ML20067B721

Revised Prairie Island Large Break LOCA Calculation Using WCAP-10924,Vol 1,Addenda 4 Model Changes

ML20067B721
Person / Time
Site:	Prairie Island
Issue date:	02/28/1991
From:	WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML20067B716	List: ML20067B714 ML20067B721
References
NUDOCS 9102110007
Download: ML20067B721 (17)

Text

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

1 i

J REVISED PRAIRIE ISLAND LARGE BREAK LOCA CALCULATION USING WCAP-10924 VOL. 1 ADD. 4 H0 DEL CHANCES Westinghouse Electric Corporation Nuclear Technology Systems Division Nuclear Safety Department safety Analysis Technology .

February 1991 9102110007 910:05 PDR ADOCK 05000282 P PDR 0375NicSAl 2/5/91:1 I

l. Introduction This document reports the results of an analysis that was performed to demonstrate that Prairie Island, meets the requirements of Appendix K and 10CFR$0.46 for Large Break Loss-of-Coilarit-Accidents (LOCA) (Reference 1).

11 Method of Analysis The analysis was performed using the Westinghouse I.nrge Break LOCA Best-Estin,iie Methodology as given in (Reference L) and as amended in Reference 6. lho Westinghouse Best-Estimate Methodolooy was developed consistent wit , guidelines set forth in the SECY-B*-472 s document (Reference 3) These guidelines provide for the use of realistic models and sumptions, 8 th th) exception of specific models and assumptions required by

, <ix 4. 1.ie technical basis for the use of this mrdel is discussed in detail ti Reference 2.

The Best Estimate Methodology is comprised of the WCOBRA/ TRIS ind COCO computer codes (Datertrces 2 and 4, respectively). The WCOBRA/ TRAC colt was used to gontrate thq cockiete transient (blowdown through reflood) systte Sydraulics as wen as tor cladet 9 thermal analysis. The COCO code was usoo to rensrate the containment press'i a response to the mass and energy release fron 0 3 break.

This contaitiment ocessure curve was used as an input to the WCOBRA/iRAC ;'de.

The fuel parameters used as input for the LOCA analysis were generated using the Westinghouse fuel performtnce code (PAD 3.3) (Reference 5). The fuel parae9tarG iriput to the code were 'at ' inning-of-life (maximum densification) values.

9/

The awalysis was j erformed using the four channel core model developed in Ref'erence 2 for the Oc4 double-ended celd leg guillotine (DECLG) break. The transient was conside-(g te be termina ted if the hot rod cladding temperature scgan to decline and the 1. iocted ECCS flows exceeded the break flow.

0375N:tsAt 2/5/91:2

Ill. Results and Conclusions Table I shows the time sequence of events for the Large Break LOCA transients.

Table 2 provides a brief summary of the important results of the LOCA analyses for this chiculation. Figures 1 through 8 show important transient results for the limiting 0.4 DECLG break (four channel core model). Note on these figures that the break occurs at time 0.0 (the results from -20.0 to 0.0 are from the 1 steady state). Figure 1 shoss the core pressure during the transient.

Figure 2 shows the vapor and liquid mass flowrate at the top of the hot assembly. Figures 3 and 4 show the collapsed liquid level in the downtomer and core bot assembly channel, respectively, indicating the refilling of the vessel. Figures 5 and 6 show the flow of ECCS Nater into the cold leg (accumulator and high head safety injection flow) with Figure 7 showing the flow of low head safety injection into the upper plenum (UPI flow). Figure 8 shows the resulting peak cladding temperaturo for the 0.4 DECLG break as a function of time for each of the five feel rods modeled. Rod 1 is the hot rod in the hot asseWj channel, Rod 2 is iae hot assembly average rod, Rods 3 and 4 represent average assemblies in the center of the core and Rod 5 represents the lower power assemblies at the edge of the core. The safety injection (SI) 4 system was assumed to be delivering to the RCS five seconds after the generation of a safety injection signal. This five second delay includes the time required for developing full flow from the SI pumps. No additional delay was required for diesel startup and sequencing since the analisis assumed reactor coolant pumps remain in-operation in conjunction with no loss of offsite power. Sensitivity studies (Reference 2) show that this assumption results in the worst peak cladding temperature. Minimum safeguards ECCS capability and operability has also been assumed.

No additicnal penalties were required for upper plenum injection _ since this model properly models the location of the RHR flow. This analysis result is below the 22000 F Acceptance Criteria limit established by Appendix K of 10CFR50.46 (Reference 1),

c375NtC$Al*2/s/913 1

REFERENCES

\

..L 1.

" Acceptance Criteria for Emergency-Core Cooling Systems for Light WaterS Cooled Nuclear Power Reactors:

10CFR50.40 and Appendix K of 10CFR50.46,"

Federal Reoister,.Vol. 39, No. 3, January-4, 1974.

2._

Dederer, S.I., et, al., Westinahouse Laroe-Break LOCA Best-Estimate i Methodoloov, Volumes 1 and 2 WCAP-10924-P/A (Proprietary Version),

December, 1988.

3.

NRC Staff Report, " Emergency Core Cooling System Analysis Methods,"

USNRC-SECY-83-472, November,1983.

4.

Bordelon, F. M , and C. T. Murphy, Containment Dressure Analysis Code 100C0), WCAP-8327 (Proprietary Version,) WCAP-8326 (Non-Proprietary Version), June,1974.

5.

Wertinchouse Revised Pad Code Thermal Safety Model_, WCAP-8720, Addendum 2 '

(Proprietary), and WCAP-8785 (Non-Proprietary).

6._ Nissley, M. E.,

et. al'., Eg.stinahouse Larae Break LOCA Best-Estimatt Methodoloav, Vol .1, Add. 4, WCAP-10924-P, (Proprietary Version),

August 1990 0375N CS412/5/91:6

L j.

. c. .

-TABLE 1 e LARGE BREAK TIME SEQUENCE OF EVENTS

~

Four Channel Core ,

0.4 DECLG EVENT (seconds)

Start 0.0 Reactor Trip signal 's .1 Safety injection (S.I.) Signal 2.0 High Head S.I. Begins 7.0 Accumulator Injection 9.0 Blowdown PCT Occurs 10.0

-Low Head S.I. Begins 21.0 End of Bypass 26.2 Hot Rod Burst 27.8 Hot Assembly Average Rod Burst 34.7 Bottom of Core Recovery 35.5 Accumulator Water Empty. 45.7 l Accumulators Nitrogen 71.0 l Injection Ends  ;

l Reflood PCT 0ccurs 83,21 I

1 l

0375NCsAl 2/5/91:5

TABLE 2

. LARGE BREAK RESULTS Four Channel Core 0.4 DECLG EVENT Peak Cladding Temp., OF 2109.

Peak Clad Temp. Location, Ft. 7.375 Local Ir/ Water Reaction (max), % 8.265

- Local Ir/ Water Reaction 7.625 Location ft.

Total Zr/ Water Reaction, % < 0.3 Hot hod Burst Time, Sec. 27.8 Hot Rod Burst Location, Ft. 7.75 Hot Assembly Burst-Time, Sec. 34.7

+ Hot Assembly Burst location, Ft. 7.75

- Hot Assembly % Blockage 27.85 Calculation . Input Values:

NSSS Power, Hwt, 102% of 1650.

Peak Linear Power, kw/ft, 102% of 14.862 Peaking Factor. 2.4 Accumulator Water Volume 1270.

(Cubic Ft. Per Ta-1, Nominal)

Accumulator Pressure, psia- 754.7 s

. Number of Safety injection Pumps 3 (Operating (1 RHR + 2 HHSI])

Steam Generator Tubes Plugged 10%

0375NicSAl 2/5/9116 l

l-F: E3EUGE 1551*

CMANNEL 10.h0DE 7 TCP OF CCCI 25)0.

2250.

20)0.

5 1750.

m O.

~

l5'i.- l J

l

@ 1250. -

m  !

u f i F-

~

10: ).

• j ru.-

l o 3.1 250.

i

),

-25. O. 25, 50, 75. 100. :25. 150. 175. 200.

T ! t1E (SECONOS1 PLOT NO. 36 FIGURE 1. CORE PRESSURE DURING THE TRANSIENT 1

e

m . _ _ __

4 I

LIQU!O. vcFO . ert E ts i c c ! fi E D HA?! rLOW g "- g *- t...v , - C H c tif. E t,

- 12 NODE 7 (Hot AssEng; 3 1 - L ! :; U ! O F 0W. 2-yep;c r;gy, . g . ,7 p a . ,, g g gg,,, _ ,

a :. .

~

t I

w i'-

o W

c CL 1D+

a O

, .- 5' I. Il 4 L.;

t , 3 ,

-! A.

- ~g-v~}y,gv.

A w; 3 C + a.. ~ .

C i l' g/,  ! V I

.) * \f

-5 i I Q

w f

.in.

e l

~15 )

-20

- 5 0 25 50 75 100 125 150 1r5 200 TIME ISECONO3) CASE T i l

l l

l l

l l

l i

, FIGURE 2. VAPOR, LIQUID, AND ENTRAINED MASS FLOW RATE AT l THE TOP 0F THE HOT ASSEMBLY l

4 LICUID LEVEL 0 0 W N C 0 t1 E P LIQUID LEVEL C H anile L 5 2 3.7 6 14 15 22 23 29 30 37.39 43 44 10 Ei ^ - BOTTOM 0F COLD LEG ELEVATION w /

L

/

EJ T r s,x y

'H Wg c 15 +

2 '

i

= l /

i - ,.! /

I l'

t fI i I

5+

J ,

0

-55 0 25 50 75 100 125 153 175 200 i I t1 E ISECONOSI PLOT NO. e 2 FIGURE 3. COLLAPSED LIQUID LEVEL IN DOWNCOMER l

l I

w L!OUID LEVEL CORE L10V!O LEVEL CHANNEL 12. (HOT A S S E t1B L Y 1 12 10 ,

C b

6t a

H.-

C 6 b

3 et ff 2

(7,

'l 0

-25 0 25 50 75 100 125 150 175 200 T ! 11E tSECONDS1 FLOT NO. e7 I

FIGURE 4. COLLAPSED LIQUID LEVEL IN HOT ASSEMBLY  !

l 1

I i

l

s TOTAL FLOW cCCUNULATOP TO INTACT COLD LEO COMPONENT 10. CELL 3

[ 3000.

• N 2500. \

- \

. /

u u

f/J g 20 0 0. +

= >

$ 1500, f i

)

6 l J

1000.  !

X O

~

500.+ '

i 0.

-500.

-25. O. 25, 50. 75, 100. 125. 150.  !?S. 100.

TIME (SECON05) PLOT NO. Jo FIGURE 5. ACCUMULATOR FLOW n

t T r.i i ; ..

r'.--

- w e. j 7-. . ,o;;-; u - ...-.

- ,. _ e.

. . n-L-.,.,

.- < . : 11 t.. 5 I t' 15 0 +

n u

u <

0 100 N

r r.0

.J c0+

.,n / ,

3 '

/ l C

J -l  !

}--

L. 0 I

i a:

I

.O- -E o. + .

l x (

I i

--100

-150

-200

-25 0 25 50 ~5 100 125 150 17; gec e

TIME ISECONDS) ;c?E i 2 FIGURE 6. HIGH HEAD S1 FLOW

I l

TQiAL FLOW R4R FLQW UFcER FLENUM COMPQNENT 14. CELL 1 300

- 15 0 + ,

u r l ./ ,

a E }0 t -

i

.)

n C i J l 6 150 i W -

x  !

-. . 0+ ,

C I i

i t

50 i

l I

100 125 150 175 200 0.. -

0

c

-- CO

- 75

.3 '

PLC' 0' 43 tME iSE QNO3)

FIGURE 7. LOW HEAD RHR FLOW i I

HSP AFPK CASE T E005 1 THEU $

ELEVATION: 18 9. 7 4 [7+ 3 75 NT) 2200.

2000.

~

18 0 0 . + "

u LJ

@ 1600.+

8 -

A - - -

1400. _

b 1200, \/

c b 10 0 0. + -

c r

d 800. h 600. N J 400. -

200.

-25. O. 25. 50. 75. 100. 125. 150. 175. 200.

TIME (SECONOSI go FIGURE 8. CLADDING TEMPERATURE AT PCT ELEVATION

Core floa Difference Observed in WCOBRA/ TRAC Calculations WCOBRA/ TRAC calculations are initiated from a steady-state system condition which the code calculates to establish the vessel and loop flows as well as the_ fluid temperature distribution in the primary coolant system. There are a_ large number of parameters which must be specified to obtain a valid steady state such as reactor power, pump fl ows , steam generator tube plugging levels, fuel temperatures and internal fuel rod gas pressure, just to name a few. These parameters can vary slightly plant-to-plant. The fuel rod- inbrmation 'is obtained from the PAD fuel rod code, while the reactor system hydraulic conditions are obtained from primary side and secondary system calculations which have been benchmarked to plant data as well as scaled hydraulic tests, WCOBRA/ TRAC can yield slightly different values for the different parameters due to the different computational techniques and methods of modeling the reactor system between the reference plant calculations and WCOBRA/ TRAC modeling, it has been recognized that such differences can exist, hence acceptance ' criteria were established in WCAP-10924 to minimize these efforts such that consistent results would be obtained. The plant parameters which had to be matched were divided into two groups. The relative importance for large break LOCA of the parameters in each group were discussed in Chapter 3 of Volume 2 of WCAP-10924. The first level variables contain the reactor heat source parameters such as power, fuel temperature, coolant temperatures, peak kw/ft, and reactor pump delta-P and were held to a very tight tolerance. All power parameters were specifically made to exactly match the desired value or were made conservative. This criterion is given in Table 2-1 'on page 2-32 of WCAP-10924 Volume 2, Revision 2. Most of the parameters had to be matched within +1%, -0%.

.:LEH1/2/5/91 1

7 t

c,- The secondary- parameters included the reactor pressure drop values, core; flow, .and .the--ratios 'of. pressure drops to ensure the proper hydraulic resistance distribution between the loops, and the reactor vessel. The

-  : pressure drop Linformation= generally has .a higher uncertainty.and is more -

difficult -to match because of the uncertainty in.the reference calculations.

as well- as the WCOBRA/ TRAC modelling. It should also be noted that the accuracy. of the secondary parameters is sacrificed to obtain a more accurate ,

fitL :to the first level parameter such as fluid temperatures. However, even the secondary 11evel variables' are required to be within 5% as shown,i_n  !

LTable r 2-1 on page :2-32 of WCAP-10924, Volume 2, for a valid steady state calculation.

.The difference between'the original WCAP-10924,' Volume 2, Revision " Prairie Island calculation and' the revised calculation using the Addendum 4-to ,

WCAP-10924 is two-fold. First there' is some _possible difference due' to the-

' decay power effects of_ correcting the decay-heat error which is the basis for. Addendum 4. ItLis expected that the decay _ heat effect is smallisince the timeMin question ist very early in.the transient and the-integrated

'effect of the' decay heat curve = difference would be very small.

The 'second change between the two calculations is the method used to match-

~

the ' reference reactor system pressure -drop and flow information during steady -state._ The revised calculation directly calculated the form loss pressure'. drop from the WCOBRA/ TRAC ~ code output'_to compare with the estimated j

- unrecoverable - -- pres sure losses _given irt theJreference reactor coolant . system.

s ccalculations. This 'is a more accurate method since the unrecoverable pressure losses _ are- directly compared. -TheLoriginal calculation compared

~

component -pressure drops, not specific: form losses, and1the form losses were then adjusted to match the calculated pressure . loss.

i e

J:LEH1/2/$/91 - 2

. ~ .., . . . - - . -- . . . . .. _ . -. _ ~ , _ _ - - _ _ 'l

l

, in the revised method, the velocity head effects are accounted for in a more accurate and systematic fashion, Both techniques yielded secondary level variables- that were within the 5% guidelines given in WCAP-10924, Volume 2, Revision 2, which had been reviewed and discussed with the NRC. The primary difference between the- two calculations is the ratis of_the lower plenum pressure drop to the vessel pressure drop, This ratio is slightly less

('1%) for the revised calculations which will result in slightly more downflow through the reactor vessel during blowdown, While this preshre drop adjustment, which is well within the allowable variation, may result in momentarily more core downflow, other such adjustments could have easily resulted in slightly reduced core flow. Since the reactor system pressure drops are difficult to exactly predict, the allowable tolerance of 25% is reasonable and should result in only small changes run-to-run with a minimal effect on the final peak cladding temperature, l J:LEH1/2/s/91

• 3

{

l