ML19261F143

Analysis of Small Breaks in Reactor Coolant Pump Discharge Piping for B&W Lowered Loop 177 FA Plants.

ML19261F143
Person / Time
Site:	Three Mile Island
Issue date:	04/27/1978
From:	BABCOCK & WILCOX CO.
To:
Shared Package
ML19261F140	List: ML19261F139 ML19261F143
References
NUDOCS 7910240802
Download: ML19261F143 (15)

Text

.'

ANALYSIS Or SMALL BPIAKS IN THE PIACTOR COOLANT PUMP DISCHARGE PIPING TOR THE B&W LOWEPID LOOP 177 FA PLANTS APRIL 24, 1978 7010240 l 0 2 p 1485 526

1. Increduction On April 14, 1978, B&W reported that previous s=all break analyses had not been based on the worst break location. This report indicated that the worst case break had now been determined to be at the reactor coolant pump

- discharge. A spectrum of small breaks has been examined for the B&W 177-FA lowered loop plants using the small break evaluation model described in BAW-10104, Rev 3, "B&W's ECCS Evaluation Model." These results sh t that it is necessary to use operetor action during the early stages of the pos-tulated accident, to effectively =itigate the accident consequences and

=cet the criteria of 10 CFR 50.46. Operator action is used to achieve suf-ficient and balanced flow through all four HPI injection lines. This re-port shows that operation up to at least 2568 Mit is possible within the criteria of 10 CFR 50.46 and Appendix K.

2. Evaluation 2.1. Method of Analvsis The analysis method used for this evaluati)n is that described in Chapter 5 of BAR-10104, Rev 3, "B&W's ECCS Evaluation Model." Specifically, the model, except for break size, break location, and core power, is the same as utilized in Appendix C of BAW-10103A, Rev 3, "ECCS Analysis of B&W's 177-FA Lowered-Loop NSS." The analysis uses the CRAFT 2 code to develop th'e history of the reactor coolant system hydrodynamics. The CRAFT model uses 19 nodes to simulate the reactor coolant system, two nodes for the secondary system, and one node for the reactor building. A schematic diagram of the model is shown in Figure 1 along with the node descriptions.

Control volumes (nodes) in and around the vessel are all connected by a pair of flow paths to permit counter-current flow. The break is assumed to be located at the bottom of the cold leg piping between the reactor coolant pucp discharge and the reactor vessel. The Wilson, Grenda and Patterson average bubble rise model is used for all nodes. Uithin the core region, however, a multiplier of 2.38 is applied to the calculated bubble rise velocity.

Appendix F of BAW-1010. demonstrates that a multiplier of 2.38 in CRAFT 2 gives a mixture height within +2% of that predicted by F0AM. Thus, no FOAM analysis vill be needed if the CRAFT 2 mixture level remains above cae co:e by 21 of the active lengrn.

The following assumptions are made for :caditicus and s/ sten responses during the accident:

1485 327

a. The reactor is operating at 102% of the steady-state power level of 2568 MWt. For breaks greater than 0.1 f t, the analysis utilized a power level of 102% of 2772 MWt.

b. The leak occurs instantaneously, and a discharge coefficient of 1.0 is used for the entire analysis. Bernoulli's equation was used for the subcooled portion of the transient, while Moody's correlation was used in the two-phase portion.

c. No offsite power is available,

d. The reactor trips on low pressure at 1900 psia.

e. The rafety rods begin entering the core after a 0.5 second delay from the time the reactor trip signal is reached.

f. The RC pumps trip and coast down coincident with reactor trip.

g. One complete train of the emergency safeguards system fails to operate, leaving two CFTs and only one HPI and one LPI system available for pumped injection to mitigate the consequences of a cold leg break.

h. The auxiliary feedwater (FW) system is assu=ed to be available during the transient. Its main function is to remove heat from the upper half of the steam generator during the initial stages of the transient.

When the secondary side of the steam generator becomes a source of heat to the primary system, the assu=ption of auxiliary FW maximizes the energy that cust be relieved.

1. ESFAS signal error band is considered in the analysis to signal the actuation of the HPI system.

j. The peak linear heat generation rate in the hot pin is the maximum allowed by the Technical Specifications at the 10.5 ft 1cvel.

k. Operator action is taken to increase the HPI flows to the intact cold legs at 10 minutes following the ECCS initiation signal. This assump-tion is explained more fully below and in section 3.

As most of the breaks evaluated in this spectrum showed core uncovery, temperature calculations were necessary. Once core uncovery occurs a spatial swell distribution aa . lysis is necessary to assure that n-;y the ccie covered by nixture is included in the swell level. BM.' us e- a FOAM code. The code was utilized under the same cssumptions as descr ' above with the following additicas:

~, -

1485 328

1. The power shape shown in Figure 6 was used but implemented with a radial peaking factor of 1.0. This represents the average channel condition which is appropriate for use in swell level calculations.

2. Steam production due to heat from the primary =etal, core and lower plenum flashing, was conservatively underpredicted. Although the CRAFT model accurately predicted these effects, full credit was not included in the FOAM simulation as a conservative computational con-venience. This simulation, therefore, underpredicts both the swell level and the steaming rate. Consequently, more core uncovery and lower coolant flow are used in the heat-up evaluation.

The heat-up calculation was performed using the THETA code in the manner described in section 5 of BAW-10104. The following additional assumptions are utilized in the THETA evaluation:

1. The pcuer shape of Figure 5 was used with a radial power factor of 1.8.

This maximizes steam superheating and sets the peak local power at 10.5 ft at the technical specification LOCA limit.

2. Coolant flow and mixture level were taken directly from the FOAM calcu-lations.

3. End of life pin pressures were used to conservatively predict the inci-dence of fuel pin rupture.

2.2. High Pressure Injection System Performance The previous arrangement of the HPI system allowed for one pump to inject into the reactor coolant system (RCS) at two locations. As one injection point could be in the region of the break, 50% of the one HPI flow could fail to penetrate the reactor vessel. This flow would, therefore, not be ava'_l-able to provide core cooling. The proposed operator action, section 3, vill provide four points of penetration of the RCS. Therefore, only 25% of the HPI flow would be lost.

Since the flow from one HPI pump will now be distributed to four injection points and to assure conservatism in allowing for injection line loss dif-fe rences , this analysis assumes 30% of the HPI is injected into the broken cold leg. The implemented action starts at ; ninutes af ter an ECCS signal and is concluded 15 minutes af ter the cf..r l. The resultaat HP1 flev can he m u ervatively represented as a linear camp froa 5 to 15 minutes. This

}29 was simulated in our present CRAFT code as a step function at 650 seconds (600 seconds for action, 50 seconds for ECCS signal) . This is illustrated in Figure 7.

2.3. Break Spectrum and Results All evaluations reported in this analysis assume the high pressure injection performances as described in section 2.2. Breaks of 0.3, 0.2, 0.15, 0.1, 0.07, and 0.04 ft2 were evaluated. The evaluation of a 0.5 ft2 break was reported in BAW-10103A, Rev 3, and shows co=plete core covery at all times and thus no temperature excursion. The 0.5 ft break results are independent of HP1 flow and remain valid.

Figure 2 shows the RCS pressure transient for each break. As shown, each ac-cident initiates CFT flow within 2000 seconds except for the 0.04 ft2 break.

Figure 3 shows (CRAFT) mixture height as a function of time for each break of the spectrum. As can be seen, breaks of approximatcly 0.3 f t and larger 2

than approximately 0.04 ft uncover part of the core. Various uncovery levels #q

'O and times are observed but all trends are consistent throughout the spectrum.

The 0.04 ft break achieves a match up of effective ECCS (the HPI injected into the intact cold legs) with the core decay heat and the RCS metal heat at 2500 seconds. After 2500 seconds the mixture level will rise in the core due to excess HPI injection. As the 0.04 ft break has a level of 14 feet at this time the core never uncovers and no temperature excursica occurs. For breaks smaller than 0.04, the match up will occur at approximately the same time and the core mixture levels will drop slower; thus, for all smaller breaks the core will remain covered.

Figure 4 shows the time duration of uncovery for four core elevations as a function of break size. These results are from CRAFT. As can be seen, the maximum degree of uncovery and the maximum time of uncovery occur for the 0.15 ft break and is the worst case break. This break can thus be identi-ficd as the worst case. A similar uncovery occurs for the 0.1 and 0.07 ft breaks. The 0.07, 0.10, and 0.15 ft breaks ere analyzed for te=perature respense. The results are shown in Figure 5 and are well within the criteria of 10 CFR 50.46. They provide positive assurance that all breaks of the spec-trum are within acceptance criteria.

1485 330

n e 0.2 ft break was not evaluated for three reasons:

1. The uncovery tire is approximate 1.y 2/3 of that for the 0.15 ft2 break.

2. The depth of uncovery is only 1/2 of that for the 0.15 ft b reak. The mini =um core level is only 11 feet as compared to 10 feet for the 0.15 ft break.

3. The decay heat rate will be approximately the same as for the 0.15 f t b re ak.

Thus , the 0.2 f t case is well bounded by the 0.15 fe case.

Local metal water reaction is shown below the te=perature curves on Figure 5.

The highest value is 2.8% for the 0.15 ft 2b reak. This value is well belev the local oxidation limit for the large breaks utilized in BAW-10103 for the whole-core cetal-water reaction calculation. Thus, the whole-core metal-water reaction results given in section 8 of BAW-10103 is conservative for scall breaks. The degree of clad damage is bounded by the large break results which produca higher clad temperatures. Thus, all criteria of 10 CFR 50.46 are cet. This analysis is conservative for =any reasons as detailed in the write-up and =cets all evaluation criteria. This analysis shows that all 177 lowered loop plants ceet the criteria of 10 CFR 50.46 if operated at or below 2568 MNt power and in conjunction with the specified operator action.

3. Ocerator Action Tne ECCS analysis used as a basis for this report assumes that the operative HPI train (one train is lost due to a single active failure) provides emer-gency core cooling water to the RC loop containing the break. It is conser-vatively assumed that the break is on the lower portion of the reactor cool-anc pump discharge piping resulting in the total loss to the system of 50%

of the available HPI flow. Acceptable nitigation of the accident requires core than the 50 % of this flow from one HPI pump. If, following the LOCA, it is assumed that one train of HPI does not start it is necessary to take operator action to achieve a flow split wherein no more than 30% of the re-raining pump's flow goes into the cold leg containing the break. The follow-ing is a description of the action required for a typical plant.

1. Upon ESFAS signal check for flow through both HPI trains.

1485 331

2. If no flow in one train:

-- cpen pump header cross-connect valves

- check HPI valve position and open if closed

3. Secure flow through nor=al makeup line if flow is indicated

4. Thsottle HPI valver as required to balance flow ar.d meet run out limits The above actions initiated at five minutes and completed within 15 minutes subsequent to the ESFAS actuation ensures adequate HPI flow for accident miti-gation. In the analysis, credit is taken for the HPI flow as the HPI injec-tion valves are opened. Figure 7 shows the calculated HPI flow for a typical plant as a function of time for a 10 minute valve opening. As shown in Figure 7, the majority of the HPI pump capacity would be delivered with a partial valve opening. For the small break analysis, a linear flow versus valve posi-tion response was simulated by a step function increase,10 minutes af ter ESEAS actuation.

4. Evaluation of Other B&W Supplied Plants

a. Davis-Besse - The D3-1, 2 and 3 Plants have been analyzed for a spectrum of small breaks a the RCP discharge in accordance with an approved small break evaluation model. This analysis is reported in BAW-10075A, Rev 1, lbrch 1976. In addition, the Davis-Besse 1, 2, and 3 units have a split high press'are injection and makeup system design. The Davis-Besse HPI pumps, therefore, have considerably higher capacity at the system pres-sures experienced.

b. 205 and 145 FA - These plants have been analyzed for a spectrum of small breaks at the RCP discharge in accordance with an approved small break evaluation model. These analyses are reported in 3AW-10074A, Rev 1, and B A', -10062 A, Rev 1, March 19 76. In addition, the 205 and 145 FA HPI sys-tens contain cross connects between the two HPI trains downstream of the H?! inj ection valves. These cross cenaccts effectively achieve the same flow split as the operator action assumed in the current 177 FA lowered loop analysis and the flow split is achieved when the HPI pump is started.

c. All 36'J supplied plants except the 177 FA lowered loop plants have raised loops and thus do not trap a large volume of coolant in the cold leg.

3e raised loop design allows this coolant to drain into the core for core covering and cooling by boiloff.

9

~j-1485 333

Figure -1. CRAFT 2 Noding Diagra:n for Sc:all Break 1@' ,, 1@'

3 23 33 E. is

@ 13

-) - g @ 12 (0 L tv s< - G ,,

<< s% 3 >

I gg g$

h# 20 I'

13 l i I

9 se..e 1 D... N.t Sh.we ..

g

.A.41

....1.....

1: u 11 g . i . c.. .. - .

, ca g i. .u.. ,... ,.

@ u> '

,. ...... u ...,...

...c.

.... e .. . . . . .. . , . . . .

5,r.y 57..

Nod e No. Identification Path No. Identification 1 Downcomer 1,2 Core 2 Lower Plenum 3,4,18,19 Hot Leg Piping 3 Core, Core Sypass, Upper 5,20 Hot Leg, Upper Plenum, Upper Eead 6,21 SG Tubes 4,14 Hot Leg Piping 7.22 5,15 fl Lower Head Steam Generator Upper 8 Lore Bypass Pud, SG Tubes (Upper Half) 9,13,24 Cold Leg Piping 6,16 Jw Tubes (Lower Half) 10,14,25 Pumps 8,18 SG Lower Be,-i 11,12,15,16,26,27 Cold Leg Piping 9.11.19 Cold Leg Piping (Pump Suction) 17,31 Downcomer 10,12,20 Cold Leg Piping (Pump Discharge) 23 LPI l' Upper Downcomer 28,29 Upper Downcomer (Above the C of Naz:le Belt)

~ 30 Pressurizer 21 Pressurizer 32 Vent Valve 22 Containment 33,34 Leck & Return Path 35,36 EPI 37 Containment Sprays 1485 334

o o

' Il N f

i I I o o

i

? - ec

. I s I

$ I o o

l

=

- ~ - I ~

su- ~w ,= ~-w c ~e l I

I m ---= E I t ei &&&d& -

t o I

1 o i: iI IX l f

l j ,

fi

.I  :

I l .

I o i T , o N

l X -

a I

w I

} 1 e -

i x o I o W }

0 -

/ X

/ - 8" Hg

[ l ' E N

. j j X l o

~

~

w // / -

8 e x 8

/  !

/ /

y-

o

_ o

/ /f I 1

i 5 D

l l

,l

, /-

i

/

/

./

o o

/ -

. /

I ,l ' *i/' .

i o

/. "

- o

/

h N j,lf('..

/s --

'l

, ,4#'

e: o o a o o o o o o o o O o m N d r-i VISd '380SS3dd 1485 335

FIGURE 3. CORE MIXTURE HEIGHT (CRAFT)

)ig .......... 0.3 FT2 1 ---- 0.2 Fi2 i -X - 0.15 FT i 0.1 FT 2 I 2 I

0.07 FT I .- 0.M FT2 1

1 25 1 gg

\\*

l

\

il i.

i 20

}\gg S

.,. . = y z :

~ m.

'A.s. ~-~.

5

^

'D i %, ; \

3

• 1~ N g

.014 LEVEL OFF 2500 $ -

_/ N /

tn 'O

7,'u 'x ,/ N fv_ _. # % _ m u E 10 V -

~ K-

)

u -

m S I I I I I I I ' ' -

0 200 400 600 800 1000 1200 1400 1600 1800 2000 TIME, S

FIGURE 4. CORE UNC0VERY TIMES VERSUS BREAK SIZES 600 G

d 400 -

m

$ 11.5 FT m

200 -

11.0 FT O.5 FT

=

10.0 FT O _ c 3 2 4O O 0.1 0.2 0.3 0.5 BREAK SIZE, FT2 M

M CD LD u .

U N

I

FIGURE 5. PEAK CLAODING TEMPERATURE O

2000 -

(1760F) u.

2.8%

f 1500 ~

RE IN (1440F)

{

, (1320F)

[ 0.15 FT2 '

.07 FT S

~ 1000 -

g

=

0.1 FT2 g c

k

_, 500

• a en u

U 0 e i e i i e i i 0 200 400 600 800 1000 1200 1400 1600 1800

, TIME, S i

FIGURE 6. AXIAL POWER SHAPE I

2.0 l

I O

x

=

m ca 5 l2 7 g 1.0 _

1

! 3 i

e

.n.

• m 0.0 1 I i  ;  ;

I w

O 2 4 6 8 10 12 u

u ELEVATION IN CORE m

a

FIGURE 7.

FLOW VS VALVE POSITION (Based on Actual Cy VS Valve Posi tion for Oconee HPI Valves) 100 -

l 90 -

CALCULATED CURVE > h 80 -

I 70 -

l FLOW RESPONSE 5 60 - FOR SMALL BREAK d l ANALYSIS d 50 -

l-5 1 40 -- l l

30 -

1 I

20 -

l 1

~

n 10 l co I tn 0 I I I I I I I I I u 5 6 7 8 9 10 11 12 13 14 15

• TIME, MIN ca