ML20245H689

Forwards Info in Response to Requesting Informal Discovery.Related Correspondence

ML20245H689
Person / Time
Site:	Limerick
Issue date:	08/10/1989
From:	Wetterhahn M CONNER & WETTERHAHN, PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC
To:	Elliott C POSWISTILO, ELLIOT & ELLIOT
References
CON-#389-9032 OL-2, NUDOCS 8908170242
Download: ML20245H689 (48)

Text

-

gy J

e a s t R e E W 5'8 d

[Q(KElEh LAW OFFICES CONNER Sc WETTERHAHN, P.C.

pr.NPC 17 4 7 P E N N SYINA N I A AV E N U E, N. W.

TSOY B. CONNER, J R.

. MA=M J. WETTERHANN mossat w.mADem

' NILS N. NIC=OLS c s===Amo a. amen nOmr*"

Au9ust 10, 1989

/. u.

or cov===t woa,.333s00 CABLE ADDRFER' ATOML Aw FEDERAL EXPRESS I

Charles W. Elliott, Esq.

Poswistilo, Elliott & Elliott i

Suite 201 1101 Northampton Street L.

Easton, Pennsylvania 18042 In the Matter of Philadelphia Electric Company

{

(Limerick Generating Station, Units 1 and 2) b

^

Docket Nos. 50-352 and 50-353 - O ('

(Severe Accident Mitigation Design Alternatives)

Dear Mr. Elliott:

[

In response to your June 28, 1989 letter requesting i-informal discovery in the captioned matter, enclosed you will find the requested information.

This completes the response to all outstanding informal discovery requests.

l The provision of this information as informal discovery should not be deemed an admission that the Company would be required to produce it under the NRC's Rules of Practice or that the information is relevant or material.

Its provision is also without prejudice to any objection that may be made regarding its admissibility as evidence in this proceeding.

Sincerely, Mark J. Wetterhahn p

Counsel for Philadelphia Electric Company MJW:sdd I

Enclosures cc:

Service List i

8905170242 B9001' ~ kg PDR ADOCK 05000; q

G P.

b

E.1 foi Responses to LEA's June 27, 1989 Letter 1.

Modification of November 1988 internal events PRA results to reflect a

turbine trip frequency of 2.55

-scrams / year.

Limerick Unit 1 has been in operation since early 1986 with a very low number of scrams.

In 3.33 full calendar years of operation, the unit has had only 8 turbine trips; four werg associated with manual scrams and four were automatic scrams.

The only revision to the transient initiator fre-quencies utilized in the November 1988 update is for turbine trip which is from the above values for a Bayesian update with a noninformative prior.

While no manual shutdowns have occurred, the earlier value was not changed since it is a very small contributor to total core damage frequency.

The

low frequency initiators (<1/ year) were also left at their generic values although full use of plant specific data would lead to lower frequencies.

-2.

Revised fire PRA lto. reflect Rev. 11 of Limerick Fire' Protection Evaluation Report, latest plant.-logic models (11/88 PRA update) ' and initiator. frequency and suppression probability from Sandia Fire Risk Scoping Study..

The Limerick Generating Station. fire risk analysis, originally developed in the 1983 Severe Accident Risk Assessment and later updated in SARA Supplement 2,

was revised to reflect the most current available information.

These revisions included the following items:

Updated fire initiation frequencies Revised time-based probabilities of fire sup-pression Updated LGS internal event and fault tree models Incorporation of the Fire Protection Evaluation Report (FPER), Rev. 11 information Reexamination of critical fire locations within the fire areas of interest The methodology employed in the fire analysis update was the same as that used in the original SARA work.

The fire analysis is based on an integrated event

' tree / fault tree approach.

Figure 2-1 shows an example fire progression event tree.

One fire progression event tree was quantified for each initiating fire type and for each fire area.

Each branch point was quantified taking into account the dependencies between them.

Failure events B, D,

and F represent the conditional core damage probability given the amount of equipment postulated to be damaged by the fire at that point in the fire progression scenario.

These condi-tional core damage probabilities were derived from the LGS i

uA__._m

__a-mm___-mm--_._,.m m

_,,_._:.m._

internal cvents fault and event trees.

As such, the CD and B sequences from the fire. progression event tree represent multiple random failure scenarios which result in core damage.

For the CEF sequences (fire growth stage 3) all shutdown methods are damaged (for fire areas evaluated here);

hence, core damage occurs without any additional failures once the fire propagates to fire growth stage 3.

The calculation of each of the branch point probabilities and the final results of the fire analysis update are presented below.

Since the original fire initiation frequencies were developed for SARA, significant additional data has been made available.

This data was examined by Sandia National Laboratory and reported in NUREG/CR-5088, " Fire Risk Scoping Study:

Investigation of Nuclear Pceer Plant Fire

Risk, Including Previously Unaddressed Issues," along with revised fire initiation frequency estimates.

The frequencies reported in NUREG/CR-5088 were directly used in the LGS analysis.

Table 2-1 summarizes the fire initiation frequencies used in the update for each fire area and fire type.

The

values, particularly those for

cables, are considered conservative since no credit was taken for the presence of IEEE-383 rated cables at LGS.

The time based probability of fire suppression, events C and E on the fire progression event trees, were signifi-cantly different between the SARA analysis and NUREG/CR-5088.

In both of the previous quantifications, i

4. a these events were solely based en historical fire -sup-pression data.

The curve provided in Figure

-4.1-1

'of

- NUREG/CR-5088 and included here as Figure 2-2 is based on historical data which is noted to be primarily associated with manual detection'and suppression events.

In most cases of concern 'in the update, automatic detection as well as partial coverage. automatic suppression systems are avail-able.

Since the automatic suppression systems for the areas of concern only provide partial coverage and the initial response time is relatively short (e.g.,

0-10 minutes) the probability of event -C used in the quantification was

.conservat ve y taken directly from the curve in Figure 2-2 i

l for cable and transient combustible initiated fires.

The available suppression time was taken from SARA and is 10 minutes for al'1 fire areas except fi.re area 2,

where the time is zero.

Panel or cabinet initiated fires were treated differently since they have not historically propagated outside the cabinet in which they ignited.

This probability was calculated based on one fire in 38 which did propagate outside the cabinet based upon NUREG/CR-5088.

In addition, I

this was reduced by a

factor of 2,

as suggested in NUREG/CR-3493, "A Review of the Limerick Generating Station Severe Accident Risk Assessment,"

to account for the i

t exclusive us of IEEE-383 rated cable.

Event E was quantified differently from the original SARA and NUREG/CR-5088 studies.

Event E, which represents a significantly longer response time (e.c.,

1-3 hours), was 1

4 g ;-

1 calculated for each applicable scenario using representative values from past PRA experience with fire suppression system-

[

- modeling and with human errors.

To quantify event E,

the following cases were defined for the fire areas of interest.

Case 1 Failure to suppress fire before damage to protected' cables 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> fire. wrap on cables - Fire Area 2 Case 2 Failure to suppress fire before damage to protected cables - 1 quadrant, 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> fire wrap on cables 2 quadrants, 1

hour fire

wrap, water

curtain, and automatic pre action sprinklers Fire Area 44 Case 3 Failure to suppress fire before damage to protected cables - 1 quadrant, 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> fire wrap or I hour fire wrap and within sprinkler coverage - 2 quadrants, I hour fire wrap and water curtain - Fire Area 45 Case 4 Failure to suppress fire before damage to protected-cables - 1 quadrant, 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> fire wrap or 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> fire wrap and within sprinkler coverage - Fire' Area 47 These cases were quantified by using the following equations:

P -Case 1 =

P E

3 P -Case 2 =

[P Q

• P ] + [P O

• P

• PC*P]

g y

3 2

y g

g P -Case 3 =

[P Q

• P C

• P

• P] + [P Q * (1 - P C)

~

E y

g y

g y

g

• P] + [PO*P

• P C) 3 2

y g

P] + [P Q * (1 - P C)

P -Case 4 =

[PQ*PC*P E

y g

y g

y g

• P33 Failure of 3 hr fire barrier = 1E-3 where:

P

=

3 (SARA Supplement 2)

P

=.

Failure of I hour fire barrier = 0.18 1

(NUREG/CR-5088) i a

--_---------------J

).

Failure to manually initiate water PC

=

g curtain = 3E-3 (Basic human error probability) l Failure of automatic sprinklers to P

=

g suppress fire = 2E-2 (typically in range i

of IE-2 to 2E-2) l Percentage of sprinkler coverage in

(

PC

=

g quadrant = 0.5 (estimate based on j

layout drawings in FPER, Rev. 11) l Fire occurrence in 1 quadrant = 0.25 PQ

=

y Fire occurrence in 2 quadrants = 0.5 PC

=

2 The results of the quantification of events C and E are summarized in Table 2-1 along with the fire initiation frequencies.

To reevaluate the LGS internal event and fault tree quantification, the systems or equipment which are postulat-ed to be unavailable were identified in the same manner as in SARA.

However, the basis used in the update was the Fire Protection Evaluation Report, Limerick Generating Station (FPER),

Rev.

11.

Table 2-2 summarizes the systems or equipment assumed to be damaged and hence, unavailable for each fire growth stage and fire area.

In the calculation of the probability of event B using components which are assumed to be failed by the fire initiating events are assigned values of 1.0, and new minimal cut-sets determined and evaluated.

As noted previ-ously, event B actually represents multiple scenarios which lead to core damage given the initial unavailability of systems assumed to be damaged by the fire.

Table 2-3 identifies the systems assumed to be unavailable due to each i

l

)

1 l l

fire and the results of the calculation of conditional core damage frequency.

The probability of event D was calculated in a similar manner to that used for event B.

The conditional core damage frequencies determined in this manner are provided in Table 2-4.

Since all shutdown methods are lost for fire growth stage 3, event F has a probability of 1.0.

The final results were obtained by evaluating the fire progression event trees for each fire initiator and area using the values reported in Tables 2-1 through 2-4 and are i

summarized in Table 2-5 by fire area and sequence type.

These results, explicitly for 4 fire areas give a total cdf for these areas of 3.84E-06.

A value of 0.4E-06 was judged appropriate for the total for other areas giving a total cdf due to fires of 4.2E-06/ reactor year.

F i

r 9

8 4

8 8

8 8

8 e

3 8

4 4

4 4

4 4

4 4

4 4

2 2

7 7

7 7

5 5

5 4

4 4

Area Pn P

P P

P P

l P

a n

n C

n n

C n

C C

I l

1 1

a l

l a

1 1

a P

a b

b 0

b a

b E

1 1

1 l

1 1

l D

1 l

n l

F 0

0 0

e 0

0 e

2 0

e e

e i

B B

B

/

B B

/

0 D

/

l

/

2 2

2 T

2 2

T 2

2 T

s T

re 0

C C

1 1

0 C

2 2

C 4

3 4

4 3

2 1

0 3

I E

1 1

1 3

1 1

4 2

1 5

8 1

F 1

1 1

7 1

1 0

2 1

9 9

9 i

0 0

0 0

0 0

0 0

0 0

0 0

r E

E E

E E

E E

E E

E E

E e

3 3

3 4

3 3

4 3

3 4

3 4

r 0

0 0

0 0

0 0

0 0

0 0

0 F

e q

1 1

1 7

1 1

7 1

1 7

1 1

E v

3 3

3 3

3 3

3 3

3 3

3 0

e 0

0 0

0 0

0 0

0 0

0 0

0 n

E E

E E

E E

E E

E E

E E

t

+

0 0

0 0

0 0

0 0

0 0

0 0

C 2

2 2

1 2

2 1

2 2

1 2

0 S

5 5

8 8

8 2

2 2

1 1

5 E

v 7

7 7

7 4

4 4

6 6

6 0

0 e

5 5

5 5

5 5

5 0

0 0

0 0

n E

E E

E E

E E

E E

E E

E t

0 0

0 0

0 0

0 0

0 0

0 0

E 4

4 4

4 4

4 4

4 4

4 3

3 l

J

Table 2-2 EFFECT OF FIRE GROWTH STAGES ON THE LGS SHUTDCWN METHODS Cable Firee Transient Comb. Fires Pent.1 Fires Fire Aree 2 Fire Growtn Stage 1 Lose PCS (use MSN Same as Cable Fires Same as Cable Fires closure ET)

Fire Growtn Stage 2 All Ecuipment Lost Same as Cacle Fires Same as Caels Fires Except that associated with Shutcown Method B

Fire Growtn Stage 3 All Shutcown Methocs All Shutcown Methocs All Shutcown MethcK:s Lost Lost Lost Fire Area a4 Fire Growth Stage 1 One Shutcown Methoc Same as Cables Pnt 10C201 casables only Lost, assume a1 are RC:C, Pnl 10C202 & 10C203 equalty procable disaele onfy HPCI Fire Growin Stage 2 50 % camage au but Same as Cables Same as Cables SOM A. 50% camage all but SCM O Fire Growtn Stage 3 Au Shut:0wn Metnocs All Shutcown Metnocs All Shutcown Meinocs Lost Lost Lost l

Fire Area 45 Fire Growth Stage 1 One Shutcown Methoc Same as Caeles Pnl 1CB2.ll3 disables AC;C &

Lest. assume an are

'C' Loop of LPC lRHR, Pnl ecually precaole 1C834 c:saele MPCI & 'C' l

Loop of LPC:/RHR Fire Growtn Stage 2 50 % camage an but Same as Caeles Same as Cables SOM A. 50% camage an eut SCM B Fire Grewth Stage 3 All Shutc0wn Methocs A!! Shutcown Methocs All Shutcown Methocs Lost Lest Lost

{

.,r. a r.e 4, Fire Growtn Stage 1 One Shutcown Metnod Same as Caeles Pnt 1C8204 cisacies ad Crv 4 Lest, assume all are comp., 2nd 108213 c:sacies 'A' ecually procacte Loco of LPC iRHR/CS, Pnt 1C8214 cisacle 'B' Loop of LPC iRHR,CS Fire Growtn Stage 2 50 % camage all but Same as Cables Same as Cactes l

SOM A. 50% camage l

all but SCM S Fire Growtn Stage 3 All Shutccan Metnocs All Shutcewn Metnocs All Shutco*n Metnocs Lcst

,l Lest Lov E_________________

Table 2-3 b M Y OF CALCULATED VALUES FOR EVENT B.

SYSTEM /

SHUTDOWN METHOD CONDITIONAL FIRE AREA INITIATOR UNAVAILABLE CDF 2

CABLES, PANELS, PCS 9.0E-6 TRANSIENT COMBUSTIBLES 44

CABLES, A

1.1E-5 TRANSIENT-B 9.3E-6 COMBUSTIBLES C

2.0E-6 D

2.0E-6 PANEL FIRES RCIC 5.6E-7 HPCI 4.9E-7 45

CABLES, A

1.1E-5 TRANSIENT B

9.3E-6 COMBUST ~BLES C

2.0E-6 D

2.0E-6 PANEL FIRES RCIC,LPCI/RHR "C"

5.6E-7 1

HPCI, LPCf/RHR "D"

4.9E-7 47

CABLES, A

1.1E-5 TRANSIENT B

9.3E-6 l_

COMBUSTIBLES C

2.0E !

D 2.0E-G PANEL FIRES DIVISION 4 POWER 2.0E-6 RHR & CS "A"

2.0E-6 RRR & CS "B"

2.0E-6

)

TABLE 2-4

SUMMARY

OF EVENT D it aire i

ecre m onai = F i

s.a nce

,,r. aree s2 Caot e/1C 1.5g c3 an s2 Cante/TC 2.96E 05 ouX s2 Cante/TC 9.40E 05 OWFuWEcc s2 Cacle/TC 5.11E M QWFdet s2 Caote/TC 2.23E-05 antFdcc s2 Penets 1.56E C3 aN s2 Penets 2.98E 05 auk

1. 2 Penets 9.40E 05 oWreWwice s2 PeneLs

$. tit 04 GWFdce s2

• ar:e t s 2.23E 05 ouWrdee s64 caste /1C 3.oOE-c3 aN E4 Cacte/TC 6.60E 05 ouK s&4 Cabte/TC

1. ICE 05 pan s', &

Cacta/tC 2.10E * %

GWFdec 844 Caote/TC 2.30E 06 M ce S&4 Pet 100201 7.10E 03 a#

s&&

Pnt 10D201 1.20E * %

QuX s44 Pat 100201 3.20E 05 Ma#

e44 Pet 10c201 2.10E 06 GWFd ce s44 Pat 100201 2.30E 06 Pdec s44 Pnt 100202.3 3.60E 03 EN s&4 Pnt 10D202.3 6.4CE 05 oux sL4 Pet 1(2202.3 1.60E 05 pan 844 Pat 100202.3 2.10E 06 GWFdes s&4 Pet 100202,3 2.3CE-06 PWEce se5 Cacte/1C 6.3M 06 aN 845 Cante/TC 8.22E 06 our 845 Cable /TC 1.96E %

PotN e45 Caete/TC 2.07E 06 oWFdcc 845 Cacie/TC 2.31E-06 Pd ec

1. 45 Pni 104223 3.93E 03 otN s45 Pet 1C5223 6.50E 05 cux 845 Pnt 108223 1.77E-05 PotN 845 Pet 108223 2.07E 06 QWFdec e45 Pnt 108223 2.31E-06 PWEcc

1. 65 Pnt 108224 3.60E 03 EN e45 Pet 106224 6.40E 05 oux e45 Pet 104226 1.60E 05 PotN s45 Pnt 104226 2.10E 06 GWFdec 845 Pnt 105226 2.30E 06 PWEcc 867 Cacts/TC e.36E 04 an

$47 Cacts/TC 8.22E 06 oui s47 Canie/TC 1.96E-06 PotN 84 7 Cante/TC 2.07E 06 OWFd ec e47 Caste /TC 2.31E 06 Puce s47 Pnt 10E2%

e.36E %

a#

M7 Pnt 1052%

8.22E 06 as s47 Pnt 108204 1.96E-06 pan E7 Pet 10s7%

2.07E 06 QWF uevice 867 ent 108204 2.31E 06 PWEcc M7 Pmt 108213 4.nt 06 aN E7 Pnt 108213 9.02E 06 ouX s&T Pnt 108213 2.13E 06 pan 847 Pet 10E213 2.06E-06 OWF uW'Ecc e47 Pmt 108213 2.30E 06 PWEcc M7

• nt 108214 3.98E 04 an 847 Pet 10E216 7.41E 06 oux 847 Pm4 908216 1.79E 06 pan 867 Pnt 108216

2. 07E -06 QWF W ec e47

• nt tes214 2.31E 06

'WFec

TABLE 2-5 FIRE ANALYSIS UPDATE RESULTS Fire Area 2 Freovency Fir, Area 44 Freauency 8.21E 07 F44-ouv 1.88E-06 F2 etN 2.51E 08 F44 oux 3.51E-08 F2-Qux 1.79E-07 F64 pouv 7.81E.39 F2 oWFdec F2 oWFwvEcc 3.30E 06 F44 hFdec t.02E 09 6.82E-09 F44 oWFwvEcc 7.08E-12 F2 ouwF dcc F44 ofuEcc 2.15E 11 Total 72 1.07E 06 F44 PwEcc 1.09E 09 f otst Ft.4 1.92E 06 Fire Ares 47 Frecuency l

Fire Ares 45 Frecuency 5.07E 07 F47 ouv 3.19E 07 F45-Quv 6.98E-09 F47.our 8.63E 09 F45-Qux 1.05E 09 F47 PQUV 6.13E 10 F45-MUV 6.93E 10 747 oWFdec 6.90E 10 F45-QWFdec F45-oWF wdec 2.56E 12 F47 oWFwvEcc 9.10E 11 7.40E 10 F47 **4 cc 7.23E 10 F45 PWEcc 5.16E 07 total F47 3.30E 07 Total F45 3.34E 06 Total Fire aetated Core Damage FreoJuncy Fire aetsted seoance Fremencies 4tt initiators QUV

}

QUE

}

DQUV

{ QWf eccc I

OWF =Avvi ce i AWf erd C 9 Of =i CC I PWICC 3.53E-06 7.58E-08 9.47E-09 1,32E 07 3.31E 08 6.S2E-09 2.15E 11 2.55E 09 L

g e.

/.S l

s t-1 i

l..

. S r.

1-r,

. r.,

1-

.a t

b

i..-

. J, i -

.z. y a i.

< w

-.5 i

. i 5 _..

-x

1 i

- >~ >..

< s s p.

' P: -

r

.e.

en=

(

~

C

=

m a

, m.

c s-,am..

et a L -

, me

=

C a.

c. :. - 4

C C-

- > m-2 c.

_5 u.-

r-4 h

a.

C C.

t' C,

C-

c. >x ta.

q*

vg

=.*.v,,,_-

m >

e

',.-. m ;

.1.;

-=

K.. s.,

}

?. -

i I

t 1

.s-

- l' s

as

! E a a s..

s..

a

,n,

-a

! - e n E "'

f-l t

-z v.

e i-p., a 1,

s

,- - > z -

e.*=

4 a -

e>

-s,-z.

r 3

g

.2..

e=

.i t

s g..*sq.

s -

e c

I h

.g e.

.e w.wfic

= >

,>T S

E es

• '".j h,

h.

h b

m...

l".'

P e

2 >

i

<

• eb s.

it g

b -

i 3

2 -

P.. *"."

CK N

?.

O s

nn a

ab g

n. k 1

e8

.m.

==.

. $,. 1. *>-*

C'j R.

4 g*

' E e.

m 6

g

• E

$ * & * 'I E

2 *

a. "' f a.

c,.,-,,

5 -

q -

e 'e."'*

I

-g i

f 3*

'3

~ _-

y f

g l

l a

Z. *-

E ea

,e g ** I e bba I

E gg, lI

-*\\

=

l 5.!

e.

---

1.--_._m._.~__m

4 Cumulative Probability (%)

m w

.a, tn m

N m

u lD

~

l' N

1

~\\

LM

\\ ;_

l

\\"_

i v_

e iX i

i i A i

S 3

• c - t

\\

N, g ! u, a

E

\\

m YB

,o _

s w

on I

I I

I lNIe C

i i

e i

T_i f

i t

4 Ye i i

e e

I i

e i

Vi i

{

i i

e t

A.

O N

C i

i m

i I

i i

s t

i i

c CD

3.

Revised seismic PRA to account for fragilities based on actual LGS equipment seismic quantification data and a more recent assessment of ceramic insulator fragility and analysis of recoverable electrical system failures (circuit breaker trips).

The method used to update the seismic portion of the Limerick SARA analysis is outlined below.

The latest plant PRA model was reviewed and differences between it and the original Limerick PRA model used in the Limerick SARA analysis were determined.

The seismic model was revised to reflect the latest plant model.

Recoverable circuit breaker trips caused by relay chatter were added to the seismic model.

The seismic model was revised to reflect updated fragility information.

Seismic sequences were quantified to determine the effect of various model changes.

The original Limerick SARA analysis review showed that five seismic sequences were included in the dominant contributors to core-damage frequency.

Thest five are listed in Table 3-1 with their SARA core-damage frequency.

The review of the current internal event model and post SARA LGS seismic assessments revealed changes since the original Limerick PRA and SARA analyses:

Several random failure probability and seismic fragility values had changed The original loss of offsite power event tree had been expanded into a larger tree consisting of five support states in the Nov.

1988 LGS PRA update.

The components whose random failure probability or sessmic fragility value changed are listed in Table 3-2.

Parameter values used in the SARA and updated analyses are shown.

The random failure probabilities were updated from the current Limerick internal events model.

The SARA seismic induced loss of offsite power event tree was compared with the current internal events loss of offsite power (LOOP) event tree.

It was determined that the SARA seismic induced loss of offsite power event trre dominant sequence equations were the same as those that vould be obtained after modifying the tree to reflect the new support state LOOP event tree.

However, one change was required to make the equations identical.

The change was to add an Emergency Service Water common cause event wherever the Diesel Generator common cause event occurred in the seismic sequence Boolean equations.

This change was made to the SARA seismic equations.

Two other changes were made to the seismic sequence Bcolean equations.

The first added electrical fault events that were recoverable by operator action.

This modeled recoverable failures of the electrical system due to earth-quake induced faults, e.g.,

circuit breaker trip followed by failure of the operator to reset the breaker.

The second change was to allow credit for recovery of diesel generator HVAC faults.

Probabilities for non-recovery of recoverable electrical faults and non-recovery of diesel generator HVAC was estimated to be 0.2.

Once the equations had been modified and new fragility and random failure probability values had been developed,

- _ _ _ -. 4 the sequences were quantified.

The mean results of the final sequence quantification are presented in Table 3-1.

4 TABLE 3-1 LIMERICK SEISMIC DAMAGE SEQUENCES 3 ARA Current Mean Mean Description frequency frequency

- Sequence (yr 1)

(yr 1)

TgEgUX Seismically initated loss of 3.1E-6 1.8E-6 offsite power, followed by failure of high pressure injection and failure of timely depressurization Failure of shear walls in the 9.6E-7 8.6E-7 TgR3 reactor enclosure leading to a loss of all emergency core cooling (seismic initiated) 4.8E-7 Seismic failure of the reactor 8.0E-7 TgRPV vessel upper lateral support Seismically induced loss of 5.4E-7 1.6E-7 T EgCgC2 S

offsite power with control rods failing to insert followed by failure of the boron injection system Failure of shear walls in the 1.4E-7 1.2-7 TgR CB3 reactor enclosure leading to a loss of all AC and DC pova.r and failure of the control rods to insert.

Total 5.5E-6 3.4E-6 i

r

~

TABLE 3-2 RLVISED FRAGILITY VALUES AND RANDOM FAILURE PROBABILITIES FRAGILITY PARAMETERS SARA Values Revised Values Components A

BR BU A

BR BU 1.

4.16 kV bus /SG~

1.49 0.36 0.43 2.60 0.35 0.42 (non-recoverable) 1.66 0.26 0.49 2.

4160-480 V transformer 3.

480 bus /SG 1.46 0.38 0.44 3.95 0.35 0.57

.(non-recoverable) 4.81 0.24 0.31 4.

480 V MCC (non-recoverable) 4.43 0.35 0.74 5.

125 V DC fuse box 4.43 0.35 0.74 6.

125 V DC distri-bution panel 4.30 0.26 0.78 7.

250 V DC MCC (non-recoverable) 8.

DG circuit brkr 1.56 0.32 0.41 2.60 0.35 0.42 9.-

SLC test tank 0.71 0.27 0.37 4.34 0.24 0.28

10. N2 accumulators 0.80 0.27 0.20 4.02 0.31 0.48

11. RER heat exchgrs 1.09 0.32 0.34 1.44 0.31 0.45 1.33 0.35 0.38

12. 4160 V Switchgear (recoverable) 2.75 0.24 0.35

13. 480.V MCC (recoverable) 1.50 0.35 0.44

14. 480 V SG (recoverable)'

0.B3 0.26 0.43

15. 250 V DC MCC (recoverable)

16. Offsite Power 0.20 0.20 0.25 0.30 0.25 0.50 RANDOM FAILURE PROBABILITIES SARA Values Revised Values component / Event Unavail.

Error Unavail.

Error (median)

Factor (median)

Factor

1. Diesel Gen. Common Cause 1 0E-3 3

3.3E-4 3

3.4E-4 3

2. Emergency Service Water Common Cause

3. HPCI System 7.9E-2 2

7.1E-2 2

4. RCIC System 6.6E-2 2.3 5.7E-2 2.3

5. ADS System 7.5E-4 10 1.5E-4 10 2.0E-1 6.

Recovery of DG HVAC 2.0E-1 7.

Recovery of Recoverable Electrical Faults

-- 11'-

'4.

Revised flooding PRA reflecting results'of detailed flood protection analysis, the logic models of the 11/88' PRA, and the occurrence of spurious fire suppression ini-tiation as discussed in the Sandia Fire Risk Scoping Study.

The core damage frequency (cdf),from accident sequences-initiated by -internal flooding at LGS was originally calculated in SARA.

This study was then. updated in the November 1988 LGS PRA, to account for the revised transient event trees.

In both SARA and the Nov. 1988 update, conser-vative assumptions were used to obtain an upper bound estimate of the core damage frequency.

Subsequently, I

calculations were performed to eliminate some of the conservatism made in these prior flooding analyses.

The detailed study by Bechtel Corporation on flooding at LGS, Moderate Energy Pipe Break Analysis Report (Rev.

2)

(December 1984)', was used together with a more realistic modelling of the accident sequences.

Results from the PBA shows that the dominant sequences i

from a flood in'the turbine enclosure are T oUV and T 00X f

f t

with cdfs'of 7.0E-8 and 3.5E-8 per year, respectively.

The

)

PRA analysis is base /. on a 0.016 per year frequency of interna'. floods in the turbine building (based on industry

)

data).

This analysis also assumed that all turbine building

)

floods will lead to a lo.ss of feedwater transient with the probability of failure to recover feedwater in the l

short-term set to 1.0 and probability to recover feedwater J

l in the long-term set to 0.3.

1

a 2

i

~

' P The cdf from the. PRA analysis was. reduced by ~ taking into. account the fraction of turbine enclosure floods that are severe enough to initiate a transient event, as'well as severe enough to degrade the Power Conversion System (PCS)

'to.the extent that it will no longer be available in the short-term.

This reduces the initiating event frequency (and thus the cdf) by approximately a factor'of between 3 and 5.

For example, the Sandia Fire Risk Scoping Study, NUREG/CR-5088, provides data to show that flooding from the-spurious actuation of the fire system will degrade a plant.

system in about 25% of all of the transients initiated.

Therefore, the total cdf from flooding in the turbine enclosure has been reduced to approximately 3E-8 per year.

A review of the PRA analysis shows that except-for three areas, flooding in the reactor enclosure will not present a problem in terms of core damage.

The three areas of concern are labelled RB-FLil, RB-Flit and RB-FL-15.

A description of these areas is provided in Section 3.7 of the Nov. 1988 updated PRA.

In the PRA, a flood in RB-FLll was assumed to lead to a loss of condenser vacuum transient with the following systems degraded:

HPCI, RCIC, RHR, LPCI, and LPCS (pumps A and C).

The dominant accident sequences were T 0UV with a l

cv cdf of 1.27E-6 per year and T 0UX with a cdf of 9.86E-8 per cv year.

A more realistic study was performed for flooding in RB-FLil using work done in the Moderate Energy Pipe Break

- _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ l~

Analysis Report as the main source of information.

This study differs from the PRA study in ' that it considers maximum flood heights, flooding alarms, and more realistic j

initiator frequencies.

The flooding analysis shows that the only pipe break that will cause a direct flooding problem (splashing and spraying not included) in RB-FL11 is a HPCI line break which produces a break flow rate of 563 cfm.

The largest break flow ate from a non-HPCI line break is 131 cfm which will result in a 2-inch equilibrium fle'd height in RB-FLll.

There are no flood sources outside of RB-FLll that will affect this area.

A HPCI line break will result in a flood height of 9 inches with operator action in 20 minutes (Equilibrium flood height is approximately 24 inches without operator action).

Above a flood height of 9 inches, the systems that will be degraded include HPCI, RCIC, RHR, CS Loops A and C,

ar.d LPCI.

To aid operator action, a flooding alarm was in-

]

stalled in this area, with the alarm setpoint at 3-inch flood height.

Since the pipe break is postulated to occur during system operation, it is assumed that the HPCI system is most likely to be in operation in the test mode.

Under i

such conditions, it is expected that the operator would i

assume that any signal from the flooding alarm is caused by J

a rupture in the HPCI system.

As discussed above, the HPCI line could cause flooding if there is a pipe rupture during i

testing.

The frequency of pipe rupture during testing is

]

)

g.j

.]. calculated. - by ' a ssuming ' a test. interval of one every 3 4

months, attest duration of 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />, and a rupture probabil--

ity of 8.6E-10 per pipe section per hour.

This yields a pipe rupture frequency of 1E-8 per year.

Another cause of flooding would be from major HPCI maintenance actions where

- the system is not properly isolated.

This frequency is equal to the product of the frequency of HPCI in major-maintenance, the probability that power.'is not removed from the isolation valves, and the probability that the operator will not maintain suppression pool or Condensate Storage Tank. (CST) ' isolation.

Using LER rates for turbine driven pump failure events, the frequency of HPCI in major mainte-1.ance is estimated to be 0.08/yr (this generic number is conservative compared to actual LGS experience).

The probability that power is not removed from the isolation valves, and the probability that operators would not main-tain isolation of the water sources are both assigned values of 0.01 based on similar analyses from other PRAs.

There-fore,-the-frequency of flooding from HPCI in major mainte-nance is 8.OE-6 per year.

Adding this to the pipe rupture l

frequency to produce a total flood frequency from HPCI piping will yield a total of approximately 8.0E-6 per year.

If there is flooding from the HPCI line (either due to maintenance or pipe rupture), operator action is necessary to prevent the wateer level from reaching a height of 9 inches where major plant systems can be degraded.

Flooding alarms in the area are set for a 3-inch flood height.

A f

l ': -

3-inch' flood height occurs atlapproximately 5 minutes after L

1 the start L of the flood, and a 9-inch' flood height occurs approximately 15 minutes later.

Therefore, the operator has 15 minutes between the time the flood is annunciated and the time when safety systems might be degraded.

The operator L

error probability for this situation is taken to be approxi-mately 0.1..

The total probability that the flood remains unisolated given a flooding condition is the sum of the above operator error probability and the probability that the alarms fail..

Since the latter probability is assigned a value of 0.003 based on industry data, the total conditional probability of an unisolated flood is 0.1 + 0.003 or approx-imately 0.1'.

It is assumed that a flood in the RB-FL11 would cause a plant scram that would behave in the manner of a turbine trip transient.

Utilizing the turbine trip event tree, conditional cdfs were calculated with the following systems out-of-service:

HPCI, RCIC, RHR, LPCI, and the LPCS Pumps A and C.

Taking into account the initiating event frequency of 8.0E-6 per year, and the probability that the flood is unisolated of 0.1, the following accident sequence fre-quencies were obtained (for the 3 dominant sequences)

T QUV 3.0E-10 T QWFWE 4.2E-11 T OUWFWE 4.2E-11 t

cc The effect of clogging of drains has been considered.

At worst, the impact would be to eliminate the chance of

l (.

successful operator mitigation, thereby increasing the above values by a factor of 10.

The analysis discussed above dealt with the effects of direct flooding (i.e.,

water accumulation).

However, the l

effects from splashing must also be considered.

For this, it is assumed that the components affected by splashing are the same as those affected by direct flooding.

The two major differences between the calculation here and that done for direct flooding are:

(1) the initiating event frequency will increase because flood sources will no longer be limited to just the HPCI line, and (2) cperator action cannot be assumed because flooding alarms will not activate in a splashing scenario.

The frequency of pipe rupture for tl e service water piping and all ECCS piping is calculated in Section 3.7.6.1.2.3 of the Nov. 1988 updated PRA.

This frequency is 1.64E-S per year.

The flood frequency from the maintenance of each system is then calculated as follows:

Flood Frequency Assumed Probability from Frequency Power not Probability Mainte-of Major Removed Water nance Maintenance from Source Not Actions System (per year)

Valves Isolated (per year)

HPCI

.08

.01 TOI BE-6 LPCI

.08

.01

.001 BE-7 LPCS

.04

.01

.001 4E-7 RCIC

.08

.01

.01 BE-6 Service Wtr

.04

.01

.001 4E-7 Total 1.76E-5

-_~ - _

Therefore, the total initiat'.ag event frequency from pipe rupture and maintenance actions is 1.64E-5 + 1.76E-5 or 3.4E-5 per year.

Using the above initiating event frequency together with a turbine trip transient event tree (with HPCI, RCIC, RHR, LPCI, and CS Loops A and C out-of-service) tl.3 dominant accident sequences are:

1.3E-8 per year T QUV

=

1.7E-9 per year T QWFWE

=

T QUWFWE 1.7E-9 per year

=

c These results are conservative since there is no single pipe break location where the splashing would disable all of the above systems.

Another source of splashing is from the spurious actuation of '.he fire system, which'would affect MCC 10D203 and MCC 10D202 which might in turn degrade the performance of the HPCI system.

From the Fire Risk Scoping

Study, the number of transients initiated by the spurious actuation of fire systems (breaks and leaks included) is 2.3E-2 per plant per year.

If this is multiplied by a factor of 0.25 to account

{

for the incidents severe enough to degrade at least one system, and by a factor of 0.08 to account for the ratio of the fire system flow rate in RB-FL11 to the total fire system flow rate, the initiating event frequency is 4.6E-4 per year.

)

. l Again if a turbine trip transient is assumed, this time with only the HPCI system degraded, the dominant sequences and their frequencies are:

1.BE-10 per year T OtJX

=

8.7E-11 per year T QUV

=

The dominant sequences from flooding in RB-FL14 are l

T OW (cdf =

2. E-0 8 per year) and T OUV (cdf = 5.1E-09 cy cy per - year).

For this analysis a loss of condenser vacuum transient was assumed with failure of the RCIC system as well as the main feedwater system.

Since the cdfs are already relatively low, a detailed analysis was not done.

Accor65g to the PRA, the dominant accident sequence from a flood in RB-FL15 is T OUV with edfs ranging from ns 1.0E-07 pe'; year to 6.6E-10 per year depending on the flooding scenario.

For a 12-inch flood, MCC 10B213 and MCC 10B214 as well as load center 10B204 are assumed disabled, thus failing LPCI (loops A, B and D) and core spray (loops A and B).

For the 36-inch flood, HPCI is also assumed to fail.

Also assumed is a probability of 0.9 that the opera-ter will cerminate the flood before it reaches 36 inches.

A review of the " Moderate Energy Pipe Break Analysis" shows that the terminal strips for load center 10B204 have been relocated to a higher elevation in the panel to prevent any credible flooding damage.

The report also indicates that MCC 10B213 and 10B214 might not be affected by flooding effects.

Nonetheless, if it is assumed that MCC 10B213 and 10B214 fail, thus failing loops A and B of LPCI and LPCS,

'the probability for failure of the low pressure ECCS (i.e.,

function V) as 5.22E-03.

This contrasts with the value of 1.0 ~ that was conservatively used in the November 1988-LGS PRA.

It.can therefore be concluded that'CDF from flooding in this area for the most restrictive case will decrease to approximately lE-09 if the more realistic probability for V failure is-used.

i In summary, an' updated flooding analysis was performed j

to eliminate some'of the conservatism made in the original i

flooding-analysis presented in: the Nov.

1988 LGS' PRA.

Flooding in the control structure or diesel generator enclosure.was not reanalyzed.

Results of the present update-are presented below.

Core Damage Frequency from Area Internal Floods (per year)

Turbine Enclosure 3E-08 D.G. Enclosure 5E-10 Reactor Enclosure RB-FLil 2E-08 RB-FLl4 3E-08 RB-FL15 1E-09 Control Structure 1E-09 Total BE-08 i

l

l-

l. 1 l

f 5.

Complete listing of accidcat sequences the PECO response attachment only covers the first 24 sequences.

The complete listing of accident sequences on which the June 23, 1989 responses to NRC Staff questions are based is a

provided by the followings i

Table 5-1 Internal events Table 5-2 Fire Table 5-3 Seismic The 24 dominant sequences listed in the June 23, 1989 response are the result of integrating the atove lists.

It should be noted that beyond those fire areas-listed, other fire areas were not explicitly calculated.

For seismic initiators, the lower frequency sequences were also not evaluated.

Internal flooding sequences are not provided because their contribution is very small.

i

(

TABLE 5-1 REVISED ACCIDENT SEQUENCE INTERNAL EVENTS kT FREQUENCY l

I I

p t:

j.

p i

a.

r l

TESSIO- '. 7. 34E -007 12.4%

TE50SF DGORmC TCVS14 6.69E-007 11.0%

TCVOUV TE5518 4.90E-OO7 8.;%

TE30SF;DGIOSF SDG? 05: 1?DG10

TES06 3.79E-OO7 s. 4 *.

TEEc:.

TCVS15 0.01E-007 5.6%

TCVOUX TM514 0.06E-007 5.2%

TMOUV TCP 511 2.04E-00

O.4%

TCFCLHU

TTS14 1.77E-OO7

.0%

T TCU'.

.TE1500 1.68E-OO7 2.5%

TE;UHU7i TMS15

1. 57-E-oO7. -

0.6%

TMGUA TMP 511 1.

!E-OO7 0.1%

TMFILHU TMSG14 1.00E-OO7-2.1% : TMSCU'>

TTPPSO; 1.16E-OO7 2. 0".

TTFeu VRSO:

1.COE-007 1. 7 *'.

VP1 TFS14 9.55E-OOB

1. c;.

TFOUV

TE1519-9.50E-008

1. e% - TE1UHURV TE5544 9.06E-OOB 1.5%

TESUHURV

'TE4544 8.5;E-OOB.

1.4%

E4UHURV TTS15 8.06E-OOO 1.4%

TTOUs TCP*'SO3 7.10E-OOB 1.2%

TCPOU 51S16 6.COE-OO8 1.0%

sioux-

-TE5540 5.68E-008

1. 0*;

TESUMOSFIDGORmC TTFPS11 5.5 E-OOG O.0%

TTPPLHU' TMSS15' 5.04E-COB o. G *;

TMSOUX TTF151; 4.94E-OOO 0.8%

TTP10HOIU~

TFS15 4.69E-OO8 0.8%

TFOUX TE3S 7 4.57E-008 0.8%

TE!URUHV TTPPS18' 4.51E-000

0. 8*.

TTFFLHIU' TMPOSO; 4.31E-GOO O.7%

TMP:U' TEOS 7 4.03E-OOO O.7%

TECUHURV TEP2511 0.84E-COB O.67.

'TEPOLHU' TE5 SOS 3.79E-Oo8 0. 6 *.

TESUHOSFIDG :SF50G50EF10DG10 TICMSOO

!.4TE-008 0.6%

TICMU S1500 3.!!E-COS 0.e%

S10V TTP1514 2.81E-008

0. 5*;

TTP1UHOIV TMSSo5 1.4 E-CAS 0.0%

TMSOWFWEicc TTSO5 1.69E-008

0. ;*.

TTOWFWE(c.-

a508 1.64E-60E

^ 5%

c.D T! CMS 11 1.6;E-00E 0.

TICMLHU' TCVSAS 1.eGE-00E

e. 7%

T;'. QWFwE t.

TCP Sle 1.40E-008 0. *<.

TCPOLHi TES~s*7 1.09E-008 a. ~ *.

TE5WFW TE So!

1.!GE-000 0.;%

T E;'uFL TEF0507 1.24E-00e

..f.

TE500 TICMSIG 1.0!E-008 v. *.

TICML.i 'J TIS 14 1.35E-005 0. *.

T i t UV

'TMFPSo9 1.01E-098 O. ; *..

TMFFLHU' TE 322 1.25E-OG8 o. ;*.

TECUHOSFIDG ;m V TE05:2 1.21E-008

.5. 2 %

TE!UFOSF;DGIFmCv

r a

TCP S 5 1.14E-OOO O.2%

TCPOUHU' TTPPSO:

1.13E-OOB O.2%

TTPPLHIXU' TCPSO7 1.11E-OOB O.0%

TCPCOC1 S1517 1.08E-OOB O.0%

S1D TE5519-1.OOE-OOB' O.0%

.TE50SPODGOOSF 5DG5 CST TICMS~5-9.74E-OO9 0.0%

TICMC10' TCP Sir 8.94E-OO9 0.2%

TCPOLHV

- TMP 516 8.50E-OO9 O.1%

TMP2LHX' TE2517 8.08E-OO9 0.1%

TECUHOSPODGOOSF10DG10V

. TMSOS 7.75E-OO9 0.1%'

TMOWFWE(cc)

TTPPSO 7.10E-OO9 0.1%

TTPPLHX*

ASO~

7.09E-OO9 0.1%

AJE(ce)

TE1SO4 6.91E-OO9 0.1%

TE1WFWE(cc)

TMPOS 5 6.86E-OO9 0.1%

TMPOUhU' ASO6-6.8;E-OO9 0.1%

AI

- ASO7 6.83E-OO9 0.1%

AV.

TMPSO7 6.73E-OO9 0.1%

TMPCCC1 TCP S 6 6.45E-OO9 0.1%

TCPOUHV TTPPUS01 6.45E-OO9 0.1%

TTPPULHLLIU' TFSOS 6.01E-OO9 0.1%

TFOWFWE(cc)

TTPPSO4 5.81E-OO9 0.1%

TTPPV TCP 519 5.79E-OO9 0.1%

TCPOLHLLU'

'FPSO7 5.59E-OO9 0.1%

TFPUHU' TIS 15 5.5BE-009 0.1%

TIQUX TFPS25 5.56E-OO9 0.1%

TFPC10 TMP 510 S.41E-OO9 0.1%

TMPOLHV TCP SO4 5.27E-OO9 0.1%

TCPOV TECSOS 4.70E-OO9 0.1%

TECUHURX TCPPSO; 4.68E-OO9 0.1%

TCPPU' TCVS18 4.05E-OO9 0.1%

TCVPWE(cc)

ASO9 4.10E-OO9 0.1%

ACm TCPPS10 4.10E-OO9 0.1%

TCFPLHXU' TEFFSO9 4.07E-OO9 0.1%

TEPPLHU' TMFOS 6 3.91E-OO9 0.1%

TMP2UHV TE0509 0.90E-OO9 0.1%

TE3URUHX TTP1528 3.87E-009 0.1%

TTP1UHUROIU' TTFPUSO~ 0.84E-OO9 0.1%

TTPPULHLLIX TCP 510 3.74E-OO9 0.1%

TCP2LHW TTPSOB 0.73E-OO9 0.1%

TTPCOC1LTC TMSSO7 0.55E-OO9 0.1%

TMSOWFWW(v)E(cc)

S:515 3.53E-OO9 0.1%

SCOUX TTPPUSO; 3.34E-OO9 0.1%

TTFPUU' TICMS2 3.00E-OO9 0.1%

TICMLHIXU' TMP519 3.30E-OO9 0.1%

TMP2LHLLU' TCPCSci 3.25E-009 0.1%

TCPOLHLLX TFPSOB 3.00E-OO9 0.1%

TFPUHV ASOS 0.COE-OO9 0.1%

AJW(v)E(cc)

TCPOSCO 3.19E-OO9 0.1%

TCPOLHLLV TMPOSO4 0.19E-009 0.1%

TMPOV

1L.

TTSO7 5.09E-OO9 0.1%

TTOWFWW(v)E(cc) g..

LTCVS31 0.9BE-OO9 0.1%

TCVPGUV

-TCVSO7 2.95E-009 0.0%. TCVGWFWW(v)E(cc)

TEPOSO5 0.94E-OO9 0.0%

TEP2UHU' TE4519 0.89E-OO9 0.0%

TE40SPODGORMCV TMPPSO3.~2.BCE-OO9 0.0%.TMPPU' TECSO7-0.BOE-OO9 0.0%

TECCCC1 TTPPS10- 2.76E-OO9-

'O.0%

TTPPLHV TTP1SO 0.72E-OO9 0.0%

TTP1W TEPOS16 2.65E-OO9 O.0%

TEPOLHX' TMS1B 2.60E-009 0.0%

TMPWE(cc)

TCPPSOS 0.60E-OO9 0.0%.TCPPX TCPPSO4 0.58E-009 0.0%

TCPPV TTPSOO O.55E-OO9 0.0%

TTPM

31515' O.54E-OO9 0.0%

S10UV TCPOSO9 0.50E-OO9 0.0%

TCPOUHX' TMPPS10 0.48E-OO9 0.0% 'TMPPLHXU' TISOS 2.46E-OO9 0.0%

TICWFWE(cc)

TISOO.

2.46E-OO9 0.0%

TIC *C

TTPPSO7 0.07E-OO9 0.0% -TTPPXU' TTPPSOO O.57E-009 0.0%

TTPPW TTP1500. 0.07E-OO9 0.0%

TTP1UHUROIY TTP1SO4 0.06E-OO9 0.0%

TTP1X' TMPOS10 C.06E-OO9 O.0%

TMPOLHW TTPr :.19 -O.25E-OO9 0.0%

TTFPLHIV TTP1SO9 2.00E-OO9 0.0%

TTF1UHUROIV TTP1S1":

0.16E-OO9 0.0%

TTP1UHOIW TICMSOO 2.09E-OO9 0.0%

TICMLHX' TTP1517 0.08E-009 0.0%

TTP1UHOX' TMPOSO1

.1.96E-OO9 0.0%

TMPILHLLX TMPOSCO 1.87E-OO9 0.0%

TMPOLHLLV TCPOSO2-1.79E-009 0.0%

TCPOW TCVSOO 1.74E-OO9 0.0%

TCVPOWE(cc)

TICMSO4 1.70E-OO9 0.0%

TICMV TEPOS10 1.68E-OO9 0.0%

TEPOLHV TEPOSO6 1.67E-OO9 0.0%

TEP2UHV TMPPSOS 1.59E-OO9 0.0%

TMPPX TTATSSOO 1.56E-OO9 0.0%

TTATSCMR

-TMPPSO4 1.56E-OO9 0.0%

TMPPV-j TCPPS10 1.50E-OO9 0.0%

TCPPLHXW TMPOSO9 1.50E-OO9 0.0%

TMPOUHX' TTPPUS34 1.48E-OO9 O.0%

TTPPULHLLX' TMSO7 1.40E-OO9 0.0%

TMQWFWW(v)E(ce)

TCVS00 1.38E-OO9 0.0%

TCVPOUX TMS01 1.07E-OO9 0.0%

TMPQUV l

TE1SO6 1.07E-OO9 0.0%

TE1WFWW(v)E(cc)

TTPRBSO8 1.27E-OO9 0.0%

TTPRBI TEPOS10 1.17E-OO9 O.0%

TEPOLHW TTFPS10 1.10E-009 0.0%

TTPPLHW

1TFSO7 1.11E-OO9 0.0% -TFQWFWW(v)E(cc)

TMPOSO:

1 08E-OO9 0.0%

TMPOW TEP 519 1.01E-009 0.0%

TEPOLHLLU' TEP2SO4'.9.91E-010 0.0%. TEPOV TE2516~

9.64E-010 0.0% :TEOUHOSPODGODSF10DG10W TCPPS10 9.49E-0101 0.0%

TCPPLHV L

i TMPPS10; 9.27E-010 0.0%

TMPPLHXW TTPPS17: 9.COE-010.

0.0%

TTPPLHIW TCP 514 B.74E-010 0.0%

TCPOLHXW.

TEPPSCO-B.51E-010 0.0%

TEFPU' TTPPSCS 8.15E-010 0.0%

TTPPLHLLX TTPPS30 B.15E-010

.O.0%

TTPPLHLLIX-TICMS12 8.15E-010 0.0%

TICMLHV TCVSCO.

8.01E-010 0.0% 'TCVPWW(v)E(cc)

TMS ~

7.96E-010 0.0%

TMPOWE(ce)

.TESSO9 7.87E-010 0.0%

TE5UHOSP2DG20SP5DG5 CST l

TTPP500 7.77E-010 0.0% 'TTFPLHLLIV TTPPS27 '7.77E-010 0.0% ~TTPPLHLLV TCPPS 5 7.74E-010 0.0%

TCPPUHX TEPPS13 7.72E-010 0.0%

TEPPLHXU' TEPOS24 7.70E-010 0.0%

TEPOUHW 3

TCPPS24 7.6;E-010-O.0%

TCFPUHV TE2523 7.55E-010 0.0%

TECUHOSPODG2FmCX

]l TCPCSO

.7.44E-010 0.0%

TCPOLHLLX' S 516 7.;OE-010 0.0%

SCD TICMSO7 7.OCE-010 0.0%

TICMXU' l

TICMSO 6.09E-010 0.0%

TICMW J

TMS 2 6.81E-010 0.0%

TMPOUX TICM519 e.65E-010 0.0%

TICMLHIV TEFOSOC 6.55E-010 0.0%

TEPOUHX*

TFPSCO 6.51E-010 0.0%

TFPUHFU' TTSIB 6.49E-010 0.0%

TTPWE(cc)

TECSO8 6.COE-010 0.0%

TECDGee

-l T /P1 SOS 6.16E-010 0.0%

TTP1UHX TEPOSO1 6.01E-010 0.0%

TEPOLHLLX TCPPS19 5.84E-010 0.0%

TCPPLHLLX TMPPS10 5.74E-010 0.0%

TMPPLHV TEPOSCO 5.77E-010 0.0%

TEPILHLLV

}

TCPPS18-5.57E-010 0.0%

TCPPLHLLV TEPOSO:.5.41E-010 0.0%

TEFCW TCP S!4 '5.04E-010 0.0%

TCP2UHURX TMPOS14 5.OBE-010 0.0%

TMPOLHXW

.TCPOS 4 5.06E-010 0.0%

TCPOUHW TEOSO3 5.14E-010 0.0%

TECUROSPODG RmCX J

TE1510 4.9 E-010 0.0%.TE1UHWFWE(cc)

TMSCO 4.85E-010 0.0%

TMPWW(v)E(cc)

TEPPSO4 4.94E-010 0.0%

TEPPV TEOS1B 4.81E-010 0.0%

TECUHOSPODGOCSP10DG10x TMPPSOS 4.68E-010 0.0%

TMPPUHX i

o,

9 i

ll j

ITMPPS24 4.6:E-010 0.0%'

TMPPUHV 7 ISO 7 4.54E-010 0.0%_ TIQWFWW(v)E(cc) h_

'TMP 520.'4.50E-010

_O.0%.

TMPOLHLLX*

4 TICMS 6 4.OOE-010 0.0% iTICMLHLLU' 2 i >- TCPPSOS.:,0.97E-010' O.0%. TCPPLHW F

f

'TFPS 0.86E-010

.O.0%

TFPUHPX TCPSO-5.80E-010 10.0%

TCPM

~

TTPPSO4; O.74E-010

'O.0%

TTPPLHLLX*

.TFPS 1:

10.69E-010-0.0%

TFPUHPV

-TMPPS19 0.50E-010 0.0%

TMPPLHLLX-TISO1 0.09E-010f 0.0%

TIC'GWFWE(cc)-

TMPPS18' O.07E-010 0.' O%

TMPPLHLLV

.TICM510 'O.00E-010 0.0% LTICMLHW TCVS 5 0.20E-010 0.0%

TCVPGWW(v)E(cc)

TMPOS 4 0.10E-010-0.0%

TMPOUHURX TEAS 45-0.10E-010-

10. 0% - TE4UHURX TMPSO4- 0.OBE-010 0.0%

TMPOUHW 1TEPPS10 0.9BE-010 0.0%-

TEPPLHXW TTPPUSOO'O.85E-010 0.0%

TTPPULHLLIW TICMS17 0.7 E-010 0.0%. TICMLHIW TEP 514 'O.71E-010

-0.0%.TEP LHXW L

.TFPS16 0.66E-010' O.0%

TFPUHURX TE4S 0 0.53E-010 0.0%

TE40SF DG FMCX

'TCP 518 2.54E-010 0.0%

TCPOLHLLW

.{

TFPS15 2.45E-010 0.0%

TFPUHURV TICMSOS' O.40E-010' O.0%

TICMLHLLX TICMSOC O.40E-010 0.0%

TICMLHLLIX TFPSO6

'O.40E-010 0.0%

TFPUHW

.TMPPSOS 2.40E-010 0.0%

TMPPLHW TFPS11 0.08E-010 0.0%

TFFUHX' TCATSSOO s00E-010 0.0%

TCATSCMR TTPPSO1' O.00E-010 C.0%

TTPPLHIXW TICMSO:

2.29E-010 0.0%

TICMLHLLIV TICMS 7 0.09E-010 0.0%

TICMLHLLV TFPSO2

..00E-010 0.0%

TFPW TTP1527 0.14E-010 0.0%

TTP1UHURGIW TCPPSO:

2.06E-010 0.0%

TCPPW TEPPS24 2.OOE-010 0.0%

TEPPUHV TFPSO4

~1.87E-010 0.0%

TFPX' TEPPS10 1.79E-010 0.0%

TEPPLHV TTP1501.1.65E-010 0.0%

TTP1UHUROX" TECS14 1.50E-010-0.0%

TECUHOSPODGOOSP10W TFPSCO 1.49E-010 0.0%

TFPUHPX' TMS 5 1.47E-0101 0.0%

TMPQWW(v)E(cc)

TISOcc 1.46E-010 0.0%

TIC'QUV TMP 518 1.44E-010 0.0%

TMP2LHLLW TMATESCO 1.41E-010 oO.0%

TMATSCMR TEPOSCO 1.0EE-010 0.0%

TEPOLHLLX' TE5545 1.29E-010 O'.0%

TESUHURX

a H

\\

i 4

TMPP500:.1.25E-010 0.0%

Tt1PPW.

TEPPSOB. 1.00E-010:

0.0% :TEPPLHW.

TTATSS07 1.00E-010 0.0%. TTATSCER

.TTSOO 1.19E-010l OLO%

.TTFWW(v)E(cc)

TE0509 1.10E-010~

0.0% LTECUHOSPOWFW

- TFATSS05 1.17E-010:

0.0%

TFATSCMR 51505;. 1.15E-010 0.0%.

SiOWFWE(cc) c TEPPS18 1.10E-010 0.0%

TEPPLHLLV 5:514

,9.87E-011 10.0%,SOQUV

TE151 9.07E-011.

0.0%

TE1UHWFWW(v)Eice)

TEPOS18 47.75E-011 0.0%

TEPOLHLLW

~TECSOO.

7.40E-011 0.0% 'TECM S2505 7.28E-011 0.0%

SOCWFWE(cc)

TTPPUSO2'6.91E-011 0.0%.TTFPUW TEPP505 6.87E-011 0.0%

TEPPX TISCO 6.04E-011-0.0%

TIC'OWFWW(v)E(cc)

TFPSO~

6.18E-011 0.0%

TFPU' LTIS01 6.14E-011 0.0%

TIC'OUX TEPP502 ~6.01E-011 0.0%

TEFFW TE5506 5.4:E-011' o0.0%

TE50SF2WFW'

'TEOS 6 5.04E-011 0.0%

TEOUHURWFW TE2510 4.87E-011 0.0%

TECUHCSFCDG WFW TE5517 4.86E-011 0.0%- TE50SPODGCCSP5DG?OSF10W

T 518 4.B~E-011 0.0%

TFFWE(cc)

TEOSC1 4.4BE-011 0.0%

TECUHOSPODGIFmCWFU TE2506 4.41E-011 0.0%

TECUHWFW LTESSO~

a.06E-011 0.0%

TESWFW TE4SO3 2.97E-011 0.0%

TE4WFW TECPS 5 0.87E-011 0.0%'

TEPPUHX TFPS19 2.85E-011 0.0%

TFPUHPW 31507 1.8 E-011 0.0%

S10WFWW(v)E(cc)

TESS12 1.81E-011 0.0%

TE50SFCDG20SP5WFW TCATSSO7 1.78E-011 0.0%

TCATSCER TTS01 '

1.75E-011 0.0%

TTPQUV TEFFSlo 1.56E-011 0.0%

TEPPLHLLX TFPS10' 1.51E-011 0.0%

TFPUHURW TE5SO9 1.4SE-011 0.0%

TE50SPODG2WFW TE1516 1.44E-011 0.0%

TE1UHURWFWE(cc) 50507 1.~4E-011 0.0%

SOQWFWW(v)E(cc)

TCVS11 1.15E-011 0.0%

TCVOUWFWE(cc)

TMATS507 1.08E-011 0.0%

TMATSCER TTS 0 1.03E-011 0.0%

TTPOWE(cc)

TFS31 9,55E-012 0.0%

TFPOUV LTFATSSO7 8.91E-012 0.0%

TFATSCER TFSCO 8.90E-010 0.0%

TFPWW(v)E(ce)

TTS11 8.56E-012 0.0%

TTOUWFWE(cc) 51512 8.1~E-010 0.0%

S10UWFWE(cc)

TTPSO6 7.05E-010 0.0%

TTPCIM TTS 7.04E-010 0.0%

TTPOUX

- - - ~. _. _. _ _ _. _ _ _ _ _

e a

P TES 5 4= 6.97E-010 0.0%

TEPOUHURX TEhS15 6.48E-010 0.0%

TE50SP DG 05PSDG5WFW

. TF5 -

5.54E-010 0.0%, TFPGWE(c:)

TMS11 -

5.0 E-01 0.0%.TMOUWFWE(c )

TFS 4.67E-01:

0.0%

TFFOUX.

TFS11-0.89E-01 0.0%

TFOUWFWE(cc)

TMSS11 0.77E-010 0.0%

TMSOUWFWE(c

)

TE1518-0.65E-012 0.0%

TE1UHURWFWW(v)E(c:)

fTCVS1; 0.11E-01 0.0%

TCVCUWFWW(vie (c:)

TCVS S 2.05E-01; 0.0%

TCVPGUWE(c-)

T-

'1.90E-012 0.0%

TTPOWW(v)ctc 1 Th:*[5 s

'1.58E-010 0.0%

TTOUWFWW(viEte:)

sigl4 1.50E-010 0.0%

S1GUWFWW(v) tic:)

CFsos

1.0 0E-01

0.0%

TCPCOM

- TF5:5

.1.00E-010 0 ~. C %

TFPCWW(v)EIC:)

TMS1; 9.27E 0.0%

TMGUWFWW(v)E(c:>

TMS23 9.06E-01; 0.0%

TMPOUWE(c )

T I 511' S.75E-01; 0.0%

TICUWFWE(c:)

-.g 7 B.19E-010 0.0%

TIC'QUWFWE(c:s Th250-7.16E-01; 0.0%

TFOUWFWW(v/E( :)

~TM551-6.c E 0.0%

TMSCUWFWW(v)E( :i

'TMP506 6.21E-010 0.0%

TMPCOM TE2306 4.1TE 0.0%

TECCOM TCVS 0 3.E!E 0.0%

TCVPCUWW(v)E(0:)

4519

!.45E-01.

0.0%

ACeK TMS 0 1.67E-010 0.0%

TMPCUWW(v)Ei::)

T si-1.61E 0.0%

TIQUWFWWiviEi:: )

.,..= _ e.

4 5.1e 0.0%

TIC'GUWFWW(v*E(::

A517 5.64E-014 0.0%

ACED Asi:

..5E-014 0.0%

ACeJE( :)

4s15 2.01E-014 0.0%

Acel 4513 2.01E-014 0.0%

ACeV Asis

.1.41E-014 0.0%

ACeCm As r.4 1.04E-014 0.0%

ACeJW(viEic:)

TF529 6.14E-015 0.0%

TFFCUWE(c:)

TTSOS 5.55E-015 0.0%

TTF CUWE t c: )

TFs O 1.1 E-015 0.0%

TFPOUWW(v)Eice)

TT5 0 6.54E-013 0.0%

TTF OUWW ( v i E i c : -)

i I

a l

i l

TABLE 5-2 FIRE ANALYSIS UPDATE RESULTS Fire area 2 Freauenev Fire are..

,,eauene,

F2-QUV 8.21E*07 F44-Quv 1.SEE 06 FZ-Qux.

2.51E 08 F44 oux 3.51E*08 F2 awFdec 1.79E 01 744 Pouv 7.StE-09 F2 owfwwvEcc 3.30E*08 F44.cwFdes t.02E 09 F2-QLf4Fwf Gc 6.82E 09 F44.owFwvEcc 7,;gg.12 F44 cFwEcc 2.15E 11 Totet F2 1.07E 06 F44 PwEcc 1.09E 09 Total - F44 1.92E 06 Fire Aree 45 Frecuency Fire area 47 Krecuency F45 ouv 5.07E 07 F47-QuV 3.19E 07 F45-Qux 6.98E 09 F47-Qux 8.63E 09 F45 PQUv 1.05E 09 F47 Pouv 6.'3E 10 F45-owFdce 6.93E-10 747-QwF dce 6.90E-10 F45 OwFwwvEcc 2.56E 12 F47 owFwvEcc 9.10E 11 F45-PW ec 7.40E 10 F47 PWEcc 7.23E 10 Totat - F45 5.16E-07 totat 847 3.30E 07 Total Fire nelated Core Damese Freemncy 3.34E 06 Fire aetated seo s ce Freauenetes Att Initiators ouv i

aus I eeuw i ows.dce i ;wf *vice

! uf = ce

=ic:

• *E :: '

3.53E 06 7.58E 08 9.e7E 09 1.S2E 07 3.31E 08

6. 321. M 2.t5E M 2.!!E M

TABLE 5-3 LIMERICK SEISMIC DAMAGE SEQUENCES SARA Current Mean Mean Sequence Description frequency frequency (yr 1)

(yr 1) l l

l TgEgUX Seismically initated loss of 3.1E-6 1.8E-6 offsite power, followed by failure of high pressure injection and failure of timely depressurization TgRB Failure of shear walls in the 9.6E-7 8.6E-7 reactor enclosure leading to a loss of all emergency core cooling (seismic initiated)

T RPV Seismic failure of the reactor 3.0E-7 4.8E-7 S

vessel upper lateral support TgEgC C32 Seismically induced loss of 5.4E-7 1.6E-7 offsite power with control rods failing to insert followed by failure of the boron injection system TRCSB3 Failure of shear valls in the 1.4E-7 1.2-7 reactor enclosure leading to a loss of all AC and DC power and failure of the control rods to insert.

Total 5.5E-6 3.4E-6 l

. 6.

Consequence calculations carried out beyond 50 miles based on the SARA source term and consequence analy-sis; this information should be directly available from the CRAC2 printouts.

Based on the results of CRAC2 analyses for the dominant release category (OPREL), the 500

• 111e population dose is approximately 1.75 times that for 50 miles.

i l

l l

l

vr

- 22'-

7..

Documentation. describing " adjustments" made to PRA results'.to account for the " benefit" of ' spray or injection into drywell following core melt.

In the SARA analysis, no credit was taken for operation of - the existing drywell sprays in mitigating the conse-quences of an accident.

A review was performed of the.

dominant accident sequences in each class'to determine for i

E which sequences Lie drywell sprays would be available for mitigation.

The dominant core damage sequences and their associated accident classes are shown on Table 1-2 of the June 3,

1989 response.

The estimated probability of the I

' dry ;11 sprays being available for each dominant non-seismic

-sequence is shown on Table 7-1.

Dominant sequences 11, 17 and 19 are Class 1 transient initiated sequences with loss of high pressure injection and failure to depressurize.

Since AC power and the low pres-sure. systems are available, there is a high probability (.9) s.g' g~

that the drywell spray system is available.

Dominant sequences 5,

8 and 10 are loss of offsite i

power initiated sequences.

Spray availability was' de-termined by the probability that AC power would be restored prior to containment failure or battery depletion for these sequences.

This probability was estimated to be 0.5 (the i

range for individual sequences was from 0.

to.76).

Dominant sequences 1,

4, 6,

7, 12, 13, 15 and 21 are l

1 characterized by failure of the low and high pressure core coolant injection systems.

Because of the dependencies j

l

_ _ _ _ _ - _ - _ - _ _ _ _ between'the low pressure injection systems and _?2 drywell spray: system. there is a relatively high probability that the drywell spray system will also be unavailable.

It was estimated that for only about~ 1/3 of the sequences with failure of the low pressure systems would the drywel.'. sprays be available.

The overall Class 1 probability of spray availability was determined from a frequency weighted average of the above dominant. sequence probabilities.

Psp (.9) (7 x-10

) + (.5) (2.3 x l,-6) + (.33) (5.8 x 10-6)

~

7 x 10

+ 2.'3 x 10-6 + 5.8 x 10-6

~

.42

=

i The probability of spray availability for Class 2 sequences was judged to be very low because of dependencies between j

the containment heat removal system and the drywell spray i

system.

Class 3 sequences are similar in character to Class 1 l

sequences.

Consequently, the same probability

(.42) for 1

spray availability is assumed for Class 3 sequences as for 1

I Class 1 sequences.

i For Class 4 (ATWS) and for Class S (vessel rupture) j containment failure ocr.urs before core melt.

In the Limerick PRA and in SARA, containment failure was assumed to result in the loss of core coolant injection.

For a

l

-[

.--24.-

o.

consistency, it is assumed.that drywell spray capability is a2so'. lost due to' containment failure.

Using a' probability of.42 for the spray availability 1

l

-for. Class 1 and.3 sequences and assuming that the drywell-sprays prevent con'tainment overpressure and overtemperature.

I-failure results in the modified Limerick. risk values. shown in Table 2-2 of the June 23, 1989 PECO submittal.

l 1

I d

I

p' TABLE 7-1 ESTIMATED PROBABILITY OF. DRYWELL SPRAY AVAILABILITY DOMINANT SEQUENCE PROBABILITY OF CLASS NUMBERfS)

FREOUENCY SPRAY AVAILABILITY l

- l' 11, 17, 19 7E-07

.9 l

1 5,

8, 10 2.3E-06

.5 1

1,4,6,7,12, 5.8E-06

.33 13,15,21 2

'16 o,

4 14,20,23 o,

S 24 o,

i

8.

' Justification for location underground of the Dedicated Suppression Pool Cooling System, and comparative l

analysis of the impact of this decision on the cost of the I

system versus an above-ground location.

Due to NPSH requirements, the pump structure has to be located approximately forty feet underground.

No analysis was done for an above-ground location because of NPSH i

requirements.

I j

. i_^ Nj; l s-

-~26'-

9...

Documentation supporting the ' assumption that the Dedicated Suppression Pool Cooling System cannot mitigate Class 4 ATWS sequences.

It is - unlikely that - the Dedicated Suppression Pool Cooling System will-be effective in mitigating.- ATWS sequences.

The design heat removal capacity of this system (W 45 MWt) is far.below the heat production rate during.ah ATWS (N 10% of full' core power or 330 MWt).

Hence, it is

~

unlikely-that this system will prevent containment overpressure failure or core melt.

Furthermore, this system provides no mitigation of the radionuclides released during the accident.

l;*.

- 2]

L l

H 10..

Documentation of all PECO cost-benefit analyses which considered combinations of the SAMDAs which it evalu-atedi for example, any - analysis ' of the combination of the Dedicated Suppression Pool-Cooling System and the Enhanced Drywell Spray System, - which are intended to be operated l'

together, but which were evaluated separately in PECO's response.

i.

As stated on page 2-11 of the June 23, 1989 response to NRC request for additional information, the enhanced drywell spray system was evaluated in conjunction with the Dedicated Suppression Pool Cooling System.

No other combination of SAMDAs has been evaluated.

1 e