ML17309A436
| ML17309A436 | |
| Person / Time | |
|---|---|
| Site: | Point Beach, Turkey Point, Ginna, Robinson, 05000000 |
| Issue date: | 02/02/1990 |
| From: | Novak T NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | Rossi C Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML17309A437 | List: |
| References | |
| AEOD-T904, NUDOCS 9002120044 | |
| Download: ML17309A436 (14) | |
Text
FEB 02 1890 MEMORANDUM FOR:
Charles E. Rossi, Director Division of Operational Events Assessment Office of Nuclear Reactor Regulation FROM:
SUBJECT:
Thomas M. Nova k, Director Division of Safety Programs Office for Analysis and Evaluation of Operational Data DESIGN DEFICIENCY OF WESTINGHOUSE SAFETY INJECTION BLOCK SWITCH On April 18,
- 1989, we issued the enclosed Technical Review Report, AEOD/T904, on a design deficiency of the Safety Injection (SI) block switch used at Point Beach Nuclear Plant, Unit 1.
The existing scheme uses a single "block/unblock" manual selector switch for both SI trains.
A licensee gnalgsis indicates that a single failure of this switch can block low pressurizer pressure or low steam-
'line pressue SI signal in both trains.
On August 19, 1988, the Wisconsin Electric Power
- Company, the owner of the Point Beach Nuclear Plant, issued LER 88-007 detai ling the deficiency they had found in the existirg scheme.
They have resolved the issue by installing two selector
- switches, one for each train.
At our initiation, Westinghouse, the NSSS vendor, has investigated the generic aspects of this single failure issue.
Westinghouse has identified three other plants - Ginna, Turkey Point Units 3 8 4, and and Robinson Unit 2 which use similar block switches.
Westinghouse has informed the licensees of these three plants of the problem and its proposed resolution (see enc'tosed copy of Westinghouse letters on this issue).
This completes our study on this issue.
Enclosures:
As stated Thomas M. Novak, Director Division of Safety Programs Office for Analysis and Evaluation of Operational Data CQOIK g OOL l'l IGlO.
- IOO, lOO i OQ'A Gl<
CQOK
,OA
'o-a N L'istribution:
'Central Fi'le,'.
ROAB R/F CSP R/F SMazumdar hlChiramal JRosenthal KBlack MWi 1 1 iams ROA SP:AEOD SMazumdar:as h
1 /L f1'(
Veenaroya CHaughney HFau 1 kner CHeltemes EJordan MTay 1 or BBrady ROAB:DSP:AEOD MChiramal
'we R
e:DSP:AEOD JRosenthal
~A w/vo D:DSP:AEOD VBenaroya 2 i I ~9O p p(b)
QgctACl'~
AC~
xtisl~4 telu~
w( J)~
~qt(~Ph t
- AEOD T
vak 2AM90
0 S
II I
F
~,(j
)
~
NestIngh ouse Electric Corporation Energy Systems RGE-89-647 Nuclear and Advanced Techrology Oivislrl Box 355 Rnaburgh Peeeylvania 15230 0355 October 12, l989 NS-OPLS-OPL-I-89-517 Hr. R. Eliasz Rochester Gas and Electric Corp.
49 East Avenue Rochester, NY 14649 ROCHESTER GAS Ir ELECTRIC CORPORATION RE GINNA STATION
Dear Nr. Eliasz:
The purpose of this letter is to provide confirmation and formal notification to you as follow-up to our telephone conversation of August 17, 1989.
The issue involves the use of a single switch to control the block/unblock function for both trains of safety in)ection.
During a control rooa design review at the Po1nt Beach Nuclear Plant, the use of a single aanual SI block/unblock switch for both SI trains was questioned.
A subsequent review led to the conclusion that a single failure of the switch (Westinghouse OT2) could block either the automatic low pressurizer pressure or the low steamline pressure SI signal in both trains.
Point Beach prepared LER 88-007 (attached) on 9/16/88 describing in detail their review and conclusion.
Westinghouse was contacted by the NK informing us of the issue and requesting our review.
Mestinghouse has reviewed both the LER and the postulated switch failure mechanism and agrees that a single failure could cause the blocking of both SI trains.
four contact blocks are stacked in series and operated by a single switch Iechanise.
Ef the upper contact block internals stick it would cause both contacts to remain in the block position.
Nore recent Westinghouse designs have provided one switch per train and Point Beach is planning to do the, same.
i1
RGE-89-647 NS-OPLS-OPL-I-89-517 October 12, 1989 Page 2
If automatic initiation of SI is blocked by a failed switch, an annunciator will alert the operator so that SI may be initiated manually.
The operator should also observe a failed switch during a normal cooldown for the same reason.
For most events, pressurizer pressure and steamline pressure SI are also backed up by SI initiation on containment pressure.
These c~upied with the 'low probability of failure of a control board switch (10- /yr.)
provide sufficient Justification for continued operation until a convenient time to add a separate switch.
Our records (Dwg. ¹110E059 Sheet 3 Rev.
- 10) indicate that the above mentioned block functions were implemented with a single switch at Ginna.
Although not an immediate safety concern, Westinghouse recommends that design changes be developed to provide separate block switches for each train.
If you have any questions or comments, please contact the undersigned.
j.8c 7mmmmm~
S. DiToamaso/
Attachment cc: G. Mrobel, HQ D. Lewis, EUFS
- 6. Link. HQ 1L, lA 1L, 1A 1L plA Very truly yours, p
NESTNGHOUSE ELECTRIC CORPORATION S,
. Swigart, Pro ct Nanager New York Area Customer Prospects Department
Nesttngho use Electric Corporation Energy Systems Nuclear and Advanced Technology Divlslon 8ox 355 Pittsburgh Pennsylvania 15230 0355 October 26, 1989 FPL-89-884 NS-OPLS-OPL-II-89-779 Hr. D. A. Chancy, Director Nuclear Licensing Department Florida Power 5 Light Company P. 0.
Box 14000 700 Universe Blvd Juno Beach, Florida 33408 Attention: Hr. P. L. Pace FLORIDA POMER 5 LIGHT COMPANY TURKEY POINT UNITS 3 5 4
Dear Hr. Hale:
The purpose of this letter is to provide confirmation and formal notification to you as follow-up to our telephone conversation of August 17, 1989.
The issue involves the use of a single switch to control the block/unblock function for both trains of safety injection.
During a control room design review at the Point Beach Nuclear Plant, the use of a single manual SI block/unblock switch for both SI trains was questioned.
A subsequent review led to the conclusion that a single failure of the switch (g OT2) could block either the automatic low pressurizer pressure or the low steamline pressure SI signal in both trains.
Point Beach prepared LER 88-007 (attached) on 9/16/88 describing in detail their review and conclusion.
M was contacted by the NRC informing us of the issue and requesting our review.
g <IJUFjSSg M has reviewed both the LER and the postulated switch failure mechanism and agrees that a single failure could cause the blocking of both SI trains.
Four contact blocks are stacked in series and operated by a single switch mechanism.
If the upper contact block internals stick it would cause both contacts to remain in the block position.
More recent M designs have provided one switch per train and Point Beach is p'lanning to do the same.
FPL-89-884 NS-OPLS-OPL-II-89-779 Page 2
A T
M If automatic initiation of SI is blocked by a failed switch, an annunciator will alert the operator so that SI may initiated manually.
The operator should also observe a failed switch during a normal cooldown for the same reason.
For most events, pressurizer pressure and steamline pressure SI are also backed up by SI initiation on containment pressure.
These c~upl~d with the low probability of failure of a control board switch (10 10 /yr.) provide sufficient justification for continued operation until a convenient time to add a separate switch.
Our records (Dwg. fllOE188, sh 5, Rev.
- 10) indicate that the above mentioned block functions were implemented with a single switch on your plant.
Although not an immediate safety concern, Westinghouse recommends that design changes be developed to provide separate block switches for each train.
If you have any questions, please contact the undersigned.
Yery truly yours, WESTINGHOUSE EL CTRIC CORPORATION a.
A.
. J. Richards, Manager Florida Power L Light Project
~
A A
IH gg4 r
Nestlngh ouse Bectrlc Corporation Energy Systems Bet 355 Pttt~ Peesyhranil 15230 0355 Hr. R. E. Morgan General Manager H. B. Robinson SEG Plant Carolina Power 3 Light Company P. 0. Box 790 Hartsville, NC 29550 CPL-89-633 October 13, 1989 NS-OPLS-OPL-II-89-,751 CAROLINA POMER 5 LIGHT COHPANY H. B.
ROBINSON UNIT 2
Dear Hr. morgan:
The purpose of this letter is to provide confirmation and formal notification to you as follow-up to our telephone conversation of August 17, 1989.
The issue involves the use of a single switch to control the block/unblock function for both trains of safety injection.
KQSHEE During a control room design review at the Point Beach Nuclear Plant, the use of a sing'ie manual SI block/unblock switch for both SI trains was questioned.
A subsequent review led to the conclusion that a single failure of the switch {llestinghouse OT2) could block either the automatic low pressurizer pressure or the low steamline pressure SI signal in both trains.
Point Beach prepared LER 88-007 {attached) on 9/16/88 describing fn detail their review and conclusion.
westinghouse was contacted by the NRC informing us of the issue and requesting our review.
QIKllKI9H westinghouse has reviewed both the LER and the postulated switch failure mechanism and agrees that a single failure could cause the blocking of both SI trains.
Four contact blocks are stacked in series and operated by a single switch mechanism, If the upper contact block internals stick it would cause botf1 contacts to remain in the block position.
Nore recent Westinghouse designs have provided one switch per train and Point Beach is planning to do the same.
"4t
Nr. R. E. @organ CPL-89-633 October 13, 1989 If automatic initiation of SI is blocked by a failed switch, an annunciator will alert the operator so that SI may be initiated Ianually.
The operator should also observe a failed switch during a normal cooldown for the same reason.
For most events, pressuriier pressure and steamline pressure SI are also backed up by SI initiation on containment pressure.
These co~pled with the low probability of failure'f a control board switch (10- 5/yr.)
provide sufficient Justification for continued operation until a convenient time to add a separate switch.
Our records (Dwg. Nl)OE198 Sheet 6 Rev. )2) indicate that the above mentioned block functions were implemented with a single switch at H. B. Robinson.
Although not an isssediate safety concern, Mestin~house recommends that design changes be developed to provide separate b ock switches for each train.
If you have any questions or comments, please contact the undersigned.
Very truly yours, WESTINGHOUSE ELECTRIC CORPORATION 6.
. Percival, a ager Carolina Area Customer Prospects Department
/)as cc:
R. E. Morgan (COL - HBR)
)L C. R. Diet@ (CPEL - HHR)
)L J. H. Curley (COL - HBR)
)L D. H. Boatwright
{CP8iL - HHR)
IL B. H. Slone
{CPSL - HBR) 1L M. J.
Flanagan (CPIL - HBR)
)L L. I. Loflin (CPSL) 1L R. H. Parsons (CP8L)
)L T. 8. Clements (CPSL)
)L C. M. Crawford (CP8L)
)L R. L. Sanders (COL)
)L J.
F. Nevill {CPEL)
)L R. J. Nuth (M - HBR)
)L
- 6. S. Meingarten (I - Raleigh)
)L
t
~