ML20024B299

Draft of .04 Square Feet Pump Discharge Break

ML20024B299
Person / Time
Site:	Crane
Issue date:	04/12/1978
From:	BABCOCK & WILCOX CO.
To:
References
TASK-06, TASK-07, TASK-6, TASK-7, TASK-GB GPU-2430, NUDOCS 8307080313
Download: ML20024B299 (4)

Text

- - - - - - - - - - - - - - - - _ - - - - - - - - - - -

r 4*

GPu 293o 4/12/73 V

D 2 A F 7, c

.04 ft, Pues Mscharte 3reak

/

O s

As a result of small break analyses on ene 203FA plants, a conccrn was raised that the worst small break on the 1777A plants =ay not have been evaluated. As a result of this cencers, an analysis of a small break LOCA (0.04 square foot) at the reactor coolant pus:p discharge has just been completed. De results show that the core is uncovered for an extended 2

period of time. ne.most recent previous analysis of the.04 f t break for the 177FA lowered loop plant is reported in 3AW-10102A, Rev. 3. "ZCCS Analysis of 3&W's 177FA Lowered-Loop NSS " July 1977. Both 10103 analysis and the most recent small break analyses were performed with the small brea'k evaluation model described in BAW-10104A, Rev. 3. "B&W's ECCS Ivaluation Model."

Previous analyses of the.04 f t break for the 177TA lowered loop plant showed acceptable results. nese analyses (3AW-10052 and 3AW-10103), however, s

s assumed the break to be at the RC? suction. nese earlier investigations (3AW-10052) dealt with the diff,arances between pu=p discharge and suction breaks. He investigation, however, was performed for a 0.1 square foot break.

Die break being larger allovs the RCS to depressuri:e to a value at voich the LPI and CFT systems are functional. ne result was that the pu=p suction was identified as a worst case. ne re=aining break specertus cases vers then perfor=ed for pu=p suction breaks. It is now evident.that the very sc:all breaks, those for which the EPI is the only systen injecting water, should be 7 1",:,,/,:

-w.-

i-%-

done at the pump discharge.

f ne recently cos:pleted analysis assumes a reactor trip and flev coastdova.

a..

during which time the fuel sensible heat is reenved. Bis portion of the OJ-

~...

5-

...~J

'~" "s^ ;.;[

transient lasts about 100 seconds. During the next several hundred seconds, C

<. a staan pockets form in the upper, or trapped, regicos of the RC3 (reactor CD g

coolant systen) and loop draining occurs. No core tec:perature transient will result until loop draining is finished. For this accident, at about 1700 seconds l

305 0 2 8' OJ

~

[ 1 5 N.. o~=o~.

~-

~

o 8307080313 780412

~~

PDR ADOCK 05000289

~~

P HOL

4 s

~

L 2

r (for a 2772.We power), the fluid levels have fallen to the height of the O

reactor vessel nozzle belt and a boiling pot situatien exists. In this mode.

attigation of the accident requires injection of water at a este equal to or greater than boil of f.

Boil off rates in excess of the injection rates viu cause fluid levels to drop. This analysis showed that injection rates did not equal boil off rates until 8200 seconds for the 0.04 square foot pump discharge break.

As mainreion of Figure 1 shows why the pump discharge break is a worst case. Consider a break at the pump suction. Two dis vould nor=aily be actuated, but in the evaluation, only one is 411 cued because of single failure.

o Still, because of the pump veir effect, all DI vater will flow to the reactor vessel. Thus,100% of the actuated DI can be used for core cooling. This flow is sufficient to provide continuous core cooling. Now consider a break at the puso discharge, specifically en the icver half of the pipe. Any D I

-s) vater injected into the broken cold les vill pass by the rupture prior to vessel penetration. This flow can pass out the break and thus not be utilized for core cooling. Because of the single failure, the HFI is the unbroken loop is assumed not to function. The HFI attached to the broken loop will inject 501 into the intact les and 50% into the broken leg. Thus, only 50%

of one HFI is available for core cooling. This is insufficient to prevent

~

uncovering the core.

The.04 ft break appears to be very near the largest size break in which only the HFI system would be utilized, and thus the most liniting small break s

of this category.

N

'5&*4 has identified and is actively evaluating ;ocential solutions to this small 10CA probles. To date, these soluticus primarily deal with potenfial ways to increase effective HFI flow (flow-to the RC cold legs not centaining the IECA) or to depressurize the RC systen to obtain 121 flow. All solutione 3 05 L~O 2 8' 11 E15'2T e/.

~-

qag e.ee e.

4

,g,a 5

]--

y

. m.. -

y

~ ~ ~ *=

m

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

$ ~ ~ ~ ~'

^ ~ - ~ ~ - - - - ~ -

• 4

?

l.

l i

break at the pu=p discharge. 4 less of offsi:e power and a assume a.04 ft 3ree single active failure which results in the loss of one EFI trais.

O f

1 potential solutions are being

• mined:

l l

Opening the EFI discharge cross-connects and injection valves to permit (1) more of the flow from the operating pu=p to enter the reactor coolant system Actuating a standby HPI pump by connecting it to the operating aM s>7 (2) power source Opening the at:nospheric dump valves to more rapidly depressurize the (3) primary syste.,, thereby resulting in additional injection from the cotu L

flooding and low pressure injection systems.

The results of the evaluation to date have shown that securing flow from a second HFI pump (itam 2 above) or steam generator cooldown (item 3 above),

I starting at 20 minutes in the IAC1, will keep the core covered with a mixtuzs m

v and thus result is no temperature excursion.

L By 4/14/78, these solutions will have been more fully invescigated and S&W further decisions should be possible regarding corrective actions.

t believes that plant operations should be unaffected during the brief period of time required to define corrective actions because of the extreme 1T low

-r probability that a break of this specific size vt11 occur at the bottom portion of the pump discharge piping at the same time one HPI string is assumed not to function because of single failure assumptions.

i

%e*

t P

9 le-305[028'2) s 1s y, g7..

= ~

d

' ' " - ~*-

64e

• -6&

___-__m_

.e _5. 6 ST3 iE 8 E O'ISOE

^

4 3 i

.ate

s..

ym

.O a

Se r W-

-.4, s

h

.-uw g

g

. E =i'., {

wa

~

s u

=.

T i

.t, h

J

~s c

=

T t

)Q u

j g

r f

i e

y B

f

.j 3

-g

xi 2.p4

u... #

9 u

S.

3 A

d a

x i

i s

..e8

.g...*...

s i

i 1

e g

s a.

......v y

t,.

8 C

i i

i, o

.e.

f p

...f}.

• g

e...

.L 3.,.

a

.\\

=

.= s

~ = =

g

$5

,a"

• e w

E

~I t

t W

=.

C

~

l

- \\

b l

a I

g l

... - l 3

il i

e *>.

TR @

N

• 2

_E

'c

9..

a Eu U

i I

e e

3, 9

... +-

............--.--...~...p.

^'

~-

___ _ _ _ _ _ _