ML20024B298

Requests Managers of Plant Integration & QA Review Preliminary Rept of Safety Concern, PCS 10-78 Re Incorrect & Incomplete LOCA Analysis of Small Breaks

ML20024B298
Person / Time
Site:	Crane
Issue date:	04/25/1978
From:	Taylor J BABCOCK & WILCOX CO.
To:	Karrasch B, Mackinney A BABCOCK & WILCOX CO.
References
TASK-06, TASK-6, TASK-GB GPU-2429, NUDOCS 8307080310
Download: ML20024B298 (9)

Text

~

~~

,e UMf

..L

,F,. j.

yg.h THE BABCCCX & WILCOX CCNPANY P M ER GENERATION GROUP

3. A. Karrasch. Manager, Plant Integration A. L. McKinnev. hnater Cuality Assurance m

~s.

Froe J. H. Taylor, Manager. Licensing (1317) (/

/A.

Cu s t.

,,,,m All 177 TA - 14vered loop

/

File No.

203 ;4,4

/

~

or Ref.

Psc 10-78 Subj.

Prelimir.ary Report of Safety Concern PSC 10-78 DateApril 15, 3973 gn.

P5c 10-78 concerns an incorrect ami incouplete LOCA Analysis of saal.1 breaks.

i

~

Pursuant to procedure NPC 1707-01 Licensing has coupleted its evaluation of the subject FSC and concludes that this concern does represent a L

defect as defined by 10 CT121.

p Due to the nature of this concern. the attached report has already been e

sent to the NRC.

The investigation of this concern le in progress.

technical details and possibly the overall results of this matter r.ay change The se a result of this investigation.

S All results to date however confirm the basic nature of the concernt i.e., a violation of the criteria of 10 CT130.46 in the absence of some corrective action.

The Manager. Plant Integration, and the Manager. Quality Assurance, are requested to review the attached report, signify concurrecce or non-concurrer.ca eign, date ami return this sheet to Licensing within one week of t date; a detailed explanation should accompany any non-concurrence.he above Licensing. require additional information, H. Bailey (Ext. 2678) is the contact inShould you g

Plant Inteersefon Snaear Actfon t

Concurrence P

Mowncurrence Signature Sate CA '8.av.aeer Aerton Concurrecca Non-concurrence

~

Signature _.

cate Os.)%, Yd!

?I

.y 5-w, t v. aw q

.%4 Attachment e(* urw,1. UOf~

f;$ih CCs E. 1..Kane D. Mars 2acord Center 9

9 9.4 p

o l

3 05 [0 2 4.' y 15 No~0i ze 8307080310 790425 PDR ADOCK 05000289

. P

. ' V.* ".**: _... =* L. *:'. ^. ' ~ - ~

~-:

- - - - - - - - - - ~

L J

f..*

EVALUATICN OF 177FA LC'.TRED !.CCP ECCS CCNCEE.4 R

This report docu=ents the evaiustion of a concern wherein

$ L it was postulated that for 35W 177FA lowered loop plants, the i

analysis pres'ented in 3AW-10103A, "ECCS Analysis of 33'i's 177FA j

Lowered-I. cop NSS," may be nonconservative for a s=all break j

In the reactor coolant pu=p discharge. The purpose of this report is to docu=ent the background and reasoning that led to the conclusion that this matter is reportable under 10 CFR 21.

[

)

f.' '

in m

.o v

r c.

2..

w

.s a

r.

[,

3 05[0 2 Ar5i E 15 2~Q T--

4

[

f._.-

m.

- *6 es e e

e-940nh*=*ee

9 I

?

1 IDENTIFICATIC*i The preliminsry safety concern proposed that a s=211 break un the dischstge side of'the reactor coolant punp =sy produce

,3 unacceptable results based upon assumptions used in the analyses.

The affected plants is:1ude:

Occace 1, 2 *Ed 3

[

Three Mile Island 1, 2

• /.

Arkansas Nucicar Cne - 1 I.

Crystal Riter 3 L-Mid12nd 1, 2 g

Rancho Seco a

.2 u

Q

~

v s

3 05 [n _4' 61 E 15 W__ o' E, l

02- --

. ~.

e.mo e u

.em

- +.

- ~-- - -- - - ~ ~ - - ~

O.

h ANALYSIS OF OCCUR."E? CE

~

p Recent analyses of sma11' breaks with cross sectional areas 2

of approxi=stely 0.c4 f t have indicated a change in calculated results, compared to those previously reported in 3AW-10052 and BAW.10103A, Rev. 3 for the 34N 177FA lowered loop plants.

These recently calculated results indicate violation of the ECCS acceptance criteria of 10 CFR 50.46 under certain unique cendi-ticas. These conditions, or analysis assu=ptions, are:

Y 1.

Break si:e: The break si:e must be on the order of =0.04 ft2 to that system depressurization,(no operator action) to fh pressures at which the LPI system beco=es operative,is i.

very slow.

y f

2.

Break location: The break.must occur in the cold leg i

P ping between the high pressure injection no::le and the L

reactor vessel inlet no::le and must be oriented at the r

bottom of the cold leg piping.

(This break location mini =i:es the effectiveness of the HPI flow is that a 7_,/

Portion of the total HPI flow can be lost directly out the break.)

.3.

A Loss of Offsite Power: With this assu=ption, operation -

of the HPI systes is dependent on emergency power supplies.

Only 2 of the 3 available MU/HPI pu=ps are, in general.

supplied with emergency power, and therefore only 2 pu=ps can be assumed available following ESFAS sctuation. With

.offsite power, the additional flow frem the spare HPI

. pump would be available to the operator for accident

. mitigation.

4.

A single active failure: A single failure,which =ust be

-either the diesel or a component (pu=p, valve, etc.) of the HPI syster. occurs, so that the HPI train supplying ~

I water to the intact RC loop is lost.

v 5.

No operator action.

r 305(,0 2. 4*7.1

~

E15 TSc7 n.

....~

. -.. ~. - - - - - - - - - - - - ~ ~ ~ -

- - - - - ~ ~ -

~~

" ' ~ ~ ~ ~

)

g ANALYSIS OF CCCURSE.'!CS (Cont'd)

/"$

The combination of these conditions; small LCCA, specific locatied and orientation, no offsite power, single failure, and no corrective operator action is extremely unlikely. Addis tionally, identified conservatisms within the LOCA evaluation criteria create the unfavorable result. That is to say if these factors we,re evaluated realistically, no adverse conse -

quences would result. Therefore, although we believe this analysis is reportable : under 10 CPR Part 21,no

~

4

. compromise of public safoty has been shown.

Of the five conditions presented above, only the break

' location has been changed in,. this recent analysis. Previously,

• the small break spectrum anilyses were performed. for postulated breaks in the cold leg pipes between the SG and the RC pumps.

The break location in the previous analysis was based on 2

the results of a sensitivity study for a 0.1 ft break which

~

l indicated more ' severe consequences for breaks at th's pump suction as compared to the pump discharge. For break sizes 2

3 0.1 ft, the RCS will depressuri:e due to large leakage rates

• to a value where the CFT and LPI systems become functional.

'It is evident now that the 0.1 ft2 break. location study. is not

. valid for smaller break si:es where the CFT and LPI system do._

-asot become operative for long. periods of time.

.An examination of Figure 1 shows why the pump discharge

..bre'ak is a worse case for breaks where the HPI is the predominant protective system. Consider a break at the pump suction. Two RPIs would normally be actuated, but in the evaluation..only one is allowed because of single failure. Still, because of the pump geometry and HPI no::le location, more HPI water will.

~

. flow to the reactor vessel. Thus,100% of the actuated HPI can 4

.be used for core cooling when the mixture level gets below the RCP casing. This f1'ow is sufficient to provide continuous core.

cooling. Now consider a break at the pu=p discharge. Any HPI water injected into the broken cold leg will pass by the rupture a

prior to vessel penetration. This flow will be directly swept out the break and thus not be utili:od for core cooling.

Because 3

.I_

3 05 }0 2, 0>8!

E 15 T9 97

-~e.

ew s

w v

,--p*g*

wr o

,,------*g rw

.ai-p

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

O k

A.4ALYSIS OF OCCURRENCE (Cont'd)

/~5 of the single failure, the HPI in the unbroken loop will not be f

functioning.

The HPI attached to the broken loop will inject 50%.into the intact leg and 501 into the broken leg. Thus, only 50% of one !!PI is available for care cooling.

The.04 ft break at this very unique location appears to be 2

f very near the la.rgest si:e break in which only the HPI system would be utill:ed, and thus the =ost limiting small break of this l

The loss of 50% of the available HPI flow due' to the category.

combination of conservative analysis assu=ptions for this break results in core mixtur,e levels at the top of core at approximately 1700 ' econds following the p,ostulated event. At this time into s

the transient, the reactor coolant system behavior is analogous to a boiling pot.

In this = ode, accident mitigation requires injection of water at a rate equal to or greater than boil off.

With no operator actions and the present analysis assu=ptions, the

~

required match between injection rates and core boil off rates l

]) will not occur until approxi=ately 3:00 seconds.

I Insufficient core cooling is thus currently predicted for a sufficient period to lead to a violation of the criteria of

~

10 CFR 50.46.

The lengthf of a pas,tulated transient o'f this nature' and the abundance of accident indicators (Iow RC pressure, low pressuriter level, high R3 te=perature and pressure, high R3 radiation icvel, etc.) and equipment status (ESFAS actuation. HPI flow indication, etc.), however, provide ample ti=e for operator-initiated corrective actions.

s..

' s.- -

s

~

305}024,'9.'

E 15 j 't o 9

-^ 1

-2

^

- e-N e

1. 4.>

mem e w w

. - - ~..

m L

lH i

CORRECTIVE ACTIO.T O

B4W has identified and is actively evaluating potential solutions to this small break LCCA problem with our customers.

i To data, these solutions pri=arily deal with potential va'ys to increase effective HpI flow (flow to the RC cold legs not containing the brea);) or to depressuri:e the RC system to r

obtain LPI flow. All solutions assume a.04, ft break at the 2

Pump discharge, a loss of offsite power and a sin:le active failure' which results in the loss of one H7I train. Four Potential solutions are being examined:

t 1.

Opening the HPI discharge crosssconnects and infection valves to permit more "of the flow from the operating pu=p t

to enter the reactor coolant system in the unbroken cold legs.

s 2.

Actuating.a standby HPI pump by connecting it to the operating auxiliary power source.

(

, 3.

Opening the atmospheric dump valves to more rapidly

]-

depressurize the pri=ary system, thereby resulting in additien-al'. Injection frem the core flooding and low pressure injection I

systens 4.

Opening the pressuriter relief valve to depresurite the'

~

primary system in order to activate the LPI system.

l The results of the e~ aluation to date have shown that actions '.'

v 243,1f initiated. by 20 minutes. provide continueus core coverage and result is no~ cladding temperature excursion.

Action 1 will

, rovide continuous core coverage for power levels up to 2563 Mft.

p For higher paver levels, temperature excursions may occur.

Once l

specific plant conditions are accounced for these te=peratures are likely to be below 2200F. M: tion number 4 will nest likely '

provide acceptable mitigation but has not yet been specifically analy:ed.

i i

t

~l

~

305}025'0.!

E 15 4 9 o 9

~

I

~

. ' ~ ~ ~ ~.

. : ~...... a.....

O L

I 34W believes there is an extre=ely low probability Q

that a break of this specific size will occur at the bottom portion of the pump discharge piping at the sa=e time one HPI string is assu=ed not to function because of single failure assumptions. Therefore ve believe that nor=al plant "

operations should continue while the investigations continue.

We further believe that operating procedure changes can readily be imple=ented that will provide adequate protection

[

to the core in the unlikely event of s=211 break LCCA described

,herein.

r l

/

i.

REPORTA3ILITT This concern is reportable under Federal Regulation 10 CFR Part 21 for 34'.f 177 FA lowered loop plants.

f It does not affect 145 FA plants, 205 FA plants or Davis 3 esse 1, 2 or 3 since sna11 break analyses at the pump discharge have been complete'd and are acceptable. These acceptable analyses are reported in:

BAW-10074A Rev.1, "!!ultinode Analysis of S=all 3reaks For 34W's 205-Fuel Assembly Nuclear Plants E

~

With Internal Vent Valves" 3AW-10075A Rev.' 1. "Ifultinode Analysis of Small 3reaks For 34W's 177-Fuel Assembly Nuclear Pla' ts With a

Raised Loop Arrangement and Internals Vent Valves" 3AW-10062A Rev.1, "Witinode Analysis of Small Breaks For 34W's 145-Fuel Assembly Nuclear Plants With Internal Vent Yalves"

..v i I t

l i

!s

}

e-

= p...

I 3 05. 0 2 5_- If

[

E 15 }.$ a 0 ( =-

.-1

l

'g 1

';,i1i yi ';;. :;; I?

1 i,

1

- 'i.

il i

.{

e ll

r.

ig.

s*

c-C GI I

t s

N@

co e

s h A A P

r f A

M T

t A

U H y

.t R

f

)

~

O.

R C

F

..s b

R t., 7,

(

F a

(

n.-

I U

BD t

a r i t

es J t

m.

ac kh F e_.

e r

F,

)w,

'e (t

b rt s

(L c.

v.

I f

e v

U-gw-a,

.C t

o gA h

.r t

g a..

r e

r

),

n E

n..

)S g

B(T E

w M

M"

• l R

F E

I (f

.D*

t R

r q

sC

.t P

U M

v.

j N

R

~

~

mG et t

I s

m S

u(

r r s u

y rt e

A r t

r A io y

M

)

e t

n

)

B r

e a

k

/

i 0 O n MO. lG O. N'.s ddn 1 9(o f

)

t r

I 0

E n

.t e

i l

l

.{

j

~I i, ) l_

e t

li

'f14,Ii i

'?!

i.;*

'f[

i