ML20154D836
| ML20154D836 | |
| Person / Time | |
|---|---|
| Issue date: | 12/31/1987 |
| From: | Ross D NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Jordan E NRC |
| Shared Package | |
| ML20154D839 | List: |
| References | |
| REF-GTECI-093, REF-GTECI-NI, TASK-093, TASK-93, TASK-OR IEB-85-001, IEB-85-1, NUDOCS 8809160144 | |
| Download: ML20154D836 (4) | |
Text
~
. ,n M
'f! E M' El
- [#* *'*%?. ...
UNITE 3 STATES E ' yy ' i
- E NUCLE AR FtEGULATORY COMMISSION WASHINGTON D. C. 20555 J k , , ) lM s '.m / /-
N , ,l', . / DEC 31 IM f Yd N ""*
PEMORANDUM RE: E. Jordan
, FROM: D. Ross
SUBJECT:
REVIEW OF GI 93, "STEAM BINDING OF AUXILIARY FEEDWATER PL7pS", AT CRGR MEETING #127, DECEMBER 23, 1987
Enclosure:
.. Memo Ernst to Ross, dated December 30, 1987 Revision 4 to the CRGR Charter (4-23-87), page 8, paragraph D, notes that:
"It is the responsibility of each men 1ber to assure that the minutes accurately reflect his views."
Since I did not agree with the consensus of the CRGR, I request that this coevnent memorandum be included as part of the minutes ogting #127.
GI-93 was reviewed by CRGR on December 23, 1987. Some of the AFW steam 4
binding history is: ;
- 1. The history of steam binding of AFW pumps dates back at least 1
to 1983.
- 2. Rulletin IE 85-01 was issued on 10-29-85, on this topic.
(This bulletin was reviewed by CRGR at meeting #81).
- 3. The IE 85-01 noted licensee actions as:
a) develop procedures for ronitoring fluid conditions in AFW system b) develop procedures for recognizing steam binding ard restoring AFW system to "operable" 4
c) procedural controls remain in effect until completion
, of hardware rodification to suostantially reduce ;
{ the likelihood of steam _ binding (emphasis added) or until GI 93 resolved.
The proposed resolution of GI 93 is a generic letter, which indefinitely extends items a and b above, in paragraph 3, but is silent on (c).
i The generic letter notes that, for plants with a high back-leakage event rate, the trenitoring system with control room alanns was instrumental in providing early warning and timely corrective action, !
) 8009160144 000111 1 PDR REVCP NRCCRCR NEETINQ127 PNU i
)
i 2
1
! ne GL concludes that the renitoring requircrrents of the bulletin rust
- be
- entinued (indefinitely) to keep the risk low. If, the GL continues, nlfunctioning of chcck valves cccurs, the rrcnttoring frequency nust be i r.c rea sed. Again, it is silent on hardware modification.
The regulatory analysis (p. 7) notes the INPO SOER-06-3 (issued 10-15-86)on <
t this topic. Among other things, the follow-up to this industry-sponsored i SOER tasked utilities to implement appropriate changas in their check valve ,
j raintenance and t~
~
orograms, and to identify the high risk check valves i and include ths 7 gram (this requirerent is more than in the proposed GL).
The regulitcry N -
. .rting on p. 14, discustes what it refers to as group l B pla , i.e *
% have experianced multiple instances of backleakage j
! into A: r a r.o include Farley (both units), McGuire (both units), i
- Catawba it ' + m' Diablo Canyon 2. Apparently there has not yet been ;
any atta . At Farley, the underlying problem was check valve !
l ieakage, i, che regulatory analysis, this was the result of bad '
maintenar:ce , aw be fixed. Along these lines I called M. Ernst in R:!I to inqu . to the status in his regfoC His response is provided in
- the e9 closure. He provided details for Farley, McGuire, Catawba, and Shearon Hareis. Check valve repairs, modifications, or replacements have been e'fected at these plar.ts such that they probably are no longer B plants. Thus I believe they did what the IE-85-01 bulletin wanted dere; the GL should do no less.
At McGuire, there was an installation error, and possibly poor design.
At Catawba ** there was a pattern of almost continuous leakage through one or core interfacing check valves. At Diablo Canyon 2 there was a leak in a weld '
en an interfacing check valve disc; a weld repair apparently fixed this. The enclosure is a more up-to-date assessment of the B-plant status. ,
1 l l I
I l i i
- "must" is used several times in the GL; it is not clear to me how something '
i] that "aust" be done is assured sans a licensing connitirent. !
l !
- This information that I extract from the regulatory analysis should be l qualified to the extent that the analysis was dated February 87, and thus '
- reflects the plant status as of about 1 year ago.
\
l ,
l l
i 1
a~
3 Tre distressing factor in the Categorv B plants is that, as discussed in the re;ulatory analysis and noted f ri !EB 85-01 (para. 3, p 3) hardware rcdifications 4 te substantially reducc the likelinced of steam binding are needed, re ertheless the GL is silent o7 this tooic. It seemingly tolerates e likelihood of leakage en the basis that it would be detected; this is to me a reduction in defense depth. (Indeed, on p.17, of the regulatory anal two alternatives do not include a hardware modification as an option.)ysis the I do not fault the risk analysis; it well may be correct. And, on p. 23, the 1
analysis suggests a reduction in probability of leakage once the industry program on check valve reliability is implemented. Although the B plants i have incorporated continual monitoring systems with control room alann, there is no requirement to do so, either in IEB 85-01 or in the proposed GL.
i Considering the safety function of the AFW, it does not seem prudent for the NRC to tolerate the category B plants and simply leave the repair, replacement, and proper maintenance to the option of the licensee.
Fortenately, it appears that the bt.ck. leakages are fixed, now, i Perhaps 50.59(2) applies, regiNNng an unreviewed safety question and the "probability of occurence of an accident or malfunctior, of equipment important
- to safety", or other portion of para (b). Perhaps it is GDC 34. on the j subject of residual Fect removal. Ferhaps we could cite Section XVI.
Corrective Actions, of App. B to Part 50. Although the regulations are not specific ir providing a basis for rcouiring high reliability of cocponents such as pumps nr valves, there is no doubt that GCC 1* applies, thus
. establishing the general requirement for high reliability where the safety 1
function of the cceponents require it. Fore importantly, the basis of
- regulating by the NRC has been defense-in depth that requires among other tHngs that first you design and fabricate to high structural and functional standards and then assume failures not having known failures.
l Whatever the regulatory basis citation, I believe that the resolution of this i GI should have included as requirements:
j !) the spirit of para. 3 of IEB 8!-01: "hardware rodificatiens te substantially reduce the likelibccd l of steam birding" for category 8 plants, and 8
]'
- !t is relevant and interesting to note that in an NRR letter to Jordan j (wrt CRGR) cn the subject of 3 proposed bulletin on reactor trip breakers, j NRR cites G001 in its clause on requiring that "structures, systems. A i ccrperents important to safety be designed, fabricated, erected, and tested
! to quality standards comensurate with the importance of the safety functions
] to be r 'orred." If licensees are to assure compliance with this criterion j for re6 # trip breakers, then why not the AFW interfacing vavlss?
i i
i
_ . _ . - - -- _ . _ , _ _,-_-. - _-. . . _ _ _ ._,,.-- _ _ n -~,
I 4 l
- 2) ccntinuous tronitoring alarrs in control recr; for B plants, pending the completion cf (1) above.
i > addition, I have sone continuing c:ncern abcut use of procedures alcr.e :
! ("touchee-feelee") for plants (even the A plants) that will cperate perhaps -
i 40-60 years more. At the least, AEOD should do another case study in about '
3 years to see if there are any residual or new B plants.
l
,I ~ j i e -
i D. . oss f
'i Deputy Director for Research t
- Office of Nuclear Regulatory Research
1 l
Enclosure:
, As stated [
. t f
cc: CRGR Nembers_ ,_
i
< i 4
f i
I l
i j i 1 !
i i 3 t I
i !
f !
I