ML20039C810
| ML20039C810 | |
| Person / Time | |
|---|---|
| Site: | Shoreham File:Long Island Lighting Company icon.png |
| Issue date: | 12/03/1981 |
| From: | Mattson R Office of Nuclear Reactor Regulation |
| To: | Eisenhut D Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20039C806 | List: |
| References | |
| NUDOCS 8112300266 | |
| Download: ML20039C810 (5) | |
Text
.
a > coq n
j%,
UNITED STATES NUCLEAR REGULATORY COMMISSION y
t l
WASHINGTON, D. C. 20555
\\*..../
DEC 3 1981 MEMORANDUM FOR: Darrell G. Eisenhut, Director Division of Licensing FROM:
Roger J. Mattson, Director Division of Systems Integration
SUBJECT:
BOARD NOTIFICATION - BWR CORE SPRAY DISTRIBUTION CONCERNS 1.
Item for Notification The staff has been informed by the ACRS that Japanese spray distribution tests of a simulated BWR/5 configuration have shown that central bundles receive low core spray flow due to maldistribution. We have also discussed the Japanese tests with the General Electric Company.
The Japanese tests were conducted for utilities which have kept the data proprietary and very little specific information is available 0 in the U. S.
The test data obtained was for tests in steam of a 60 sector of a full scale BWR core. The results were very similar to g
those obtained from 30 sector tests in steam in BWR/6 and BWR/4 (similar to BWR/5) configurations under the joint NRC/EPRI/GE Refill /Reflood program in Lynn, Massachusetts in 1979-80. The BWR/6 test results are discussed in staff SERs.
0 The Lynn data are believed to be atypical of a BWR 360 configu-ration.
Thisconr.lusionisbasedonknowngesignatypicalities, on data from air-water tests of a BWR/6 360 configuration and data from tests with other variously sized sectors which have shown that BWR/6 spray overlaps in the center of the core causing high flow to central bundles. This overlap does not occur in a sector test since thenozzlesfromtheopp.ositesectorwhichwouldprovidetheove5-lapping flow are not available. We would expect the Japanese 60 sector test to suffer the same deficiency.
We understand that the Japanese core spray test program for the BWR/5 configuration consists of the following:
(I) single header operation in all tests (two are available in all BWR/5 reactors) 0 (2) 60 sector tests in steam 0
(3) 360 tests in steam with onl to inadequate steam supply) y every sixth nozzle operating (due (4) 360 tests in air 8112300266 811211 PDR ADOCK 05000322 PDR l
l
D 3%
..Darrell G. Eisenhut Althougg no specific data are available, we have also been told that the 360 tests by the Japanese with 5/6 of the spray nozzles blocked g
gave similar results to the 60 sector tests. This could be in-terpreted to infer that our previous conclusions concerning the atypicality of low central bundle flow are incorrect or are not applicable to the BWR/4 or 5 configurations.
This data is important to us and we are asking our Office of International Programs to actively seek it from the Japanese govern-ment.
Since credit is taken for core spray heat transfer consistent with an assumption of a minimum spray flow to all bundles in the General Electric ECCS Evaluation Models, this information may be of interest to licensing boards.
2.
Relevancy and Materiality The Ja anese tests are designed to simulate a BWR/5.
Because of similt..ty between BWR/4 and BWR/5 designs, the new data also may have implications for BWR/4.
Thereisalsogomepossibilitythat the new data contradict conclusions from 360 air-water tests in the U. S. for a BWR/6 configuration.
Tha test results relate to the validity of core spray cooling assumptions employed in the Appendix K LOCA Evaluation Model for licensed BWRs.
3.
Significance of Test Information The Japanese data are not the first to show low flow to some sections of BWR cores. Thus, as described above, the staff has previously considered the effect of low core sprel flow to individual channels on calculated peak clad temperature (FCT).
In our evaluation of NED0-20566 Amendment 3, " General Electric Company Analytical Model for Loss-of-Coolant Analysis in Accordance with 10 CFR 50 Appendix K -
Effect of Steam Environment on Core Spray Distribution," it was concluded that minimum spray flow to any channel followiy a LOCA would not be less than half of the design flow that was demonstrated to be available by tests and calculations.
The tests and calculations did not include steam effects on nozzle spray patterns and flow rate.
Based on measurements of minimum bundle spray flow for each BWR size and type for one sparger only, in air, the minimum flow for BWR/2 through BWR/5 designs was calculated to be 1.3 times the flow necessary to remove decay heat by vaporization (reference flow).
Thus, the steam effects on spray distribution would not result in less than 0.65 times the minimum reference flow (or 1.3 times with both spray spargers operating). BWR FLECHT data (Ref. 1) show
=
C 3%
Darrell G. Eisenhut little degradation in heat transfer for flow as low as 0.38 times the reference flow, or approximately one gpm. As far is we have begn told, the minimum flow observed for any bundle in the Japanese 60 sector tests was one gpm. The heat transfer coefficients in GE's ECCS Evaluation Model are based on the FLECHT data, and a minimum bundle flow of one gpm would justify the hegt transfer coefficient for core spray cooling (1.5 Btu /hr-ft - F) used in that 2
Model.
To demonstrate conservatisms in the ECCS Evaluation Model and in response to questions raised by the NRC staff, the effect of reducing the core spray cooling to zero (i.e., no credit for core spray cooling) has been evaluated by the General Electric Company for one BWR/3 and one BWR/5.
For the reactors evaluated, it was found that the peak clad temperature limit in 10 CFR Part 50.46 of 2200F was not exceeded even if credit was not taken for core spray cooling.
It was further shown that a core spray maldistribution can result in lower peak clad temperature if the spray flow is assumed to flow down peripheral channels so as to increase the bottom reflood rate.
These results, however, are not necessarily valid for all BWRs since they are dependent on plant specific reflood rates and on the available margin to the peak clad temperature limit (2200F) in previous calculations.
The staff concludes that the new information from the Japanese tests does not pose a safety concern for operating reactors for the following reasons:
(a) Only preliminary and incomplete data are available from the Japanese tests and it is impossible to draw final conclusions from them at this time.
(b) Core pray flow maldistributions resulting in flows on the order of one gpm per bundle (apparently consistent with those g
obtained in the Japanese 60 sector tests) would remain con-sistent with the core spray cooling assumptions employed in the present ECCS Evaluation Model of GE.
(c) Evaluation of one BWR/3 and one BWtt/5 plant have shown that at least some plants can meet the LOCA core cooling criteria without taking credit for any spray cooling, although this may not be true for all plants.
The staff intends to pursue this matter by attempting to obtain access to details of the Japanese tests via our international information exchange programs.
In addition, the General Electric Company has been asked to brief the ACRS subcomittee on ECCS concerning its awareness and interpretation of the Japanese data.
Darrell G. Eisenhut DEC 3 193l 4,
Relation to Projects The Japanese tests relate to BWR/4 and BWR/5 reactor types.
It is reconsnended that appropriate boards be notified of the existence of the data and of staff's plans to try to obtain the data for evaluation of their effect on the GE ECCS Evaluation Model.
\\
-Q Roger. h;tson Director t
Division Df Systems Integration
DEC 3 1961 Reference 1.
APED-5529, " Core Spray and Core Flooding Heat Transfer Effectiveness in a Full-Scale Boiling Water Reactor Bundle," June 1979. F. A. Schraub and J. E. Leonard.
6 b
)
1 I
l' I'
. - _ _ _ -