ML17285B113

Advises of Release of Subchannel Thermal Hydraulic Computer Code & Requests Review of VIPRE-01 Mod-02 & Issuance of SER for PWR & BWR Applications.Bwr Qualification Analyses Completed & Rept on BWR Applications Encl

ML17285B113
Person / Time
Site:	Columbia
Issue date:	03/28/1990
From:	Yung Y WASHINGTON PUBLIC POWER SUPPLY SYSTEM
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
EANF-90-0053, EANF-90-53, NUDOCS 9003230055
Download: ML17285B113 (125)

Text

ACCELERATED DISTRIBUTION DEMONSTIRATION SYSIHM f

REGULATORY INFORMATION DISTRIBUTXON SYSTEM (RIDS)

ACCESSXON NBR:9003230055 DOC.DATE: 90/03/28 NOTARIZED: NO DO'CKET FACIL:50-397 WPPSS Nuclear Project, Unit 2, Washington Public Powe 05000397 AUTH. NAME AUTHOR AFFILIATXON YUNG,Y.Y. Washington Public Power Supply System RECXP.NAME RECIPIENT AFFILIATION Document Control Branch (Document Control Desk)

SUBJECT:

Advises of release 6 request for NRC review of VXPRE-01 Mod-02.

DISTRIBUTION CODE: T010D COPIES RECEIVED:LTR ENCL SIZE:

TITLE: Non-Proprietary Topical Report NOTES: g~~~+

RECIPXENT COPIES RECIPIENT COPIES ID CODE/NAME LTTR ENCL ID CODE/NAME LTTR ENCL INTERNAL: ACRS 10 10 ASLBP 1 1 NRR POTEAT,E 1 0 NRR WILSON,V. 1 1 NRR/DLPQ/LPEB10 1 0 NRR/DST SE2 1 0 NUDOCS-ABSTRACT 1 1 B 1. 0 OGC/HDS2 1 0 REG F L 01 1 1 RES 1 1 1 0 RGN2 1 0 RGN3 1 0 RGN4 1 0 RGN5 1 0 EXTERNAL: NRC PDR 1 1 NSIC 1 1 NOTE TO ALL "RIDS" RECIPIENTS:

PLEASE HELP US TO REDUCE WASIEf CONTACI'HE.DOCUMENT CONTROL DESK, ROOM Pl-37 (EXT, 20079) TO ELIMINATEYOUR NAME FROM DISIRIBUTION LISIS FOR DOCUMENTS YOU DON'T NEED!

TOTAL NUMBER OF COPIES REQUIRED: LTTR 27 ENCL 17

t 0 0 l I l'

C c'I I,

1 J

WASHINGTON PUBLIC POWER SUPPLY SYSTEM P.O. Box 968 ~ 3000 George Washington Way ~ Richland, Washington 99352 February 28; 1990 EANF-90-0053 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington; D.C. 20555 Gentlemen:

Subject:

NOTIFICATION OF RELEASE AND RE(}UEST FOR NRC REVIEW OF VIPRE-01 MOD-02 Refer ences: 1. Letter from J. A. Blaisdell (UGRA) to H. R. Denton (NRC),

December 17, 1984

2. Letter from C. E. Rossi (NRC) to J. A. Blaisdell (UGRA);

Acceptance for Referencing of Licensing Topical Report, EPRI NP-2511-CCM, "VIPRE-01: A Thermal-Hydraulic Analysis Code for Reactor Cores"', Volumes 1; 2; 3, and 4; May 1; 1986.

VIPRE-01 is a subchannel thermal-hydraulic computer code that was developed by Battell'e Pacific Northwest Laboratories under the sponsorship of the Electric Power Research Institute (EPRI). Because the EPRI charter does not allow for long term support of a released computer code; the VIPRE-01 Maintenance Group was formed in 1987 to provide funding and guidance for ongoing support and code maintenance. The VIPRE-01 Maintenance Group is a code user organization that currently consists of seventeen member utilities; the membership list is shown in Attachment l.

VIPRE-Ol was submitted for NRC review by the Utility Group for Regulatory Applications (UGRA) in 1985 (Reference 1). An SER (Reference 2) was issued by the NRC approving VIPRE-01 for PWR applications for heat transfer regimes up to the point of critical heat flux. As discussed in the SER, VIPRE-01 is maintained under a guality Assurance program at Battelle. Code errors reported by code users are handled in accordance with this (tA program, which provides for user notification and release of new code versions incorporating error corr ecti ons.

The VIPRE-01 NOD-02 has recently been released for use. This version, which was prepared in accordance with the VIPRE-01 gA pr ogram, includes 77 changes from the VIPRE-Ol NOD-01 version. VIPRE-01 NOD-01 is the version that was reviewed and approved in Reference 2. A report (Attachment 2) describing the effect of the 77 changes is included in this submittal. A copy of the complete error/change log will be submitted if required.

9003230055 900328 PDR ADOCK 05000397 P PDC

l Page Two NOTIFICATION OF RELEASE AND REQUEST FOR NRC REVIEW OF VIPRE"01 MOD-02 The VIPRE-01 computer code'ocumentation (EPRI NP-2511-CCM; Volumes 1 through

4) was submitted to the NRC previously.'t your request, we will submit the appropriate number of copies of the updated code documentation. It is the desire of the VIPRE-01 Maintenance Group to have the NRC review VIPRE-Ol MOD-02 and issue an SER for PWR and BHR applicati'ons.

The original VIPRE-01 SER did not address BWR applications. However, BWR qualification analyses have now been completed, and a report on BWR applications (Attachment 3) is included in this submittal.

A review of this new code version is appropriate due to several near term utility applications, including:

Date Submittal Mid 1990 Oconee Nuclear Station - reload thermal-hydraulic analysis, Duke Power Company Early 1992 HNP-2 Cycle 8 reload safety analysis; Washington Public Power Supply System (WPPSS)

Late 1992 Comanche Peak Unit 1 - Cycle 3 reload safety analysis,'U Electric If you have any questions or comments, please contact me at (509) 372-5195.

Sincerely, Y. Y. Yung, Chairman VIPRE-Ol Maintenance Group Washington Public Power Supply System PO Box 968 Richland, HA 99352 YYY:bw Attachments: 1) VIPRE-01 Maintenance Group Membership List

2) EPRI Report Summarizing Changes

3) HPPSS BWR Analysis Report cc: R.C. Jones, NRC S. Maier, TU Electric D. Koontz, Duke Power Company G.S. Srikantiah, EPRI J. Cuta, Battelle Pacific Northwest Laboratories VIPRE Maintenance Group

ATTACHMENT 1 VIPRE-01 Maintenance Group Members Baltimore Gas 8 Electric

2. Commonwealth Edison

3. Duke Power Company Florida Power 8 Light Company

5. General Public Utilities

6. Houston Lighting 8 Power

7. Illinois Power Company

8. Middle South Services

9. New York Power, Authority

10. Northeast Utility Service Co.

11. Northern States Power Company

12. Public Service Electric 8 Gas Company 13.' Tennessee Valley Authority,'U

14. Electric

15. Union Electric Company'olf

16. Creek Nuclear Operating Corporation

17. Washington Public Power Supply System

~:

t

Summar of Changes in VIPRE-01 between MOD-01 and MOD-02 The revisions to VIPRE-01 between MOD-01, released in 1985, and the current version, Mod-02, released in October of 1989, included 77 individual changes.

These changes consisted for the most part of minor error corrections to the code and enhancements of code output features. The one major new modification was to include an optional drift-flux model"to improve the code's capabilities to calculate the evolution of the void fraction profile in transients with two-phase flow conditions.

Twenty-two of the 77 changes were corrections to the ASP post-processor graphics program, which reads a binary file created by the VIPRE code, and makes CALCOMP plot files. These changes make the ASP program more useful, but do not affect the results or performance of the VIPRE code itself.

Of the remaining 55 changes, 28 were changes to enhance the appearance of the output -- e.g., expand variable print-out fields, add new or expanded informative diagnostic messages, write out additional information for the user -- or to update relevant information on the output banner page and case title line. The nature of these. changes was such'hat, although it is desirable to have them ia the new version, they have no real impact on the, performance of the code..

The other 27 changes in MOD-02 were changes that fix problems which actually make the code fail in execution, or which'esult in different answers from those obtained in MOD-01. Only eleven of the changes were for problems that cause the code to fail. In most cases, these errors occurred when certain options were used in the code in unforseen ways, or in untested combinations, and since the calculation actually failed, the user was in no danger of assuming .the code had given him an answer which consisted of bad resul,ts.

The remaining 16.changes, however, do result in some difference in the performance of the code, when comparing results from MOD-02 to those obtained for the same problem in MOD-01. In all cases, the differences are small, or confined to a certain narrow region of a calculation. The following table summarizes briefly the changes that can result in differences in the code results.

Table 1: MOD-02 Chan es which affect Code Results CH101 BAWg2 CHF correlation's nonuniform axial power correction should be applied even if quality is below -0.03, which is below the range of the correlation.

(This is consistent with standard usage of the correlation.) Results in small change in DHBR values at high subcooling.

CH102 Corrects pressure solution for water tube channel modeled using optional input in group BWRG. Results in small change in water tube flow rate.

CH110 Corrects the application of the usineu axial power profile option. With a cold inlet length the total power input was incorrect in MOD-01 for this case.

Results in change in total power input, with consequent change in fluid conditions. Magnitude of the difference between MOD-01 and MOD-02 results will depend on the problem.

CH116 Fix for coding error in the water tube channel model. Results in a small mass flow difference, which is on the order-of machine round-off.

CH119 Fixes error in Y'erm in the Bowring MSC-2 CHF correlation, which results in small changes in DNBR values. Differences are noticeable only for cases with abnormally large crossflow velocities.

CH125 and CH128 Fix errors in coefficients of zircaloy thermal conductivity equation, which cause small changes (on the order of 58) in clad and fuel temperatures.

CH127 F>xes energy error in boiling transients when the snbcooled boiling models are used. Results in differences in the evolution of the void profile during the transient. Magnitude.of'the differences between MOD-01 and MOD-02 results is problem-dependent.

CH130 Fixes. heat transfer input flag 'THSP'o that't gives Thorn-plus-Dittus-Boelter for boiling heat transfer coefficient, as it is supposed to, rather than Thorn plus user-specified single-phase heat transfer coefficient. Results will be different only in cases where the user selected something other than Dittus-Boelter for single-phase heat transfer, and the differences will'be.

noticeable only if the alternate correlation gives numbers significantly different from what Dittus-Boelter would. predict for the same conditions.

CH143 Fix to correct an overwrite that occurs in the energy solution if the bandwidth for a problem is larger than the number of channels (Note:

requires a perverse disregard for the recommended efficient channel number it convention to encounter this error.) Results in differences in enthalpy; the magnitude of the effect will be problem depended. In any even moderately efficient channel numbering scheme, the bandwidth will be much smaller than the number of channels, so this is an extremely rare problem.

CH147 Fix to correct the user-specified gap conductance table interpolation when

'the axial locations named in the input correspond exactly to the locations of the node boundaries. Results in shifting of the node locations of the specified gap conductance values, and can result in differences in fuel and clad temperatures. Magnitude of the effect will be problem dependent.

CH148 Fix to reset heat transfer coefficient derivatives to zero between calculation for each rod in the conduction solution. If this is not done, round-off or internal iteration error can sometimes cause a spurious heat flux from a rod with an adiabatic boundary condition. Results could change if the spurious

heat flux were large enough to constitute a significant energy source, but in most cases it will be essentially zero, even without the fix.

CH158 Fix so that the vapor density is evaluated at the film temperature, rather than at the superheated vapor temperature, in the application of the

~

Groeneveld-Delorme film boiling heat transfer correlation. Results in differences in the heat transfer coefficient value, and consequently the surface temperature and rate of heat transfer to the fluid, but affects only heat transfer in the film boiling (post-CHF) region when the Groeneveld-Delorme correlation is selected for heat trans'fer in that region.

CH159 Fix to include the iteration to determine the critical boiling length for the EPRI-2 CPR correlation. Results in different critical power predictions when the EPRI-2 CPR correlation is used, and yields a better comparison to data.

CH174 Fix to make Levy subcooled boiling model use the Dittus Boe-lter single-phase heat transfer correlation even when the user selects an alternate correlation for the single-phase regle; ..Results in different subcooled quality, predictions only when the user selects something other than Dittus-Boelter (the VIPRE default) in the single'-phase region, and the differences will be noticeable only if the alternate correlation 'results in significantly different heat transfer coefficients, compared to Dittus-Boelter.

CH176 'l F>x on the default for the minimum number of iterations for calculations with the pressure drop boundary condition option. For transients, the minimum number of iterations needs to be 3 rather than 2, or the code might think is converged when it is not. Affect on results will be problem-dependent, it with most cases unaffected, because they will probably be taking. more than the minimum number of iteration, anyway.

ATTACHMENT 3 9003230055

1

'l l

'i II'I II

VERIFICATION OF VIPRE-Ol FOR BWP, ANALYSIS By Y. Y. Yung September 1987 WASHINGTON PUBLIC POWER SUPPLY SYSTEM Rich1and, Washington

a, TABLE OF. CONTENTS SECTION PAGE

1.0 INTRODUCTION

2.0 BWR CORE MODELS FOR VIPRE-01 3.0 SENSITIVITY STUDIES FOR VIPRE"Ol BWR CORE MODELS 13 4.0 BENCHMARK STUDIES 31 5.0 FSAR TRANSIENT CALCULATION COMPARISON 45

6.0 CONCLUSION

55

1 I

(

I

VERIFICATIOH OF VIPRF-01 FOR BWR ANALYSIS 0

1. 0 IHTRODUCTI OH VIPRE-01( is a thermal-hydraulics code designed for steady state and transient analysis of nuclear reactor cores under normal operating condi-tions and a wide range of assumed abnormal events and transients. It was developed by Battelle Horthwest under EPRI sponsorship to be a tool for evaluating nuclear reactor safety limits, '.ncluding minimum departure for nucleate boiling ratio (MDHBR), critical power ratio, (CPR), fuel and clad temperatures, and coolant state. Developed from the COBRA(

family of codes, VIPRE-01, underwent a long per iod of testing and verifi-cation by a working group composed of EPRI-member utilities from May 1981 to January 1984. The code was first released in January 1984, as YIPRE-01, MOD-O, and was subjected to an independent design review. The code and its documentation were revised in response to design review com-ments, and a new version, YIPRE-01, MOD-1, was published in t1ay 1985. At that time, a Utility Group for Regulatory Application (UGRA) was formed

.to prepare a submittal to the Huclear Regulatory Commission (HRC) re-questing generic approval of VIPRE-01 for licensing calculations. A sub-mittal addressing PWR application has been approved, with the HRC Safety Evaluation Report 'SER) issued in April 1986.

The work reported here is in support of BWR applications of VIPRE-01.

Core models were developed for YIPRF.-01 based on Washington Nuclear Plant Ho. 2 (WHP-2), a BWR/5 reactor with 8x8 fuel and a design power level of

f 3323 MWt. Sensitivity studies were performed for the core model, and transient calculations were performed to compare with selected transients from the WNP-2 FSAP~ ~. The core models are described in detail in Section 2.0. The sensitivity studies performed with the core models are discussed in Section 3.0. Calculations to benchmark the VIPRE results against GE design calculations for core pressure drop, FSAR steady state results and WNP-2 core follow calculations are presented in Section 4.0.

The FSAR transient calculation comparisons are presented in Section 5.0.

Conclusion is provided in Section 6.0 and references are in Section 7.0.

2.0 BWR Core Models For VIPRF.-01 The reactor core of WPPSS Nuclear Plant No. 2 at Hanford was used in this evaluation of the VIPRE-01 code's capabilities for modeling BWR thermal-hydraulic conditions. A complete description of the plant is available in the FSAR (see Reference 6). The core of WNP-2 consists of 764 assem-blies of Bx8 fuel, and a bypass region containing 185 cruciform control rod structures. Figure 2.1 shows a diagram of a one quarter cross sec-tion of the WNP-2 core. The core region begins at the inlet orifice, where flow enters from the lower plenum, and extends to the top of the fuel assemblies, to just above the upper tie plate.

Two basic core models were examined in these calculations. The first was a relatively simple 4-channel model, consisting of two channels modeling a single bundle each the hot central bundle and the hot peripheral bundle and two large lumped channels; one modeling the remaining 671 central bundles, and one modeling the remaining 91 peripheral bundles.

II l

I I

I

I The rods are similarly lumped, with one rod representing all the fuel rods contained in the bundles comprising lumped channels. Figure 2.2 illustrates this radial noding scheme for the four channel model. The second VIPRE model for the llHP-2 was a detailed model of a one quarter section of symmetry, with one channel for each assembly. Figure 2.1 illustrates the channel and rod numbering scheme for this model.

Both of the VIPRE-01 models of the HNP-2 core considered here model the full axial length of the core region, from inlet orifices to the upper tie plate. The fuel assemblies are modeled with one dimensional chan-nels, with a single average flow area, velocity, and density at a given axial location.

lilith this approach, VIPRE-01 can correctly calculate the axial distribu-tion of the radially averaged void fraction, density, and enthalpy along a fuel channel. A more detailed subchannel model of the flow field with-in the fuel assembly is unnecessary, and would in fact be counter-produc-tive, as the three equation homogeneous formulation of the conservation equations in VIPRE is not capable of solving for the correct radial dis-tribution of void in fully-developed annular two-phase flow. This is a generic limitation of the homogeneous model, and its effect on the code's accuracy in two-phase flow calculations is fully documented in Volume 4 of the VIPPE code manual (see Reference 1). The results presented in Volume 4 show that VIPRE is fully applicable to two-phase flow analysis when. the flow field can be averaged in the radial direction.

l

/ll f,

f 1

l I

I

,[i l

For these calculations, the bypass region of the core is not included in the VIPRE model. Instead, it is assumed that 10.4 percent of the rated core flow is diverted to the bypass. This assumption is made for convenience only. VIPRE-01 is capable of modeling the bypass region of a BWR, and calculating the amount of bypass flow through the various leak-age paths at or near the core inlet. However, this option requires de-tailed input describing the lower core structure, and the effective hy-draulic losses of individual leakage paths. This information is not always easily obtained, and would in any case be plant specific. For the purposes of these calculations, which are to demonstrate capability rather than produce specific results for a particular analysis, the as-sumption of 10.4 percent bypass flow is quite sufficient.

The inlet flow distribution among the channels modeling the fuel assem-blies is calculated in YIPRE assuming a uniform pressure drop across the core. With this boundary condition option, the inlet flow to each chan-nel is adjusted iteratively until the pressure drop in each channel is the same. (The total specified inlet flow is conserved, however.) The pressure drop in a given channel is calculated considering the total hy-draulic losses, including gravity head, wall friction, and local pressure losses. Wall friction was determined using the Blasius smooth-tube cor-relation, f= 0.32 Re (which is the VIPRE-01 default). Local pressure losses were modeled for the inlet orifice, lower tie plate, seven spacer grids, and the upper tie plate. Figure 2.3 shows the axial locations of the local pressure losses, and Table 2.1 lists the loss co-efficient values 'sed in VIPRE to model the losses.

l J

I k

The conduction model was not used in these calculations. The fuel rods were modeled using the "dummy" rod option in VIPRE, in which the power generated in the fuel is used to define a heat flux boundary. condition at the rod surface. For the 4-channel model, the axial and radial power distributions were obtained from values repor ted in the FSAR (see Pefer-ence 6). For the 191-channel quarter core model, radial and axial power distributions were obtained from a calculation using the SIMULATE-2(

core neutronics code. SIMULATE-2 produced a radial power factor for each assembly in the quarter core model, as shown in Figure 2.4. An axial power profile was generated for each of eight radial rings in the core.

Assemblies within each ring were assumed to have the same axial power profile. Figure 2.5 diagrams the rings and indicates by corrspondence with the quarter core diagram in Figure 2. which assemblies are included in each ring.

The operating conditions used as boundary conditions for the steady state calculations are listed in Table 2.2. These are the rated conditions from the FSAR. It should be noted, however, that the value used for the inlet flow was the active core flow, rather than the total rated flow, where active flow = (total flow - bypass flow).

Critical power calculations were performed using the General Electric proprietary critical quality boiling length correlation, GEXL . The correlation was programmed into VIPRE using the dummy subroutine CPR3,

/

provided for that purpose, and is accessed by the option "t1IHE" in the VIPRE input for CPR correlation selection. The bundle R-factors required as input for the GEXL correlation, which are functions of the radial power distribution within the fuel bundles were obtained by linear inter-polation of tabulated R-factor values in the l<HP-2 Process Computer Data Bank (see Reference 7).

ll 1

f

TABLE 2.1 PRESSURE LOSS COEFFICIENTS FOR WNP-2 CORE CHANNELS Structure Loss Coefficient Inlet Orifice Central 29.25 Peripheral 213.52 Lower Tie Plate 7.86 Upper Tie Plate 1.46 Spacer Grids 1.24 TABLE 2.2 OPERATING CONDITIONS FOR VIPRE-01 CALCULATIONS WITH MODELS OF WNP-2 Parameter Yalue Core Power 3323 HWt Inlet Flow Rate Rated 108 5 X 106 ibm/hr Active 97.2 X 106 ibm/hr*

Inlet Enthalpy 527.6 btu/lb System Pressure 1035.0 psia

• MNP-2 Process Computer Data Bank (Reference 7) gives the core bypass flow as a tabular function of the total core flow entering the core inlet plenum. At the rated total core flow (108.5 X 106 ibm/hr), the bypass flow is 11.3 X 106 ibm/hr which is 10.4 percent of total core flow.

I' II I

1

Channel no.

Rod no.

1 2 3 4 5 6 7 8 9 10 1'1 12 13 14 15 30 31 26 24 22 76 20 19 91 18 106 120 12 0

06 06 167 04 185 186 187 188 189 190 191

~ nOl ocg 0 0 0 0e e0 t0 0m e0 0 N u) co m 0 01 9 C9 W 9- IA 9- Cg 9

W CO T Ol 9- Ol 0 OI C9 N Ol Ol Ol I

Ol OJ Figure 2.1 191 Channels guarder Core Nodalization

II Channel No. 4 S

2 01 Rod No.

Channel 1 models one bundle, the hottest peripheral bundle Channel 2 models one bundle, the hottest central bundle Channel 3 models 91 peripheral bundles Channel 4 models 671 central bundles Rod 1 models 62 fuel rods'2 Rod 2 models fuel rods Rod 3 models 91 X 62 = 5,642 fuel rods Rod 4 models 671 X 62 = 41,602 fuel rods Figure 2.2 Four Channels Full Core Nodalization

II g

l

Node 31 30 UTP 29 TAF 28 27 SPA SPA 20 SPA SPA 15 SPA 10 SPA SPA 6

6 4 BAF LTP~

ORIF 1

UTP '- Upper Tie Plate TAF - Top of Active Fuel SPA - Spacer BAF - Bottom of Active Fuel LTP - Lower Tie Plate ORIF - Orifice Figure 2.3 Axial View of Bundle/Channel l

i I

4

(

0. 828 1. 080 1. 113 0. 945 0. 942 i. 123 l. 082 0. 834 0. 832 1. 097 i. 162 j. 167 i. 033 0. 815 0. 312

1. 080 1. 167 l. 286 1. 164 1. 252 1. 202 1. 255 '1. 072 l. 146 l. 154 l. 261 1. 188 i. 053 Q. 819 0. 313

i. 113 1. 286 1. 242 1. 346 1. 249 1. 350 i. 202 i. 193 i. 093 1. 256 1. 180 i. 192 l. 047 0. 812 0. 308

0. '945 l. 164 l. 346 i. 238 1. 341 1. 260 1. 272 0. 942 0. 996 1. 122 1. 243 1. 153 l. 012 0. 788 O. 298 O. 942 l. 252 1. 249 1. 341 1. 251 1. 377 l. 1'99 1. 039 0. 943 1. 214 l. 224 1. 134 0. 983 0. 756 0. 280

i. 123 i. 202 l. 350 1. 260 1. 377 1. 295 l. 371 1. 191 1. 257 1. 191 l. 233 1. 123 0.'951 0. 701 0. 247

1. 082 i. 255 1. 202 1. 272 i. 199 1. 371 l. 285 1. 357 1. 227 l. 280 l. 194 i. 062 0. 858 0. 583 0. 198

0. 834 l. 072 l. 193 0. 942 1. 039 1. 191 1. 357 l. 228 1. 272 l. 214 1. 110 0. 935 0. 696 0. 294

0. 832 l. 146 1. 093 O. 996 0. 943 i. 257 1. 227 1. 272 'i. 197 1. 116 0. 962 0. 752 0. 324

1. 097 1. 154 1. 256 1. 122 1. 214 l. 191 l.'280 1. 214 '1. 116 0. 964 0. 751 0. 336 0. 185

l. 162 l. 261 i. 180 1. 243 l. 224 i. 233 1. 194 i. 110 0. 962 0. 751 0. 353

i. 167 1. 188 1. 192 i. 153 1. 134 l. 123 l. 062 0. 935 0. 752 0. 336 33 l. 053 1. 047 1. 012 0. 983 0. 951 0. 858 0. 6'96 0. 324 0. 185 5 0. 81'9 0. 812 0. 788 0. 756 0. 701 0. 583 0. 294

0. 313 0. 308 0. 298 0. 280 0. 247 0. 198 Figure 2.4 quarter Core Radial Power Factors Calculated With SIMULATE-2 l

t t'

l I

1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 2 2 2 3 3 4 4 5 5 6 6 7 7 8 8 2 2 3 3 3 4 4 5 5 6 6 7 7 8 8 3 3 3 3 4 4 4 5 5 6 6 7 7 8 8 3 3 3 4 4 4 5 5 5 6 6 7 7 8 8 4 4 4 4 4 5 5 5 6 6 7 7 7 8 8 4 4 4 4 5 5 5 6 6 6 7 7 8 8 8 5 5 5 5 5 5 6 6 6 7 7 8 8 8 5 5 5 5 5 6 6 6 7 7 7 8 8 6 6 6 6 6 6 6 7 7 7 8 8 8 66666777788 7 777777888 7777778888 88888888

8888888, Figure 2.5 quarter Core Pings Definition

[f

'l r

3.0 SENSITIVITY STUDIES FOR VIPRE-01 BMR CORE MODELS Sensitivity studies were performed to assess the effects of various modeling options and approaches for BWR analysis using VIPRE-Ol. Of pri-mary interest was the radial noding study, to determine if the simple 4-channel core model would yeild results as good as those obtainable with the detailed quarter core 191-channel model. It is reasonable to expect that the two models would yield similar results, given the nature of the flow field in a BMR core, which consists of an array of isolated chan-nels, wherein the inlet flow distribution provides the only means of altering the local flow. It was necessary to determine that lumping large numbers of assemblies into one or two channels would not materially affect the inlet flow to the hot assemblies in the center and periphery of the core.

The results presented in Section 3.1 show that the 4-channel model is quite adequate for BWR analysis with VIPRE, and yields essentially the same hot channel results as the detailed quarter core model. Since the 4-channel model is much less expensive to run, it alone was used in the subsequent sensitivity studies to investigate the effects of different modeling options on the BWR results. These sensitivity studies examined the effects of axial noding, water properties options, hot wall correc-tion for viscosity, local pressure option, and two phase flow correlation selection, and are discussed in detail in Sections 3.2 through 3.7.

t 3.1 Radial Noding Sensitivity Radial noding sensitivity was examined by comparing hot channel exit temperatures and core pressure drop calculations obtained with the 4-channel and 191-channel models of WNP-2 in VIPRE-01. Nineteen different cases were run, for varying combinations of power level and flow rate. The results are summarized in Table 3.1. For each case, the hot channel exit temperatures and core pressure drops pre-dicted with the two different models are essentially identical.

These results are consisted with the results of radial noding studies presented in Volume 4 of the VIPRE documentation (see Refer-ence 1). Boding studies in a PWR core geometry showed a similar insensitivity to core lumping pattern, when the hot region of the core was modeled with adequate detail. In a BWR core, where flow redistribution due to crossflow cannot occur at all, the overall core lumping pattern is even less important, when the hot assemblies are modeled as a single channel each.

TABLE 3.1 RADIAL NODING SENSITIVITY STUDY -

SUMMARY

OF RESULTS Case Totalc Hot Channel Core No. Power Flow MCPR Exit Tem ( F) Pressure Dro X of '4 of OCa FCb ()C FC QC FC ated) . ated) 1 100.0 100.0 1.42 1.42 548.3 548.4 22.12 22.41 2 76.0 99.0 1.89 1.89 544.1 544.3 20.92 20.89 3 68.6 59.7 1.87 1.87 541.3 541.5 10.63 10.60 4 91.1 83.1 1.56 1.57 542.0 542.2 17.33 17.29 5 86.7 98.0 1.60 1.60 540.3 540.5 21.42 21.39 6 97.2 99.5 1.48 1.48 543.5 543.7 23.07 23.03 7 99.5 100.0 1.46 1.46 545.2 545.4 22.83 22.81 8 95.5 98.5 1.51 1.51 544.2 544.4 22.27 22.23 9 48.8 50.0 2.31 2.31 538.8 538.9 8.41 8.37 10 98.1 98,2 1.46 1.47 544.4 544.6 21.89 21.86 11 100.2 98.8 1.39 1.39 546.1 546.2 22. 26. 22.24 12 100.0 97.2 1.43 1.43 545.4 545.6 21.98 21.94 99.8 95.7 1.48 1.48 544.5 544.7 21.05 21.02 14 96.1 88.9 1.50 1.50 544,7 544.9 19.03 19.01 15 57.7 32.6 1.71 1.72 543.3 543.5 5.55 5.50 16 70.4 53.0 1.68 1.70 544.6 544.8 9.90 9.87 17 71.8 52.1 1.66 1.66 545.5 545.7 9 79 9.75 18 71.4 50.7 1.67 1.67 543.7 541.4 9.48 9.53 19 71.6 50.7 1.63 1.63 544.1 544.3 9.50 9.48 NOTES:

aguarter Core 191-Channel Model bFull Core 4-Channel Model cTotal flows are given here. VIPRE calculations used active flows, where active flows = (total flows - bypass flows). Bypass flows which are func-tions of core flows were obtained by linear interpolation of tabulated by-pass flow values in the WNP-2 Process Computer Data Bank (see Reference 7).

I I

3.2 Axial Noding Sensitivity Sensitivity to axial noding was investigated bv examining the effect of three different axial noding patterns on the MCPR, exit flow rate, and void fraction in the hot central channel using the 4-channel model. The only effect the axial node size might be expected to have on the results of a calculation in what is essentially a one dimensional channel is in the averaging effect on the axial power distribution. Finer noding might make minor differences in the enthalpy and void fraction distributions, as the non-uniform axial power distribution is more accurately resolved.

8ut it should make no real difference in the overall results.

Three different axial node sizes were used in this sehsitivity study.

The base case used six-inch axial nodes. Case I used coarser noding, with ten-inch axial nodes, and Case 2 used slightly finer noding, with five-inch axial nodes. (In both of these cases, six-inch axial nodes were used for the first 36 inches of the axial length. This length con-sisted of the unheated inlet region, and only the first 18 inches of active fuel where the fluid would be subcooled, and thus would have little or no effect on the void distribution.)

I I

t I

l I

t II f

The results obtained for these three cases are compared in Table 3.2.

The differences in the flow solution, as evidenced by the values for pressure drop, exit flow, and void fraction, are very small. The MCPR is approximately three percent lower for the coarse noding (Case 2), but there is no discernable change in MCPR for the finer noding (Case 3).

These results indicate that six-inch nodes quite adequately resolve the axial distributions in the core, and is a suitable approximation for B>JR analysis.

TABLE 3.2 AXIAL NODING SENSITIVITY STUDY -

SUMMARY

OF RESULTS Hot Channel Core Computer Case Flow Rate Exit Void Pressure Time No. Axial Nodalization HCPR (ibm/sec) Fraction Drop (psi) (SRU) 31 (31X6") 1.316 34.821 0.7851 23.78 17.339 21 (6X6' 15v10") 1.274 34.850 0.7845 23.72 14.430 36 (6X6" + 30X5") 1.316 34.858 0.7850 23.72 18.906

3.3 Water Properties Generation In a YIPRE calculation, the physical properties of the working fluid must be described. These water properties may be specified in the VIPRE model in three ways. First, a water properties table may be created by VIPRE. When using this option, the user specifies the range of the table which is needed for a given problem and the num-ber of entries that the table will contain over this range. The EPRI water properties functions (Reference 11) are used to generate the points in this table. In performing its solution, VIPRE inter-polates on this table for any needed property. In the second method, YIPRE used the EPRI water properties functions directly for each property needed for the solution. (This method eliminates any table interpolation errors.) In the third method, the user speci-fies property tables by input. User input is required if a fluid other than water is being modeled. Also, user input is necessary if the operating conditions of a problem are outside the range of the EPRI water properties functions.

The EPRI water properties functions are very accurate over a wide pressure and enthalpy range. This pressure range spans from below atmospheric pressure to above the critical pressure. The enthalpy range is from 200 to 830 btu/ibm. The operating conditions which are used in VIPRE analyses for WNP-2 are well within the range of the EPRI functions.

'I I

t

The option for user input of the water properties table is not con-sidered for two reasons. The first is that the EPRI functions are very accurate over their range of applicability. Secondly, the po ten tial for input error i s increased.

Two cases were considered in the water properties sensitivity study.

In Case 1, water properties were determined from a 26-entry table (table spacing = 10 btu/ibm) generated using EPRI functions. In Case 2, the direct evaluation of the EPRI functions was used to de-termine properties. Any differences in the results obtained for these two cases would have to be ascribed to differences between the property values interpolated from the table, and those obtained from evaluation of the EPRI functions for the local fluid conditions in each cell.

The results obtained for these two cases are shown in Table 3.3.

The difference in saturation temperature is only about 0.1 percent which is smaller than the error band in the properties function it-self. These results show that linear interpolation in the water properties table is as accurate as direct evaluation of the func-tions at BWR operating conditions when the table is specified with a reasonable enthalpy increment (i.e., on the order of 10 btu/ibm).

Within these constraints, the VIPRE solution is insensitive to the method of water properties generation. This is consistent with the results in VIPRE-01, Volume 4 (see Reference 1).

TABLE 3.3 WATER PROPEPTIES GENERATION SENSITIVITY STUDY -

SUMMARY

OR PESULTS Hot Channel Exit Core Computer Case Temp Flow Rate Void Pressure Time No. MCPR ( F) (ibm/sec) Fraction Drop (psi) (SRU) 1.337 548.37 34.669 0.7862 24.11 16.854 1.337 548.99 34.667 0.7861 24.09 16.961 3.4 Rohsenow-Clark Viscosity Model The Rohsenow-Clark model modifies the axial friction factor (f) to account for the effect on wall drag of the fluid viscosity variation near a heated surface. This model adjusts f to include the effect of decreased fluid viscosity near the fuel rod surface. This is usually a small effect, and the default in VIPRE is to ignore it, and evalute friction factor based on the'bulk fluid temperature in a node.

However, in BWR analysis, the adjustment to wall drag will be felt more strongly in the hot channels than in average channels (in the 4-channel core model), resulting in a different flow distribution to yield a uniform pressure drop. The magnitude of this effect was investigated by performing two calculations at normal operating con-ditions (see Table 2.2), one with the Rohsenow-Clark model, and one case without.

The Rohsenow-Clark model adjusts the friction factor by the relation.

I l

I II

fRC -f L1.0

- +

p-w

(

N "bul k

) - 1 0)j (3-1)

Where:

Friction factor from Blasius relation Heated perimeter of channel P Wetted perimater of channel Fluid viscosity evaluated at bulk temperature uWALL = Fluid viscosity evaluated at rod surface temperature, TWALL, where TWALL is determined as

+ q WALL bulk where:

I Linear heat generation rate in the node

( btu/hr- ft) h = Surface heat transfer coefficient (btu/hr-ft - F) t l

P

It is clear from this formulation that the adjustment must be rela-tively small for cases within the normal range of wall superheat expected at BWR operating conditions, and the results. in Table 3.4 for the two cases show this quite clearly. There is no discernable effect on the hot channel conditions or the core pressure drop, and the MCPR for both cases is the same. The only significant differ-ence between the two cases is the slight increase in computer time required to evaluate the equation defining the Rohsenhow-Clark model. These results indicate that the default in VIPRE of neglect-ing the heated wall viscosity effect is reasonable for BWR analysis.

TABLE 3.4 ROHSENOW-CLARK MODEL SENSITIVITY STUDY Hot Channel Exit Core Flow Rate Void Pressure Computer Rohsenow-Clark Model MCPR (ibm/sec) Fraction ~II Time (SRU)

No 1.337 34.669 0.7862 24.11 16.854 Yes 1.337 34.668 0.7862 24.11 17.335 3.5 Local Pressure Option In a VIPRE calculation, fluid properties are normally calculated based on a reference (system) pressure which is the core exit pres-sure. This method assumes that the pressure drop in the system is small compared to the system pressure itself. This assumption may not be valid for cases with low system pressure or large pressure 0 drops. When using the local pressure option, fluid properties are I

il I

I l

'l

calculated at the average pressure at each axial level rather than at the uniform system (channel exit) pressure. The use 'of this op-tion results in a more accurate solution for analyses at low system pressure and high pressure drop conditions.

For normal BWR operating conditions, the core pressure drop is on the order of 15 to 25 psi, which is less than three percent of the nominal operating pressure of 1000 psia. The approximation error due to assuming a constant reference pressure should be very small.

To check this assumption, two cases were run, one with the local pressure option active, and one with constant reference pressure.

The results of these runs are summarized in Table 3.5, and show that there is virtually no difference between the two cases. There is no change in the exit temperature or void fraction, and only a 0.3 per-cent change in the core pressure drop. The t!CPR is identical for the two cases.

For most BWR applications, the core pressure drop is small compared to system pressure. One should not use local pressure option since it adds significantly (56 percent in this study) to the computation time. VIPRE-01, Volume 4 (see Reference 1), presents low pressure cases where the local pressure option is necessary, but these are not relevent to BWR analysis in normal operating ranges.

r 4

TABLE 3.5 LOCAL PRESSURE OPTION SENSITIVITY STUDY Hot Channel Exit Core P

Local 0 i lt Temp fF)

Flow Rate f1b/ I Void i ~fbi Pressure Computer i tSllt Yes 1.337 548.37 34.670 0.7862 24.05 26.307 No 1.337 548.37 34.669 0.7862 24.11 16.854 3.6 Numerical Solution Schemes VIPRE has two solution options which can be selected by user input.

The UPFLOW solution solves for the pressure gradient in each channel at an axial level through a combination of the axial and lateral momentum equations. The new channel pressure gradients are used to update the pressures. The new lateral pressure difference is then applied to the lateral momentum to yield the new crossflows. The new axial flows are computed with the continuity equation, using the flows at the last axial level and the new crossflows. In RFCIRC, tentative axial flows and crossflows are obtained at each level with the respective momentum equations, using information from the last iter ation. The tentative flows and pressures are then adjusted to satisfy continuity by a Newton-Raphson procedure.

RECIRC allows reverse and recirculating flow (i.e., instances where crossflows become large compared to axial flow). UPFLOW can be used only when flow is positive and predominantly in the axial direction.

I(

<r,

UPFLOW has two options: ITERATIVE and DIRECT.

In UPFLOW, a set of linear equations for energy at each axial level of the form (3-2) is established where [Aj is a nonsymmetric, banded, and diagonally dominant matrix with H X N elements where N is the number of chan-nels. A similar expression is used for pressure gradient. Using the DIRECT technique, h an zP are solved for in Equation 3-2 hy direct elimination (Gaussian elimination). In the ITERATIVE techni-que, a Gauss-Seidel inner iteration with successive over-relaxation is used to solve the equation for h and hP.

Hence, there are three solution schemes to choose from:

1. UPFLOW - ITERATIVE (or simply ITERATIVE).

2. UPFLOW - DIRECT (or DIRECT).

3. RECIRC.

For the present WHP-2 model, there is no crossflow, so in effect, the DIRECT and ITERATIVE schemes are the same and only the DIRECT option is used in this study. The PECIRC scheme is required when bypass regions, leakage paths, or water tubes are modeled.

Each solution scheme has an associated set of required input in the form of:

Maximum allowable number of external iterations.

Minimum allowable number of internal iterations.

Convergence criteria.

Damping/acceleration factors.

Results For nominal operating conditions in the WHP-2 core (see Table 2.2) are shown in Table 3.6. These calculations used the rec-ommended default input in VIPRE for the DIRECT and RECIRC options.

The different solution schemes yield the same results. The solution with the RECIRC scheme, however, took nearly 30 percent more compu-ter time, a result consistent with the run-time studies reported in Volume 4 of the VIPRE-02 documentation (see Reference 1). The DIRECT (UPFLOW) solution is the more economical of the two for 8WR analysis, if leakage paths and water tubes are not included in the model. (When these features are modeled, the RECIRC solution is the only option available.)

~t l

TABLE 3.6 SOLUTION METHOD SENSITIVITY STUDY Hot Channel Exit Core Temp. Flow Rate Void Pressure Computer Solution Method >ICPR ( F) (Ibm/sec) Fraction Drop ( si) Time (SRU)

Direct 1. 337 548.4 34.669 0.7862 24.11 16.892 Recirc 1.337 548.4 34.669 0.7862 24.11 21.912 3.7 Flow Correlation Sensitivity For BWR Core Models Two-phase flow effects are modeled in VIPRE-01 by emperical correla-tions for subcooled boiling, bulk void/quality relationship, and two-phase friction factor multiplier. These features are important in BWR analysis primarily because of their effect on friction and gravity head pressure losses, which can affect the inlet flow dis-tribution required to yield a uniform core pressure drop.

The correlations for two phase flow available in VIPRE all yield reasonable agreement with two-phase flow data at BWR operating con-

,ditions, as shown by data comparisons reported in Volume 4 of the VIPRE-01 documentation (see Reference 2). EPRI models for subcooled boiling, bulk void/quality, and two-phase friction factor multiplier were used in the sensitivity calculations in Sections 3.1 through 3.b. The purpose of this study is to evaluate the effect of the l

f

different options in VIPRE on the results of a BWR analysis, parti-cularly on the HCPR. Four cases were selected, using different com-binations of the subcooled boiling and void/quality relationships available in VIPRE. The correlation selection for each case is shown in Table 3.7.

Case 1 uses the EPRI void model, which is the default in VIPRE.

Case 2 uses the EPRI void/quality relation, but with homogeneous quality only, neglecting subcooled boiling effects. Case 3 utilizes the Levy model which is widely used in core thermal-hydraulic analy-sis in PWR applications. Case 4 uses an optional slip model with subcooled boiling effects accounted for by using the flowing quality predicted by Levy's model.

For all cases, the EPPI friction factor multiplier is used to model the effect of two-phase flow on friction pressure drop. There are four optional formulations in VIPRE to account for this phenomenon, but the data comparisons reported in Volume 4 of the VIPRE documen-tation (see Reference 1) show that the EPRI correlation is very accurate at BMR operating 'conditions, and is the recommended corre-lation for all ranges of application of the code.

The results for the four cases are shown in Table 3.8 . There are minor differences in hot channel exit flow and void fraction among these cases, and very small differences in core pressure drop. But the MCPR values obtained in the four cases are essentially the same.

Of particular interest is the extremely small difference in the re-suits for Case 1 and Case 2, where the only difference in the two cases is the absence of a subcooled boiling model in Case 2. This result indicates that subcooled boiling has a relatively insignifi-cant effect for BWR conditions. The behavior is dominated by bulk boiling, and a reasonable bulk void/quality relation will produce a reasonable flow field.

TABLE 3.7 FLOW CORRELATION SENSITIVITY STUDY -

SUMMARY

OF CASES Subcooled Bulk Void/ Two-Phase Case Boiling Quality Friction Factor Multi lier EPRI EPRI EPRI Hone EPRI F.PRI Levy Zuber-Findaly EPRI Levy SMIT EPRI TABLE 3.8 FLOW CORRELATION SENSITIVITY STUDY -

SUMMARY

OF RFSULTS Hot Channel Exit Core Flow Rate Void Pressure Computer Case MCPR (ibm/sec) Fraction Dro (psi) Time (SRU) 1.337 34.669 0.7862 24.11 16.854 1.337 34.672 0.7862 24.05 16.245 1.339 34.938 0.7863 23.88 14.989 1.339 34.900 0.7256 23.57 15.251

)j k

4.0 BENCHMARK STUDIES Calculations were made using the 4-channel full core model with YIPRE-01 to benchmark the code's results for steady state BWR analysis. These calculations confirm the accuracy and applicability of YIPRE to BWR an-alysis with comparisons to core pressure drop design data( ) provided by GE for WNP-2, to steady state FSAP. results, and to core follow analy-sis of WNP-2. For these comparisons to be meaningful, however, it was necessary to quantify the effect of variations in inlet orifice loss co-efficient on the overall results.

Yalues for the WHP-2 core analysis were obtained from the Process Compu-ter Data Bank (see Reference 7). To determine the effect of different orifice loss coefficients, a calculation was made with loss coefficients obtained from the FIBWR verification report (see Reference 8). Table 4.1 lists the two sets of loss coefficients. The operating conditions for these calculations were for nominal operation (see Table 2.2). The re-sults of the calculations are summarized in Table 4.2. The change in hot channel exit flow rate is less than one percent, even though the differ-ence between the central orifice loss coefficient is nearly five percent, and that for the peripheral orifices is almost 20 percent. The HCPR is changed by less than 0.1 percent which is insignicant.

These results indicate that reasonably close values of orifice loss co-efficients will produce essentially the same flow solution. Meaningful comparisons can be made between calculations with somewhat different ori-fice loss coefficients.

I, 4.1 Core Pressure Drop Com arisons The core pressure drop data for WNP-2 provided by GE was obtained using the GE proprietary code ISCORE. This code is a parallel flow path computer program for steady state BWR reactor core thermal hy-draulic analysis, and has been used in design and licensing calcula-tions of all GE Class 3 through 6 BWR plants. The data was provided for 23 cases, as core pressure drop as a function of rated power and flow, with no other boundary conditions defined.

Calculations were made with VIPRE-Ol using the 4-channel full core model for the 23 cases, assuming a range of system pressures of 900, 1000, and 1050 psia, and inlet subcooling of 20 and 30 btu/ibm for each case. As one might expect, the results for a given case were very similar, regardless of the pressure or inlet subcooling. The VIPPE results obtained at the nominal conditions of 1000 psia, and 20 btu/ibm subcooling are compared to the GE calculations for each case in Table 4.2. The VIPRE results are in very good agreement with the GE calculations for the full range of power and flow rates given. Figure 4.1 shows these results graphically, with a plot of core pressure drop as a function of rated power for various rated flow rates. The VIPRE results fall neatly on the lines of the GE calculations.

TABLE 4.1 LOCAL LOSS COEFFIEHCIENTS UNCERTAINTY STUDY -

SUMMARY

OF CASES Orifice Loss Coefficient Center Peripheral Case Assembly Assembly Source of Data 29.25 213.52 Process Computer Data Bank 30.77 170.99 FPRI Report NP-1923 TABLE 4.2 LOCAL LOSS COEFFICIENT UNCERTANITY STUDY -

SUMMARY

OF RESULTS Hot Channel Exit Core Flow Rate Void Pressure Computer Case MCPR (ibm/sec) Fraction Drop (psi) Time (SRU) 1.337 34.669 0.7862 24.11 16.854 1.336 34.460 0.7874 24.26 17.248 TABLE 4.3 CORE PRESSURE DROPS FOR MANFORD .2 CORE AT VARIOUS POWERS AHP FLOW RATES Core Core Core Power Core F1ow Pressure Pressure

( X Rated) (/ Rated) Drop (PSID) Drop (PSID)

VIPRE 100.0 120.0 32.34 31.38 100.0 100.0 24.74 24.40 100.0 75.0 16.66 16.82 100.0 50.0 10.10 10.70 75.0 100.0 22.84 22.60 50.0 100.0 21.01 20.89 25.0 100.0 19.32 19.45 0.0 100.0 17.41 19.14 80.0 100.0 23.22 22.95 60.0 100.0 21.72 21.56 40.0 100.0 20.29 20.25 100.0 80.0 18.15 18.20 80.0 80.n 16.98 17.04 60.0 80.0 15.80 15.93 40.0 80.0 14.71 14.95 100.n 60.0 12.54 12.99 80.0 60.0 11.73 12.07 60.0 60.0 10.90 11.25 40.0 60.0 10.14 10.55 100.0 40.0 7.87 8.66 80.0 40.0 7.46 7.98 60.0 40.0 7.00 7.41 40.0 40.0 6.58 7.00 l

I 1

ir t

f f

'E CORE PRESSUPE DROP (PSID>

38.

25

%+4+ VIPRE 28

'8,

18. 28. 38. 48. SB. 68. 78. 88. 98. 188.

CORE POWER (i RATED)

Figure 4.1 I

I I

1 r

4.2 FSAR Comparison The void and quality axial distributions for the core average and the hot channel are reported in the FSAR for WNP-2 under normal operating conditions. The radial distribution of core flow is also reported, along with the core pressure drops. The core conditions, including axial and radial power distributions, for which these re-sults were obtained are listed in Table 4.4. (These conditions are from Chapter 4 of Reference 2.)

VIPRE calculations for these conditions were made using the 4-channel full core model for the operating conditions given in Table 4.4. Table 4.5 compares the core flow distribution calculated in VIPRE with the FSAR results. The VIPRE results are within a few percent of the FSAR results, in all regions of the core. Table 4.6 compares the pressure drop values calculated in VIPPE to the FSAR values, and the average pressure drop across each type of orifice is essentially the same.

The axial distributions of quality and void calculated in YIPRE for the core average and for the hot channel are compared to the FSAR results in Table 4.7 and 4.8. The comparisons are also shown graph-ically for the core average in Figures 4.2 and 4.3. The VIPRE re-sults are in very good agreement with the FSAR'alculations.

ti 4.3 Core Follow Comparison Core follow analysis for QfP-2 is performed using SIMULATE-2, a three dimensional steady state nodal analysis program. The output from SIMULATE-2 can be used to define nodal powers and R-factors for various rated flow and power conditions in the core, representing different state points throughout the initial fuel cycle. This in-formation can be used to generate input for VIPRE to benchmark the VIPRE code against the thermal hydraulic solution produced by SIMULATE-2.

A total of 21 cases were selected from the SIMULATE-2 analysis for comparison with VIPRE-01. These cases represented various power and flow conditions, with pressure varying from 950 to 1020 psia and inlet subcooling varying from 18 to 32 btu/ibm. VIPRE calculations were performed for these 21 cases using the 191-channel quarter core model. This model was used, rather than the simplier 4-channel core model, because it has the same channel layout as the SIMULATE-2 core model, which greatly simplified the task of generating the VIPRE input.

'The results of the VIPRE-01 calculations are compared to the SIMULATE-2 calculations in Table 4.9 which shows core pressure drop, MCPR, hot channel flow rate, and core average void fraction for each case.

I

'I(

J I;

L t

C I

The core pressure drop values obtained with the two codes are essen-tially identical, and the hlCPR values are within two percent of each other for nearly all cases. The largest differences occur in cases with flow at approximately 50 percent of the rated value. The hot bundles flow predicted by both codes is very close for each case, the lar gest difference again occurring in cases with approximately 50 percent rated flow, which is probably the main cause of the same trend observed in the MCPR comparison. The core average void frac-tion is essentially the same for the two calculations in each case.

This comparison indicates that the thermal hydraulic solution in VIPRE is at least as good as that in SIh<ljLATE-2, and will produce similar results for similar conditions.

jl II I

TABLE 4.4 CORE CONDITIONS FOR FSAR STEADY STATE ANALYSIS Parameter Value Core Power 3323 h1Wt Active Core Flow Rate 26750.0 1 bm/sec Inlet Subcool ing 20.18 btu/1 bm System Pressure 1035.0 psia Radial Power Distribution Hot Central Assembly 1.400 Hot Peripheral Assembly 0.950 Average Central Assembly 1.040 Average Peripheral Assembly 0.700 AXIAL POWER DISTRIBUTION Node Axial Power Factor Bottom of Core ~

] 0.38 2 0.69 3 0.93 4 1.10 5 1.21 6 1.30 7 1.47 8 1.51 9 1.49 10 1.44 11 1.36 12 1.28 13 1.16 14 1.06 15 1.01 16 0.97 17 0.94 18 0.97 19 0.96 20 0.91 21 0.77 22 0.59 23 0.38 Top of Core 24 0.12

TABLE 4.5 CORE FLOW DISTRIBUTION Orifice Zone Description Relative ,en ra en ra enp era erip era Assembly Flow Mot ~Avera e Mot Average FSAR 0.928 1.059 0.547 0.572 VIPRE 0.987 1.060 0 qaO 0.563 TABLE 4.6 CORE PRESSURE DROP FSAR YIPRF.

Total Core Pressure Drop, psi 24.74 24.11 Average Orifice Pressure Drop Central Region, psi 6.03 7.96 Peripheral Region, psi 16.54 16.54 I

)

I i

TABLE 4.7 FLOW (}UALITY OISTRIBUTION Core Average Maximum Channel (Average ttode Value) (End of trode Value)

Hode FSAR VIPRE FSAR VIPRE Bottom 1 0.000 0.000 0.000 0.000 2 0.000 0.000 0.000 0.004 3 0.000 0.000 0.003 0.014 4 0.001 0.000 0.012 0.026 5 0.004 0.001 0.026 0.040 6 0.010 0.006 0.042 0.057 7 0.019 0.017 0.061 0.075 8 0.029 0.028 0.081 0.093 9 0.040 0.039 0.100 0.111 10 0.051 0.050 0.119 0.129 11 0.062 0.061 0.137 0.145 12 0.072 0.071 0.153 0.161 13 0.080 0.080 0.169 0.175 14 0.089 0.088 0.182 0.188 15 0.097 0.096 0.196 0.202 16 0.105 0.104 0.208 0.214 17 0. 112 0.111 0.221 0.226 18 0.119 0.118 0.233 0.239 19 0.126 0.126 0.246 0.251 20 0.134 0.133 0.258 0.262 21 0.140 0.139 0.268 0.271 22 0.145 0.144 0.275 0.277 23 0.149 0.148 0.280 0.281 Top 24 0.151 0.150 0.282 0.281

TABLE 4.8 VOID DISTRIBUTION Core Average maximum Channel (Avera e Node Value) (End of Node Value)

Node FSAR VIPRE FSAR VIPRE Bottom 0.000 0.000 '.000 0.006 0.000 0.000 0.007 0.072 0.007 0.005 0.073 0.188 0.041 0.052 0.183 0.290 5 0.104 0.137 0.287 0.371 6 0.179 0.215 0.371 0.440 7 0.254 0.285 0.443 0.498 8 0.323 0.347 0. 500 0.543 9 0.379 0.394 0.545 0.580 10 0.425 0.433 0.582 0.610 11 0.462 0.467 0.611 0.635 12 0.492 0.493 0.636 0.656 13'4 0.517 0.517 0.655 0.674 0.537 0.536 0.672 0.692 15 0.555 0.553 0.686 0.703 16 0.571 0.568 0.699 0.719 17 0.585 0.582 0.711 0.729 18 0.598 0.595 0.722 0.737 19 0.610 0.607 0.733 0.747 20 0.622 0.619 0.742 0.756 21 0.631 0.628 0.750 0.763 22 0.638 0.636 0.756 0.767 23 0.644 0.641 0.759 0.770 Top 24 0.646 0.643 0.760 0.770 42-

TABLE 4.9 CORE FOLLOll DATA COMPARISON Core Pressure Core Average Power Fl ow Drop (psid) MCPR Bundle Flow (lb/sec) Void Fraction Case (%~ated) (X~ated j S IMa QC~ SIM QC zCPR/CPRd SIN QC zF/Flowe SIM QC 1 100.0 100.0 22.31 22.12 1.40 1.42 1.2% 34.47 35. 33 2.4% 0.42 0.42 2 62.0 88.0 17.05 17.34 2.22 2.25 1 . 5'X 30.72 31.39 2.1% 0.36 0.37 3 75.2 94.4 19.21 19.55 1.82 1.85 1.97> 32.51 34

'3.

2.5% 0.36 0.36 76.0 99.0 20.41 20.92 1.86 1.89 1.8% 34.46 35.31 2.4% 0.35 0.35 5 68.6 59.7 11.06 10.63 1.83 1.87 2. 5% 20.41 21.06 3.1% 0.48 0.49 6 91.1 83.1 17.77 17.33 1.53 1.56 2. 0% 28e12 28.95 2.8% 0.48 0.49 7 86.7 98.0 21.27 21.42 1.57 1.60 1.8% 33.04 34.07 3.0% 0.41 0.41 8 97.2 99. 5 22.53 23.07 1.48 1.48 0.07% 34.06 34.89 2.4% 0.45 0.48 9 99.5 100.0 22.77 22.83 1.44 1.46 1.4% 34.17 35.00 2.4%, 0.45 0.46 10 95.5 98. 5 22.22 22.27 1.49 1.51 1.4'X 33.62 34.49 2.5% 0.46 0.47 ll 48.8 50.0 8.56 8.41 2.22 2.31 3.9X 17.67 18.24 3.2X 0.39 0.39 12 98.1 98.2 21.84 21.89 1.44 1.46 1.6'f. 33.41 34.26 2.5% 0.43 0.44 13 100.2 98.8 22.24 22.26 1.37 1.39 1.9% 33.30 34.25 2.8% 0.44 0.45 14 100.0 97.2 22.04 21.98 1.40 1.43 1.5'f. 32.85 33.85 3.0% 0.47 0.47 15 99.8 95.7 21.17 21.05 1.46 1.48 1.6% 32.80 33.67 2.6% 0.43 0.44 16 96.1 88.9 19.35 19.03 1.47 1.50 1.8'X 30.05 31.09 3. 3'X 0.46 0.47 17 57.7 32.6 5.80 5.55 1.65 1.71 3. 5'X 11.53 11.94 3.4% 0.46 0.46 18 70.4 53.0 9.40 9.08 1.62 1.68 3.1% 18.09 18.86 4.1% 0.39 0.38 19 -

71.8 52.1 9.41 8.98 1.60 1.66 3. 9'X 17.58 18.45 4.8'X 0.43 0.42 20 71.4 50.7 9.06 8.69 1.60 1.67 4. 0% 17.18 17.96 4.3% 0.42 0.41 21 71.6 50.7 9.12 8.71 1.57 1.64 4.0'X 17.02 17.82 4.5% 0.45 0.44 NOTES:

'a ~ SIMULATE-2 results.

b; VIPRE Quarter Core Model results.

c~ Bundle flow in the most limiting bundle.

d. CPR(QC) CPR(SIN)

e. Flow(QC) - Flow(SIM) ow

CORE AVERAGE FLOW QUALITY DISTRIBUTION

~ FI OM QUALITY 8 28

~

8,15 8.18 8.85 2 3 4 5 6 7 8 9 18 fi 12 13 14 15 16 17 18 19 28 21 22 23 24 AXIAL NODE Figure 4.2 CORE AVERAGE VOID DISTRIBUTION VOID FRACTION FSAR gyp PRE

8. 88 8,68 8.28 1 2 3 4 5 6 7 8 9 18 li 12 13 14 15 16 17 8 19 28 21 22 23 24 AXIAL NODE Figure 4.3 44

I 5.0 FSAR TRANSIENT CALCULATIONl COMPARISON A major BWR application of YIPRE-Ol is to calculate the transient hCPR.

The transient sCPR is defined as the difference between the initial CPR and the minimum CPR during transient. For each fuel cycle, the limiting transient aCPR (i.e., the maximum transient dCPR) is used to determine the MCPR operating limi,t (i.e., the MCPR operating limit = MCPR safety limit + the limiting transient hCPR). The results of calculations per-formed to benchmark the VIPRE-01 code with WNP-2 FSAR Chapter 15, aCPR data are presented in this section. Six transients are analyzed with the 4-channel full core model. They are:

1. Loss of feedwater heater, manual control.

2. Feedwater controller failure, maximum'emand.

3. Generator load rejection, bypass-on.

4. Generator load rejection, bypass-off.

5. Turbine trip, bypass-on.

6. Turbine trip, bypass-off.

CPR is evaluated with the GEXL correlation. Initial conditions and forc-ing functions (system pressure, inlet enthalpy, core flow, and core power versus time) were obtained from tables and figures in the FSAR. They are

t v

shown in Table 5.1 and Figures 5.1 through 5.6. An R-factor of 1.05 and a chopped cosine axial power shape with a peak/average ratio of 1.4 is assumed in these calculations. (The R factor and axial power profile used in GE calculations were not provided in the FSAR.) For each trans-ient, the hot channel initial power level is adjusted so that the minimum CPR during the transient equals the safety limit, 1.06 . A comparison of aCPR computed with VIPRE-01 and the GE ODYN Code is presented in Table 5.2. These comparison show that VIPRE-01 predictions of BWR fuel bundle aCPR are conservative and thus acceptable for BWR licensing analyses.

TABLE 5.1 INITIAL CONDITIONS FOR TRAHSIEHTS Thermal Power Level, MWt 3468.0 Core Flow, lbs/hr 108.5 X 106 Vessel Core Pressure, psig 1031.0 Core Coolant Inlet Enthalpy, Btu/lb 529.3 Core Leakage Flow, / 11.84 46

TRANSlENT FORCING FUNCTIONS RELATIVE VALUE, 1.2 0.8 0.6 POWER 0.4 PRESSURE INLET ENTHALPY 0.2 0.0 10 20 30 40 50 60 70 TIME (SEC)

Fi gure 5.1 Loss of Feedwater Heater, Manual Control TRANSlENT FORClNG FUNCTlONS 1.2

'D 1.0 0.9 K POWER 0.8 FLOW PRESSURE 0.? INLET ENTHALPY 0.6 16 18 8 10 12 14 TIME (SEC)

Figure 5.2 Feedwater Controller Failure, Maximum Demand II 1

f l l

II I

I

TRANSIENT FORCING FUNCTIONS 1.20 1.15 1.10 1.05 1.00 0.95 POWER 0.90 FLOW PRESSURE 0.85 0.80 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 TIME (SEC)

Figure 5.3 Generator Load Rejection, Bypass-On TRANSIENT FORCING FUNCTIONS 1.20 1.15 1.10 1.05 1.00 K

0.95 0.90 POWER FLOW PRESSURE 0.85 INLET ENTHALPY 0.80 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 TIME (SEC)

Figure 5.4 Generator Load Rejection, Bypass-Off

1.00 0.95

) 090 g 0.85

~ 0.80 0.75 POWER FLOW PRESSURE 0.70 INLET ENTHALPY 0.65 0.60 00 0.3 0.6 0.9 1.2 1.5 1.8 2.4 2.7 3.0 TIME (SEC)

Figure 5.5 Turbine Trip, Bypass-On TRANSIENT FORCING FUNCTIONS 1.15 1.10 1.05 1.00 0.95 0.90 POWER FLOW 0.85 PRESSURE 0.80 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

'IME (SEC)

Figure 5.6 Turbine Trip, Bypass-Off

TABLE 5.2 TRANSIENT ANALYSIS RESULTS gCPR(l)

Transient Loss of Feedwater Heater, Manual Control 0.16 0.259 Feedwater Controller Failure, Maximum Demand 0.08 0.104 Generator Load Rejection, Bypass-On 0.04 0.075 Generator Load Rejection, Bypass-Off 0.09 0.122 Turbine Trip, Bypass-On 0.06 0.085 Turbine Trip, Bypass-Off 0.08 0.105

~ ~aCPR; =- Initial

~ ~

CPR - ~ ~ ~

Minimum CPR" During Transient

A ll I

h h

5.1 Time Step Sensitivity Time step sensitivity study reported in VIPRE-01, /olume 4 (Refer-ence I), shows that subcooled boiling models (LEVY and EPRI) in VIPRE are not suitable in boiling transients where the courant num-ber is less than unity. Thus, all the preceding transient runs were made with the EPRI void model without subcooled boiling. Time step sizes were selected to follow the forcing functions.

To check the sensitivity of VIPRE-01 to time step, the Generator Load, Rejection, Bypass-On transient was run with several different time step sizes. Since the changes in forcing functions occur over 0.1 second, that is the largest time step that can be used to follow the forcing functions accurately. Time steps of 0.2, 0.1, and 0.05 seconds w'ere used to cover a reasonable range for this parameter MCPR values versus time for each time step are shown on Table 5.3.

Note that the transient MCPR values for 0.1 and 0.05 seconds time steps are virtually identical. However, the 0.2 time step misses the minimum point during transient and is unacceptably nonconserva-tive. The results indicated that time step size equal or less than the largest time step that can follow the forcing function accu-rately should be selected. This is consistent with the results pre-sented in VIPRE-01, Volume 4 (Reference I).

4 TABLE 5.3 TIME STEP SIZE SENSITIVITY STUDY GENERATOR LOAD REJECTIOH, BYPASS-OH MCPR With GEXL Correlation Time (Sec) dt = 0.2 Sec at = 0.1 Sec = 0.05 Sec 0.0 1.136 1.136 1.136 0.05 1.136 0.1 1.136 1.136 0.15 1.136 0.2 1. 136 1.136 1.136 0.25 1.136 0.3 1. 136 1.136 0.35 1.119 0.4 1. 112 1.103 1.105 0.45 1.113 0.5 1.123 1.122 0.55 1,123 0.6 1.128 1.126 1.125 0.65 1.126 0.7 1.126 1.122 0.75 1.118 0.8 1.114 1.115 1.114 0.85 1.118 0 9 1.113 1.112 0.95 1.107.

1.0 1.099 1.101 1.101 1.05 1.072 1.1 1.060 1.061 1.15 1.068 1.2 1.073 ,1.075 1.076 1.25 1.079 1.3 1.081 1.083 1.35 1.086 1.4 1.086 1.088 1.089 1.45 1.094 1.5 1.103 1.103 1.55 1.114 1.6 1.122 1.123 1.123 1.65 1.125 1.7 1.127 1.126 1.75 1.126 1.8 1.131 1.130 1,131 C

6 .0 CONCLUSION In evaluating the VIPRE-01 Code's capacity for modeling BWR cores, a full core, 4-channel model was developed for Washington Public Power Supply System Nuclear Plant No. 2 (WNP-2) reactor core. This development in-cludes the noding sensitivity study and a series of generic sensitivity studies testing VIPRE options, correlations, and input parameters.

Benchmark studies were performed to confirm the accuracy and adequacy of VIPRE-Ol for BWR thermal-hydraulic analyses. These studies showed that the core pressure drop, void, quality, and flow distribution, MCPR, and transient sCPR calculated by VIPRE-Ol were in good agreement with vendor, FSAR, and SIMULATE-2 Code results. On the basis of this evaluation, it is concluded that VIPRE-01 is acceptable for BWR licensing calculations.

4 l[

II

7.0 REFERENCES

1. VIPRF-Ol, Computer Code Manuals, FPRI HP-2511-CCM, Volumes 1 Through 4, Revision 2, July 1985

2. D. S. Rowe. COBRA-IIC: A Di ital Computer Program For Steady State and Transient Thermal Hydraulic Analysis of Rod Bundle Nuclear Fuel Elements. Richland, Washington: Pacific Northwest Laboratory, March 1973. BHWL-1695

3. T. L. George, et al. COBRA-WC: A Verson of COBRA For Single Phase Multi-Assembly Thermal Hydraulic Transient Analysis. Richland, Washington: Pacific Northwest Laboratory, July 1980. PNL-3259

4. C. L. Wheeler, et al. COBRA-IV-I: An Interum Version of COBRA For Thermal llydraulic Analysis of Rod Bundle Nuclear Fuel Elements and Cores. Richland, 'Washington: Pacific Northwest Laboratory, Inarch 1973. BNWL-1962

5. Safety Evaluation Report on EPRI NP-2511-CCM VIPRF.-01 For PWR Application

6. WPPSS Nuclear Project No. 2 Final Safety Analysis Report

7. Process Computer Data Bank, 23A1372

8. FIBWR Verification and (}ualification Peport, EPRI HP- 923, July 1981

<i l

I f

I

9. SIMULATE-2 Computer Code User's Manual, EPRI RP 710-1, July 1982

10. General Electric Thermal Analysis Basis (GETAB):. Data, Correlation, and Design Application, General Electric Company, November 1973 (NEDO-10958)

11. RETRAN-02, "A Program For Transient Thermal-Hydraulic Analysis of Complex Fluid Flow Systems," EPRI NP-1850, Volume 1, May 1981

12. Letter, B. F. Rubin to David Chan, "Core Pressure Drop Calculations at Various Powers and Flow Rates," GEWP-2-84-5134, September 18, 1984

1

[

Jl

~

<s VERIFICATION OF VIPRE-01 FOR BWR ANALYSIS SUPPLEMENT 1 (01/19/90)

Analysis in this report were performed with VIPRE-01 NOD01. However; the conclusion is also applicable to VIPRE-01 MOD02 since the changes made in VIPRE-01 NOD02 had no effect on the calculations. This was confirmed by repeating selected calculations with YIPPEE-Ol NOD02 and identical results were found.

4 Oi

)