ML20040E168
| ML20040E168 | |
| Person / Time | |
|---|---|
| Issue date: | 09/18/1981 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Boehnert P Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1392, NUDOCS 8202030207 | |
| Download: ML20040E168 (4) | |
Text
O
~
LOS AN* ELES: SCHOOL CF EN!!NEEi!NG g
AND APPLIED SCIENCE S4)/2 /- 8-$A i
T0:
Paul Boehnert 6, T - / 3 (/ 0LJ FR0ft:
Ivan Catton
SUBJECT:
BWR BDHT Review Group Meeting 18 Septenber,1981, Willste Building The review group is improperly named the BDHT Review Group. The review group considered BWR/LOCA ECCS research programs in a general way.
NRC sponsored BWR thermal hydraulic research includes experimental studies using the Two-Loop Test Appar atus (TLTA), a Full Integral 5,imula-tion Test (FIST), and a refill /reflood apparatus. BWR TRAC code develop-ment and assessment rounds out the BWR research effort. Most of the work is under joint sponsorship of NRC, LPRI and GE.
Past studies with TLTA nave shown a great deal of conservatism in the BE EM code and that the physical phenomena of a LBLOCA and a SBLOCA are similar.
Comparison with predictions show that details of the ther-mal hydraulics are important if good predictions are to be obtained.
Separate effects studies needed to obtain sufficient knowledge to predict the details were not evident and will probably not be obtained from the new f acility.
Two Loop Test Apparatus (TLTA). The TLTA has reached the end of its usefulness.
In general for the LBLOCA it was fourd that heat transfer from the heaters to whatever medium surrounded them was good.
CHF was only observed for very high heating. The result being that the system enters the later stages of a LOCA pretty cool. Comparisons with Appendix K show predictions to be con ?rvative by several hundred degrees (*F).
The peak clad temperature never exceeded 1000*F. The primary reason for this is the lengthy period where Appendix K requires a zero heat transfer coef fi ci ent.
It appears to me as if there is still a great deal of specu-lation about what is actually happening.
Depressurization rates were found to be slower than expected. This lead to board notification; not a safety issue, rather a questioning of ones ability to cal culate.
One argument as to why the slow depressuriza-8202030207 810918 PDR ACRS CT-1392 PDR UNIVERSITY OF CALIFORNIAHLetterhead for interdepartmental me)
l LOS AN':7 LES: SCHOOL CF EN 7!NEERIN3 l
AND APPLIED SCIENCE I
tion occurs is that ECC water causes the jet pumps to c.over and eli-minates a path for steam flow out cf the vessel. Another argument is that more water in the core results in more steam which results in a decreased rate of pressure decay.
It seems that the TLTA did not have sufficient instrumentation and the codes still have too much trouble modeling CCFL and ECC penetration to establish what is I.ctually occuring.
There is, however, no doubt about the excessive conservatism in the rule.
There are too many scaling compromises for tr.e TLTA to be a " good system" for SBLOCA investigation.
For example, the annulus area is too large and the lower plenum is too shallow. On some tests, ECC water flowed down the heated. channel and up the bypass. This occured late in the test so it is a question of understanding, not of safety. Quali t a-tively, the SBLOCA looks like a LBLOCA with a stretched time scale. The codes do not predict the phenomena probably because CCFL is not handled as well as it might be. The codes did not predict coolant getting into the core much less predicting a flow reversal.
For the most part the SBLOCA tests results were relatively unexciting.
RELAP4/M006 predicts most of the observed phenomena as far as it goes.
It becomes unreasonably expensive once subcooling is a considera-tion. Timing of various events is off and when jet pump uncovery occurs is missed. This makes the predicted behavior quite different than the observed.
BWR Full Integral Simulation Test (FIST).
The TLTA is being upgraded to improve the simulation capability for SBLOCA's and non-LUCA transients.
The upgrade is actually a complete rework of the existing f acility. The new f acility will, hopefully, correct TLTA s'caling compromises.
It will be full height ana contain one 8 x 8 bundle of simulated fuel rods. The facility will be a significant improvement to TLTA.
It was noted that "no justification could be found to require multiple' bundle simulation."
This find should be weighed against the Japanese two bundle f acility re-sults showing different behavior, the multi-dimensional flow observed above the core as shown by the Lynn facility results anc the results of Lahey's four bunale freon facility.
It's not clear how FIST will add to data that could result from the Japanese program.
UNIVER$17Y OF CALIFORNI A-(Letterbead for interdepartanental use)
~
AND AFFLIED EC3ENCE t
Six tests are planned for the first phase of FIST testing. Ther e
\\
will be 2 LBLOCAs to allow a comparison of the new facility with TLTA.
This will be followed by tests representing an IBLOCA:, a steam line break and two SBLOCAs.
The testing is scheduled to start during the last quarter of CY 1982 and to end in Mid-CY 1983.
A later phase of FIST use is still being worked out.
It will pro-bably include ATWS simulation, separate effects testing of ATWS, core uncovery and naturai circuIation, LOFW and turbine trip. At present, I do not know how FIST will be used for ATWS simulation. A o itical review is needed.
There are no plans to do anytning to alleviate the system heat loss.
Dr. Sozzi indicated tnat ne was not concerned about it.
In that FIST will surely be part of the " Required Problem Program" and vendor codes do not have two-sided heat slabs, the heat loss question should be serious-ly addressed before tne system is f abricated.
ihe agonies Semiscale has had with the heat loss problem should be incentive enough for the FIST program.
In my view a system heat balance should be possible or the utilities should not be required to make calculations of its performance.
BWR REFILL /REFLOOD PROGRAM.
The objective of this program is to improve understanding of refill-reflood phenomena, to develop and qualify advanced best estimate LOCA codes and to provide data for assessing licensing codes.
Data for this program comes primarily from the 30 deg. Steam Sector Test Facility (SSTF). TRAC is being used as the framework for a BWR best esti-l mate code.
SSTF work on core spray distribution is completed with the refill /
reflood work being in progress. Tne purpose of tne refill /reflood pro-gram is to evaluate steam-water mixing and supply data for qualifying best estimate models.
It was found that a Wallace type CCFL crtterion gave good predictions of the SSIF results when there was only one channel.
Data from tests with multiple channels show more drainage to the lower plenum. The reason l
is probably becaust some channels, the low power edge bundles, are full l
of liquid whereas others are steam only or a mix of water and steam. The l
bypass channels fill very fast. The spray distribution over the bundles UNIVERSITY OF CA1.IFORNIA--(htterhead for interdepartreestal use)
e LOS ANCLES: SCHOOL GF EN!IMEERIN3 AND APPLIED SCIENC3 n
and the bundle flow behavior results in a highly three-dimensional problem.
The TRAC-BWR does not have a model for the spray distribution. A user selects a spray distribution using the SSTF data as a basis. With this reasonable predictive capability was shown.
I r
UNIVER$1TY OF CALIFORNIAMiateertwed go, gg,g,%g 4
-