ML20246F277

Safety Evaluation Supporting Mods to ECCS Evaluation Model, Including Changes to FLECHT-based Reflood Heat Transfer Correlation,Steam Cooling Model & post-critical Heat Flux Heat Transfer Model

ML20246F277
Person / Time
Site:	Yankee Rowe
Issue date:	07/11/1989
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20246F269	List: ML20246F266 ML20246F277
References
NUDOCS 8907130196
Download: ML20246F277 (6)

Text

93a -

/ 'o,, UNITED STATES L * .!' n

(, NUCLEAR REGULATORY COMMISSION g g WASHINGTON, D. C. 20555

%...../

EHCLOSURE SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATING TO ECCS EVALUATION MODEL LICENSE NO. DPR-3 YANKEE ATOMIC ELECTRIC COMPANY YANKEE NUCLEAR POWER STATION DOCKET NO.50-029

1.0 INTRODUCTION

On January 5,1988, Yankee Atomic EMctric Company (YAEC) proposed revisions to the ECCS evaluation model for the Yankee Nuclear Power Station (YNPS) (Ref.

1). On May 2, 1989, YAEC provided supplemental information and modified its proposal (Ref.2). Pevisions were made to the FLECHT-based reflood heat transfer model, the steem cooling model, and the post-CHF heat transfer model. Our evaluation of these revisions and their conformance to the requirements of Appendix K to 10 CFR 50 is provided below.

2.0 EVALUATION 2.1 FLECHT-Based Reflood Heat Transfer Model For reflooding rates greater than one inch per second,Section I.D.5 of Appendix K to 10 CFR 50 requires that reflood heat transfer coefficients be ,

based on applictble experimental data. Such data have been obtained in the FLECHT(Ref.3)andFLECHT-SEASET(Pef.4)testprogramsandhavebeenusedby other PWR vendors as the basis for acceptable reflood heat transfer models.

The currently approved YAEC reflood heat transfer model is based on the Westinghouse FLECHT heat transfer correlation in Reference 3 with neuitiphers applied to make the correlation a best estimate prediction of the experimental data. The currently approved set of multipliers is referred to as the ENC-2 8907130196 990711 FDR ADOCK 05000029 P PDC

e.

6 2

FLECHT multipliers. In addition, the correlation is based upon the assumption that the heat transfer coefficient at a given elevation is dependent on the integrated energy distribution up to, that elevation. This approach allows the correlation to be applied to core geometry and axial power distributions different from that used in the FLECHT tests. Because the Yankee Nuclear Power Station (YNPS) fuel rod diameters differed from that used in the FLECHT tests, the calculated heat transfer coefficients are multiplied by 0.8 for use in the LOCA analysis.

YAEC proposed to modify its reflood heat transfer model by replacing the ENC-2 FLECHT multipliers with multipliers based upon the FLECHT-SEASET data. This data was used because the fuel rod ciaraeters were more representative of the YNPS fuel. Thus, YAEC also proposed removal of the 0.8 multiplier.

To justify the model, YAEC performed benchmarks of the revised model to the FLECHT-SEASET data. The specific test conditions for these benchmarks envelope the range of reflood conditions expected for YNPS. The results of the benchitarks showed that the model was generally conservative.

Although the rod diameters in the FLECHT-SEASET tests are more representative of the YNpS fuel, we questioned the removal of the 0.8 multiplier as the YNPS fcel hydraulic aianeter was different from that in the FLECHT-SEASET tests.

In Reference 2, YAEC modified its model to account for differences in rod bundle geometry. This modification is based upon preserving the total energy per unit flow area betwe.n YNPS and the FLECHT-SEASET rods. This approach was adopted in the FLECHT-SEASET overlap tests which examined the effect of Lordle geometry on reflood heat transfer. These tests results verified that bundle gectretry effects are small when this scaling approach is used.

l Since the proposed model generally produced conservative results in corrparison to the experimental data and the approach taken to account for bundle geometry effects has been experimentally verified, we conclude that YAEC's proposed changes to its FLECHT-based reflood heat transfer raodel corcplies with Section I.D.5 of Appendix K to 10 CFR 50.

1

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

l l

l

{

l 3

i 2.2 Steam Cooling Model 1 i

i For reflooding rates of less than on,e inch per second,Section I.D.5 of Appendix K to 10 CFR 50 requires that heat transfer coefficient be calculated assuming steam cooling only and that the effects of flow blockage on local i stearn ficw and heat transfer be accounted fer. The currently approved YAEC steam cooling ir.odel calculates an equivalent steam flow which, when used in  !

the Dittus Boelter correlation, yields the sarne heat transfer coefficient just below the blockage plane as that obtained with the FLECHT-based reflood heat transfer correlation. To account for flow blockage effects, the steam flow is reduced by a fraction which series with distance from the blockage plare. The reduced steam flow is used in the Dittus-Boelter correlation to yield the heat transfer coefficient. The fluid energy solution above the blockage plane is also modified to account for blockage effects.

The FLECHT-SEASET exper.frents have shown that as flooding rates are' decreased below unie inch per second, the reflood heat transfer behavior is not different from that observed at the higher flooding rates. That is, there is no abrupt change in the heat transfer coefficient at any of the flooding rates examined. The FLECHT-SEASET experiments have also shown that blockage does r.ot result in heat transfer degradation (Ref. 5). As a result of these experimental observations, YAEC concluded that its current raodel was overly conservative and submitted a revised model.

The revised YAEC steam cooling model functions as follows:

l

1. The coolant flow is assumed to be saturated steam.

2. The local heat transfer coefficients are calculated using the FLECHT-basec heat tratisfer tredel.

l

3. Flow bypass in the blockage region is calculated usir g the currently approved flow diversion model.

, 4 a

4 ,

4. The local heat transfer coefficients are modified to account for the flow bypass effects.

5. The local heat transfer coeffic'ients are enhanced due to the effect of increaseo turbulence of the steam phase due to boundary layer separation caused by the flow blockage, i

The net effect of the new model is to result in heat transfer coefficients which are closer to, but no larger than, that calculated by the FLECHT-based reflood heat transfer model.

For floocing rates less than one inch per second,.the staff finds that the YAEC model assumes steam cooling only and accounts for the effect of flow blockage on local steam flow and heat transfer. In addition, the model will yield heat transfer coefficients which are less than that obtained using the unblocked FLECHT-based reflood heat transfer model. Since the new steam cooling model will not predict the improved heat transfer observed in the FLECHT-SEASET blocked bundle tests, the revised model will yield conservative peak cladding temperature results. Thus, we find the model meets the requirements of Section I.D.5 of Appendix K to 10 CFR 50.

2.3 Post-CHF Heat Transfer Model The current YAEC ECCS evaluation model uses the Dougall-Roshenow correlation l for post-CHF heat transfer. This model is no longer specified as an I acceptable post-CHF model in Section I.C.5 of Appendix K to 10 CFR 50. In fact,Section I.C.5.c of Appendix K now states that if model changes reduce the calculated peak cladding temperature by at least 50*F the Dougall-Rohsenow correlation can not be used under conditions where 1onconservative heat transfer coefficients may result. YAEC believes that the proposed model changes result in a peak cladding temperature decrease in excess of 50*F.

l To address the Appendix K requirement, YAEC has modified its selection of post-CHF heat transfer correlations. The new logic uses the Groeneveld 5.7 l

l

5 correlation for pressures greater than 500 psia. Below 500 psia, the film boiling heat transfer coefficient will be the minimum of that calculated by either the Groeneveld 5.7 or Dougall,-Roshenow correlation. As described in Reference 6, the Groeneveld 5.7 correlation is prevented from misuse near its

. low pressure singularity.

The staff finds that the proposed modification to the post-CHF heat transfer model meets the requirements of Section I.C.5.c of Appendix K to 10 CFR 50.

In addition, the Groeneveld 5.7 is listed as an acceptable post-CHF correlation in Section I.C.5 of Appendix K. Therefore, the staff finds the proposed model change acceptable.

3.0 SUM. MARY AND CONCLUSIONS YAEC has proposed several modifications to the YNPS ECCS evaluation model.

Modifications were made to the FLECHT-based reflood heat transfer correlation, the steam cooling model, and the post-CHF heat transfer model. The staff finds that these modificaticas satisfy the applicable requirements in Appendix K to 10 CFR 50. Therefore, we find these modifications acceptable.

4.0 REFERENCES

1. Letter, G. Papanic (YAEC) to NRC, "LOCA Reflood Heat Transfer Models,"

January 5,1988.

2. Letter, G. Papanic (YAEC) to NRC, "YAEC Response to NRC Review of Revised Reflood Heat Transfer Model for YNPS LOCA Analysis," May 2,1989.

3. F. F. Cadek, et al., "PWR FLECHT Final Report Supplement," WCAP-7931, October 1972.

4. M. F. Loftus, et al., "PWR FLECHT-SEASET Unblocked Bundle, Forced and Gravity Reflood Task Data Report, Volume 2, Appendix C,"

NRC/EPRI/ Westinghouse Report No. 7, September 1981.

l

- --- --------- _ - _ _ a

s 6

5. L. E. Hochreiter, et al., " Analysis of FLECHT-SEASET 163-Rod Blocked Bundle Data Using COBRA-TF," NRC/EPRI/ Westinghouse Report No. 15, April 1985.

l

6. "WREM: Water Reactor Evaluation Model (Revision 1)," NUREG-75-056, May 1975.

Principal Contributor: R. Jones l

l i

4 1

I

__ _ _ _ _ _ _ _ _ .