ML20206F182

Marked-up Analysis of Station Blackout Transient W/Reactor Coolant Pump Seal Leakage for Seabrook Nuclear Plant

ML20206F182
Person / Time
Site:	Seabrook, 05000000
Issue date:	06/06/1985
From:	Bayless P EG&G IDAHO, INC.
To:	NRC
Shared Package
ML20204G644	List: ML20206C066 ML20206E142 ML20206F071 ML20206F086 ML20206F182 ML20206F313 ML20206G821 ML20206G928 ML20206H081 ML20206H113 ML20206H144 ML20206J252 ML20206J270 ML20206J368 ML20206J403 ML20206J600
References
RTR-NUREG-1150 TRC-74-85, NUDOCS 8704140170
Download: ML20206F182 (25)

Text

v ,

, .. .a . ,. . - _

J .-

Jp*

I 3.33 Attachment 1  ;

June 6, 1985 4 , .

TRC-74-85

, ANALYSIS OF A STATION BLACKOUT TRANSIENT WITH REACTOR COOLANT

. PUMP SEAL LEAKAGE FOR THE SEABROOK NUCLEAR PLANT Paul D. Bayless '

Introduction An analysis of a station blackout transient with loss of all emergency feedwater (THLB' sequence) for the Seabrook plant was performed as part of the Severe Accident Sequence Analysis Program using the RELAPS and SCDAP computer codes.1 The NRC then requested that the transient be l repeated with reactor coolant pump seal leakage taken into consideration.'

A RELAP5/M002 calculation was performed that modeled leakage through the seals of each of the four reactor coolant pumps.

Leakage through the seals may be important because it may allow the .

)

loop seals to clear. If the loop seals are cleared of liquid, superheated l l

steam could flow through the loop piping. Fission products from failed fuel rods could be carried by the steam flow and deposited on the cooler steam generator tubes. Heat transferred from the steam and the fission products may weaken the tubes. Pressure surges caused by interactions between molten core material and the liquid in *.he reactor vessel lower l plenum could cause the steem generator tubes to fail, praviding a path for fission products to leave the containment.

i Boundary Conditions i

The boundary conditions for this calculation were the same as for the base case calculation described in Reference 1. with the exception of the pump seal leakage. The area of the pump seal leak path was assumed to stay at its steady state value until 2700 s into the transient. At that time. the leakage flow area we.s increased to a maximum' expected value for the rest of the transient.

1

~

ruo4140170 87040s PDR NUREG 1150 C PDR _

. . , . . . .; - - . . .-.z,,..,. ,

.. . 7 7.. . ., .;. , , , .. .

o ,

, y The eady state flow area allowed a leak rate ofg1/s (20 gpm)

~

pg) with the cold leg liquid at 15.51 MPa (2250 psia) and 56,5 Kg . . c.

.(557 *F). Themaximumflowareaallowedaleakagemf.,j0.04s#(475gpm).

pg,,gump with the liquid at 15.51 MPa (2250 psia) and 565 K (557 'F). r Yhese values reprasent maximum leakage flows expected for Westinghouse 'I -

pumps. The 30.0 1/s (475 gpm) flow was calculated by Westinghouse, while an independent consultant calculated a v'alue of 26.5 1/s (420 gpm).2

.w . .

. Small-gse experiments performed by Atomic Energy of Canada Limited

, showed a minimum tine ts. seal _failuresof;pbout ,one hour.2 However, the experimenters cautioned against extrapolating these results to full scale ,

pump seals. Analyses perforwed for_the..IDCOR program assumed..that the pump seals failed it 2700 's'.3 It was felt that using 2700 s for the

%y.,timeoffullsealfailureinthisanalysiswouldleadtoaconservative

$ ulation in that the loop seals would be more likely to clear than if the pump seal failure occurred later. There are g ata_that indicate the pump seals would fail before 2700"s.

Results Figures 1 through 18 present important parameters from the RELAP5 calculation. In most cases, the corresponding parameters from the (praviously reportrd) base case calculation are also presented. Table 1 l presents a sequence of e<ents comparison hotween the TMLB' calculations with and without reactor coolant pump seal leakage. A discussion of the plant transient behavior follows.

At transient initiation, the reactor scrammed, and the reactor coolant and feedwater flow coastdowns began. The core power decreased .

more rapidly than the coolant flow. causing the average coolant

- temperature to decrease. The associated shrinkage of the coolant drew liquid from the pressurizer, expending the steam volume, which in turn caused the pressure to decrease. As the reactor coolant pump coastdown ,

)

ended, natural circulation flow through the loops was established. 1 2

The pressurizer pressure and level slowly decreased until 2700 s, the result of the mass lost through the reactor coolant pump seals. The oscillationsin,thehtwoparameters,aswellasinthereactorcoolant

,u 7 r. .

systemtempert.tures,werecausedbythesteamgeneratorpressurevprying between the relief valve opening and closing pressures. At 2700 s, the leak rate increased, and the pressurizer level and pressure decreased more rapidly. The saturgion pressure of the liquid in the core and hot legs was reached near 2850 s. The depressurization stopped for a short period as the liquid began to boil, then continued. The depressurization rate decreased when the pressurizer emptied near 3020 s.

The fluid draini gn from the pressurizer raised the temperature of the fluid in the single loop hot leg. The increased heat transfer in the steam generator caused the secondary liquid to boil off faster in the single loop than in the combined (three) loop steam generator. Shortly after the cold legs reached saturation, natural circulation flow through the loops stopped. Natural circulation flow ended at about 3730 s in the single loop, and nearly 400 s later in the combined loop. The loss of flow lowered the heat transfer rate, and the liquid level on the secondary side of the steam generator decreased more slowly. All of the secondary side liquid had boiled off by 5200 s in the single loop and by 5460 s in the combined loop.

The lower heat transfer rate in the steam generators together with the boiling (in the core) of the 1 squid draining from the cold legs caused the reactor coolant system pressure to start increasing shortly after 4000 s. The pressure dropped briefly near 4200 and 4500 s because steam was condensing in the cold legs. Some of the liquid in the steam generator tubes drained, forcing cooler water from the loop seals into the cold legs, where some steam was condensed. By 5100 s. most of the liquid in the cold legs had drained. The higher quality of the fluid in the cold legs resulted in higher volumetric and energy flows out the pump seals, which caused the reactor coolant system to depressurize. The depressurization continued until the liquid remaining in the steam l generator tubes drained about 450 s later.

3 y -

.~

. l l

Even though the hot and cold legs had drained, some liquid remained in the steam generatior tubes. This liquid holdup was probably a i code-related phenomenon that would not occur in an actual plant. When ,

I this liquid drained at about 5550 s. three phenomena resulted that

! contributed to the pressure increase. The heat transfer rate to the secondary side of the steam generators decreased because the liquid was replaced by steam.- The liquid draining lowered the void fraction in the cold legs. reducing the volumetric and energy flow rates out the pump i seals. Finally. the liquid drained to the reactor vessel where it could be boiled in the core.

4 i

The ' pressure increased at a lower rate after about 6500 s because the

> cold legs had again drained. increasing the energy and vulumetric flows through the pump seal leak path. This was not sufficient to cause a depressurization, as it had earlier, because of the reduced heat transfar rate in the steam generators.

Shortly before 7000 s. sufficient mass had been lost that the top portion of the core began to heat up. The power-operated relief valves on the pressurizer opened only once, at about 7200 s. After that time, the i

pump seal leakage, coupled with a lower boiling rate in the core (bect.use

~

of the decreasing liquid level), was able to reduce tne reactor cooian b system pressure. By 7800 s the maximum fuel cladding temperature exceeded 1000 K (1340 'F) and calculation of cladding oxidation would have begun in the SCDAP (or linked SCDAP/RELAPS) computer code. The energy released

! by the oxidation reaction would slow, or pcssibly step.,the ,

depressurization. The calculation was terminated near 8600 s when the maximum steam temperature reached 1500 K (2240 'F).

At 7800 s there was about 3 m (10 ft) of liquid on the pump side of f

each loop seal and 2 m (7 ft) on the. steam, generator side. At a corresponding time in the base case, each side of the loop seals contained about 3 m (10 ft) of liquid. The presence of this much liquid in the loop seal prevented superheated steam from flowing around the loop. In fact.

the steam in the hot legs and steam generator tubes was at the saturation temperature. Some superheated steam had been drawn into the hot legs when 4

i

1 0

4 the power-operated relief valves opened, but it had since cooled. The lack of relief valve flow prevented superheated steam from entering the hot legs from the reactor vessel. The pump seal leak drew superheated steam from the upper pler.um into.the cold legs via the_ upper head region of the reactor vessel. The steam temperature when fuel cladding oxidation  ;

would have begun was80-110 K (150-200 *F) cooler in the pump seal leak case cold leg than it was in the base case hot leg. The lower temperature was caused by heat transfer to the structural metal in the upper head and upper portion of the downcomer. i Conclusions e

including reactor coolant pump seal leakage in a TMLB' station blackout transient for the Seabrook plant shortened the time to core 47l**

heatup and fuel cladding oxidation by 1389 s compared to the base transient. Sufficientliquidremainedinthel()psealstoprevent superheated steam from flowing through the loop piping. Specifically, the steam in the steam generator tubes was at the saturation temperature, and the tubes would not be expected to fail based on his analysis. The lack of relief valve cycling kept superheated steam from being drawn into the 2 hot legs, while the pump seal leakage did draw superheated steam into the cold legs.

I With a large enough pump seal leak, then, a piping failure caused by <E ~~~'

4 overheating would be mora likely to occur in the cold leg than in the hot f leg. The opposite would be true if the majority of the flow were out the relief valves rather.than through the pump seals.

j A TMLB' calculation modeling pump seal leakage will be preformed using the linked SCDAP/RELAP5/ TRAP-MELT computer code. This integrated analysis should determine whether or not the loop seals clear, and assist l in the evaluation of the reactor coolant system pressure boundary l integrity prior to the reactor vessel melt-through.

1 o

i 5

i e=* ..

References

1. P. D. Bayless and R. Chambers. Analysis of a Station Blackout Transient at the Seabrook Nuclear Power Plant, EGG-NTP-6700, September 1984

2. Presented at a Nuclear Regulatory Commission meeting on pump shaft seal integrity. Washington, D.C., December 13, 1984.

3. Zior Nuclear Generatino Station, 10COR Task 23.1 Integrated Containment Analysis.

j 6

~~ W-

_. __ _ , , _ . . . 1__ . _ _ _ __ . . _ . _ . _ _ . . _ . _ _ _

1 TWL S,

. Table 1. Sequence of Events for the Pump Seal Leak and Base Cases.

4 Time (s) l Event W. A/*? Leak case Base case (d* IN g)

Transient initiation ' 3 N2 "*FI 0 0 Leak flow area increased 1 2700 a j g 2850 6514 Hot legs reach saturation temperature p3 a

Pressurizer empty 3032 Natural circulation ends 3728b,4116e 6797 6- $

Steam generators empty 5196b ,5460c ,5008 ' -/f W Fuel cladding heatup begins 6984 e 8289 Maximum fuel cladding temperature , ,,gy J{*'

exceeds 1000 K (1340 *F) 7776

• s** 9060 '

a. Did not occur

b. Single loop (loop with pressurizer)

c. Combined loop g 3 .h~ I"r3 O

d 7

l )

D 9

0

- 0 0

- - 0 1

o <4 ccL s

/

0 l 0 I

, " 0 8

, * . s e a e

• r c t , - u e

o u o, 0 s

s s k

g l 0 e a c I 0 r B a p -

6 l

b V ) r e

O n s z e

i t

o a

J- p (

e i

s r s u a c

sc t

s m s i e k T r a k 0 P e o 0 l o a ,'$ l .

c oc I

r a 0 1 l b t, 4 a a c n ,) ,

e r s e

e -

S = , u 1 D i

g p i

n

!: F u e'l -

s P s -

E V A h9 v 0 I

~' s- pa y l 0 0

i

- 2 d b^

3, ^

^

^

^

{^^

n a

i 0

C s 0 s 2 1 1 25 ,!in.

a c

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

A .o c

l Seabrook station blackout

.1 100 , ,  ; c) ,

i i

= o P 50 - -

g - 1ssst e s issssss e -- r e-n, 4

l 4

I i

0 I 4 I. 1 _. l 0 2000 4000 6000 8000 10000 i

Time (s)

I Figure 2. Pressurizer collapsed liquid level.

i

A-Pump seal leak case O-Base case 2

J

P

~

Seabrook station blackout .

640 l -

l l l O

620 - _

l n

. 6 t

j 600 n-- __

5 5 H

- - - g .__ _ -g_

I I l l 560 0 2000 4000 6000 0000 10000 Time (s)

Figure 3. Single loop hot leg liquid temperature.

A-Pump seal leak case O-Base case

-_ - _ _________ _ - __ - __ _ - _ _ _ _ _ _ _ _ __ _ _ ^ - - - - ^ ^ - - - - - - * - _ _ _ _ _ _ _ _ _ , _ _

5 '

e Seabrook ctation blackout 1000 ,

, i l l I

't iV n

800 - --

v x

i

.5 2

h i

m - Am L ~

f N 600 W P 400 0 2000 4000 6000 8000 10000 i

Time (s)

Figure 4. Single loop hot leg vapor temperature.

A-Pump seal leak case O -Base case i

9 .

e Seabrook station blackout 650 l I I I I I 62s - -

O

. Q v

i 3

y 600 a r- -

1.

5 u -

s75 -

I I I I l sm 0 2000 4000 6000 8000 10000 I Time (s)

Figure 5. Pump seal leak case single loop hot leg fluid temperatures.

A-Liquid 0-Vapor

a W

e Seabrook station blackout 640 I I l l l 620 - -

E 6 3

y r% - -

h

- G u

580 - -

I I I I ss0

' O 2000 4000 6000 0000 10000 Time ( s)

F15ure 5. Single loop cold leg liquid temperature.

A-Pump seal leak case O-Base case

, O

• t Seabrook station blackout 000 ,

, I I I I d,

n 700 - -

i x u

j e

i E

(

x Q

i 800 _ _

, p- ::=:s:::::_ :=:s .

i i

)

see 1 - I --_ I I _

i 2000 4000 6000 O 0000 10000

^

Time (s) i l Figure 7. Pump seal leak case single loop cold leg fluid temperatures, j A-Liquid O-Vapor l

4 i

Scabr00k station blackout 4

-l 1.5 l l l l 0

J -

1.0 -

?.

N E

La 5

tr -

0.5 -

r a 6

. _a

, 5; b.

i; =

a

! r l(- .O .

' g C i O e.e - - -

O 1 -

2 I  !  ! I -

-0.5 0 2em 4000 sees seee lease Time (e) 4 Figure 8. Pump seal leak case hot leg riass flow rates.

A-Single loop O-Combined loop i

_ _ _ . _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ m _m _ . _.

a Seabrook station blackout ,

35 -

l 1 I I I O

10 - -

E l v l

'f 1 fpI r

?a- M E .3 5 -- -

1 0 O 1O 10 !

0 2000 4000 6000 8000 10000 Time ( s )

Figure 9. Combined loop steam generator downcomer collapsed liquid level.

A-Pump seal leak case O-Base case L

1 l , ,

a ,

9 Seabrook station blackout 15 I I I I

O

/

10 -

E l v

m k, 7  ;

J jr

'E I'

. u y -

J -

r 5 -

2

[ I I I O iO l e -

e 2ees 4ees 8000 8000 lesee Time (s)

Figure 10. Single loop steam generator downcomer collapsed liquid level, 6-Pump seal leak case O -Base case il ,.

,t

1 ,

i <

Seabrook station blackout i

9 i I I I i 1

i t

5 $

7 -

l~

m o

i 6 0 2000 4000 6000 0000 10000 Time (s) i Figure 11. Pump seal leak case steam generator pressures.

A-Combined loop O -single loop

4 0

0 m 0

- - 0 _

1 _

0 _

0 i

0

% 8 t

s e

a _

r p t

f w o o o

u 0 l

f l

o e k 0 s l _

c I 0 s g a 6 a n l _

b ) mi _

S n s k - _

o ( a i e O t

a e ml l

p t _

s i a o T e ol _

k 0 s o /

0 p d o I ,

0 m e r u n b 4 P i a b e m S .

2 o .

1 C e

lI u r A u

g i

0 F I

0 0

2

- - 0 5 0 5 0 _

7 5 2 _

O%e 2& r

f G _

i i: i.

l 0

0

. 0

- 0 1

- n i

o

- t c

- 0 a 0 r

'r 0 f y 8 d i

n o v

r g

, e

-g l d

l t 0 o u c o 0 p k

I 0 o c 6 o a ) l l

b s d

(

n e o n i e i b

t a

m m t i o s l 0 T c k 0 e o s I

0 o a r 4 c b k a a e e S l l

a e

s 0 p I 0 m 1

0 u 2 P _

3 1

e r

u g

i F

- 0 0 5 0 1 0 0 _

EY,at 3 i

mo

- . : ' , i  : 1

a Seabrook station blackout

~

l l l N ,g r t

E u I a

i il 0.5 -

2 eo

~

I .

]

O.0 0 2000 (4000 6000 8000 10000 y , . }, n,,u. e. t\ lo-

. Time ( s )

Figure 14. Pump seal leak case single loop cold leg void fraction.

O ca i

-l Seabrook station blackout ,

1 1

1.0 i i i  ; i  ;

1 1

i E

z e

!' II 0.5 - -

2 0 -

1 I

O.0 0 0 O- d O 0 2000 4000 6000 8000 10000 Time (e) .

Figure 15. Top core volune void fraction.

A-Pump seal leak case O-Base case

0 0

0

- 0 1

e r

0 u l 0 t a

A - 0 r 8 e p

m e

t e e c s a

M -

'2 c t -

u 0 e e o

k c

a

, -~I 0 0

s s g B n -

a l

6 i

)

O b d n

s d a

o - (

l e i c s t

a e a c

t m l e

s i

- u k 0 T f a k e o

o 0 el r

m .I 0 m 4 u l b l a a

e o

v s e

S e g r n o u c P mu p 0 T oA I 0 .

0 6 2 1 e

r

. .w u g

i _

F _

' - - >e 0 4 .

0 0 0 0 0 0 0 0 0 5 0 .

5 2 1 1

,x* h%ka t

Mu l, .i! i,. , ,4 i  :

I !i , Il  ! ; l

l lI t

0 0

I 0

- - 0 1

1 e

r u

0 t a

I 0 r I

^

0 e p

8 m e

- t e

c a e f s r a t

m u c s

u 0 e o I 0 g s k I 0 n a c i B a 6 d -

l ) d b s l a O n ( c o e i

t e l e a s

t a

i m u c f

s T k 0 e a k

o I 0 mu l e

o r

I c 0 4

l o l b v a a e e e s S r o p c mu mP c t t

o -

A 0 o l

I 0 B 0

2 7 1

e r

u g

i F

- - g w O _

0 0

0 0

7 0 0 0

s ,

oo e=1ki-e _

s

l! . l i

,s< -

0 -

0 -

0

- 0 _

1

~

' 0

( I 0 O 0 0

~ e

t. r u

t a

r e e

p a s

^

m c t

k l

b u

o c

a pI 0 0

0 6

)

s t

e r a o

p -

a v

B O

e s-n 0 (

e i

o e mu e

s t m l a

- t a i T

o v

c s 0 e a k

0 r e c

I k  ;

o 0 o l o 4 c r l

- b a

p a o e e T s S . p 0 m 1 u P

^ e -

r 0 u A

I 0 g 0 i F

2 e

e 0 0 0 0 0 0 0 5 0 5 1 1 mxv=2.k !-s

~

,