ML19329C276

Forwards Response to NRC 750225 Requests for Addl Info Re ECCS Analysis

ML19329C276
Person / Time
Site:	Davis Besse
Issue date:	05/17/1976
From:	Roe L TOLEDO EDISON CO.
To:	Rusche R Office of Nuclear Reactor Regulation
References
NUDOCS 8002120987
Download: ML19329C276 (49)

Text

__ .- . _ . - - - _ --.

., , . jegslatory Docket 5g .

TDLEDO EDISON Docket 50-346 LOWELL E. ROE vie. p,....

v.ca,u o .. s mei ass-sa42 May 17, 1976 s 6v h er> .

/' y pocn'et i.f, W

v C

\Y

%O e==

Mr. Benard C. Rusche, Director , y*d' Office of Nuclear Reactor Regulation United States Nuclear Regulatory Commission 'g l

Washington, D. C. 20555 '

• '  % I v

Dear Mr. Rusche:

i

/  ;

Enclosed are the responses to the Requests for Additional Information '

transmitted in your letter dated February 25, 1975 concerning the ECCS analysis for Davis-Besse, Unit 1. These responses are being submitted to the NRC prior to May 21, 1976 as promised to the NRC Licensing l Project Manager, Mr. Leon Engle. J Yours very truly, r'

y,*

s -.

r - i. -n t.- \

,% ) ' 'd N i 1

dh e/5 2'3  ;. ~ ) ,;~J

~

f.,s- E . s. ,;[JQ x,, + A -,,

, ' 's ' =r, J'

\

,Mj , /~~$,s =t \ "D

\

j l

l l

l SU3S THE TOLEDO EDiSCN COMPANY ECISCN PLAZ.A 300 MACISON AVENUE TOLEDO, OHIO +3652 fooN M f7

. , .s _

Regulatory Docket CV, . ' f' ' '$b -'

D-B

1. Justify the selection of initial pin pressure and oxide layer for the CFT line break (referenced in BAW-10105 to FSAR). Explain not considering the Case 1 power shape previously shown to produce a higher PCT. What value of LHGR was assumed and why?

RESPONSE

Revision 14 to the Davis-Besse 1 FSAR presents the results of the CFT line break. Beginning-of-lif e (BOL) conditions were used for the initial pin pressure (1280 psi) and oxide layer thicknesses

_5 (1.5x10 inches on both the inside and outside cladding surfaces).

No rupture can occur during the flow controlled portion of the transient as the cladding temperature is maintained within a few degrees of the saturation temperature. Once the mixture height falls below the top of the core during the quiescent portion of the transient, the temperature of the uncovered portion of the cladding increases and rupture occurs. This situation is analogous to large i break analyses where rupture occurs during adiabatic heatup. The time-in-life (burnup) studies reported in Section 5.2 of BAW-10102 and Section 5.4 of BAW-10103 shows that so long as rupture does not occur during blowdown, the BOL conditions maximize the peak clad -

ding temperature. Therefore, the values used for pin pressure and oxide layer thicknesses for the CFT line break yields the highest cladding temperature.

The power shape used for analyzing the CFT line break was chosen to maximize the peak cladding temperature. As explained previously, the cladding temperature at any elevation is maintained within a few degrees of saturation until that elevation is uncovered. The use in cladding temperature after uncovering increases with the length of time that the elevation is uncovered. Therefore, peak power locations in the upper portions of the core must be analyzed

, for small breaks. The Case 1 power shape (Revision 2 of FSAR) has its peak power located at 3 feet. Using Figure 6-44 of Revision 14 of the FSAR, this elevation would only be uncovered for 30 seconds.

For the power shape used in the analysis, the peak power location (10.5 ft) is uncovervi for 110 seconds. The larger uncovery time for the power shape analyzed will result in the maximum peak cladding temperature and is appropriately conservative for this accident.

11

.  ; ,m -.

'T D-B In order to ensure a conservative analysis for the CFT line break, a spectrum of possible power shapes expected during normal ,

operation was examined and the most severe case was chosen. The criteria used for this choice were:

1. The peak power must occur at the highest possible elevation in order to maximize the uncovery time.

2. The maximum peak power expected during normal operation at this elevation must be used.

Using these criteria, the power shape used in the CFT analysis was obtained. This shape has a maximum LHGR of 12.88 kw/ft which occurs at the 10.5 foot elevation.

.i

~

1-2

,. - , _ . - - . , , , -s. r - +1e--- _. y

l .

a. , m -

D-B

2. With regard to the single failure analysis in your letter dated September 5, 1975:

a.) The core flooding line isolation valves CF1A and CFIB will be required to have power disconnected and breakers locked open.

b.) Attachment I states that if valve HP1556 spuriously closed during the injection phase of a LOCA, there will be no effect on RPI capability. To confirm that your evaluation was complete, provide the details of your study which considered this spurious closure during a small break which allows RCS pressure to remain above the cut-off head of the HPI pumps for such a time as to compromise pump integrity (due to the loss of the lh-inch bypass lines). Provide the time that you assumed it would take before pump damage would occur and relate this situation to its effect on the capability to meet the criteria of 10 CFR 50.46.

RESPONSE

a. As documented in FSAR subsection 6.3.2.15, after the core flooding tank isolation valves CFIA and CF1B are fully open, the breaker of the combination starter of each isolation valve will be manually tripped open and padlocked. The tripped position of the breakers will be monitored on the main control board by one blue indicating light for each breaker. With the source of power to the motor operator padlocked, there is no possibility of the valves closing with the reactor at power.

b. The spurious closure of valve HP1556 has been precluded by re-quiring that the air supply to the valve will normally be removed, assuring that the valve will remain in the open position. Should it be necessary to close'the valve, it would be necessary to physically connect the air supply to the valve.

/

2-1

D-B

3. Your July 9, 1975 letter indicated thca, with the exception of decay heat suction valves DH-11 and DH-12, no critical equipment is affected by post-LOCA flooding. Provide the level of water (above the containment floor) assumed for the LOCA and include the calcu-lations upon which this value is based. Also, the statement is made that a water-tight " trench" will enclose valves DH-11 and DR- ,

12. Provide a description, with diagrams, of the trench and discuss the surveillance planned to ensure that this installation remains water-tight throughout the reactor lifetime.

RESPONSE

The maximum water level inside the containment has been conservatively calculated to be at elevation 572 feet-2 inches after a LOCA. The water level was calculated using the following water sources: ,

1. Reactor coolant system - Assuming the pressurizer contains 100 ft of water above the normal level, and 60 percent of the water volume in the reactor vesse.1 is lost to the sump, and all other primary system components are assumed to blow to the sump.

2. Core flood tanks are assumed at their normal level and are blown to the sump.

3. The borated water storage tank is assumed full.

4. The makeup tank is assumed to be discharged into the containment.

When all the above water sources are taken into3 account the total volume of water in the containment is 86,100 ft .

The area inside containment which is ut .Lzed by structures and equipment was calculated by evaluating er.ch area and component.

The following areas and components were considered:

1. Reactor vessel cavity - All area above elevation 561 feet was assumed full and the volume of the reactor vessel was-calculated using the volume from the outside of the in-sulation.

2. Volume of the incore instrument tunnel was assumed to be

~

20 percent full to provide for the incore instrument tubes, stairs, and tube supports.

3. Volume of the pipe tunnel was assumed 15 percent full of piping.

4. The normal sump and access tunnel was considered 10 percent full of equipment.

5. The refueling canal was considered 10 percent full of equipment.-

3-1

D-B

6. .The area under the steam generator supports was .

considered 40 percent full of structural steel.

7. All wall volumes were calculated and subtracted from the open area.

8. All equipment above elevation 565 feet, which is located at elevation 565 feet, was assumed conservatively to be located on the floor at elevation 565 feet.

9. Af ter all closed areas were totaled,10 percent of that figure was added for conservatism.

The water-tight cover provided over valves DH-11 and DH-12 is constructed of 1/4-inch steel plate with all sides and access covers sealed with 1-inch wide and 1/8-inch thick asbestos gasket material (see drawing, sheets 1 through 3, attached) compressed by bolting. An 8-inch line is provided in the cover to prevent cover 4

damage due to excessive Ap during blowdown. The opening for the 1 vent is located at elevation 573 feet-2 inches, which is high enough to prevent possible flooding durir.g a LOCA with the maximum floor level of 572 feet-2 inches.

At each refueling, the cover will be inspected to assure that all gasket material is intact and all bolts are tight.

I 3-2

kM .D-8

^

o. , $

. - . , ,n *r- -<

e.-.e,

. s ' '.,'.,' 5 37 m

s

• .'9,- y m -

7;o'

_ _ ' ' , 4'

• o*

1 i 1LOh , .!

16' '-5>

l 13

-v m ,y ,

, , i

. ' r3 Re .Lp ,

u. 51.S/

wA ln A- i . -4*

7 "w . i .J - N ,( e t.p3 r ' '

lT

~

i ,'

7 NJ Nd- ' >

yR "- (S! NR a

'; o C !! S ,

i 9

en  %- '

f ,

g- ,. . ' b u o' '

84 PIPE /

HS ' /3,//---je,nr!:':o str .va 7. ', .'i,

, 8. se m

3  :,

_c f__._..______,

g ,

2

./ n m

! se c.a WL -

WG a;-cm.J,)$- ,t

-sol- {;ili 2

s j m_-. o n2 . c i LR4 I ~<

,m n9s B' k il 52;

. , ,i

. eca 't e. r r 4 i-

.' 3-n ml,s!m _ m .m , g 0..

m , .. ,

. F -, - o o.m i . 4 ( E d.

o'; i i

gbo*

. M'"

• i ,,

y

/.s f r .*.

i

-zm j-r

3) _ _ ;: M e! *f#s e -

2 sm'y4 4.n. 'l o.

,it, ,/li x lVc":aunmed -(gDi 3

,j :. ,3 a m:xj:ST4_. j JarocGricurg,j f, q:

.?.:

Akero /

p s 4>, . --w---------;

3f g *,

t c*  : Lp gj dm %

7.

• 'm .o 4(CL / dtS4 :l m -

~

. 1 l w U T

. ACS4 51 ti 4 C 5.4 I d'h

_y '

h, s, I t-4 P'[ t . ,  ?. f'S -

r*

,; . . a ..d 4f S.4_- -

ul 4C64 ,y l,

k-z

' n~. +,. ,

A Ni i ej r,:. cp

?- -

u m ,e7) m\l. 9,:e n, svn -v u ;,.* gy -- - -

h h 'Y r.r,>

uA h.h ;fD .

r g's L y.i 4m e, >

4

..FT y o,4- l g'oi 9 ,

sin

^ <

l h -_psipd.- ,-

-,i 0

Q .

\ f?'1 PIF'E d RlI}

w1.o.satvc I I3* t .f J;3' l'.3 2:2

• REF. DV!G. C-G35 7 g.

DECAY HE AT VALVE PIT COVER PLAN D/WIS - PECS_E -UNIT - 1 SHEET l Or 3

pp he 70 CECT ,@ .

D-g

e .*,* lA b .

& O. CONC. WALL U. ?$ % ~~~~'~~^l- r E. Q' FM MACH. CCMtts EL.EO.iTic' f{"1 *~ l \'La*ptegg 4 OD NEMM

• ~

i NOTE ~ ' '-

Af t4GvAeLE AANE!. gg SEAL ME f.O b)#. TO\ .,,

V g' e * .4

' ' * " ' '

• A,,, e--

' ' ' ~ E

?' C, E b 2 AT INSIDE ALLINST A T O/i:L/5c ut. T' J. .V. ?' li *

. . -- c j/ 4 TOFL EL.SGS:O ./ .ral /

.. SEAL WELO GC TO E. C .l s'AS -ESTOS -

gggy;g 7. , g;_,3 T ^- 8 .

PIM C 7A P rcR

~TS9 MfrCH. EcRsws

-- ... . D. FlATE

• f '56 WEJ ITS ,

". N ~ '

• • ~ '

g , 'q 2. TO l l g v.312 N GE ---- NOT E .f..*_.

'G C B 2, A L L A C CUN D '; _ l',t CHK'O. 2 M W eE IN CONC. Wot L OFNG. >

') s WELDED TO BEAMS N.S WATERTIGIIT Boot' h Op* ..' RECVth E O To HO L O

--.PLA TE S . FL/I T.'

W/CLhMPS

% j( ? p 4 Ib' I.D % P. SPLIT Sl E EVE,

~/t'r  ;. - /

( T1i () .SEAL sect. m WELO SIMILAR SECTlON To @

4-fg 7D_E , g _ ./__ ,

gHjt/ . N.T. S .

. ElSS99' E '

?-

f! * .T -

~ TYR\

/16* f ,,,

CU T HCA 7. C. C AND WELO TO EliEVE GN b.O. PIY EL. SSb'-o*

}%  :

3:>* a.*, % a.

USE '4 s &* hibibTOS GAShiT FOR WATEF CCNtl. ( EXPANSICN ALL ARouND. FACE OP WALL Ph'. 6 8 C E SECTION @ <ge,t WEtc.

USE 'i ASGGSTOS i,',-

L .GASKETSFCA NA7fR'Y '

% ..._  % PIPE. TIGHT CONNECTIONG ,

' CHICO '=<-

N N _ .

.'.- . WATER TISitT EcOT .

, _ *t.fo.,_

, , 3

. WITH CLANP3 . \j - - - - --- .

R o .,

J/ '

\- .a

._x y . ',. / qy 9  :

p T 4'Citiro.12. /

.- w =. .:-- = = , ., . ..f %

. --- y->

-s, ,' ggg

, ,v ,_

'~' ~ ~ ~ ~ ~~ '

'4 PLATE SLE EW ,

SPLIT ATDC Allot.J

' (SEAL WELO , .

. ER ECT'CN OF PLh7'. kI

SE E PLAM FOR DIA. QQN '

NEXIST. PIPE. .

{h5llCLESfonO.d y5417 ;,ngu y n .m..

b g ::10 AMCHORS TO E C cN I?"Cih.'.

REF. DWG. C-60 5 m a To cesT es m o.

SECTION (A3

, c . ,4 . -ss u , SECTION emw-M

,voArv,pEczarnar=:S a y

DECAY HEAT VALVE PIT COVER-SECTIONS

& DFTAll' R-

~ SHEET 2 CF 3

.. '. . ' [ ' W .' I I ' !

i.

O-8 ,

O' '

., .;_ , ** Al WELD) ,, y :D

- e e',l'e, h

' T. O. FL. El. SGS ' O' f

. . . g_,?. a

.e- , . . _.

', 's L. _,Lf'W-5 ' -

• • ' .. r . ', Q '

fQ /.'uV _ '3cicts ,,

I-. - - L' # C O A

W(X/C i' = %' L l'G ,

[se,i '

, ,' 6 C CRd'# ,

I b* u3'u %~ CLIP L .

~*

~ QV W '

'bVGADTdiC'G  %~

2* LG. FOR 4 C'S. NWA LL CC Vt. VE PTT

$=3'ab$^O;G"Lf_([lPAtJGl_g_

• (2) 4b,6 WEr.Ir '

s ON EACI: SIDE, FCR 6 C & b kc.

.. ANCHoit[4'4'E14 SED.

\ Ito' ELBOW SECTION /DN .

. b'* I:O'  :., l 160* ELLOW .

f3 rt MS%_

, n n -

12

~

3f t

i Q DOLT h,r d ' C TP A P To '

f,' a 4' o IQAi.L' T'y;2). Q e ;.jf s .g7 <-2 -h_, l o ' Fxi ANCHofl5,4N* E/4EED. jd v)

Cm g- ,2 SEAL WELD } y .,

pg,g, cS c y go  %' - ~ PIT FLR. El!!ac

. . . , p.A *

.p . . r : t.-

-g. ;  ;  ;'

[

*

s. ,-

q -

.a i,.,a <':}*{-- ,

i u. w U

SECTION

['j N. T. S ,

s ~., .j

. Y*. .Ls i *$

• 4 ' S TPAP D '

w. * ;: L', W

_ f"

(

( A$$ Fil 14ACHINE Scrii/s

% * *'e:.

• . ?h'*'

• Q F. DWu,,, C - G 35 -

zu A7 .D LL6i'.C' c1 6'O.C CYliid~fLATE

..u __

s .  %, c.np. :t t

l REMOVAULE PhNEL

%.# wt> rT ., gy-- -(

pi to*

DETAIL l I; O '

M 2

- l- W /gfj'oS 6*t7

. DorL t i Tho Fer'

. .L.,. -u tstACH ECGEWG g .$: . A,4 _ ,_ 6,,, _ _ 6 . , g 6%. _. f7. NOTEt , .,i j 1---.~ l l l I '

USE COVER PLATE FOR : :. V Q br //.S

" LOCA71 lG. P.H. SCREWG  %. AShEGToS GAjKET

?f. e.

)!

v .1 k.i*._'~-M.'^r:52Y

. 1 2, e y e n s w+rip.'c"=w"a

. q . . .i. 3, , s . .,

j; g,l ; til gv SCFist/$ m's4r.".se"

,o, y

ii i

CTEA'. PLNrE SECTION /Th ve. .e1 .'la:,: ll- 'C%y " . t s- \s" '/

. u; . o' i e .

,, o h, r

_%, i

> _.. :.. ,:: G 1 L

l< , mix.w ccvCn DECAY HEAT VBJ.VE V, .

':. h, 'l ll 4s, 3 !k. id PIT COVER-SECTICNS 2y ..  : -:.e . . .v: . x i J .'.yi \ < c" w a- - x2' ". "# & DETAIL 5

-SHEET 3 OF 3 D.E T. /.:d.I. _. . .T.

. m , . , m. . <. ,. .,,,,,7.,

~

D-B 4a. Provide an analysis of a break in the idle pump discharge.

RESPONSF The partial loop LCCA analysis was performed assuming the worst 2

cas break (8.55 ft DE, C =1.0)

D reported in BAW-10105. A break at the active pump discharge of the loop with the idle pump was analyzed and reported by B&W in response to Question List #10 pertaining to BAW-10105. Further analysis of a break at the idle pump discharge is not necessary for the following reasons:

1. The maximum LHGR for the partial loop analysis is 84.7%

of the LOCA limit value reported in Section 7 of BAW-10105. The peak cladding temperature calculated for the partial loop case analyzed is considerably lower than the 6-foot LOCA limit case in BAR-10105, 1675F vs. 2166F respectively. With a margin of 525F relative to the 2200F criteria for the break location analyzed and the reduction in LHGR for 3 pump operatin, a break at the idle pump discharge is not expected to violate the acceptance criteria for ECCS (10CRF50.46).

2. The core flow distribution for the Davis-Besse 3-pump case analyzed is similar to the core flow for the same 3 pump case reported in Appendix A of BAW-10103, Rev. 1.

Both of these cases have positive flow until =10 seconds, and thereafter the core flow remains negative. Since the flow distributions are similar for the same break location, the core flow for a break at the idle pump discharge for Davis-Besse should be similar to that reported in BAR-10103 for the break at the idle pump discharge. The idle pump discharge break in BAW-10103 has a lower peak cladding temperature than the break at the active pump discharge of the idle loop. Therefore, it is not necessary to analyze a break at the idle pump discharge for Davis-Besse as the same conclusions should be obtained due to the flow similarity.

i l

l I

l 1

l 4-1 l l

1

\

m -

D-B 4b. Explain the double peak in cladding temperature under 20 seconds and explain why the 1st peak is more pronounced in this analysis

~ relative to the 4-pump break spectrum and relative to the 3-pump analyses for.other category plants.

RESPONSE

With 3-pumps operating, 43% of the total RC flow (which is only 75%

of 4-pump flow) passes through the active pump in the loop with the idle pump. Placement of the break at the discharge of this active pump results in a substantial degradation in positive core flow during blowdown relative to the 4-pump case. This reduction in core flow results in a more pronounced first peak in cladding temperature during blowdown.

The first peak in cladding temperature is more pronounced for the Category II plants relative to the Category I plants because of the differences in the plant designs. With the raised loop design of the Category II plants, the hot leg has a greater height relative to the Category I plants. Also, there are only 4 vent valves in the Category II plants compared to 8 vent valves for the Category I plants. These differences in design causes the Category II plants to have a less positive core flow and a more pronounced first peak in cladding temperature. Comparing Figure 6-15 of BAW-10105 and Figure 6-4 of BAW-10103, the same phenomena was observed for the 4-pump cases.

The second peak in cladding temperature before 20 seconds is caused by the reduction in heat transfer that occurs during the transition from positive to negative core flow. This peak is typical for large break analysis of all B&W's plants.

J 4-2

, ,+ * * ~ ~

.,. ., m .

D-B 4c. Provide assurance that the PCT versus break size curve in BAW-10105 would not be significantly altered by partial loop operation.

RESPONSE

The partial loop analysis was performed assuming the worst case 2

break (8.55 ft DE, C D= 1.0) reported in BAW-10105. Historically, the above break has resulted in the highest cladding temperature for LOCA analysis. In general as the break size decreases, the duration of the blowdown increases which results in decreased maximum cladding temperature. Table 6-1 of BAW-10105, Summary of Break Spectrum verifies this statement, i.e., the maximum cladding temperature decreased 354F when the discharge coefficient for a 8.55 2

ft DE break was changed from 1.0 to 0.6.

Since -core flow is similar for 2

breaks in the same location as shown in BAW-10105, i.e. , 8.55 f t DE break at PD with C #""8 "3 *

• D to 0.6, core flow for partial loop breaks in the same location with varying CD should be similar. Therefore, with similar flows, the PCT versus break size curve should exhibit the same trend, i.e.,

decreasing PCT with break size.

O Since the PCT for the partial loop analysis is 491F less than that given for the worst break in BAW-10105, smaller breaks will exhibit 0

large margins of safety relative to the 2200F criteria.

4-3 i

,,- ., n -

v D-B 4d. - Submit the LOCA parameters of interest identified in the " minimum Requirements for Break Spectrum Submittals" dated April 25, 1975.

RESPONSE

The following are additional LOCA parameters of interest for the B&W category 2 partial loop LOCA analysis.

3-Pumps, Break At Active Pump In Loop With Idle Pump (CRAFT Run PP201 (9N) 2 Figure 1. Reactor Vessel Pressure for 8.55 ft DE Break at Pump Discharge During Partial Loop Operation, CD"

• 2 Figure 2. Core Water Level for 8.55 ft DE Break at Pump Discharge During Partial Loop Operation, C = .0 D

2 Figure 3. Downcomer Water Level for 8.55 ft DE Break at Pump Dis-charge During Partial Loop Operation, C "

• D 2

Figure 4. Total Power for 8.55 f t DE Break at Pump Discharge During Partial Loop Operation, C "

• D 2

Figure 5. Containment Pressure for 8.55 ft DE Break at Pump Dis-charge During Partial Loop Operation, C = .0 D

l Computer Data for the Figures Fig. No. Version Name Version Date Run Name Run Date 1 CRAFT 2, Version SPP 4/17/75 PP201 (9N) 8/28/75 2 -REFLOOD2, (Loop) 12/20/74 PR201 (20) 8/30/75 3 REFLOOD2, (Loop) 12/20/74 PR201 (20) 8/30/75 4 CRAFT 2, Version SPP 4/17/75 PP201 (9N) 8/28/75 5 CONTEMPT, Version 15 11/15/74 PC201 (92) 9/05/75 Other Codes Used (Figures provided in original report)

Version Name Version Date Run Name Run Date THETAlB, Version 6F 1/23/75 PT201 (FN) 9/03/75 4-4

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

.. .. u: . 2

,Cm.-.,.,.,........

sauns6 a e...e co

! i -

l l l_

__ j _. j- . '! . .

I i .

l

_l

• l i _.

.. ..I. .

. ._ L.

i r l

l l

'  ; l

- ; ._ ! . _ i .

I_ .u.

!  ; j .

2500-- ,

!i ' ' ~ ' '

l e

i . . , l i ii i i

.. l . , g

.. _a . _ . . . _ _ l .__ j _._ _  : ,

6 -

2 l l Figure 1: Reactor Vessel Pressure for 8.55 ft j l .

i

j i l D.E. Break at Pumo Discharge During f i , i. '-

a .

l i Partial Loon Operation, C D=1.0 j g j i

l ,

y;

• s l

l j

_j _

i e ' .i _ _. , .I i I

. . . . _ . .._.__!_ ._._1___ _. _ f _: .i l!  ! I I I  : _ .l .! . . i l '. ] .

r i . j i i ,

. i J

-/soo.-

Tl i  ;

I I i i . i i i i i , i

. o, .

<* ' i i.

1 .  ; ,i .j. t , i i

.i. .

l i . .i .

q -- . .

_I_  ! .._ i I i  !

i

• ]_ m i'

_. f _

i . , .

i, ,

i

_ l i . . .

g -l - - . ,..  ! . L..

i

!  ! . .I '

i- ._,

c l l

- l .

i

• l 1 .

I .

g _f g ..

a i

t

.I  !..!.  :

j I

I m i  :

i

. . . . I . i .. . .

i . .

i l I :i-- t ,

! i

! l.  !: . i

}

l l i  ; l

--500,-

' l , ,

l i

r .

1

. i i i i . 1 .

1 l a.  : -

i ,

i  ;  :

i

, __ _o,o, ao .% o . e.-_ Go 9o i . /2,o i rio /s.o 2/,o z.t. o .

' i i i i .

I Li  : I  ! i I -

l Time, sec  !

.i , i .

' I -

1  : .  ;

l l  !.

- -.-. - - - - t . . . . - - . . .

I.....,.................o...

.u . . . . . . . . . . .

...... . ..... ~ ,

. , i I

i. i , , g.

i

' I g i

_ _ . _ _ _ .  : _ L_ _!._ ._ I' _. __. _ . :. . .! _I___i.. '

.t _ _ . i _.

i  !

e  ;

s I

- l i

._.i. J ' j i  ; i 2 I

. .. l

' l 8'O

~~~ -

. L l .  : 1 i i ,

j- ~ ,

i

. 1. _ __ _

2 3 .. !

j..

Figure 2: Core Water Level for 8.55 ft D.E. Break j '

at Pumo Discharae During Partial Loon l -

.__;.l_l._ _ _ .. . __ _ . Operation, CD=1,0 l_,j .

i 1 .  ! , i . i  !

! . _ _ ._ . g, p _. ,

' .. _. l

__L_ _._.'_

. l .. ' .. - -

_$ i I l . . _ . .

I  !  ; j l i i  !

. i

i I  !  ! -l !1 l

. i i I i -  : *

, j i i  ! i ,

I i  !. l I. ..i 1 i I.

g _ ' '.f,0 -

! i  ! l i l i i i l I i l  !. !I . .Il .

1, O

i i  %  !! j ' ~[ i ,

.  : l i l '; I- i '

l l '

S $  ;  : .

l t -

i ,  ; .

1

.a . -._ _- ._ . ! _ .  ;

......_- i. . 6 i -

i.l

' I -

o D i I  !  ! . 1, . i  ! l

.' l .i l i

m i . . .

i

! i i i

._. ~ ._ 4,0 I

o f.'

o I

I. i_l .

i, . .ii 4 i

l 1

! t l .'

l I. 'l .i

.  ; . t

! t . -

i . i ,

l l , .

?

i e i

'. , I

,.. _ 2 f) .

+

t i l I  : ,. ,

' i l

.._ ._._. O,0

.f .- .

I I

,I . ,

g 0.0 25,0 .L 50,0 15,0 * /M,0 /25,0 /50. 0 _

l l i -

l >

! i l  !

Time After E08, sec

._...__._..__.___i__. - l 4 l '

I -

. ~ . . . . . . . ~ .

i l

i i l .

I i

_l _ .. . .  ; i..l_l l. l -  ! I

_. . _. . L _ .

I  ! i  !!  : Ii !_ t l I -

l

. I i i i i l i  !  ! i l i l l '

t 2

~

Figure 3: Downcomer Water Level for 8.55 ft -

.. ...i.. _ ._ _ . __

D.E. Break at Pump Discharae During l l ._ -

Parcial Loop Operation, C D .0 l i l .;

._7_,.._ .

.' _./f,0 .

I I f I . .. ! _J _ ._ .,

i i .

l I

l l

c. _. . .

M l  !

f.

! , I  ; i i I i

f. l 5

1' i.>  ; i

. i l .

i

. . i I l l l 1 1  ;  : i

i. . - l l t

i

_ b;jgg. l I  !  !

i

! l l !l li I I !_.__. .. . _1 . .

I O

, e .

.I i i i

j '

i j {

q g ,

_. l i ,

. f u +

- + = .

l

- .e -

l l

l

.  !- ! i l .

l 5

t l

l l i, i I u i I ,

, i , ,

t l

- 8 i

! g,o . j._. l_ __ _

' l l

l I_._I. I  !

8_  !. l i

E l ' ' I i - i '

l l  ! l l i .!  ! .

I  : ,

o 1 6 -

i i '

I i o - l 1 i '

l l l l .

l

. 1 i. l .

- - ~

l l l l i. .' i.  !.  :

. j . .  ! >

j 1 i .

l '

._....:y,0 _ ._i_. - - __. .

l I  ! i

. i  !

l  ; i  ! '

!l l .

i -

l l

l .I l t i . .

l .ao j -

i  ; .-

l O.O 250 _l 50,0 '

75, 0 /g9.0 fj5',0 ffo, O l l t l  ! l l i j . i  ; t  ; 1 i -

I  :  ?

s .

Time After E08, sec I

i i - i -

l

' I J .' .

1 l

. . ._ a l

!....................s.....

. . , . ... 4

' + L .

i  ! l l -

_. . . . . . . _ . . . _ . _._1_ _ _ . . . _ -

. i

- l I  :

. i ,_a .. . _ .. .

I i > I I

! , l .

! ,  ! i 1' .

l' t i l i l I -  ! l I i  !

/,d ,

. . . i . . j j

- _. _ - . . _ l __ - . l .__ _ ; _. 2 Fiaure 4: Total Power for 8.55 ft 0.E~. Break

_ _. _ p l, ___ .'.__

at Pumo Discharon During Partial Loop i Doeration, C =1.0  ! -

1 --

l l D '

i i l t . - -

i l i p

6 - . _ _ . . _ _ _ - . _

I i l

i , i t i . l . .

2.  !._ ..,3_._ _4 3

... f l

.i. l .i j i .

[

i

. . . . , ' g 5 L l l  ; l l.

e

! i 1 l l t

i I  ! t t I l } .

I i ' ' '

• I . ,

I *

, . I .l *

• l 'e6. i . i l_. I __ i i  !  ! ._ l ..

- I l

I '

! O

% g s-

1. . l. .

l  !

1 2 . !. y! a e 3 1 i

i j e . I

e D 3

. . , , t 4 . s ..

o . _ .

o. . i

• l e

• . I  : l l _l. i  ! .

I  ; *,

f I I 3 L L_ t.

8 -

l  : _t.. ,

- .. ' t i e I i . -

.no.~* , , ,

l .

! . 1 ol'- . .

O '

.j  ;

r . i ,

t . . .

I I 8

.. ; _. . _. _ i.

8 _.

'  : i ' ,

. i .

' l l '

i  !

l ' ' -

J_ ... _i.__. . ~

' ' - e

1 1  ;  ! .

. !.  !.  !  ! i. i i ,

_... I. _ ,  ; }

,2 . . '

i ' 3 l

t i g

,. I I i

> l i i . t

' t 8< O  ! .

l 240 i

o.o 340 - 6o 9,0 , /?.o ,

Mo /B.0 260

- e i i t i i e

} . . , ,

• e

- Time, sec .

. _ . _ . i L. i t . ._ _

,_ 1. . .

.__1._... .. . . . . -

. . . ._J

4 0

' 1 i 28.0 l 26.0 .

I  ! l 2

Figure 5: Containment Pressure for 8.55 ft 0.E.  !

24.0 .._,__ Break at Pump Discharge During Partial ll Loop Operation, CD=1.0 l 22.0 - .u .

. , -)

n

\

l w 20.0 _ _.

[ l 1o i

x.  !

-I 11 -

]_ _,l [ *' l

3 L.

18.0 __

g y

/ 3_

i I M

16.0 / \ _._ ll o

4, g _ __.. , ,

• 8 u- 14.C ,

/

r___.

\ ,

_L -l.

i l ,!.,l i 'r 5 I

o -

N a 'li I i w I

2 12.C / > I w _ _ , -

r l -l i i l Z ii

!! I l 10.C l -

~r 5 -

,7 o

z 8C I

l i

i t

l.

'li

'l

)

L) - - -

l l /  !

l

~

i i / t . ,

6.C .

i

/i '

l 6 i ,

/.i  ; .

i

' i I

4.C  : ./ . i

, T' -  ;

• / ,

i -

i . . I

} .

t F

l i l l  ! i i l l

2.C .. . . -. . a . . _u . . .

I 6 .

.. . ( .

0.C z

.r' -

L di .

I

!0,

! l O MD'.. 10 10' 10' Time After Rupture, sec.

D-B 4e. Explain the basis for the initial power level assumption of 77% for 3 pump operation.

RESPONSE

With only one reactor coolant pump idle, the nominal core power trip is set by Technical Specifications to be 75% of rated power.

Two percent is added to the nominal power to include the effect of instrument uncertainty, thus the initial power level is 77% for 3 pump operation.

i 4

i 4-10 l

L:

i D-B 4f. It is stated that the containment building pressure calculated by CONTEMPT is similar to the worst case in BAW-10105. Why didn't the lower initial core flow and power level for 3-pump operation result in a lower containment pressure?

RESPONSE

The maximum containment pressure for the 8.55 ft DE break at the PD reported in BAW-10105 is 22.16 psig.

The maximum containment pressure for the 3-pump operation is 21.68 psig. Reducing the initial core flow and power level to the 3-pump operating levels, reduced the maximum containment building pressure by 0.48 psi. The partial pump containment pressure transient was used in the REFLOOD analysis.

t 4

I 4-11

D-B 4g. Provide the core-wide metal-water reaction for 3-pump operation.

RESPONSE

The core wide metal-water reaction rate presented in Section 9 of

, BAW-10105 was calculated utilizing pertinent data from the LOCA limite analysis and covers all potential power distribution. The LOCA limits analysis was performed by using the worst break, which was shown to be 8.55 ft 2DE rupture at the pump discharge with a 2

g=1.0. As shown in Fig. 9-3, the 8.55 f t DE break maximized time at temperature. Comparing this cladding temperature transient to the 3-pump operation cladding temperature transient clearly shows 2

that the 8.55 f t DE break at the pump discharge with a C =1.0 D

has a higher time at temperature. Hence, the highest core-wide metal-2 water reaction rate would be calculated for the 8.55 ft double-ended rupture. Because of B&W's analytical technique, the core-wide metal-water reaction rate need be calculated only once.

i l

l 4-12 l

1

I D-B 4h. Submit the values of initial pin pressure and oxide layers assumed ,

and justify the selection of these values.

RESPONSE

The initial pin pressure and oxide layers used for the partial ~ loop

analysis are the beginning of life (BOL) values of 1280 psia and 0.000015 inches, (inside and outside surfaces of cladding), respectively.

f As shown in Section 5.5 of BAW-10105, the BOL values maximize the cladding temperature response. Further justification is provided by the time-in-life studies presented in BAR-10102 and BAW-10103.

These studies show that if rupture does not occur during blowdown, which is the case of the partial loop analysis, BOL values maximize the cladding temperature response.

i 4-13

l D-B

- 5. F ; ovide your schedule for submitting the proposed Technical Specifi-cations affected by the LOCA analysis.

RESPONSE

The proposed Davis-Besse Nuclear Power Station Unit 1 Technical Specifications have been submitted to the NRC following the format of the B&W Standardized Technical Specifications. The technical specifications which have been submitted do reflect the LOCA analysis and no changes to the proposed technical specifications

.resulting from the LOCA analysis are expected.

f i

, 5-1 .

l

3 ,

t

, D-B

6. Provide the passive failure analysis committed for January (see your September 5, 1975 letter).

RESPONSE

We have evaluated the potential for passive failures of fluid systems during long term cooling following a LOCA as well as single failures of active components. Single failures of passive components in electrical systems are assumed in designing against a single fail *ure.

Durin6 the long term cooling mode, the ECCS would operate at pressures and temperatures well below their design basis which are outlined in FSAR Tables 6-11 and 6-12; therefore, the propensity for failure is reduced.

Each component in the system has.been evaluated and a potential leakage rate established as documented in FSAR subsection 15.4.6.5.

Components will be tested in accordance with technical specification. The resultant doses-have been evaluated in Table 15.4.6-1 of the FSAR. At the leakage rate determined, it will take greater than 172 days to change containment level one inch.

It is therefore concluded that th- design includes such margin so as to provide high assurance that the potential for passive failures is significantly reduced.

l 1

6-1

, D-B 7a. With regard to the ability of Davis-Besse 1 to cope with poten-tially high boron concentrations in the long term af ter a LOCA, the staff notes that Toledo Edison Company has referenced B&W topical report BAW-10105 (see letter dated July 21, 1975). The following additional information is required:

a. More recent boron dilution design proposals on such dockets as WPPSS and Oconee have the advantage of greater simplicity relative to the multi-mode piping networks described in the topical report. Also, it is the staff's position that Mode 1 (forced circulation through the decay heat drop line) should not be attempted as a method to control boron concentration in the core during long-term cooling. The success of this mode is not ensured because of the possibility of gas or steam entrainment in the decay heat suction nozzle. Such gas or steam entrainment can result in severe damage to the decay heat removal pump. Long-term heat removal requirements can exist for long durations (days or months) after the accident and continuous ope ation of one train of the decay heat re-moval system is required. In the event of equipment malfunction in this train, no method is available to remove the decay heat if the other train has been previously damaged. For the same reason, step 7 on page 10-7 should also not be attempted.

It is preferred that a simple design exist for baron dilution whereby operator involvement with major ECCS components that fulfill the primary role of long-term heat removal is kept simple and to a minimum. Accordingly, discuss alternate means to provide dilution of boron during the long-term after a LOCA.

RESPONSE

The methods described in B&W topical report BAW-10105 to limit concentration of boric acid in the core following a LOCA have been modified to simplify implementation. The revised methods are described below:

1. Bypass lines are currently being installed around the two decay _ heat suction lines' isolation valves located outside the sontainment. These bypass lines will allow the operator to initiate a 40 gpm (minimum) drain flow from the RCS hot leg pipe to the Low Pressure Injection (LPI) system. (The LPI system will continue to recir-culate from the containment sump at design flow rates.)

Each bypass line will have a fixed orifice size to pass 1 40 gpm (minimum) at the post LOCA conditions, a normally open manual isolation valve and a flow meter. The operator would open both decay heat letdown line isolation valves located inside the containment to put this flow path into operation. The controls for these valves are located in the control room.

2. As described in BAW-10105, the operator can open the con-tainment isolation valves in the pressurizer spray line from the LPI system to the RCS hot leg through the pressurizer. The controls for these isolation valves are i- also located in the control room.

7-1

a

!~ ,

D-B Following a LOCA, the boric acid concentration in the core will be

~

controlled to acceptable levels if either of the L~.e.) methods, described above, are put into operation within the first seven days. It is expected that the operator would act to implement both i schemes within a few hcurs following a LOCA.

Postulated failures and the means to ensure boron dilution flow are discussed in response to Question 7c.

4 s

4 1

4 i

a 3

7-2 1

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

<~

t D-B 7b.. Temperature indicators are not satisfactory instrumentation to verify that a minimum flow rate of 40 gpm is maintained. The staff requires flow rate indicators which will clearly show the operator that this minimum flow rate is achieved and maintained over the long term.

1

RESPONSE

A bypass line is provided around valves DH-1517 and DH-1513 with flow indicators that will provide the operator with indication in the control room of flow from the hot leg into the suction of the i decay heat removal pumps. The lines are sized to provide 45 gpm to each decay heat removal train.

4 9

7-3

4 i ,

D-B 7c. Discuss common power supply problems and the procedure to restore a loss of power to essential valves. Also, address possible access problems due to high doses should such a power loss occur after the shift to the recirculation mode.

RESPONSE

There are four valves that the operator would open in order to provide the flow paths for the boron dilution modes of operation, as listed below:

Line Valve Channel MCC Valve Location MCC Location Auxiliary pressurizer HV 2735 1 E11B Inside containment E1. 585' spray outside mechanical penetration room No. 303 Auxiliary pressurizer HV 2736 2 Fila Outside containment El. 603' in spray electrical penetration room No. 427 Decay heat suction HV DH 11 2 Fila Inside containment El. 603' in electrical penetration room No. 427 Cecay heat suction HV DH 12 1 E11B Inside containment El. 585' out-side mechanical penetration room No. 303 The power supply failures postulated were: control cable or power cable failure to a single valve, a diesel or 4 kV bus failure; a failure of one of the above mentioned McC's or unit substation.

Should a control cable or power cable to one of the above valves fail, the alternate path would not be affected. Boron dilution flow is, therefore, assured with no further action required.

Should one of the two redundant emergency diesel generators fail or should 4 kV bus C1 or D1 fail, it would be necessary to provide power to at least one of the affected two valves, one or each alternate flow path, from the unaffected power supply. This would be accomplished by running a temporary cable from unit substation El to unit substation F1 using spare breakers. This will provide power from the healthy to the affected unit substation. All work would be done in the low voltage switchgear rooms on elevation 603 feet.

Should a failure of MCC E11B or MCC Fila occur or should a failure of unit substation El or F1 occur, it would be necessary to provide power to at least one of the affected two valves from the unaffected source. A spare breaker on healthy unit substation El or F1, a spare magnetic or manual reversing motor start from stock, and temporary cables would be required.

7-4

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

.. ,. ~

D-B The temporary cables would be used to connect the load center breaker to the motor starter and from the motor starter to the affected MCC. The power leads of the cable feeding the valve motor would be disconnected at the MCC end and connected to the new

temporary cable. This operation is anticipated to take no more than one hour. The valve will be opened with this new temporary power. The opening will be confirmed by valve position indication.

Using conservative assumptions based on the TID-14844 LOCA source term parameters,-the dose rate has been calculated to be in the a range of 5-10 R/hr during the 7-day period following the LOCA at

! both MCC E11B and 2.5 to 5 R/hr at MCC Fila, principally due to the '

radiation source from the containment spray system at MCC E11B and 250 mr/hr to 500 mr/hr at MCC Fila during the recirculation mode of operation from tha containment emergency sump. A more realistic estimate, based on gap activity, would be a factor of approximately 10 lower, i.e., 500 mr/hr to 1 R/hr at MCC E11B and 250 mr/hr to 500 mr/hr at MCC Fila. This dose rate should not preclude the work as outlined above.

It is probable that, even given the situation of the power failures discussed above, at least one flow path can be established within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. An early assessment of the nature of the failure would be expected since the operators would attempt to accomplish the boron dilution modes in a short period of time after the LOCA, well within a 24-hour period. A flow path can definately be established within the 7 days required as per the B&W analysis.

7-5 l

6. .

n, a- D-B

[ 7d. Discuss the. capability to test the dilution systems.

RESPONSE

The auxiliary spray line will be tested by its use for pressure control during every shutdown when the reactor coolant pumps have been stopped. The bypass lines around valves DH-1517 and DH-1518 will be tested to assure proper flow during shutdown cooling operations while refueling.

1 i

i

+

7-6

D-B 7e. Discuss the feasibility of gravity draining from the hot leg to the sump.

RESPONSE

Providing a gravity drain flow path from the hot leg to the emergency sump is not feasible at this stage in plant construction since the lead time for motor-operated valves is in excess of 62 weeks. An alternate flow path from the hot leg has been provided to drain a minimum of 45 gpm to the suction of each decay heat removal pump from the hot leg. A redundant path has been provided from the emergency sump through the auxiliary pressurizer spray line.

7-7

} ,. - -

D-B 7f. Indicate the feasibility of monitoring boron concentration levels during the long term.

RESPONSE

A common seismic Class I sample line is provided which is connected to the discharge line of each decay heat removal pump. The isolation valves for the sample lines are located in an accessible area and can be used to draw a sample from either pump and analyzed for boron concentration.

.7-8 l

.,. n -

l D-B 4

8a.

~

With regard to the REFLOOD code resistance values in Table 4-2 used for loop venting calculations, insufficient information exists to j support the values selected.

Identify eaci parameter which has been derived from actual measure-ments made ou plant systems, components, models and/or prototypes.

Provide calculations to ahow how these measured parameters were converted to the K-factar presented in Table 4-2.

RESPONSE

i There are four K-factors in the Toledo REFLOOD model which have been derived from either actual measurements made on plant systems, components, models and/:r prototypes. .They are the K-factor for the primary pumps, the vent valves, the path from the reactor vessel inlet to the lower plenum, and the path fram the reactor vessel upper plenum to the outlet nozzle.

The primary pump flow coefficient (K-factor) used in the REFLOOD model is the forward flow locked rotor coefficient which is equal to'10.74. This coefficient was calculated from vendor supplied zone maps based on the area of the cold leg piping. The quadrant curve is given in terms of % head and % flow. At the intersection of 100% flow and zero speed, the percent of rated head is 1.04% for the locked rotor.

In REFLOOD, the pressure drop across a pump is determined from the expression P= K W!W!

1. , psi 288 FP ^FP g

i Defining the term K as 1

K = Head in feet (Flow in gpm) 1 relate K to K as 81 K = 2.317 x 10 K which is based on the cold leg suction area and is independent of fluid density. Using this relationship and the parameters obtained from the quadrant curve (rated head and flow are 358 ft of H O2 and 90670 gpm), we'get 6

K ='(1.09) (358) (2.3717 x 10 ) = 10.75 (9.067 x 10 )

8-1

D-B For the REFLOOD model, the pump resistance is combined with the piping resistance for the paths connecting nodes 7 and 8. Since the area for this path in REFLOOD is an average area over the length of the path, the pump resistance must be modified to reflect the average area shown in Table 4-2 of FAW-101057 Maintaining the P across the pump, which is proportioned to K/A , we obtain a pump A

resistance of 23.68. This represents 91% of the total resistance of the path from nodes 7 to 8. The remainder of the resistance is due to friction in the piping. The K-factor in the other REFLOOD pump paths, node 7 to 9 and node 7P to 9P, are calculated in a similar manner.

The flow-through the vent valve is actuated when the pressure difference across the valve exceeds 0.25 psi. The flowrate is determined by the orifice equation, 2

W= 288pg A 3

P K

The method and documentation for the vent valve K-factor is ex-

} tensively discussed in Appendix B of BAW-10034 (Rev. 3) and Section 3.5 of BAW-10091-Supp. 1. A value of 4.2 is used as the vent valve K-factor in the REFLOOD code as specified in Section 4.3.6.4 of BAW-10104.

The meth'od commonly used to calculate the reactor vessel pressure drop requires calculation of the pressute drop for individual vessel components. A considerable portion of the pressure losses are attributed to form losses. The actual form losses are determined from one of the three methods described below, depending on the geometry modeled.

1. Abrupt expansions and contractions which are not L/D dependent: For these losses the k values are obtained from Kays and Londong, Compact Heat Exchangers (The National Press, 1955).

2. Contractrions and expansions which are L/D-dependent, well rounded, and gradual contractions and expansions:

For these losses the k values are obtained from Idel' chik, " Handbook of Hydraulic Resistance-Coefficients of Local Resistance and of Friction", AEC-TR-6630 (1966) 8-2

. , m -

D-B

3. Turn losses within a fluid body where no total physical bondary exists: For these losses the k values are deternined by the velocity vector method outlined in the Yankee Atomic Electric Company Research Report YAEC-74 by Berringer and Bishop, " Pressure Drop, Flow Dfstribution and Mixing Studies for a Model Hetrogeneous Reactor Vessel" (1959).

In the REFLOOD model these form losses are combined with the frictional losses and a single k-factor is obtained.

Hydraulic testing of the 1/6 scale Duke Reactor Vessel as documented in BAW-10012 has yielded pressure drop and flow data. This data has been used to obtaine a check on calcu-lational methods for reactor vessel pressure drops. In the hydraulic tests of the Duke Vessel Model, pressure drops were obtained for the vessel inlet to core inlet paths and core outlet to vessel outlet paths. The calculated pressure drops are within 2% of the experimental obtained values.

Data irca operating B&W plants (Oconee, Three Mile Island, Russelville, and Rancho Seco) provide additional verification for the methods used to calculate reactor vessel k-factors.

The Oconee CRAFT model (BAW-10103) calculates a reactor vessel nozzle-to-nozzle pressure drop of 60 psi. The measured data, 6

adjusted to 137.9x10 lb/hr flow rate, indicates a vessel nozzle-to-nozzle pressure drop of 59.7 psi. Since the vessel calculated pressure losses are within 0.5% of measured values, the methods employed for calculating reactor vessel pressure drops are justified. The applicability of these methods are further assured as the Davis-Besse reactor vessel is almost identical to the Oconee reactor vessel.

The following REFLOOD flow path resistances are based on the above explained pressure drop calculational methods.

8-3 i

l l

m -

D-B Node Area, ft Resistance, k From H 1 5 7.574 2.003 1 SP 7.574 2.003 8 4 6.257 2.8897 9P 4 12.514 2.8897 3 2 45.89 6.233 l

l l

l 8-4

,. -s ^

D-B 8b. For each flow path shown in Table 4-2, justify the appropriateness of the flow resistance for each Category 2 plant. For example, it is not clear that the most conservative areas were selected to^

serve as a generic calculation applicable to all current Category 2 plants. It is our position that confirmation of these values are required for each as-built plant at the operating license stage.

RESPONSE

The REFLOOD code models the reactor coolant system with 13 control volumes representing the primary system and one control volume representing the secondary side of each steam generator (Figure 4-3 of BAW-10105). The normal flow path resistances and areas, shown in Table 4-2 BAW-10105, are based on the CRAFT steady-state flow conditions in the reactor coolant system. To determine the flow path resistances for the CRAFT code, a SAVER code model of the reactor coolant system is developed. The SAVER code calculates steady state pressure drops for piping networks using established methods. This code is described in BAW-10072A.

Because the REFLOOD model contains fewer nodes than CRAFT, methods have been established to reduce the output from CRAFT for input to REFLOOD. Table 28-7b shows which CRAFT flow paths were combined for each BW 30D path, i.e., CRAFT flow paths 40 and 42 are combined to REFLOOD flow path, (node 5 to 6).

The last column of Table 8, the ratio of K/A , is a direct measure of unrecoverable pressure drops, as the values of k or stea alone have little meaning. An example is presented below, sh, sing the calculational procedure used to obtain the REFLOOD ave? age areas and resistances. For REFLOOD path 5 to 6 we have the following:

CRAFT Area, A path length / area R=K Flow-Paths K g 2 L/A, ft ~1 A 2

40 (from loss).8246 12.40 1.649 .009436 (friction loss) .62633 42 (friction loss) 7.2899 31.94 .6542 .0071458 8-5

,- m -

D-B To calculate the average area over the length of the flow path, the equivalent REFLOOD path length / area, (L/A) , is needed first.

For a series path (L/A)S6 is just the sum of (kfA)'s.

Therefore:

~

(L/A)S6 " ( A)40 + (L/A)42 = 2.3032 ft The average REFLOOD flow area is equivalent to; A

56 " - (L A)g (Ag) / (L/A)S6 thus A

56 " ( A)40 (A)40 + (L ^}42 (A)42 / (L/A)56 2

A = 17.95 ft 56

' For a series path, the equivalent REFLOOD resistance K is as follows:

56 R R and K =R 56 " - g 56 56 *

• 56 thus R 56 " " 40 + R 42 = .016582 therefore K = (.016582) (17.95)2 ,3,343 56

The same calculational procedure is used for all the other paths.

8-6 l-1 l

l l

l

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

D-B J. .To date three nuclear steam plants fall within category 2 of the B&W product.line:

B&W-

. Contract No. Customer and site Docket No.

I i 620-0014- Toledo Edison Co. 50-346 NSS 14 Davis-Besse 1 620-0025 Toledo Edison Co. 50-500 NSS-25 Davis-Besse 2 620-0026 Toledo Edison Co. 50-501 NSS Davis-Besse 3 1

These B&W plants are identical in design except for the type of fuel. assembly. Davis-Besse 1 utilizes the Mark B (15x15) design whereas Davis-Besse 2 and 3 will proceed to the never Mark C (17x17) fuel assembly design. Thermal calculations documented in BAW-10105 have been performed based on the fuel characterists s of the Davis-Besse 1 (Mark B) fuel; however, the hydraulic response of the reactor coolant system should be applicable to all plants. The only deviation is a small difference in pressure drop across the core between the Mark B and Mark C fuel assembly.

i l

1 4

.- 8-7

, .~,A. - - , -- , - . , . . . ,

4 D-B Table 8 Flow Path Resistance for REFLOOD Reflood Node CRAF' Flow Paths (a) REFLOOD 1

(Ref: BAW-10105 Area, Resistance (K/A )

rig. 4-1,2) Ft K Ft i

1 5 63, 38, 34, 30, 24 7.574 2.003 .0349 78, 76 5 6 40, 42 17.95 5.343 .0166 6 7 44 25.52 9.0375 .0139 7~ 8 46, 48, 30 6.3495 26.01(b) .6452 7 9 46, 51, 52, 74 6.359 26.05( ) .6442 8 4 36 6.257 2.8897 .0738 I:

1- SP Same as (1) to (5) 7.574 2.003 .0349 SP 6P Same as (5) to (6) 17.95 5.343 .0166 6P 7P Same as (6) to (7) 25.52 9.0375 .0139 7P 9P Areas twice flow 12.699 26.01(b) .1613

path 46, 48, 50 9P~ 4 Areas Twice flow 12.514 2.8897 .0185 l . path (8) to (4)

3. 2 32, 64, 1 45.89 6.233 .00296 2 1 Core Flow Paths 67.301 12.068 .00266 1 '4 4 Vent Valves, 4.276 4.2(c) .2297 CFT 3 CFT surge line .7213 6.423 12.345 (a): Average areas over length of path (b) includes locked ro. tor pump resistance (c) vent valves.

i l

1 4

8-8

D-B Sc. To allow a greater understanding of the effect of these resistance on reflood rate, re-submit Table 4-2 with the flow paths listed in decreasing order of importance to peak cladding temperature cal-culations. Provide the specific sensitivity study (peak cladding temperature versus K-factor) for the first, middle, and last value.

RESPONSE

Table 8-c shows the REFLOOD flow paths listed in decreasing order of importance to peak cladding temperature calculations. To avoid duplication of flow path types only one loop is presented in the table.

Three sensitivity studies have been performed to demonstrate the sensitivity of flooding rate and thus peak cladding temperature to a change in flow path k-factor.

The assumptions regarding the status of the reactor coolant pumps during the reflood transient was examic-d in Section 5.3 of BAW-10105.

CajeulationsutilizingtheREFLOODcodewereperformedforan8.55 ft DE break at the pump discharge assuming locked and free-spinning rotors for the primary system coolant pumps. The flow coefficient (k-factors) for the forward flow, locked and free-spinning, are 10.74 and 2.89, respectively. For the locked rotor, the value of the total k-factor is as shown in Table 8 for flow path 7-8. For the free spinning rotor, the total k-factor for path 7 to 8 is 8.703 and K/A2 is .2159. Figure 5-6aof BAW-10105 showe the reflooding rates for the two cases examined. The additional resistance of the locked rotors reduces the loop venting capability of the RC system and results in lower flooding rates. At 50 seconds, the flooding rates for the locked and free spinning rotors are 1.44 and 1.78 inch /second, respectively. This reduction in flooding rate, approximately 28%, with the pump rotors assumed to be locked will obviously increase cladding temperatures. The forward flow, locked rotor k-factor was used in BAW-10105 to ensure a conservative calculation.

8-9

~

D-B The response to Question 74 of BAW-10091, Supp. I demonstrates the effect on peak cladding temperature to a 28% increase in the vent valve k-factor from 3.9 to 5.0. The peak cladding temperature was shown to increase only 77F.

That analysis was performed with the REFLOOD no-loop model. Since the Toledo plant will vent steam through the loops and vent valves, the effect of a change in the vent valve resistance should be less severe than that shown avove.

The results of the third sensitivity study are given in Table 8c-2 and shows the effect of increasing the REFLOOD k-factor by 10% for the path connecting nodes 5 & 6. This k-factor represents a middle value from Table 8c-1. When the increased k-factor is incorporated into REFLOOD, a reduction of approximately 0.005 in/sec in flooding rates is observed. This slight reduction in flooding rate would have no significant effect on the peak cladding temperature. For all K/A values of approximately the same magnitude or less than that for the path connecting nodes 5 and 6, reasonable variations in the k-factor would have no significant effect on cladding temperatures.

8-10

^

. - s

- e .

D-B Table 8c-1(a)

REFLOOD Flow Path Resistance Listed in Decreasing Order of Importance to Peak Cladding Temperature Calculations Reflood Node REFLOOD Area Resistance (K/A )

From H f 2

g gt

-1 7 8 6.3495 26.01 .6452 1 4 4.276 4.2 .2297 8 4 6.257 2.8897 .0738 1 5 7.574 2.003 .0349 5 6 17.95 5.343 .0166 6 7 25.52 9.0375 .0139 3

3 2 45.89 6.233 .00296 2 1- 67.301 12.068 .00266 (a) REFLOOD flow paths in the uploop have not been repeated as shown in Table 4-2 of BAW-10105.

8-11

,. . s D-B Table 8c-2 Comparison of Flooding Rates Due to a 10% increase in the k-factor for Reflood Path (Nede 5 to 6)

REFLOOD Run TR42 4 OK 0VL From BAW-10105 K-factor for REFLOOD 5.343 5.8773 Path (Node 5 to 6)

Time After Blowdown sec. Average Flood Rate, in/sec 0 to 34.02 0.0 0.0 34.02 to 44.02 2.512 2.507 44.02 to 54.02 1.847 1.842 54.02 to 64.02 1.582 1.578 64.02 to 84.02 1.447 1.443 84.02 to 124.82 1.345 1.342 8-12 1

1 l

l l

l l

D-B

9. It is noted that no additional flow resistance was added to the cold legs due to the HPI pumps injecting ECC water during reflood.

Evaluate the effect of an additional 0.25 psi cold leg AP upon the reflood rate and cladding temperature. For the LOCA limit analysis, compare the existing time at which the reflood rate goes below 1 in/see to the new time calculated using the additional cold leg resistance.

RESPONSE

To evaluate the impact of an additional 0.25 psi cold leg AP ypon the reflood transient, the 10 - foot LOCA limit case (8.55 ft' DE Break at Pump Discharge, C = 1.0, 17 Kw/ft at the 10 foot elevation) wasexaminedunderthefolSowingconditions:

1. One HPI pump available with injection points in the broken loop.

2. Two HPI pumps available with injection points in both the broken and unbroken loop.

In both of the above cases, one-half of one HPI pumps injection was assumed to exit the break and be unavailable for core cooling purposes. The assumptions regarding the other ECCS systems remained the same as reported in BAW-10105; that is, one LPI pump and two core flooding tanks were assumed active during the transient.

The sensitivity of the availability of the high pressure injection system was examined because of the conflicting impact of additional ECCS injection versus the additional cold leg AP. In general, the CF system provides the rapid refill of the reactor vessel lower head after tha end of the blowdown transient. Once reflood commences the ECCS fills the downcomer annulus up to the elevation of the cold leg nozzle in a short period of time. Generally, the downcomer level is maintained at this level, even after the CFT's are empty, for the remainder of the transient by the pumped injection systems.

During the time period that the downcomer is filling, however, the pumped injection systems become available and aid in the refill process. Thus, the additional HPI injection (no credit for HPI was taken in RAW-10105) can have a positive effect on this part of the reflood transient. Once the downcomer fills, the excess ECCS fluids is assumed to spill out the break. At the same time, the additional 0.25 psi AP has the potential to have a negative effect on the entire reflood transient because of the increased resistance to steam venting. By examining both HPI - system availabilit

• conditions, the worst case can be determined.

Table 9-1 show a comparison of time average flooding rates between the reflood transients utilizing the normal assumption as used in BAW-10105 and the case where both HPI pumps are available with the additional AP in both loops. The case where only one HPI pump was assumed available showed less of an impact on the reflood transient 9-1

L .

D-B and does not represent the worst case. As shown in Table 9-1, essentially no difference is observed in the first 11 seconds into reflood due to the increased injection flow. During subsequent time intervals, however, a slight reduction in flooding rate was observed in the calculation. Over the 241 second reflood period shown, the average flood rate decreased by approximately 0.6% which is not expected to increase the cladding temperature more than a few degrees.

Prior to hot spot quench for the 10 foot LOCA limit case the flood-ing rates drop below 1 in/sec due to the partial equilization of the downcomer and core liquid levels. For the same cases shown in Tables 9-1, this occurs at 264.8 seconds (BAW-10105 asaumptions) and 263.6 seconds after the end of blowdown with the additional 0.25 psi cold leg AP. The subsequent steam cooling period was also extended from 12.8 second to 14.8 seconds when the 0.25 psi cold leg AP was used. This 2 second increase in the steam cooling period would only slightly impact the cladding temperature just prior to the hot spot quench, and would have little or no impact on the overall transient because peak cladding temperatures occur earlier in the transient. Further discussion of the impact of the 0.25 cold leg AP and also steam cooling is presented in B&W responses to NRC Question List 28.

9-2

~

l I

I Table 9.1 Time Average Flooding Rates i

Flooding Rates ,

l . Time Intervals (in/s)

After the End With 0.25 psi of Blowdown Normal Assumptions AP & 2 HPI (s) l 0 0 i 0-9

! 9-20 2.5624 2.5622 1.7430 1.7185 20-50 1.4330 1.4281 50-100 1.1820 1.1781 100-250 1

)

s 1

i .

. ,-t-- -- v-- ---ww-,w, e, v--- , , . , , - + - - , ----+..pme,.,,n. . . - - - - - > ~ , . . ,

• 9- , ---aw- er *c>t--e 1-

• l .' ~ ~

o J

D-B

10. Justify that the assumed CFT line resistance is appropriate for Davis-Besse 1. Provide the L/D's for the CFT line for Davis-Besse 1 and include entrance and exit losses.

d

RESPONSE

The core flood line resistance (k-factor) utilized in the LOCA analysis reported in BAW-10105 was 6.423. This value was based on a conservative estimate of the core flod line L/D, by the archi-tectural engineer for DB-1 prior to the actual analysis, and appropriate entrance and exit losses. Specifically, this k-factor was composed of the following resistances:

1

1. K = 4.673, based on an average line L/D (14-inch, schedule 140 pipe) of 354 and a friction factor (f) of 0.0132.

2. K = 0.15, for a rounded entrance loss based on Idle' check.

3. K = 1.0, for an exit loss into the reactor vs .i.

4. K = 0.6, for the flow restrictor in the CFT nozzle.

By summing the above, the core flood line k-factor of 6.423 is obtained.

Subsequent to the submittal of BAW-10105, a detailed examination of the core flood lines has been completed by the architectural engineer.

The resulting L/D values for each CFT line are as follows:

DB-1 L/D Values

• South Tauk North Tank Pipe 72 67.25 Elbows 130 140.

'"I" Fitting 20 20.

Valves-Gate 13 13.

Checks 65.6 65.6 Total. 300.6 305.9

• Based on'14-inch, schedule 140 pipe.

10-1

$ O' ^ ,

l D-B Utilizing this new data, an average L/D for the CFT line is 303.25.

Assuming a friction factor of 0.0132 and the entrance and exit losses specified in items 2-4 above, the actual CFT line k-factor for Davis-Besse 1 is 5.753. Thus, the CFT line resistance utilized in BAW-10105 is conservative, as demonstrated by the result of CFT line resistance study in BAW-10091, for Davis-Besse 1.

10-2 i