ML20198H466
| ML20198H466 | |
| Person / Time | |
|---|---|
| Issue date: | 01/24/1986 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1482, NUDOCS 8601310037 | |
| Download: ML20198H466 (384) | |
Text
OR G NA _
M ## )
l O Uh11SD STATES NUCLEAR REGULATORY COMMISSION l
IN THE MATTER OF: DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON EMERGENCY CORE COOLING SYSTEMS e
4 i
O .
LOCATION: PALO ALTO, CALIFORNIA PAGES: 164 - 339 DATE: FRIDAY, JANUARY 24, 1986 ,
ARR8 M M M 292pdemoveimm ACRS0ffice ACE-FEDERAL REPORTERS, INC.
O U OfficialReporters 444 North CapitolStreet Washington, D.C. 20001 (202) 347-3700 1 seat.
4]; t,pa,I~ ! c-g ,g, NATION % TOE COVERAGE
ym.m,~yp < 2 i
s
.4 3
ii
)
. 1 i
e! (
i i
1
-s i
(
t i
i
.j 1
ls y-~n_~aucun , ; a: ,:w.- :wwa + w+~**'**" l i
1, 1
TRO4 01 DF75TE B. h1IITE ,
l RETURN ORIGINAL 'IO B. hlIITE, !
a ACRS-1016-H !
j I i
6 s
Thanks!
1
- Barbara Jo hhite g I
, ,. 4-1407 <
w 1 mwww. _ . --
)
- l 2
, 4
'i a
J 1
d 3
'1e
-1
.1 i
),
-a
. j f
1 i
i 1
4;. n !
164 o
A h,v 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3
4 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 5
6 7
8 9 Auditorium Electrical Power Research 10 Institute 3420 Hillview Avenue 11 Palo Alto, California 12 Friday, January 24, 1986
(^T 13
(-
%/
G 14 The meeting of the Advisory Committee was 15 reconvened, pursuant to adjournment, at 8:30 a.m.
16 17 Present:
18 DAVID A. WARD, Chairman CARLYLE MICHELSON, Member >
19 JESSE C. EBERSOLE, Member CHARLES WYLIE, Member 20 H. ETHERINGTON, Member IVAN CATTON, Consultant 21 .
C.L. TIEN, Consultant THEO THEOFANUS, Consultant 12 VIRGIL SCHROCK, Consultant PAUL BOEHNERT, Designated Federal Employee 23 24 25 y-O u
J
..: : . _a a w._.____. _____ .._ . e . , . . . _ . . _ _ . . . . _ . . . _ _ _ _ . . _ ..
PUBLIC NOTICE BY THE UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS FRIDAY, JANUARY 24, 1986 The contents of this stenographic transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards (ACRS), as reported herein, is an uncorrected record of the discussions recorded at the meeting held on the above date.
No member of the ACRS Staff and no participant at O this meeting accepts any responsibility for errors or inaccuracies of statement or data contained in this transcript.
t
---.0.-.--~-
. . u : :~. . . .- a.- -.. - .w - . - . . . . . . . . . . - - - .. ..
165 1 EEREEERIEEE 2 8:30 a.m.
3 CHAIRMAN WARD: We are reconvened. And our 4 first speaker this morning is Tom Larson. )
5 MR. LARSON: Good morning. My name is Tom 6 Larson. I have been asked to give you the current status on 1
7 the so-called IST Scaling Report. And at the meeting last l
8 June at Alliance, I think we referred to this as Scaling 9 Report No. 2.
10 What I will do today in the allotted time is 11 give you an idea of what the status of the report is, the 12 content, and as time permits we will take a look at some of
( ,)
.1' 13 the results that are given in this report.
14 Clearly, there is too much material in the 15 handout to go over in any detail.in the allotted time we have 16 this morning. So I will be skipping over some of these 17 slides.
1g (Slide.) ,
l 19 MR. LARSON: The presentation will proceed 20 along the lines shown on this slide. I will first discuss 21 the objectives, content, and the status of the report, get 22 into the results a little bit, some of the scaling analysis 23 that was done on the two facilities, try to point out some of
/O
')
' 24 the problematic areas involving comparing the data, then go 25 into some of the potential methods that I think may be used l
L
166
,, 1 to compare the data from the three facilities, talk a little
)~' 2 bit about counterpart experiments, and then some conclusions 3 that were arrived at as a result of this work.
4 (Slide.)
5 MR. LARSON: If you recall back in June when 6 this report was first discussed, or the concept of this report 7 was first discussed, uhe statement was made that there was no 8 document at that time that really outlined the objectives of 9 each facility in the program or gave a description of the 10 scaling philosophy or the geometric parameters in any'of the 11 three IST facilities -- at least not all in one spot.
12 Now, each of the facilities have a document that
~
Os/ 13 describes the machine and some of the philosophy behind the m
14 design. T,here was a desire to get this all conglomerated as 15 really a data base in one spot for ease of use. That then 16 constituted one of the objectives of this report, simply a 17 data base to provide a concise, accurate data bank for 18 geometric parameters, scaling philosophy and hardware 19 limitations, that sort of thing, for each of the facilities.
20 Also, at that same time, since each of these 21 facilities has different capabilities, to provide some kind 22 of an analysis of different phenomenalogical areas for what 23 we might expect in terms of so-called scaling distortions.
) 24 Also, one of the objectives was to provide some
! 25 insight as to how we can relate the data from each of the
-..~
167 1 three f acilities, primarily because we know each of the
( 2 facilities is a different design, differently operates under 3 different conditions, et cetera.
4 And finally, again in light of those differences :
5 in operating conditions and geometric parameters, there was a 6 desire to give some insight on, hey, how can we run a counter-7 parc experiment to examine knat the true effects -- or at 8 least get an idea of what the effects of those different 9 operating conditions and scaling criteria were on the 10 phenomena occuring at each of the facilities.
11 (Slide.)
12 MR. LARSON: This is simply a condensed table O
kl 13 of contents for the report. I don't want to spend any time 14 on it, really, than to indicate the different areas that are 15 covered in the report.
16 It includes the facility and program description, 17 the section on scaling methodology describing the available 18 scaling methods, the methods used in each of the IST 19 facilities, brief discussions on what I think some of the 20 atypicalities and limitations of each of the facilities area, 21 and a scaling evaluation.
22 What that really means is that I have gone and 23 -- where I could with a pencil and a piece of paper -- looked
() 24 at areas that may or may not be distorted just because of the C 15 facility design and/or operating conditions -- things like
168 1 fluorogine flooding, et cetera.
( 2 There is a section on the relation of facility 3 results, some of the implications of operation at these 4 nontypical pressures, potential methodologies for comparison 5 on a global basis and also on local bases, and discussions 6 on potential counterpart experiments. And, finally, some 7 conclusions. l l
8 (Slide.)
9 ' MR . LARSON: The report status, as I understand 10 it, each of the ACRS members has a draft copy of the report 11 for their review. It is currently undergoing what I call 12 inouse review also -- inhouse being peer review and our
{O - 13 typical editing people are working with the report. Copies 14 have been sent to EPRI and also to NRC for their review.
15 The schedule at the current point in time is, 16 I plan to incorporate comments next month and, depending on 17 the extent of the comments, if they are not too extensive then 18 we will try to publish this report in March -- probably late 19 March.
20 MR. BECHNER: Tom, would you like to mention 21 the date that you would like to have comments by?
22 MR. LARSON: The desired date for comments in 23 order to meet this schedule is February 10 The comments can
() 24 be sent back to me or to Richard Lee at NRC.
c' (Slide.)
25
169 l
1 MR. LARSON: The next slide in your handout is
( 2 simply a facility description slide. And I think everyone 3 here is probably quite familiar with the three facilities after having heard discussions on them yesterday. l 4 So I won't cover l 5 that.
6 What I would like to do now is move right into 7 the results and briefly go through some of the things that I g have looked out in the context of putting this report together, 9 (Slide.)
10 MR. LARSON: First I would like to briefly 11 address the facility scaling methods. Again, I think just 12 about everyone in this room has been exposed to the scaling 13 on each of the facilities. There is probably not reason to Cs- )
14 spend much time there. But it is good reference material. ,
15 Discuss a little bit the limitations and 16 atypicalities of the facilities. And I will give you an 17 indication of the kinds or the way in which material is 18 Presented in the report so that one at a quick glance can get 19 an idea of what the geometric limitations and atypicalities l
20 of the facilities are.
21 Briefly describe the scaling analysis and 22 evaluation that has been done to date, and conclusions that 23 I have reached from looking at those areas, discuss some of
() 24 the problem areas and then go into the data comparison methods 25 and the counterpart experiments.
=
170 1 (Slide-)
( 2 MR. LARSON: Briefly then, in terms of scaling 3 we have heard yesterday that MIST is a modified volume-scale 4 facility, tall, skinny, looks a lot like the facilities in the 5 past such as the various versions of the semi-scale facility, 6 operates at full pressure.
7 The volume scale factor is 1/817. In the ideal 8 case that is also the area ratio, area scale and the N-scale, 9 if you will, the number of rods, tubes, et cetera.
10 Both Maryland and SRI on the other hand are 11 reduced-pipe facilities designed according to what we have 12 come to call the Ishii scaling criteria. This scaling criteria.
Q is a more general statement of scaling and similitude than
[ ys 13 14 is the modified volume scaling critera. And in fact, volume 15 scaling is a subset of the Ishii criteria.
16 So the. fact that these_ facilities are reduced 17 in height is really not a new concept. It is just that Ishii 18 put it all together and indicated that if one desired to build 19 a reduced-height facility, then he had the leeway to pick a 20 different scale for the area, whereas we do not have that 21 degree of freedom in volume scaling.
22 Hence, the Maryland facility is a little over 23 one-fourth height scale, SRI is one-fourth height scale.
() 24 Maryland has a much larger area scale, 1/110 relative to the e' 25 SRI facility. Either of these two facilities operate at full
171 1 pressure.
--. f~% -
F 2 (Slide.) !
3 MR. LARSON: The next slide just gives a 4 comparison of some of the geometric scales and the full power 5 capabilities in these three facilities. As I said, MIST is 6 full height volume scale. This also gives you an idea of what 7 the volume relationships between the three facilities are, 8 Maryland and MIST being about the same volume.
9 In area scales and total core power capabilities, l l
10 note that none of the three facilities operate at full power.
11 They all operate at some percent of full power, MIST being 12 about 10 percent power. And Maryland and SRI, as we will see on the next slide, are -- depending on whether you are talking l (() ) 13 14 about the two-phase scale or the single-phase scale --
15 somewhere between 5 and roughly 15 percent scale power.
i 16 The pressure capabilities, MIST, full pressure.
17 Maryland, about 20 megapasicals, 300 PSI. SRI, about .7 18 megapasicals, roughly 100 PSI.
j 19 CliAIRMAN WARD: Tom, those volume numbers don't
-20 seem to square with each other. If you look at the volume 21 ratios and the actual volumes you have got shown there, 22 compared to the plant. I am just not understanding something, 23 I guess.
) 24 MR. LARSON: They are all relative to the
('L 25 plant.
.m.- _- , . _,,
172 1 CHAIRMAN WARD: Yes, but for example, you have f-('"' 2 got the volume of MIST and Maryland about the same, .57 and 3 0.6.
4 MR. LARSON: This number is calculated based on 5 the area scale and the length scale. The actual volume I 6 think is a bit distorted.
7 CHAIRMAN WARD: All right, so it is imperfect l 8 then.
9 MR. LARSON: It is not too far off. This 10 number, based on the actual volume of the facility, should 11 be I think something on the order of 1/1500, maybe.
12 CHAIRMAN WARD: Yes.
A '
k;J
~
13 MR. LARSON: That Js the confusion factor.
14 CHAIRMAN WARD: All right.
15 ,
MR. TIEN: In the MIST you show the volume ratio 16 to the nuclear plant at about I would say 550. And the 17 volume scale is 1 to 817. Your length scale is 1 to 1, your 18 area scale is 817 to 1. But the volume ratio doesn't come out 19 the same.
20 MR. LARSON: Well, again, there is a volume l
21 distortion in MIST. It is considerably larger than it should 22 be. I should not say considerably. But it is on the order 23 of, I think, 30-50 percent larger than ideally scaled. This 24 is an ideal number. Likewise, this.
C' 25 MR. MICHELSON: Let me ask you. You scaled the
173 1 length according to the plant. What do you do about scaling
( 2 the relative elevations of certain critical points, like high 3 points or low points or whatever? j l
4 The SRI-2 is not scaled because the pressurizer ]
5 is the highest point in the system. In the real plant, it 6 is nowhere near the highest point in the system. Yet you 7 claim that you scaled the length at one-fourth. So there g must be some shifting around of relative elevations that are 9 no longer similar to the full-scale plan.
10 MR. LARSON: The intent is to maintain that 11 ratio of all of the elevations. Now, there are distortions 12 of ---
O A/ m 13 MR. MICHELSON: But you didn't on SRI-2.
.k 14 MR. LARSON: There are distortions in some areas, 15 I thought SRI-2 was quite close.
16 MR. MICHELSON: I asked the question yesterday 17 because the drawing shows the pressurizer at the high point of 18 the system. And they verified that it is the high point in 19 the system.
20 MR. LARSON: That is Maryla,nd.
21 MR. MICHELSON: Maybe I am incorrect. It might 22 be Maryland -- it must be Maryland then. I will have to look 23 at the drawing. But at any rate ---
O(_j 24 MR. LARSON: But the same question?
25 MR. MICHELSON: But it apparently was the high
174 1 point of the system, whereas in the plant the pressurizer is
( 2 not the high point.
3 MR. LARSON: Carl, that report has several 4 tables that compare what I call actual over ideal parameters.
5 One is a detailed examination of the elevations from high point.
6 and low point of every system ---
7 MR. MICHELSON: But they have the low point g right, as I recall, but not the high point. They brought it 9 down. All right, that is the Maryland, yes. So it's a little 10 deceptive when you claim you've got areas to length scaling 11 when in the real world it can't be.
12 MR. LARSON: There are distortions in certain rs
,(,/ .
13 aspects of each of these facilities. And that is one of the 14 reasons for this report is that there are those distortions 15 and are they really of a consequence. You can't answer a lot i
16 of those questions right now. ~You can simply go through and 17 look at the geometry and say ,well, there is a potential here 18 for a distortion because this is the high point in this 19 facility and in the plant it is really something else.
20 But what can I say about that at this point?
21 Nothing, other than to point out that that is a distortion.
22 That is something that we need to keep in mind when we are 23 analyzing the data and running the experiments.
() 24 MR. MICHELSON: Maybe there are some other
@ 25 distortions.
- . - , - - - - - - - - - - - -- -Q n-- ,- - - - - - - . . - - - - - - , - , -. , - . - - - , - , - - . -
175 1 MR. LARSON: Sure.
( 2 MR. MICHELSON: That I think might be critical.
3 MR. ETHERINGTON: Is this just reporting the 4 ratios that exist or were the ratios established on the basis 5 of the findings of the authors of this paper? Do you follow 6 me?
7 MR. LARSON: These ratios (indicating)?
8 MR. ETHERINGTON: Yes.
9 MR. LARSON: These are by and large the ideal 10 ratios.
11 MR. ETHERINGTON: I see.
12 MR. LARSON: The intent was to scale to these
()
r _/ 13 numbers. Now, each facility has certain distorti.ons because 14 you perhaps could not design a component ---
15 MR. ETHERINGTON: But did you have input to 16 their selection, is really all I am asking?
17 MR. LARSON: No.
18 MR. ETHERINGTON: I see.
19 MR. LARSON: Well, not really. The basis for 20 some of these numbers, the 1 over 817 on MIST, for example, 21 is that B&W has over the years built and operated 19 tube 22 steam generators for separate effects testing.
23 They wanted to use those generators on the MIST
(~)s
(_ 24 facility. So that 19 tube steam generator then sets the
(' 25 volume scale if you want to use that steam generator on a new
176 1 facility.
( 2 MR. TIEN: Could I ask just about the height 3 scaling distortions, say the University of Maryland facility.
4 When they built the facility, did they conceive those 5 distortions? And then what are the consequences? Or what?
6 Are you going to also discuss what are the 7 consequences of these scaling distortions?
3 MR. LARSON: To a certain extent there is a 9 discussion of the consequences of some of the scaling 10 distortions, yes.
11 But to answer a particular question like Carl 12 brought up at this time, all a person can do is speculate and rI)
\c 13 do some calculatican. And, really, to answer that question 14 you have to run some experiments and ask yourself if you are 15 getting results in those experiments that look like something 16 maybe you did not expect.
17 And then you have to ask the question, hey, 18 is that because of this scaling distortion or not. Now, the 19 question as asked a couple of days ago, can a person take some 20 of these tables that are in 'tdua report and go down through 21 them and try to rate the overall impact of some of the scaling 22 distortions?
23 And I think the answer is yes. But by and large
/~T
(_/ 24 there is a lot of subjectivity in doing that. Some things 25 you can attack with a pencil and a piece of paper and come to
l 177 1 conclusions. But by and large a lot of that is subjective, b.- 2 really.
3 The intent of this is to point the finger at 4 areas that may exhibit distortions, and then caution one to 5 watch out for that sort of thing when you are running the 6 experiments, or analyzing the data.
7 MR. SCHROUK: Tom, is there a rationale that 8 created the given distortion, that is the question? -
9 MR. LARSON: Is it distorted for a reason?
10 MR. SCHROUK: Yes.
11 MR. EBERSOLE: Let me add to that. These three 12 experiments, were they deliberately designed with these O 13 aberrations to fit into voids of experimenta' tion not 14 performable by the other two?
15 In short, was this a preplanned set with the 16 configurations deliberately established to interlace with 17 the others?
18 MR. LARSON: Not ---
19 MR. EBERSOLE: And following on then, what 20 general region of experimentation is each intended to do which 21 the others can't?
22 MR. LARSON: The first question is, why was 23 this length scale picked for Maryland, or why was this one 24 picked for SRI?
C' 25 MR. EBERSOLE: Yes.
178 1 MR. LARSON: If you look in detail at Ishii's
?-
( -)2 2 scaling criteria, or if you look in detail at the B&W plant, 3 what you find is that the hot leg is kind of what we call a 4 limiting component. It is the pressure drop in that long 5 vertical hot leg that seems to be a deterministic feature in 6 your facility design.
7 Now, in MIST, to modify volume scale facility.
8 the hot legs are oversized because of frictional pressure-drop 9 considerations. If you want it to be full height, it has got 10 to be oversized or else there is a factor of 40 too much 11 friction pressure drop.
12 If you analyze the hot leg frictional pressure O
(X_/ 13 drops, what you will find is that there is a solution between n-14 area scale -- there is a relationship, I should say, between 15 area scale and the length scale that gives you about the 16 right pressure drop, the FL/D losses.
17 SRI is fairly close to that proper combination.
18 In other words, if you will plot the area scale versus length 19 scale, there is a solution that says, all right, for this 20 length scale and this area scale I've got about the right 21 friction pressure drop including K factors and FL/D.
22 Now, Maryland is a bit over-sized. But you 23 could put orifices in there. And so this one-fourth number r
() 24 and this 1/324 number have a basis in consideration for 25 frictional pressure drop characteristics in the hot leg.
179 1 Because it is a limiting component, 50 feet high.
( 2 MR. CATTON: But it is still a great deal of 3 compromise. If you were doing ideal scaling, as you indicate 4 ideal scaling, the SRI-2 would be 1/64 on the volume scale 5 because if you are going to do scaling properly you maintain 6 geometric similitude. And that was just thrown out.
7 The reason is the power. They can't get -- if 8 they take it as 1/64, then the required core power is too high 9 So there's a lot of compromise in here.
10 What they're doing is, they are making it 11 shorter and fatter, which is good. But it still ain't short 12 and fat enough to give you ideal similitude.
n
? )
(^
w
' 13 Now, what I would like to hear about are, what 14 -- this is much better than we have had in the past, as you 15 look across the column of three facilities. There are still 16 things that have been done to the upper plenum, the lower 17 plenum, the downcomers -- all these things that according to 18 some may have a big impact on the small break or some of the 19 transients were you are slowly draining down your system 20 because of levels and other things.
21 I think this is an area that needs some focus 22 in your report. And I don't see it there.
13 MR. LARSON: I don't quite understand your )
(^3 )
\ 24 point? I mean, the sign of the internal ---
J
( 25 l
MR. CATTON: Well, this thing is tall and l
l
180 1 skinny and even the short and fattest one is still tall and
( 2 skinny. There is some impact -- or there may be some impact 3 of that.
4 Because if you recollect what Ishii did. Ishii 5 said, gee, what I want to do is I want to find out what the 6 diameter has to be just to get-away from certain kinds of 7 phenomena that may be created by being too skinny.
8 lie didn' t ask about other kinds of things that 9 could occur because it is still tall and skinny. For example, 1
10 the bubbly flow that we see from the SAI tests.
11 Ilow are these things going to impact on the 12 data that is obtained relative to a plant? The distortion of llh 13 the upper internals and the distortion of the lower plenum --
14 particularly the upper plenum, what is that going to do if
, 15 you have stratified flow in that horizontal one in the hot l
16 1097 17 If it is shorter, it is too fat, or all of these 18 things, how are those going to play a role? The thermo-19 dynamicss part, which is I think what your report is focused 20 on, is great. But there is this other aapoet.
21 MR. TIEN: I think the answer also to the 22 carlier points there is a tendency always to say, well, maybe 23 we can -- you know, we have to wait for the test until we know
() 24 what is going on.
C: 25 It is a scaling distortion. I don't completely
181 1 disagree with that. But I think a lot of thinking should be
( 2 being done before thinking about what are the possible impacts.
3 Because this will have some bearings on what kind of test you 4 like to make. j 5 Instead of just going to go make a test and 6 then analyze what are the scaling distortions, I think there 7 is an iteration that you have to go before you do the test.
3 You have to really think about all of this scaling distortions ,
9 what are the possible impact and major impact, and what 10 phenomena. And there you do certain kind of test.
11 MR. IARSON: I don't think there is anybody here 12 that would disagree with either of those two comments. The
- 13 only problem is, there is nearly an infinite number of things 14 that can be looked at beforehand, so I'd appreciate any feed-15 back on what this group thinks the most important things are.
16 certainly, the report can be added to and 17 include those kinds of things. And getting back to things 18 like the L/D ef fect or the L/D distortion on, for example, 19 development stratified flow in this horizontal section and 20 the effect of distortions in the upper plenum internals on 21 the development of that stratified flow, hot leg, those kinds 22 of things.
23 MR. CATTON: I understand you can't do anything 24 about them. But on the other hand, if the codes cannot handle 25 them at this point, I am not sure what direction you take or
182 1 what good the data does.
(2 2 Because you are going to have to analyze these 3 experiments. If you broaght in a different phenomena as a 4 result of the scaling distortions, then they damn well better 1 5 think about being able to model it in their codes. Or they i
6 are never going to bring the package together, j
(
7 MR. LARSON: Right.
8 MR. CATTON: That is why it needs some discussiory 9 -- not that the discussion will lead to a change. We 10 obviously can't. These things exist. Your analysis should t 11 have been done before the systens were designed.
12 But that is a day late and a dollar short.
O N_/ 13 MR. BECKNER: Ivan, when you comment, can you 14 give a list of specifics? Because there are a number of 15 things that we have addressed that were brought up as concerns 16 carlier and we have got separate effects, or at least if we 17 haven't addressed them, we have acknowledged that they can't 18 be done.
19 So when you make comments, could you list some 20 specific phenomena?
21 MR. CATTON: Well, I did just now. The 22 stratified flow in the upper plenum. It may not be an impact, 13 but it would be nice to have that included as part of the
( 24 analysis that Tom is putting together.
~'
25 MR. BUCKNER: All right, I am not sure I know i .
183 1 what that means. So we need to understand that.
( 2 MR. CATTON: I will try to amplify on that when 3 I write up our report on the meeting.
4 MR. EBERSOLE: Let me ask a question. Isn't it 5 true that at these shut-down power levels that the heat flux 6 is so low that if you've got.almost any perceptible level of 7 true quality coolant above the clotting, you've got it made?
8 MR. LARSON: If I've got any perceptible what?
9 MR. EBERSOLE: Two-phase fluid above the core.
10 MR. LARSON: The core is covered?
11 MR. EBERSOLE: The core is covered and you've 12 got margins of safety running out your ears, isn't that right?
13 MR. LARSON: You should not get core heating.
{
14 MR. EDERSOLE: Lut that is the bottom line?
15 MR. LARSON: Yes.
16 MR. EDERSOLE: And all the peripheral informa-17 tion you have to support that conclusion, you can kind of 18 massago around. As long as you have that ultimate conclusion, 19 isn't that right?
20 ,
MR. LhRSON: As long as the core's covered, 21 there is ---
22 MR. EDERSOLE: I caid covered in a controlled 23 context. I didn't know about what quality, but I'm saying j
() 24 with even low quality fluid you've got it made.
MR. LARSON:
25 liigh quality, yes.
1
184 l l
1 MR. EDERSOLE: No, I am talking about lots of 2 vapcr.
3 MR. LARSON: That's high quality.
4 MR. EBERSOLI: Right, you're right. Sure, high 5 quality.
6 MR. LARSON: Yes, I think that is true. Any 7 time the collapsed level in the core is ---
g MR. EBERSOLE: Well, that is what we arn all 9 after. You know, the golden nugget we are after is -- and the 10 rest of this stuff is peripheral to that.
11 MR. LARSON: This is all the issues that the IST 12 program is to addresn. Core coolability I cuppose is kind of lh 13 the underlying theme of the whole ---
14 MR. EBERSOLE: Sure.
15 MR. LARSON: --- program. But there are other 16 effects that are of interest like natural circulation, stall,
,17 when do you got to BCM ---
18 MR. EBERSOLE: Well, yes.
19 MR. LARSON: --- a phenomena in the system may 20 lead to operator confusion, may lead the operator to do 21 somothing that ho shouldn't to jeopardize ---
22 MR. EBERSOLE: Ye s .
23 MR. LARSON: --- the core cooling.
() 24 MR. EBERSOLE: It is to convince the operator 25 that he is all right, that it's okay.
185 ,
1 MR. MICHELSON: One is very much interested in
( 2 the pressure history that one can anticipate in the event, t
1 3 Because many systems don't even have high-pressure makeup. So 4 although the core is covered, the pressure history may be 5 adverse such that you will never -- you will eventually dry l
6 out at very high pressure. And you don't want to do that.
7 MR. LARSON: You can't get water in ---
8 MR. MICHELSON: You can't get water in, so it is 9 very important to be abic to calculate and predict this 10 pressure history. And isn't that what this is all about, 11 making sure we understand the pressure history we know the 12 inventory makeup rates that we can expect?
<~y (I
J 13 Bdcause although there is a point in time when l
r 14 the core is cooled, fine. The pressure may be so high that
!$ you aren't replacing inventory. So before long, you just sit
)
i 16 there at high pressure and dry out.
I 17 So all of this Mickey Mouse is juct to try to 18 make sure ve understand the prescure history of the event, 19 isn't it?
20 If you are sure to got the very low pressure, 21 you would have no problem. Any power company could put the 22 water in.
23 MM. LBERSOLE: The ultimate solution is blow-dowr .
O (J 24 MR . LARSON : The overall theme of the IST
( 25 program iu to provide the data that will give un the data bane
186 1 :o make sure that the codes are capable of calculating
()b '
2 responsibly the pressure history and whatever other history.
3 MR. MICHELSON: But the inventory replenishment '
4 will be correct so it will keep the core cool.
5 MR EBERSOLE: Yos.
6 MR. MICHELSON: But that is why you worry about 7 natural circulation because it affects the pressure because it 8 affects the rate at which heat is removed. If you have no 9 circulation, obviously the core is going to heat up very 10 quickly, even if it's covered.
11 You have 2,500 pounds pressure with a full core, 12 MR. EBERSOLE: The core level devices in the 13 vessel would forecast if you were gaining or losing.
{
14 MR. MICHELSON: With 2,500 pounds, you don't 15 r eally care whether it is gaining or losing . There is no way 16 to make it up.
17 MR. EBERSOLC: You would bc losing and you would 18 see it.
19 CHAIRMAN WARD:I think what Tom is saying, the 20 strategy is to use the codes for that -- for those evaluations, 21 And the experimental program is only to build the codes into 22 credible tools.
23 MR. MICHELSON: Right, to make sure they are o
U 24 right.
(~.
25 MR. SCHROCK: There is another point I want to
187 I
i 1 comment on. And that is, the inventory distribution.
r-) - .
(%) 2 llave you given consideration to that, yet, as to 3 how these different f acilities compare with the prototype?
4 MR. LARSON: Where is the water?
5 MR. SCHROCK: Where is the inventory in the 6 system as a function of time?
7 MR. LARSON: As a function of time?
8 MR. SCHROCK: Yes. The scaling dietortions '
9 probably produce dif ferences among these various facilities.
10 And they are each undoubtedly different from what will occur 11 in a full-scale plant.
12 How do we know how to gauge the importance of
) 13 that -- of those differences, if we did know what they are? '
14 I mean, eventually measurements, calculations of these 15 f acilities will tell us what the distributions are. But ---
16 MR. LARSON: Eventually, I am sure. But at 17 this point --- ,
18 MR. SCilROCK: llow does that compare with the 19 plant would that impact on phenomena that are occurring?
20 MR. LARSON: Well, at this point in time, if 21 one just looks at the geometry of each facility, sure there 22 are volume distortions.
23 MR. SCl! ROCK: It is not just a question of the
() 24 relative volumes of different components, but how the water
(' 25 in redistributed during the transients, how the void ---
188-
} MR. LARSON: What if the void fraction is wrong 2 in this facility because ---
3 MR. SCHROCK: Because of it went into the 4 pressurizer and pushed water out of the pressurizer. That 5 would influence what stagnation state is driving the critical 6; discharge.
7 MR. LARSON: Truc. It would also influence what 8 the hydrostatic head is. There are some thoughts on that and 9 some concerns. But at this point, I have looked at the void 10 distribution in the hot leg and made some conclusions about 11 tho effects of low pressure on that. And we will see that in 12 a moment.
A
'(J 13 , But all I can do is point to that and say that 14 could be a problem. Right now, I can't project out without a 15 lot of computer code calculations and say that will be a 16 devestating effect on the scaling. I just can' t do that.
17 I can only speculate and say that it could be 18 a problem because there is a void distortion there. Whether 19 or not I will ever see it in the data, if it is significant, 20 I don't know yet. It is just something to point to. But 21 that is a good point.
22 MR. MICl!ELSON: Well, have you looked at the 23 volume of the pressurizer in ratio to the balance of loop (Q
_j 24 volume?
25 MR. LARSON: Each of the facilities scale e .
1R9 l
1 reasonably well.
(Ok 2 MR. MICHELSON: You are saying that in each 3 facility the pressurizer volume scale is correct?
4 MR. LARSON: Yes.
5 MR. MICHELSON: I thought there was some --
6 quite a difference'in pressurizers. Well, I may be mistaken.
7 CHAIRMAN WARD: It is scaled correctly as far 8 as it being a right fraction of total system volume, I~think 9 is what you are saying.
10 MR. MICHELSON: Yes. And apparently it is.
11 CHAIRMAN WARD: Although it is in the wrong 12 location, at least in the Maryland facility.
{g] 13 , MR. LARSON: In any of these facilities on a 14 component like that, it is very difficult to get the right 15 area and length and volume all at the same time and not 16 introduce some other complication like too much metal heat 17 transfer or too small of an area for phase separation.
18 So, generally, the conscious thought in 19 pressurizer design in these facilities was to get the right 20 water mass and vapor mass.
21 MR. MICHELSON: To what extent was there any 22 attempt to simulate the conductance of the difuser in the 23 pressurizer?
O
'i
' 24 MR. LARSON: Conductance in the ---
(.$ 25 MR. MICHELSON: Yeah.
l l
190 l l
, 1 MR. LARSON: In a thermal sense? Or the 6
4 2 mechanic sense?
3 MR. MICHELSON: In a hydraulic sense. Because 4 water is flowing in and out of the pressurizer during the 5 test.
6 CHAIRMAN WARD : Do you mean the flow resistance 7 of it?
8 MR. MICHELSON: Yes, the conductance of it.
9 MR. LARSON: Perhaps Randy or someone else can 10 answer that question. But to my knowledge, if the surge-line 11 resistance is to scale, I know on the MIST facility they are 12 going to some length to try to model the fact that there is 13 a difuser there.
C;'
14 MR. MICHELSON: I found the discussion of the 15 surge line, but I did not see whether that -- it wasn't clear 16 that that included the difuser conductance or not.
17 MR. CARTER: We went through a discussion a 18 little bit yesterday about the difuser design. And I do have 19 people checking on the screens.
20 MR. MICHELSON: But the question now is, to 21 what extent whatever you decided the difuser design is, how 22 is that modeled into these various test facilities?
23 MR. CARTER: In MIST the difuser design and its 24 corresponding scale resistance is included.
(";
25 MR. MICHELSON: And that is scaled to the real
191 1 plant?
( 2 MR. CARTER: Right.
3 MR. MICHELSON: Thank you. I don't know about 4 the other two facilities, if anybody wishes to comment whether 5 they have done any scaling of the difuser.
6 MR. HSU: We did not.
7 MR. MICHELSON: You did not?
8 MR. HSU: No.
9 MR. MICHELSON: It may not be important either.
10 I am not saying that. I am just asking 11 (Slide.)
12 MR. LARSON: That brings up a good point though, 13 Carl. Just pointing a finger at another example, something 14 that I had not even thought of and an infinite number of 15 items like that that one can worry about from the scaling 16 point -- and that is why I would appreciate what your thoughts 17 are on things that are lacking in that report. So that we can 18 go and fix it. Or at least attempt to fix it.
19 The next slide simply shows the comparison of 20 scales for velocity, time and power for each of the facilities .
21 The two different numbers for each of the Maryland and SRI 22 values are simply because of the fact that they operate at 23 reduced pressure relative to MIST.
24 Note that MIST velocity, time scale and power C' 25 scale are as listed. What we see here is that for Maryland
9 192 1 and SRI, the single-phase and two-phase scales for velocity,
<w I
3; / 2 time and power are all different. Again, that is simply a 3 consequence of the scale equations and the difference in the l
4 scale equations for single-phase and two-phase flow. l 5 And it is one of the complicating factors of 6 operating at reduced pressure. If one is desiring to perform 7 a simulation of what he might expect in either MIST or plant.
8 MR. MICHELSON: How is the by-pass conductance 9 between the hot leg and the core barrel handle, is that a 10 small hole somewhere in these facilities to simulate that?
11 MR. LARSON: The by-pass between?
12 MR. EICHELSON: The hot leg -- you know, the
{) 13 14 hot leg has to have a sleeve fitting in the core barrel and there is a pretty good size gap in the real plan. I am just 15 wondering how that gap, which is a continuous by-pass as 16 opposed to a vent valve, a flat opening and close, is that 17 continuous by-pass incorporated in these loops? Or is that 18 thought not to be important?
19 MR. HSU: When were were designing our core 20 reaction in the core, we did coordinate that with Babcock 21 Wilcox with the possibility of that bypass. But they already 22 had designed the present one, so we decided what we'd do is 23 we will run the test without bypass.
p
() 24 But the next time when we open the can we can 25 provide the bypass. So this is an actually case in point
193 l l
1 that we already had designed. But we also had a coordination
( 2 and we took that into consideration for future testing.
3 Also, we would provide a source, whether the 4 presence or absence of that bypass would mean anything or 5 not.
6 MR. MICHELSON: One of the key questions on 7 bypasses, of course, is how big is it. It is very difficult 8 to get into the pass from the vent because from one plant to 9 another there was quite a difference in the setup.
10 In some cases, the bypasses are. In some 11 cases, they're not.
4 12 MR. CARTER: Just a couple of comments. It A
might be easier if I could show you this figure, the
{j 13 14 situation.
15 The direct bypass from, say, the downcomer into 16 the hot leg amounts to about 2 percent.
17 MR. MICHELSON: That's the whole flow in the 18 real world.
19 MR. CARTER: And that is with the pumps in 20 operation.
21 MR. MICHELSON: That was kind of a magic 22 number. Now, the question is, in the real world after 23 fabricating d'.d you end up with 2 percent, at least in the 24 test number s or something?
($( ) 25 MR. CARTER: These are design bases.
194 1 MR. MICHELSON: That was the design basis for
( 2 the 2 percent bypass. But in the real world, how much has it 3 amounted to? I understood that it varies from time to time 4 and the gap is quite.large in some cases.
5 MR. CARTER: The other thing we did consider is 6 the flow would go.up through the core, through the flow holes 7 in the barrel and out the top hot leg. But first the flow 8 would go up to the guy rods and then back down into the hot 9 leg. That split's about 95 to 10.
10 The total is 5 percent, or the 2 percent that 11 we talked about some other associated with the vent valve. So 12 you have a split of 85/10. In MIST the way we have that model
)
13 is, there is a flow plate on top of the inner cylinder barrel 1,4 that allows the flow to come up and then back down to the hot 15 leg as well as the direct flow holes to the hot leg.
16 MR. MICHELSON: That is built in?
17 MR. CARTER: That is built in to MIST.
18 MR. MICHELSON: But you can't determine the 19 sensitivity, you cant vary that?
20 MR. CARTER: You can very it, but with some 21 difficulty. I mean, you've got to take the upper head off but 22 it can be done. That is one of the things that do exist in 23 the MIST facility.
24 MR. KIM: Looking at those two facilities with
(~( ) 25 the two different scalings compromises the one, the single-l
. \
195 1 phase and another for two-phase. By the same token, when you
([ ) 2 say MIST and another prototype -- if you look at those two 3 facilities, with two different numbers, one for single-phase 4 and another for two-phase.
5 By the same token, if you would like to incor-6 porate MIST to the prototype, wouldn't you have to have the 7 two numbers from MIST too?
8 MR. LARSON: No. MIST operates at full pressure ,
9 So we will see in a moment that the scale equations are 10 property groups on the single-phase and on the two-phase scale 11 equations. But if you are operating at full pressure, all of 12 those property groups are unity, they're both single-phase and 13 ' two-phase.
{ ll ,
14 If they are not operating at full pressure, 15 since the property groups on the single-phase scale equations 16 and on the two-phase scale equations are different, then you 17 get these two numbers -- two different numbers.
18 MR. KIM: One more question. And that is, how 19 do you compromise the operating of these two smaller facili-20 ties when you conform the single-phase or the two-phase mode, 21 it seems like there is some discontinuancy.
22 Do you have any recommendations?
23 MR. LARSON: We will see the discontinuity in a
(~N 24 minute. I think that Mr. Hsu has some data from his facility nG ks 25 where he can speak to the ultimate impact of that
.\
196 1 miscontinuity, particularly the power.
(3 h 2 That is the thing that I have been most con-3 cerned about, because there is a very pronounced discontinuity 4 in the power scale. But you can see it right here.
5 My thoughts were that, well, on a transient, if 6 it comes along and you are in single-phase operation, then 7 go to two-phase, you suddenly realize, hey, I've got to turn 8 the power down or I'm going to have trouble.
9 And Y. Y. has verified this. The pressure 10 starts to go back up if you don't go to a two-phase power 11 scale.
12 But my thought was, when you went to the two-13 phase power scale, you had introduced a considerable C( )
14 perturbation in the system which may cause core collapse and 15 other gyrations in the system. And Y. Y. has said that so 16 far he has not seen any bad aspects of that sudden power 17 change on his facility.
18 But I have a recommendation later on as to what 19 I would do about this discontinuous scale.
20 MR. EBERSOLE: May I ask, on the aspect of 21 power, if you had an asterisk next to power, what would you 22 define the power as, as you have it up there?
23 MR. LARSON: Core power.
24 MR. EBERSOLE: I know, but in what context?
J E,f) 25 Percent of what? 1/817 of what?
I
197 1 MR. LARSON: Of the plant power. On the slide n
k(_3/ 2 before that we saw MIST at about 330 kw. power.
3 MR. EBERSOLE: In the tripped condition?
4 MR. LARSON: Pardon me?
5 MR. EBERSOLE: In the tripped condition, X-6 minutes after ---
7 CHAIRMAN WARD: Five percent, Jesse.
8 MR. EBERSOLE: Wait a minute, I think it is 9 different. It is more than 5 percent, isn't it?
10 MR. LARSON: MIST is about 10 percent, 330 kw.
11 MR. EBERSOLE: What I am saying is, power is 12 different in each one of them, isn't it?
ll 13 MR. LARSON: Sure.
14 MR. EBERSOLE: So I would just --- .
15 MR. LARSON: Just to calibrate you ---
16 MR. EBERSOLE: Yes.
17 MR. LARSON: --- SRI, for example, that ratio 18 corresponds to roughly 5-6 percent power, scale full power, 19 in that facility.
20 MR. EBERSOLE: Yes.
l 21 MR. LARSON: This number corresponds to about l
l 22 17 percent of scale full power. Now, these numbers look like 23 they are backwards, but they are not. The reason being, if f') 24 I use the single-phase power scale and calculated power of 25 full power value for SRI single-phase, I would say, all
198 1 right, SRI needs about 1800 kw.
L_/ 2 MR. EBERSOLE: I guess I don't understand how l 3 you would juggle power according to the phase of the fluid 4 when in fact, if there is.anything that is going to be un-5 changeable in experimental conditions, it is the power level.
6 MR. LARSON: Power level, right.
7 MR. EBERSOLE: So here you are juggling power 8 to fit the cooling condition, and in real life you can never 9 do that.
10 MR. LARSON: We will see the scale equation in 11 a moment and I think maybe it will be a'little clearer. It
)
12 is a consequence of being at pressure that is not proto-ll 13 typical, reduced pressure.
- 14 You can't have the same delta T in these low-15 pressure facilities across the core and beta is different.
1 f 16 IIFG is different.
17 MR. EBERSOLE: That compensates for that by 18 adjusting power down as you go to two-phase, is that right?
19 MR. LARSON: Exactly. That is the consequence.
20 MR. EDERSOLE: So it is another bias on the 21 power just like the experiment is itself.
22 MR. LARSON: I would not call it a bias on the 23 power. It is just a scale equation, that is what it has to 24 be in two-phase and this is what it has to be in single-
])
25 phase. If I am going to be scaled and try to maintain all
u +
199
} of the similarity parameters. You would never get to all of
()lh 2 the similarity parameters, of course, but it is just what the 3 scale equations say.
4 Now, there is a discontinuity at the intersectior t 5 between single-phase and two-phase for water.
6 MR. EBERSOLE: I guess what bothers me is you 7 are diddling with power, when in real life you cannot.
g MR. LARSON: But here we are diddling with the 9 nower to simulate what it is in the plant. I mean, it is not 10 changing a whole lot from the plant, but here you have to 11 change it in ordeE to simulate that constant value.
12 (Slide.)
(()
s-13 MR. LARSON: In terms of limitations and 14 atypicalitics, Jesse shared one of the conerns I have had 15 all along on that. And that is what to do with that power.
16 And Y. Y. has some information that says that may not be a 17 problem.
]8 Pressure -- reduced pressure to me in one sense 19 ' of the word is a limitation and/or an atypicality in these 20 facilities. Now, I've given this presentation several times 21 in the past to people that misunderstood or misinterpreted 22 what I really want to say about the effects of pressure.
23 I know all of us here know full well you can
(") 24 never achieve 100 percent similitude in any of these facilities 25 relative to the plant. But one of the things that I have been
200 1 interested in looking at here are what are the things that 1 A
C1) 2 need to watch out for so that I can after the fact go do more 3 analysis -- or even before the fact, if I know exactly what I 4 am looking for -- do more analysis to make myself feel more 5 comfortable with reduced pressure at operation and simulation 6 back to some reference condition.
7 So, I am really not trying to get to a point 8 here why I am devising a transient in these reduced pressure 9 facilities that is 100 percent in terms of similitude relative 10 to the reference base.
11 I will say some more about power and pressure 12 scaling later. I just wanted to give you an example here of
{]) 13 something that Carl asked a' moment ago. Where can I go and 14 look for ratios of this actual two ideal components. And a 15 part of the report has got some tables that give the ratios ,
16 between ideal and actual components for things like 17 geometric parameters and hydraulic resistances, and a number 18 of other things. ,
19 (Slide.)
20 MR. LARSON: To answer your question, Carl, 21 about the pressurizer, you can come in here and look and see 22 that both MIST and SRI are very closely scaled according to j 23 ideal value for the pressurizer volumes. Maryland is about l t
() 24 15 percent high.
25 Now, I can come across this table for any of l
(
201 1 these parameters in here and look at that, and then ask (fw(_) 2 myself the question, well, gee, does that 15 percent distor- l l
3 tion, is that a bad thing or a good thing. l 4 And the answer is, well, at this point I can 5 only speculate. But it is something that I may want to look 6 at later on after we have data. What is the effect of that 7 over volume in the pressurizer.
g (Slide.)
9 MR. LARSON: The scaling analysis and evaluation 10 that I spoke of earlier, the things that I was able to look 11 at in some detail with pencil and paper are list here, and 12 include things like hot leg flow regime, void. fraction, flood-And also to a certain extent,
('(
u
) 13 ing, et cetera, et cetera.
14 the pressure and the power.
15 These first six items here, I just want to 16 summarize with my conclusions. Then we will spend a few 17 moments talking about the pressure and analysis of the power 18 distortions.
19 (Slide.)
i 20 CHAIRMAN WARD: Tom, before you get into that, 21 let me ask one more question about the equipment arrangement.
22 ANd it was pointed out that the pressurizer in Maryland was 23 not in the right place.
What about the raised loop-lower loop question?
) 24 l
l L('~ 25 llave you looked at that in all of these?
202 1 MR. LARSON: These are all scaled to the lower
( 2 plenum.
3 CHAIRMAN WARD: Isn't Bellafonte a raised loop 4 or is it a lower loop?
5 MR. MICHELSON: It is a raised loop.
6 CHAIRMAN WARD: So three of the nine -- I guess 7 there are nine plants in the country -- are raised loop, and
. 8 have you looked at problems there might be associated with 9 that?
10 MR. LARSON: The difference in elevation of 11 those, no.
12 CHAIRMAN WARD: I guess the question is, will 13 the codes be able to, you know, credibly predict the behavior
{)
14 in a raised loop plant when they are validated by lower loop 15 experimental data? Have you thought about that?
16 MR. LARSON: Gerda was a raised loop, so we do 17 have a certain, data base. ANd Otis was a raised loop also.
18 So ---
19 CHAIRMAN WARD: It seems that that is a question 20 that needs to be addressed, I think, in your overall 21 evaluation.
22 MR. LARSON: All right, point taken. Ultimately ,
23 when we have got data from MIST and there were comparisons g.-)
f~ 24 made back to Gerda and Otis, we will be in a position then to C~ 25 make a more definitive statement about are the responses of 9
r 203 1 those two facilities different. And is it because one is a
(_ 2 2 by 4 loop and one is a single loop, or is it because of the 3 elevation, for instance?
4 Conclusions from the scaling analysis and 5 evaluation. Ilot leg flow regime, as we saw yesterday at the 6 SAI test, according to calcuations using available flow regime 7 maps for both vertical and horizontal maps, the conclusion was 8 made that in all of the IST facilities we expect to see bubbly 9 flow in the vertical pipes and stratified flow in the 10 horizontal.
11 , Now, that is simply based on an energy balance 12 on the core assumption. And 90 percent of the steam generator lh 13 in the core goes through the vent yhlves. And simply plotting 14 a lot plot of JG/JF on the flow regime map, the point here 15 being that we do have three facilities, each- of them has a 16 different diameter in the hot legs.
17 We expected perhaps to see differences in flow 18 regime behavior there, which we saw yesterday. It can be an 19 important thing in terms of void distribution, what that 20 might do to the overall driving heads in the system.
21 But the conclusion reached here is that, for the 22 typical nomal type transients, we should see the same flow 23 regime in all three facilities and in fact similar flow I) 24 regimes is what we expect in the plant.
25 so just based on flow regime behavior, one may
i 204 1 not expect any significant distortions in facility behavior 1 (p 2 with respect to flow regime, i
3 MR. EBERSOLE: Let me ask you a question ---
4 MR. CATTON: Well, let me ask. You split the 5 flow between two steam generators and come to this conclusion.
6 Is that conclusion still the same if you have one steam 7 generator out, which is a reasonable ---
g MR. LARSON: You effectively then double the 9 superficial velocity so there is a point when you will cross 10 over into the slug flow regime at extremely -- not extremely 11 high, but at high JF.
12 Now, you can get to that same problem, Ivan, if for some reason the vent valves aren't passing as much flow (lh 13 14 as we think they are.
15 MR. CATTON: That's right.
16 MR. LARSON: Now, if 50 percent of the vent 17 valves are locked closed, then you reach some different 18 conclusions. You could get into the slug flow.
19 MR. CATTON: Or if it's 80 percent instead of 20 90, then you double -- they are doubling the superficial steam 21 velocity also.
22 MR. LARSON: I looked at 50 percent. And there 23 was a potential for some slug flow there, just barely.
MR. THEOFANUS: What was it?
) 24 C(' 25 MR. LARSON: Fifty percent of the steam
O 205 1 generated by the core goes to through the vent vales. And the (f) 2 superficial velocities in the hot leg are high enough to just l
1 3 get you a slight flow.
4 4 MR. THEOFANUS: Isn't that done if you are 80 5 percent? If you are getting 50 percent, you get ---
6 MR. LARSON: When I say 90 percent, I mean most 7 of the flow goes through the vent valves. Most of the steam 3 flow goes through the vent vales so you have lower JG/JFs in 9 the hot leg, then you do at 50 percent of the flow going 10 through the vent valve, the steam flow.
11 MR. THEOFANUS: So that covers the 80 or 70.
12 MR. CATTON: I said 80 because that should 13 double the steam flow. I am just remembering the diagram wo
{ {} ,
14 saw from SAI. If I move a factor of two over to the right, 15 I'm into the slug flow if the JF is low enough. In the lower 16 region.
17 MR. CATTON: Just another comment. I know on 18 the curse they presented yesterday they assmed a 4 percent 19 scale power. I am not sure in your calculations what you 20 assume. You pass 4 percent in roughly one minute after your 21 steam. So you are in a fairly rapid decrease in power in 22 the early part.
23 MR. LARSON: My calculations were all done at 24 about 5 percent power. And that is a factor that is too
[}
b 15 conservative.
l l
l 206 I
1 MR. EBERSOLE: That was my question. I wish I
(;,) 2 you would qualify flow in the context of what conditions of 3 whatever flow you are using obtained. Is it as you said 4 4 percent or 5 percent? Will it break at a certain place or in 5 another place or no break at all? And is it a variable l
6 rather than a constance as you imply, and will eventually get 7 so slow that new problems arise?
g MR. LARSON: Eventually, as the decay goes 9 down.
10 MR. EBERSOLE: Will it eventually result, for 11 instance, in gas blocking at the top of the U-bend as the 12 velocity decays and approaches the bubble rise velocity?
h 13 MR. LARSON: That is a possibility. We have 14 seen that.
15 MR. EBERSOLE: Your troubles occur late in 16 time rather than early.
17 MR. LARSON: That is as the loop starts.
18 MR. EBERSOLE: When you say flow, you are 19 leaving me floating there.
20 MR. LARSON: When I say flow here, I am talking 21 about a certain combination of power and the resultant 22 superficial velocity.
23 MR. EBERSOLE: We know that, but I did not know
(~s: 24 what it was.
s (L 25 MR. THEOFANUS: That was the source of the full I
i
i 207 1 pressure, right?
l b(L_/ 2 MR. LARSON: The reference pressure I am using l
3 here is 1000 PSI in the plant and 7.7 -- 100 PSI and 300 PSI, l 4 respectively, 5 MR. EBERSOLE: Is there a LOCA someplace which 6 is changing the velocity?
7 MR. LARSON: THere are just steady state.
8 calculations..
9 MR. EBERSOLE: With no break?
10 MR. LARSON: Right.
11 MR. EBERSOLE: Just post-trip cooling?
12 MR. LARSON: Right.
[ ) 13 MR. THEOFANUS: So if you were to look at a 14 small break as it depressurizes, at some point this conclusion 15 will still be valid? If you look at 500 PSI, for example?
16 MR. LARSON: You can do the same kings of 17 calculations at different pressures,and reach the same 18 conclusions.
19 MR. THEOFANUS: Right. But would you reach 20 the same conclusion? ..
21 MR. LARSON: Yes. But certainly there is a .
22 breakpoint when ---
23 21R. THEOFANUS: It seems to me at some point 24 you want to reach the same conclusion, some pressure would 15 be in the breakdown. The similarities between ---
t
e 20n 1 MR. LARSON: In terms of flow regimes?
t V> 2 MR. THEOFANUS: Yes, especially --- !
3 MR. MICHELSON: The void fraction is a function l
4 of pressure.
5 MR. CATTON: I would think you'd take.a look at 6 a particular transient and take a look at it.
7 MR. EBERSOLE: I thought this was just a 8 snapshot somewhere.
9 MR. LARSON: This is just a snapshot.
10 MR. EBERSOLE: We need the range, don't we?
11 MR. THEOFANUS: Especially in small breaks. If 12 you want to make sure in a small break. You see, that is
{ ]) 13 where the most likely situation that you worry about, the 14 stagnating. And if you really want to know when you are going 15 to stagnate and how, you have to follow this kind of approach.
16 But, remember, it is not snapshots. It is 17 lower pressures and different powers. Especially the lower 18 pressures. Because the flow regime is going to be the thing 19 that is going to tell you whether you have good similarity 20 with respect to the level swell as we had yesterday, to flow 21 over.
22 MR. LARSON: If you look at the Tatil-Doktor I 23 map though, between the bubbly flow and the slow flow for 1 c} 24 25 these diameters there is not a very significant pressure effect.
209 1 If you look at the boundary at 100 PSI and 7
(i ) 2 1000 PSI, 300 PSI and the range between, the boundary is very 3 insensitive to the pressure. So I don't expect the boundary 4 to change much. The diameters are constant, so the only thing 5 that is going to change is the LOCI of JG/JF depending on the 6 power you show them.
I 7 MR. THEOFANUS: Well, maybe I am not explaining 8 myself correctly. Really, the ultimate indication to assure i 1
9 ourselves that what regimes we have in the facility will l
10 actually be the same with the flow regime in the reactor ---
11 MR. LARSON: Right.
12 MR. THEOFANUS: Therefore, as we heard yesterday
{ ll 13 there is a difference between those two things just because 14 of the size of the pipe.
15 MR. LARSON: Right. That is what I am saying 16 here, Theo. The regime boundary is virtually a constant --
17 virtually the same for these three conditions or four 18 conditions, MIST, plant, SRI and Maryland.
19 For conservative power conditions and a 20 reasonable -- at least what we think is a reasonable assumption 21 of how much the vapor generated in the core goes through the 22 vent valve, then we predict bubbly flow in the vertical and 23 stratified flow in the horizontal.
24 Now, if I reduced my velocity -- superficial G{:} 25 velocities in the hot leg even more, that does not put me in
210 1 a different flow regime. I am still in a bubbly flow.
()lh 2 MR. THEOFANUS: If you use the velocity, yes.
3 You still will be in the bubbly flow. But how about if you 4 increase the velocity?
5 MR. LARSON: I can start it with conservatively i 6 high values of the superficial velocity. Even then I am in 7 bubbly flow. Now, given ---
8 MR. THEOPANUS: Was it in high pressure, though?
9 If you increase the' velocity ---
10 MR. LARSON: That is at a high pressure, 100 11 PSI. That is at 50 PSI.
12 MR. THEOFANUS: What is superficial velocity?
i a
') 13 Was it higher than the low pressure? I think we-a,re missing 14 something here. If we are at the lower velocity for the 15 same power, you are going to get much higher superficial gas 16 velocity. You have to.
17 And it seems to me that you would be approaching 18 a sliding regime at some point, if you have to.
19 MR. CATTON: But you are scaling the power 20 down for the other facilities at lower pressure, aren't you?
21 MR. LARSON: Yes.
22 MR. CATTON: So if it is in a given facility 23 and it reduces the pressure, he gets into that problem.
1 24 MR. THEOPANUS: But the power is scaled only 25 with respect to physical properties, as we have heard before. I i
211 1 And also with respect to -- is that also taken into account,
( ,
)' 2 the density?
3 MR. LARSON: Sure.
4 MR. Ti!EOFANUS: The gas density, the power 5 dropping? DUt how come then you have a fixed power reduction?
6 It seems to me that that would become a function of pressure.
7 In a small break situation, your power should keep on changing 8 and you give me only one reduction.
9 MR. LARSON: I am looking at a conservatively 10 high value of power.
11 MR. THEOFANUS: Well, I don't understand how you 12 can do that. You have such a wide range of pressures you have ll 13 t o cover in a small break situation. How can you do it with 14 a fixed power ratio?
15 MR. LARSON: I said it was conservative, to 16 calculate one locus. I mean, you can calculate any number of 17 loci Js for power-pressure combinations. I think that is 18 Wat you are suggesting.
19 MR. THEOFANUS: I guess what you are saying is 20 that you were able to pick a power ratio no matter what your 21 regime is in. Is that what you are saying? Because it is 22 too conversttive, the power.
23 MR. LARSON: Yes. It is still in bubbly flow
(')
24 for the given assumptions. j 25 MR. THEOFANUS: And always will be a bubbly flow, l
1 j
212 ,
1 for the given assumptions.
n C7! ,/ 2 MR. THEOPANUS: And always will be a bubbly flow 3 even at core pressure, that is what you are claiming in all 4 the experiments as well as in the reactor?
5 MR. LARSON: Yes.
6 MR. MICHELSON: I guess I must be missing a, 7 point then. Because ultimately youir objective always is to 8 transport the heat from the core to the steam generator.
9 That is, even though 90 percent of the flow is 10 bypassed, the only flow that takes heat from the system is to 11 flow off the steam generator. And that, for a given energy 12 level, as you drop the pressure the velocity must increase.
r'(
v )
13 You have to, or you won't transport the heat out there.
14 So it looks to me as if as you drop the pressure 15 these things get higher and higher velocity and the flow 16 regime is going to change. But I don't know. I just don't 17 understand it. But that's all right.
18 MR. LARSON: I understand your concerns and we 19 will do something in the report to make it more clear. The 20 transient situations rather than the snapshots.
21 MR. SCHROCK: I have one last question which is 22 related to that. You have those different power scaling 23 f actors for single-phase and two-phase . And then that would
(~T 24 seem to imply, from your point of view, that the way to
(. 25 operate this experiment would be to start the transient with
213 ,
j l
1 one power and as you go into this thing, from the single phase l (f)) 2 into the two phase, you can suddenly change the power to a 3 different basis. But that is not the plan for the operation 4 of the experiment, is it? or is it? )
5 MR. LARSON: My recommendation is to not do that ,
6 But I think what Y. Y. has tried so far is to change the 7 Power when you get the two phase.
8 !
Is that not correct?
9 MR. HSU: Yes.
10 MR. Sct1 ROCK: It is the plan. By a factor of 11 more than 3.
12 MR. LARSON: At even higher than that. I have a' 13 power tracer here later that you will see that.
b(~T 14 MR. THEOFANUS: One change of power of a 15 continuous change in power?
16 MR. LARSON: Both.
17 MR. TIIEOFANUS: Is there going to be one power 18 change or continuous?
19 MR. LARSON: Continuous.
20 MR. IISU : When we do these phase, we step down 21 once. We do not preclude any -- or exclude any possibility 22 in the BCM when we learn how to operate, whether we need to 23 continue we don't know yet.
24 The thing about our operation is, we allow a t 25 certain time to learn the machine, learn how to do it.
214 1 MR. THEOFANUS: It seems to me that if you were
/ ~.
(h. 2 going to depressurize, if you want to maintain similarity with 3 the full-scale, it sounds to me like you almost have to 4 continually vary your power.
5 MR. LARSON: It depends on the pressure 6 transient that results.
7 MR. THEOFANUS: Of course. But I can see that 8 there will be some pressure transients that you would have to 9 do that. If you are to maintain a similar flow regimes.
10 MR. LARSON: Sure.
11 (Slide.)
12 MR. LARSON: Virgil, this is the power traces.
() 13 There are three different concerns here, or three different 14 situations -- not concerns. But I have speculated a pressure 15 transient in the model facility that will reproduce P/PO 16 ratio equal to 1.
17 In other words, I am forcing the pressure 18 transient on the model facility that maintains the same 19 normalized pressure in its own time scale as existed in some 20 reference point which was the MIST, 310,000.
21 MR. EBERSOLE: That pressure transient can come 22 about two eays. One is to depressurize the primary by a small 23 LOCA. The other is to shock-cool the primary from the 24 secondary. But you lose no inventory.
(1( ) 25 Does your case cover either or both of these or
215 1 what?
([) 2 MR. LARSON: This is the case for the small 3 break LOCA.
4 MR. EBERSOLE: At some particular place?
5 MR. LARSON: The cold leg.
6 MR. EBERSOLE: Well, you see, all of these are 7 qualifiers that you don't mention. And certainly, we are 8 going to start with the fact that the plant has to live 9 through a 1000 secondary prepressurizations.
10 MR. LARSON: I am ahead of myself. And that 11 is why. First of all, no break. That is a work without a 12 break, with a secondary transient. You probably have seen
{ll 13 14 this more times than you wanted to.
In SR1, the single-phase power capability is 15 somewhere around 80kw, 70-80 kw. We have to use that power to 16 cale the single-phase conditions with the reference Delta T.
I'7 We do that, and then follow and normalize the KD for single-18 phase.
19 The power trace this until we get out to about 20 -- this is plant time, now. For this particular reference 21 transient which is a 10 percent power or 10 cc squared break 22 simulation, we go two-phase or roughly 190 seconds.
23 Now at that point in time, we shift to the two-24 phase power scale. You should see, to maintain similitude and C) 25 power the sudden step change in power ratio as shown here.
q
216 1 Now, depending on what reference pressures we
([) 2 are using, you could get either of these three curves.
3 MR. THEOFANUS: That shows that effect then, to 4 drop the power.
5 MR. LARSON: Yes. I am not advocating that that 6 is the right way to control the power. I am just saying that 7 if one follows a scaling rationship, this is the power of g transient that should result.
9 MR. BECHNER: But that is starting pressure, not 10 transient pressure. You are not going to be varying the 11 power based on what the pressure does on the test.
12 MR. LARSON: I hope not.
-('N 13 MR. BECHNER: That is what I want to clarify.
Si~A 14 MR. LARSON: You are exactly correct. If the 15 pressure does something different than what I want it to do 16 based on the way I calculated this power curve, then by right 17 I should have a feedback system that controls the power to 18 match the pressure in the model facility so I can scale back 19 the reference pressure, j 20 MR. BECHNER: That is, if you are scaling for 21 some reason. I l
22 MR. LARSON: What? )
23 MR. BECHNER: If your scaling has been 24 imperfect for some reason.
('} 25 CHAIRMAN WARD: But it is in the two small 1
217 1 facilities in this respect.
O _ . _
hij 2 MR. DECIINER: Maybe I have a misunderstanding.
3 If you have a reference pressure of, say, 300 pounds and the 4 scaling is perfect and the transient goes, and you exist as 5 you think it should, then the power doesn't change other than 6 this step.
7 MR. LARSON: It has to follow the K.
3 MR. DECilNER: I guess what I am saying, if you 9 have a reference pressure of 300 and things get very different 10 and you are way off from what you thought you would be, then 11 in theory you should change.
12 MR. LARSON: Right, and that is what Theo said.
13 CIIAIRMAN WARD: But you are suggecting that that 14 change isn't very important, it is really just a step. If you 15 make the step change, that that is a good enough approximation 16 MR. BECIINER: I think that is still the bottom 17 line. But I think the real answer is, we do have a problem 18 here to see what these facilities do.
19 MR. SCIIROCK : What about the stored energy 20 aspect? llave you considered whether that is important here or 21 not?
22 MR. LARSON: In the core?
23 MR. SCliROCK: No, do to the step change in power 24 in the core .
C) 25 MR. LARSON: The stored energy in the rods O
218 1 themselves?
(AJ 2 MR. SCHROCK: Yes. The power really needs the 3 heat transfer and the power input.
4 MR. LARSON: I am not worried about that, no.
5 We saw Y. Y.'s data yesterday. The original concern I have 6 with a discontinuous change like that was -- well, in the 7 first place, one has to recognize when this facility goes 8 two-phase in order to affect that step change.
9 In the second place, if you recognize that the 10 system nas gone two-phase and there is void somewhere now 11 I have made the step change in power, what does that do in 12 terms of -- does the void collapse suddenly because they have CI .lh 13 now perturbed the system?
14 And Y. Y.'s data suggests that, no, you don't 15 get any adverse probicms from that step change in pcwcr.
16 MR. MICHELSON: How do you handle the stored 17 energy in the galance of the system? The vessel walls and so 18 furth?
19 MR. LARSON: You don't.
1 20 MR. MICHELSON: That is also a potentially 21 important situation that the void will collapse on a given 22 area. You don't attempt to put a tracing on the control or 23 that sort of thing?
O 24 MR. LARSON: Well, in the MIST facility there C"(_) 25 is tracing of the pipes. But neither Maryland or SRI have
219 1 tracing.
n h3_) 2 MR. SCHROCK: I guess your scaling considerationn 3 are showing that you would not experience temperature changes 4 in the loop that are anywhere like proportional to the power 5 change that you want to impose at this transition from single-6 phase to two-phase?
7 MR. LARSON: That's right.
3 MR. THEOFANUS: In that respect, the distortion 9 actually helps you. It gives you an overlapping damper.
10 MR. HSU: In fact, that is what happened. We 11 figure the reason of the no void collapse is because of stored 12 energy is a smooth transition.
{h 13 But we have to -- the dif ficult part -- is try 14 to sort out that distribution.
15 MR. THEOFANUS: One more thing. How is the 16 change if you were to change the speed going through the vent 17 valves from 9 to 10 to even 50-50?
18 MR. LARSON: How about the power trace?
19 MR. THEOFANUS: Yes.
20 MR. LARSON: Would that change?
21 MR. THEOFANUS: Would that change at all?
22 MR..LARSON: No, that is based solely on the 23 property groups. It really has nothing to do with the
( 24 fraction of vapor generated. That is assumed to go through q'c) 25 the rod.
220 1 MR. TliEOFANUS: How is the flow going to change?
m
( _j 2 The flow regimes?
3 MR. LARSON: How would they change? Based on 4 which of these power traces I selected?
5 MR. THEOFANUS: Yes.
6 MR. LARSON: Again, these powers here -- this 7 point is actually scaled to about 3.8 percent power. That is g why I said I was using 4 on that s.teady state energy balance 9 sbapshot. These powers are all less than this.
10 So, I guess, based on that I would still expect 11 to see only bubbly flow.
12 CHAIRMAN WARD: Let me play " chairman' for a h 13 minute here. We have used up all of Tom's time.
14 MR. LARSON: May I go to the bottom line?
15 CHAIRMAN WARD: Obviously, there is a lot of 16 interest in what you are saying. But we do have another 17 presentation. ,
18 MR. MICHELSON: May I ask a question until he 19 gets here. What happens if the vent valves freeze up in the 20 real world and all of them are opened?
21 CHAIRMAN WARD: That is what the codes are for.
22 MR. MICHELSON: Well, no, that is what some of 23 Theo's question is, of course. Now the pipe has to go way
() 24 down.
25 MR. LARSON: Then you have got a significant e
??1 1 distortion in model facilities relative to the plant in a n
(s- 2 number of areas -- flooding, flow regime.
3 MR. MICHELSON: Are you attempting to answers 4 as to the change in situation as to freeze up vent valves?
5 Because you can't test these things very easily.
6 MR. LARSON: I have been attempting to address 7 that in the report from the steady state basis. And MIST is 8 going to run some experiments where they will lock the vent 9 valves closed, and we will have the data base that the codes 10 can ---
11 MR. MICIIELSON: Thank you.
12 MR. LARSON: I am going to go right to the 13 bottom line.
14 MR. THEOFANUS: May I say something? If I 15 promise no to interrupt anymore, I would like to hear the --
16 if he can really do it in about five or 10 minutes, I think 17 it is very, very important to run through this.
18 If I promise not to interrupt, can you let him 19 finish?
20 CilAIRMAN WARD: IIe was volunteering to jump to 21 the bottom line. And I said you would like to hear a little 22 more.
23 Let's do that. Why don't I give you until 24 10 o' clock. But give yourself time to cover the. bottom line.
{,()
25 MR. LARSON: All right. I'have made:some
222 1 conclusions about the void quality. By virtue of the fact 1
(( ) 2 that two of the facilities run at reduced pressure, we know 3 that the quality, the core outlet, the phase change number 4 scale and the other number scale -- the quality, the core 5 outlet is going to be much, much less in those facilities 6 than it is in the reference plant.
7 That is the product of the scaling relation-g ships and how it works. The ultimate consequence of that 9 is that I believe there will be some void distortions in the 10 hot legs on both of the low pressure facilities for qualities 11 less than about 20 percent in the reference plant.
12 This grouping translates to about 20 percent
() 13 of the plant. Now, the TRAC calculation that I have been 14 using as a reference indicates that throughout the majority 15 of the transient, the core outlet quality is in the order of 16 20 percent.
17 And this void distortion comes about because 18 there is an assumption in Ishii's scaling criteria that says 19 that by and large the global slip is more important than the 20 local slip.
21 But there is a point -- and it happens to be 22 about 20 percent quality -- there is a point below which the 23 local and the global slip are of about equal importance.
f~) 24 And it turns out that the PGJ is different 25 enough in the low pressure facilities with the small pipe
223 1 diameters that it causes considerable effect on the void h 2 fraction for these lower qualities. And it may be as much as 3 10, 20, 30 percent.
4 Again, the ultimate consequence of that, I 5 don't know. Actually, it has a potential to effect driving 6 heads and that so -t of thing . But it is something that we 7 will just have to look at when we get the facility data.
8 Flooding distortions will be present if the P
9 vent valves are frozen shut, to address what Carl has asked.
10 All the pressure drops in these facilities I think will scale 11 quite nicely by this ratio provided that the resistance, the 12 mass flux and this quality parameter are maintained.
r]()
V 13 Now, again, from the scaling relationships and 14 some mixing calculations that I have done, I have concluded 15 that if temperature HPI is used in these low pressure 16 facilities, then there is going to be some distortions in 17 the cold leg subcooling. There will be excess subcooling 18 primarily because in order to maintain the subcooling number 19 there, which is required by their scaling relationships, you 20 have to heat the HPI water considerably.
21 If we don't use that, then we are going to get 22 excess subcooling. So the conclusion on break areas is simply 23 that at low pressure the ratio between some critical mass 24 flux in your reference plant and that in your model facility CD() 25 is different for subsool flow than it is for saturated flow.
l l
224 1 All it means is you might need multiple break
( 2 areas if you are extremely serious about trying to do your 3 best at reproducing a pressure transient in some reference.
4 (Slide.)
5 MR. LARSON: The next slide shows the scaling 6 relationships that I have referred to several times in the 7 past. All I want to point out here is something that I have 8 made reference to here several times.
9 The property groups on these scaling relation-10 ships, this happens to be the one for volumetric heat 11 generation rate.
12 The property groups for single-phase are quite
- ) 13 different than the property groups for two-phase. Now, the 14 reason at full pressure -- this is of no consequence. It is 15 simply the fact that at full pressure we assume we are 16 using materials and the like that are the same in the model 17 facility relative to your reference facility.
18 Hence, this property grouping is all one.
19 Likewise it is all one for the two-phase group, because we 20 are operating at typical pressure conditions. So we get the 21 same scaling relationship inverse of the length ration for 22 volumetric heat generation rate.
23 Now, it is easy to see here that if we are not 24 operating at a cold pressure, then these groupings will have
{j ) different numberical values. And that is why we saw in the 25
225 1 previous slide the fact that there was one scale for single-m (fm) 2 phase and a different scale for two-phase.
3 What is the consequence of that? Well, simply l 4 there are property ratio multipliers on the scale, and we 5 have seen that single-phase is not the same as two-phase. .If 6 p ressure changes, psi changes. Theo made reference to that.
7 Now, additionally, scale relationships require 8 that the phase change and the subcooling number be specified.
9 When we scale the high pressure facility, we usually say, all 10 right, we have got some reference condition at which we 11 specify the phase-change number and the subcooling.
12 We assume that during the transient, the h 13 relative ratio of the subcooling and the power in the model 14 facility stay the same relative to the reference as it was at 15 my reference condition.
16 We hope we can do that same sort of thing for 17 the reduced pressure f acilities. If we can't, then we get 18 back to what Theo was saying. We may have to change -- or in 19 theory, anyway -- we may have to change the power as a 20 function of time to match the pressure that exists in the 21 model facility, if you want to maintain similitude.
22 (Slide.)
23 MR. LARSON: This slide just reiterates again
()
e 24 some of the concerns, or however you want to phrase it, of the
! 25 fact that these fluid property groups exist on the scale
226 1 equations,
(,n_) 2 We know that if we are at full pressure, the i 3 pais are 1, the pressure. ratio model to plant is not 1 and 4 the psis are not one. For. steady. state it is not a problem.
l 5 By steady state here, I mean constant pressure. i 6 Because that psi then will be a constant. The pressure 7 changes will be a different constant.
8 For transient, that pressure ratio model to 9 plant may not be a constant unless we can somehow figure out 10 exactly the pressure transient we want to somehow induce on 11 that model facility.
12 The reference variables on subcooling to phase (ll 13 change do change with time. Hence, that property group is*
14 really a function of pressure and time.
15 It could be, if we were really worried about 16 maintaining similitude in the best of fashion. So that there 17 are some questions for transient experiments.
18 Firstly, is there a way you can operate the 19 facility to force that property group to be a constant and 20 get out of our seeming problematic area? How do these 21 reference variables change in time? Is that significant? Do 22 I have to worry about that?
23 If I can figure out what makes these psis
('i 24 constant, then how do I model the power and the subcooling in
(!?'
25 the model facility? And the reduced pressure facility.
227 1 (Slide.)
I() 2 MR. LARSON: The next slide just shows the 3 methodology I used to try to devise an ideal transient in these 4 reduced pressure facilities. I am not advocating that this 5 is the way to do it.
6 But I needed something that would tell me under 7 ideal circumstances, here i5 what I would like to get out of 8 these facilities. And then I wanted to take that result and 9 go off and do some more scaling analysis to try to arrive at 10 a method to scale a break and maintain the time scale and 11 make some conclusions about the need for heating the HPI.
12 Really, all I did was take the Los Alamos 13 calculation for the MIST nominal transient and also utilize
[)
14 the plant calculation that they provided me -- use that 15 pressure transient as a reference -- and ask the question, 16 hey, if in the model facility, after I pick a reference 17 pressure, if I force that model f acility transient to do the 18 same, normalize pressure transient as it was in this 19 calculation, what happens to the side groups, the property l ,.
20 ratio groups?
l 21 (Slide.)
i 22 MR. LARSON: The next slide simply shows a l
23 couple of different pressure transients that resulted from
( 24 going through that kind of analysis. The first one simply
(]) 25 says -- the xs here and the triangles here -- simply say, all 1
l l
L
228 1 right, I forced somehow on this model facility a normalized 2 pressure transient that is the same as this from the plant.
3 This is the plant transient (indicating). This 4 other transient here simply was devised by saying, all right, 5 let's force the quality scale multiplier to be a constant 6 throughout this transient. That produces this pressure 7 transient.
g Now, they are not tremendously different. And 9 that is important, I think. Because if we look on the next 10 slide ---
11 (Slide.)
12 MR. IARSON: These are simply numerical values.
13 These are all for the University of Maryland facility, by 14 the way. These are numerical values for those scale equation 15 multipliers.
16 Now, what this is really saying is, here is the 17 quality scale -- rho /g over delta rho ratio. The reference 13 here is the start phase in that plant calculation whi.ch is 19 about 1600 PSIs to plant reference pressure and 300 PSI --
20 actually, it is about 280 PSIs -- is the reference pressure 21 here for the Maryland facility.
22 What we see here is a time variation on these 23 property groups. To me, this is not tremendously significant.
h 24 What it means is that if we get close -- close, meaning within C- 25 the bounds of how the pressure varied on that previous slide
229 1 in the low pressure facilities -- then scaling the core power
( ,) 2 in the subcooling will be relatively easy. At least, the 3 power feed.
4 Because this is a 15 or 20 percent change. We 5 can't really measure it much more acurately than that, 6 probably. So there is not a large variation in the property 7 group.
8 So what this led me to believe is that if we 9 get close on our normalized pressure transient, we are in 10 pretty good shape.
11 And I think Y. Y. will have a slightly different 12 view of that opinion here after I get done.
{h 13 We have already seen the power traces. All this 14 is is that you have to be careful, you need to pick your 15 references pressures. Ilow you pick your reference pressures 16 will determine how this curve looks like. And it is not 17 insignificant because these differ by a factor of 2, 2-1/2 to 18 3.
l 19 (Slide.)
20 Fm. LARSON : As I said earlier, the reason for 21 going through these kinds of gyrations was simply because I 22 wanted to go back and given the perfect transient in these 23 facilities, look at some additional things like mixing, what
^
( ) 24 the effects of these different pressure operating conditions s
25 on the mass flux were, and how to define a break area that
230 g maintained time scale and all the other things that I was es (j j 2 worried about in terms of similitude.
3 (Slide.)
4 MR. LARSON: This slide is a produce really of 5 something I said before. And that can simply look at what 6 the subcooling number requires for HPI scaling. But it is 7 quite significant at low pressure.
g You have got to heat the HPI up considerably 9 in the low pressure facilities. You can't just use room-10 temperature water. Now, that is to maintain similitude for 11 the scaling of the HPI subgroup.
12 I am not saying you have to do that. I am saying that it is something that needs to be locked a para-
, ((ll 13 14 metrically in the facility to see if it is a really big 15 effect or not. I suspect it is a fairly significant effect.
16 We will see the ultimate consequence of that in a moment.
17 These ar,e just mixing calculations for the loop 18 using Theo's model and shows the significant effect of HPI 19 temperature on stalled loop mixing.
20 Now, one can take this one step further and ask 21 the question, well, given that the loop is flowing, suddenly 22 it stalls, you are putting this cold HPI in, or maybe you are 23 putting this warm HPI in, what might that do to my critical 24 fl0W7
)
25 Because there is a considerable difference in 9
231 1 the degree of stratification from the plant with the MIST
(}I g ,-') 2 condition to the reduced pressure facilities, with properly 3 scaled !!PI subcooling, which for Maryland is about 422K, and 4 for SRI is something in the order of 405K.
5 That is really hot water. Probably it is not 6 physically containable in those facilities. It is just a 7 problem to get that high. But that is what the exact 8 similitude when using the HPI water would require.
9 CIIAIRMAN WARD: Tom, why don't you go on and 10 say a few words about the counterpart tests, just a summary 11 of that ---
12 t1R . LARSON: All right.
O 13 CHAIRMAN WARD: --- and then'get on to your
{..r.J 14 conclusions?
i 15 MR. LARSON: All right.
16 (Slide.)
17 MR. LARSON: The data comparison techniques, I 18 would like to just say a couple of words on that before I get 19 on to the counterpart test. First, there are at least four 20 dif ferent viable methods of comparing the data from these 21 different facilities.
22 You can use the scaling relationships under
- 23 certain circumstances. The void scale, the quality scale, all
('
() 24 those things giving a reasonably steady state situation, we 25 can use in comparison of the data.
232 1 Equilibrium plots I think will be quite useful.
' (f,;'~)/
2 It will tell us how to run the experiments in addition to ways 3 of comparing the data and making statements about what may 4 happen in our reference condition based on what happened in 5 the low pressure facility or the MIST facility, for that 6 matter.
7 There is some portentially very interesting g applications of dimensional groupings to these data. Both a 9 method that the Japanese have written a paper, and also Novak-10 Zuber's presented some dimensional groupings that I think have 11 a lot of potential for at least categorizing these experiments ,
12 Then finally, thermohydraulic codes, I believe, A
,/ 13 ultimately are going to be the way that we arrive at a lot 14 of conclusions based on the data from these three different 15 facilities relative to the plant.
16 As far as the counterpart experiments, the need 17 for counterparts I don't think anyone can argue with.
13 (Slide.)
19 MR. LARSON: You need a data base for determin-l 20 ing and evaluating the system response differences that may 21 occur because of differences in scaling criteria used to 22 design the facility.
23 Plus, that same data base will help, I think,
() 24 to determine what the most reasonable methods of comparing
(:~ 25 the data from the facility to facility are. And, obviously, l l
1 i -
1
233 1 we are going to have to take the data and a potential method O
khrl 2 and iterate several times to refine the method before we 3 ultimately arrive at a good method -- a good transform, I 4 guess, is one way of putting it.
5 The counterpart tests -- I think the mapping 6 tests that are planned in the MIST facility are the best test 7 to meet the objectives as listed here.
8 I have also specified a transient experiment 9 that will be much more difficult to make a counterpart out of 10 that, but nevertheless I think it is something we need to do.
11 The mapping tests in MIST are really pseudo-12 steady state experiences that are designed to determine what
(~% >
13 the system conditions are at transitions between these
{,)
14 important small break LOCA events, like the single-phse, two-15 phase and natural circulation, natural circulation stall and 16 eventually boiler condenser mode. And there are a number of 17 other different phenomena also.
18 Ultimately, we will be able to run these 19 experiments on a counterpart basis in each facility- And I 20 think build maps of system inventory versus timing, perhaps --
21 normalized time, perhaps, of conditions at which we got these 22 transitions.
23 And we can do that using some of these
() 24 o dimensionless groups that are -- I hope using some of these (i ,J dimensionless grc that are listed in the report.
234 l
1 The mapping tests will be conducted with constant
( 2 boundary conditions. That is one reason that they are the 3 most prime candidate for mapping experiments. The only thing l 4 that will be changing here is the system investory, and of 5 course the pressure under certain circumstances.
6 Like I said, the system mass is the only control 7 variable. Pressure will be an uncontrolled variable, g depending on what is actually going on in the system.
9 (Slide.)
10 MR. LARSON: As far as the transient counterpart 11 mapping test, the most likely and logical selection for 12 transient counterpart at this point in time is the MIST ,
/~S 13 nominal transient.
3~. )
14 And that really is because it has had a lot of 15 analysis code-wise already. Y. Y. has been using it and it 16 is kind of a reference condition for his initial characteris-17 tics for the facility.
18 The initial condition should be quite similar 19 to the mapping test -- the steady state mapping test. There 20 are some differences in steam generator control that can be 21 handled in the low pressure facility.
22 There are some questions about how to do this 23 transient experiment, though. And that is, what pressure c)
l \_) 24 scale should I select, how should I select the initial
! ~ .
25 pressure, what should the break area be -- I have got some 1
1 l
l
..- =.-: -
- .-..:.-. .a -
235
,_s 1 recoumendations on that -- and what about the HPI conditions,
(>') 2 the combinations of flow, subcooling, et cetera, to match on '
3 a coanterpart basis the mistransient as best I can.
4 What this boils down to is, I have really got 5 some concerns about the fact that in the low pressure facili-6 ties we cannot match both mass inventory and energy 7 inventory at the same time. And that is where parametric i
8 experiments are going to have to be done to determine the 9 best method of conduct really is.
10 (Slide.)
11 MR. LARSON: Conclusions, rapidly. From a l 12 geometric sense, I think all the IST facilities are well- -
{ 13 scaled. They are all related by the Ishii general scaling 14 criteria. .
6 15 I really don't expect that there are any 16 significant local phenomena distortions that alter the 17 facility response. There are some exceptions to that, like ,
18 I said.
19 If the vent valves are closed, then there could 20 be flooding distortions. There may be some void distortions '
21 in the hot leg. The ultimate significance is not clear at 22 this point. This is just something that we have to look at '
23 and analyze.
24 The mixing is another thing that is strongly
(** ,
25 dependent on HPI temperature. Again, the ultimate significance
236 1 is not clear. I would expect it could be fairly a domineering (6 2 factor in terms of trying to maintain similitude. On the 3 other hand, it may not be as severe as I think.
4 The pressure can be scaled to the property 5 groups. We see that the single-phase and two-phase scales 6 are diacontinuous for reduced pressure. At this point in 7 time, I still recommend it for mapping tests and the transient 8 test to be conducted that the two-phase power scale be used 9 in lieu of changing the power in discontinuous fashion.
10 Now, Y. Y.'s data will make sure that that will 11 not be nesessary. But in terms of specifying the initial 12 conditions for the counterpart experiments, I have used only g)
</ 13 the two-phase power scale and just allowed the single-phase 14 part of the transient or the mapping test to be distorted.
15 (Slide.)
16 MR. LARSON: We are primarily interested in the 17 two-phase part anyway. As far as data comparison, I think 18 Lt will be reasonably straightforward for constant pressure.
19 For changing power and pressure, it gets more complicated 20 simply because of the effect of the potential change in the 21 property groups. And they are significant because they 22 determine the scale factors for the power.
23 The comparison methodoligies, there are several p)
(_ 24 different possibilities. I don't know which is going to work.
k' 25 None of them, probably, until we get some data and refine l
237 1 the methods.
(.O
.. 2 The codes I think should be an integral part )
3 of this whole analysis effort, and in fact are. I think this 4 is the part that has been left out of this presentation in l 5 the past. And it is assumed on my part that the codes 6 ultimately will be used and would be the ultimate transforma-7 tion mechanism.
8 But there is a lot to be learned by looking at 9 things from the viewpoints I have taken here, too.
10 (Slide.)
11 MR. LARSON: The last slide, counterpart 12 testing. I think the steady state mapping tests and any 13 other steady state experiments are going to be the most 14 useful.
I 15 The transients, in terms of actually trying to 16 simulate some referenced condition or some references 17 transient, are going to be complicated by break size, iiPI 18 flow and temperature, and time scale requirements. And 19 additionally, I think the steam generator effects -- heat 20 transfer.
21 In order to do that transient -- and for that 22 matter, the steady state counterpart test, I think -- will 23 require some trial and error experimentation to actually r^s
( / 24 learn how to run these facilities and the best method to do a C' 25 counterpart.
238 1 That concludes my presentation.
2 CHAIRMAN WARD: All right, thank you, Tom.
3 And before we take a break, Y. Y. has a brief 4 comment he would like to make on the scaling.
5 MR. HSU: Thank you, gentlemen, for giving me 6 a little time. My purpose in this presentation is to amplify, 7 clarify a little bit of what Tom said.
8 First, is about this -- Tom was doing all of 9 this based upon the assumption of a pressure ratio.
10 (Slide.)
11 MR. HSU: This is between the models, and in ,
12 our experience, and as I showed yesterday to you, that we
[)
c-13 could have a pressure ratio -- cur pressure ratio as shallower 14 than the plant pressure ratio.
15 This is ours and this is plant (indicating).
16 So the result is, we have a slower change of pressure. And 17 that means when you have a property ratio, we could be 18 closer to constant ratio. But with this kind of a situation, 19 be even closer to constant. That is the nice thing about it.
20 And then Tom mentioned void collapse. We have 21 done quite a bit of test and the void do not collapse. And 22 as I mentioned to Theo., that was, we suspect, because the 23 stored heat instead of curve like this generally makes it
( 24 smooth.
('= ,
25 Then, finally, when you reduce the two-phase
239 3 flow, whatever the steam generator that we have is not quite 2 -- is overpower. So you have to watch out how to control the 3 secondary side before closing it.
4 Now, that is about one thing. And may I ---
5 MR. SCIIROCK: Y. Y.,.can you clarify for me 6 what this D means? If you are in a single-phase first, when 7 y u reduce the power you have not a void. So why is there g an initial void collapse?
9 MR. IISU : If you do not have all stored heat 10 and also you lost your power, and then you are going out at 11 the same time, lose your fluid, and then what you need is a 32 supply additional vapor -- we supply additional vapor, you need.
13 pow r.
14 And that power actually is the one -- when you 15 have two-phase, you only need a small amount of power in your 16 reduced pressure. And then you need a small amount of power 17 to generate a lot of not a void fraction.
gg And that is why we need the lower power when 19 you do this and Tom worries about when you also lost your heat 20 source while you maintain your heat zinc, or else you lost 21 your entire heat, then the void will collapse.
22 We found that it does not really happen that way.
23 MR. SCl! ROCK: I still don't understand, but I 24 will talk to you later.
C' 25 MR, liSU : Another thing I think that probably we
I 240 i
I probably have an idealistic case of all of the parameter O
(b- 2 ratios and realistically we know that we do not achieve.
l 3 In other words, he was hoping all of the paths 4 would be following this exact same paths, as time goes on 5 between the two facilities. Since you cannot realistically 6 achieve that, the next best thing is to hope that all of the 7 physical phenomena are preserved and the sequence are 8 preserved.
9 So, in other words, if you traverse from one 10 region to another, or one phenomena to another, you traverse 11 different bodies but you only traverse the body from one side 12 to another.
13 In that way then, you requirement is no longer 14 on those ratios but over a range. Itere is an example.
15 (Slide.)
16 MR. !!SU: If you are talking about the low
, 17 transition, then you don't have to flow or follow the same 18 path from one pattern to another, as long as you save some 19 bubbly to slug, or whatever, you are passing from one to 20 another, then you are at range for flow pattern-wise'usually 21 a range of 10.
22 And you have a much more latitude. And the 23 same thing with ---
O(_/ 24 MR. Ti!EOFANUS : At the same time or different h 25 times?
a
.? 41 1 MR. IISU t Times? l (hl_/ 2 MR. Ti!EOFANUS: Do you want to be crossing at l
3 the same time or different times? 1 4 MR. IISU : No, the time , that is nothing. We 5 would like to see another translation like what I was saying 6 instead of exactly the time. We should use some other 7 parameters to translate the time. But as long as it goes 8 through that thing.
9 And then about the phase change number and the 10 subnumber which is very important, but to my mind they are 11 important two things. One is, for the flow instability. And 12 another is for flow scaling that Zuber was talking about.
{( ) 13 Again, because of that, you don,'t have to 14 preserve them exactly the same. For the stability, all you 15 need is within the range the transit from stable to stable 16 and vice versa. As long as you cross the boundary, you 17 d,n't have to cross at the same point.
18 And the global one in Zuber's equation shows 19 all you need is the difference. There was a missing sub-r 20 here, by the way. As long as there was a difference isn't 21 preserved , and then that's okay.
22 And so you don't have to have individually 23 preserved ratio. The difference ratio is preserved. And,
) 24 finally, the counter-curve flow. As long as you have a K c' . 25 star preserved on one side of the counter-curve instead of
-, - - - , --- ~ , - - , -
242 1 everyone's going the same way -- these are all examples --
n (h_) 2 and of course you worry about J-star because you might be 3 there.
4 It should not be a concern to us. And these 5 are the cases -- what I'm saying, we could be a little bit mor e 6 forgiving.
7 (slide.)
g MR. IISU : Now, yesterday the chairman asked 9 questions about what do you visualize are roles are in the 10 individual facility. And also in the facility comparison.
11 And also it. amplified Tom.
12 But the way I see, there were three expectations
(() 13 -- three levels of expoctation for interfacility comparison.
14 The fir.at one is a direct one-to-one comparison. That you 15 can only hope for with MIST. There is no way you can hope 16 for it from the smaller or lower pressure facility.
17 So the next step is that we are trying to find 18 the translation of steam, or like Tom said, comparison 19 methodology, so we can translate information to some one-to-20 one mapping to get another e.apar) son curve.
21 This is the goal. This is the effort which I 22 do. And whether we can achieve it or not remains to be seen.
23 But the bottom line position is still that we can use the
()
24 code, as long as the code can predict each facility using 25 same code. Then that is already a very good achievement.
. I 243 1 MR. CATTON: The codes have to predict each
( 2 facility, don't they, to make the program a success?
3 MR. IISU: Yes, that is what I am saying. This 4 is a bottom line. And that we think -- that is our format.
5 And I am hoping we can achieve this part.
6 And this will go beyond what this MIST is. If 7 we can achieve this, then the achievement goes more broader 8 in the generic.
9 MR. CATTON: Well, it gives you much more 10 confidence in what you are doing.
11 MR .IISU : Yes.
12 (Slide.)
( 13 .
MR. IISU : Then the finally, there is the role 14 of the facility is, first of all, to provide data for the 15 code assessment, to make sure the code works. And the second 16 part is to provide the data for interfacility comparison 17 and especially if we have some translation of scaling.
18 And then the third one, people did not talk 19 about it but I would like to stress, and that is the 20 sensitivity study. The very fact that we are different could.
21 be an asset. That is, if we are different.
22 Our ratios are different from one. And if we 23 come up result with three of them, then we can show that O
i) s_ 24 particular group is with a sensitive or crucial or not. And C:' we can eliminate a lot of frivolous worries and make it a 25
244 1 sensitive one. So that particular one I think is the one that t
h}' 2 our smaller facilities can beat.
3 And then finally is that we can use each one 4 to do the mapping data on the flow plot which Tom did not 5 show. Or we can use it for again to draw up scaling method-6 ology again.
7 This one transits beyond the generic uses.
8 Thank you.
9 MR. TIIEOFANUS: Trying to anticipate some of 10 the difficulties, I appreciate that. My personal opinion is 11 that what you show there for either facility, I see that you 12 are -- it's a consequence of what you are trying to do, you
{O 13 are going back again picking all of the possibilities for 14 quick operations.
15 In a way, I feel that this -- I hope that this 16 is only taking us into a kind of throw-back position. But 17 what happened here is very healthy in the sense that you are 18 looking and we are going to compare it.
19 And this process that Tom Larson was talking 20 about is extremely important and I hope that this position 21 is a very good reaction. And that there will be more thinking 22 and more sophisticated approach in actual scaling.
23 Just a comment.
24 cilAIRMAN WARD: But that is what your No. 2 .
C 25 proposal is trying to do.
l l
245 1 MR. IISU : Yes. In the very beginning there are
( 2 PeoP l e who always expect this. And I don't think any reduced ,
3 pressure can achieve this at all. And people always fall back 4 and say, well, we have a code. And that is realistic and that 1
5 is true too.
6 But we are a little bit more ambitious, a little 7 bit more optimistic, hoping that we can move that direction a 8 little bit more.
9 MR. TiiEOPANUS: To make that clear, yes.
10 CHAIRMAN WARD: Thank you, Y. Y.
11 At this time we will take a 10-minute break.
12 //
O
((J 13 //
. 14 //
15 //
16 //
17 //
18 //
19 //
20 //
21 //
22 //
13 //
() 24 //
25 //
,,w ==w. . . t . ., av , -x,1uammu, s.s ,s, ;;a.a:. ,, ;. . n .
1 246 1 MR. WARD: On the record.
2 Our next speaker is Mr. Knight.
3 STATEMENT OF THAD KNIGHT, LOS ALAMOS NATIONAL 4 LABORATORY 5 MR. KNIGHT: My name is Thad Knight. ;
6 (Slide) 7 MR. WARD: We are running a little behind and 8 some of our parti.cipants have." drop dead" times early in 9 the afternoon. So I have asked Mr. Knight to shorten up to his presentation, particularly the first section. But he 11 will have to get cooperation from the rest of us in doing 12 that.
(,) 13 MR. KNIGHT: This should prove interesting. I am 14 going to show you some of the results of the calculations 15 that we have done and those calculations will be done most 16 recently bya fellow by the name of Jim Steiner from our 17 organization, and in the past by Bob Fujita. That provides 18 a litle bit of continuity in people that you are seeing from 19 us.
20 I would also like to acknowledge the help of 21 EPRI in getting me around the time crunch in preparing some 22 slides.
23 (Slide) 24 At the last meeting before: you there were 25 presented to you a nominal phase calculation from RELAP-5 ks)
o 2 247 (ll 1 and a nominal case calculation for TRAC. And there were 2 some differences in the steady state conditions, specifically 3 differences in the steady state flows in the primary. And 4 the question came up why do we have those differences.
5 I want to go through a little bit more detail. But 6 now I've been asked to be thoroughly brief about it. I 7 would like to give an indication of where those differences 8 arise from. They do not arise out of the core because the 9 core, the user through input specifies the power distribu-10 tion.
11 And if you are noting details, loading detail is 12 sufficient then there is not a problem in specifying the.
13 thermal center in the core. The other thing is that Jim
- i, )
(>
k 14 Gloudamans of B&W has looked at thermal properties of the 15 fluid in the two codes and concluded they were essentially 16 the same.
17 So we are off to the steam generator models. Both 18 those represent the primary side of the steam generator 19 with two channels. There's the three tube section which is 20 wetted by the auxiliary feedwater and a sixteen tube section 21 that is considered to be unwetted by the auxiliary feed-12 water.
23 The secondary side is represented differently 24 between the two codes. RELAP-5 uses a single channel with 25 a special heat transfer logic to wet the two primary side
- . a .a a ..~. i 2 a m - .r-.a A - . - - w s ....----.:. .a--. -..
248 3
/~'i/
.(L[ I tubes to the secondary channel. TRAC uses two channels 2 with cross-flows and a separate channel for both the wetted 3 tube and the unwetted tube. And, like I said, there are 4 heat transfer differences.
5 MR. CATTON: Are those resulting heat transfer 6 differences or coefficient: heat transfers?
7 MR. KNIGHT: They are differences in logic 1
8 correlations, little emphasis on correlations getting you l 9 to the heat transfer differences. And they are differences 10 in the coefficient ultimately.
11 (Slide) )
12 What this slide shows you is a conceptual j f) 13 description of the secondary steam generator models for 14 TRAC and RELAP. I have shown the TRAC model on the left 15 and the RELAP model on the right. Let me first address 16 the RELAP model.
17 Again, there is a three tube primary and a 18 sixteen tube primary to differentiate between vet and dry 19 tubes and then a single channel for the secondary. You've 20 got heat slabs that conduct the primary sixteen tubes to 21 the secondary so that the heat slabs that conduct the I 22 three tub primary to the secondary channel. In TRAC we've 23 got, again, the same sixteen tube and three tube primary 24 split, but now we have the sixteen tube and three tube 25 secondary split with heat slabs conducting the primary to O
G
249 4
1 the secondary.
2 We have also modeled two cross weld junctions, 3 one that is close to the pool at the bottom and then one 4 at a fairly high elevation. What I have done here is I have 5 described heat transfer coefficients in terms of how they 6 are different; those properties that are the same, such 7 as velocity, I have left out.
8 Let me move forward to the words and not spend 9 any more time looking at that slide.
10 (Slide) 11 The primary side heat transfer both use the 12 Dittus-Boelter correlation for the fixed connection heat
('S 13 transfer to single phase liquid. The primary side is v
14 essentially the same for the two sides. You get a primary 15 difference because of the slight difference between the 16 two codes.
17 MR. CATTON: Do you make the --
18 MR. KNIGHT: On the primary side that does not 19 make any sense. On the secondary side the answer is no, 20 we don't. And we're going to use the correlation. In the 21 core I'm talking about in this steam generator.
22 MR. SCHROCK: In the steam generator, I'm sorry.
23 MR. CATTON: I was thinking of the secondary 24 side. But go ahead.
p 25 (Slide) y' t
n
o 5 250
( 1 MR. TEEOEANUS: Is it important, the heat transfer 2 coefficient, or do you use it hot enough so you don't care 3 what it is?
4 MR. KNIGHT: In terms of this problem, where you ar e 5 concerned about natural circulation flows and the driving 6 heads for those flows, it is important because it affects 7 the elevation at which you are getting the energy out of 8 the steam generator.
9 MR. THEOFANUS: I see.
10 Well, if that is important then don't you worry 11 about the kind of thing that, as we heard yesterday, that 12 one would expect to have this mixed conversion effects or
(~s 13 a slow flow distortion..of the velocity profile and occasio-14 nally in the heat transfer coefficients?
15 MR. KNIGHT: I'll be talking about the stuff that 16 we saw in SAI or some of the stuff that was in the movie.
17 MR. THEOFANUS: I was talking about the stuff 18 that Y.Y. --
19 MR. HSU: No, that is just a single phase on 20 both sides.
21 MR. THEOFANUS: Are you on two phase?
22 MR. KNIGHT: Right now we are primary is liquid /
23 solid and the secondary is two phase.
24 MR. THEOFANUS: Well, so on the primary side I 25 would imagine the heat coefficient is going to be limited by rm V
(E
- 251 6
(g 1 that and I would expect it to be affected by the natural 2 conversion effects in the lower velocities. And we know 3 that.
4 MR. KNIGHT: The codes have some logic built in 5 to account for differences between natural convection 6 versus forced convection. It is a very simplistic logic, 7 if I am understanding your question correctly.
8 MR. TIEN: It is mixed convection.
9 MR. THEOFANUS: I'm talking"about mixed convection.
10 I'm talking about your previous slide where you said both 11 sides use this water. Does that always use that? I guess 12 you led me to believe that you always use that.
13 tiR. KNIGHT: No, I'm sorry. It does not always
,.(
ki J 14 use this Dittus-Boelter. I am looking at a snapshot at 15 the end of the steady state calculation. When I go into 16 a transient mode then I end up getting heat transfer 17 coefficients that come in that the Dittus-Boelter under 18 certain conditions out of the natural convection correlation 19 and other conditions, raw condensation, if I get into a l 20 reverse --
i 21 MR. THEOFANUS: For the single phase it is higher i 22 than the first convection and higher in the natural 23 convection. How do you handle that?
24 MR. KNIGHT: We do not do anything special. When 25 we go into single phase heat transfer to the wall and where
- O Q.,)
- 252 7
( 1 we are dumping the heat from the liquid to the wall the 2 code picks up Dittus-Boelter and then it does a check on l
l 3 the natural convection correlation and takes the maximum l 4 of the two.
5 So that when you get down to zero velocities 6 you do not have a problem.
7 MR. THEOFANUS: It may be off quite a bit, it 8 may be off by a factor.
9 MR. KNIGHT: I don't disagree with that, Theo.
10 The reason why I'm up here presenting is to talk about 11 differences between TRAC and RELAP and try to understand 12 why the steady state conditions between the two codes are
(~') 13 different. I am not really addressing the code relative (r./'
3? 14 to reality at this point, if you will.
15 MR. THEOFANUS: That might be --
16 MR. KNIGHT: It is worth looking at.
17 MR. THEOFANUS: I didn't even know it was so 18 important as you simply have indicated.
19 MR. TIEN: I think that can be checked very l
20 quickly, just what the flow condition relative to -- you l 21 immediately see whether that fact. But I would like to ask 22 that the dimension of this viscosity that is included in 23 your correlation.
24 MR. KNIGHT: I think so.
25 MR. TIEN: The viscosity ratio?
(b s
C
- _ _ - - - - - . . - - - . = ~ = = _ _ - _ . _ _ . - - - . - - - .
8
( 1 MR. KNIGHT: Yes.
l 2 MR. TIEN: The answer was no, simply people said ;
3 not included. If you have a liquid that could be off a 4 very significant factor. The viscosity ratio.
5 MR. KNIGHT: I don't know if there is a ratio.
6 There is a viscosity term in there.
7 MR. TIEN: Yes, that is correct.
8 MR. KNIGHT: From the secondary side, looking at 9 the unwetted tubes -- and, again, I'm looking at a very 10 local snapshot -- in the pool both codes are using -- and 11 it openly sets the outlet, primary side outlet conditions 12 for these channels to essentially the same value. Because
(")
13 we specified the secondary side pres sure.
~-
14 Above the pool, again, we see the two codes using 15 the Dittus-Boelter for forced induction of the heat transfer 16 to steam for TRAC because the TRAC usually goes to this 17 correlation because void fraction is 1.0. The sink 18 temperature is T(vapor), which is greater than TSAT and 19 we see superheats which are as much as 26K.
20 (Slide) 21 Continuing on with the unwetted tubes on the 22 secondary side, RELAP-5 has a special auxiliary feedwater 23 wetting model incorporated into the code which for the 24 unwetted channel is forcing the code to Dittus-Boelter, 25 even though the local void fraction is not necessarily 1.
\J'
254
(- 1 (T 1 The sink temperature is assumed to be T(vapor) 2 where the T (vapor) is approximately TSAT. So there is some 3 superheat. The maximum that we saw is 8.4K. The net result 4 of this is to reduce the amount of heat transfer in the 5 unwetted channel above the pool in TRAC as well as in 6 RELAP-5. We are using the same correlations. We got about.
7 the same vapor velocities over there, but TRAC has a 8 smaller delta-T so that we got less heat transfer in the 9 unwetted channel. Thatis not the big actor, though.
10 (Slide) 11 CATTON: So the heat transfer in the unwetted channel is 12 relatively unimportant so an error of 20 to 30 percent in 13 the heat coefficient would not make a big difference?
gg 14 MR. KNIGHT: I don't think an error on that 15 magnitude is all that important. I'm not telling you that 16 it is an insignificant heat transfer that is taking place 17 there. The pool soaks it up. j 18 '
When you get into the wetted channels pool heat 19 transfer phenomena is the same as before. That pool down i
20 there at the bottom, again, basically forces the outlet 21 temperature on the primary side to be the same between the !
i 22 two codes. So ultimately what we are talking about is a l 23 difference in the hot leg temperature.
24 Because TRAC calculates it at a lower steady state 25 mass flow rate on the primary side and it forces the delta-T
\J
I l
255 10 .
i v
) I to be higher in TRAC than RELAP. Above the pool the TRAC 2 void fraction are less than they get -- for the snapshot 3 that I looked at all the way up to wetted channel they are 4 less than 1 but they are very high.
5 The code initially selects Chen Nucleate Boiling 6 Correlations. And, again, based on the local conditions 7 that the code sees the check is made on void fractions and 8 they are high. The code also calculates the force 9 convection heat transfer coefficient from the Dittus-Boelter, 10 the single phase vapor, and interpolates between those 11 two points based on the void fraction.
12 So as my void fraction approaches 1 I get less 13 l and less contribution cut of the Chen Nucleate Boiling I'
(.k .
14 1
Correlation. So I'm forcing -- at the high elevations 1 -
15 am forcing my heat transfer coefficients down.
16 MR. CATTON: But ,this is the thin film running 17 down the tube, isn't it?
18 MR. KNIGHT: That's right.
19 .Am. CATTON: That would not be boiling unless the l
20 file was thick.
21 MR. KNIGHT: That's right. I 12 MR. CATTON: So you really need to know the film 23 thickness of the delta-T?
24 MR. KNIGHT: You need to know the film thickness 25 of the delta-T and decide which.
(~s
.)
e
l 256 l 11 1 MR. CATTON: And if the delta-T is less than 2 10 degrees then the boiling probably won't be important.
3 MR. KNIGHT: Yes. l 4 But, you know, we are coming at this from the 5 standpoint of the code that had a heat transfer set of 6 correlations and logic built into it and we have not 7 modified those correlations at the current time for 8 application to MIST and others. When we applied the 9 correlations as they existed in TRAC to the Otis Test, 10 which was a high-steam generator, we got reasonably good 11 answers and a reasonably good comparison.
12 I think we showed you some of that in the ggg 13 previous meeting. It is only when'the steam generator was R' 14 reduced in elevation relative to the core that these 15 differences in the heat transfer begin to show up as being 1G important.
17 The auxiliary feedwater wetting model in RELAP-5 18 above the pool attempts to represent what I call the " thin 19 film nucleate boiling." It is not quite the terminology ,
l 20 but they try to represent the film on the tubes. They !
21 l
assume that if there is liquid there it is subcooled, it 22 is on the tube, and they have a special correlation for 23 that.
24 When the liquid goes saturated, again, they are 25 using the Chen Nucleate Boiling just as TRAC is, but they (v I
- .. w : .
. a..- a:.uw-.. .w.a. r.--w w : - -: . - - - - . ~ . - . . . . ~ - . - l l
l 257 g2 - --
h7 1 are not doing the interpolation between Chen at a high 2
void fraction and a void fraction of 1 as TRAC is doing.
3 ~ So the effect of this is to give RELAP-5 a higher heat 4
transfer coefficient than what we are getting out of the 5 TRAC code.
6 The net result is that the TRAC calculation 7 above the pool on the wetted channel can produce, does 8 produce, lower heat transfer coefficients than RELAP-5.
9 (Slide) 10 A summary: TRAC tends to force more~o'f the heat 11 transfer down toward the pool than the corresponding 12 calculation in RELAP-5. The effect of this shift in the C 13 heat transfer i.s to lower the primary side thermal center 14 in TRAC relative to RELAP-5. The change in the thermal 15 center gives me a lower flow.
16 And I think if you look at the next page in the 17 handout that is what you see.
18 (Slide) 19 The bottom line for us right now with regard to 20 TRAC, and I assume it is with regard to RELAP-5, is we are 21 going to make no changes to the code until MIST produces 22 data indicating the correct system. At that point when you 23 get that data out of MIST and it tells you what the right 24 answer is we may or may not need separate effects data to 25 help us develop a superior auxiliary feedwater model.
c
258 13
)
((~/ 1 Right now we are not going to go in and change 2 the code. We are not going to go take TRAC code or the 3 RELAP-5 auxiliary feedwater wetting model, but if that 4 looks like it is the superior model when the MIST data 5 comes out we may incorporate it. If the MIST data indicates 6 that there are problems with that model also then you may 7 need to go into a development plan that requires some data '
8 that we don't have at the current time.
9 MR. CATTON: Well, your method presently is 10 physically incorrect. You are just-hoping that it is 11 close for some reason.
12 MR. KNIGHT: Our model right now has deficiencies.
r~S 13 MR. CATTON: It is physically incorrect. It is
- 14 not thin film. It- is nucleate boiling, that's different.
15 MR. KNIGHT: That's right.
16 MR. CATTON: If you're lucky the heat transfer 17 coefficients in both cases are so high they won't matter.
18 MR. KNIGHT: That's right.
19 But the RELAP model has the same problem except 20 with the --
21 MR. CATTON: Then the comment goes to that RELAP-5 22 model as well.
l 23 MR. KNIGHT: True. '
24 All I'm saying at this point is that neither code 25 in my opinion is physically correct in this. We are going e
A o
14 259
( 1 to wait and see whether or not it is a significant thing 2 one way or the other. If the RELAP model gets you closer 3 to the answer without having to go -- acceptably close to 4 the answer without having to go into a development plan 5 then we will do that.
6 MR. CATTON: Let me offer you something. You 7 don't need a separate effects data gathering. I mean, you 8 don't have to run any experiments. This kind of data exists 9 from the Oak Ridge Seawater Conversion Studies in great 10 quantities. They have vertical tube evaporators and that 11 is what you have here. They have thin films running down 12 them.
13 They are categorized as a function of the amount g
c~ 14 of running water down the tube, film thicknesses, all kinds ,
15 o'f things are_already available.
16 MR. KLINGENFUS: Klingenfus-from B&U.
17 The RELAP correlation is in there. It is a 18 following film heat transfer correlation. It is a so-called 19 liquid falling in and it does take into consideration the 20 film thickness as well as it takes into consideration the 21 temperature between the wall and the film temperature itself.,
22 Only when it becomes saturated does it switch over.
23 MR. CATTON: But the point is you should not 24 switch over unless the film is thick and the temperature 25 difference is high enough. And the data to decide this v
C 9
15 '
260 (ll 1 exists. It is not a mystery, it has existed for twenty 2 years.
3 MR. KLINGENFUS: I just wanted to make sure that 4 there wasn't a misunderstanding.
5 MR. KNIGHT: We're not saying that the data 6 doesn't exist. Part of the problem that we get concerned 7 about is the very thing that the SAI experiment is looking 8 at in terms of spreading. MIST has considerably more 9 instrumentation in the steam generator than we have had 10 previously and we can get an idea about how far around that 11 wetting can go.
12 And so you need not only a model that deals with 13 a thin film on the wall, but you also need a model that C.i
'k"})
14 deals with the wetting across the wet side to the dry side 15 of the steam generator. And we would like to treat that '
16 as a package in the event that we have to. If we don't 17 have to then we don't want to go into a development plan.
18 NR. SCHROCK: Could I make a comment about this, 19 not wanting to go into a development plan? It seems to me 20 when you know very clearly you have inadeqdacies it is not ;
I 21 so much the comparison of RELAP-5 versus TRAC in terms of i 12 these inadequacies. They exist in both of the codes and 23 we have known that for a long time. There was a lot of 24 resistance to ' hanging them c
25 But.the comparison is really between the large
('^i
p 261 16
( 1 break relations that are needed and the small break 2 relations that are needed and the transition from the 3 state of the codes from large break to the state of the 4 codes for the small break never really got at those heat 5 transfer packages for either one of the codes.
6 And we consistently have this resistance to fixing 7 them in a realistic way given the state of the knowledge 8 of those heat transfer relationships. Why can't we do that?
9 MR. KNIGHT: We can do it. I am not saying that 10 there is a resistance on our part to making changes in the 11 code. What we do have, though, is a list of changes that 12 we have seen from other experiment programs, not just MIST
/^i n" .;
13 but other things that we are doing, that, given a limited 14 amount of money,.we have to go in and prioritize which ones 15 are more important to the questions that the code must 16 address.
17 MR. SCHROCK: I don't buy it. I just can't 18 believe that we are going to have successful codes that 19 contain known serious deficiences in the heat transfer 20 package. It is more important application and we will have 21 to grant this transient in the small break. Transients 12 are in fact important.
23 MR. KNIGHT: These transients in small break ,
24 applications are important. But keep in mind that MIST is 25 being initialized under conditions in actual circulation.
(s s)
262 17 (k,_) 1 MR. SCHROCK: My comment was far more generic 2 than MIST or any one of these individual things. It is the 3 end result, it is the code, or isn't it going to be a really 4 useful tool. And when it is properly assessed against the 5 data that come out of these programs in 2-D and 3-D and 6 all of the programs can it really do the extrapolation to 7 the full-scale plant.
8 I don't see how in the world we're going to 9 argue that on the basis of.the code that contains arbitrary 10 interpolations between two heat transfer coefficients 11 both of which are out of context for the present application.
12 I mean, that is nonsense.
- fs 13 MR. KNIGHT: I agree with you.
(0 14 MR. BECKNER: Mr. Chairman, may I address this?
15 It seems to be coming up agaiit and again. I would like'to 16 try to explain what we are doing with the codes. The 17 codes have been frozen by mandate, the definitions are 18 frozen. Sometimes it's not clear,but they have been frozen.
19 And there were basically two very good reasons, I think, 20 for freezing.
21 One is that we are trying to assess the codes, it 22 is a moving target. Jesse yesterday said it would be'very 23 nice to have a code that can predict MIST, University of 24 Maryland and whatever to give confidence the way it has 25 been in the past when you never had that because you never O
LJ h
_ ~ . - - . - . _
r 18 263 )
( 1 knew what version of the code you were using.
2 So the first reason to freeze the codes is to try 3 to assess something and know what you are assessing. The 1
l 4 second reason is sort of indirectly related, is that in 5 the past we have identified a deficiency such as this and 6 we have quickly gone out and wired something in to fix it.
7 And that has been successful and sometimes it has been 8 unsuccessful.
9 We do plan to fix these things. The code is 10 frozen for a certain period of time then there is a plan to 11 release a new version of the code.- And I think it is 12 Probably better to do this in a more systematic manner. We
(' ' ; 13 ha'e two examples. .
_J
~
14 One is the flow regime map in the hot leg and 15 the other is this auxiliary feedwater problem. We have data 16 coming in that will assist us in both of these areas and 17 we do plan to make use of that data. But I think it is 18 premature to quickly ram something into the code and then 19 later finally find that model only is applicable to MIST.
1 20 ' So we will make these corrections but we are going 21 to do it in an orderly manner, I believe.
22 MR. SCHROCK: I'm still not convinced because 1
23 we have not seen that happen in the past. And the test of 24 whether the heat transfer is or isn't important in a given 25 transient is always based on very global interpretation of
.~ j
264 19
(.) 1 an integral test versus the code prediction. There is 2 always noding sensitivity and there is always some fussing 3 around with the noding in order to get the correspondence 4 which does finally emerge.
5 And then a conclusion is that, well, the heat 6 transfer probably wasn't that important during that 7 particular transient so the package doesn't need to be 8 changed. I can't believe that the cost of installing the 9 right heat transfer relationships is really so great that 10 it needs to be avoided kind of at all cost.
11 And we don't need new tests for MIST, University 12 of Maryland or SRI-2 to establish that those particular 13 things that were just discussed in both of the codes are L' 14 just dead wrong. We don't need anything new to establish 15 that.
16 Ivan has pointed to some available experimental 17 data. A little further looking would reveal still more 18 available experimental data. There are better things that 19 can be done and waiting until we have an integral test i
20 result, which I can guarantee we are going to be interpre- l l
21 ting in a way_that says, "Okay, those packages are not 22 that important to the global response of the plant and 23 therefore we will leave them alone, even though we know that 24 they are wrong."
25 The fallacy in that kind of thinking is that every xcs /
l l
20
- 265
( 1 time you subject the test -- or you subject the code, 2 rather, to an application to a new problem you are going 3 to have a new surprise because now there are going to be 4 new sensitivities to these which were relatively insensitive 5 in previous applications.
6 I think it is a mistake.
7 MR. BECKNER: I think we are in agreement. I 8 think it is a matter of how to do it. We are not going to 9 neglect these things, I don't believe.
10 MR. WARD: But I think Virgil is suggesting that 11 he would like to see you make it more explicit differentia-12 tion between code changes that are resulting from new r~N 13 information generated in the current experimental program a
(z[ i 14 and code changes where you have identified something in 15 the analysis that the physics is wrong in the code.
. 16 And I guess he would like to see you differentiate 17 them. Maybe there could be a different schedule for those 18 two types.
19 MR. BECKNER: That is a gray area we make. good use 20 of as far as what is in error.
21 MR. WARD: Yes.
22 MR. MICHELSON: Is it an error in logic or physics?
23 MR. KNIGHT: Virgil, we are not ignoring those 24 problems. The development effort is going on, all be it at 25 a reduced level than what it has in the past for obvious x
hm/
. . ._.:. :... .a a.u.a..: ..=..: ~ .-.=.. m _.a u w - _- ._.=__;..~.r...-..---.
- 266 21 O
(bw 1 reasons. But we are not ignoring that and we have plans 2 to release future versions of TRAC that will address some 3 of the errors and problems that we have seen in the code.
4 We have to have a certain prioritization of those 5 problems in order to spread a limited resource. And this 6 problem, while I agree that the physics is not right, when 7 we do our plant calculation we don't really get into this 8 problem. Because the plant always starts off as it being 9 pumped and the pump sets the flow. The plant starts off 10 with a normal feedwater injection and you don't get into 11 the problem where these things result in a significant 12 deficiency in the overall calculation. ,
()
13 And I am, again, beating around the bush that you i
14 said you didn't want me to beat around, but because of those 15 things -- you know, this thing right now is taking a 16 slightly lower priority to us than the problem calculating 17 the early rewets, for example.
18 MR. CATTON: But this is part of the basis for the 19 Follow-on Program. Maybe if you do it right now you won't 20 need the.Eollow-on Program. I don't think it is cost 21 effective to delay it.
22 By the way, it is not just Virgil who feels this 23 way. I feel very strongly.
24 MR. BECKNER: I think, Ivan, it is because we 25 kind of quick-fixed this before thinking and we're very
()
)
1
22 267
,- \ -- .- -
P" I sorry. This is a case of fine tuning.
2 MR. CATTON: This is a case where it is not )
3 tuning, it is fixing the physics. And that is very, very I
4 different. It is very different than going in and adjusting 5 the noding or something to get the answer to fall more in 6 line.
7 MR. BECKNER: I don't think we know what the 8 physics are necessarily. We know that what is in there is 9 wrong.
10 MR. CATTON: And we know that it is a film Il running down the tubes. Now, you ought to have that kind 12 of heat transfer built into it, or maybe you ought not be
( 13 using the code until you do. I think using the code that
(:- )
14 has physics that is incorrect -- it seems to be important 15 with respect to this thermal center -- is just wrong.
16 I mean, if we had a graduate student doing that 17 we would flunk him, we would throw him out.
18 MR. BECKNEh: I've already graduated.
19 MR. TIEN: I think Dave mentioned very clearly, ,
20 and I feel very strongly alsc, I think that this attitude l
21 like any code changes all can fall into one category, some 22 elements of the code changes that you should change, you 23 know. We develop information is such that the physics is 24 wrong you should change. !
25 There are others not very totally settled elements, V
(~
e
268 23 n
(h- 1 I agree, and maybe you should freeze that for the moment 2 and see what is going on. But if you just don't change 3 those clearly enough and then you say, "Well , we check with 4 heat test and so on, that is meaningless, I think. That 5 will mislead people.
6 MR. KNIGHT: Let me restate what I said up front.
7 We are not going to change TRAC today to look like RELAP 8 because we don't feel that what RELAP did gets me all the 9 way to the answer. We are not saying that, you know, we're 10 going to ignore this problem. We are going to consider it.
11 But we have limited resources to attack a whole 12 range of problems and at some point we have to sit down with
(}(~ 13 14 our funding agent, the NRC, and set priorities for fixing all of this stuff. And that is the process that we are 15 going through now.
16 I just can't bring myself, though, to go into 17 TRAC and make the TRAC heat transfer model look like RELAP.
18 MR. CATTON: I don't think anybody is asking you 19 to do that, nobody is proposing that.
20 MR. KNIGHT: I realize that.
21 But on the other hand, that was the statement 22 that I made. I am not going to do that. I did not say 23 I was not going to ultimately fix the code.
24 MR. CATTON: When you use your frozen code doc.you 25 change nodalization?
\c 1 (
O
24 269
( ) MR. KNIGHT: Within the noding studies, yes. l 2 MR. WARD: I think we have made the point here.
l 3 MR. CATTON: We are not penetrating.
4 MR. WARD: Let's move on to thenext point.
5 MR. KNIGHT: I am almost afraid to say this. The 6 TRAC code with regard to that that is being used to do the 7 MIST analysis, we have not made any changes for the code 8 models based on the results from GERDA, OTIS or MIST. And 9 when I say that I mean we are not adjusting the code models 10 to give us the detailed answers that the data is showing 11 us.
12 MR. THEOFANUS: Excuse me.
13 How different are the velocities that you mentioned 14 to do this, roughly?
15 MR. KNIGHT: Ue are about a third lower than 16 RELAP.
17 MR. THEOFANUS: Thirty percent lower?
18 MR. KNIGHT: It turns out that the way --
19 MR. THEOFANUS: Is that duplication of the thirty 1
20 percent -- is there any other duplications there? i
- 2) MR. KNIGHT: The way the current nominal test is 12 specified the two codes give essentially the same timing of 13 events.
24 MR. THEOFANUS: Thank you.
25 i MR. KNIGHT: We have found code logic errors and V
O
270 25 1 FORTRAN errors that we have been correcting as we go along.
2 This is allowed under the concept of the frozen code in 3 using in applications and clearly all of that stuff is 4 factored into the next version of the code that is going to 5 be released to the public. These errors have been corrected.
6 (Slide) 7 MR. BECKNER: Those code errors were released 8 immediately?
9 MR. KNIGHT: At this point we are making releases 10 at least quarterly.
11 MR. BECKNER: We are not waiting for two years?
12 MR. KNIGHT; We're not waiting for the two years 13 to get them out. But we like to get some experience in-14 house with them before we release them to the public.
15 TRAC input for MIST. We have tried to use only 16 physical modeling techniques. We don't know of any place 17 where we have just thoroughly distorted the geometry of the 18 facilit.y in order to get a desired answer. The input 19 geometry is based directly on the facility geometry. The l
\
20 trip and cor. trol functions are based on the planned :
21 operation of the facility. j 22 That is, again, we are not trying to prove --
23 through input to contrive input and come up with a desired 24 answer for the test. We are trying to base the input solely 25 on the other codes.
I 1
271 26
(/~l J
1 (Slide) 2 Noding sensitivity studies. We have done a few 3 of these. We think that it is a valid part of using the 4 code, at least until you get to that point where you have 5 user guidelines that are sufficiently detailed and 6 sufficiently general that you can go and look up what the 7 noding should be for a given application.
8 We have looked at a detailed upper-plenum i cylinder. The basic analyses that we have done for a long 10 time did not include that upper-plenum cylinder representa-11 tion. B&W has had long discussions with us about how 12 important that upper-plenum cylinder is to the calculation 13 of their flow of stuff into the hot leg.
{L/} 14 We have done some analyses looking at that. We 15 have tried -- we have incorporated a detailed model of the 16 upp'er-plenum cylinder and it shows that the events shift 17 in time. It is a significant shift but it is not really 18 changing the calculated events themselv6s, 19 In the 1-D components the detailed noding of the 20 upper-plenum cylinder requires very small cells, on the 21 order of about 20 cc, in order to get all of the flow paths !
22 hooked up correctly. These small cells adversely impact the 25 time-step size because the levels tend to reside and as the 24 level oscillates up and down a little bit the code tries to 25 track that level and that gives us a small time-step size.
272 27
( } 1 MR. MICHELSON: Is that upper plenum modeled at 2 all in MIST?
3 MR. KNIGHT: Yes, there is a representation in 4 the upper-plenum cylinder. OTIS did not have it.
5 MR. MICHELSON: And SRI doesn't try to do anything 6 like that?
7 MR. KNIGHT: I'm not sure.
8 MR. SURSOCK: Yes, there is in SRI. The upper-9 plenum is the rods. It has a dry rod that represents the 10 control rods.
11 MR. MICHELSON: But we are talking about here is 12 the c'onfinement of the upper-plenum region, is that right?
r~s 13 Is that modeled in the SRI?
(Y ]
14 MR. SURSOCK: Yes, you have to correct flow path 15 as far as distances between the upper head in the upper-16 plenum and the upper-plenum and the hot leg.
17 MR. MICHELSON: That is in there then?
18 MR. SURSOCK: Yes.
19 MR. MICHELSON: How about in Maryland, is it in 20 there?
21 MR. SURSOCK: Pardon me?
22 MR. MICHELSON: Is the upper-plenum modeling 23 incorporated in the Maryland facility as well, or is that 24 just an open --
25 MR. SURSOCK: No, we have large rods, we have A
(.Q s
I l
28 273 l
1 resistance.
2 MR. MICHELSON: I don't want to take up any more 3 time now.
4 MR. KNIGHT: I don't'.want to dwell on it, but 5 because of these small cells, 20 ces is af.prettp~small 6 mmount of liquid to try to track. We have seen adverse 7 effects on trying to run some of the other calculations 8 that were on the loop, so we had backed off to what we 9 were using before pending an investigation of using a 10 multi-dimensional modeling with a more flexible track 11 vessel component to represent that because of the small 12 time steps that we were seeing coming out due to these
(~x 13 things, these small celle.
f .)
L- 14 The problem in the vessel itself we don't think 15 will be a problem and it will allow us to hook things 16 together more directly and more physically.
17 MR. THEOFANUS: Let me ask you this: The way I 18 perceive your presentation of this particular issue is that 19 it is an organization problem. Is that all or are there I
20 also physics problems having to do with after you hook l
21 things up what goes this way, what goes that way? Are there l 12 some significant physics there that maybe we are missing?
23 MR. KNIGHT: I do not believe personally right 24 now that we are missing any of the physics problems. The n
25 flows tend to be low enough that you are not going to get, n
(3'
.~...:2_. ._m.._~.-2..._._.m__...__,._......a..: .. . __ ._ . _ ..;.. u .; __
29
( 1 you know, just the stuff penetrating through significantly 2 different voids.
3 MR. THEOFANUS: But when you hook up another 4 component how do you do it now. if you have a T there?
5 MR. KNIGHT: I will get a mixture and it is the
- 6 void fraction upstream.
7 MR, THEOFANUS: And you do that both going up 8 and going to the side?
9 MR. KNIGHT: That's right.
10 MR. THEOFANUS: How do you know that this is 11 right? It seems to me that you have all of your liquids 12 spilling over the side and a very significant separation
(^)
- J 13 of the two qualities the void fraction going up.
c1 14 MR. KNIGHT: I don't know that it is right in all 15 cases.
16 MR. THEOFANUS: But you know it is wrong in all 17 cases if you're doing that.
18 MR. KNIGHT: No, it is not wrong in all cases.
19 MR. THEOFANUS: You can't do this if the mixture 20 is coming off the side, there is no way in the world. I 21 MR. KNIGHT: Low flow is not the only case.
22 MR. THEOFANUS: But that is the case that you 23 brought up for this particular problem here. That would be 24 wrong every single time you did it. At no one time would s
25 it be right. And off by a lot, not by a little bit.
U
)
..aw._ ..~.wa..... c; .:o aw ww.u .. = -
a -
a 30 275 O
('"# 1 MR. KNIGHT: We have done this in OTIS in much 2 the same way and the margins that we get there say that 3 we are getting an answer that is approximately correct. It 4 doesn't overlay the data but the data is a little bit --
5 MR. THEOFANUS: When you know there is something 6 wrong, I mean, you have figure going up this way and 7 you have enough off the side like that you know that you 8 are doing something wrong. I don't care if you got this 9 in all other facilities you used it.
10 I don't try to say that to emphasize the previous 11 one because you know it happens. Maybe I'm not exactly 12 as extreme as the ot;ters, but in this particular case it g~
() .
13 is important where the void is separate. And you know
- 14 that what you are doing is wrong and I think you are 15 approaching this problem. I just wanted to state that I 16 look at this.
17 MR. KNIGHT: I'm not approaching it as being 18 only an organization problem.
19 MR. THEOFANUS: That's what I said and that's 20 what you told me. You told me you understand the physics 21 and it-is only an organization problem.
22 MR. KNIGHT: If I told you that then I did not say 23 it correctly. What I told you is that I did not represent 24 the upper-plenum annulus and the upper-plenum annulus is 25 an important geometrical factor in this facility. And this 1
t (2) :
- z. ... . . . . . .. . . . . . .a..... . a .. ~ . . 2 .
. - - . - ~u 276 1
1 is simply looking at that, it is not really looking at how 2 the donor stuff is in the side legs of the -- going into the 3 hot legs. -
4 And two models with and without the detailed 5 upper-plenum annulus modeling I used the same model.
6 MR. THEOFANUS: You have even more. You compound 7 your problem if you put this upper-plenum annulus and you 8 keep doing what you are doing in this framework. I think 9 you are going to compound the errors of downering two-phase 10 mixture in different directions. Really, you are going to 11 compound them.
12 MR. KNIGHT: We know that.
f) 13 (Slide) 14 We have done studies looking at the noding around 15 the cold leg injection and the mixing of the OTIS noding 16 tests that we save looked at. This centers around the cold 17 leg injection and it revealed that we can come up with a 18 noding that is unacceptable. That is, that it does not 19 give us the right mixing effects and condensation effects.
20 And what that noding shows is that if we put one 21 cell for the cold leg and we inject the HPI into and 22 conduct that to the downcomer that that answer gives us a 23 bad comparison to data. If I put more than one cell -- and 24 at this point it is two to four -- I get essentially the
,s 25 same answer.
t I v
't
. _ u m._u. .;__ _. _. _ _ _ . . .__ . . . . . . . _ _ _ _ _ _ - . ..u .. _ _ _ _ . . _ , _ _
32 277
(;' 1 And what I am looking for, again, is the noding 2 sensitivity in trying to have a noding framework that is 3 converted to an answer. And this noding sas consistent Qith 4 this finding.
5 (Slide) 6 Two-channel modeling in the steam generator results 7 from the OTIS analysis show the importance of better 8 representing the flow field and the heat transfer in the 9 OTSG secondary. The approach in the 1-D components has been 10 to split the secondary and the primary for our code into 11 two channels based on the data from the vendor.
12 We got information from B&W that told us jll 13 approximately three tubes were wet and sixteen tubes were
. l' 14 dry. Further experience in MIST with the lowered elevation 15 steam generator indicated the importance of the cross-flow IC connections.
17 We believe that the ultimate calculational 18 resolution of this problem will require at the very least 19 multi-dimensional models. This gets around to the question 1
20 of how do you predict the plant when you don't have full-21 scale plant data.
12 MR. THEOFANUS: How easy is it to go in and make 23 it so that in a case like this your donor liquid only is
'24 coming off to the side and what is left basically will --
- 25 is that something -- can you go into the core and describe U
9
33 278 (h 1 this formula instead of describing void fractions?
2 MR. KNIGHT: Yes.
3 MR. THEOFANUS: You can do this to put this as 4 an input?
5 MR. KNIGHT: No, not as in input. Subsequently 6 within the framework of the numerics you can alter the 7 donoring instead of taking volume average and properties 8 and donoring them up through the tier down through the 9 side leg. It requires a model to do that, yes.
10 MR. THEOFANUS: So you have to go in and actually 11 change all of the equations, right?
12 MR. KNIGHT: No, I don't have to change all of f x, 13 the equations. What I have to change is the information 14 that tells it what to flow through, that par'ticular 15 interface.
16 MR. THEOFANUS: Is that a monumental effort or 17 one year, one day, what is it?
18 MR. KNIGHT: It is more than a man-month.
19 MR. THEOFANUS: That would be the obvious thing 20 to do. You can bracket on one end and here you have one 21 going to the side. You have an equal void fraction donorer.
22 MR. KNIGHT: If you want to go to the point where 23 you have forced complete separation.
24 MR. THEOFANUS: But not the separation. I want to 25 force that the liquid goes to the side; wWhat is left, which rx ;
O i
o 34
( 1 is liquid and vapor, going up.
2 MR. KNIGHT: That I could do fairly directly.
3 MR. THEOFANUS: How do you do it?
4 MR. KNIGHT: There are input flags that allow us 5 to switch that.
6 MR. THEOFANUS: You might want to think hbout 7 that because I think that bracket would be heavier. Below 8 a certain flow criterion you could spill over and the 9 two-phase mixture, what is left, will go up.
10 MR. KNIGHT: The argument that we have had with 11 B&W is that the flow through the vent valve is pure vapor 12 or a two-phase mixture. And what you are arguing for is 13 a two-phase mixture versus single-phase liquid.
ggg I
14 MR. THEOFANUS: Yes.
15 At least I would bracket. And I don't know how 16 they get onto that with a two-phase mixture but it could 17 be maybe --
18 MR. KNIGHT: We are arguing that it is single-19 phase vapor.
20 MR. THEOFANUS: Going on to the vent valves?
I 21 MR. KNIGHT: Yes.
22 MR. CATTON: But the level ~in the core is all 23 right, it is only steam.
24 MR. THEOFANUS: Of course.
25 MR. KNIGHT: True.
f .
. ,Y
($
1
280 35 tq ] ,/ 1 MR. THEOFANUS: No, that is obvious. Of course I
2 that is the case. But now we are talking about where you 3 actually have a two-phase separation at the upper-plenum l
4 and I think that you can bracket and analyze it by doing i l
l 5 this calculation. And maybe you can run it and you can do 6 it, but let's do it.
7 MR. KNIGHT: We've had a lot of discussion in the 8 past on the multi-dimensional or annular configuration of 9 the top of the MIST downcomer.
10 (Slide) 11 We have run calculations looking at 1-D versus 12 3-D modelings of the MIST downcomer behavior. The results 13 of these calculations show that there are no significant 14 multi effects. That is, velocities are always in the same 15 direction and approximately the same magnitude. And we did 16 not see temperature distributions around the downcomer.
17 (Slide) 18 At this point what I would like to do is describe 19 to you very briefly and repidly the latest calculation of 20 this MIST Nominal Test 310000. This slide gives you some 21 of the information pertinent to what this test really is. I 12 don't really want to belabor it other than to say that the 23 initial core power is 3.9 percent. That represents 3.5 24 percent of scale full power plus an additional .4 percent 25 to make up for heat loses, or unbalanced heat losses.
O
,,/
(L)
+
36 281 1 (Slide) 2 The next slide which, again, is there for 3 completeness -- and I don't particularly want to read it 4 unless there are comments about it -- describes the 5 operations of this facility starting with natural steady 6 state conditions and natural circulation. The test is 7 initiated by opening the discharge lead valve and then the 8 cold leg discharge leak valve.
9 And on a local level you trip on the HPI, initiate 10 core power decay, and begin the automatic control of the 11 reactor vessel vent valves, and start full auxiliary 12 feedwater flow to build the steam generator secondary from 13 the five foot level through the 31.6 foot level. ,
C- ) 14 (Slide) 15 The nominal test conditions specifications for 16 this test, the primary system pressure is adjusted to 17 obtain 12.2 K in subcooling in the hot legs. The steam 18 generator secondary pressure is maintained at just below 19 7 ?Ta, or certainly more than 1000 psi. Core power is 20 128.7 kilowatts. l 21 And then you see some conditions of levels and 22 fluid temperatures.
23 (Slide) 24 MR. EBERSOLE: Say, we're going to do a small 25 break experiment. You're not going to do less than
()
U
o 37 282
( 1 secondary feedwater first, which is going to be historically 2 the thing that's going to happen a hundred times more often.-
3 MR. KNIGHT: We are only in the process of doing 4 a total of five different pretest predictions for the 5 MIST facility. One of those tests is the nominal and that 6 is where we have concentrated the effort in order to help 7 B&W prescribe what that nominal test is. Then we are 8 going to do two feed and bleed calculations.
9 One is full HPI with a 20 minute delay availabil-10 ity to the HPI. The other is the so-called evaluation 11 model HPI flow with a norman availabilty in terms of time.
12 And that EM HPI test is approximately half the full HPI. So 13 we are in the process of looking at feed and bleed tests
- 14 which presuppose up front a loss of feedwater.
~
15 MR. EBERSOLE: Okay.
16 MR. KNIGHT: This is a .. "mponent schematic of our 17 MIST input. Again, I don't want to dwell on it at length, 18 but to say that there are 66 components, a total of 19 approximately 260 cells, I think. We have used the new l l
20 film components in a variety of those stations in order to l
21 get around problems that we had with D-components in l
22 coupling pipes together.
23 We have also subdivided components in order to --
24 in locations like here --
25 (Indicating) J
,O k /'
283 38
( 1 -- and in the hot legs in order to get a representation of 2 the uncompensated local heat loses that are associated with 3 some of the instruments, specifically the cool TC's, so 4 that we are representing this system as 80 at the outside l
i 5 battery, with the exception of the heat losses representing 6 that are not made up by the guard heating.
7 MR. MICHELSON: The number 6 at the top of the 1
8 upper head, what is that?
9 MR. KNIGHT: This is just a requirement within 10 our code that the upper head goes up to here and then we 11 have to tenninate that component with something, it has 12 to have an end. And that end in this case is a zero
'~
/N 13 velocity fill component. ,
U ~
14 MR. MICHELSON: Maybe Randy can answer.
15 MR. KNIGHT: The upper head here is defined as .
16 the vent valve area of the vessel because that is what he 17 shows the vent valves coming off of,'the upper head.
18 MR. CARTER: Right.
19 MR. MICHELSON: By his definition of the upper 1
20 head. Now, where is the volume which is above the upper i
21 head, which is the defined volume I talke/ lbout? '
22 MR. CARTER: It's above the vents.
23 MR. KNIGHT: It is in the model.. Excuse me, this 24 is a component schematic. I'm showing you just the
! 25 individual components, I'm not showing you individual cells (G.)
(
-- _ x.u -
..-..w .
-w m a . - a w: . -
= --.--. - -.a~. . . a 284 39
( 1 within the components.
2 MR. MICHELSON: But that is not treated as one 3 cell?
4 MR. KNIGHT: This is not one cell. There are 5 four cells here with,I think, there are two cells above 6 the vent valve.
7 MR. MICHELSON: I have you now. I thought it 8 was treated as one cell. Thank you.
9 MR. SCHROCK: Is the vent valve as far above the 10 hot leg as that schematic implies?
11 MR. KNIGHT: When you say as far above, the hot 12 leg is connected in here and it gets a little bit crowded
(~'; 13 so it is hard to show relative --
14 MR. SCHROCK: What I am after is how does the 15 code really look at that elevation in relationship?
16 MR. KNIGHT: This elevation from the hot leg 17 nozzle center line to the center line of the vent valve, 18 that elevation is preserved in the code.
19 MR. THEOFANUS: What was it, do you remember?
20 MR. KNIGHT: Standing.here before you I don't 21 know. It is whatever the MIST facility is.
22 (Slide) 23 Let me skip the next slide and then show you 24 very briefly just a couple of the precalculated results. It 25 turns out at this point that we have run this calculation, (7,i
%.J
40 285 (h I this basic calculation twice. We have run the nominal 2 test a number of times more than that in the process of 3 helping B&W clarify this test. The difference between the 4 current calculation and the previous calculation is that 5 we had made an input error relating to the operation of the 6 vent valves and the vent valve resistance.
7 And this plot is included in order to give you the 8 relationship to the old, which this is, this is the 9 previous calculation, and the current calculation. And you 10 can see the effects of changing the vent valve resistance 11 by a factor of 16.
12 And this plot also has annotated onto it several 13 heat points in time.
14 (Slide) 15 Keepi ng in mind that the previous plot is for the 16 old calculation for where the resistance was low, this is 17 for the current calculation where the vent valve resistance,
,18 vent valve operation, is correct. You will notice that there 19 is some difference here. Other than that the two plots ,
20 look quite similar. l l
21 There is a forward-shifting in time in this plot
~
12 as a direct result of changing the resistance. I have also i
23 shown an additional curve -- well, actually two curves here 24 -- which are the control for the steam generator secondaries.
25 And you see that the steam generator secondary depressurized q
b:;
l 41 286 !
( 1 below the control pressure for a period of time.
2 The broken loop depressurizes below the control 3 pressure for a very limited period of time, reflecting the 4 fact that you maintain flow of the broken loop for a longer 5 period. And the intact loop you get a very early interrup-6 tion of natural circulation flow and the secondary 7 depressurizes far below the control point in response to 8 the injection of highly subcooled auxiliary feedwater as 9 you go through a filling phase from the five foot level 10 to the 31 foot leve.
11 Once the fill for the secondary is complete then 12 that pressure remains fairly constant until it bubbles 13 back up.
ggg .
14 (Slide) 15 This plot shows you mass flows in the reactor 16 vessel. The mass flow coming out of the core exit, the 17 mass flow is going into each of the two hot legs and then 18 the flow through the vent valve. In our model, because the 19 1-D nature of it, we have combined all of that vent valve 1
20 piping into a single piping line and a single valve. l l
21 What you see if you follow all of this is that l 22 after about 800 seconds the flow into both hot legs is 23 essentially zero. You do see some oscillation from one 24 hot leg to the other from this period of time. The vent' 25 valve flow is relieving almost all of the core exit flow.
,m.
C.
287 42
( 1 And you say, "Well, here you're got a case of the vent 2 valve flow being in excess of the core exit flow," and it 3 reflects the fact that you are getting some backflow of 4 liquid from the hot legs into the upper plenum.
5 MR. MICHELSON: Is the heat being removed if the 6 hot leg flows are zero?
7 MR. KNIGHT: The heat during this portion of the 8 transient is being removed through mixing of the two-phase 9 mixture calculated to flow through the vent valves with 10 cold HPI in the downcomer of the cold legs.
11 MR. MICHELSON: So it is just the capability of 12 the injection of water to perturb the energy for a particular r3 13 length of time?
{d j 14 MR. KNIGHT: We will get out beyond 3000 seconds 15 a boiler condensor mode transfer that will dump a great deal 16 of energy to the steam generator.
17 MR. MICHELSON: This is a cold leg break?
18 MR. KNIGHT: This is a cold leg discharge break.
19 MR. MICHELSON: Thank you.
20 MR. KNIGHT: The thing that you ought to be aware l 21 of is that the calculated void fraction in the vent valve 22 flow from this point on is approximately -- well, it is 23 in the range of 50 to 60 percent. If I go one cell above 24 I've got fully voided conditions, I've got no liquid up 25 there.
O.
( )
C' I
p 43 288 (h 1 And one cell below I see approximately the same 2 sorts of void fractions.
3 (Slide) 4 MR. THEOFANUS: That is pretty steady then?
5 MR. KNIGHT: Yes.
6 This plot shows you the calculated intact and 7 broken loop hot leg flows. He are looking at a location in 8 the hot legs. It is actually up very close to the top of the 9 U-bend. You see that the intact loop interrupts very 10 early, you see a couple of occasional spillovers of liquid 11 but not very much. It is a very rapid interruption after 12 the hot water from the pressurizer dumps in and you get 13 flashing that way.
14 The broken loop hot leg flow continues out to 15 approximately 800 seconds at which point it interrupts. The 16 difference between this calculation and the previous 17 calculation with the low vent valve resistance is that there 18 was an interruption, a transient interruption of the 19 broken loop hot leg flow at about 500 second.
20 So this flow would have come down and then as the 21 system heated up in response to the loss of heat sink the i 22 levels swell and you get a spillover and you get another 23 surge.
24 MR. THEOFANUS: What kind of results are you 25 getting from the upper downcomer in this period of tlne when n
m o
289 l 44
( ) 1 you said you had one fraction about 40 to 50 percent, is 2 that what you said?
3 MR. KNIGHT: In the upper plenum the flow going 4 through the vent valve, or entering the vent valve, has a 5 void fraction of 50 to 60 percent. That is a donor value 6 that we talked about before from the top.
7 MR. THEOFANUS: What is in the upper valve, in a the upper downcomer?
9 MR. KNIGHT: There is voiding over there also. At 10 this point I weuld be hard-pressed to tell you the number. I 11 can get that number, I just don't have it with me.
12 MR. THEOFANUS: I think that with the close 13 examination of these kinds of numbers you can apply the
(
-(J")
% 14 question of heat in this calculation. Because when you are 15 at these kind of void fractions you have a tremendous slip 16 between the vapor. It is a fitback thing here. If you 17 fit the wrong thing in it you are likely to get the wrong 18 thing out of it.
i 19 And I don't know if that is a stable or unstable i 20 feedback. I don't know if you spill it -- you won't be 21 condensing it, you don't have the energy, you are going to 22 be boiling more. And that is going to effect, now, your 23 void fractions in the upper plenum. It could be that it 24 has a stable behavior in the sense that indeed it is going 25 to bring you where you are going to void that region and you g
290 45
- n. I j ) have vapor coming in.
2 But it ought to be looked at from the realistic 3 flow regime and the physics for this flow.
4 MR. KNIGHT: Yes.
5 MR. WARD: Can you finish up in three or four more 6 minutes?
7 MR. KNIGHT: Yes, I will be even quicker than 8 that, I think.
9 (Slide) 10 This table gives you a sequence of events. The 11 ones that are a little bit more formal are the values from 12 the previous calculations. The numbers that have been typed 13 out to the side are the values from the current calculations.
(G> 14 I would like to caution you that these pipe numbers are 15 numbers that are basically being read off of plots, whereas 16 the bold numbers have a more inherent accuracy because we ,
17 were getting those numbers out of printouts.
18 But what you see is the effect of the increased 19 resistance in the vent valve and that is increased going to 20 a desired value for the vent valve resistance is to slightly 21 reduce the time to get to a particular point in the calcula-22 tion. That is, the calculation is being compressed forward 23 in time. The end times are just when we decided to termi-24 nate the calculation.
25 (slide) gh
47 291 A
() 1 The conclusions are still valid. The current 2 results are not signicantly different from previcus pretest 3 predictions. Basically what we have been doing for some 4 time is just changing the time of the events in the calcu-5 lation.
6 In the current case of the two calculations we 7 are doing things slightly forward. The MIST pretest l l
l 8 calculations is all phenomena that are expected to occur in l 9 a B&W PWR and as observed in the OTIS test. We see draining to of the pressurizer, intermittent loop circulation, pool and 11 and high auxiliary feedwater, boiler, condenser heat l
12 transfer modes, and a refilling of the primary system.
,_s 13 We have not completed the MIST calculation for CT) 14 the complete refilling of the primary but that is a 15 requirement in order to do the pump-bump pretest prediction 16 that we are supposed to do. The MIST calculations are 17 large asymmetric loop behavior. We see the large loop 18 continue to flow for a period of time and we see a fairly 19 large flow loop existing from one cold leg on that broken 20 loop to the other cold leg on that broken loop as a result 21 of the break in one of the legs there 22 And in our opinion the hot leg vent valves should 23 be opened after the steam generator primaries are refilled l 24 to aid in faster refilling of the primary system and earlier 25 completion of the test. Thank you.
(
z u w = = u a a d e =<--.s..a ;~.o .. . : :.-.-..~ . - . . . . . . - . - . . . . .
CERTIFICATE OF OFFICIAL REPORTER I
l This is to certify that the attached proceedings before the l
UNITED STATES NUCLEAR REGULATORY CO2GiISSION in the matter of: l NAME OF PROCEEDING: i ACRS ECCS SUBCOMMITTE5 MEETING 1
DOCKET NO.:
PLACE: - Palo Alto, California 24 January 1986
-DATE:
were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear gulatory Commission.
(Sigt).
(TYPED) JAMES W. HIGGINS Official Reporter Reporter's Affiliation l
I i
1 l
1 l
i l
, l l
- O
_a a w:nuw m ... a.- . --~- -- -r.- - --~~- - --- - - - -
gd 1 292 Sids 16 g-) 1 MR. KLINGENFUS: For those of you tha t I have (bi 2 not met, I am John Klingenfus. I work for B&W in Lynchburg, 3 Virginia, and I have been performing the Mist Pre-Test Pre-4 dictions, as well as doing some Otis Post-Test Presictions, 5 and today I would like to at least show you what we have been 6 doing.
7 There are six transients included in this 8 package of information that I have just given you. Obviously, 9 in the time available there is no way we can cover in a very 10 detailed or at least enough to get a basic understanding.
11 So what I thought I might do is go through and 12 explain to you the transients that we have performed and try f\
(}[ 13 to maybe cover two of them in a little more detail.
14 First of all, I think I will change slightly 15 the order wnich your slides are in.
16 (New slide.)
17 one of the Agenda Topics today was to describe 18 which or what the RELAP Code that we are using in B&W for 19 these predictions actually consisted of. Ihe Code itself is 20 the RELAP time onto Cycle 36.0 that was transmitted to B&W 21 fromPG&G.
22 We incorporated updates that they sent us to A
23 incorporate the Code up to Cycle 36.1.
24 Following that incorporation of those updates,
~
25 we also added a couple of Models in there, one of which we
- . .- a - w- - - a a .a- - =--.a=--.-_
- --- -
2 293 1 have already heard talked about earlier today, the high eleva-
-s 2 tion.
3 This model incorporates details which incorporat e 4 the wetting as a function of the inward auxiliary feedwater 5 flow rate, as well as the following film mass flux flowing 6 down the tubes. That also takes into consideration elevation 7 effects and what the actual wetting profile looks like.
8 That model also includes the heat transfer 9 upgrade which we also talked tout briefly earlier with the 10 following heat transfer correlation.
11 There's also a noncondensible gas heat transfer 12 degradation model that has been added to the Code. Tnis model O 13 we haven't used for anything in the pre-test predictions, so 14 I won't go into it, but that is included in this version of 15 the code that we are using.
16 There are two other changes that are in this 17 code. one was during the debug of Benchmark Test, we found 18 that there was a discontinuity between the subcool liquid and 19 forced natural liquid and forced natural circulation correla-20 tions. That discontinuity was removed and I believe that dis-21 continuity was with respect to all mass flux. So I think it 22 was just to make sure that there was a smooth transition 23 between the two correlations in there.
24 There was one other change that was made to the o 25 code, and this was with respect to the time-step control. We
._ m. = ~ . ~ . - . . -......:..--............ . - - - ._ a 294 9
1 were experiencing some problems with the Code artificially T'
2 cutting the time step way back during periods of stagnant flow.
3 with stagnant pressure response whre the system, the velocities 4 were very small and tending to oscillate about zero. l 5 What we found was that the Code was trying to 6 resolve the velocity flip-flops, as we call them, and was 7 cutting the time-step way back in order to try to resolve that 8 energy exchange there.
9 What we did, we put a check in there that basically 10 checked on the magnitude of the energy exchange and if it was 11 below, we continued with the same time step, and we found that 12 the results of that Model gave us much improved run time and 13 f {}p# did not affect the results at all.
14 MR.S'CHROCK:On your next to the last item, you 15 have removed a discontinuity that exists in nature. You have 16 ignored the mix convection region and so why does one need i
17 to remove the discontinuity that does in fact exist?
18 MR. KLINGENFUS: The discontinuity we experienced 19 there, and I must admit I'm not that aware of what actually 20 was removed or what actually took place there, because I was 21 not involved in it, But my understanding was that there was 22 a fairly large discontinuity in the heat transfer coefficient 23 at that time, that that caused problems in the Code solution n
'- '} 24
(' 25 techniques which were similar to what were talked about before, A lot of interpolations are used between correlations in order
. _ . , _______A_ - - - _ _ _ - -
. . . = . - . - - ~ ..--.:...-~-.n .-
- - - - -...a.------ J 4 295 1 to try to cmooth the phenomena and what takes place in there C~.2 2 so it does not cut the time-step back.
3 MR. CATTON: Well, here is an example of the 4 circumstance where mixed convection is what makes the transi-5 tion from one to the other for you, and you need to use a 6 particular parameter to decide how fast that transition is 7 taking placo. If you just sta3oth, that's not right.
8 MR. KLINGENFUS : I did not say it was right. I 9 am just crying to elaborate what was in this Code.
10 MR. CATTON: This is an amazing businesc. No- I 11 thing really matters very much.
12 MR. T IEN: I don't think you will every have
( /
{] ' 13 a solution without knowing the key parameters. YoO've got to 14 have that.
15 I do have another question.
16 You say you are not going to measure the non-17 condensible heat transfer degradation model. Cocid you mention 18 a little bit what you are adding into the Code? ,
19 MR. KLINGENFUS: I really don't want to talk ,
20 about that now. It is not in context of what we are trying to 21 describe today. I think our time is better spent talking 22 about what we have actually done within that Code and this 23 Project.
('x
- )
,l ' 24 MR. TIEN: Is that based on what B&W had pre-k_ ~
25 viously?
___.ma._._. ;
5 296 i
,_ 1 MR. KLINGENFUS: Yes.
4 4
(27 2 MR. TIEN: That was based on a correlation 3 30 or 40 years ago.
4 MR. KLi'7GENFUS: I'm not familiar with that 5 model. Like I say, we have not used it.
6 MR. CATTON: My recollection of the discussion 7 of this that took place some time ago, is that this is wrong, 8 too.
9 MR.SCHROCK:One last question:
10 Is 36.1 the frozen version?
11 MR. JOHNSON: Yes, it is. Actually, 36.0 is the 12 first one, is the frozen one. We can take some liberties with 13 it. And then .1, .2, .3 represent error corrections.
{
14 MR. SCHROCK: That is like freezing of an impure 15 substance.
16 MR. KLINGENFUS: The six Transients thera 17 included in your package today, there are three of them that 33 are Post-Test Predictions of Otis data.
19 The B&W Owners Group has sponsored two of these 20 Presictions, using the Otis data in order to try to build 21 models that we would use for MIST and Plant Input checks for 22 our calculations that we would be performing in the future.
13 The B&W Owners Group funded one Otis Test 2202 AA ,
24 This Test is a test where we have 15 square centimeter, scaled C 25 15 square centimeter cold leg suction leak, and in this case
._-. __ - . . - . . . --- . . - - -- -- ==_u m .
297 6
1 it was a perturbation in which they were isolating the
( 2 pressurizer in order to try to isolate the pressurizer effects ,
3 This test came in combination wit.h a test that (
4 was identical to it, only it did not isolate the pressurizer.
5 This is the one we chose to predict all of the 6' 0:is transmissals.
7 Let me back up and say that it is: ftr those of 8 you whc are not that familiar with Otis. Otis is a single-9 loop facility, began a natural circulation somewhere in the 10 MIST with initialization of how it worked.
11 The Otis facility was a raised-loop plant, which g,2 is a question I heard asked before today, where MIST is a p
13 low-urge loop plant, but it has two loops, two by four.
14 The other prediction that the B&W Owners Group 15 sponsored was loop cool-down test 220899, and this test we 16 did not have a leak, f
17 The transient was initialized by starting the 18 core power to K RAMP and initiating HPI. Once the system 19 pressure rose to 2300 PSI, the FORV's at that point, the PORV' s 20 were manually opened and held open for the rest of the tran-21 sient.
22 There we am looking for upper head cooling as 23 well as overall loop cooling, in this case.
k_ 24 That test we used a switch from the auxilliary C 25 feedwater, we switched to low feedwater to try to minimize the
l 298 7 l 1 steam generator heat transfer. In that case we could more
) 2 or less isolate the feed and bleed effects.
3 MR. MICHELSON: In the case of Otis, can you 4 isolate the pressurizer?
5 MR. KLINGENFUS: Yes. We hve an isolation valve.
6 MR. MICHELSON: Is the valve right at the hot lec 7 itself?
8 MR. KLINGENFUS: Yes, I think it is within a foot 9 of that, very close to the hot leg.
10 Those two tests, like I said, were sponsored by 11 the B&W Owners Group.
12 There was a third Post-Test Prediction that was g 13 performed and that was funded by Toledo Edison, and that was 14 done as support for the RELAP-5 Calculations. That test was 15 230299. This is a very interesting test in that it covers 16 various aspects of heat removal from the reactor system.
17 The test is a scaled 10 squared centimeter cold 18 leg pump suction leak and it also diverted auxiliary fuel 19 water from high to low in order to minimize the steam genera-20 tor heat removal in this case. However, later on in f.he 21 transient, after the feed and bleed portion of the transient 22 was calculated, we went into an oscillator intervention mode 23 in which the operator initialized auxiliary feed water; high 24 auxiliary feed water again went into a high elevation botier-25 condenses. mode (BCM), as we describe it, and following that,
i 8
299 1 1 we g t into loop refill. This transient covers such a various
(] 2 array :of heat transfer that I think we may want to look at 3 this in a few minutes. ;
4 (New slide.)
5 The actual MIST transients that we have per-6 formed, we have done three of them. The first one was the 7 Nominal Case that Taad was just presenting that TRAC did.
8 The Nominal case we were using to try to get some information 9 as to test procedure writing. Also to make sure that the loop 10 was going to do what we expected it to do. That was the ini-11 tial test that we ran. Also we ran two other tests. These were 12 Pre-Test Predictions that we wanted to perform, but also we
/-
{ ,)) 13 chose two that would be rather limiting so far as the facility 14 was concerned. One of them was the 50 squared centimeter 15 break, and the other one was a large steam generator tube 16 rupture.
, 17 The nominal case, like that described, there
]g was a 10 squared centimeter cold leg pump discharge break, 19 and that included full HPI flow. It had a period where yea 20 have intermittent circulation and then a loss of circulation, 21 end you get into finally, during that period where you have 22 lost circulation, you have heat leak, HPI cooling, and then 23 you have a pool BCM, where the primary levels drop down to
/ \
2j 24 below the secondary level and exposed that condensation 25 surface.
e
9 300 n
1 That test, I think we will cover today.
(- 2 But all of the plots and description of the 3 test are included in your handout for your review later.
4 The 50 squared centimeter cold leg pump discharge 5 leak that we performed, this one we used Evaluation Model 6 HPI which is equivalent to approximately one HPI, scale of 7 one HPI pump from the plant.
8 In this case we have again leak HPI cooling.
9 We got into a high elevation of DCM because of the large 10 inventory loss, the steam generator, the primary side voided 11 out before the auxiliary feed water filled the tubes on the 12 secondary side to the 95 percent range.
r's .
{) 13 At the time that the steam. on the primary side 14 got down within the tube region, at that point in time we 15
, experience,d a high elevation BCM and a rapid depressurization 16 of the loop.
17 Also, later on in time, as the level continued '
18 to drop after the auxiliary feed water had terminated once 19 it filled the secondary level of 95 percent, we got into the <
20 pool boiler condenser mode, where the levels, primary and 21 secondary levels overlap.
22 The third test that is included from MIST 23 a Pre-Test Prediction, was a scaled double-ended 10-tube 24 rupture low in the steam generator. The leak organs, or the
(()
25 way the leak is set up in MIST ic it is a separate leak
~-
10 301 1 circuit that runs from the lower point of the steam generator G
h 2 into the secondary side tube region, so we have a separate 3 pipe that runs outside. It has the leak orifice in it.
4 In this case we have actually scaled the area 5 that tends to rupture, but because we don't have the effect 6 of the 50-feet of tube in the degradation of that flow down 7 that length, it really represents more like a 13 to 15-tube g double-ended rupture in the plant.
9 So it is a little bit larger than a scaled 10 10-tube rupture in the plant so far as the leak flow is 11 concerned.
12 That one I think might be of interest because
{ 13 we haven't seen anything related to otis MIST,anything similar 14 to that. So I think we will look at that, also.
15 (New slide.)
16 There was some discussion earlier, and I happen 17 to have a slide of the reactor vessel, upper, or reactor 18 vessel itself, including the upper head region. Some of the 19 models that we've used for MIST, otis, and the plant, we 20 have looked at in detail what the noding scheme is needed in 21 the reactor vessel up ahead following the separation, like 22 I said, in order to get the right phases flowing up that hot 23 leg and through the reactor vessel vent valve.
O 24 8esice11v, es vom heve three mesor f1ow geths C 25 coming from the core, you have small flow holes that go
11 302 1 directly into the hot leg. You also have large portholes,
( 2 flow holes, high up in the upper plenum cylinders that comes 3 down and out the hot leg.
4 MR. THEOFANUS: How small are they?
5 MR. KLINGENFUS: I think they are one-inch 6 diameter holes, and there are 40-some-odd of them. I'm not 7 s u re.
I MR. THEOFANUS: Forty in front of each hot leg, 9 or altogether?
10 MR.KLINGENFUS: In front of both hot legs.
11 MR. SCHROCK: What is the relative flow resis-12 tance for the two paths?
13 MR. KLINGENFUS: That is what I started to give
{>
14 you. The flow comes up the core and out here, it's about 15 twenty percent forced flow. Up through here is about 70 16 percent through the large holes, and then there is about ten 17 percent that goes up into the upper bead, and then comes down gg through holes that are located in the upper head region up
.. 19 here, down into the outlet and out the hot leg.
20 MR. SCHROCK: That is the single phase? -
21 MR. KLINGENFUS: Forced flow. Yes. With the 22 pumps on.
23 Those three patterns, flow paths, are modeled
_) 24 in MIST, and we do have the internal plenum assembly here b(A 25 that acts as an efficient separator for the flow that comes
. 1
12 303 1 up and out these holes.
2 Also, the vent valve is located here on that 3 outlying anulus region.
4 Other places where we have looked at and done 5 noding sensitivity .othsr then the upper vessel, we have looked 6 at the cold leg nozzle region, similar to TRAC, to try to come 7 up with what we feel is an optimum model there.
8 Also looked at the hot leg, the U-bend model.
9 So we have done noding sensitivities in these 10 areas, and we have used the same models from Otis to MIST to 11 the plant, and in one respect we have tried to freeze the 12 model as much as possible so we don't look at noding implica-j) 13 tions on each of the transients in trying to compare from one 14 facility to the next.
15 (New slide.)
16 I think now we will look very briefly at the 17 Otis Test 230299 and this is probably the third test included 18 in your Otis.
19 M.R THEOPANUS: You haven't told us what the 20 logical conclusion of this is.
21 MR. KLINGENFUS: We will see the noding scheme 22 here in a minute for what we have used.
23 The Otis Test, like I said, 220899 was funded
( 24 by Davis Besse and their RELAP bench-marking effort.
(1:} 25 A more detailed event description of this test 1
13 304 1 I kind of covered before. I don't think I need to go through l
(-)s 2 it again, but it is lo squared centimeter cold leg discharge l
3 -- excuse me -- not discharge -- suction break. !
4 And that was opened at time zero initiating 5 the vent. It was closed at 30 minutes. So at that point in 6 time we didn't have any leak. We waited for the system to 7 start to represssurize for a while and the PORV was opened 8 and we went into feed and bleed cooling. The head flow was 9 a scale, full low head, EPI flow similar to the actual Davis 10 Besse plant head flow.
11 Like I said, the PORV activation was included.
12 I think everything up there is self-explanatory 13 and the initializacion of how the auxiliary feed water then
{)
14 to low feed water in trying to minimize the steam generator 15 heat transfer that went on. And then later on we do switch 16 to operator intervention where we initiate the high auxiliary 17 feed water BCM. Then we did complete this calculation through 18 to loop refill and recovery of single phase natural circula-19 tion.
20 (New slide.)
21 The actual model shown here was what was used 22 for this benchmark. I don't know if you want me to go into 23 a whole lot of detail. You see noding diagram similar to
( 24 this before. Basically we have a region in the reactor vessel G) 25 up ahead that we found to be a very efficient phase separator
14 305 1 and also produces the phases that we expect. It maintains
{d.s / 2 liquid going out the hot leg until the time when you see that 3 level, MIST level drop down to the hot leg. And also allows 4 for steam to flow to the reactor vessel ventvalves as opposed 5 to liquid when there is the level dropping down below that 6 reactor vessel ventvalve.
7 So this region up here we found to give us a 8 good representation and we use the sane model in MIST and the 9 plant, also.
10 The hot leg U-bend medel we have also used in 11 both and the cold leg nozzle region with a double flow path 12 in here that gives us the mechanism in order to look at counter
() 13 liquid flow, even though RELAP is not the Code you want to use 14 for mixing calculation, it tends to give you a reasonble repre-15 sentation in that area.
16 MR. THEOFANUS: You went too fast for this very 17 important area.
18 How do you handle cross-loads, and how do you 19 handle mixing?
20 MR. KLINGENFUS: We use cross-flow junctions in 21 some of these areas up here. Cross-flow here and here (indi-22 cating) in order to try to get the phase separation up there.
23 MR. THEOFANUS: Can you be more specific?
24 MR. KLINGENFUS: So far as what? I guess I don't O([) 15 understand what you are looking for there, specifically.
__ . __. . ._ ..-._m 4 .. . _ . . . _ _ _ . . . _ . _ . . __
15 306 1 MR. THEOFANUS: What kind of flow do you let
( 2 go over the side? How do you, what is your prescription for 3 letting the flow go to the side?
4 MR. KLINGENFUS: We don't have any way of forcing 5 which flow goes out.
6 MR. THEOFANUS: So you are forcing the same 7
thing, the --
8 MR. KLINENFUS:
It is based on what the up-9 stream donor cell is, but this noding scheme tends to allow 10 liquid and steam from the core being generated in the core to 11 flow up. The steam preferentially tends to flow up into the 12 reactor vessel, upper. head.
O e !
7' 13 MR. THEOFANUS: Why? It seems like you're hav-14 ing the same thing going both ways.
15 MR. KLINGENFUS: Just because of the elevation 16 difference. In cross-flow you have equal pressures there.
17 The steam tends to want to flow up as opposed to across. In 18 th& model. The cross-flow model helps us in that respect.
19 Usually in that region the steam will prefer-20 entially flow up.
21 THEOFANUS: So then you are not using a donor, 22 you're not using what Taad was using before which was based 23 on the fraction in the cell, basically just the same thing to
,/
' 24 the side and up. You are not maintaining a separation scheme b- 25 there.
._.a . - ~ - s., . : ~_~_ .u- ~ . . : .- : - .. .. .:- .. . . a . . :--.a - - - - . - -- -
16 307 1 MR. KLINGENFUS: It is designed for separation, l 1
i f:~9
% 2 yes. )
1 3 MR. THEOFANUS: Is that the normal model that 4 is used in RELAP-5?
5 MR. KLINGENFUS: Yes. We have developed this 6 model.
7 MR. THEOPANUS: You develop?
8 MR. KLINGENFUS: Yes.
9 MR. THEOFANUS: So that you have a modification 10 of RELAP-5?
11 MR. KLINGENFUS: This is a user-input model, 12 using the normal RELAP junctions and components that are in
{}y#' 13 there. We have investigated noding schemes and looked at 14 how their phase separation occurs during the transient.
15 MR. THEOFANUS: That is what I was trying to 16 get to before. I did not know you,were going to address that .
17 But I wish you would give us some more detail of what kind 18 of model you were using.
19 MR. KLINGENFUS: That was not the specific 20 thrust of my presentation today. I mainly came to try to 21 enlighten you on what transients have been performed with 22 RELAP using what information was avalable from Otis and then 23 trying to translate what we learned there into MIST and A
V 24 also --
25 MR. THEOFANUS: But you see it does not do us l
C
. _ . _ . . - . _.. w . ..~ . _. -. ._ .w. . - -,.m um_a a -
l 17 j 308 1 much good if you come here and you show us all of this if you (O - 2 don't tell us how you generated the analysis. It is almost 3 meaningless. '
4 MR. KLINGENFUS: We have gone through it in con-5 siderable detail, not with you, unfortunately. I'm given an l
6 hour. It's very difficult to go through a lot of details like 7 that and also show you results.
8 MR. THEOPANUS: Have you documented it in any way?
9 MR. KLINGENFUS: It will be documented. This was 10 documented by Davis Besse. The two Otis transients that were 11 benchmarked by the B&W Owners Group I think is in the form of 12 documentation right now. MIST will be i,ncluded in the final
\
{ ,7) 13 report in the MIST Pre-Test Predictions.
14 MR. THEOFANUS: Have the documents been trans-15 mitted to the ACRS? Are they available?
16 MR. KLINGENFUS: I will have to check with the 17 Owners on that and see what we can do.
18 The initial conditions here, I won't go over 19 them in detail. The natural circulation pressure at 2200 20 PSI, pressurizer level, as shown there. We have both test 21 and RELAP there.
22 Our initial condition, we are real close to the 23 actual natural circulation hot leg temperatures that we see e~s (x/ )
24 there. RELAP at six ten points, seven with the actual test i
25 having between six ch eight and six ten.
_ _ _ _ _ _ _ _ . _ _ . ... _ . _.._._......_....m._.-_.-
18 309 1 (New slide.)
Q#
h7 2 The next slide is the sequence of events. I 3 will skip that and describe it as I go through the pressure 4 predictions.
5 (New slide.)
6
, This slide we have the data from the test in-7 cluded as the data slide here.
I MR. THEOPANUS: Excuse me. You said also that 9 you did some special modifications for getting the HBI mixing 10 with RELAP and I thought you said it does a good job. It is 11 not meant to do that, but it still does a good job. What do 12 you mean?
{]- 13 MR.KLINGENFUS: No. The thing that we use, we 14 use a double flow pattern technique at the cold leg nozzle.
15 That allows liquid, liquid counter-current flow there at the 16 nozzle. The hot fluid flowing through the reactor vessel vent 17 v lve tend to flow back up into the cold leg, mixing at the gg HPI site. While you get relatively colder fluid running down 39 through the lower junction at that cold leg nozzle down into 20 the lower downcomer. Granted in cells the size that we are i
21 using here, you are not looking at microscopic mixing charac-22 teristics, but it is something that gives you a reasonable 23 representation, especially when you get situations where you are A
[d 24 voiding in the upper downcomer. It allows you to get a pretty 25 decent representation of leakage, NPI cooling, especially when
-.. .. . .-..... .. .. -.. .. . . .. _ .. u - - . . . . a ...~ ..--...._: _ aa. ~
19 310 7s 1 the steam flows through the reactor vessel vent valve, progresses A
(J/ 2 l
up to the HPI site and is condensed there.
3 MR. THEOFANUS: I don't see an organization 4 diagram.here. I see a node and then I see two junctions.
5 MR. KLINGENFUS: Two junctions. That 's what I'm 6 talking about.
7 MR. THEOFANUS: Is the node, the node has a 8 uniform property site. If you put two parts to it, it doesn't' 9 do much to it.
10 MR.KLINGENFUS: But the Delta P across those 11 junctions is a function of the upstream and downstream cells.
12 One goes up, the other goes down.
(') *
{# 13 MR. THEOFANUS: And the two downstream cells 14 is one seventy-five and one eighty?
15 MR. KLINGENFUS: I think so.
16 MR. THEOFANUS: Those represent the two cells 17 in the downcomer?
18 MR. KLINGENFUS: Those two cells are split right 19 up the center line of the cold leg nozzle.
20 MR. THEOFANUS: The cold leg nozzle.
21 MR. ELINGENFUS: This is the cold leg nozzle 22 here at the center line, right between these two cells (indi-23 cating). Typically you see when the reactor vessel vent valve s 24 is open, flows through it into the up or downcomer, tierming 25 this upper region here, if it is liquid, or collecting steam
20 311 1 if it is two-phase, and then the warmer fluid tends to flow
( 2 back up into the cold fluid nozzle, mixing with the HPI at i i
I 3 this location and then you get a relatively mixed temperature 4 coming down into the lower downcomer. It doesn't give you a 5 real cold temperature here like you expect or like you would ha ve:
6 'if yo6'did not have a model similar to this.
7 It is more consistent with what we expect to g occur given the --
9 MR. THEOFANUS: You said it is fixed?
10 MR. KLINGENFUS: It is a noding sensitivity study 11 that we ran in order to come with the best model that we would 12 have in that area.
O (l 13 (New slide.)
14 The two pressures as shown here again the dash 15 is the data that follows the RELAP prediction. The secondary 16 pressure basically it was just controlled for the whole tran-17 sient. The data shows some slight perturbations here. It was 18 not controlled as tight as when we controlled RELAP. But as 19 I recall the first thing we did was to switch from high 20 auxiliary feed water to low auxiliary feed water, so we had 21 relatively little primary to cecondary heat transference, 22 especially when you lose circulation across the top of the 23 U-bend. So these perturbations really don't have much 24 impact right here.
' (I(:)
25 Ihe actual primary pressure now, you see,the
21 312 1 leak opens, sub-coolflow down, coming down to the system
( 2 saturates. System saturates and you get into the intermittent 3 natural circulation, and finally lose natural circulation here.
4 You probably note here that there is a big bump 5 in the actual data that occurred. Upon investigation of this 6 bump, it appears that something happened to the reactor vessel 7 vent valve at that time and it closed for some time period.
8 I really did not understand, given the Delta P 9
across the vent valve, why it would close and basically what 10 happened, when the vent valve closed, it tended to divert all 11 of the steam flow that was going out the reactor vessel vent 12 valve up the hot leg and it bolstered the level in the hot leg
{h 13 and gave us a pretty decent spill-over once it reached the 14 top here and you see the result in primary, secondary heat 15 transfer here, which we did not see in RELAP. But this was 16 just a perturbation that tended to, it lasted only for a short 17 period of time. The integrated effecb on the loop were very 18 small, however, because it is just a momentary spill-over and 19 washed the temperatures out and then back to where we belonged 20 after that time. -
21 When we came out of there RELAP was within about 22 10 to 20 PSI. The actual test predictions.
23 MR. THEOFANUS: You said intermittent natural
~
/T (j 24 circulation. You are showing here natural circulation inter-25 ruption. I want to make sure I understand you. Do you mean --
9
22 313
,m 1 MR. KLINGENFUS: That is the time that we lost k_ 2 all circulation over the U-bend.
3 MR. THEOPANUS: Intermittent means that you had 4 several interruptions.
5 MR. KLINGENFUS: There was intermittent, there 6 was interruption and then a little spill-over.
7 MR. THEOFANUS: I see only one spill-over there.
g MR. KLINGENFUS: That is a large spill-over, 9 like I said, related to when the vent valve -- There are small 10 spill-overs here that you can't resolve just because of the 11 pressure response. They did not resolve in a lot of primary, 12 secondary heat transfer because of the low feed water. This one
{
13 was large order of magnitude and that is the reason why you.
14 see that.
15 MR. EBERSOLE: What is this spill-over we are 16 talking about? Is it the top of the downcomer?
17 MR. KLINGENFUS: It is the U-bend itself. The 18 levels had dropped down the hot leg.
19 MR. EBERSOLE: I'm sorry. I meant the U-bend.
20 MR. KLINGENFUS: The levels had dropped down in 21 the hot leg.
22 MR. EBERSOLE: What would that differential 23 be equivalent to"that you had at that time between hot and j 24 cold?
25 MR. KLINGENFUS: We have the collapse level in
\
23 314 1 here. Basically what it was, it was the effect of the volume
' J.
2 of steam accumulating in the upper reactor vessel and steam 3 that was passing out into the hot leg as opposed to going 4
through and being condensed on the HPI. So we were collecting 5
steam volume down there which displaced the liquid in the 6
reactor vessel upper head up high enough to give us a spill-7 over in that U-bend.
8 That is what physically took place in the test.
9 MR. EBERSOLE: Well, at the U-bend proper, what 10 sort of level differential?
11 MR. KLINGENFUS: We'll look at it in a minute.
12 I don't remember what it was.
{}; 13 Now, continuing on here, following our completee 14 interruption of natural circulation, RELAP ended up about 15 20 or 25 PSI lower than the actual test prediction, which is 16 very reasonable. We were very pleased with it. At that point 17 in time you have to consider what this test was. It was a jg low head HPI and the portion of the curve that we are on, 19 9iven the pressure that we leveled off at, was a very steep 20 portion of that HPI curve. And to give you a feeling for how 21 steep it was, roughly a 10 PSI change resulted in the two 22 percent in HPI head flow. Because what happened, when we 23 star ted out slightly below pressure, we got more HPI flow, g,
() 24 which gave us more potential to condense the steam coming V
25 through the reactor vessel vent valve and that allowed us to l
j
24 315 1 depressurize the system a little bit faster than what the
( 2 actual test- data showed, and it kept compounding itself as 3 the pressure got lower yet and more HPI flow, until we get out 4 to the 30-minutes in time. At that point in time we isolate 5 the leak. If you start to repressurize the system, both 6 at data and RELAP until, I think it was two minutes later, two 7 or three minutes later, then the PORV was manually opened and 8 at that time we go into a feed and bleed cooling phase. You 9 see, the depressurization over the time period where you are 10 throwing steam out the PORV, the data was not quite as steep 11 s RELAP was and there is a couple of reasons for that. One 12 is the effect of us getting more HPI flow during this time fa
.) 13 period and we had a'little less steam space in the system, so 14 RELAP tended to phase a little more dynamically because of the 15 smaller steam volume located there. Once it went to liquid 16 only after PORV, at this time RELAP leveled off and we happened 17 to end up at the same pressure, following this feed and bfeed 18 cooling phase. At one hour in time,at that point in time, the 19 operator initiated high auxiliary feed water. He filled the 20 level to, he was supposed to fill it to ten feet on the secon-21 dary side. He really over-filled it ever so slightly,and that 22 is one of the reasons why you see his pressure drop that much 23 more. He maintained primary to secondary heat transfer a little
(~)
(_/ 24 bit longer.
C 25 We didn't follow actually wha t was done. We
25 316
,y 1 held it more consistently with what the actual test specs 5
CL'; / 2 called for. So ours stopped a little early.
3 Following that excellent primary-secondary heat 4 transfer in its high elevation BCM, the auxiliary feed water 5 terminates and we have now relatively cold temperatures up in 6 the U-bend resulting, and we're sitting there at TSAT from 7 our lower pressure both in the plant and at RELAP. And that 8 cold metal up there gave us the capacity to condense a lot of 9 steam in that trapped bubble up there and it allowed us, as we 10 were filling the loop up with HEI and without PORV or leak in 11 the system, allowed us to condense that steam, push the level 12 up, finally get a spill-over. Once we got a spill-over, that's 13
{ it. You condense all of the steam because the system is rea-14 sonably sub-cool elsewhere and you recover natural circulation, 15 bringing the pressure down.
16 Again here you see the time shift, the phase 17 shift here, and that is because the integrated HPI flow 18 differences.
19 MR. ETHERINGTON: What is the Heat Sink for con-20 densing the steam? Is it the cold steel?
l 21 MR. KLINGENFUS: In the U-bend region itself, 22 during this boiler condenser heat transfer, you are pulling 23 two-phase flow over the U-bend and this is a significant time r8 1
_) 24 period where that two-phase flow over the U-bend, you cool the !
l ( ()
25 metal in that region, get down to saturation at its lowest l
e a .x :x:.W.:.22::L.z.u:.......= ._.
. . u.. _ .~...a,- aA " a". . - . . .-
26 317
~
1 pressure here or real close to it, then as you repressurize,that.
2 metal is sub-cooled.
l 1
3 So it provides a nice Heat Sink to condense that 4 trapped bubble in the U-bend.
5 CHAIRMAN WARD: I will have to ask you to vind 6 it up in about nine or ten minutes.
7 MR. KLINGENFUS: I won't really talk about it, 8 the collapse levels are shown here. You were interested in 9 v.at the actual collapse level was, the difference was when 10 that spill-over occurred. It is about this time period right 11 here. It dropped down to, looks like about seven feet or so 12 below and then it swelled up and gave us the spill-over.
(];. 13 (New slide .-)
14 Also included in there are leak flow plots and 15 reactor vessel collapse level plots, and this I will show 16 just because of the interest in what the reactor vessel itself 17 was doing.
18 This is the collapse level, so we are really not i
4
.. 19 looking at frothing and mixture level. But RELAP did a pretty 20 good job within about a foot or so of predicting what the 21 actual collapse level was in the reactor vessel. So that told 22 us we were doing pretty good with that model that we had of 23 the reactor vessel upper head giving the phase separation.
i (1) Where is the top of the core?
i _
24 MR. CATTON:
( .;
25 MR. KLINGENFUS: I think it was minus six feet,
..-.. ... ~ .
~ . . - . . . . . . . - . . . .. . . -. . - _ _ .
i I
I 27 318 )
l 1 something like that.
( 2 So the level itself is way above the top of the l
3 core and the mixture level tended to reside between the reactor 4 vessel vent valve and the hot leg nozzle. So we never really 5 exposed the hot leg nozzle to pure steam flow in this case. The 6 model worked very well in tha't respect.
7 The conclusions basically I will just say, and 8 we won't look at them, the prediction was very good,especially 9 when you consider the delicate nature of this test. We could 10 really miss the phase separation in the reactor vessel upper 11 head. If we had put too much steam up the hot leg, you were 12 going to see a trapped steam bubble forming up there that is
/~
-) 13 going to tend to repressurize the system and not give you very 14 good predictions with the actual test.
15 MR. THEOFANUS: We have to -- Did you have to 16 play a. lap with the model?
17 MR. KLINGENFUS: No. This model, it has been 18 set up and we have used the detail upper head noding and it 19 worked fairly well.
l l
20 MR. THEOFANUS: I'm saying do you have to play 21 a lot with the model.
22 MR. KLINGENFUS: In order to get it originally l
23 when we were looking at the noding sensitivities, yes, it took O
i.) us some time in order to come up with a reasonable model that 24 25 gave us a nice phase separation. But this model has performed
28 319 1 well for us.
2 THEOFANUS: Why is it that the people don't know 3 about this? Or do you know about this?
4 MR. KNIGHT: Why don't we know about what?
5 MR. THEOFANUS: They seem to have solved this 6 problem as I was asking before, this problem of phase separa-7 tion. It looks like they are doing it satisfactory, if you 8 believe the data and your comparisons.
9 Are you familiar with the work that has been done ?
10 MR. KNIGHT: I would submit that our comparisons 11 on Otis look comparable to the comparisons that he shows.
12 MR. THEOFANUS: Aside from that, are you familiar
(- ,
13
(/ with this particular phase separation that they have?
14 MR. KNIGHT: Yes.
15 Are we talking the phase separation in upper 16 plenum?
17 MR. THEOFANUS: Yes. They are basically doing 18 dif ferent thing s . One goes this way and one goes that way 19 than what I thought you were doing.
20 MR. KNIGHT: I don't think that they are, Theo.
.. 21 They ha ve a numerical scheme, a donoring scheme that is very 22 much the same as ours. In our Code we can preferentially 23 trap the vapor up high because we don't bend at all. All I'm
("'8 k.J 24 talking about is that in a given time step what goes up is
(< 25 also what is going to go out to the side. Well, once the
- ;- . n >. .a a w m : . - - -.a . - - a .a a. m - -. . - - . ..~ - . . ~ ~ . . - >
29 320 p 1 liquid goes up, it can fall back down.
k 2 MR. THEOFANUS: I understand that, but apparently 3 that is not what they are doing.
4 MR. KNIGHT: That is what they're doing.
5 MR. THEOFANUS: Tney are phase separating.
6 MR. KLINGENFANUS: The cross junction tends to 7 provide you with a different mechanism just because it is 8 getting - .
9 MR. KNIGHT: They have a different scheme by 10 which they couple together the cells than what we do. Every-11 thing in ours is tied to the cell center and their calculation 12 is tied to a cell interface, and because of that you can 13 bracket your junction up or down, but that also creates prob-14 lems in terms of how you connect, for example, hot legs to 15 P l enums.
16 MR. THEOFANUS: Okay.
17 MR. KLINGENFUS: You can see we are running jg pretty short of time here. There are other plenums and I guess i 19 we really don't have time to go thrm. We have time to lock 20 at the steam generator, steam rupture briefly?
mido 17 21 CHAIRMAN WARD: Yes, why don't you do that.
22 Briefly.
23 MR. KLINGENFUS: Like I was telling you before O 24 there was a scale tended to double and to rupture that really lb
- 6. 25 came closer to being more like a 13 to 15 series, sometimas <
t j
. . . . .. . .:- . . . ~ . - . . . . .... w . . . .. a- . -. . . a.w - .. w - .-..
30 321 1 _
I rupture in the plant just because of the 53 feet degrading i
2 and actual flow that we had. This model includes the uncom- l 3 pensated low velocity heat losses we have in that. It also 4 includes the core augmentation to offset those heat losses.
5 We use full HPI capacity for this test. We had an automatic l 6 steam valve controller that is trying to control secondary 7 level, maintain it between ten and twenty feet on the secondary ,
8 side. We had no auxiliary feed water going to the steam 9 generator after the leak initiation which was at times zero.
I 10 We also had an automatic 100-degree Fahrenheit I l
11 for cool down for steam generator A.
12 (New slide.)
! )
The model used for this, and this is just one
(' 13 14 of the loops, again you see here the same upper head model -
15 that we used, the same U-bend model, same full leg nozzle.
Ib So we are preserving this model. The leak ac- ,
17 tually occurs in the B generator and it occurs froi the lower ;
18 plenum of the steam generator primary side and goes into the l
19 tabe region. '
20 ME. EBERSOLE: Well, you are feeding and bleeding I 21 the open atmosphere here, aren't you?
22 MR. KLINGENFUS: Open atmosphere?
23 MR. EBERSOLE: Right through the secondary.
r~x
\ s) 24 MR. KLINGENFUS: It's into the secondary itself.
h',
25 The primary to secondary leak.
e
n.~. a u... . . . .. ....--... - - : a :. - .:.-- = ---:-.- . . . . . .. ---- . a.:. . - - -
31 322 1
] MR. EBERSOLE: Well, it is going to go out the
( ,J i
2 secondcry safety PRV's.
3 MR. KLING'2NFUS : We have in this, we have steam 4 flow valves. Okay. And the steam flow, or we have a capacity 1
5 to control secondary pressure and we intend to control secon- i 1
6 dary pressure to get cool down rate because it goes thrcugh ,
7 to the condenser. We also have the capacity to medel what would g be a perfect by-pass-type valve.
9 MR. EBERSOLE: You're invoking the continued ,
10 presence of the condenser?
11 MR. KLINGENFUS: Yes.
12 MR. EBERSOLE: We don't do that, though, life 13 gets sticky.
14 MR. KLINGENFUS: This is MIST that we are simu-15 lating here.
16 MR. EBERSOLE: Then it has to go out to atmos-17 phere.
18 MR. KLINGENFUS: Yes.
19 MR.EBERSOLE: And do you open the secondary PRV 20 to let it go?
21 MR. KLINGENFUS: Again, like I said, this is MIST.
22 This is not the plant calculation.
l 23 (New slide.)
l 24 The initial conditions in the slide you will find C~ right there and for your information it is included back two 25
- .r gg .
~
. 323 c
1 transients before.
s_
() 2 There ic a summary of initial conditions for 3 Default Three and the MIST Test back there. They specified 4 initial high primary pressure twenty-one f'ifty, specified 5 second9ry pressere full power, the pressurizer level at five 6 feet, and the secondary level at five feet, something, sea watc r, 7 MR.EDERSOLE: One thing about this, you say this E is based on MIST, this is not,a plant calculation?
9 Mn. KLINGENFUS: Yes.
I9 MR. EBERSOLE: I'm trying to couple the plant 11 performance to the MIST performance. Would you do in fact --
12 Well, in you; experiment you simply invoke th'a presence of i
13 the cendenser. Right?
14 MR. KLINGENFUS: Right.
15 MR. EBERSOLS: Had you not done that, what -
16 would have been the MIST experiment?
17 MR. KLINGENFUS- Other than to isolate the i 18 secondary side. I'm not sure. I'm not that familiar with 19 the actual . . .
20 MR. EBERSOLE: Well, you know the typical D&W 21 problem is that it throws you off primary out into the world, 22 into the open world, for this transient you're talking about.
23 It would be a prodigio 2s flow rate.
24 MR. KLINGENFUG: That is something that is diffi-25 cult for me to answer, b- .uae I'm not familiar with the
33 c , 324 1 actual plant configuration.
f_\
5 C; d/ 2 MR. EBERSOLE: Well, it can jump straight through 3 the exits of the PRV's on the secondary side into the open 4 atmosphere, and that can lead to some other problems.
5 MR. BECKNER: This is one MIST test. And we
}
6' have one test that would isolate the steam generator and go out 7 on the safeties.
8 MR. EBERSOLE: All right. But the safeties are 9 set at 1100, or thereabout.
10 MR. KLINGENFUS: There are a sequence of events 11 as we go through here that you might want to go back and look 12 at later.
(m
('w.k- 13 (New slide.) .
14 This is a pre-test prediction so we have no data 15 ,to compare with. Initially sub-core blow-down phase, we end 16 up losing circulation in the inpact loop where the pressurizer 17 is located in the impact loop which tends to form a steam 18 bubble in the hot leg U-bend and cuts off circulation early 19 in that loop. We continue depressurizing via all the heat 20 transfers going on in the effective generator. The leak is 21 flowing in from the bottom. It is essentially saturating the 22 liquid, slightly voiding as it comes into the bottom. You've 23 got a significant heat transfer going on while you have circu-rs 24 lation there, and all of that primary to secondary flow is
( (_) 25 providing you a nice cooling mechanism for the primary side.
- r. ~ = . . . - . _ . . .- .. .. . - . . . . . - . ---..-.~-.-.a.-<a _
- -. m.
34 325 1 It is kind of interesting. Eventually we lose circulatian in x 2 the B-loop and the pressure flattens out and then starts to 3 increase ever so slightly. During that time period, the steam 4 accumulates even more so in the U-bends of that impact loop 5 and eventually we get to the point here about this time where 6 the steam on that loop, primary side, has gotten down into the 7 region.
8 Well, during this time we kind of leaked primary-9 to secondary, or secondary level has built up such that the 10 mixture level is basically to the top of the tubes. se now we 11 have got a steam generator tube rupture leak BCM. No auxiliary 12 feed water. The leak fluid from the primary is cooling
() steam as it comes into the top. The pressure dif-13 the primary
{'
14 ference between primary-secondary provides that mechanism, 15 saturated steam on both. Two-phase conditions then on the 16 secondary side. so we've got significant heat transfer going 17 on in the affected generator.
18 The other generator, however, is not doing 19 hardly anything at all, and in fact, out here at this point in 20 time, Generator A becomes a heat source, because its pressure 21 jumps above the primary, gradually climbs above the primary.
22 We have two-phase and intermittend natural cir-23 culation during this time phase in the affected loop. We 24 finally lose circulation here and we maintain our heat trans-C 25 fer via that leak BcM that is going on up there.
_ .-x=. . ., _ . . .
35 326 1 MR.MICHELSON: Since your generator has become
( 2 a heat source, are you starting to void the tubes on the 3 primary side in that generator?
4 MR. KLINGENFUS: The tubes are highly voided.
5 We have some collapse levels that we will be looking at.
6 Back here in this time period, it was the time 7 in which we actually filled up that secondary side, the mixture 8 height all the way to the top of the tube region in the affected 9 steam generator.
10 MR. MICHELSON: Actually you are filling the 11 steam line too and I guess it won't do that.
12 MR. KLINGENFUS: In this case it would. We don' t.
13 have the steam lines simulation as the plant does. We don't
(
14 have the anulus on the outside. The hot leg U-bend flow 15 you see when you lose circulation here and not here. This is 16 effect of some leaked BCM pulling two-phase flow over the 17 top of the U-bend.
18 The reactor vesscl collapse level, we see the 19 reactor vessel collapse level here on the top. The downcomer
~
20 level on the bottom; basically is what happens is that the 21 downcomer collapse level drops down into the cold leg nozzle 22 region and the reactor vessel collapse level drops down to 23 where the mixture level at this point in time is at the hot
>g
'/ 24 leg nozzle. So we expose the steam to the hot legs themselves.
~
25 The collapse level is at the top of the core but the core is e
36 327 '
1 adequately covered in cool, but the mixture level is still s' 2 above three feet over the top of it, or I guess four feet.
3 If you look at the hot leg collapse levels, 4 they are asymmetric in this case because of the one steam 5 generator acting as, not doing anything. This is the impact 6 loop, the good loop, steam generator. If you look at the 7 primary collapse level and secondary collapse level, the 3 secondary fills up to its 95 percent on the operating range, 9 and you hold it at a constant level. The primary level drops 10 down to the top of the tubes during this time period and you 11 have some boiler condenser mode going on in that steam genera-12 tor. One it reaches the 52 feet, which is the top of the e-l 13 tubes, after you fill the secondary side up, y'ou lose your
(}I) s 14 primary-secondary heat transfer for the most part. There is 15 very little pressure difference. The primary continues down, 16 and down and down. At about this point in time that hot leg 17 itself empties completely. The hot leg side. This is the 18 steam generator level, but that is the hot leg level right 19 there at that elevation.
20 This is the time at which that steam generator 21 also becomes a heat source.
22 The other hot leg, on the other hand --
23 MR. MICHELSON: It becomes a pressurizer for 24 the plant; doesn't it?
25 MR. KLINGENFUS: It indeed is a large steam
, ~
. . - _ ... ... n . ~ a .- . y .~ s .. .. . ~ - - .an ... - . . - . . - . ... .. .
37 328 1 space. The pressure is being controlled more by the large DJ 2 leak throwing a lot of volume out the system and the condensa-3 tion of steam via HPI and the natural steam generator heat 4 transfer is going on in the bad generator.
5 The affected loop, you see, the primary collapse 6 levels come down to the tube region here. You never see them 7 drop below the top of the tubes. And the reason why is because 8 of what's going on in the secondary side. The leak is initiatec 9 here at that time. The level starts up very quickly. When we 10 reach 20 feet that is the level at which we try to steam the 11 generator, control the level between ten and twenty feet. We 12 are unable to do it. We don't have enough steam relief capacity .
( )
13 we flll up- we eventually get to two-phase condition in the ku)/e f.
14 steam line. At that point in time the generator is completely 15 full. Not until we build the level up higher do we get the 16 right two-phase flow of the top we can handle the leak flow.
17 Eventually we level off our level we then con-3g tinue on out as the primary-secondary pressure differential 19 gets smaller. Our leak flow tends to decrease even though it 20 stays saturated liquid. So you maintain a large primary to 21 secondary loop flow the whole transient.
22 This is what is rather surprising, the asymmetry 23 that we see between the two loops here. And I will try to i7,)
.)- 24 superimpose these collapse levels both on the same scale.
kD 25 You see the collapse level in the affected loop i
i j
.w. - w . = = w -- .
= a- - - -
38 329 g- I way up here at 52. The other one way down here below 20.
x ,,
~.' 2 The significant Delta P in order to support that 3 level. There is a significant condensation going on in that 4
loop and in the Code that is the driving force for supporting 5 that level, that condensation.
6 I will say, however, that that asymmetry is more 7 significant than what we expected. But overall this was 8 of concern so far as maintaining the core covering pool with 9 the large inventory loss in the 10-tube rupture down low and 10 then take single phase conditions at the leak side.
11 But we did not have any trouble so far as the 12 core collapse levsL remained above the core at all times and (I it was adequately covered in cool.
('7) 13 14 So that is basically it. That is the queue 15 of the six transients we have here. The others are included 16 for you to look at later. If you have any questions, I will be 17 glad for you to gise me a call, whatever. I'll be glad to 18 try to explain them to you.
19 MR. EBERSOLE: What is the present state of the 20 emergency procedures now? I am thinking about the Indian Point 21 case where they deliberately avoided throwing this stuff into 22 the condenser and elected to throw it outdoors instead.
23 MR. KLINGENFUS: I am not the person to address O
k' 24 that.
25 CHAIRMAN WARD: Thank you very much, John.
1 I
l
. ~ . . . , . . . . . . . . _: ...__na_._._,.
39 330
(~T 1 Our final speaker is Bill Beckner.
)
2 MR. BECKNER: I am supposed to talk about how 3 we are going to coordinate the analysis efforts,and I will 4 briefly talk about that. And you are being handed out an 5 updated Test Matrix which I will consider in a moment.
6 Before I do that, I want to try to put in per-7 spective what we're trying to do with the analysis efforts 3 and this relates to the fact that we have had a 1ot of ques-9 tions today concerning the Codes and what we are doing with 10 them,and so forth.
11 We have been having difficulty addressing these 12 primarily because we did not have the time, nor at least (f , 13 planned to address some of these Code questions. And so what I 14 would recommend that if you do have continued Code questions, 15 why, invite us back and we will bring particular people like 16 Shotkin (phonetic), really people who are more familiar with 17 the Code development and the Code assessment efforts.
18 What we are trying to do in the analysis effort 19 here is really be a user of the Code. We are m; king use of the 20 Code to try to help us conduct the tests at this point in time ,
21 And that is why basically you see a lot of Pre-Test Predic-22 tions. And I will give you some rationale as far as why we 23 pick certain tests.
,' ~# 24 It is r.ot necessarily for Code assessment, at b It is to help us run the test to make 25 this point in time.
..,.. A . .__m._ . . . _ . ~ , _ _ ~ . _ . . _ _ _ -.-_. _ _ - . _ - m..___ . . . . . . . _ . _ . _ _ . . _ _ .
40 331 1 sure we don't get some surpriras and so forth.
(E[ 2 With that way of brief introduction, I will show 3 you, starting with the calculation side of analysis, anyway.
4 (Slide.)
5 This has been changed slightly so I have redone 6 it. I want to let you know where we are, first as far as the y first MIST test,that we have done pre-test calculations for 8 all three Codes, RELAP, TRAC and RETRAN and redoing the MIST 9 nominal test. RELAPS is complete. RETRAN is in progress, and 10 as Taad indicated, we are we are still doing some input stu-11 dies on the. nominal case of TRAC. I hope we'll get by with 12 only one more test for this calculation of this test.
()
( .; 13 The next series of tests here to corae are tests 14 that we plan to do with TRAD. Basically all of these are in 15 progress except for the pump bump as Stan indicated, he has
- s. 16 got to finish one of the tests before we can do that.
17 The next series are the RELAP5 calculations and 18 is the change that I want to highlignt.
I 19 John Klingenfus showed you this one and as you I
20 will remember from our previous charts I've shown you before, 21 where basically this one hac been deleted and this one has 22 been added.
l l 23 The reason for that, again, when we asked B&W 24 to go back and review the calculation plans, one of the guide-h 25 lines is whether we need to help plan the test, what are the i
)
. ~ -ma. . - . . . - - . - _ . . - - .. ..- . -. ..- . . - . - - . . - . . ~ . .
41 332 1 tests that are potentially, may be a problem in the conduct, ;
i ic 2 and basically that is why this decision was made here, to l 3 switch these two.
4 As it turned out, this was a relatively unevent-5 ful test, but again it was felt that this potentially might 6 have been more facility challenging than this test.
7 These two calculations down here were en the 8 Matrix before I got TBD, which means "to be determined" right 9 now. Basically what we have asked B&W to do is to quickly go 10 back and re-evaluate where they are as far as resources, as 11 far as calculations, so we can. make a decision as to whether 12 we want to use the remaining resources for either continuing 13 Pre-test calculations, or whether we want to save it for post-(
14 test calculations.
15 So this just provides an update of the MIST 16 Pre-Test Calculation Matrix which you have seen before. We 17 will have a similar Post-Test Calculation Matrix, but we will 18 not develop that until we start seeing what the data looks 19 like.
20 This is our primary means that we are using to 21 coordinate the Code calculations. This is coordinate as op-22 posed to what Tien was talking about when he said it would 23 interact. Our interaction effort has been primarily through (D
\~' 24 meetings such as this, and particularly when we met earlier (3 25 in the week with our PMG meetings that are held quarterly.
l
_ . _ . . . . _ _ _ _ _ . _..___.__.,_..m_ .m._ -
-_ m_ - . _ . _ . _ _ _ _ _ _ . _ . . . . . . _ _ . . . _ _
42 333 1 I tend to agree with Tien. We should have inter-2 action. We've had interaction. We've had problems in under-
.. 3 standing the problem with the different circulation flows.
4 But we will try to keep the Code developers 5 together, as best we can.
6 In addition to these calculational plans, we 7 do have some other calculational plans. Los Alamos is develop-8 ing a University of Maryland deck, and also an SOR II deck.
9 That's been on the back burner because we have tried to get all 10 of these calculations done before the test starts. But we 11 have those plans.
gs 12 In addition, the University of Maryland will be
! l using RELAP5. I understand they do have an IBM version of it
({ ' 13 14 but they don't have the latest. I think in March they will be 15 getting in effect RELAP5 MOD 2. They will be using under
.. 16 the Maryland program they do have a deck, I bel ieve.
17 The other Code analysis that we plan to do is 18 we do plan to, as I told you before, try to do some in-house 19 sensitivities studies, look at the feed water. We know from 20 plant data that it is about a ten percent wetting but we don't 21 know the precisic. on that number or the sensitivity on that 22 number. And so we will be trying to do some in-house calcula-23 tions. Richard has gotten a B&W model and the B&W version l
/'T l z i 24 of RELAP5, the one you see here, at Idaho, so they will be ;
l - '
25 using that.
. .. u .---..- ..-..- .. . .. u--. .: u..-..- = a.: ..-- : ~ - _ ~ .- --. .
43 334 1 MR. TIEN: Will the other IST facilities, do you
-i k/ 2 have some plans to do some pre-test calculations?
3 MR. BECKNER: Not pre-test.
4 MR. TIEN: That would be helpful for to even 5 see a very limited number of pre-test calculations for other 6 facilities.
7 MR. BECKNER: The problem with that suggestion 8 I guess is that what you are seeing, particularly with the 9 University of Maryland, I think they will have the same prob-10 lems with SOR, too, is that we are really investigating how 11 we could operate those facilities right now. ,
12 So therefore we don't have a firm test matrix.
n
[] ' 13 Because of that, pre-test predictions are really 14 impossible to do. It would be nice, but until we feel our 15 way through and see how these facilities should be operated, 16 we really can't do that.
17 The other part of the coordination and analysis 18 is the non-Code effort.'And let me say again one thing that 19 has become obvious as a result ofthese meetings at the ACRS ;
20 and PMG meetings is that we seem to ccc.e a closure point 21 or a point where we should be closing on the hot leg studies,
.. 22 we going to get the SRI 13 Y.Y. has a lot of work and ISNII has a lot of
(-)
_' J 24 work so as far as the coordination analysis, that is an area
(;
15 which I think we need to sit down and write a plan now to see
44 335
- 1 if what our individual pieces, and modify them and see where 2 they fit.
l 3 MR. HSU: To answer that, we do not do pre-test 4 calculations but we do do scoping calculations to determine 5 the parameters for ISNII.
6 MR. TIEN: I like that. But I was thinking if you 7 can do the Code calculation, then you have a real good compari-8 son. I think the scoping calculation is very important. I 9 like to see if SRI do something like that,also.
10 MR. BECKNER: That is basically my speel on 11 analysis coordination. I would like to maybe close with a 12 couple of comments on things that are rather important.
fh r(e) 13 First of all, Tommy Larson is trying to get this h
14 scaling report finished up. Whether or not he will get all his 13 comments on the 10th I'm not sure, but again I would like to 16 have, particularly Ivan, if you want to make some comments 17 by the 10th, let us know. And other than that, thank you, and 18 I think we've had some good suggestions here today.
19 CHAIRMAN WARD: Could you, I guess I still don't 20 understand the resolution of the concerns we expressed in a 21 previous meeting about the number of wedded tubes. I thought i
22 we were going to hear something from Thad Knight today about 23 sensitivity studies which --
O(- 24 MR. BECKNER: Are you talking about the plant or C 25 the test facility?
45 3 36 f s- 1. -
CHAIRMAN WARD: All of the test facilities are J
Q..j) 2 committed to an assumption about the number of tubes that are 3 Wedded by Aux-Feed Flow and we have never really heard a good 4 justification for that assumption.
5 There was going to be, I thought, a sensitivity 6 study to give us some handle on whether that assumption is 7 really important or not.
8 MR. BECKNER: Correct. That is what we are talk-9 ing about we are trying to do right now at this point in time.
10 You are correct. The plant data shows about, I 11 believe, ten percent or so of the tubes are wedded, but we 12 don't know how good that data is and how precise that number 13
( is.
14 CHAIRMAN WARD: That 's right . But you are getting, 15 I mean, that study has not been done yet. If the sensitivity 16 study says it is important, then what do you do? I mean, I
' SAIC results, experiments, figure into 17 guess, how do the 18 that?
19 MR. BECENER: Those have in effect been tuned 20 to a 10 percent number.
31 MR. ETHERINGTON: They get it at different levels ,
23 too, don't they, to work the same way?
23 MR. BECKNER: I'm not sure if we understand FO ex ctly how they are wedded. We know roughly from plant k/ 24 G 25 d ta that about a ten percent of them seem to be wedded near i
4
m~ . . ;.. . . . . _ .
46 337 1 the top, but that is a problem. And what we plan to do again
( 2 is to try to investigate what the sensitivity to that number 3 is.
4 CHAIRMAN WARD: That is coming kind of late in 5 the day.
6 MR. BECKNER: You're right, but basically it is 7 resources to do the job.
8 The other thing is that the facilities are in 9 effect fixed and if we find that it is not ten percent but it 10 is twenty percent, hopefully not thirty, but if it's not close 11 why we have a facility that will be different in that regard 12 from the plant.
(~') -
(]f' 13 CHAIRMAN WARD: That strikes me as d problem.
14 MR. BECKNER: It is a problem, I agree. And 15 again, it is a problem with resources.
16 MR. CATTON: You may be forced into separate effE ctg
.. 17 testing. And then you bridge the gap with the Code.
18 MR. BECKNER: Right. If we find that it is very 19 sensitive, that is some ten or eleven percent or something 20 like that, we may be forced into both experiments and Code 21 in calculation.
22 MR. CATTON: And the SAIC tests seem to indicate 23 that the fractional wedding changes as you pass that tube O'u 24
(' 25 sheet.
MR. BECKNER: That could very well be. And what
. - - - . . - . . . . . . , . .. +_ , .
47 338 1 we may have seen --
O (7 2 MR. CATTON: And that will lower the thermal )
3 center.
4 MR. BECKNER: And we may have seen in the plant 5 some effective things but we don't know how we them.
6 MR.EBERSOLE: Let me ask a question which refers 7 to the Davis Besse incident and the somewhat interesting I numerous transits that move very fast on the systems. Is thero i
9 a growing sort of realization that maybe we need bleed-feed 10 for these plants a little bit more than the great big wet 11 boilers?
12 MR. DECKNER: I think from my perspective, which tfsu 13 is resource and not NRR, is that NRR is become very interested 14 in finding out how we can calculate it and obviously they 15 are interested in bleed-feed. So from my perspective in 16 research where we have been trying to figure out how well we 17 can calculate bleed and feed and I think only then will NRR 18 know whether their concerns or their potential fixes are 19 valid.
20 MR. EBERSOLE: Well, you know, of course, we.
21 can't get it no matter how well you cancalculate it until you 22 resolve the fact that the PRV's are not qualified items to do 23 it with.
( 24 MR. BECKNER: That is another problem. ~
It's
- r. ';.
~
25 going to do it twice.
I
. . . . - . . . . . . - . . , . - - . - . . . .-.....-...u..--_.w - . - - a_
48 339
<- 1 MR. EBERSOLE: Right. -- -
"~
2 CHAIRMAN WARD: Any other questions'for Bill?
3 (No response.)
4 CHAIRMAN WARD: Th'ank you very much.
5 We will break now for lunch and come back in 6 an hour. We really just need the Committee and Consultants 7 back for the discussions that we want to have and we would 8 like to have the Reporter back for that, also.
9 so let us break for lunch.
I 10 (' Whereupon, at 12:55 p.m. the meeting was 11 adjourned for lunch, to reconvene at 1:55 p.m. of the 12 same day.)
(]J 13 14 15 16 17 18 19 20 21 22 23 O' 24
(.'
25
.--.. a _ - . . . - . .. -- .-
.:.---....~.: . . :: - .. . . . . . _
49 340
, 1 AFTERNOON SESSION 2 1:55 p.m.
3 CHAIRMAN WARD: Let us reconvene.
4 What we want to talk about now was indicated on 5 the Agenda. Two things: The Review of Hydrodynamic Loads 6 Issues that was talked about some months ago. And the desire 7 on the part of some of the Sub-Committee Members and consul-8 tants to learn'about the details of that.
9 Then a discussion of~Overall Subcommittee 10 Activity.
11 MR. THEOFANUS: What do you mean by Hydrodynamic 12 Loads?
)
13 CHAIRMAN WARD: For BWR's. BWR-3 particularly,
( .}:'
14 wasn't it?
15 MR. BOENNERT: It was issued as BWR-3, the 16 Sub-Committee Meetings.
17 CHAIRMAN WARD: Let us take them for a moment in 18 reverse order and look at the Overall Sub-Committee Activity 19 which kind of deals with the philosophy, I guess, of what the 20 Sub-Committee ought to be trying to do, what the full Committee 21 should be trying to do and what the Agency and the National 22 Research Community ought to be trying to do in the area of 23 what we are interested in here, hydraulics related to power O 24 plantsi (P ~
25 I think we are in sort of a little bit of a 1
, . - , - - - - , - - . - - - - - - -- ---n, - - - -
.. .~ . . .
50 341 1 unique situation, perhaps we're at the threshhold of a new era, 2 and I think we may even have moved into that new era and 3 beyond the threshhold. Perhaps it is a less interesting era 4 for some of us than the previous one, but let me just remind 5 you of a few things. And again I'm probably kind of preaching 6 to the choir when I remind you of these. Many of you know 7 much more about it than I do.
g But just to help set a context for some discus-9 sion.
10 The NRC Research Prograin originally was just 11 completely dominated by thermal hydraulic issues and parti-12 cularly because of the ECCS hearing and development of the, 13 y u know, extreme regulations in the area of requirements for
, g4 plants to deal with Large Break LOCA, I mean, there grew up 15 research activity and a body of specialists and activities 16 that was probably in this area that was probably most sophis-17 ticated technical area that involved with the power plants.
18 And I think the ACRS in those days played a 19 major role in that activity and in particular there has 20 traditionally been, in supporting the ECCS a very strong com-21 plement of consultants to deal with really the complex and 22 highly sophisticated questions.
23 Well, for many years the NRC research budget O
O 24 and activities in NRR indirect.ly related to that continued to 25 grow and even back four years ago, five years ago, there would l
n
51 342 1 be, the research budget reached a peak, at least a dollar limit
( 2 of something like $230 million, and although certainly all of 3 that wasn't in thermal hydraulics research, most of the growth 4 to that high number had been in other areas, thermal hydraulic 5 research was still pretty prominent in there.
6 For the last two years, things have been changing ,
7 and changing a lot. This year, for a budget we're looking at 8 for '87, the total research budget is down to, I don't know, 9 something like, instead of 230, it's down to less than a hun-10 dred million, probably down to 95, it's probably going to be 11 less than that. And it is no lenger dominated by the thermal 12 hydraulics research. That is still a substantial part of it.
,o
<? 13 I think it is a total of 15 milli 6n, or something like that.
Q..y ,)
14 But the projections for the future are that it 15 is going to continue to decrease, I mean the total is going to 16 continue to decrease and possibly the thermal hydraulic frac-17 tion is going to continue to decrease.
18 Right now I think the biggest fraction of the 19 total research budget of the last couple of years has been in 20 the severe accident area, a Resource term, and other parts of 21 that.
Risk Assessment has been taking something pretty 22 big, about like thermal hydraulics. It's in the $15 million 23
(~') range, I think.
z 24
- ,~ s
' But for the future, what we see, you know, I 25
52 343 1 think it's obvious that there is going to be decreasing
( 2 Government-funded activity in this area. Unfortunately there 3 isn't any evidence that privately funded activity is going to 4 pick up the slack. But I think there is a couple of reasons.
5 Of course, one is the general thrust of the current Administra-6 tion and Congress to decrease Federal spending, and that's a 7 reality that is probably going to be with us for a long period 8 of time.
9 And of course, as taxpayers, most of us are happy 10 with that.
11 The other is, you know, decreasing interest in 12 nuclear power on the part of the Administration and the con-
,13 gress, but also a perception that a lot of safety problems in
{]flh 14 nuclear power have been solved, and '
particularly that the 15 several hundreds of millions of dollars that have been spent 16 on research and related activities in the thermal hydraulics 17 area has been usefully spent in that a lot of these questions 18 which are very fundamental to the thermal hydraulics area 19 fundamental to nuclear power plant safety have been answered.
20 Not perfectly, but well enough. That is the perception. And 21 I think it is those combinations of things that are luading to 22 the expectation that there is going to be less and less of this 23 money in the future, n
')./
24 Well, I think what the ACRS is faced with is 25 having to advise the commission on exactly what. give the
53 344 1
Commission its best technical safety advice on how research
( ll 2 resources should be used in the fact of this general trend.
3 I don' t think the ACRS and the NRC or anything can hope or 4 help to reverse the general trend.
5 S I think our task is to provide them with the 6 best, soundest technical safety advice we can. And I think 7 that is what we really look to this Committee for and parti-g cularly to our wise Consultants who have the ability to under-9 stand this really very complex and sophisticated technical 10 area, 11 MR. SCHROCK: Can we ask questions as you do this 12 CHAIRMAN WARD: Yes. I'm going to quit in just 13
} )( ) a moment. So what I would like to try to do in the future is 14 sort of orient our thinking about this area toward what really i 15 is most needed, what's possible, and we can't expect to fill 16 in all of the blanks.
b I think we cantexpect to reach perfec-17 tion in the state of understanding of these issues. But we 18 are concerned that there are some serious gaps that are left.
19 And I would like for us to try to concentrate on identifying 20 what those are and developing proposals for dealing with them.
21 That is really all I've got to say. .
r 22 MR. SCHROCK: I was just going to ask about this 23 impact on the thermal hydraulics, the thermal hydraulics ,
24 h
' research budgeting level of the decision.to go after analysis
) 'r^3 25 of severe core damage. Was that something that the Commission I
l i
1
345 1 pursued on a recommendation of ACRS, or did they, ACRS, go b 2 along with that decision as it was developed in research?
3 CHAIRMAN WARD: I don't know if I can answer 4 that. I don't think the ACRS --
5 MR. SCHROCK: I think that is a major mistake.
6 CHAIRMAN WARD: I don't think the ACRS furnished 7 the leadership in going in that direction.
8 MR. SCHROCK: But it wasn't on the urging from 9 ACRS?
10 CHAIRMAN WARD: We certainly went along with it.
11 MR.MICHELSON: We were certainly aware.
12 MR. EBERSOLE: I don't think we fought fully
( 13 against it. ,
14 MR. SCHROCK: I don't know how the other people 15 feel, but my feeling is that you are not really in a different 16 subject area for severe core damage. That is mainly a thermal 17 hydraulic's tool. It is fairly complicated by temperature 18 reactions and moving boundary problems and a lot of complica-19 tions that we haven't any more classical form of reactor safety 20 thermal hydraulics.
21 MR. EBERSOLE: I think that is a new Large LOCA.
22 MR. SCHROCK: It could compromise this program 23 in several ways. I keep coming back to what I think is more 24 than just a perception on my part. I think it is an honest to 25 God fact;that is that people who work in this as contractors
346 1 and I see people within the NRC who are guiding the programs
( )J 2 and so forth, have to follow the line which is developed at a 3 higher level then NRC, and thsy have to go along with 4 that. And there's very little feedback that tells the manage-5 ment, "No, the path isn't the best path." And you end up with 6 -- I don't know. I don't want to mention names, but I just 7 very recently have been told by one guy,- "You know, we 8 have spent a couple of hundred million bucks and God damn it 9 we can't say now that that problem isn't adequately understood 10 because we should have been able to do it with that amount of 11 money," and so forth.
12 Well, you know, there's merit to that argument.
v?) s 13 But that's not a technical judgment of where we y"')
14 are and where we ought to be and I really thhk we need to 15 have a stronger input in positions of this kind to people who 16 are able to make decisions on what directions are going to be.
17 I think the budget constaints have pushed all of the programs 18 into postures where they should not be.
19 This business that we discussed this morning 20 with Thad about the Codes, there's no better example than 21 that. It always comes back to do what you can do within that 22 budgetary constraint. It's not that the budgets have been all 23 that small, but still the right things are being avoided 24 because there is a pre-determined budget level which isn't 25 based on defense of a technical need.
as 56 347 1
Now, where does the technical need really come 2 inte play in determining --
g 3 CHAIRMAN WARD: But isn't that the boundary condi--
4 tion we have, though ? I mean, we can, you know,we, the ACRS, 5 can do, I think, if we did it e.ffectively, should have a large 6 influence on the allocations of the total research budget and 7 how it is spent, and I think we have had some influence on 8 that for better or worse in the past.
9 I don't think really there is a major influence 10 we can have on the total pot that is there, and I think we 11 almost have to, you know, we have some, we write letters to 12 Congress and we carry on --
- 13 MR. MICHELSON
- That is a given.
14 CHAIRMAN WARD: We're a gnat, I think, as far 15 as influence there. But where we can have an influence and 16 where we should and perhaps we haven't done as well as we 17 could, is accepting that as a boundary condition, okay, now 18 what should we be doing now? You guys suggested that maybe 19 too much money has been spent on severe acciden'c research.
20 MR. SCHROCK: I mean, there is a lot of fiction 21 in our analysis of reactor transients and accident sequences, 22 a lot of fiction.
23 MR. MICHELSON: Now, it depends on who you talk 24 to, of course, as to whether severe accdent work, and there 25 is another group in ACRS who is very much interested in the
iv 348 1 severe accident and doesn't want it attacked at all, in fact, (h-) (./
2 they would like more money. And you will find other constitu-3 encies with other problems. That is why we go through our 4 research bud 3 et every year and try to provide some advice.
5 MR. SCHROCK: I'm looking at it from a technical 6 point of -'.ew. What I'm saying --
7 MR. MICHELSON: Every group. looks at it tech-8 nically.
9 MR. SCHROCK: There is a tremendous amount of 10 fiction in these computer codes which then people want to refer 11 to as first principle. codes.
12 MR. CATTON: Particularly the severc accident
- 13 code.
- \ l
%d 14 MR. SCHROCK: I'm talking about the ones before 15 you get to severe accident. But then, once you get to the 16 severe accident problem, now there is just no pessibility of 17 carrying through the kind of first principle analysis that 18 people are trying to do. So what is the end result of it all 19 going to be. I think if you have your choice between doing 20 that and doing a better job on how to analyze present systems 21 and future systems, I think that is another thing that gets 22 forgotten is the companies are going to come up with new con-23 cepts and they will need to be very carefully analyzed before 24 the decision to issue the permits to build them and operate 25 them, instead of the way we have done in the current generatior ,.
=
- 349 1
1 So, given that situation, it seems to me that a l
( 2 very strong case can be made that you are much better off in 3 getting you analysis goals sharpened up for designing systems 4 and analyzing their off-design operation and so forth, that 5 will give you very strong assurance that you never get into 6 the unwanted severe core damage area.
7 That is far better as a philosophy than saying 8 that "No, we really can't over really assure. We have to 9 analyze the consequences of what will happen if we get there."
10 And the second analysis ought to be done on a 11 much lower level and a different kind of, well -- i 12 CHAIRMAN WARD: But the apostles of the
']t'OS 13 defense in depth say we have to do the latter, that we have to 14 provide a containment and a means to mitigate, just in case --
15 MR. SCHROCK: I'm a hundred percent in favor of 16 containment. I would like to that designed, well designed.
17 But they never work.
I'g CHAIRMAN WARD: That's right. There were not any; 19 criteria for which containment had been designed has got 20 nothing to cb with severe accident.
21 MR. SCHROCK: Yes. '
22 CHAIRMAN WARD: I think that's what all of the 23 activity about severe accident has been about is to try to 24 catch up to date,despite the f act that they haven't been CO 25 designed for this thing explicitly, to the thing they're
- gg mu
- 350 1 supposed to cope with. You know. What are we going to do about
(>^^ 2 it? It's like Jesse said, these accidents on the new large O
3 break LOCA,well,back about 15 years ago, or whenever,there were 4 claims that the ECCS system could cope with a large break LOCA 5 and that was it,and the Judge said," Prove it." And $700 6 million later there is kind of proof that we can cope with it.
7 MR.CATTON: I would say that $700 million 6 la ter that NRR does have the calculations.
9 MR. SCHROCK: Out of necessity.
10 CHAIRMAN WARD: I think certainly people are 11 more ccnfident that the ECCS system will cope with it than they 12 would be ten years ago. But that is where we are with severe
/ 'N . 13 accidents,I think. ,
V_( )
14 We say the containment can cope with it, the 15 severe accident, and the Judge hasn't said it yet, but somebody 16 else is saying," Prove it." And that is what the activity is .
17 , about.
18 MA.EBERSOLE: Yes, but how much of the research 19 should go into that versus what Bert says is the permit area, 20 and I think he's right. Not much,not a large percent. I will 21 arbitrarily say it ought to be something like' twenty-eight. -
22 MR. SCHROCK: Yes,I would hope.
23 MR. EBERSOLE: Twenty-eighty.
24 CHAIRMAN WARD: About twenty-five seventy five.
25 MR. EBERSOLE: I wouldn't struggle with that.
~- .
S 351 r
1 MR. CATTON: Eve.n seventy-thirty.
'h 2 CHAIRMAN WARD: And there we go.
3 MR. THEOFANUS: I think that the first thing 4 is to try to look at the limited resources and think how much 5 we can get from this point and that point, because I think 6 that puts it in a real context. I think the approach ought 7 to be to look into what needs to be done in a realistic way 8 and prioritize the issues in a focused way, in a pre-supportive 9 way,and find out the time to go into the needs. Often that 10 is the case because things that have to be done ha /o to be 11 done,and that can be just as equally involved with basic 12 thermal hydraulics within the design base and also off the
% 13 design basis.
14 Now, having said that,and having had the context 15 today which you provided to this meeting or this portion of 16 the meeting, I am thinking that it is a rather big horrendous 17 question what is trying to be discussed here and I see our-18 selves going through this soul-searching which is very 19 valuable to go through anyway. So I would like to suggest that 20 we develop an approach to it, a way to go at it. The approach 21 would be specific goalsand different people might think it is 22 appropriate for this or this, or it could be even take a step 23 before if as I understood you to say some other time c.y letter l
24 to you and I think Schrock's letter is going to be --
(N 25 CHAIRMAN WARD: Definitely. The thingr that you've w---___.....-._.._.. _.
~
352 1 said in meetings.
.Og 2 xR. rutorAsus, rhet.e right.
3 And I think in .other words maybe make a list 4 of things that some of us have seen, maybe not being done in 5 exactly the way we would like to see them, find out what is 6 the relationship of this equation or difficulty, and then see 7 if there are any ways that one can approach them in a reasonabl e 8 way.
9 But what I'm suggesting is to go ahead and 10 organize in this way. Otherwise, you may be having a lot of 11 comments back and forth and there will be nothing to point to.
12 There will be no conclusion at the end.
13 MR. EBERSoLE:
t] I have a terrible problem with the -
14 severe accident program because every so often as they come up, 15 but they come up,rnd I have expressed that you have failed in 16 your ECCS cooling mode and having failed you have permanently 17 failed and you will never reinitiate it and so they proceed gg from that point of arbitrary definition to come to the degrada-19 tien of the core,and never recognize that the prerogative of 20 the operator at any time he has the chance to get it he's going yg to pour water on it,some percent toward total degradation, but 22 n b dy knows where or how often.
23 And that to me is a total unreality not to recog-24 nizer that he has the potential for intercepting a progressive GG 25 severe core accident, maybe making a very large mistake, as he
l 353 1 intercepts. But I think it is generally acknowledged now that 2 if you are already into a severe LOCA, no matter, you are 3 always going to try to stop it., irrespective of what.
4 MR. THEOFANUS: Yes. Always you try to get water 5 into it.
6 MR. EBERSOLE: There's no other way. But that is 7 not even part of the analysis now.
8 MR. THEOFANUS: I know it is not, but it is 9 getting more and more recognized in the severe accident. Not 10 only before' the stuff gets out of containment, but you talked 11 about mitigating the cycle.
12 MR. EBERSOLE: Well, you still pour water on it.
13 CHAIRMAN WARD: I guess Theo's suggestion that we 14 e uld, I don't think we are in any position to do that today, 15 but we could have some sort of interactive meeting and develop 16 n agenda for safety research that this group of people thinks 17 should be followed.
18 MR. THEOFANUS: No,I'm not referring to that.
19 Maybe I did not make myself clear. This is only,I think, just 20 putting a list of items is only part of the outcome of what I 21 expect to see. I think for some of us and especially myself 22 i I feel a great deal of frustration to have seen this Committec 23 or Sub-committee not coming to grips with specific problems 24 and seeing them all the way through to resolution and see us 25 floundering. We ' ll t.9 '- about th is tomorrow. We ' ll talk
38 354 1 about this next day, depending on who is at the meeting,who
.p
" 2 is sitting up there. Always we're getting random comments
%)3 3 back and forth. We don't see things in a goal-oriented way 4 where there's an item there and you try to provide a specific 5 definitive answer and see the answer and that's it and close 6 the book. I'd like to think in terms of specific goals. Be-7 cause only- then one feels one 's time is well spent.
8 And we say a bunch of things here at meetings 9 and then we go and two years later we find things, some things 10 have been done, some have not been done. I don't say that 11 the Sub-committee should be strong, I mean everybody into 12 following their will. But I would like to think that somehow 13 among rational people any specific issue should have eventually (7J <
)
14 a specific resolution. And I don't see, even in cases where 15 in my opinion the resolution actually has been technically 16 attacked, and I can give you specific examples. Look at 17 pressure suppression damage. Look at PDS. Again, again, and 18 again. The moment the issue comes in, it is very fashionable, 19 everybody jumps into it, a lot of money being spent left and 20 right, as long as it is fashionable. At some point some people 21 get the idea that most of it has been done, most of the bases 22 have been covered. Suddenly they kind of turn off the faucets.
23 not only the monetary faucets, but also the intellectual ;
1
_ f, 24 faucets. And the intellectual faucets should not be stopped Us, 25 at that point. Because if you shut them of f at that point - --
l l
. __. .______l
'64 355 1 basically you have wasted most of your money that you put 2 there. And if we could learn one thing from the last boon-3 doggle, some people think it was a boondoggle, was that althouc h 4 you may be spending money in one area, what you learn can be 5 be very beneficial later on in another area.
6 And what I submit to you is that if it was not 7 for this millions of dollars that some people say look at 8 the money you have spent and we are in the middle of a TMI 9 and the transient fashionable after that,and we will not have 10 the faintest idea how it worked,and the fact of anything going 11 in these transient areas makes things even more difficult to 12 calculate, which I think still too few people appreciate the
{] 13 difficulties here.
14 But if we didn't have the background of the 15 large LocA's we wouldn't have the faintest idea how to start.
16 What I'm saying here is again the same thing, 17 is how can you spend the money, throw money at the problem 18 because you think that by money you're going to s61ve the 19 problem. At some point we have some kind of halfway idea.
20 Yes, that is what this is and everything is set up and if 21 thines are not brought to an orderly completion, to a document 22 to be closed, even to this date,for example, you can't find d cument that puts down LOCA in its full perspective and 23 24 puts down very clearly what has been done about it, what is 25 not known about it, what is going to be studied and what is
~~
356 g the outcome of that.
( 2 So you can't come to a meeting like this and 3 Taad stands up thereand I have a lot of respect for the work 4 these people have done because they actually pushed forward 5 the state of the art tremendously,and we say, well,the heat 6 transfer equation is wrong. It comes in our way. You can't 7 do much with it. On the other hand if that's part of the LOCA, 8
the Large LOCA, for example, this was to be put down, compila-9 tions were to be put down, subject in detail, it goes to reso-10 lution. At least a limitation should show tht they are using 11 the wrong heat coefficient and maybe the occasion as far as 12 Large LOCA should be clarified, or there's wrong input, or 7m 13 wrong correlation.
- w. . ), ,)
14 So later on'we take the same one and we are 15 going to try to use it in another situation,as Taad said, 16 this mixed condition is important. Well, for a Large LOCA you 17 don't care about this thing, but what you try to do, the 18 transit analysis, the whole answer is there.
19 So somebody ought to be able to catch it right 20 at the-very, very beginning to come to this computation and 21 a few hours after those calculations had been done, like we 22 have been today, here to a meeting like this, they are using 23 the wrong heat transfer coefficient. They ought to have
, ,e3 24 picked it up before they leave the actual study input. To do L) 25 a calculation for a small break operation.
65 66 .
357 1 So the whole thing gives me an impression that
(]lll 2 it is a very randomized, chaotic, the ACRS is basically the 3 only hope to provide come order in all of this.
4 MR. KLINGENFUS: It seems to me that the Commit-5 tee's recommendations on the budget should reflect the views 6 of this Sub-committee and its consultants.
7 Maybe we should look and see where our recommendt .-
8 tions were defective and whether they could have been modifiec ..
9 MR. THEOFANUS: Well, the best -- it is only 10 again a part of the picture. I think all the money they spend --
11 MR. KLINGENFUS: Well, it is money and how shoulc.
g2 that money best be used. What is our recommendation?
MR. TIEN: I think you have apoint, but I think
[~ 13 14 we have togobeyond that. I think we are still largely driven 15 by budgetary allocations, even at the meetings I think they 16 are structured in such a way to reveal particular projects 17 bs in progress because somehow I think we have to change that 18 or balance it a litt:2 better so that we input the budgetary 19 allocation. I'm not talking about the boundary conditions. I 20 am coming from the Congress and others. But internal budgetary 21 allocation or research priorities, say severe accident versus 22 other thermal hydraulic aspects.
23 I think we have not really spent much time in 24 that kind of decision so that we can really intelligently U_ f')
25 or wisely or convincingly influence the whole budgetary internal
67 358 1 allocation.
2 I think we need perhaps really two kinds of 3 sub-Committee meetings, one to review, maybe like a Matrix-4 type of, one is to review the specific projects, but then 5 we've got a situation always asking the same questions. Just 6 like you have the same problems, same questions repeated again, 7 I think we need another kind of meeting once in g' a while prepa re ourselves to really have an overall 1ook and 9 a broader look to have all different projects, say what are the 10 important problems. Some of them are common, just like a Code 11 changes and all of these are very common problems, but we 12 are always coming back to the same problem with all the dif-
(( 13 ferent meetings.
14 I think we need some kind of meeting to really 15 cover a more wider perspective and then we can perhaps discuss 16 in terms of the research priorities overall, on an overall 17 basis. In that way I think we can really influence. Without 18 that kind of discussion, I don't think we can ever influence 19 effectively the internal allocation of the budget.
20 We've got to prepare ourselves. Have something 21 Prepared and then we can really find the budgetary plan come 22 out.
23 MR. ETHERINGTON: I certainly don't think we 24 can just wait until there is a routine review, because I 25 don't think a separate letter to the Commission, apart from
68 359 1 any other phase of the budget would carry some weight,although q
k1/ 2 not as much .. . . a c . .. it would have done twenty years 3 ago.
4 MR. TIEN: We really did not have any discussion 5 about it.
6 MR. MICHELSON: But that type of discussion did 7 take place. It did not take place with you present, but it 8 did with the research, the , safety research program sub-commit-9 tee meetings, which I did attend and at which we heard all of 10 the same things you are saying, and the human factors were 11 saying about their work, fundamental qualifications, all the 12 way down the line. Everybody's work is always the most impor-(~/) 13 tant. !
l 14 MR. TIEN: I think you've got,I'm not talking 15 about that level. I'm talking about --
16 MR. MICHELSON: That is where the decisions really 17 had to be made. If you want to spend money on work, human jg factors is a major consideration but which we now have zero 19 Opportunity. You're not even concerned about people any more.
20 MR. TIEN: I'm talking about a substantive --
21 MR. MICHELSON: You are talking about substructure 22 and cooling the core and I'm worrying about the operators and 23 the men. For which we are spending nothing.
24 MR. THEOE' ANUS : This is why I don't want to put the
' ~
25 emphasis in money. I don't want to put the emphasis into
360 I slicing. I think we can do some of that. If I can say, becauto 2 I want to be understood,I see right here we all need an inte-i { })
3 gration and resolution,and I think a very good example of how 4 influential this sub-committee is in affecting how well 5
integration has done is the MIST Program.
6 If it was not for the strong standard between 7 all of us here with this MIST thing two years ago, this 8 present level of coordination,and many of us were very compli-9 mentary about would not have come about. It would have been 10 just like it was in New York. Because we pushed it, we went 11 to meetings with those guys, sat down, and we sat down at a 12 blackboard and put those things there by hand. Sometimes they "y g 13 had to be taken by the hand and carried through and up the
~m i
%)
14 hill to effect this kind of integration and resolution.
15 I think it is a very, very good example of how 16 influential we can be.
17 Just by asking the right questions and asking 18 them for the right kinds of goals to be met and say,okay, for 19 example, we want you to come tomorrow and give us a full asses ;-
20 ment of the Large Break LocA or a small part of it, maybe, 21 whatever it might be, and do only a very thorough description 22 of the history, of the results, of the data, the distillation 23 of it, the conclusion, a set of limitations in the analysis 24 and the importance of the limitations so when you push them 25 to that limit they will deliver. They will put 'it out in the
W, t
361 1 open here, it will be debated, and then all of us will be
- - 2 better for it. I think that is where I am seeing the main 3 difficulty in what we are doing. Because we are doing a piece-4 meal approach, a little bit here, a little bit there, and 5 we keep spinning our wheels and we never resolve any item in 6 a definitive way.
7 MR. TIEN: We have n 't really integrated.
8 MR. TIIEOFANUS : .Even if you asked me today about 9 how the pressure containments data, there is no place to go 10 there and find in a nice way, this is the data base, this is 11 the kinds of problems we are worring about. This is the 12 support for it. This is the analysis. These are the loads.
8
, 13 So if you look at the last thing -- I had a lot of problems --
14 I told the Committee never to listen -- with the staff report.
p 15 They don't even tell you --
16 MR. SCIIROCK; This is what we did in January 17 and we talked tout integration-centered concept and the rolo 18 of the NRC staff in doing the synthesis of the information 19 gleaned from research.
20 MR. TIIEOFANUS: We talked some of it. I'm talking 21 more generically now. Because if we take these guys, I don't 22 believe we're going to do it because they were supposed to 23 have been doing it all along. What I'm saying is that unless 24 integration also is attempted, at least to be forced in some c' 0 25 way by this Sub-Committoo, I don't see it is coming.
71 362 1 MR. SCHROCK: How are you going to force them?
( 2 MR. THEOFANUS: By asking them questions.
3 MR. SCHROCK: Who are you going to put the ques-4 tion to?
5 MR. THEOFANUS: You put it to both,the NRC 6 staff. They are supposed to get together. They are supposed 7 to coordinate and they come here and give you the particular 8 subject, sub-integration, or whatever it is that they are doing .
9 MR. SCHROCK: Won't they say that they have their 10 priorities? And they don't have time to do it.
11 MR. THEOFANUS: I think that's a point, but 12 there is never a time that is right. It never happens and I V
13 think we are all the worse for it.
14 MR. SCHROCK: Then' what I'm wondering is how the 15 Sub-Committee could force them to re-assign the priorities so 16 that they would do it.
17 MR.THEOFANUS: The prime e:: ample is MIST. If 18 the Sub-Committee had not pushed them the way they pushed 19 them, we would not have this result that we have today.
20 What this paper that you saw which is so revealing 21 in many, many fundamental ways,I mean in essential ways.
22 When he tells you he's got to increase the temperature by this, 23 so much to meet that, that is very important. That would not '
24 have been here if it wasn't for our pushing, persistent push- i 25 ing.
)
1 l- __ --
ns 3 363 1 That is what I am saying.
2 But for some reason we are able to focus on 3 this problem. I don't know how, but we managed to focus and 4 some how we did not manage to focus there.
5 S what I am saying, this Sub-Committee, in order 6 to be effective and in order for us to feel that we are not 7 w sting our time, I guess, we ought to try to maintain some g focus and discuss the bigger picture, the smaller picture, 9 and smaller picture, so you know how everything fits tcgether.
10 MR. MICHELSON: Well, my exposure to other areas 11 of research, this one is probably handled better than most in 12 this regard.
13 What you are saying is nice. It's a real logical
(_
14 way of doing the job. It's a tidy way to do it.
15 MR. THEOFANUS: Dut it can be done.
16 MR. MICHELSON: But by and large the programs that 17 can be done by the NRC have not been tied up before they go 18 off on something else,'and there is a lot of loose ends each 19 time. The argument you generally get is that they don't have 20 the people, the resources, and so forth, to sit down and 21 tidy them up.
22 Now, the question is how important is it to 23 spend your time to finish up a few things versus many things 24 in a semi-finished state.
25 MR. THEOFANUS: This is what I'm saying. I'm
73 364 1 saying that it is very important to finish these right for
(]!lh 2 two reasons. One is that the end, in the final analysis, is 3 the economical way, because you don't dissipate yourself by 4 going back and re-inventing the wheel. And I will give you an 5 example. A specific example. I think it is important for us 6
to speak in terms of specific examples.
7 PDS. PDS went through a rather complete analysis 8
and everybody thought that yes, we have it covered. And suddenly 9 Ivan says, "Well, we have come to the fifth difference in 10 risk because of this failing to understand the details of 11 stagnation.
12 Well, as a matter of fact, we have -- and that
. (' ll) 13 went on on the other side of the fence because I had been 14 working with those guys you know. And we identified this 15 difficulty about a year and a half ago. And yet that difficulty 16 has not been put in a real clear way in documents so that 17 everybody knows it.
18 That is where there is a difficulty and even in 19 the later communication with eher people somehow this escaped.
20 A year later Ivan comes along, he reviews it 21 and sure enough he says,"There is a big problem over there."
22 MR. CATTON: And it is still there.
23 MR. THEOPANUGt It is still there.
l; 24 MR. CATTON: No one is doing anything about it.
v 25 MR.MICHELSON: You're probably right.
~
365 1 MR. THEOFANUS: Let me finish.
)
2 Then what happens is, in the meanwhile we lost 3 all of the people that had actually done those analyses and 4 they have gone to other programs.
5 When I was screaming and I was screaming when in 6 the probject, I was telling them don't put those God-damn 7 reports out. Extend the thing by three months and then address 8 this problem, solve it. Because that is what, it's going to 9 compromise the whole calculation. It will compromise every-10 thing else.
gg They did not listen to it. They thought,well, we are doing it for several reasons. We can close it here.
33
,M We war supp s d to start a program with SAI and Ken Williams 13
- )
y to address that on the outside in a specific way. For a year 15 n w we are trying to see a way, a financial way of doing that, 16 because of difficulties it hasn't happened yet. It is going 17 to be happening in the next month.
18 So what you are seeing here is again because of 19 lack of coordination and lack of tidiness, there is a lot of 20 dissipation in the process. And they called an ACRS meeting, 21 Ivan spent his time setting it, gave a presentation, now this 22 research are supposed to give a written answer to Ivan's 23 problems and all of this participation should have been in the 24 first place. That's the important thing. Finish it. Close it ,
25 The Commission said it was resolved, once the
c .
75 -
366 ido 19 1 Commission says it is resolve, the money faucet just gets
( g) 2 turned off tight. Even if it's right in the middle of their 3 programs, if they just can it.
4 That's wrong.
5 MR. CATTON: I think the ACRS should take a 6 stronger role and say, " Wait a minute."
7 MR. EBERSOLE: I don't think the Commission should 8 turn it off. I don't think it should be turned off.
9 MR. CATTON: They didn't turn you off. They said 10 the icsuo is resolved and they turned off all of the rest of 11 the NRC.
12 CHAIRMAN WARD: The ACRS understood that well
(]}gg 13 enough and agreed that that was a mistake, you know. We could 14 have, we cortainly would have said something, but we didn't.
15 Either we did not understand it well enough, or it is not 16 true.
17 MR.MICHELSON: Well, there were not enough mem-18 bers to support it.
19 MR. CATTON: Well, the ACRS involvement, as least 20 t'o my recollection, the ACRS involvement in that particular 21 issue waned strongly after there was a program set in placo.
22 They said, "Here is what we are going to do," and that that 23 was sort of the and of the interest. At least it was the end 24 of the involvement of any of the people outside of the Com-25 mittoo itself.
m____
367 1 MR. TIEN: I fcol from th0t thic typa of Sub-I -
2 Committee is very common. So again this is primarily the reasor 3 this project, I still come back to that, but how to reverse 4 it somehow, how the balance of this, I think it requires dis-5 cussions of a lot of information preparation.
6 I still feel the meetings I attend nost are very 7 specific project-oriented and in particular not enough discus-8 sion.
9 Right now the discussion I find very, very good, 10 interesting. But not enough discussion of this type.
11 So that we cannot p,ermeate or influence maybe 12 even to the ACRS level. I think ACRS should also have more
/>- 13 discussion about the problem so it is not just a very quick 14 decision of verbal agreement of budget dnd so on.
15 That is what I am really talking.
16 I think we need a different level 'f kind c of 17 cross-discussion instead of just those very specific project-18 oriented discussions.
19 MR. THEOFANUS: There's another side to it.
20 I'm going to say another side. You know, I maintain that 21 tomorrow all of the people that have been involved with li-22 censing were to just disappear, and you put new technical 23 people in their positions, we would have an awful lot of new 24 questions that will cause a re-inventing of the wheel again, 25 because of this lack of thorough documentation.
1 l
l
368 g Becturn what is happening ie, as long as you
( )- 2 the particular person is involved, go over things and do a Lm 3 lot of work, to satisfy themselves that certain things are okay .
4 If that is not tidied up,as you say, and put down into a good 5 technical base there for other people to come in, then it is 6 found that other people are going to come in will either have t.o 7 ignore important things that have been overlooked or other-8 wise will bring in a whole new idea.
9 The vessel suppressing today was a very good 10 example. I want to be specific. When we became involved in 11 this in 1972,already there was a MARK 1 out there and a MARK 2 '
12 that had been approved and supposedly it had all been worked g3 up through the staff and everything had been blessed. And I
%)^
g4 was called in to get MARK.3 and there lots and lots of problemn 15 with it. And I went back and looked at some of the documents 16 that they had produced from Vallecitos and so on for the 17 MARK 1 and it was clear that they had not done their job as 18 far as documents.
19 And then _ started asking all of the' relevant 20 and irrelevant questions at the same time, because I found 21 if it was work, it was new to me, and you ask a lot of ques-22 tions. And of course you know the result of that. I think 23 if all of us were to disappear today and put new people in 24 our shoes, the whole thing is going to start all over again 25 because nothing has been documented in a satisfactory way.
I r . . . . . . . . . . . . . _ . _ . . _ . .
/
369 1 If you look at the whole ownarc Program you would
( ^
2 think documents in piles and who is going to go through it and 3 sort through everything else to find out what is going on in 4 the pressure containment. And then the staff put together that 5 little report supposed to have the loading definition. They 6 give no basis. They give no justification. No nothing. So
' 7 here you take this line and that is your load.
8 This is just like it took all of that money, all 9 of that human effort end went and threw it away.
10 CHAIRMAN WARD: What I hear you saying, Theo, is 11 that a major thrust of the NRC research effort should be to 12 document things so well that if we all ' died or they all died--
13 MR. THEOFANUS: Absolutely.
h 14 CHAIRMAN WARD: I can't take th'at very seriously.
15 I have trouble about that.
16 MR. THEOFANUS: We can argue about that.
17 CHAIRMAN WARD: Is there any human activity in 18 which that sort of documentation is done or really believed 19 to be useful?
20 MR. THEOFANUS: Nuclear power and safety I think 21 is not exactly the ordinary human activity.
22 CHAIRMAN WARD: That's true.
23 MR. THEOPANUS: I think that if you look at 24 technology and how it involved over the years, that is the only 25 way that youcan maintain quality assurance, because they depend m._ . - , . .. -
'5h
'I 370 I so much on people reviewing things, looking at accidents as
( gg 2 ill defined, you know, the big plan. What happened to you?
3 A million things can happen to it. You've got to approach in 4 an orderly way otherwise you will never obtain anywhere near 5 to being halfway comfortable and never achieve completeness.
6 While completenasa is never completed per se, 7 but if you follow this approach you are going to be closer 8 to it. If you follow an orderly approach, at least you will 9 have everything on the way.
10 No effort gets lost because every step brings 11 you one more stop closer.
12 That's all I am saying.
] 13 Now, given the f act that the documentation 14 part is the easy part and it is the part that is a resource.
15 You are talking about a few essentials in resource. It's a 16 crime not to do that and that is what burns me up.
17 some of this data, in many cases, specific 18 examples, these are gone, these are lost.
19 MR. CATTON: They're gone.
20 CHAIRMAN WARD: What?
21 MR. CATTON: The BMW blowdown heat transfer 22 program went off for a year and the quarterly report, semi-23 annual report, it is gone.
24 MR. THEOFANUS: What you have is a pile-up. What 25 happens then, you burn them,and then you never congot to them.
371 1 Then it's gone.
g 2 MR. CATTON: No synthesis was ever done.
3 MR. THEOFANUS: There's no reason for this, 4 really.
5 MR. CATTON: I agree with you, but I think beca':so 6 -- I agree to a point. I think you have to balance documenta-7 tien effort which is certainly needed.
8 MR. THEOFANUS: We have zero.
9 MR. TIEN: Then is a way of, a file to make an 10 offectivo documentation. Now I think we carry it to an 11 extreme, of course;I think that will be perhaps, it is not the 12 right direction. Eventually they will have to freeze the 13 whole, froze it for a while, otherwise you get into complete g
14 chaos. I think there are ways perhaps NRC can document many of 15 these technical areas in a bottor way, especially in a con-16 sistent fashion. Perhaps they should have an issue of volumes, 17 cortain major aspects of phenomena and have some kind of a 18 serios so that people can come back.
19 I always find that is a problem. The same phono-20 mona you can have very different versions from different 21 sourcos, and then it takes a tremendous amount of timo and 22 offort to got even, find different various sources.
23 I think there is a nood of that kind. But I 24 think that really we have to discuss how canwo make such a 25 documentation. In fact, not to go overboard.
D 372 1 MR. THEOFANUS: I think you have to worry about gg 2 going overboard when you have done enough of it.
3 MR. cATTON: There is a basic problem with all
' 4 of this. It's really not,the documentation is lacking primarily 5 because management leadership in research was lacking. You 6 tell a guy and you give him some money to work, go off and 7 werk on some nice problem. He does it. He presents his paper 8 to local conferences and so on and so forth, but he never 9 supplies the sponsor with the documentation that his sponsor 10 had asked for at the outset.
11 And the sponsor says,"Oh, well, what the hell."
12 And then it never happens. It is a management problem.
]y- 13 I think the path, the requirements that are 14 initially stated for these things are all right. It is not a 15 matter of how to do it. We know what they should have done.
16 They just did not do it. And further, sometimes when the 17 requests have gone out for the documentation to the National 18 labs, they have just been ignored. And I'm not sure what you 19 can do about that.
20 But it is clearly not a technical issue and it's 23 not a how to document is s ue .k It is a management issue within 22 RES.
23 MR. TIIEOFANUS: I think someone has to be pushed,
_ 24 A very good example is MIST. I said that I wanted this and
( IJ this done and that is what they did. They had to get up and 25
rus 373 1 present it. And that is why I say this committee can play a
!A 2 role in this. If anyone is criticized, if you find it unsatis-l\-ggg 3 factory, you say, "Come back in six months an do it again."
4 Somebody has to be pushed to go through the painful step of 5 doing that integration. It is not only integrating one con-6 tractor's work, but integrating several things and making sense 7 out of it.
8 MR. CATTON: This is exceptional,too, Theo, 9 because I think the management both of NRC and EPRI from thin 10 were better made programs in the past and the two managers were 11 receptive to the criticism.
12 Look at FLEC. That whole program, the same criti-
-'Y 3 13 cisms were leveled but ignored because of the person who
-' l 14 managed it.
15 MR. THEOFANUS: Well, there is a difference.
16 All of those other things who don't remember us putting our 17 foot down as we put it in this MIST.
13 MR.CATTON: That's true. -
19 MR. THEOFANUS: I think we've only done it with 20 FLEC but we have done it with MIST and that is the way that 21 we should play our role.I feel that we have not conceived 22 that type of thing with FLEK,for example, or heat cransfer.
23 Onand on and on. It just keeps on going.
, 24 MR. TIEN: Well, you can have a program, say the 25 documentation wassome better and that was one integration
374 1 effort, but again I think we know the important documentation, 2 but this happens in many other industries. Basically you have 3 the same problem in Defense. You have the same problem. Many 4 efforts actually fail. But after the put a tremendous amount 5 of money,I think it is very important to know exactly how you 6 are going to document. That is what I --
7 MR. CATTON: I don't see any difference beccuse 8 the party that supplies the document, if it isn't supplied, 9 they don't get paid. That's the big difference.
10 MR.TIEN: But still the knowledge is lost. Not 11 hardware.
12 ,
MR. THEOFANU5: Here you are dealing with, you 13 have a different bundle here than you have in the space indus-14 try.
15 CHAIRMAN WARD: Let me me if I can perceive and 16 suggest a course we might take. You'know, first of all, taa 17 Committee la in the business of giviaf advice and we are on a 18 line organization in research or anything that carries 19 it out and finishes, anything like that. We are in the busi-29 ness of advising and exhorting and suggesting and reacting 21 to something and that sort of thing.
22 MR. THEOFANUS: But we can be expected and to 23 expect something be done.
24 CHAIRMAN WARD: Those are powerful tools and 25 they have been used in the past and I think what we are
84 375 1 talking about is how best to use them.
( ll 2 Now, in this new era 50 million bucks to 3 spend a year. This is an important area. And it isn't all 4 research. I mean there are other activities and other staff 5 that are related and really have got to be considered like 6 this. But I think this, the Committee as a whole and this 7 Sub-Committee can be in a position to influence it.
8 In research we have tended to be a locked into 9 these annual reviews where we look at elements and sub-elementa 10 and all of the stuff like Tien is saying, and I think the 11 rest of the Committee, the Committee as a whole has come to 12 the conclusion that that does have some problems and we tend
( l '; 13 to get too bureaucratic in our responses and strictly reactive
.a ,
14 and I think what I am hearing is that in this area we ought to 15 take more of a leadership role and not just react.
16 Well, I think it goes back into this Sub-Com-17 mittee developing as I said earlier, Theo disagrees, but 18 developing an agenda for how we think out in the future 19 the resources available in this area should be used.
20 MR. THEOFANUS: I don't disagree with that.
21 CHAIRMAN WARD: I think --
22 MR. THEOFANUS: I said that is part of it. I'm 23 saying that is a bigger problem. What you are suggesting shou 1 3 24 be done but it is part of the problem.
G.,r^)
J 25 CHAIRMAN WARD: I see. Your concern is one
kJ 376-I class of thing that that agenda would deal with.
(]gg 2 MR. THEOFANUS: The other way around. That 3 genda is a subset of my --
4 MR. MICHELSON: He is trying to go beyond the 5 Committee's area of interest I gather. Is that right?
6 MR. CATTON: One of the big problems I have seen 7 in the past within this area is that NRR's needs have not been 3 satisfied by RES. This as just stated. I mean there has 9 been evern letters written on this and as a result at times --
10 CHAIRMAN WARD: At times.
11 MR. CATTON: Well, even now to a certain extent.
12 MR. MICHELSON: That's no longer true. There
(. ] ^; 13 are cases where research is supporting something the NRR would 14 like to do, like Phase 2 of MIST.
15 MR. CATTON: That could well be.
16 CHAIRM1.N WARD: That's what Ivan said.
17 MR. CATTON: It goes both ways. As a result of Ig this, when OMB or Congress or something makes inquiries and 19 says," Gee, what is RES doing?" he sayd, "How the hell do I 20 know?" What that leads to is a re'uction d in the budget, 21 because clearly if you are not supporting your customer, you 22 ought not be getting money.
23 I think we ought to look into that. Wo ought to 24 as a sub-committee in this area ought to make sure that RES gg 25 is mocting the needs of NRR, and if not, why.
86 377 I Brian Sharon has stepped back from support of 2 some of this because he is not getting what he needs out of 3 NRR -- I mean out of RES. And as a result the budget gets 4 cut because if you don't need semi-scale that is five or ten 5 million dollars.
6 I think we should look into that.
7 If RES cannot support the needs of NRR, then 8 RES should not exist.
9 MR. THEOFANUS: That's right.
10 MR. CA'1 TON : I don't think NRR, the people that 11 I know, are at all pleased with this severo accident part, 12 and as a result they basically are supporting research in 13 that area on their own. -
14 Well, if they have to do that, why do you nood 15 Rascarch. And this just, I think this is the reason that the 16 budget has boon cut by Congress so dramatically, and it gets 17 back to the managomont problem.
18 MR. GCIIROCK: There is some risk, though, in 19 laying it all on RES. It scams to mo that it is too extromo 20 and of ton NRR has not had a very clear view of what their 21 noods aro.
22 MR. CATTON: That's cortainly truo, but thoro 23 is no moro customor.
24 MR. SCl! ROCK: And ovon when they think they havo 25 they hevo not communicated it very well at othor timos.
V 378 1 MR. CATTON: Then RES should interprot-it. They 2 are the customer.
3 MR. SCHROCK: So I think that this Committoo is 4 looking at that issue as an area of responsibility for the 5 Sub-Committoo. It will have to be viewed in both ways.
6 How to make that interaction moro productive and have a common 7 interest than it has in the past, and that is a very legitimatc 8 comment.
9 MR. TIEN: This Committoo should be mora activo.
10 That's the thing.
11 CilAIRMAN WARD: I think what I'd like to try to 12 do la to havo, I mean our product will and up being,a product
't 13 of the full Committoo will be a letter of recommendation for
.u~.
14 -- this is what I propose:
15 That the Committoo develop a letter recommanding 16 the directions for uso of Agoney resources in thermal hydrau-17 lica, safoty area, over the next whatever years. And, of courno ,
18 that for the Committoo to writo such a lotter. That lottor has 19 to como frcm this Sub-Committoo.
20 So rather than havo Roscarch review mootings, 21 wo aro just roacting to what the staff han proposed beenuse 22 it in that time of the year. I guess what I'd liko to do in 23 propono that wo have a serion of two or throu, probably at gg 24 loant throo mootings whoro wo, an a Sub-Committoo, without a 25 lot of outsido pronontationn from peoplo, try to develop, I
88 379 1 call them agenda, just an agenda for identifying what we think
( g) 2 are the major safety problems in the area, and identify w ays 3 that they can be addressed and develop that sort of fairly 4 comprehensivo set of recommendations for the Commissioners. ,,
5 Does that scom to the sort of thing that is 6 needed? I mean --
7 MR. THEOFANUS: Not what I -- I think I have a 8 difference of how you think about this.
9 CHAIRMAN WARD: I'm having a hell of a time 10 understanding you, frankly. I really am.
11 MR. THEOFANUS: Maybo you ought to spend a little 12 more time with this kind of discussion.
13 CI! AIRMAN WARD: That is what I'm proposing to 14 do in the futuro. That is why I am saying what I'm saying.
15 MR. THEOFANUS: But yoi think that is not a l
You think that coming up with someone also 16 short deal.
17 now says and our problems are done. I think thatthis is 18 rather moro of'a question. Wo don't have any safety problems 19 right now. Let's face it. Is it a question of maintaining 20 some activity that is important to be maintainod? In it a 21 question of orderly closing somo activitios? Is it a question 22 of totally documenting somo of the things? Nobody is going 23 to como and say, "You'vo got to solvo this problem by tomorrow 24 othoruiuo -"
b 25 CHAIRMAN WARD: Thoso are examplos of the sort l
a 380 1 of thing.I don't know if I agree with you on all of them g 2 but one consideration is that even though there is no out-3 standing problem identi~fied right now, we think it is very, 4 very important to maintain the body of expertise in this area .
5 And this is how we suggest we do it. If that is what you think.
6 MR. THEOFANUS: Because you said safety issues 7 and I interpreted that to be in a narrow sense, a safety issue, 8 CHAIRMAN WARD: That's all we're here for. If 9 it isn't a safety issue I don't want to touch it with a ten-10 foot pole.
11 . MR. THEOFANUS: You mean developing a body of 12 documentation is a safety issue?
33 CHAIRMAN WARD: If thinking is important to 34 protect the safety of the --
15 MR. THEOFANUS: It may be semantics but it is 16 n t a safety issue.
17 A safety issue is something concrete and speci-18 fic.
19 CHAIRMAN WARD: I don't -- that's not what the 20 dictionary says.
21 I mean issue in a broad sense, not some specific 22 narrow body.
23 MR. ETHERINGTON: I think a safety issue is 24 one from forty-three and I think --
25 ME. EDERSOLE: In the case of that PDS problem,
90 381 1 I was disappointed and bothered by the fuzzy way that that
(]lll 2 w s, t the conclusion at the last maeting when Ivan came in.
3 He said it dribbled its way on into Paul's unilateral decision .
4 So what?
5 CHAIRMAN WARD: No, he did say petered out.
6 That's right.
7 MR. EBERSOLE: That is what bothers me. It was g not ended there. It has not ended.
9 MR. MICHELSON: There was no strone Committee 10 support to proceed in an orderly fashion,as I understand it.
gg CHAIRMAN WARD: As I understand,most of the 12 Committee members did not understand.
g"jl 13 MR. EBERSOLE: Well, we're not gdng to have v
14 them,anyway, or something to that effect.
15 CHAIRMAN WARD: One member volunteered and com-16 mitted to come back to the next meeting with that proposal 17 as to how the Committee should reconsider. And that is what 18 we are doing. I agree with you. It petered out. Nobody 19 knew what the hell it was.
20 MR.MICHELSON: Well, there is no small body of 21 support for it unless there's a convincing argument that can 22 he presented and I haven't heard a convincing argument, 23 although I have a great sympathy for it. But to be honest,
(~]
j 24 nobody has set it up and made a convincing case, 25 MR. EBERSOLE: There are two pieces of making
91 382 1 a convincing argument. There is a recipient who it may be (lll 2 impossible to convince of anything. And then there is the 3 convincer who may not make a good enough try.
4 CHAIRMAN WARD: Yes, there are two sides.
5 MR. EBERSOLE: More often than not, the one to 6 be convinced can alway say, "You did not work hard enough."
7 CHAIRMAN WARD: I don't know.
8 MR. EBERSOLE: That is the boss's prerogative,
, 9 "You did not convince me."
10 MR. TEDT: Sometimes that is highly true.
11 MR. CATTON: The result of all this is it's 12 partly my fault for the presentation?
-s e's y3
}j j MR. THEOFANUS: My understanding is that RES Id and NRR are putting together an answer in writing to some of 15 the questions that were raised, and this is supposed to.be 16 transmitted to the ACRS in the near future.
17 As far as I know, that is the outcome.
18 . So I'm awaiting the answer.
19 CHAIRMAN WARD: Let me say something frankly:
20 I would say that Committee members often think, 21 get the impression that when a question is raised, the Con-l 22 sultants' answer inevitably is: "Well, it is pretty good but 13 I think we need to do a couple of more tests, or a couple of l
fj 24 more calculations.'" And that is almost always --
~
25 MR. EBERSOLE: That is the easy way to go.
i 1
e
6 1 383 1 CHAIRMAN WARD: And that is not always very 2 helpful.
3 MR. EBERSOLE: No, it isn't.
4 MR. THEOFANUS: Well, you can't say that here.
5 Some of us here are 15 years experience in this 6
Co mmittee and I challenge you to go back at the trascripts.
7 CHAIRMAN WARD: On the other hand, the Consultants 8 might say, "I am being frank. " The Consultants may say, "Well, 9 I think the Committee members are so thick tney can't under-10 stand what I'm saying." And there's truth on both sides.
11 I'm saying that there is a problem.
12 MR. EBERSOLE: I always hear it when I hear it
("y-
%+
13 and I say in reply, "That is a coward 's view to say, 'I need 14, to have more.'" It's just an easy way out and maybe it is the 15 right way, but it might just be an easy way to postpone.
16 MR. THEOFANUS: Why do you say that?
17 MR. EBERSOLE: I hear that so much,"I must know 18 more. I must know more." It suggests, you know, there is never 19 any finality about anything.
20 MR. THEOFANUS: But you are saying that about 21 Consultants.
22 MR. EBERSOLE: No. I'm biking about ourselves 23 even.
24 CHAIRMAN WARD: Yes, we do the same thing. That's 25 what we tell the Commission. Our standard to the Commission n
0- m 384 1 is that, "It's pretty good but probably you ought to do 2 these thing s . "
3 MR. EBERSOLE: It's a comforting thing to say.
4 It avoids you being tripped, trapped. It is a mere threat 5 against it.
6 CHAIRMAN WARD: Everybody wants to get on the 7 road, I think, but I would suggest -- Really what I am pro-8 posing to do is that we have a series of meetings. I don't 9 think this will be a one-shot deal. Maybe this could be a 10 pattern for the future. But I tbink we have to, we have to 11 Put some discipline on ourselves that will come to some con-12 clusion and some agreement. .
13 one thing you Consultants don't see z.s often
.Cm(l J) 14 is how the Committee works on Saturday mornings. Saturday 15 mornings is where the rubber really meets the road. The road 16 for the Committee is that do it all. That is where we write 17 our letters and we argue and fuss and fume and people get mad.
18 I usually end up calling up my wife at about 10:30 at the 19 break on Saturday morning, and I call her up, and I say, 20 "Maud, am I pissed off'" That's a regular thing. I'm always 21 pissed off. But there is a real discipline there in that, you 22 know, a sort of collegial opinion comes out of that. Sometimes 23 it is so watered down it doesn't mean anything, but a lot of 24 times it is kind of valuable.
U,
r'~'s 25 I think if we are going to come up with useful
w 385 1 advice for the full Committee to develop --
( ll 2 MR. ETHERINGTON: Sometimes there are 14 addi-3 tional comments.
4 CHAIRbOW WARD: But not always. Usually there 5 are not. I think that is remarkable. But we need to go through 6 some sort of discipline process.
7 ,
So I'd like to suggest that we start that and 8 try to develop something and one thing I would suggest is 9 that if we could each furnish Paul within the next month or 10 so a list and the list might just be a list of nice things, 11 they could just be sentences or short paragraphs. I don't 12 mean you have to knock yourselves out with full -- But the 13 give things that you consider most important issues for which
.f...7(',)
n_
14 the NRC should be spending taxpayers money in the area of, 15 well, I think we are talking about the general area of thermal 16 hydraulics related to safety. And if we get those from every-17 body, maybe Paul and I can put them together and we will 18 structure kind of a couple of meetings around trying to talk 19 about that first, and pare it down to some kind of rational 20 Proposal, at least, to ultimately get to the Commissioners.
2} MR.IIEN: Do you view documentation as one of 22 the items?
23 CHAIRMAN WARD: Certainly. Yes. I mean, if you gm 24 think -- Probably if you are as specific as possible, it will j-.
s
)
a 25 help.
~
386 1 MR. EBERSOLE: From the beginning I am cer tain 2 that I would like this group to take a somewhat diverse and 3 different view of the severe accident LOCA.
4 MR. MICHELSON: I think that is a separate 5 subject. It is a separate Sub-Committee.
6 MR. EBERSOLE: It is not a thermal hydraulic 7 problem and to leve it to the thermal hydraulic group as their 8 sole endeavor is not right.
9 MR.MICHELSON: Where do you draw the lim ?
10 I can name lots of problems in other areas but, 11 you know . . .
12 CHAIRMAN WARD: Yes.
13 MR. THEOFANUS: Also another aspect of it that 14 I wrote in my letter and I want to bring up and it has to do 15 with the technical quality of meetings. That might have to do 16 with the way that the meetings are laid out.
17 Maybe we try to cover too much and go over too 18 little. But, for example, many times, like 70 percent of our 19 meetings are moderated,which you don't have to have somebody 20 tell you because it'is written already somewhere. And then 21 when we get into the real substance of things, then we get 22 short of time and things -- the people come in and they are 23 not prepared and after a while they get accustomed to us 24 not getting into depth. So they are not bringing any depth 1 25 with them, and that is another thing that I think would be
387 1 very influential in maintaining a certain level of standa rd, 2 a certain standard in the technical community. That is National 3 laboratories and inventors when they come in and they know 4 they have to stand up there and really stand on their own two 5 feet. That is helpful to them in an indirect way.
6 CHAIRMAN WARD: Yes.
7 MR. THEOFANUS: I don't know if there is anything 8 that can be done to help that, but certainly it should be 9 considered.
10 MR. CATTON: I think we ought to broaden our 11 view how we consider these issues. Many of these things 12 talked about should be included. I think we found ourselves 13 focusing more on two-hase equations than we ought to. I think
(
14 this committee ought to broaden. The thermal hydraulic 15 aspect might well --
16 CHAIRMAN WARD: As a matter of fact, we will work 17 on that, that specific thing. A more logical --
18 MR. TIEN: Maybe joint meetings sometimes and 19 discussion.
20 MR.MICHELSON: Blow down disability also is
.. 21 now a sub-committee.
22 I don't disagree. But you are saying there's a 23 whole bunch.
24 CHAIRMAN WARD: I guess the way we are trending ;
to break down the committees, one would be something that 25
m 388 1 might be called thermal hydraulic phenomena, which would be
( g 2 more ECCS than fluid mechanics. And the other would be 3 more equipment related. A committee related to the heat trans-4 port systems in the plant or something like that. Get into 5 more of the hardware, and I think that is probably that we are 6 going to tend to go.
7 MR. MICHELSON: There 's another whole discussion,
.. g Even your new listing is still how to get the KAYHEE 9 removed as a separate sub-committee. That wasn't clear to me 10 why that would exist as a separate sub-committee other than 11 tradition.
12 MR. EBERSOLE: NRC has done a horrible job 13
(;] trailing the B&W users and putting some constraints on that 14 work and defining the mode of operation is getting us into 15 trouble. Can't get it's interface with the thermal hydrau-16 lics. It is a tight relation to it.
?
17 Well, thank you. I would like CldAIRMAN WARD:
18 to get written reports from you fellows on the meeting and 19 in particular there is sort of a date for comments on Larson's 20 report. If you get the 10th of February to incorporate in 21 that, I'd appreciate it.
22 MR. BOENNERT: Next Sub-Committee meeting 23 on the Monday of a UHI Deletion. It's February 26 in Washing-
, ,,x 24 ton.
OJ 25 CHAIRMAN WARD: The hearing is closed.
98 389 1 1 (Whereupon, at 3:05 o' clock p.m., the hearing Cl x 2 in the above-entitled matter was closed.)
3 4
5 6 .,
7
~
8
.9 10 11 12
',: I3 14 15 16 17 18 19 20 21 22 23
!O O u
- 25 ,
i
CERTIFICATE OF OFFICIAL REPORTER i
(> )
i This is to certify that the attached proceedings before the UNITED STATES NUCLEAR REGULATORY COMMISSION in the matter of:
NAME OF PROCEEDING: ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON EMERGENCY CORE COOLING SYSTEMS DOCKET NO.:
PLACE: PALO ALTO, CALIFORNIA DATE: THURSDAY, JANUARY 23, 1986 ,
were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatcry Commission.
(sigt) 8/ 2 (TYPED [ [
JIM HIGGINS Official Reporter Ace-Federal Reporters Inc.
Reporter's Affiliation, O
,_ j\ y (fr*
O i
1 OBJECTIVES i
e 10 CNARACTERIZE TWO-MME R5 REGIMES IN A CAWT CAIE ET LEG GEtBETRY e Vertical Section e Entrance region (horizontal section)
O e i . er - ...en.
e TO Faw ROW CARRYOVER AS A FUNCTim W CE. LAP 5ED LIQUID LEVEL l
O i
l l
\. I
--_--.a - - +.a - --- a O I .
!I r 1 .
g*
l -
@ _$ .i @I
,' , <-]
=
! L3
! 1: -)~~
i t f!
( '
i
~!' !!!
' ,g : I
- -[d!::: ::
I O m -
I (I '
3
-! I
!- = I P I i,7 " '
i
=[
}
. l 11 q I
_a w , -
a I e O 2
. =
- , m 3
I .a U
l e 4 e
l w -
El g
h
\l O 0! -
[ I s
l - - _ _ - _ . - _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ . _ _ _ _ _ - . _ , . _ _ _ _ _ _ _ _ _ , _ _ _ _ _ _ _ _ _ . ___ ._. _--_.- _ ___.
O O O -
l l
l M4 VENT l
P __~- ,
TOP TANG IAP 3 70 Top y,g, 'Ch$'ng ,y3 PREE$uRE Prpg a
IM #3 TO Dreg
[ ^'
vis
, kyig v143
' Vf73 Vig V10 ASEAsop'g, 88ENT TANK Vf4 # TOTALilER FLOW "818 m (14g,g TO .-Vli M DRAIN DETAILS S & C (3: t' PlPeteG, C: 2* PtPtNG)
DETAIL A N
e V5 Ir REOu TOR y,
V8 IL SEE DETAIL' i vie
- - ~ ~ - - --
lvai _#- 'i3 WATER WATER TANK 2 Puesp A
pe PRQ5uRE N_ kER T LE COtePRESSOR TANK 1 V ENT U RIE_
RE R R SEE DETAIL h SEE DETAIL
~ W i
TO DRAIN a
f ):
hf' Figure 2-2 EPRI/SAIC Test facility (4-inch pipe) i
O o O -
10.0 Plant
' Bubbly / Slug Transition (high pressure steam-water
! 1.0 --
Experiment (air-water)
- Plar.t Experiment 7 g lhigh pressure g y (air-water) Churn / Slug to
- ( steam-water
~ Annular Transition T
u J
0.1 -
8 8 8 I ' lit! i i i i i n ii , , , ,,,,,l , , , , , , , ,
0.01 4 0.01 01 0.98 14.57 e u
jg (m/s) ll Figure 6-3 Comparison of Transition Lines for Plant vs. Experiment i
i
C O O AIR FLOW P. ATE IDENTIFICAT10f4 e c o a
' - y h /
e
~ ( /
/
, BUBBLY 7
6 A 6 A O M O [el a /
/
0 4 .ie. TEST 3E S
/
/
m y , a o A O N ON O p 0 6 4
^ "
y SLUG OR CHURN / g
-e, / -
/ y
>- - / u t -
/ w u -
/ -
, - / 5 s _, A a o o O s Os O /
/
0 $ O $
i d SLUG # w
/ <
do -
2 w CliURN/ SLUG o,
/
S BUBBLY / SLUG TO ANfiULAR / '
TRANSITION /
o
. .i _ TRANSITION / EXPERIMENTAL DATA LEGEND b
- (EXPERIt;ENT) (EX PE R If1Ef!T )
w
/ g
$ ~
~
o sueety o stuo h ,
O cnunn
~
0 ammutan
. sata seconoso
- T AITEL ANO DusLLER I e mi . . . . ...I . . . . ...I . . . .....I . . . . . i i . I e ei ei i is ios AIR FLOW SUPERFICIAL VELOCITY (m/s), j 9
Figure 7-5 Comparison of Experimentally Observed Flow Regimes in the Vertical Section of the 10.2cm Hot Leg with Taitel and Dukler (1) Analytical Results I l
= = = = - = * - ee .=-o %
-e g==- =
. , . . .a .
_ _ _ ___ - -__m_
Table 6-4 FLOW REGIME CHAR?.CTERIZATION AND CARRYOVER (12-inch pipe)
TENTATIVE TEST MATRIX Air Water Air Flow Water Flow Superficial Superficial Velocity Velocity Test (cfe) (m3/ s) (gpm) (1/s) (m/s) (ft/s) (=/s) (ft/s)
TEST 8A 4.5 2.lE-3 0 0 0.03 0.1 0 0 TEST 88 9- 4.2E-3 0.06 0.2 TEST 8C 18 8.5E-3 0.12 0.3d TEST 80 90 4.2E-2 U.60 1.90 TEST 8E 180 8.5E-2 1.16 3.82 TEST 9A 4.5 2.1E-3 45 2.84 0.03 0.1 0.04 0.13 TEST 98 9 4.3E-3 0.u6 0.2 TEST 9C 18 8.6E-3 0.12 0.38 TEST 90 90 4.3E-2 0.60 1.90 TEST 9E 180 8.6E-2 1.16 3.82 TEST 9F 1.8 8.6E-4 0.01 0.04 TEST 10A 4.5 2.lE-3 70 4.42 0.03 0.1 0.06 0.2 l
TEST 108 9 4. 3E-3 0.06 U.2 TEST 10C 18 8.6E-3 0.12 0.38 l TEST 100 90 4. 3E-2 0.60 1.90 TEST 10E 180 8.6E-2 1.16 3.82 ,
l l
4 O
. ' ' 1 !er. I
~
O s
/
m s 3
/ s s s 0 s m / / /
m /
m m 0 m 6 )
1 6 2 6 =
0
. 1 0 I m 4 0
. 1 7 1 0 O 9 8 0 .]
=
=: I ' a.. T1 T1 -
2
=
o j j
H
, r o o 8 H f
' /
0 h n
(
i o
T t H a G v I e l
E H E
_- ,.- v2 T1T 6 .
D s E
0 Z v l
l n o
O A M
R O
N i
t c
a r
F 4
' d 0 i E 0 C B o 9 9 A V 9 9 9 T 0 S 1 E -
T 7 Pc. Tf 2
0 e
r u
g i
F m - _- - - ~ - 4 0
1 1 0
0 , 0 1 0 0
.BpWeu_
O y
l;l ,li ll!l1! !! I1
1.0 Curve fit to 10.2 cm.
Experimental Data t
n, V
Zuber-Findlay (2_0) ,,,,,,,,,
j j= 0.23 m/s jj = 0.04 m/s
.8
- @ jt = 0.47 m/s .. - -
+j = g 0.23 m/s .....
- j = 0. 04 m/s ,,,,.. .. -
e ... _,,... - .....- ....... -
.6 - -
= -
O_ ,
e o -
2 -
u- . .
o .-
o
> .4 o
.2 -
jf = 0.47 m/s I
i l
0.0 . , , . .
0.0 .5 1.0 1.5 2.0 2.5 3.0 i
GAS SUPERFICIAL VELOCITY j (m/s)
O i Figure 7-8 Comparison of Average Void Fraction in the Vertical Section of the Hot leg with Zuber-Findlay (2g) Predictions 7-16 m
1.0 Curve fit to 30.5 cm
(] Experimental Data
(/
Zuber-Findlay (20) . . . . . . . . . .
ejj = 0.0 m/s
.8 _ + de = 0.04 m/s
- d = g 0.06 m/s j, = 0.04 m/s jg = 0.0 m/s t
..'l......,,,,,,,**....
.6 .
8 -
t ..-
o N
q *
... . - r: ..,.......... """""""*"
e o
4 -
, .... 9
. o
~~l
. ..~
2 ..
J, , 0.og m/s
.~.
0.0 ,' e
.4 .6 .8 1.0 1.2 0.0 .2 GAS SUPERFICIAL VELOCITY jg (m/s) b Figure 7-12 Comparison of Average Void Fraction in the Vertical Section of the Hot leg with Zuber-Findlay (20) Predictions i
t I
7-25 L .______ _ _ . _ _
f(
m c
)
0
- 2. AT =
1
.s/
0A m 1D 2
F)
Oh 9 m c 1 c En =
A Gi x 5T N - a s
. A A4 m, e 0D R( p 3 g i .
[
) j Fh x P _
Oc ,
m n N c Ei x G - a -
5 N2 A1 x m, .
R(
g 0
[ N i 1 /
j g
3 d
X 0 j n
( a Y m T c
\'\x
' I -
i\
s C 2 O
\ L E
0 1
V 1 iY, \ V r
V o H
G U Ei n
w3 s\' x L A
I C
I l
f e
O F v
\r\s' e R
H '
-Ois R E L T
R ld 1 P
, ,i'
',1 i 0 S i u
r.ib \
0 S q P A i G L
%0 iN 49
- -N'\ A L
N d
s e
p
-iN O I
S l a
N l E o
- ep e
p M C 4 _
'a l i p
i p
D I
f o
rl s m m 0 n
/
c c N o 2 5 s
- m i 0 0 0 1 r 1 3 0 a
. i 0 p 0 r r m D _.- o o 0 o E f f C TA . ,
CCE _ - m m EOG J 4 7 5 PLN _ -
1 8 1 XBA . -
ESR 3 2 7
= = e r
o g u g
H H i
F
- 1 0
0 0
- ~
- ~ - 0 0
0 6 2
- ~ 0 1
22~ d> 8aU aE%d8 SMNe@
u.E i
O O O i
INTEGRAL SYSTEMS TEST (IST) PROGRAM i
SCALING REPORT STATUS /RESULTS T. K. LARSON
/
b am ACRS ECCS Subcomittee Meeting j gg7/ggg Palo Alto, California eat 7AWFEXAtt7 l 44808A7087 January 23-211, 1986 I
i PRESENTATION OUTLINE e Report Objectives / Content / Status e Results Scaling Analysis Problem Areas Potential Methods of Data Comparison
" Counterpart" Experiments e Sumary/ Conc l u s i ons/Recommendat i ons i
9 O O
l
, l o O -
f o _
d e t c n u e s d o m s
i e r s t p e i s l a s -
i t a c a r f
a d o
d f l n a a g r n S g s i E e e t V t i s I n t e T I s i t -
C r l e
O E T "
i J S t c t B I e a r O m f a p
f a r g r T o n e R a p
O n i t n
P o t a u E i c R t i l o p y r e c i h t r "
r p e s c o m i n n s s o s o o a e o e y n d l g l t t e
/ i s a h g
h g m o
n h d n n o p n o a i i n i a i s s e t g t g n n h p a n n g t i a n i i
/
t i n l i i l e e g n
e a s m a d d m c e i c i i i u S D L S v v l c o o a o r r c D - - - - P P s o e e O -
, ;il ! :!ji! ;' i! I r! I l!!I l , l
!l l 4
9 Report Content e Program ond Facility Description IST Program Facility Description and Objectives e Scaling Methodology Scaling Methods Facility Scaling Methods Facility Limitations and Atypicalities Scaling Evaluation e Relation of IST Facility Results Implications of Operation at Nontypical Pressures Methodology for comparison e " Counterpart" Experiments Steady-State Transient e Conclusions O O O
- __ -___ _ __-_ -______ - __. . -- ._ - - . - _ - - .-. __ . _ - _ - _ = _ .
i O O O 1
REPORT STATUS l
t l e Current Draft Completed 12/86 Reviews Underway
. In-House (peer and tecnnical editing)
. EPRI
. NRC e Schedule !
- Incorporate Casunents 2/86 Publish 3/86 l
6 i
i l
r i
t
l t
l y
l Facility Description
- MIST
! > 2 x 4 Loop, Four Pumps, Two OTSG
> External Pipe Downcomer l > Four External RVVV I > Scaled Core l
- UMCP .
. > 2 x 4 Loop, No Pumps, Two OTSG
> Internal Annular Downcomer, Eight RVVV
> Core Not Scaled Gemetrically l
- SRI-2 I > 2 x 4 Loop, Four Pumps, Two OTSG s 1 > Internal Annular Downcomer, Four RVVV j > Core not Scaled Geometricaily ;
l l HAK00144
_l a...1 .i [,.li,b il..-.-
L i I h.. 3.1 -
L.
~
O O O RESULTS e Foc111ty Scaling Methods
'e Limitations and Atypicalities e Scaling Analysis / Evaluation e Problem Areas e Potential Comparison Methods e " Counterpart" Experiments
,. ,w- m --we - - w -m
Facility Scaling
- MIST
- > Modified Volume Scaling
> b = 1/817 = Va = Ng = ag
>O=ta g = 0'n' = Pg = 1
- UMCP .
> Reduced Height Scaling (Ishii)
> Q g = 1/4.4
>an = 1/110
>Pg*1
~
> Reduced height Scaling (Ishii) l > 9 a = 1/4 i
>an = 1/324
> Pa # 1 M AK00160
- 0. 7
O O O PLANT-MIST UMCP_ SRl-2
- Volume (M 3) 0.57 0.6 0.2 314 Volume Scale 1/817 1/500 1/1296 1 Length Scale 1 1/4.4 1/4 1 Area Scale 1/817 1/110 1/324 ~
1
, Core Power (kw) 330 200 88 1.39 E5 "
Pressure (Max)(MPa) 15.02 2.1 0.7 15.02 1
f
- a. At 5% decay heat M AK00147
- . . - - . . =_ _- , __ .,, _
MIST UMCP" SRI-2a Scale Velocity 1 0.26, 0.48 0.28, 0.5 Time 1 0.88, 0.48 0.88, 0.5 Power 1/817 1/810, 1/683 1/1542,.1/4909 l a. First number for single-phase, second number for two-phase. Both numbers based on property ratios at constant pressure.
M AK00 H5 r - - -
NI 1IL __U E*1gl i- ; a f- - - -. - ----- Till
l
- O O l O I
FACILITY LIMITATIONS AND ATYPICALITIES
, e Power l
l e Pressure l
j e Actual / Ideal Comparisons
- Geometric Parameters (flow areas, volumes, etc.)
Component Elevations Hydraulic Resistance I
1 i
i 1
i f
l 1
1 . . _ _ _ _ _ - - - _ _ a
g O B 12/09/85 TSB Doc 0460H Disk 0040H Job A2435 Proof 1 ~ dah LEJ TABLE 12. COMPARISON OF ACTUAL / IDEAL PARAMETER VALUES FOR IST FACILITIES O
Parameter MIST SRI-2 UMCP Operating Pressure 1. 0.046 0.14 Total Primary Volume 1.5 0.82 0.96
- Total Core Powera 0.1 0.05 (le) 0.059 (le) 0.176 (24) 0.052 (20)
Vessel Downcomer Volume 1.6 0.995 0.79 Flow area 1.1 1.14 1.56 Length 0.9 0.91 0.49 Gap 4.1 1.14 1.32 Vessel Lower Plenum Volume 1.05 1.02 0.26 Flow area -- -- --
Length 0.08 -- --
Vessel Upper Plenum / Distribution Annulus Volume 0.97 0.78 0.67 Flow area -- -- l 77 Length -- -- --
Core Region Volume 0.84 1.66 1.48 Flow area 1.1 1.9 3.05 Length 1.0 0.89 0.73 Og 1.0 4.5 12.8 Number rods 1.0 0.16 0.046 Rod diameter 1.0 1.45 2.3 Pitch 1.0 2.64 6.16 Heat transfer area 1.0 0.2 0.07 Volumetric heat generator rate 0.74b 0.67 (14)c 0.67 (le)d 2.33 (20) 0.47 (24)
, Vessel Upper Head l Volume 1.04 1.25 1.16 Flow area -- -- --
Length -- -- --
Hot Leg (1 of 2)
Volume 3.17 1.23 1.13 Flow area 3.4 1.00 1.09 Length 0.92 1.12 1.0 I. D. -
1.84 1.03 1.04 Pressurizer Total volume 1.0 1.04 1.15 h
Flow area 3.64 0.99 2.4 Length 0.267 1.9 0.45 l Diameter 1.88 0.99 1.55 5
. O O O SCALING ANALYSIS / EVALUATION i
,e Hot Leg Flow Regime 1
! e Hot Leg Vold Froction
! e Hot Leg Flooding e Pressure Drop l e Mixing o Critical Moss Flux e Pressure
! e Power i
i I
I J ,
i
C SCALING ANALYSIS / EVALUATION CONCLUSIONS e llL Flow Regime Bubbly flow in vertical Stratified flow in horizontal e llL Vold Fraction / Quality X
R
<< 1 Vold distortion for Xaf/f < 4 e Flooding distortions in HL If RVVV closed
! e Pressure drops scale by [R R if G, R', and )Dif /f3 scaled e Excess subcooling if room temperature llPI used in reduced pressure fccilities e Multiple break areas O O O
8 12/09/85 TS8 Ooc 0460H Disk 0040H Job A2435 Proof I _ dah LEJ O TABLE 26.
acaeRaLIze0 ISHII #ATURAL CIRCULATION SCALING CRITERIA PROPOSED BY Primary Relations Sincle-Phase Two-Phase 9 #SR C pg,g U RA o R I A #f I f
R p 1
R Must be selected Must be selected ag Must be selected Must be selected R RR R R t
R S AT )-1/2 7 , ,
gg S R AT, ) oR O y o
/032's '(*a')
i 2Sgh (pc/ #
l y
7 Ah ggg --
[ #c f c l Ap R Secondarv Relations a #s Ps 9 5R pC R AT R *R I aE R oR pp R R i
- O R R R
- R #R R PR "R 2 -2 R R O R
4 C fph \ \
R #sR PsR (AT,) aRlIoR 1 op /R #f,R 3 O ,
R
~
20
P PRESSURE e PropertyRatioMultipliers(Tg)onScaleEquations
- e, Single-phaseinotsameasTwo-PhaseT e If Pressure Changes T Changes i
4 e NPCH and NSUB must be specified d
l I
l i
\. O e e
O
~
O O i
Recognize :
- 4, n (fluid property groups) multipliers on scale eqns.
o If Pg = 1 [ full pressure capability?, p,, g = 1
- If Pg
- 1
-S,a*1 t
- For steady-state experiments - Pg = constant
- For transients
> Pg may not be constant ..
> Reference variables change with time
> pi, g is f(P,t)
MGH00906
i l
4 Questions for Transient Experiments o How to force S;, g. = constant ?
l ,
j
- How do reference variables (Nsub, Npch:1 change in time ?
e if 4, g 1:t? = constant what should model power [q: 1
! and subcooling (Ahsubf be?
4 j
MGH00907 j . e e e
O O O
~
Methodo ogy o LANL plant ca culation for nominal MIST transient (310000?
- P(t) .
I '
-Npch't?
-Nsub (t) o Specualte that if P(t)/P o R-1 Then f;,a m constant a Check speculation .
o Require
-N pch R
- -N sub g 1
o Compute q(t) and Ahsub(t) (gives cold leg fluid temperature) -
o Examine mixing, critical flow, break scaling, etc.
! o Equilibrium plots TOM t. AR SON 1085-4 8 l __
0 3 A e W O e o e c 5 N " & O o i 4
i i 4 l
@ O 5 ;
=
3 4
- = o ",
_ I I _
g m.la'. M
~~
l$
-2, r
-h 6 e 48 I
_ . x ., o ,. _
- l. .=. 3=3 e
,c en .
CL, %#'
o *1;
/ an O 1 /,
lF5 N
3e
. 2 2' It 3
< il m (f
a a -
- T i
272 m .*
e o ::""e
- O
/>
4 - t~2
/ _w,
~
l I i i . , ,
i ,
e .a_
- o w
CD 00 5 C @ t M N "
c i
(edW) e;nsseJd O
s 0
e e
O
- i e
m N 9 Q $
0 0 C oC 3 i 6 6 i ic a i i i y
i i I. i i 4 i a< =
- l e
l[ ] It 3E Q
c.
R e
y e .
13 it 2: o-i C.
/ 5cc ET
=
112 it n
x w
e ;q c, I
si - c.
7 C .:,
- u :c o:t 4 19 it >= -=
a
. . V 8:=
2 I ew E G@
e c :: 1: :c C-C 34
- \ -
@ E5 ~
i e u
- u u -=
I l 32
~~
2 :c-c
- d x N j
u-.
O e
g
? N O CD
- 9 M o ~
" " " O O O O O O
S 6
,,--...--,-,..n.. . , .- - -- - . . , . - - - - - . - . , . , . - - -
l i
j i
J l
4 EWJ
' te m N
(
o J-
- +,,m
- a s - ,,, ( J' ,) =t.,
- a w I
,, a a -.- l; *as i
e l
3 x 15 .
i v Based on seestant property ralle (at .6894 Ira to that at 6.894 Ira)
'A
" = = , __
N_ _
5 .
j Rogstre to stasch % for }#*9'8 8' **8*h % for sus pressere wasseest enweed pressere tramesent I I l i g
9 1900 2006 gege 4000 5000
\'
Plant time (s) musus IIqure 23. Scaled tore power for SRI assuming full P/Po pressure transient j
matched from plant calculation.
h
O i .
E O
K G3
- 3 L -
- act 01 LA.
E x
m m
ba)
E g
m u
$2l3 e e e O
1
1.0 --
Cold Leg Temperature Stratification .
i ii i i Plant j 0.8 -
-- -- - --- Mist i I -
-- UMCP 405K i 394K ' -
.- S RI 374K ,
289K - !
0.6 \1 422K -
y
.g l g . _ . . .
N
, j o :
V '
! 289K 0.4 [/
a n O.2 -
i j -
i :
' 8 '
- f. _
- i
0.0 200 300 400 500 600 Temperature (K) 10,unsoms.u
- 9 O O
_ m
e e
o 2 !
o g
g W I
, x . m i-
.c.
~
t u o -
E .:
q ~1
- 5 2 o i, e'
=7 y o -
f M-Sa 5
.S C'-
a.
0 @ $*
5 g :s; e
- e k c 55 O e x n ue Us z -
e g
gje G G G --
@C n a o % g ;;
I$2-o g ne n
N e si!
231 I
t e ;
~
j N- _
= c :
C
- o o
- WlH.DJo59 0
9
P BREAK SCALING e Fractional Mass Change
~- [R d R A R e AB,R g.
R 1
d B,UMCP
= 2.82 m (0.111 in) dB, SRI = 2.57 m (0.101 in) l ,
b i
l 1
- O e
. . 1 O O O
~
i i
I t i
l l t e Single-to Two-Phase Discontinuity I e Pressure con be Scaled Through T e Subcooling Difficult to Scale l
1 l
j l
i i
3
f DATA COMPARIS0N
.e Scoling RelGllonShips ;
e Equilibrium Plots i
I e Dimenslanless Groupings >
4 e T/II Codes 6
5 b
l O O O e
o l 0 o i
i l !
Steady State Experiment Premise :
0 =1 IFF N PCH = N SUB " l Pg g R R j Pg ,
l X Ap m g Ap Pr 9 Ap jg i ____i; _
- ~~
i Pg fil Pg Pg Pg ji i
mo del jg+j model X Ap ~~
ig l Pg ji i model i
Umodel I R t = t model I d9 R
) AP = AP model /(PR OI R M AK00142 I
I -
i i :
I l.
I !
! SSG-SOU I
s* g6** I n
kcm 463* gw x
- T l l 3 ,a 41 Y
j
~
$ f x <' nn _
i a
m ,
en 121~
2 j Isas l
i e . . .
25e 300 4F# 500 See I tot t
l Presswo (kPa) m ass-ar
} Figure 27. SRI equilibrium plot for 305 K (90 F) IIPI temperature and I leak area of 5.21 E-2 cm2 (5.608 E-2 ft2).
l.
I i
- G e .
O .
0 ,
9 ,
^ 6 ,
- o. ,
- "s n
[ 7n.
i 0
4 6
- g a e
r o't e.
' a
' " i a ,. 0
't s
t s
a 0
5 k
l a
e t
a.
s os .
- 0 d
. n i
4 a 6 1
- =
0 ,
s 9 l
- 4 )a )g A
0 P 'G .
4 (k ()
/2 t
4 er ,f 0 u 55 s rE -
y ' 9 s o O t r
o c
3 e
r f6 t0 o
6 i
e '
0 4
P l 8 p(
a r 3 s2 t
u no r
p -
i r
m 0 b2 e '
9 i -
l E t --
2 i d
- u9 i
u q 4, e
rl ef 0 1 4 I wg 2 Hf ole S o P d 4 .
%lo 0 3C y '
9 1
9 2
/
e x* r u
- - - f' n
0 4
1 i
F g
0 0 0 0 0 0 0 O 0
0 0
0 g0 g
0 0 7
0 6
0 6
0 4
0 3
0 2
0 1
1 1 0 8 1
Em5 c 3oE O
il l ll! ;4 ; i
C 12/18/65 TSB Ooc 0442H Disk 0090H Job A2435 Proof 2 _ dah LEJ The above discussion shows that the potential for comparison of low pressure scaled facility results to plant behavior via the concept of an equilibrium plot. The case with which this can be accomplished depends on combinations of local phenomena which influence the ability in the model factitty to satisfy the similarity criteria. It was shown that " correct" string of the break will most likely be an iterative process. Also, it was shown that it may be necessary to heat the HPI fluid in the low pressure facilities to produce a more typical response on an equilibrium plot.
4.2.2.2 Global Dim?nsionless Groupinos. If a system is considered to be a simple control volume with mass and energy crossing the boundaries and energy input (via core heat addition for example) then a simple mass and energy balance yields the following dimsenionless equations (see Appendix D).
.G v h, - u ) +v +sjqv _
g Bu -
av 3
au, v g
. y(E) + sj 6 r 6) . (97)
U di = G' v2 (90) where 9 l s' . ! .
(99)
O A "o c 3
D
.v O
c" #o g A i f
t .
' . (100)
O 16 i
)
'I >
C 12/18/85 TS8 Doc 0442H Otsk 0090H Job A?435 Proof 2 _ dah L E.1 Based on conservation equations for a global control volume for a
,]
system as shown in Figure 30. the following relationships were suggested as similarity relationships for steady-state situations:
Q
- H j= . (103)
>- in
\~' . .
Qin ~ 9eut * ~Nin (hy-h3 g ,
in fg Zuber notes that Hj represents a kinematic relationship and U 2 is a thermal expansion. For a transient, an elasticity group (mechanical expansion) is posed as H (105) f inf %a 3
pg O
Where a represents an average acoustic velocity for the system, t is a time scale, and AP represents a pressure change from one state to another. It was suggested that the time scale could be defined as a geometric mean i.e.
,. (106) t = Yin 90ut So that the mechanical expansion becomes Q
M AP
'3
- Q in pf4 2
- Zuber suggestej that these similarity relationships can be applied to components or to an entire system. For example, if Equation (104) is O written for the co/o (with C 0) neglecting heat loss (goog) it reduces b to Equation (A-10) 1.e.,
79
C 12/18/85 TSB Doc 0442H Disk 0090H Job A2435 Proof 2 _ dah LEJ Therefore, it is reasonable to scale the power for the reduced pressure h facilities per the two-phase relationship (see Table 26) in order to avoid problems associated with the single-phase to two-phase power discontinuity mentioned previously.
. The implications of the first item are that the reference pressure in the reduced pressure facilities should be selected so that latitude is allowed for pressure to increase after saturation conditions are achieved.
The hot leg temperature may then be specified so that it corresponds to the saturation temperature at the reference pressure. The power can be specified using the two-phase relationships and a reference AT computed by equating the single-and two-phase relationships for power. Although there may be numerous other methods for specifying the initial conditions, the following procedure is used here.
- 1. Selectareferencemodelsaturationpressureandcalculate/
property groups.
O
- 2. Calculate property groups at MIST saturation pressure.
- 3. Specify cold leg temperature from N SUB' 4 Compute power required from two-phase power relation.
- 5. Compute reference of ratio by equating le and 2$ power relations.
- 6. Calculate reduced pressure system AT assuming MIST initial AT is known.
- 7. Calculate hot leg temperature.'
iIt/a.Mvf ,
J. Notethat this is an tctm n t ';: process since the hot leg temperature should be close to the saturation temperature at the reference saturation g
pressure.
89
O O
~
~O COUNTERPART TESTS e Provide data base to examine methods of interfocility com arlson e Data base for determining /evoluoting system response differences attributed to scaling e Mapping tests e Nominal MIST transtent l
i
G.
~ tn M
C C
GJ auC 8
af m
C u
=
m I C
C M C aC C
'C .C 6 L M C u
C m
M C C C C O C -s
- w m W e e
- D U C >
c C 6 C U C U w
> C E 6 QJ O O > l w D C I in C -
> lll3 I
m C m m .C in
>== CJ C to C w E LLJ - C
~ E S E w C w w Z w C m g" g u C
m E ,
O'
q
~l B 12/09/85 TSB Doc 0460H Disk 0040H Job A2435 Proof 1 _ dah LEJ !
Q TdBLE28. POSSIBLE MAPPING TEST COUNTERPART CONDITIONS MIST UMCP SRI Core Power (RW) 119 78.7 10.64 Initial Pressure (MPa) 12 2.07 0.689 HotLegTemperature(K) 580-586 476 426 Saturation Pressure (MPa) 9.6-10.2 1.54 0.514 Cold Leg Temperature (K) 561 468 420.5 Secondary Pressure (MPa) 6.96 1.35 0.42 Secondary Level (M) 10.06 2.29 2.5 HPI/ Break Conditions Desired HPI Temperature (K) 306 420 410 0.106-0.126 0.372-0.52 0.14-0.18 HPIFlow(t/s)2)
Break Area (mm 2
1.224 12.3 7.16 3.02 Break Diameter (m ) 1.248 3.9 Alternate 1 HPI Temperature (K) -- 366 366
-- 0.176 0.044
% HPI BreakFlow (t/s)2)
Area (m -- 6.4 2.4 Break Diameter (m) -- 2.85 1.76 Alternate 2 HPITemperature(K) -- 339 339
-- 0.14 0.03 HPIFlow(t/s)2)
Break Area (m -- 5.2 1.7 Break Diameter (m) -- 2.57 1.47 O
21
l l
l l
l 1
Oi i
l 1
l
\
m C I O
U u V) t/)
GJ U H C C
Cn C -
- O C -
O CJ w V)
C V) - C E C C C c) -
H C G w C w M w U cJ -
< C C C L U A C U - L C C M C - V) D O LAJ C - C m .ad: U
& r-A G CJ C en C E M C > - O CJ -
- 3 - U M G G Q.
O H - - to CL, CCi :I:
U Co E CJ
- - 0 3 H I* Co M C 1 1 I E
L&J
- S S S S M
E M
O
O O .
O CONCLUSIONS e All IST facilities generally well scaled
- s. Facilities related by general scaling criteria ,
e No significant local phenomena distortions expected e Cold leg mixing strongly dependent on HPl temperature e Pressure can be scaled through property groups e Single-phase and two-phase scales discontinuous for reduced pressure operation e Use of two-phase power scale recommended to avoid discontinuous power changes
O l
e e b r o u l s l s s a i e e e v s r r l a y p u o l
) s r a a t s t n d
n e t a a
' a r n d t
t p a f n s c r o e
C o n o
c d
n a
i f
i f t t r O
(
n s a a p
S r r g e y N o e i i O f w s t r l I o p y i a s
a r
S d l g
U r a i s L a g l b e e C w n p i c t N
r i s e n O o g s s s n i C
f n o e o t a i i g
p e e h h t b b
_ g c a o t i r l n l d a r o e l l u
t r
f o p u
d o
r e
i w o n s o h f s
h s
o d r t e
f s y e g i t s
i l t M d n r e a y e e a
p i v i c t r
n o
l a
m e
d o
m t l e s r n c o a p p i e i C l m o r v f I I
e o r a p
e e /
a R C P S R T t m o
a C D - - - - - -
e' e O
!!l ll
O O
- O l
CONCLUSIONS (Cont'd.)
l e Counterpart Testing
- Steady-state most straightforward and useful
- Transient conplicated by break size, HPI flow, HPI temperature and time scale requirements
- Trial and error necessary to find "best" methods of conduct
-C
~
~~
l G&
~
O SCALING RATIOS THE REQUIREMENTS THAT 4/h= 1 IS T00 STRIGENT IF / IS A PARAMETER RELATING TO REGIME MAP OR STABILITY MAP.
EG. 1. FROUDE NO. IN A REGIME MAP CAN HAVE VARIATION IN THE RANGE OF 10 WHILE STILL BRACKET THE SAME FLOW PATTERN 2~ N PCH SUB THUS AS LONG AS THEY COVER THE SAME GENERAL STABLE / UNSTABLE REGIONS, ONE CAN TOLERATE SOME VARIATION O 3. IF ONE USES ZUBER'S GLOBAL SCALING CONCEPT, THEN EVEN IF N PCH '
SUB (N '
PCH SUB I4. J g MAY NOT NOT BE 1, BUT AS LONG AS K g IS PRESERVED, CCFL IS SCLAED t
O l
O ROLE OF FACILITIES
- 1. TO PROVIDE DATA FOR CODE ASSESSMENT
- 2. TO PROVIDE DATA FOR IllTER-FACILITY COMPARISON BASED UPON SCALING f1ETHODOLOGY
- 3. TO PROVIDE OPPORTUNITY FOR SENSITIVITY STUDIES
- 4. TO PROVIDE " MAPPING DATA" ON FLOW REGIME,
([)
~~ ~ ~ '
" EQUILIBRIUM FLOW PLOT" ETC. TO CHECK SCALING METHODOLOGY R
l O
1 l
O THREE LEVELS OF EXPECTATION FOR INTER-FACILITY DATA C0f1PARISDN
- 1. DIRECT COMPARISON BETWEEN TWO FACILITIES -
CAN ONLY BE EXPECTED FOR WELL-SCALED FACILITY WITH FULL PREESURE TEST CONDITION
- 2. TRANSLATION OF TEST RESULT THROUGH SCALING METHODOLOGY, SUCH AS " COMPRESSED TIME SCALE",
"P/P g vs. WA(g" ETC. , THE HOPE IS THAT SCALING f1ETHODOLOGY CAN BE DEVELOPED 10 ENABLE US TO THIS IS AN ADDITIONAL, O PERFORM SUCH TRANSLATION.
BUT MORE BASIC MISSION OF VARIOUS REDUCED-SCALED FACILITY
- 3. COMPARIS0N OF DATA THROUGH COMPUTER CODES -
THIS IS AN ACHIEVABLE G0AL AND IS THE BASIC EXPECTATION, WHILE " LEVEL 2, SCALING TRANSLATION" BEING AN HOPED-FOR G0AL.
O
O LOS ALAMOS SUPPORT OF THE MIST EXPERIMENT PROGRAM THAD KNIGHT JAMES STEINER O
SAFETY CODE DEVELOPMENT LOS ALAMOS NATIONAL LABORATORY ACRS MEETING PALO ALTO, CA i
JANUARY 24, 1986 i
O JPS/cs/3923ST6A
TRACE AND RELAP5 DIFFERENCES STEAM GENERATOR MODELS BOTH CODES REPRESENT THE PRIMARY SIDE WITH TWO CHANNELS
- 3 TUBES IN A WETTED CHANNEL 16 iUBES IN AN UNWETTED CHANNEL THE SECONDARY SIDE IS REPRESENTED DIFFERENTLY BY THE TWO CODES:
RELAP5 USES A SINGLE CHANNEL O -
TRAC USES TWO CHANNELS WITH CROSS FLOWS HEAT TRANSFER DIFFERENCES O
JPS/cs/3923ST6A
TRAC RELAPS h, = f (a , T, , t )
hp = f (T,) h, = f(T A T,ot) y Ty ~T sot i f a< 1 3l
, hp = f(T,)
4
\( 7 N
(- *-g*i 7ph= f(T,)
d h, = f (a ,T,) ifa< 1 h, = f(T y ) ifa=1 )
_c c _
hp = f (T,)
h, = f (a ,Ty,T,)
osA:smies e e e
O PRIMARY-SIDE HEAT TRANSFER STEADY STATE BOTH CODES USE THE DITTUS-BOELTER CORRELATION FOR FORCED CONVECTION HEAT TRANSFER TO SINGLE-PHASE LIQUID.
THE PRIMARY-SIDE REPRESENTATION IS ESSENTIALLY THE SAME FOR THE TWO CODES.
O i
I l
2 l
O JPS/cs/3923ST6A
SECONDARY-SIDE HEAT TRANSFER UNWETTED TUBES ,
STEADY STATE IN THE POOL THE CODES BOTH USE THE CHEN NUCLEATE-BOILING CORRELATION. THE POOL ACTS AS A LARGE HEAT SINK FOR THE PRIMARY AND THUS SETS THE PRIMARY OUTLET FLOW AT THE SAME TEMPERATURE FOR BOTH CODES.
ABOVE THE POOL, BOTH CODES USE THE DITTUS-BOELTER CORRELATION FOR FORCED CONVECTION HEAT TRANSFER TO Q STEAM.
FOR TRAC, THE CODE USES DITTUS-BOELTER AB0VE THE POOL BECAUSE THE LOCAL VOID FRACTION IS 1.0. THE SINK TEMPERATURE IS T (VAPOR) > TSAT. WE OBSERVE VAPOR SUPERHEATS OF AS MUCH AS 26K.
O l JPS/cs/3923ST6A
O SECONDARY-SIDE HEAT TRANSFER UNWETTED TUBES STEADY STATE (CONTINUED) k FOR RELAP5, THE AUXil,lARY-FEEDWATER WETTING MODEL FORCES THE UNWETTED CHANNEL TO DITTUS-B0ELTER HEAT TRANSFER TO VAPOR ALTHOUGH THE SECONDARY VOID FRACTIONS ARE s 1.0. THE SINK TEMPERATURE IS T (VAPOR). MAXIMUM OBSERVED SUPERHEAT IS 8.4K.
Q NET RESULT IS THAT THE TRAC CALCULATION ABOVE THE POOL PRODUCES A HIGHER SINK TEMPERATURE.
O JPS/cs/3923ST6A
SECONDARY-SIDE HEAT TRANSFER WETTED TUBES STEADY STATE THE POOL HEAT TRANSFER PHENOMENA ARE TREATED THE SAME AS IN THE PREVIOUS DISCUSSION ON UNWETTED-TUBE HEAT TRANSFER. THEREFORE, THE PRIMARY-SIDE COLD-LEG TEMPERATURES ARE ESSENTIALLY THE SAME.
AB0VE THE POOL, THE VOID FRACTIONS IN TRAC ARE s 1.0 AND THE CODE INITIALLY SELECTS THE CHEN NUCLEATE-A CHECK IS THEN MADE ON VOID
() BOILING CORRELATION.
FRACTION. AT HIGH VOID FRACTIONS, THE CODE ALSO CALCULATES FORCED-CONVECTION WITH DITTUS-80ELTER TO SINGLE-PHASE VAPOR AND INTERPOLATES THE HEAT-TRANSFER COEFFICIENT BASED ON VOID FRACTION.
O JPS/cs/3923ST6A 1 s __ - _
O SECONDARY-SIDE HEAT TRANSFER WETTED TUBES STEADY-STATE (CONTINUED) l THE AFW WETTING MODEL IN RELAPS AB0VE THE POOL ATTEMPTS TO REPRESENT THIN-FILM NUCLEATE BOILING HEAT TRANSFER WITH A SPECIAL THIN-FILM CORRELATION WHEN THE LIQUID IS SUBC00 LED AND CHEN NUCLEATE BOILING WHEN THE LIQUID IS SATURATED.
NET RESULT IS THAT THE TRAC CALCULATION ABOVE THE POOL
[])
CAN PRODUCE LOWER HEAT-TRANSFER COEFFICIENTS THAN RELAP5.
! ($)
JPS/cs/3923ST6A 1
O
SUMMARY
SECONDARY-SIDE HEAT TRANSFER TRAC TENDS TO FORCE MORE OF THE HEAT TRANSFER DOWN TOWARD THE POOL RELATIVE TO RELAP5.
THE EFFECT OF THIS SHIFT IN HEAT TRANSFER IS TO LOWER THE PRIMARY-SIDE THERMAL CENTER IN TRAC RELATIVE TO RELAP5.
(THE THERMAL CENTER IN THE CORES IS NOT AFFECTED
(~j) BECAUSE THE AXlAL POWER DISTRIBUTION IS FIXED THROUGH INPUT.)
1 l
l i
l l
($)
4 JPS/cB/3923ST6A
1
SUMMARY
SECONDARY-SIDE HEAT TRANSFER (CONTINUED)
BECAUSE OF THE REDUCED THERMAL-CENTER ELEVATION IN THE TRAC STEAM-GENERATOR PRIMARY. TRAC PRODUCES LOWER PRIMARY FLOWS IN NATURAL CIRCULATION.
WE PROPOSE MAKING NO CHANGES IN EITHER CODE UNTIL MIST
- PRODUCES DATA INDICATING THE CORRECT SYSTEM BEHAVIOR.
. AT THAT TIME MORE SEPARATE-EFFECTS DATA MAY OR MAY NOT BE NEEDED TO PRODUCE A SUPERIOR AFW WETTING MODEL.
O
- O JPS/cs/3923ST6A
O TRAC Code WE HAVE NOT MADE ANY CHANGES TO CODE MODELS BASED ON OUR ANALYSES FOR GERDA, OTIS. AND MIST.
(
WE HAVE FOUND CODE ERRORS (LOGIC AND FORTRAN) AS WE USE
, THE CODE IN PREVIOUSLY UNTESTED AREAS. THESE ERRORS 2
HAVE BEEN CORRECTED.
i o
I E 9
's 1
O -
2 JPS/cs/3923ST6A
+ ,
i Q TRAC INPUT FOR MIST WE HAVE USED ONLY PHYSICAL MODELING TECHN10VES.
! THE INPUT GE0 METRY IS BASED DIRECTLY ON FACILITY GEOMETRY.
l THE TRIP AND CONTROL FUNCTIONS ARE BASED ON THE PLANNED I OPERATION OF THE FACILITY.
l 1
i O i
l i i
I ,
i i
O i JPS/cs/3923ST6A ;
4 i
l O N0 DING SENSiTivi1Y STUDIES DETAILED UPPER-PLENUM CYLINDER - RESULTS SHOW THAT THE CALCULATION IS SENSITIVE TO THE INCREASED DETAIL IN THE N0 DING WITH CALCULATED EVENTS SHIFTING IN TIME BUT WITH NO CHANGE IN THE CALCULATED EVENTS THEMSELVES.
IN THE l-D COMPONENTS, THE DETAILED N0 DING OF THE UPPER-PLENUM CYLINDER REQUIRES VERY SMALL CELLS (ABOUT 20CC) IN ORDER TO MODEL CORRECTLY ALL OF THE FLOW PATHS. THESE SMALL CELLS ADVERSELY IMPACT THE TIME-t STEP SIZE BECAUSE LEVELS TEND TO RESIDE AT THE SAME ELEVATIONS AS THE SMALL CELLS. WE WILL INVESTIGATE A MULTI-DIMENSIONAL MODELING BASED ON THE MORE FLEXIBLE VESSEL COMPONENT.
l l
l l
1 0
1 :
JPS/cs/3925STEA
I n
U N0 DING SENSITIVITY STUDIES (CONTINUED) ,
STUDIES OF THE COLD-LEG INJECTION AND MIXING IN OTIS - ,
N0 DING SENSITIVITY STUDIES OF THE COLD-LEG INJECTION AND MIXING IN OTIS REVEALED A MINIMUM ACCEPTABLE N0 DING (ONE CELL BETWEEN THE INJECTION POINT AND THE ,
2 DOWNCOMER). OUR MIST N0 DING IS CONSISTENT WITH THIS
~
FINDING.
- O i
l l
t O .
JPS/cs/3923ST6A 1
l l
l N0 DING SENSITIVITY STUDIES (CONTINUED) l l
l TWO-CHANNEL MODELING 0F THE OTSG - THE RESULTS FROM OUR j I
OTIS ANALYSES SHOW THE IMPORTANCE OF BETTER REPRESENTING THE FLOW FIELD AND HEAT TRANSFER IN THE OTSG SECONDARY. THE APPROACH IN THE l-D COMPONENTS HAS BEEN TO SPLIT THE SECONDARY (AND THE PRIMARY) INTO TWO CHANNEL BASED ON DATA FROM THE VENDOR.
FURTHER EXPERIENCE WITH MIST SHOWS THE IMPORTANCE OF CROSS FLOWS.
(v~3 CALCULATIONAL RESOLUTION OF THIS PROBLEM WILL REQUIRE AT THE VERY LEAST MULTIDIMENSIONAL TOOLS.
l O
JPS/cs/3923ST6A
. . . - . _ ~ - - . - - _ - _ - - . _. ._. - . . . _ . _.
!'O N0 DING SENSITIVITY STUDIES t (CONTINUED) i
- l-D VS. 3-D MIST 00WNCOMER BEHAVIOR - WE HAVE RUN CORRESPONDING CALCULATIONS WITH 1-D AND 3-D REPRESENTATIONS OF THE MIST UPPER 00WNCOMER. WE DID NOT OBSERVE ANY SIGNIFICANT 3-D BEHAVIOR IN THE FLOW
! FIELD CALCULATED WITH THE 3-D VESSEL COMPONENT.
INTEGRAL BEHAVIORS WERE VERY SIMILAR.
O 1
i -
JPS/cs/3923ST6A l
O .O 0 .
M S Tes- 3'0000 -
L L
- Nominal case test
- Initial natural circulation conditions
- Steam-generator AFW injected at upper minimum wetting j nozzles AFW head-flow characteristics from Fig. 6.3 in MIST f acility specifications 1
1 1
- Initial core power 3.9 % of scaled full power
!
- Scaled 10 cm2 cold leg discharge leak in Loop B il e HPI at cold leg discharge locations l
l
- S t eam-genera t or-secondary levels increased from 1.524 m t o .9.632 m a t f u ll AFW flow i
l
- Steam generator-secondary pressure controlled to simulate
!. 55 K/hr cooldown ,
- Reactor vesseI veni valve in automatic controf mode with open/close set points of 862 and 276 Po j
- Hot leg high point vents not utilized
o 8 O .
Summary of Expected MIST Operating Procedures .
i for Test 310000 i
- Estoblish steady-state conditions during notural c ir cu i a ti on mode i
I e i n iti a t e transient by opening the cold leg discharge leak l valve i
l
- Perf orm the f ollowing actions when the pressurizer liquid level decreases t o 0.305 m r e l a ti ve to the bottom of the p r essuri ze r l
- Trip HPI sys ? ems at all CL discharge l oca ti ons l
- Initiate core power decay 1
Begin automatic control of RVVV (initially in closed
>s iti on) l
! f ort f ull AFW flow to increase the secondary liquid l
ieveIs from 1.524 m to 9.632 m o Maintain SG secondary level at 9.632 m f or duration of
- tr ans i en t
- Terminate test after the hot leg u-bends are refilled and natural circulation is r e-es t ab li shed
- Nominal initial Test Conditions for MIST Test 310000
! System pressure Vary to obtain 12.2 K
) subcooling SG secondary pressure 6.96 MPa (1010 psia)
Core power 128.7 kw 4
l SG secondary liquid level 1.52 m (5.0 f t) i l
Pressurizer liquid level 0.76 m (2.5 f t) i AFW fluid temperature 310.9 K (100 F) l HPI fluid temperature 299.8 K (80 F)
Core flood iank fluid iemperaf ure 299.8 K (80 F)
< Core flood tank pressure 4.24 MPa (615 psia) 4
s O
g ,
g ( q................g
$$ ; i i i <
E_ _-
f
$ 8 b, i
I mm m m en
< 2
_E 8 a @ g @
r Lr pl !... ,,,,,, _ :_ _ ..
co lc=. ____.___. __ _._
-i Eo u -.J 13 ll L
/g
/ l l "l
is.
m e e e/ '
/e e e I J j
'/ / 8 l j........--
- E u= E ,
3 .e
- O ,
co
- @ -,-i
@ v (g) u (g) u
-o J
'. Z)
(g
(
T_
> l(g)
. . ,,2:n - . $ _ . .__ _ _.
j.$ j
- - ! .i --F-------J j 5 g i I a li 3R g ga
$ l 9 E m w a @ @a f i+ mm m on 1
- ~
~ ([r e
g <
4 g . . . . . . . . . . . . . . . .p-r Oh o $
O d O .
- Calculated Steady-State Conditions f or MIST Test 310000 Upper plenum pressure 13.14 MPa (1905 psia)
Primary fluid subcooling 12.3 K (22.1 F)
SG secondary liquid level 1.60 m (5.25 f t)
Pressurizer liquid level 0.76 m (2.52 f t)
Hot ieg fluid iemperature 592.1 K (606.0 F)
Cold leg fluid temperature 560.7 K (550.0 F)
Loop mass fl ow 0.325 kg/s (0.71 lbm/s)
- 3 _ e .
TRAC-PF1 PRut.si PREDICTION OF MIST TEST 310000 .
PRIMARY AND SECONDARY PRESSURES 14000000 ' ' ' ' ' ' '
-2000 flot teg/ upper head fittid scttunitted o PRIMARY -1800 12000000- -
_ a INTACT SEC -1600 10000000- /
= End of NC in -1400 g both toops Q.
Irtterruption
- Nigh ,AF l P BCM u 0000000~ , of NC rn Loop B BCM ut Loop B -1200 T.-
3 SG and flow High AF BCM in Loop A SG.
~
y evtculation in Loop B tart of pronary system refill v k
e cold legs M N
- d- -1000 6000000- -
_ Flow circulation SG ptunaries _ggg in loop A cold refilled tego irritiated _N W
~
4000000- - -600
--SG secondaries fitted to 9.63 m
-400 2000000 , , , , , , , , ,
0 500 1000 1500- 2000 2500 3000 3500 4000 4500 5000 Time (S)
e O O TRAC-PF1 PRETEST PREDICTION OF MIST TEST 310000 PRIMARY AND SECONDARY PRESSURES 14000000 , , , , , , , ,
)
- l. I 12000000- _
a INTACT SEC 10000000_ + BROKEN SEC_
i
! o x BREAK 32 l S-e
- $8000000-
- BREAK 69 _
E ,
E i i o.
6000000- _
- ; ; - O ~
4000000- _
i i
2000000 , , , , , , , ,
0 500 1000 1500 2000 2500 3000 3500 4000 4500 4
Time (S)
O i i i i 8
4 . . .
n 8 d a - a u
_ "o @ 'X _8 o f 8 $ N_f ? :
o 9 5
~
M o- ~8 o
kE 4 +
x D a
o s 3: _ e _o h
- u. > m l
~
P, 8
$m q N3 O "9 m ;! E h3 _
l _ F Eh d M.m i l'w2E _ _8 S
EJ n .
i
~ ~
o l, o
~S
~'
I i "4
6 l
'^
T 4 4 o
4
- =
o 9
o x
o N
o o a o
M o
9 o
n V
(S/63) mou sson lolol ,
o
TRAC-PF1 PRETEST PREDICTION OF MIST TEST 310000 s
HOT-LEG U-BEND MASS FLOW O.6 , , , , , , , ,
4 l
0.5 - o INTACT -
i a BROKEN O.4 - 4 8
] [ 0.3 - n r
o '
l E l
i 0
0.2 - h( -
l 1 l
i j 0.1 - -
0.0 - '
- :-^ ^ -
- - =^I- *= 2 ^^ q -
-0.1 , , , , , , , ,
- O 500 1000 1500 2000 2500 3000 3500 4000 4500 71 IImO (S)
L l __ ____ __ ____
^
O O O ,
Sequence of Predicted Events for MIST Test 310000
Event , Time (s) n m (s) l Cold leg leak initiated 0.0 >
Pressurizer level decreased to 0.305 m 48.1 so I initiation of core power deco , HPI, automatic VVV control, and refilling of G secondary) l i
Loop A hot-leg u-bend fluid saturated 87.0 9a Loop B hot-leg u-bend fluid saturated 100.0 103 l Interruption of natural circulation in Loop A 168.0 170 i Pressurizer empty 243.0 243 I SG secondory filled to 9.63 m 495.0 509 i l Initial interruption of natural circulation in Loop B 488.0 n/n l Comple t e interruption of natural circulation in Loop B 855.0 840 Both SG secondory pressures become equal 2975.0 ?959
.I
- BCM commenced in Loop A SG 3216.0 1000 Fluid upstream of leak saturated 3314.0 1289 j Stort of primary system rafill 3393.0 3300 i Terminated coIcuIation 5000.0 4200 t
\
O'..
! O O l Conclusions of MIST Pretest Prediction No. 11 l
i
!
- Current results not significantiv diff erent from '
l previous MIST pretest ca l cu l a ti ons
- MIST pretest calculation exhibits all phenomena l that is expected to occur in a B&W PWR and i as observed in the OTIS tests l = Draining of pressurizer and loop saturation
= Intermittent loop circulation
- *PooI and high-auxiliory-f eed boiler-condenser hea t-tr ans f e r modes i = Primary system r e fill i
- e MIST calculation exhibits large assymetric loop behavior
- Hot leg vent valves should be opened after SG l p ri ma ri es r e fill ed to aid in f as t e r r e filli ng o f
- the primary system and earlier comp I e ti on of
- the test i
- i. _-_ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _
/ :v6" j,6 O
MIST PRE-TEST CALCULATIONS TEST NO. / TYPE.
CODE STATUS 310000 SSLOCA(NOMINAL) RELAP5 COMPLETE TRAC. SENSITI\llTY STUDIES RETRAN IN PROGRESS 310$O3 SSLOCA (ASYM. sG) TRAC. lhi PROGRESS M OIot 6BLocA (POMP BOMP) TRAC NOT STAR.TED 330201 FEEMBLEED (EM ECCS) TRAC IN PROGRESS 330302 FEED / BLEED (DELAY HPI) TRAC IN PROGRESS 320201 sSLOCA C50 GM*> RELAP5 COMPLETE 340403 MTR (SG.150LATEb> RELAP5 CELETED 340502 3r.7R(t0W ELEVATlDN %1R) RELAP5 COMPLETE S40100 SCTR (SG 1.2NEL NOMINAL) RELAPS TBD 340504 MTR/SLB RELAP5 TBp O
l
! l 1 K r < < L,
- .i -3 -
- i
- l i O ;
I
( .
4 t
RELAP5/ MOD 2 IST ANALYSIS TESTS i
8
} ACRS ECCS SUBCOMMITTEE MEETING i
i; PALO ALT 0, CALIFORNIA JANUARY 23-24, 1986 %
4 1
l !
lo Bv JOHN KLINGENFUS i BABC0CK & WILC0X i
I I
L i 1
i i :
j . l I
I
t t O
RELAP5/ MOD 2 IST ANALYSIS TASKS I. OTIS POST-TEST PREDICTIONS e FUNDED BY B&W OWNERS GROUP
- OTIS TEST 2202AA--A SCALED 15 CM2 CLPS LEAK WITH PRESSURIZER ISOLATION
- OTIS TEST 220899--A FEED AND BLEED C00LDOWN TEST USING PORV ACTUATION AND HPI WITH MINIMAL SG HEAT TRANSFER O FUNDED BY TOLED0 EDISON (DAVIS-BESSE)
O 2
- OTIS TEST 230299--A SCALED 10 CM CLPS LEAK WITH MINIMAL SG HEAT TRANSFER AND SCALED FULL-CAPACITY LOW-HEAD HPI FLOW, THIS TEST INCLUDED LEAK ~ ISOLATION, FEED AND BLEED COOLING, HIGH ELEVATION BCM, AND LOOP REFILL WITH REC 0VERY OF SINGLE-PHASE NATURAL CIRCULATION, O
, o l
O l
l l
l OTIS TEST 2202AA FUNDED BY B8W OWNERS GROUP PERFORMED BY MARK RINCKEL, B&W O
o a Ficure 1-2. PELAPS N0 ding Diagram O
9 AW M/.,
950 120-9 STM P 650 120 6 440 &A J 374 370 120-10 r , , 450 l d 130
- ' 200-8 9 0 J 3 TUBE f20-5 200-7 140-1 654 150-1 g
200-6 16 TU8E
% 150-2 140-2 653 200-5 120 4 140-3 652 150-2 200-4 200-3 140-4 635 150 4 f20-3 200-2 140-5 530-5 150-5 200-1 A -\
150-6 N _-
\ 120-2 630-4
~
140-6 210 g 210 21f
-1 210 -3 2
- \
\
140-7 630-3 150-7 \
\
g 120-1 140-8 630-2 150-8 \
b 140-9 630-t 150-9 550 tic r \ a
" 160-2 '"
N f53 l OE 171
. ,gg -
180 h 55 \ _ 160-1 460 t.EAK 480 470 IOS*1
'c:,"'
rpg eac > _
y 490 $
N h
- g 185-6 Sf0-3 5'0-2 E
f85-8 5t .:
g 9
1-5
i i Table 3-1. Test 2202AA OTIS RELAPS/ MOD 2 Initial Conditions O Test MOD 2 Model Core Power, % 4.17% 4.17%
Scaled full power Primary Press., psia 2207 2200 PZR Liq. Height, ft. 20.3 20.4 SG Sec. Level, ft. 5.7 5.7 PRI Hot Leg Temp., F 609.2 610.8 PRI Cold Leg Temp., F 571.5 571.3 PRI Mass Flow, % 5.60 5.68 Scaled full flow Scaling factor =
_1__
1686 Scaled full power = 2.13 MWt
- Scaled full flow = 25.9 lbm/s O
3-7 Babcock & WilcOE J MCDermott comparty
r .
Table 3-2. Secuence of Events O
Experiment MOD 2 Simul. Experiment Event Ficure (3-11 (min.) Ficures 3-2 thru 3-7 (sec)
Steady 0-11.4 0-110. ----
State Leak 11.4 110. 110.
Actuation Przr. 12.8 194. 194.
Isolation Core Power 13.75 251. 251.
Reduction High Pressure 13.75 251. 251.
Injection Begin Secondary 13.92 251. 261.
Ramp to 38 ft 1st Interruption 15 min. 360. 326.
of Circulation BCM-High Elevation 22.5 675. 775.0 lll Pool BCM 34. N/A Not Shown Transient ----
1350. 1350.
Terminates O
3-8 gag g gggggg a McDermott company
. i Figure 3-1. Primary and Secondary Pressure (Observed)
O' Test Initiation is at 11.4 Minutes 4c::
1 INDEx VIA3 ;
'RV KP9 4VP82J I STMK DR aspazo f )
J S00. .. l l
l l
i 30:0 PHASE 1 ,'
PHASE 2 25:: -
i
- PHASE 3 t DRZR. ISOLATED u 200: #
I 5
e } SPILLOVER ,
d ;
b .m HIGH ELEV. BC!t E -
6 .---- 1 iS:: .
I PQQL BCM
,:e3 2
\. / /
F ;
- N .: .
50 0/HR DEPRES$URIZATION INITIATED sac .
l i
o, , , , . , i is ao is oo ao oo as a: :: ce ss.oc 4: ec OTIS TIME (MIN ) 0-0755 42. 19-APR-34 3-9
I e 8
e 93 '9 e
/ .
. a
. n._
t e
l C (
.e .
t 5 'l me PS c) 20 -8
,v_, k , -
o , y
<M D
W n< :
Ng ' . ,
g
.9
~
Ud ,'
eH ww N
g .'*
a 25
-m 8
OM ,e e
I c N'
w g
,y
.' x a
g a*
f T u Q 8 M
m 13 W<
a Od na O C C
, l a O'00k2 0'0002 0*Ob51 0'C021 0*dO9 0'00F (ENI/J8T d 3-10
ll
~ .
0 O 0 0
5 1
aj-0
, 0 0
2 1
2 e
M C
5 1
n o
g i
t c 0 u
Sl 0 pv e 0 me uL '-
9
)
P r C So I t E
Ta S Or I e
O An Ae 2G 0
E M
I 2m 2a tt e 0. T sS 0 e E 0 TA 6 C
3 e
r u R 0 g T i R 0 F - 0 D 3 S
TP A S L EE 0 T R a C a
0 I
0 98 g ., odm 9o O > O _i ooN MT>)_.WHZo o
t
I 0
0 0
5 1
n 0 0
0 y 2 1
v n
i o
t c y u '
s y
S p
m 0 u 0 P 0 w 9 So Il 1 TF O C R E m
k Aa T S Ae 2L 0
2A 2C R
O 5
g E
SH I C 2 TR o - 0 0
- 0. T T5 6 O1 5
o a 3
e r
u g
m su i
f 0
0 0
3 A
m 1" I 0
0 L -
9o ro g
n 1 i c 7?
h sd 7 w w6
9 C] 8 9
13 0 13 0
/. g i
o r
iE / u M
. N-s C .
U
$ }
E s n 9
- c ..o O
j . '. -
o
- m. s- u D
m '?
Y ,i -
og ,
L1 O II m
25 O
- l. '
' r n
~y y' l e N w*
i< h 8 m8 % /
Cu om t i f
s' l C N[
A ,'
,u c
h e o-en t/ c t :
ca 8 e>
Cl D<
U ua
, A t-- W D o 0
9 0 i , , .
C'00k2 0*0002 0~0091 C'002! C'009 G*C0F t2NI/J971 d Q
3-15
O OTIS 2202AA CONCLUSIONS 8 THE RELAP5/ MOD 2 PREDICTION OF OTIS TEST 2202AA WAS VERY RESPECTABLE. THE CODE PREDICTION DURING THE INITIAL BLOWDOWN, THE INTERMITTENT CIRCULATION, AND THE BEGINNING OF THE HIGH ELEVATION BCM PHASES WAS EXCELLENT, 4 THE ONLY MAJOR DEVIATION OCCURRED AT APPROXIMATELY 875 SECONDS. AT THAT TIME THE SECONDARY PRESSURE CONTROLLER RAPIDLY DECREASED THE PRESSURE TO 975 PSIA. THIS ACTION CAUSED A LARGE LEVEL SWELL PLUS FLASHED A CONSIDERABLE lll AMOUNT OF POOL INVENTORY WHICH REINITIATED AFW AND OVER-PREDICTED THE PRIMARY TO SECONDARY HEAT TRANSFER.
8 A BOUNDING CALCULATION (FIGURE 3-7) IN WHICH THE SG PRESSURE WAS NOT BLOWN-DOWN RESULTED IN AN UNDER-PREDICTION OF THE HEAT TRANSFER. A GRADUAL PRESSURE DECREASE WOULD PROBABLY REPRODUCE THE ACTUAL TEST RESULTS.
O
1
- O .
RELAP5/ MOD 2 POST-TEST PREDICTION OF OTIS TEST 220899 i !
O FUNDED BY B&W OWNERS GROUP PERFORMED BY MARTY PARECE, B&W Y
i J
i r
. O l 1 i e--w e v. -,e
-w=*m---ww r ywwe . - - - - --cyw9-ym pww --
,w,,y m.pm,=wwwwwmcw www
O OTIS 220899 POST-TEST PREDICTION 0 THIS TEST IS A FEED AND BLEED C00LDOWN TRANSIENT USING THE HPI AND PORV WITH MINIMAL SG HEAT REMOVAL.
8 TEST IS INITIALIZED WITH HIGH AFW, UPON INITIATION, AFW IS DIVERTED TO LOW INJECTION TO MINIMIZE HEAT TRANSFER.
6 FULL HPI CAPACITY HEAD FLOW IS UTILIZED. g O
_Ficure 1-2. RELAPS Noding Diagram AFW $ ($v 120-9 STM P 650 120-10 440 &, 320-6 374 370 v , , 450 g ,3n
) ) 200 8 0 n 9 i ruse 120-5, 140-t 654 150 1 /
16 TUBE 200-6 140-2 653 150 2 200-5 120-4 f40-3 652 150-2 200-4 ,
0-3 14 4 635 !30 4 200 2 f20-3 140 5 630-5 150-5 200-t 140-6 630 4 150-6 g 120-2 N
140-7 630-3 150-7 g 4
\
\
i f20-f' 140-8 63C-2 150-8
\
140 9 630-1 150-9 550
-\ ~ "'
160-2 2
?7? .f4L.j 5' 153 1 171 00 l 180 h55\ _ 160-t 460 185-f 520 5 LEAK 480 470 9 mpt t -
m
"'O h-
$.' W D
~
(Cone J
185-6 510 3 M S10-2 185 8 -
5'0 1
=-g -
0 1-5
r O
OTIS TEST 220899 INITI AL CONDITIONS ACTUAL TEST RELAP5/ MOD 2 PRIMARY PRESSURE, PSIA 2180.* 2178.7 SECONDARY PRESSURE, PSIA 1200.* 1200.2 HOT LEG TEMPERATURE, F 610.0 610.6 CORE POWER, % FP 4.2* 4.2 PRESSURIZER LEVEL, FT 18.3* 18.3 SG SECONDARY LEVEL, FT 5.4* 5.53 IEI CORE FLOW, LBM/S 1.44 1.48 HOT LEG SUSC00 LING, F 38.1 37.5
\
'SPECIFIED llI i 1
~_- I
O OTIS TEST 220899 SEQUENCE OF EVENTS ACTUAL TEST RELAP TEST DAS TIME DAS TIME EVENT (MIN) (MIN)
DIVERT AFW FROM HIGH TO 10.1 10.1 LOW, RESET SECONDARY LEVEL AND PRESSURE TURN ON HPI 10.7 10.7 CORE POWER DECAY RAMP 10.8 10.8 PRIMARY PRESSURE t2300, 13.0 13.0 OPEN PORV REACTOR VESSEL UPPER 22.0 35.9 HEAD SATURATES PRESSURIZER FILLS WITH 24.0 '
23.4 L10VID TEST CALCULATION TERMINATED --
60.0 0
OTIS TEST 220899 lHPI/PORV C00LINGD 2400 2300- -RELAP5/M002 a OTIS DATA 2200-210F 2
a2*
g' 1900-2 8 1800-E a a iW a, $'""1,1.
> 0 a
h 1600- oooooo oo,,
,g *aoooooo, 150P *o o o,O o O
$ 0 0 0 0 C 1300, i i i i i i i i i 5 10 15 20 25 30 35 40 45 50 55 60 Time After DAS Actuation. Minutes O O O .
O O O
~
l OTIS TEST 220899 i
3
- ELAPS /M002 J
- OTIS DATA
- 6-l
'EPER LEVEL TAP i
oo***
J 40 -
k i
m i
a2 35-i m
-s 1
b N
2- ,
.c4 c_ o
- 3 i #
m b~ o m o t
' a.
- i g_
1
! 15 i i i i i i i i i i i i i i i 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 Time After DAS Actuation. Minutes l
l i OTIS TEST 220899 6
l - KLAPS/M002 i f o OTIS DATA
- u-cn 610-Qi l
l E %
, k .
y 600-i i 590-i i y 2
1,1'
- .5 580-
' "J ,
l N A"<
g 570-l E 560- '
- O <
i 550 i i i i i i i i i i j 5 10 15 20 25 30 35 40 45 50 55 60
! Time After DAS Actuation, Minutes l
- O e.
l
O O O
~
1 i
OTIS TEST 220899 600
, g AP5/M002 o OTIS DATA g 595-U) .
$$ 590-i C I
U8 0o o
O 585- # ,
ci EE f 580- > "o l
.3 ,
.3 i 575- ,
u.
O o
- j 570-T3 o 8 565-
=
l 560. i i i i i i i i i 5 10 15 20 25 3G 35 40 45 50 55 60 i Time After DAS Actuation, Minutes
F O\
- 8 2
$ J o
e o m #
0 "O
su !
o m
$ C3
= e me
-w "
~ o B
& o C o C
.m 0
o m E -,e x cn ! 5 ; C CD o u O E' w C3 Q "
NC o
C'U o
w w ru cy o a s a
a g ?
m d OU H
m O o
O m a N
w
"" 4 r/>
O U.J o O
z
~
z 3 u <
p 8 d a" m
=
_mo w ! W c e
H e 5 a
E e
- u s SQ O l
' e e
";~
3 l
ll d i l
I l
=
l' -
o o
o m o I I
I
,I m I I o m l
l l
=i l
m o l
m i i i 7 7 H 'taAa 1 pinb!1 pasdettoo .Ag O
O O O
~
OTIS TEST 220899 1.8 RELAPS/M002 1.6- o OTIS DATA 1.4-E 1.2-E i_
.8- hp
- 2-B
- e o_
5 10 15 b0 b5 30 35 k0 IS 50 $5 60 Time After DAS Actuation, Minutes l
1 i
OTIS TEST 220899 620
- ELAPS /M002
' a OTIS DATA 610-E 600-a , ,
{ 590- ,.
% \ ,
, k 580-i--
g 570- s 8
o E 560- N' 8
550 i i i i i i i i i 5 10 15 20 25 30 35 40 45 50 55 60 Time After DAS Actuation, Minutes O O O .
9 '
O CONCLUSIONS--0 TIS TEST 220899 8 RELAPS/ MOD 2 CLOSELY PREDICTED THE LOOP C00LDOWN WITH FEED AND BLEED COOLING.
8 THE SMALL DIFFERENCES BETWEEN THE PREDICTION AND THE TEST DATA CAN BE ATTRIBUTED TO:
1
- 1. THE START OF THE RVVV CYCLING TIME.
- 2. THE TIME AT WHICH THE VOLUMETRIC FLUX OUT OF THE PORV IS EQUAL TO THE EXPANSION OF THE RCS FLUID AND INLET HPI FLOW. THE RELAP CALCULATION IS INFLUENCED BY THE HEIGHT OF THE UPPER PRESSURIZER VOLUME.
O
l OTIS TEST 230299 FUNDED BY TOLEDO EDISON (DAVIS-BESSE)
O O
KEY EVENT DESCRIPTION OF TEST #230299 0
0 SCALED 10 CM2 COLD LEG SUCTION LOW POINT BREAK WHICH IS ISOLATED:AFTER 30 MINUTES.
8 HPI HEAD FLOW IS SCALED TO THE DAVIS BESSE HPI-LPI
" PIGGYBACK MODE".
\
t MANUAL PORV ACTUATION AFTER LEAK ISOLATION TO PROVIDED FEED AND BLEED COOLING.
9 MANUAL PORV ISOLATION AFTER 30 MINUTES TO BEGIN REFILL PHASE.
8 AFTER INITIALIZATION WITH HIGH-ELEVATION AFW, THE AFW IS DIVERTED TO LOW ELEVATION TO MINIMIZE THE SG HEAT REMOVAL UP TO AND INCLUDING THE FEED AND BLEED COOLING PHASE. WHEN THE PORV IS CLOSED, THE HIGH ELEVATION AFW IS UTILIZED TO FILL THE SECONDARY LEVEL TO 12 FEET RESULTING IN A HIGH-ELEVATION BOILER-CONDENSER MODE (BCM) 0F HEAT REMOVAL.
8 LOOP REFILL FOLLOWING BCM.
O
l l
1 0 a ;.k 4G $e/
t
=
N>
950 120-9 Sm P 650 FIGURE 2 OTIS Nodin Diagram for Test 23029 120-6 370 120-10 374 c,34.) '30
) ) f3o-3
- ~0 p p p 3 TUBE 200-7 120-5 140-1 654 150-1 {
16 TUBE 200-6 140-2 654 150-2 200-5 PZr.
120-4 140-3 653 150-2 200-4 l Hot 1
~
140-4 652 150-4 200-2 140-5 630-5 150-5 200-1 act ~\
140-6 630-4 150-6 \ . 20-2 in g 211
\
\
140-7 630-3 150-7
\
\
g 120-1 140-8 630-2 150-8
\
o \ _
140-9 630-1 150-9 Cold Leg 550 95~1 f3-2. 173 x
[ f~e s-f 160-2 fgg , p~ ~c j73 no -f.,
l
~
153 N. g, Q* 100 h55\ _ 160-1 460
- WI 185-1 520-5 LEAK 483 470 l LOM p sp; 185-2 h A fta 490 g-h CORE te5-5 E i 185-6 510-3 f Downcomer n 185-8 -
510-2 510 '
yI l -
- 9 12
Tabla 1. Initial Ccndittens RELAPS/ Test 230299 MOD 2 Planned Actual Core power (1 of full power,11 full power = 24.1 kW), 4.17%
includes 0.5% to replace losses to ambient. 4.2 1 0.1% 4.17 Natural circulation x x x Primary pressure., psia 2204.6 2200 t 50 2200 Pressurizer liquid height, (f t from SGLTSUF) 18.96 16.6 1 2 19.0 Pressurizer main and guard heaters adjusted for an x approximately adiabatic pressurizer x x RVUHV and HLHPV (reactor vessel upper head vent x and hot leg high point vent) closed x x RVVV (reactor vessel vent valve) in automatic x x x (dif ferential-pressure) control with open/close setpoints of 0.25 and 0.125 psid X x x AFW at 100F injected at the upper elevation X
- using the minimum-wetting nozzle x x SG secondary (collapsed) liquid level (with constant level control) 5.66 5i1 5.7 Hot leg fluid temperature, F 610.7 610 1 2 608-610 llPI and leak systems are not yet in use. x Primary non-condensible gas additions are x x not to be tested Initialization is continued until a suitable system steady state is obtained:
Pressurizer metal temperatures, F -
650 1 10 652-681 The RVVV is not cycling x x x The steam generator fluid temperatures are varying x x less than 10F/hr; exception: cycile secondary fluid x temperature variations associated with high AFW injection, and with internal circulation within the secondary liquid pool, are acceptable.
e G # -
_se p I
() SEQUENCE OF EVENTS FOR 230299 TIME,(SEC) ACTION
- 0. OPEN 10 CM2 CLS LEAK.
150. INCREASE SG SECONDARY PRESSURE TO 1350 AND BEGIN GRADUAL DEPRESSUR-IZATION. DIVERT AFW FROM HIGH TO LOW INJECTION CONTROLLING TO 3 FEET.
BEGIN FULL CAPACITY LOW-HEAD HPI FLOW.
180. BEGIN CORE POWER RAMP.
1800. ISOLATE THE CLS LEAK.
1980. OPEN PORV.
)
3650. ACTUATE HIGH ELEVATION AFW CONTROLLING TO 12 FEET. CLOSE PORV.
O
SGI;00 at.01.83 .n D '/ *sm . 8 9 f, LuMKliU: 1G092EG KiB OUUPn C%$M9APJ'W >
OTIS TEST 230299-0B LOW HERD HPl WITH LERK ISOLRTION PRIMARY RND SECONDARY SYSTEM PRESSURES o
8 s
1 1 g
d29 -
0- OPRIM PRES DATR o -o SEC PRES DATR
- t. A PRIM PRESSURE e o spill- over .
.SEC PRESSURE
~4 o l _
m8 O_ - Ch
- Open PORV
~
0 BCM Recover NC O ..0 NC Interr uption }' IBbrQ)n g g
))' o- ,
/ ,,
r-
~
ce Closed CLS ..ea k 01 Pzr .
O' d r g Full 9
' O 00-o' f,
m 'S a .M o_ g w ,y C~ 4700.0 N
'l c \_ - ,
m 9
8 m '
0.0 000.0 1000.0 2700.0 3600.0 4500.0 g TIMg(SEC1 g ,
PLO! 1300 al.ti.25 40 7 8t f.. a335 XE-CC l l CtCLOLE h*O WILC0( C0 DISSPLA 0.0 GT1S TEST 230299-0B LOW RD HPT WITh LERK ISOLRTION HOT LEG,PZR,SG PRIMARY COLLRPSED LEVELS 9
8 1
o -oHL LEVEL DATA a
o -o SGP ZCOL DATR c; _.. A- - A PZR ZCOL DRTR HL ZCOL CRL x x SGP ZCOL CRL c o PZR ZCOL CRL
[ ' o
% ci ~
y mQ g.
o
@~ D
[J (
~
>- O. .O------
0, j[
n _m.
e c -
J O O D' C- > 50'00 0-o a
9 R ..
a o A .A 0- .-s ; a: a ,
^ v .
0.0 900.0 1600.0 2700.0 3600.0 4500.0 TIME [SEC)
i LO: 100 i!.60 2) ed 0 9 6us, son J00-CC I I laiccotcI Fto WILf04 C0 CisFfta 3.0 '
OTIS TEST 230299-DB LOW HERD HPI WITh LERK ISOLRTTON CLPS LERK RND HPi FLOW LBM/S E?
d s
Q o -oRVE LERK DATR r e oLERK FLOW
- 9 m
- t. AHPI FLOW
[c-L_ [ ]
EI O D C3 9 d 2 O x d W '^ y v
c LS Leak LJ -
$ _n w o l
ci /
PORV F g j _.
I S o1 D l S
- d. , a u oc c = .
0.0 900.0 1800.0 2700.0 3600.0 4500.0 g TItg[SEC) g .
l'LO! 1600 al.53 s9 ThuR J AUG. a 335 J00-C00% fABC CX f1ND ElLC04 CO DISSPL fl 3.0 OTIS TEST 230299-DB LOW RD HPI WITH LERK ISOLATION RV COLLRPSED LEVEL 0
[-< l $.
[ O U1 o a a RV ZCOL DATR d* p e o RV ZCOL CRL g; g ni
_a O
U U1 a.
L rn -
G-
_.]
__]
O -O oO C1
]
h m
[Jc .o y - o m
LJ d 1 0
-D' e t m ~ os n- " A Q. 7 W a
9 T'
.0 900.0 1800.0 2700.0 3600.0 1500.0 TIME fSEC)
CONCLUSIONS ll) 0 OTIS TEST #230299 INCLUDED THE FOLLOWING PHASES:
- SINGLE PHASE NATURAL CIRCULA110N
- INTERMITTENT LOOP FLOWS AS THE HL U-BENDS VOID UNTIL CIRCULATION INTERRUPTS COMPLETELY
- STAGNANT LOOP LEAK-HPI COOLING
- FEED AND BLEED COOLING AFTER LEAK ISOLATION
- HIGH-ELEVATION BCM
- LOOP REFILL 0 THE B&W VERSION OF RELAP5/ MOD 2 PROVIDED A GOOD OVERALL PREDICTION OF OTIS TEST #230299, ESPECIALLY CONSIDERING THE SENSITIVE NATURE OF THE TEST REGARDING: llI
- RV UPPER HEAD PHASE SEPARATION
- COLD LEG COUNTER-CURRENT LIQUID FLOW AND MIXING
- STEEP SLOPE AND SHUT-OFF HEAD OF THE LOW HEAD HPI PUMPS O
&-a A a s s- -r-u a .. a _
k a -
t t O
MIST hiOMINAL CASE (310000)
PRE-TEST PREDICTION WITH RELAPS/ MOD 2 VERSION 3.0B O
O
0 i SIMULATED MIST TEST 310000 FOR PMG #10 g 2
8 SCALED 10 CM CLPD LEAK 0 FULL HPI WITH AUTOMATIC THROTTLING S FINAL SYNCHRONIZED SG SECONDARY PRESSURE CONTROL 0 UNCOMPENSATED LOCALIZED HEAT LOSSES MODELLED.
8 CORE POWER AUGMENTED TO OFFSET HEAT LOSSES.
8 RVVV LOSSES TUNED TO ACTUAL HARDWARE TESTS (6.5 TIMES PLANT g IDEAL RESISTANCE).
9 INITIAL PRIMARY PRESSURE ADJUSTED TO OBTAIN 22*F SUBC00 LING IN THE HOT LEGS.
O
10 12 112-1 12 2 FIGURE 2 RELAP5/ MOD 2 MIST LOOP A STIAM 115 1 110-5 N0 DING DIAGRAM IN [ +
115-2 Y
I 680-1 120-1 110 4 HOT LEG A 683 4
SitAM 120-1 tlNE 73 20-2 120-3 i
4 140-1 6S4 150-1 110-3 950l Y -
7 ggg 140-2 653 INLET 1s0-2
+ + Y CORE i SEC FLOOD tau 140-3 652
- 150-3 110-2 8 t
+ -
+ =- ,m 140-4 635 150 4 5" 2" M g
- _ g.- 110-1 I 04 6 30-5 + 107 370 620-3 42 -
370-2 s g95-1 / ,
+ 107 1 360-2 I 370-1 165 170 US 195-3 / #
140-6 630-4 $ . 380 %0 1 175 a 19 415-3 " 105 g -
150-6 105 7 g ,, , y3, g 195-2 ][ 915
--(- -
Y -
Y W; 390-1 3%
355
, 355-3 M ,3 910 +
920
- 10 630-3 150-7 PRts5UR12tR 330 1 350-3
-)- -
l 620-2 -
HEAltD 160-2 330-2 350-2 140-8 630-2 150-E REGION - , __g_-.
3 k 330-3 Y -
I Y ~
e 345 160-1 I40-9 630-1 150-9 3 TUBES 330-4 620-1 340 158-1 335 - I I 65 TO 3;J. NV gp7gggsl l153-2 N M*
- 153-1 153
, gg3,3 ft I5 TO 178,CtA2 10 158. (LAI T UG A2 155-1 155- [ (OLD LtG A1 51144 GENERATOR A O O O
O O O
~
212-1 217
'12-2 210-5 Steam 215-1 kUtN5' E In ;-l 215-2 3,,,,
--+- 720~1 '210-4 Hot Le9 8 FIGURE 3 RELAPS/ MOD 2 MIST LOOP B 78 tine 220 N0 DING DIAGRAM m_ J, zig-2 E20~3 240-1 754 970 250-1 210-3 4 - ; _.
-4 _
ET 240-2 753 250-2
+ -
I -
Y Core 240-3 752 g 5ec. 250-3 210-2 Tube Tank
- Bundle 4 4 Ryyy 4 0; 960 240-4 735 250-4
- -50 K y- 384 4 - _
240 5 730-5 253-5 207-1 370-2 370 ll
~
285
~
207 l -
-M 1 360-2 I 370-1
- 300 380 MO \ 775,3_ y95,3 /f 730-4 250-6 5 M
~
240-6 ya m-e y m -1 925
)
r M -2 ]p
--(-- -
7 Y -
+ 205-1 390-2 390 355 260-3 6
33 280-3 240-7 730-3 250-7 -
7 OI 930 + 320-1 940 350-3 p ]
330 1 Y 4 ; 720-2 _
4 e 9F1 + HPI r 350-2 2W2 3- 280-2 Heated - 10 730-2 250-8 pegion
- CLO 240-8 l'8k 330-3
- 3 4 _
4 _
4 .I 3 Tubes _
. 345 260-1
+
730-1 720-1 250-q 280-1 240-9 340 330-4 335 258 1 278 ic w av I 725 . I 253-2 253-1 Reactor 253 IO 16 fubes . vessel 258 D = Komer 255 To 278. CL82 CLS teak Ta 258. CLBt g ; :: :
Cald Leo 81 Cold leg 82 Steam Generator 8
3 2 -
' 0 0 ,,
Y 8 2 A 1 81 8 A R 7
, 1 I G 5 U 1 I 9 I p O
1 0
1
- 2 I
P m 5 9 H 9
1 9 1
A i
' 2 P ' -
O 5 9
O i L Il .
0 1 2 3 4 .
T O - - - - R S
M E0K R0N OtA Ia 3
g ;0 3
0 ; 0 3 3 0
3 I
M IMA R CrI J
3 3 3 3 V
RM O
C 2G s 3 '2 ND DAI O
5 7
5 7 ; J.
4 MD 1 1 3
/ 5 -
SG 5 3 O P N 7 3 T _
AI 1 0 I L D 1 P E 0 9 H R N 5
0 0 2 7 9 1 1
8 E 5 R 1 J
l -
G I 5 4 0
1 l
A F 6 i j es 0 t 1
1 6 9 G -
- 1 h f t i D
L m M R
O L
1 5 5 1 5 4
- 3 3 1 )
- 6 0 3 7
3 3 2 1 5 - - -
I M v 5 a'5 0 0,i 50
- 3 3
5 3 3 I" 2
- 7 0
6 2 3 yt 0
1 0 7
3 8 3
a %
3 O
0N gI [O lI 1 AG A - EE 0 HR G 9 E 5 3 L 0 1
T O
H 7
0 1 '5 0 5 4 -
3 2 1 1 0 *8 0 l 0 ,a 0 ,' 0 + 7
- 1 1 1 1 1 0 1 1 1 1 1 1
$" R 0
0 t 1 ?
1 4 I
- R 2 1
- u 5 A 2
5 E L 5 4 3 , 1 5 B C 1
- U 1 $ 0 2 5 5
5 5
t a T .
4 1 1 1 1 P 8 3 7 2 b, 4 4 4 4 2 1 2 -
- 3 O 2 1 3
- - 2 3 4 6 7 F 5 T 2 1 1
- - - - - - 9 1 1 0 0 -
0 0 0 0 -
2 2 , 0: 0 ,e 0 : 5 5 5 e' 5 a' 0 l 1 1 5 5 5 1 51 1 1 1 1 1 1
1 2
- 3
-[-
- - .+ 5 - - - ~ _
,51 1
1 1
- - - . 3 5
1 5
5 1
i g 3 2 1
- ' A 0 t 0 '* 0 ,' 0 R 2 5 Wt C 1
0 8
0 2
- 9 F h AI Sh I
g 2
6
?
6 2
6 O
T A
R 6 1 # ,5 1
3 5 E 6 g N 2 1 E 5 4 3 - G 5 - 0 0 3 4 3 2 0 0 0 0 M 9 E 8 y, 5 : 5 , 5 3 6 e 3 3 3 e' 3
,' 6 3
- 1 A 6 6 6 6 6 6 6 6 6 - I 5 T 5 S 1
E g
R L y UO - - - _ - - _ ~ ~ g a$R M g aST AE 3 4 6 8 9 tE N EN 1 2 - - - -
tRO Il St 0 : 0 ,,0 : 0 i a* 0 4
e' 0 ,* 0 l 5PC 4 O
4 4 4 4 4 1 I 1 1 1 e 1 I 1 AI L
C 8.
51 0
1
O O O 210, 737 212-1
'12-2 210-5 215-1 Steam
${$' 790 -I- -I-n mul 215-2
21 -4 '*t L'8 8 FIGURE 3 RELAPS/ MOD 2 MIST LOOP B 5:ea. 780 tene gro N0 DIE DIAGRAM
_n-, -
y zio-2
?I0-3 s
140-1 754 970 250-1 l
)_ - ; _
+ 210-3 753 IntET 250-2 240-2
+ -
I -
Y Core 240-3 752 g 5cc. 250-3 N Tube 210-2 Tank
- Bundle 4 _ ._
4
$ RVVV 240-4 735 250-4 X 960 Y -
7
^
Y 210-1 240-5 730-5 253-5 207-1 370 2 ll 4 370
$'I N "
e 720-3 207 285 9 _ _
4 1 0-2 I 370-1 300 380 360 NN 775,3_ j9g,7 /[
Y 240-6
~
730-4 Y ~
250-6
'I 05 W m -z T m -1 390-2 390 355 260-3 gn 320
-m[) 9 280-3 240-7 730-3 250-7 - -
j M-I 930 + No-1 940 350-3 Y jr 330-I Y
--g.-. -
l 720-2 _
4 a Hpg + HPI Heated - 350-2 260-2 J510 280-2 730-2 2%8 Region m = cto _4_
240-8
\
350-1 M Leak 330-3 Y 4 . p. 4 _
4 5 3 fubes ,
e 345 4
I30-I 720-1 25049 260-1 280-t 240-9 258-1 278 Ill , _
e .NJ, RV 253 1 253-2 Re ac tor 253 16 Tubes yrssel 0" "0*
a 2'i8 To 258. Citt _, - 255
- To 218. CL82
- :-: fLSleek Did leg BI Cold teg 82 Steam Generator 8 w--- = _ +.m - e .- e i .e _
- - . __m m : .. + .. .. , , e- .. _-.
()
RELAP5 CALCULATED OR SPECIFIED INITIAL CONDITIONS MIST TESTS 310000 AND MIST TEST i PARAMETER 320201 340302 1
PRIMARY PRESSURE, PSIA 1700. 2150.*
SECONDARY PRESSURE, PSIA 1010.* 1010.*
- HOT LEG TEMPERATURE, F 588.1 588.9 2
SG EXIT TEMPERATURE, F 549.2 549.2 i
O CORE POWER, BTU /S 122.2* 122.2*
PRESSURIZER LEVEL, FT 2,50* 5.00*
SG SECONDARY LEVEL, FT 5.00* 5.00*
CORE FLOW, LBM/S 2.16 2.15 HOT LEG SUBC00 LING, F 22.0* 57.8 UNCOMPENSATED HEAT LOSSES, 11.0 11.0 BTU /S
- SPECIFIED O
O MIST TEST 310000 SEQUENCE OF EVENTS EVENT TIME (SEC)
LEAK OPENS 0.0 PRESSURIZER LEVEL T0 1.0 FT (INITIATE 56.5 CORE POWER RAMP, HPI, RVVV AUTOMATIC CONTROL, SG SECONDARY REFILL)
H0T LEG A SATURATION 90.0 HOT LEG B SATURATION 425.0 HOT LEG A INTERRUPTION 150, O
SG SECONDARY FILLED TO 31.6 FT 500, HOT LEG B INTERRUPTION 1050.
FEED CYCLE BCM STARTS 2825.
REFILL BEGINS 3075.
CALCULATION TERMINATED 4000.
O
7 1
MIST RELAP5/M002 TEST = 0000-PRE-TEST PRC0lQfION _
i j WITh L_006LlZED HERT LOSSES AND CORE AUGMENTflTION o
n f i -
I of
~-
$ Loop A Loss of,
- sat. cire- o -a PZR PRESSURE !
e rirst o -oSG8 PRESSURE
- interruption complete a a SGB PRESSURE y ,f lcib. " ' -
1 i
- Wg
~
1 d cycle BCM, l %
un a g
- id r
C1 @ ' - - - -
-% r--- g *'
- h] 8- '"
3 p' W '
f S* ; M '" % }
l 3,, . . -
a .s .
i 1 i
\ i t
Cb . -
i j 0.0 '300.0 1600.0 2700.0 3600.0 4500.0 TIME ISEC1 i i
-. . _ . . . , - - - _ _ - _ _ _ _ . ~ . - . . . _ - , .. . . - - . - . -. - - ... .. .. . . . - - , .-
I I l 0
0 0
g 5
~ 4 CA N N ]
O 4
I O _
0 I
T -_ 0 C T 0 I R C a 3 6
D T -
E N R E P M C :.
G T U S R C :.
E T E .
R W
_ 0 E O _ -
C z0 C O _
,0 R i LW 7 P D FO 2 N L C :.
i O A K f' c O H W E O S EI O _
_ s O E LPL _ C .
hF e
i 3 S _
0 S D _
2 v O
_ P4T L/F _
c t
mTES s T C1 C _
ao^
3 :.
- 0. i
,0 4_
r c R 0 T E -- .
$ c6 1 H - V w 2 ooA _ I D P D
Q_l
. l .
c D _
f C :
u a M E c / Z I a o 5 L t C :.
a F - o A A T u
e E L C O Q G : .0 t
R L A -
,0 0
T h _ 9 S T C :.
r e I I _
M W c
~m - c a
~
3_ -
m s
_ n___ _
- h. _ &__ - 4 ,0 : 0 c I
<o g eg .
9g eg t
gd wr r-
- w. s rm1_ z] u. - Q. E. C'1ZG MCJ
(( 3_ e ra m
c 1
muk inuoi to ' oitna 5.c ^ " ' =
^
kosn&c cuo u wi s m. E m-tia
~~
~~ - ^
MIST RELRPS/ MOD 2 TEST 0000 PRE-TEST PREDICTION WITh LOCRLIZED HERT LOSSES AND CORE AUGMENTATION 9
M o y l- l o aPZR COL LEVEL o-o RV COL LEVEL g c, a -^DC COL LEVEL
, . .'6 .,- _
' ~
w d
m J
v W@ 44.LW s o m e " C"_
aR -
VW -
9- _
n_ Top of Core ,_,
e d
a o
cr 9 y O~
a Z
C o ;
o y _. -.
tr s
, t 9
a c a c-e- c n n c 2 n e a e o a 9.0 900.0 1800.0 2700.0 3600.0 4500.0 TIME [SEC)
but <w C 1. '29. zo Ptw 6 fW, 49M JOB- CT II CHCLGLE F10 i lt_t0* CD 0157vt 2.9 J
)
MIST RELRPS/M002 TEST n310_000 FRE-TEST PREDICfl0N j i
HITh LOCALI_Z_ED HERT LOSSES RND CORE RUGME_NTATION STERM GENERATOR R COLLRPSED LEVELS ;
o !
g.-
i i
n, o m SGR PRIM LEVEL l 9 e o SGR SEC LEVEL i i
e>
f, Top of Tubes ._,
E
_J O
- o. -.._ _ -_ _J K .-
Nm* !
l W
l3" in ,
l 1,
95% !
n:
j ilh C O -- 0 0 0 0 *+CWF, W Level * {
d O
o l
l, f
c 9
o_. , ,
0.0 900.0 1800.0 2700.0 3600.0 4500.0 O TiMetSEci e.
"scor Tr6c c u i.o. m s ea, .m 2a-cw men mo aua co oisv u 9.2 MIST _RELRP5/ MOD 2 TEST 10000 PRE-TEST PREDICTION WITh LOC 6L{ ZED HEAT LDSSES RND CORE RUGMENTATION 9
g --__- _
t-.
$o '
"]jb -q o o SGB PRIM LEVEL 0--e SGB SEC LEVEL LJ .. -
T - -
4 Top of Tubes --
a-o ,_ _
\
y_
i 95%
N M #6MM Level ~
w ft' (4J Z
N - --. -. - - - - . - - . ...- ~--- - - _ - - _
T IJ o
9 o - __
0.0 900.0 1600.0 2700.0 3600.0 4500.0 TIME ISEC)
T;., ,a w. a u. o v. s n, ,vs a-con ,nua na ca.u us. on w u, 3.a MIST RELRPS/M002 TEST v310000 PRE-TEST PREDICTION WITh LOCALIZED HERT LOSSES RND CORE RUGMENTRT10N
<a i[
u, s
r cgj__ _ _
W o
__]
L.
a 'j - i- u ku d' a c -e' c' ' c t- -c-d- c h ' #^ t ^ - e- '
c' - e z
L]
cn a
co L1 w
_J __. _ _ _ _ _
g s~
o o r eHLR UBEND FLOW o- oHLB UBEND FLOW 1__ _ _ - --
T 9
7 -
0.0 900.0 1000.0 2700.0 3600.0 4500.0
$ Tli$[SEC1 g ,
i i
i O
b J
MIST TEST 310000 CONCLUSIONS i
0 OVERALL RESULTS OF CURRENT PREDICTION ARE CONSISTANT WITH .
t i
THE REDBL5 DESIGN VERIFICATION ANALYSIS.
! O THE CURRENT PREDICTION CONTAINS LARGER LOOP ASYMMETRIES. :
f E
l
!O i
k f
1 J
l n
i O
t ;
i
8 9 O
MIST TEST 320201 PRE-TEST PREDICTION WITH RELAP5/ MOD 2 VERSION 3.0B e
i l
l 9
r O
MIST 320201 TEST DESCRIPTION i -
2 0 SCALED 50 CM CLPD LEAK 0 EVALUATION MODEL (EM) HPI FLOW i e SYNCHRONIZED SG SECONDARY PRESSURE CONTROL i
1 0 UNCOMPENSATED LOCALIZED HEAT LOSSES MODELED 3
0 CORE POWER AUGMENTED TO 0FFSET HEAT LOSSES i O O CALCULATION PERFORMED BEYOND MINIMUM CORE INVENTORY i
i i
1 i
t 0
l 4
O MIST TEST 320201 SEQUENCE OF EVENTS EVENT TIME (SEC)
LEAK OPENS 0.0 PRESSURIZER LEVEL TO 1.0 FT (INITIATE CORE 13.1 POWER RAMP, HPI, RVVV AUTOMATIC CONTROL, SG SECONDARY REFILL)
HOT LEG A SATURATION 35.0 HOT LEG B SATURATION 40.0 HOT LEG A INTERRUPTION 65.0 (gg HOT LEG B INTERRUPTION 85.0 HIGH ELEVATION BCM BEGINS 160.
COMPLETE LOSS OF NATURAL CIRCULATION 250.
SG SECONDARY LEVEL TO 31.6 FT 450.
POOL BCM STARTS 1425.
REFILL BEGINS (WITH LEAK HIGHLY VOIDED) 2200.
CALCULATION TERMINATED 3600.
O
"-- ~
kor iac ~ .7.c2.o ro,4 s 2, , ion ;oe-ccv: caecocx ,.e :ittor co cissets s.o-O O MIST RELAPS/ MOD 2 TEST n320201 PRE-TEST PREDICTION O ~
WITh LOCRLIZED HERT LOSSES AND CORE RUGMENTATION--EM HP_I 9
8-8 c' o 8-e cc complete o o PZR PRESSURE G Loss of Nat. C o SGR PRESSURE c' 9
/ cire.
A aSGB PRESSURE g y- CFT PRESSURE g _
m ' Pool
] BCM ff c, M
5 N~ .
cer riow f-*
'n 4 Begins
_. J N g , , , . - - -
g Sec. Level 0 95% _
s-i 0.0 000.0 1600.0 2400.0 3200.0 4000.0 TIME (SEC1
f o.
o O
a I
ZZ OO
_J J
-, L L-O
'i QQ <
f zz z r MU U o y mm ,
, Da a.
s i, o Z (1 W
- JJ n O a
- Er
% H OO 1 c H Ci s
m y' r
ao :
M c_o ,
7 o
D c
0 -
o c;
I a_ o i ?a z
a U
b o
N g m I G
-i ui m < r
~
r l y . : i
$ $ m qi l, 9H c LJ m , o o
=
H o *
- l
_) <
8 b,
S m
. O c S O y u
a r r- i s -
g)
! W o <
9
[
_J d
- 3 M _j G: 9
'
- C 8 H o -
w
@ o s J Y r c' Y
$ p ,
2 3
I JI
-2 !
. X . _ ._ , 7d l g
( ! .
e , b- 't i 6 i .
'y g 9'I 9'C C'C 9'0- 9'!- g E
S/W91 SM0lJ GN]Sn S31 104 W p
-4 ci D
-- a - - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - _ . ._ J
61 m a .ru.30 a s . .oas m-ccu amu .nuax to cissn.as.s O O MIST.RELRP5/ MOD 2 TEST 9520201 PRE-TEST PREDICTION
~O WITH LOCALIZED HERT LOSSES R_ND CORE RUGMENTATION--EM HPI l l
~ !
2 i i
m o
u, x l 1 r m !
J c oCLPD LERK FLOW i.
z o- o1/4 HPI FLOW jl 3"'g_ _ _ _ _.
^ ^CFT FLOW ,
Q- ,
c '
o z ,
Tm y ci-x LJ
_J o
Q- Total IIPI
_J _ _ ___ . ___ pM +, E c; --
O e O'l,*ik'o b1/#u'd L*b."-wr u g g g g { g g g g
~ !
l Y- i i
0.0 000.0 1600.0 2400.0 3200.0 4000.0 TIME [SEC1
++++-m -=m 9
o I h I J o o oJJ - U
>uu N 4 No u>> j o, o Joo e I JJ c.
J 0 z r a1J e a a U Uoa o p i ao -
8 Z a
. N>O exo a m a
W x [C QO < }
ct H < a g z OO< -
g kl U E i- C D a b C 9 cc y 8
' x n a \- a -
U U O y m
$ o
$n b <
- a e,
m ,
u 5 g M S D m a o. H e w U 8 B o g e J -
o a a m H
- O [
e 3
a U i
r I' N ( -.
g ya % l
! b
$ U b c
-p G
_a R
H O \
\ 8 I'
U m 3 ' G f
r 3 9 H
'1 C
I
' I i
< c', b i i : Jli d
- O'5E 0*SE 0*12 0 "r l O' 4 0'C O'4-e E
IJ S13A3103SdBTl03 BZd GNb 30 A8 O S
a e- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ .__ . _ _ _ -
% u. .. n.o uus 8 n. ass man ume. a maux us un n n u.u '
O o O -
MIST RELRPS/M002 TEST dYl0201 PRE-TEST PREDICTION -
WITH LOCALIZED HERT LOSSES AND CORE RUGMENTATION--EM HPI STERM GENERATOR R COLLRPSED LEVELS 9
8 4
a p- A o oSGR PRIM LEVEL -
e o SGR SEC LEVEL b Top of --*
g Tubes u
9 f
ma t
g a
= = -
m :L o
9 g__ - _ _ _ _ _
c 9
0 0.0 000.0 1600.0 2400.0 3200.0 1000.0 TIME (SEC)
na, usu a v..v u, , ,m . , u. nu- ex u,ws m ,wsr uu ..sas.>
MIST RELRPS/M002 TEST w320201 PRE-TEST PREDICTION WITH LOCALIZED HEAT LOSSES AND CORE RUGMENTRTION---EM HPI R
8 s
k l
<n
>8- o oSGB PRIM LEVEL M o o SGB SEC LEVEL LJ Top of Tubes v3 ct (C
o o --.
o O O 'hqq L 1_3 H
CE cr kJ z
kl o (D g _ ___ _ _ _ _ _ _ _ _ _ _ _
r4 (I-kJ
(
9 o- ,
0.0 000.0 1600.0 2400.0 3200.0 4000.0 g TItgISEC) g ,
r O
MIST TEST 320201 CONCLUSIONS 8 CORE IS ALWAYS COVERED AND COOLED.
8 AT APPR0XIMATELY 1800 SECONDS, 3/4 0F THE HPI IS CAPABLE OF ABSORBING THE CORE DECAY HEAT IN A SINGLE PASS BOILER MODE.
8 N0 UNEXPECTED RESULTS.
l t
l lO i
i
! l
l el MIST TEST 340302 PRE-TEST PREDICTION WITH RELAP5/ MOD 2 VE_SION R 3.0B e:
6
-4 -- m - _ m._ -- .aaa Ap._ -w - -,---. -_---+4 - . m j
MIST 340302 TEST DESCRIPTION 0 SCALED 10-TUBE DOUBLE-ENDED RUPTURE AT THE BOTTOM OF SG-B 6 UNCOMPENSATED LOCALIZED HEAT LOSSES MODELED O CORE POWER AUGMENTED TO 0FFSET UNCOMPENSATED HEAT LOSSES
! O FULL HPI CAPACITY e AUTOMATIC STEAM VALVE CONTROLLER IN THE AFFECTED SG TO
(~)
, ATTEMPT TO MAINTAIN THE LEVEL BETWEEN 10 AND 20 FEET 0 NO AUXILIARY FEEDWATER- TO BROKEN SG AFTER LEAK INITIATION l
O AUTOMATIC 100 F/HR C00LDOWN FOR SG-A
1 MIST TEST 340302 SEQUENCE OF Ei'ENTS llk EVENT TIME (SEC)
LEAK OPENS (BEGIN CORE POWER DECAY RAMP, PLACE 0.0 RVVV ON AUTOMATIC, TERMINATE AFW TO SG-B)
CORE EXIT SUBC00 LING 550 F (INITIATE FULL HPI, 12.0 SG-A SECONDARY LEVEL TO 31.6 FT, 100 F/HR SECONDARY C00LDOWN)
HOT LEG A SATURATION 85.0 SG-B SECONDARY LEVEL 220 FT (START FULL STEAM 105.
FLOW TO CONTROL LEVEL)
HOT LEG A INTERRUPTION 130. (l)
HOT LEG B SATURATION 185.
HOT LEG B INTERRUPTION 265.
SG-A SECONDARY FILLED TO 31.6 FT 465.
COMPLETE LOSS OF NATURAL CIRCULATION 760.
HLA EMPTIES 1400.
l REFILL BEGINS ~ 2650.
CALCULATION TERMINATED 2700.
4
brsoc c2.un us 2 m, im m-coa. saecoc< = accox co o us,ta 2.2 O O MIST RELRP5/M002 TEST x540302 PRE-TEST PREDICTION O :
SGIR WITH LOCRL12ED HERT LOSSES BND CORE auGt1ENTRTION 9
8
- G o
O.
8-1,oss of Nat. Circ-in Loop ^
$; / -- a o PZR pggsSUR -
First Loss osga PRESS 0R
" of Nat. Circ. ^~" sgg PRESSURE h in LOOP O ' CFT pggssuRE
% o. - -
h 8' til 3 Complete loss f 8 of Nat. Circ
- g_A Reverse bo Ileat Transfer CFT FLOW t5 g y
yo ' ^ "Begins k
.. %- % w u
A A SG-A filled A o to 31.6 f A A g, c_. --
9 o- i 0.0 s00.0 1200.0 1800.0 2400.0 Tliig tSEC1
M I 13C C3.">3 a7 Id.S 7 Site, 60 % J0(rCDit urcLuLa tro c31Lt0< CO OtY#LH 9.0 MIST RELRP5/M002 TEST n340302 PRE-TEST PREDICTION SGTR WITh LOCALIZED HERT LOSSES AND CORE RUGMENTRTION U)
O g _ _ _ _ _ _ .. _
o 1 0 o eSGTR LERK FLOW o o 1/4 HPI FLOW EJ ^ CFT FLOW z
CD g.
.J .__
In O G.
9
~_ g _ _ _ _ _ _ _ _ _ _ _ _ _
.N
~ '
a-g- Total IIPI L]
x H
^ ^ ^ ^ ^ ^
O 3 D o
m O
O w ha mwL t.
?- i 2400.0 3000.0 0.0 600.0 1200.0 1600.0
$ Titg[SEC1 g }
tai m. u.u a u, ,m.., m.m. _ _s.. ..m m o. m ,2.2 MIST _RELRPS/M002 TEST n 0302 PRE-TEST _F_REDICTION .
SGTR WITH LOCALIZED HERT LOSSES AND CORE RUGMENTATION
,7 _ _ _ . _
\
.s __ _
o P f q e ,
i s S-- f I 9 k bi _,. . . -
_Jl J I l .
i m
c3 0 - . _ .. _ ..
D e
o s J a- '
$3 r o--a hLR UBEND FLOW o o HLB UBEND FLOW
. =-
?- i 0.0 fiOO.0 1200.0 1000.0 2400.0 3000.0 TIME fSEC1
ha in C,'5.5 TES 7 we, a3M JOB-CDit uht/_'Ax wie citr ut LO G!$'#LR J.0 MIST RELAPS/ MOD 2 TEST #340302 PRE-TEST PREDICTION SGTR HITh LOCALIZED HERT LOSSES AND CORE RUGMENTRTION 9
g oo c OPZR COL LEVEL 3 o o RV COL LEVEL -
,_. 10 -
' ^ ^DC COL LEVEL w
IIL Nozzle
@"g_
' ^
rN m-m=qg 2'"" %
u A~ Top of
- a. W rwM4= . ::c
= % core ~
59 g ,. . _ _ _ _ ._ _
O O
Cr N
EL L3
( _ - -- - -
L;l O
C3 O C o O O O O O O O O O
E~ >
0.0 600.0 1200.0 1600.0 2400.0 3000.0 9 ri @ sECi g i
-- - - m
u , ,., . n . . , .n . . . - -- sa e c. .. - ~o v. o . - m MIST RELRP5/M00.2 TEST b O302 PRE-TEST PREDICTION O- "
SGTR WITh LOC 6LIZED hERT LOSSES AND CORE RUGMENTATION STERM GENERATOR R COLLRPSED LEVELS o.
8
[
9 o-o oSGB PRIM LEVEL g c>-e SG6 SEC LEVEL t_
Top of Tubes _.
t!
8 J9 g o, __
95%
C g
N + hh Levei_
O e
h (
g . __ _ _ _ _._ ..____.
9 c .
0.0 600.0 1200.0 1000.0 2400.0 3000.0 TIME ISEC1
Ri.iz =<m.a ns r #u, .us a-m i,- us &wa co mew 2.2 MIST RELRPS/M002 TEST v340302 PRE-TEST PREDICTION SGFR WITh LOCRLIZED HERT LOSSES RND CORE RUGNENTRTION 9
8 s
k U W
_.] o
\\
lJ cf .. - . -
k c _.
El Top of
] _- O c RA c_ <%. .= c - -
.m Tubes -a w
CL T
_1 do ,% y W -+
u o_ -
O Two-Phase Cr Flow in S
c steam Cr Line 4 _ _ _ _ .
u z
tj o a oSGB PRIM LEVEL ^~
g-- -
o--e SGB SEC LEVEL r
C Steam Valve -- Desired Col.
fd Ful1 Open Leve1 Band m m 9
o-.
- 0.0 600.0 1200.0 1600.0 2400.0 30b0.0 ~
$ TIrgISEc1 g ;
. 7, o o N
MIST TEST 340302 CONCLUSIONS 9 CORE IS ALWAYS ADEQUATELY COVERED AND COOLED DESPITE THE LARGE INVENTORY LOSS, 9 THE HOT LEG LEVEL ASYMMETRY IS LARGER THAN EXPECTED, HOWEVER, THE BROKEN LOOP HEAT TRANSFER (STEAM CONDENSATION)
PRODUCES A CONSIDERABLE DRIVING FORCE,
/
i e RELAPS/ MOD 2 IST ANALYSIS TASKS (CONTINUED)
II. O FUNDED BY MIST PROGRAM
- MIST TEST 310000--THE NOMINAL SCALED 10 CM2 CLPD LEAK USING FULL HPI CAPACITY AND l
INCLUDING LEAK-HPI COOLING AND POOL BCM
- MIST TEST 320201--A SCALED 50 CM2 CLPD LEAK USING " EVALUATION MODEL" HPI CAPACITY AND INCLUDING LEAK-HPI COOLING, HIGH ELEVATION BCM, AND P00L BCM O - MIST TEST 340302--A SCALED DOUBLE-ENDED 10-TUBE SGTR LOW IN THE SG USING FULL HPI CAPACITY O
F R
?
}
B&W RELAP5/ MOD 2_VERSICfl 3.08 ORIGIN: RELAP5/ MOD 2 CYCLE 36.0 UPDATES: CYCLE 36.1 ADDITIONS:
- 1. HIGH ELEVATION AFW WETTING AND HEAT TRANSFER MODEL j
- 2. A NONCONDENSIBLE GAS HEAT TRANSFER DEGRADATION MODEL O CHANGES:
- 1. A DISCONTINUITY BETWEEN THE SUBC00 LED
~
LIQUID FORCED AND NATURAL CONVECTI0N HEAT TRANSFER COEFFICIENTS WAS REMOVED.
- 2. THE VELOCITY FLIP-FLOP TIME STEP REDUCTION CONTROL WAS MODIFIED.
I
! o
>