ML19294C152
| ML19294C152 | |
| Person / Time | |
|---|---|
| Issue date: | 09/24/1979 |
| From: | Gary Young Advisory Committee on Reactor Safeguards |
| To: | Bender M Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-SM-0166, ACRS-SM-166, NUDOCS 8003070207 | |
| Download: ML19294C152 (6) | |
Text
,
,Me a an k'o
,g$9W 8/D h UNITED STATES NUCLEAR REGULATORY COMMISSION J7 b 4 '*
N 3'
i g
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS e
W ASHINGT ON, D. C. 20555 o
-/
us September 24, 1979 n
r i
Mr. Myer Bender ACRS GENERIC ITEM, VACUUM RELIEF VALVES CONTROLLING BYPASS PATHS ON BWR PRESSURE SUPPRESSION CONTAINMENTS As requested during the 233rd session of the ACRS, a study was done on the recent experience at operating BWRs relative to vacuum relief valves controlling bypass paths on BWR pressure suppression contain-ments.
This generic item was initially identified in ACRS letter, Siess to Schlesinger, dated Dec. 18,1972, " Status of Generic Items Relating to Light-Water Reactors."
It was listed as formally resolved in ACRS letter, Kerr to Anders, dated fiarch 12, 1975 The problem was identified with the Mark I design for pressure suppres-sion containment. The Mark I design (Figure 1) for pressure suppression during a LOCA requires the escaping steam from the drywell to be bubbled through the pressure suppression pool for condensation. The condensation of the steam prevents containment overpessurization.
The torus-to-drywell vacuum breaker valves (Figure 2) are designed to swing open when the pressure in the torus exceeds that of the drywell by 0.5 psi.
However, during early commercial operation of the Mark I BWRs, it was found that the vacuum breakers often stuck ooen or partially open. This condition would allow partial bypass of the LOCA steam flow directly to the torus, bypassing the pressure suppression pool. The result, if the bypass leakage is enough, could be contain-ment overpressurization.
Ammu.
/
/
MTTJNS agt t
~
supe t,,sio,s Mttsott a
pee 6
=W-
,GA.
Nunt 1.
man I Tonus 4amuseer (rmud.
800807 o *L O']
2.
// /
l
./.]
,,..t
'..,..~..,..;l m
.., '...,3' wwtu.
,,g
&ct L
E
/ L/
f_ s is///
/r Ficant 2. Tecus To Deye hun Bacxxce,Tvrw i.
tcom (tws. scaied).
C4un. Wwt The resolution of this generic itr.m was to require frequent inservice inspection of the vacuum relief valves for BWR pressure suppression Tests are run monthly to determine operability and containments.
leak tightness of these valves.
To determine the recent experiences with these valves, an LER study At present, only Mark I design containments are in Mark II and Mark III designs are still in the was performed.
commercial operation. Table 1 (attached) lists all the operating Mark I construction stage. Table 2 (attached) lists all the reported failures of the vacuum breakers to operate properly since January 1,1975.
BWRs.
The most common reported failure (12 out of 25) was for a valve to hang This failure is partially open during the monthly stroke testing.immediately re shutting the plant down for repair of the valve.
identified almost immediately, since it occurs during testing.
The second type of reported failure is identified when the drywell The majority pressure cannot be maintained above the torus pressure.
of Mark I containments are inerted with nitrogen gas during r Again, this operation. leak-tightness of the vacuum breakers is also monitored.
type problem occurs and is identified almost immeJiately.
3.
The third type of reported failure was related to startup testing at 7
Hatch-2. The vacuum breakers were prevented from closing by magnets which interfered with the normal closing of the valves.
This problem was identified and corrected during startup testing and is not con-sidered a problem expected to occur when the reactor is in commercial operation.
E The total number of torus-to-drywell vacuum breakers in service at Mark I BWRs is between 220 and 300. The total number of failures is 25 since 1975. Of those,12 were related to startup testing leaving All 25 failures 13 failures that occurred during commercial operation.
were identified promptly and corrected according to technical specifi-cations.
Figure 3 shows the failure rate by year since 1975 k
.9' 3ew n!.-
3 y s-5 6-g i,.
6 g
It o
2
..i.
ien
.. r.
. to i.,.
na Fic.urec 3 EcPaxcv hecx or swr Tows -To-c.ywcu-w w T4,c e 7. c4oss VAcwm Snana v4cn3 Poe Vcsrc.
we om so stone restmc.
+12 or twest is raswats The operating Mark I containments were analyzed by the licensees to The analyzed determine the r:aximum allowable bypasg leakage path size.This is equal to a 6 to 8 i value is approximately 0.2 to 0.3 ft.
The leak detection capability open pipe between the to.'us and drywell.
of the routine leak test (test of maintaining drywell pressure above torus' pressure) will identify a leak equal to or greater than a one-inch
4.
plate orifice.
This is equivalent to roughly 2 to 3% of the allowable maximum leakage.
Therefore, the technical specifications are conservative, and significant leaks can be tolerated through the vacuum breakers in spite of the low limits set hy the technical specifications.
The LER study revealed a large number of reported position indication failures on the vacuum breakers. These position indicators are required by technical specifications to assist in monitoring the position of the valves.
Once the position indication is lost or becomes unreliable, the valves are verified open or closed by the drywell pressure test. There-fore, none of the position indicator failures were reported here unless the valves were also found to be open.
In conclusion, the number of recent failures of vacuum breakers between the torus and drywell of pressure suppression containments has been relatively few.
All of the failures were promptly found and corrected in accordance with the technical specifications.
M.
Garry G. Young (
ACRS Fellow Attachments:
- 1. Table 1. Operating BWR Mark I Containment Nuclear Power Plants
- 2. Table 2. LERs on BWR Torus-to-drywell Vacuum Breaker Failures Since January 1,1975 cc: ACRS Members ACRS Technical Staff ACRS Fellows
g TABLE 1 OPERATING BWR MARK I CONTAINMENT NUCLEAR POWER PLANTS
- Megawatts Date of Date of Unit Name (Electric)
Initial Criticality Commercial Operation 1065 8/73 R/74 Browns Ferry 1 Browns Ferry 2 1065 7/74 3/75 Browns Ferry 3 1065 8/76 3/77 Brunswick 1 821 10/76 3/77 Brunswick 2 821 3/75 11/75 Cooper Station 778 2/74 7/74 Dresden 2 794 1/70 7/70 Dresden 3 794 1 /71 11/71 Duane Arnold 538 3/74 2/75 Fitzpatrick 821 11/74 7/75 Hatch 1 786 9/74 12/75 Hatch 2 784 7/78 Millstone 1 660 10/70 3/71 Monticello 1 545 12/70 6/ 71 Nine Mile Point 1 620 9/69 12/69 Oyster Creek 1 650 5/69 12/69 Peach Bottom 2 1065 9/73 7/74 Peack Bottom 3 1065 8/74 12/74 Pilgrim 655 6/72 12/72 Quad Cities 1 789 1 0/71 2/73 Quad Cities 2 789 4/72 3/73 Vermont Yankee 1 514 3/72 11/72
- As of July,1979 (NUREG-0020, Vol. 3, N, 7).
- Not in commercial operation as of July,1979.
TABLE 2 LERs ON BWR TORUS-TO-DRYWELL VACUUM BREAKER FAILURES SINCE JANUARY 1,1975 Unit Name Event Date Comment Dresden 2 12/77 1 valve failed to close during testing
- Dresden 2 5/78 2 valves failed to close during testing
- Hatch 2 10/78 12 valves were found open during testing
- Millstone 1 1/76 1 valve failed to close during testing
- Millstone 1 8/77 3 valves failed to close during testing
- Monticello 1 2/78 1 valve failed to close during testing
- Oyster Creek 1 9/76 1 valve failed to close during testing
- Peach Bottom 3 7/77 1 valve was found open during testing
- Quad Cities 1 12/77 1 valve failed to close during testing
- Quad Cities 1 5/78 1 valve failed ic close during testing
- f Quad Cities 1 5/79 1 valve failed to close during testing
- These valves were closed prior to testing, shaft binding during surveillance testing caused them to stick open.
These valves were prevented from closing by retaining magnets used for set-point adjustment during startup.
This valve was found open when drywell nitrogen pressure could not be maintained.
.