ML20078S044
| ML20078S044 | |
| Person / Time | |
|---|---|
| Site: | 05200004 |
| Issue date: | 01/16/1995 |
| From: | Delvin S, Shiralkar B, Jacqueline Thompson GENERAL ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20078S040 | List: |
| References | |
| NUDOCS 9502230277 | |
| Download: ML20078S044 (11) | |
Text
n Enclosure to MFN No. 021-95
+ .
, SBWR Drywell to Wetwell Vacuum Breaker Valve White Paper 1.0 Purpose ;
The treatment of check valves, in passive reactors such as the SBWR, as active or passive
' components is design and application dependent. If the reliability of the component can be ,
demonstrated, the component can be treated as passive. Passive components are j exempted from application ofsingle failure criteria.' The ANSI /ANS Standard 58.9-1981, :
endorsed for application to ALWRs, allows ext:mption from single failure criteria if the !
function of the component can be demonstrated despite any credible condition. J In . !
SECY-94-084," Policy and Technical Issues Associated With Regulatory Treatment of Non-Safety Systems in Passive Plant Designs", March 28,1994, the USNRC staff recommended that passive reactors, such as the SBWR, treat check valves as active i components subject to meeting single failure criteria, unless proper function can be i demonstrated and documented. The staffposition is that, "In determining an exemption ;
to single failure consideration for a particular check valve application, the plant designer ;
shall perform a comprehensive evaluation of check valve test data or operational data for !
the similar check valve designs in similar applications and operating environments to .
demonstrate that the probability of failure is comparable to those of passive components. ;
A failure probability on the order of 1E-4 per year or less would be low enough to be .!
considered a passive failure." The SBWR wetwell to drywell vacuum breaker is a check j valve' designed and tested to comply with SECY-94-084 and ANSI /ANS Standard 58.9- ,
1981 criteria for passive check valves. The purpose of this document is to outline the - !
program that has been undertaken to show that the SBWR vacuum breaker valve meets ;
the passive component criteria and to demonstrate its reliability for all credible events. !
2.0 Background i In the SBWR, vacuum breaker valve leak tightness and reliability are critical because the !
Passive Containment Cooling System (PCCS) depends on the pressure difference between the wetwell and drywell to perform its containment cooling function. Excessive vacuum .
. breaker leakage, or failure to close, would result in degradation of heat removal capability. l The SBWR design ensures that the SBWR vacuum breaker reliability and leak tightness. !
are commensurate with its safety function. !
I 2.1 Function The vacuum breaker reduces the challenges for negative drywell pressure !
. exceeding the containment liner negative pressure service limit of three pounds per square -
inch. By limiting drywell negative pressure, the vacuum breaker also prevents the - :
suppression pool water from rising to the level of the spillover holes in the main vents. If sufficient pool water spilled to the lower drywell, the PCCS vents could uncover, creating - [
a potential suppression pool bypass path. l t
I l
h 9502230277 950216 [
, . . . ~_m __ - . _ -.~.. , m
' 25 Installation The vacuum breaker is located on the floor of the drywell slab which -
forms the boundary between the drywell and wetwell. The vacuum breaker passes the noncondensibles through a vent pipe in the floor as soon as the negative pressure !
difference exceeds the valve lift pressure ofone half pounds per square inch (0.5 psid). )
1 i
2.3 Operation There are two scenarios which could lead to vacuum breaker openmgs. ]
The first is due to inadvertent actuation of drywell sprays or flooding. The other is a Loss- !
of- Coolant Accident (LOCA), with or without drywell sprays or flooding. !
In the first scenario, the critical vacuum breaker function is to open on demand. There are i three vacuum breakers. Only two are required to control drywell negative pressure; therefore, the vacuum breaker system meets the single failure criterion for valve opening. ]
In a loss of coolant accident, noncondensibles are forced into the wetwell air space during !
the initial blowdown. The reduction of steaming from the reactor pressure vessel following J actuation of the Gravity Driven Cooling System (GDCS) has the potential to reduce the drywell pressure and open the vacuum breakers. For a GDCS line break where flow from l the broken line spills into the dryweII, or if flooding or containment spray is initiated, -
steam in the drywell is also condensed, further reducing the drywell pressure and making vacuum breaker openings more likely.
Because the vacuum breakers remain closed during the initial blowdown, they are protected against debris resulting from the blow down. Vacuum breaker opening occurs later in the LOCA transient during a quiescent period, with small pressure differences between the drywell and the wetwell. During the quiescent period, the vacuum breaker is protected from debris by inlet and outlet screens.
In normal operation, the vacuum breakers provide a wetwell to drywell path for inerting nitrogen supplied by the Containment Atmospheric Control System (CACS). The inlet screens protect the valve from debris coming from the wetwell. i 3.0 Vacuum Breaker Valve Operating History
-Historically, BWRs have used swing check valves to provide wetwell to drywell vacuum breaking. Swing check valves have a pivot pin that rotates with the disk. Gravity acts to close the valve but the closure force is reduced as the valve closes. Despite the proven attributes of swing check valves, the SBWR design team determined that a specially designed vacuum breaker valve would be desirable to meet the more stringent leak -
tightness and reliability criteria for passive check valves in the SBWR.
1 I
l 4.0 SBWR Vacuum Breaker Valve Design Requirements To meet passive criterion, the SBWR team developed the following design requirements:
4 i
2
4.1 Reliability The valve reliability goal was established by the SBWR Probabilistic Risk Analysis (PRA). Using an assumed failure rate of three every ten thousand (3E- l 4/ demand) to open or close and seal on demand, the SBWR safety objectives can be met assuming three vacuum breakers arranged in parallel. The NRC Staff reliability goaI of _
IE-4 failures per year or less is not as stringent as the failure on demand goal of 3E-4/ demand from the SBWR probability risk assessment calculations because the vacuum breaker safety related demand rate is much less than once a year. The reliability goal established the extent of required reliability testing. The valve was designed to a failure rate far lower than assumed in the PRA.
4.2 Leak Tightness For conservatism, the total effective leak area between the drywell ,
and wetwell of the SBWR was set at a maximum allowable.ofone square centimeter. .j Since the SBWR drywell liner is designed to be highly leak tight, the vacuum breakers represent the largest potential fraction of this leakage. The vacuum breaker leak area design goal was set at two percent of the allowable drywell to wetwell leakage area, or 0.02 square centimeters per valve maximum. Although the allowable leak rate of the SBWR vacuum breaker is similar to vacuum breakers used on other BWRs, the smaller number of vacuum breakers,3 versus typically 10, and the improved design of the SBWR vacuum breaker will result in a significant improvement in leak relaibility. The maximum leakage allowable is applicable during a design basis accident (DBA) that occurs on tie last day of the sixty year life of plant operation and with the valve non-metallic .
components at the end of their six year specified life.
4.3 Passive Operation The valve does not depend on external sources of power and actuates by differential pressure. The valve is normally held shut by gravity. During a los3 l of coolant accident the positive drywell to wetwell pressure also acts to close the valve. I The valve opens momentarily in the event of a negative pressure transient. !
i 4.4 Resistance to Debris In service, valve failures due to debris on the seats and bearings is a significant contributor to valve failure. The SBWR vacuum breaker was designed to l resist failure due to any credible debris. Credible debris is defined as material that could i pass through the vacuum breaker inlet or outlet screens. j 4.5 Stability Check valves tend to chatter at low flow rates. Chatter results in seat and I bearing wear and has sometimes caused failure of swing check pivot pins leading to separation of the disk. The SBWR vacuam breaker valve is designed to minimize chatter. I 4.6 Service Testing The SBWR vacuum breaker has provisions for leak testing during l outages by pressure decay measurements. The bearing friction drag will also be measured l by measuring the force required to lift the disk. !
4.7 Position Monitoring In service, the position of the valve disk on the seat is capable of being monitored by sensors and annunciated in the control room.
I 3
- yp' " m - -" - - -- ~" ' ' - ^
y;tf _ W Wlx -
, '4.8 Missile and Jet Protection The valve has shields providing protection to internals
' from small missiles andjets.
y 7 ' .5.0 Vacuum Breaker Design Features The design requirements are met by designing a valve with features, illustrated in Figures I and 2, that address the design requirements as described below:
5.1 Poppet Design Simplicity of design is used to ensure maximum design reliability.
The valve design selected has one moving pan, a venical lift poppet disk with two bearings to maintain stem-disk alignment. Closing force is provided by gravity and drywell to wetwell pressure differential. Opening force is provided by wetwell to drywell vacuum negative pressure differential. The venical poppet design was selected because h
the full force ofgravity is applied throughout the stroke providing more positive seating than afforded by a swing check valve. The engagement motion and force are uniform over the perimeter of the seat. The bearings are venical and do not support the weight of the disk as on a swing check valve, thereby minimizing bearing drag.
5.2 Double Barrier Seal Design Les Oghtness is achieved by use of a non-metallic main seal and a backup hard seal with maximum allowable leak areas of 0.02 and 0.2 square centimeters respectively. The double seal desgn provides protection from obstructions that could lodge on either seat without exceeding allowable leakage. If a panicle of the maximum size that can pass through the inlet or outlet screen holes lodges on the hard seat, the primary seal will still be in contact with the seating surface allowing the lip seal to maintain leak tightness. If on the other hand the panicle lodges on the soft seat, the lip seal can deform around the panicle and still maintain leak tightness. Since either seat can act independently to maintain leak tightness, the seal design provides single seat failure protection. After over seventy two hours at accident temperatures the primary seal will take a compression set; however, the compression set does not prevent sealing and will not occur early in the DBA when debris lodgement is more likely to occur.
5.3 Anti-Chatter Ring To prevent valve chatter that can cause excessive seat and
' bearing wear, the SBWR vacuum breaker was designed with an anti-chatter ring around the disk to ensure the seating pressure is significantly lower than the lift pressure.' The anti-chatter ring deflects flow downward, creating lift that softens disk-to-seat impact and provides damping by energy absorptior 5.4 Inlet and Outlet Shields The valve is equipped with inlet and outlet shields fabricated from perforated stainless steel sheet. The perforations are sized to prevent the entrance of a panicle large enough to create leakage through the valve that exceeds maximum allowable if the particle or panicles lodge anywhere on the seat. The shields L.,
also protect the valve internals from jets and missiles.
4
i' 5.5 Disk Position Sensors To provide the plant operator with confirmation that the disk is ,
seated securely enough to be leak tight, four non safety-related position sensors are !
located around the disk perimeter. The use of four position sensors set to measure small displacement deviations allows for the failure of any one sensor without loss of position !
sensing capability.
6.0 Testing In order to demonstrate that the vacuum breaker valve design meets the passive check !
valve reliability criteria, a prototype valve was fabricated and extensively tested.
6.1 Leak Tightness The initial testing verified the leak rate performance of the valve in i the newly manufactured condition. The valve primary soft lip seal had zero leakage (bubble tight) and the hard seat had a leak area equivalent to 0.0002 square centimeters.
The hard seat leak rate was approximately one thousandth the maximum allowable.
6.2 Performance Testing The valve was then placed in the flow test facility illustrated in Figure 4 to confirm lift pressure, flow rate and chatter performance. The valve flow capacity was adjusted by increasing the stroke. The valve performed smoothly, lifting at one half pounds per square inch (350 mm of H2O) and closing / opening with a minimum of .
vibration with the anti-chatter ring installed.
6.3 Design Basis Accident Leak Tightness In order to demonstrate that the valve could !
meet design requirements at the end ofits sixty year qualified life, the valve was aged by j radiation, high temperature, and vibration the equivalent of sixty years. The valve was then leak tested with steam for over eighty hours in the DBA test chamber illustrated in Figure 3 at temperatures and pressures enveloping the SBWR design basis loss of coolant l
accident conditions. A periodic spray of cold water to simulate the thermal shock cf containment spray was applied. The valve had zero steam leakage over the duration of the test. The valve was then removed fro.n the steam chamber and leak tested with air.
Valve leakage with air remained zero. The end oflife leak tightness requirements were exceeded. Figure 5 shows that the main sealleak rate was essentially zero until debris was deliberately introduced into the valve (Case E) after DBA testing was completed.
6.4 Reliability Testing The objective of the valve reliability test was to confirm that the valve would open and close on demand a sufficient number of times to demonstrate the l
reliability goal. The SBWR Probabilistic Risk Assessment assumed vacuum breaker wavailability is 3E-4/ demand. Bayesian reliability calculations showed that three thousand test cycles without failure would provide confirmation of the assumed unavialability. Prior to beginning three thousand cycles of testing, four pounds of sandblasting grit were passed through the valve to conservatively simulate grit that could collect on the valve bearings during sixty years of service. The valve was coated with oil to ensure the grit would adhere to the bearing surfaces. The valve was then cycled three thousand times. As shown in Figure 6, there was no significant change in the lift pressure l
5
oi % rate. The valve was then leak rate tested. Although the primary soft lip seal was degraded by embedded grit, the maximum allowable leak rate value was not exceeded.
( Figure 5, Case F) The reliability test confirmed that the valve was rugged, reliable, and met leak tightness requireme.r.s afler being exposed to conditions more severe than design basis service.
7.0 Conclusion The SBWR design team has designed a new valve to meet the stringent leak tightness and reliability criteria for check valves in passive plants To meet reliability criteria, the valve was designed with features that eliminate the problems experienced with swing check type vacuum breakers. An extensive performance and reliability test campaign demonstrated that all dnign criteria were met or exceeded. For opening, the SBWR vacuum breakers meet single failure criterion because only two of three valves need open to control negative pressure. To address the single greatest threat to valve leak tightness, judged to be debris lodged on the seat, the seat was designed to highly resistant to debris. Valve reliability testing demonstrated that failure to close is not a credible event. The SBWR vacuum breaker thus meets ANSI /ANS- 58.9 and the NRC staff criteria for passive components in passive plants.
Prepared by T.L.Thom s6n Reviewed by ,
B. Shiralkar Accepted by_
'.Dehin
/_'I!/6[9[
/
Accepted by_ b T.McIntyre" '
t.,
j Approved by .
P.N(arhott l
l 6 l
p.
)
1 2 l
vhq/)/,9f * , h,;" 77/V//,.///{1 l
%9Mi g, ' Ibe%I%
l l
t l v
--2r5 EE2'----l l
J S 6 5? ' h5 .----
4 f 22/? ;/M \ S
\
/ l '
l DRY l t 5 l WELL l
~~~ ~
LEGEND A
$ [ ' , .,'s,,d$'..
,, ' #- """]
- 7 ,", 1. DAMPER COVER
. ,',~'- 2. VB BONNET h ,' , '
',,,' ,,h14' , ', ' #, , ,
s l
e
- 3. LEAK TIGHT TEST FLANGE
- 4. OUTLET SCREEN
, 's ,' ; , -) ' , , s' ,'
l
, 5. VB BODY (4 NOZZLES)
I -,
' ' 6. CABLE PENETRATIONS
- 7. PROXIMITY FROBES (5 ITEMS) AND 13 . 9 ACCELEROMETER
' l.
l 8. REACTOR BUILDING FLOOR l 10 l 9. VB SUPPORTING STANDPlPE l 10. INLETSCREEN l 11. DISK STEM UPPER BEARING l
l
- 12. LOWER BEARING WITH ANTIROTATION l l DEVICE l 13. VB SEALING DISK l ! 14. SOFT SEAL e
- 15. HARD SEAL 12 l 11 16. ANTICHATTERING DISK WET e
- 17. BALLAST WEIGHT WELL
- 18. DAMPER Figure 1: Vacuum Breaker Design Features *"'^
f 7
t A~>
- ;7 75
. ~ '
,;. -j
- 4 ik . .-
'j-
,~
j 7
. *\ ,
\
i l .
\ .
l
,i R 8ure 2: Vacuum Breaker yagye
'l i
~
r l
4
- pr 2.
I, ,
' h ;
- 19 ,
% 1j, I
rr ,
IN ri p
-j .-
l '
$l .4
, - t- e -
? k j
~ : " i g l g j _ .e
.m &
~; >
~ ~ .: ~ ,
i .;pQ ,3 l ~
Rgure 3: DBA Test Chamber i
I
r.
5 < 4
>' 4 '
a 7 > t==":1 l a 8 9 10
,, (13.5 ft) r a1m ;
, 3m
^/ ; (10 ft) .
.. ~ .. l f/ (3.2 ft) (m s > ;
y ,
l
-w... - )
N, , _ _ . _ v N N k t
/\ /[ , *
- 2.5m :l; em 2m
' l;
) 3m
'l 1m. 0.5m ,
(8.2 ft) (13.2 ft) (39.5 ft)
(10 ft) (10 ft) __1.5mq)
(4.9 ft (1.6 ft)
/
1 3 y7- ? .O . O j f <_. L
[ [
'/ ( __[
)
N N $
l } l l ',
I l h % % H E "
~
-O O H
- 24m i _ I (79 ft)
LEGEND A
- 1. MAIN VB !"EEDING FAN (NOMINAL PERFORMANCES 8
- FLOW: 80.000 m Ar(2.84 x 10e ft8 Ar)
- HEAD: 800 mm H 2O (31.5 INCHES OF WATER)
- SPEED: 1.490 rpm
- POWER: 220 kW)
- 2. AUXILIA?Y VB 8FEEDING FAN. 3
" LIFT PRESSURE" AND " LOW FLOW" TESTS
- FLOW: 500 m Mr (17700 ft Mr)
- HEAD: 525 mm H 2O (21 INCHES OF WATER)
- 3. MAIN FAN FLOW DIVERTER BY-PASS
- 4. MAIN FAN SUCTION REGULATING VALVE
- 5. ANTIVIBRATIONJOINT
- 6. BY-PASS FLOW CALIBRATION GATE
- 7. BY PASS FLOW REGULATION VALVE
- 8. FLOW VEINS STRAIGHTENER
- 9. PIPE CONNECTION (RECTANGULAR - CIRCULAR) 8
- 10. PIPE METER (*PITOP TUBE)(FROM 0 TO 80.000 m Ar)
- 11. PRESSURE TRANSDUCER (FROM 0 TO 1.000 mm Hg0 (40 INCHES OF WATER)
- 12. VB PROTOTYPE
- 13. ANTI-NOISE STRUCTURE
- 14. VB SUPPORTING STRUCTURE OPERATIVE FLOOR
- 15. PIPE SUPPORT Figure 4: SBWR Vacuum Breaker (VB) Prototype Qualification Tests Performance and Reliability Test Stand (Fiat Ciei Test Hall)
BG581-2 9
~-
r; e
y . ;
.t l
i l 700 600 -
MAXIMUM ACCEPTABLE LEAK RATES <
s l
500 -
UNCERTAINTY FIELD TEST PHASES CONDITIONS A) BASELINE DATA !
' ~ Al AFTER RADIATION AGING .
g _ B) AFTER THERMAL AGING C) AFTER DYNAMIC AGING i l . ! D) AFTER DBA SIMULATION
!- S . E) AFTER GRIT INGESTION (DIRTY SURFACES)
(
w F) AFTER REUABluTY TEST (DIRTY SURFACF.S)
- F) AFTER REUABluTY TEST (CLEAN SURFACES) g 399 F7 AFTER REUABluTY (CLEAN SURFACES AND SEAUNG l R PASTE RENEWED ON THE BACK OF THE SEAL)
- l
\
1 O
200 -
_b
+
g I
D l l
I 100 .-
@ C A'
A
, . , , , , , , , , , , i 0 -
l 2 3 4 5 6 7 8 9 10 11 12 13 14 15 DIFFERENTIAL PRESSURE (psi) l Figure 5: VB Systematic Checks Primary Soft Seal Leak Tests Results BG581-5 l
l 10
r-h f
e 0.7 '
W 0
SG m 200 -
0.6 -
STABluZED FLOW
'y ,30.5 -
" 5 -
E UFT PRESSURE g [ 0.4 i
g 100 M V
0.3 -
u w CLOSING PRESSURE 3a g -
n
- U 0.2 - ,
?
0- 0.1 ' , , , ,
1000 1500 ~2000 2500 m -
l NUMBER OF STROKES l
Figure 6: VB Reliability Test Summary Results BG581-4 i
i P
11
'