ML20133G972
| ML20133G972 | |
| Person / Time | |
|---|---|
| Issue date: | 07/01/1985 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-2324, NUDOCS 8508090141 | |
| Download: ML20133G972 (8) | |
Text
'. Ades- 2349' k~ -
}/)/ of'65f 5
' DATE ISSUED: 7/1/85
- 1,1 tl SS ACRS SUBCOMMITTEE MEETING MINUTES ON EMERGENCY CORE COOLING SYSTEMS (ECCS)
JUNE 12-13, 1985 ALLIANCE, OHIO PURPOSE: The purpose of the meeting was to continue discussions of the joint NRC/B&WOG/EPRI/B&W IST Program and related programs. A tour of the MIST Facility was conducted on the afternoon of June 12, 1985.
MEETING ATTENDEES:
ACRS NRC/ Contractors D. A. Ward, Subcommittee Chairman W. Beckner, RES C. Michelson, Member R. Lee, RES I. Catton, Consultant R. Jones, NRR V. Schrock, Consultant L. Shotkin, RES A. Cappucci, Staff T. Knight, LANL T. Larson, INEL INDUSTRY H. Carter, B&W J. Cudlin, B&W J. Gloudemans, B&W R. Turner, B&W A. Hossier, WPPSS ;
J. Sursock, EPRI D. Sallet, University of Maryland A. Hashemi, SAIC HIGHLIGHTS,. AGREEMENTS AND REQUESTS
- 1. Two reports are underway for the IST program. One report will address how experimental and analytical programs will be used to address the MIST atypicalities relative to major technical issues such as 8508090141 850701 DESIGICED ORIGI2 L PDR ACRS 2324 PDR Cortiflod By_
~
l .-- _- -
L i
- 1
- ECCS Meeting Minutes June 12-13, 1985 f
interrupted natural circulation, estab'aishment of boiler condenser mode, reestablishment of natural circulation, long tenn cooling, interloop interactions, and combined primary and secondary side blowdown. The other report will examine each experimental facility's scaling tech-niques and provide a common comparison for tests. The reports are expected to be released sometime in August or September 1985.
- 2. The integration of the separate effect test data into the IST analyses should be complete sometime in 1986 with the interfacility tests taking place over the 1985/1986 time frame. The IST support projects are expected to be complete by the end of 1985 and the MIST Program complet-ed by the end of 1986.
- 3. In response to questions cor.cerning primary / secondary side blowdown Mr.
Sursock explained that the MIST Facility could not reproduce the heat transfer characteristics of a double ended break (DEGB) of a main steamline. The major problem is the limitation of the MIST Facility to remove steam on the secondary side. However, a break less than a DEGB could be postulated and then extrapolated to a DEGB using computer codes. The arguments related to the above discussion have not been published, therefore it is considered open until the experiment and tne code analysis are complete.
- 4. During discussions related to scaling, Mr. Michelson asked if geometric scaling was considered in addition to thermo-hydraulic scaling.
T. Larson indicated that it was and it should be addressed in the reports discussed above.
- 5. The University of Maryland Facility has eight passive vent valves on the reactor vessel. The retarding forces on the vent valve can be changed creating an absolute seal if desired or allowing them to remain par-tially open. The valves have been tested to determine their single phase flow rate but the two-phase characteristics have not been
\ ECCS Meeting Minutes June 12-13, 1985 obtained. It was pointed out that the University of Maryland Facility would exam the effects of transients on the vent valve and MIST would ,,
examine the effects of vent valves o,n transients.
- 6. The Subcommittee raised concerns during Dr. Sursock's discussion of transfer functions used in the modeling of the SRI-II facility.
Dr. Schrock suggested that many thousands of transfer functions would have to be examined in order to determine system instabilities. Dr.
Sursock disagreed indicating that only the most adverse situations from the stability viewpoint would have to be examined. He stated that only a few of those existed. The Subcommittee suggested that loop to loop transfers should be evaluated in addition to the common downcomer element. There could be an opportunity for feedback from one cold leg to another while the SRI-II model only allows feedback through the comon downcomer. Dr. Sursock indicated that this area would be looked into.
- 7. Dr. Sursock pointed out that the results of an experiment carried out in the 1/3 scaled B&W hot leg geometry (two-phase flow regimes) to compare collapsed liquid level in 4-inch and 12-inch pipe show that slug flow doesn't occur in the 4-inch line, only bubbly flow. Also, pronounced slug flow appears to be formed out of the range expected for small break LOCAs.
- 8. Dr. Catton raised questions related to the scale up of the OTIS Facility tests (19 tube steam generator). He suggested that full wetting of steam generator tubes could occur at OTIS where it might not in a full scale steam generator. Dr. Catton suggested that this was another area that scaling studies could be directed to. In response to Dr. Catton's coments, Dr. Knight indicated that in order to predict steam generator response to some split of wetted and non-wetted steam generator tubes separate effects tests would probably be required.
L.s
' ECCS Meeting Minutes June 12-13, 1985
- 9. Mr. R. Jones,'NRR, pointed out that the results of the OTIS/GERDA tests would not be used directly in NRR. NRR will be studying the computer codes as the link between the test data and the actual plant.
- 10. Mr. Jones, NRR briefed the Subcommittee on the recent total loss of feedwater event at Davis Besse. At about 1:35 am on June 9, 1985 Davis Besse was operating at 90% power in a stable mode. One main feed pump was on automatic with the other pump on manual because of apparent governor problems. The main feed pump on automatic tripped and the plant attempted a run back. Within 30 seconds the RCS pressure had increased to 2300 psi (reactor at 80% power) causing a high pressure reactor trip followed by turbine trip. About the same time a spurious trip of the steam feedwater rupture control system occurred. At about 1:40 am the operator manually initiated the auxiliary feedwater pumps, but they tripped imediately on overspeed. Over roughly the next five minutes, the steam generators started to dry out, with complete dryout (water level at approximately seven to eight inches) by 1:50 am. The PORVs were cycled approximately three times. Mr. Jones also indicated that that the operator initiated the HPI pumps manually, but this was not a significant actor in the event because of the 1700 psi HPI shutoff head. However, he suggested that this was a proper action if only to set up for feed and bleed. About 1:50 the operator manually started the half capacity motor driven auxiliary startup feedwater pump, recovered steam driven auxiliary feedwater pumps by 1:53, and the plant was under normal control by 2:00 am.
- 11. In response to Subcomittee questions related to costs for OTIS/ MIST type and plant type calculations, Dr. Knight pointed out that the MIST /0 TIS calculations were 1-dimensional models while the plant type calcul'ations use 3-dimensional models. The plant calculations run slower, therefore longer for a given transient of finite length, i However, for a facility like SRI-II the code run time appears to go up as the scale cf the facility goes down. He concluded that a plant
\ ECCS Meeting Minutes June 12-13, 1985 4
analysis would run faster than a 1-dimensional analysis running essen-tially the same transient at MIST.
- 12. Dr. Catton questioned why B&W chose to use RELAP-5/ MOD-2 instead of other codes. His concerns centered around the 1-dimensional modeling of the downcomer. Mr. Cudlin, B&W, pointed out that they believed from the standpoint of system level effects that the downcomer was essentially 1-dimensional. Dr. Catton pointed out that the downcomer might not be 1-dimensional and asked B&W if they had plans for other tests if it turned out not to be. Mr. Cudlin indicated they did not. Dr. Catton suggested that the B&W view was narrow in that the ultimate goal is to understand the phenomenology in the plant.
- 13. In response to previous questions related to steam generator modeling, Dr. Knight indicated that LANL was taking at face value the data from B&W which indicates that three tubes were wet and 16 were dry. He further explained that the envisioned model would not lead to predictive capability in that area because of an inability to analytically split wet and dry tubes. The wet and dry tube separation would be specified on input.
- 14. Dr. Knight stated that thermally stratified flow is inherently beyond the capability of the TRAC and RELAP codes. The problem appears to be a hotter liquid displacing a colder liquid due to buoyancy effects during counter-current flow. This can be caused by HPI into the cold leg or cold liquid penetrating pump suction driving the oscillations.
- 15. Mr. Ward questioned the extent to which the MIST Facility actually models the plant from the TRAC Code perspective. Dr. Knight indicated that precisely modeling the plant was was not of concern because the facility ultimately sees the same type of phenomena as the plant.
Dr. Schrock pointed out that getting the same phenomena was not ade-
' quately represented in results presented so far at the meeting.
l u- .
ECCS Meeting Minutes June 12-13, 1985
- 16. In discussions related to the phenomena of higher pressure (secondary pressure) in the broken loop than the unbroken loop due to thermal decoupling of the unbroken loop because of flow stagnation, Mr. Michelson suggested that the operator would become confused as to which loop was broken. The reason for this confusion is that the operator normally relies on pressure indications to determine break locations.
- 17. Dr. Knight discussed the conclusions of the plant comparison analysis I
stating that MIST and plant transients show similar behavior. The pressures and loop voiding compare reasonably well. There appears to be
- an offset in plant pressures but is of no concern to LANL. The cold leg temperatures differ because of mass flow differences and plant cold leg flow circulation.
- 18. The Subcommittee questioned the importance of single channel modeling vs. dual channel modeling in the steam generators, i.e.: some tubes wetted and some dry. Dr. Catton suggested that a model needs to be developed to evaluate this phenomena. Flooding at steam generator tube spacer grids and supports needs to be calculated. RES stated that there was not a first principle model to predict auxiliary feedwater sprays.
- 19. Mr. Ward requested concluding comments from the Subcommittee, these are presented below.
Dr. Schrock - Studies of vent valve behavior need to be emphasized with the University of Maryland experiment having the most poten-tial for gaining further insight. The ultimate scaling has to be to the full size plant which is essentially represented by the code.
There may not be enough emphasis in reviewing codes to determine they meet this scaling step. Because of discussions related to modeling steam generators and the first principle calculation characterization of the TRAC Code more emphasis needs to be placed
~.
'- ECCS Meeting Minutes June 12-13, 1985 on this modeling. The codes should be able to generate highly accurate steady state data. If they cannot then doubt is cast upon their ability to calculate transients. More attention is needed in this area. There are areas in the power plant comparisons where the phenomena appeared not be as similar as indicated by Dr.
Knight. For example the broken loop voiding time in MIST appeared to be a lot longer than in the plant and has rather different characteristics. This is significant to the ability of the codes to follow the details of system behavior.
Dr. Schrock indicated there is merit to a follow-on full power testing program at MIST. However, in view of high cost estimates, full power testing should be reconsidered in context of the future general purpose Integral Test Facility.
- Dr. Catton - Doesn't believe full power testing is necessary.
Scaling the proper parameters and running in two different pressure regimes will give the same information. Steam generator modeling needs attention. Tuning of codes to small scale facilities may lead to trouble in extrapolating to full scale facilities. This is shown by plant comparison studies of the importance of multi-dimensional downcomers. Research should evaluate whether the codes can handle a multi-dimensional downcomer with a 1-dimensional core.
Multi- dimensional characteristics of vent valve behavior emphasize the need for using both the University of Maryland and SRI-II in any code assessment. MIST hardware is well in hand. Scaling studies will yield useful information.
- Mr. Michelson - Minor concerns about the vent valve arrangement at the MIST Facility. Will have to await results to determine the importance. Partial answer will come from the University of
ECCS Meeting Minutes June 12-13, 1985 Maryland; Recommends reducing the noise level in the MIST Facility Control Room. Question on full power testing needs to be answered.
- Mr. Ward - More comfortable with the integration of programs and facilities than a year ago. TRAC calculations of the University of Maryland experiments should be considered. Weaknesses in MIST to model asymmetric effects such as downcomer and vent valves. Steam generator auxiliary spray in an important issue. Need to determine if it is a safety concern. Suggests sensitivity analysis. Code assessment for B&W should not get lost in the overall code assess-ment effort.
- 20. Mr. Ward requested feedback on Subcommittee comments from NRC/RES within a few weeks. He specifically mentioned comments on the TRAC analysis of the University of Maryland transient and the steam generator auxiliary.
feedwater distribution question. -
-t NOTE: Additional meeting details can be obtained from a transcript of this meeting available in the NRC Public Document Room,1717 H Street, N.W., Washington, DC, or can be purchased from Ann Riley and Associates, Ltd., 1625 I Street, NW, Suite 921, Washington, DC.
l i
4 e o 9
l l