ML20150E909

Forwards Results of an Analysis Performed to Investigate the Effects of LPCI Diversion on ECCS Performance.Rept Was Incl as Part of Amend 13 Originally Submitted in .Encl Analysis Will Be Submitted in Amend 14

ML20150E909
Person / Time
Site:	Black Fox
Issue date:	12/18/1978
From:	Ewing T PUBLIC SERVICE CO. OF OKLAHOMA
To:	Varga S Office of Nuclear Reactor Regulation
References
NUDOCS 7812220115
Download: ML20150E909 (17)

Text

6212 DIN 8-013-900 4 PUBLIC SERVICE COMPANY OF OKLAHOMA A CENTRAL AND SOUTH WEST COMPANY P O BOX 201/ TULSA. OKLAHOMA 74102 / (918) 583-3611 '

Public Service Company.of Oklahoma December 18, 1978 Black Fox Station File: 6212.125.3500.21L LPCI Diversion Effects On ECCS Performance Docket STN 50-556 and STN 50-557 Office of Nuclear Reactor Regulation Division of Project Management Light Water Reactors Branch No. 4 U. S. Nuclear Regulatory Commission Washington, D. C. 20555 Attn: Steven A. Varga, Chief Centlemen:

Public Service Company of Oklahoma is resubmitting the results of an analysis performed to investigate the effects of LPCI diversion on ECCS performance for Black Fox Station Units 1 and 2.

This report was included as a part of Amendment 13 after being initially submitted by letter. November 7, 1978 (our 6212 DIN 8-013-292). Two telephone conversations, November 16 and. November 17, between the NRC Staff and PSO relative to information contained in the report and information requested in the NRC letter dated October 11, 1978, revealed that additional expansion and clarification was needed. The report is therefore being provided in a format which more closely follows the sequence of questions in the NRC letter and also includes information to clarify the original responses.

The attached analyais will be submitted'in Amendment 14 to the BFS PSAR, to be filed prior to Construction Permit issuance.

Very traly yours, T. N. Ewing Manager, BF, clear Project TNE:so

' Attachment i xc: See Attached Service List 781mous Y 8 og SlE

} j CENTRAL AND SOUTH WEST SYSTEM

' o k'^" *" B"t'=ttrn Elo,cpic Power

/ O

'gp i E*S'ch 'a"A'*ht b&cM!c!?

c *P "Y gst Tegs Utihtios

glACKFOXSTATIONSERVICE-LIST XC: L. Dow Davis, Esquirs . Joseph R.' Farris,- Esquire William D. Paton, Esquire John R. Woodard, III. Esquire

' Colleen Woodhear, E quire Green, Feldman, HE Woodard Counsel for NRC Staff - 816 Enterprise "

U. S. Nuclear _ Regulatory Commission Tulsa Oklahoma Washington, D. C. 20555

~

Andrew T. Dalton

- Mr. ' Cecil Thomas . 1437 South Main 3F U. S. Nuclear Regulatory Commission' Tulsa, Oklahoma Phillips Building 7920 Norfolk' Avenue - Mrs. Ilene H. Young Bethesda, Maryland 20014 3900 Cashion Place Oklahoma City, Oklah 3112 Mr. Jan A. Norris ,

Environmental Projects Branch 3 Mr. Lawrence Burrel .

U.S. Nucler P.egulatory Commission Route 1, Box 197 Phillips Building Fairview, Oklahoma 73737 7920 Norfolk- Avenue Bethesda, Maryland 20014 Mrs. Carrie Dickerson Citizens Action for Safe Energy, Inc.

Mr. William G. Hubacek P. O. Box 924 U.S. Nuclear Regulatory Conrnission Claremore, Oklahoma 74107 Office of Inspection and Enforcement Region IV 611 Ryan Plaza Drive, Suite 1000 Arlington, Texas 76012 Mr. Gerald F. Diddle General Manager

' Associated Electric Cooperative, Inc.

P. 0. Box 754-Springfield, Missouri 65B01 Mr. Maynard Human General Manager Western Fanners Electric Cooperative P. 0. Box 429 Anadarko, Oklahoma 73005 Michael I. Miller, Esq. 1 Isham, Lincoln & Beale I One 1st National Plaza l Suite 4200 1 Chicago, Illinois 60603 Mr. Joseph Gallo Isham, Lincoln & Beale 1050 17th Street N.W.

Washington, D. C. 20036 ,

l i

w.*e.e.. p w ,--gaaqb w a w ,

d ANALYSIS OF LPCI DIVERSION EFFECTS ON ECCS PERFORMANCE FOR BLACK FOX STATION UNITS 1 AND 2 DECEMBER 2, 1978 DOCKET STN 50-556 STN 50-557

(

+s A .A.e e- aar .2 Jim ;a y J_ 4 _ a ,s _

• • • 1 # 4 . B*

,r PURPOSE This analysis' was' performed to investigate the effect on the ECCS analysis for the Black Fox Nuclear Power Station of diverting low pressure coolant injection (LPCI) pumps to the containment spray mode ten minutes after a loss-of-coolant accident (LOCA) initiation. .

Automatic diversion of LPCI flow to containment spray has been provided in response to an NRC requirement to assure containment integrity for postulated high steam flow bypassing the suppression pool, Such flow diversion would occur only if a high containment pressure ( 9 psig)'

signal is present after ten minutes. The assumption of sufficient bypassing to cause such a pressure has been shown by GE to be extremely conservative and unrealisticl. The results of the drywell cracking study referenced in footnote 1 showed that for small breaks, which are of interest in this study, no through wall c.'acking of the drywell would occur where the structure has only minor prior cracking damage. Therefore, no bypass is expected to occur due.to cracking of the drywell wall and the 9 psig containment pressure will.not be reached.

CONCLUSION The results show that the worst single failure / break type combination is the high . pressure core spray (HPCS) line break (approximately .02 ft2 )

assuming the fai wre of the low pressure core spray (LPCS) diesel generator (D/G) which powers one LPCS pump and one LPCI pump. This single failure / break type combination yields the highest peak cladding temperature (approximately 20850F) of all the cases affec:ed by LPCI diversion at ten minutes. The '

peak cladding temperatures experienced by the cases affected by LPCI diversion are below the limits established in 10 CFR 50.46 (22000F).

The maximum cladding oxidation is less that 2%, well below the 17%

limit. The maximum hydrogen generation is less than 0.17%, well below the 1% limit.

I NED0-10977 Drywell Integrity Study: Investigation of Potential Cracking in BWR/6 Mark III Containment ,

S c- - . . - . - . _ . - . - - -

t .- ':

ASSUMPTIONS- ,

1) A maximum of two LPCI pumps (specifically LPCI "A" and LPCI "B")

can be fully diverted .at ten minutes ' to the containment spray mode.

'(NOTE: LPCI "A" shares an emergency diesel generator with the LPCS; LPCI "B" and "C" share an emergency diesel generator. The pump; associated with LPCI "C" cannot be diverted to containment-

. sp rays ) .

2) .The standard FSAR assumption of one automatic depressurization l system (ADS) valve failure combined with the worst additional

- single failure was retained because this assumption is built into ,

the present model. In addition, failure to account for this ADS valve failure would result in limitations on the operation of

~ Black Fox plant which could affect plant availability. This bounding' >

assumption y4 elds conservatively higher ' calculated. peak cladding

~

temperatures (PCTs) by approximately 1000F.

3) Approved Appendix K _ analysis _ models were used,- except that some LPCI flow to the reactor vessel was stopped ten minutes after the accident. -

RESPONSE TO SPECIFIC NRC CONCERNS Only those accident cases which are not reflooded to the hot node before ten minutes are affected by the assumed LPCI diversion. Once the core ,

has been reflooded, only one ECCS pump is necessary to keep the' core covered. Thus, the breaks affected include small breaks less than approximately .02 ft2 (depending on the break location).and outside steam line breaks (0SLB). The effect of 'the assumed LPCI diversion on the OSLB is small and is discussed in a later section of this report.

1 The following break locations were considered: A) core spray line, B) recirculation line, C) feedwater line, D) the steam line, and E) LPCI line. A brief summary of each analysis is provided below.

t L.._

A. Core' Spray Line Break (HPCS Line) - It is conservatively assumed that no flow enters the vessel through the broken line independent of the' break size. For this. case, the failure of the diesel generator. associated with LPCS and LPCI "A" is the worst . single ' failure since all credit for core spray ' cooling is eliminated. The ECC systems remaining before diversions.

are 2 LPCI + ADS and 1 LPCI + ADS after diversion at ten minutes. Because in both cases the reflooding time' is based on only subcooled LPCI flow reflooding the vessel, there is a longer reflooding time associated with the diverted case with

. reduced ECCS flow. The results of this investigation are shown in Figure 1. Because the temperature increase from the

~

non-diverted case is a result of a loss of reflooding flow f rom 1 LPCI pump, intermediate cases (loss of part of the flow) will experience intermediate (lower) temperature increases.

This' particular failure / break type combination was the most adversely affected by the assumed LPCI diversion. .However the peak cladding temperatures are still below the limit of l 22000F.

B. Recirculation Line Break - Because this investigation is primarily concerned with small breaks, the failure of the HPCS, for non-core spray line breaks, is the worst single failure for this study. If the HPCS were operable, the break sizes being analyzed would reflood earlier than ten minutes with the very small break sizes never uncovering. For a recirculation line break, the worst single fail'ure is consequently the HPCS failure. The ECCS remaining before diversion are 3 LPCI + LPCS + ADS and, after diversion,1 LPCI + LPCS + ADS.

Since in the diverted case the remaining LPCI flow is not enough' to significantly quench the voids in the lower plenum, the mixture in the lower plenum will reflood with a higher .

voided mixture. This higher void fraction for the diverted

e i

case more than' offsets the reduction in ECCS flow entering the vessel'.due to this diversion of LPCI. Hence, there is a net

~

reduction in PCT due to a shorter reflooding time and the

' recirculation line break without diversion which has already been reported is bounding relative to a line break with diversion.

A representative break (.01 ft )2was analyzed which confirmed these results. The results of this investigation are shown in Table 1. 'Intennediate cases (diversion of less than the full

' flow from two pumps) should result in smaller temperature

' decreases.

C &' D. Feedwater' and Steam Line Breaks - For these breaks, the worst single failure is the HPCS failure, as' described under Item B.

The ECCS . remaining before diversion are 3 LPCI + LPCS + ADS and after diversion 1 LPCI + LPCS + ADS. For the diverted case, there will be a reduction in calculated PCT- for the same reasons discussed for the recirculation line break. A representa-tive break (i.e. 01 ft ) was 2 again analyzed which confirmed the anticipated results. The res'ults of this investigation are shown in Table 1. For both cases, insignificant decreases in calculated PCT result = from LPCI diversion. The outside (isolated) steam line break was also. considered with similar res ul ts.

E. LPCI Line Break - As in the case of the core spray _ line break, it is conservatively assumed that no flow enters the vessel through the broken line independent of the size. For this break, the worst single failure is the HPCS failure, as described previously. The ECCS remaining before diversion are 2 LPCI +

LPCS + ADS and, after diversion. LPCS + ADS (if the break is in line "C") or LPCS + LPCI + ADS (if the break is in line "A"/or "B"). In either case there is insufficient LPCI flow to significantly quench the voids in the lower plenum. Therefore, the core will reflood with a voided mixture. This higher void fraction can more than offset the reduction in ECCS flow entering the vessel .due to diversion of LPCI. In these cases

= the calculated PCTs are extremely low' and changes in PCT in  ;

either direction would result in calculated PCT's which are still far below the limit. As above, the .01 ft2 break was analyzed and the results of both diverted cases are shown in Table 1.

RESPONSE TO QUESTION ON OPERATOR ACTION The operation of the ECC systems including diversion of LPCI.to contain-.

ment sprays requires no operator action for at least 10 minutes following accident initiation. Ten minutes is the present licensing basis for operator manual action time following automatic actuation of the ECC. ,

system. .There is no requirement either in 10CFR50.46 or GDC 35 for assuming no operator action 20 minutes after the initiation of the accident. Ten minutes continues to be the licensing basis used and supported by General Electric. It is also the basis for the containment performance evaluation as it has been for other BWR plants. ,

RESPONSE TO QUESTION (1)'

The system provided.for diversion of LPCI flow is a safety grade system.

Consequently,'it has a high reliability in performing its intended function. Postulation of a -failure of this system to perform its function in combination with another single failure is not required under GDC 35 or 10 CFR 50.46.

RESPONSE T0 QUESTION (2)

As noted in items B, C, and D, for non-ECCS line breaks there is a decrease in' the PCT in going from the non-diverted.to the diverted case.

As noted in E, the' PCT for the LPCI break is far below that for the limiting break. For LPCS line breaks, with the worst single failure of

the HPCS, there are 3 LPCI pumps available before diversion and one LPCI available after diversion.

o

I

• r 4

For the HPCS line break with failure of-the LPCS/LPCI diesel only 2 LPCI's are available before diversion and 1 LPCI' is available after diversion.

Consequently, the HPCS line break represents the most limiting break

. location when evaluating LOCA with diversion. '

RESPONSE TO OVESTION (3)

The diesel generator failure for the LPCS/LPCI is more limiting than the diesel generator failure for the 2 LPCI's because the LPCS, as opposed ,

to the LPCI, will rapidly reflood the core with a voided mixture. With a voided mixture, the swollen level inside the lower plenum is higher and reflooding can occur sooner. The effect of LPCI flow is to quench voids in the lower plenuin therefore requiring more water to reflood the hot node. Consequently, the reflooding time with just LPCI available is longer than that with LPCS available, and the loss of LPCS relative to LPCI would have a more adverse impact on reflooding capability. An analysis was performed for this limiting break and a failure of the LPCI/LPCI diesel generator and the calculated PCT's were substantially 4 below those for the assumed worst single failure. It should be noted that CCFL effects on LPCS which have been included in this analysis are small for this break, and consequently, the LPCS diesel failure still represents the worst single failure.

RESPONSE TO QUESTION (4)

Figure 1 is a plot of peak cladding temperature vs. break area for an HPCS line break assuming an LPCS diesel generator failure. Both diverted >

and non-diverted cases are shown. The curve for flow diversion was generated using data points at break areas of 005, .01, .016, .02, .03,

.04, .05, and .06 ft 2, Figures 4a through 4e are plots of water level inside the shroud, reactor vessel pressure, convective heat transfer coefficient, peak clad temperature, and LPCI flow versus time.

T

RESPONSE TO QUESTION-(5)

The maximum temperature for the assumed LPCI diversion case for any given break location occurs at approximately that break size where the LPCI system would normally inject flow into the vessel starting at 600 ~ seconds (i.e. the assumed LPCI diversion time). Bigger breaks get some reflooding benefit from the LPCI pumps before diversion. Smaller breaks have the same ECC' systems available as this maximum break, but the smaller break area has a lower calculated PCT. This follows'from the fact that: (1) The core is uncovered for shorter periods for smaller breaks since less mass is lost through the break during the blow %in from the time the reactor water level trip setpoint is reached until the time ADS is actuated 120 seconds later allowing the LPCI to operate, and (2) the decay heat is lower at the time of uncovery for smaller breaks.

~

For breaks smaller than the critical break size'of approximately .02 ft 2, the LPCI system would normally inject flow into the vessel at some time af ter 600 seconds. Therefore, a longer LPCI diversion time would have correspondingly smaller breaks where the maximum PCT would occur.

From' the above argument two points arise: (a) Maximum PCT .from' diversion will occur at a smaller break for longer diversion times, (b) the smaller the break the lower the calculated PCT as discussed previously. Consequently, diversion at times greater than 10 minutes will have less severe consequences than diversion of 10 minutes.

It should also be noted that the dotted line curve of Figure 1 is bounding for small breaks ( approximately 0.02 ft )2since only one LPCI pump (the minimum possible since LPCI pump "C" does not divert) is assumed to operate for these breaks. Consequently, if one assumed a new scenario whereby diversion occurred at times later than ten minutes, the Figure 1 curve would peak at a smaller break area and the PCT at that peak would be the same as that on the dotted line curve for that break area.

^- -' ~~^

.- . ..:4 _ . . - .

1 TABLE 1

~

THE EFFECT ON THE PCT OF DIVERTING LPCI FLOW AT 2

10 MINUTES FOR VARIOUS .01 FT BREAK TYPES BREAK PCT PCT TYPE NO DIVERSION WITH DIVERSION Recirculation Line 948*F 877 F Feedwater Line 917'F 836 F ,

Inside Steam Line 920 F 831*F LPCI Line- 834 F 804 964 F F2)(II) t NOTE: (1) PCT if break occurs in LPCI line "A" or "B" (2) PCT if break occurs in LPCI line "C" i

250o REFLOODING TIME < to MINUTES 20oo -

/ f\ 7 p' / \

. N  %

E Break Area E

- .005 1788

~

.0 6 20 3

.02 2085

< .03 1859 E

.06 1747

{. .05 1760 e

y

'Ig 2000 -

SMEAK SIZE FOR WHICH LPCIINJECTS

! u FLOW tNTO THE VESSEL AT to MINUTES -

W 9ASE CASE (NO DIVERSIONI meme amme ammma 1 LPCI REMAINING AFTER 1 P DIVERSION AT to MINUTfB 500 -

i 7

5 1 I I I I I o.co2 0.00s oot o.02 o os o.1 0.2 03 BREAK AREA (ftI l Figure 1. Peak Cladding Temperature versus Break Area for an HPCS Line Break Assuming an LPCS Diesel-Generator Failure. Systems Remaining: 2 LPCI + ADS

!' . !t !i i M

I

)

g e3 r3 o(

C2 I

M I

E t R f U

S 02 S

E R 0 P

=

H Ga I e MrA a

k ga ne I i r M wB o

l ,

l e or F u l

)

c di o

s ua

( oF E r M hG I SD _

T eS hC tP -

L 0 e I 0 d ,

0 ik sa ne I r B

l ee vn .

ei LL .

ry m

ea .

t r ap _

WS _

0 .

I 0 a 3

4 _

e _

r u _

g i

F F F

A A T B

- - O 0 0 9 4 23y I i?

',l ' <

4 1200 1 -

soo -

w I

w E

n I

I t

40 -

t i

4 2

i 1 I I '

o o goo 900 1200 1900 7IME iseel Figure 4b. Reactor Vessel Pressure 7ollowing a HIGH PRESSURE Core Spray Line Break, LPCS DG Failure; Break Area = 0.02 f t2 (SBM).

t

i .

l

, t r

i E j.

o.

"r k

i 3 -

E ,

I T

M i i

w 8

. .L 5 m u.

' I E

1 1

to.ooo 2s 1 I IE i i o 300 e00 900 1200 1900 TIME isoc)

Figure 4c. Convective Heat Transfer Coefficient Following a HIGH PRESSURE Core Spray Line Break, LPCS DC Failure, Break Area = 0.02 ft 2 (SBM).

4 1 i i !!' ,: 1

= . . _

M B

I e

y _

a r

p S _

I M e) -

1 rM _

oB _

CS _

( _

E _

R2 Ut -

Sf S _

E2 _

R0 -

P I

0 G=

I Ha _

e _

I 0

0 ar -

9 g

A nk _

i a _

we or lB _

l _

) o , _

c Fe _

o s r

(

E eu _

rl _

MI ui _

T t a aF _

r _

eG -

0 pD 1 0 m -

8 eS -

TC _

P -

n i ,

dk d a ae l r CB

- ke

_ an ei PL 1

n d 4

e r

u g

i F

- - L o 0 0 0

m - 0 3

. t m;4E E uw6*

ewY i

• ! 4

t ' .. f

' d I

M 1

k a

e r

B e

n i

L y

a r

p S

e .

r) oM CB S

0 E(

l 0 R 0

1 U2-S t Sf E

R2 P0 H0 G

i I =

c H o

(

s a E ae M r

. I gA T n ik wa oe l r lB o

F ,

e er t u al Ri a

0 wF l 0 5

o l G FD IS CC PP LL e

4 e

r u

g i

F

- - O 0 0 o 0 0 0 0 2 2 32 E3s.b h

,