ML18023A881

Letter Transmitting Report on Deficiency in Box Column Supports for the Susquehanna Nuclear Plant.

ML18023A881
Person / Time
Site:	Susquehanna
Issue date:	08/29/1975
From:	Curtis N Pennsylvania Power & Light Co
To:	O'Reilly J NRC Region 1
References
PLA-82
Download: ML18023A881 (40)

Text

TWO NORTH NINTH STREET, AL LENTOWN, PA. 18101 PHONE~ (215) S21 5151 August 29, 1975'r.

J. P. O'Reilly Director - Region I U.S. Nuclear Regulatory Commission 931 Park Avenue King of Prussia, Pennsylvania 19406 SUSQUEHANNA STEAM ELECTRIC STATION DEFXNITXVE REPORT OF DEFICIENCIES IN BOX COLUMN SUPPORTS DOCKET NOS.: 50-387 and 50-388 LICENSE NOS.: CPPR-101 and CPPR-102 ERs 100450/100508 FILE 840-4 PLA-82

Dear Mr. 0'keilly:

This letter transmits the sub)ect report which supplements PLA-74, dated July 25, 19JI'5 and is intended to complete PPEL's responsibility in reporting the deficiencies in fabricated structural steel members (Box Column Supports) which had been located in position for installa-tion in Susquehanna Unit 1 k 2 structures.

As indicated in the attached report, the decision has been made to redesign the structures and to replace all eight units, including those found not to contain defects. The replacement units employ

'aterial changes and weld )oint configurations which will facilitate fabrication and should preclude the recurrence of the lamellar tearing phenomenon. experienced in original units.

Xn eddition to those corrective measures identified in PLA-74 and. the attached Bechtel detailed report, PPRL has concurred in an increase in Bechtel surveillance activities at the vendor's fabrication facility (Bethlehem Steel Company's Pottstown, PA plant).

Additionally, receiving and preinstallation activities at the SSES ite will be aligned to perform a final visual examination before the new units are permanently committed. to the plant structure.

TATe await your advice should you have any further questions or concerns on this matter.

Very truly yours lh,&

N. W. Curtis

'Vice President - Engineering 5 Construction S400 ARS:b PENNSYLVANIA POWER 4 L I GH T CO)APANY

TWO NORTH NINTH STREET, ALLENTOWN, PA. 18101 PHONEME {215) 821 ~ 5151 Mr. J. P. O'Reilly PLA-82 Pa e 2 Sworn to and. subscribed. before me this l975 Notary Public My commission expires:

cc: Mr. Donald F. Knuth Director Office of Inspection and Enforcement U.S. Nuclear Regulatory Commission Washington, D.C. 20555 PENNSYLVANIA POWER 8 L I GH T COMPANY

REPORT ON DEFICIENCY IN BOX COLUMN SUPPORTS FOR THE SUSQUEHANNA NUCLEAR PLANT

~@cd)y~

4uap g )g75

~QUEI NNg p~

Prepared by  : G. Shah Checked by  : J. N. Mulay Approved by  : M. J. Lidl BECHTEL POWER CORPORATXON San Francisco California (B-2la)

TABLE OF CONTENTS 1.0 Introduction 2.0 Description of Deficiency 2.1 Box Column Base Description 2.2 Description of First Crack 2.3 Visual and NDE Inspection 3.0 Investigation 3.1 NDE Shop 3.2 NDE Field 3.3 Analytical Investigations 3.4 Lamellar Tearing 3.5 Conclusion 4.0 Remedial Measures Appendices:

A. Design Criteria B. /DE Reports C. Design Considerations Figures 1 Original Column-Base Detail 2 Redesigned Column-Base Detail

1.0 Introduction This report is prepared in accordance with the requirements described in Title 10, Code of Federal Regulations, Part 50, Section 50.55(e).

Hairline cracks were found in major weldments at the base of box columns. The significance of these linear indications has been evaluated and it is concluded that the box column bases which were found to be defective, represent a deficiency in construction (fabrication) which will require extensive repair or rede-sign and replacement, and which, were the defi-ciencies to have remained uncorrecrted, could have an adverse effect on the structure in the performance of the intended safety function.

2.0 Description of Deficiency 2.1 Box Column Base Description Four (4) box colums are provided in the Reactor Building of each unit.

These box columns support the con-crete girders which form the walls of the spent fuel pool and dryer and separator pool. They also support the intermediate floors. Figure 1 shows the details of the box column and its base.

~ As shown in Figure 1 the column is 3'-0" square and made of four steel plates welded together with partial penetration welds. Top.of the column is rigidly connected to the steel box girder which is provided below the con-crete fuel pool girders.

At the bottom, stiffeners and base plates are welded to the column to transmit the design loads. As shown in Figure 1, the base plate stiffeners are welded to the column with full penetration butt welds, while fillet weld is used to con-nect them with the base plate. Nominal fillet weld is provided between the base plate and the column. In order to re-duce the base plate thickness, a system (B-2la)

I page 2 of main and secondary stiffener is provided.

The design criteria for box column base is discussed in Appendix A of this report.

2.2 Description of First Crack During the course of random visual inspec-tion of installed structures of Unit l, PL's Quality Assurance engineers noted the appearance of welds on four (4) box column bases and questioned their compliance with ANS Dl.l. PL Deficiency Report No. 0022 was issued to document the condition.

Subsequently, closer inspection revealed the existence of a visible hairline crack at the toe of a full penetration weld between a stif-fener and the box column S-27.5. Bechtel NCR (Nonconformance Report) 376 was issued request-ing further non-destructive examination. Xt was decided to carry out the ultrasonic exami-nation (UT) to establish the extent of the crack+

The UT examination was carried out by Bechtel's NDE Subcontractor, Peabody Testing Company. The results of UT examination revealed that the crack extended into the base metal. The crack was ap-proximately 37" long and located at approximately 1-3/8" from the face of the column. Appendix B shows the details of UT examination.

2.3 Visual and NDE Xnspection Further visual examinations of welds revealed a similar subject condition on box column Support Q-36. The UT examination revealed a- crack varied from surface to l-l/'4" depth and was approximately 27" long. NCR 377 documented the defective condi-tion of column base Q-36.

Since identical design details were used on all box columns, tion on all it was decided remaining seven to perform UT examina-columns. Bethlehem Steel Corporation, Pottstown Plant, carried out the task. The UT examination was carried out on all the columns in the field. \

page 3 Four additional column bases Q-22, Q-36, S-30.5-

~

and S-36 were found to have additional indica-tions (interpreted subsequently as lamellar tear-ing) in the base material of the box column. Two additional NCRs 401 and 402 were issued to do-cument these defects.

Appendix B describes the details of UT. NDE of the columns indicated that three (3) columns did not show any defects. Before these defects were observed, small cranes had been installed on top of erected columns for lifting reinforcing bars.

The loads applied to these columns through the cranes are considerably smaller than the inservice column loads.

3.0 Investigations 3.1 NDE Shop The full penetration butt welds joining the stiffeners to the box columns were 1008 UT examined in the shop of Bethlehem Steel Corporation prior to shipment. All fillet-welds were 100% examined by Magnetic Particle.

Testing (MT). The column bases were also post

-weld heat treated as specified, after fabrica-tion in the shop.

3.2 NDE Field The UT exami'nation that was performed'in the, field has been described in detail under Sec-tion 2.3 of this report.

The limited amount of magnetic particle (MT) testing was also performed for fillet welds joining the stiffeners to the base plates.

HT examination was performed since analytical investigations described in Section 3.3 below indicated if one of the two welds (; tiffener to box column or stiffener to base j1ate) is adequate, the load can be transferred satis-factorily.

3.3 Analytical Investigations

'nalytical investigation was made to check the design loads could be adequately transfer-if red to the foundation fro'm the column even

Page 4 with the observed linear indications in the base metal. The following are the analytical considerations:

(a) Column base will be subjected to axial loads only.

(b) The column stiffeners where linear indications were found, would be subjected to compressive and shear stresses from the base plate loading.

(c) Compressive stresses can be transferred through the ade-'uately cracks.

(d) It is questionable ses can be adequately if shear stres-transferred from stiffener to the column un-der the existing condition. However, the base plate along with stiffeners would be able to transfer shear forces.

From the above mentioned analytical considera-tions, it was concluded that the box column base would adequately transfer the load with the existing =linear indications provided these indications would not propagate.

3.4 Lamellar Tearing Bechtel's metallurgical staff was further'consulted to identify the observed linear indications and to determine if these indications would propagate through the application of external design loads.

Metallurgical staff indicated that the observed de-fects were classical examples of lamellar tearings.

These are caused by the shrinkage of we. 3 metal which induced stresses in the base meta,',. in a direc-tion transverse to the material "grain"." Lamellar tearing phenomenon has received attention in the industry fairly recently, and therefore, several as-pects of this phenomenon are not yet fully defined and understood.

Further evaluation by metallurgists from Bechtel and Bethlehem Steel Corporation indicated that (B-21a)

8I page 5 ost robably the lamellar tears would not pro.

pagate any further since the s tress due to weld hrinkage is relieved due to 1am ellar tears. T h'xs conclusion is again augmented by the fact that rect tensile load will not be applied to the af-fected area.

3.5 Conclusion Since it xsis nonot pos possible to establish conclusively that the lamellar tears wall not p ro p a g ate under service loa d s, wee have ave to conclude that the observe defects are reportable. If lamellar tearss 'wou 'would have propaga t e d ex extensively, then this deficiency might have adversely affected the structura ral inte-g-

rity of the plant.

4.0 Remedial Measures Even though on a purely theoretical bass.s, xt could be demonstrate d thata thee column bases would function ade-uatel , it was considered prudent to repa'ir or re 'ect the existing column 1 s. Because b ases. e of the indeterminate nature oof thee lamellar tearing phenomenon andd tthee in-herent risks in attempting repairs, s it was decided to replace all eight (8) column support bases with new sup-ports having a different design detail.

the column base considerably reduces the possibility of lamellar tear s.. App endix C of this report out-lines the details of redesign of the base connec ion.

shows the revised column base details.

The rationale of redesigned base can be summarize 'zed as follows:

{a)) Th e we ld detail for the revised detail i.", such that the possibility of lamellar tears hav e been greatly reduced.

(b) Neither base plate nor box column is subjected to thorough thickness loading during application of design loads.

(c) .The redesigned base of box column will adequately transfer the design loads during service life of the plant.

(B-2la)

DESIGN CRITERIA l.0 Loads Box columns are designed to support following loads:

(a) During construction Box column supports the crane load, load of intermediate foaming and of the wet concrete of the fuel pool concrete girders.

(b) During plant operation.

Box column is designed to transfer loads of intermediate floors and fuel pool concrete girders'

' Loading Combinations and Allowable Stresses:

Th c d e t ai'l e d design criteria is outlinedll in Appendixx "C" "Structural Design and Loading Criteria of Prelim'inar Safety Analysis Report (B-2la)

APPENDIX B ULTRASONIC EXAl1INATINONS

J s l I

I

'a~

ll g h1 f 4t

4 V

1' ij N

l

i~ ~ ~

l S. l

~ ~

C e J

~ P

0 dl e N h

1I tl

'W

~

~

I ~

~ I

~ ~

0 I

J I

r S

b

~ ~.

h \ 4

~ 4 44 4 /feet' \ p g S

~ 4 ) e b

' 4

~ le bb ft04 II 4 ~ ~

l 44l r <<'4 4 4 err 4. 444 x 4 4 ~ b 4

44 4

lt

~ ll 4

e 4 4 4 4 ~

I

~

4 ~ 'e

~~

4 b ~

4

~ 4 '4, ~ 44 Il ~ 4 ~

~ g * '

4 44 ~ 4 4

P'

~ ra

~ ' ~

4 '

4 e 'ee' e'4

~ 4,

I 1

~ ~

~ ~ 4 ~

Ib

~

p

~ ~ ~

~ ~

~

4

' ~

P P

0 kl

~ b I'

~ ~

/I 4

~ ~

~

~

~ ~

'0 g 0

~ ~ e 4 4

I"

'I 0

I h h Ei y ~ 00 00 ~ 0 C

0

~ 0 0

0

~

0 N ~ II ~

~ 0 h 0

Cl

~~

~ ~

0

~ ~

4 4 le N. N 1

~ N N

1 0

~ IE ~ Cl 0 0 ~ N

'I I N 0 0

I ~

4 0

~ ' 0 01 ~ ~ 4 4 0 0 ~

~ 0 0 0~

0

~ 0 ~ ~

~ I NN

~I 1

0

~ '0 4

~ E ~ 4 N

~ 'I '4 ~ ~ '0 ~

7 0 44 \ ~ I \ ~0 e

~, ~

0 0

0 4

Ch

'0 r NN* ~

Cl

~ ~

~

~ ~ ~

i k

~ ~

I I

It I V P I

V h

I II I W it t V

W t

II I

I

s

~

~

C ll a

~ a

~ I I ~

4 l

J P

e T

I V

II I

V I

tl V

li'

'V

~

'h

~ ~

~

~ t ~

g

~ S ~ ~

APPENDIX C DESIGN CONSIDERATIONS

U J \I L

UI II t ~

~ ~

~ ~

~ Ul 4 ~

g I ~ ~

~ ~

. Jl 1.0 Original- Column Base I

I

-Following assumptions were made in the design:

(a) 'he box column was assumed to be a continuous member,'ith a pinned base and supported laterally at intermediate levels. Therefore, the column base was designed mainly for axial loads.

(b) Because of the rigid connection with box girders at the top, moments are transmit-

'ted to the column top depending on the re-lative stiffenesses of the column and the girder.

(c) Column load is distributed uniformly over the entire area of the base plate.

(d) "

Stiffeners and base plate act as a composite section for transmitting design loads..

2.0 Redesigned Column Base The 'maj'or change made in the redesign of box column base is elimination'f stiffeners for the base plate.

The base plate utilizes'the original erection bolts as shown in Figure l.

(B-21a)

FIGURE 1 ORIGINAL COLUMN BASE DETAIL

"P~% 4 tt I 441 t~

'0 II

'0 0 0

/p S//FF A'S ~

2'-// /O g COB.

l* 1 0 4h'Cb'OjP /ALT l

DETAIL'

~ / 0 P

3' 3 O 3 L>

PEAN SVR ABOu7 80X'Nu,yA I I( I. I II II 1 ns// ~

lf I) I/ I zd 1 II I I

II. I II I

yP

~ a q q t'HOOF

~

,' Ig C2~c Sf/FF +

~rk /rp/ 4

/ STIPE /P.

'p.

(3 ~ Gr.'o//I'/<

BrtS/'

~

~~EC7/C'O/

/g -./ 0 r

FIGURE, I ORI GINAL COLUMN BASE DETAILS

FIGURE ?

REDESIGNED COLUMN BASE DETAILS

~

t II D

Se

\

0 I

A LT II 4h

'ILP 41 eb g r ~ C 0