ML20134Q271

Provides Supplemental Guidance on Benchmarking Vipre Code to Validate Implementation & User Application of Changes to Accomodate Use of Lithium Burnable Poison Rods for Production of Tritium

ML20134Q271
Person / Time
Issue date:	02/24/1997
From:	Martin T NRC (Affiliation Not Assigned)
To:	Sohinki S ENERGY, DEPT. OF
References
PROJECT-697 NUDOCS 9702260361
Download: ML20134Q271 (4)

Text

{{#Wiki_filter:l 9 tioq

1. g UNITED STATES Y

S NUCLEAR REGULATORY COMMISSION i b f WASHINGTON, D.C. 20666 4001 \\*****/ February 24, 1997 i l Stephen M. Schinki, Director Office of Commercial Light Water Reactor Production Defense Programs Department of Energy Washington, DC 20585

SUBJECT:

SUPPLEMENTAL GUIDANCE ON BENCHMARKING THE VIPRE CODE TO VALIDATE T IMPLEMENTATION AND USER APPLICATION OF CHANGES TO ACCOMMODATE THE USE OF LITHIUM BURNABLE POISON RODS FOR THE PRODUCTION OF TRITIUM

Dear Mr. Sohinki:

By letter dated February 4, 1997, the staff transmitted the NRC safety evaluations which provide the conditions for acceptable use of the VIPRE-01 thermal hydraulic code to you for guidance. Upon review of your letter dated February 7, 1997, the staff concludes that the use of the VIPRE code needs to ) be further justified and benchmarked for the specific use described in the your " Report on the Evaluation of the Tritium Producing Burnable Absorber Rod Lead Test Assembly." In particular, the staff requests that detail on the modeling assumptions, choice of flow models and correlations, heat transfer correlations, CHF correlation aM DNBR limit be presented and justified for the spec:fied use. In addition, benchmarking of the VIPRE results to other NRC-approved codes should be included. i contains the NRC Safety Evaluation for use of the VIPRE-01 code at i %.abrook, " Acceptance for Referencing of YAEC-1849P, ' Thermal-Hydraulic Analysis Methodology Using VIPRE-01 for PWR Applications,' for the Seabrook Station, Unit No. 1," dated August 15, 1994. This should provide more guidance regarding the information needed for review of the use of the VIPRE code for PWR licensing applications. In addition, Enclosure 2 contains the summary and abstract of NUREG/CR-4394, " Rod Bundle Filns Boiling and Steam Cooling Data Base and Correlation Evaluation," which discusses the conservatisms of several commonly used film boiling and pure vapor heat transfer coefficients which may also be of assistance in the evaluation, gTDb I 1 pgol W .-n 9702260361 970224 l ea; PROJ PDR @ 'fl r MC FR.E CENTER COM

9 i- ' February 24, 1997 If you have any questions regarding this information, please contact the project manager, J. H. Wilson, at (301) 415-1108. Sincerely, David B. Matthews/for Thomas T. Martin, Director Division of Reactor Program Management Office of Nuclear Reactor Regulation Project No. 697

Enclosures:

as stated DISTRIBUTION: Central File PGEB R/F AThadani ACRS FGillespie Glainas PUBLIC FMiraglia BSheron JMitchell RMartin KRapp, RII KKavanagh LPhillips RArchitzel GHolahan BBoger MVirgilio DOCUMENT NAME: P:\\TRTVIPRE.INF .i Os/ OFFICE PM:PGEB:DRPM C:SRX h C:PGEB lh D: DRPH /// / NAME JHWilson:s h 3 JLyons' DMatthe # TMart V' f.,/\\ DATE 2//f/97 '27" ~ 2/1,6 /97 2/M/97 2/14 /97f, 0FFICIAL RECORD COPY 1 e fj@ % EON l

. February 24, 1997 If you have any questions regarding this information, please contact the project manager, J. H. Wilson, at (301) 415-1108. Sincerely, David B. Matthews/for i Thomas T. Martin, Director 1 Division of Reactor Program Management Office of Nuclear Reactor Regulation Project No. 697

Enclosures:

as stated DISTRIBUTION: Central File PGEB R/F AThadani ACRS FGillespie Glainas PUBLIC FMiraglia BSheron JMitchell RMartin KRapp, RII KKavanagh LPhillips RArchitzel GHolahan BBoger MVirgilio DOCUMENT NAME: P:\\TRTVIPRE.INF ,f I~) s ) 0FFICE PM:PGEB:DRPM C:SRX63 C:PGEB 'ili D:DRPl t /// [ NAME JHWilson:s h > JLyons' DMattbedf TMart W LA DATE 2//f/97 'F" 2/p/97 2/%i/97 2/f /97 ( 0FFICIAL RECORD COPY i l l l 4 J l

' February 24, 1997 If you have any questions regarding this information, please contact the project manager, J. H. Wilson, at (301) 415-1108. Sincerely, Thomas T. Martin, Director Division of Reactor Program Management Office of Nuclear Reactor Regulation Project No. 697 i

Enclosures:

as stated cc: see next page l l l i i I i l I

l g W. p 't l UNITED STATES

-{

g j NUCLEA' R P.EGULATORY COMMISSION Q,, t WA$HINGtoN. D.C. 30665-4001 ,\\ *...,/, l August 15, 1994 3 l Docket No. 50-443 l Serial No. SEA-94-021 Mr. Ted C. Feigenbaum Senior Vice President and Chief Nuclear Officer North Atlantic Energy Service Corporation l Post Office Box 300 Seabrook, New Hampshire 03874

SUBJECT:

ACCEPTANCE FOR REFERENCING OF YAEC-1849P, " THERMAL-HYORAULIC ANALYSIS METH000LdGY USING VIPRE-01 FOR PWR APPLICATIONS", FOR THE SEABROOK STATION,' UNIT NO. 1 (TAC M86958) On February 2, 1993, North Atlantic Energy Service Corporation (North i Atlantic) submitted for review three proprietary reports (YAEC-1849P, YAEC-1854P, and YAEC-1856P) relating to core reload analyses methodologies. j North Atlantic has proposed to apply the methodologies described in these reports to support future operations at Seabrook Station, Unit No.1 (Seabrook). We have completed our review of YAEC-1849P which describes the methodology for core thermal-hydraulics analyses based on the VIPRE-01 code. The staff finds YAEC-1849P acceptable for referencing in licensing applications for Seabrook to the extent specified and under the limitations stated in YAEC-1849P and the enclosed Nuclear Regulatory, Commission safety evaluation. The enclosed safety evaluation defines the basis for accepting YAEC-1849P for application to Seabrook. I If the NRC staff's criteria or regulations change such that the conclusions about the acceptability of the YAEC-1849P are invalidated, Yankee Atomic Electric Company (Yankee) should revise and resubmit the respective documentation or a justification should be submitted for the continued effective applicability of YAEC-1849P without revision. The use of YAEC-1856P was approved previously (letter A. W. De Agazio to T. C. Feigenbaum, August 8, 1994). Our review of YAEC-1854P will be discussed in other correspondence. NRC FILE CE!HER COPY n 'unaiegrOSOOOH390 noe S 1

DR A

k d) ( PDR '\\\\

G Mr. Ted C. Feigenbaum August 15, 1994 The staff was assisted in this review by International Technical Services g (ITS) Inc. under Contract No NRC-03-90-027, [ JCN No. L1318, Task Order No, 021. Our safety evaluation (Enclosure 1) is based on the ITS Technical Evaluation Report (ITS/NRC/94-1) (Enclosure 2). j Sincerely, 1 Albert W. De Agazio, S Project Manager Project Directorate I-4 1 Division of Reactor Projects - !/11 Office of Nuclear Reactor Regulation

Enclosures:

l 1. Safety Evaluation i i ( 2. ITS Technical Evaluation Report t l cc w/ enclosures: 1 See next page } i i i ) i

= ...~ f ('.

1. 40p k

UNITED STATES 2 NUCLEAR REGULATORY COMMISSION f WASHINGf oN. D.C. 30eu 4001 \\.....,/ 'i y 1 ?p 3, SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION h RELATED TO THE REVIEW AND APPROVAL OF YAEC-1849P - A THERMAL-HYDRAULIC O ANALYSIS METHODOLOGY USING VIPRE-01 FOR PWR APPLICATIONS o { j!I EQRTH ATLANTIC ENERGY SERVICE CORPORATION. ET Al SEABROOK STATION. UNIT NO. 1 ~ DOCKET NO. 50-443 a j ? i \\

1.0 INTRODUCTION

I l OnFebruary2,1993,Nort5AtlanticEnergyServiceCorporation(North j Atlantic) submitted for review (Ref.1) Yankee Atomic Electric Company i (Yankee) report YAEC-18499 (Ref. 2). Additional supporting information was i submitted on January 14, 1994 (Ref. 3). References 2 and 3 contain Yankee i documentation related to the development of a core thermal-hydraulic l methodology using the VIPRE-01 computer code (Ref. 4). The VIPRE-01 code j would replace the COBRA-!!!C code in an NRC-approved methodology for the Yankee Atomic Power Station and the Maine Yankee Atomic Power Station (Ref. 5, 6, 7, 8). North Atlantic intends to use VIPRE-01 for the Seabrook Station, Unit No. 1 (Seabrook) Departure from Nucleate Bolling (DNB) analysis. In the methodology submitted by North Atlantic, uncertainties are applied to the DNBR limit calculations for Seabrook'using a statistical, rather than a deterministic, method by adapting the Westinghouse Revised Thermal Design Procedure (RTDP) (Ref. 9) as part of the DNB design basis approach. The RTDP method is a thermal-hydraulic analysis technique which computes DNB margin by statistically combining associated uncertainties. The stated objectives of the YAEC-1849P are to: provide a description of the extension of NRC-approved Yankee e subchannel analysis methodology to the VIPRE-01 code for application to Seabrook, and 5 document the development of a fuel-design-specific WRB-1 correlation DNBR limit based on RTDP methodology to be used for Seabrook applications. 4 40steOsas_,40s 3 W hDR ADOCK 05000443 PDR EW

}t.*,-

.p

' I 1 Towards these goals, the submitted documentation presents a plant specific geometric representation of the core, a selection of thermal-hydraulic models and correlations, and a description of the RTOP methodology using VIPRE-01 for g Seabrook. l Since this is the first submittal based on use of the Yankee VIPRE-01 computer code, the review was also' performed to assure fulfillment of VIPRE-01 safety evaluation report requirements (Ref. 4) and to assure conformity to the RTOP 2 (Ref. 9) requirements. I s l 2.0

SUMMARY

OF YAEC-1489P \\ References 2 and 3 document descriptions of Yankee's VIPRE-01-based subchannel analysis methodology in support of core reloads for Seabrook and Maine Yankee. The 56 Ject methodology is an extension of the NRC approved methodology using COBRA-IllC for use in Yankee and Maine Yankee applications. Descriptions are provided of plant specific core models, selections of thermal-hydraulic correlations, and Seabrook specific application of the Westinghouse RTOP 4 methodology for determination of a safety limit on DNB. YAEC-1849P discusses applications of the methodology to Seabrook and other nuclear power plants. This SE is limit =J to the application of the methodology to Seabrook. l Application to other facilities will be discussed elsewhere. 2.1 VIPRE-01 Comouter Code t VIPRE-01hasbeenpreviofslyreviewedandapprovedforapplicationto pressurized water reactors (PWR) in steady-state and transient analyses with heat transfer regimes up to the critical heat flux (CHF). The NRC safety i evaluation (SE) on VIPRE-01 (Ref. 4) includes conditions requiring each user to document and submit to the NRC for approval its procedure for using VIPRE-01 and to ;rovide justifications for the specific modeling assumptions made, the choice of particular two-phase flow models and correlations, heat transfer correlations, CHF correlation and DNBR limit, and input values of plant specific data such as turbulent mixing coefficient and grid loss coefficient including defaults. In References 2 and 3, the selection of the core model, use of certain thermal-hydraulic correlations, and other key input selections were described and justified. Details related to model selection are provided in Section 3.3 of this SE. 2.2 Revised Thermal Desion Procedure Methodolooy TheRIDPwasdevelopedbyWestinghousetocombinhthesystemandcorrelation i uncertainties associated with the statistical vice deterministic prediction of ONB. The RTDP methodology was approved by the NRC in 1989 with certain restrictions and is documented in WCAP-ll397-P-A. These restrictions require the user to provide justification for any changes in DNB correlation, THINC-IV i 1 i correlations, or parameter values listed in WCAP-ll397-P-A that are outside of previously demonstrated acceptable ranges. t 4

1..

Y

g l 3 n

l 1

3.0 EVALUATION I 3.1 VIPRE Model Descriotion VIPRE-01 correlations, in luding CHF correlations planned for use by Yankee for Seabrook is summarized in Table 3.3.1 of YAEC-1849P. The correlation selection is ides.tical to that used with the COBRA-IllC code which was reviewed and approved previously by the NRC, except for the turbulent mixing coefficients and CHF correlation. 1 Selection of each model or correlation is based upon one of the following considerations: (i) selection of some correlations /models are known to be insensitive to the Departure from Nucleate Boiling Ratio (DNBR) results; (ii) the selection is' consistent with the current Final Saftty Analysis Report (FSAR) methodology; (iii) the applicability range is appropriate; or (iv) sensitivity studies have demonstrated that some correlation selections were more conservative than those recommended in the VIPRE-01 manuals. l Justification of the adequacy of the overall modeling approach was provided through benchmark calculations. This approach is acceptable to the staff. 3.1.1 Turbulent Mixino ' The lateral momentum equation requires two parameters: a turbulent momentum factor (FTM); and a turbulent mixing coefficient. The FTM describes the efficiency of the momentum mixing. A conservative value was selected for FTH although the minimum DNBR is known to be insensitive to this parameter. The turbulent mixing coefficient is an important parameter since it determines the flow mixing rate. The value used for the mixing coefficient for Seabrook is documented in the current updated FSAR and was experimentally determined to be conservative based upon tests conducted using the 17x17 geometry and mixing vane grids on 26 inch spacing. 3.1.2 Critical Heat Flux Correlations The VIPRE-01 SE requires that use of a new CHF correlation for VIPRE be qualified. The use of WRB-1 with VIPRE has been approved for other licensees. The WRB-1 correlation is used in the active fuel region above the first mixing vane grid. In an attempt to develop a Seabrook specific WRB-1 correlatiop limit, Yankee selected applicable (by fuel type) sets of test data from the original WRB data base to obtain a DNBR of 1.16 compared to the standard 'WRB-1 limit of 1.17. The staff, however did not accept this lower value because it is based on only a subset of ti data. Yankee has agreed to use'the currently approved DNBR limit for WRB-1 of 1.17, including the associated correlation statistics. This is acceptable to the staff. The W-3 correlation will be used in situations where the W-3 (R-Grid) and WRB-1 correlations are not applicable. The use of the W-3 correlation has been approved previously for use with VIPRE-01. This is acceptable to the staff.

ws i , m i' b' O I k ! p l 4- - y h

(

{. 3.1.3 Other Model Selections Yankee did not provide detailed justification for the selection of parameters i 0 or nodalizations discussed in this section. However, Yankee specifically i flc addressed items in the VIPRE-01 SE to demonstrate that it had followed l suggestions and recommendations cited in the SE in addition to the limitations described in the SE. This approach is acceptable to the staff. In addition, Yankee asserted that the ultimate justification is provided though the benchmark analyses discussed in Section 3.2 of this report. The staff agrees i with this view. f 3.1.3.1 VIPRE Nodalizations The core nodalizations afe presented in Reference 3. The radial and axial nodes used are presentedt in tables of VIPRE-01 Model Descriptions in Appendix i 8 of YAEC-1849P. These hodalizations were reviewed and found acceptable. l 3.1.3.2 Fuel Rod Modelina i j \\ The Conduction Ro l model is used to compute the rod surface heat flux; when the heat flux is known, the Dummy Rod model is used. When the Conduction Rod model is used, the gap conductance is determined using the approved FROSSTEY 1 i code. The radial nodalization is similar to that used with the approved CHIC-KIN methodology currently used by Yankee for Maine Yankee, +his is i acceptable. { 3.1.3.3 Inlet Flow Distribution Yankee stated that the method used for determination of inlet flow maldistribution factor is consistent with the approved application of Yankee's subchannel methodology. lThe flow reduction factor is computed for each core to assess the effect of core-wide inlet flow maldistribution and of " mixed cores" on enthalpy rise in hot assembly locations for a range of power distributions. This flow reduction factor is applied to determine a conservative inlet flow for the subchannel model in the hot assembly, this is acceptable. 3.1.3.4 Power Distributions The radial and axial power distributions are computed using an approved s( methodology using a combination of computer codes accounting for fuels, burnup, reactivity feedback, presence of burnable poisons and control rod position. i.- 3.1.3.5 Numerical Solution Technioue The direct solution meth most of the situations. gd was used for all of the Yankee plant models for Yankee will determine the use of the direct or RECIRC solution technique basedlupon whether the axial flow / crossflow ratio criteria ismet,thisisacceptable. P e i

\\ i 4. l o 4 f,- 3.2 Seabrook Demonstration Analyses i c. E Four cases were analyzed 4for the purpose of comparison with Westinghouse f THINC-!Y and THINC-Ill r These cases were: (1) technical specification k comparison,(2)Seabrook/psults. Core-1 steady state minimum departure f t boiling ratto (MDNBR) vs l ower level, (3) Loss of Flow accident run as a p l $[ transient case, and (4) boss of Flow accident run as a series of static cases. Inorderforthecomparij!on to be meaningful, VIPRE-01 was run using the W-3 i ~ j correlation with a DNB 1Imit of 1.3. Seabrook Station Technical Specification k" curves which define the Peactor core safety limits in terms of RCS average i c temperature versus rated, thermal power were generated using VIPRE-01. In l f Cases 1 and 2, Yankee's computed curves agree well between VIPRE and THINC-IV results. In Case 3, VIPRE transient results showed that VIPRE-01 was consistently more conservative than the THINC-Ill results. Yankee stated that against THINC-IV, the VIPRE results would be expected to be more comparable. i When run as a series of steady state cases for the same loss of flow event, comparable results from those in Case 3 were obtained and comparison between j the two code results were equally good. 3.3 Revised Thermal Desion Procedure The traditional method f uncertaintiesthatenterhqraccountingforthedesignandmodeling into the determination of a DNBR assumes that key i input parameters to the core thermal-hydraulic code are simultaneously at their worst level of uncertainty (a " deterministic method"). The RTDP was developed to remove excess conservatism of the deterministic method in that l variatinns in plant operWting parameters, nuclear and thermal parameters, feel j fabrication parameters, a'nd DNB correlation predictions are considered j statistically to obtain s DNB uncertainty factor. The methodology, therefore, i assumes that the uncertainty associated with the DNB correlation can be statistically combined with the system uncertainties associated with seven 3 independent parameters. e 3.3.1 RTDP Parameter Statistical Distributions The first step in development of the RTDP limits is to determine uncertainty distributions associated with seven independent parameters in RTDP. Uncertainties due to the computer codes used (VIPRE-01 and RETRAN) are also incorporated. Yankee's use of a 4% unce'rtainty due to VIPRE-01 is an arbitrary selection. A 5% uncertainty for use in statistical DNB methodology has been approved previously by the NRC. uncertainty has changed,T'here is no reason to believe that the code since such approval. Similarly, Yankee's use of a 1% uncertainty in transient" analysis performed with RETRAN is equally arbitrary. BasedupontheonlybenchlmarkanalysisprovidedcomparingRETRANcalculat Supporting data was not provided to justify the uncertainty values used. t with plant data, at least' 3% should be attributed for the RETRAN transient code /model uncertainty..:Therefore the total value for code related uncertainty should be 8% instead of the 5% used by Yankee. Yankee stated 3 (Ref.10) that the use of about 8% in the total code uncertainties does not A. k f, o

3.; of / 'l p 15 affect the RTDP-based safety analysis limits. If the RTDP limits need to be re-evaluated in the future', the values stated above should be used for code uncertainties unless model improvements can be demonstrated to reduce uncertainties. 3.3.2 Determination of Sensitivity Factors Sensitivity factors are determined over a wide range of statepoints covering

operating' conditions with MDNBR values near the expected design limit DNBR using RTDP. Sensitivity factors are defined as the percentage change in the VIPRE-Ol calculated MDNBR resulting from a 1% change in a RTDP parameter.

The most limiting statepoint was selected as the one for which the sensitivities resulted in the highest design-limit DNBR using RTDR. This approach is acceptable. 3.3.3 RTDP DNBR Limits and Penalties The RTDP DNBR limits are determined from standard deviation values, the i independent parameters, the maximum sensitivity factors, and WRB-1 correlation ( statistics. The followingiDNBR penalties are applied to the computed RTDP i DNBR limit to ensure a margin of safety: approximately 1% penalty for the fuel rod bowing effect, 1 3% DNBR penalty to account for lower plenum RCS flow anomaly, and 5% DNBR as margin available to offset potential future unidentified non-conservatisms, including cycle-to-cycle variations. This approach is acceptable to the staff.

4. (-

CONCLUSION AND LIMITATIONS I YAEC-1849P, together with the information submitted with Reference 3 contain sufficient information to satisfy the SE requirements for VIPRE-01 and the RTDP. Furthermore the usejof VIPRE-01 with appropriate CHF correlations for Seabrook is acceptable. The Yankee VIPRE-01 model for Main Steam Line Break l and Rod Ejection analyses was reviewed previously (Ref.10), and therefore, i was not included in this review. The application of RTDP with VIPRE-01 for calculation of DNB limits jfor Seabrook is also acceptable subject to l restrictions presented belpw: l 1. Yankee continues to be subject to VIPRE-01 and RTDP SE requirements l should any situation cited in the SE conditions arise to change the j applicability of the current set of code / correlation /model combinations, t i 2. Use of the WRB-1 CHF correlation with VIPRE-01 is found acceptable l with a DNBR limit of 1.17 instead of 1.16, and is limited to 26 inch j spacing grids, l 4 ll U l il i

- ~. t ~ . q:

• (

.y. 3. The uncertainties associated with VIPRE and RETRAN shall be 5% and 3%, respectively, and 4. Although the approach to the Yankee thermal-hydraulic subchannel methodology, as described in VAEC-1849P and Reference 3, is generally applicable to othpr PWR plants, this SE is applicable only to Seabrook due to the correlations and models selected as well as the useofspecificuqcertaintiesanddistributionsbaseduponplant specific data. ]

5.0 REFERENCES

1. Letter from T.C. Feigenbaum, North Atlantic Energy Systems, to USNRC, " Request for NRC Reviqw and Approval of Analysis Methodologies to be Applied to Seabrook,"'!Fabruary 2,1993. 2. " Thermal-Hydraulic Analysis Methodology Using VIPRE-01 for PWR Applications," YAEC-1849P, October 1992. 3. Letter from T.C. Feigenbaum, North Atlantic Energy Systems, to NRC, " Response to Request for Additional Information (TAC M86957 and TAC M86958)," March 9, 1994. 4. Letter from C.E. Rossi NRC to J. A. Blaisdell (UGRA), " Acceptance for Referencing of Licensi:ng Topical Report VIPRE-01: A Thermal-Hydraulic Code for Reactor Corej, EPRI NP-2511-CCM, Vols.1-4," May 1,1986. 5. " Maine Yankee Core Thdrmal-Hydraulic Model Using COBRA-IIIC," YAEC-Il02, I June 1976. I 6. "A lhermal-Hydraulic Analytical Model Using COBRA !!!C," YAEC-1058, May 1974. 7. "DNBR Limit Methodology and Application to the Maine Yankee Plant," l YAEC-1296P, January 1982. l 8. Letter from R. A. Clark NRC, to J. H. Garrity YAEC, Safety Evaluation by the Office of Nuclear Reactor Regulation on " Topical Report YAEC-1296P "DNBR Limit Methodology and Application to the Maine Yankee Plant," March 9, 1983. l 9. Letter from A.C. Thadani NRC, to W. J. Johnson, Westinghouse, " Acceptance for Referencing of licensir.g Topical Report WCAP-ll397, " Revised Thermal Design Procedure"," Januarf 17, 1989. " STAR Methodology Aphlicatian for PWR's Control Rod Ejection, Main Steam 10. Line 8reak," YAEC-1752-A, October 1990. Principal Contributor: Limbros Lois Date: August 15. 1994 q l

2 1 l=5 1 I i NUREG/CR-4394 l ORNL/TM-9628 t 4 1 l dAK RIDGE NATIONAL l LABORATORY Rod Bundle Film Boil.ing and l Steam Cooling Data Base n MAR 1"fN MARIETF"A and Correlation Evaluat. ion G. L Yoder i e 4 l t i 1 i Prepared for the U.S. Nuclear Regulatory Commission ) Office of Nuclear Regulatory Research Under interagency Agreements DOE 40-55175 and 40-552 75 l l 6 ~e6-tosto w e60831 7 [k j (PERATEDBY PDN NUREQ MARilN MARIETTA ENERGY SYSTEMS. IN; CR-4394 R PDR FOR THi UNITED STATES l OEPARTMENT OF ENERGY l

f5 ix k A 't

SUMMARY

l i

• L i

f i In the process, of boiling on vertical heated rods, vapor generated O at the surface is swept away by the flowing liquid (forced convection) or l buoyancy forces (pool boiling). As the rods are traversed axially f rom l bottom to top, even'tually one of two conditions must occur. Either the surface will become dry due to vapor void generation, or the. surface will become dry because the critical heat flux (CHF) }oint has been reached. j In either case, the transition between the wet and dry portions of a rod also marks a transition between heat transfer regiees. If the rods are wet, surface heat transfer is very good, with correspondingly low rod i surface temperatures. If the surface of the rod is dry, heat transfer f rom the rod to the vapor is relatively poor, resulting in high rod sur-i face temperatures. The object of this report is to document a data base which has been assembled from experimental rod bundle data taken in the post-CHF regime. The data base includes data from 10 film boiling references and 5 steam cooling references in total, over 20,000 data points are included. These data have been used to evaluate several film boiling and steam cooling heat transfer correlations. Film Boiling Dougall-Rohsenow Dougall-Rohsenow (wall Prandt1 Number) Condie-Bengston IV Groeneveld 5.7 j Groeneveld 5.9 Groeneveld-Delorme Steam Cooling Dittus-Boelter Dittus-Boelter (film properties) Sozer Comparisons indicated that the Condie-Bengston IV correlation predicted best in the film boiling regime. While all three of the steam cooling correlations examined performed similarly, the Dittus-Boelter (film) l correlation and Sozer correlation predicted most conservatively. 6 O J c

~ -. .i ~ p i RODj BUNDLE FILM BOILING AND STEAM COOLING DATA BASE AND CORRELATION EVALUATION Craydon L. Yoder i O Engineering Technology Division Oak Ridge National Laboratory Oak Ridge, Tene.c.use 37831 )i ABSTRACT 4 Film boiling and steam cooling data from several rod bun-i die experiments have been compiled from fourteen references. Steam cooling data are presented for five references, while t film boiling data are presented for ten references (two of these for tube geometry). The compilation includes experi-mental parameters necessary for characterization of the local + rod heat transfer. These data have been compared to several commonly used 4 film boiling and pure vapor heat transfer coefficients. } Results show that the Dougall-Rohsenow correlation tends to overpredict film boiling heat transfer, while Condie-Bengston IV and the Groeneveld 5. series correlations do a reasonable 1 job of predicting film boiling heat fluxes. Of the three vapor correlations examined, the Dittus-Boelter correlation is least j conservative, and the Soter correlation most conservative. [ i 1. INTRODUCTION l In the process of boiling on vertical heated rods, vapor generated at the surface is swept away by the flowing liquid (forced convection) or buoyancy forces (pool boiling). As the rods are traversed axially from bottom to cop, eventually one of two conditions must occur. Either the j surf ace will become dry due to vapor void generation, or the surf ace will i become dry because the critical heat flux (CHF) point has been reached. i in either case, the transition between the wet and dry portions of a rod i also marks a transition between heat transfer regimes. If the rods are { wet, surface heat transfer is very good, with correspondingly low rod surface temperatures. If the surface of the rod is dry, heat transfer from the rod to the vapor is relatively poor, resulting in high rod surface temperatures. j If dryout of the rod occurs due to vapor void generation (i.e., the void fraction and quality become one, steam cooling exists above the dry- } out point. If CHF occurs with quality less than one, film boiling exists above dryout, and although the surface of the rods are dry due to very l-high temperatures, liquid is still present in the flow. A ) ) ] ? I

4l n 2 = \\ High surface temperature under dry rod conditions has made film boiling a flow pattern which has been studied extensively over the 3 years. Steam cooling has also been investigated, although to a lesser extent, since single phase flow behavior is much easier to understand and i predict than two phase behavior. Until recently, the major portion of 1 i both two phase film boiling and single phase steam cooling heat transfer I data was acquired using tubular test sectior.s. usta of this type can be found in many references,1-5 and evaluation of several heat transfer c - relations has been perfprmed using tube data.6,7 Within the last ten years or so, larger facilities incorporating bundles of rods have been ) ) used to gather film boi ing and steam cooling data. Data from several experiments have been g thered and evaluation of this rod bundle data is i { documented in this report. A few general ground rules for data acceptance can be stated. Rod surface conditions (heat flux, temperature) and local flow conditions (quality, mass flow, temperature) must be provided by the experimenter or I be calculable, in order to determine the local heat transfer. The limited scope of this investigation did not allow transient bundle cal-culations to be performed when only bundle boundary conditions or local i instrument responses were available. The data base is assembled from bundle heat transfer data except where experimenters specifically asked that their tube data be included in the data base. Other criteria were also used and will be explained on a specific basis in Sect. 2. Data j gathered during the course of this investigation will be placed on the i INEL data bank. ) Data from fourteen references are included in the data base and are compared to several commonly used correlations. t; Film Boiling Dougall-RohsenowB Dougall-Rohsenow9 (wall Prandt1 Number) Condie-Bengston IV10 1 Groeneveld 5.7 1 Groeneveld 5.911 Groeneveld-Delorme 12 Steam Cooling Dittus-Boelter13 Dittus-Boelter (film properties) Soz e r l Five sets of data were acquired under steam cooling conditions while ten mets of data were taken under film boiling conditions (two of these in tube geometry). A sunanary of individual investigations and overall film boiling and steam cooling data ranges is presented in Table 1.1. This table also indicates the number of rods used in' the bundle, the type of data gathered during testing (film boiling or steam cooling), an indica-tion of whether heat transfer data is available near the bundle spacer grida, the type of test performed (steady state or transient), and the type of heating within the bundle (uniform or non-uniform), as well as the range of conditions covered during testing.

. _.... - - -.. _ ~. -. .. _ _. ~ m_._.-- 9..".* g f atrie I.I Rod bundle itIn botting/ steam cnoting data sources Wetted Ref %o. Data Grid -Test -- Bursd tir - - " Yd ' *" I I Pressure : theat tt r" diameter '- ~--- II"' II"" heatec no. rods type info? Type heating (MPa) (t) (ku/myn (kg/e7,3 g,,) p,,g,,,,, r ratto Ad..r n a I% 7 Film No SS UA-t'R J tie-i sso l iini-int Mi s.o M-80 U.917. 0.709 1.5. 1.55 t.res neveld le i Film No SS Fuet 4tMS-I tins t lfMI-2(uW1 6.5 40-60 1.2) 3.94 elements Mataner 17 19 Fite h 5% UA-NR tik ki-19s M t hSo-2h50 7.n 20 50 0.755 t.43 Mataner 14 19 Flin l!A-NR Mfkl-I b p s 7t h)- 2t t99 3.-M. 60-60 0.5 14 1.7 Y r=te r 19 64 Film Yes 5% UA-l'R lbe-96o . su-Mi w i 4.5-33. 60-120 1.0% t.) Ya.le r .M 64 F.5 Ves 5% t' A-t*R 300-470 40-260 1.M-M.5 Itn-Itpre t.on 1.) j gnktam 23 64 % tram Yes SS t!A-UR 1%-4o lie-t 3 2.b-7.1 3.44, t.) w Anklan 22 %4 4teme Yes 54 l'A-UR 310-R0 %- D) 4.1-7.2 3.ub t.) w..n4 21 868 Steam Yes 54 MA-UR o.%-% l-2 ) it. ) 1.tR -l.23 4,rets 2 h4 File Yes T 4 4-t'R tho-s tehl t ir - t o9es 5.2-12.4 3n-850 1.00 t.) te, 25

=i t Film Yes T

NA-t?R t re-So - l :.o n te-o.3 20-300 1.3g -g,yg o mauasant 14 Tutte File No 55 UA 6al- )os 150-%50

t..M-t e.. ;

W-120 8.08 i e with 27 Tube Ftte No SS CT to-4ami 25-1688 0.te.) 10-80 t.25 3 t.ni t us 2M 28 Steam Yes SS MA-UR O.4-6.% 3-17.% e s.14 0.887 1.55 F svetwit s CT - ont rolled temperature ) IA - unifore aulatly tR - unifore radlally NA - nemun i f ora an t a ll y NR - nonuniform radtally %s - steady state T - t rans tent i f

v...

4 i. 0 Data within the data base can be used to assess and develop corre-lations used to predict fuel rod behavior under film boiling or steam cooling conditions. Enough information was included for each data point that this should be a relatively painless exercise. The data base was designed to require only a water-steam thermodynamic and physical prop-erties package to accomplish this task. Section 2 describes each experiment individually, while section 3 shows details of the data parameter ranges. Section 4 describes the correlations evaluated and shows the results of the evaluations. Con-elusions are presented il Section 5. The appendix describes the format j used in tabulating the data for the INEL data bank. l i i I I .I i F i I

Project No. 697 DOE Tritium Program cc: Max Clausen Office of Commercial Light-Water 1 Reactor Production Tritium Project Office U.S. Department of Energy 1000 Independence Avenue, SW Washington, DC 20585 DP-60 Records Management Office of Commercial Light-Water Reactor Production Tritium Project Office U.S. Department of Energy l 1000 Independence Avenue, SW Washington, DC 20585 Jerry L. Ethridge, Sr. Program Manager Environmental Technology Division Pacific Northwest National Laboratory Battelle Blvd. P.O. Box 999 Richland, WA 99352 4 l l}}