ML20151D146
| ML20151D146 | |
| Person / Time | |
|---|---|
| Issue date: | 03/09/1988 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Boehnert P Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1923, NUDOCS 8804130417 | |
| Download: ML20151D146 (2) | |
Text
CT-/M3 UNIT E U S1 ATEs w-
[}
')
NUCLEAR HEGUL ATORY COMMISSION l
- '- M f.
c ADVISORY COMMITTE E ON REACTOR sat EGUARDS b
/
v ASm% f 0N, r). C. 20555 o
March 9, 1988 MEPORA2M TOPM"'"'""*PtM *EMim t g
FROM:
l Ivan Catton d
SUBJECT:
TECHNICAL PROGRAM COMMITTEE MEETING - CODE SCALING APPLICABILITY AND UNCERTAINTY (CSAU) METHODOLOGY -
NICHOLSON LANE BUILDING, FEBRUARY 22-23, 1988 Scre interesting "lessens" are being learned from the CSAU effort and are to be compiled in a special docunent for future guidance.
For e x a r.ple :
(a) Much of the data obtained by RES from bypass studies is no longer any good because the reports are incomplete.
To insure that the data will be useful, pre-test calculations should be done to make sure all the needed measurements are being made.
This should be followed by an analysis of the experiment.
If the test does not justify the pre-and post-test analysis within a short time, then it does not need to be run.
Much of the Severe Accident experi-mental program might be eliminated if this criteria were imposed.
(b)
Plant code input decks assembled some time ago were found to have modeled flow bypass blockages that do not exist. Apparently, they were built into the input deck and the users never checked. The i
modeled blockages are one of the reasons bypass is so conservative.
The lesson here is that one can never be too sure about what others have done.
NRC should have an audit procedure for its own codes as well as the vendor codes.
(c) During the initial stages of ECC injection, TRAC uses one hour of CREY time to calculate for one serend of real time.
You are paying a high price for very little.
The message here is that general purpose codes like TRAC or RELAPS may not be a good substitute for thinking.
In many cases, one should write the code for the specif-ic task at hand, i
These and a nurber of other lessons have been learned while exercising the CSAU process. Whether or not they will be heeded by RES is another matter.
I would hope that the lessons will be given serious consid-eration before any more money is spent on code development.
8804130417 800309 PDR ACRS CT-1923 PDR i
i
t 7
[
CSAU Methodology 2/22-23/88 Mtg March 9, 1988 4
l i
I Some time was spent discussing the coments of the Members of the Peer Review Group.
For the most part, there were no surprises. Typical conments were the followir,g:
' TRAC is a glorified correlation exercised against a great deal of i
data.
- It is very dangerous to extrapolate beyond your current data base.
' CSAU arguments are defensible only because data is available.
- Sone small-scale data should be used and the comparisons with data shculd be shown.
- The n.ettedolos, offers no criteria for establishing when expert j,
opinion is so immature that the proposed approach is irapplicable.
- tcw ranked phcrcrena with high uncertainty combined with high 5
ranked phenomena were not well addressed.
1
- There is concern about the skills within NRC to assess the efforts of others rcpeating the demonstrated process. This concern points to the need to be more prescriptive in exercising the process.
- Nodalization is a concern,to some of the Peer Review Group members.
I Two concerns resulting from the Peer Review are being addressed.
The first, the question of cross products (s uncertainties in interacting parameters)ynergistic effects due to
, was addressed and found to be inconsequential. This is an example of why there is concern about the capability of others within NRC to carry out the process.
The members I
of the TPG have had enough experience with LOCA thermal hydraulics that they :ould rake such judgments.
In seine cases, this experience came from actually running experirents and reviewing, in detail, the data from others' experiments.
The second concern was raised by Westinghouse.
They noted that the treatment given the hot channel was incorrect because it put a hot pin into a channel with core average j
thermal hydraulics.
Screhow, this was overlooked when the set of TRAC runs was set up. An extra ring of nodes will be put into the vessel, at i
some expense, to address this question.
The results will hopefully be j
available at the next TPG meeting.
The next stcp in the CSAU process is to address the reflood peak clad
]
temperature.
To do this, uncertainties in the bypass modeling must be addressed.
Early data from BCL and CREARE are not of much use because i
the reports were to some extent incomplete and too much time has passed.
TRAC calculations show strange behavior.
The interfacial dr g is too j
high and the fluid just refuses to fall into the lower plenum until the y
steam flow from the core is very low. After a great deal of discussion, 1
l 1
l
CSAU Methodology 2/22-23/88 Mtg March 9, 1988 it was decided that the interfacial shear could be locally adjusted.
This means that the values used in the downcomer can be different than they are elsewhere. A multiplier will be used and will be varied over a range dictated by the UPTF full scale downcomer data. Without the UPTF data, this uncertainty would be very difficult to quantify. The other
]
parateter of importance during the bypass phase of the accident is 9
condensation.
Here T3AC uses an oversimplified approach based on a i
constant Stanton numbar. A multiplier will be used here as well.
The l
problem is what to use.
UPTF will be a help, but a great deal of judgement will also be necessary.
Gerry tellouche will address this 1
problem before the next TPG reeting.
l i
f 4
l I
)
1 i
I i
l 1,.
--