ML20053A679
| ML20053A679 | |
| Person / Time | |
|---|---|
| Issue date: | 04/12/1982 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Boehnert P Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1438, NUDOCS 8205270117 | |
| Download: ML20053A679 (5) | |
Text
~
27-W3a" a
9 S
-t T0:
Paul Boehnert
.12 April 1982 um su:aw n 11 FROM:
Ivan Catton ICC 4
SUBJ:
Review of Selected Expe
.p ms,10 March 1982 - Bethesda Six experimental programs planned or underway were reviewed. Five of the programs are in response to the need for further understanding of the themal hydraulic shock question. The other addresses the water hammer issue.
The themal shock program involves analysis as well as experiments. The analysis aspect of the program involves personnel from three national laboratories (ORNL, INEL and BNL), one private company (SAI), and the use of at least four major computer codes (RELAP5, TRAC-P, SOLA and COMMIX).
On the order of 15 man-years will be devoted to the effort. ORNL will play a key role in coordinating these efforts.
From the brief description of the program I have heard and my preconceived notions, I can only conclude that the governments money is not being well spent.
The THEF at INEL is underutilized at present. This has lead RES to look for programs that will utilize it.
The programs suggested (pressurizer and candy cane studies) need some attention.
It would, however, be better to use the f
skilled personnel at INEL to address the problems properly rather than try to use a facility.
Future problems will surely arise where the facility can be used.
Semiscale continues to be a valuable tool.
Most surprises (TMI, GINNA, etc.)
have been themal-hydraulic and most of them can be reproduced on Semiscale.
This tells me that Semiscale is a valuable tool. Whether or not a 2 x 4 Semiscale is needed, or something else, deserves more of our attention as its long tem future does not appear too secure. Most ambiguities faced by operators are thermal-hydraulic in nature. An interesting future use of Semiscale might be to look at whether existing PWR instruments present ambiguous information on a given state and what is needed to remove the ambiguity.
Themal Mixing. The most important and least understood aspect of the themal shock concern is mixirig of thermally stratified flows in the primary systems.
EPRI,in a joint program with NRC-RES, is funding Creare to conduct an experimental program to investigate themal mixing. The earlier 1/5 scale will be replaced with a 1/2 scale facility. The facility mimics a 90* sector of the downcomer and some portions of the cold leg and core barrel. The scaling parameters are Reynolds number, Grashoff number and Prandtl number.
There is no way that scaling can be accomplished. This means a great deal of understanding of the physical phenomena will be needed if one is to use the results with confidence.
EPRI plans to use the data from the tests to validate the COMMIX code (an ANL code originally developed for LMFBR subchannel analysis) and its over-simplistic turbulent mixing model for thermal mixing calculations.
It's not clear that one will be able to separate numerical diffusion induced themal mixing from that predicted by the simple turbulent mixing model. This area of study has been of interest to meterologists for a number of years. They have had only limited success even with very powerful computational support such as is found at NCAR.
DESIGNATED ORIGINAL
$6g 8 2 0 5 2 7 0 l l).
cerunea 27
Catton to Boehner2 page 2 4/12/82 h
Pressurizer Behavior. The advanced codes have demonstrated a lack of ability to represent pressurizer behavior during rapid re-pressurization.
This is the other half of the thermal shock question. During repressuri-zation, the water-steam interface temperature will be at a satur6 tion temperature that follows the pressure rise. This leads to stable stratification with conduction in a thin surface layer governing the condensation rate.
It is not surprising that the nodal representation in the codes does not yield good results.
It was proposed that the blowdown loop at INEL be used to carry out a series of experiments aimed at yielding some kind of a heat transfer coefficient to couple the steam space to the upper node in the pressurizer model.
It is my view that this approach will only yield limited success.
There is a need for analysis before the experimental program is initiated.
Simple conduction models could keep track of the history of repressurization properly and if added to the top pressurizer liquid node would stand a chance of being successful. Then the experiments could be used to determine an effective fluid thermal conductivity as a function of fill rate and level.
Candy-Cane. Behavior. The concern is whether or not we know what void fraction and flow rate leads to cessation of natural circulation in a B&W reactor. The INEL THEF is really not needed for an experimental study of loss-of-natural circulation. Air witn water or freon in a glass pipe would yield much more definitive infonnation for model construction.
Further, EPRI has reported work on this problem in Report No.2067 (or 2068) dated October 1981.
MIT Experimentation. Results of an interesting study of steam-bubble collapse induced water hammer were reported. The goal of the study was to obtain a criterion for bubble collapse. The goal was achieved and a model developed so that a flow rate and subcooling boundary to water hammer could be obtained.
With this criteria, based on slug flow transition, one can insure that water hammer will not occur. The criteria could be written into technical specifications instead of administrative procedu es using words like " fill slowly".
Semiscale Feed /Stean Line Breaks. This series of tests is in response to a request for specific data from NRR and must be delivered by a certain date (that I missed). Three feed line breaks are planned with the largest giving the same steam generator time to drain as a 100% break in a CE system.
The other two breaks will be 50% and 20%. The break will be in the bottom so that the discharge is liquid.
The largest steam line break will be equivalent to a 100% MSL break in a Westinghouse system.
One test will be with off-site power and the other wi t hout.- A third test will be for a 2.6% break to simulate a stuck open safety valve. All three tests will have hot standby as an initial condition.
Steam separators may play an important part in the steam generator behavier following a steam line break. This requires that some attention be given
Catton to Boehnert
~
5 page 3 4/12/82 to how well the Semiscale steam generator works and is modeled. There is also some concern about the 1-D nature of the Semiscale steam generator during a rapid blowdown that may be different than a full scale steam generator.
Further, the extra metal in the Semiscale steam generator may delay cooling of the primary system making the experimental data less valuable.
Good modeling should be able to account for both these effects.
2 x 4 Configuration Testing. RES is struggling to decide whether or not they need an integral facility to represent a CE or B&W NSSS. One of the arguments for needing the facility is that flow phase splits between active and inactive loops cannot be calculated, whereas the size may be so small that separation may not be scalable.
The Semiscale facility evolved from LBLOCA concerns and many aspects important to SBLOCAs and transients are not a part of it.
It may be time to "do it over" with a new emphasis on more timely concerns.
l
Catton to Boehnert page 2 4/12/82
\\
Pressurizer Behavior. The advanced codes have demonstrated a lack of ability to represent pressurizer behavior during rapid re-pressurization.
This is the other half of the thermal shock question. During repressuri-zation, the water-steam interface temperature will be at a saturation temperature that follows the pressure rise. This leads to stable stratification with conduction in a thin surface layer governing the condensation rate.
It is not surprising that the nodal representation in the codes does not yield good results.
It was proposed that the blowdown loop at INEL be used to carry out a series of experiments aimed at yielding some kind of a heat transfer coefficient to couple the steam space to the upper node in the pressurizer model.
It is my view that this approach will only yield limited success.
There is a need for analysis before the experimental program is initiated.
Simple conduction models could keep track of the history of repressurization properly and if added to the top pressurizer liquid node would stand a chance of being successful. Then the experiments could oe used to determine an effective fluid thermal conductivity as a function of fill rate and level.
Candy-Cane Behavior. The concern is whether or not we know what void fraction and flow rate leads to cessation of natural circulation in a B&W reactor. The INEL THEF is really not needed for an experimental study of loss-of-natural circulation. Air with water or freon in a glass pipe would yield much more definitive infomation for model construction.
Further, EPRI has reported work on this problem in Report No.2067 (or 2068) dated October 1981.
MIT Experimentation. Results of an interesting study of steam-bubble collapse induced water hammer were reported. The goal of the study was to obtain a criterion for bubble collapse. The goal was achieved and a model developed so that a flow rate and subcooling boundary to water l
hammer could be obtained.
With this criteria, based on slug flow transition, one can insure that water hammer will not occur. The criteria could be written into technical specifications instead of administrative procedures using words like " fill siowly".
Semiscale Feed / Steam Line Breaks. This series of tests is in response to a request for specific data from NRR and must be delivered by a certain date (that I missed). Three feed line breaks are planned with the largest giving the same steam generator time to drain as a 100% break in a CE system.
I l
The other two breaks will be 50% and 20%. The break will be in the bottom so that the discharge is liquid.
l The largest steam line break will be equivalent to a 100% MSL break in a Westinghouse system. One test will be with off-site power and the other without.- A third test will be for a 2.6% break to simulate a stuck open safety valve. All three tests will have hot standby as an initial condition.
Steam separators may play an important part in the steam generator behavior following a steam line break. This requires that some attention be given
5 g
Catton to Boehnert page 3 4/12/82 s.
to how well the Semiscale steam generator works and is modeled. There is also some concern about the 1-D nature of the Semiscale steam generator during a rapid blowdown that may be different than a full scale steam generator.
Further, the extra metal in the Semiscale steam generator may delay cooling of the primary system making the experimental data less valuable.
Good modeling should be able to account for both these effects.
2 x 4 Configuration Testing. RES is struggling to decide whether or not they need an integral facility to represent a CE or B&W NSSS. One of the arguments for needing the facility is that flow phase splits between active and inactive loops cannot be calculated, whereas the size may be so small that separation may not be scalable.
The Semiscale facility evolved from LBLOCA concerns and many aspects important to SBLOCAs and transients are not a part of it.
It may be time to "do it over" with a new emphasis on more timely concerns.
1 s
--,rs.
-v-..
-