ML18227D283

Response to Additional Questions for Supplemental Information Regarding Steam Generator Tube Integrity

ML18227D283
Person / Time
Site:	Turkey Point
Issue date:	02/02/1977
From:	Robert E. Uhrig Florida Power & Light Co
To:	Lear G Office of Nuclear Reactor Regulation
References
L-77-40
Download: ML18227D283 (18)

Text

NRC FORIs 195 I2 76) I NRC DISTRIBUTION FQR PART 50 DOCKET MATERIAL U.S. NUCLEAR REGULATORY ~MISSION DOCKET NVMBE 5O-2 FILE NUMBER

/251 TO:.- 'FROM: DATE OF DOCUMENT Florida Power & Light Company 2/2/77 Mr;"George Lear r

Miami, Florida DATE RECEIVED

~ t r r Mr, Robert E. Uhri 2/7/77 fggCl'TE R 'CINOTORIZED PROP (NPUT FORM NUMBER OF COPIES RECEIVED JQ;0 R I G INA L /KUNG LASS IF I E D Three sighed

,-'. ~ Ht:otv. 37. co ~es'ncl.;recvd.

DESCRIPTION ENCLOSURE

~~ ~

h Ltr..re our 1/25/77 ltr. and their 1/21/77 r ltr. ~ . ~ trans the following: Consists . of supplemental information '

'oncerning ~S earn Geperator Tube Integrity.. ~

ZCKXOVrLEDGED PLANT NAME: (1-P) ,((9-P)

Turkey Point Units t

3 & 4 Dp Npp ggjgpvg SAFETY FOR ACTION/INFORMATION 2 7 77 RJL

" ASSIGNED AD:

gg . ~-Lear C MANA v, Elliott . ~ PROJECT MANAGER L C ASS . Parrish- LIC ASST INTERNAL DISTRIBUTION

,REG FXLE SYSTEMS SAFETY PLANT SYSTEMS S TE SAFE

~g

HEINEMAN SCHROEDER TEDESCO QN'0 OELD GOSSXCK & STAFF --ENGXNEERING IPPOLXTO ..- .

. MIPC ERNST CASE KNIGHT BALLARD HANAUER SI1WEIL OPERATXNG REACTORS SPANGLER HARLESS PAVLXCK STELLO SITE TECH PROJECT MANAGEMENT REACTOR. SAFE OPFRATING EXSENHUT.

CH GA19IILL STEPP BOYD . ROSS P ~ COLLINS 0 HULMAN HOUSTON ROSZTOCZY PFTERSON CHECK BUTLER SITE ANALYSIS MELTZ VOLLMER HELTEMES AT&X BUNCH SKOVIIOLT SALTZMAN- J ~ COLLINS RUTBFRG KREGER

. EXTERNAL DISTRIBUTION CO/ OL NUMBER LPDR'IC:

NAT~ LAB'EG MOKIAYEKM V.IE ULR KSON OR NSIC: LA PDR ASLB: CONSULTANTS:

ACRS CYS WeueNe/ E

FLORIDA POWER & LIGHT COMPANY February 2, 1977 L-77-40 Office of Nuclear Reactor Regulation Attention: Mr. George Lear, Chief 9> Lgo Operating Reactors Branch I3 g(cU~(0 Division of Operating Reactors S0jt U. S. Nuclear .Regulatory Commission <

Washington, D. C. 20555 u,S.iiiIP~~ISSI0ii ~,.

I ~isa CO

Dear Mr. Lear:

'4 Re: Turkey Point Units 3 and 4 Docket Nos. 50-250 and 50-251 Steam Generator Tube Integrity Su lemental Information Your letter of January 25, 1977 formally requested additional information regarding steam generator tube integrity at Turkey Point. All of the questions in your letter have been answered by our submittals L-76-432, L-76-434, L-77-3,', and,L-77-30.

Subsequent to your January,25 letter, we received two additional f

questions from your'taff. The answers to the two additional questions are attached.

As stated in our letter L-77-29 of January 21, 1977, Florida Power & Light Company has submitted considerable information on the subject of steam generator tube integrity as it to Turkey Point Units 3 and 4. Based on the information we.

applies have submitted through today, we request that you grant approval for continued operation of Turkey Point Unit 4 beyond the time limit specified in Amendment 20 to Operating License DPR-41.

Based on our operating history from December 3, 1976 to date, and operational considerations, we woul approval by 5:00 p.m. on February 8,

'o and assuming continued operation at 100% rated power, we will reach the limit on February 9, 1977. Because of load management have your Very truly yours, Robert E. Uhrig Vice President REU/MAS/cpc Attachment 8t C'P//8 g5 6

'~op, QC786 Dg+

cc: Mr. Norman C. Moseley, Region Robert Lowenstein, Esquire II PEOPLE... SERVING PEOPLE

E ATTACHMENT Turkey Point Units 3 a 4 Steam Generator Tube Integrity Su elemental Information Sup ort Plate Ex ansion a.) We have correlated support plate expansion vith actual months of operati by means of a finite element model which utilizes a pseudo-thermal expansion technique. This technique has been used to simulate plate.

behavior up to full closure and several months beyond. The expansion has been correlated with actual months of operation. In order to do this, relationships between field data, effective months (EM's) and, results of the finite element anlysis must be established. Since the denting phenomenon .extends over the entire plate, there's good correl-atipn between measured denting and expansion of a dented plate.

Although our finite element -model .is not detailed. enough to yield dent ing rates, it does quantify the extent of flow slot closure for a pre-scribed expansion rate. Since the amount of closure over an extended period of EM's is available from field data, a relationship between model closure and EM's can be established. The rate of expansion-is'(

independent of boundary effects, insertion of blocking devices, and tim The procedure for calculating the rate of expansion per EM-is as follow For a plate expansion of . 014 (hot leg)/. 010 (cold leg) in/in applied to the updated plate model (Figure 1)

• the average flow slot. closure is .675 inches.

• The model shown in Figure 1 is for a 51 top support plate. Several comments are necessary regarding applicability of this model to a 44 bottom support plate. With regard to the differences between a 44 and 51 plate all significant parameters are approximately 5% smaller for a 44 than a 51. Thus, the two plates are virtually identical with regard to relative parameters and. 5% different. in overall size. The effect. of this on our results would be insignificant. The bottom plate is suppor differently than the top plate in both units However, the difference can be accounted for by transposing support locations about the diameter perpendicular to the tube lane. This can be accomplished simply by flipping the results of the finite element, number the flow slots in the model as 1-6 analysis. That is, from left to right for if we the to plate, they become 6-1 for the bottom plate.

i I

~ ~

For the actual plate the maximum (most conservative) rate of closure for a bottom support plate flow slot is

.24 inches/EM Thus .675 inches of closure represents 2.8 EM's and the plate expansion equivalent of a single month is

. 014 = ..005 in/in on the hot leg side and 2.8

. 010 2.8 .0036 in/in on the cold. leg side The .hot and cold leg expansions will be denoted as follows:

.005/.0036 in/in Based on field data taken during the last outage, the average opening was 1.5". Using a rate of closure of .24 in/month, the expected open-ing at the end of the current 2 month period is 1.0" This is equivale to a 1.75" average closure. Figure 2 shows the strain intensity plots for, the plate at. an average closure of 1.75", and a pseudo-thermal expa sion of .041/.029 in/in.

The rate of plate expansion occurs on a'ery local level and will not-change after complete closure of the flow slots. Thus, for each addi-tional month of operation beyond closure, an additional plate expansion of

.0050/.0036 in/in should be utilized.

Our results show closure occurring at

.064/.046 in/im expansion.

Thus, based on the expansion rate of

.005/.0036 in/in/month the number of operating months between the current: closure (at two mont subsequent to the December start-up) and. full closure is 4.6 zaonths.

More conservatively, we can consider the worst, slot which had an -openin 7/8" at the last inspection as being representative of all slots. The expected time to full closure for that slot, is

~ .

(. 875 2 (. 24)) l. 24 = l. 6 EM's beyond the end of the current two effective month period.

Figure 3 shows the strain intensity plots for the plate with an expansio of

".08l/.058 in/in or 3.4 EM's after closure.

The two plots indicate that while expansion continues, the continued, operation to closure, then more than 3 EM's beyond that, does not alter the strain intensity patterns. That is, the areas .of excessive hard spots are currently wel3. bounded, and will remain so for more than months a'dditional operation.

NOTE: An "effective month" is defined as operation of the reactor coolant system above 350 F.

Tube Loadin j 1n order to quantify the additional load on the tubes, due to the possible tendency of the plate to buckle after full closure, j.t j.s necessary to determine the load required. to prevent buckling of support plate. He have combined two analyses from Timoshenko's "Theory o Elastic Stability" to arrive at a very conservative load,,

which is still quite small.

Consider the case which generates the highest in-place load before buckling occurs, and ther'efore would require the largest transverse load to prevent buckling. This is the case of a circular plate with clamped edges. From Timoshenko, page 390:

(N ),14. 68 D a2 where E*h3 D

12 (1-v* )

v* = .42@

h = 75'I (Nr)c = 1215 lbs/in-Now let us consider a unit strip with clamped edges of length 2a m2D

~cr (Timoshenko, page 390) 817 lbs.

4 0 ~

For the case of where-a beam on an

~

Ncr = g2 E*

L elastic foundation (Timoshenko, page 94)-

I= 1/12 bh = 1/12 (1) (.75) = .035 in for a 1" strip of plate.

L = "reduced" length of the strip or beam as described on pages 96 and 97 of Timoshenko.

7f 2 E*

Ncr The'ctual length in question is R = 2a = 116" and L/R = .51 is the ratio of the "reduced" or equivalent beam length to the The reduction is due to the effect of the transverse support actual'ength.

on the beam, in our case created by the tubes. Table 2-5 in Timoshenko relates this ratio to the force per unit. length of the beam for a unit deflection 8. The stiffness of the tubes is such that the most conser-vative evaluation of tube loads would not. cause axial deformation of Thus, assuming a 1" deflection in order to calculate the reactive fore'e is conservative. From Table 2-5 for L/R = .51:

4R /16 E* I = 18.5 P

18.5 x 16 x 7.8 x 106 x 035 1164

.45 lbs per linear inch for a 1" strip or

.45 lb/in2 For the whole plate this translates into 4755 lbs.

Xf we add a further conservatism and assume that only 100 tubes take all of the load, the load per tube is 48 lbs/tube.

However, let us go one step further in the direction of: conservatism an use a buckling load of 1215 pounds from the circular plate results, but use the strip model to obtain 4.

1 L2 >2 x;26 x 30 x 106 x 035 1215 L 47 1" L/R = .42 50 18.5 x .45

= 1.22 lbs/in 2

~

and the total load on the tube bundle is 12,850 pounds.

Again, assuming that only 100 tubes interact, the load per tube is 129 lbs/tube This translates into a stress of 129/p ( 43752 38752) or E

1,000 psi

1 Hadst Characteris ties

l. Hot/Cold Side Expansion Bias

2. Improved Perforated Area ~aterial Behavior (Anisotropic)

3. Elastic Behavior oir Channels at Support Locations

4. scrapper Stiffness Incorporated at Periphery

~

5 60.

'OT LEG 30, l5, COLD LEG I ~

~

, I

-'7'i. 0 X

=GO. 0 ~'a. 0=3K 0 a.O Figure 1. Finite Element l!odel.

CONTOURS REPRESENT  % OF i~fAXINEf STRAIN INTENSITY:

1-10%

'2-50%

3-60%

4-70%

5-80%

.041/.029 in/in Expansion Load 0 03 0

Figure 2. In-Plane Strain Intensity At 1.75" Closure.

CONTOURS REPRESENT  % OF i&GH~aQP;1 STRAIN INTENSITY:

1-10%

'-50%

3-60%

4-70%

5-80%

.081/.058 in/in Expansion Load G >4 Figure 3. In-Plane Str in Intensity 3 ~~months After. Closure.

~ I