ML20052C573
| ML20052C573 | |
| Person / Time | |
|---|---|
| Issue date: | 01/04/1982 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Boehnert P Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1406, NUDOCS 8205050190 | |
| Download: ML20052C573 (3) | |
Text
min m Wissi G h poe v.wd pw e7-Nu ao en D#s8,1011p,1stgl e48 i
6 m
A q
N g
n _':3 ~
m g
January 4,1982
- d 9pgg h 7
m t,; ;h o
- 2, sun ~t
/aul Boehnert TO:
P A
/
FROM:
Ivan Catton p' \\ '
SUBJECT:
ECCS Subcommittee Meeting Los Alamos, 2 Dec 1981 Preliminary discussion with Lou Shotkin about the various code ment programs were somewhat incomplete.
In the area of containment codes there are three efforts:
- 1) development of the CONTAIN code of SANDIA;
- 2) development of a code to calculate H2 distribution at LASL; and
- 3) modification of COBRA-TF to do containment calcula There are also codes under development in Germany that are available the U.S.
There is a need to review the containment code program to deter-mine where it is headed.
It's not clear to me that three efforts plus an international agreement are needed.
The thermal shock (overcooling transient) question seems to have ini-tiated some code development.
A code based on SOLA is under development at LASL and a code called COMMIX (developed by NRC) is being used by E This is being done without really knowing what we need.
As much as I would like to believe that the local thermal hydraulic phenomena (heat transfer coefficient) is dominant I am not convinced.
Mixing processes are possibly more important.
A large difference in the code development effort will result if you attempt to accurately determine the heat transfer coefficient
. Some aspects of this question were addressed at Professor Shewmon's subcommittee meeting on overcooling. I think the ECCS or FD subcommittee needs to hear a great deal more about RES plans before a code development program is j
launched.
RELAP5 code development seems to be pretty much completed.
Predictions all look good.
The presentation by Dr. Ransom leeds me to believe that one can build a model to predict most any process.
The question becomes--
DESIGNAT"3 ORIGINAL 8205050190 820104 PDR ACRS fertified By
[A CT-1406 PDR "L.
L l.
~
f
)
Ivan Catton to Paul Boehnert January 4, 1982 Page 2 is NRR happy with the product? The importance of the analyst cannot be understated and it is time that RELAP5 be turned over to NRR so that the analysts (customer) can become competent in its use.
The version of TRAC now under development at LASL secms to be pretty much complete.
Further developmental efforts will probably not be cost effective. The new fast computational algorithm was of concern to me.
l Now that I understand that it only computes fast when nothing is happening, my concerns are eliminated.
It is taking advantage of the quasi-steady nature of a slow transient. Mass conservation tests show that this is no longer a problem.
Energy conservation tests, however, have not been carried out. Such checks should be carried out to test the numerics and to insure that the analyst does not make an input mistake.
I would like to see automatic energy and mass conservation checks that would warn the analyst if something is amiss built into all the codes.
(
An effort is underway to improve reflood modeling. Water behavior below the quench front plays a role by feeding the entrainment process a larger amount of liquid where the quench front lags because the pins are hotter. Hence, there is more entrained liquid and better precursory cool-ing in the high power part of the core. This is a multi-dimensional pro-I blem both in the two-phase region above the quench front and the coolant below the quench front.
Further, the behavior in the upper plenum and at the upper core support plate yield a spatially dependent boundary condition for the core flow. With all these phenomena in mind, I don't understand the rationale for a more sophisticated one-dimensional reflood model.
A valuable outgrowth of the 2D/3D program has been the experience gained in running TRAC. The results shown us by Ken Williams were very good.
I'm not sure whether or not this necessarily justifies the 2D/3D program.
It was also interesting to find out that a large number of cal-culations have been made for NRR. This is the type of effort that matures a code and should be encouraged.
I don't see what we have gained from the " code assessment program."
I think it's time to take a look at what we have. In my opinion, codes such as RELAP5 and TRAC are highly sophisticated packages of correlations l
a..
Ivan Catton to Paul Boehnert 4
January 4,1982 Page 3 and as such are only as good as the correlations in them.
Further improv-ments will be achieved when better experiments, or re-interpretation of present experiments, are conducted. This is an evolutionary process. A bigger question is the eventual user. He needs to be trained to use the code properly and it's about time we get on with it. I think use of RELAPS with LOFT and Semiscale experiments is an example of the maturing process.
TRAC should replace RELAP5 as a support tool for LOFT and Semiscale.
The TRAC-BWR code is being used in this way as a part of the BWR research pro-The code development efforts should become consulting efforts in gram.
support of the code users and should maintain the codes while keeping close watch on various experimental programs.
Code assessment as planned does not seem to be cost effective.
s i
- ~=w u
_,