 .Roark, R.J., Formulas for Stress and Strain,

+., McGraw-Hill. .

6. URS/J.A. Blume and Associates Report, " Seismic Acceleration Floor Response Spectra for the Reactor Building at Oyster Creek Nuclear Power Plant" dated December 1981. y I 4

\

s l

. - . - _ . ._._ - .