ML20024A618
| ML20024A618 | |
| Person / Time | |
|---|---|
| Issue date: | 05/23/1983 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Boehnert P Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1593, NUDOCS 8306210422 | |
| Download: ML20024A618 (3) | |
Text
{
MEMORANDUM 23 May 1983 T0:
. Paul Boehnert FROM:
Ivan Catton
SUBJECT:
ECCS Subcommittee Meeting Washington, D.C., 10 May 1983
GENERAL COMMENT
S The ECCS Subcommittee was. presented information on seven different topics.
The conclusions I reached following summarizing my notes may be the result of insufficient information. Conclusions I reached for each of the topics are as follows:
The GE SAFER code for BWR-BE calculations must be able to handle mu 1.
tiple channels if it is ever to fulfill its stated objectives.
The B&W OTSG AFW spray does not need to be a good heat removal method 2.
for core uncovery to he avoided if the AFW and Feed and Bleed can be Hence assured and proper instructions are given to the operators.
it is recommended that they be safety grade systems.
RELAP5 analysis of phenomena that involve condensation are highly 3.
Such analysis is being performed for the staff as part of suspect.
their review of the B&W ATOGs. The technical basis for portions of 4
the ATOGs may be weak.
The use of PORVs for SGTR mitigation at plants with OTSGs is necessary.
4.
For CE plants, in my opinion, it is probably necessary.
Resolutior. of USI A-1 (Waterhammer) is incomplete. Waterhammer result-5.
ing in large amplitude pipe motion and its effect on attachments was overlooked.
6.
Resolution of USI A-43 was accomplished in depth. The staff did a good job.
The planned B&W Integral System may be inadeouate to achieve its stated 7.
The staff needs,to present a matrix of tests and analysis that goals.
clearly demonstrates how the goals will be achieved.
SPECIFIC COMMENTS 1.
SAFER Code Review. The SAFER Code, developed by GE to replace their.
8 BWR-EM code, is a measurable improvement over the code presently used by GE for GE has proposed to do a best estimate (BE) calculation licensing calculations.
with SAFER and by comparison with TRAC-BWR and data from FIST or elsewhere esta-blish an " adder" for the PCT. The " adder" is to account for conservatisms in SAFER and add a safety factor to obtain a value of PCT for licensing.
In my view this approach is unsatisfactory. The staff suggested the following alter-nate approach:
i I
g,g g ORIGI M
[
.s 8306210422 830523 PDR ACRS Certified 37
[
CT-1593 PDR
f ECCS Subcommittee Meeting (cont'd.)
May 23, 1983 3
1.
Establish an " adder" to the SAFER BE results that will give the 95% bound on PCT 2.
Calculate PCT using a BE code with Appendix K " legal" requirements Add the " adder" to a BE PCT without Appendix K limits to obtain a 3.
BE 95% PCT 4.
Use the larger of the two values of PCT to determine whether or not the legal requirements have been met.
It There are a number of problems with the approach suggested by the staff.
will be a major undertaking to sort out Appendix K.
In some cases the heat transfer coefficient is specified, in some a correlation is named and in others What is in the law is thought un-it only requires that one assures conservatism.How the second step is treated will be der some cases to lack conservatism.
crucial if meeting legal requirements is intended.
Cer-Obtaining a BE calculation with the SAFER code will be difficult.
tain observed physical phenomena are not calculable using the SAFER code be-cause it has a one-channel core.
Even though we all may be convinced that SAFER will yield a conservative calculation of the PCT, and I am, we cannot The single-channel representation of the core in quantify the uncertainties.
SAFER does not allow the more monotone behavior cf the T/H variables in the hot channel resulting from the peripheral channel draining.
Rather, the single-channel drains the above core region periodically. We all can cite cases where parts of the solution "look okay" whereas some of the details are nonsense.
SAFER cust have a multiple channel capability if it is ever to be considered a BE code Without multiple channel capability tne uncertainties will never be quantifiable.
2.
B&W Auxilliary Feed Spray Effectiveness. The AFW system can fill the Once at the 95% level, the steam generator from 50 to 95% in fifteen minutes.Further, the HPIs can be used cooling is sufficient to prevent core uncovery.
in a Feed and Bleed mode. When you look at the range of events that might occur, This means that both you find that you will need either the AFW or the HPI.
should be qualified to assure availability.
During this portion of the meeting we heard a lengthy presentation on The some ill-conceived experiments that did not yield anything meaningful.
poor heat transfer analysis did not need to be presented.
3.
RELAPS Analysis of Boiler-Condenser Cooling. The calculations yielded a number of very improbable results.
Showing us a series of bad results implies It appears as if RELAP5 condensation models something about the staff or us.This leads me to believe that the RELAPS calculations need m are incorrect.
As a result,
. more serious review before they can be given any consideration.
The tech-the staff use of RELAPS in their B&W ATOG review may not be adequate.
nical basis for the AT0G may be weak. The only way one can rationalize going forward with the ATOG is that they take advantage of the AFW and HPI without resorting to what they think the reactor T/H parameter behavior will be like when it is in the Boiler-Condenser mode.
4.
Use of DDRVs For SGTR Mitigation. The arguments given by the staff about the need for a PORV for plants with OTSGs were convincing. Giving CE credit for a safety grade pressurizer spray seems to remove the requirement for 7
. ~..
t s
.P0RVs for their plants.
I have not heard arguments yet that convince me that the CE position should be accepted. There are a number of arguments for PORVs.
One of the most convincing postulates a steam bubble'in the reactor head.
Un -
der such circumstances the sprays immediately become ineffective. The ECCS sub-commettee should meet with the staff and CE to resolve this issue.
S.
Resolution of USI A-1, Waterhamer. For the most part the staff has It wasn't, however, as thorough as it might have been. The done a good job.
type of Waterhammer event that lead to the atmospheric dump valves being used It to reject decay heat during the TMI-2 incident was completely overlooked.
seems to me that there ought to be a requirement for at least an inspection to sce that air lines, electrical connections and instrument piping can survive large pipe motion (there is no question about the pipe itself) on pipes where' valve or pump control has safety implications.
6.
Resolution of USI A-43, Containment Emergency Sump Performance.
Resolution of USI A-43 has been accomplished in depth.
I have followed this program since its inception and must congratulate the staff on a job well done.
7.
B&W Integral System. The need for an integral system was clearly demonstrated by the staff presentation of RELAP5 calculations of a OTSG in the condensation mode. My views, however, remain unchanged from those I expressed following the 8 Feb 1983 ECCS Sbete meeting on the GERDA upgrade. The big ques-tion is always one of scaling. Exact similitude has been shown to be impossible time and time again. What one would, therefore, like to do is optimize the scaling in some fashion with parametric studies of various parts of the system being done as,a separate effects effort. The present effort starts by using that the time scale must be unchanged and all elevations must be maintained.
This leads to severe geometric distortion. As we all know, geometric similitude is a cardinal rule for scaling. Reflecting on all the questions and debate (and dollars sometimes wasted) resulting from semi-scale, LOFT and other integral facilities, it seems to me that freezing a design because "we don't know what else to do" is without justification.
Let's not spend the money until we have a matrix of tests and analysis that will yield the desired result. To emphasize my view, I will discuss several aspects of the B&W Integral Facility.
The best way to discuss the planned facility is to ask oneself if it will yield answers to the questions posed for it to resolve.
First, single-phase natural circulation does not require a facility for its study. We can predict natural circulation behavior more accurately than the small scale facility can reproduce a full size plant's behavior. Questions of loss-of-natural-circulation can only be answered if the hot leg U-bend is properly scaled and its not clear Recovery of single-phase natural circulation by pressurization that it is.
is not properly scaled unless U-bend volume to pipe area is properly scaled and I don't believe that it is. Conduction in the pipe walls plays a role and its scaling will not be possible unless some fancy new pipe designs are developed (double wall with a pressurized annulus). Two-phase flow in c 2 1/2 inch dia-meter, 70 ft. high pipe will not be the same as a 48 inch diameter, 70 ft. high pipe. Aspect ratio will play some role (maybe minor). The experiments where one steam generator cools differently than the others will lead to (or could) loop to loop oscillations that require that all the vent valves be adequately modeled.
In the proposed system all four loops interact with the core via a single downcomer and vent valve. This is inadequate for reproducing the anti-cipated loop-to-loop behavior.
As can be seen, many of the questions that lead to a perceived reed for the facility may not be answered. Further, the facility as presently conceived A clear path had better may lead to the codes emphasizing the wrong phenomena.
be found soon or we will be subjected to several years of rationalizations and explanations.
.