ML20052C783
| ML20052C783 | |
| Person / Time | |
|---|---|
| Issue date: | 01/06/1982 |
| From: | Garlid K Advisory Committee on Reactor Safeguards |
| To: | Plesset M Advisory Committee on Reactor Safeguards |
| Shared Package | |
| ML20052C775 | List: |
| References | |
| ACRS-CT-1408, NUDOCS 8205050529 | |
| Download: ML20052C783 (3) | |
Text
._
l UNIVERSITY OF WASHINGTON SEATTLE, WASHINGTON 98195 January 6, 1982 College of Engineering Department of Nuclear Engineering Dr. Milton Plesset, Chairman ECCS Subcomittee Advisory Comittee on Reactor Safeguards U. S. Nuclear Regulatory Comission Washington, D. C. 20555
Dear Dr. Plesset:
l Here is my report on the meeting of the ECCS Subcomittee in Los Alamos j
December 2 and 3, 1981.
I regret that it did not reach you before Christmas.
t As you know I have not been closely involved with either the work of the Subcomittee or with the progress of caiculational methods and experimental programs for more than two years. As a result I am not well-qualified to com-ment on specific technical details that were presented or discussed.
I would like, however, to report the changes in emphasis that I observed and my impres-sion of the strengths and weaknesses of the work being conducted.
Computer Code Development It is obvious that a significant change in emphasis has occurred away from large-break LOCAs and toward small breaks and transients.
I believe this is a healthy development. The increased emphasis on best-estimate calculations as contrasted with evaluation model calculations is also healthy.
In my opinion the use of best estimate results, together with prediction and comparison with a variety of experimental results, is a sound way of developing a more rational basis for licensing and safety requirements. The increased sophistication and speed of the codes being developed is impressive.
In all of these ways, there-
. fore, I believe the code development program is proceeding well.
There cre three areas, however, in which I believe there could be improve-ment:
1.
The transient codes under development, i.e., TRAC-P, TRAC-B, and RELAP 5, all use point kinetic neutronic models. While this is probably adequate in most cases I believe there will be situations where it is not, but I have seen little evidence that the limitations have been seriously addressed. To what extent the limitations of using point kinetics can be experimentally tested is uncertain, because situations where spatial de-pendence is important will normally occur in large reactors undergoing transients and these are rarely available for experimentation. I believe, however, the limitations imposed by using point kinetics can at least be explored analytically and should be so that they are understood by every-one involved.
ghh pDR CT4 400 Benson Ha!!, BF-10 / Telephone: (206) 543-2754
Dr. Milton Plesset January 6,1982 2.
It was encouraging to learn that TRAC-P had predicted at least one surprise, i.e., the GPWR loop-flow reversal (L. Shotkin presentation),
that was later observed experimentally. Considering the effort that has been invested ir, computer codes, however, it is somewhat disconcerting that more surprises have not been uncovered.
I believe there should be concentrated effort to use codes with this specific objective in mind, and would suggest that someone, or a team, qualified by virtue of experi-ence with operating plants and human foibles, be assigned the task of exploring, via computer codes, possible scenarious that might lead to unexpected results.
Perhaps this is already being done, but I have the impression that experirrents with which code calculations are compared are carefully planned in an evolutionary sequence where major surprises are neither desired or expected, and that most calculationc follow the same kind of sequence.
The codes are impressive creations that in my opinion are not being exploited as fully as they might be.
3.
It is important that code development continue in the direction of incorporating human and operator intervention in the analysis of reactor accidents and transients. The extent to which this is now being done is not totally clear, but both NRC Staff members and contractors appear to be aware of the desirability of doing this.
I believe the need is par-ticularly acute in order to appraise the use of comercial simulators for training operators.
In my experience simulator portrayals of accident sequences are often accepted by utility management as accurate and com-plete representations of the accidents' course when, in fact, the models used are grossly simplified.
If operators, as a result of such training experience, are conditioned to expect certain cause and effect relations based on a simulation that has subtle but major shortcomings, the final results in an actual situation could be serious.
Experimental Programs Here also my coments may reflect ignorance of material that has been pre-sented to the Subcommittee or to the full Comittee on other occasions.
l 1.
It is my impression that the experimental programs in Semiscale, the 2D/3D facilities, and in Flecht Seaset are being competently conducted and are yielding useful results, but that the rationale for choosing program priorities is not clearly spelled out.
I believe there is too much emphasis on setting specific timetables and meeting them relative to digesting and distilling the information obtained to determine what phenomena or combina-ations are (a) well, (b) onl And of those in category (b) y reasonably well, or (c) poorly-understood.
and (c), which have the most serious safety implications. An overall perspective is either lacking or has not been clearly presented. Such a. perspective is needed so that specific experi-mental programs can be placed within a framework that is based on a philo-sophy of obtaining better knowledge in areas where ignorance or uncertainty are potentially most dangerous.
l e
3 t
Dr. Milton Plesset January 6.1982 2.
It was my impression that program rationale and perspective were better defined in the case of BWR experimental programs: RES/GE FIST facility; the Refill /Reflood Program; and the Spray Distribution Program.
Unresolved Safety Issues Presentations concerning Containment Emergency Sump Perfonnance and of Water Hamer were useful and showed an intelligent approach to their resolution.
Recomendations 1.
I would suggest that the Subcomittee hold at least occasional meetings de-voted to particular phenomena rather than to experimental plans and programs.
In this way the group could better assess the state of understanding that pre-vails. Two examples come to mind: natural circulation and non-equilibrium effects. Both are important to the understanding and prediction of accident sequences, and it would be useful to assess how well such phenomena are under-stood and can be predicted. Another area is the question of whether and how it is possible to simulate the variety of forms human intervention might take during the course of a transient or accidefit situation.
2.
A second recomendation would be that the NRC Staff devote more effort in the area of operating reactor transients that are related to ECCS systems and the degree to which they are understood or might have been predicted using cur-rent computer codes.
In my opinion the present program under-emphasizes actual operating systems, whose complexity and scope are substantially different from research facilities. More involvement of people with operating and system ex-perience would help guarantee that the concentration of effort on thermal-hydrau-lics does not obscure more mundane areas that may have a greater impact on the safety of the power generation system as a whole.
Best Wishes for the New Year.
Sincerely yours, ff Kermit L. Garlid Professor
-..