ML19224D401
| ML19224D401 | |
| Person / Time | |
|---|---|
| Site: | Brunswick  |
| Issue date: | 06/07/1979 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19224D402 | List: |
| References | |
| NUDOCS 7907120058 | |
| Download: ML19224D401 (3) | |
Text
.
BRUNSWICK STEAM ELECTRIC PLANT, UNITS 1 AND 2 REVIEW OF PIPING REANALYSIS PER IsE BULLETIN 79-07 SAFETY EVALUATION REPORT INTRODLCTION/ BACKGROUND In their April 24, 1979 response to I&E Bulletin 79-07 Carolina Power and Light Company (CP&L) stated that the recircu.lation and main steam piping nad been analyzed by GE using a computer code that comoined directional seismic responses by algebraic summation. All other safety related piping was analyzed by UE6C using a computer code that combined directional seismic responses oy algebraic summation. CPaL has supplied supplemental information on this suDject at meetings with the NRC staff and in letters dated May 15, 21, 22, 29, and June 4,1979.
DISCUSSION CP6L has stated that a reanalysis of all af fected piping in the "as-uuilt" condition will be completed with the results showing all piping stresses remaining below their allcwable values, as specified in the BSEP FSAR Dy July 21,1979. Additionally, all loads on attachea equipment (nozzle loads) will De acceptable.
Upon completicn of modifications to certain pipe supports whicn were determined to De originally underdesigned, all pipe supports attacned to safety related piping or equipment in the plant will De "operacle" and within FSAR criteri a.
These modifications will be completed prior to return to power operation.
The recirculation and mainsteam lines were reanalyzed oy GE using PISYS. The responses f rom two directions, the most disadvan' ageous comoination of one horizontal with the vertical, were comoined oy the absolute sum and the results were within FSAR allowable. Refer to Evaluation Section for reanalyses done oy UEsc.
Tne licensee's response to IE oulletin 79-04 states that no VELAN swing check valves are in any of the affected piping. Further, IE Bulletin 79-U2 was addressed unen piping support modifications were found to De necessary.
The licensee has stated that the reanalysis has no effect on pipe Dreak criteria since the postulated creak was analyzed to occur at any point on tr e pipe, inside or outside containment.
EiaLUATION The reanalysis technique employed was a lumped mass responsa spectra modal analysis.
Inis dynamic analysis procedure is an acceptacle methoa. The aosolute comoination of responses in two directions is also acceptacle to tne staff.
7r4 JU4 257 7 9 07120tSJFJS"
. Once the support modifications are complete, the af fected piping stresses, attached equipment loads, and support designs will all De in accordance with FSAR criteria and acceptaole to the staf f.
UE&C has reanalyzed category 1 (pressure coundary) and category 2 (other safety related) lines ; sing Square Root of Sum of Squares (SRSS) load comoination instead of algeeraic summation. This analysis employed a conservative f actor of 2 to convert hen the use of SRSS metnodology for a 2D analysis was fron OBE to DBE.
a questioned Decause it did not conform to FSAR commitment, the licensee applied a conservative factor of 1.3d to the SRSL results to convert to aosolute sums, which is an acceptable load comDination method with a 2D analysis, in sucn cases, credit for conservatism in the OBE/DbE relationship was taken (a more realistic f actor of 1.2 was used instead of 2).
When this exercise was completed, one of the first 39 reanalyzed lines was found to exceed total allowable stress by 21, out was still iess than 0.9 Sy as permitted by the FS AR and was found acceptacle.
For tne remaining 411 unreanalyzed lines, SRSS stresses were estimated f rom the algecraic summation stresses by applying a f actor of 1.5.
The SRSS results were then converted to absolute sums for use with tne 2D analysis by applyirg a factor of 1.3o.
Credit was again taken for the conservatism in the OBE/DBE relationship.
When this exercise was completed, 39 additional lines were found suspect.
SRSS stresses were computed for these lines wnich eliminated the factor of 1.5 Hcwever, several of these lines still gave stresses in excess of code allowaole. To resolve this proolem, the licensee recomputed the total stress using coincident point values instead of maximum values. Tnis metnod is more realistic and is acceptacle to the staf f.
The new total stresses were all within code allowable.
The aoove procedure anicn the licensee toom in completing tne piping seismic stress reanalysis under IE Sulletin 79-07 is acceptable to the staff.
ae find the licensee's responses concerning 15E Sulletin 79-U2 and 79-04 acceptacle.
fne reanalysis nas no effect on BSEP pipe creak criteria commi tted to in the FSAR.
The staf f still nas some concern as to whetner the reanalyses effort rdflects the true as-built conditions in the plant.
However, CPsL has completed a walk-down of tha oiping and supcorts inside tne dry well to verify that the as-cuil t condition hc: indeed oeen utilized.
Additionally, tne licensee nas commi tted to a field verif1:ation of all lines on octh units by June 15, 1979.
ae fi nd tni s commi tment acceptable.
CODE VER:F: CATION Tne licensee has indicated that tne following computer procrams were used in the reanalysis of this plant:
)Ui
[<
... PISYS - General Electric Company ADLPIPE United Engineers and Constructors SE has stated that the code PISYS comoines the responses due to seismic i..ulti-axial excitation by absolute summation; the modal response due to eacn excitation are combined oy methods i2s specified in Regulatory Guide 1.92.
A Fortran listing of the dynamic response calculations section of PISYS has oeen suemitted by GE and these statements nave been certified and c on f i rme d.
GE is also presently solving a set of NRC generated benchmark proolems as part of the code verification ef fort.
UEC has indicated tnat the code ADLPIP-II combines tne responses due to seismic excitation by the SRSS method when used with the response spectra technique. This has been confirmeo by examining the code listing and oy verifying tne code by solving a set of Denchmark proolems.
In addition to satisfying the code verification requirements, the licensee
.~as also d$ reed to provide the NRC two proolems for cor finaatory analysis.
These confirmatory problems will be solved independent y oy consultants to l
the NRC at Brookhaven National Laboratory. The models suomitted for these piping probleus will De confirmed by the licensee as ccrresponding te the "as-ouil t" Condi tion.
We fin 0 i'.ese Commi tments accep*.able.
CONCLUSION dased on the discussion and evaluation presented aDove, we conclude tnat the requirements set forth in laE Bulletin 79-07 are adequately satisfied to allow resumption of operation upon completion of the modifications identified in our meeting of June 4, 19T3.
b.o k
. ", r/
<v