ANALYSIS OF A TWO INCH SURGE LINE BREAK RESULTING IN A POSTULATED

PRESSURIZED THERMAL SHOCK VESSEL FAILURE IN THE CALVERT CLIFFS POWER PLANT

SUMMARY - A 2" surge line break in Calvert Cliffs was postulated to result in a PTS vessel

failure after nominal PTS conditions were achieved. The principle purpose of this analysis was

to provide mass and energy release information to be used in pressurized thermal shock (PTS) containment failure analyses. Two subsidiary purposes were:

1. To provide RCS pressure information that might be helpful in assessing the potential for

RCS internal damage

2. To provide information about the timing of possible core damage after the vessel failure

A two inch surge line break was analyzed using RELAP5/MOD3.2.2 (version "Relapon") starting

with a large break input deck that was developed for IPE success criteria analyses and

improved for Appendix K rule change studies. After the annulus water temperature reached

3000F, the vessel was assumed to fail. The problem was then restarted at the closest restart time assuming a vertical split in the vessel.

PROBLEM DEFINITION AND DEVELOPMENT - The particular LBLOCA input deck (calsb33.i)

was chosen as the starting point because it had a multi-region downcomer (12 axial regions, 6

azimuthal regions and 1 radial region). Six of the axial regions spanned the same elevation as

a 20 axial region core. The core also had a hot channel equivalent to one fuel assembly which

also included the heat structure for an additional hot pin. Figure 1 shows the vessel noding

used for this problem. In order to perform a suitable PTS calculation, certain trips were

modified to assure full pumped ECCS and aux. feed operation. The large cold leg break was

replaced by a surge line break near the bottom of the surge line. After the 300°F vessel failure

criterion was reached, the problem was restarted at the next subsequent restart time with a 12

square foot break extending from the top to bottom of the active core (See Figure 1). The

vessel break consisted of 12 one square foot junctions discharging from each azimuthal face of the volumes shown in Figure 1.

Six vessel cavity volumes at the break junction elevations served as break repositories. A 10

volume vessel cavity (including the 6 break repositories) was developed based on information

received from the utility (BG&E) for a previous direct containment heating (DCH) study. Other

containment components included a tunnel to the sump (lower containment) and a large volume

upper containment. A large containment wall heat structure was in place from the original deck, but no containment mitigation systems were used or available. The containment volume was

arbitrarily increased to 4 million cubic feet to limit the containment pressure rise.

SYSTEM RESPONSE PRIOR TO VESSEL BREAK - The surge line break was initiated at time

zero. The RCPs were manually tripped at 150 and 200 seconds according to small break

procedures. Natural circulation was lost at about 1000 seconds (See Figure 2). Vessel

pressure in the annulus is shown in Figure 3. Figure 4 shows the collapsed core and annulus

water levels. As is expected in a hot leg side break, more than sufficient water level is available to maintain core cooling, especially with full ECCS.

Attachment

In earlier versions of this calculation, numerically driven flows were observed. This phenomenon was discussed by Dave Bessette (Note to Farouk Eltawila, et al., 1/25/01) as part of the review of NRC PTS calculations for Oconee. The imposition of high reverse flow pump resistance was used for Calvert Cliffs as was done for Oconee. Figure 5 shows average loop flows from the termination of natural circulation until just before the vessel break after the loop flow "fix" was invoked. As can be seen no "numerically driven flows" are present. Figure 6 shows the water level in the containment components. The BG&E information says that at least one foot of water is required in the sump before sufficient NPSH is available for recirculation. This calculation indicates that this level is not present until after vessel failure. However, without containment mitigation systems, that important contribution to sump inventory is not considered. Figure 7 shows that the average of 24 annulus water temperatures closest to the core midplane reach 300°F by 3400 seconds after the surge line break. At that time the vessel break was initiated. Use of the pressures in Figure 3 and temperatures in Figure 4 should provide information to determine if the 300°F criterion used in this analysis is appropriate.

SYSTEM RESPONSE AFTER VESSEL BREAK - The next 8 figures are presented to provide boundary condition information for the PTS containment failure analysis. As shown in Figure 8, the vessel break blowdown is over within 10 seconds. Figure 9 shows that the vessel break flow increases somewhat after the blowdown when excess ECCS water spills out of the break. The high flow between 3420 and 3460 seconds is due to SIT injection. Figure 10 shows a break mass flow of -1350 lbs/sec., due to pumped injection spilling out the vessel break. Figure 11 is the vessel break enthalpy flow rate during the vessel failure blowdown. Figure 12 is the vessel break enthalpy flow rate for the entire post-vessel failure period of the transient. Figures 13, 14 and 15 are presented in a form that may be useable for some containment codes. For these figures, the enthalpy flow rate was divided by the mass flow rate to provide a specific enthalpy of the break flow. When the net flow is near zero, countercurrent flow may be present or flows may be in different directions for each of the break junctions. Under these conditions of "mixed flow", a single specific enthalpy to be applied to a single break flow appears oscillatory and sometimes negative, which of course is impossible. Therefore in Figures 13 and 14, a "faired" line is drawn for the short time period between 3409 and 3422 seconds.

The next three plots are provided to assist in assessing the potential for vessel component and containment failure. Figures 16 and 17 show the pressure drop across the core barrel nearest the mid-point of the vessel split. Figure 17 shows that the maximum AP occurs within 30 milliseconds after the vessel split. Figure 18 shows that the lowest vessel cavity volume reaches a pressure of nearly 160 psia within 2 seconds of the vessel split.

The balance of the figures describe the onset of core damage that is calculated to occur within 800 seconds after the vessel split. Figure 19 shows the cladding temperature behavior of the three pins represented in this calculation. The single hot pin has a peak linear heat rate of 14.5 kw/ft at the 10 ft. elevation. The hot assembly is represented by a pin that has an average power of all the remaining pins in the hot assembly. The average core is represented by a pin that has an average power of all the remaining pins in the reactor. The calculation terminates in RELAP5 when the cladding temperature reaches melting, which is the limit of the conductivity property table. As shown in Figure 19, this is about 3500'F. For both the hot pin and the hot assembly the influence of the zirconium-steam reaction can be seen by the sharp upturn in cladding temperature. The average pin, however, has not yet reached a temperature where metal-water effects are significant. Extrapolating the average pin behavior would indicate that approximately 500 more seconds would be needed to approach clad melting for that pin. It should be noted that the solid body radiation model was not invoked in RELAP5 as it would normally be in SCDAP or MELCOR.

The cladding temperature heat up begins after termination of accumulator injection (@ 3456 seconds). The cladding temperature behavior is dependent on the water level in the core. Figure 20 is a plot of core and annulus collapsed liquid level relative to the bottom of the core. A dotted line is drawn to approximate the mixture level in the core. This line is based on individual void fractions in the average core channel. The collapsed level in the core is very dependent on the collapsed level in the annulus. For this problem, the annulus collapsed level is dependent on the vessel break characterization. The intent of developing 12 one square foot break junctions from the vessel to the cavity over the length of the core was to simulate a longitudinal 12 square foot vessel split. The results appear to be reasonable. After the system has depressurized, the break is a very low energy process in which water is spilled out of the bottom region of the break. Figure 21 shows the individual liquid flow rates from all 12 break junctions during this time period. Also shown is the total break flow from all 12 junctions. The four lowest junctions each contribute about 320 lbs/sec for a total of 1280 lbs/sec or about 95% of the break flow. Figure 21 also shows that the total vessel break flow is about 1340 lbs/sec., which is equal to the total pumped ECCS flow with all pumps operating. Thus most of the water being pumped into the RCS is being spilled out the break, which is to be expected in this large vessel break.

There is a concern however, that RELAP5 may not be calculating the annulus water level properly for this kind of spillage. Annulus water level is a strong determinant of core liquid level

and the time available before core damage begins. Figure 20 shows that the annulus collapsed water level during this time period is about 2 feet. This height is relative to the bottom of the

vessel split. If a rectangular split is assumed, the Francis formula for constrained weir flow can

be applied and solved for the height of water above the crest of the weir. By assuming upstream velocities of zero and a maximum equal to the average spillage velocity, the water

height is calculated to be between 3.5 and 4.3 ft.. Thus on this very simplified basis, RELAP5

does not appear to be "overestimating" the water height in the annulus. It should be noted that

weir formulae do not account for upstream pressure effects or other complexities that would

exist in the annulus, especially in the break region. A good deal more assessment of vessel

splits is needed including the nature of the break flow over a very wide range of conditions.

CFD calculations could be helpful in scoping some of the issues related to annulus water level behavior.

[Ojwoup

009 wfulU8d 48m01l

LO-00S __ 90- s0- V70 co0- F0- 10-009

01

-t'09*t'

-IGSSOA JE)uui

0LI

.0119 6Z- 8z- Z : 9- g

-- - -- - - -- - - - - - -- - - - - - -- - - - - - -- - - -- - -9-- - - -

iiitj -------- .....J................... ------cL 9L L

LjL ýl9 I~

sflnfuu\/ uo!beH IeqlflW!zV X!S[L9

wL:Jbu!G bu!PON snIlnuuV Si~d SjIIIO P)j'8A0 L ainb§j

'0- LO-0- e0

0 o c

CD!9'

.ieddfl

.jeddn

.........

0CS9

9

0

9

0L

CD

CD

m CD

03

0L7

98681

3869*

3868*

3868*

3868*

LLT

LUT~

zI

Calvert Cliffs 2" Surge Line Break PTS EventFigure 2. Average Core Inlet Flow

30000

25000

20000

15000

10000

- mflowj-505010000]

0 500 1000 1500 2000 2500 3000 3500 4000 Time (s)

coreflow

E -Q

4'Z

m

0r

0 1L

5000

0

Calvert Cliffs 2" Surge Line Break PTS Event Figure 3. Pressure in Vessel Annulus @ Core Mid-plane

2500

2000

C, (.

1.3_

1500

1000

500

00 500 1000 1500 2000 2500 3000 3500 4000

Time (s) vessores

Calvert Cliffs 2" Surge Line Break PTS Event Figure 4. Vessel Collapsed Water Levels

Average Core

0 500 1000 1500 2000 Time (s)

TAF

2500 3000 3500 4000

veslevels02

25

20

15

..-

10

5

0

Calvert Cliffs 2" Surge Line Break PTS Event Figure 5. Loop Flows*

Surge Line Break

SG 0154 SG,

10 F2•0

148 12

LOOP 1 LOOP2

ECCS ECCS 145 3\0245

ROP 151 251RP 35 42

ECCS ECCS

175 ý275 3\0 3ý0r

RCP 181 F2871 DRCP

35 42

* Loop flows are average flows from 1100 to 3100 seconds.

M are junction numbers. Arrows and adjacent numbers indicate direction and magnitude of flows.

Ioo1flows

Calvert Cliffs 2" Surge Line Break PTS EventFigure 6. Sump, Tunnel & Cavity Water Levels

4.5

4

3.5

3

£ 2.5

-J 2

1.5

1

0.5

0

'I i!

IIII

- voidf-890030000

................ voidf-891090000

-- voidf-89401 0000

sumDIvI

Sump_Level required for recirculation

0 500 1000 1500 2000 2500 3000 3500 4000 Time (s)

,i

Calvert Cliffs 2" Surge Line Break PTS EventFigure 7. Average Annulus Liquid Temperature*

I I I

LLI

>

600

550

500

450

400

350

300

250

200

150

100

50

0

* Average of 24 temperatures around core mid-plane

0 500 1000 1500 2000 2500 3000 3500 4000 Time (s)

annliatem=

I - tempf-571070000

a_

L.

1.J

0E

E 0

fiterion Toi ii-fiatfing vesseIlailare

. .l l l l l l i l l l l l . . l i I l l. . i . I i l i l l l lIl. . I . l r . L I

Calvert Cliffs 2" Surge Line Break PTS Event Figure 8. Vessel Break Mass Flow (Ibs/sec) LIIva

160000

140000 ( 0

120000 co

= 100000 (D

80000 0

z 60000

40000

20000

O . 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410

Time (s)vesbrkflo03

Calvert Cliffs 2" Surge Line Break PTS Event Figure 9. Vessel Break Mass Flow (Ibs/sec) LIIIIvarI98O

7000 ..... . ..... I......... .... [.... ......... ......... . ... ..... I.....I.....I.....

6000 uU) (D 0/)

5000

4000

3)

o 3000

2000

1000

3390 3400 3410 3420 3430 3440 3450 3460 3470 3480 3490 3500 Time (s)

vesbrkflo02

Calvert Cliffs 2" Surge Line Break PTS Event -s cntrlvar-1980

Figure 10. Vessel Break Mass Flow (Ibs/sec)

2000

1800

1600

1400

1200

1000

800

600

0

U)

400

200

0 , 3300 3400 3500 3600 3700 3800 3900 4000 4100 4200

Time (s) vesbrkflo01

Calvert Cliffs 2" Surge Line Break PTS Event Figure 11. Vessel Break Enthalpy Flow (BTU/sec)

4e+07

3.5e+07

3e+07

2.5e+07

2e+07

S1.5e+07 W

1 e+07

5e+06

0 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410

Time (s)enthflow

Calvert Cliffs 2" Surge Line Break PTS Event Fir I- cntrlvar-1981

Figure 12. Vessel Break Enthalpy Flow (BTU/sec)

500000

450000

400000

350000

300000

250000

200000

150000

100000

50000

0 3300 3400 3500 3600 3700 3800

Time (s)3900 4000 4100

enthflow02

0

I->13

0

-" Cwl

4200

Calvert Cliffs 2" Surge Line Break PTS Event Figure 13. Vessel Break Enthalpy (BTU/Ib) 11111

600

550 CO I

500 a):") Interpolation for "mixed flow"

450 >

.. 3 400 D Q

250

U) 1300

UJ_ 50

3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 Time (s)

br kentnh02

Calvert Cliffs 2" Surge Line Break PTS Event Figure 14. Vessel Break Enthalpy (BTU/Ib) I vaI198O

3400 3405 3410 3415 3420 3425 3430 3435 3440 Time (s)

brkenthOl

600

550

500

450

400

350

300

250

200

150

100

50

co

w C,

Q

al)

Q3.

co

0 L 3395

Calvert Cliffs 2" Surge Line Break PTS Event I cntrlvar-1980I Figure 15. Vessel Break Enthalpy (BTU/Ib)

80

75

,. 70

"" 65 C.

W 60 0

55

50

45 3300 3400 3500 3600 3700 3800 3900 4000 4100 4200

Time (s) brkenth03

Calvert Cliffs 2" Surge Line Break PTS EventFigure 16. Pressure Drop Across Core Barrel I - P-51 10000

0 33 I I ..... I I 340 I 3403 .... -15 3 , , ...0 .. 3 0 3 0 3.10

3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410Time (s)

corebaridoo0

300

250

-. 200 "0

0

o 150

U)

0 100

50

Calvert Cliffs 2" Surge Line Break PTS EventFigure 17. Pressure Drop Across Core Barrel

3400 3400.02 3400.04 Time (s)

300

250

200

150

100

C,o 0t

n CO,

p-51 9010000

3400.08

50

0 .... 3399.98 3400.06

corebarldD02

Calvert Cliffs 2" Surge Line Break PTS Event p-891010000

Figure 18. Cavity Pressures -p-891020000 -- D-891030000

1 6 0 . .. .. .. .. I .. . . .- I I . . . . . 1 .. . . ... I . .... , . ..-.. . . . I, p -8910' Slower - p-8910!

2 1 ............... ... P - 8 9 1 0 S-- p-8910(

___P-891 0 140 ('L p-89101 S....... p-8910 • "-I ... p-89111

cf) upper >1

120

•. 100 .a

• •80 co)

i- 60

40 I

20

0 . . . . . . . . . . . . . . . . . . . . . I

3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 Time (s)

cavpres

40000 50000 60000 70000 80000 90000 00000

Calvert Cliffs 2" Surge Line Break PTS Event -httemp-515401809

Figure 19. Hot Spot Cladding Temperatures for Each Pin -httemp-515301809 httemp-515101809

o... It Hot

Rn

3500

3000

"Average Pin

3400 3500 3600 3700 Time

3800 3900 4000 4100 (s)

cladtemD

/Hote kssembly

U) .- I

a)

E

D_

H.9O

.D 02

2500

2000

1500

1000

500

0 0. 3300 4200

Calvert Cliffs 2" Surge Line Break PTS Event Figure 20. Vessel Collapsed Water Levels F

20

15

"5 10 _J

5

3300

Core froth level approximation

Average Core

3400 3500 3600 3700 3800 Time (s)

3900 4000 4100 4200

veslevelsOl

Calvert Cliffs 2" Surge Line Break PTS Event Figure 21. Vessel Break Liquid Mass Flows

1600 .

1500

1400

1300 total vessel break flow

1200 =-1340 lbs/sec 1100 75

S1000 >

-n 900

* 800

700 0

S600 total liquid flow from bottom four break junctions

500 =-1280 Ibs/sec

300

200

100 t0alal'uid flow fro= r•ercning eight Drea qJunctions lsec1 2 347 8 910

0

3300 3400 3500 3600 3700 3800 3900 4000 4100Time (s)

vesbrkliqflo

4200