ML20024C812

Requests Resolution of Operator Interruption of HPI During low-pressure Transients.Memos Dtd 780209 & 18 Encl

ML20024C812
Person / Time
Site:	Three Mile Island
Issue date:	08/03/1978
From:	Hallman D BABCOCK & WILCOX CO.
To:	Karrasch B BABCOCK & WILCOX CO.
References
TASK-03, TASK-3, TASK-GB GPU-0571, GPU-571, NUDOCS 8307130355
Download: ML20024C812 (4)

Text

t .

g S' 4

4

-h 4a

, . 1

. t .

- i f

~ ,

9 e

.A f

• t- I , n. l

1. , - i+

• e e -

. . p.

.4.

,, 4

/ e y

= s s . '

S .. .,

b ' ,

t' <t

-t 9

• s , ""

,*~

e g' #

' # *ae, a n a y y s s

N

.. 1

' 94 . s j

- I-.

-' .a- .

e p np - f,* j.

I 'N.  %- , , .

• l:

s ,

. , . . , i ,

• ?: .

7. _ u 8 ,(

= g

.* g, e .

4 ,,

E 4-e

• y 6 1

s 4-

,.T.

' t

.c 9

- - . + *

}

e

~#

s' -

i v.

4,.

4 e.

6 V

e a O 4 4 ,

(

1. M l re.~ ,

e * * 'r' s

, ' 1 s s' a $e O,

- i ,

)

-g a 'y ,.

9 ~

6 T

• e\

t A-.w.,

, C n ; ,, .

S- -

g _

& g w v em ,

• e 9

, j .

g g e,

8

' 3' . #- *

, *'* < t ,

- , e: - ,

o

, r ,

(,

6

~

F 1

- - 7

. J.

- . e o

el g.

4e e

4 3

4 6

.a f

e kfhf0cnosooogg, 305 780803 Q-P HOL y_

. s. ..

h $' O THE Bn6 COCK & WILCox CO:@ Ally I, h, $$uoj

}

'  % .B. *n. street' POWER GE :ERJTid ! GT:00? n. :1. Quan To l .l. F. Walrces

h. A. Karrasch. .enager, v Plant Integration i From sos 463.s D. F. Hallman. Manager. Plant Performance Services Section (L149)

File No. .

Cust. or Ref.

Date.

Subj .

Cperator Interruption of High Pressure injection (5PI) 'hugust.3.1978

. .. w .a,.

l ru. i . . . . .. . .. . .. .. .. .. ...%. . .. ,

Enf erences: (1) 3. M. Dunn to J. Taylor, sama subject, 7ebruary 9,1970.

(2) 3. M. Dunn to J. Taylor, sans subject, February 16, 1973.

References 1 and 2 (setached) recoc:nsnd a change in BW's philosophy for RPI systen use during low-pressure transinnts. Basically, thay recomnend 1 caving the epi punps on, once HPI has been initiated, until it can be deter =ined thac

• tbe hot leg temperature is more than 50*F below isat 'for the RC3 pressure.

Neclear Service believes this ccde can cause the N.S (including the pressuricer) t'. rough the to go solid. The pressurizar reliefs vill lif t, with a water durgs discharge piping into the quench tank.

We believe the fol.lowi=g incidents should. be evalusted:

1. If the pressuhi:er goes solid with cne or mors RPI punps centinuing to operate, would there be a pressure spike before the reliefs open which could esuse damage to the RCS?

2. tihat damage would the vntar surge chrough the relief valve disch:rge piping and l

quench tank cause?

~

To date, Nuclear Sarvice has net notified our oversting plants to change EPI policy Yet, the consistent with References 1 and 2 because of our above-stated questions.

references suggest the possibility of uncovering the core if present EPI policy is continued.

tic requect that Integration resolve the issue of how the UPI system should ba usd.

l

' We are available to help as needed.

n $.e. .c..

<.';!:, , u ..

D. F. ilnlinsn nmt r..u CONFIDENTIAL

.u. t.n,w .u .

I EL ONLY T2591.s l

\ --_

\

l

_.m_s.--- . . . . _ . . . . . _ _ . . . . . . . . _ . . . . . . .

-o

3 f s .

"7 f h Mg" '.~ ,lsJ r.-

. ' 'ig ,

, f. ' t:.r .J.'.SCOCX & WILCOX COMP A.oy -

t mi R GEi!ERATict! GROUP

! To } . /,

Ji= Taylor, Mananor'. Licensina 3!3 - ' N l

• ** **3 8 BertM.Dunn. Manager.EccsAnalysis(2138)[

- fil' M8.

hust. -

or Ref.

Subj . . . . .

. Date. . . . .. ..

Operator Interruption of High Pressure' Injection- i february 16. 1978

m. s ... i . .. . .. .= d = ,.

l .

In review of my earlie: ce=o on this subject, dated February 9,1978, Tield Service has recoiunended the following procedure for ter=i=ating high pressure t injection folicuing a LOCA. . ,

l 1. Low presrure injection has been ac:uatad and is flevir.g at a rate in ex-

< cess of the high pressure injec:Lon capahility and that situation has been stable for a period of c1=e (10 minutes). Sa=a a.s previously stated. , ,

s, . s.

2. At X sir.utes following the initiation of high pressura injection, ter=ina-tion is allowed provided the hot ic;*ta=perature 1:dic2: ion plus app;cpri-ate ins::u=ent c ro: is =cre than 50 y below clie saturstien)co-.perature corresponding to the resc:or coolant sysca= pressure icss instru=ent er:cr.

I is a time lag to preven: the ter=ination of the high pressure injectie:

ta=ediactly following its initiation. I: requi:es fur:sse work to define

. its specific v:tlue, but it-is probable that 10 minu:es .ill be adequate.

The v.eed for the deley is ths: nor=21 operating ec:11:Lons are within the abcvs criteria and thus it is conceivable that the high pressure in-jection would be ter=inated during the inicizi phase c:l a s=all LOCA.

I find that this sche =a is accepestle f := tha s::ndpol=: cf p:ever.:ing ad-verso long range proble=s and is easier to i=pic=ent. Tharefore, I wish to i modify the procedure requested in =y first =e=o to the o=a identified hera.

l l

cc: E.*'. . Svanson i J-D.H. Roy ..~-

3.A. Ks :as-h b, n.,'S 7 'g' H.A. Bailey ,i ' j^, , [

J. Ke117 E.R. Kane ~

f ,l l

/* ,/.0";y

)

J.D. Ass: ,

{ R.L. Piet=:n / .

l J.D. ?!.inny l T. Scot l

T25913 l

l t

i t .

. . O Q',f % q.f" xp?'

. CAP. COCK !. WILCOX CCMPANY , 's-

.UtlER GEllERATION GROUP

/b//J I To l

• 1 Jim Taylor, Hansace. Licensing

,from .

f Eart ?!. Dunn. !!anager. ECCS Analysis (2133) /M .M ses m.s Cust. File Ho.

or Raf.

Subj. .

Date .

Coerster Interruntion of Hizh Pressure Inhee:ien Februarv 9. 1878 g m. , i .n., . 9 This =e=n addresses a serious concern within ECCS Analysis about the po-cential for opera:or ac: ion to ter=inate high pressure injection follev-ing the initial stage of a LOCA. Successful ECCS operacica during s-m

. breaks depends on the accu =ulated reae:or coolant systen inventory as well as the ICCS injection :s:e. As such, it is mands:ory thst full injeccion flow be maintained from the pois: of emergency safety fe::ures se':uation system (ESTAS) actuation us:11 the high pressure injectica rata can fully compensate for the reac:or hes: load. As the injection rata depends on the reactor coolan: system pressure, the ti=e at which a cc pensating match-up occurs is variable and cannet be specified as a fined r)u=ber.

It is quite possible, for ena:ple, that the high pressure i=jec 1:= =27 successfcily catch up with all hea: sources at ti:n e and t'st due to sys--

ten pressurization be inadequs:a at sece later ti=e c2-The direct concar here rose out of the recen: inciden: a: ' Toledo. During the accident tha operator cer=inated high pressure injecticn due to an ap-psrent sys:e= racovery indies:ed by high level vichin the p:sssurizar.

This actica would have been acceptable only af ter the primary sys:es hcd been in a subcooled sta:e. Analysis of the dats fre: the transien: cu -

rently indicates cha: the sys:es was is a two-phase-s:sta and as such did not centsin sufficient c paci:7 to allow high pressure inje-:: ion :ar= ins-tion. This became eviden: a: sece 20 to 30 =inu:es following cer= ins: ion

~

of inje::1on when :he p:cssuri:cr Icvel again coll:psed and injaction h:d co be reini: dated. During the 10 :o 30 =inu:es of noninjectica fiev they l vere centinuously losis; i=por:=: fluid inven:ory even che.:;h the pres-surizer indicated high level. I believa 1: fertunsce that !aledo was at an exerc=cly low power and ex:re=ely Icv hurnup. Had :his .cv:n: occurrad in a rese:or s: full power vi:h other than insignificant dernup 1: is quite possible, perhaps probable, that core uncovery and pessible fuel damose vould have csulted.

The inciden: points out that we hsve not sup : lied suf ficien: 1:: forms: ion '

l to reactor eper:ters in the ::cs of recovery fre LCCA. T>.e following j rule is based on :n se:c=pt :o slicu ec:minscion of high proscure injec-tion caly at a cine when the resctor cockn: sc:caisin: / sub:coled

ste and :he pres::uri:c is inlica:ing :c luss: a normal L vel for small brenhc. Such coa. .9 :.- near.in:.:.: n11 .: . .w i:y a:. . t : ... .6 -c2 that durin; ..ny Nicu on c :nsien: vould be no sorse chan :he initisi neciden:.  :. :hurcfore, receem.ad chat c7ers:in; pr. cecure be wri::.h I to allev for :ctmin:clen of high pressure injec:lon uncier cae follovi{..? T25920 l two con.:Itlon;. only:

I l

fg-NMg,-)N =O . k - "

• \

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

y (g. . .. . .

!m W 5 -

t 1 t

4 P

1. Low pressure injection has been accusted and is flowin; at a rate in ,

excess of the high pressure injection espebility c=d that situscion has been stable for a period of ci=a (10 minutes) .

, j 1

2. Systen pressure has recovered to ucral operating pressure (2200 or 2250 psig) at d system campert.;uw within the hot le; is less than or equal to the normal operating condition (605 F or 630*F) . .

I believe this is a very , serious matter and deserves our prespc attention and correction. -

32/lc cc: I.V. Swasson D.E. Key .

3.A. Karrasch E.A. Emiley .

J. Kelly E.R. Kane

• J.D. Agsr 1.L. Pite=ss '

J.D. Phinny

) -

T. Scott O .

a 6

l q,

1 8 t J -<

T25921

., -- - - - - , - , , , - , , . - . - - - , - , - , - - - - - , - , - , - - - , - - - - - - - - - - , . ,