ML20009B404
| ML20009B404 | |
| Person / Time | |
|---|---|
| Site: | McGuire, Mcguire  |
| Issue date: | 07/06/1981 |
| From: | Parker W, Wylie K DUKE POWER CO. |
| To: | Adensam E, Harold Denton Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20009B403 | List: |
| References | |
| NUDOCS 8107150378 | |
| Download: ML20009B404 (3) | |
Text
.
. ATTACHMENT 3 5..
~
c DUKE POWER COMPANY July 6,1981 Mr. Harold R. Denton, Director Office of Nuclear Regulation U.S. Nuclear Regulatory Commission Washington, D.C.
20555 Attention:
Ms. E. G. Adensam, Licensing Branch.4 Re:
McGuire Nuclear Station Docket Nos. 50-369,370 Desr Mr. Denton:
A quest' ion has arisen with regard to the use of D. G. O'Brien connectors in the McGuire containment. The pupose of this' Letter is to confirm the qualification of these connectors is adequate for plant opera-that It is our position that the connectors in tion to 100% of rated power.
the electrical penetration assemblies,(EPj) are fully qualified for normal and postulated accident ' environments on the bisis of in-service original developmental and qualification testing in accordance with IEEE Standard 317-1972 and IEEE Standard 323-1971 performed by D. G.'O'Brien, Inc. (D.G.O.) as documented in D.G.O. Engineering Report Number 252 dated March 21, 1977 as revised August 15, 1977, previously submitted The adequacy of these tests is evident by the more. severe to NRC.
exposure of the inside connectors to the accident steam environment and environment; i.e., 340'F the margin included in the specified test temperature for the test in lieu of 327 F and 240 F respectively for the Addit' anally three maximum steam line break and LOCA conditions.
consecutive accident cycles were performed on the "K" type instrumenta-tion without failures. Also, all the EPA's for the station were
, qualified by testing five different sizes of EPA's selected to envelope those installed in the plant.
The question concerning the McGu, ire connectors stems f rom recent qualifitation testing conducted at Sandia Nationst Lacoratory on several connectors like those installed in McGuire.
This.. testing was conducted by the NRC as part of a qualification verification program initiated Duke Power Company has cooperated with the NRC several years ago.
Staff in this effort.
There were four (4) variations between the Sandia tests and the original qualification program conducted by D. G. O'Brien:
1.
The Sandia test profile was shorter than the o iginal qua'lification -
profile by 30 Iours at the 250 F plateau
~
8107150378 810707 PDR ADOCK G5000369 P.
PDR.
-4
.This difference makes the sandia test slightly less severe from the
'~
standpoint. ' Both profiles enec; pass the postu-time at temperature Lated accident peak temperatures and hold phese tempc ratures for times in excess of the postulated 7a*ccident profiles.,
inr.luded a junction box on the inside chamber end 2.-
The Sandia testThis simUlaterthe actuatinstartled configtsrationA-of the EPA.
D. G.'O'Brien test did not include.this box and.therefore allowed the inside_ connector to immediately see the fullfsteam temperature.
showed there was a considerable lag time between the The Sandia. test introduct.i.on of the steam and the gotht where the' connector tempera-ture actually reached the chamber temperature C 20-30 minutes)
I,;$
thermally agF3' the penetration at 300 F for 168 hours0.00194 days <br />0.0467 hours <br />2.777778e-4 weeks <br />6.3924e-5 months <br /> 3.
The Sandia test The D. G. O'Brien test in a fully assembled and torgoed condition.
did not include thermal aging of th'e cable grommet.
4.
-In the Sandia test the outside cont &inment end of the penetration was covered throughout'the test with an ins'ulating blanket..
D. G. O'Brien did not include. insulation in their test'.
The Sandia test did not a* Low for the normal heat sink provided by the outside containment'end of the penetration being exposed to a The D. G.
cuch lesser temperature than the accident temperature.
O'Brien test did provide for this.
The lack of the heat sink causes the connectors to see a h'i' her total temperature than would g
be seen in the as installed conditio0 r
t the should be noted that neither of these tests fully represen, It sink provided by the penetrations being bolted onto a flange heat In both tests the sleeve.which is part of the steel containment.
chacber.
penetration flance was bolted directly to the heated test Normally heat would be lost through the flange and not introduced A mocking up of the actual irstallation would further into it.
reduce the total temperature experienced by the connectors.
These failures Several failures resultedsfrcm the Sandia test program.
can be attributed to the accelerated aging method that was ecployed.
The thermal aging of these connectors in an assembled conditien at the
. introduces a failure mode which would not exist This additional elevated remerature under normal conditions during the life of the plant.
f ailure mode consisted of an expansion of the grommet material (Silicone The subsequent RTV) which caused damage to the cable insulation.
retorquing of these connectors further introduced a siteation which would This retorquing further accele-not exist in the as installed condition.
rated this failure mode when the pen-tration was exposed te the accident Duke would note that tests of several connectors c6nduc'ted by.
profile.
D. G. O'Brien for another f acility resulted in f ailures siailar to the It is inasmuch as the same failure mode was observed.
Sandia experience Duke Power Company's position that such failures would not occur under postulated accident cc.sditions, but rather, are the result of theRegardless, what-artificialty of the accerterated/ thermal aging process.
ever the ultimate disposition of the aging question may be, there is no:
immediate-safety concern, due to the fact that McGuire has yet to commence operation.
p.-
s.
I Notwithstanding the above, it is recognized that there are several' aspects of the Sandia test which merit further investigation in order to better understand the failure mechanism observed. Duke Power Company has planned additional testing to furthei veiify the adeiluac9 of the:
A more complete description of the :t qualification of these connectors.
planned testing will be submitted by July L5, 1981.
bh' 8
William O. Parker, Jr.-
l Y
O 6
0 9
4 6
5 ii
_ -.._