ML20024D337

Summary of ACRS Subcommittee on ECCS 830217-18 Meetings W/ GE in San Jose,Ca Re Review of New GE Safer/Gestr ECCS Evaluation Model Licensing Code.Supporting Documentation Encl

ML20024D337
Person / Time
Issue date:	02/22/1983
From:	Advisory Committee on Reactor Safeguards
To:	Advisory Committee on Reactor Safeguards
References
ACRS-2106, NUDOCS 8308040553
Download: ML20024D337 (19)

Text

r fl G P3 5 A M M D:

2/22/83 U

f3 a

MhAUl 3J ihit M

3 ACRS ECCS SUBCOMMITTEE MEETING SAN JOSE, CA FEBRUARY 17-18, 1983 PURPOSE The purpose of the meeting was to continue review of the new GE SAFER /GESTER ECCS Evaluation ibdel licensing code. The meeting was closed to the public to protect GE proprietary information.

ATTENDEES:

Principal meeting attendees included:

GE, ACRS G. Sherwood D. Ward, Chairman D. Der.nison J. Ebersole, Member M. Almagir

1. Catton, Consultant H. Pfefferlin M. Plesset, Consultant S. Congdon V. Schrock, Consultant T. Lee C. Tien, Consultant J. Findley Z. Zudans, Consultant Y. Anderson P. Boehnert, Designated Federal Employee D. Humon MEETING HIGHLIGHTS, AGREEMENTS, AND REQUESTS Dr. H. Pfefferlin opened the meeting by stating the purpose, i.e., to 1.

respond to Subcommittee questions related to the GE realistic ECCS He reviewed the highlights of the SAFER /GESTR code review approach.

(Figure 1), noting that NRC approval of these codes is expected in March 1983.

GE's approach is two pronged:

(1) decay heat exemption request to us.e the 1979 ANS Standard (worth 2-400*F PCT), and (3) the SAFER /GESTER Evaluation Model (worth 700-1000'F PCT). The NRC ha completed technical review of the decay heat exemption and use of 8308040553 830222 DESIGNATED OiUGIIML PDR ACRS Certitled Dy-Q p~

PDR s

, A.

ECCS Meeting Fcbruary 17-18, 1983 the 1979 Standard will be reviewed on a individual licensee basis.

l To date, no licensee hs formally requested use of the new Standard.

Additional discussion on this point indicated that GE believes that increased fuel utilization will be the chief benefit of the use of SAFER / GESTER and the new decay heat Standard.

2.

1 3.

Mr. T. Lee detailed the above sensitivity studies. Figure 5 shows a comparison of the 1979 curve over a typical BWR operating range and the Appendix K curve.

Dr. Schrock requested the details of the vari-able power histories in order to determine the validity of the 1979 curve sensitivity. He expressed concern that using a generic curve may obscure critical details of the calculations, since each " curve" is in effect a point calculation. Dr. Schrock urged GE to provide the NRC Staff the details of these calculations for their review.

Further, Dr. Schrock suggested GE examine the ratio of peak to average power density in a shutdown reactor vs the same ratio for an operating reactor since they can be different.

Sensitivity studies show that the limiting bundle exposure of interest for ECCS analysis is 10-15 GWO/T (hot bundle).

The most sensitive operating parameter between the ECCS limiting hundle and the remainder of the core is the exposure time. The sensitivity of the limiting bundle is very small (3.7%) compared to the rest of the core. GE concludes that e e=

m

. February 17-18, 1983 ECCS Meeting Figure 7 a single reference curve (Figure 6) is an acceptable procedure.

Dr. Catton is the GE reference curve for applying the 1979 Standard.

asked why GE used an enveloping curve.

GE used an average value since they are doing nominal calculations. There will be an Adder put on this curve to account for uncertainties which will be discussed below.

4.

l i

I

ECCS Meeting February 17-18, 1983 5.

The details of the GE TRAC B02 development effort was discussed by Mr. J. Andersen.

This is a cooperative effort between GE and EG8G. NRC, GE and EPRI are joint sponsors. Figures 14 and 15 list the improvements incorporated by GE into TRAC BD1 Version 12 which is the latest official EGSG released version.

Mr. Anderson detailed the major GE models/ improvements made to the code..These include:

(1) consistent use of flow regime map, (2) interfacial shear, (3) CCFL, (4) heat transfer improvements, (5) jet pump model, (6) separator and dryer models, (7) upper plenum model, (8) two-phase level model, and (9) improved numerics.

Mr. Anderson

.o.

ECCS Meeting February 17-18, 1983 proceeded to detail the above models/ improvements. The Subcommittee expressed general satisfaction with the TRAC work discussed above.

6.

Mr. N. Almagir reviewed the TRAC B02 qualification effort. TRAC B02 is being used to benchmark the SAFER /GESTER EM code.

Mr. Almagir presented an update of further TRAC calculations of separate effects tests that were discussed at the December 1-2, 1982 ECCS Subcom-mittee meeting.

These calculations included: a TLTA (two loop test apparatus) peak power low flow (60% of normal) ECCS test, a SSTF system response test simulating a BWR/6 DBA, and the Peac'h Bottom turbine trip test. Specific highlights of the presentation include:

• Drs. Catton and Plesset urged GE to participate in the international Standard Problem Program, (ISPP) noting this program is an excellant training aid for the code users as well as a good test of GE's code skills.

GE indicated that they feel the ISPP is not as productive as it should be, but they are following the sprit of the ISPP in testing their code against a broad spectrum of transients and facilities.

• There was extensive discussion of the TRAC code's ability to calculate the details of the TLTA thermal hydraulics.

It was agreed that while accuracy was lacking, the code does well on a global basis.

• The determination of the TRAC nodilization scheme was described.

By use of a closed loop starting with model comparisons against data, through nodilization sensitivity studies, and finally, feeding back to the code models - and considering nodilization experience - GE has, they believe, incorporated the optimum nodilization scheme for their uses.

There was extensive discussion of the use of just two sets of nodes in the sensitivity studies. Drs. Catton and Tien felt that three sets of nodes was the minimum nunber needed for these sensitivity studies. GE said they regard the sensitivity studies as confirming i

their understanding of their modeling of the phenomena, i.e., the e

e. -

ECCS Mesting Februa ry 17-18, 1983 second nodilization run is regarded as confirming the choice of the first number of nodes used.

It was suggested that NRC and/or INEL should fund a more extensive nodilization sensitivity study. Figure 16 shows GE's final choice for the BWR/6 nodilization scheme for LOCA calculations.

7.

4 8.

fiessrs D. Hamon and B. Shiralkar provided an update of the SAFER LOCA calculations for typical BWR/4 and BWR/6 plants.

Figure 18 lists the analysis assumptions. Figures 19-21 summarize the results of LB and SB calculations.

PCT's shown were below 1000*F. Given these PCT values, tir. Ebersole asked if GE is consideirng requesting relief on the crash-start requirement for the plant diesels. GE said this matter is under consideration.

The BWR/6 results are shown in Figures 22 and 23. All the BWR/6 PCTs shown were also under 1000*F.

The SAFER /GESTER application methodology (Adder) was detailed by Mr. Shiralkar.

The methodology has three main parts.

These are:

6

l ECCS Meeting February 17-18, 1983

• LOCA events are analyzed with nominal input values in SAFER /GESTER.

• Nominal PCT is increased by an " Adder" to obtain an upper bound PCT for design evaluation.

• The " Adder" encompasses Appendix K specified values as well as other uncertainties combined in a statistical manner.

Mr. Shiralkar noted that, at present, the NRC Staff has indicated there may be a legal problem with the statistical combination of the Appendix K specified parameters.

The Adder method consists of two parts:

• Calibrate SAFER modeling uncertainties vs. "BWR-LOCA Experiments"

("BWR-LOCA experiment" = TRAC prediction for BWR, corrected to account for TRAC bias and uncertainty).

• Quantify effects on PCT due to plant parameter uncertainties by performing sensitivity studies with SAFER.

The adder is based on: test data, TRAC predictions of the test data, TRAC plant benchnark calcultions and SAFER /GESTER plant calculations.

GE noted that the experimental results are the " anchor" for the adder.

Figures 24 and 26 detail the actual adder calculations. Dr. Schrock questioned the GE assumption that there are no scaling concerns relative to the LOCA experiments that needs.to be accounted for in the bias cal-culations. GE agreed scaling concerns are always there but they feel sufficiently confident that, based on SSTF data for example, they have sufficient basic knowledge of the phenomena. GE also noted that the FIST (Full Integral System Test) facility will be used to help confirm the scale-up of TLTA data. Dr. Plesset and Mr. Rard suggested that GE I

i e

ECCS Meeting February 17-18, 1983 perform an adder calculation independent of of the Appendix K require-ments to determine the true uncertainty in the experiments and calcula-tions.

It was also noted that the SRSS method of combination is not strictly technically correct since all the parameters being combined are not independent of each other. GE agreed, but said that if a Monte Carlo analysis was done, for example, the difference would not be great.

The basic adder elements were discussed and their methodology detailed.

These elements and associated objectives (given in parenthesis) are:

(1) TRAC-B02 calibration vs experiments, (assess TRAC B02 bias 3nd uncertainty),

(2) SAFER /GESTR comparisons vs TRAC B02 for BWR Transients (calibrate SAFER bias and uncertainty for BWR calcualtions), and (3) SAFER /GESTR sensitivity studies to quantify plant uncertainties (quantify PCT changes due to plant parameters uncertainties and Appendix K requirements).

The results of the adder calculations are given in Figure 27.

For the BWR/4 and BWR/6 plants, the total calculated PCTs are 1234'F and 1045*F, respectively.

9.

Mr. Ward indicatd that NRC should complete their review by late March and following another Subcommittee meeting, full Committee review in May appears appropriate.

The Chairman also requested written comments from the Consultants on the meeting discussion.

10.

The meeting was adjourned at 4:10 p.m., February 18, 1983.

NOTE:

No Transcript Available - Closed Meeting.

~

GE ECCS APPROACH STATUS UPDATE o

PRESENTED DVERALL ECCS APPROACH TO ACRS-ECCS SUBC0>1MITTEE AllG.1981 e

SilBMITTED SAFER /GESTR REALISTIC MODEL TO DEC. 1981 NRC FOR REVIEU e

SUBf1ITTED GESSAR II DECAY HEAT EXEf1PTION DEC. 1981 TECHflICAL BASIS T0 f1RC FOR REVIEl!

e 11ET WITH NRC Of1 SAFER /GESTR APPLICATION JAtl. 1982 APPROACH e

PRESEf!TED DECAY HEAT EXEf1PTI0fl TECHf!ICAL JllflE 1982 DETAILS TO ACRS-ECCS SUBCOMMITTEE e

PRESEllTED SAFER TLTA QUALIFICATI0fl RESULTS AllG.1982 AfID APPLICAT10ft PLAtlS Tn NRC e

PRESErlTED GE LOCA MODEL TECHfilCAL DETAILS DEC. 1982 TO ACRS-ECCS SUBCOMMITTEE e

PRESErlTED Fif8AL SAFER RESULTS TO flRC ON JAf'. 19P3 OIJAl.IFICATION, APPLICAT10fl AND TYPICAL BREAK SPECTRUf1S e

RESP 0flD TO ACRS Ol1ESTIONS FEB. 1983 l

SAFER /GESTR APPROVAL EXPECTED MARCH 1983 DKD: pes /103A-4 2/16/33 f/-)al

c..

COMPARISON OF APPENDIX K

~

AND 1979 ANS STANDARD F

APPENDIX K ANSI /ANS 5.1(1979)

DIFFERENCE-GE APPLICATION (1979GE-APPK)/1979GE All fissions in U235 fissions in U235,U238,Pu239

~200 MeV/ fission calculated MeV/ fission

-5% to -10%

infinite operating time actual operating history actinide decay-actinide decay-

<1%

i U239,Np239 chain U239,Np239 plus others activation of structure

<1%

and poisons correction for neutron

<1%

capture Legislated adder

-26% to -31%

i 4

TOTAL

-28% to -38%

l @S The 1979 standard includes an evaluation of data uncertainties (7%-12% 2cr). To be included in overall oeck clad temoerature adder.

GENERAL ELECTRIC DECAY HEAT BASIS UTILIZING ANSI /ANS 5.1 1979 2-P(t)=

D n (t)+D, (t,T)+A(t,T)+A s (t,T)

Q(T) t

=

Time ofter shutdown T

=lrradiation time Q(T)= Available energy / fission D y (t)= Fission heat produced by delayed neutrons l

D, (t,T)= Fission product decay heat A(t,T)= Heat produced by actinide decay A s(t,T)= Heat produced by decay of activated structure and poison materials

p-F SS ON 3 RO JUC-~

J ECAY TERM DEPEND ON REACTOR POWER AND POSITION IN CORE i

t

' 3-D SIMU'_A.TT h l ABSORPTION OPERATING l[

COLE CORRECTION HISTORY DECAY HEAT I

3 p

sf CURVES P(t') T r Q(t) )S r,(t') f,(t+T-t') dt' cpu v

D

= G(t,T)

) k Q(t')

f

^

O i

MEV/ FISSION REl.ATIVE NUMBER B'JNDLE PHYE!1i OF FISSIONS IN s

CODE l

U235,U238,PU239 V

DEPEND ON FUEL ENRICHMENT AND STEAM VOID FRACTION l

\\

IN PRINCIPLE, EACH POINT IN THE REACTOR HAS A UNIQUE DECAY HEAT CURVE lF/6$

CCFL CORRELATION CONSIDERATIONS L. General Momentum Flux Forri;:

2 5/

.,, fg t i /y

_ft l1 4+*.. I 3I.

1 h ic..

Critical Velocity Relations:

1.

b (Wallis)

V gc

. Pg[f (Kutateladze)

F V

=K gt Kutateladze Implies Characteristic Dimension L = F L y v\\'#2' KI#

Characteristic Dimension is Size Dependent L -+ Diameter (for small dimensions)

L

--. Kutateladze (for large dimensions) l l

l<Mb %,- _ _.

P l

l

?

SIDE ENTRY ORIFICE AND FUEL SUPPORT E.

C O

s sesma8

+

3,.,s

,o

\\.s,

/..e Y

h[ h g'

I u'\\

[J:- &.,

-==

_.+e w '

pu W "'

<s j ir

~

,s.

\\

x N

.m

/

k 9

\\ /l 3

e

' \\,

\\

\\\\

gxsts

[F/6 ti j eg2 :r...

,.: =.,z7 ::w

. _.c ;

,.m, I

O cu nnsL f

y;

{

e

=. r FutLSUPPCM CCRE SUM i

~

/

,,,,&Tutt a

guto sxmouO FUEL SUPPORT ASSEMBLY l

{

,y TRAC B02

SUMMARY

OF GE UPDATES IMPLEMENTED INTO TRACBD1 VERS TRACBD1 Version 12 (TRACIB002) generated 5/20.82.

GE models implemented into TRAC 1B002 (14081 lines of update) 1.

Two-Phase level tracking model 2.

Steam separator model 3.

Improvements to interfocial shear Improvements to interfocial wall and heat transfer

()

4.

including consistent use of flow regime map 5.

Improved prediction of reversible pressure losses 6.

Manentum source for vessel

• 7.

Virtual mass for vessel

• 8.

Forward and reverse flow local losses

• including liculd and vapor phase losses of ID c0Bponents 9.

Improved thermal radiation' 10.

Improved jet pump model Correction to choked flow model and vessel donor 11.

O celling for small velocities 1

e s'

Sumory of GE Updates Implemented into TRACBD1 Version 12 Icontinued) 12.

Steam dryer model 13.

Upp2r plenum model 14.

Se9mented loading on CDC 7600 15.

Predictor-corrector numerical method.

O

$

• Developed previously for TRACBD1 Version 3.

O

{FM S

~~

TRAC BWRh NobALInATtoA (wcA APPLICATiede) g x

I I

I i

i l

1 l

t I

g 8

i I

l I

, _a _. r a _ L

.t., _I i

MAIN I

I I

s u aan 4

Lials.

a i

7 l

\\

~

4 t.

.l._

4 i

FEE WATER g

l 4

l s

tecs c

I I

HPCS

~ ~

y- - - "

3 ;

=.n..n..n.,n..n

~

3 l

l I

l LPer h

E z

. _ l _. L_ l _, ', __l.

i i

i i

I I

l--.

. ll f

I I

, '_n. _n ' YL,

I T

g=4v l *a l REclRcutATimi j

l t,,j g E- -l-g - h --

t-l u

a e

i I

I.i I

I l

l I

s

-p.;, ; p.. j I.

_l..i I Uli i

I i

i l

t I

e

'l t

~~

9*

~

N N

\\

~

7 f.__

N T\\

I

,0 010):

\\. \\

\\

'n /

/,

x x

/

\\

N

/

~

.i

[

l

,$s m

fPJ6 lA j

. _ m

= _ -.

~ - - _

t n.

IMPLICATIONS OF' SAFER C RE REPRESENTATION 7

EARLY TRANSIENT - No compromise

=

Similar core-wide behavior MIDDLE TRANSIENT.-Average and high power bundles show right regimes

-High power bundle cooled by dispersed droplet flow

-PCT occurs, before lower plenum filling LATE TRANSIENT

- Upper plenum mass buildup and subcooling

- Periodic CCFL breakdown at top and bottom of average bundles

- Oscillatory refilling of lower.

plenum and core e

.