ML20148D707
| ML20148D707 | |
| Person / Time | |
|---|---|
| Issue date: | 10/17/1978 |
| From: | Schroeder F Office of Nuclear Reactor Regulation |
| To: | Vassallo D Office of Nuclear Reactor Regulation |
| References | |
| TASK-AS, TASK-BN-78-24 BN--78-24, NUDOCS 7811020459 | |
| Download: ML20148D707 (5) | |
Text
. .
sit e Pb g
- UNITED STATES 8 ' ^
NUCLEAR REGULATORY COMMISSION
$? :E WASHING TON, D. C. 20555
% ,, # OCT 171978 MEMORANDUM FOR: D. B. Vassallo, Assistant Director for Light Water Reactors, DPM FROM: Frank Schroeder, Acting Assistant Director for Reactor Safety, DSS
SUBJECT:
BOARD NOTIFICATION - RECENT BWR BD/ECC EXPERIMENTS The Blowdown / Emergency Core Cooling (BD/ECC) program is a cooperative experimental research program jointly funded by the Electric Power Research Institute (EPRI), General Electric (GE) and the Nuclear Regulatory Comnission (NRC). Tests are conducted by GE under this program in the two loop test apparatus (TLTA) in San Jose, California.
The purposes of the program are:
- 1. Obtain and evaluate basic BD/ECC data from test system configurations which have calculated performance characteristics similar to a BWR with 8 x 8 fuel bundles during a hypothetical LOCA.
- 2. Determine the A gree to which models for BWR system and fuel bundles-describe the observed phenomena, and as necessary, develop improved models which are generally useful in improved LOCA analysis methods.
The TLTA configuration used for the BD/ECC is scaled to a BWR/6 design (624 bundles) and includes the following major components: (1) pressure vessel and internals, (2) an 8 x 8 heated bundle, (3) two recirculation loops,(4) ECC systems (HPCS, LPCS, LPCI), (5) Automatic Depressurization System,(6) Auxiliary systems.
During August of 1978, test number 6405 was conducted; the test had an average power bundle with low ECC injection flow. Results of the test were compared with those from test 6007 which had the same initial conditions, but no ECC injection. The comparison was presented in the monthly report issued in September,1978 and in a program management group meeting on September 21 and 22,1978. The comparison showed that the system depressurized more slowly with ECC injection than without ECC injection. Since the slower depressurization with ECC injection l was contrary to intuitive expectations, GE was requested to discuss I the test results and implications with the NRC. This discussion took l place on October 10, 1978.
Contact:
M. W. Hodges, NRR, X27588 7 fl/co459
OCT 171978 D. B. Vassallo The overall results of the test, when ccmpared with the results of tests conducted without ECC injection, clearly indicated the benefits of ECC injection. Dryout was definitely delayed for most bundle elevations and maximum cladding temperature was lower with ECC injection than without injection. The tests also indicated higher heat transfer rates than those used in licensing calculations.
The test result which raises questions is the slower depressurization rate for the test with ECC injection. Since no pre-test calculations had been performed by GE, as part of the program to predict test behavior, a direct comparison between the test results and those that would be predicted by the calculational techniques '-d for safety analysis is not yet available. Preliminary post-tr :alculations performed for NRC with RELAP-4 at INEL indiccte t! eme trend on depressurization rate as observed in the tests, but not to the same extent.
Preliminary calcuiations performed by the staff and similar calculaticns performed by GE suggest that the slower depressurization is due to larger vapor generation with ECC inject.%n than without ECC injection.
The staff calculation indicates that the vapor generation significantly exceeds that predicted from either the GE proprietary vaporization correlation used to calculate counter current flow limiting (CCFL) conditions by the REFLOOD code or by the average core heat transfer model in SAFE. SAFE and REFLOOD are part of tk approved ECCS evaluation model.
The energy required to increase the rate of vapor generation could come from nither the heated core or from the stored energy in the structural elements of the lower plenum. The preliminary calculation done by the staff cannot distinguish between the two sources. If the extra steam
- i. generated in the core, then the vaporization correlation and the SAFE heat transfer model predict too little vapor generation and could l mean that non-conservative assumptions have been made about actuation of Li 1 and CCFL breakdown in LOCA analyses. If a large portion of the l
extra steam is generated elsewhere, then little or no change will be required in the currently approved licensing models REFLOOD and SAFE.
Because of scaling restrictions, the steam generation in the lower plenum of TLTA is larger than would be predicted in a BWR if TLTA scaling were used. This compiicates the interpretation of the new test data.
Several actions are ongoing to assess the significance of the recent test results:
- 1. GE is analyzing the test data in detail as part of the cooperative research program. This effort should help identify the various sources of steam.
I
OCT 171978 D. B. Vassallo 2. GE has been requested to perform calculations with the ECCS evaluation model to test its essential features against the experimental data for tests with and without ECC injection. The calculation is expected to take apprcximately four months and will check the capability of the evaluation model to predict the observed phenomenon.
- 3. GE has also been requested to perform a study of the sensitivity of the post-LOLA peak cladding temperature to increased steaming in REFLOOD and SAFE.
- 4. The staff will perform its own calculations to attempt to predict the observed pnp omenon.
We expect to review these work products and complete our evaluation of the significance of the test results in March,1979.
NRR Office Letter No.19 calls for a determination of the safety significance of new information by evaluating "whether this information could reasonably be regarded as putting a new or different light upon an issue before Boards or as raising a new issue". The information from these experiments is that the steam generation rate in the test assembly may be significantly underpredicted by the vaporization correlation and core average heat transfer model used in the GE ECCS evaluation model.
If there is an error in the model, it could put a nes; or different light on the behavior of BWR blowdown and eflood modeling in that the presence of water in the bundles may reqa fe modification to correlations used in the model. More information from a source or sources external to the NRC staff is required before that need can be sssessed. Therefore, I conclude that appropriate boards should be notified. In accordan.e with NRR Office Letter No.19 requirements, I am providing you the follnwing:
- 1. the item for notification;
- 2. considerations regarding relevancy and materiality;
- 3. statement on perceived significance,
- 4. relation to projects.
i
- 1. The Item l
Comparison of TLTA tests 6405 and 6007, with ECC injection and ;
without ECC injection respectively, showed a slower depressurization l with ECC injection than without ECC injection. The difference is possibly due to increased steam generation in the test with ECC injection. The steam generation appears to be significantly larger
^
D. B. Vassallo than predicted from the GE vaporization correlation and the core average heat transfer model.
- 2. Relevancy and Materiality The vaporization correlation and the core average heat transfer model are relevant to all BWR's. There are some questions relating to scaling atypicalities; the TLTA lower plenum surface to volume ratio is incorrect, there is only one heated bundle and the jet pumps are too short.
- 3. Significance The results of the tt ' clearly indicate the benefits of ECC injection.
The tests also indicate higher heat transfer than used in licensing caiculations. However, the tests indicate that the vaporization correlation in REFLOOD and the core average heat transfer model in SAFE may significantly underpredict the vapor generation k the core. Further analysis may show the need for modifyina correlations. Because a deficiency ir, the treatment of st generation is only suspected, not verified, and because otoe..
parts of the ECCS evaluation model are known from previous test, tu be conservative (e.g., the evaluation model predicts premature dryout and underpredicts hot bundle heat transfer) the staff concludes that continued use of the currently approved evaluation model is acceptable and appropriate while the significance of the test results is being assessed. Also, it is the present judgement of GE that the tests with ECC injection demonstrate that the evaluation models and the licensing analyses are sufficiently conservative.
- 4. Relation to Projectl This relates to BWR's in general.
I recorted that appropriate boards be notified of the existence of the new + :ta and of the staff's plans to evaluate its significance regarding the continued acceptability of the GE ECCS evaluation model by the end of March,1979.
l
/, l. t H t .\.
l Frank Schroeder, Acting 3 1stant Director j for React (r Safety l Division of Systems Safety l
cc: See next page.
f cc: R. Mattson V. Stello R. Boyd D. Eisenhut D. Ross S. Hanauer T. Murley T. Engelhardt H. Denton B. Grimes P. Check T. Novak Z. Rosztoczy G. Mazetis F. Coffman L. Phillips P. Norian R. Frahm G. Holaban J. Guttmann R. Woods W. Hodges NRC POR y@ ,