ML20084N219

Summarizes Incident Rept,Turkey Point Unit 3,Valve Headers. Rept DC-94A Being Written by Parameter,Inc Re Header Failure Investigation

ML20084N219
Person / Time
Site:	Turkey Point, 05000000
Issue date:	03/03/1973
From:	Lofy R PARAMETER, INC.
To:	Beratan L US ATOMIC ENERGY COMMISSION (AEC)
Shared Package
ML20084N179	List: ML20084N182 ML20084N198 ML20084N212 ML20084N219 ML20084N247 ML20084N251 ML20084N422
References
720303, NUDOCS 8306020664
Download: ML20084N219 (3)

Text

,,,,,. _._-..

_. - _ - . . . . .-. .- . _ _ . ~ - _ _. _ - . - . - __ -_

g . #,. . . . , _ _

q c yg J

l

] '

March a 3, 1972, 1

l -

j i 1

l l ,

LAan Beratso, Teeknical Support Branch, CD i \

i REVIEW OF " INCIDENT REPorr, TURKEY POINT UNIT 3, . VALVE HEADERS," DhJED -

i DECEDBER,1971,= BY FIARIDA POWER AND LIGHT C0le'ANY AND EVALUATION OF l REVISED HEADER DESIGN.

T l

Referescar AEC Contract (AT-11-1)-1658-A PAR: 11-72-a

. Assignment No. DC-94 l Introduction 1

f A report .on the. above seject DC.-94-A, is being prepared by Parameters, j Inc. to supplement the earlier preliminary . report AEC--94 which dealt with the initial Header Failure-investigation. It will be submitted to the

. Division of Compliance about March 7. . This memorandum is' intended os an advance summary of the' salient determination of the report. They are.

. listed below essentially as will be contained in the summary of findings section of the report. h. evaluation of the metallurgical studies con-

, ducted by Battelle-Columbus, Appendix 5-A, will not be included in Para-mun' apn.

• l , 1. The Florida Power and Light Incident Report does not provide enough detailed information for a thorough independent' review of thq stresses t reported for the original or revised header designs. To provide a i relative assessment of both ' designs. Attachments 1, 2 and 3, to j Report DO-94a,' in which ~ arbitrary loading conditions have been used

, for comparison purposes, have been prepared. These analyses also l provide a basis for evaluating the relative strength of the' main i

steamline with the header as the branch connection. . The findings 2, j ,

3, 4 and 5 below result from the analyses.

, 2. The maximum principle stress determined by the analysis of Atta' chment 'l l was found to be 38% higher than the 145,000 psi reported in Table

! 4.3-2 of the FPL incident report. . This stress occurs at the' point of

failure.

! 3. Assuming an exhaust stack configuration identical to the original design (24 inches offset), the revised header asseshlies would meet allowable stress limits for the discharge condition as evaluated by ASME 'Section III methods using Section VIII allowable stresses. It i appears ' that the horizontal extension and outboard supports for the

exhaust pipe would not have been needed on this basis. Such a design ff) t could not be proved out however, on the basis of Section I or Section i

ZII - 15 .S' . ,

M. ' ..

ann 8306020664 730604 )

1 PDR S ADOCK 05000250

g. PDR k om > L. .....t...... .

Form AEC-Ste (Rev.9 $3) AECM 0240 o u s covenwest mumm orres 1,7, .44e . se i

n,,,-,e.,-- ,---,-, ,,.,... - - - -.. , , -- - . ,~.---,..,,..-,,..n,. v. . , - -- , , , , , - - - - n,,,,,re,- -,n,-,,-r-,-,,

r- -~ ~ 2  %. ..

' "' ~ ~ " ~ ~ ' ~- --

I 3

J

4. Upea diseharge, the safety valves will experience a momentarily i

abalamond horisontal reaction before the steam reaches the vertical stack. The magnitude of this reaction is unknown. In the analysis of Attachment 2, stresses at the header shell of the revised desip were calculated for the horizontal reaction load of 20,000 pounds.

Stresses are within allorables when evaluated by Section III methods using Section VIII allowables. It should be noted that the revised desip is restrained from horisontal translation only by the header to sta==14== attachment for this condition.

5. Attachment 3 to report DC-94a smalyses the stresses in the main steamline and at the header to weldolet juncture for the reaction described in finding number 4 above. Both valves in one heeder branch are assumed to discharge siamitaneously. Stresses are within

• limits as evaluated by Section III methods using Section VIII and B31.1 allowables.

6. The incident report (Section 3.2) records the discharge of one or l

possibly two safety valves in the 3C steam header on November 23, 1971. It is probable that the very high stresses that have been established to be associated with valve discharge, resulted in some permanent deformation or material damsp to the header. The results of detailed nondestreetive emandnations and dimensional measurements of this header would have been an appropriate matter for presentation in FPL's incident report.

7. The safety valve manufacturer (Dresser-Consolidated) reeemmends that a horisontal offset of no more than 24 inches be used in attaching a vertical discharge pipe, the prima consideration being stresses is .

the valve neck under the discharge condition. The revised desip of the ezhaust pipe for Turkey Point 3 has a five foot horisontal run which is externally supported to prevent the discharge reaction fres acting on the valve body and header. - This extra pipe length could, however, affect the capacity of the safety valve. The valve annu-facturer's approval of the installation should be recorded as part of the redesign documentation.

8. The FPL report Section (4.3) indicates that the original header support was desiped for groes sta==1ine motion under normal plat operating conditions. The desip did not provide for the motion realised during hot functional tests when the main steam isolation valves were closed. In addition to a statement that the now desip accepts stemm11pe motion during all modes of operation, it would be logical to also provide assurance that the sta==14=am can accept the relative motion realised when one or another loop is inoperative.

omes > . . .

W SullteAME >

. DATE > ... . . . . . . . . . . . . . . . . . . . _ - . _ _

Form AEC-St8 (Rev.9-53) AECM 0240

• u s coVERNMENT PmNTaNG OFFICE 1971.416-468

. PT" M '["

i DLk .

4 b -

• P " Y T'< -M'N' 'E"MT> # '## "

.,:_. . r y

(. 1

) 1 3-i t

9. In view of the ' fact that the design analysis work on the original and

,' revis,ed header configuration was not attached to the FPL report, 4

it would have been appropriate to identify the authors and organi-sations (presismsbly Boehtel) who performed the engineering analysis of the original and revised designs and to reference the report which contained detailed work.

10. The FPL incident report essentially agrees with the findings contained DC-94a the Parameter, Inc., preliminary report to the Division of Compliance.

11. A more meaningful review of the analysis of the incident and header redesign'in terms of applicable codes and design bases could be made

' if the detailed engineering works performed for FPL were available.

It is not expected that such a review would change the above stated findings appreciably.

l Richard A. Lofy Parameter, Inc.

Elk Grove, Wiscemain bec: J.B.Henderson, CO W.J. Collins, CO ~

L.L.Beratan, CO W.C.Seidle, 00:II l

h l

,. omer > -

SumeAME > _ . . . . --

f g., DATE > . ..

) Form AEC-Ste (Rev. 9-53) AECM 0240 o u s cove =NueNT mNvuda OMCs 1971 416-.468 l

I l

. _ . . -_.