Text

Draft, Evaluation of Risk of Induced Steam Generator Tube Rupture, for Callaway Plant
	ML20207A740
Person / Time
Site:	Callaway
Issue date:	03/30/1999
From:	Connelly K, Walz M; UNION ELECTRIC CO.
To:	;
Shared Package
ML20207A737	List: ML20207A732; ML20207A740;
References
	NUDOCS 9905270241
	Download: ML20207A740 (21)
	v • d • e

6 o

A DR&ff AmorenUE CALLAWAY PLANT Evaluation of Risk of Induced Steam Generator Tube Rupture March,1999 Prepared by:

K. G. Connelly M. D. Walz DRA F f 9905270241 990520 c

DR ADOCK 05000403 PDR j

wdTarwr1ghavgamew c;

DilAFT CALLAWAY PLANT Evaluation of Risii of Induced Steam Generator Tube Rupture introduction During review of AmerenUE's license amendment application for steam generator tube electro-sleeving, NRC questioned the frequency, for Callaway, of having an induced steam generator tube rupture (SGTR) under severe accident conditions.

The postulated requisite conditions for an induced SGTR are core damage, loss of liquid inventory on both the primary and secondary sides of the steam generators, and the reactor coolant system (RCS) intact. This case will be referred to as Case 1.

During discussions with NRC, a second (Case 2) scenario was postulated which could result in induced steam generator tube ruptures. This case involved:

successful feedwater injection initially, followed later by a loss of feedwater, e

and hence loss of liquid inventory on the secondary side of the steam generators, an RCP seal LOCA with no RCS makeup, and e

core damage.

e In this case, the RCP seal LOCA could cause the RCS to depressurize to a point at which the accumulators would begin to inject.

This would result in the generation of steam in the damaged core, forcing hot gasses into the steam generator tubes. As RCS pressure rises, due to steam generation in the core, the accumulator discharge check valves will close, terminating injection into the damaged core, Pressure would then drop in the RCS, due to the RCP LOCAs, and the accumulators will again inject. As this process repeats itself, it is postulated that induced SGTRs could occur.

The frequencies of occurrence, at Callaway, for each of these postulated cases are discussed below.

Evaluations

)

Case 1:

To determine a frequency for Case 1, described above, the Callaway core da. mage sequencos were screened to identify those that entailed core damage due to failure of feedwater and feed and bleed core cooling, in these sequences, the steam generators would have no inventory on either the primary or l

DRA F T

I

.}

DRAFT 1

secondary c! des, and the RCS would remain intact for some period of time.

Attachrnent 1 provides a listing of those sequences meeting these criteria, along i

with their frequencies of occurrence.

l A simple event tree (Attachment 2) was developed to evaluate the induced SGTR frequency corresponding to Case 1. The first event, Dsa, represents the Case 1 requisites for induced SGTR. The frequency of Dan is the sum of the frequencies of the sequences on Attachment 1, or 4.5gE-6 per year.

The second event tree heading, FPinj, factors into the event tree the possibility that the operators, following both the Emergency Operating Procedures and the Severe Accident Management Guidelines, will inject firewater into the SGs (secondary side) using the plant fire protection system.

If feedwater is successfully injected, resulting in liquid inventory on the SG secondary side, induced SGTR would be prevented from occurring. The probability that firewster injection into the SGs fails was estimated to be 0.1g5. The simple calculation of this value is shown on Attachment 1.

The 0.195 value is based upon an assumed human error probability of 0.1, and estimates of the unavailability of the firewster suppression system under both loss of offsite power (LOSP) and non-LOSP conditions.

l The final event tree heading, Lseat, depicts whether or not the loop seal clears.

l If the lot.,p see! does not clear, hot gasses from the molten core materials cannot l

(

circulate through the SGs; honoe, an induced SGTR would not occur. A split l (

fraction of 0.5 was assigned to this event.

The Case 1 event tree was quantified manually.

The resulting frequency estimates for conditions requisite for induced SGTR are summarized in the table below.

Results of Case 1 Induced SGTR Frecuency Quantification 1

l l

Dan

FPird Yes (with no creditforloop seal) 8.g5E 7 Dan

FPinj

Lseal Yes (with credit forloop seal) 4.48E-7 The estimated frequencies, in the table above, of RCS conditions wh*ch'could result in induced SGTR, are below the NRC espoused threshold of 1E-6 per year.

C CRAFT

v.y DRA F T m

case 2:

This section discusses estimation of a frequency for the Case 2 Induced SGTR scenario, described above.

This scenario involves successful feedwater injection, initially, with feedwater failure at some later time. An RCP seal LOCA and core damage are also requirements.

In the Cellaway PRA, the feedwater mission time, used to calculate feedwater unavailabilities, was selected to achieve a stable endstate with respect to core cooling.

Those accident sequences in which feedwater injection is not successful for the selected mission time can result in loss of decay heat removal, and are included in the Case 1 evaluation.

The scenario, modeled in the Callaway PRA, in which feedwater is initially successful, and then falls, is a station blackout sequence in which the turbine-driven auxiliary feedwater pump (TDAFP) train operates successfully, but then fails at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , due to battery depletion and no AC power recovery. This scenario is delineated below.

1.LOSP

2. Failure of both diesel-generators to power their respective 4160 VAC bus.

3. Successful feedwater injection via the TDAFP train.'

4. Operators fall to cool and depressurize the RCS. (Note here that successful RCS cooldown and depressurization occurs prior to core melt; therefore, the C

postulated Case 2 induced SGTR would not occur.)

~

5. AC power is not recovered within 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months .

The NUPRA code and the updated Callaway Level 1 PRA was used to quantify the frequency of the above series of events. The frequency determined was 2.75E-7 y(',

For the Case 2 scenario to be valid, an RCP seal LOCA, of a certain size range, must occur such that the RCS depressurizes, to a pressure below which the accumulators will inject, after TDAFP injection has failed and the SGs are dry.

MAAP runs were performed to establish the required seal LOCA size range.

(These runs and the results are discussed below.) A conclusion of the MAAP runs was that the most probable RCP seal failure mode (and corresponding leakage rate of 21 gpm por RCP) will not provide for RCS pressures that would allow accumulator injection, prior to the postulated failure of the reactor coolant system. The probability of the 21 gpm per RCP leakage rate is 0.9. (Therefore, leakage rates other than this have a cumulative probability of occurrence of 0.1.)

Therefore, the frequency of induced SGTR, given the Case 2 requirements is d

below 2.75E-8 yr (i.e., (2.75E 7) * (0.1)). Also, note that this estimate does not credit injection of firewater into the secondary side of the steam generators following failure of the turbine-driven auxiliary feedwater pump.

OfAf f

~

q DRAF7 m

MAAP Runs k.

e The Case 2 scenario, described above, was evaluated with the MAAP severe accident thermal-hydraulic code. The MAAP model was based upon several cases presented in the Callaway IPE submittal.

I The basic scenario is outlined below:

1. Station blackout occurs at time zero.

2. RCP seal LOCA occurs at 10 minutes due to a loss of seal injection and cooling.

3. AFW is successful initially via the turbine driven pump.

4. Power recovery is not successful. AFW is assumed to fall at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months due to depletion of station batteries and subsequent loss of control and instrumentation.

5. MAAP assumes the reactor vessel falls, after a user defined delay time, after the lower core support plate falls. The IPE typically assumed a 60 second delay. For the cases presented here, the delay time (TTRX) was set such that the vessel never failed.

Several sizes of RCP seal LOCAs were evaluated to span the range of anticipated leak rates. provides plots of pressurizer pressure versus time for four cases.

(.

These cases are summarized below:

CASE 14Ba - 21.4 gpm/ pump seal LOCA was modeled. The RCS pressure decreased initially due to the seal LOCA and the removal of decay heat by the i

secondary. The RCS pressure stabilized at the secondary pressure. AFW was lost at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . As the steam generators emptied, the RCS pressure began to Increase. The steam generators went dry at 13.6 Murs. The core became uncovered at 14.2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months due to primary fluid loss through the seal LOCA. The RCS pressure had now increased to the safety valve setpoint. The core support plate failed at 16.2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months and the corium dropped to the lower vessel head. The RCS pressure begins a slow descent as pressure is relieved via the seal LOCA.

The accumulators would inject at about 30 hours3.472222e-4 days 0.00833 hours 4.960317e-5 weeks 1.1415e-5 months into the event. By this time, the corium would have been in contact with the lower vessel head for about 14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months .

CASE 14Fa - 60 gpm/ pump seal LOCA was modeled. The RCS pressure decreased initially due to the seal LOCA and the removal of decay heat by the secondary. The RCS pressure stabilized at the secondary pressure. AFW was lost at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . The core became uncovered at g.3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months due to primary fluid loss through the seal LOCA. As the steam generators emptied and the core heats up, the RCS pressure began to increase. The core support plate failed at 11.g hours and the corium dropped to the lower vessel head. The RCS pressure DRA F T

m DRA F T spikes several times as the water remaining below the core flashes to steam d

when the corium descends. The steam generators went dry at 13.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months . The LL RCS pressure begins a slow descent as pressure is relieved via the seal LOCA.

The accumulators would inject at about 17 hours1.967593e-4 days 0.00472 hours 2.810847e-5 weeks 6.4685e-6 months into the event. By this time, the corium would have been in contact with the lower vessel head for about 5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months .

CASE 14Ea - 183 gpm/ pump seal LOCA was modeled. The RCS pressure decreased initially due to the seal LOCA and the removal of decay heat by the secondary. The RCS pressure stabilized at the secondary pressure. The core i

became uncovered at 3.3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months due to primary fluid loss through the seal LOCA.

I The core support plate failed at 5.0 hours0 days 0 hours 0 weeks 0 months and the corium dropped to the lower vessel head. The RCS pressure spikes as the water remaining below the core flashes to steam when the corium descends. The RCS pressure begins a slow descent as pressure is relieved via the seal LOCA and removal of decay heat by the secondary. The accumulators inject at about 7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months into the event. AFW i

was lost at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . The accumulators are depleted at about 20 hours2.314815e-4 days 0.00556 hours 3.306878e-5 weeks 7.61e-6 months . The steam generators retain significant inventory throughout the event.

l CASE 14C - 480 gpm/ pump seal LOCA was modeled. Also, the reactor vessel was allowed to fail 80 seconds after core support plate failure.

The RCS pressure decreased initially due to the seal LOCA and the removal of decay heat by the secondary. The RCS pressure stabilized at the secondary pressure. The core became uncovered at 1.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months due to primary fluid loss through the seal 1

LOCA. The accumulators inject at 2.3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months into the event. The core support

(~(

plate failed at 2.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months and the corium dropped to the lower vessel head. The reactor vessel falls 60 seconds later. The RCS pressure begins a fast descent and the corium relocates into the containment. The accumulators deplete at 2.7

)

hours. AFW was lost at 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months .

The steam generators retain significant inventory throughout the event. If the vessel had not been allowed to fail, the results for this case would resemble that for CASE 14Es.

~

As can be seen, the smaller seal LOCA cases (which represent the highest probability seal LOCAs) show that the accumulators will inject after the RCS depressurizes and after AFW is lost. However, the corium is in contact with the lower vessel head for an extended period of time, which would result in failure of the lower head prior to accumulator injection.

The larger seal LOCA cases show that the primary depressurization occurs early in the event and results in accumulator igiection occurring prior to steam generator dryout.

DRAFT

~

- ~

vm r.

DIA F T w

{C' Conclusion

{.

As discussed above, both the Case 1 and Case 2 scenarios, under which induced steam generator tube ruptures are postulated to occur, have estimated frequencies below 1E-6 yr'.

l l

1 1

1 M

w DRA F T

~ ~ $MWh with Loss of DHR (FW and FIB)

S(1)S05 6.25E-10

{'

S(1)$10 1.83E-10 S(1)S11 0.00E+00 0RA F T S(2)S10 0.00E+00 S(2)S11 0.00E+00 S(2)S20 1.82E-10 S(3)SO4 1.23E-09 S(3)S05 2.31E-09 S(3)S12 4.95E-08 T(1)S03 5.36E 08 3.12E 06 is freque.7ey of sequences with LOSP.

T(1)S04 1.69E-08 T(1)S05 7.51E-08 T(1S)S22 1.03E-08 0.68 conditional prob that loss of DHR la LOSP-initiated.

T(1S)S23 1.36E-07 T(1S)S24 5.51 E-08 0.32 conditional prob that loss of DHR is not LOSP-initiated.

T(1S)S25 1.22E-08 T(1S)S28 2.75E-06 T(2)S03 5.69E-08 T(2)SO4 1.09E-07 T(2)S05 2.38E-08 T(3)SO4 5.84E-08 T(3)S05 1.34E-07 T(3)S06 3.43E-08 TAT 1S03 0.00E+00 P(FPinj) = 0.1 + [0.32(0.02) + 0.68(0.13)) = 0.195 TAT 1012 0.00E+00 TAT 1S13 2.49E-10 F(Dsa)

P(FPinj) = (4.59E 6) * (0.195) = 8,95E-7 TAT 1S22 0.00E+00

(

TAT 1S23 1.05E-10 F(Osa)

P(FPinj)

SF(Lseal) = (8.95E-7) * (0.5) = 4.48E-7 TAT 3S03 1.27E-10 TAT 3S12 9.32E 10 TAT 3S13 1.23E 08 TAT 3S22 0.00E+00 TAT 3S23 5.21 E-09 TNK1S03 2.51E-08 TNK1SO4 4.71E-09 TNK1S05 2.06E-07 TNK4S03 2.02E-08 TNK4SO4 4.61 E-09 TNK4S05 3.76E-08 T(SG)S20 1.48E-09 T(SG)S21 2.39E-10 T(SG)S22 5.95E-10 T(SG)S37 2.39E-09 T(SW)S19 4.16E 49 T(SW)S20 1.61E-09 T(SW)S21 3.02E-08 T(SW)S22 6.43E-07 T(C)S15 0.00E+00 T(C)S16 0.00E+00

(

T(C)S17 0.00E+00 T(C)S18 0.00E+00 T(C)S19 1.15E 10 T3S2S10 0.00E+00 1of2 DRA F T O/Yorch,,, m f 7

[" - " " jeabenEes with Loss of DHR (FW and F/BJ t

TSS2811 0.00Et00 T382820 0.00E+00 j

e' T182810 0.00E+00 0RA F T l l T182S11 0.00E+00 k

T182S20 0.00E+00 T2TCS14 0.00E+00 T2TCS15 0.00E+00 1

12TCS16 0.00E+00 T2TCS17 0.00E+00 T3TCS14 0.00E+00 T37CS15 0.00E+00 T3TCS16 0.00E+00 T3TCS17 0.00E+00 T1TCS14 0.00E+00 T1TCS15 0.00E+00 T1TCS16

- 1.21E-08 T1TCS17 1.25E-10 NK47CS14 0.00E+00 NK4TCS15 0.00E+00 NK4TCS16. 0.00E+00 NK4TCS17 0.00E+00 Total: 4.59E OS 2 012 DRA F T

v N

e 9

0 I

(I, t

ae.=

g 3

$i:

WI ug l

gI 5

Rg I!

k ::

J J

I 58, i

g-E I

E I

ji

.w.-

8 s :*

$l it d 1

I I I4 il; Ii l

4 N

>=

W W

ed, i

4 98 as O

Ig l

IA I

s E

I'I5

- g

(

j k

E

$*n f

~

v' a

1'/;O?a s'.'6(lGe"A~4 O M arja m e 4 Z_.

.I 00 0

5 tn-o p

0 le 0

S.

54 n

oi -

tce -

jn X

0 i

R o -

0 r

T t

0 a -

4 T

lum-

,n u -

m c

c k

A -

k 0

o 0

o 5

)

a 3

B C 4 y r

1 Ea 0

d Ao 5

S n 0

0 Cc 3

( e TS Uo T

ON

)w F

K x

0 t

(

0 A

CW e

5 R

2 m B

AF i

D LA T

B s

Nr h

O 0

8 I

0 T

T 0

F Ap 2

t A

Tm ros R

Su pi l

D pa f P l Se uF l

Ar t

e ea 0

Wp 0

r loP A A C

5 1

L L C y

AO C L l

d a

e 0

0 e

r e

S v

y 0

t s

o 1

t m

c p

o n

m t,L p

U E

g W

e G

F 4

o S

A 0

r 1

C 0

2 5

[

00 0

0 0

5 0

5 O.

3 2

2 1

1

?E_ EE3E G' E

i

;i l i

I

a 3

~

)

0 0

~

0 5

(

0 0

5 4

y n-0 X

^_

0 R

T 04 T,

n w

o d

00 I

lo 5

)

3 t

ao W__

FC 41 y r

Ea Sd 0

u_

0 An 0

W_ _ ~ ~

(C o 3

c Te U S w__

x Oo h_~~

KN 0

T 0

C A W, F

.e 5

A 2

v.

D LF R

BA B

~

Nr s

Oh 0

I 0

T8 0

A y

2 Tp p

t Sm m

Yu E

P S/

t G

AWr 0

e 0

Ap 5

ro s LA pR 1

pa L C uF _

b A O

,Y gSe C L t

\\

ea oP _

l r

0 la C

0 e

d 0

e S

r 1

e to_

s m

o\\

v W __

l y

cn g

u F

0 e

A_

0 6

r 0

o c

5 00 0

0 0

M 0

5 0

S O

2 2

1 1

g t33 ga.

jll)il'

8 0

0 C

0 5

('

0 0

,I 54 X

0 R

0 T

4

,n w

o d

0 0

l W

o 5

)

3 a o t

EC n_

i o_

N 4 y p

1 r l

e_

Ea o.

h S

Sd 0

n n.

0 Ao it a

0 Cc l

3 TS j_

e c

(

e e

k _

c UOo o_

A

)r T

r N

t 0 h F

a K

l

(

u 0

A C W, e

m_

5 R

u.

2 m T

A F B

Oo.

i c

L A c _

T B

A s

Nr 2

d O h 0

8 0

IT 0

A p, 2

Tm Su

^

w-YP Ar 0

Wp e

0 A

5 LA 1

LC t

AO so CL L

l W /,f N 0

a 0

e F

0 A

S 1

m hB y

g tr 3

os pi 0

8 l

pa 0

l 1

uF 5

'S e d

t ea ea r l oP C

C I

0 0

5 0

5 Q

2 2

1 1

n$ae:ELgt i!

,il!iI

l, ii

I 9

$y t

0 T{3 0

0 5

0 0

54 0

0 0

n 4

w o

lk oo 0

0 C

53 Cry

)

4 a 1 d En So 0

Ac 0

C. e 30 S

(

To t

UN e -

o T

O p -

)

K W, r

F le -

0 h T

(

F CF S

0 A

e R

A AA n -

5 o

2 m D

Bh i -

i m <.

L r

T tc e

N 8 jn -

i O p, ro 0

ITm a

t 0

Au lm-u 0

TP 2

Sr u

c e

c -

Yp t

A s

A o

WA L

0 C

W 0

A O 5

F L

1 L

A LAl a

l Ce es S

se 0

m Vls 0

r ia 0

y oF 4

1 g

t c

tr 0

d os a

8 e

pi e

l r

pa R

4 e

uF -

v S e /-

0 o

et 0

c a

r l n

oP 5

U/

C 4

e roC 00 0

0 0,

0 0

0 O

5 0

5 2

2 1

1 7TeS !En. $

i

!i

i j

I) 5

)

.s

\\

i i

Fax from Union Electric Company on April 27,1999," Issues Related to Qualification of Delta-LERF Due to Installation of Electrosleeves" ATTACHMENT 2

APR 27 '99 12:50PM E GQEsMK1 W.M d

ELECTRO 8LEEVE LICENSE AMENDMENT

!ssues Related to Quantification of Al FRF Due to Installation of Electrosleeves

1) Approximately 25% (1.63E 4 yr') of the total'CDF (6.61E-6 yr') of the candidate sequences for induced SGTR are non-station blackout oore damage sequences. These sequences could be mitigated by injection of firewster into one (1) steam generator, pursuant to FR-H.1. Core damage would be precluded if firewster were successfully injected into a steam generator. This success path is not credited in the CDF estimate noted.

above. Using an estimated probability of 0.13 for failure to inject firewster into a steam generator, the non-SBO contribution is reduced to 2.12E-7 yr'.

2) The highest frequency (2.75E-6 yr') SBO core damage sequence, which could contribute to the frequency ofinduced SGTR, is T(1S)S26. This is a station blackout, followed by failure of the turbine-driven auxiliary feed pump, and failure to recover AC power within 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months . in actuality, core damage will begin at approximately 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months . The 1-hour timeframe was chosen, for CDF quantification, to provide approximately 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months for the operators to start and align mitigating systems, such as ECCS, to prevent core damage. For quantification of the potential for induced SGTR, use of the full 2-hour timeframe is more appropriate. At 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months , core uncovery begins. However, the SGs (secondary side) are still pressurized. If AC power is recovered

(,

within 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months , operators can (nject into the RCS with the ECCS, and begin a controlled injection with the motor-driven auxiliary feed water pumps into the SGs.

NUREG-1032 is used in the Callaway PRA to determine probabilities of AC power recovery following station blackout. The probability of failure to recover AC power within i hour, used in the quantification of CDF for T(1S)S26, is 0.389. The probability for failure to recover within 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months is 0.222. If the 2-hour timeframe is credited the frequency of T(1S)S26 is reduced from 2.75E-6 yr' to 1.57E-6 yr Removing three additional conservatisms that exist in the calculation can credibly reduce the 1.57E-6 yr' value further. These three conservatisms are:

a) The diesel-generator (DGN) fault trees model a direct dependency of the DGNs on diesel building HVAC. However, AmerenUE has previously determined that this dependency does not actually exist, unless the outside ambient temperature is greater than 65'F, Using data in the Callaway FSAR Site Addendum, the probability that the outside ambient temperature is greater than 65'F is, conservatively,0.583.

APR 27 '99 12:SePM uE NUCLEAR p,3 J

[*

b) Test and maintenance unavailability point estimates used in the updated Callaway PRA are based on data from the IPE (1/87 - 5/90) and Cycles 7, 8 and 9 (11/93 - 5/98). A review of this data shows that trains of equipment were out of service for test / maintenance less time in Cycles 7, 8 and 9 than during the IPE data collection period. Accordingly, more representative test / maintenance point estimates, for the diesel-generators, turbine-driven auxiliary feed water pump and essential service water l

pumps, can be derived using the Cycles 7,8 and 9 data only.

c) A review of cutsets for sequence T(1S)S26 reveals cutsets with both the turbine-driven auxiliary feed water pump and a diesel-generator in test /

maintenance. In fact, the plant configuration risk matrix (contained in plant procedures), developed pursuant to (a)(3) of the Maintenance Rule.

identifies simultaneous outages of the turbine-driven auxiliary feed water pump and a diesel-generator as being undesirable from a plant risk perspective. Accordingly, concurrent planned outages of this equipment would not be undertaken.

T(1S)S26 further reduces the frequency, from 1.57E-6 yr'y ca Factoring the above three conservatisms into the freq'uenc to 9.58E-7 y('.

In addition to the above discussion, MAAP runs for this sequence show that the steam generator tubes will not reach the requisite temperature for induced SGTR. (Refer to item 5, below.)

3) The second highest frequency SBO CD sequence, under evaluation for Induced SGTR potential, is T(1S)S10. This sequence is a SBO, successful operation of the turt>lne-driven auxiliary feed water pump, successful cooldown and depressurl::stion, but AC power is not restored within 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months .

The CDF for this sequence is 1.93E-6 yr'.

MAAP runs for Callaway show that, given this sequence, the time at which core uncovery begins is actually 20.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months . The failure to recover probability for 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months is 4.44E-3 (based on NUREG-1032). NUREG-1032 only provides data for times up to 18 hours2.083333e-4 days 0.005 hours 2.97619e-5 weeks 6.849e-6 months . If 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months is used as a surrogate for the actual 20.7 hour8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months timeframe, the failure to recover AC power probability is

< 1.22E-3. Therefore, the frequency for this sequence can be reduced to <

5.31E-7 yr5.

Also, per item 5, below, MAAP runs show that the SG tubes will not reach the temperatures required for induced SGTR.

4) The RCP seal LOCA model, used in the Callaway PRA for sequences in which seal injection and cooling is lost, assigns a 90% probability to a leakage rate of 21.4 gpm per RCP. This leakage rate is insufficient in most scenarios to depressurize the RCS and thereby prevent induced SGTR from occurring.

F-APR 27 '99 12:59Prl UE NUCLEAR p,4

'..I However, larger postulated RCP seal leakage rates would depressurize the RC8 such that induced SGTR would not occur. It is important to note that Callaway CDFs are calculated assuming either (1) a 21.4 gpm per RCP l

LOCA or (2) a 90% probability of a seal LOCA of that size.

5) The discussion ofincreases in ALERF for the purposes of the electrosleeve amendment should be limited to proposed scenarios where the steam generator tube temperatures fall between the failure temperature of an Alloy-600 tube (~1400'F with no tube degradation) and the failure temperature of an electrosfeeved tube (~1100'F with a 2 inch axial flaw). At temperatures above 1400*F, no increase in ALERF occurs since any tube (including the original Alloy-600 tube) is predicted to fail. At temperatures below 1100'F, no

' increase in ALERF occurs since no tubes are predicted to fail. This narrow temperature band limits the scenarios which would lead to an increase in ALERF.

The MAAP code was used to estimate the steam generator tube temperatures for the sequences identified in items 2 and 3 above. Figure 1 provides the steam generator hot leg tube temperatures for sequence T(1S)S26. Figure 2 provides the steam generator hot leg tube temperatures for sequence T(1S)S10. Both figures show that the steam generator tube temperatures remain below 1100*F and thus result in no increase in ALERF.

Conclusions Based on analysis of issues related to quantification of ALERF due to installation of Electrosleeves, the following conclusions are drawn.

The frequency of Callaway core damage sequences that may preceed induced d

SGTR is very small (appioximately 1.70E-6 yr ). In addition, this frequency would have is be multiplied by the conditional lneremental probability of induced tube rupture of Electrosleeved tubes. MAAP runs for Callaway show that, for the dominant relevant two station blackout core damage sequences, the steam generator tubes would not reach the required temperature for Induced rupture of either the Electrosleeved tubes, or Alloy-600 tubes. Therefore, the conditional probability of induced tube rupture can be taken to be zero.

4 i

(

j I

F-EM 27 %$ 12:bba S!g Rugyg1 te.g3 l

l 1

I 8

S t

8 W

8 e

8

.E g

e=

/

8 I I

.N.I

(

g BI 8

n.

R gg.

fa 1

(

s 8

3 n

'g

(

vs T

8 8

I I

E R

I I

N

!E

~

d) angeJedwel seesqni 1H DS

~

G N '99 12:59PM UC NUCLEAR P.6 8

E 1

l 8

1 g

l l

s

r 4

8 g "d

.W i

ru a

8 5

N IE.

I 8

l N$

e 8

8 e

g

.R R

L Ig 1

i s

d T

i h

i a

'8 O

1 N

8.

8 s

5 8

8 8

8

'8 8

l l

R I

I c_

~ m.%

+

w-m e e -4

+

y

.,.+4.

- - + +.

"r e

s LIST OF TELECON PARTICIPANTS APRll 29,1999, 9:00 a.m. (eastern time)

Union Electric Company K. Connelly A. Passwater D. Shafer N. Staten NRC M. Gray S.Long G. Parry ATTACHMENT 3

!

ML20207A740

Text

Navigation menu

Search