ML20153G209
| ML20153G209 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 04/30/1988 |
| From: | NORTHEAST NUCLEAR ENERGY CO. |
| To: | |
| Shared Package | |
| ML20153G190 | List: |
| References | |
| NUDOCS 8805110169 | |
| Download: ML20153G209 (54) | |
Text
D_ocket No. 50-245 812132 l
1 l
(
l Millstone Unit No. 1 l
Probabilistic Risk Assessment of Various Test Methods and Design Options l
Concerning Specific Containment Penetrations 1
i l
l l
l l
l l
l l
April 1988 l
8805110169 000429 ADOCK 050 g 5 l
PDR l
P l
Attachmont 2 t
B12132/Page 1 1
Containment Penetration Evaluations Descriptkn Millstone Unit 1
is requesting exemptions on certain existing penetrations in containment tha do not fully meet the requirements of Appendix J of 10CFR50.
Following any core related
- accident, containment integrity plays an important role in the retention of primary system radioactivity.
The degree to which containment integrity can be maintained determines the overall public risk associated with such accidents.
Appendix J is concerned with requirements on the leak-tightness of penetrations.
The purpose of the following calculations is to establish a basis for the comparison of risk posed by various test methods and design l
options concerning specific containment penetrations.
A l
l quantification of the residual public risk, permitted by the exemption l
request, is presented for each penetration.
Methodology and Major Assumptions l
All of the affected penetrations were evaluated to quantif) t.he impact 1
on public health and safety due to any exe=ption being granted.
j Public risk is explicitly calculated using PRA models which are based on the following assumptions.
i 4
L
B12132/Page 2 l
i 1.
It is conservatively assumed that containment will remain intact l
following all accident sequences.
This assumption is made in i
order to calculate the maximum public risk reduction that coulc be achieved by being in full compliance with Appendix J.
As noted in 1
the Hillstone Unit 1 PSS, the total core melt frequency of 1
i 5 7 x 10-4/yr is attributable to accidents caused by fires and internal events.
The PSS quantified all types of accident events, including thoce which go beyond design basis such as station blackout.
2.
The off-site public consequence _ due to an unmitigated core melt l
accident with containment failure is estimated to be 3 x 106 l
man-rem.
This valua represents an upper bound in the public safety l
impact of any core melt accident since it is based on an early t
)
l core melt with failed containment.
j t
I i
l 3
For valves on penetrations that do not comply with Appendix J type
{
4 i
i C testing, a leakage rate of 10% containment volume per day is assumed for each valve.
This value is approximately eight times higher than the maxinum allowable containment leakage rate of 1.2%
l per day for all valves.
l 4.
Where testing on valves fully complies with Appendi:2 J, the rate of leakage is assumed to be:
i 4
1 i
a) the average of the as-found leak rate for t he. last S l
local leak rate (LLRT) tests or i
t
B12132/Page 3 b) the administrative limit of 18.8 sofh, whichever is greater.
5.
For systema that are water sealed (i.e. LPCI, Core Spray), it is conservatively assumed that 8% of the water leakage will flash to steam and create airborne activity.
This is based on the assump-tion that drywell sump and torus water be remain at Tsat condi-tions and 43 psig.
6.
Normally closed manual valves, check valv6s, or remote operated valves in series with a particular valve of interest are assumed to reduce leakage in the pathway by a factor of five.
In other-words, the valve leakage probability is 0.2 given that the valve closes or is already closed.
No credit is taken for leak
[
isolation by closing normally open manual valves.
a J
7.
A reduction factor of 10 is applied to containment release paths I
through water because of filtration, scrubbing, carry-over fractions, and plate-out.
l 8.
Systems that are closed inside and/or outside containment need a I
breach in their boundary for a release of radioactivity to the j
environment.
For systems where euch a breach would be detectable prior to an accident, no breach is initially assumed.
Following i
I an accident, seismically qualified piping is assumed to remain
)
l intact but nor.-seismic piping is assumed to rupture with a proba-bility of 0.1.
.t'
,j;
+
-t B12132/? age 4 9
4.11 public risk calculations assume that if a va'ive were to be tested according to Appendix J,
there would be 'zero leakage 9
through the valve as a result of testing.
This assumption only-applies 'to those valves that are n_ot currently tested lto Appendix
( J requirements, c
- 10. The following failure rates are used for valve failure to close'on demand '5ased on the Millstone I.P.S.S.):
s o
Motor operated valve 3 x 10 3 demar.d
/
o Check valve 6.6 x 10-4/ demand l
t o
Air operated 'talve 3 75 x 10-4/ demand I
o Solenoid operated
{
l process valve 1.25 x 10-3/ demand
)
l 11.
The following failure rates are used to determine the proba-bility of external leakage on pumps and valves.
o pumps 6.36 x 10.6/ hour
[
o clieck valves 6.5 x 10-8/ hour o
air operatec valves 3 8 x 10-7/ hour l
o all other valves 6.9 x 10-8/ hour Quantification A.
Penetrations X-18 and X-19:
Drywell Sump Line Isolation Valves
.Dwg.
No. 25202-26010 Sheet 7)
S-
r Attcchm:nt 2 B12132/Page 5 Presently, there are no test connections to permit Type C Appendix 1
J testing of the cont:inment isolation valves in these penetra-tions.
Each penetration has 2 air operated containment isolation valves associated with it.
a) X-18:
air operated valves SS-3 and SS-4 in series 1
outside cor.tainment 4
b) X-19:
air operated valves SS-13 and SS-14 In series out-side containment The isolation valves associated wf.th each penetration are normally closed unless the sumps are being pumped down.
Plant records show that either sump could be pumped down as often as every 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> when the plant is at power- (i.e. surps are pumped o'2t on periodic basis to determine drywell leakage rate).
Therefore, it is conservatively assumed that both sets of isolation valves could be open when an acci-dent occurs.
For design basis LOCA events, the drywell sumps would remain covered by water snd only a liquid type leak could occur.
All other accidents may result in a dry sump condition which could produce a gaseous type leak. To simplify the analysis, it will be assumed that all accidents lead to a dry sump condition.
This will produce a bounding estimate of public risk.
4
.\\ttichment2-B12132/Page 6 The quantification assumes that any leakage associated with the 2 sets of penotration isolation valves could either be external or ihternal to the valves. Consequently > two scenarios are presented.
o External Valve Leakage In order to get a radioactive release to the environment during post core melt conditions, either one of the two isolation valves on each penetration would have to leak externally.
It can be shown that the worst case occurs when the valve closest to the drywell (i.e. first valve). leaks.
If the first valve does not leak externally, then significant external leakage on the second valve could only occur if the first valve leaks internally or fails to close.
'This is represented by the three cases discussed below:
Case 1 the first air operated valve leaks externally.
Although no LLRT testing is performed on the valves, external leakage can be quantified by taking credit for the contain-ment integrated leak rate test which is perfortaed 3 times per'Techhical Specifications.
over a 10 year period as Assuming that the longest period between tests is 5 years, the external leakage failure rate from assumption 11 and the untested valve leakage rate from assumption 3, valve leakage is quantified by:
t
.c
,.e a
o Attachmsnt 2 3
B12132/Page 7 i
e-e P(A0V1_ leakage)'=1/2(38x10'-7/hr.)(Syr.)(8760hr./yr)(0.1dh
= 8.32-x 10-4 x
Using the public ' safety impact model developed for ISt.P, the public risk is:
4 i
PR = CMF x T x C x P(A0V1 leakage) x Ri where CMF is the total core melt frequency of 5 7 x 10-4/yr, from the Millstone Unit 1 PSS; T is the remaining 25 years of operation left on the plant: C is the offsite consequence as per assumption 2; p(A0V1 leakage) is defined above; and Ri is a reduction factor of 0.1 used to credit the SBGT system in the Reactor Building.
Substituting timese values into the above equation for public risk yields:
PR = (5.7 x 10-4/yr. )(25 yr.)(3 -x 10+6 man-rem)(8.39. x 10-4)(0.1)
G6 3 56 man-rem
=
Case 2 - the second air operated valve leaks externally and the first valve letalis internally.
Again, the integratsd leak rate test on contsinment can be ~ credited for detecting external leakage on the second' valve.
Internal leakage is assumed for the first valve _ (as per assumption 3) and the second valve is credited with decreasing leakage by a factor A
1
-,u.
..n--
..r.-
t
+
3:
1 B12132/Page 8
- e
- t of 5 (seo assumption 6).
External; leakage from 'the second 4
v3.1ve is quantified by:
- st 1-
,.g
.p(A0V2 leakage) = (0.10)(1/2)(3.8 x 10-7/hr)(5.yr.)(6760 hr./yr.)
(0.2)
= 1.66 x 10-4 Public risk is determined by:
,PR - (5.7 x 10-4/yr.)(25 yr.)(3 x 106 man-rem)(1'.66 x 10-4)(0.1)
= 0 715 man-rem where all values, except 1.66 x 10-4 for p(A0V2 1%v. age),
were previously defined in Case 1.
3 1
Case' 3 - the second air operated valve leaks externally and the first valve fails to close.
The 'first valve has a -
failure rate as per asstanption 10 and external leakage for the second valve is assumed equal to that for the first (i.e.
8.32 x 10-4 from Case 1).
Valve external leakage is~ quanti-fied by:
- f P(A0V2 leakage) = (3 75'x 10-4)(8 32 x 10-4)
= 3 12 x 10-7
[I a
m.
+
+
yr Attcchm:nt 2 B12132/Page 9
.c y
1 g
4,
g Public risk is determined by:
1 Q
PR =-(5 7 x 10-4/yr.)(25 yr.)(3 x 106' man-rera)(3 12 x 10-7)(0.1)
' 1.29 x 10-3 man-rec 2
=
+
a y
The maximum positive impact that Appendix J Type C testing can have on ' valve ~ external ' leakage is - to reduce the public risk by 4.28 man-rem for eachisot of isolation valves.. This*
is based on assumed zero leakage followinga the implementation of such testing as noted by assumption 9 o
Internal Valve Leakage Internal leakage through either pair of isolation valves could result in a radioactive release to the envircnment if i
there is a breach in the downstream radwaste. system. -In this analysis, it is conservatively assumed that system leakage automatically occurs even if the non-seismic piping - does Mot -
rupture post accident.
Public risk is quantified for valve 4
internal leakage by the following three cases, t
Cs,se 1 - the first air operated valve leaks internally after closure and the second valve also leaks after closure.
j 1
Leakage through the first valve (closest to the drywell) is assigned a value of 0.10 as per assumption 3 The second valve is credited with reducing leakage as per a:sumption 6 l
l o
q A
~
s N.
Atttchm:nt 21 B12132/Page 10 a
.i.
g' F
' I and is -assigned a' value of 0.2.
Consequently, the proba-
. bility of leakage through both isolation valves is gifen by:
. s.
{
Y}
P (A0V1 & A0V2 leakage) =>(0.1)(0.2) g
= 2.0 x 10-2 f
- )
t' Public risk as deterinined1by. the public safety impact 'model C
is shown below:
R
' PR = CMF x T x C x p(A0V1 & A0V2 leakage) x R1xR2 i
6 s
where CMF, T and C are defined in Case
.1 for external' leakage.
The probability of leakage, p(A0V1.& A0V2 leakage),
is,given above and Rj represents the reduction factor (0.'1) per Assumption 7 Reduction factor R2 is assumed to be 0 33 due to the long tortuous path that radioactive material has e
to travel through before it leaks out of the-radwaste system.
It should be noted that R2 bounds the ' probability of piping rupture which is 0.t as per assumption 8..
The resulting pubMc risk is thus:
PH = (5 7 x 10-4/yr.)(25 yr.)(3'x 106 man-rem)(2.0 x 10-2 (0.1)(0 33)
= 28.2 raan-rem f
1
.0
r 1
Attech nt 2 B12132/Page 11 Case 1 - first air operated valve leaks internally and the second valve fails to close.
As noted in Case 1 above, valve internal leakage is assumed to be 0.10 and valve failure to close is 3 75 x 10-4 from assumption 10.
The probabilit!y of internal leakage through both valves is given by:
(A0V1 & A0V2 laakags) = (0.10)(3 75 x 10-4)
= 3 75 x 10-5 Public risk is determined by:
PR = (5.7 x 10-4/yr.)(25 yr.)(3 x 106 man-rem)(3 75 x 10-5)
(0.1)(0 33)
= 5.16 x 10-2 man-rem Case 3 - the first air operated valve fails to close and the second valve leaks internally.
This is identical to Case 2 above except that the order of failures is reversed.
The public risk for Case 3 is, theratore, equal to 0.0516 man-rem.
l The total public risk due to tne internal valve leakage associated with each set of isolation valves is 28 3 man-rem.
Combined with the public risk due to external leakage, the maximum possible risk reduction for each set of isolation
s
,y
. g Attachm*nt 2 B12132/Page 12
?
valves is 32.6 cman-rem.
However, it should be noted that, the public risk reduction would only be 3 3 man-re.n if water r
were present in the drywell sump; i.e.,
factor of 10 reduction due to path through water).
3 B.
Penetrations X-204A, B and C:
Torus.to LPCI/ Core Spray Pump Suction Line Isolation Valves (Dwg. No. 25202-26008)
All of the above torus penetrations are part of closed loop systems that are sealed with water.
Each penetration on the torus communicates with a common header which supplies water to 4 LPCI and 2 Core Spray pumps.
Presently, there is no provision to allow Appendix J Type C testing of containment isolation valves on the suction side of each pump.
The motor operated isolation valves v
that are associated with penetrations 204A, B and C are as follows:
o MOV 1-LP-2A (LPCI pump A suction) o MOV 1-LP-2B (LPCI pump B. suction) o MOV 1-LP-2C (LPCI pump C suction) o MOV 1-LP-2D (LPCI pump D suction) o MOV 1-CS.2A (Core spray pump A suction) o MOV 1-CS-2B (Core spray pump B suction)
Following an accident the torus will remain filled with water.
If not, then it may be assumed that primary containment (i.e. the torus) has failed and the question of valvs leak-tightness is 9
4
'a
--m--,-
-. - ~
w
~,.,,,,
,n.-
e Attcchmsnt 2 B12132/Page 13 immaterial._
Conseduently all valve leakage, whether externalfor r
internal, is assumed to be of the liquid type and no ' direct
-+
gaseous'. type leaks are assumed for theso : valves.
Quantification of public risk 'is based on both external and internal valve
\\,:
leakage as described below.
i o
External Valve Leakage In order to get a
radioact.1ve release outside containment, any one of the 6 isolation valves would have to first leak externally.
Radiation would largely remain confined to the Reactor-Building unless significant airborne activity were generated asz a result of the leakaga.
It is assumed (see assumption 5) that as mucli as 8% of the radioactive water would flash to steam and create a release to the envirordnent.
Although it can be shown that the MOV's in question are constantly exposed to approximately 5 psig pressure, no eredit is taken for this demonstration of leak tightness on a continuous basis.
The probability of valve external b
leakage is quantified by taking credit for the contain-ment integrated leak rate. test described in Part A.;
Given that a valve leaks externally, the untestod valve leakage rate from assumption 3 is assigned and 8% of the leakage assumed to go airborne..
The probability of external MOV leakage is given by:
i t
i.
(,;;
- 7
- 5
-Attachmsnt 2 3
B12132/Page 14 U
s p(MOV leakage) = (1/2)(6.9 x 10-8/hr.)(5yr.)(8760 hr./yr.)
(0.10)(.08)
= 1.21 x 10-5 0
t where 6.9,x 10-8/hr 1s the eaternal valve leakage rate,.5 years is the ' assumed time between containment': integrated leak rate tests (i.e.
not LLRT's), 0.10 is the untested valve leakage rate, and 0.08 is the fraction of leakage which goes airbcene.
2 Using the public safety impact model described earlier, the public risk is determine'd by:
PR = CMF x T x C x p(MOV leakage) x Ri where CMF = 5.7 x 10-4/yr., T = 25 years, C = 3 x 106 man rem as discussed in Part A.
The > probability of MOV leakage is 1.21 x 10-5 from above and the reduction 1
factor through water, Ri, is 0.1 as per Assamption 7 I
Substituting thene values, 4
PR.= (5 7 x 10 /yr)(25 yr.)(3 x 106 man-rem)(1.21 x 10-5)(o,1)
= 5.16 x 10-2 man-rem for each of the 6 MOVs u
0
.g
v-n.
=
Attachmsnt 2 3
B12132/Page.15 o
Internal Valve Leakage '
Since the LPCI _and Core Spray systems are closed, internal valve leakage is meaningless 'without some addi-tional ' breach in - the system.-
Assumption 8 notes that seismically qualified piping. is' assumed to remain intact s.
post core melt and, therefore, breaches in the_ piping are not postulated.
L 3
Components that are located downstream of the pump suction MOV's include pumps and other valves, in that 4
order. -Allvalvesareconsta$tlymaintainedatpressures-above that required for Appendix J Type C testing.
For example, the keep-full systems for LPCI and. Core Spray maintain pressures of 45 psig and 150, respectively.
If external valve leakage were to occur it would be detected by:
a) Reactor Building sump trending I
b) Once per shift. walk-throughs by plant equipment operators
~
c) Monthly surveillance testing'of the systems Consequently, valve external leakage is not postulated downstream of the MOV's ira question.
However, because
'i.
+
. y,s Attachmsnt 2 B12132/Page 16
+
there is check valve on the discharge side of each pump, the pumps are never exposed do the full keep-fill system pressure.
In the-event of MOV internal leakage post s
accident, it is assumed that a systsm breach could take place via the LPCI and Core Spray pumps.
Following an accident, it is ' highly unlikely that the valves on the low-pressure pump suction lines would ever be closed.
- However, to. determine the impact of Appendix J on internal valve leakage, it will be assumed that the pumps are shut down and isolated after external pump leakage is observed.
The probability of internal valve leakage leading to a radioactive release is quantified as follows:
z Leakage through the valve could either be due to its,
failure to close, which would result in 1'005 ' leakage (i.e. factor of 1.0), or its inability to remain leak-tight.
For the latter conditian, an untested valve i
leakage rate of 10% (i.e. factor of 0.10) is assigned per i
assumption 3 given that the valve closes.
This can be expressed probabilistically by (0.10 +
1.0" 3x 10-3) where the valve failure to close probability is taken by assumption 10.
The probability of external pump leakage is determined by assumption 11 and an assumed mission time of 30 days (720 hours0.00833 days <br />0.2 hours <br />0.00119 weeks <br />2.7396e-4 months <br />).
It is also assumed that 8%
l i
)
Attichm:nt 2 B12132/Page 17 of the leakage goes airborne (assumption 5).
The proba-bility of internal leakage, system breach, and airborne activity is given by:
p(release) = (6 36 x 10-6/hr)(720 hr.)(0.10 + 1.08 3 x 10-3)
(0.08)
= 3 77 x 10-5 Public risk is determined by:
6 man-rem)(3 79 x 10-5)(9.1)
PR = (5 7 x 10-4/yr.)(25 yr.)(3 x 10
= 0.162 man-rem for each of the 6 MOVs Altogether, the combined public risk associated with internal and external leakage from each MOV is 0.214 man-rem.
If it is assumed that an Appendix J Type C test of the MOV's in question would result in zero leakage, the above public risk of 0.214 man-rem per MOV could be averted.
C.
Penetrations X-210A and B:
Torus Test Line to LPCI/ Core Spray Pump Discharge Side (Dwg. No. 25202-26008).
The above penetrations are associated with closed loop systems that are sealed with water as described in Part B.
Each penetra-4
7
+
. c ;. f. -
Attachmsnt 2 B12132/Page 18
+-
c.
L tion involves a test line that goes through the top section of the m:
torus and termitiates at point below the surface of the suppression pool water.
,At the present time', there is no provision to allow-Appendix J : Type C testing of the isolation valves on the test o
line. The valves associated with penetrations X-210A and B are:
I 4
o MOV 1-LP-43A (LPCI A test line to torus) o MOV 1-LP-43B (LPCI B test line to torus) o' CV 1-LP-24A (LPCI pump A mini-flow line) o CV 1-LP-24B (LPCI pump B mini-flow lina) o CV 1-LP-240 (LPCI pump C mini-flow line) o CV 1-LP-24D (LPCI pump D mini-flow line) o CV 1-CS-14A (Core spray A torus teot line) f o
CV 1-CS-14B (Core spray B torus test line) 3 l
Because of the different configurations involving the above isolation valves, they are analyzed in separate groups as described below:
1.
LPCI Test Line to Torus - MOVs 1-LP-43A and B; Following an accident event in which the normai shutdown cooling i
systems (e.g.
main condenser, isolation condenser and shutdown cooling) are all unavailable, the LPCI System is used to remove long term decay heat.
One of the cooling flow path options is to align the output side of the LPCI heat exchangers to the torus via i
the test line.
In-this particular line-up, torus water is removed I
p
-O
)
s
Attachm:nt 2 B12132/Page 19 by the LPCI pumps and forced through the heat exchangers where it is cooled and returned to the torus.
It is extremely doubtful whether a test line would ever be isolated after being lined-up for use in a post com melt situa-tion.
The only credible situation where such isolation might take place 10 after all. 4 LPCI pumps failed to continu:e running after they started.
Because of the LPCI crosstie, discharge flow from all 4 pumps and their associated heat exchangers would be directed through one of the two test lines.
Thus, only one test line actually needs to be considered.
For this analysis, thougn, it is assumed that both test lines would initially be placed into service and then isolated post core melt in order to evaluate any potential benefits of Appendix J testing.
Quantification of public risk is based on both external and internal valve leakage as outlined below.
o External Valve Leakage Appendix J Type C testing during refuel outages would produce no measurable impact on valve external leakage detection.
This is because such tests subject the valves to 43 psis internal pressure while the normal keep full pressure on the valve is approximately 45 psig.
Conse-quently, the affected valves (i.e. 1-LP-43A and B) are in i
1
i q
s 1
~Y~ ~Attachm:nt 2 B12132/Ptge 20 a constant test mode for external leakage which would be -
. detected by:
t a) ' Reactor Building sump trending E
4 b) Once per shift walk-throughs by blant. equipment operators.
j c) Monthly surveillance testing with the torus.stest lines.
d.
Therefore, the public risk associated with external valve leakage is estimated to be zero.
4 o
Internal Valve Leakage i
Because the LPCI system is normally olosed, internal.
valve leakage is meaningless without some external brei.ch in the system.
Part B notes that all piping or components located downstream of the-LPCI pumps' discharge are not considered to be ' sources of en ternal l
leakage.
- Hcwever, the pumps and their associated
)
isolation valves (on the auction side)' are considered to i
be valid locat_rns for a potential system breach.
This analysis assur.es that internal leakage through HOV 1-LP-43A(43B) could result in external leakage to the Reactor Building if the LPCI pumps or their isolation
. =
Attachm:nt 2 B12132/Page 21 A
valves developed a breach.
Following a core melt acci-4 dent, such an event could only take place if all 4 LPCI pumps failed to run and the following failures occurred:
a) Leakage past - inboard MOV LP-44A(44B).
The reduction factor for this MOV is the sum of its failure to--
close probability (i.e. 3x 10-3 as per assumption
- 10) and the reduction factor of 0.2 (assumption 6),
or 0.203 I
b) Assumed internal leakage through outboard MOV LP-43A (43B). Either 1005 leakage occurs if the valve fails to close or 10% leakage occurs given that it closes.
This can be expressed by (0.10 + 1.0# 3x 10-3) as shown for internal valve leakage in part B.
c) Leakage past globe stop check LP-9A (9B).
The flow reduction factor of 0.2 is summed ~ with the check valve failure to close probability of 6.6 x 10-4 (assumption 10) for a total value of 0.201 as a reduction factor.
L d) Leakage past either one of two LPCI pump discharge check valves LP-3A ar.d 30 (LP-3b and 3D).
The flow reduction factor of 0.2 is summed with tno valve failure to close probability of 6.6 x 10 L ' for each valve.
Since leakage past either check valve will create leakage in the flowpath, the total reduction factor is 2(0.201).
1
. - -.. ~ -
9 Attachmsnt 2
~
B12132/Page 22 i
e) Leakage ' from LPCI pumps ' A and C (B and D) or the associated isolation ' valves LP-2A and 2C (LP-2B and-
^
2D).
From part B,
the probability of external leakage' for a pump is (6.36: x 10 6/hr.)(720 hr.) or 4.58 x 10-3 Similarly,'the probability of external 3
leakage for a valve is (5/2)(6.9 x 10-8/hr.) (4.28 x 4
10-3. Since external leakage can 10 /hr.) or 1.51 x be due to either LPCI pump or its associated isola-c' tion valve, the total probability of external leakage I
in the flowpath is 2(4.58 x 10-3 + 1.51 x 10-3).
The probability of, internal valve leakage (i.e. MOV i
LP-43A or B) leading to torus water leakage in the Reactor Building is quantified by failures (a) thru (e) as follows:
p(leakage) = (0.203)(S.10 + 1.08 3 0 x 10-3)(0.201)(28 0.201) s i
(2)(4.58 x 10-3 + 1.51 x 10-3)
= 2.06 x110-5 4
Using the public safety impact model which was defined earlier, public risk is determined by:
T
i 4-j- '
Attachmon?, 2 B12132/Page 23 PR = CMF x T xCxp(leakage).xR1xN2 l
where CMF=5.7 x 10-4/yr., T=25 yr., and C=3 x 106 man-rem A.
The probability _
leakage,.
as defined in part of p(leakage),;. is 2.06 x 10-5 from above.
The' reduction i
- i A
factors R1 and R2 are 0.1 for the flow path through water and 0.08 for the percentage of -leakage that goes airborne, respectively (see part B).
Substituting the above values into the equation for public risk yields:
t PR = (5.7 x.10-4/yr.)(25 yr.)(3 x 106 man-rem)(2.05 x 10-5) i (0.1)(0.08)
= 7 05 x 10-3 man-rem for internal leakage on l
Altogether, the combined public risk associated with internal ar.d external leakage from each MOV is 7 05 x 10-3 man-rem.
This value is reasonably low enough to be considered as zero risk to the public.
At best, Appendix J Type C te$ ting would avert only 7 05 x 10-3 man-rem for each valve if it.. is assumed that such testing would result in zero valve leakage.
2.
LPCI Pump Minimum Flow Line Caeck Valves 1-LP-24A,B,C and D.
These check valves are located on the LPCI pump minimum flow lines 1
l which direct pump flow to the torus via the torus test line described earlier (Part C, Section 1).
If a LPCI loop is not t
e -
n - e i
e--
- -. +
4
<4,---e-
--m.,, -
u.
m
--4--
g
AttachmGnt.2 B12132/Pa'ge 24 operating, these check ' valves can act as a containment boundary for penetrations X-210A and B.
Valven LP-24A and 24C are associ-ated with LPCI Loop A while LP-24B and D are associated with Loop B.
a In order for the check valves to be challenged as containment boundaries post accident, both LPCI pumps in each loop would have to fail.
Although highly unlikely, this analysis assumes total failure of all 4 pumps to evaluate the potential benefits of Appendix J testing on the check valves.
Quantification of public risk is based on both external and internal valve leakage as described below.
e o
External Valve Leakage
)
The LPCI system operability test, which is performed monthly, verifies that the LPCI pumps acheive a discharge pressure of exactly 80 psig.
During the course of the a
test, block valves in the minime flow line (i.e. LP-26A i
and 26B) are closed with the result that the. subject i
check valve internals are exposed to full ;nunp discharge pressure.
Consequently, all 4 check valves are "leak i
tested" monthly to pressures exceeding that for Appendix J tests which are only conducted each refuel (i.e.
18 months).
p 1:
u-
-i ;
- y
- t :
Attachmsnt 2 i
B12132/Page 25 The LPCI operability test is considere.d to be a valid 4
test-for external leakage on the miriimum flow-line check valves because:
a) Operators must be present during the test and they are asked to verify that system leakage is not' present. -
b) Abnormal. system leakage as a result of the test would be independently ' verified by Reac' tor Building' sump-trending.
.)
c) The monthly test exposes the check valves to internal pressures that are far in excess of those required for Appendix J testing (i.e. monthly test is 80 psig-vs. 43 psig for Appendix J).
For the reasons stated above, it is concluded that Appendix J testing of the check valves on the minimum flow line will produce no measurable reduction in public risk.
A zero man-rem _ risk reduction is, therefore, nasigned to Appendix J external leak testing of these valves.
1 e
I i
i l'
Attachmsnt 2.
i B12132/Page 26 o
Internal Valve Leakage
)
j 1
Section 1 in this part ' of the analysis ' notes that the LPCI system is normally closed and that ' internal valve
,e leakage alone would not produce any leakage into the.
Redactor Building.
The same section also notes that - the LPCI pumps and their isolation valves are considered to be the only sources of. a potential system breach.
Internal back-leakage from t.ho torus through the minimum 4
flow line check valves could only. present a leakage r
r problem if the following failures occurred:
)
a) Total loss of both LPCI pumps within a loop.
Unless I
.t such a failure occurred, the mini-flow check valves would not be challenged.
The analysis conservatively assumes that such an event will always occur with a r
probability of 1.0 after a core melt accident.
4 b) Internal back-leakage through any one check valve in j
l a loop.
Either 100% leakage occurs if the valve fails to close (probability of 6.6 x 10-4 per assump-i l
tion 10) or 10% leakage occurs if the valve does closa.
The probability of internal check valve leakage is given by (0.10
+ 1.08 6.6 x 10-4) as shown in part B.
u a
,y.
L 3-.
.Attachmsnt 2 B12132/Page 27 i
l
=
t c) External system leakage via LPCI pumps ' A or C (B or D) and isolation valves LP-2A -or 2C (LP-2B or 2D).
Since external leakage can be attributed to either.
LPCI pump or,its associated isolation valve within a loop, the total probability of such leakage in the loop flowpath is (2)(4.58 x 10-3 + 1,51 x x 10-3) as shown in Section 1.
The probability of internal check valve leakage that results in leakage to the Reactor Building is given by failures (a) thru (c) as shown below:
p(leakage) = (1.0)(0.10 + 1.08 6.6 x 10-4)
(2)(4.58 x 10-3 + 1.51 x 10-3)
= 1.23 x 10-3 for each check valve The public risk due to internal valve leakage is quanti-fled by using the same values that were used in Section i
1, except that p(leakage) = 1.23 x 10-3 from above:
PR = (5 7 x 10-4/yr)(25 yr.)(3x106 man-rem)(1.23 x 10-3)
(0.1)(0.08) 1
= 0.42 man-rem for internal leakage thru each check valve 3
I
Attachze0% 2 B12132/Page 28
(
The total benefit due.to Appendix J testing of the mini-flow check valves would be a public risk-reduction of 0.42 man-rem per valve.
This is based on the assumption that Appendix J
-Type C testing would guarantee zero leakage for each of the 4 4
4 3
Core Spray Test Line to Torus - Check Valves 1-CS-14A and 14B.
i These check valves are located on the Core Spray pump tost lines which direct flow to the torus via the test lines described 2.
earlier in part C.
If the Core Spray pumps do not operate post.
- i accidetit, these check valves can act as a containment boundary for penetrations X-210A and B.
Although such an event is not expected, this analysis assumes failure of both Core Spray pumps in order to evaluate the potential benefit of Appendix J testing on the valves.
Quantification of the aasociated public risk is based on external and internal check valve leakage as described below.
l i
il o
External Valve Leakage The monthly Core Spray system operability test is considered to be a valid substitute for Appendix J tests that would detect external leakage on check valves CS-14A j
and 14B. The reasons for this statement are given below:
4 4
i
- ... ~, -.
+
l Attachm:nt 2 l
.B12132/Page 29 y
a) Appendix J Type C testing subjects the valves to an y.
internal' pressure of 43 psig once per refuel (i.e.
i-approximately every 18 months).
I b) The Core Spray operability test exposes the valves to an internal pressur6 that is at least 43 ' psis or greater.
As noted by the Core Spray acceptance.
r criteria, pump discharge pressure must be in the f
range cf 261 to 298 psig with flow at 3600 spm.,
t r
h c) The operators are specifically directed by procedure' to observe check valves CS-14A and 14B during the operability test.
Any external valve leakage would be detected at this time.
i d) Abnormal system leakage would be independently veri-fled by Reactor Building sump trending.
Based on the above, it is concluded that the monthly operability test is at least equivalent to Appendix J testing for external leakage.
A zero man-rem reduction l
is therefore assigned to Appendix J testing of the eht.ck 3
' valves.
]
l l
l i
l i
1 1
.I
,.. J
..m,
m Attachm:nt 2 4-B12132/Page 30 o
Internal Valve Leakage O
It was noted earlier that the Core Spray system is closed and a keep-full maintains system pressure at 150 psig.
I Also, internal val'Ts leakage is considered meaningless without some additional breach in the system to create an cxternal flowpath.
Based on the discussion of Section 1, it is' assumed that the Core Spray pumps and their associ-ated isolation valves (i.e.
CS-2A & 2B) are the only
]
credib.'.e sources of external leakage.
l 1
Following a core melt accident, it is assumed that both Core Spray pumps fail to run.
Internal leakage of torus 3
water through the test line check valves (i.e. CS-14A &
I 14D) would result in external leakage to the Reactor Building if the following failures cocurred:
a) Internal back-leakage through the test line. check valve in either Core Spray loop.
As shown in Section 2,
the probability of such leakage is (0.10 + 1.08 6.6 x 10-4) for an individual check valve.
1 b) Leakage past valves in either one of two' lines that i
connect the test line discharge check valve with its associated Core Spray pump and -isolation valve.
In the first line, there is a motor operated valve and a j
check valve in series. The probability of leakuge I
d
Attacharnt 2 B12132/Page-31' l
through both these valves is (0.203* 0.201) where a
s O.203 and 0.201 represent the probability of leakage past the MOV and check valve, respectively (see Sec-a' tion 1).
I The second line has only one check valve whose proba-c' bility of leakage is 0.201.
The oc,mbined probability i
of leakage through either the first or second line is 3
4 (0.2038 0.201) + (.201), or 0.242.
t c
c) Leakage from Core Spray pump A (B) and its associated a
isolation valve CS-2A (2B).
From Section 1,
the i
probability of external leakage for a pump is 4.58 x 10-3 and the probability of external leakage for an isolation valve is 1.51 x 10-3 The combined proba-1 i
bility of leakage frora the pump and the valve is (4.58 x 10-3 + 1.51 x 10-3).
l The probability of internal leakage in a check valve i
(i.e.
CS-14A or B) leading to a leak in the Reactor j
1 i
Building is quantified -by failures (a) thru (c) as 2
l j
shown below:
I I
p(leakage) = (0.10 + 1.08 6.6 x 10-4);0.242)(4.58 x 10-3 l
+ 1.51 x 10-3)
= 1.48 x 10-4 I
i i
i i
~..
. ~,,,..
i
)
Att:chment 2 B12132/Page 32 Uaing the public safety impact model defined earlier, public risk is det' ermined by:
i PR = (5 7 x 10-4/yr.)(25 yr.)(3 x 106 man-rem)(1.48 x 10-4)
(0.1)(0.08) i
= 5 06 x 10-2 man-rem (for each check valve) 1.48 x 10-4, l
where all values except p(leakage)
=
from above, were taken from Section 1.
The combined public risk dre-to external and internal leakage associated with check valve CS-14A (14B) is 0.051 man-rem.
If Appendix J
testing were implemented for the check valves, this is the maximum public risk reducticn that could be achieved per valve, assumias that Appendix J testing results in zero valve leakage.
D) Penetration X-15:
Reactor Water Cleanup System Return Line Isola-tion Valve (Dwg. No. 25202-29130 Sheet 1)
The 3 actor Water Cleanuo (RWCU) system is a normally running svru 4hich o.'
tes to remove impurities from the primary system j
"' 7-
' i <.s suction from the "A" roeirculation loop and
]
Recotor Coolant System (RCS) via feedwater
<s' D
Attochme:. J B12132/Page 33 line "A" which is inside containment (i.e.
the drywell).
Pene-tration X-15 is located on the RWCU return line and can be isolated by either the check valve (CU-29) inside containment or the motor operated valve (CU-28) which is outside containmente At the present time, only the motor operated valve is Appendix J testable. This analysis evaluates the potential benefits of Appendix J Type C testing for check valve CU-29 by quantifying the public risk associated with external and internal valve leakege as described below, c
External Valve Leakage When the RWCU is operating with the plar.t at power, the internal pressure on valve CU-29 is appro.;imately 1130 psig.
Any leakage out of the valve is contained in the drywell and would be detected by periodic sump trending.
Even if the valve were leaking at the maximum rate for unidentifiable leakage (i.e.
2.5 gpm as per Technical Specifications), this would correspond to an insignifi-cant atount of external leakage at 43 0-18 (i.e.
Appendix J
test pressure).
Consequently, proposed Appendix J testing for external leakage or vtive CU-29 produces no reduction in public, risk.
y
. Atttehaznt 2
. B12132/Page 34-I o
Internal Velve Leakage 9
'+'.
The RWCU is considered to be a closed syste'n which.is sealed with water.
During plant operation, the system is i
i
+
i cer.tinuously running and any leakage out of the system i
would be detected as follows:
a) Once per shift walk-throughs by pl'nt equipment 4
-s*
operators.
i b) Reactor Building sump trending c) RWCU system 1 solation due to grosa, leakage.
The I
system automatically isolates on signals of low flow through the filters or high temperature following the 1
non-regenerative heat exchangers.
Either one of l
)
these signals would be generated by excessive leakage out of the system.
1 i
l All of the RWCU system from penetration X-15 back to the cleanup recirculation pumps meets seismic requirements.
This section of the system is also designed for 1250 psig
^
operation.
If a break were to occur in the icw pressure i
piping upstream of the cleanup pumps' discharge, I
containment isolation for penetiation X-15 could be achieved at multiple locations. Isolation valves on the 1
.m I
..g.
sAtttchment 2 B12132/Fase 35 high pressure secti?n.of. RWCU can be used - to isolat'e.the I
i system at the points described below:
+
f a) Check valve CU-29 inside the drywell near penetratio X-15.
s J
+
b) Motor ' operated valve (MOV) CU-28 located outside, the
- 1 L
drywell near penetration X-15.
t I
q c) Check valve CU-27 upstream of the regenerative heat i
exchangers but downstream from the cleanup pumps.
l 1
d) Check valve CU-22 on the bypass line and MOV's CU-20A and B on the two cleanup pump discharge lines.
Each of these valtes can isolate one of the 3 lines where 4
i j
it connects to the main flow line upstream of CU.-27 i
l Since only one cleanup pump is operating when the RWCU is in service, either MOV CU-20A or CU-20B is closed.
j I
e) Pump discharge check valves.CU-19A and-19B along with j
normally closed M07 CU-21 on the bypass line.
Thed j
valves are located just upstreans of the valves that were described for ;) above.
5 I
Following a core melt accident, internal leakage t!. cough ~
I check valve CII-29 would not automatically result in a
'I m
m-7 Attnorment 2.
B12132/Page 36
.i release of radioactivity.
Leakage would have to be present at each of the 4 points (tescribed above and there t
would have to be a breach in the low pressure piping tts I
1-well.
The-hWCU system is assumed to be closed -and l
leakage fres prior to any accident and only the non N i.
seismio low pressure piping is assumed to rupture post
' accident (see assumption 8).
It is further assum d that the reactor vessel contains only water and no gaseous leakage would occur from the RWCU system.
If the vessel-l were devoid of water, then it can - be assumed that all-ECCS injecticn has failed and containment failure is 3
8
)
imminent.
1
+
In order for internal leakage through check valve CU-29 a
to result in reactor water leakage out. of the RWCU I
system, the following failures would have to occur:
i a)
[nternal leakage from the reactor vessel through j
check valve CU-29 The probability of internal leakage is (0.10 + 1.0# 6.6 x 10-4)'from section 1 of
~
Part C.
i i
b) Leakage past MOV CU-28.
Since this valve is tested by Appendix J Type C tests, the rate of leakage is taken from assumption 4 (i.e.
18.8 sofh or 0.046%
)
]
leakage).
Either 100% leakage occurs if the valve fails to close or 0.046% leakage occurs given that i
j
,-o.
y
+
'I Atttchment 2 B12132/Page-37 the valve does close.
The probability of leakage is-(4.6 x 10-4 + 1.0' 3x 10-3) as shown for internal valve leakage in Part 3.
4 c) Leakage past check valve CU-27.
The probability of leakage. past the check valve is 0.201 as shown in
~
Section-1 of Part C.
j
'd) Leakage past valves in any one of 3 lines that feed into the main RK'CU return line upstream of check valve CU-27 The first line has check valve CU-22 e
and MOV CU-21 in series.
Both valves on this bypass-i t
line are normally closed and can be assigned a reduc-j tion factor of. 0.2 each as per assumption 6.
The j
j probability of leakage through both valves is thus (0.2)2 op o,04, i
I Leakage could also occur on the non-running pumps discharge line which also has a normally closed check valve (CU-19A or 19B) and a normally closed MOV (CU-20A or 20B) in series.
The probability of leakage through both closed valves is 0.Gh from above.
The i
i third line has an "open" check valve (CU-19A or 19'J) f since it contains the pump that is running price to j
the accident.
As noted above, the probability of
^
back leakage past a check valve is 0.201. The total i
y 4
.+
4 C
. Lttschment 2 B12132/Page 38 probability of leakage due to the 3 lines can be i
expreseed as (0 34 + 0.04 + 0.201) or 0.281.
+
e) Leakage out of the low pressure section of the RWCU syst'em is assumed to occur with a probability of_ 0.1 as per assumption 8.
The probability _ of check valve CU-29 internal leakage that results in RWCU external leakage is quantified by i
failures (a) thru (e) as follows:
4 i
.j p(leakage) = (0.10 + 1.08 6.6 x 10-4)(4.6 x 10-4 + 1.08 3 x 10-3)(0.201)(0.281)(0.1) 4 4.
i
= 1 97 x 10-6 t
i
+
Using the public safety impact model whxch was defined earlier., public risk due to internal. check valve leakage
^
l from CU-29 is detern.ined by:
i PP = CMF x T x C x p(leakage) x R1xR2 Where CMF=5.7 x 10-4/yr.. Tr25 yr., and C=3 x 106 man-rem as W
defined in Part A.
The probability of leakage,. p(leakage),
is 1 97 x 10-6 from above.
The reduction factors R3 and R2 4
j are 0.1 for release path through water and 0.08 for the O
percentage of leakage that goes airborne, respectively - (see I
l l
i
.l m
w B12132/Page 39 Part B).
Substituting these values into the above ec,uation for public risk yields:
PR = (5.7 x 10-4/yr.)(25 yr.)(3 x 106 man-rem)(1.19 x 10-6)(0,3)(0,08)
= 6.73 x 10-4 man-rom Based on the above, the maxteum expected risk reduction that Appendix J testing of CU-29 would provide is 6.73. 10-4 man-rem for bath external and internal leak testing.
This value is low enough to be considered a zere-reduction in public risk.
E) Penetrations X-23 and X-24:
Reactor Building Closed Cooling Water (RBCCW) Isolation Valves (Dwg. No.
25202-26006).
The RBCCW system is in continuous operation while the plant is at i
power, circulating cooling water to heat loads inside and outside of containment.
The two penetrations of concern involve an isola-tion check vaire (RC-6) inside containment on the inlet line at X-23 and a motor operated valve (RC-15) outside containment on the return line at X-24.
Presently, the check valve does not meet Appendix J requirements.
Additionally, the MOV does not meet Appendix J Type C testing requirements.
y Attcchtent 2 B12132/Page 40 The purpose of this analysis is to determine the potential bene-fits of being in full compliance with Appendix J on check valve
{
RC-6 and MOV RC-15 The public risk reduction associated with full compliance is quantified separately for each valve as shown below.
1.
Penetration X-23:
Check Valve RC-6 This analysis will quantify the public risk reduction that is associated with being in compliance with Appendix J.
The potential benefits of full compliance are quantified below based on eliminating external and internal valve leakage for RC-6.
o External Valve Leakage The RECCW system is a closed loop operating at an internal pressure of 150 psig inside containment.
Any external leakage from check valve RC-6 would be detected in the following manner:
a) Unidentified drywell Jeakage based on sump trending.
As noted in Part A,
the drywell sumps are pumped doun approximately every 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> to determine leakage.
If 7 7-6 developed a leak, it would be noted by increasing
B12132/Page 41 e
e drywell leakage which is limited to 2.5 gpm by Technical Specification.
b) RBCCW surge tank level.- Any decrease in tank level, due to check valve external
- leakage, would be noted.
Decreasing surge tank level in conjunction with increasing drywell leakage would indicate a breach in the RBCCW system.
Based on the above, the RBCCW system is in a constant "leak test" mode at pressures that are significantly higher (i.e.
3 times) then those required for Appendix J Type C testing.
Appendix J testing is only performed once per refuel (i.e. 18 months) and offers no advantage over the present on-line method of leak detection.
Consequently, Appendiy J testing produces no measurable reduction in public risk due to external check valve leakage.
o Internal Valve Leakage Internal leakage though check valve RC-6, by itself, would not pose a protelem unless the RBCCW loop insido containment were breached and additionel valves also experienced internal leakage.
If this were the case, containment atmosphere could possibly enter the BBCCW system and travel through the system outside containment, eventually making its way to the surge
~
Attochment 2 B12132/Page 42*
.y c.-.
i tank vent.
As noted in the preceding - analysis, the RBCCW I
J system is in a constant Jeak test mode and any system breach would be detected prior to an accident.
Following an i
accident, either the RBCCW pumps continue to run or they are i
immediately restarted as per procedure for loss of of.* site i
power events.
System leakage which could possibly develop b
post accident inside containment would, therefore, be out of the system and not into it.
{
The RBCCW system is seismically designed inside and outside of containment, except for the' drywell air handling units.
This analysis assumes that a RBCCW system breach would only j
i occur due to a saismic induced failure of all air handling l
j units inside containment, following a, design basis earthquake l
(DBE ).
The DBE is also assumed to cause a loss of offsite
(
4 power and the failure of other non-seismic equipmeat.
l 1
However, it is assumed that all seismically designed equip-4 ment would be available post event.
The Millstone Unit 3 PSS show that the frequency of a 0.17g J
DBE is 2.04 x 10-4/yr. for the Millstone site.
The DBE frequency is used along with the loss of offsite power (LNP) event tree from the Millstone Unit 1 PSS to calculate the i
frequency of core melt due to such an event.
This approach is considered to be more of a best estimate than using the core melt frequency attributed to all internal events (i.e.
I 1
t
x
+
1&
B12132/Pago 43 5 7 x 10-4/yr.) and applying - Assumption 8 for the RBCCW air
.j I
handling units.
i The LNP event tree for Millstone 1 was simplified to reflect
]
the assumed failure et. non-seissio equipment. - The dominant core melt sequences from this tree are given by the following Boolean expression:
i i'
(N# LTC) + (FW# 38 LTC) + (FW8 LP) f where FW represents the unavailability of the Feedwater/FWCI system, LP is the unavailability of the low pressure ECC l
systems (i.e. LPCI and Core Spray), and LTC is the unavail-
{
ability of long term cooling.
The unavailability of these 4
systems was given a factor of 2 increase to account for any
~
]
decrease in reliability following the DBE and the results of I
this are shown below I
j i
FW = 0.2 and 5 = 1.0 - FW LP = 0.02 and H = 1.0 - LP J
LTC = OJ4 and LTC = 10 - LTC l
Substituting these values into the above equation yields the I
i conditional probability of core melt!
v
(
i I
..c -
Attcchment 2 B12132/Pcgs 44 p(CM) = (0.80' O.04) + (0.2# 0 985 0.04) + (0.2e o,34)
= 4.78 x 10-2 The frequency of core melt due to,a DBE can now be calculated l
by multiplying the frequency of the DBE initiating event (i.e. 2.04 x 10-4/yr.) by the conditional probability oracore melt (i.e. 4.78 x 10-2) 1 j
C.M.F. = (2.04 x 10-4/hr)(4.78 x 10-2) 9 76 x 10-6fyr, Following a DBE induced core melt accident, it is at,sumed that the drywell air handijn>3 units will fail and cause a j
i breach in that section of the RBCCW system which is inside
{
j containment.
Oiven the breach, the following failu=es would have to occur before contair. ment atmosphere could communicate with the RBCCW surge tank as described earlier:
i
)
I a) Internal leakage from the 'drywell atmosphere through check valve RC-6.
The probability of such leakage is (0.10 + 1.0a 6.6 x 10-4) from Section 1 of Part C.
i b) Leakage past check valves (i.e. RC-2A or 2B) on the RBCCW 1
pump discharge side.
The probability of leakage past i
either check valve is (0.201 + 0.201).
1 4
l n
.a Attschment 2 B12132/Page 45 The probability of containment atmosphere communicating with the RBCCW surge tank is thus:
p(lea' cage) = (0.10 + 1.08 6.6 x 10-4)(0.201 + 0.201) 4.05 x 10-2 Using the public safety input nodel, public risk due to internal leakage in RC 6 is determined by:
PR = Clf x T x C x p(leakage) x R1xR2 Where CIE=9.76 x 10-6/yr. for DBE events, T=25 yr and C=3 x 106 man-rem; the probability of leakage through the RBCCW is p(leakage)=4.05 x 10-2; R =0.1 is the reduction factor due to 1
SBGI: R=0.33 is the reduction factor assumed for scrubbing and plate-out due to the long tortuous path that radioactive material has to make before it exits tne surge tank vent.
Substituting these values into the equation for public risk yields:
PR = (9 76 x 10-6/yr)(25)(3 x 106 man-rem)(4.05 x 10-2)
(0.1)(0 33)
= 0.978 can-rem The maximum benefit that could be expectea from full compli-ance with Appendix J for check valve RC-6 is a public
r B12132/Pego 46 l
risk reduction of 0 98 man-rem.
This value includes the risk due to external and internal valve leakage, assuming that Appendix J testing would result in zero valve leakage.
l
- Howevor, it saould be noted that the man-rem reduction l
associated wich Appendix J is estimated to be zero if the RBCCW. system were seismically upgraded inside containment.
No breach is postulated for a seismically qualified system af ter a DBE and, thus, no release path to environment would exist.
2.
Penetration X-24: Motor Cperated Valve RC-15 Presently, MOV RC-15 is not in compliance with the requirements for Appendix J
Type C testing.
The following analysis quantifies the public risk reduction that could be achieved by Appendix J tecting for exte. cal l
and internal valve leakage.
o External Valve Leakage l
As noted earlier in Section 1, the RBCCW is a closed system which operates at a pressure of ISO psig.
The MOV in ques-t!on is located outside containment and any leakage frori this valve would be detected by:
a) Once per shift walk-throughs l
l
\\
~ ~
i B12132/Pcge 47 b) Reactor Building sump trending c) RBCCW surge tank level alarms Based on the above, MOV RC-15 is considered to bo in a constant leak test mode at pressures greater than those required by Appendix J.
It is therefore concluded that Appendix J testing for external valve leakage would produce no measurable impact on public ris's reduction.
I o
Internal Valve Leakage In the event that containment atmosphere enters the RBCCW system piping following an accident, internal Itakage through valve RC-15 would eventually result in a radioactive release via the surge tank.
Such an event could occur in the wake of a DBE which is assumed to rupture the RBCCW system inside containment by failing the dryvell air handling units.
The public risk associated with this event is determined by:
?R = CMF x T x C x p(leakage) x R1 and R2 where CMF, T,
C, R1 and RP are defined in the previous section for check valve s.C-6.
The probability of leakage (0.10 + 1.08 3 x 10-3) through MOV RC-15 is p(leakage)
=
l
i Attbehment'2 B12132/Pcgs 48 r
j-4 I
as defined ' j n Part B. ' Substituting 'for the terms 'in the above equation yieldst d
6 (9 76 x 10-6/yr.)(25 yr.)(3 x 10 man-rem)(0.10 +
PR =
1.0' 3 x 10-3)(0.1)(0 33)
= 2.49 man-rem i
The maximum reduction in public risk that could be achieved by implementing Appendix J testing for external l
and internal leakage on valve RC-15 is 2.49 r.an-rem.-
l However, as noted earlier in this section, the Appendix J i
]
man-rem reduction would be zero if RBCCW were seismically l~
upgraded inside containment.
l j
j F.
Solenoid Operated Valves IC-6 and IC-7:
+'
Isolation Condenser Vent to the Main Steam System (Dwg.
No.
l 25202-29131 Sheet 2).
i i
When the Isolation Condenser (IC) is in its normal etandby mode, both valves on the steam inlet line are open and one of the two I
valves on the condensate return line is closed.
Non-condensable gases are removed from the IC steam inlet by a Wnt which is connected back to the "A"
Solenoid operated process valves (SOV's) on the vent line (i.e.
IC-6 and IC-7) i close on a group 1 isolatior., a grwup 4 isolation, or whenever the i
i, I
1 i-
N.
} 'Q '
g-g.
Attcchment 2 ja-B12132/Page 50 exposed'to RCS pressure).
As noted in the~ preceding section, non-seismic equipment is exoected to fail after a DB'..
This j
analysis assumes that.a DBE would cause breaches in the main i
condenser would be kept in
~
steam system and the isolation service with significant fuel damage present.
Given these x
i assumptione, the public risk associated with internal leakage i
through SOV'a IC-6 and 7 can be quantified by 'the following s
three cases.
SOV's IC-6 P.nd 7 both leak internally after-Case 1:
successful closure.- Leakage through the first valve (closest l
s l
to the IC) is assigned a value of 0.0432 as per assumption 3 1
r a
The second valve is credited with reducing leakage by a J
factor of 5 (Assumption 6) and is assigned a reduction factor
[
of 0.2.
The probability of leakage through both valves is given by:
l l
0 p(SOV1 & SOV2 leakage) = (0.10)(0.2)
)
1 a 2.0 x 10-2 1
i Public risk can now be determined by the public safety impact l1 i
model as shown below:
i, i
PR = CMF x T x C x p(SOV1 % SOV2 leakasa) j i
8 i
I i
r
]
J
I Attrohment 2 B12132/Pcg3 49 IC is placed into service.
These SOV's act as the boundary for the IC closed loop.
At the present time, neither one of the vent line SOV's is checked for leakage by Appendix J Type C testing requirements.
The poten-tial for reduction in public risk due to Appendix J testing is quantified below for external and internal valve leakage.
o External Valve Leakage Because the valves on the IC steam inlet line are normally open, the vent line and its associated SOV's, IC-6 and 7, are constantly exposed to full reactor pressure.
Any external leakage throJgh the vent line SOV's would be detected by plant equip. tent operators during once per shift wtlk-throughs.
Consequently, the SOV's are in constant test at pressures far exceeding those which are required for once per refuel Appendix J tests.
It is therefore concluded that Appendix J tests for valve external leakage would produce no measurable reduction in public risk.
o Internal Valve Leakage Internal leakage through both SOV's would only present a probles a'
the IC were used after significant fuel damage occurred and there waa an accompanying breach in the main steam system (system is a closed loop which is normally i
Atttchment 2 B12132/Pego 51 where CMF, T and C are defined in preceding Section E ~ and p(SOV1 & SOV2 leakage) is given above.
The resulting public risk is thus:
PR = (9 76 x 10-6/yr.)(25 yr.)(3 x 106 man-rem)(2.0 x 10-2)
= 14.6 man-rem Case 2 - The first SOV leaks internally given that it closes and the second SOV fails to close.
As noted in Case 1 above, valve internal leakage is assumed to be 0.10 and valve failure to close is 1.25 x 10-3 from Assumption 10.
The probability of internal leakage through both valves is given by:
I p(SOV1 & SOV2 leakage) = (0.10)(1.25 x 10-3) i
= 1.25 x 10-4 Public risk is determined by:
PR = (9 75 x 10-6/yr.)(25 yr.)(3 x 106 man-rem >(5.4 x 10-5) i
= 9.15 x 10-2 can-rem Case 3 - The first valve fails to close and the second valve leaks internally. This is identical to Case 2 except that
Attcchment 2 B12132/ Pag 3 52 the order of valve failures is reversed.
Accordingly, the public risk for Case 3 is also 9.15 x W-2 man-rem.
The total public risk associated with SOV's IC-6 and IC-7 is 14.8 man-rem.
This level of risk could only be nyerted if it is assumed that Appendix J type C leak testing would guarantee zero valve leakage post test.
Table 1 summarizes the maximum public risk reduction, in man-rems, that could be achieved for all of the valves that are presently not included in Appendix J Type C tests.
1
c O
I l
1 l
l Table 1 Maximum Public Risk Reduction Due to Implementation of Appendix J Type C Testingl Penetration Associated Reduction in Number Valves (s)
Public Risk X-18 A0V's: SS-3 and SS-4 32.6 man-rem 2 X-19 A0V's.
SS-13 and SS-14 32.6 man-rem 2 X-204A,B&C MOV:
LP-2A 0.214 man-rem HOV:
LP-2B 0.214 man-rem MOV: LP-2C 0.214 man-rem MOV:
LP-2b 0.214 man-rem MOV:
CS-2A 0.214 man-rem MOV:
CS-2B 0.214 man-rem X-210A & B MOV: LP-43A 0 man-rem MOV: LP-43B 0 man-rem CV: LP-24A 0.42 man-rem CV: LP-24B 0.42 man-rem CV: LP-24C 0.42 man-rem CV: LP-24D 0.42 man-rem CV: CS-14A 0.051 man-rem CV:
CS-14B 0.051 man-rom X-15 CV: CU-29 0 man-rem X-23 CV: RC-6 0 98 man-rem 3 X-24 MOV:
RC 15 2.49 man-rem 3 IC Vent SOV's:
IC-6 and IC-7 14.8 man-rem 1 ssumes valve leakage is zero after Appendix J test performed.
j A
2Reduction in public risk would only be 3 3 man-rem if the drywellsump had water in it; 1.e., 32.6 man-rem reduction is based on a dry sump.
3 Reduction in public risk would be zero if that portion of the RBCCW system which is inside containment were seismically qualified.