ML20195F704

Safety Evaluation Accepting Topical Rept NEDC-24154P, Supplement 1,for Referencing in Licensing Applications to Extent Specified & Under Limitations Delineated in Rept

ML20195F704
Person / Time
Issue date:	11/17/1998
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20195F688	List: ML20195F704
References
NUDOCS 9811200062
Download: ML20195F704 (10)

Text

.

~. - _... -. - _ _ ~

.... -.. ~. -.

>S Ricoq l

p

\\

UNITED STATES s

j NUCLEAR REGULATORY COMMISSION l

WASHINGTON, D.C. 20656-0001

\\...../

SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION OF NEDC-24154P. SUPPLEMENT 1. "OUAllFICATION OF THE ONE DIMENSIONAL l

CORE TRANSIENT MODEL FOR BOILING WATER REACTORS."

1 INTRODUCTION The staff has reviewed the request by General Electric Nuclear Energy (GENE) to extend the range of applicability for the previously approved ODYN code (ref.1). The staff also considered a revision submitted on April 26,1998 (ref. 2). This proposal would allow ODYN to be used for l

Anticipated Transients Without Scram (ATWS) and non-pressurization Anticipated Operation Occurrence (AOO) evaluations. The basic model is unchanged from the original code which the staff reviewed and found acceptable in 1981 (ref. 3). This review, therefore, will focus on the consideration of the new features of the code which are summarized below:

a. a soluble boron transport model;

b. a model of the r3 activity effect of soluble boron; and l

c. a model to account for changes in the weighting function used to collapse 1D kinetics j

parameters during long transients.

2 DISCUSSION OF MODEL CHANGES This section will serve to describe the new features of ODYN.

2.1 Boron Transoort Model The boron transport model consists of the addition of one continuity equation for the mass of the boron. This equation is formulated with concentration as the dependent variable and is solved by first-order, upwind differencing. The assumptions inherent in this model are:

1

a. the boron moves in solution with the bulk fluid and at its velocity;

b. the boron does not interfere with the material properties of the bulk fluid; and

c. the boron only transports with the liquid phase.

The model accounts for boron stratification by using user input mixing efficiencies for different regions in the pressure vessel.

2.2 Boron Reactivity Model The modelis formulated to relate the boron concentration to its effect on the multiplication l

factor, input coefficients are derived from lattice physics calculations as a function of density for both controlled and uncontrolled nodes, including the effect of boron in the bypass region.

9811200062 981117 PDR TOPRP ENVGENE l

C PM ENCLOSURE l

r

2 The effective boron density is calculated by homogenizing the bypass and in-channel boron i

over the control cell. The volume in the peripheral bypass is used to calculate the bypass concentration but it is not used in the averaging process. The reactivity effect of boron is formulated as a quadratic function.

2.3 Optional Nuclear Cross Section Fits A modelis used in ODYN to relate two different sets of cross section inputs by linear interpolation. Typically, the first set will be derived from rated conditions and the second from an assumed new quasi-steady, natural circulation condition. This model allows for the kinetics input to be representative of the actual reactor state being considered. The argument proposed by GE is that for long transients such as ATWS, the reactor will transition from its original condition to a new, quasi-steady natural circulation condition and that this transition will effect the values of the collapsed 1D parameters.

3 EVALUATION OF MODEL CHANGES The following section contains the results of the staff review of the GE proposal.

3.1 Boron Transoort Madd The staff review consisted of an examination of the comparison presented by GE of ODYN calculations to test data obtained in the GE Vallecitos 1/6 Scale boron mixing tests. This validation confirmed that the numerical modelin ODYN is appropriate because it used mixing efficiency data derived from the Vallecitos facility as input to calculations which were used to derive mixing efficiencies for comparison to the Vallecitos data. The comparisons, documented in the report, show that the modelis performing satisfactorily. This calculation serves to validate the numerical model because the use of the Vallecitos data in the input allows for the calculated concentrations to be essentially de-coupled from the mixing phenomena.

3.2 Boron Reactivity Model The staff review consisted of examining the model features presented in the document and performing some audit calculations (refer the Appendix A) to confirm that the model conservatively predicts boron reactivity. The model accounts for boron reactivity from boron both in the channel and in the bypass region by homogenizing the boron concentration over the total free volume of the control cell (a control cell consists of the channel and its associated bypass area) and then using this concentration to calculate the boron worth. The staff questioned this model because the averaging process does not account for the differences in the neutron slowing process in the bypass and channel regions and this could introduce a bias into the results. For example, although the bypass accounts for roughly 30% of the volume of a control cell, it has a largo effect on the thermalization of neutrons in the system. This fact is confirmed by the ctaff's audit calculations described in Appendix A. The possibility of a bias comes from the fact that the averaging process used in ODYN could increase the average boron in the bypass and artificially increase the boron reactivity effect. However, as long as the boron mass in the bypass and the channel are equivalent, the model will conservatively account

3 for boron reactivity because it will tend to reduce the effective bypass boron concentration.

GENE provided information to the staff in the form of a table of boron mass in the bypass and i

channel for the cases documented in the report (ref.1) and this table showed that, for these cases, the boron mass in the bypass was either equivalent or greater than the in-channel mass.

Therefore, the staff concludes that the model as proposed should lead to a conservative prediction of boron reactiivity.

3.3 Dotional Nuclear Cross Sections The staff review consisted of examining 2D radial power distributions from previous analyses and questioning GENE about the effect of the model. The previous analyses were performed by the staff during the review of NEDC-32523P and are briefly discussed in Appendix B to this report. The results indicated that the radial power distribution and, therefore, the radial weighting function, are not very sensitive to the current reactor state. This observation concurs with GENE results.

4 COMPARISON OF ODYN TO PLANT TRANSIENTS l

GENE included in the report (ref.1) several integral benchmark cases using ODYN.

l Benchmarks of ODYN to the following data are presented:

a, level data from a loss of all feedwater test in a BWR/6

b. flow, power and pump speed data from a flow demand step change test in a 251 in i.d. BWR/5

c. flow, power and pump speed data from a pump upshift test in a 251 in. i.d. BWR/5

d. core flow and water level data from a pump trip in an MG-set BWR/4 These benchmarks demonstrate the code's ability to successfully predict plant results for low level or low flow conditions necessary for the proposed extension of ODYN application which were not considered as part of the original review (Ref. 3). For example, ATWS simulations require a code which can predict the collapsed level in the downcomer over a wide range and the benchmark presented in the report (ref.1) demonstrates that ODYN can accurate;y predict the water level as a function of time during an actual BWR/6 transient.

4.1 Acolication Scoce of ODYN GENE proposes to expand the scope of ODYN applications to include ATWS and non-pressurization transients. The LTR under review contains ODYN results to validate its application to ATWS and a discussion and example evaluation of the proposed method for calculating non-pressurization transient ACPR. GENE proposes to follow a conservative method for non-pressurization transient ACPR calculations discussed in section 5 of the report l

(ref.1). Prior to the current modifications ODYN was fully capable of predicting non-

[

pressurization transients and with the current modifications it is also capable of predicting ATWS conditions. During the course of generation of the LTR and the staff review, GENE identified the following restrictions on ODYN ATWS applications:

l l

(

i 4

a. The downcomer level must remain above the jet pump suction and no prolonged level in the active channel is allowed:

b. The duration of the simulation after the upper plenum subcools should be limited.

c. The mass in the separators should not remain zero and, therefore, the code is restricted to applications where the water level remains at or above the top of active l

fuel plus 5 feet.

1

d. The code is not presently qualified to perform stability calculations.

e. No lower plenum voiding is allowed.

5 CONCLUSIONS The staff has reviewed the material presented in references 1 and 2 and concluded that ODYN is acceptable for applications to both ATWS and non-pressurization transients provided that it is applied consistent with the restrictions contained in section 4 cf this safety evaluation.

6 REFERENCES

1. Letter from W. Marquino (GENE) to T.J. Kim (USNRC) submitting NEDC-24154P, Supplement 1 " Qualification of the One-Dimensional Core Transient Model for Boiling Water Reactors," January 13,1998.

2. Letter from W. Marquino (GENE) to T.J. Kim (USNRC) submitting a revision to NEDC-24154, Supplement 1, April 26,1998.

3. Letter from R. Tedesco (USNRC) to D. Sherwood (GENE), " Acceptance for Referencing General Electric Licensing Topical Report NEDO-24154," February 4,1981.

4. M.C. Brady, M. Takano, M.D. DeHart, H. Okuno, A. Nouri and E. Sartori, " Findings of the OECD/NEA Study on Burnup Credit."

5. Memo from T. Ulses (USNRC) to Distribution," Validation of Lattice Physics Methods,"

September 1998.

6. " Gamma Scan Measurements at Quad Cities Nuclear Station Unit 1 Following Cycle 2,"

NP-214, Electric Power Research Institute, Palo Atto, CA, July,1976.

Appendix A - Lattice Physics Calculations

5 A.1 Summary of Results The staff evaluated the effect of boron homogenization on a GE9B fuel assembly in a C-lattice.

The assembly chosen was derived from the OECD burnup credit code validation program and is described in Ref. 4. This analysis determined the effect on K., from homogenizing an equivalent mass of boron in the bypass, the total control cell flow area and the active channel, respectively. The analysis shows that when the boron is averaged over the control cell that it is worth less than ifit is only homogenized into the bypass. Conversely, the staff also showed that homogenizing the boron from the active channel over the control cell increased its worth.

Table A.1 summarizes the staff's results. Since the homogenization process used in this analysis essentially mimics the one proposed by GE for use in ODYN, the staff concluaes that this method could introduce a bias into the predicted boron reactivity.

Boron Location Different Void Fractions - K.

0%

40%

70 %

Bypass 0.9377011 0.9060468 0.8891168 Entire Domain 1.0027180 0.0805285 0.9659925 Active Channel 1.0358780 1.0160160 1.0017660 l

l l

A.2 Description of Methods J

l The staff used the New ESC-Based Weighting Transport (NEWT) code to calculate the infinite i

multiplication factor for a GE9B assembly described in Ref. 4. NEWT is a new method being i

developed by ORNL for use with the SCALE system. NEWT validation is described in Ref. 5 and it comprises comparisons to KENO-VI calculations, XSDRN calculations and the Quad l

Cities gamma scan data (Ref. 6). NEWT uses a characteristic method for spatial discretization

(

and, as such, it requires no homogenization which leads to very accurate results. All calculations were performed with an S quadrature and a P scattering expansion using a 44 4

3 group ENDF/B-V derived library. An example input deck is included.

=shell l

In -sf SRTNDIR/ newt $TMPDIR/ newt end

=csasi OECD Phase IllB BWR Benchmark,40% void fraction, O B/U,100 gm b-10 in bypass 44groupndf5 latticecell

' Fuel group no.1 u-234 101.0443E-5 900 end u-235 101.1284E-3 900 end u-236 10 6.9317E-6 900 end t

u-238 10 2.1606E-2 900 end l

l-

LO-300L99'C=(C)uep 639'O Jepu!IXo C=seJ LO-300L99'C=(g)uep 629'O Japu!IXo g=se; elep ajouJ pua 9 Ogg L 9 L 990'L E9 L yoll eJenbs i

d dwoo pua I

pua 699 E-St629'L 0 9 o pue 699 E-39990'C 0 9 4 JoieJepoW, pus 699 9-300bL'E 01 OL-4 pua 699 66'O4 o24

ssedXg, pua 699 E-32963'# 0 9 JZ pua 699 #-3909P'L 0 9 e; pua 699 9-3L699'l 0 9 Jo Su!ppelo, pua 006 #-30020'C 0 9 09L-p6 pua 006 P-306CL'C 0 9 99L-p6 pua 006 P-39490'E 0 9 29L-p6 pua 006 >-39210 0 0 9 99L-p6 pua 006 P-3LVL2'2 0 9 99L-p6 pua 006 9-30932 C 0 9 #9L-p6 pua 006 E-39609't 0 9 o pua 006 E-36940'E 0 9 BCg-n pua 006 9-3920L'# 0 9 903-n pua 006 #-31909'9 0 9 90g-n pua 006 9-3#E99'9 0 9 #Cg-n (po; pe) 9 ou dnoJ6 lend, pua 006 2-3 L6P9'r 0 V o pua 006 E-39023'E O # BCg-n pua 006 9-3090P'E O # 9Cg-n pua 006 P-39962'9 0 # GCg-n pua 006 9-39002*r 0 # PCg-n p ou dno;6 ang, l

pua 006 2-3r6t9'# 0 C o pua 006 E-39#03'E O C BCg-n pua 006 9-3049C'# 0 C 902-n pua 006 #-34906'9 0 C 90g-n pua 006 9-399#2'9 0 C #Cg-n C ou dnoJ6 eng, t

pua 006 E-BZ6#9't 0 2 o pus 006 2-3106L'E 0 2 BCg-n pua 000 9-3Lotp9 0 g 90g-n pua 006 P-Br063'8 0 3 9Cg n pua 006 9-30829'l 0 E #Cg-n j

g 'ou dnoJ6 and.

l pua 006 g-3y0gg y 0 L o pua 006 2-3909VZ 0 L SCg-n 9

1 7

res=4 cylinder 0.529 dan (4)=3.85100E-01 res=5 cylinder 0.529 dan (5)=3.85100E-01 end more end

= newt OECD BWR Test problem 100 gm b-10 in bypass read parm l

fillmat=7 collapse =yes prtbroad=no kguess=1.08 prtflux=yes prtburn=no prtmxtab=no end parm read esc accel =yes sn=4

-epsinner=1.e-4 epsouter=1.e-4 epseigen=1.e-4

- end esc read materials 10 0.0 " Fuel type 1" end 2 0 0.0 " Fuel type 2" end 3 0 0.0 " Fuel type 3" end 4 0 0.0 " Fuel type 4" end 5 0 0.0 " Fuel type 5, with Gd203" end 6 0 0.0 " Cladding" end

. 7 3 0.0 " Bypass" end 8 3 0.0 " Moderator" end end materials read collapse 9r1 17r2 9r3 9r4 end collapse read geom xgrid 0.0 7.621 128 ygrid 0.0 7.621 128 repeat 1.63 1.915 1.915 5 13 23 22-32 31 mulcylin 1 1 0.529 " Fuel material for fuel rod 1" end cylinder 6 0.615 " Clad material for fuel rod 1" end end repeat 1.63 1.915 1.915 5 03 02-11.

20 30 mulcylin 1; 2 0.529 " Fuel material for fuel rod 2" end cylinder '. 6 0.615 " Clad material for fuel rod 2" end 2

I i

l L

8 end repeat 1.63 1.915 1.915 2 01 l

'10

. mulcylin 1 3 0.529 " Fuel material for fuel rad 3" end cylinder 6 0.615 " Clad material for fuel rod 3" end end i'

repeat 1.63 1.915 1.915 1 00 i

mulcylin 1_4 0.529 " Fuel material for fuel rod 4" end cylinder 6 0.615 " Clad material for fuel rod 4" end end repeat 1.63 1.915 1.915 2 y

12 21 mulcylin 6. 5 0.529 " Fuel material for fuel rod 5" end cylinder 6 ' O.615 " Clad material for fuel rod 5" end end l-cylinder 7 7.620 7.6201.500 " Central water hole" end cylinder 6 7.620 7.6201.600 " Central guide tube" end cuboid 8 1.100 1.10014.140 14.140 " water inside channel can" end cuboid 6 0.840 0.84014.400 14.400 "zircaloy channel can" end end geom end l

i L

t I

i I -,,

l l

9 l

l Appendix B - Reactor Power Distributions The staff performed some 3D TRAC-8 (a combination of TRAC-BF1 and NESTLE 5.02) l simulations during the review of NEDC-32523P to examine the effect of BWR power uprate en l

the peak vessel pressure. These results also provide us with a capability to examine the effect l

changes in core power and flow have on the radial power distribution. The es ent studied was a MSIV closure ATWS, The model consisted of 101 individual CHAN components and 191 radial i

neutronic nodes providing 1/8 hydraulic and 1/4 neutronics symmetry, respectively. The results presented in this appendix as Figure B.1 consist of radial power distributions at time zero and 25 seconds following the valve closure. One will note that there is very little change in the two distributions.

T1rne= 50.0000 Edit b 50 Total Power (MWt) 958.61 Maxmimum Powered node at (13, 7) with value of 1.37 0

1 2

3 4

5 6

7 8

9 10 11 12 13 14 15 l'O.396 0.531 0.653 0.436 0.449 0.610 0.801 0.652 0.718 0.921 1.239 1.251 1.278 1.328 0.976 2 0.531 0.684 0.578 0.669 0.638 0.772 0.655 0.655 0.683 1.059 1.138 1.340 1.362 1.347 0.971 3 0.653 0.578 0.696 0.582 0.740 0.733 0.809 0.601 0.782 0.926 1.282 1.342 1.364 1.347 0.966 4 0.436 0.669 0.582 0.471 0.500 0.015 0.812 0.748 0.803 1.148 1.150 1.266 1.282 1.327 0.974 5 0.449 0.638 0.740 0.500 0.543 0.793 0.996 0.834 0.899 1.097 1.299 1.268 1.281 1.323 0.971 6 0.610 0.772 0.733 0.815 0.793 0.976 0.893 1.092 1.057 1.238 1.135 1.359 1.361 1.336 0.972 7 0.001 0.655 0.809 0.812 0.996 0.893 1.089 0.985 1.196 1.123 1.296 1.366 1.373 1.151 0.978 8 0.652 0.655 0.601 0.748 0.834 1.092 0.985 0.940 0.953 1.147 1.086 1.335 1.340 1.007 0.000 9 0.718 0.683 0.782 0.803 0.899 1.057 1.196 0.953 0.943 1.048 1.180 1.272 1.164 0.000 0.000 j

10 0.921 1.059 0.926 1.148 1.097 1.238 1.123 1.147 1.048 1.094 1.014 0.947 0.981 0.000 0.000 11 1.239 1.138 1.282 1.150 1.299 1.135 1.296 1.086 1.100 1.014 0.868 0.000 0.000 0.000 0.000 i

12 1.251 1.340 1.342 1.266 1.268 1.359 1.366 1.335 1.272 0.947 0.000 0.000 0.000 0.000 0.000 l

13 1.278 1.362 1.364 1.282 1.281 1.361 1.373 1.340 1.164 0.981 0.000 0.000 0.000 0.000 0.000 l

14 1.328 1.347 1.347 1.327 1.323 1.336 1.151 1.007 0.000 0.000 0.000 0.000 0.000 0.000 0.000 15 0.976 0.971 0.966 0.974 0.971 0.972 0.978 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Time = 75.1100 Edit #=

167 Tu:a1 Power (MWt) 576.49 Maxmimum Powered node at ( 7,13, with value of 1.39 0

1 2

3 4

5 6

7 8

9 10 11 12 13 14 15 l

1 0.369 0.501 0.623 0.409 0.425 0.590 0.708 0.642 0.714 0.917 1.233 1.234 1.268 1.336 1.028 2 0.501 0.d52 0.548 0.642 0.616 0.754 0.640 0.642 0.673 1.058 1.130 1,339 1.365 1.357 1.023 3 0.623 0.548 0.668 0.558 0.719 0.716 0.798 0.587 0.774 0.922 1.281 1.340 1.368 1.358 1.018 4 0.409 0.642 0.558 0.447 0.478 0.802 0.002 0.743 0.802 1.150 1.143 1.253 1.273 1.337 1.028 5 0.425 0.616 0.719 0.478 0.522 0.782 0.989 0.830 0.897 1.097 1.298 1.256 1.274 1.334 1.025 6 0.590 0.754 0.716 0.802 0.782 0.968 0,885 1.091 1.054 1.236 1.129 1.361 1.367 1.352 1.028 7 0.788 0.640 0.798 0.802 0.989 0.885 1.a87 0.980 1.191 1.114 1.295 1.373 1.386 1.168 1.037 8 0.642 G.642 0.587 0.743 0.830 1.091 0.980 0.931 0.941 1.128 1.071 1.340 1.357 1.062 0.000 9 0.714 0.673 0.774 0.802 0.897 1.054 1.191 0.941 0.930 1.029 1.168 1.284 1.214 0.000 0.000 10 0.917 1.058 0.922 1.150 1.097 1.236 1.114 1.128 1.029 1.085 1.015 0.992 1.039 0 000 0.000 l

11 1.233 1.130 1.281 1.143 1.298 1.129 1.295 1.071 1.168 1.015 0.905 0.000 0.000 0.000 0.000 t

12 1.234 1.339 1.340 1.253 1.256 1.361 1.373 1.340 1 284 0.992 0.000 0.000 0.000 0.000 0.000 13 1.268 1.365 1.368 1.273 1.274 1.367 1.386 1.357 1.214 1.039 0.000 0.000 0.000 0.000 0.000 14 1.336 1 357 1.358 1.337 1.334 1.352 1.168 1.062 0.000 0.000 0.000 0.000 0.000 0.000 0.000 15 1.028 1.023 1.018 1.028 1.025 1.028 1.037 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 l

Figure B.1 Radial Power Distributions During MSIV Closure Transient from TRAC /B-NESTLE at time zero and 25 seconds.

i E

5' I

I' t

l u

GE Nuclear Energy cc:

Gary L~ Sozzi, Manager l

Technical and Modification Services GE Nuclear Energy 175 Curtner Avenue San Jose, CA 95125

]

George B. Stramback GE Nuclear Energy 175 Curtner Avenue San Jose, CA 95125 i

David W. Reigle GE Nuclear Energy 175 Curtner Avenue -

San Jose, CA 95125 -

James F. Klapproth GE Nuclear Energy 175 Curtner Avenue San Jose, CA 95125 i

i Re!ph J. Reda, Manager Fuel and Facility Licensing General Electric Company P.O. Box 780 Wilmington, NC 28402 --

]

Glenn A. Watford, Manager

' Fuel Engineering GE Nuclear Energy P.O. Box 780 -

l Wilmington, NC 28402 James L. Rash GE Nuclear Energy P.O. Box 780 Wilmington, NC 28402 l

l ll.

P i

F-A