ML20197F557
| ML20197F557 | |
| Person / Time | |
|---|---|
| Issue date: | 04/17/1981 |
| From: | Ernst M Office of Nuclear Reactor Regulation |
| To: | Muller D Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20197F561 | List: |
| References | |
| FOIA-86-504 NUDOCS 8105080329 | |
| Download: ML20197F557 (15) | |
Text
%dA "
f'
...m.......
Central File
~-
SPEB File April if' 1981 44 jff k
ENORANDUM FOR: Lester S. Rubenstein Assistant Director for Core and Containment Systens, Division of Systers Integration, NRR FRON:
Malcolm L. Ernst, Assistant Director for Technology, Division of Safety Technology, NRP.
SUBJECT:
PROPOSED POSITION REGARDING CONTAINMENT PURCE/ VENT SYSTEMS
References:
1.
MePe te L. Rubenstein, from M. Ernst " Proposed Position Regarding Centainment Purge /Yent Systens," dated December 2,1900.
2.
Menn to H. R. Denton, thru D. F. Ross, ferm L. Rubenstein, same subject, not yet dated.
3.
Memo to H. R. Denton, from R. M. Bernero, 'ALAB Decision 603, dated July 30, 1990 on Station Blackout At St. Lucie Unit 2 " dated August 22, 1980.
In my menerandum of December 2,1980, (Reference 1) I advised you that the Safety Program Evaluation Branch (SPEB) was perforrina a risk assessment evalua-Q.
tien of the three positions regarding the design and use of containment purce I
systems, i.e., (1) the SRP 6.2.4 and BTP 6-4 position, as presently written; (2) a position limiting purging to 90 hours0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> a year: and (3) the proposed codi-l fication pemitting a continuous purging through a small three valve bypass system (Reference 2). The SPEB was assisted in this evaluation by members of the Reliability and Risk Assessment Branch (RRAB) and the Applied Statistics Branch of the Office of Management and Program Analysis.
This memorandum reports the results of the completed evaluation. A sunmary of our evaluation is presented below. Following this are sections which. discuss j
our assessment and the underlying assumptions which were used in performing the evaluation, and present our conclusions and recommendations.
The evaluation took longer than originally anticipated. One reason was other high priority assignnents. Another reason was that we deciddd to estimate the upper and lower rounds of the risk assessment as accurately af possible. This required a considerable theoretical effort by the Applied Statistics Branch.
Suveiary The three different containment purge systems listed above were evaluated. The evaluation basically consisted of calculating the probability of a LOCA followed by undesirable sequences in which there was either a direct path or leakage path throuch the containment to the environment. Some of the sequences considered involved failure of the ECCS and therefore major core damace and high conse-quences.
In other sequences the ECCS worked successfully and the consequences are less severe. Since there is mlatively little data of the failure probabil-ity of the large diameter butterfly valves used in purge systems, a sensitivity
/r/@,
fg2oJh@3.27g/
y,
t Lester S. Rubenstein
-?-
(
study was performed to determine the impact of valve failure probability on the overall probability of direct or leakage paths. Best estimate, upper and inwer 1 ound probabilities were calculated for the various sequances.
The results of the evaluation confires the adequacy of the present position listed in SRP 6.2 and BTP CSB 6 4 at least for several years while definitive dsta regarding the failure probability of large purge butterfly valves are being obtained. Even for an assumed failure rate as great as 0.1 per dement',
the upper bound failures that result in eigher a direct path, or a leakare path from the containment is less than 10 per year for a LOCA that results in a degraded or severely damaged core. For the same assuned valve failure rate
(.1 per demand) with a successfully pitigated L0gA, the upper bound probahtlity
. of a direct path from the containment is 5 x 10 per reactor year, but the consequences would be less severe because of the successful ECCS operation.
Based on the results of our evaluation, we recorriend that the positions des-cribed in SRP 6.24 and BTP CSB 6-4 be continued to be used as the licensing basis for purge systees. However, we further recomend that a program be initiated to obtain data on the failure rate of large butterfly valves and that the leakage tegting recocoendations rtde as a result of the resolution of Gemeric Issue B-20 be incorporated in the revision to the SRP currently in preparation and be implemented on operating plants as soon as possible.
Discussion The first step in the risk evaluation was to develop event trees which include all the potential failures which could result in a hich consequence event. The initiating event for each event tree is a loss-of-coolant accident (LOCA).
Potential success-failure paths include purae system operation at the time of the LOCA, enntainment isolation signal and containment isolation valve operabil-ity, containment isolation valve leakage ECCS perfonrance and operator action to intervene and correct for selected failures. The evaluation included both upper and lower bound calculations of the probability of specific cochinations of events.
Table 1 presents the upper and lower bound values assumed for the probability of specific events which were used in our analysis. About half of the individual reliability values used were obtained directly from the Reactor Safety Study (WASH-1400). The remaining values were determined based on engineering judge-ment, plant operating experience, the individual engineers knowledge and extra-polation of WASH-1400 values for similar but not identical equipment.
Figures 1-3 present selected portions of the complete event tree which depicts all the combinations of individual events necessary to produce a complete spec-trum of possible end effects. Figures 1 and 2 depict the possible sequences which could occur assuming the purge system is in use (containment isolation valves open) at the time a LOCA occurs. Figure 1 consists of that portion of l
l l
l Lester S. Rubenstein,
the event tree that assumes that containment isolation signal A fails. Figure 2 follows the possible sequences that could occur if containment isolation signal A operates as designed. When the containment purce system is not in use the purge system containment isolation valves will be in the closed position. We did not consider failures which could open the valves which are already closed at the time of a LOCA. However, we did consider the possibility of excessive leakage of both containment isolation valves in series in deterwining the overall probability of a LOCA in which excessive leakage through the purge system is experienced. As can be seen from Figures 1 and 2 the possible sequences reduce to three distinct results: (1) a LOCA with a direct path to the environment through two open purge system containment isolation valves; j
(2) a LOCA with a leak path to the environment and (3) a LOCA with no path to the environment. Figure 3 shows the event tree that considers the first l
two of these distinct situations and the sequences for the situations where the ECCS does and does not successfully sitigate the LOCA.
The following three endesirable sequences leading to significant conseopences were defined:
(1) LOCA with major fuel damsoe and excessive containment purge valve leakaoe; (2) LOCA with majn-fuel damage and loss of containment integrity by failure of the contalment purge system to isolate, and j
g (3) LOCA with acceptable ECCS performance but loss of containment integrity by f
failure of the contaiment purge system to isolate.
It should be noted that the Reactor Safety study reported in WASH-1400 concluded that only accidents which resulted in core melt (major fuel damage) produced the major risk events for nuclear plants. However, the reactions to the THI accident indicate that a mitigated LOCA with loss of containment integrity (i.e., purge system failed open), although not considered to have a severe enough release to the public to be considered a major risk in. WASH-1400, would clearly be an undesirable event with political overtones which would warrant some measures to avoid. Therefore, sequence 3, above, was included in our analysis of unacceptable consequences.
Generic Issue Bdo " Containment Leakage Due to Seal Deterioration studied the problem of excessive seal leakage for valves utilizing res111 ant seals, primarily the large butterfly valves used as contaiment isolation valves in containment purge and vent systems. Information gathered on the leakage performance of these valves has shown that when they fail to meet their containment local leak rate testino requirements, they. frequently exceed the maximum allowable containment leakage rate' by factors as high as ten times. Appendix J to 10 CFR Part 50 requires that Type C local leakaoe rate tests be performed during each reactor shutdown for reTueline but in no case at intervals greater than two years.
Instances of cross leakane failures of containnent purge system butterfly valves have been found as a result of Type C tests conducted in accordance with 10 CFR Appendix J.
For the purpose of this evaluation we have l
I Lester S. Rubenstein..
assumed leakage rate failures of containment isolation valves to be gross failures (i.e., leakage rates seny time the maximum allowable containment leakage rate used in the radiological consequence analyses performed for the plant). Gross leakage failures combined with a LOCA resulting in raajor core damage, while not having as severe a consequence as failure to isolate the purge systen (for the same ever.t), would result in offsite doses exceedino 10 CFR Part 100 guidelines by factors as high as 10. Therefo e we have classified the secuence involvino a LOCA resulting in major cor3 dameoe in combination with excessive purge valve leakage (sequence 1 above) as an undesirable secuence.
The event trees shown in Figures 1 through 3 and the probabilities of individual events shown in Table 1 were then used to determine the probability of occurrence of the above three undesirchle sequences.
In performing the calculations we found that directly applicable data for the' failure rate of large butterfly valves of the type typically used in contain-ment purge systems is not readily available. The probability of. failure of con-ventional type valves (i.e., plug, gate, etc.) determined from nuclear industry data in the WASH-1400 study was 3 x 103 to 3 x 10-4 / demand. These data were obtained from tests of safety system valves which were performed for operability testing per ASME code requirements. We talked with NRR personnel knowledgeable of the mechanical design of purge system butterfly valves, their operational history and the NRR requirements for establishing the operability of purge h
system isolation valves to be used while the plant is in operating modet 1, 2, 3 or 4.
Based onzthese conversations we used a failure rate of 3 x 10-Z 3 x 10-3 / demand for those butterfly valves whose operability has been estab-lished to the NRR staff's satisfaction. We and the NRR personnel whom we con-i sulted believe this to be an apprcpriate and probably conservative judgment.
However, as noted below, the limited data available would indicate a valve failure rate which falls in the upper end of this range.
The "oD hour per year" position and the "three valve bypass system" position t
both seem to anticipate that the probability of mechanical valve failure is very high. Based on this supposition, the "90 hour0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> per year" position requires that purge systems which utilize laroe qualified butterfly valves additionally limit the use of the system to about 1%/ year. The CSB alternate position proposes to redesign to effect containment purging through a system which utilizes three igualified valves smaller in size and of different desion for both mechanical closure and seal characteristics. In recognition of this apparent feeling that the mechanical reliability of the purge valves is low, we
(
performed a sensitivity study to investigate the effect of valve failure rate l
on the overall probability of the undesirable sequences defined above. Spect-l fically, we repeated the analysis of the probability of the three undesirable l
events assuming that the p bability of mechanical failure of the qualified i
butterfly valves is (1) 10- / demand and (2) 5 x 10- / demand.
l t
_, ~,. _ _. _
s Lester S. Rubenstein With recard to the probabil ties assumed for mechanical failure of the qualified butterfly valves we were able to find very little data. However, we were able to obtain some data which was compiled by INEL from NPRDS data and reported in an interim report currently being reviewed by NF.P. staff personnel (Sumary of Nuclear Plant $sfety Related Pump and Valve Engineering and Failure Data, EGG-EA-5253, October 1980). This report indicates that of 539 tests on butter-fly valves used in safety related nuclear facility systems there were 75 reported failures. We contacted one of the authors of the report to get a further break-down of the failures and were told that 60 of the 75 failures were excessive leakage failures and the other 15 were various other failures of which about 10-12 appear to be mechanical failures. If one were to place a high confidence in this data, then a probability of mechanical failure of about 2 x 10-2 / demand and a probability of leakage failure of about 1 x 10 / demand could be derived. s i
Calcula'tional Results and Conclusions
,g -
Table b presents apper bound, lower bound, and mean probabilities of occurrencek s of th( sequences leading to three undesirable consequences for the use of con '
tainment purge systems which are designed and operated in accordance with (1) SRP 6.2.4 and BTP CSB 6-4; (2) the '90 hour0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> per year" position; and (3) the CSB proposed alternative three valve bypass systee" position. The table also indicates the probability of occurrence of the three undesirable sequences as a function of the assumed failure rate for large butterfly purge system isolation o
valves for the three diffrent assugptions of failure gate; i.e., mechanical valve WW failure is:
(1) 3 x 10- to 3 x 10
/ demand; (2) 10- / demand; and (3) 5 x lel demand.
The probability calculations for the occurrence of the three undesirable sequences considered operator intervention to close the containment purse system isolation valves in the event the containment isolation signal failed resultinn in open purcr system penetrations. We found that the assumption of successful operator interven-tion or non-intervention nede no significant difference in the probability of the three undesirable events. The probability of the three undesirable eventsLis dominated by the probability of purge system isolation valve mechanical failure and/or leakage failure.
3 The PAS has taken the following approach to determining the need to make ir' prove-ments in plant designs.
If the probability of occurrence of an event which trou1d be a rejor contributor to overall plant risk is: (1) less than 105 / reactor year l
the design is acceptable; (2) between 1 gland 146/ reactor year the design is accep-l table for a few years while an overall balanced approach to a design or operationa i
change is developed; (3) greater than 194 / reactor year a more timely improvement l
1s needed. The PAS has used the mean probabilities of occurrence of specific events in applying these guidelines.
_-_--,_-m-.--
,-,,,,-----r--.----
-.------w-.-.-,--,,--,
7, - - -. -, -
Lester S. Runensctin t Examination of the results must be tempered by the realization that.the con-t sequences of the events with major fuel damage are significantly greater than the consequences for those sequences where the LOCA is successfully. mitigated.
The results presented in Table 2 show that use of the "90 hour0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> per year" posi-tion or the "three valve bypass system" position reduces the probability of occurrence of the three undesirable events by one or two orders of magnitude.
However, if the mean probabiiaties of occurrence of the three undesirable events are compared to the above stated PAS guidelines, then all three of the positions are acpeptable for assumed failure rates of large butterfly valves the failure rate of large as great as 10
/ demand.
If one arbitrarily assume,sj / demand (one failure qualified butterfly valves to be as great as 5 x 10 for every two demands) then the meagprobability of a mitigated LOCA with a failed open purge system is 5 x 10
/ reactor year which is in the PAS category of being acceptable for a few years while a balanced approach to design or operational changes is developed. Farther, since the mitigated LOCA does not present a large real rigk (as opposed to the potential public perception), a probability of 5 x 10 - / reactor year may be acceptable indefinately.
As can be seen in Table 2 there is a large spread between upper bound and lower bound probabilities and this is turn is reflected in the mean values which have been calculated. The Lewis Report cautioned against making decisions based on the mean probability of events detemined in the WASH-1400 reports where values used We for each event in a sequence have not all been verified by appropriate data.
h therefore felt that it might be prudent to consider the upper bound estimates of the probability of the three undesirable events before making a firm recommen-dation.
We requested the assistance of the Applied Statistics Branch (ASB) from the Office of Management and Program Analysis to develop the upper and lower bound probabilities.
Specifically, we asked Mr. Abramson to develop a technique whereby we would determine a multiplication factor to apply to a product of upper (or lower) bound probabilities to assure that the product has the same degree of conservatism as that afforded by each of the constituent events.
The technique developed by the ASB involves the use of a multiplier to be applied to the value derived by the product of upper (or lower) bounds. The multiplier (M) is defined by the following equation:
y M
=e where: A=[(Inf)-
nf) j j
i f $= upper bound value mean value 1
for each term in the product of upper (or lower) bounds.
Lester S. Rubenstein B The ASB has promised to provide us with a report that docunents 'the assunptions made in the derivation of the above relationship. We will provide you with a i
copy of the report when we receive it.
{
l The upper (and lower) bound values of events which were derived from the Reactor 8
Safety Study (WASH-1400) are reported to be 95 percentile values. Assumine that the values which the SPE8 and RRAB derived using techniques similar to WASH-1400 methods are also 95 percentile values, we used the ASP technique to calculate 95 percentile estimates of the upper and lower bound probahilities of the three undesirable sequences. These are the values shown in Table 2.
If one coepares the upper bound estimates of the probabilities of the three undesirable sequences (see Table 2) with the PAS ouidelines for decisiors based on the mean probability of significant events, one would be led to the following conclusions:
1.
If the probability of secheical failun of large qualified butterfly valves is in the range of 3 x 10 to 3 x 10
/ demand any of the three positions are acceptable;
-1 2.
If the probability of failure of large qualified butterfly valves is 10
/
demand, the BTP CSB 6-4 position is acceptable for at least several years and the other two positions are acceptable; h
3.
If the probhbility of failure of large qualified butterfly valves is as great as S x 10
/ demand then the BTP CSB 6-4 position is acceptable for only a limited time while a better solution is developed and the "three valve bypass system" alternative position is an acceptable solution. The '90 hour0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> per year" position is acceptable for the high consequence unmitigated LOCA event and is in the marginal range for the mitigated LOCA.
l
- As can be seen the conclusions which may be drawn are very dependent upon the failure rate of qualified butterfly valves if upper bound probabilities are considered.
It is our understanding that knowledgeable merhers of the staff believe the failure rate of2qua11fied betterfly valves is probably less than the assumed band of 3 x 10 to 3 x 10
/ demand. The limited data wh.tyh we we able3to obtain would tend to fall into the higher end of the 3 x 10 to 3 x 10 band.
Throughout the analysis which we have performed we used the assumption that the probability of excessee leaket) failures for large qualified butterfly valves is in the range of 10 to 10
/ demand. Again, we could find very little data l
to support or derty this assumption. Thel report that we used indicated failure rate leading to excessive leakage of 10
/ demand.
None of the positions which we reviewed had arty definitive requirements regarding leakage testing frequency.
Leak testing frequency can and should be used to assure the proper maintenance of res111 ant seals typically used in butterfly valve applications. Generic l
i
t Lester S. Rubenstein,
Issue B-70, " Containment Leakage Ove to Seal Deterioration" studied the problem of leakaoe for msiliant seals including priearily large butterfly valves.
Generic Issue B-70 has been completed and the conclusions and recommendations approved by the appropriate groups within FRR. Generic Issue t-70 recewender' that the following clarification be added to position B.4 of BTP CSB 6-4:
"The leakage integrity tests of the isolation valves in the containment vent and purge lines shall be conducted following each cycling of the month nor less often thre,e systees, but no more often than once each isolation valves in thes once each six sonths."
This change, which we understand has been approved by the proper parties, has yet to be implemented for tre operating plants and has not been incorporated into BTP CSB 6-4.
Recomendations Having considered the Lewis Report cautions on the use of use of mean probabi-lities of occurrence in decision making and the conservatism in decision making based on upper bound estimates of probability, the SPEB makes the following recomendations:
and operation of nuclear power plants. Licensees / applicants should coewnit to the design and/or use of purge systems which will minimize the use of the systems consistent with the recomendations of BTP CSB 6-4 paragraphs B.7 and B.3.
2.
Initiate a progran to make a valid engineering assessment of the failure rate to be expected for large butterfly valves which meet the NRR positions ngarding valve operability; 3.
Incorporate the Generic Issue B-70 recomended change in Position B.4 of BTP CSB 6-4.regarding leakane testing in the revision to the SRP currently under preparation.
4.
Implement the leakage testing requirements recoasnended by Generic Issue B-20 at all operating plants as soon as possible.
5.
Following completion of the evaluation of recommendation 2, above, recon-I sider the continued use of the SRP 6.2.4 and BTP CSB 6-4 position using the best engineering assessment of large qualified butterfly valve failure rates.
s Malcolm L. Ernst Assistant Director for Technology Division of Safety Technology Office of Nuclear Reactor Regulation Enclosures and ecs:
See next page
r 9-Lester S. Rubenstein Er. closures:
As Stated ces w/ enclosures:
H. Denton E. Case T. Murley D. Ross W. Butler A. Thadant P. Hearn J. Shapaker D. Shia.
W. Pasadag L. Abramson R. Baer W. M11 stead G
~
TABLE 1 April 17,1981 PROBABILITIES USED Uppe M
.l0_,gBound Lower Bou'od fy,r;,,,
_10-4 /yr.
CIS failure 1.
I train 8 x 10'3/ demand 5 x 10'4/ emand 2.
2nd train 10-2/ demand 1.5 x 10'g/ demand given failure of ist train Containment Isolation Valve (CIV)
Fails to close of demand 1.
Small process system type valve a.
one valve
-3 b.
2nd valve 3x10
/ demand 3 x 10
/ demand
-4 3rd valve given 10'{/ demand c.
10.}/ demand
/ demand 10
/ demand failure of two 10 2.
Large qualified butter-g fly valves a.
one valve
-2 b.
2nd valve given 3 x)l0
/ demand
-3
-Q 3x{/ demand
-[
0
/ demand 7/
10' / demand 9
failure of 1st 10 -
c3 #
Containment Isolation Valve Leaks exc(CIV) essively
- 1.. snall process system type valve a.
one valve
-2 b.
2nd given 1st leaks 10 -2/ demand
-3 3rd given 1 & 2 leak
-2/ demand 10.3/ demand 10 c.
10 3/ demand 10
/ demand 10
/ demand 2.
large butterfly valve a.
one valve
-1 b.
2nd given 1st leaks 10 1/ demand
-2 10 1 / demand 10
/ demand 10
/ demand Operator manually closes 2.5 x 10'I / demand 10'I / demand CIV having detected CIS failure ShorttermECCS(inj) fails 3 x 10 -2/ demand 3 x 10-3 / demand resulting in core damage e
~
-2 April 17,1981 LOCA Upper Bound Lower Bound 10~2 /yr.
10 - l-/yr.
Long Term ECCS (recir.)
fails resulting in core damage
-2 a.
with containment 3 x 10 / demand 4 x 10
/ demand
-3 integrity (expected PcandHp)
-1
-2 b.
without containment 2.5 x 10
/ demand 10
/ demand integrity (low Pc and H 0) 2 Containment Purge System in use 1
1 a.
SRP 6.2.4/BTP 6-4 4.x 10
/yr.
10
/yr.
b.
90 hr./yr. Position 10-2 /yr.
10-2/yr c.
Alternate Position 1.0
/yr.
4 x 10-j /yr.
o O
t
.m.
TAPLE 2 April 17,1981 i
l'.0CA with Major Fuel Damage Mitigated LOCA Direct Pct 0 to i
""I' i
Direct Path to Environ Leak Path To Environ Upper tog Ltser Upper Log Lower Uppe
.ig
....e r Bound Mean Bound Bound Mean Bound bound. ::ean dvund Assuming the Probability of Failure to Close = 3 x 10 3 x 10-3/ Demand for Large Butt.erfly Val ver.
l l
2x10-7 4x10 1x10 1 x10-6 2x10-8 j 5x10-IU lx10-6 '7x10-410 " !
SRP 6.2.4/BTP 6-4 4
90 Hr./Yr. Positior 8x10-9 2x10-10 5x10-lx10-6 '2x10-8 4x10-10 7210-8 3x10-3 lx10-10 CSB Al ternative Pos 2x10-8 4x10-10,1 x10-I 4x10- 9 8x10-2x10-12 2x10- [6x10-S2x10' j
Assuming the Probability of Failure to Close = iO-I/ Demand for Large Butterfly Valve.
Q r
SRP 6.2.4/BTP 6-4 5x10-6 j lx10-7 2x10-9 2x10-6 5x10-8 lx10-9 5x10-5 2x10-6 1 x10- 7
=
90 Hr.Yr. Positin:;2x10-7 5x10-9 lx10-E l x10-6 2x10-8 3x10-10 3x10-8x10-8 2x10-9 CSB Alternative 2x10-8 4x10-10 1x10 I 4x10-9 8x10-I 2x10-12 2x10- 7 6x10-9 2x10-10 Positter I
Assuming the ProbabilSty of Failure to Close = 5 x 10-I/ Demand for Large Butterfly Valves SRP 6.2.4/BTP 6-4 5x10-5 3x10-6 2x10-5x10-6 2x10-7 5x10-9 5x10-4 5x10-5 5x10-6 90 Hr.Yr. Position 2x10 2x10- #
2x10-lx10-6 4{j0-8
),) 0- 9 3x10-5 3x10-6 3x10
~
2x10 4x M NO 4xM 8xM 2xi[ 2x10 6M 2xd CSB Alternative 7,; tion-
... ~ -.
Probability of Specific End Effects - /Reac r Ye k h fr-o(
f N
plkf spp
l
_e.-
Cdnt.
Cont.. Cont..
Cont.
REMARKS
'or i sol a tion i sol ation i sol ation, Cont.
LOCA Purge Con t.
olatiorisolation 'Effget of occurs system isol in use signa A signal B ecleo A valeo B ealce A voleo B operatoe actior i
)
fails fails fails to fails to closos closos on pathway to close close but leaks but leaks environs YES YES m DPWOI YES M DPWOI YES DPWOI NO DPWOI YES YES LPWOI YES NO NPWOI YES LPWOI YES NO NPWOI YES LPWOI k
NO NPWOI E
N I
LPWOI
-3 NPWOI j
YES NPWOI
_^'~__}
No
- r YES NO NPWOI A
N DPWOI YES
_ JQ _ : e DPWOI ggg;;.-;.
DNOI NO DPWOI
- k. T*[NM YES LPWOI 1
-. c.;- ; g NO NO NPWOI esesses.....,
,-.--w-._
YES LPWOI km/h84w; NO NO NO NPWO!
E 4
ummumme Asesesse.....
vr LPWOI s
YES h
NO LPWOI T
YES NPWOI NO,
NO NPWOI NO k*
NPWO!
80805005308333 g
YES I
N
% Direct Path
. - _O __
SEE i
i
' FIG. 2 -
Laak Peth
,,,m sensessesse..e No Path -
4 DPWOI - Direct path with ' perator intervention to close containment isolation valve (s) o f,
LPWOI - Excess leakage path with cperator intervention to close cor.tainment isolation valve (s) g NPWOI - No path with operator intervention to close containment isolation valve (s) i FIGURE 1
~-
LOCA Purge Con Co'n2. '
' Cont.
' Cont.
nt.
' Cont.
REMARKS occur!
system isol a 61or isolatic 11solatior isolation isolatior isolation Effect of in use signal A signal B valve A valve B valve A valve B operator cctior fails fails fails to fails to closes closes on pathway to close close but leak! but leaks environs YES See
, - - - - FIG. 1 YES DPWOI YES YES NO DPWOI i
Y S
DNOI NO
-NO DPWOI YES YES N
l YES LPWOI l
YES m
r I
l YES YES O
I
"".ars..n
- PWOI
=
"'m"'"'i NO l
NO NPWOI lCEEEEEE800ggg
~
I YES
(
LPWOI j
l NO NO NPWOI l
l 5
a nnMu,,
M
=~
no l
N0 sununansi NO NPWOI Asunnune
}
YES DPWOI
..n ;..,.
?hyNh I
i NO V"
DPWOI
~, -
>. $~&varp-'t
[
YES DPWOI
- > y c~,e..
- -4 l'-h dh LPWOI L'~
E k
l NO a n,, g,,,,,
NPWOI L-* Ash 4G l
m gg
-~
~~
NO NO NPWO!
..un.n.n.,
)
YES LPWOI I
LPWOI I
NO ussied4n.
NPWOI I
NO I'"""""'?: n n e'.i.0 NPWOI sues i
YES LPWOI NO I"'"'""'
......E1,..
NNOI,
,o
[
esusniassu:nsessen9 ens NPWO!
N amuummum Direct Path r:
- Leak Path enseessusu No Path DPWO! - Direct path with operator intervention to close containment isolation valve (s)
LPWOI - Excess leakage path with operator intervention to close containment isolation I
valve (s)
NPWOI - No path with operator intervention to close containment isolation valve (s) l FIGURE 2
.. ~..
Probability of Short Term Long Tom LOCA with ECC" s ils ECCS Fails a
Direct Path to Environs Direct Path LOCA with Major Fuel Damage l
l VES Direct Path LOCA with Major
(
Fuel Damage NO NO Direct Path Mitigated LOCA O
Probability oi Short Term Long Term i,
~'
LOCA with ECCS Fails ECCS Fails Excessive Leal Path to Environs V't Leak Path LOCA with Major Fuel Damage
_ ___ 9 ", - ? c- [.
+
bb("h'$NIN-h
~
h;*e - gez +'i d r:
% eak Path LOCA with Major YES L
-4 FN-Fuel Damage
~j L_-#NGd?f6 t
., i:-~.. p-;
b Ahdq-O NO Leak Path Kitigated LOCA
~"
(not a high risk event)
FIGURE 3 -
m I
N, d