ML20054J527
| ML20054J527 | |
| Person / Time | |
|---|---|
| Site: | Clinch River |
| Issue date: | 06/25/1982 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1108, NUDOCS 8206290177 | |
| Download: ML20054J527 (400) | |
Text
. , . .
ICCI. EAR REGUI.ATORY COM!CSSICN
> . 3,.
ey nQC lI 1 b
'l g d -}-. . ..
ba:Y;z : T.h .
i- ,.-
In Pe Mat:::er cf: ADVISORY COMMITTEE ON REACTOR SAFEGOARDS .
SUBCOMMITTEE ON CLINCH RIVER BREEDER . .' 3 [I
~-
REACTOR / SITE EVALUATION ...t .
s .-
g...
e: f'
=.
_ ,.- 8
. w,_ . , "-:
m June 25, 1982 'paggg: 342 - 588 ;*. ' I, ..
AT: Washington, D. C. . . [c :-
y ,
n .
ft .
Y( ., ,
g*, x, k.p1
] ,
- ) F 4- .
,,,,,, , 3 ,.
, s a ..
'- t v:
400 Vi_gisia Ave., S.W. Washing ==, D. C. 20024 .- .
s.f4 Talaphena: (202) 554-2345 fl. .,.,
b 8206290177 82062S b [;-
PDR ADOCh 05000 S37
.r- k[? W f T PDP
342 O ' UxITED STATES or A"ERrCA 2 NUCLEAR REGULATORY COMMISSION O 4 ADVISORY COMMITIEE ON REACTOR SArEGUARDS S
SUBCOMMITTEE ON CLINCH RIVER BREEDER 6
REACTOR / SITE EVALUATION 7
Room 1046 8 1717 H Street, N.W.
Washington, D.C.
9 F rid a y, June 25, 1982 10 The Subcommittee met, pursuant to notice, at 11 8:30 a.m., MAI CARBON (Chairman of the Subcommittee) 12 presiding.
13 O m
""tst"T-ACRS MEMBERS:
15 MAX CARBON, Chairman 16 MYE3 BENDER, Member WILLIAM KERR, Member 17 J. C. MARK, Member JESSE C. EBERSOLE, Member 18 ACRS CONSULTANTS:
19 R. HOSKER 20 W. KASTENBERG W. LIPINSKI 21 M. TRIFUNAC Z. ZUDANS 22 DESIGMATED FEDERAL EMPLOYEE, 23 PAUL BOEHNERT O 24 ACRS STArr.
25 ALDEN N. BICE O
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
343 O ' 81tt =^tozarcz
, 2 ALSO PRESENTa 3 R. STARK O 4 P. DICKSON C. CLARE D. UJIFUSA 5 N. KAUSHA C. BOASSO 6 R. HILLIARD HR. BURKHART 7
8 9
10 11 12 13 O ,,
15 16 17 18 19 20 21 22 23 O u 25 O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554-2345
344 O >aoeeto1"cs 2 HR. CAEBON: The meeting will now come to 3 order. This is a continuation of t:1e Advisory Committee 4 on Reactor Safeguards, Subcommittee on Joint CRBR site 5 evaluation.
6 My came is Carbon, the subcommittee chairman.
7 We will proceed with the meeting, and I will call on Mr.
8 Ujifusa of DOE.
9 MR. UJIFUSAs Good morning. I am Dan Ujifusa 10 from the Department of Energy. Mr. Longenecker could 11 not be here this morning and asked that I introduce 12 today's presentation.
13 We have on the agenda discussion of some of 14 the outstanding questions that the Cor.aittee and the 15 Consultants have raised in previous meetings.
16 First of all, Dr. Paul Dickson of 17 Westinghousa, the technical director for the project, 18 will discuss the CRBR design basis leak in a steam 19 generator and what we expect to be the response to 20 sodium-water reactions.
21 Then Mr. George Clare of Westinghouse will 1
i 22 discuss the plant response to station blackout, both 23 on-site and off-site.
() 24 In the afternoon, we will.have presentations 25 on liquid metal fires, both sodiur as well as NaC fires, O
l l ALDERSON REPORTING COMPANY,INC, t 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
345
() 1 and we have split this up into three parts.
2 First, Mr. Robert Hilliard from the Hanford 3 Engineering Development Laboratory will discuss the 4 phenomenology of both sodium spray and pool fires.
5 Second, Mr. Cliff Boasso of Westinghouse vill 6 discuss some of the engineered safety features, such as 7 cell liners in the containment and catch pans in the 8 steam generator program which we believe will minimize 9 the impact of these liquid metal fires.
10 Finally, Mr. Boasco will then summarize the 11 statements that are being made in the PSER in Chapter 12 15, the accident analysis, which gives an evaluation or 13 analysis of our sodium fires.
O 14 In addition, I think Dr. Ninu Kausha, who is 15 also sitting in the front row, who is the assistant 16 director for engineering in the project, will answer any 17 sdditional questions that have been asked.
18 That pretty much summarizes today 's 19 presentatons. Are there any questions?
20 MR. CARBONS Any questions of Dan? No, then 21 let's proceed right on.
22 MR. DICKSON: If you compare the title of that 23 slide, which is Large Sodium / Water Reaction Design
() 24 Basis, with the subject on the agenda you will see a 25 slight difference. I was not at the meeting at which O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
l 346 l
() 1 this question was asked and I thought the question was 2 how do we form our large sodium / water reactor design 3 basis.
4 I found from both the agenda and the questions j 5 yesterday that you are probably more interested in how 6 ve deal with all sorts of potential sodium / water 7 reactions in the steam generator. So, I will try to 8 touch on that subject, realizing that I do not have 9 slides to cover all of it and that I have to recall a 10 lot of numbers from memory. But I will do the best I 11 can.
12 I am sure all of you know what a steam 13 generator looks like, but that is it. Water enters from 14 the bottom, exits from the top after going through two 15 sheets of some 750 tubes, and sodium enters down this 16 path, up near the bottom of thL two sheets and out 17 through that path.
18 MR. CARBON: This shows the steam generator at 19 a right angle there at the top. I had been under the 20 impression it had a 45-degree angle. Is this correct?
21 MR. DICKSON: This is correct, it is a 22 90-degree angle.
23 MR. MARK: Could you help me? The tubes, the A
(_) 24 sodium is coming from the top down at the botton?
25 MR. DICKSON: Yes.
O ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554-2345
347 I
() 1 MR. MARK Water comes from the bottom up, 2 steam to the top.
3 MR. DICKSON: That is correct. l
)
4 MR. MARKS In what way are those tubes 5 different from the steam generators that apply on light 6 water reactors, are they thicker walls, thinner walls, 7 better steel, or what?
8 HR. DICKSON: I really do not know enough 9 about the details of light water reactor steam 10 generators other than the fact that the wavings are 11 welded. It is a change from the standard practice and 12 was done based on the experience of the number of leaks 13 one gets in the normal weld. I do not have a slide on O 14 it, but there are several ways of connecting tubes to 15 tube sheets.
16 One way is to bring the tube clear through the 17 tube sheet and you veld at the top. That clearly is out 18 for sodium because you have a crevice then formed where 19 sodium and any resction product can catch and cause 20 corrosion.
21 Ea rlier, in some British steam generators, 22 they simply welded at the bottom with a hole drilled 23 through the tube sheet and did experience some leaks'
( 24 with it. The problem, as we saw it, was that it was 25 difficult to inspect these welds to be certain they were
(
ALDERSON REPORTING COMPANY,INC, i 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 1
348
() 1 correct.
2 Our tube sheet is made with nipples, a 3 position for every tube so that the end of the tube is
[}
4 in effect an integral part of the tube sheet and a 5 butt-weld is made with the tube and that nipple extended 6 down. That butt-weld can be thoroughly inspected.
7 MR. CARBON Excuse me, would you mind drawing 8 that on the board?
9 MR. MARK: I was going to say that this leaves 10 me a little bit unclear. Do we have experience with it 11 being handled and done in the way you are going to tell 12 us, if it works or does not work?
13 MR. DICKSON: Yes, we have the MSG that was O 14 built that.way, and was tested. Now, it.was not tested 15 for a 30-year life. But there have been other 16 experiences with ra ther long lives. I do not know of 17 any specific case where this exact steam generator 18 configuration would make the claim that it has a 30-year 19 experience.
20 MR. MARKS I mean, there is some experience in 21 Britains there is some experience in France, and there 22 is some experience here. I presume you are taking 23 advantage of that to the extent possible.
24 MR. DICKSON: Absolutely. This is, 25 incidentally, in our IHX the tube goes through the tube O
ALDERSON REPORTING COMPANY, INC, 400 VIRGINtA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
349 O ' =aeet eaa t vet $ea et tae too-2 MR. CARBON: That is what the IHXs look like.
3 MR. DICKSON: That is correct.
4 In the case of the British experience the tube 5 was butted as shown there against the tube sheet that 6 had a hole drilled through it, and welded at that 7 point. There is a certain amount of inspection you can 6 do there, ultransonic and what not; but you cannot get a 9 complete radiograph.
10 HR. CARBON: Is that what they had on BFR 11 where they had all kinds of problems?
12 NR. DICKSON: Yes, sir. In our case, and I 13 believe others are adopting it now too - I do not say O 14 there is anything unique at this juncture, it is just 15 that the state-of-the-art is advancing - the tube sheet 16 is fabricated with an integral portion of it forming the 17 end of what becomes the tube after the weld is 18 completed. The tube is then butted up.against it and 19 welded.
20 Clearly, that is a more expensive operation 21 both to perform this and to perform the butting and 22 velding because it takes a coniderable amount of 23 alignment.
24 However, you can do a radiograph inspection as 25 well as visual, ultrasonic, dipenetrant. So, you can O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
._ - _ _ _ _ _ _ _ _ _ _ _ _ _ . )
350 h I have a thoroughly inspected weld and that is very l
l 2 clearly tha major requirement.
1 gg 3 HR. ZUDANS: These are single pipes, not 4 double.
5 MR. DICKSON: That is correct.
6 MR. ZUDANS: Now, in EBR-2 they are double 7 vall pipes and they are welded similar to this one here, 8 except that they did not machine its nipple, they formed 9 it by a special process in welding.
10 MR. DICKSON: Yes.
11 MR. ZUDANS: And that is the experience that 12 is available.
13 HR. DICKSON: The EBR-2 is quite successf ul.
14 MR. ZUDANSs Almost twenty years, but it is a 15 double-wall pipe.
16 MR. DICKSONa Yes.
17 HR. MARKS Was this approach then, the notion 18 of randoen and wild leaks between the secondary sodium 19 and the steam as much held down?
20 MR. DICKSON: Yes, sir. That does not mean 21 that we are going to claim that we have a completely 22 leak-free cystem for the entire life of the plant.
23 MR. ZUDANS: All of these tubes bend in that I
24 dog-leg over.
25 MR. DICKSON: That is correct.
It ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
351 0 1 MR. ZUDANSs One tube sheet to another at 90 2 degrees.
3 MR. DICKSONs That is correct.
4 MR. ZUDANS That is not obvious from that
~
5 picture, though.
6 MR. DICKSON: It is from this one, I believe.
7 MR. ZUDANS: Oh, yes.
8 MR. BENDERS Just to get things clarified a 9 little bit I think there are two points that we could 10 develop a little better. The tube sheet velds have been 11 the probles, I think, in the steam generator today. But 12 part of the questions that need to be addressed are the 13 question of tube failures in other places in that tube O 14 sheet.
15 Are we saying that the potential for f ailure 16 in other places is somewhere near zero; is it lesser or 17 greater than in water-cooled reactors?
18 MR. DICKSONs I believe it is lesser. I base 19 that on my belief that sodium is far less corrosive 20 substance than water. In our allowance for the tube 21 thinning - and we do anticipate it to decrease in 22 thickness - it starts at the beginning of life at .109 23 inches. He have an end-of-life allowance and that is 24 where it was analyzed for in the worst cases at .077 25 inches.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE S.W., WASHINGTON, D.C. 20024 (202) 554-2345
352
() 1 The majority of that decrease in thickness is 2 on the water side. The sodium was tot very damaging to 3 the tubes.
)
4 nn. BENDER: Well, I think that is also true 5 in the water-cooled system. But most of the problems 6 have turned out to be on the secondary side of the 7 system.
8 MR. DICKSON: That is where the steam is, 9 which is more erosive than water.
10 HR. BENDER: Well, the point I am trying to 11 make really is, it would be worthwhile to know what 12 sechanisms for failure have to be identified and how you 13 are dealing with them. Whether they are more or less is O 14 not really all that important. But we need to know what 15 the modes migh t be.
16 HR. DICKSONa We have done whatever we can to 17 minimize the number of leaks. Ve do anticipate that 18 there may be leaks and they may be either a t the welds 19 or at some flaw.
20 The experience is, based upon other 21 sodium / water steam generators, the majority of the leaks 22 do occur at the welds. But wherever they occur, yes.
23 HR. DOCHERTYa Pa t Docherty from Westinghouse.
24 An important distinction is that most LWR 25 experience in tube failure that is related to corrosion O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
353
() 1 resulted from a deposit of corrosion products in the 2 pool boiling situation on the outside of the tube.
3 In this situation you have boiling on the 4 inside of the tube with higher velocity. So, the 5 expectation is that those higher velocities that are 6 maintained inside the tube would prevent the build-up o f 7 corrosion products and you would not have the kind of .
8 corrosion-in-the-tube problems that you see in LWRs.
9 HR. DICKSON: You simply will not have to corrosion on the outside of the tubes like you did in 11 the water.
12 MR. DOCHERTY: The distinction is that the 13 boiling occurs on the inside of the tube in CRBR with 14 higher velocity, you have more of a pool-boiling type of 15 situation on the outside of the tube in LWRs.
16 HR. KERRs It is worth remembering, however, 17 that much of this corrosion that has been observed in 18 LWRs was not anticipated.
19 HR. DOCHERTYs That is correct.
20 MR. MARKS Is the pattern that you described 21 for us similar or different from that? That, I.believe, 22 has been rather successfully used in EBR-2?
23 3R. DICKSON: It is different in the EBR-2,
() 24 the double wall steam generator.
25 NR. MARK: So, it is better protected?
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
354 O "a otcxso". 'e=-
2 HR. MARK And I believe they have not had a 3 lot of probleras.
4 HR. DICKSON: That is correct.
5 MR. MARK And you would expect that there 6 would be less problems.
7 MR. DICKSONs We beieve so. But that does not 8 sean we do not allow for leaks, as I am going to get 9 into next.
10 MR. BENDER: What is the composition of the 11 tubing?
12 HR. DICKSON: Three and a quarter Chrome-Moly.
13 MR. BENDER: How much experience do we have on 14 that?
15 HR. DICKSON It is a normal alloy, so there 16 is a considerable amount of experience with that 17 particular metal. In that particular application it is 18 one of the two standard alloys used in sodium / water 19 steam generators. It has been used by the English, the 20 French, the Russians - the Russians use something a 21 little different.
22 MR. BENDER: And how many operating hours can 23 we claim for that which we are talking about? Well, O 24 den.,try to guess.
25 MR. DICKSON: I can't venture a quess.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., $.W., WASHINGTON, D.C. 20024 (202) 554-2345
355 O ' na. azaota: r tata* en t xiaa or i rorm tioa 2 would eventually be helpful, I think, in trying to 3 develop whatever case l's being made for its integrity.
4 I am sure this question will not die just because we 5 have stopped asking questions here, other people will 6 ask.
7 MR. DICKSON: Yes, and we have that answer, 8 although I do not have it with me as to how many 9 operating hours. But we are not trying to make the 10 claim that we will never get a leak in any way, that we 11 know how to prevent them en tirely.
12 MR. BENDER: The point I have tried to develop 13 earlier, and I will try once more just to put it in O 14 con te xt . It is not just the tube sheet we are asking 15 about. The whole tube has a potential to leak. So far 18 the problems have been with tube sheets and you probably 17 engineered your way out of them - we hope you have.
18 But we still have to look at a period of 19 operation that is 20 to 30 years and the potential has 20 to be considered over that length of time.
21 MB. DICKSONa Yes, sirs that is correct. But 22 let me note, we start out with a thick wall, 109 mills.
23 It does reduce over time. End of life is on the order i 24 of 77 mills. We plan to inspect the tubes throughout 25 the life of this plant periodically so that we are from O
f ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
I 356 l
() 1 an overall gross erosion-corrosion-degredation 2 standpoint in reasonable shape. ;
- 3 We would anticipate leaks as a result of 4 flaws. We do not expect to have leaks as a result of a 5 general corrosion because the plan could follow that, 6 monitor that and see to it that that is not excessive.
7 The reason that I made the point with the 8 change in the tube-sheet weld was because it was related 9 to the number of problems that people.have had in 10 sodium / water steam generators, such as BFR. I wanted to 11 be certain that that was in the proper context that that 12 was a problem that we understand and have designed out.
13 We do not anticipate that problem. We do anticipate O 14 some flaws.
15 MR. CARBON: Go right on.
16 MR. DICKSON: Yes, sir. This is somewhat of a 17 cartoon of just the sodium portion of the steam 18 generator system. The sodium from the IHX enters the 19 superheater down-flow, separates in two paths, going 20 into three evaporators which are rejoined ~ at this point, 21 back to the pump and then onto the IHX.
22 At the outlet of the superheater there is a 23 leak detector which has two meters to sensor oxygen and
() 24 one to sense hydrogen. A leak detector at this outlet 25 of this evaporator - not shown on here only for l
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
357 lh 1 si3plicity, there is another line coming over to this 2 leak detector from this other evaporator. So, they are 3 both monitored by one leak detector. l 4 In addition, in the vent lines that run from ,
5 the evaporator and the superhea ter back to the expansion 6 tank gas line, there is also a leak detector there. So, 7 each evaporator superheater is being monitored by two 8 leak detectors.
9 Experience is that leaks that are smaller than 10 about two times ten to the minus fifth pounds per second 11 will tend to self-plug. They will come and go and in 12 any event, when you shut the plant down you cannot find 13 the leak if it is smaller than that to begin with.
O 14 ER. CARBON: What size is that?
15 HR. DICKSON: Two times ten to the minus fifth 16 pounds per second. That is roughly the lower level of 17 detectibility of our leak detection system. The 18 low-level detection range is considered f rom ten to the 19 minus sixth to two times ten to the minus fifth. But it 20 takes a long time to detect the leak.
21 If, however, a confirmed leak is found even in 22 that range - and let me refer to a table that I wish I 23 had a viewgraph of but do not. A confirmed leak in that 24 range will call on the operator to reduce the 9,over to 25 40 percent and to continue the search for the leak O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
358
() 1 location.
2 If the leak gets larger than two times ten to 3 the minus fifth - and the range of two times ten to the 4 minus fifth is 6.5 times ten to the minus 3 pounds per 5 second, the operator is to initiate shutdown and blow 6 down the affected module.
7 If the leak gets larger than 6.5 times ten to 8 the minus 3 he is to initiate a reactor scram, initiate 9 new flow-down of all three modules. All those are 10 operator intervention actions.
11 The alarm is both on the level for the 12 intermediate and the high and on the rate of change for 13 the low, the intermediate, and the high.
i 14 If the operator f ails to take action and the 15 leak continues to progress unti1 it gets into the range 16 of a tenth of a pound to a couple of pounds per second, 17 it will begin to pressurize this entire secondary 18 system. If the system is held at 93 psi in order to 19 protect the IHX and be certain that the secondary sodium 20 is always at a higher pressure than the primary sodium, 21 so, it is held at 93 psi anyway.
I 22 If the pressure begins to rise because of a 23 leak in either evaporator or superheater and reaches 150
() 24 psi, this rupture disc will blow and off-load the 25 pressure into the sodium dump tank.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE S.W., WASHINGTON, D.C. 20024 (202) 554 2345
359 1
There is a sensor on the down-stream of the 2 rupture disc. If it blows, it automatically dumps the q 3 water steam system, isolates and dumps the water steam Q
4 system. There is no automatic scram out of that signal, 5 but a scram w5.ll follow shortly thereafter because of 6 the steam feed mismatch. So, the system will scram.
7 Yesterday it was brought up what are the 8 consequences on the core. That event has been analyzed 9 with the assumption that at the time of the blow-down 10 the IHX is completely isolated and is no longer removing 11 any heat. The scram occurs within seven seconds. The
, 12 sodium takes about 60 seconds to get from the IHX back l
l 13 ts the reactor. So, the system is cooling down, the j 14 fuel is cooling off for the first 60 seconds.
, 15 When the hot sodium gets around to this icop, 16 hot primary, and gets to the reactor inlet, it mixes 17 more or less with the other two loops. If the mixing is 18 complete, th e temerature rise of the f uel cladding is 90 19 degrees Fahrenheit and it only rises from 810 to 900 20 degrees.
21 If it is does not six at all, and assume it 22 only goes up one-third path , the temperature rises about 23 270 degrees and the cladding temperature goes up to 24 about 1115, which is well under its normal operating 25 point.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
360
() 1 The point is that from a core standpoint this 2 is a non-event. From a standpoint of any radiological 3 hazard the maximum leak that can be allowed in the IHX 4 is six pounds per hour, and the leak would normally be 5 from the intermediate sodium to the primary because that i 6 is the pressurized sid e .
7 In this event, the intermediate sodium becomes 8 low pressu're. There would be some back leakage of the 9 primary sodium into the secondary sodium but it never 10 gets out of the system. It ends up either in the sodium 11 dump tank or trapped in one of the places that is not 12 drained.
13 HR. BENDERS Paul, could we get a few working O 14 conditions cleared up? First, the steam working 15 pressure is what, the steam system pressure ?
16 MR. DICKSON: It is 1450 psig.
17 HR. BENDERa And the intermediate cooling l 18 circuit is in what pressure, 200 psi, or less?
( 19 HR. DICKSON: A couple hundred. Yes, less, 20 maybe 300 - 200 to 300 psi maybe.
21 NR. BENDER: And what you are saying is'that 22 if a leak occurs in the steam system and imposes a 23 pressure on the intermediate sodium circuit, the rupture
() 24 disc is suppoed to blow?
25 ER. DICKSONa This rupture disk here will blow l
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
361 O t 'So ==1-2 MR. CARBON: Straighten me out right there, 3 that is in the intermediate circuit and it is going to 4 blow at 150. But I understood you 'to say the pressure 5 was around 300. So, I am mixed up.
6 MR. DICKSON: Yes, and that is why I 7 hesitated, I should have been more clear. This 8 expansion tank is pressurized to 93 psi normally. Then, 9 of course, the pump adds more pressure towards the IHX, 10 so that it goes up - and I was grasping in my mind what 11 it is but I think it is on the order of 100 psi or so.
12 So, we are talking something in the order of 200 to 300 13 in the IHI.
14 HR. BENDERa Now, the protection that goes 15 with this rupture disc has to'do with the rate at which 16 pressure might be imposed on the intermediate circuit by 17 pressure from the steam system through the leak. So, 18 the size of the leak and the way in which it imposes 19 pressure is part of the assumption.
20 What are you presuming in size that goes with 21 this rupture?
22 MR. DICKSON: That is a tenth of a pound per 23 second, to two to three pounds per second.
24 MR. BENDER: I see. So, that is about the 25 size leak that is tolerable with the present system. Is O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
362 I
() 1 that what yo are saying?
L 2 HR. DICKSON: No, sir.
! 3 MR. BENDER: What is a tolerable leak?
[}
4 HR. DICKSON4 All that one does is pop this l 5 rupture disc and put sodium into the sodium dump tank, 6 dump the water and scram the reactor. Now, I had not 7 gotten to the next slide.
8 ER. BENDER: That is all right, we are just 9 trying to get ourselves up to speed.
10 HR. EBERSOLE: May I ask a question about the 11 rupture disc? A good many years ago I had some 12 experience with them and I learned that if they are 13 dependent on simply hot stress to do their thing, then O 14 they are probably in a state of creep.
15 NR. DICKSON: In a state of creep?
16 MR. EBERSOLE: They are in creep as they stand 17 there, waiting to be relieved. Are these disks simply 18 broken by excess pressure, are they piloted by any 19 device?
20 MR. DICKSON: They are not piloted.
21 MR. EBERSOLE: They are not piloted. What is 22 the accuracy of their response in the range within which 23 you expect them to do their function, then, in view of 24 the fact that they are like fuses, you cannot test them?
25 MR. DICKSON4 That is correct, they are like O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINI A AVE, S.W, WASHINGTON, D.C. 20024 (202) 554-2345
363
) 1 fuses. Let me address these rupture discs which I am 2 going to get to. They are made of Inconel 600, and the 3 rreason they are is so that they are not under creep at 4 the pressures we are talking about.
5 MR. EBERSOLEs What is the thickness of the 6 disc?
7 MR. DICKSON: I don't know.
8 MR. KAUSHA: Ninu Kausha. I do not know the 9 thickness but the point needs to be made that we were to suffering a type disc which popped and then there is a 11 sharp edge that cuts the device as they pop in the 12 reverse direction. So, they are not dependent on 13 rupture.
D 14 MR. DICKSON: That is a good point. They 15 simply hava the flex and they will be popped. And I 16 should note that they are all double discs with a leak 17 detector in between, so should there be any leakage of 18 either disc --
19 HR. EBERSOLE: Do you know what the tolerance 20 on their performance is?
21 MR. DICKSON: It is not very tight, I know 22 that. I believe this one that pops at 325 is plus or 23 minus 25.
24 MR. EBERSOLE: I think that is extraordinarily 25 close.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
364 k 1 MR. DICKSON I know that th ey have been .
2 tested and have been found to be quite reliable.
3 MR. EBERSOLEa Thank yuo.
g 4 MR. DICKSON: All right, now we are going to 5 larger leaks. If we get a larger leak - and I will talk 6 later about the scenario that gets us to the largest -
7 possible leak I can think of.
8 Here we are talking about a leak th'at is 9 equivalent -- well, if you get a leak much over a couple =
10 of pounds per second, you will pressurize the system so 11 rapidly that this may not get to 150 psi before one of .
12 these systems sees 325 psi. -
13 In that instance,'you rupture what is called ID 14 the sodium / water reaction product system and begin to 15 void the sodium and the sodium / water reaction products _
16 into the reaction products separation pens.,
17 There is a rupture disk here that bends to 18 atmosphere on a hydrogen ignitor that will burn any of 19 the hydrogen products, and when that rupture disc is 20 ruptured it not only isolates and blows down the water 21 side, it also trips a secondary protection system trip 22 automatically and the primary will follow from the -
23 isolation blow-down of the steam water system. -
Il 24 The system has been flushed with nitrogen and 25 the check valve ultimately closes to prevent any ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345 _
l l
l 365 l
() 1 back-flow of air.
2 These tanks are sized to take the full amount 3 of sodium plus the sodium / water reaction products with 4 about a 20-percent margin.
5 MR. BENDERS How does this nitrogen flushing 6 action come about?
7 MR. DICKSON: I am really not familiar with 8 the nitrogen system. I know that it wants to hold the 9 system to the 300 psi.
10 MR. BENDER: If there is an over-pressure on 11 the system it just comes in automatically as the sodium 12 is depleted?
13 MR. DICKSON: Yes, sir, it is triggered O 14 automatically. The normal is Argon and.there is a 15 linkage between this argon, this argon, and the vent 16 lines on each of the evaporators, and of course the 17 pump. That is the normal cover gas. The nitrogen is 18 thrown in when we need large quantities of gas for an 19 event like this.
I 20 MR. BENDERS There is a gas over-pressure on
, 21 the expansion tank, but is that the source through which l
22 the nitrogen is introduced?
23 MR. DICKSON: No, sir, that is an Argon source.
() 24 MR. BENDER: Well, where would the nitrogen be 25 introduced that would do that to do the purging? It O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
366
() 1 looks like we have an infinite number of opportunities 2 there.
3 It is not that important, we can find out 4 about it some othnr time.
5 MR. DICKSON: I have a line drawing. It is 6 introduced - and I do not know exactly where it is, I 7 would have to get the line drawing - near the top of 8 each of these.
9 MR. EBERSOLE In the context of removing to decay energy none of this system up here is safety 11 grade, is it?
to MR. DICKSON: A good deal of it is safety 13 g ra de , yes, sir.
- O 14 MR. EBERSOLE: I notice it is all dependent on 15 the presence, then, of feedwater and non-leakage, and it 16 is a single-track system. So, I have to say that all of 17 this can disappear in the context of loop removal of 18 decay.
19 MR. BENDER: I don't think we are seeing all 20 of the system.
21 MR. EBERSOLE: Oh, I know that.
22 MR. DICKSON: Once this is triggered, you have 23 lost decay heat removal. From a core standpoint the
) 24 event, when you pop these rupture discs, is the same as 25 when you pop this.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W, WASHINGTON, D.C. 20024 (202) 554-2345
367 I 1 MR. EBERSOLE4 So, this is one of three loops.
2 MR. DICKSON: I should have explained it, this 3 is one of three loops.
4 MR. EBERSOLE: Yes, but it had the common 5 denominator which is the steam system.
6 MR. DICKSON: Yes, sir.
7 MR. EBERSOLEa And wha t do you do, 8 sectionalize that?
9 MR. DICKSON: Yes, sir. This is isolated.
10 These th ree share a steam drum and this steam goes to a 11 steam header and that is isolated; and those are safety 12 grade.
13 MR. EBERSOLEa And that steam header has the 9 14 opportunity of becoming super-saturated with f eedwater 15 or else boiling down completely.
16 MR. DICKSON: Right.
17 MR. EBERSOLEa At extremes of operation, 18 either too much or too little water.
19 MR. DICKSON: I am not sure I understand the 20 question.
. 21 MR. EBERSOLEs Well, on the one hand you can 22 have a steam header failure which will depressurize the 23 steam system, all right? It is a standard LWR accident.
24 MR. DICKSON: Yes.
25 MR. EBERSOLEa On the other hand you can have I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON. D.C. 20024 (202) 554-2345
368
() 1 an influx of main feedwater which will flood it. Are 2 you prepared for both extremes?
3 HR. DICKSON Yes, we are. I am not prepared 4 to talk about them today because that is quite a way's 5 from the subject.
6 I can say that the feedwater is controlled 7 into a drum, into a feed drum. The steam from the drum, 8 saturated steam, goes to the steam generator, through 9 the steam generator, and out.
10 -
The only way you could hit this with cold 11 feedwater is by getting too high a level in the drum.
12 There are both alarms and safety trips that will isolate 13 this drum with safety-grade equipment from any feedwater O 14 if the level goes too high to protect against that, for 15 instance.
16 If the header of the steam system breaks or 17 something goes wrong with the turbine, all three 18 superheater outlets are isolated - again with 19 safety-grade equipment - automatic 11y. There is a l
20 separate system - I do not know whether you are going to 21 talk about it, George. Yes, you do talk about it today, 22 the separate system.
23 That when we lost the turbine, that means
() 24 removing heat, a separate system to remove heat both 25 throught venting and what is called the protected O
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
369
() 1 air-cooled condensers for venting for short term, to get
- 2 cid of some of the heat in the first decay, and for long 3 term protected air-cooled condensers with a water supply 4 from a protected water storage tank. The system becomes 5 self-contained without any reliance on whether or not j
6 you lost the final steam outlet.
7 But that is going to be covered in more detail 8 later, so I won't touch on tht.
l 9 HR. EBERSOLE: Thank you.
10 ER ZUDANSs On these reaction product 11 separation tanks, then, they are able to accomodate the 12 entire intermediate sodium loop volume plus, of course, 13 whatever reaction products are with the water.
O 14 How much of sodium is assumed to have been 15 interacting with water before the tanks are filled 16 because it overfinws if you continued feeding steam, 17 water, into the sodium.
18 MR. DICKSONa Well, the water gets isolated.
19 So, what is taken is the total amount of water --
20 HR. ZUDANS: That was there before isolation.
21 HR. DICKSON Was there before isolation. And 22 there is a double isolation. There is an isolation that 23 is just as you enter here from the reserve pump and that 24 is a single valve. So, there is a saf ety-grade double 25 valve further up.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
370
() 1 So, the system is really designed to handle
] 2 just the water that is in the steam generators and 3 superheaters and the inter-connected piping, and not 4 adding in the steam drum. It is sized, however, to S handle the steam drum as well.
6 HR. ZUDANSa The steam then is isoitted f rom 7 the outlet evaporator and inlet after the pump?
8 HR. DICKSON That is correct.
9 HR. ZUDAMS: And a superheater is isolated 10 right at the superheater or some place near the headers?
11 HR. DICTSON: Yes, right near the superheater.
12 HR. ZUDANSt So, that is all the volume that 13 you allow for.
14 HR. DICKSON: But we throw in the volume of 15 the steam drum in case one of these should fail.
16 HR. ZUDANS: They are sized for that.
17 HR. DICKSONa If I said all of the 18 intermediate sodium, that is not quite correct, it 19 cannot all fit there.
20 HR. ZUDANSs That is what you said.
21 HR. DICKSON: Then let me correct that. It 22 cannot all get there. That is all that can get there.
23 HR. EBERSOLE4 How do you isolate the other
() 24 two intermediate loops from all the excitement that may 25 be taking place in this one? Do you have valves?
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
371
() 1 MR. DICKSON: No, they are not involved except 2 on the steam side.
3 NR. EBERSOLE: No, I as talking about on the 4 sodium side. What keeps troublos or failures in this 5 side from progressing to a general loss of secondary l
l 6 inventory of sodium? i 7 MR. DICKSON: There is a single reactor with 8 three loops. Each loop goes to a separate intermediate 9 heat exchanger, and each intermediate heat exchanger has to its own second intermediate loop.
11 HR. ZUDANS: A closed loop, not connected.
12 ER. DICKSONa This loop is completely 13 separated from the secondary loops.
O 14 MR. ZUDANSa They are connected to the steam 15 side.
16 HR. DICKSON: Yes. I wish I had brought the 17 overall schematic of the plant.
l 18 NR. EBERSOLE: That is OK, I get the picture.
19 Thank you.
20 MR. DICKSON: So, its only interaction with 21 the core and with the sodium, as I nentioned, has to do 22 with the temperature eff ect which is trivial.
23 MR. EBERSOLE: Yes.
( 24 HR. DICKSON: And the amount of leaking 25 sodium. Now, in this particular case the sodium could O
AWERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
372
() 1 leak across the IHX. Assume you also have a leak and 2 have had a lesk in the IHX at the time of this incident.
3 Some of it does get to the reaction part of 4 the separation tank and does in fact go out the vent. I 5 do not have the numbers with me but I believe it is 6 something like .006 milli-Rea whole body dose. It is a 7 trivial event, completely enveloped by a leak in the 8 system which would allow primary sodium then to get to 9 floor of the steam generator builiding.
10 MR. ZUDANS: I have another question in my 11 mind that bugs me. Why do you have a pump in the hot 12 leg of the primary sodium and not in the cold leg? Just 13 out of curiosity.
O 14 HR. DICKSON: Well, one reason for it that is 15 very handy is, that leaves the ofthe reactor vessel at 16 atmospheric pressure because with the pump drawing down 17 on the reactor vessel, that can be lef t to the 18 atmospheric and the remainder of the system is lef t 19 pressurized.
i 20 NR. ZUDANS: But you have to cope with 21 260-degree high temperature.
22 MR. DICKSON: Yes.
23 HR. ZUDANS: With that piece of equipment you
() 24 are subjedet to creep at those temperatures.
l 25 HR. DICKSON: That is correct.
(
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE. S.W., WASHINGTON, D.C. 20024 (202) 554 2345 1
373 I will give you another reason.
(]) 1 If we are 2 attempting to make Clinch River prototypic of larger 3 plants so that if we may go to larger plants the 4 sesle-up will not be as dramatic, then a ho t-leg pump in 5 Clinch River is not too greatly different than a 6 cold-leg pump in a larger plant where it has to go to 7 larger sizes. Thst was also a consideration.
6 MR. ZUDANS: I see. I understand what you say.
9 MR. BENDERS I want to summarize what we have i
10 been told so far. What you have are some rupture discs 11 and the presumption is that the rupture disc will unload 12 the sodium system into these collector tanks, the dump 13 tanks, as you call them. And it is the intent to say 14 that that is where the bulk releases of sodium will
, 15 deposit if you have an over-pressure event of some sort?
16 MR. DICKSON: I would characterize it not as 17 an assumption, I would characterize it as designed that 18 way, and there is no place else for the sodium to go.
19 MR. BENDER 4 OK.
20 MR. DICKSON: We have a high-pressure system, 21 completely bottled-up high-pressure system on an open 22 rupture disc and a straight pressure here. The sodium 23 and anything else that is in there is going to go out
() 24 that 18-inch line and into that separation tank.
25 MR. BENDER 4 Well, I don't want to have this
)
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345 L
374 l
(lh 1 statement misinterpteted but that was the intent in a 2 number of wa ter-cooled reactors that I know about, for 3 the tanks to collect the water. It did not always 4 happen that way, but, that is life.
5 MR. DICKSONs Well, I cannot think of a 6 different failure mode. I have on one side a thousand 7 pounds of steam; I have on the other side 325 psi 8 sodium; I have water in sodium reacting in the middle of 9 that system and expanding. And I have an 18-inch line to that is open.
11 And nothing else is anywhere close to its 12 rupture pressure. We are talking 300 psi.
13 MR. KASTENBERGa Paul, bef ore you take tha t O 14 off, what is the significance of the two valves on the 15 dump dank line?
16 HR. DICKSON: If you want to dump this 17 intermediate system for whatever reason you hold the 18 pressure in that position, the stuck-open valves.
19 MR. KASTENBERGa And the dump tank is in the 20 containment or outside? I guess the whole thing is 21 outside.
22 ER. DICKSON: Let me correct myself, that is 23 not correct. There are five dump lines - this actually
) 24 shows three, this schematic. They go to the low points l
25 in the system. You would not dump through here at all.
l I
xs
(
ALDERSON REPORTING COMPANY INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
375
() 1 These are normally open, you do not dump through there.
2 These are normally open valves and the idea is that they
() 3 can be cloed if you wawnt to cloe off the dump tank 4 system from the remainder of the system.
5 But under normal operation they are open, so 6 that the atmosphere here is the same atmosphere as that 7 in the expansion tank. The dump lines, and there are 8 five in number, come from low points.
9 Now that I corrected my erroneous statement, 10 repeat the second question.
11 MR. KASTENBERGs This whole system is outside 12 of the containment,-right?
13 MR. DICKSON: Yes, sir.
14 HR. ZUDANS Now you have confused me. You 15 say those valves are normally open, what is the point in 16 the rupture disc, then?
17 MR. DICKSON: I've got something wrong. Help 18 me, George.
19 HR. CLARE: George Clare, Westinghouse.
20 You were half right the first time and half 21 right the second time.
22 (Laugher.)
23 MR. CLAREa All of the valves you pointed to 24 there from the sodium dump tank are normally closed.
25 g MR. DICKSON: These are normally closed?
G V 1 t
4 ALDERSON REPORTING COMPANY,INC, 400 VIRGINI % AVE, S W., WASHINGTON. D.C. 20024 (202) 554-2345 L - - .-. - - ,
376 k 1 MR. CLARE: Right. The correction you made 2 that the draining comes down through what is indicated 3 there as three valves is correct. The two upper valves ggg 4 are also normally closed. They would open at the time 5 you drain the codium, and what that allows is, the gas 6 that is initially in the dump tank to fall back into the 7 intermediate heat transport system and replace the 8 volume from which the sodium is flowed.
9 MR. DICKSON You cannot drain sodium into 10 there if this is a closed tank. So, these would be 11 filled with sodium coming from low points in the lines 12 during a drain. So in order to let the gas out, it goes 13 out there instead of leeting out the atmosphere with the O 14 vacuum system.
15 MR. ZUDANS: But that still leave sthe 16 question of that rupture disk at that location because 17 it connects the gas. It is intended for communications 18 when you begin to dump the sodium dump time. So, what 19 is the point of that disc?
20 MR. DICKSONa No, it is really pressure 21 relief, it relieves pressure. Whether or not it only 22 saw Argone going through or also Argone sodium would not 23 matter. It holds the pressure of the system at not 24 greater than 150 psi.
25 MR. ZUDANS: If the pressure is built up in O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554 2345
377
() 1 that system because of th e reaction that took place, you 2 would connect the sodium dump pump gas space to the 3 expansion tank gas space.
(])
4 HR. DICKSON: Tha t is correct.
5 HR. STARK: That means some of the gas from 6 the expansion tank would move the sodium dump tank, or 7 wherever they are dumping things. Does it mean that 6 that expansion normally it is used to pressure liquid 9 sodium, and that is the end of it? What do they do, 10 then open the velds for dumping?
11 HR. DICKSON: Yes. But then that is a 12 manually oprated thing, not an automatic thing.
13 HR. ZUDANS: So, your original statement that 14 the rupture disc is the one that provokes the dumping of 15 sodium is incorrect?
16 HR. DICKSON: No, sir. I don't believe I made 17 that statement. I hope I did not.
18 The sodium is never automatically dumped. The 19 rupture disc dumps the water automatically.
20 HR. ZUDANS: Well, we don't have that part of 21 the picture, you don't have any water lines.
l 22 HR. DICKSOWs There is a sensor here - I took 23 the water lines off to make it simpler. There is a 24 sensor here that senses this rupture disc failure and 25 dum ps, isola tes and dumps the water side of the steam O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W. WASHINGTON, D.C. 20024 (202) 554-2345
378
() 1 modules.
2 MR. ZUDANS: And the rupture disc, the only 3 thing it does in sodium, it connects the gas space in 4 these two tanks, and equalizes the pressure.
5 MR. DICKSON: If the leak is small enough --
6 large enough to rise to 150 psi but small enough that it 7 is not much more than that and not rising rapidly, it is 8 conceivable that this would rupture, off-load the 9 pressure, dump all the water end steam, te rmina te the 10 event, terminate the generation of hydrogen.
11 HR. ZUDANS4 How does the water-dumping device 12 know that the disc ruptured?
13 HR. DICKSON: Beg your pardon?
O 14 MR. ZUDANSs How does the water get the signal 15 to be dumped if this disc ruptures?
16 HR. DICKSON: I don't know what that sensor is 17 that senses the rupture disc rupturing.
18 Do you know, Ninu, is it a pressure sensor?
19 Mr. Kaushala I am not positive. I believe it 20 is both pressure and flow.
21 HR. EBERSOLEs Your sodium dump tank looks a l
22 little bit like the dump tank on the boiling water 23 reactors which did not dump a few months ago because
() 24 they were already filled with something else. How do 25 you know that is always empty?
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
l i
l 379
( 1 MR. DICKSON: I don't think that we have a 2 level meter in here, but it is normally empty.
() 3 MR. EBERSOLE: Well, do you know the boiler 4 problem?
5 MR. DICKSON: No, I am not familiar with the 6 boiler problem. I have a hard time believing this looks 7 like a boiling water reactor.
8 HR. EBERSOLE: Well, that part of it looks 9 like the control rod dump tank on a boiler.
10 HR. DICKSON: This system has enough sodium in 11 it, that is all that is put in, that if this is filled 12 the loop is empty and vice versa.
13 MR. EBERSOLEa But does it have to be vented O 14 to fill it?
15 MR. DICKSONa I would think it would fill very 16 slowly if it was not vented. Obviously, if you did not 17 vent this and you simply tried to drain, it would really 18 get drainage of a sort but you would have backflow of 19 vaterline gases.
20 MR. EBERSOLE: Which might be very slow and 21 inefficient, if at all. So, do you have safety-grade 22 vents on that thing which are guaranteed such that that 23 in fact never does go much above some stated pressure?
24 HR. ZUDANSs They don ' t have any vents at all 25 to the outside. The only vent you have is the extension.
([) !
l l
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W WASHINGTON D.C. 20024 (202) 554-2345
380
() 1 MR. EBERSOLEa That is it right there, where 2 you have two valves in a series there, which seems to 3 defeat their liability function of venting, instead of 4 two valves in parallel which would presumably enhance it.
5 MR. DICKSONa I am not sure I follow what you 6 are talking about.
7 MR. EBERSOLEa I am just looking at the 8 reliability of interchanging gases for sodium in the 9 sodium dump tank.
10 MR. DICKSONs What is the accident you are 11 postulating?
12 MR. EBERSOLE: You need to dump, all right?
13 You have to dump, it is an emergency.
h 14 MR. DICKSON: No, sir.
15 MR. EBERSOLEa I am saying, you have something 16 happen where you have to dump because of pressure rise 17 or whatever.
18 Now, then you attempt to dump but the sodium 19 dump tank is filled with gas and it will not vent. Does 20 that bother you any?
21 HR. DICKSON: I am not certain that I can 22 follow that argument too much because you a re talking 23 about both of those valves not being able to open.
( 24 3R. EBERSOLE: Either one, rather.
25 MR. DICKSON: But I want to emphasize the O
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
389 1 point that this dump, the sodium to the dump tank, is 2 never triggered by any automatic signal. It is an 3 operator action only. So, it is not considered a part 4 of the safety problem. When in a sodium-water reaction 5 ve get rid of the water.
6 If this thing has ruptures or if we have a 7 leak that the operator is taking action on, so he has -
8 blown the system down, we vill not' dump until the 9 temperature of the sodium is down below 800 degrees.
10 The sodium is held there. The sodium is dumped at a 11 significant length of time after any such accident.
12 MR. EBERSOLE: So, sodium dumping is not a 13 safety function.
14 MR. DICKSON: Sodium dumping is not a safety 15 function, it is a maintenance function. So, if one of 18 these valves failed to open - and I see no reason why it 17 should, but if should fail, the operator has plenty of 18 time to take action. There is never an immediate sodium 19 dump.
20 You cannot dump very fast, either, it takes 20 21 minutes, actually 30, to completely drain the thing; 209 22 minutes until the last of these three are drained of 23 sodium.
24 The water, on the other hand, dumps rapidly, 25 being at high pressure. It drops down to 300 psi in 30 ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
382
(') I seconds. -
2 MR. ZUDANS: So, the rupture disc is to
, 3 actuate this vent quickly in the case of sodium-water 4 reaction, and in that case.you still have to provide 5 manual opening of those sodium vents, dump vents.
6 HR. DICKSON: Yes.
7 MR. ZUDANS: However, you do have a vent path 8 already provided by this rupture.
9 HR. DICKSON: Yes.
10 HR. ZUDANS: Thank you.
11 MR. DICKSON: The whole purpose of this is so 12 that we do not involve these because if you did not have 13 a way of relieving pressure from a relatively large O 14 leak, .1 to two pounds per second, except through this 15 path, than any leak of that size would involve this, 16 which is a lot more effort to correct the system after 17 the accident than it is here. Here you just take the 18 sodium back, clean it, and put it back in the system.
19 Here, you are cutting out reaction product 20 separation tanks for the plant, removing them.
21 MR. ZUDANS: I don't quite see the 22 difference. The first sodium dump also could have i
23 reaction product because that is the same circuit.
24 HR. DICKSON: It could, but you would not 25 expect it to have as much. It is made to take sodium O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
383 llh 1 more thar. one time. The whole system is kept trace 2 heated so that it can take sodium. This is not. These 3 tapes are fair thermal shock and they are designed for 4 not a large number. It is assumed that the clean-up 5 after these are involved is significantly different.
6 MR. ZUDANS: They both dump the same 7 intermediate sodium loop. They both dump the same loop, 8 it is just that there are different initiation signals.
9 HR. DICKSONa Yes. But what causes that to different initiation is quite significant. If it is a 11 small leak, the kind that will rupture this rupture disc 12 and not rupture these, that is not greater than a couple 13 of pounds per second. The amount of sodium-water O 14 reaction product one has is not that excessive.
15 HR. ZUDANS: And at what point would this 16 other reaction product separation dump system begin to 17 f unction ?
18 HR. DICKSON: At around two to three pounds a 19 second. The pressure will go up more rapidly than the 20 the 325 psi here before it gets to 150 psi here. Under 21 that scenario the rupture disk closest to the system 22 will fail.
23 MR. ZUDANS: And then you would dump from this 24 reaction product separation tank and also, of course, 25 you already would have been dumping into the other one.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
384
() 1 MR. DICKSON: No.
2 MR. ZUDANS: Sure because that opens earlier 3 than this.
4 MR. DICKSON: No.
5 MR. ZUDANS Oh, it is a manual action.
6 MR. DICKSON: These dumps, this sodium you 7 mean?
8 MR. ZUDANS: Right.
9 HR. DICKSON: They are manual, they are not 10 being dumped.
11 MR. ZUDANSa But you would have broken that 12 rupture disc by then.
13 HR. DICKSON: It is possible since a slow leak 14 vill cause this to break first.
15 NR. ZUDANS Right.
16 HR. DICKSON: To reach a point where both this 17 one and this one breaks, but for the very large leaks 18 this one will break first.
19 MR. ZUDANSs I don't quite see why if that is 20 set at a higher pressure than the other one.
21 MR. DICKSON: That's right.
22 HR. ZUDANS: So, why would it break first, 23 then?
() 24 HR. DICKSON: Because if you have a leak 25 greater than two to three pounds a second, the pressure O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
385 7()
l I builds up here f aster than it can get back through ;
l 2 wherever it has to go through, back to this pump, l 1
3 through this loop right here.
,O 4 MR. ZUDANS4 So, that is a long path to get 5 there?
6 HR. DICKSON: Yes, sir.
7 ER. CARBON: I have a question. Your main 8 thrust is to get rid of the water, right, in case of a 9 leak in the evaporatgor. Where does the water go? None 10 of this sketch shows that.
11 MR. DICKSON: That is right. As I said, I was l 12 not quite prepared to discuss what we are discussing.
l 13 It goes to a water duap then.
l 14 HR. CARBON: And the system parallel to all 15 this. So, the main safety thrust is to get the water 16 out of there as quickly as you can, as I see it.
17 MR. DICKSONs That is correct.
18 ER. CARBON: Your reaction product separator l 19 tanks, what do they truly do for you safety wise?
20 MR. DICKSON: They are placed to off-load the 21 pressure without dumping it all over the floor. You
! 22 have had an expanding volume of hydrogen.
() 24 HR. DICKSON: You cannot get it out very 25 readily through any path except through one that goes
'O I
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W WASHINGTON, D.C. 20024 (202) 554-2345
386 O i de n.
2 MR. CARBON: So, they serve, then, as a 3 separation for the hydrogen so you can then get it out 4 and get rid of it.
5 MR. DICKSON: Yes. If you just put, for 6 example, this rupture disc at the top of the evaporator 7 and relied on the fact that the hydrogen will ultimately 8 tend to rise, it of course . has brought a great amount of 9 sodium and sodium-wa ter reaction product. So, you put 10 it through a tank --
11 NR. BENDER: ~Where does the hydrogen go, Paul, 12 where it is supposed to go?
13 MR. DICKSON: It goes out this vent line O 14 through this rupture disc, through a hydrogen igniter.
15 HR. BENDER: And then it just burns outside 16 containment somewhere.
17 HR. DICKSON: It is burned outside the 18 containment, yes. And that was the source of the .006 19 milli Rem of radiation you have on the outside.
20 MR. EBERSOLE: When all this happens to this 21 loop you have lost one-third of your deep synch, right?
22 MR. DICKSON: Yes, sir.
23 MR. EBERSOLE Now, you are depending on the 24 other two thirds for shutdown?
25 MR. DICKSON: Yes, sir.
O I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
387 1 MR. EBERSOLEs And you are dependent in the I 2 context that you had to, by a mechanical action and 3 sectionalization of steam feedvater and whatever, use 4 the remaining two thirds as your head synch. You did it 5 through a number of mechanical responses.
6 NE. DICKSON: Yes.
7 MR. EBERSOLE: You do not have a dedica ted 8 system.
9 MR. DICKSON: Yes, sir, we do.,
10 MR. EBERSOLE: Other than beyond this?
11 MR. DICKSON: Yes, sir. -
12 MR. EBERSOLE: OK, don't bother.
13 MR. DICKSON: Two of them.
14 MR. EBERSOLEa G r ea t , fine.
15 MR. CARBON: But actualy you don't have to 18 have any mechanical action, do you?
17 HR. DICKSON All I need to do is have these 18 safety-grade isolations belts work.
19 MR. CARBON: To keep the steam from the other 20 headers, the main steam from the turbine to feed back 21 into there; that is the only thing.
22 MR. DICKSON: Right.
23 MR. EBERSOLEa And to hold steam pressure, I 24 guess, in the other two-thirds of the system.
25 MR. DICKSON: Well, actually I don't even have O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
388 l 1 to have that because I can isolate the other two.
2 MR. EBERSOLEs And then fall back on your 3 dedicated system.
4 MR. DICKSON Well, one of the two, yes.
5 HR. CARBON: Would you summarize again how 6 rapidly you can get rid of the water when you have a 7 leak of the various sizes?
8 NR. DICKSON: The water pressure will drop to 9 300 psi.
10 HR. CARBON: It is normally about 1400?
11 HR. DICKSON: Yes, 1450. It drops to 300 in 12 less than 30 seconds. Once this rupture disc opens, 13 depending on the size of the leak, but for a largest 14 possible leak which makes it the f astest and the most 15 over-pressurizing the system, it is cleared, the 16 effective steam generator evaporator is cleared in about 17 three seconds and the line is cleared in three to five 18 seconds.
19 MR. CARBON: So I guess what I just heard was 20 that if you have like tube ruptures in the steam 21 generator, in any part of the evaporator superheater or 22 wherever, you have the water effectively out in about 23 three seconds from the break here.
24 ER. DICKSON: Well, there is still some water 25 going into it because the water on the other side is D
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
)
389 llh 1 still at some 300 psi. It is dropping down 200 to 300 2 psi in 30 seconds. So, there is still some water coming 3 in. But what it coming into it is primarily a region of 4 raction products and water vapor because that expanding 5 bubble is blowing the sodium out. I 6 MR. EBERSOLEs If you have a depressurization 7 on the steam side, is there a chance of getting water in 8 the superheater and experiencing a severe thernal shock 9 as a result of that?
10 MR. DICKSON: Do'I get a depressurization?
11 MR. EBERSOLE: On a secondary site, and then 12 that is followed by expansion of water into the 13 superheater, like a main steamline failure.
O 14 MR. DICKSON: Like a main steamline failure?
15 MR. EBERSOLE: Yes. Do you then get a thermal 16 shock derived from flushing the superheater boiler with '
17 water?
18 MR. DICKSON: No. This evaporator, or boiler 19 as you call it, has water coming from the steam drum 20 anyway. It is water, not steam.
21 MR. EBERSOLE: Yes.
22 MR. DICKSON: This is getting steam from the 23 steam drum and it can only get water from the steam drum 24 if the steam drum fills too full.
25 MR. EBERSOLE: Is that a p]ossibility?
O ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE. S.W., WASHINGTON D.C. 20024 (202) 554 2345
390
() 1 MR. DICKSON: The steam drum has sensors that 2 will cut off feed and trip the reactor if it gets too 3 full.
[}
4 MR. EBERSOLE: Does that have to be safety 5 grade, then?
6 MR. DICKSON: Yes, sir.
7 MR. EBERSOLEa So, you really do not want to 8 say that you ever have water in the superheater.
9 MR. DICKSONs Th at is righ t.
10 HR. EBERSOLE: Because something would happen, 11 I gather, not very good if that were to occur; or would 12 it matter?
13 MR. DICKSON: It would be a life-degrading O 14 event for the superheater.
15 -
MR. EBERSOLE: Thank you.
16 MR. DICKSON: I have taken my whole hour and I 17 don 't know whether I should go on because we spent a lot 18 of time on the system.
19 What I said I wanted to talk about was - and 20 this is the part that I had planned to talk about - was 21 how we get a large water reaction and what our design 22 basis sodium-water reaction is. You can get a tube 23 rupcure from a sudden rapid propagation of a flaw.
24 MR. CARBON: We do want to hear this, so, 25 don 't necessarily rush.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
391
() 1 MR. DICKSONs All right.
2 Or you can get the growth of a small leak l I
3 followed by tube failure. Now, as I will show you
(]) ,
4 later, this is enveloping to this event. So, in either l l
5 case we have a sudden rapid rupture of a tube.
6 The large leak is presumed to develop in this 7 manner. We have first a leak on the order of two times 8 ten to the minus fifth, the level I talked about i 9 earlier. The level at which they don't continue to to leak, necessarily, they plug. They go away when you 11 shut down so you canot find them.
12 MR. CARBON: What is that in terms of GPM?
13 MR. DICKSONs Well, that would be about three O 14 and-a-half times ten to the minus sixth GPM.
15 MR. CARBON: Very tiny.
16 MR. DICKSON: Yes. If they get much smaller 17 than that, the clean-up system just cleans up the 18 hydrogen and oxygen and you have a hard time knowing you 19 have one. But somewhere below there you are detectable, 20 but not very easily and not with a high probability.
21 If this continues, it will begin to erode.
22 You get brief intermittent leaks. They keep getting 23 larger, developing a larger crater until it finally gets l
O 24 the crater down to where,it thins the wall and then you l
25 will get a rupture.
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W, WASHINGTON. D.C. 20024 (202) 554 2345
392
) 1 The time from step one to step four can be 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, or days, or months. They could stay in this form
() 3 for a long time. '
4 The time f rom here to here is on the order of 5 a minute or less. Now, this leak at this point has gone 6 up to about three times ten to the minus two pounds per 7 second.
8 ER. CARBON 4 I guess it is ten to the minus 9 two GPM, or something like that.
10 MR. DICKSONa Yes, divided by seven.
11 HR. BENDER: How do these time progressions 12 get established, Paul?
13 . MR. DICKSON: These are bench-scale 14 experiments.
15 MR. BENDER: Based on tests on the rate at 16 which the corrosion phenomenon progressed?
17 NR. DICKSON: Yes. At this juncture this is 18 very readily detectable and over the threshold of what 19 the operator knows is an intermediate leak and he has 20 alarms, and he should do something. fle at that level in 21 fact should initiate a reactor scram and loop 22 blow-down. And here I am talking about water blow-down 23 of all three modules.
24 MR. CARBON: Just a minute. You seem to show 25 the erosion taking place on the sodium side. What am I O
ALDERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE, S.W, WASHINGTON. D.C. 20024 (202) 554 2345
393 I 1 missing there? The first step shows steam on the under 2 side, 3 MR. DICKSON: Well, the steam is at higher g
4 pressure than the sodium. So, it is working its way 5 through and reacting with the sodium right at that point.
6 MR. CARBON 4 OK, and that is causing the 7 erosion.
8 MR. ZUDANS: On these tests that were 9 performed, was tha t more than one tube and a propagation 10 to the next tube was observed? -
11 MR. DICKSON: Yes, I am going to get to that.
12 MR. ZUDANSa I am sorry.
13 MR. DICKSON: What we have talked about now is 14 the development of a small leak. And that small leak is 15 the kind of leak that action should be taken. If it is 16 not, that is the sort of leak that is not leading to a 17 larger one.
18 Well, if action is taken there is no problem.
19 What we are talking about now is the tube-to-tube 20 propagation. We assume we have a precursor and it has a 21 plaint f ront, it is heating up another tube.
22 Three ways it can act on that other tube. You 23 either get an overheating in which you get a pressure 24 rupture which occurs in a few seconds; or you can get 25 wastage which occurs in tens of seconds. Once a l
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
394 i
lll 1 pressure rupture occurs there is a very rapid change -I 2 use the term microseconds, it might be a little more 3 than that, in the leak rate.
4 MR. CARBON: Excuse me, this is overheating on 5 the next tube, the second tuba?
6 MR. DICKSON: Yes, on this tube.
7 MR. CARBON: And that is caused by wha t?
i 8 MR. DICKSON: There is water blowing into the 9 sodium, it is hottest at the sodium-water interface, 10 water inside and sodium-water out here.
11 MR. CARBON: And wastage is simply erosion?
12 MR. DICKSONa Yes. Now, in most experiments 13 wha t ha s actually happened - and I am not going to show O 14 it this way - what actually happens is, this first 15 falure is caused by wastage and it is only a little bit 16 larger, and it leads to a second failure.
17 But we don't worry about that. What we do is 18 assume that that next failure ruptures from overheating 19 and typically the ruptures are like this, one to one 20 and-a-half inch long with a 45-degree gap at the most.
21 The tearing quits as soon as you get into the ductile 22 region.
23 You have an overheated region right here and 24 when it fails from the presesure inside, it starts to 25 tear. Of course, the pressure immediately drops inside O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
395
() 1 and the tear cannot go beyond the ductile point.
2 At that juncture, with a primary failure, we 3 have a system that idBks like this. We have a very
)
4 dynamic environment and a reaction zone that still 5 exists. It is a lot less stable reaction zone now, you 6 no longer have that little -- it is less likely to 7 propagate when the primary crater is large. That is why 8 I mentioned one of the things that has happened, one 9 failure, a small failure here, led to a large one.
10 But if you assume a very large failure that 11 involves a lot of tubes, it is much less likely to 12 propagate until you get a third tube 13 ER. EBERSOLEa Since you mentioned micro O 14 seconds, the local pressures in that region, the sodium 15 system,,become very high long before the leak disc sees 18 anything bacause of inertial and other effects. Do you 17 calculate the distribution pressure as a function of 18 time to the rupture disc?
20 HR. EBERSOLE Yes.
21 MR. DICKSON: Yes, and let me make two 22 points. One of th e reasons tha t we are using this 23 particular precursor is that we assume that this leak
( 24 has been going on just long enough to raie the pressure ,
25 of the whole system to 150 psi just below where that ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
396 lll 1 rupture disc ruptures, and then this fails. That is the 2 worst case you can have. That is why I said, this case 3 envelopes the sudden rupture to a flaw. If you get a 4 sudden rupture to a fla w it will be done a t the normal 5 pressure in the syctem which is down by 50 psi or so.
6 MR. EBERSOLEa Why is that because you have 7 now got the rupture disc where its time constant is very 8 short.
9 Well, what I am getting at, is there not a 10 pressure profile throughout the system where the higher 11 end of it is significant relative to --
12 MR. DICKSON: Yes, sir. We do do a time 13 history analysis of the explosive wave that is generated O 14 by this.,
15 MR. EBERSOLE4 OK. -
16 NR. DICKSONs For all of this, through all of 17 the piping, through the IHX and we had disclosure in 18 fact due to the fact you were making, I believe, the 19 last time, that the wave traveling along is higher at 20 this point thaa it is at this point, and that pipe sees 21 a differential pressure process as opposed to simply the 22 wave pressure.
23 dR. EBER50LE: What is the terminal pressure Ih 24 the system can tolerate?
25 HR. DICKSON: Emercency conditions are in the G
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
397
() 1 order of 700 psi.
2 MR. EBERSOLE: Thank you.
- 3 MR. DICKSON4 Their steady-state design 4 pressure is 325.
5 MR. ZUDANS: I am lost again. You say that
- 6 this occurs when the rupture disc is already ruptured?
7 MR. DICKSON: No, sir. If I mislead you, let i'
8 me correct myself.
' 9 We take this event in such a way that we 10 assume that this little failure here is gradually 4
11 pressurizing the system to bring it all to 150 psi i
12 before this one goes. So that the pressura wave that 13 this one lays in is starting n a system that is already 4 14 at 150 psi.
15 MR. ZUDANS: That means that the discs did not 16 rupture yet.
! 17 MR. DICKSONs That is correct. The discs are 18 ruptured by this pressure wave.
19 MR. ZUDANS: So, you just built a base 20 pressure. .
21 MR. DICKSON: That is correct. That is why 22 this small leak, precursor leak causing this f ailure, i
23 envelopes the complete rupture, completes the rupture of l()
24 a given tube because that will be done with the system j 25 at normal pressure.
l l
l l
l ALDERSON REPORTING COMPANY,INC, l 400 VIRGtNIA AVE., S.W, WASHINGTON, D.C. 20024 (202) 554 2345
398 1 We will go on, then, to potential for second 2 failures.
3 MR. CARBON: What terminology? We have had 4 the leak , th e f ailure and the --
5 MR. DICKSON: Primary failure?
6 MR. CARBON: Primarily failure being the first?
7 MR. DICKSONs The first major leak.
8 MR. CARBON: Well, the first major leak is in 9 the first tube to ao s is it not?
10 MR. DICKSON: You mean this one?
11 MR. CARBON: Yes.
12 MR. DICKSONs Well, this one is the kind that 13 would ultimately rupture. The kind that first should 14 have been detected and corrective action taken.
15 Second, if it was not detected and corrective 16 action taken, if it had not caused a larger rupture, the 17 rupture disc from the tank that goes at 150 psi would 18 have relieved the pressure and we would never have 19 involved the rupture discs for these SWRers.
20 When this one goes it raises the pr essure 21 quite suddenly and dramatically and the SWRer system is 22 involved.
23 MR. EBERSOLE: But it did not go out radially O 24 as rou had it there, but reeched over end eet the second 25 tube as a result, then you have two tubes going. Does O
ALDERSON REPoRTINd COMPANY,INC, 400 VIRGINIA AVE., S,W., WASHINGTON, D.C. 20024 (202) 554 2345
399
() 1 that alter your problem much?
2 MR. DICKSON: No. You are talking about this?
3 MR. EBERSOLE: That one reaching over there.
4 MR. DICKSONa Over to this one.
5 MR. EBERSOLE: Right. In other words, you had 6 a larger flame front.
7 MR. DICKSON: That is right. I will get to 8 that. We assume that this ruptures as a full 9 double-ended guillotine, so that is the same as if it 10 just severed. That flow is 12 and-a-half pounds per 11 second in the evaporator and there is no way you can 12 make a flame front out of that, a big ruptured hole and 13 a very rapidly-moving expansion.
O 14 It can rupture smaller and do just exactly 15 wha t you said for the flame front on the next one. In 16 that case you will get the second rupture but in those 17 cases you neither are as large as this one. So, 18 assuming that this one goes completely and disruptively 19 is a more limiting and conservative case than assuming 20 successive of smaller rupture 21 When we say we are designed for a full 22 equivalent double-ended guillotine it does not matter l
23 whether it is two half DEGs or one DEG in one tube.
( 24 To go on. The flame front has kind of 25 disappeared in this large wave but it is possible to t
()
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
400
() 1 involve yet another tube, a secondary tube. We allowed 2 for a secondary tube involvement.
3 MR. KASTENBERGs Why only one?
4 MR. DICKSON: I am sorry.
5 MR. KASTENBERG: Why only one secondary tube?
6 MR. DICKSONa OK, we will give you another one.
7 MR. KASTENBERG I guess what I am asking, 8 really, is there some experiment that was done that 9 gives you an idea of when tubes in a bundle generally 10 vill go?
11 MR. DICKSON: Yes.
12 MR. KASTENBERGa Or do you just postulate?
13 MR. DICKSON: No, we are not just postulating, O
\l 14 there were a lot of experiments, French ones, Japanese, 15 and ourselves. There have been 34 large leak tests and 18 they are mos tly nonproductivic, I an afraid I have to 17 admit. And only four times have there been secondary 18 tube failures. In nine United States tests that were 19 specific, one test produced a secondary failure.
20 Now, in the case of a secondary failure in the 21 U.S. test there was a stagnant water system. So, it.vas 22 prototypic in dimensions and materials but it was not 23 prototypic in the flow. The stagnant water system
() 24 tended to lead progression.
25 In the other three where there were secondary O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
401 lh 1 failures there was both stagnant sodium and stagnant 2 vater. In that test series there have been two that had 3 what we refer to not tertiary failures but more than one 4 secondary failure, but many seconds downstream. In 5 fact, all of these were somewhat downstream.
6 MR. ZUDANS: And right after your first 7 primary f ailure you would have already had to dump the 8 system.
9 MR. DICKSON: It depends on how fast the 10 secondary --
11 MR. ZUDANS: What I an driving at is, it 12 really does not matter how many more failures because 13 the system would be dumped and isolated.
D 14 MR. DICKSONa Beyond two it absolutely does 15 not matter, that is correct.
16 MR. CARBON: I have a question back here yet, 17 before this slide. Of those nine prototypic tests there 18 was one secondary failure of just one tube?
19 MR. DICKSON: There was one secondary failure 20 that did occur in the time frame of the accident before 21 the velds were opened. There were actually several 22 other tubes that failed because the water side was 23 bottled up and the system heated up and the water 24 pressure continued to rise with nowhere to go, and 25 several of the tubes failed.
I ALDERSON REPORTING COMPANY,INC, 400 WiGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
402
] 1 But as far as the test results aste concerned, 2 only one was really involved in the accident where a 3 rupture disc blev.
D, 4 MR. EBERSOLE: You show one relief disc per 5 steam generstor. Yet, in the long-term sense they are 6 in parallel. Do you have just one per steam generator?
i l
7 MR. DICKSONa Well, they are double rupttwe '
~
8 discs.
9 MR. EBERSOLE: I don't mean that, I am talking 10 in the constext of relief. There is one per steam 11 g en e ra to r?
12 MR. DICKSON One per module, yes.
13 MR. EBERSOLEa Does the distant one which is 14 redundant - I presume considered redundant for relief -
15 respond to this f ast accident?
16 MR. DICKSON: On the fast accident they all go.
17 MR. EBERSOLEa The distant one will serve as a 18 redundant.
19 MR. DICKSONa Yes, sir.
20 MR. EBERSOLE: Thank you.
21 MR. CARBONa I don't understand your 22 explanation about the water pressure building up because 23 the water is exiting the ruptured tubes and going out r 24 into the sodium.
25 MR. DICKSON But the non-ruptured tubes still ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
- 403 1 contain water and still contain caputured water.
2 MR. CARBON When you say " captured water,"
3 they have their normal rate?
4 MR. DICKSON: No, it was not a flowing water 5 system, they were all closed. They are all closed tubes.
6 MR. CARBON Oh, I see. It was not very 7 prototypic, then.
8 HR. DICKSON: No, it was a f ailure propagation 9 experiment, it was not prototypic in the sense of 10 flowing sodium and flowing water. There had been no 11 secondary failures in any system that had flowing sodium 12 and flowing water. To go on about these --
13 MR. ZUDANS: I guess that created another I 14 question in my mind. Those tests that are closed tubes 15 filled with water under a given pressure, under a 16 certain pressure, and if that tube would f ail the 17 pressure would immediately be lost.
18 MR. DICKSON: Yes.
19 MR. ZUDANS But there was really no kind of a 20 driving that you would have with your steam line being 21 at 1400 psi, having a steam drum which is capable to 22 expand?
23 MR. DICKSON Each tube is individually sealed I 24 off.
25 MR. ZUDANSa So you would expect less h
ALDERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
404 =
O i prooeoetion then in the others.
2 Did not the Atomic International test these 3 steam generators with real conditions in a loop?
4 MR. DICKSON: We have one in LNEC that is full 5 of sodium right now.
6 MR. ZUDANS: Full sized?
7 MR. DICKSON: Full sized, yes.
8 MR. ZUDANSa That was not tested for this 9 matter.
10 MR. DICKSON: I sure hope it is not.
11 MR. FBERSOLE: Well, how many tests have you '
12 done where you threatened tube f ailure which are f ull of 13 flowing sodium?
O 14 MR. DICKSON: None.
15 MR. EBERSOLE: Well, when you say there have 16 been no failures, you can't have any failures if there 17 have been no tests.
18 MR. DICKSON: I thought how many failures have 19 we had.
E 20 MR. EBERSOLE: You said there have been no 21 failures, and I am asking, in how many tests have there 22 been failires.
23 MR. DICKSON: I believe the number is seven of 24 those.
25 MR. EBERSOLE: Which did have flowing water?
O i ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
405 lll 1 MR. DICKSON: I believe that is the number.
2 MR. EBERSOLE: They did have flowing water.
3 MR. DICKSON: Yes.
4 MR. CARBONS I am mixed up here. I thought 5 you said there were no tests with normal flowing water 6 through the tubes.
7 MR. DICKSON: That had secondary failures.
8 MR. CARBON: How many tests have there been 9 where there was normal flowing water and normal flowing to sodium?
11 MR. DICKSON: I belive the number -- normal 12 flowing water and normal flowing sodium?
13 HR. CARBON: At normal pressures, under 9 14 prototypic conditions.
15 MR. DICKSON: Let me go back. Normal flowing 16 water and not flowing sodium there have been a fair 17 number of tests we have conducted that way, seven in 18 this country.
19 I do not know of any that have had both normal 20 flowing water and normal flowing sodium. Can anybody 21 help me on that? No one seems to recall. I have the 22 complete list of tests and the conditions.
23 MR. CARBON: Well, I don't care about the 24 exact number, I was trying to get some feel for it.
25 There have been a few, then, with flowing water and O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
406 h 1 stagnant sodium and none with flowing sodium and flowing 2 vster.
l 3 MR. DICKSON: None that I know of that would 4 be prototypic.
5 MR. CARBON: Does everyone feel real 6 comfortable with that? I personally, based on my 7 knowledge of the last half hour would not.
8 MR. DICKSON: Comfortable to the extent we 9 feel it is a worst case. If the water is now flowing 10 you certainly are more likely to get a f ailure than if 11 you had flowing water cooling it.
12 MR. CARBON: But conversely, if your water is 13 flowing you have the opportunity for a large amount of I
14 vater that comes pouring out into the sodium.
15 MR. DICKSONs Well, I should have noted that 16 in these tests there is a intentioni initial f ailure of 17 whatever size is desired. A small test with a small 18 hole drilled, or a large test that is suddenly opened to 19 expose the water. Thst tube has always had flowing 20 water in all the tests.
21 MR. CARBONa Oh, so that the water can come 22 out, the full amount, into the sodium to cause this 23 secondary effect.
I 24 MR. DICKSON: Yes.
25 MR. ZUDANS: I thought you just said the tubes I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554-2345
407 ,
1
() 1 were capped and had no supply of water.
2 MR. DICKSON: Except the one that was 3 intentionally failed.
4 MR. ZUDANS4 Now it explains why you did not 5 have secondary failures with the pressurized secondary 6 failure tubes. The only source of supply would be the 7 one that is being tested.
8 MR. DICKSON: Will you get the list of tests, 9 George? I do not have them.
10 MR. CARBON: Let's just take a short break.
11 (Whereupon, at 10 a.m . a ten-minu te recess wa s 12 taken.)
13 MR. CARBONa Let us proceed.
O 14 MR. DICKSON: This is a configuration if some 15 of the tests that I was not able to describe without a 16 picture, apparently. There is flowing water though this 17 tube and it has a pre-generated failure point of small 18 sizes to the large, full double-ended guillotine. There l
19 is a sleeve covering this.
20 MR. CARBON: Say tha t again, will you? It has j 21 a wht?
l 22 MR. DICKSON: It has a pre-machiend failure 23 point, either small for the small test or in some cases
( 24 a large test, a double-ended guillotine size.
25 MR. CARBON: When you say double-ended l
t
()
I
(
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D C. 20024 (202) 554-2345
408
() 1 guillotine size, you mean that it is such that you 2 rupture it in the fashion comparable to it?
3 MR. DICKSON: Yes, that is correct. It is a 4 size hole that is greater than this cross-sectional area 5 plus this cross-sectional area, so that the flow 6 simulates a maximum flow that you can get into such a 7 tube.
8 MR. CARBON From a double-ended guillotine.
9 MR. DICKSON: That is correct, the maximum to flow from any size break, what you get through the two 11 ends, which is a double-ended guillotine.
12 The sleeve is suddenly pulled aside in the 13 three British tests where they got secondary tube O 14 f ailures and not any of these other tubes failed. The 15 rupture on it was actually down near the bottom of the 16 sodium and the water was capped, as shown here. It not 17 only did not flow, it could not escape as it expanded.
18 One test that was like this ended up with more 19 than one secondary failure. That was one that was 20 completely capped and the full system raised to a 21 pressure at which ultimately several tubes popped over a 22 span of two minutes.
23 But for the quick-time test the most frequent
) 24 result is no s,econda ry ruptures at all, they do not 25 rupture. And I was not able to find in my list of tests O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
409 lll 1 how many actually had flowing water and how many had 2 stagnant water. I only know that in the one test where 3 it was more than one failure there was completely capped 4 water.
5 MR. ZUDANS: It says on your sketch brown, 6 secondary tube water. Were not those tubes filled with 7 sodium, the capped tubes?
8 MR. DICKSON: No, sir. The sodium is in 9 yellow.
10 HR. ZUD AN S : Yes, but that says water on the 11 legend that is written up there on the right-hand side 12 at the top.
13 MR. DICKSON: This is water.
O 14 MR. ZUDANS: Yes.
15- MR. DICKSON: The orange is water and the 16 yellow is sodium.
17 MR. ZUDANS: Oh, yes, correct.
18 MR. DICKSON: And the red is also water.
19 MR. ZUDANS OK.
20 MR. DICKSON: Water pressure.
21 MR. ZUDANS: The only thing that I guess you 22 would have in the sketch, the secondary failure would 23 not be backed up by a pressure in that tube, it would 24 relieve the pressure.
25 MR. DICKSON: That is correct.
I ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W WASHINGTON, D.C. 20024 (202) 554 2345
410 g 1 MR. EBERSOLE: This was not a pneumatically 2 pressurized tank over here to continue to feed?
1 3 MR. DICKSON: Yes, it continues to feed.
4 MR. EBERSOLE: Oh, it does?
5 MR. DICKSON: Yes, sir.
6 MR. CARBON: It would seem that if your sodium 7 vere flowing and sweeping a way your reaction products 8 you would get at least a different circumstance, 9 experiment, something, result, than if the sodium were 10 still. -
11 MR. DICKSON: Yes, sir.
12 MR. CARBONS Is it considered to be worse if 13 the sodium is stagnant in terms of added failure?
14 MR. DICKSON: Yes, sir.
15 MR. CARBON: Why is that?
16 MR. DICKSON: Because you could concentrate a 17 small leak against a neighboring tube in stagnant 18 sodium, whereas if the sodium is flowing it would of 19 course vibrate and not remain quite as stagnant. So, 20 this is a worst case.
21 MR. CARBON: That certainly would seem true 22 insof ar as it goes. But also it seems that there would 23 be a reverse effect or different effect in that if you I 24 sweep the reaction products away, you can get more flame 25 and one would tend to balance the other.
I ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
411
() 1 MR. DICKSCN I am not sure I follow how you 2 would get more flame. As this expands, if it is a small 3 leak, the flame front stays right there.
4 MR. CARBON: Yes, but if you fill up the flame 5 front, so to speak, with products, NAO-2 or whatever, 6 then you do not have any more flame.
7 MR. DICKSON: Yes,that is correct. For very 8 small leaks that have this kind of flame there were 9 tests influenced as well, with very, very slow erosion 10 of the next tube.
11 MR. BENDER: Is this work reported somewhere 12 so we could go look at it?
13 MR. DICKSON: Yes, there is a variety of 14 reports.
15 MR. BENDER: Do you have some consolidated 16 report with you that we could look at?
17 MR. DICKSONs There is a consolidated report 18 of the American tests, except for the last one.
19 MR. BENDERS Why don't you identify it for us, 20 so that we can take a look at it ourselves?
l 21 MR. DICKSONs I could provide them right now.
22 MR. BENDER: Well, if we put them in the 23 record, then we will know what they are.
() 24 MR. EBERSOLE Dr. Carbon's remark is mindful 25 of the fact that you can start a fire in space because ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
412
(} 1 the convection products simply stick around the flame, 2 and you do get bigger fires.
3 MR. DICKSON: In the case of a large leak,
()
I 4 those tests where we had a full double-ended guillotine i 5 rupture, the sodium does leave by virtue of rupturing a 6 rupture disc so that it does move.
7 MR. CARBONS It moves, but you don't have new 8 sodium coming in to react with water.
9 MR. DICKSON: There would not be in the case 10 of a large BEG in any event. The flame front is moving 11 in both directions away from the water, and the water 12 that can react has to react at the surface of the flame 13 front that is pushing sodium in both directions.
( 14 MR. CARBON: Can you be sure of that, though, 15 when your bulk sodium is traveling at some relatively 16 high velocity, has a lot of inertia?
17 MR. DICKSON: The bulk sodium is moving slow 18 relative to that bubble expansion. The bubble expands 19 in both directions, both sgainst and with the direction 20 of sodium flow, for those large ruptures.
21 MR. CARBON: You were going to identify that 22 report.
23 MR. DICKSON: Yes, sir. There are three of
() 24 them. J. C. Amos, et al., evaluation of LLTR Series 2 25 test, A-6 results, prepared for United States Department ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
413
() 1 of Energy under contract DEAT03-76SF0030, June '81.
2 D. E. Knittle, et al., the same title, except 3 A-7 results; same contract, September 1981.
4 J. J. Reginbal, et al., same title, except 5 A-8; same contract, Feb rua ry '82.
6 J. C. Amos, et al., same title, except A-3 7 results; the same contract except it is given as Work i
8 Package AF151005, WPT No. SG07, and no date.
9 Those are the major summary reports. I have 10 another 13 references of some of the small leak reports.
11 MR. CARBONa That ought to take care of it.
12 Is there any reference to the last test that you spoke 13 of?
14 MR. DICKSON: The results are still under 15 evaluation. It was completed in April and is not 4
16 finished with the post-test evaluation and analysis.
17 MR. CARBON: Paul, would you send copies of 18 those to all the members who have been here, Mike and 19 us, and Dave Okrent, and Zenon I am sure will want them 20 and Walt. Do you want any, Mike?
21 MR. TRIFUNAC: No, thank you.
22 MR. CARBON: Go shead.
23 MR. DICKSON: All right, in the case of where
() 24 we have had secondary tube failures, in four of the 34 25 large leak tests in several different countries, as I 1
O ALDERSON REPORTING COMPANY,1.'C, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 202) 554-2345
414
() 1 mentioned, the conditions were conducive to propagation; 2 there was no sodium flow.
3 The times of the second failure - and that 4 would be the second failure of one of those tubes, 5 ranged from two and-a-half, three and-a-half, four 6 and-a-half to 17 seconds.
7 The two and-a-half second failure was in a 8 vall thickness of 55 mills, and that compared to our 9 end-of-lif e wall thickness of about 77 mills or a 10 beginning-of-life wall thickness of 109.
11 There has also been an analysis - and it is 12 not very good but it is certainly on the conservative 13 side. What was taken was simply a flame front against O 14 the plate. There was no heat conduction in the hot 15 spots, so obviously no flowing water. Simply, how fast 16 do you heat up the wall of a tube with no coolant.
17 Steady state, so there is no flame fluctuation.
18 The reaction temperature was taken at a metric 19 maximum of 2700 degrees Fahrenheit. The flame 20 temperatures are normally much lower than that and the 21 highest that has been measured was in the range of 2200 22 to 2300 degrees Fahrenheit.
23 The results of this test, or this analysis
() 24 rather, is that you could expect the temperature of the 25 tube to rise to the point you no longer sustain the O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
415 1 pressure in something on the order of a half second to a 2 second, depending on the hea t transf er coef ficient. The 3 PSAR reports this as .3 seconds in the evaporator and .4 PO 4 in the superheater, but that used a completely arbitrary 5 and excessively large heat tranfer coefficient of ,
I 6 10,000. For the normal range it would be a half second 7 to one second.
8 But as I say, this is on the conservative side 9 completely, but certainly puts you in the range of the 10 two and-a-half second failure that happened in one case.
11 So, as we see it, ther largest event that is 12 really plausible would be as precursor that would raise 13 the sodium pressure to 150 psig for the whole system, 14 and in a primary failure of about half the equivalent 15 DEG - and I guess I did not mention that earlier - all -
16 secondary failures that we have seen are less than a 17 half in equivalent DEG with one exception, and that was 18 in the one that I said had several failures subsequent 19 ot the sweep-out of the sodium.
20 NR. CARBONS What is DEG7 i
21 HR. DICKSON: Double-ended guillotine.
22 So, most of them in fact have been more in the 23 order of 25 to 30 percent of DEG , of the la rge O 24 fa11eres. And rou 1 ht expect then a secondarr fa11ere 25 after that, some two and-a-half seconds later. The ALDERSON REPORTING COMPANY,INC,
- 400 VIRGINTA AVE, S.W., WASHINGTON. D.C. 20024 (202) 554 2345
416 l 1 earliest we have seen in experiments are not much 2 greater than analysis.
3 MR. CARBON: Once again to be sure I 4 understand that, the precursor tube is the first to go 5 and it breaks; and the primary tube, then, is assumed to 6 break at the same time. So, you have two at the same 7 time.
8 MR. DICKSON: No, this precursor has been 9 putting a flame front on the primary. This is a small 10 leak on the order of a pound a second, raising the 11 pressure of the system to 150 psi and then this one 12 breaks suddenly.
13 MR. CARBON: And it is the first one, then --
14 MR. DICKSON: The first one that will initiate 15 the SWRers event. As I said, if this failure does not 16 occur, this will ultimately trip the system at the 150 17 psi I pointed to.
18 MR. ZUDANS: Now, if the system pressure is 19 tripped by this first precursor before the failure 20 occurs in the tube, you would be essentially in the same 21 condition with a lower sodium pressure.
22 It would appear to me th ey would not really 23 make any great deal of difference in terms of ultimate l
lh 24 need to dump the whole system anyway.
l 25 MR. DICKSON: Maybe not in terms of need to D
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
417
() 1 dump the whole system.
2 MR. ZUDANSs Well, the second line or third 3 line of failure vill dump the system automatically.
O 4 MR. DICKSON: If this trips, it dumps the 5 water. It dumps the water, not the sodium.
6 MR. ZUDANSs It does dump the water?
7 MR. DICKSON: It does dump the wa ter. If the 8 rupture disc goes, the water dumps. Now, obviously, as 9 soon as the water pressure starts to go off, this leak 10 begins to reduce and whatever it has been heating up.
11 MR. ZUDANS: So, then you would have a lesser 12 event because you would not have the feed of the water.
13 MR. DICKSON: That is right.
14 MR. ZUDANS: That would mean that it would be 15 adventageous to have this precursor act on the water 16 side.
17 MR. DICKSON: Iou mean have it trip the 18 rupture disc?
19 MR. ZUDANS: Right.
20 MR. DICKSONs Oh, yes. In that case it is an 21 event but it is not a design basis enveloping event.
22 MR. ZUDANS: That is right.
23 MR. DICKSONs Wha t we are searching f or is,
(]) 24 how bad can it be that we have to design for. In fact, 25 a leak of the kind we are talking about here of a pound O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
r418 O - =ecoaa i= < ener ea#r o== 1e x aa the overetor 2 should have responded to this long before it got to that 3 size. They should have responded and taken action.
4 MR. ZUDANS: And this pound per second leak, 5 how long would it take to build up the pressure to a e rupture pressure?
7 HR. DICKSON: When this one goes?
8 MR. ZUDANS: On the next one, the upper one.
9 The first line.
10 MR. DICKSON: The upper one?
11 MR. ZUDANS Yes.
12 HR. DICKSON. It is a matter of seconds.
13 MR. ZUDANSs A matter of seconds.
14 MR. DICKSON: At a pound a second.
15 HR. ZUDANS: So, you cannot really expect 16 anything to be done by operators during tha t time.
17 HR. DICKSON: Not when it got that high, that 18 size.
19 Actually, you have somewhat of an almost 20 impossible sitation. The size we are talking about 21 where it is going to cause this large failure on the 22 order of a pound a second, it should raise the pressure 23 of the system fast enough to rupture the rupture disc --
24 MR. ZUDANS: That is correct.
25 MR. DICKSON: -- much earlier than it can O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W, WASHINGTON, D.C. 20024 (202) 554-2345
h 1 cause a failure here.
2 MR. ZUDANS4 Now, when it ruptures the rupture 3 disc, how much time does it take to start the signal and 4 perform the dump of a water?
5 HR. DICKSON: Within 30 seconds the water is 6 down to 300 psi.
7 MR. ZUDANS: That is then too late because the 8 secondary falure, if it occurs, will long since*have 9 occurred.
10 MR. DICKSON: But almost instantaneously the 11 pressure begins to drop, within a second it is down 12 significantly.
13 MR. ZUDANSa But it is still auch higher than D 14 the sodium pressure.
15 HR. DICKSON: Yes.
I 16 MR. ZUDANS: So, it may not be such a great 17 help.
18 MR. DICKS 0ha Well, it certainly will do one 19 thing for you, if this broke it would be starting as 20 pressure wave in the low-pressure system.
21 MR. ZUDANS: That is right.
22 MR. DICKSON: So, with that as a possible, let 23 me describe what our design basis sodium-wa ter reaction 24 event is.
25 Gi ven the fact that the number of experiments D .
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554 2345
420
() 1 we have is certainly not exorbitant and the complete 2 understanding of this is understood but not subject to 3 analyses, we take this precursor - and this is really 4 the unrealistic thing, it raises the sodium pressure to 5 325 psig, just below the rupture disc of the SWRers much 6 above 150 over here.
7 Then we get the primary failure of one 8 equivalent DEG at time zero, and we take a full 9 equivalent DEG even though none had been seen in any 10 experiments.
11 We then take the secondary failure which, as 12 you saw, has been seen occasionally but not norrally, 13 and we take that and another equivalent DEG one second O 14 later. And then just again because we do not really 15 understand all this and you asked what about another 16 one, we will give you another one, a third equivalent 17 DEG at two seconds after the first one.
18 We do not consider any further ones because 19 within three seconds of this, following this sequence, 20 the sodium is swept out of the evaporator that is l 21 involved or some heater that is involved.
22 MR. EBERS01Es Tell me because you have 23 started this as 325 psig, it is just like a reactor
() 24 flux, the trip level; you are already very near your 25 trip-set point of the rupture disc. All it takes is O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
I 421
) 1 just a little flip and it is done.
I 2 So, you have preconditioned this to '
3 instantaneous release.
4 HR. DICKSON: On the contrary. As you 5 men tioned ea rlier, we have to do a dynamic analysis. If 6 you hold it to a lower pressure and then let the event 7 occur, you get much higher loads in the pipe, or much 8 lower loads in your piping system.
9 This gives you the worst because even though 10 the rupture disc ruptures, what you are tracking is a 11 sonic wave --
12 HR. EBERSOLE: It is the waves, OK.
13 HR. DICKSON: And that rupture disc is not O 14 doing you very much good with a sonic wave going down 15 the reflection it meets coming around the other side.
16 You are past the rupture.
17 HR. EBERSOLE: Yes. So, you know this is the 18 worst end of the spectrum.
19 HR. DICKSON: That is right. If this were a 20 slow event what you say is correct, but it is nots it is 21 a milli-second type of event.
22 HR. CARBON: You said that three tubes would 23 be the maximum number because by two seconds the sodium 24 is already swept out. I thought it was the water that 25 you were sweeping out and getting rid of.
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
422
() 1 MR. DICKSON: We do not take the sodium out 2 with any sodium dump. That is a manual operation later 3 if you decide you need it.
4 What is sweeping the sodium out of here is 5 this expanding gas bubble. Wnat we have going in at 6 this juncture is 12 and-a-half pounds per second from 7 the equivalent DEO plus about a pound per second from 8 the precursor, 13 pounds per second, generating a very 9 large hydrogen bubble in a hurry.
10 .
MR. CARBON: So, you end up with like 35, 40 11 pounds of sodium per second. You say then there is a 12 big bubble or not?
13 MR. DICKSON: Yes, it is expanding so that it O 14 cleans a six-foot long tube in a matter of three seconds.
15 MR. ZUDANSa Supposing the precursor causes 16 this 150 psi pressure rupture disc to break, that would 17 provide additional volume for the sodium to expand 18 before it builds up the pressure. How would that affect 19 this sweeping out on a primariy failure, subsequent to 20 primary failure? Because you have to build the pressure 21 to make the other disc to what?
22 MR. DICKSON: 325 is what breaks the SWRet 23 discs.
24 MR. ZUDANS: And how big does the bubble have 25 to grow before it can build that pressure?
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
f423 O 1 HR. DIcxS0*. The shock weve coming off of 2 this turbine DEG will break that.
3 HR. ZUDANS: The shock wave will break it? I 4 don't know. I guess I don't know where it is located.
5 HR. EBERSOLE: After all this is over and has 6 settled down, what are the complications, if any, in 7 resto ring the system to a safe condition; how long does 8 it take? What does it imply you have to do? Is it a 9 very messy event?
10 HR. DICKSON: Yes, sir. If you pop the SWRers 11 rupture disc for this large sodium-water leak it is 12 considered a faulted event or that loop of piping, an 13 emergency event, a faulted event that affected the O 14 module; and emergency event for the remainder of the 15 loop. You must clean up the whole SWHer system.
16 HR. EBERSOLE: And how long would that take?
17 HR. DICKSON: I don't have a time, but to go 18 in and remove a tank and replace it; check all the 19 components that were involved in the emergency event to l 20 make certain that they have not-sustained damage, and 21 replace the steam generator. I believe you are talking 22 a month.
23 HR. EBERSOLEs I really am getting to the 24 point of the nervous characteristic of the system and 25 aspect to recovery to a normal working state.
O ALDERSoN REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
, 424 g 1 MR. DICKSON: We provide this safety 2 capability'for all eventualities. You should never get 3 to this stage. Any small leak, before it rose to that 9 '4 size, will be clearly detected. We have redundant, bio 5 detection systems. We have operators that should act, 6 cesponsible individuals that have plenty of time to 7 act. We should never get to othis event.
8 MR. EBERSOLE: Have you, or do you intend to 9 do a PRA study to estimate the frequency of this event 10 in the context of whether this whole show is a practical 11 show or not?
.12 MR. DICKSON We are doing PRA studies. When 13 you say whether this whole show is a practical show, you 14 are talking about.the potential down-time involved?
15 HR. EBE350LE: All that.
16 MR. DICKSON: Those really fall under what I 17 c$nsider the reliability availability studies and yes, 18 ve are doing those, too. As I said, this is a highly 19 unlikelqievent. We designed for it.
20 MR. EBERSOLE4 Yes, there are a number of 21 events of less consequence in front of this, I guess, in 22 other words, the spectrum. Do you display this sort of 23 thing to convince yourself'you have a practical h 24 experience?
25 MR. DICKSON4 Yes, sir. As I said, the small 9 '
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D C. 20024 (202) 554-2345
425
() I leak is part of our anticipated event scenario. The 2 large leak is a highly unlikely event.
3 MR. EBERS01Es To your planning that was based 4 on a detection of the leak, tha t was really at the 5 pin-hole stage or thereabouts. So, there is no real 6 pressure propagated; is that correct ?
7 ER. DIOKSON: That's right. All the evidence 8 is -- well, you do have to allow for this.
9 HR. ZUDANS While describing the clean-up to procedure you said you would replace the steam 11 generator. Do you plan to replace it in case of a 12 single tube failure?
13 MR. DICKSON: Not in the case of a single tube O 14 failure we would not repice it. We would simply plug 15 the tube. After a SWRets event I do not believe that it 16 is acceptable to leave the steam generator.
17 NR. ZUDANS But you only have subjected, say, 18 three or four tubes to some damage, according to this 19 assumption. Why would you replace three or four tubes 20 with 800 other tubes?
21 MR. DICKSON4 You are correct, it is possible 22 that you would not have to remove the steam generator.
23 MR. ZUDANSs I think it is awf ul to get into
() 24 this mess.
25 HR. DICKSONa I think that is such a terrible ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
426 l
() 1 mess, it is one that you do not anticipate and it is a 2 significant kind of event. But even if you did not have 3 to replace the steam generator --
4 MR. CARBON: Maybe we better go ahead, time is 5 running out.
6 MR. TRIFUNACs Can I ask a question?
4 7 MR. CARBON: Yes.
8 MR. TRIFUNAC4 These tests are done in a 9 static manner, there is no vibration going on.
- 10 MR. DICKSON That is right.
11 HR. TRIFUNAC: Now, supposing that you have a 12 very fine fracture or a weakness point somewhere, do you 13 have any idea how much worse the situation is compared O 14 to the crack --
15 MR6 DICKSONs From a crack?
16 NR. TRIFUNACs So that time time which you are 17 talking about in which an operator would see something 18 and detect a small leak would be that auch more reduced?
19 MR. DICKSON: I suppose that potential 20 exists. That is under the scenario that I mentioned 21 earlier of a flaw gradually propagatign to a sudden 22 rupture tha t induces this event. It is not quite as bad 23 as when you had the first precursor because it raised
() 24 the systen pressure. But it would fall under this 25 category.
- O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W, WASHINGTON, D.C. 20024 (202) 554 2345
427 lll 1 MR. TRIFUNACs If you look, for example, at a 2 typical pipe that is welded and has a bend and so forth, 3 what sort of vibtstion would be associated with the 4 portion where a fracture may occur? I am sure that it 5 is not just supported at one end and the other, there 6 are also intermediate supports.
7 Do we have any idea of what frequency?
8 MR. DICKSON: Yes, we do. There is a variety 9 of frequencies, depending upon the particular supports.
10 In fact, the seismic analysis of this system has to be 11 done with the time history of the system in order to 12 take account of the fact that not only do we have a 13 given frequency but the fact that if they are not D 14 rigidly supported, say around a bend point, that bend 15 point has a wide range of frequencies. Most of them are 16 not in resonance with the tube shell, some of them are 17 because it spans it.
18 If you do t response spectrum linear analysis 19 you find a sizeable margin of safety in the average 20 tube, but you also find tubes that are supposedly in 21 frequency and are giving a different kind of response.
22 They cannot stay in frequency because they are not in 23 f act locked, they will move off of their support and h
m/ 24 detune themselves.
25 That is why a dynamic time history analysis q
ALDERSON REPORTING COMPANY,INC, l
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 I
428
() 1 must be done and is presently under way. I do not know 2 the results, obviously because it is not done. But if I 3 base it on exerience of when we have done the time 4 history anslyses compared with prior response f actor 5 analyses, I would expect to find a margin of safety 6 better than 80 percent and that is at and of lifes at 7 the beginning of life it is more like a hundred percent.
8 NR. TRIFUNACs Could you comment on what would 9 be the frequency range associated with, say, a dominant 10 smount of vibration that you find in the analysis?
11 NR. DICKSON: I just know in that one region, 12 that is the only one that I can recall, and it runs in 13 the order of about three or four hertz to over 70.
14 NR. TRIFUNACs So, it goes out to very high 15 frequencies.
16 NR. DICKSON: Yes. But that is a wide range.
17 I do not know what the frequency contents between the i 18 supports are.
l 19 HR. ZUDANSs One quicky on design. If I 20 remember correctly, the tubes in the bent portion do not 21 see the sodium, there is another head in between that 22 blocks the sodium.
23 NR. DICKSON The stagnant sodium.
() 24 NR. ZUDANS Yes. Do you have the details of 25 that other separating head that is in the straight ALDERSON REPORTING COMPANY,INC, 400 VIRGINTA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
+
429 ll 1 portion?
2 MR. DICKSON: It does not really separate.
3 MR. ZUDANSs No?
4 MR. DICKSON: No.
5 MR. ZUDANS: I thought that this bent portion 6 was filled with a cover gas.
7 MR. DICKSON: No. It is filled with stagnant 8 sodium.
9 MR. ZUDANS: Filled with stagnant sodium.
10 In that case that particular transition is not 11 attached in any way to the tubes, it is freely 12 flo a ting . Now, there is another head before the curve 13 begins.
14 MR. DICKSON: Well, it is not really a 15 transition in the way you are thinking of it. There is 16 a shroud. The sodium enters, has to float up for a 17 while to get over the shroud where it can get into --
18 MR. ZUDANS: And beyond that shroud there is a 19 closure of some kind.
20 MR. DICKSON: No, sir.
21 MR. ZUDANS: I see.
22 MR. DICKSON: The only thing that you have is 23 spacers, spacer plates that act as somewhat closures.
I 24 They perform restrictions to the flow but they are 25 really spacer plates to hold the tubes apart. They are h
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
430
() 1 not in there to prevent the flow of sodium. There are 2 spacer plates even with the sodium closed.
3 NR. ZUDANS: So, anyway, there is no physical 4 attachment in tubes of any kind that would provide other 5 spots or something like it.
6 HR. DICKSON: The tubes go through spacer 7 platess that allow some action.
8 HR. ZUDANS OK, thank you.
9 NR. DICKSONs Well, to quickly summarize, then.
10 We have a precursor that has gone on for some 11 tens of seconds to pressurize the system. In the one 12 case, in our design basis, we pressurize at 325. The 13 aost plausible is something below 150. The more O 14 plausible event is about half an EDG or less - when you 15 take the upper limit of wha t you generally see. In two 16 and-a-half seconds, though, you get another one. This 17 is the event 'that you would think would be a reasonable 18 envelope of what you might expect in one of the worst 19 cases.
20 But what we hae taken for a design basis event 21 is this first one to a full DEG, the precursor is 22 pressurized to 325 psi; the second one, one second 23 later; and the third one, one more second later, that is
() 24 what is analyzed.
25 Compare that with what foreigners have done, O
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S W., WASHINGTON, D.C. 20024 (202) 554-2345
431 i
() 1 and you are always asking the question - but of course 2 you would not ask if I did not have it today "What is 3 your comparison with foreign experience?"
4 The U.K. takes the DEG all right, and they 5 take three of them at or.e second intervals, but not as a 6 licensing design basis accident. If you ask me any more 7 about that, I just told you all I know.
8 Germany takes one; France takes one - and I 9 have to put a caveat on that. That, I know, is 10 specified for Super Phoenix and I do not what they use 11 for Phoenix.
12 Japan uses four. We do not know their 13 interval. Again, they say only one for a licensing O 14 purpose but four for sizing, SWPers, dump tanks, and 15 what not. -
16 And we tak e three at one-second intervals. By 17 "we" I mean Clinch River.
18 HR. ZUDANS4 Do you have any information on 19 the French experience when they had problems with the 20 steam generator?
21 NR. DICKSON: The leak apparently was not so 22 large. But them somehow instead of back-flushing with 23 nitrogen they got air in there and caused themselves
() 24 even more damage. Does anybody recall reading that?
25 I read it. It was a damaging lea'. but it was l
l
()
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
432 I- 1 not a DEG type of leak. They shut down and then, ;
2 instead of getting a nitrogen back-flux they somehow got 3 air in there and that complicated the problem. I g
4 MR. BENDER: What kind of experience do the 5 Russians have?
6 MR. DICKSON: I know the Russians had several 7 large experiments. I do not know what sort of 8 experience they have had with steam generators.
9 MR. BENDER: I had the impression that they 10 had had some leakage problems.
11 MR. DICKSON: I remember. It seems to me that 12 I heard of one steam generator problem they had on 13 BOR-60. But I don't recs 11 and I have not heard of any I
14 on their larger plants.
. 15 Well, by way of summary - and I really used up 18 the time - our large event sodium-water reaction is 17 conservative both experimentally and analytically - both 18 from the standpoint of pressure we take, the size of the 19 first f ailure and the timing and size of the second 20 failure. The very existence of the third f ailure has 21 never been seen and no experiments have been done and 22 com pa red to foreign design basis events.
23 Thank you.
24 MR. CARBON: I have some general questions.
25 In terms of sensitivity, how far could you go beyond I
ALDERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
433
() I three tubes before you would start having something that 2 would sffect the core?
3 MR. DICKSON: It never affects the core. You 4 still have two loops to :) move the decay heat.
5 MR. CARBONS What I am really saying is, how 6 isolated is this? Excuse me, go ahead.
7 MR. DICKSON: It is isolated to the extent 8 that the only thing the core sees is'when the IHX of the 9 steam generator system that is involved quits removing 10 heat. That sodium comes back warm, so that the core 11 which is cooling down will begit to heat up again. It 12 hea ts up either something between 90 and 270 degrees, 13 depending on how much mixing one gets in the plenum.
O 14 In either case, even if it heats up to the 15 full 265 to 270 degrees, it does not reach the stay 16 state temperature. So, there is no problem with the core 17 NR. CARBON: My question is really a little 18 bit different. Could you postulate a number of tubes 19 with a guillotine break such that a large pressure would 20 build up?
21 Could you have enough tubes breaking that you 22 could build up enough pressure in the intermediate heat 23 exchanger that this pressure would damage it and reflect
() 24 through to the core under any conceivable scenario?
25 MR. DICKSON: The pressure you get in the IHX O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
434
() 1 depends on whether or not this scenario we used is in 2 the evaporator or the superheater. In the worst case I 3 believe it is a little over 400 psi, well under the
'( }
4 700-pound design pressure.
5 If you had more breaks involved than our 6 design basis, it is obvious the pressure would go up.
7 It would not go up linearly because that expanding 8 bubble simply cannot react the extra portion of water.
9 So I really, without analyzing it, could not 10 tell you at what point we would have a problem with a 11 potential breakage of an IHX. We do believe that our 12 design basis event is meant to envelope anything you can 13 expect.
. 14 NR. CARBON: Yes, I know, but I as just trying 15 to get a feeling for the sensitivity of it.
16 NR. DICKSON: One thing I did not add, there 17 are two other conservatisms in the analyses. The
- 18 analysis of the pressure wave that goes down the tube 19 was done by Transrep, which has been verified by some of l
l 20 these tests; verified to the extent that it predicts 21 well along straight planes; yet under-predicts the 22 dissipation going around elbows.
23 In fact, the dissipation could be 10 to 30 24 percent experimentally and Transrep comes up with only 25 three because it really does not go around an elbow, it O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINTA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
l 435
() 1 is a one-dimensional code. What it does is simply have i
2 a loss as it goes down the pipe. l 3 So, that is another conservatism. And another 4 one is that we assume 65 percent of the water is 5 reactive to sodium, where experimentally it is more like 6 50 to 55 percent.
7 MR. CARBON: If you have a thousand pounds 8 pressure in the secondary side there which reacted on 9 the intermediate heat exchanger, I guess you said the to intermediate heat exchanger would tolerate something 11 like 700 pounds or some such thing.
12 MR. DICKSON: That is the so-called emergency 13 limit. If you have a thousand pounds you would have O 14 exceeded the limits and we would have a hard time 15 proving to the NRC that we should continue to operate 16 afterwards. But it would not rupture, not in reality.
17 HR. CARBON: Are the steam generators built to 18 tolerate the SEE?
19 HR. DICKSON: Yes, sir.
20 HR. CARBON: Some of Dr. Trif unac 's 21 calcualtions will indicate that the SSE is not all that 22 unlikely. How much beyond .25 G would it likely 23 tolerate?
() 24 HR. DICKSON: We did a margin report on how
' 25 auch beyond that we could stand, and I don' t recall O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20 6 (202) 554 2345
436 I 1 whether or not we looked at the steam generators. We 2 had not done the analysis even at the .25 G other than, 3 as I mentioned, to simplif y analyses with response g
4 spectrum on a linear system.
5 In that analysis at .25 G our. average tube had 6 an 8', percen t margin of safety. If I assume - and I 7 have to assume it because we have not done the analysis 8 - that when we do the time history - not a linear 9 analysis - that we get an equal margin of safety in all 10 tubes, say 80 percent at the end of life, which is over 11 a hundred percent at the beginning of life.
12 It is obvious that we could take something 13 significantly larger as an SSE without losing the steam 14 generators. Exactly how much larger is a rather lengthy 15 study. If you look at the analysis that was done in the 16 .25 G, there is a significant amount of amplification 17 going to the steam generator building because it is 18 elevated to give you the natural circulation capability.
19 In doing so it passes through a lot of 20 concrete, none of which is assumed to crack or fail in 21 the analysis. You do get some amout of damping but no 22 failure. As you get above .25 Gs, the effect on the l 23 steam generator will not go up linearly. The .5 G input 24 would not double the load because some of the concrete 25 will crack on the way up, there will be a greter amount l
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINfA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
437 O ' or a ata -
2 So, if we are talking about a margin of safety 3 of somewhere between 80 percent and a factor of two, 4 realizing that you can take something greater than .25 G 5 on a non-linear way. In a way that helps you. I would 6 say we have a significant margin.
7 MR. CARBON: We would have asked about the 8 foreign slide if you had not put it up there.
9 HR. DICKSON: I am glad you would have.
10 MR. CARBONa We will make you feel good.
11 But it is my impression that people throughout 12 the world tend to feel that the steam wa ter problem in 13 the steam generators is truly not a serious safety O 14 pro blem . Is that so ?
15 MR. DICKSON: I regard it more as a serious 16 availability problem that was talked about earlier. One 17 of the major requirements of the element of the PR 18 industry is to get its steam generators to a highly 19 reliable position so we can make these plants 20 economical. If we have very many leaks it is not going 21 to be good for the economy. I regard it more from that 22 aspect.
23 I believe that from a safety standpoint we
) 24 have provided systems and redundant systems, and even 25 further redundant systems to take care of the worst case I
U i
l l ALCERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345 t
I 438 h 1 from the standpoint of a safety problem.
2 MR. CARBON: Do you know, do other nations 3 have the same view on it, the technical people?
g 4 MR. DICKSON: I would have to judge that from 5 the fact of the U.K., for example, that a nuclear 6 sodium-water reaction is not even a licensing event. I 7 just assume from that that they must take a similar 8 view. They do want reliable steam generators.
9 HR. CARBON: Can I ask the staff, have you 10 carried out your snalysis very f ar, can you commen t on 11 everything that has been said here today?
12 MR. STARKs Richard Stark from th e staff.
13 The staff review is in process right now. We 14 are using several consultants, INEL and O'Donnell. We 15 have had some questions that we have sent to the 16 applicant. We have anothher meeting set up for the last 17 week of July, or in the last week of July, on steam 18 generators.
19 We have a way to go yet, but we still have 20 time.
21 MR. CARBON: Do you have any conclusions or 22 thoughts at this time?
23 MR. STARKa No. I think we have been doing I 24 something very similar to what you have been doing. We 25 have been asking more questions in an attempt to get our I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554-2345
439
() 1 feet on the ground before we make any conclusions. l 1
2 HR. ZUDANS: On the sodium dump time, what is l l
3 its design pressures do you remember? I 4 HR. DICKSON: No, I don't remembe r. But I 5 would assume it is on the same order as the rest of them.
8 MR. ZUDANSs The IHX is the same.
7 MR. DICKSON: Yes.
8 NR. ZUDANS It could depressurize by the 9 second break. And that the steam generator goes through 10 several floors. It is a very long hockey-stick type of 11 device. How many supports does it have along the 12 vertical axis?
13 HR. DICKSON: It has two major shell O 14 supports. One right here and- one at the bottom that 15 really is not shown, that is a restraint. This is the 16 only major vertical axis support a though there is some 17 provided at the bottom.
18 There is also a restraint here for the pipe 19 leak problem. You can get a steam line break which can 20 put loads on it.
21 HR. ZUDANSs Thank you.
22 MR. CARBONa Nike?
23 MR. TRIFUNACs Yes, I have a comment regarding
(} ,
24 the discussion of the SSE analysis. l 25 HR. DICKSON: I thought you might. j 1
i
- I ALDERSoH REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
l l 440 1
() 1 MR. TRIFUNAC: It is just a general comment 2 which has been on my mind for some time and does not 3 really just apply to steam generators but generally to j 4 high frequency equipment.
5 I think one of the problems we have, as I see 6 it now that these -- have essentially relatively flat, 7 constant acceleration f or the f requencies and are 8 somewhere higher than, say, 20 hertz or so. Recently we 9 have had an increasing number of measured very small 10 earthquakes nearby, of very small magnitude, of very 11 high accelerations and high frequency content.
12 Which would suggest that the specification of 13 the measurement that goes into the building, which is O 14 primarily motivated by the civil engineering end of 15 analysis, may be somewhat misleading for the equipment 18 analysis in the following ways 17 For example, you may get accelerations that 18 have two times or maybe three times of what "SSE" 19 accelera tion would be from the -- point of view, a very 20 small, very short-lasting excitation , say two , three i 21 seconds, not longer. But maybe two times the -- whose 22 frequencies are not of interest or damaging for the 23 concrete containment, civil engineering types of systems.
( 24 But if you look a t something that had a 25 frequency of 50 hertz as those might be, I though t I O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
441 h 1 might just mention this because this is one case where 2 we are sort of formally and literally takin g what comes 3 off of civil engineer instructions may be somewhat 4 misleading. It is not necessarily the worst case.
5 HR. DICKSON: You mentioned that the last time 6 that we should maybe more concerned about the high 7 frequency omplaments in this plant, snd that they may 8 not be treated on the same basis.
9 I wanted to ask you then - I will ask you now 10 - what is the attenuation characteristic of these high 11 frequencies through the concrete? I was thinking more 12 through the concrete.
13 HR. TRIFUNAC: We don 't know because we don't 14 do the analyses, you see. If we are doing analyses for 15 a containment we would essentially ignore the high 16 frequency because the very specification of excitation 17 function is non-existent for frequencies higher than 18 ab at 25, 30 hertz.
19 In the Reg Guide 160 there is a magic number 20 of 33 hertz.
21 HR. DICKSON: Yes.
22 MB. TRIFUNAC: But in reality all the motion 23 that we deal with has no energy outside 25 hertz. ,
t 24 HR. KERR But by its very nature a massive 25 structure is going to attenuate our frequency very much, N
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
442
() i isn't it? I do not see how one can avoid that.
2 HR. TRIFUNACs This has been suggested by n 3 number of people. In principle it sounds a very 4 rea sona ble thing to expect that it should be so. But 5 some of the measurements and experiments we have done to 6 verify this sugget no significant reduction at all in 7 the frequency range that they think this should already 8 take place in.
9 NR. KERR4 I would be suspect of the to measurements.
11 HR. TRIFUNAC4 I am sorry.
12 HR. KERR I would be suspect of that sort of 13 seasurement.
O 14 ER. TRIFUNACs Well, we have several.hundred 15 recordings on buildings of different sizes, and we 16 looked at the question of whether --
l 17 NR. KERRs You do not get attenuation f rom the i
- 18 aarth to the structure?
- 19 ER. TRIFUNACs You don't get attenuation which i 20 is significant and which can be picked up by the j 21 measurements. I am not saying it is not there.
22 If you do calculations, calculations suggest l
23 that this should take place at rela tively low
() 24 frequencies, and tha t it should be quite significant at 25 maybe 20, 30 hertz , if you do theoretical O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
443
() 1 calculations.,
2 But then you look at the neasurements and it 3 is not there. And it has to do with the fact that 4 probably the relative stiffnesses of these concrete l
5 structures are not that stiff compared to the rock they 6 a re sitting on. So, the experiments themselves will not 7 support this.
8 NR. DICKSON: Can you refer us to a paper that 9 would shed some light on this?
10 ER. TRIFUNACa There are several papers. I 11 have a paper that just came out recently in Dynamics of 12 Earthquake Engineering. That just recently came out. I j 13 can send you a copy if you like.
14 HR. DICKSON: Thank you, I would appreciate it.
15 MR. ZUDANS May I comment? The higher 16 frequency content that is introduced to the seismic 17 event will propagate through the structure without 18 amplification, essentially.
19 HR. TRIFUNAC Yes. And that is the answer 20 too , tha t you would have to look at the individual pipes.
21 HR. ZUDANS Yes.
22 MR. TRIFUNAC: The fact is that the 23 amplification of lower f requency range is significant.
() 24 They may see here G loads f, rom a seismic event that 25 might be one G or maybe even two Gs because of
()
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W WASHINGTON, D.C. 20024 (202) 554 2345
444 1 structural amplification at the lower frequency range.
2 So, it may not be very significant to worry.
3 But to bring the matter to rigid body acceleration, that 4 should be proper.
5 MR. ZUDANSa That is true, I agree. But that 6 seems like a pseudo-static force, perhaps, of one or two 7 Gs. But I am looking at a different situation where 8 you have a very short-lasting but possible comparable 9 level of acceleration that has a strong dynamic effects 10 which might tune in with some of the frequencies that we 11 are talking about here.
12 MB. CARBON: Let's move on. Jesse, did you 13 have a question?
j O 14 MR. EBERSOLEa One topic that you did not i
15 mention was; during all these transients you mentioed 16 thermal in-points and so forth, but you never did 17 mention thermal gradients. And I thought that when you 18 a;e dealing with sodium that is always a present problem.
19 Are there any thermal gradients associated 20 with this event that are significant?
21 MR. DICKSONa None are significant. As the 22 primary system scrams, the sodium begins to reduce in i'
23 te m pe ra ture .
( 24 MR. EBERSOLE: What about when you dump sodium 25 in the dump tank?
O .
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
445 lh 1 MR. DICKSON: In general the temperatures -- l l
2 MR. CARBON: Hold up a minute. The recorder 3 cannot hear.
g 4 MR. DICKSON: In general, the temperatures are 5 going down in the primary loop. Then there is a 8 gradient of some 260 degrees between the one loop that 7 is affected, remaining hot as this hot sodium goes into 8 the inner plenum of the reactor vessel, compared to the 9 other two coming in.
~
10 That is not too strong a gradient. So, your 11 dump tanks are not going into play until after the 12 operator determines that he wants to dump sodium. In 13 the case of a small leak, if he dumps sodium at all, it 14 will only be after he has let the sodium cool down to 15 some 800 degrees at least.
16 But in the event that he does want to dump 17 sodium fairly ra pidly, the dump tank and the pipes 18 leading to it are kept with trace heaters so they are 19 varmed up. I believe they are held at something like 20 500 or 600 degrees so that there will not be a great 21 thermal shock.
22 MP. EBERSOLE. ThanL you.
23 MR. DICKSON: The SWPers tanks will see a I 24 shock. They are not preheated, they will see a shock.
25 But they are throw-away items anyway.
k ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554-2345
446
) 1 MR. EBERSOLE. Thank you.
2 MR. CARBON 4 Any other questions on this 3 topic? I think we had better move on, then. Thank you, 4 Mr. Dixon.
5 MR. CLARE: Good morning, I am George Clare 6 from Westinghouse. The topic that I will be addressing 7 is one which has come up in various ways several times 8 before the subcommittee; that is the question of the 9 plant's response to a postulated blackout.
to A blackout in the term we are using it here is 11 a complete unavailability of all of our off-site power, 12 the connections with the TVA grid, and the 13 unavailability of all of our bulk AC power from our 14 emergency on-site sources which are diesel generators.
15 In addition to the fact that this has come up 16 before this subcommittee before, I am sure you are well 17 aware that it is also one of the topics of the 18 unresolved safety issues which the staff keeps a list of 19 and is working on. If you are interested, it is Number 20 844.
21 MR. MARK: Will you be telling us what the 22 experience is in attempting to start, having diesels 23 start, either yours or generally, in the course of your 24 discussion?
25 MR. CLARE: No, I was not planning on ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
447
() 1 addressing that in any detail.
2 MR. MARK: Well, I did not want a lot of 3 detail. I just wanted to know -- every time you have 4 off-site power go off, the diesels are supposed to start.
5 MR. CLARE: Yes.
6 MR. MARK What is that number?
7 MR. CLARE: I don't know what a number would 8 be that would represent the experience of either nuclear 9 power plants in general or power plants today. I know 10 that as part of the aceptance test for the plant we have 11 to start our diesels a hundred times in succession in 12 accordance with the Regulatory Guide - I forget the 13 number - in order for the staff to consider them to be O 14 sufficiently reliable.
15 MR. KERRs Since Dr. Mark expressed some 16 c u rio si ty , let me give at least one data point. I am 17 familiar with one light-water plant that has had six 18 off-site power losses, and in two cases only one diesel 19 sta rted . In the other four cases, both diesels started.
20 MR. CARBON: Fine. Let's go ahead, then, Mr.
21 Clare.
22 MR. CLARE: I will address two points. Dr.
i 23 Mark's question tended to be in the first area which I
( 24 will address only briefly, and that is what features we ,
25 have in the plant to prevent a station blackout.
O ALDERSON REPORTING COMPANY, INC.
400 VIRGINIA AVE., S.W WASHINGTON. D.C. 20024 (202) 554 2345
448
() 1 The second area I believe was the area of most 2 significant interest here which is how we would 3 accommodate a station blackout were it to occur.
4 We do have several redundant and diverse power 5 sources to prevent a station blackout. First, we do 6 have an off-site power supply which is a rather 7 extensive set of equipment. We even.have a redundant 8 viewgraph here.
9 (Laughter.)
to NR. CLARE: Our off-site power supply includes 11 four separate connections to the TVA grid. These are 12 four separate sets of transmission lines, 161 kw lines, 13 tha t come f rom four dif ferent substeations varying O 14 distances from the plant.
15 These four connections are connected through 16 three transformers, and I am referring to this which is 17 the main transformer for the plant, and these two 18 reserve transformers.
19 This is in our main generating switch yard 20 where we export power to the TVA grid. These is what we 21 call our raserve switch yard which is separate from and 22 located in distance away f rom this switch yard. That 23 gives us protection against, oh, na tural phenomena or 24 perhaps fires that might affect one and not the other.
25 And we do have as flexible set of buses by O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGIN!A AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
449
() I which we can connect the off-site power to all three of 2 the load groups which are of interest to the safter of 3 the plant. I have tried to illustrate very simply what
(])
4 our switching network is.
5 Basically you will find that Load Group 1 car 6 be connected to this transformer in this manner, or to 7 this transformer. Load Group 2 can be connected to any 8 of the transformers here, and Load Group 3 can be 9 connected to this transformer or through this 10 transformer back here.
11 Not to dwell on this in any great detail, I 12 would merely like to make the point that we do have what 13 we feel is a highly reliable connection to the TVA 14 grid. The experience with the TVA grid has been that 15 it, itself, is a highly reliable grid. There have been 16 relatively few occurrences of any loss of power to any 17 one of the substations that we are conrected to, and 18 even the ones that have occurred have been relatively 19 short in duration - generally measured in seconds, a few 20 have measured on the order of one hour.
21 MB. KERRs What do you estimate to be the 22 probability of loss of all off-site power?
23 MR. CLARE: I don't have that number., The O 24 project has made estistes of that. I don't have the 25 number with me today. Paul?
O ALD'IRSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
450 0 1 MR. DICKSONa Our design basis is 15 events in 2 the life of the plant. It is the general feeling that
() 3 that is excessive.
4 MR. CARBON: I could not hear you. Was that 5 15 loss of all off-site power events?
6 MR. DICKSON: Fifteen events, yes, sir.
7 MR. CLARE: That is the duty cycle for the 8 structural design of our e'quipment as a conservative 9 bound.
10 MR. DICKSON: Yes.
11 MR. CLARE: Now we also, of course, have 12 emergency on-site power supplies. The items of interest 13 are three seismic Category 1 diesel generators. Of 14 those threa, two of them are large diesel generators, I 15 believe the size is about 7,000 kw each; and the~ third 16 one of a different size, somewhat smaller - I do not 17 know the exact size - and also of a different 18 manufacturar which gives us some protection against 19 potential common-cause failures of the diesel generators.
20 Now, each of these diesel generators is 21 connected through an appropriate switching network to a 22 separate safety-related load group, and each of those 23 load groups is adequate to achieve and maintain a safe O 24 shttdown of the reactor. And of course what we are 25 principally concerned with here is shut-down heat
( )
i ALDERSON REPORTING COMPANY. INC.
i 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
451 1 removal.
2 Given that array of protection against the
() 3 loss of all AC power to one set of the safety-related 4 load groups, we of course feel that the likelihood of 5 this event is sufficiently low that it need not be in 6 the formally-established design basis of the plant.
7 dB. EBERSOLE: These diesel generators have 8 certain auxiliaries like water for cooling?
9 MR. CLARE4 Yes, sir.
10 MR. EBERSOLE: And other things, air systems, 11 perhaps, fire protection, etc. To what extent do ther 12 have communality in the auxiliary sense to themselves?
13 MR. CLARE The cooling water systems, the air
()
14 starting systems, all of the auxiliaries for the diesels , ,
t 1 15 are separated just as I have indicated here, all of the 16 critical ones. I
~
17 MR. EBERSOLE: They even have their own water i'
18 pumps?
- 19 MR. CLARE
- For the cooling water, yes.
20 NR. EBERSOLEs Do they go back to exchanges L, 1 J i \
' 21 that have any communality? ~',,
22 MR. CLARE: No, they do not. There are two j ff 23 emergency cooling towers, redundant emergency cool'ing. , ,,
I
\
24 towers, ona each for these diesel generators. There Y ; \
25 will be another heat exchanger for this diesel
j' s.
, /*
l i
i ALDERSON REPORTING COMPANY,INC, ,
/
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2245
- 452 i /
() 1 generator. That system is under development right now, 2 I do not know the exa'ct form.of that.
3 MR. EBERSOLE[ gAre the cooling towers for 4 those tornado-competenh?
5 HR. CLARE: Yes, and seismically qualified.
6 MR. EBERSOLE: What about the field system 7 supply, is that a common system?
8 '
MR. CLAREs The diesel fuel supply system?
9 MR. EBERSOLE: Yes.
10 MR. CLARE: There are separate day tanks for 11 each individual diesel. They are fed by what is 12 eventually a common fuel supply. But for some extended 13 period they do have separate seismically qualified f uel 14 supply.
t i .
15 MR. EBERSOLE Eventually, though, they are 1,6 looking to a common fuel supply.
.J r HR. CL AR E: I believe that is correct.
}17 10 MR. EBERSOLEs Thank you.
- i 19 HR. MARK
- Supposing you lose off-site power, 20, which I believe ve just heard might happen 15 times in 21 the course of --
i \. ,:
22 MR. DICKSON: Let me correct that., We do not
'A' 23 expect'it to happen at all, in fact, from the experience
)
O- ,
24 of TVA. Wa have designed f or it happening 15 times in
, 25 the life of the plant from the standpoint of structural
.;O 1-
)
j
, ,8 5 ALDERSON REPORTING COMPANY,INC, i r *
,-j' e '> 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 0 t ,5
453 )
() 1 evaluation. We do not expect it to happen.
2 MR. MARKS Very good. Suppose it happens, i
3 what are the circumstances under which you need the i
}
4 diesels to start suddenly, or how long can you coast 5 along perfectly happily without them starting, an hour, 6 or two hours, or two' seconds, or what?
7 MR. CLARE: The loads to which our diesel 8 generators are connected vary from the kinds of water 9 pumps that are used for the equipment cooling systems to to the pony actors for our primary and intermediate sodium 11 circuits; air-cooled condenser fans which I will mention 12 in a little while.
13 In the sense of our design basis we ask our O 14 diesel generators to be on line in ten seconds.
15 MR. MARKS That is fine. Is that necessary?
16 HR. CLARE: The presentation which I will be 17 making next --
18 MR. MARK If you are going to cover it, that 19 is all right.
20 - MR. CLARE: -- suggests that if we did not 21 have them at all we would still be able to achieve a 22 safe shutdown. So, the answer to your question is, in 23 one sense I never need the diesels.
24 HR. MARKS I would like to hear you sort of 25 explain to us, as I just heard from Professor Kerr, the ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
I 454 1 reason you need them immediately in an LWR in a large 2 LOCA. You do not have large LOCAs.
3 MR. CLARE: We certainly consider pipe leaks, 4 but even in the even t of a pipe leak we do not have any 5 active equipment that would come on line, either 6 immediately or ever, to be able to accomodate that. So, 7 that immediate need is not present, that is correct.
8 We do have a capability to accomodate a 9 station blackout in spite of the f act that we do not to consider that a design basis event for the plant. I 11 have tried to put on this viewgraph a chronological 12 valk-through of the kinds of things that one would find 13 occurring in the event of a station blackout.
14 First, the reactor shutdown system would send 15 you thfough any number of sensors the loss of the 16 off-site power. Two evident ways that that would be 17 sensed would be that we would lose power to the mala 18 motors for the primary coolant pumps. The reactor 19 shut-down system would sense that, both by the loss of 20 electric power and also by the loss of flow.
21 It would also be able to sense the loss of 22 main feedwater to the steam generators. And there 23 probably would be other signals as well to the reactor 24 shutdown system. That system will automatically shut 25 down the reactor. That system relies in nc way on there ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
455 l
() 1 being AC power from off-site or from the diesel 2 generators.
3 I mentioned that we would lose power to the 4 coolant pumps in our primary system. That would also be j
5 true about the transport system. Those loops, those 6 pumps would close down and the flow would close down.
7 But the circulation in the primary heat transport 8 system, the intermediate heat transport system, the two 9 sodium systems, and in the water system, the steam 10 generator system, would naturally circulate.
11 We have specifically designed the plant so 12 that that natural circulation is assured. This is kind l
13 of a cartoon of one ci our three loops that shows, that 14 demonstrates that we have established the elevations of 15 the individual components in such a way that natural 16 circulations occur.
17 The elevation difference between the core in 18 flame and the IHX is about 15 feet from there to the i 19 thermal center of the evaporators, which is the 20 important point in looking at natural circulation in the 21 intermediate heat transport system. It is about 30 feet 22 from the evaporator to the steam drum. It is about 13 23 feet - and I will stop there - I will talk about the
) 24 air-cooled condenser in a minute. It is one of the heat 25 sinks that we eventually consider.
O l
ALDERSoN REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
456
() 1 HR. EBERSOLE In the 0786 report there was 2 some doubt experessed as to whether you in fact did have 3 natural p ro te c tio n . And then there was a disturbing 4 implica tion , implica ting statement, that if you did not 5 have it the staff would consider pony motors as an 6 adequate replacement for it.
7 What is the explanation of that?
8 MR. CLAREs The subject of natural circulation 9 is a subject which is under review with the staff now, 10 and we have been providing some significant support to 11 their review. The project has recognized that this was 12 a concern of the NRC staff back in the earlier days of 13 revie , back in '75 and '76.
O 14 There have been a number of tests and analyses 15 performed. Perhaps the most significant of these is the 16 test of natural circulation that was performed at FFTF 17 some number of months ago, in which ther operated their 18 plant at full povar and then turned off all the electric 19 power and scrammed the plant. So that there was no i 20 forced circulation there.
21 That plant behaved wonderfully from the 22 standpoint of natural circulation. We had pre-predicted 23 the results of that test using our analytical methods 24 tha t we will be using on this plant. We found that, as l
25 we expected, our results were very conservative in terms j
(
ALDERSON REPOPTit COMPANY, INC, 400 VIRGINtA AVE., S.W., WASHINuTON, D.C. 20024 (202) 554-2345
457
() 1 of the data that was actually taken from the FFTF test.
2 We believe that that serves to valida te out 3 methods.
4 HR. KERRs I am sorry, does that mean that you 5 were not able to predict what happened very well?
I 6 HR. CLARE: No.
7 MR. KERR What does "conserva tive" mean then?
8 HR. CLAREa We performed three different sets 9 of calcultions. We performed an expected set of 10 calcula tions ; a slightly conservative set of 11 calculations, and wehat we consider to be our design 12 calculations for the plant. As expected, our nominal 13 calculations were not far off from what actually 14 happened. Our design basis calculations were.
15 Now, let me also note that our pre-test 16 predictions did not use the proper assumptions in terms 17 of such things as the pre-test history of the plant and 18 what not. So, we hve had to go back and adjust those 19 test results for the initial conditions prior to the 20 test, which one normally does. However, the methods all 21 stayed the same, the correlations, etc.
22 MR. ZUDANS On this picture, this air-cooled 23 condenser requires power to be driven.
() 24 HR. CLARE: I will discuss that in some detail.
25 MR. ZUDANS: You know my question, I have O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
458
() 1 asked that before.
2 MR. EBERSOLEa May I continue this just a 3 little bit. This statement which I found so
]
4 extraordinary is in this 0786 on page II-13. It says:
5 "Furthermore, in the event that natural circulation i
6 cannot be relied upon to provide a diverse means of 7 assuring flow in the normal heat rejection paths, the 8 applicants will be reqwuired to commit to providinga i
9 (1) motive and control power to assure that for onsite 10 eletric power system operation (assuming of fsite power 11 is not available) forced convection flow is maintained i
12 throughout the entire decay heat removal train (PHTS,
! 13 IHTS, SGS, SGAHRS) assuming a single failures" et O 14 cetera, et cetera.
15 I read this as saying that staff will accept 16 the absence of natural convection which I find 17 extraordinary. Is that a correct assumption on my part?.
18 HR. KERRs I don't think we ought to tell 19 k'estingh ou se that because they might then think they 20 have to design for natural circulation.
21 MR. STARK This is Richard Stark from the 22 staff again.
23 The staff is pursuing natural circulation and 24 would like very much to see natural circulation be 25 defensible. In that respect, the Applicant has
- O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
459
, /"N I
(,) 1 presented FFTF information. We received their report in 2 January which we are reviewing, which looks at the first 3 500 to 600 seconds.
4 That looks fairly good to us so far. But we 5 are looking at natural circulation for an extended 6 period of time and have requested more information.
7 Historically, and the site suitability report 8 includes what the staff did four or five years ago, 9 there was not the FFTF experience that confirmed natural 10 circulation, and the staff did in the past, as it does 11 right now, is pursuing options or flexibility that might 12 excist for a general reactor of this size and type.
13 That is what that report is looking at.
O 14 What if natural circulation does not work, i
15 what would a reactor of this size and type have to do to 16 satisfy the staff on station blackout or emergency heat 17 removal? What are the heat removal paths and what are 18 the options, and what must you have.
19 I think that particular report looks at 20 options and~flexibililty, and it is a true statement 21 that to date natural circulation is not accepted or 22 endorsed by the staff, but is under review. As I have 23 indiated, we have first, I believe it is 450, or 500, or 24 600 seconds. That report came in. That looked pretty 25 favorable. It needs a lot more time. So, we have O
ALDERSON REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
i l
460 f I
i
() 1 requested more information from the Applicant and got a 2 report in last week, on which I do not have any input.
3 It is a subject that we are going to look I 4 into, and we are looking into, of course. We are 5 looking into it with Brookhaven and the results will 6 fall wherever they fall, I am not going to prejudge i
i 7 them. I think the Applicant is very confident and the 8 FFTF experience increases their confidence.
9 MR. MARK Are you telling us that there are 10 circumstances under which you might be willing to agree 11 that natural circulation may be relied on?
12 MR. STARKa If we feel that the FFTF 13 informatin confirms the Applicant's analysis and their O 14 studies on natural circulation, then we feel it could be 15 relied upon, yes.
16 HR. MARK Supposing natural circulation could 17 be relied on for 6000 seconds, would you then be willing
- 18 to drop the inquiry as to what you do if it does not 19 work?
20 HR. STARK: That is correct.
21 MR. CARBON: Jesse, you appear to still have a l
22 q ue s tion . Let me see if I can help something. I think l
23 the staff encourages natural circulation design and so
( 24 on, but they are,not prepared to say that we are sure 25 tha t it will take place properly in CR.
'O l
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE S.W., WASHINGTON D.C. 20024 (202) 554 2345
461 1
() 1 MR. EBERSOLE: I guess I was looking for a 2 mandate that if we don 't get it, that is the end of the 3 road.
4 MB. ZUDANSa That is right, I would agree.
5 MR. EBERSOLE: That was the death knell of the 6 high pressure reactor, it was the absence of a 7 conviction that you could always get force convection.
8 Anyway, I will buy a fly-wheel but I wouldn't buy motive 9 power, from motors.
10 ER. CLARE: Well, we certainly believe that 11 reliance on natural circulation is appropriate.
12 HR. EBERSOLE: Would you even enter into this 13 project if you felt you were not going to get it?
O 14 MR. CLARE We obviously believe it is the 15 right way to oc, with natural circulation.
16 MR. EBERSOLE: You are proceeding on the 17 thesis that it is going to work; right?
18 MR. CLARE: We are proceeding on the thesis 19 that we vill make it work, yes.
20 MR. KASTENBERGa Can I just raise a question?
21 Wha t is your criteria for success, it is a clad 22 temperature?
23 MR. CLARE: Yes, in this particular type of 24 transient the important thing really is to assure that 25 you don't boil sodium in the core. Since you assume ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
462 1 your coolant temperature as your clad interface, you 2 could say that it is a clad temperature, yes.
3 HR. KASTENBERGs Can you have natural 4 circulation with some low level of power as well, have 5 you ascertained what your outside limit might be?
6 MR. CLARE: On neutron power?
7 HR. KASTENBERGs Yes. In other words, suppose 8 you ended up at some low power rather than decay heat.
9 HR. CLARE: Certainly, natural circulation is 10 the sort of thing that is inherent, the more you ~
11 equalize your core the more you increase your thermal 12 driving heads, and the more cooling you would get.
13 MB. KERBS Since the immediate decay heat on 14 shutdown is 7 percen t of full power, one could almost 15 certainly remove the heat at that low power.
16 HR. CLARE: Yes.
17 MR. KASTENBERGs But you do not have an 18 outside limit.
19 MR. ZUDANS: The natural circulation will not 20 start immaliately.
21 MR. CLARE: Theoretically one can design a 22 reactor with its own na tural circula tion ,
- even for power 23 operations.
24 MR. CARBONS You make the statement here that 25 no operator action is required in the second bullet.
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S W WASHINGTON, D.C. 20024 (202) 554-2345
463
() 1 How many check valves and things like that have to work, 2 and what sorts of studies have you done on the rs 3 propability tha t they will or will not work when called
(_) l 4 upon?
5 HR. CLARE: There are no check valves that 6 have to work , no isolation valves, no throttle valves 7 that have to work. There are certain valves in the 8 system.
9 And let me point out as best I can on this 10 cartoon where those would be. There is a check valve 11 which is normally open and would remain open in the cold 12 leg of each of the three primary loops.
13 ER. EBERSOLE: What are the forces that hold O 14 that open?
15 ER. CLARE: The forces that hold that'open are 16 the slope forces. However, the normal position of the 17 valve is such that some significant flow can get through 18 i t. There is some significant opening even with no 19 flow. Furthermore, there have been water tests with the i
20 valve to assure that the pressure drop is well' 21 understood and in fact those pressure drops have been 22 incorporated into our calculations.
- 23 HR. EBERSOLE Are not the potential forces
( 24 that would inhibit operation of that valve so low that i
25 you would be concerned about natural convection heads O
l
{ ALDERSON REPORTING COMPANY,INC, i
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
464
- 1 br-ing unable to accomplish any movement on that?
2 MR. CLAREs Again, it is normally wide open,
(]) 3 and it would begin to close under natural circulation.
4 HR. EBERSOLE: By gravity?
5 MR. CLARE: By gravity, yes.
6 HR. EBERSOLE: Why don't you keep your gravity 7 the other way? I mean, is not the importance of it to 8 be kept open? Which is the most important end of this 9 operation, closed or open?
10 HR. CLARE: In this particular event it is 11 open.
12 HR. EBERSOLE: Well, I mean in th e general 13 context.
14 MR. CLARE: In the general context we would 15 not have it there. If there was no an event we would 16 not want it.
17 MR. EBERSOLE: Yes, but like everything else, 18 you may have put it in at an undue price.
19 MR. CLARE: We do not believe so. Again, we 20 have completed a thorough set of water tests on check 21 valves of this type.
22 MR. EBERSOLEs I know, but the journals have 23 agreed to all those things, and now it is going to be O 24 old and sticky.
25 HR. CLARE: Well, if it is sticky it will O
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
1 1
465
(} 1 stick in the way-open position, which is very beneficial 2 from a natural circulation point of view.
3 MR. ZUDANS: What is it needed for in a closed
({}
4 position?
5 HR. CLARE: What we provide that check valve 6 for is in the event we should lose one loop, the pumping 7 power in one loop, and we would then continue pumping 8 power in the other loop.
9 Now, we have done safety analyses to show that 10 in that event we do not have to have that check valve in 11 order to maintain the integrity of the system. However, 12 it would cause, could potentially cause some structural 13 damage because of the thermal shock of pushing cold 14 sodium back around this loop backwards, that we would 15 prefer to avoid.
16 MR. EBERSOLEa That is not an assisted valve.
17 HR. CLARE: I am sorry.
18 HR. EBERSOLE: That is not an assisted valve 19 by an, external force.
20 HR. CLARE: No, that is correct.
21 ER. ZUDRNSs But you would be pumping the 22 sodium at 700 degrees that comes from the other two 23 loops that are intact, and this is your closed loop.
24 The worst that it can do is wash the IHX with the 700 25 degrees.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
466 1 HR. CLABE Well, that is absolutely correct.
2 It is normally seeing roughly a thousand degrees at that 3 top. We would just prefer to avoid that thermal
)
4 transient. Also, the other thing that would happen in 5 this type of an event is by that check va.ive closing we 6 assure a greater degree, a greater flow through the core 7 and provide that much better cooling of the core in the 8 event of a loss of a loop.
9 It is those two functions of that valve that to are beneficial.
11 HR. CARBON: Go ahead with the valves 12 pertaining to natural circulation.
13 MR. CLAREa There are no valves in the line of 14 the transport system from the intermediate heat 15 exchanger through the superheater, which I have not 16 really shown here, through either of the evaporators or 17 back to the IHI, no line valves whatsoever.
18 In the steam generator system there are 19 several check valves and isolation valves, all of which 20 are normally open; none of which will close in the type 21 of scenario we are talking about here. We do have a 22 bi-pass line around our recirculating pump in situations 23 under which we might choose to by-pass the pump or not.
I 24 But in either case natural circulation would be 25 perfectly adequate. We have looked at the pressure drop k
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
l 467 l
l
( 1 under both situations and that would not be significant 2 one way or the other.
3 MR. CARBON: There are no check valves on the
{}
4 core, in the reactor vessel.
5 HR. CLAREs There are no valves at all in 6 terms of the reactor vessel.
7 I wanted to use this vievgraph to try to 8 address some of the questions that Mr. Ebersole had on 9 overall decay heat removal schemes, if I could. This
- 10 sight be the best one I have.
11 What I will do is to try to use this viewgraph 12 very briefly and conceptually go through our decay heat 13 removal scheme, and especially how the diff erent heat O 14 removal heads are separated from one another.
15 What I have represented here is the one 16 reactor vessel, we have only one of those. Beyond that 17 everything appears in at least triplicate. We have 18 three primary heat transport loops; three intermediate 19 heat exchangers; three intermediate heat transport 20 loops; three sets of steam generator modules - there are i 21 actually three modules in each loop; we have three steam 22 drums, and for each steam d rum we have three sets of 23 emergency heat sinks.
24 Now, I am going to go into those in detail in 25 just a minute. Just assume for the moment that that ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
468 i
() 1 protected sir-cooled condenser illustrates that heat 2 sink.
3 Now, each one of these three loops is adequate 4 to remove all of the sensible decay heat after a reactor 5 shutdown.
i 6 NR. EBERSOLE: Now, these are the big loops, 7 the commercial loops?
8 NR. CLARE: Yes.
9 NR. EBERSOLE: These are the ones we saw a 10 while ago.
11 HR. CLARE: That is correct.
12 NR. EBERSOLE: Now, the steam drum is, 13 however, in normal operation commonly manif olded to the
' O 14 turbine generator.
I 15 HR. CLARE: There are three steam drums. The 16 steam to the turbine actually flows out of the 17 superheater, which is one of the steam generator I
18 modules. Each one of those superhater outlet piping l
l 19 legs has in it two isolatin valves which are intended to 20 isolate tnat loop from the other three loops.
21 Now, downstreaa of those valves one comes to a 22 header and tha t header, then, feeds steam to the steam 23 turbine. It is the outlet isola tion valves on the 24 superheater piping that provide the isolation between 25 the loops and the wa ter side, and I will show a figure O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINI A AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
469
() 1 of those in a minute.
2 HR. EBERSOLEa Well, then, if would you have a 3 steam headar failure or some upset in the commercial
("}
4 steam system, you are in one of two logics, having lost 5 one loop, right?
6 MR. CLARE: Well, no. If I have a failure in 7 the steam header, that is not essentially what I would 8 refer to as the non-safety related.
9 MR. EBERSOLEa OK, but in the header ahead of 10 that, that is one of the steam generator headers, one of 11 the input headers.
12 MR. CLAREs If I have a failure anywhere in 13 one of my steam generator systems --
0 14 MR. EBERSOLE In one out of two, now.
15 HR. CLARE: If I disabled that loop, I an in 16 one out of two.
17 HR. EBERSOLEa All right.
18 NR. CLARE: You mentioned earlier, your term 19 was a " dedicated decay heat removal system." Not l
l 20 illustra ted in any way on this viewgraph is another set 21 of equipment we refer to as a direct heat removal system.
22 MR. EBERSOLE: That is what I meant.
23 MR. CLAREa Which is r. set of small sodium 24 typing with a sodium heat exchanger within an air heat 25 exchanger, with an assortment of pumps and such, that O
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
470 l l
() 1 can be used without any dependence on the intermediate 2 system or the steam generator system.
3 MR. EBERSOLEs Is that other system in 4 duplicate or just single track?
5 MR. CLARE: It is actually a duplicate set of 6 equipment. If one were to assume that you relied on 7 only th a t system immediately upon shutdown, one would 8 have to depend on both of those sets of duplicate 9 equipment. At some point later on where one assumes 10 that any hast removal is achieved through this set of
~
11 systems, one would then be able to rely on both of them.
12 HR. EBERSOLE: Then your system, like most 13 PWRs, is virtually totally dependent on the auxiliary O 14 feed pum p system that feeds this drum; right?
15 MR. CLAREs For these three heat sinks on 16 these three loops we do depend on the auxiliary 17 feedvater system.
18 MR. EBERSOLE: And tha t is two motor driven 19 and one turbine driven.
20 MR. CLAREs That is correct.
21 MR. EBERSOLEs The turbine driven is dependent 22 on steam coming out of that drum.
23 MR. CLAREs That is right. We will go over
() 24 tha t in detail.
I 25 MR. EBERSOLE: If you don't have pressure, I O
ALDERSON REPORTING COMPANY,INC, l
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
471 1 guess you don't have it. -
2 MR. CLAREs That is right.
3 MR. EBERSOLE: Thank you.
4 MR. CLARE. Now, chronologically going then, 5 once I have continued the circulation in the primary, 6 intermediate and steam generator systems and removing g 7 the heat from the reactor, what I now need is a heat 8 sink at the end of that loop. _
9 We have such a heat sink and we refer to our 10 set of heat sinks as the steam generator auxiliary heat 11 removal system. It has three sub systems that I have ._
12 iden tified separs tely here.
13 The first is a set of power vent valves that 14 will provide a short-term heat sink. Now, let me use 15 this drawing to try to illustrate how those work.
16 Now, this figure is representative of one of 17 o ur three heat transport loops, and this is the 18 short-term heat sink on the end. Let me mention first 19 some of the isolation features we were just talking 20 about.
21 In the event of a loss of off-site power the 22 first thing that would happen would be, each of these l ,
23 three steam drums and its associated steam water system 24 would be isolated from the non-safety rela ted steam 25 turbine system, commonly named feedvater system. That ALDERSON REPORTING COMPANY. INC, 400 VIRGINTA AVE, S.W., WASHINGTON. D.C. 20024 (202) 554 2345
472
) 1 is comprised of primarily a superheater outlet isolation 2 valve - as I mentioned, there are actually two of those
{} 3 on each loop. Those are failed closed valves, when they 4 lose power they will fail closed.
5 We have a steam drua, a drain which is part of 6 our water clea n-up system, purification system. It will 7 fall closed on loss of power. And in response to one of 8 the questions that was raised earlier, I added on here 9 that we also have a fail closed set of valves, two on to each loop for the main feedwater inlets to the steam 11 drum.
12 Now, you isolated from your normal heat sink f,s 13 the steam turbine and what happens is, the steam
~-) In addition to that, as 14 pressure here begins to rise.
15 soon as one gets closure of these valves or failure of 16 the feedvater system, there would be .a steam to feed 17 flow mismatch.
18 In that event, these vents, vent valves, one 19 at each superheater outlet and one from each steam drum, 20 will be armed and open. The control for these valves is 21 driven by our instrumentation and control batteries 22 which are seismically qualified batteries for 23 specifically this purpose.
O p~/ 24 Either one of these two vent valves on any of 25 our three loops will be adequate to remove sufficient O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINtA AVE, S.W., WASHINGTON. D.C. 20024 (202) 554 2345
473 1 decay heat from the plant. These valves are controlled 2 to bring the steam pressure in the steam drum down to a 3
gg temperature of approximately 1,450 psi. This is the 4 same as saying that we will drive this water heat sink 5 to roughly 600 degrees, and that then drives the cold 6 leg, and eventually the plant would drive towards an 7 isothermal condition at 600 degrees.
8 Now, here what I have done is, I have started 9 dumping water in the atmosphere th ro ugh these vent 10 valves and unless I do something I run out of water in 11 the steam drum. So, what I do is to provide feedwater 12 through the operation of a turbine-d riven a uxiliary
,, 13 feedwater pump. That is illustrated on the next
/
14 vie vg ra ph .
15 - As I mentioned, the steam to drive the 16 turbine-driven auxiliary feedwater pump flows from the 17 thre steam drums - and I have noted here that we can get 18 steam from either the loop that I have illustrated here 19 with this steam drum or either of the other two loops, 20 through an isolation valve which fails open on loss of I 21 power, through a pressure control valve which is battery j 22 controlled to the drive for the feedwater pump which has 23 a battery-controlled government.
N 24 Now, there is one turbine and one pump. The 25 pump takes suction on a seismic Category 1 protected ALDERSoN REPORTING COMPANY,INC,
! 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
l l
474
() 1 vater storage tank. Between the storage tank and the 2 pump is a pump inlet isolation valve which is there 4 3 strictly for maintenance purposes, it is a normally open 4 valve. In adition to that, it is a fail open valve on ,
l 4
5 loss of power.
6 The feed pump has a header on the outlet which 7 allows it to distribute water to any one of three steam j 8 drums, or all three of the steam drums. The pump is 9 sized to provide sufficient flow for all these steam 10 drums. The flow to the steam drum is controlled, uslng 11 a flow-control valve - again using the battery - and the 12 control is based on what we call a low-low level set of 13 instrumentation on each steam drum which is also O 1<4 supported by the battery.
15 So, there is a control room between the steam
- l 16 drum and the flow valve, all supported by our seismic 17 Category 1 batteries. There is an additional isolation
- 18 valve on each steam drum which fails open.
19 That, then, provides make-up water to the 20 steam drum adequate to support the venting from the 21 superheater and steam drum and connect valves.
22 NR. CARBON: Wha t does "EH" stand for?
23 NR. CLAREa Electro hydraulic, that is the
, 24 type of operator.
25 NR. ZUDANSa How many of these steam vent
(
l j s ALDERSON REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
475 9
s_/ 1 feedwater turbines do you have?
2 MR. CLAREs One turbine and one pump.
3 MR. ZUDANSs That means at that the onset of
- '])
4 your natural circulation then a continuation of it, 5 totally depend on a single component.
6 MR. CLARE: Again, the scenario we are talking 7 about is failure of four connections to the grid, th ree 8 diesel generators, and now you say, well, it has taken 9 eight pieces of equipment.
10 MR. ZUDANS: That is right. Now you are left 11 to the mercy of a single piece of equipient.
12 3R. CLAREs That is correct.
13 MR. EBERSOLE: Assume because of the small l
14 plant that you have here that when you trip this turbine 15 off, the rest of the TVA syst'em won't even notice and l
l 16 you don't get any power failures at that time.
l 17 You know, the standard plan with a commercial 18 reactor is that a shock wave does travel through the 19 system and you get a coincident or near coincident 20 on-site power failure.
21 MR. CLARE: That is right. That fact is, as I 22 best understand it, part of the reasoning behind why one 23 essentially postulates loss of off-site power with f]'
24 reactor scram.
25 MR. EBERSOLEs Do you do it here?
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
l l
476
) 1 MR. CLARE: I am sorry?
2 MR. EBERSOLE: Do you do it here?
3 HR. CLARE: Certainly.
(]) This whole scenario 4 assumes that I lose all four of my grid connections and 5 all three of my on-site --
6 MR. EBERSOLEa Well, if I start this scenario 7 by bursting the steam line here on one of these two 8 loops, you say the isolation valves fail open.
9 MR. CLARE: This is a different isolation 10 valve now.
11 MR. EBERSOLE: Right. But I have broken it 12 within that network that you .see, where you have loops 2 13 and 3 coming in there.
14 MR. CLARE: Yes. If the steam line should 15 fail between this isolation valve, downstream of this 16 isolation valve --
17 HR. EBERSOLE: Well, it would not matter 18 because you said it stays open anyway.
19 MR. CLARE: That's correct. At some point the i 20 losses might be such that you would lose sufficient 21 steam that you might lose your motor force for your 22 turbine. You led me to one point I wanted to make.
l 23 Each of these items that we are talking about 24 here does have a manual back-up and the operator would 25 be able to manually close this valve. The water l
l ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
uT7
() 1 inventory in the steam drums is good for something on 2 the order of ten minutes before the steam drum would go
{} 3 dry and if you postulated some sort of severe rupture, 4 the operator would have ten minutes to find and correct 5 that.
8 MR. ZUDANS: Wat size is.the protected water 7 storage tsnk in a common sense, if you had to continue 8 with this mode of cooling?
9 NR. CLAREa The protecte? "ater storage tank ,
to the primary tank that we would depend on in this type of 11 scenario, that tank is sized for a 30-day cool down of 12 the plant.
13 Now, a normal cool-down of the plant would 14 include closure of the vent valves at some time earlier 15 than the kind of scenario that we hre talking abouts and 18 I have not walked through it all. But we think that 17 based on our calculations this inventory would be good 18 f or at least two hours.
19 NR. ZUDANS: At which time you expect to 20 recover your offsite power.
21 ER. CLARE: If one looks at the generic safety 22 issue treatment, yes, at which time you expect to 23 recover on-site power. Also for this plant, well before 24 that one would expect to be able to transfer the entire 25 heat load to the air-cooled condensers which then depend O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20C24 (202) 554-2345
478 O 1 on very little make-up water, which would then not be a l
2 problem.
(]) 3 MR. ZUDANS: But do not those air-cooled 4 condensers require some power, as I remember? I guess 5 you will come to that.
6 MR. CLARE: I will come to that, right.
7 That in fact is where we are now coming down 8 our viewgraph here. The protected air-cooled condensers 9 will automatically open to provide a long-term heat sink 10 using natural circulation on both the water and air 11 sides.
12 Again what I have here is a very simplified 13 drawing of one of the particular air-cooled condensers 14 that represents one of our three loops, one of our three 15 steam drums. Now, there is a seismic Category 1, Class 16 1-E pair of fans on each of our protected air-cooled 17 condensers. Under any of our our design-basis scenarios 18 the diesel generator would start up and the diesel would 19 provide power to the fan. The fan would operate, and in 20 that mode a single pack, one protected air-cooled 21 condenser, would be adequate to remove all of the decay
- 22 hea t f rom the plant, beginning one hour after shutdown.
l l 23 MR. ZUDANSs So, in the first hour you would
() 24 have to have that other steam operational. If you did 25 not have tha t, this would allow the temperature to rise O
ALDERSON REPORTING COMPANY,INC, j
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
479 1 anct the trouble would be uncontrolled. { .
2 MR. CLARE: The venting heat sink is required i i
] 3 for the initial period. Now, what I have just said, ,
, N 4 there is a one-hour period that is based on the j )i ', e .;
5 assumption that we have one protected air-cooled
/ -
, nr s ,
6 condenser operating under the forced circulation-air l 7 side mode.
8 Now, what we are talking about here is a 9 station blackout where it has been suggested we would to lose all three diesel. generators. In that case we do.
11 not have a fan here. In that case we would have to rely 12 on air-site circulation on the air-flow condenser.
13 Our understanding of the behavior of these 14 units is that we would remove approximately one-third of 15 the nominal rating of the unit under that kind of 16 operation. H:g 17 Therefore, what we are suggesting is that 18 something on the order of one hour - if all three of our 19 protected air-cooled cor.densers were available - then i
20 one would be able to remove the entire heat load at tha t 21 point and there would be no need for further venting for 22 a heat sink.
23 Now, let me note that that then allows us to
]' 24 terminate venting within the two hours for which this 25 protected water storage tank is available. We have not ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 l
If l p
480 1
1 yet begun to tap the condensate storage tank. That is a
[()
2 very large tank. I do not have an estimate of the
),
$ 3 amount of time that would be good for in the event we
\1 '
' '/ ' i' , 4 had to' continue venting.
,l
' ' , 5 I do not wan t to mislead you, tha t is not a
- 'i -
\ 6 seismic Category 1 storage tank but it would be s 7 available for use if it had maintained its integrity in 8 this kind of a scenario.
9 Now, that then completes the whole story. We 10 have' reached the point where we can maintain the cooling 11 in the shut-down of the plant. So, in summary, we do 12 have off-site and on-site power supplies that assure 13 that a station blackout is a very lov likelihood event, 4
i O 14 so unlikely that it is not a design basis.
15 However, we have looked at a station blackout 16 event and the combintion of the natural circulation in 17 the heat transfer systems which can remove the heat from 18 the reactor to our heat sink and the heat sink which has 19 multiple short and long-term heat sinks, using power 20 from the batteries for the necessary control, will 21 accomodate the event.
2, 22 MR. EBERSOLE May I ask a question, just a 23 standard question?
() 24 You have looked at the primary problem. Are 25 there any subtle aspects of the loss of AC power in viev iO ALDERSON REPORTING COMPANY, INC,
)
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
481
() 1 of the fact you had a very high tem pera tt;re system here 2 to deal with, and you are continuously de pendent on
() 3 moving air systems or water systems for nabient control, 4 which get out of hand very shortly when ;ou lose AC 5 power?
6 gR. CLARE: Well, we certainly are dependent 7 on power for our air handling systems, and we do try to 8 maintain our plant, areas, buildings, at fairly low 9 temperaturas. That we do, for example, in order to to protect our accurate equipment. Our por.y motors, for 11 example, in our steam generator building, depend on just 12 outside air being brought in to cool that building and 13 the instrumentation that is associated with that.
O 14 On the other hand, we performed calculations 15 for this kind of an event where we in essence lose all 16 of our air-handling systems. We are not depending on 17 any active equipment other than the turbine-driven 18 auxiliary feedwater pump and the associated control.
19 And our understanding is that the heat load will be low 20 enough that even the long-ters temperatures achieved 21 will be within the capability of those systems.
22 We specifically are qualifying much of our 23 equipment in our environmental qualification program to t 24 be able to withstand a sustained loss of cooling type of' 25 event.
(
l' l
ALDERSoN REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W WASHINGTON. D.C. 20024 (202) 554-2345
482
() 1 I will noto that even though our normal 2 operating temperstures are up around a thousand degrees, 3 under this kind of shut-down condition one would be 4 somewhere between 400 and 600 degrees, and that brings 5 the kind of consideration you are talking about being a 6 auch lesser kind of problem.
7 NR. DICKSONs I think your answer with regard 8 to the safety-related equipment may have been just a 9 little bit misleading. In the event of a loss of to off-site power for longer than two hours, there are a 11 number of equipments that would require repair or 12 replacement because of the over-temperature.
13 MR. CLARE: Certainly.
O 14 MR. DICKSON: None that involved any safety 15 aspects.
16 MR. CLARE: I was speaking strictly from the 17 saf e shutdown.
18 NR. EBERSOLE: I as thinking about not just 19 loss of of f-site power but loss of all power.
20 ER. DICKSON: That was wha t I was referring 21 to, loss of all power with subsequent heating of the 22 cells, some complements of some cells that would require 23 replacemewnt or repair.
() 24 MR. KERRs Would this be like pump seals?
25 NR. DICKSON Yes.
O ALDERSON REPORTING COMPANY. INC.
400 VIRGHIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
483
- O 1
sR. E8ERSotE. And the feedvater ouses ver se, 2 that is the steam-driven pump and its auxiliaries, can
! 3 withstand the rising ambients that you would associate 4 with this?
5 MR. CLARE: Yes. We have looked st that and 6 are confident that that turbine and its controls would 7 survive tha kind of event we are talking about.
8 HR. EBERSOLEs Frequently one has to put in, I 9 think alvars, the detection systems that detect the 10 failure of the steam supply systems of feedwater 11 turbines or main feedwater turbines, or for that matter 12 the big tu r bin e.
, 13 This is to close down the discharge of steam O 14 if you have a major stenaline failure. Sometimes it has i 15 been found that these detection systems are based on 16 thermal temperature rise and therefore they erroneously 17 respond to a rise in ambients when you lose the fans.
18 They thus cut off the steam supply to a needed steam 19 system.
20 Are your systems independent of that sort of j 21 thing?
22 MR. CLARE: Help me understand the questions.
23 NR. EBERSOLE I as saying you have isolation 24 devices to prevent continued steam flow.
25 MR. CLARE That valve, for example.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
484 ;
\
() 1 MR. EBERSOLE: That is right. It is driven by 2 tempera tures in the system sometimes which erroneously 3 respond to a loss of force convection.
4 MR. CLARE: I don 't know today what control 5 system might be associated with this particular valve 6 with respect to temperature conditions in the cell.
i 7 I am not aware that there is any system that 8 is intended to isolate that valve on sensing high 9 tem pe ra tura s . But I cannot speak with a hundred percent to assurance on that.
11 HR. EBERSOLE Well, since that valve there is 12 set to fail open, it would appear that you would have to 13 argue you are prepared for sustained steam flows out of 14 that broken pipe without any particular problem. Is 15 tha t correct?
16 NR. CLAREa Sustained for some period of time.
17 HR. EBERSOLE: Into the building?
18 MR. CLARE: Yes, for some period of time.
19 HR. EBERSOLE: What happens to condensation 20 and temperature rise on your auxiliaries out of the 21 building?
22 MR. CLARE Again, I cannot address that in 23 any quantitative way today. The operator is able to
() 24 isolate this valve from the control room, if nothing 25 else, to terminate the operation of this auxiliary O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINTA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
485
() 1 feedwater pump in the event it is turned on and 2 eventually is found to be not needed.
s 3 MR. EBERSOLEs Yes, but if you blow steam out l
4 of that to these rooms - and they may be interconnected 5 - a n ea rly event is condensation --
6 MR. CLARE: I understand quite well your 7 concern, I am just not prepared with the details here to I
8 be able to discuss it.
9 MR. BENDERS I have a question in a comparable 1
10 area. Sodium has a rela tively high f reezing point. To 11 what extent are we dependent upon power sources to make 12 sure that we do not get freezing of sodium lines where 13 we need to maintain continuity of flow?
14 MR. CLARE: We, of course, have a number of 15 auxiliary systems, filling systems, draining systems, 16 purification systems and those kinds of things. These 17 are all trace heated and many of them would be trace 18 hea ted. The heaters would be operational during the 19 plant operation in order to assure that whatever liquid 20 metal was there would not freeze.
21 For the main systems that we are talking 22 about, the primary and intermediate systems that we use 23 in his kind of a situation, they of course have trace
() 24 h ea ting which is there for initial fill considerations.
25 It would be a very extended time period, anny, many O
ALDERSON REPORTING COMPANY,INC.
40 VIRGINIA AVE, S.W, W ASHINGToN. D.C. 20024 (202) 554-2345
486 lll 1 months, before the decay heat level were to reach a 2 point whera any considera tion of heating would be 3 required at all.
4 I cannot give you a number. I know that there 5 were specific es1:ustions done for FFTF on that 6 question. On the order of a year or more would be the 7 time to raise that concern.
8 MR. BENDER I have in mind not so much the 9 primary circuitry but the secondary circuitry and relief 10 lines, and things of that sort, where you have small 11 quantities of stagnant sodium sitting because they are 12 out of the flow stream.
13 MR. CLARE: I am not sure what relief line you 9 14 may be talking about. For example the rupture discs 15 that Paul has discussed this morning?
16 HR. BENDERa Yes.
17 HR. CLARE: They are located in 18-inch piping 18 just a few feet from the steam generators themselves and 19 the conduction would be more than edequate to keep that 20 sodium liquid.
21 3R. LIPINSKI What kind of time do you have 22 before this air is discharged?
23 MR. CLAREs The batteries are nominally sized h 24 to be able to support their full load for two hours. We 25 would anticipate in the kind of event we are talking O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON. D.C. 20024 (202) 554 2345
487
() 1 about here that they would be good for considerably 2 longer than that. If nothing else, the oprators would 3 probably shed some of the less critical loads from the 4 b a t te rie s.
5 Again, we do have manual backup for all 6 control f unctions I mentioned. So, the operator in that 7 period of time would be able to maintain the system.
8 MR. LIPINSI: What about the instrumentation 9 of plant sta tus?
10 -
MR. CLAREs What instrumentation are you 11 talking about?
12 MR. LIPINSKIs Well, if you are going to run 13 this systen you would like to know something sbout the O 14 temperatures and flows to make sure it is functioning 15 normally.
16 MR. CLARE: Yes. We have a full complement of 17 accident-monitoring instrumentation. We are about to 18 bring our design effort there to some fruition. It 19 includes some considerable number of thermocouples and 20 other temperature and flow instruments throughout the 21 plant.
22 In accordance with the guidance of Regulatory 23 Guide 1.97, portions of that are connected to batteries,
( 24 portions are connected to the diesel generators other 25 portions rely solely on off-site power. So, there is a O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S W WASHINGTON D.C. 20024 (202) 554-2345
488 l l
i
) 1 full range of those things and appropriate consideration 2 to this sort of event we are discussing today is given 3 in reaching the specific design.
4 MR. ZUDANS: One quicky. In the first initial 5 phase of natural circulation when you the vent, the 6 superheatar is part of your game.
7 MR. CLARE: Yes.
8 MR. CARBONS I cannot hear you.
9 MR. ZUDANS: During the first portioa of 10 natural circulation onset when the auxiliary heat pump 11 is used, the superheater is part of circuit because the 12 vented steam goes through it.
13 On the second phase then you begin with a 0 14 protected air-cooled condenser, the superheater is out 15 of the picture because you take this steam directly from 16 the steam drum and return it back to the steam-down 17 water wash.
18 Does that mean that the superheater then will 19 come up to a full temperature without apparently a 20 secondary sodium -- and stay that way?
21 MR. CLARE: Yes, that is essentially the 22 situation. The control points on these cool vent valves 23 are such that -- I was going to say this one closes
) 24 first and now, as I think about it, I am not certain. I 25 belie ve this one closes first. Then we remove all the ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
489 O 1 hest using the evapora tor and vent only from the steam 2 drum.
3 In that instance there ought to be some 4 probably insignificant heat transfer for the steam in 5 the superheater and then this would become essentially 6 an isothernal piece of steel, would become the hot leg
~
7 of the immediate sodium.
8 MR. EBERSOLE: Just a quick addition. How 9 many DC buses do you have, batteries?
10 HR. CLARE: I believe we have three sets of 11 batteries. We have at least three. Are there any more?
12 HR. BURKHART: There are three for the diesel 13 buses.
14 HR. CARBON: The direct heat removal service 15 requires operation of one pony motor; is that correct?
16 HR. CLARE: The purpose of the pony motor, 17 circulation of primary sodium when the direct heat 18 removal. services is being used, is to provide mixing of 19 the primary sodium, the circulation of sodium through 20 the core.
21 The degree to which that is required depends 22 on serveral factors. The inlet temperature of sodium to 23 the core as well as the decay heat generation rate. So, 24 depending on where you are in a particular scenario you 25 may need three to one or no pony motors in order to O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554 2345
490
() I provide that mixing.
2 Again, in our limiting design case where we 3 assume that there is no other heat sink from the time of 4 reactor scram, our design calcuations have been based on 5 the assumption that there are three pony motors 6 o pe ra tin g.
7 MR. EBERSOLE: What are their characteristics, 8 the size, a te.?
9 NR. CLARE: The pony motors are 75 horsepower 10 motors. They are continuously operating while the plant 11 is on line. They are disconnected from the pump shaft 12 by an overriding clutch which will automatically engage 13 when the shaft speed reduces to approximately ten O 14 percent.
15 Now, even though the pony motors are sized for
.s 16 75 horepower, the actual load that they would see during 17 a ten-percent flow would be roughly 15 horsepower.
18 MR. EBERSOLE: What voltage are they?
! 19 MR. CLAREa I think they are 480 volt motors.
i 20 MR. CARBON: And they run from the diesels?
i 21 MR. CLARE. That is correct.
I 22 MR. CARBON: It is my impression that in the ,
I 23 thermal hydraulic community in general people have
() 24 questions about natural circulation during the 25 transition from the force convection to the natural l
l ()
l ALDERSON REPORTING COMPANY,INC, -
400 VIRGINIA AVE., S.W., WASH;NGTON, D.C. 20024 (202) 554 2345
491
() 1 circulation situation.
2 I gather with the confidence you have 3 expressed here today that the project just does not have 4 any of those reservations?
5 MR. CLARE: Well, we certainly understand some 6 of the concerns and we have tried to look carefully at 7 the possibility of such things as stratification, of 8 flows in the upper portion of our reactor vessel. We 9 have analyzed those carefully. We have done water tests to of that araa. -
11 Perhaps the one test that I can think of that 12 gives me personally more confidence than anything else 13 is the test that was done at FFTF, and I really cannot 14 give you auch more in teras of identification of that 15 test. But what was done was, in one of their
- secondary 16 loops they purposefully controlled the flow in that loop 17 while the plant was naturally circulated; forced the 18 flow to reverse, so that the inertia and the teaperature 19 gradients were all backwards. And then let it go and it 20 re-established natural circulation very well, was very i
21 vell behavad with no problea.
l l
22 Now, that is something, that is one data point l
23 that I find personally very comforting.
l
() 24 MR. ZUDANSs I have a little question. Does 25 FFTF have the same type of arrangement with an auxiliary O
ALDERSON REPORTING COMPANY,INC, l
400 VIRGINtA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 l
492
() 1 feed pump that operates on a steam drum, and the 2 air-cooled condensers which require natural circulation 3 by air, is that exactly the same arrangement?
4 MR. CLARE: No, it is far from the same I 5 arrangement.
6 MR. ZUDANS: What is the top end of that at 7 FFTF7 8 MR. CLAREs They don't have steam generators.
9 They have something you might .think of as being the 10 rough equivalent of our protected air-cooled condensers 11 at the end of their intermediate heat transport system.
12 So, they have s dump directly from their 13 intermediate sodium to the atmosphere. No, they do not O 14 really look anything like our protected air-cooled 15 condensers but in this type of a situation they do 16 operate by air-side natural circulation.
17 MR. ZUDANS: My concern is only about this 18 tail end of your process. There is hardly a question of 19 a primary loop or intermediate loop, they will provide 20 natural circulation.
21 At what time will they be able to cope with a 22 total decay heat, that is another question. But the 23 tail end of it where you depend on semi-active
() 24 components that may not require active power, but they 25 have to be functional - which is no't in FFTF a critical O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
493
() 1 item, it is not tested.
2 MR. CLAREs The turbine-driven auxiliary 3 feedvater pump arrangement that we have in this plant is 4 essentially identical to that used in other light-water 5 pla n ts. For example, I was just reading the safety 6 evaluation report for the Wolf Creek plant where they 7 vere discussing station blackout and, so far as I could 8 tell from that description their arrangement, using an 9 auxiliary feedvater pump, etc., is essentially identical 10 to our own.
11 It would be that experience that we are 12 relying on for those features, rather than FFTF.
- 13 ER. ZUDANS4 I don't disagree with you as long O 14 as you have that turbine.
15 But tell me one thing, supposing you got a 16 reactor scram and you lost the off-site power, how much 17 time will it take - or have you done such calculations -
4 f 18 for the natural circulation to be able to cope with a 19 total decay heat? At the very beginning it will not 20 because, first, it has not been established.
l 21 Mov, as the reactor at the beginning will
! 22 gen era te more power than it removes, it eventually will
[
23 catch up.
l () 24 ER. CLARE: Actually, if you go to the very 25 beginning of the event it is just the opposite because
(~)
ALDERSON REPORTING COMPANY,INC, l 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345 i
l l
l 494 l
l l-1 what happens is, the rods go in before the flow goes 2 down, and you actually remove more heat than you m 3 generate.
4 So, it is a fairly complica ted process there.
5 And that is to a certain extent this transition region 6 that Dr. Carbon referred to, and we have done extensive 7 analyses there. The FFTF tests, several of them, 8 involve that kind of a transition and one does go 9 through a couple of oscillations before things flatten 10 out.
11 Without having detailed curves here and going 12 into them in detail, I am not sure I can provide a much 13 better explanation than that.
G U 14 HR. ZUDANS I think at least I asked sometime 15 to show the heat balance, a snap-shot, to show do you 16 have a zero time, or clearly your pumps are costing more 17 than you need.
18 Now you are telling me that at no time during 19 this particular process do you generate more heat then 20 you remove.
21 HR. DICKSON: The decay heat is down to about 22 15 megawatts at roughly an hour, and that is also equal 23 to what the packs can handle on natural circulation, is h 24 15 megsvatts.
25 Now, you are still removing some sensible O
l ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 l
i 495 l
() 1 heat. So, you are in fact still venting somewhat after 2 an hour. But in principle, if you were not interested 3 in reducing the temperature, you could handle the whole 4 system on natural criculation after one hour and remain 5 isothermal at any point.
6 MR. ZUDANS: I believe I am more concerned l 7 about between zero time and one hour, in that period of 8 time; not what happens after one hour.
1 9 ER. DICKSON: Well, very early on you are up to around 45 megawats, very few seconds after the event.
4 11 About the time you are asking for natural circulation to 12 start it is in the range of 40 to 45 megawatts.
13 MR. CLARE: The vent valves that are provided l ( 14 on the stata drum superheater are capable of taking all 15 of the heat that can be dumped thrugh it, assuming that 16 the primary and intermediate loops are at their highest
! 17 temperature. I do not have a megawatt rating for that, 18 but it is large.
19 HR. ZUDANSs I am sure that it cannot be 20 completely analyzed right now because you have to look 21 at the numbers. But it would be nice to do the 22 different snapshots in time to see what the heat valve l
23 looks like.
n l
')
(_ 24 Clearly, when a pump goes down and a natural l
25 cirulation has not set on completely, it is unlikely to O
~
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 i
496 (g) 1 be able to remove the whole decay heat that exists at 2 that short period of time. Eventually the loop will 3 develop enough flow rate to remove the heat that is 4 generated, and from that point on you probably have 5 clear-cut sailing. I am concerned about this transition 6 period.
7 HR. DICKSON: In that initial transition 8 period the predictions for FFTF, which was followed 9 through in our analysis and then the analysis of Clinch 10 River, which gives simular results except that you do 11 have different responses.
12 The initial close-down, as George has said, is 13 slower than the drop-off in power. So, the tem pera ture 14 within the fuel cell actually drops and begins to rise 15 again, and starts back down again. .
16 The fuel assemblies and each of the blanket 17 assemblies only go through one oscillation. They drop 18 down, rise, and then drop back down. They look like 19 several oscillations if you look at the total outfit 20 because the blankets are lagging; the blankets that get ,
1 21 the lowest flow are something like three minutes. The 22 fuel has gone through this entire transient within a 23 minute after the pump has stopped. The pump goes down lll 24 in something like a minte, 20 seconds. So, within 25 another minute the peak has been reached and pull-down O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
497
(]) 1 started f rom that point. )
2 MR. ZUDANSs I see. And that peak is below r~ 0 what you can tolerate.
4 HR. DICKSONa When we were initially using 5 this natural circulation for what was still considered 4
6 an emergency event, in our initial calculations we had a 7 pump close-down. I believe, in 30 seconds minimum time 8 the pump closed down. And we also used our standard 9 three-signal values.
10 Under those circumstances the temperatures in 11 the blanket cladding did exceed the normal operating 12 temperature and did contribute to the cumulative damage 13 function.
14 Wi th the present pumps having been built and 15 tested and the close-down characteristics are known, and 16 known to be significantly slower than what was assumed, 17 that is used in the equations. Even under these 18 three-signal valves we do almost no damage to the core 19 under nstural circulation.
20 From an analyst's standpoint it is a little 21 unfortunate. The na tural circulation event used to be 22 our enveloping event to all emergency events. It no 23 longer necessarily is because it is in fact bad. But it f( ) 24 appears that on the three-signal basis we do not even do 25 damage to the core. And on our nominal basis, as George n
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
498
() 1 men tioned ea rlier, came very close to predicting what 2 they actually saw in FFTF.
3 In fact, during that first several minute 4 interval of the analysis they wero never more than 30 5 degrees away from the experimr.ntal results. The 6 transient is a very general transient.
7 MR. ZUDANS: Is a ny of this analysis discussed 8 in this PSAR or some place else?
9 MR. DICKSON: It is provided in a document 10 that is now on the docket.
11 MR. CLARE: We have several documents on the 12 general subject of natural circulation. The design 13 calculations slong with some more nominal calculations 14 for the Clinch River Breeder Reactor Plant are in the 15 document WARD-D-308. .
16 We are in the process of preparing a report 17 which will summarize all of our pre-test and post-test l
18 predictions for FFTF. The pre-test predictions were 19 submitted to the NRC prior to the testing actually being 20 performed. We can dig those up. I do not recall the 21 numbers off the top of my head.
22 3R. ZUDANS4 It would be nice to have a chance 23 to see that.
() 24 MR. CARBON: Any other questions, or can we 25 break for lunch?
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S W., WASHINGTON, D.C. 20024 (202) 554 2345
499 l
() 1 MR. LIPINSKI Before you do, I would like to 2 put something in perspective as to the probabilities, 3 the Project may not kov these numbers yet.
4 On March 30, the ACRS had a report from NRC's 5 844 Blackout Task Force. Their preliminary numbers show 6 that they are expecting about a .05 frequen cy per year 7 for a one-hour duration of loss of off-site power.
8 HR. CARBON: Loss of all off-site power?
9 HR. LIPINSKIs Right. Now, I do not have the 10 number here, but some of those events are being caused 11 by malfunctions in the local switch yards. When you 1
12 look at your power lines, that is one area. The people 13 in your own switch yard can cause you to lose power.
O 14 HR. KERR4 Did you say .05 loss?
15 NR. LIPINSKI Yes, .05 per year.
16 ER. KERRs That is much less than they assumed.
17 MR. LIPINSKIs Well, they are taking 15 over 18 life time, and life time is 40 years.
19 HR. DICKSON: Thirty.
i 20 MR. CARBON. They were not assuming them.
21 MR. KERR Their design basis assumption.
22 MR. LIPINSKIs Now, the other one is the 23 diesels themselves, the test persons certainly did some
() 24 questionaires. They came back and with 73 failures out 25 of 1,400 they concluded that the common cause of diesel e
O ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON. 0.C. 20024 (202) 554 2345
500
() I failure was 5.2 times ten minus three, which were not 2 all independent failures.
~
3 then given that you want the diesel to sta rt, 4 they are predicting .02 failure per demand. Given that 5 the diesel starts, its failure to run is 2.4 times ten 6 to the minus three per hour. And then the 7 un a vaila bility of the diesel due to test and maintenance 8 is .006. Then the svarage mean time to' repair a failed 9 diesel is 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. So, if you put this all together --
to MR. CLARE4 We are doing a PRA and we will be 11 sure to take the transcript of this meeting where you 12 have entered those numbers and be sure our analysts are 13 aware of the data you have given.
14 MR. DICKSON Let me ask, is the .05 per year 15 for one off-site power source?
18 NR. LIPINSKIa No, this is per plant, whatever 17 the regulations are. This is the generic U.S. total 18 where you have to have more than one line coming into a 19 plant. Now, some plants have four lines and some have 20 three. This does not distinguish between the lines.
21 MR. CARBONa But it is total loss of all 22 off-site power.
23 HR. LIPINSKI4 Yes.
() 24 HR. EBERSOLE: But it is highly plant 25 dependent.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554 2345
501 1 MR. LIPINSKI: Yes, it is. This is a generic 2 number.
3 MR. KERRa I thought I had seen a number much 4 bigger than that for it than the .05.
5 MR. LIPINSKI This is the number they quoted 6 in the March 30 meeting. But this was, again, a 7 one-hour duration.
8 MR. KERR That is loss and f ailure to recover 9 for one hour.
10 MR. LIPINSKI Right. Now, for shorter 11 periods the probabilities are not as low.
12 MR. CARBON: Do you have a comment, Mike?
13 Fine, let's break.
14 (Whereupon, at 12:15 p.k. a luncheon recess 15 was taken until 12:45 on the same day.)
16 17 18 19 20 21 22 23 24 25 D
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
502 0 1 arrtnaoo" srss1o=
2 (1415 p.m.)
3 Mr. Carbons Let us resume the meeting.
4 Mr. Hillised.
5 PRESENTATION BY R. HILLIARD 6 COMBUSTION CHARACTERISTICS 7 Mr. Hilliards Good afternoon. My name is Bob 8 Hilliard. I am with Hanford Engineering Development 9 LaDoratory and I have been asked to give a brief 10 discussion on the combustion characteristics of sodium 11 and NaK.
12 I have been working in the research end of 13 sodium fires and related type accident spills for the O 14 last ten years at out Hanford. I an afraid that a lot 15 of what I have to say this afternoon is probably common 16 knowledge to you, but I will go over it and if there is 17 some you haven't heard, why that will be new and 18 otherwise it is a refresher for you.
19 (Slide presentation.)
20 Even though I titled the talk Combustion 21 Characteristics of Sodium and NaK, I am really going to 22 emphasize the sodium combustion but I have a few slides 23 on Nak. I have just picked a few of the properties for 24 sodium and Nak to display, as you know, that they are 25 f rom the same chemical family and their chemical O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
503 lll 1 properties are quite similar, sodium and potassium and 2 the eutectic NaK, 78 percent Nak I have listed here.
3 The main difference I have pointed out in 4 red. As far as the combustion properties go, the low 5 s el ting point of Nak forms the eutectic at 78 percent 6 Nak. Potassium melts at minus 12 Celsius which of 7 course means at room temperature it is a liquid as l
8 opposed to a sodium which melts at about 99 degrees 9 Celsius.
10 But above the melting point for both them, 11 then they are both liquid metals and they both burn very 12 reactive with oxidants. Because they are very reactive 13 materials, I could have listed a large range of 9 14 equations that would be involved.
15 I want to emphasize the combustion in accident 16 situations. You usually either have air or in some 17 cases moisture present with the air. So I have listed 18 the dominant reactions in the presence of air. Sodium 19 reacting with half a mole of oxygen going to sodium 20 monoxide is the dominant reaction.
'i
, Exothermic reaction, 104 kilocaleries per 22 mole. Under some conditions that can go further to the 23 peroxidie. In fact, under most air type environments 24 with sodium air accidents this would be the ma te rial we 25 would find out in the low temperature regions away from O
ALDERSoN REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
504 i
() I the fire and a little bit higher heat of reaction.
2 With water vapor two reactions are possible, 3 or more than that are possible, but dominant, sodiun 4 reacting with water to give sodium hydroxide and a half 5 s mole of hydrogen or sodium'with water to give sodium 6 oxide plus a full mole of hydrogen. Which of these 7 reactions occurs depends largely on the concentrations 8 of the oxygen or water vapor and the temperatures.
9 Mr. Marks Excuse me. Hydrogen and air will 10 only burn when there is about four percent of oxygen 11 present.
12 Mr. Hilliarda Five percent.
13 Mr. Marks Four or five. What about sodium O 14 and air?
15 Mr. Hilliard: Sodium reacts with oxygen at 16 slaost any tempersture, at room temperature it will 17 react.
18 Mr. Marks Well, but it isn't the 19 temperature. It will react and burn all the way down to 20 zero oxygen.
21 Mr. Hilliard: That is essentially correct. I 22 have done experimen ts where it will use up all the 23 oxygen clear down to as low as I can measure, which is
( 24 under 1/10th percent oxygen.
! 25 Mr. Marks That is my question. Thank you.
O ALDERSON REPORTING COMPANY,INC,
[ 400 VIRGINtA AVE., S.W WASHINGTON. D.C. 20024 (202) 554-2345 L
505 I 1 Mr. Hilliards But the rate decreases 2 dramatically, a very slow rate.
3 Mr. Mark: Well, proportional to the product 4 of the concentrations, yes.
5 Mr. Hilliards Besides these four main 6 equations, there are some secondary reacti'ons that can 7 occur. Outside of the high temperature flame region of 8 the fire if sodium monoxide is the chief product of the 9 reaction in the flame zone near the sodium, as it 10 diffuses out frrther away to a lower temperature region, 11 but if there is oxygen present at above, say, 10 percent 12 it can go to the peroxide.
13 Also if there is water vapor present, of I
14 course the sodium oxide can react to form hydroxide, no 15 oxygen being liberated in that case, or the peroxide can 16 react with water vapor to form hydroxide and give back a 17 half a mole of oxygen. We ha,ve seen this reaction time 18 after time.
19 Sodium reacts reversibly with hydrogen to form 20 sodium hydride, but at the kind of temperatures we are 21 interested in sodium fires and such this is not stable 22 and associates back in this direction.
23 Of course if there is carbon dioxide present, I 24 and there is a little bit in the ordinary atmosphere, 25 about 3/100ths percent, it can form carbonate, or if I
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
506
() 1 there is more from some other source, why it will go 2 even more to the carbonate.
3 Mr. Marks Do you happen to know with hydrogen 4 in the air if there is a number analogous to that 104 or 5 1247 What is that number, hydrogen burning with oxygen?
6 Mr. Hilliards It is about a hundred, but not 7 on a mass basis. This is per mole basis.
8 Mr. Marks It is per mole.
9 Mr. Hilliards Yes.
10 Mr. Mark It is about the same sort of number.
11 Mr. Hilliards Yes, on a solar basis.
12 Mr. Marks So the pressures are about the 13 same, or the temperature rises.
14 Mr. Hilliards Well, let me show you this 15 slide. I have a couple of slides here that perhaps will 16 help compare it with some more typical fuels.
17 Now these are just nominal values, but it will 18 give you an idea of the relative amount of heat 19 liberated from sodium oxidation with some things like 20 vood. This is the maximum energy producing reaction, i
21 4800 BTUs per pound versus wood, 8,000, coal 14,000, a 22 typical gasoline, 20,000. So on a mass basis it is a 23 relatively low heat producer.
() ,
24 Then as far as some of the other 25 characteristics, in this case I have selected some O
ALDERSON REPORTING COMPANY,INC,
(
400 VIRGINTA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
507
() I hydrocarbons, typical ones, kerosene and alcohols.
2 One of the distinctions of sodium is it has a 3 very short flame length because the sodium is a fairly 4 non-volatile materials. It boils at 1620 Fahrenheit.
5 Most fires, the pool is going to be down in the region 6 of 1100 or 1200 or possibly 1300 degrees Fahrenheit. So 7 the vapor is very low vapor pressure as opposed to a 8 hydrocarbon fuel where you can get several meters flame 9 height. This is more lika chsrcoal, a very, very lov 10 flame height. The flame temperature also is lover, less 11 than 1400 C compared to all the way up to 3000 C for 12 typical hydrocarbons.
13 The moles of gas after the reaction is 14 actually less than the initial air because it consumes 15 oxygen and doesn't have gaseous products of reaction as 16 opposed to hydrocarbons which increase due to gaseous 17 reaction products.
18 The convection currents would be less because 19 of both the less heat and the lower flame temperature.
20 Also the radiant heat is auch less because of the lower 21 flame temperature and the less heat.
22 One bsd thing about sodium is it does give off 23 more aero3ol products than typical fuels because the
() 24 sodiur oxide or hydroxide nucleates to small particles.
25 So the reaction product itself is a aerosol product. I O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
508 (f 1 will show a little schematic of that in a minute.
2 Mr. Carbons What specifically do you mean by
, 3 aerosol? I know I use the word in general but yet I am
/4 not certain. That is a good question because maybe we 5 all have different ideas. When I say aerosol I am 6 really not being as true as I should be. I tend to 7 think of it as the particle itself, which can be either 8 liquid or solid, floating suspended in a gaseous medium, 9 whereas, it is really the whole system, gas with its 10 suspended mass. I tend to keep talking about the 11 serosol as the particles that are floating around when I 12 say the aerosol. I think most people do.
s 13 This is a schematic of a pool fire model of i
14 how it burns. It is somewhat a function of the 15 temperature, but I will start down here with a liquid 16 sodium pool. This can be any depth. At some 17 temperature it exhibits a vapor pressure so that sodium 18 vapor diffuses upward.
19 I am talking about a sodium mir system now.
20 0xygen is diffusing down towards the system and they
, 21 meet in a reaction zone. The distance off the surface 1
22 here is dependent on the temperature of the fuel. At K 23 low temperatures this whole zone sits right down on the 1
() 24 surface. So there is sodium diffusing and there is a
! 25 heterogeneous reaction right on the surface.
i 4 ALDERSON REPORTING COMPANY,INC, I' 400 VIRGINIA AVE., S.W WASHINGTON. D.C. 20024 (202) 554-2345
509
() 1 At higher temperatures and really te m pe ra tu res 2 that we get more interested in for accident analysis, it 3 is lif ted above the system. It is importan t only in 4 that a lot of pictures you see can be at low temperature 5 and at low temperature where the reaction occurs right l 6 on the surface we get a lot of nodules, rough. This ;
~
l 7 acts a wicks and you will get all sorts of strange l 8 things here. But at temperatures above, say, about 700 9 or 800 Fahrenheit on above that, this is usually a flat 10 surface and all the reaction products that reach the 11 pool are mixed into the pool.
12 Diffusion of sodium vapor upwards and oxygen 13 down, this is one of those dominant reaction that I
- 14 mentioned going to the monoxide. The monoxide being 15 very non-volatile nucleates to small particles in a very 16 high concentration so that they grow in size. Then they 17 either are swept out away f rom the burning. surface or 18 they settle back to the pool. I will talk about the j 19 fraction split here in a few minutes.
20 I also want to point out of course that heat 21 is liberated in this zone and goes both ways also. The 22 heat back to the pool sustains the vapor t ?.a tion of 23 sodium vapor plus any heat losses out the b6ttom and
() 24 sideways.
25 Once the sodium monoxide leaves the area by O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
510 ll 1 convection currents, it can go through some of these 2 secondary reactions. If the oxygen is above about 10 3 percent and the temperature is lower out here it will go 4 to the peroxide and be swept far away from the fire as 5 an aerosol particle.
6 There is a very parallel system for water 7 vapor. That which I just showed you is for dry air and 8 for many environments. Say in the ordinary room type of 9 at atmosphere there is water vapor or if there is a 10 water producing reaction such as hot concrete, then 11 water can be released. In any event, for the sodium 12 vater vapor model the same thing happens with the water 13 coming down and sodium vapor up. In this case we would l 14 get something like sodium hydroxide, the main difference 15 now being that you can get some hydrogen from this 16 reaction.
17 It is of interest to us and to everybody I 18 think, weil, what happens to this hydrogen. ' dell, if 19 there is an air atmosphere and actually it is just about 20 5.0 percent oxygen required for this reaction. It is 21 the combustion limits of hydrogen and oxygen.
22 As hydrogen leaves and goes out and reaches an 23 oxygen that is better than five percent, then it will 24 burn back to water because it is already above its 25 ignition tempe rature being this close to the burning l
ALDERSoN REPORTING COMPANY,INC, i 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
i 511 This water then, and of cource some of it can go l( ) 1 zone.
2 down, but most of it will a ttach to aerosol particles. l
's
- 3 These particles just love water and of pourse this will 4 go to hydroxide and leave as an aerosol
- particle.
5 The main point being here that ;ven though 6 there may be lots of water or hydrogen say coming out of 7 here if there is a sodium concrete reaction or being 8 generated by this, it will go back to water and go and 9 tie itself up with aerosol particles and hydrogen is 10 removed in that manner.
11 Mr. Zudans: Could I ask for a clarification 12 because I got a little bit confused. On this slide here 13 you showed some negstive and some positive results of
,O
~
14 reaction and here both are exothermic.
15 Mr. Hilliards Thank you. I made an error on 16 that slide.
17 Nr. Zudansa They are both exothermic 18 reactions.
19 Mr. Hilliarda Those should have minus signs 20 in front of the 45.7 and 42.2.
21 Nr. Hilliards Otherwise I hope you will tell 22 that they both are cooled down.
23 (Laughter.)
() 24 M r. Hilliard s I didn't proof read carefully 25 enough. If you would write just a minus 45.7 and 42.2.
O l
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
l 512
() 1 They are both exothermic reactions.
2 The type of aerosol product that is formed
- 3 from this type of combustion, there is sometimes some 4 confusion about whether you get sodium peroxide or 5 sodium monoxide. I have tried to show in general terms 6 thst at low oxygen concentrations you almost always get 7 the monoxide. Also, at high tam..eratures like above 650 8 to 700 Celsius we get the monoxide. But at low 9 temperatures and above about 10 percent oxygen we get to sodium peroxide.
11 We have some some experiments for some other 12 purposes actually. I hope you can see this. I show 13 this only to demonstrate. This is a sodium peroxide O 14 aerosol. It has some hydroxide mixed with it because in 15 ordinary air there is always enough ambient water vapor 18 to make some sodium hydroxide.
17 As I remember this was about two-thirds 18 peroxide and more or less the balance was hydroxide. It 19 is d ry a nd flu f f y and an operator going in after the 20 fire stirs, it goes it every which direction. It 21 settles and I think you can see here that it settles on 22 upward facing surfaces, but the verticle surfaces are 23 almost free. Actually the flux downward compared to
() 24 sideways is about a factor of a hundred to one. That is 25 because the aerosol particles have mass and i
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D C. 20024 (202) 554-2345 L
513 l
() 1 gravitational settling is more important than diffusion 2 or other forces that get them over to other surf aces.
(g 3 Mr. Mark: Are you going later to talk about V
4 the kinds of pressures that occur in one or another !
5 event?
6 3r. Hilliards I wasn't, sir, because I think 7 Cliff Boasso is going to talk about tha t specifically 8 for CRBR.
9 Mr. Marks If it will come up later then I 10 will leave my questions for later.
11 Mr. Hilliards I was really trying to be kind 12 of fundamental in my talk about combustion. I 13 Mr. Mark: Well, pressure is pretty O' 14 fundamental.
15 M r . Hillia rd a Yes. Here now for one that is 16 more richer in sodium hydroxide, this cell is a 30,000 I:7 cubic foot cell, a rather large vessel. We sprayed 18 continuously for about 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> spraying sodium into an 19 air atmosphere and at the same time we are injecting 20 steam to stimulate the release of water from heated 21 concrete. So we converted to sodium hydroxide and it 22 has essentially no hydrogen or 1/10th, 2/10ths percent 23 hydrogen formed during those types of experiments.
) 24 Then after the test, you can see this is 25 clumped up almost like wet snow. Sometimes the aerosol O
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE S.W., WASHINGTON. O.C. 20024 (202) 554-2345
514 g
I would be four or five inches deep after a test and a man 2 could walk on it like kind of icey snow as opposed to a 3 peroxide that is dry. -
4 Also then in some tests we released carbon 5 dioxide, and here again, very fluffy. The same amount 6 of sodium would pile up three or four tirnes deeper 7 because this is more compacted, sodium carbona te.
8 Mr. Ebersole: Apart from the thermal effects, .
9 what does this stuff does if it drifts out as smoke and 10 begins to permeste the electrical apparatus, et cetera?
11 Mr. Hilliard You know, I don't know, sir.
12 ile have never had any particular problem with it in the ,
13 experiments we have done and we haven't checked that 14 specifically. But we hu e talked and I have seen some 15 reports. Of course it is not good because we pick up a -
16 little bit moisture then and it becomes an electrolyte -
17 or' conductor. _
18 Mr. Ebersoles You are always in the presence 19 of some moisture. k 20 Mr. Hilliard Yes. ]
21 Mr. Ebersole. So I am just wondering if it 22 represents, say, a shorteircuiting potential throughout 23 the plant.
h 24 Mr. Hilliard4 Cliff, is anybody going to be 25 able to address that.
O =
=
M ALDERSON REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
1 515
'O i hr. 01ckson, rhie is reu1 01cxeon of 2 Westinghouse. Part of our environmental qualification 3 is to quality our components to the sodium aerosol 4 environment they are expected to see for the worst 5 design basis accident.
6 Mr. Ebersoles Does that mean they have got to 7 be practically watertight?
8 Mr. Dickson In some cases, yes, that might 9 be the solution. In other cases they don't have to be.
10 It not only acts as a conductor when it is vet, as you 11 have noted, but it also acts as an insulator wherever it 12 settles also.
13 Mr. Ebersole4 Thank you.
14 Mr. Hilliard We put equipment in these cells 15 when we have fires. Of course, we have lots of 16 samplers, solonoid valves that operate right in the 17 system, electrical lights and so on, but we haven't had 18 any particular trouble as long as we don't get too wet 19 or too hot.
20 On the kenetics of the burning, I didn't think 21 I would take time to go deeply into it, but I thought I 22 would show you this simplified approximate equation for 23 a pool fire. It says that the rate in kilograms of 24 sodium that burn per square meter per hour is 25 approximataly two times the parcent oxygen because it is O
ALDERSoN REPORTING COMPANY,INC, 400 V!RGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
516 k 1 proportional, directly proportional to the percent of 2 oxygen in the atmosphere. That is for a 21 percent 3 oxygen normal air atmosphere we would get about 42 4 kilograms of sodium per square meter per hour which is 5 about two inches of pool depth per hour. But for a low, 6 like a two percent oxygen atmosphere it would only be .
7 about four.
8 Mr. Marks Will someone, either you or someone 9 later, be able to make some comment on the kind of 10 pressure that is associated with that kind of behavior?
11 Er. H1111ards Yes. Cliff is nodding yes. He 12 will be talking about that.
13 Mr. Mark: Very good.
D 14 Mr. Zudins What is the specific weight of 15 sodium in this liquid state?
16 Mr. Hilliards The specific ---
17 Mr. Zudansa Weight.
18 Mr. Hilliard Well, at 1100 Fahrenheit it is 19 50.5 pounds per cubic foot, a little bit lighter than 20 water. That first table I gave you gav.s the density, 21 and I have forgotten, .85 or somethink like that, in 22 there. I kind of remember 50 pounds per cubic foot at 23 these temperatures like 1100 degrees. Of course, it is 24 a function of temperature.
25 If the exact rate of a sodium fire is wishes, D
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
517
() I then we can use a model that calculates by natural 2 convection the rate of diffusion to the surface or to 3 the burning zone and actually that is what the Sodium 4 Fire Code does.
5 I was going to skip over in the interest of 6 time these next few slides that get into that unless 7 there is a question on it. Actually the kenetics are 8 handled by analogy the heat transfer. I am skipping 9 over about the next four slides unless you have 10 questions on them.
11 I am going to point out that firest can occur 12 in different geometries. Actually the only difference 13 between these are the geometries, the pool fire being as O 14 it indicates a pool on the floor, spray being a finely 15 divided aerosol or spray of sodium metal falling through 16 the gas, vapor being a jet of sodium perhaps with a 17 carrier gas leaving, the columnar just being a pipe 18 break say and a column just descending through the air 19 or it can splash against a wall.
20 This next viewgraph, which didn't turn out at 21 all in your copy, we spilled somewhat over a ton of 22 sodium and there are about four types of fire going on 23 at one time here. This is a three inch pipe, and there 0) x 24 are about 300 gallons a minute of sodium at 1100 degrees 25 Fahrenheit in an air-filled call. We also had a
()
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
518
() 1 one-inch pipe directed against the wall. So we had a 2 vall with a wall film flowing down it. It splashed off l
l 3 of the floor and these larger drops actually bounce off 4 the floor. We mopped this up with water, too, to get an 5 idea of the bounce. Sodium is really very similar to 6 water in its viscosity and density.
7 So we had the combinations then of just about 8 every type of fire we have. This is a picture then of 9 the cell. This is a 3500 cubic foot cell in this case 10 and had about 10 square meters of floor surface. After 11 it used up all the oxygen in the cell the fire just went 12 out and it took about 15 or 20 minutes, like putting a 13 candle in a bottle and putting your hand over the top.
q,-)
14 It was a vented cell. It had a three-foot by 15 three-foot louver to relieve the pressure and we only 16 got, as I recall, about less than two inches of water 17 gauge pressure build-up during this particular fire.
18 But you can see the men are in there now chopping this 19 sodium. This is mostly metallic sodium left after the 20 fire and they are chopping it up to put it in the barrel 21 and eventus11y dispose of it.
22 Mr. Marks If there a large sensitivity to the 23 temperature of this sodium? Sodium comes out as a O
\- 24 liquid let's assuae. You said the temperature of the 25 sodium in that was ---
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554 2345
519
() 1 Mr. Hilliards 1100 Fahrenheit.
2 M r. Marks 1100 Fahrenheit. It doesn't really 3 matter, does it, whether it would have been 1200 or 10007 4 M r. Hillia rd s Not really.
5 Mr. Mark There is a slight increase in the 6 Cates, is that right, as the temperature should be 7 raised?
8 Mr. Hilliards The rates are mass diffusion 9 dependent more than the tem perature of the sodium. It 10 is not chemical reaction rate limited. It is mass 11 transfer of oxygen. So that the only way the 12 temperature enters in then is if you say the oxygen is 13 at some higher temperatures and convection currents are O 14 greater at higher temperatures. It is a slight increase 15 but not much.
16 Mr. Marka Okay. Well, that is what I was 17 supposing and you are telling me that is about right.
18 Mr. Hilliard That is about right. I will 19 have a viewgraph here in a minute called " Ignition 20 Temperatures." I really hate to talk about ignition 21 temperatures because there isn't any exact ignition 22 temperature for sodium. It reacts at almost all 23 -temperatures, but there is a temperature above which it 24 will increase in temperature and I will talk about that 25 in a minute.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554-2345
520
() 1 Mr. Ebersoles In that connection , are these 2 men actually chipping sodium off the floor?
3 Mr. Hilliards Yes, sir.
)
4 Mr. Ebersoles Isn't there a possibility of 5 high local temperstures that will start something in the 6 chipping process?
7 Mr. Hilliards This was done after cool down 8 to ambient. Then we poured some oil over the surface.
9 Mr. Ebersoles Oh , then that settles it. You 10 don't need to go any further. I figured you had to use 11 something.
12 (Laughter.)
13 Mr. Hilliards Actually I will back up. This O 14 is a steel liner over concrete. This was a one-foot 15 special test, one-foot deep and one-foot on the wall and 16 then a carbon steel liner was put on it. So when this 17 -heated up just f rom the sensible heat of the sodium it 18 drove water vapor out and you can see right up along 19 here that water vapor came out and reacted as those 20 equations I showed. After the oxygen was gone water l 21 continued to come out and we had sodium hydroxide.
l l 22 Sodium hydroxide then had a surface right over the top 23 of the sodium a nd it was more like that sodium hydroxide 24 experiment that I showed you. So that it was not 25 pyroforic and men could walk on it. But just while we O
ALDERSON REPORTING COMPANY,INC, l
400 VIRGINIA AVE S.W., WASHINGTON. D.C. 20024 (202) 554-2345
521
() 1 were chopping up we put some turbine oil over it and 2 that way kept it from generating any aerosol or
(} 3 overheating.
4 This slide is a composite, I really only 5 vanted to refer to the upper-right-hand corner because 6 this shows you a little bit better a small pool fire.
7 It gives you an idea. Here it is. You can cee sodium 8 being poured out, and again I think that temperature was 9 under 1000 Fahrenheit in this case. As I recall, it 10 ven t down and stayed down and ignited a few seconds 11 later. It was't an immediate ignition. But then it 12 stead over the entire surf ace and here comes the aerosol 1
- 13 up.
- 14 A man could really walk right up next to this, 15 except for one problem, that is this smoke usually comes 16 down and eventually fills the room and the man has to 17 have an air assist of some sort to breathe in it. But I 18 vanted to show it because you get a good idea of how it 19 comes off.
20 This is just a drum that we buy, 400 pounds 21 solid pack. We melt in this thing and then drive it 22 over into here. Or if we want high temperatures, we put 23 it over in a secondary tank and drive it over. This is 24 of some interest in that after some tests where some 25 reaction products have fallen on the floor and then a
()
ALDERSON REPORTING COMPANY,INC, 400 VIRGtNIA AVE S.W., WASHINGTON, o.C. 20024 (202) 554-2345
522 l
[
() 1 man disturbes it by walking on it, you can actually 2 start some of these pyrophoric reaction products afire 3 snd it makes the men kind of jump around and get a 4 little excited when they do this.
5 It really would probably not ignite any sodium 6 because sodium has so much heat capacity that it is very 7 difficult to start a fire with sodium. When firenen try 8 to show how to do fire protection or what sodium looks 9 like, they will go out an put a blow torch on an ingot 10 of sodium and th ey have a very difficult time getting it 11 to go because it has such good heat capacity that it 12 loses its heat to the ground almost as fast as they are 13 putting it in with the blow torch.
O 14 This is the viewgraph I had on sodium ignition 15 temperature. I have defined it as the minimum 16 temperature at which a self-sustaining temperature rise 17 occurs. In other words, at room temperature sodium l .
18 could probably sit there and never ignite, but it would 19 slowly react over a matter of days and probably pick up l 20 water. I have seen it condense water and form a lake of l
21 water standing over metallic sodium and just stay there 22 because it seals itself off with some crystalized sodium 23 hydroxide that separates the sodium from the water.
24 On that basis if you keep raising the 25 temperature there vill come a point that heat gain ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
523
) 1 exceeds the heat loss and from then on temperature 2 increases to some other higher stable temperature.
3 For a pool of sodium that is undisturbed, that 4 is if you had a lid over it and an inert gas and you 5 filled it and then you just removed the lid , it can be 6 and normally is up near 400 degrees Celsius and possibly 7 an even higher temperature is required before it will 8 ignite and take off. If you disturb the surface like 9 pouring additional sodium into it it is around 200 10 Celsius. .
11 I have just pointed out this extinguished 12 pool. I showed you the pictures of the where the men 13 had stepped in some of the pyrophoric reaction products 1
'^#
14 and that pyrophoric reaction product from an 15 extinguished pool can have little, fine metallic sodium 16 aerosol entrapped with some sodium peroxide and then it 17 kind of insulates itself until you disburb it and then 18 it can ignite and give off a little heat.
19 For spray drops there really isn't any good 20 value for an ignition temperature because it is a 21 function of drop size. A very, very small mist will 22 ignite at almost any temperture. Of course it has to be 23 at least 98 bef ore it will be liquid, but we put down 24 120 as kind of an average value.
25 After the reactions occur and we form an 9
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE S.W., WASHINGTON, D.C. 20024 (202) 554 2345
E 524 lh 1 aerosol there is no analytic model to predict the 2 faction that forms an aerosol of the combustion 3 product. But by many experiments it is found that pool 4 fires range from about 15 percent to about 30 percent of 5 the reacted sodium that leaves as an aerosol with high 6 temperatures and large surfaces favoring the higher 7 value and low temperatures and the smaller surface the 8 smaller value.
9 For a spray fire it is generally assumed that _
10 100 percent of all the reacted sodium will leave as an 11 aerosol, although we think it is probably somewhat less 12 than that. It might be 90 percent or something less.
13 There are two more slides. This is on the 3 14 non-radiological chemical toxicity of the product. The 15 only values available to us are what is published by the 16 American Conference of Government Industrial -
17 Hygienists. The latest one that I have available is 18 1981 that lists only for both sodium hydroxide and 19 sodium peroxide, and under the oxides they just say use 20 the same value 'eus the hydroxide, which is two milligrams j 21 per cubic seter. This would be of the suspended mass of 22 particles per cubic meter of air on a time weighted ,
23 average for an eight-hour exposure.
24 Notice this is an occupational limit and not 25 an emergency dose, but for people that are working is ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
525
() 1 sodium hydroxide factories I guess. For our purpose 2 they say you can have excursions to double that as long
'()
3 4
as the time weightei average comes out to two milligrams over an eight-hour time.
l 5 For the carbonate, both sodium and potassium, 6 there is no value given. They list it as a slight 7 irritant and we know it is not nearly as caustic or 8 harmful to the body as the hydroxide.
9 Mr. Marks If you spend eight hours with that 10 concentration what happens, you keel over?
11 Mr. Hilliard No, this is an occupational 12 permitted value where workers can work day and ---
13 Mr. Marka Okay, if you do that day after day O 14 you are still all right.
15 Mr. Hilliarda Supposedly.
. 16 Mr. Marks At what level does it become in the 17 short term lethal? At some level it must.
l
) 18 Mr. Hilliards I don 't' kno w. Usually what 19 happens is that it gets irritant on your nose and eyes 20 and you leave at some value.
21 Mr. Marks At this value you are not even 22 badly irritated?
I
! 23 Mr. Hilliard: Oh, at this value. At this 1
! 24 value you are just beginning to get where you might 25 detect that there is something different than normal in i
. ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
526
() 1 the air.
2 Mr. Lipinskis There is nothing visible in the 3 way of smoke at that point?
4 M r . Hillia rd s Yes. I think at two milligrams 5 per cubic meter there would be a visibility impairment, 8 but in a short-range room like this it wouldn't be much 7 worse than the pipe you are smoking or something like 8 that. However, at very much above two it begins to get 9 where you =an data:t it and want to leave. I think I 10 have probably breathed it up to 50 or something like 11 that and didn't want to stay too long at those kind of 12 conditions.
13 Mr. Ebersole: Does this material have a O 14 tendency to clog up filters pretty badly?
15 M r. Hilliarda Yes, it does.
18 Mr. Ebersole We have electronic modules 17 which believe it or not are ultimately dependent on 18 little motors operating through filters. When one 19 searches around for common mode failures and of course 20 if you clog up the filters the f ans don' t work any more 21 and they don't work without fans. So this smoke would 22 have to be looked at as a potential common mode failure, 23 wouldn't it?
() 24 Er. Hilliards Well, I think you need to be 25 aware that any kind of an aerosol is picked up by a O i ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
527 O
A_j 1 filter. That is the purpose of the filter and the 2 filters have some limited mass that they c1n have 3 without the pressure drop getting excessive. We have 4 done a fair amount of research on this for things like 5 standard heat filters that are made for air cleaning and 6 sand and gravel beds and other things on the pressure 7 drop as n function of these aerosols that get picked 8 up. We have information available on that.
9 Mr. Ebersoles Thank you.
~
- 10 Br. Hilliard
- Now once this aerosol gets l 11 released out into the environment and drifts downwind 12 there,isn't very much available on the kenetics, but we
- 13 do know that it only takes a few seconds at relative
!O' j 14 humidity above about 30 percent to form the hydroxide
, 15 and just a few more seconds if there is carbon dioxide 16 available, which there is in the atmosphere, and I 17 mentioned 3/100ths percent normal, but again this would j
18 be a mass transfer of carbon dioxide down to the plume 19 before this reaction can occur. So I can't really say
- 20 30 seconds." It would have to be modeled.
i 21 In loop experiments in Germany and now ongoing 22 in France and in experiments where we injected carbon 23 dioxide we know it is very fast as long as you can get 24 the reactants together.
- 25 So the point I am making is that by the time lO I
ALDERSON REPORTING COMPANY,INC.
400 VIRGINIA AVE., S.W WASHINGTON, D.C. 20024 (202) 554 2345
528
() 1 it gets very far downwind you are no rmally going to have 2 sodium carbonate. Sodium carbonate, as I listed in the 3 previous slide, there is no value given for the
(}
4 toxicity. In fact, this material of course is mined out 5 of the ground in Idaho and used very commonly, sodium 6 carbonste and' sodium ash. Eventualy going at much 7 slower, probably days and maybe weeks, you will get to 8 the bicarbonate which of course is baking soda.
9 That is about all I had in my talk, but I 10 would be glad to answer any questions.
11 Mr. Carbon : What can you say about what kind 12 of fire and how serious to where visibility drops down 13 to where the person could not see very much ?
O 14 Mr. Hilliards It impairs visibility.
15 Mr. Carbons I don't mean from the standpoint 16 of hurting the person's eyes.
17 Mr. Hilliard No, but just seeing. In this 18 big vessel I described to you we have windows so we can 19 look in and we also have known targets that are a i
20 certain distance from us.
j 21 I don 't know if you can see this or not, but
- 22 this is the suspended mass concentration in grams per i
! 23 cubic meter. ,
l 24 Mr. Cir' bon: Do we have that?
25 Mr. Hilliarda No, sir. This was an extra O
I ALDERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
529
() 1 one. I wondered if a question would come up. So I ,
1 1
2 included it. The visual range, I don't know what is 3 safe for your suspended mass concentration. Outdoors it 4 would be done around here somewhere or lower, but in an 5 enclosed room we normally measure values up around 10 or 6 20, something like this, from a big in an enclosed 7 room.
8 In those kinds of conditions the visibility is 9 two to three tenths of a meter. In other words, the 10 smoke, you can't see more than a foot or so for a big 11 fire in an enclosed room. But as it gets down to, oh, 12 like we were talking about two milligrams per cubic 13 meter, this is a hundred, so it would be way out here.
O 14 I don't know if this is fair to extrapolate.
15 By the way, I do reference Harvard Air 16 Cleaning Lab and if Dave Moore were here he would be 17 glad to see that I had his data on there also. It is 18 roughly the same ss ours.
! 19 I don't now if I have answered your question
- 20 or not, but 2t does impair visibility at high 21 concentrations.
i 22 Mr. Carbont Along the same lines, what sort 1
I 23 of practical fire might you have where In individual l 24 says, gee, this is hurting my eyes and I worry if I 25 would go blind or something and the heck with it I will
(^)
ALDERSON REPORTING COMPANY,INC, j
! 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554 2345
530
) 1 throw down my tools and get out?
2 Mr. Hilliard s Well, I guess I have 3 experienced that. I have thrown my tools and gotten out
(])
4 a time or two.
5 (Laughter) 6 Mr. Hilliarda This isn't something I could 7 quote so I didn't put it in here, but the British did a 8 pseudo experiment along this line where they had three 9 men in a cell and they built a fire and they were to mea su ring the concentration and they earphones and so on 11 so the guy could talk and give his reactions. As I 12 recall, he stayed in till, I don't know, somewhere up ,
l 13 close to 100 milligrams, but then he had to get out, 100 14 millicrams per cubic meter. It began to bother him too l
I 15 much. There were three guys actually. -
[
i 16 As I say, it is a report I had at an 17 international meeting that said not to be published. So 18 I couldn't cite it, but I tell you they have done things 19 like this. To me it meant that, well, for emergency 20 short-time there is no damage, and there isn't any limit 21 that I know of, quantitative limit.
22 Mr. Ebersoles Is that just because they l
l 23 haven't gotten around to finding out what the long-tern 24 effects are?
25 Mr. Hil11ards I don't know. There haven't O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
531
(} 1 been too msny people exposed.
2 dr. Ebersoles I mean, that used to be the 3 case with asbestos.
(])
4 Mr. Hilliard: Asbestos, I suppose they were 5 working with it day in and day out. Sodium fires, 6 people don't normally work day in and day out even us 7 with our experiments. We are pretty far apart sometimes.
8 (Laughter.)
9 Mr. Carbons Mr. Hilliard, you are doing work to on sodium still, are you not, at this time?
11 Mr. Hilliard Yes, sir.
12 Mr. Carbons It is my impression that we 13 understand cool fires pretty well, but that there are O 14 still questions or RCD needed for spray fires; is that 15 correct? What is the situation there?
16 Mr. Hilliards Yes, there is quite a bit of 17 effort going on in the world on sodium fire technology 18 at the present time.
19 Mr. Carbon Are these spray fires primarily?
20 Mr. Hilliards Mostly spray fires, yes. Code 21 development and experiments both. We have a large spray 22 fire planned for September and then some continuation in 23 this large 30,000 cubic f oot vessel, but it is primarily 24 to investigate aerosol behavior and not the fire 25 itself. We use that as a way of generating aerosol. I O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
532
() 1 think the following speaker will talk some about spray 2 fire for CRBR. But, yes, you a re right, there is a 3 great deal of interest in the world.
4 Mr. Carbon: But again for the aerosol 5 activities, to understand the aerosols better.
6 Mr. Hilliarda We are just doing it to 7 validate aerosol behavior codes, how fast they
~
8 coagulate, settle and move to surfaces. There are 9 computer codes available in the world and we wanted to 10 validate them.
11 Mr. Kastenberg: I have just a general 12 question. Are there any implications in terms of 13 radiological doses, off-site doses from accidents in 14 terms of sodium fires and aerosol behavior and so on?
15 Mr. Hilliard: I don't know if I understand 16 your question, but if there is radiological source term 17 there almost certainly will be sodium along with it 18 because sodium is the coolant and there is so much more 19 aass of sodium than there is any radiological source 20 term. This is very helpful in that aerosols are 21 depleted by settling mainly. So the more mass you can 22 put in they will coagulate with some small radioactive 23 particles attached to the sodium fire particles that
(')\
's_ 24 they settle out much more rapidly. So from that point 25 of view you would like to have lots of sodium smoke.
)
i ALDERSON REPORTING COMP ANY, INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
533
() 1 Mr. Kastenberg: Could you have the opposite 2 effect where the heat of reaction would tend to be move 3 of a driving force to give you more dispersal?
4 Mr. Hilliards Yes, that is one thing that is 5 designed I think. The following speakers are going to 6 address this more. This is getting sort of off from the 7 f undamental combustion behavior that I was talking 8 about. I don't mean to dodge you, but I think they 9 could probably answer your questions better than I can.
10 Mr. Haskers Did you say anything about what 11 kind of typical particle size range comes out in the 12 combustion process?
13 Mr. Hilliards No, I didn't, but I can speak
- O-14 to that. I have been researching it quite heavily. It 15 depends on the concentration and the size. Most of 16 those tests like I was showing you had about a five 17 micron mass median diameter with a standard deviation of 18 about two and a half, somewhere between two and three.
19 By that I mean on a log probability, a log normal 20 probability paper of a mass median, half of the mass 21 would be above about five and half below the slope of 22 the line, two to three.
23 From small fires just a little bit. Where the l
) 24 concentrations aren't as high they will be down more 1 25 like two, or something like th a t . The primary size of
!O l
l ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
534
() 1 the particles are a ftv tenths of a micron, but they 2 very rapidly conglomerate to larger particles. In fact, 3 we measured them clear up to 50 microns and 100 microns, 4 very large particles.
5 Mr. Hasker: Does that hold for the spray 6 fires, too?
7 Mr. Hilliard: Yes, even more so. They are in 8 high concentrations.
9 Mr. Ebetsole: What is the most efficient var 10 to stop a fire? What do you use?
11 Mr. Hilliards Just take away the oxygen. It 12 is the same as an y o the r fi re . You either have got to 13 get rid of the fuel, the oxygen, or cool it down where O 14 the - reaction rate is so slow that it will lose its heat 15 f aster than it qsins it. The only practical way I know 16 of is to stop the oxygen coming in.
i 17 Mr. Ebersole Does that mean you use dry - -~-
18 Mr. Carbon That must be the next topic.
19 Mr. Ebersolea Oh, okay.
20 Nr. Carbon: Thank you, Mr. Hilliard.
21 PRESENTATION OF C. BOASSO 22 LIQUID METAL FIRE METIGATION APPROACH 23 EVALUATION OF DESIGN BASIS SODIUM FIRES 24 Mr. Boassos My name is Cliff Boasso from 25 Westinghouse and the Clinch River Project.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
535
() 1 This afternoon I would like to take the 2 opportunity to talk about the liquid metal fire 3 mitigation approach that we utilize in Clinch River.
)
4 (Slide presentation.)
5 Mr. Boasso The fire mitigation approach 6 utilized in Clinch Rivers Controlled environments, 7 inerted cells, argon cover gas on our sodium systems, 8 structural features, lined cells to contain potential 9 radioactive sodium spills as well as catch pans, catch 10 pan fire suppression deck systems that I will get into 11 later on, as well as a sensitive' leak detection system 12 utilized in conjunction with operator action.
13 As I indicated, the controlled environmental O 14 features. Those cells containing radioactive liquid 15 metal are inerted with nitrogen to minimize sodium 16 burning and all liquid metal systems are supplied with 17 inert argon cover gas.
18 Structural features. We have an all welded 19 steel cell liner system in inerted cells to contain i 20 spill volumes. Behind these cells liners we have 21 insulation to protect our structural concrete from the 22 thermal loadings. We have a containment confinement 23 structure to contsin radioactive aerosols for postulated 24 accidents within the containment building. We have 25 steel catch pans in air-filed cells to contain spill O
ALDERSCN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
536
() 1 volumes. In addition, in the air-filled cells we have a 2 fire suppression deck system to minimize long-term
{) 3 4
sodium burning effects. And, as with the cell liners, ve have insulaticn under the catch pans to protect our 5 structural concrete.
6 Mr. Carbon Is the insulation there typically
'7 a type of concrete?
~
8 Mr. Boasso In the inerted cells we use a s
9 perlite lightweight insulating concrete. In the 10 air-filled cells we are using a magnesium oxide
' - 11 aggregate as the insulating material to protect the 12 structural concrete.
m
[13 Mr. Zudansa You say that all liquid metal O 14 systems are supplied with an inert gas cover.-
15 <dr. Boasso: The cover gas system.
16 Mr.'Zudans What about the steam generator on 17 top of the sodiun. I asked the same question of Mr.
'18 Dickson and he said 'no.
~
19 Mr. Boasso: I think I will have to refer that 20 question back to Dr.t!Dickson.
21 Mr. Dickson: Maybe I should get someone else 22 to answer it. Th3 answer is there is no cover gas up 23 there. It is filled.with sodium.
O k(/
l 24 25 Mr. Boasso In the steam generator itself.
The comment is made relative to the secondary systen
('
I 1
i .
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 A
537
() 1 itself. It has got an argon cover gas and expansion 2 tank.
3 Mr. Zudans: Okay. Now I understand.
(]}
4 Mr. Mark Your inerted cell has perhaps two 5 percent oxygen or something like that?
6 Mr. Boasso Yes, sir, two percent or less.
7 Two per ent would be the upper limit. The typical 8 operating range is a half a percent to one.
9 Mr. Marks And if you didn't need to burn that 10 oxygen, that one or two percent, you really don 't need ;
11 the insulation. The insulation is there only in the 12 event that you get fresh air into that cell, is that not 13 right?
O 14 Mr. Bosssos No, sir, that is not the 15 situation. The loadings that result, it is strictly 16 from the hot sodium going into the cell which requires 17 us to protect our structural concrete.
18 Mr. Marks That is from the sodium ---
19 Mr. Boasso: Just a classical heat transfer 20 from the lined call to the structural concrete.
21 Mr. Marks Okay. It has nothing to do then 22 with the fire, because that fire must be a minor thing 23 at two percent.
24 Mr. Boassos Yes, sir, extremely minor.
25 Mr. Ebersoles Isn't all of what you say on O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
538
() I this slide here pertinent just to the sodium system of 2 the primary system, that when it goes out into the 3 intermediste loops and gets into the steam generators 4 out there, all that doesn't apply, does it?
5 Mr. Boassos This is a combination of both in 6 containment as well as outside of containment. In 7 containment we have lined cells containing our sodium 8 systems which are radioactive. We also have lined cells 9 in our reactor service building because we also have to some very mildly radioactive sodium there which comes 11 from the vessel storage tank that contains potentially 12 extended fuel.
13 In the air-filled cells we have an arrangement O 14 of ca tch pans with fire suppression decks as well as a 15 few cells in the reactor service building to accommodate 16 postulated firest in an air environment. So I am 17 talking across the nuclear island complex.
18 Mr. Ebersole Well, in this requirement that 19 the LBRs hsve of looking at pipe failures outside of 20 containment, do you have any potential water / sodium 21 interactive effects as a result of pipe failures outside 22 of containment? The main problem is to keep fire and 23 sodium apart outside in the aux building.
24 Mr. Bossso: With a postulated accident in the 25 steam genera tor building, and I would like to use that
(
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
539 1
1
() 1 as a point of discussion since the water system is in 2 the vicinity of the sodium system, the resulting 3 conditions are such that the water system can survive 4 that accident. We will not have a combination sodium 5 fire accident resulting and a water accident 6 simultaneously giving you a sodium / water reaction.
7 Er. Ebersoles What about if you have a water 8 sccident which impacts on the hot sodium piping?
9 Mr. Boasso: We have shields in the steam 10 generator building to preclude potentially adverse 11 effects from a postulated steamline rupture impinging on 12 our sodium piping.
13 Mr. Ebersole: And the same for feedwater I O 14 guess, right?
15 Mr. Bossso: I believe that is the case.
16 Er. Ebersoles So the thrust of this is you 17 are trying to keep water from impacting on this sodium 18 piping. It is mandatory that you do that, right?
19 Mr. Boisso: We protect our sodium piping 20 system from potential jet impingement effects from 21 steamline ruptures or high pressure line ruptures.
22 Mr. Dickson: I might add, except where we 23 can't avoid it which is around the steam generator n
(_) .
24 building, we don't let water systems and sodium systems 25 in the same cells.
ALDERSON REPORTING COMPANY,INC, 400 VIRQlNIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
}
540
() 1 Mr. Boasso That is correct.
2 Mr. Ebersole That is where you just can't
(} 3 4
escape where you go in the cells.
Mr. Dicksons That is right.
5 Mr. Ebersoles Thank you.
6 Mr. Lipinskit Going back to your line two 7 where you say you have insulation behind the cell 8 liners, is the concrete cooled?
9 Mr. Boasso Right here?
10 Mr. Lipinskis Yes, because in the steady 11 state that concrete tempera ture eventually has to rise 12 unless you are removing heat from it.
13 Mr. Boasso: The concrete temperature will
() 14 rise. It is the ultimate heat sink and later on in my 15 presentation I will give you some typical temperature 16 rises as a result of a postulated accident in an inerted 17 cell.
18 Mr. Lipinskia But in the steady state you 19 said that insulator was split there because of the hot l
20 pipes in the room.
21 Mr. B3assos I am sorry. Say that again, 22 please.
23 Mr. Lipinski In reponse to an earlier 24 question you said that the insulation behind that cell 25 liners was there because of hot pipes in the room.
O A'.DERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
541
() 1 Mr. Boasso I as sorry. I did not mean to q
2 imply that. It is there to protect the structural 3 concrete from a posulated sodium spill accident. It is 4 not there to protect the structural concrete from the 5 thermal loading in the cell with the plant under normal 6 operating conditions.
7 Mr. Lipinskia Your earlier response misled me.
8 Mr. Boassos I am sorry if I misled you. That 9 is why the insulating concrete is there.
1 10 Mr. Ebersoles Well, when you throw hot sodium 11 into a metallic trsnk next to a concrete structure is 12 that tank anchored to the concrete?
13 Mr. Boasso. Our cell liner system is anchored 0 14 to the concrete.
15 Mr. Ebersole Now there will be an 16 extraordinary temperature differential between the 17 concrete ingredient and the liner which tends to pull 18 all the attachments loose in the liner since concrete is 19 not going to go and it is going to take it a long time 20 to heat up. What do you do?
21 Mr. Boassoa I cannot speak for the 22 architectural engineer, Burns and Roe, but our cell 23 liner system has an extensive arrangement of anchors, O
\_/ 24 Nelson studs on 15-inch centers whereby we ensure that 25 it will maintain its integrity in conjunction with the
(}
ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D C. 20024 (202) 554-2345
542
() 1 sccident. I persansliy cannot give you the results of 2 these detailed analyses, but they are done.
3 Mr. Ebersoles You say you are ready. for this 4 setallic tank to take a hot dump even though it is 5 attached to cold concrete?
, 6 Mr. Boassos Yes, sir.
7 Mr. Zudansa Let me just clarify a previous 8 presentation. In a previous presentation I think we 9 vere told there is a monitoring system that monitors the 10 gap betvean the concrete and the liner and that is used 11 in the cooling system as well to see that the concrete 12 is not overhestad. Which of the cells has that kind of 13 arrangement?
O 14 Mr. Boasso: We do not have a monitoring 15 syste's to monitor the gap between the cell liner and 16 insulating concrete. We have a nominal one-quarter inch 17 gap, and let me put this up, if I may, please.
18 Hr. Zudansa I thought there was some purge 19 system that would pump this thing out to detect eventual i
20 leaks. I think in the last meetino I was told that that 21 would be also a cooling medium.
22 Mr. Boasso: No, sir, that is not the 23 situation. The cell liner system, you have the velded 24 steel plate 3/8ths of inch carbon steel. Between the 25 cell liner and the insulating concrete we have a quarter O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
543
() I inch air osp to allow off gases from the insulating 2 conLrete to escape back through a vent system that feeds
(} 3 back into the air filled cells of the reactor 4 containment building or the reactor service building, 5 depending on where the postulated accident occurs.
6 Mr. Zudansa So at least my understanding that 7 this vent system was also an affective cooling system ---
8 Mr. Boassos Let me refer you to Dr. Dickson.
9 He was at the last meeting I believe.
10 Mr. Dickson: Well, I am going to refer you to 11 Burns and Roe who designed the system.
12 (Laughter.)
13 Mr. Burkharts I an Al Burkhart from Burns and O 14 Roe. I believe you are recalling the in-service 15 inspection testing system on the vent system that we 16 have prescribed. That is a periodic test. That is not 17 a normal operating monitoring system.
18 Mr. Zudansa Therefore, a t least to my 19 perception that this was the ultimate barrier against 20 overheatin7 is wrong obviously. Then I have to return 21 back to Walt's question of what will happen to this 22 concrete in the steady state operation? Eventually it 23 will heat up to whatever the temperature of the
- 24 particular cell is. ,
25 Mr. Burkharts Let me see if I understand your O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W, WASHINGTON. D.C. 20024 (202) 554 2345
544 llh 1 question. You are concerned in the steady state normal 2 operating condition what is the heat removal from the 3 cell?
4 Mr. Zudsnsa Supposing there is a fire that it 5 is at the hot leg temperature, 900-and-some degrees.
6 That pipe sits in a duct that is lined. Behind the 7 liner thera is concrete. If I have no heat removal 8 capability eventually that duct and the concrete vill 9 hest to tha steady state temperature of the sodium.
10 Mr. Carbons Well, your concrete is a heat 11 sink so it is not going to get up to the temperature ---
12 Mr. Zudsns: Oh, eventually it will.
13 Mr. Carbons Not if it is conducting heat into O 14 the ground.
15 Mr. Zudansa That is what I want to know.
16 Where does that heat go?
17 Mr. Burkharts I believe the ground. There is 18 an active cooling system in each cell, the recirculating 19 gas cooling system, an inert nitrogen system which takes 20 the major heat load from the pipes in the system.
21 Mr. Zudans Oh, in the interior of the cell 22 and not behind the liner.
23 Mr. Burkhart That is correct.
24 Mr. Zudans: Oh, that is all right. Why 25 didn't you say so.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
545 O 1 <taughter.)
2 Mr. Burkharts The normal operating 3 temperature of the cell is less than 150 degrees.
)
4 Mr. Zudans: That is fine. I have no more 5 p ro ble m s .
6 (Laughter.)
7 Mr. Marks As long as the pump for the 8 nitrogen keep running.
9 (taughter.)
10 Mr. Zudanst Well, they would have to do -
11 something like freeze the solid sodium.
12 Mr. Boassos In the air-filled cells we have a 13 catch pan system. We can have an open catch pan, an O 14 open catch pan with a drain to a catch pan with a fire 15 suppression deck which I will subsequently describe or 16 we can have a catch pan with a fire suppression deck.
17 The opening catch pan system, we have a 3/8th 18 inch all welded carbon steel plate that forms the pan 19 configuration. Underneath the catch pan we have 20 insulation to protect our structural concrete from the 21 thermal loadings from a sodium spill in this catch pan.
22 We have an air gap on the side of the catch pan to allow 23 gases to escape. Above the catch pan we have a steel 24 gratino to allow workmen to access cells without 25 actually walking in the catch pan. There is a lip over O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W WASHINGTON D.C. 20024 (202) 554-2345
546 i
() 1 the catch pan to preclude sodium from accumulating
- 2 behind the catch pan in conjunction with the postulated 3 spill scenario. It is basically a very simple 4 configuration.
5 In addition, some cells have a catch pan drain 6 to a fire suppression deck system. I just want to point 7 out that the catch pan, as I previously described, 2 8 instead of containing the sodium spill volume will allow 9 it to drain down to another system that has a fire 10 suppression deck which I will now describe.
11 The fire suppression deck system extends above 12 the catch pan. It is configured basically of a "0" deck 13 configuration, and by "0" deck it is a hat type
- .O 14 configuration, which has drain lines in the hat area, in 15 the troughs of the hats, between hats, to allow sodium
- 16 to drain down into the catch pan. ,
17 As the sodium pool builds up above the bottom 18 of these drain pipes, the reaction products accumulate 19 in thc drain pipes and eventuall'y will preclude oxygen 20 from going from the air environment in the cell in which 21 the postulated accident is' occurring to the sodium in 22 the catch pans.
l 23 So actually the fire suppression deck will a
( 24 snuff off the fire as a function of time and it might 25 take as long as 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> in order for the reaction O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
547
!() 1 products to accussulate in the drain pipes and preclude 2 interface of oxygen from above the "Q" deck down below 3 the "0" deck to the sodium pool.
4 We have performed testing to derive the 5 features of the fire suppression deck, the size of the 6 drains, the length of the drains, the various "0" deck 7 configuration: and are in the process right now of doing 8 a large-scale test to verify that these particular 9 features Osn perform to their designed function for a 10 large sodium spill.
11 Mr. Zudansa Have you done a small scale test 12 to show that this things works okay?
13 Mr. Boassoa Yes, sir, we have. We have run O 14 small-scale tests where we have the same basic 15 configuration that is approximatey six feet by two 16 feet. We have taken this fire suppression deck and put 17 it into a test tank that is approximately 10 feet in 18 dismeter and 20 feet high and have spilled sodium up on I 19 the order of 500 pounds onto the "0" deck to determine 3
20 the suppression characteristics of this deck.
21 Mr. Zudsnsa For example, if you vent, if hot 2
22 gases develop and you vent to equalize the pressure on 23 the top and bottom cf your fire suppression deck, why
) 24 wouldn't some oxygen come down the vents? ,
25 Mr. Boassos In addition to this particular i
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (200 554 2345
548
() I system that has been tested in the past, the large-scale 2 test, it will be investigating a hat type configuration 3 on top of the vents that operates very similar to the 4 same thing that you see on exhaust f rom a diesel 5 engine. It would allow the overpressure to escape from 6 under the fire suppression desk and then it drops back 7 down again as the pressure decreases.
8 Mr. Zudansa I see. So that would physically 9 prevent it.
10 Mr. Boassos And we have that type of feature 11 in this large-scale test that we plan at the latter part 12 of this year.
13 Hr. Ha rk s Have arrangements of the sort you O 14 just had up there a minute ago been used in other places 15 or in other contexts, and I mean in the chemical 16 industry or in France, or things like that?
17 Er. Boasso: The French have a fire 18 suppression deck system which is quite different in 19 design from ours. Ours is a very passive system. The
'20 French have a configuration where they have a flapper 21 valve type of arrangement where when the sodium spills 22 onto the top of the catch pan the valve system opens to 23 allow the sodium to drain into the catch pan and then
() 24 when the driving head is reduced then the flapper vsive 25 opens back up again. It is a trough strangement with O
ALDERSON REPORTING COMPANY,INC, 400 VIRGIMA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554 2345
549
) 1 this type of configuration.
2 I do not have a viewgraph to show you.
3 Mr. Marks No, that doesn't matter.
4 Mr. Boassoa But they do have a fire 5 suppression deck system as well as the Germans.
6 Mr. Marka They are after the same thing, 7 namely, to keep a strict limit on the amount of air that 8 can access this system.
9 Mr. Boasso Yes, sir, that is the overall 10 sporoach to minimize the transfer of oxygen from above 11 the pool down to the pool.
12 Mr. Zudansa I would like to return once more 13 to that cell that we referred to that I didn't know how
,O
- 14 it was being cooled.
15 Mr. Bosssos The inerted cell.
16 Mr. Zudansa Supposing now you spill the 17 sodium in it, hot sodium and your cooling by nitrogen 18 that was ref erred to will be no longer available. The 19 question is how long can you tolerste liquid hot sodium 20 in there before you begin to make it difficult for the 21 concrete behind the liner? Have you don't such an 22 analysis?
23 Mr. Boasso: We design our cell structures to G
k) 24 accommodate the worst case spill volume that can occur 25 in that cell. We take the plant at the maximum normal ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
550 1 l
() 1 operating condition and we postulate a leak in our 2 system. We look at the total spill volume that can 3 result from that posulated leak, we allow it to go into 4 the cell and then we evaluate the structural response of 5 the cell to that particular spill. From that point all 6 the way out until it has cooled down and it has actually 7 frozen.
8 Mr. Zudansa I"see, and how much time does it 9 take before temperatures, say, drop below 200 degrees?
10 Mr. Boassos I will be showing you a viewgraph 11 later on that will describe tha t.
I 12 Mr. Carbons You have made no mention of the 13 use of powders or anythink like that. Do they not play
)
14 a role?
15 Mr. Boasso In Clinch River the only type of
( 16 extinguishers, or extinguishments may be the appropriate 17 terminology, are portable NaK extinguishers that are 18 used for small type sodium fires. The approach that we 19 are utilizing in our plant design in the cir-filled 20 cells to mitigate large sodium fires, potential large 21 sodium firas is the fire suppression deck system that I 22 just described. We are not using the graphite powders 23 or the Marcalina that the French are using in their
() 24 particular plant design.
25 Mr. Carbon: Do they make a lot more reliance
)
l ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
551 2 () 1 on extinguishments, of powder type things, or are they 2 use them basically the same way we do?
3 Mr. Boasso: I would like to refer this
(])
4 question to Mr. Hilliard because he is more familiar 5 with the French in this particular area.
6 Correct me, Bob, if I am wrong, but I believe 7 they do have some active powder extinguishing systems in 8 the Phoenix plant.
9 Mr. Hilliard Yes, they do rely more on 10 powders. Both the Germans and French rely a little bit 11 more than we do on powders, but they also have similar 12 catch pan arrangements as Cliff described as a primary 13 method.
O 14 Mr. Carbona Do they rely significantly more 15 on powders or just a little? What I lead up to is if it 16 is a big dif f erence wha t is their ra tionale?
17 Hr. Hilliarda In my opinion, it is just a 18 little difference. They have a hose' type system for
! 19 applying it in some cases. For their big engineered 20 spills like Cliff is talking about they don't rely on 21 powders.
22 Mr. Marka What is the function of the powder?
23 M r. Hilliard: The function of powder is to s 24 exclude oxygen. The powders that have been generated 25 have additives to it that causes the flux to l
ALDERSON REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
552
) 1 temperatures that are involved so that it actually seals 2 off the surf ace of the sodium.
3 Mr. Marks It is purely an inert mechanical
}
4 function and not a chemical function.
5 Er. Hilliard The Nax powder that Cliff 6 mentioned that you called Nax. It is N-a-x. The sodium 7 carbonate Lase was met in nylons and some other things 8 to make it flowable and it actually seals.
4 9 Mr. Boasso: Now I would like to go into an 10 evaluation of design basis sodium fires in Clinch River.
11 First, I would like to look at design basis 12 spills in the interted cells. We postulate liquid metal 13 spill events in all of our inerted cells. We have a O
14 total of 19 in the reactor containment building, 16 in 15 the reactor service building and then there is one cell 16 in the steam generator building. The spill conditions 17 that we postulate, the temperatures, flow rate and 18 volume are selected to provide the maximum challenge to 19 our cell structures.
20 For these particular conditions we do spray 21 pool burning evaluations. Tha subsequent pressures and 22 temperatures and aerosol conditions are then integrated 23 back into the respective building design.
[]-
S- 24 Briefly, the range of spill conditions in our 25 inerted cells can be temperature-vise as high as 995 ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
553 e
- 1 degrees in the primary heat transport system cell or 400 2 degrees Fahrenheit in one of the ex-vessel processing
T 3 cells in the res= tor service building.
J 4 Spill rates can range from as high as 950 5 gallons per minute in a primary heat transport system 6 cell to as low as six gallons per minute in a 7 characterization cell in the reactor containment 8 building.
9 Spill volumes range from as high as 35,000 10 gallons in the primary heat transport system cell to as 11 low as 800 gallons over in the reactor service building, 12 one of the auxiliary systems cells in the reactor 7s 13 service building.
%) 14 This gives you a brief overview of the range 15 of conditions.
16 Mr. Carbon : What a:cident conditions would 17 correspond to that 250,000 pounds of sodium, a 18 guillotine break of a primary pipe or something?
19 Mr. Boasso: This spill volume results from a 20 postulated MEFS leak in our primary piping system. It 21 is a moderate energy fluid system break. It is 22 non-mechanistic in nature and it is that type of 23 non-mechanistic break as defined in the staff's branch q
x/ 24 technical position MEB-3-1.
25 Mr. Carbon: How close does that come to q
%.)
ALDERSON REPORTING COMPANY. INC.
400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
554
() 1 assuming a guillotine break?
2 Mr. Bossso The flow rate from this i
3 particular break is 950 GPH and the flow from a 4 guillotine rupture could be as high as full flow of 5 33,000 GPH.
6 Mr. Carbona I have still one more question 7 there. It is my impression that the French assumed much l 8 larger breaks. Can you comment on that?
I 9 Mr. Boassos I cannot comment as to their 10 rationale for the type break they use in their system.
11 They do use a larger break in their secondary system.
12 Mr. Carbona In their secondary system?
13 Er. Boassos Yes, in their secondary system.
O 14 Mr. Carbona Apprecisbly larger?
f 15 Mr. Boassos It is a double-ended guillotine 16 rupture, to my understanding that they utilize in their 17 secondary system.
18 We have done margin snalyses in our primary 19 cells to find out the structural margin that is 20 available even for a break as large as a double-ended 21 quillotine rupture.
l 22 As Mr. Clare indicated in the May 24th meeting 1
23 with this subcommittee, the conditions were acceptable.
24 Hr. Carbon I have forgotten that.
l 25 Mr. Boassos I believe that George went l
l l
l ALDERSON REPORTING COMPANY,INC, l
! 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
1 l
l l
555 l
() 1 through the various pressures that we see in our cells 2 from those kind of conditions.
{ 3 Mr. Zudans: I understood from previous
' 4 presentations that it didn't really rstter how much is 5 spilled because of the limited amoun t of oxygen 6 available in the cell and that was the reason for 7 assuming a smaller break and saying it makas no 8 dif ference if it is a larger break. Is this a correct 9 understanding?
10 Mr . Boasso s Let me refer that back to Mr.
11 Cla re, if I may, please.
12 Mr. Clares I don 't think I said that was the 13 rationale of why we picked the lower leak rate for the O 14 particular effects that we were talking about.
15 Mr. Zodans: That is how I read it.
16 Mr. Clare We talked about the contribution 17 of various considerations to the pressurization of the 18 cell, for example. I indicted that it indeed didn't 19 make a lot of difference relative to our understanding i
20 of the design capability of those cells what size you 21 postulate.
22 There is a rationale for why we choose th e 23 size we have chosen. We don't have our experts here 24 today to want to do that. It has to do, for example, 25 with fracture mechanics of the piping, et cetera. We O
V ALDERSON REPORTING COMPANY. IMO.
400 VIRGINtA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
556
() 1 could review that some day.
2 Mr. Zudansa That is a completely different 3 set of reasoning. But if you are able to state that the 4 larger break in the pipe would not challenge the 5 structure of breaking more, then of course the 6 assumption of a smaller break doesn't have any logic to 7 it. Now if you say the larger break cannot develop 8 because of the materials characteristics, because of the 9 design, ba:ause of the wall thickness or because of the 10 flexibility, that is another issue.
11 If that is the basis for selecting the smaller 12 break, it is really not necessary to go through that 13 route if you can tolerate larger still. 'So there has to O 14 be something that we don't know about that kept you from 15 assume a double-ended guillotine break.
16 We discussed it last time and I don't think I ,
17 walked away completed convinced of these arguments, but 18 you must have some good reason why you did not want to 19 take 33,000 gallons per minute and why you limit it to 20 950, although you say that the structures are not 21 challenged by t5at.
22 Mr. Clare: That is right, and our discussions 23 on that have been docketed and were reviewed by the 24 staff I think in our earlier licensing interactions and 25 also to a certain extent this time. The considerations O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345 l
557
() I thst I think you are referrin7 to sre those related to 2 the core.
3 Specifically I believe Dr. Kastenberg brought 4 that up in the last meeting and I indicated that, yes, 5 especially with respect to the cold leg there can't be a 6 challenge to the core somewhat beyond our current design
! 7 basis were we to change the design basis from this l
8 smaller leak to a larger leak.
9 Mr. Zudans: Okay. Now is the difference in 10 design or concept between CRBR and say Superphenix or 11 Phenix so great that they do not encounter the same 12 problem rela tive to core repair of some guillotine break 13 as you do?
14 Mr. Clares That is a very difficult question 15 to answer. Superphenix is a pool type reactor.
16 Mr. Zudans: I know. That is exactly it.
17 Mr. Clares I can't comment on what effects on 18 what leaks or. what proportions the internals of their 19 pool might have on the core.
20 Mr. Zudans: So what made you propose a 21 smaller brask flow hss nothing to do with the structures 22 where your cells are housed but mainly again because of 23 the limited amount of oxygen that you have in those
() 24 cells. I think that is the only reason for it. What 25 happens in the core is what limits you. I think I O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W., WASHINGTON. D.C. 20024 (202) 554-2345
I 558
(
() 1 recall a little bit of that, but certainly I can't say I 2 fully understood and I don't think I understand it now.
i 3 I don't know whether it is very important or not and it 4 doesn't really matter whether anybody accepts your 5 proposed method or not. I think we have to come to an 6 understanding. l 7 Mr. Carbona Do you have any further comment 8 to that?
9 Er. Clare Well, certainly if the committee 10 would like to go over the information in that area, I 11 presume that could be scheduled at some point.
12 Mr. Mark I think that there is a point here, 13 that the biggest possible break is not the one we are 14 talking of and tha rationale for why should be 15 satisfactory to talk of 1,000 gallons per minute. It 16 either needs to be elucidated or, as Zudans was just 17 saying, the fact that you could stand a bigger one is 18 clearly indicated.
19 Mr. Zudansa There could be a physical I
- 20 limitation that you must comply with and that is the 21 reason.
22 Mr. Marks It order to make it just 23 straightforward and simply, 950 is a fine number, but if i
() 24 you could have 30 times 950, then one wonders why didn't 25 they use 30 times 950.
['\
V ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
559
() 1 Mr. Carbon: We will come back 'to this later.
2 Mr. Marks It is just a matter of educational 3 simplicity.
4 Mr. Carbona Fine. Let's go on.
5 Mr. Boassos For the postulated pipe break 6 condition we analyzed both a spray fire and a long-term 7 pool thermal consequence scenario. The spray fire 8 conservatively assumes that 100 percent of the discharge 9 sodium or NaK is converted to droplets. These droplets 10 subsequently react with the available oxygen in the 11 cell, and in the case of the inerted cells we have a 12 very low oxygen concentration, less then two percent. ,
13 Mr. Carbona This refers to the spill on the O 14 previous slide; is that correct?
15 Mr. Bosssoa Yes, cir.
18 With the low oxygen concentration combustion j
17 is negligible compared to the transfer of the sensible 18 heat to the gas atmosphers. Heat is basically 19 transmitted by classical convection, conduction and 20 radiation from the burning droplets to the gas 21 environment, in this particular case being the nitrogen, 22 and then subsequently from that particular environment 23 to the cell structures and then subsequently to the 1
( 24 structural concrete. ,
25 Mr. Marks When you say that the chemical heat O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
560 l
() 1 is negligible compared to the sensible physical heat, 2 you are referring I presume to the 995 degrees sodium 3 temperature?
4 Mr. Bossso I am referring to the amount of 5 burning tha t one would get in an inerted cell with an 6 oxygen concentration on the order of two percent.
7 Mr. Marks But you are referring to the sodium
, 8 cosing in st some temperature or other, close to 1,000-I 9 degrees Fahrenheit and not the 400 degree Fahrenheit 10 which was on one of your earlier slides.
11 Mr. Boassos In either case if the sodium were 12 at 1,000 degrees Fahrenheit or at 400 degrees 13 Fahrenheit, the combustion would still be negligible O 14 with respect to the low oxygen concentration. I don't 15 think I understand your question.
16 Mr. Marks Well, I am not questioning what you l
17 say. It is just that it isn't clear from the slide what 18 you are saying, whether this statement is true only for 19 hot sodium or ---
20 Mr. Boassos Oh, I beg your pardon. Here I 21 was using the example of 1,000 degrees. I beg your 22 pardon. It is for either one.
, 23 Mr. Marks Thank you.
() 24 Mr. Boasso As Mr. Hilliard indicated, pool 25 fires occur when 11guld sodium is greater than 400
!O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
561
() 1 degrees Fahrenheit and when we have an accumulation of a 2 pool of sodium on a cell #1oor. The burning occurs at 3 the pool surface. Again we have heat transfer via 4 convection, conduction and radiation from the pool 5 surface to the gas atmosphere and from the gas atmoshere 6 to the cell structures, and we also have heat transfer 7 via conduction from the sodium poor to cell structures.
l 8 Lov oxygen concentrations, whether the sodium 9 be at 1,000 degrees Fahrenheit or 400 degrees 10 Fahrenheit, the combustion is negligible compared to the -
11 transfer of the sensible heat to the gas atmosphere.
12 I will show in the next slide the thermal 13 consequences. During the spray phase the peak gas 14 pressure and temperature occur shortly within ten 15 minutes after initiation of the sodium discharge, 16 insignificant burning in the inerted cells due to the 17 low oxygen concentration and typically the spray phase l 18 concludes at the end of sodium discharge which is less 19 than ten hours. It will conclude at the end of sodium 20 discharge which typically occurs within less than ten 21 hours2.430556e-4 days <br />0.00583 hours <br />3.472222e-5 weeks <br />7.9905e-6 months <br />.
22 Mr. Mark. Doesn't the oxygen deficient have a 23 separate and possibly different limit? I mean if you
'() 24 have got two percent oxygen, then you can't burn very 25 long because there is no oxygen left.
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
562 1
l 1
i 1 Mr. Boasso: In that context the burning is 2 very miniss1. It is basically just classical hea t 3 transfer from the sodium droplets to the gas 4 environment. It is really not a burning situation in 5 the inettad cells. We have very little burning.
6 Whereas in the oxygen rich air-filled cells it will be a 7 different situation.
8 Mr. Marks Well, ten times different.
9 Mr. Boassos Examples of temperatures during 10 the spray phase for a postulated leak in the primary 11 heat transport system cell. Very quickly, the 12 temperature goes up to somethink or the order of 700 13 degrees and then the temperature is lowered as a result Os 14 of the reactor getting a plant protection scram on a lov 15 sodium level of approximately four and a half minutes 18 into the scenario. Then the temperature will continue 17 to rise very gradually with the plant being on pony 18 motor flow until the sodium in that particular loop is 19 exhausted.
, 20 Hr. Marks This is a curve or an inerted cell?
21 Nr. Bossso An inerted primary heat transport 22 system cell, the cell in which we postulate 35,000 23 gallons of sodium being spilled, yes, sir.
O 24 nr. zudans, what are the corresponding liner 25 and the concrete temperatures? Do you have a slide for O
J ALDERSON REPORTING COMPANY,INC, 400 VIRGLNIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
563
(} 1 2
that?
Mr. Boassoa Yes, sir, I do. As you would 3 expect, is this the corresponding pressure profile.
O 4 Pesk pressure is just below 15 psig, coming down and ]
5 following the hydraulics of the situation and then 6 coming down below 10 psig and then terminating at 7 something lika on 10 psig at the end of the scenario s 8 which is like 500 minutes.
9 The long-tern pool thermal consequences in 10 inerted cells. We perform the evaluations for extended 11 time frames to determine concrete response 12 characteristics on the order of from ten's to hundred's 13 of hours as the situation may dictate.
( 14 The spray end-point thermal conditions serve 15 as the initial boundary conditions for the pool 16 evaluation. Essentially no burning occurs because the 17 oxygen has been consumed during the spray phase and we 18 have a classical heat transfer problem. The cell gas 19 pressure and temperature conditions decrease.
20 The corresonding concrete conditions are as 21 follows. As time progresses into the scenario the cell 22 liner temperature decreases with time. For the primary 23 hest transport system cell postulated accident it is
() 24 down almost just above 200 degrees at about 70 hours8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br /> ,
25 into the s:enario.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
A T
564 lll 1 As we dump heat into our structural concrete 2 the structural concrete comes up in tempera ture just 3 below 200 degrees and then flattens out as a function of 4 tine. A t this particular temperature this concrete node 5 is one and A half inches into structural concrete.
6 Mr. Marks That temperature is drawn assuming 7 the benefit of the perlite insulating concate.
7 8 Er. Boassoa Yes, sir. The perlite insulating 9 concrete is significantly higher than our structural 10 concrete.
11 Mr. Zudansa And it takes ten hours to 12 actually ---
13 'Mr. Boassoa Something on the order of about 14 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> to reach its peak temperature of about 200 15 degrees. -
16 Mr. Zudans: I assume since this was a 17 calculation you have some details on the time f rom zero 18 to ten hours both on the perlite concrete and the 19 concrete and the liner. Is that a correct assumption?
20 Mr. Boassos Yes, that information is 21 available.
22 Mr. Zudsns: It is sn inch and a half inside 23 but on the surface it will be higher.
lll 24 Mr. Boasso: The surface of the concrete would 25 still be way down in this region. The perlite concrete O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE.,'S.W., WASHINGTON, D.C. 20024 (202) 554-2345
565 l would be higher but not the structural concrete. I
(])
1 2 don 't have that information available.
3 Mr. Zudans: And then somebody would have 4 analyzed what it means structurally?
5 Mr. Boasso: The complete scenario is analyzed 6 by Burns and Roe.
7 Mr. Carbon: Does the structural concrete 8 start out at about 100 degrees or something like that?
9 Mr. Baasso Something on that order. So you to can almost take this curve and extrapolate up in this 11 time. It is going to follow that same slope.
12 Mr. Zudans: This includes the cooling of that 13 cell?
) 14 Mr. Boassos We postulate the accident with 15 the plant at normal operating conditions with the cell 16 inert gas cooling system functioning in its normal mode 17 of opera tion.
18 Mr. Zudansa You I assume also have looked at 19 what would happan if you lost that cooling by a nitrogen 20 atmosphere?
21 Mr . Bsasso : The effect would be minimal as 22 far as challenging our structural concrete because the 23 heat load into that concrete would be very low.
() 24 Aerosol generation in inerted cells is a 25 result of the low oxygen concentration. We have very O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
l l
566 I i
l l
() 1 small amounts if serosols being developed. No more than 2 600 pounds of aerosols are developed and this particular 3 limit occurs in the primary heat transport system cell.
i 4 The aerosols are contained within the inerted 5 cell boundary and we hypothesize a release of all of the 6 radioactive aerosols resulting from the postulated 7 sodium fire event either to the RCB or the RSB 8 atmosphere and this particular assumption results in 9 doses well within 10 CFR 100 guidelines.
10 Next I would like to discuss design basis 11 spill events and sir-filled cells. Briefly we have six 12 air-filled cells in the reactor service building and we 13 have a total of 17 cells in the steam generator i building. A range of postulated spill conditions, 15 temperatures are from 940 degrees in the intermediate 16 heat transport system to 400 degrees in the ex-vessel 17 processing system, spill rates from 1,000 GPH in the 18 intermediate heat transport system to 45 GPF in the 19 ex-vessel processing system, and spill volumes from 20 39,000 gallons in the intermediate heat tra nsport syt:?.em 21 to as small as 800 gallons in the ex-vessel processing 22 system.
23 Again the approach that we utilize. We, first
() 24 of all, evaluate a spray fire. As I indicated before, 25 we assumed that 100 percent of the discharge of liquid O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
567 I
() 1 metal is converted to droplets. These droplets in turn 2 react with the available oxygen in the cell and now we 3 have a high oxygen concentration which will result in 4 significant combustion and in significant transfer of 5 heat to the gas atmosphere. Heat is transferred by 6 convection, conduction and rsdistion from the burning 7 droplets to the gas environment and from the gas 8 environment to the cell structures.
9 Mr. Carbons These cells by definition are gas 10 tight? It is simply that they have air in them instead 11 of nitrogen; is that the case?
12 Mr. Boassos They are air-filled cells.
13 Hr. Carbons And they are tight such that no l
Ci 14 more gas leaks in?
15 Mr. Boassos No, they are not tight. Gas can 16 leak in. It is possible for gas to leak back into the 17 cell.
18 Mr. Carbona I presume that in the f
19 calculations that is allowed for?
20 Mr. Boassos That is accounted for, yes, sir.
21 I would like to emphasize that during the 22 spray phase the building would be slightly pressurized 23 precluding the ingress or oxygen back into the cell, but
() 24 upon concluding the spray phase, when you go through 25 your long-term cool down mode, then you could possibly O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
568 I have some ingress of oxygen from the external
(])
2 environment back into the building and that is accounted 3 for in the analysis.
4 Again, pool fires occur when the liquid sodium 5 is greater than 400 degrees Fahrenheit. In the 6 air-filled cells of the steam generator building and 7 reactor service building we can have the liquid sodium 8 accumulating either in catch pans or on the surface of 9 the fire suppression deck. Consequently, we have 10 burning at the pool surface and on the surface of the 11 fire suppression deck.
12 We have the same classical heat transfer, 13 convection, conduction and radiation from the pool
} 14 surface to the gas atmosphere, and from the gas 15 atmosphere the to the call structures. In addition, we 16 have heat transfer via conduction from the sodium pool 17 to the cell structures.
18 As a result of the high oxygen concentration, 19 combustion at the pool surface is significant in the l
20 transfer of heat to the gas atmosphere and is included 21 in the analysis. As I mentioned earlier, pool burning 22 is suppressed by the fire suppression deck within 36 23 hours2.662037e-4 days <br />0.00639 hours <br />3.80291e-5 weeks <br />8.7515e-6 months <br />.
() 24 A brief summary of spray fire thermal 25 consequences. The peak gas pressure and temperature O
v ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
I 569
() 1 occur shortly within two minutes of initiation of the 2 sodium discharge. We have significant burning with the 3 generation of dense aerosols due to the high oxygen O 4 concentration. The spray concludes at the end of the 5 solium dis:harge which is on the order of less than six 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> for postulated breaks in air-filled cells.
7 The pool fire consequences, they extend over 8 extended time periods, ten's of hours with burning under I
i 9 the fire suppression deck suppressed within 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br />.
to The spray burning effects are integegrated into an
~
11 overall spray pool fire analysis. Significant burning 12 will occur. Approximately 4,000 gallons of the 13 discharged sodium burns in the steam generator building 14 and this is a combination of burning from both the spray 15 and pool phases of the fire. And during the long-term 16 pool thermal evaluation the cell pressure and 17 tempera ture conditions decrease.
18 In the steam generator building for a 19 postulated IHTS pipe leak the peak temperature is on the 20 orier of 540 degrees Fahrenheit. It occurs about 200 21 seconds into the scenario and then decreases with time 22 to something on the order of 250 degrees Fahrenheit at 23 10,000 seconds.
() 24 Mr. Carbons What is it physically that causes 25 the temperature to decrease in about three minutes?
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554-2345
570
() 1 Er. Boassos I am sorry?
2 Nr. Carbons What is it physically that causes 3 the temperature to decrease in about three minutes?
4 Mr. Boassos The oxygen is being consumed in 5 the building a s a f unction of time. In addition, as I 6 will get into in another viewgraph, we vent gases from 7 the building to the external environment which also 8 contributes to this decrease in temperature. We are 9 allowing the hot gases to escape from the building 10 through a vent release system.
11 The peak floor temperature under the catch pan 12 fire suppression deck system in the steam generator 13 building for the postulated IHTS pipe leak is something
} 14 on the order of 275 degrees about six hours into the 15 scenario and then runs out to about 200 degrees 16 Fahrenheit at 50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br /> into the scenario. This is two 17 feet into the structural concrete.
18 Mr. Zudansa let's see, you said, if I 19 understood correctly, that the spray phase fire 20 concludes at the end of sodium discharge in six hours.
21 Mr. Boassos Yes.
22 Mr. Zudans And the previous slide you showed 23 that within 200 seconds the temperature begins to drop.
() 24 Where am I missing the connection.
25 Er. Boasso I think when I get into my next O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (7's2) 554 2345
571
() 1 viewgraph I will show you that in conjunction with this 2 accident we are venting gases from the building 3 immediately.
4 Hr. Zudans: But you still are burning in the 5 same atmosphere at the same rate.
6 Mr. Boassos That is true.
7 Mr. Zudans: So if it went up in the beginning 8 there is no reason for it not to stay there at that 9 combustion tempera ture.
10 Mr. Clare: I might be able to help interpret 11 the question or clarif y a suggested answer. The early 12 slide with the gas temperature which of course is very 13 auch controlled by the spray burning, this curve is the 14 concrete temperature below the deck.
15 Mr. Zudanst No, no, this is the cell gas 16 temperature.
- 17 Er. Clare Right, that is the cell gas 18 temperature where the spray fire is taking place. This 19 temperature on this vievgraph ---
20 Mr. Zudansa I am not looking at this one. I 21 an only comparing this one, which at 200 seconds began 22 to drop, with the slide that describes the spray fire 23 saying that the spray phase concludes that sodium
() 24 discharge in six hours which makes me assume that it 25 kept on burning for six hours and therefore I can't see ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554 2345
572 why the cell gas temperature would go down in two
(]) 1 2 minutes.
3 Mr. Clares I believe that has to do with the 4 rate at which the sodium is being discharged and th e 5 conduction of heat into the cell structures. Perhaps 6 Cliff can go back to the slide of the sodium discharge 7 rate and you can see some comparison there.
8 Mr. Zudans: There was no statement that the 9 sodium discharge rate changed so many gallons per minute.
10 Mr. Boassoa Yes, it does change. We have a 11 thousand GPM discharge flow rate for approximately eight 12 minutes into the scenario at which time we go to a 13 situation where the sodium is being discharged as a 14 result of static head only.
15 Mr. Zudans: Well, if you had a thousand 16 gallons per minute for eight minutes then there is no 17 resson for the temperature to begin dropping in two 18 minutes. It maybe not as drastic as six hours, but 19 there is still some discrepancy which I would like to 20 understand. Maybe the label on the other slide should 21 be hours rather than seconds.
22 Mr. Boassos let me go back to that slide. I 23 vant to make sure I understand your question. Is this O 24 the ener 25 Mr. Zudans: Yes, the last slide.
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 m.-
573
() 1 Mr. Boasso: The spray phase concludes at l 2 roughly six hours, but during the spray phase the oxygen 3 concentration is being depleted very rapidly.
4 Mr. Zudanst You have the air-filled cell here.
5 Mr. Boasso We have an air-filled cell, but 6 the building is pressurized during the accident and we 7 are consuming oxygen very, very rapidly. As a 8 consequence, even though we have a spray simulation over 9 a significant time frame, we have relatively little 10 burning from that spray phase. As a matter of fact, a 11 few minutes into the scenario we begin to approximate 12 those conditions that we have in the inerted cell. The 13 oxygen is consumed very rapidly.
O 14 Mr. Zudans: Well, I guess that could explain 15 it.
16 Nr. Boassos That coupled with the venting 17 system that I will describe to you.
18 Ihe oxygen is not maintained at 21 percent 19 throughout the scenario. It depletes very rapidly. I 20 don't have a viewgraph to show you that decrease.
21 Mr. Zudans: Whatever you had in the cell at 22 that percentage, after it gets conumed you don't admit 23 new oxygen from the surrounding building.
() 24 Mr. Boassoa Not until you get into the 25 long-term aspects of the scenario a t which point the O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
l 574
() 1 discharge from the sodium system has been concluded and 2 you are looking at the pool burnino under the fire 3 suppression deck. The peak pressures will have already a been achieved. {
5 Br. Carbons What accident are you simulating 6 where you allow the sodium to discharge for six hours?
7 3r. Boasso: We are postulating in the hot leg 8 of the IHTS loop a moderate energy fluid system break, 9 as I discussed previously, in the piping system and it 10 has a cross-sectional area of one-fourth the diameter of 11 the pipe times the thickness of the pipe. We postulate 12 that type of break in the piping system and then as a 13 result of the hydraulics in the system, the pumping head i
14 plus the cover gas pressure that is in. the system, we 15 have a resulting 1,000 GPH flow rate initially.
16 Er. Carbon: And you would expect to continue 17 discharging for six hours?
18 Hr. Boassos No, sir. The 1,000 GPH flow rate 19 continues f or approximately eight and a half minutes, at 20 which time we have a plant protection system scram of 21 the plant. We shut down the complete plant as well as 22 that particular pump in the affected loop and then in 23 the affected loop the resulting flow rates go down to l
l i
() 24 something like on the order of 300 GPH and lower from 25 static head driving considerations only. There is no O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
575
() 1 more active head from the pump or the cover gas system.
2 Mr. Carbon: And you would not have isolation 3 valves or anything like that?
4 Mr. Boasso No, sir, no credit for isolation 5 valves.
6 Mr. Lipinski Could you go back to that 7 previous slide which you took off.
8 Nr. Boasso: With the concrete?
9 Mr. Lipinskia Yes, with the 24-inch depth.
10 Mr. Boasso: Sure.
11 Mr. Lipinskia What is unique about that I 12 24-inch depth that you selected? Could you have taken a 13 selection of six hours and given the temperature as a O 14 function of depth because this shows it is running 15 around 270 for a peak. So what temperature is a 16 function of depth from zero down?
,17 Mr . Boasso: I do not have that information 18 available. I would like to ask Burns and Roe if they I
i 19 can recall the structural concrete limitations as a 20 function of temperature, the point being that even at 21 these temperatures and above this particular depth into 1
i 22 the concrete and at shallower depths we are still within l
23 our code allowables on our structural concrete.
j() 24 Mr. Lipinskia Okay. Well, you are going from l 25 270 up.
I CE)
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
576 1
I Mr. Boasso: For short-term considerations I
({}
1 2 tha t is still acceptable.
3 Mr. Lipinski: What is the limit on the
() 4 concrete temperature?
)
5 Mr. Bossso I am sorry, I can't answer that ;
6 question.
7 Mr. Zudans: It is more like 200 degrees.
8 Mr. Boassos For short-term effects I believe 9 the limitation is higher, something on the order of 350 10 or 300.
11 Mr. Clares The project has performed a number 12 of test programs with concrete at various temperatures 13 and we used those as well as the data in the standard
} 14 codes, et cetera. The full range of concrete 15 temperaturas from the surf ace to the two feet you see 16 here and below are calculated. We just didn't happen to 17 bring those curves. There is a family of curves.
18 Mr. Zedsns: It is kind of obvious that they 19 kind of increase hyperbolically to the surface 20 temperature and therefore you probably will have five or
, 21 six inches of the surface of the concrete pulled away 1
22 and become a dust.
23 Mr. Clares Well, the insulation I believe
() 24 would protect that from occuring, but certainly ---
25 Mr. Zudsnsa You got this temperature 24 l
O ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
577
() 1 inches deep, and the temperature is rising as you go 2 towards the surface. That means the insulation is all 3 hested up to the cell temperature already and it is 4 ineffective because it does not feed heat very fast. It 5 is still effective. But you have a surface layer on the 6 concrete that is probsbly in s very bad shape. I would 7 like to see those numbers.
8 Mr. Clare I believe we have reported those 9 numbers in our PSAR, Chapter 15. Again, we just didn't 10 bring them with us on vievgraphs today.
11 Mr. Zudsns: Well, there is a Burns and Roe 12 structural person here. Why doesn't he tell us.
13 Mr. Clare We have a Burns and Roe licensing 14 person here.
15 -
Mr. Zudans: Who cares.
16 (Laughter.)
17 Mr. Burkharts I am not a structural person.
18 I just want to confirm though that those numbers are 19 calculated and it is my understanding that the i
l 20 allowables are higher than the typical long-term 200 1
21 degrees throughout the structure both under emergency 22 conditions and for local effects. They are quite a bit 23 higher than the number that you are thinking about which
() 24 is 200 degrees Fahrenheit. I am confident they are in 25 Chapter 15 of the PSAR. I just don 't know them off hand.
l ALDERSoN REPORTING COMPANY,INC, 400 VIRGINTA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
578 Mr. Zudans: Well, I would begin to get
(])
1 2 worried even for a short time if it was in the 500 3 degree range or 400, somewheres in that range. From O 4 this data it is suspicious that it might get that high.
5 Why don't you let us know next time what the picture 6 looks like.
7 Mr. Burkharta We can.
8 Mr. Zudans: And also bring with you the 9 assurances that it will not become a dust at the surface.
10 Mr. Burkhart A dust at the surface or become 11 an aerosol or something like that.
12 Mr. Zudansa That it dehydrates and has no 13 structural strength and your liner anchors will just
() 14 come out like if you didn't have any.
15 Mr. Burkharts I understand your concern. I 16 believe it has been addressed. I just don't have the 17 figures.
18 Mr. Kaushals I wouldn't want to leave the 19 impression that concrete turns into dust.
i 20 Mr. Zudans: As a potential.
21 Mr. Kaushals With due respect, there have 22 been tests that have been done with long-range holding 23 of structural concete at temperatures up to 600 degrees
() 24 and above and it does not turn into dust.
25 Mr. Boassoa As you would expect in the i
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345
579 air-filled cells, the aerosol generation is indeed
(]) 1 2 significant. We can have thousands of pounds of 3 aerosols develop in an air-filled cell in conjunction 4 with a postulated spill event.
5 As I alluded to earlier, aerosols are vented l
6 from the steam generator building to maintain peak 7 pressures within acceptable limits. In my next 4 8 viewgraph I will describe that system. This particular 9 system limits aerosol releases to a maximum of 630 10 pounds.
11 For a postulated pipe break in the IHTS piping l 12 system areosols develop, are ingested into the building 13 HVAC system, are subsequently detected by two sets of 14 safety-related smoke detectors which in turn deactivate 15 the building HVAC system, the inlet fire dampers are 16 closed, the exhaust dampers are closed, two relief vents 17 are open which allows the aerosols and pressure to be 18 released from the building through a dedicated vent
. 19 stack.
20 Five minutes into the scenario these 21 particular dampers will be closed as a result of a logic 22 which is actuated by the safety-related smoke detectors 1 23 and then the building will be maintained in a closed
)() 24 condition th roughout the remainder of the scenario.
25 Mr. Lipinski What about the steam vent i
)
ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554-2345
580
() 1 louvers, what is their purpose there at the top?
2 Mr. Boassos The steam vent louvers are 3 incorporated into the building design to preclude 4 potential overpressurization of the building in 5 conjunction with a steamline rupture. The building 6 pressure is three psi, and in conjunction with a 7 postulated steamline rupture these steam vent louvers 8 will open.
9 Mr. Zudsnsa In this aerosol atmosphere that to is not at a high temperature is there a significant 11 amount of aquipment that has to function?
12 Nr. Boasso: We take no credit for any of the 13 equipment functioning in this loop' d uring the accident.
14 These particular detectors perform'their safety function 15 almost, I wouldn't say immediately, but extremely early 16 into the scenario. As soon as the aerosols begin to get 17 into the HVAC exhaust duct they send a signal to close 18 these particular dampers, the inlet to the HVAC system 19 as well as the exhaust, and once these particular 20 dampers are closed they remain closed.
21 Mr. Zudans: They fail closed if they fail and 22 the rest of it is kind of a passive system. What about 23 those closure dampers up there?
() 24 Mr. Boassos These are motor operated systems 25 tha t will have to be qualified for the environment which O
ALDERSON REPORTING COMPANY, INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554 2345
I l
581 l l
4
{} 1 they will be exposed to.
2 Mr. Haskers What happens to the control room 3 ventilating system? Is that closed down at that point, j
}
4 too?
5 Mr. Boasso The control room system is 8 isolated in conjunction with an accident of this nature.
7 Mr. Zudsnsa Just one more thing, if I may.
8 This building is sdjacent to the shield building, one of 9 the wa'11s. Is it correct as it is sketched here?
1 10 Mr. Boassos No, this is an approximation.
11 Mr. Zudans: But looks at_the shield building, 12 right?
1 13 Mr. Boassos This is the reactor containment
() 14 building and then adjacent to the reactor containment 15 building we have the intermediate bay and then that 16 feeds into the high bay where I am postulating this 17 accident.
18 Mr. Zudans: I am talking about the actual 19 containment concrate building. It is really not the 20 shield building. It is the confinement building.
21 Mr. Boassos This is the reactor containment 22 build in g .
23 Mr. Zudansa Is that wall directly exposed to
() 24 the atmosphere?
l 25 Mr. Boassoa This wall here?
i i
l
()
I ALDERSON REPORTING COMPANY. INC, 400 VIRGINIA AVE, S.W., WASHINGTON, D.C. 20024 (202) 554-2345
\
582 Nr. Zudans: Down below and within the room.
({} 1 2 Tha t is righ t.
c 3 Mr. Boasso I believe it is. I would have to f3 V 4 refer that to George Clare. The question is the 5 confinement building as we come into the intermediate 6 bay, does it see the environment in th'e intermediate bay?
7 Mr. Claret Yes.
8 Mr. Boassoa Yes.
9 Mr. Zudans: And this one doesn't have any 10 insulation on the outside? No insulating concrete, no 11 liner, nothing. That is this going to do in that high 12 temperature environment?
13 Mr. Clares Well, for those cells where there 14 are sodium systems for which this containment building 15 forms one wall, those cells will have catch pans and 16 there will be insulation behind the catch pans to 17 pretect the concrete from the thermal effects of the 18 sodium pool.
19 Mr. Zudansa But we are not talking about the 20 cell here. We are talking about the building atmosphere 21 which is at the same temperature as the cell becaus'e you 22 begin to exhaust not gases here and those hot gases will 23 fill this steam generator building in essence until you
() 24 relieve them soseplace and that hot gas will wash your 25 confinement building outside wall which has neither O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON D.C. 20024 (202) 554-2345
583 insulation nor liner I assume.
f]) 1 2 Mr. Clare That is correct.
3 Mr. Zudans What will happen to that O 4 structure? Has that been analyzed?
5 Mr. Clare Yes. A family of curves has been 8 generated describing the temperature of that wall which 7 is one of the walls of the cell. It is treated like 8 every other sei'smic category one wall of all those 9 cells. It is analyzea for those conditions and it will 10 be designed to withstand it.
11 Mr. Zudanss If you already have done that 12 analysis I would like to see the results because it is a 13 hot spot in a cold building. What it will do is it will 14 expand locally. I.t is not going to break down, but it 15 will crack the adjacent sections, potentially crack the 16 adjacent sections and open them up.
17 Mr. Clarea The moments created by the thermal 18 stresses have been analyzed.
19 Mr. Zudans Well, it is more of a local 20 behavior than a beam type behavior that I am concerned 21 with. If I a take a rectangular section on a 22 cylindrical wall and heat it up, it acts like a plug 23 that wants to expand. Nothing happens to the plug l
() 24 itself very likely, but it may crack adjacent sections l 25 and it cannot be analyzed as a beam. It would have to O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W, WASHINGTON, D.C. 20024 (202) 554-2345
584
(} 1 2
be analyzed as a shell and I am sure your people know how to do it. But just look at it, okay?
3 Mr. Clarea We will check it, yes, sir.
O 4 Nr. Boassos A summary of the evaluations.
5 The plant cell structures are designed to accommodate a 6 conservative spectrum of the design basis liquid metal 7 spill events. There is no impact on the capability of 8 the plant to maintain a safe shutdown condition in 9 conjunction with these postulated spill events and there 10 is no impa=t on public health and safety as a result of 11 the postulated design basis liquid metal spill events.
12 The conclusions on liquid metal fires. The 13 phenomenology of liquid metal fires has been extensively
() 14 investigated and we believe is well understood. An 15 appropriate spectrum of postulated spill events have 16 been selected to challenge our plant structures and we 17 have performed conservative evaluations of these events 18 which have in turn resulted in no unacceptable 19 consequences.
20 Mr. Carbons We have discussed in some of our 21 subcommittee meetings the fact that the French are 22 building the Esmeralda facility and they are going to go 23 through experiments with very large spills. What is it
() 24 tha t they feel they don't know? Why do they feel the 25 need to do this when apparently we don't have this O -
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE, S.W, WASHINGTON, D.C. 20024 (202) 554-2345
585 concern as well.
(])
1 2 Mr. Boassos The facility that the French have 3 got, Esmeralda at Cadarache, they will be doing a
)
4 large-scale pool fired test. There was a meeting in 5 Richland, Washington, in May, a meeting of the 6 International Working Group on Fast Reactor Safety at 7 which the French participated. They described the 8 proposed testing at Esmeralda and I was personally very
~
9 interestei ss to why they were doing a large-scale pool 10 fire test.
11 They told me that for their postulated pipe 12 breaks in their plant they do not assume a spray fire.
13 In some magical fashion they go from the break condition
) 14 to a pool fire. As a consequence they are only going to 15 be looking at pool fires at Esmeralda at this time.
16 The larger scale test that they have planned 17 at Esmeralda is 20 tons of sodium. We have a test 18 planned at the latter part of this year in which we will 19 be looking at both spray aspects as well as long-term 20 pool aspects with sodium spills up to 22 tons.
21 Mr. Carbona So then they are not going beyond 22 what we are already doing.
23 Hr. Boassos No, they are not going beyond
() 24 what we are presently doing. My impression is that both 25 the French and the Germans have not devoted sufficient O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W., WASHINGTON. D.C. 20024 (202) 554 2345
586 attention to the effects of spray fires in their plant
(]) 1 2 design. !
3 Mr. Csrbon Do we exchange information with 4 them? Do we have any agreement of swaping your data for 5 data or anything like that?
6 Mr. Boasso: There is an exchange of 7 information. As to the amount that we exchange, I can't 8 say. I don ' t know. But there is a program between this 9 country and the international body in the LMFBR industry to for an exchange of information. But as to the amount 11 tha t is exchanged , Dr. Carbon, I can' t say.
12 Nr. Hoskers One other question. You had 13 mentioned on the doses from aerosols generated in the 14 inerted cells that you got doses well within the 100 15 guidelines in the containment service building. How did 16 rou do the model of those? Do you take credit for the 17 conglomeration and deposition of the particles coming 18 out or do you just stir everything uniformly through the 19 building and calculate it or what?
20 Mr. Boasso: We take all of the aerosols that 21 can possible accumulate in the inerted cell in which the 22 accident has been postulated to occur. We release them 23 to either the RSB or the reactor containment building
() 24 environment and then we take credit for the containment 25 structure or the con tainment isolation system in O
ALDERSON REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W, WASHINGTON, D.C. 20024 (202) 554 2345
587
() 1 performing its particular function and then those
- 2 aerosols that get out of the containment building in 3 conjunction with the isolation time sequence that we 4 must go through, that is utilized in calculating the l 5 doses.
6 I believe for the accident in the primary heat 7 transport system cell we release something on the order 8 of maybe a few kilograms of sodium maximum.
, 9 Hr. Carbon Are there any other questions 10 that we mi2ht have?
11 (No response.)
12 Mr. Carbona I guess not. If I understand the 13 agenda correctly, that has taken us to the end of it.
14 Mr. Boasso Yes, sir.
15 Mr. Carbona I guess we are perhaps at the 16 point of adjournment.
17 I thank you. I think these have been real
. 18 nice presentations. It has been two very worthwhile 19 days and we thank you for your time and effort and
- 20 everything.
21 Mr. Dicksona Thank you for your attention.
- 22 Mr. Carbon Our pleasure.
23 If thera are no other points, I am about the
() 24 j 25 4
i ALDERSoN REPORTING COMPANY,INC, 400 VIRGINIA AVE., S.W WASHINGTON, D.C. 20024 (202) 554-2345
i j 588 4
i i
!O
' aa tae v 1 a c 11 1t auit -
j 2 ( Whereupon, at 3:15 p.m. , the ACRS CRBR/ Site 4
j 3 Evaluation Subcommittee meeting concluded.)
1 j 4 * *
- i i
l 5 i
j 6 i
) 7 i
j 8 i
j 9 i
i j 10
! r 11
{
1 j 12 i
13 lO
! 15 f
- 16
! - 17 1
d 18 19 20 21 22 l
23
, O 24 25 lO ALDERSON REPORTING COMPANY,INC, i 400 VIRGINIA AVE., S.W., WASHINGTON, D.C. 20024 (202) 554-2345 1
- TUCLEAR REGULATORY CO.'s!ISSION O
This is to certify that the attached proceedings before'the O
in the matter Cf: ACRS/ Subcommittee on Clinch River Breeder Reactor /
Site Evaluation
- Date of Proceeding: June 25, 1982 Docket tiumber:
Place of Proceeding: Washington, D. C.
wore held as herein appehrs, and that this is the original transcript thereof for the file of the Ccamission.
- u. v_ nannnn Official Reporter (Typed)
A!!44fL Official Reporter (Signature)
~
O
'O i
1 l
MF"*?ML stiwMu:'Cltf CSF*SICN
. This is Oc cS.Mif'f that th& attached prOctatiing! 'Cefert the O
in the sattar cf:. ACRS/ Subcommittee on Clinch River Breeder Reactor /
Site Evaluation Casa cf Prcceeding: June 25, 1982 Ocekat llusher:
Flace Cf Prcceeding: Washington, D. C.
wars held as herwis appears, and tha't. this is the original 0: anse:-ip thersef fc:- the- file of the Cc=::11ssicc.
Mary C. Simons Official Esper:4r (!7:ed)
%, A rri M Officizi Repertar (514:scu: e)
O O
l
i O o o ,
l CRBRP SITE SUITABILITY l BRIEFING FOR
- ADVISORY COMMITTEE ON l
4 REACTOR SAFEGUARDS (ACRS) .
CRBRP SUBCOMMITTEE STATION BLACKOUT '
l
) PRESENTED BY l GEORGE H. CLARE LICENSING MANAGER l WESTINGHOUSE-OR l CRBRP PROJECT JUNE 25,1982 , .
1
O O O "CRBRP BLACKOUT"
- FEATURES TO PREVENT STATION BLACKOUT
- ACCOMMODATION OF STATION BLACKOUT l
- 6 82-288414
! SEVERAL REDUNDANT AND DhVERSE i
! POWER SOURCES PREVENT STATION BLACKOUT l
OFFSITE POWER
- 4 CONNECTIONS TO THE TVA GRID i y
i
- 3 TRANSFORMERS WITH DIVERSITY
- FLEXIBLE BUSS CONNECTION 1 l
ONSITE POWER l
!
- 3 SEISMIC CATEGORY I DIESEL GENERATORS .
ONE OF DIFFERENT SIZE AND MANUFACTURE j
- 3 GROUPS OF SAFETY RELATED LOADS l -
EACH LOAD GROUP IS ADEQUATE TO ACHIEVE l
AND MAINTAIN SAFE SHUTDOWN >
! STATION BLACKOUT IS NOT A DESIGN BASIS FOR CRBRP 6 82 2885 15
CRBRP NORM AL AC POWER SUPPLY TVA GRID TVA GRID 12 13 14 11
/ /
1 / /
I N
MZ $Z MAIN $Z ZM GEN.
i
! I I I [ [
( I
(
LOAD LOAD LOAD GROUP GROUP GROUP 1 2 3
~~* -
\
o 0 e
- l CRBRP
! EMERGENCY ONSITE AC POWER
- DIESEL GENERATORS rm rm th D D EV l
i
) .) )
l I
i I
SAFETY RELATED SAFETY RELATED SAFETY RELATED LOAD GROUP LOAD GROUP LOAD GROUP
! 1 2 3
~ ~
O O O ,
CRBRP CAN ACCOMMODATE A STATION BLACKOUT USING NATURAL CIRCULATION HTS FLOWS AND THE SGAHRS i
i
- THE RSS WILL AUTOMATICALLY SHUTDOWN THE REACTOR !
I UPON LOSS OF OFFSITE POWER. ;
i
- PHTS, lHTS AND SGS NATURAL CIRCULATION !S ASSURED
- BY DESIGN. NO OPERATOR ACTION IS REQUIRED.
l
- SGAHRS POWER VENT VALVES AUTOMATICALLY PROVIDE j SHORT TERM HEAT SINKS. i
!
- FEEDWATER IS PROVIDED BY AUTOMATIC OPERATION OF
! THE TURBINE DRIVEN AUXILIARY FEEDWATER PUMP l DRAWING FROM THE PROTECTED WATER STORAGE TANK.
l
- PROTECTED AIR COOLED CONDENSERS AUTOMATICALLY PROVIDE LONG TERM HEAT SINKS USING WATER AND AIR
- SIDE NATURAL CIRCULATION. i I
i
{
6 82 2885 16
THE ELEVATION DIFFERENCES IN MAJOR COMPONENTS PROVIDE A NATURAL l
CIRCULATION CAPABILITY FOR CRBRP
- PROTECTED
+
W
- AIR COOLED CONDENSER REACTOR CONTAINMENT &
4 THERMAL FLOOR _
STEAM CENTERS THERMAL M _
CENTERS [
REACTOR =
J+ _
CORE lHX+
MIDPLANE = *
[+ v v
REACTOR CONTAINMENT STEAM GENERATOR BUILDING BUILDING F81259715
CRBRP STATION BLACKOUT-i VENTING l
! n '
VENT CONTROL V :
! VALVE BATTERY E/H !
l CONTROLLED
>d SUPERHEATER l OUTLET ISOL. VALVE j
( E/H FAILS CLOSED r t SUPERHEATER ,
I I BATTERY CONTROLLED E/H i VENT CONTROL VALVE i
l i
s
) [ STEAM y -
) y j
l ( DRUM l E/H V STEAM DRUM DRAIN A FAILS CLOSED l TYPICAL OF 3 LOOPS ,
7
STATION BLACKOUT AUXILIARY FEEDWATER CONDENSATE STORAGE ~
ALTERNATE V INLET
{/d LOOP 2-.
LOOP 3-,
g/:(
pISOL. VALVE MANUAL
)(
' ' PWST TURBINE PRESS VALVE CONTROL VALVE !'
PUMP FAILS OPEN / BATTERY CONTROLLED INLET
~~ ~ ~ ~ ~ ~ ~
/' A NO M k LY OPEN, LOOP 2 +
FAILS OPEN LOOP 3+ N ;
C ___
)C [} TDAFW BATTERY CONTROLLED GOVERNOR
@DRUMJAM]
( p LLL ON IEtC u BATTERY d b AFW FLOW STEAM EXHAUST l SOLATION VALVE CONT LV E TO ATMOSPHERE B Ep FAILS OPEN CONTROLLED
STATION BLACKOUT-PROTECTED AIR COOLED CONDENSER d
ssss//sss////// gjg _
OUTLET LOUVER BATTERY CONTROLLED siiis iiisisisis gjy _
INLET LOUVER a
AIR SIDE NATURAL CIRCULATION 5 MW CAPACITY PER PACC STEAM DRUM 1 OF 3 LOOPS
SUMMARY
l
- THE OFFS!TE AND ONSITE POWER SUPPLIES j ASSURE THAT STATION BLACKOUT IS A VERY LOW LIKELlHOOD EVENT; BLACKOUT IS BEYOND THE
! DESIGN BASIS l
l IN THE EVENT OF STATION BLACKOUT:
- THE HEAT TRANSPORT SYSTEMS WOULD I NATURALLY CIRCULATE AND TRANSPORT i ADEQUATE HEAT FROM THE REACTOR CORE TO i THE SGAHRS.
- THE SGAHRS AUTOMATICALLY PROVIDES
! MULTIPLE SHORT AND LONG TERM HEAT SINKS USING BATTERIES FOR NECESSARY CONTROL POWER.
.. n .. ,,
i BRIEFING FOR l
4 ADVISORY COMMITTEE ON "
l REACTOR SAFEGUARDS (ACRS) l CRBRP SU'BCOMMITTEE
\
l i
LARGE SODIUM / WATER
- REACTION DESIGN BASIS PRESENTED BY
l PAUL W. DICKSON l TECHNICAL DIRECTOR j WESTINGHOUSE
! CRBRP PROJECT l
JUNE 25,1982
. u ner ,
o o o i CLINCH RIVER BREEDER REACTOR PLANT HEAT TRANSPORT SYSTEM EVAPORATORS INTERMEDIATE h HEAT m , EXCHANGER ,
- -lNTERMEDIATE l b= _ PUMP SUPERHEATER- 5
[ -
- VALV p :
, p [ e, g
~ -
RIMAR PUMP / f e
REACTOR VESSEL#
a-o o o STEAM GENERATOR y N \ dk I y e
,.g ,
5 t k
%gy iT j
m l
g I Ut i i Kh,$$th, I lu
,e,
$. . . p
,c-,$- -
l-c?
! \
1 l
1}gb I-f l
I a
_._s_.-__ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
4 6 l, $bV y n .:.
n v 4
- g 3. 7(%
1 w-4 u.
-NI_
o)F t{ i g ,
- ' I
? . .' ,
i g
P ' -
.l h ;__
- ; i,
.,4,; ((k7 f M; s' U $P g y t ; e,.
p j' 4 gjlly,h54 q ! g g !y ,
u
,;s .. .
'2 ;
. \ q
- a_ _
STEAM GENERATOR SCHEMATIC CONE OF THREE LOOPSD l,
SUPERHEATER l FROM ' ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, .,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
G, l
IHX r EXPANSION " f
, INTERMEDIATE]
TANK r
d%
- PUMP -
i=i= i _,
n f1 Y ,,,,,,,,,
^ "
i!?x
'M"^' " 4 l f 3 j
l J g e
p A
' w " " " " '" " " " " " " m " " " '{n '""""
ZZZ27: SODIUM JEE[ SODIUM / WATER r.]
- ' # "'*]'
, r y
i REACTION PRESSURE r
! : RELIEF SUBSYSTEM 5 g m.. . __ m .m_.,
SODIUM s.,,,x j S LEAK DETECTOR TANK I I I I SEPARATOR TANKS
] $ RUPTURE DISC j
i l
POTENTIAL INITIAL FAILURE MECHANISM l LEADING TO LARGE SODIUM WATER REACTION
)
- TUBE RUPTURE FROM SUDDEN RAPID PROPAGATION OF A LARGE FLAW i
f
- GROWTH OF A SMALL LEAK FOLLOWED BY TUBE-
! TO-TUBE FAILURE ,
i j
l l
j l
1
)
l
.n m7 3
{ .
O O o l
DEVELOPMENT OF A LARGE LEAK FROM A SMALL STEAM LEAK IN 2%Cr-1 Mo TUBING EXPOSED TO SODIUM Na VERY SMALL INITIAL H 2O LEAK FLOW F//// ~10-2 g/s (~-2 x 10-5 lbs/sec)
STEP 1 0.110" (.2Scm)/ LEAKAGE PROBABLY PLUGS, OR IS g
NO HIGHER PRIOR TO STEP 5 m STEAM BELOW REACTION / EROSION EROSION BEGINS, BRIEF INTERMITTENT LEAKS, STEP 2 PROBABLY PLUGGED FOR LONG PERIODS DEVELOPMENT OF LARGE CRATER STEP 3 LEAKAGE PATH MAY OPEN FOR LONGER PERIODS CRATER NEARS STEA.M SIDE LEAKAGE CONTINUOUS BUT ,
STEP 4 !
VARIABLE: STILL NO HIGHER THAN ELAPSED TIME FROM STEP 1: STEP 1 HOURS, DAYS TO MONTHS
- l. 0.15" l RAPID EROSION AT INNER WALL (0.38cm),
STEP 5 RAPID INCREASE IN LEAKAGE TO 15 g/s (3 x 10 2 lbs/sec) i H H0.05" (0.13cm) '
ELAPSED TIME FROM STEP 4: ONE MINUTE OR LESS (SUPERHEATER CONDITIONSI
_.___a. m i
l l
l l
l O ;
O V' n 1 >
VO' n 1 ,
C1 i ;
W I j V E5
> ,A, V , a, ,
,A' V r nV r r
nV I 8
w =g!
( l t
V l
l V
1 4
O r i
i
! PRIMARY TUBE FAILURE l
- OVERHEATING - PRESSURE RUPTURE l
OCCURS IN A FEW SECONDS ACHIEVES FULL SIZE IN MICROSECONDS
- WASTAGE l -
OCCURS IN TENS-OF-SECONDS i
i l
l l
1
..>2..,.
1 CHARACTERISTICS OF PRIMARY TUBE FAILURES
- OVERHEATING - PRESSURE RUPTURE l
- AXIAL SPLIT IN LOCALLY OVERHEATED REGION
- DUCTILE MATERIAL ARRESTS TEARING OUTSIDE OF OVERHEATED REGION
- EXPERIMENTAL RESULTS ~50% EDEG
- 1 TO 1-1/2 INCHES LONG MAXIMUM GAP 45 6 82 2887 7
llj ll l ;
b S
E H
C E N R I U 1 T O T
P 1 U n u O R '
A:
E / "
R 8 U W
/
M -
/
5 S W e :
S /P )
E M R AU P GMI X
5A 4
M
(
O
[l flll l' l
I I I l A ,Q, V O g g >m
__J i i nO O D hE U bpU [h-nv vo U n d>n U
. U p(V' u
O On
<U i l i l O A $
- m l
<V,o. VO A V E$
SR 4 i i l m
\ \ '
pVn VA i !
V U .
i
PROPAGATION OF FAILURE (PRIMARY TO SECONDARYD
- PRIMARY FAILURE CAUSES DYNAMIC ENVIRONMENT
- BUT REACTION ZONE STILL EXISTS MORE TUBES
- LESS STABLE REACTION ZONE
- LESS LIKELY TO PROPAGATE IF PRIMARY FAILURE IS LARGE
m O
g=5
< g b a O O e O
O O g
O
=>
O $$
E' i e
O o o i
! SWR EXPERIMENT
SUMMARY
1
!
- 34 LARGE LEAK TESTS; MOSTLY NON-PROTOTYPIC l
- SECONDARY FAILURE IN 4 TESTS ONLY
- NINE U.S. TESTS (LLTR) SPECIFICALLY CRBRP ;
l PROTOTYPIC-ONE TEST PRODUCED SECONDARY :
l FAILURES !
i i
) ~
! T
! a sz 2ssi is
u u -e l SECONDARY TUBE FAILURES l
l
- OCCURRED IN 4 OF 34 LARGE LEAK TESTS l
- CONDITIONS CONDUCTIVE TO PROPAGATION-NOT
! REPRESENTATIVE OF CRBRP
!
- NO Na FLOW i
- TIMES TO SECOND FAILURE l -
2.5 SECONDS 3.5 SECONDS i
j -
4.5 SECONDS 17 SECONDS
- 2.5 SECOND FAILURE l -
WALL THICKNESS 0.055" l
CRBRP EOL WALL THICKNESS 0.077" CRBRP INITIAL WALL THICKNESS 0.109" l
TUBE-TO-TUBE FAILURE WORST CASE ANALYSIS -
- ONE DIMENSIONAL - (NO HEAT CONOUCTION FROM HOT SPOT)
- STEADY STATE - (NO FLAME FLUCTUATION)
- STOICHIOMETRIC REACTION TEMPERATURE
- RESULTS
- ~1/2 SEC TO 1 SEC DEPENDING ON HEAT TRANSFER COEFFICIENT 6 82 2887 IS
-~ -
PLAUSIBLE LARGE SWR EVENT
- PRIMARY FAILURE ~50% EDEG @ T = 0 l
- SECONDARY FAILURE ~50% EDEG @ 2.5 SEC i
l l
I I
i 4
l .
i i
j o u m, ,.
DESIGN BASIS SODIUM WATER REACTION EVENT .
l I
- PRIMARY FAILURE - 1 EDEG @ T = 0
- SECONDARY FAILURE - 1 EDEG @ 1 SEC. )
- TERTIARY FAILURE - 1 EDEG @ 2 SEC. l l
l e e2 2eens
O O O PLAUSIBLE EVENT VS DESIGN BASIS EVENT WATER INJECTION (NUMBER OF TUBES - EDEG) 4 NOTE: DESIGN BASIS EVENT INCLUDES _' HTS 4
PRESSURIZATION TO 325 psig. PLAUSIBLE EVENT i INCLUDES lHTS PRESSURIZATION TO 150 psig l 3 -
DESIGN BASIS EVENT l ---- PLAUSIBLE EVENT l 2 -
l PRECURSOR LEAK ,
- (TENS OF SECONDS) i 1 r---------------------------
I I
j ,_____ __ _ ___ ____J 1
l 0 [- .
0 1 2 3 4 5 l
! TIME (SEC) i
- . m == =
\-
O o o l COMPARISON WITH FOREIGN SWR l DESIGN EVENTS l
! NUMBER INTERVAL
, FAILU RE OF BETWEEN l COUNTRY SIZE FAILURES FAILURES l
- UK 1 EDEG 3* 1 SEC
- GERMANY 1 EDEG 1 NA l
i
!
- FRANCE 1 EDEG 1 NA I
- JAPAN 1 EDEG 4** UNKNOWN
- US 1 EDEG 3 1 SEC
- NOT A LICENSING DESIGN BASIS ACCIDENT l **ONLY (1D ONE FOR LICENSING PURPOSES S-82-2887-19
O o o l
SUMMARY
- SWR LARGE EVENT CONSERVATIVE BOTH
! EXPERIMENTALLY & ANALYTICALLY i
PRECURSOR PRESSURE SIZE OF FIRST FAILURE TIMING AND SIZE OF SECOND FAILURE EXISTENCE OF THIRD FAILURE i
COMPARED WITH FOREIGN DESIGN BASIS EVENTS
!l l
)
I i
o o o f.
COv 3 S~~::0\ C- A R AC-~ E R::S~~::CS 4
J
- 0 S03:: v A\) \a <
i i
i i
4 R.K. HILLIARD l
- JUNE 25, 1982 i HANFORD ENGINEERING DEVELOPMENT LABORATORY s
1 h
~
O O O .
SE_ EC~ D 3-YS::CA_ 3R03ER-~::ES 0 S0X:Uv , 30 ASS::UV A\) Na<
NaK Na (78%K) K MELTING POINT CC) 97.8 -12 63.7 NORMAL BOILING POINT CC) 883 778 757
, DENSITY AT 600 C (g/cm8) 0.81 0.73 0.70 HEAT CAPACITY AT 600 C Ccal/g cm8 ) 0.30 0.21 0.18 I THERMAL CONDUCTIVITY AT 600 C (watt /cm c) 0.63 0.26 0.38 VISCOSITY AT 600 C (cp) 0.21 0.16 0.15 N
1
~ . .
W
)
s d
l o
o /m 7 2 l
T l
a c + + #' 2 C o t 4 +
A (k /
i E g -
R j Nw h
- A N N E I
O + t TS o K #m m c O O. t 0 0 B w o 9 # a, M NE 0
a, z h V N N O 6 N R *~
C Y Na T E ~ +
X =
T O +- k s 4 N
A N
H T
I og 0 E H w
T A0
- /,
/
l W g I W # +
J M S N + + s y O O ) n eu) u D
I T a m ru a
r N la g A C
N N A c A E 2 2 E 2 K R l i! l! l
0 l
\ D
\O Q m
I l
+-
g 49 h u b :lt:
e o l
l s k n d
V 0" $ N 4w l 4~ N w
oc Ifw f i
'O
, s f ,37 D o Oa Q" .
3
- -s -# :t -4w $
! % 4 + +
! O 4 +w
! U o O 0 4
- m ~~4 l0 4
i
O O O 1
r l
- EA~~S O COV 3UST::0\ Or VAR::0US TUE_S 1
HEAT OF COMBUSTION j MJ / kg Blu / lb i
- S0 X:UV C0 Na 2 0 2 ) l'.3 z800
! WOO) '9 8000
!4 CO A _ , 3::~~U VIN OUS 32 'z000 ,
GASO _::s E 27 20000 .
I h
COV 3A R::SO\ Or 3UR\::\G C- ARAC ~ER::S~::CS 1
07 S0X:U V A\) -Y)ROCAR30\S l
HYDROCARBON SODIUM FUELS FLAME HEIGHT 1 - 5 cm 0.5 - 5 M FLAME TEMPERATURE < 1400 C 2000 - 3000 C MOLES OF GAS AFTER LESS THAN INCREASE DUE TO REACTION INITIAL AIR FORMATION OF H2O AND CO 2 CONVECTION CURRENTS LESS GREATER RADIANT HEAT MUCH LESS MUCH GREATER AEROSOL FORMATION MORE LESS
, 7 l
1 i s s
m t w
o O
.s .. >
m.
1 oa x .- x g , .: ...- .
x-- a e t r l e . ..
b & s' =
j a -
o ,-
d .. a.
i O o G I i
Q -ty gd 'h R . -
R +
m
== :. t .
- R <
e . , .- -
\
m -Id - a' o
i %
+ '; .
) .
o" . ..4..
!O
, <3 - .
A m
..~.- ,
- g . .
- .-
)
W ,.
,e i
- 3E 3
o m M ,o 2
l Q 3 i
q sm e 3a p
$13 a- t > a m
at
~3 jO gb d SG <e% Eg og ex o o 1 5g 5 a 6 0
$k O" N k EN l
i
O O O .
.50DIUFl/WATEtt VAPDR BURN /NG M002L.
2 Naon (AERosoQ a
nato 1 +
secononnY y + o, -+ y o RE AC T/ 0 Al 2 2-cu ot > s rd UPPES 4L ) k HEAT DI F FUS 10M gZOg) y2) g g
' r zone pg,,gy :..:... ::. . : .: , :
anc.rion :.._ Nag) + Mi o,, 4 MaOHw + { H, aona .
Na DIFFUS10Al ygg EONE NaOH HEAT V y l
~
LIQUID _
So D IU M oa
J
~
b SODilm POOL FIRE BURNING RATE l
l
- -APPROXIMATE RATE = (2.0)(10 2
) gNA 2
MH l
l EXACT RATE --- USE SOFIRE CODE .
i(
}
G
, ~D .
^
- JO SODIUM 0XIDATION RATE EQUATION
~ .
R=4x c A3 (C ,- Cs)
WHERE R = REACTION RATE, mot s
x c = MASS TRANSFER COEFFICIENT FOR OXYGEN, cM s
O A3 = AREA, cx 2 C = CONCENTRATION OF 0 2 IN BULK CELL GAS, not cM'3 .
AT s = CONCENTRATION OF 02 C
l -3 S0DIUM SURFACE, not cm l
l
, - - , -. - - - - - - , - --,v ,
~ - -
1' ll i . .
4
~
O .
i HEAT AND MASS TRANSFER NiALOGY i
i i HEAT TRANSFER
\
Nu = 0.14 (GR - PR)
Nu = NUSSELT NUMBER t
GR = GRASHOF NUMBER
.O PR = PRANDTL NUMBER i
i MASS TRANSFER
! SH = 0.14 (GR - Sc)3
\
K'L
! Sa = SHERWOOD NUMBER = g Sc = SCHMIDT NUMBER i
1 lO t .
s
-m<nw-o,e ny m-w wwwW,----m,- u-
. c
- ll MASS TRANSFER EQUATION ;
I (NATURAL CONVECTION FROM UPWARD I
FAC1HG SQUARE SURFACES) kLc ,2gg T, - T[ " , -
p = 0.14 i T v ,A z rv where L = characteristic dimension of heated surface, cm Dy = molecular diffusiv
, = gas density, g en qty of oxygen in nitrogen, cm2s-1 O g = gravitatienal acceleration, cm s-2 TA = thernal coefficient of thermal expension = 1/Tavg. "K-I T.s = temperature of sodium surface, OK
- == terperature gas viscosity, of gas g em* f{< 3 rom surface, OK l
i
~
O i
i i
t .
\
- _ - _ - - - . _ _ . - , . . . , . , - . - - - - - , - . , _ . ...-.,--.r. .,_...._-..,.-,..-.--_---,,-.._n.., - - . - . . _ . - - - . - - ,
O O O .
EFFECT OF CEL_ OXYGEN CONCENTRATION ON SODIUM / AIR REACTION PRODUCTS DOMINANT REACTION PRODUCT
- GAS LOW 02 CONC. HIGH 02 CONC.
TEMPERATURE (<~10%) (>~10%)
<~700-C Na2 0 Na2 0 2
>~700 C Na2 O Na2 0 l
- NaOH AND Na 2CO, WILL FORM TO THE EXTENT THAT WATER VAPOR ARE PRESENT.
AND CO, C l 1
'{., '
,? ..'; } * "
y8 - , , ., , , - .,. ;7 Eas , ...s . i , i n <o .'. : d - ~. N g J,3 g
- ,. a e pe g -r "W < n.,
S * ,R 4 .W. , M ,a.,YeEk-$* ,
v.W. , . r}';$,,.&^~ E_, 5 h" N
%i.5sNh, Ol .,&wesy*p$$?k,.L. Q Q ' c,Nf* t* u th' i.D ~ibh. .. . h$$f>*'&h.Nhk,fMp M'h *si? M.4[- x lF,
- s v e *'t nN c e h v3r' m * -
.y*. ,. . >
,.p2r ")
}d.g< ^h.~ *" "4 g@ .
Q' -. s . }h*b g r*- . . . g{M* W;a p. x %w'g?7h*fi J gs
- pk',ea j
- gg%9w
_ . c.
J%^
3 s 1.- %;,v"['3' .4 :
- ,w.,l 4
- '%+y.t. la. . 'a g' r-
. c a
s
' _ ,W 9**~ P, ',h,, p'f p n'+- ,
%u&2,'.,
e .a ,;. %@. ,ar~. ;/,' j
- f*M$ Wi>>+r:
1*h. , A~b~% ') . ' f+
n; . k: . - >:. . A. ' ' ib. ?.
- ~. ,
Y.?
S J.s .;t%kh, f,Q f. N,%~W{%., Wi .n& ' b: s a 7'%.' .
.. ! '. ' f. ?
-- y? .f
^
V' 4, s - = ~ ,
i.c ' . ? , u^Qe ~'." $:' .W. ~.% s . ,~ .w V%]A vi s ma a .v.,4y i mp+- wy ?p'"p , m. .u h .
w
.r
%q gy on- r e n,-
e c;k.Y - . -- .' %
1 w v _ &.;~o* . +;':# . ^" ~ ens e-- !
as o
a
+ - a u, v: ; 4 s ,;j,,1 .s. -*,;,. ., .- ,
- o. s%.y : ,a -
ei "
f,& . p. t. . .y m , 4 ,e -
=. - f
- y , . c ,,, " 3 in
- . s r
- v. .(w' v: r-e,v"% ?
,a' w. .'.s. n n:w *p. m.%,
- s*8 Cam[m . y y .x v
-, w w nw e. .y .s ?
- '. ^
j; 3, . .d F5%r%t,4,+eWrWt* ***".%sS*9*.N* f ' N^*- '*:*-
~
. O C 1,. -** - . .' '. T ,,
,t'~ ' * . , - . - - . .. . -
- W U.=ff'.,.,3 9' t ~. ~
- qw w -a k T i -
n JA%
k*
''.;</***1.,., --
. ' i ,1 m./P% M e m ,m pma t.o.% > W - Q ',.. At v y s. .- n: tey - .. %.. . c4 ) . - .'M:k .s..*sVi
.4.' , . . -
.s .., . .
.* a, -,n.*W.M
+.%4
.#.n.,, A' "!- g .w Ac* +s= ta%.-.%...w.f a
,- g n e w .,%.gg = mr, ....,,.,i.
ar , e *w s .e .4, .
- -.4
-z g.w%,.4*
.-. m. m. ,Pa p+,4 s.,. J,
~.' . c . c. - , .
c.
E$, 'p. M
,.. w.,. -
-+ a+o.s.
.~,. r.. . + . . .* 3 v. - .
.g Nn*$i~ NT*'frd i'%yh. 3*W **W
~
,8*-
e '{ ' i 7 - @ 1 *y* *a-U*.,% n.e.x
- f.g****W.1M. y ay*f @'M>
7 +' ,'
- .. -- -ki i w . . . ..2.. ; . ; g ~n~', ya ~ .- -
- 9. . ~~,, +w,*x,.
. .., d ..
,s% .N6.,.M ,.v.ey,s 9
, + . . , -.e.
n m.3~. ,
mn,.m..
~ ,.,s p
- n ..
.; ,,
- _a ,
l b ..L...w. rq
. s. y ppF*~' ' s ,t ,. . . ze, w. ~m*f".vn.7y K, r \c% v* p* .
t'~ t WCVAL9ggtsWVi,WWM , y . . Q dis
> p- >
- ' .x.:.c . m .
.. m . . . .
v =
- s e t w _. . . . > s
_ 4, , o.,
- y. + . %. . o' - b . ..,
S. _. _ . . . . . . . ~ . - e- ,
1% .e+ - -
g a v34 . gp - ~ ' - - *A. . . .- - -p
- P-+ + + - l
,_ f . .s a .
)4 \ %n .. .\
-s . * . _ . e-
, + .* : g w
<vss , ,
g
, .,, . ,. ..t_
.,-3.,-~._, / f.
m' .N. . g e.. i ) .
- 'O t: rn # *b'-
W g
. *"' y L_ J sL ' a. p- _ . .;p
. ._. s
- b. '-
- y %\ .
t- . ,
' *.~
g j ~
- g,,..A4 g49- _.e . - g
. {Q A lg'Yj'y'[ W,R * *, 'h
- -
- pee..' G'**?' ~ c * '-*- ' s e.p "Nm.d / 3s fE f>%,t a ? 4l '
& If' .i
.,p,,w y, ,
, , - ., p+.1 r' ; : e..
g.,g w..;=..,....,.m
,. . s m ..
- a. \ y'. /
,4
. . u p; , y .,p%
j . 3: , , , .: < ,
l[ .es. = ? q ,
%e e .? pe . sy - -4m . e i . p . , .
. e *1 m o 4r 7 , r*., <*
% , =
+,
, s. s<
a
- a%p
,* ,f a, +
s ,-
4 +4 * ,f / g. ," e
- + . 4.
...g.,A.
+A? g v v.
.+-
s e.-- ,e t > ,. , , _ .M p g. g C
- \ . , . .
j ..
.. 'mg w,n 3'.4 4
" g '_ , * * * * , ., ./ -
5g
.+ , , > 8'y
.g j.. f n* '
,A e .. w.
..,e 4 gr ' .* * < . .j . .' -4 \
- \ . .; =.' .
J .. . n - . . *% <
'p
.'!: A i .; :h
> M i N " " #*,* *w' ~%, M r~*M*'~
7 g n,q- '
'"QN
, - w.. 3
. .ww , w,s.9, or
. ,:;*? /' ,~..- . m ..r . 9. .~.. p 9
- w..
- . ., : ~- . .u .
m vv-
? " '..
u..
m c.)r#.sg *, . .,,b.
,v
.n
-y u a
R N.,.' V Y
w $'Qb. 74 h1 .: 9 .. 52- '
7 '4 y D**,"* Ng Ska .p p n%.w : *a: . # p kW
. g . d'h E . 4w#LT M hp.V V
.m $.. :, n.p[ Nu 8 x 3 . rc .M. *% TY9 ,,;.. u 1r .a M AMr e .eu e $ " .' u.4
-' ', r. ' gyI.'.,y
- 4 ,M Y.'*_;
..NM., U,
- $N' r M % T~ j j. .yA 3p.a+ w. .y;-~ A'UM, ,d*.~ .-s k see ..r,. m_ #. 7 ..w ,
u;
. g. , , . ;~ p *. - e . - p%
a g - %+,,- r. W. ... .< .~~-
- r.. -~ ,. ,<
a .w< y Aw&, p p .m% e y
- m 4 . *~ n n.I,k &. MltG.cd-L.~r:i*:' $ , S :S:Y ? . . ,
4, n * % .
-) Y %
h-ma cc:' f> s.i-
.* ..A
- ma-A m m .m W Ne m M N .^-
, , - m . , ,
.r } w, r. -. * -
y . . . . _
n'. r1 . emad
,~3 r + , 0rW , > ~ h. , .w e .. .. . , , r,. .
e%.a,.p -,,,ap D'.**...,,,*,E'1.-*: . ,y . dw m,, w nnd , .ye s.
. .* Ud w , &a U,
+ $we we'. . * ,.~at %* . , , % *Ji ,
- e .*' *.ee su e . ths
% g ; ;:=v.9 wi.-
~2:
&w.%*% m.s . i.e M .',;l.A.Y.%,c r l's~ ', c".'. ',n.. .. n, ;:.i neww ym ,.. W ~r ~f.M. -.q W'..a , s y.-?." p. ,
,.- , m .z. y n s .. w w,w .. ..
. - ~ ,, ,4, :. c a-. n . is.-
~-ogu p s. *>yM~W .u.- x. .
.w w.. . ..ns . . . ~.y ,n. A. s
, m ,,.$ -
. m.. o 3~~
?
s ..m..m
.w:mtQ%.yV. m -
- &m a, ~ 'i,w y '
q(ykQ.[
i .A -
YM. 7 D 'NY Y%~.yk q sW% ++'; meh -" .T."
'qr .m e
c G . > , .y . Q T. p i 'b*QA .*
!WE#f 1 D
,. .e. s,s m. c a. ..,--3 . s ...m. .n . .e.. - hn 4^ f,.b. .. .N.i[r O.Jw. g ,,, () (/. _r ,~+ ~.,_C ,v., - . * ,~. . . p% .. . . ,
o> ,. .
- h. . . ..:- A.J .
- s.w .-
.4 .
b p. .n4 . n-s. r ,
e y
,.:f&
~9. ,
n
..P %.f' g -
. . . ~9 - y.g ?-" yy r .
..,.z.
,.w.
. 4s.- vk **
- r.
+9r , , J&.L W .3 %. f 4 7" a*a~pWt , ' . . m y, , .== : b .44 - . <
..7 ;v. 1..yr.'% <f m,,y abg + ' . + , .c *.'r %,4 m. ' - f- , y
-. y
,' b.ge. . . A. * ',,, a - .
" / c_.h..
$ .m
{ pc'r,, La%d yge. st* t %_e os ..1+1.*
^ ' .-
.f9 . e-4 Y_ -
gWsy
, .es% g
-.c e,,#
. .g .* 4 . m +
1
+ 4.P44e.r*** eed b
~ 4 6.e .
e- -,.y c' '
O .MM
.c. . p . . Jv. 3u..,
. a.m gped ' k d e.fDO M ,88
- y e ,eSe'. 6 %' . ,.. n p e - -
1 + .y-..
, .m.,.v. >%
- s. + e . f .
' . y . g A ..e.e - . .
y ,.w.
f f> 9.. ,
, ws . ~ . +..-+~*w<
.s 4. : 4*p.
- =n
' . r L, .o L
, p s&.. :g. .,c. .
sr-s<.k y 3.g 4 ,
,0,,,,
- s. c .
'. ..ut . ; a ru - 1 i '
D k'*19 s n.JD ge* v t.s m. 4 *4~**;'_..,.---T*.
4, . *m w ***, p .
g' - - - 4 . wt.ee'. % +w v.%. -t %s *u.--
.t.. .
....m.. , . . .
. ,s,g . ,p._srp ,,.-..,..
g.e. .'**W'."=.*A*.N..*'~.*s- .a -- A,.en .j p .qi .
.av- =* s
.4.e,, , 6
,~
3 ,-%.~.
,y se u., .A._.$- ve g. .
- o. .s .
y eg , y .,3, y y eamm..,,4. g # a .w- .--s p e. .= . U s -
.a t . ., e< < %..m g $ ,-
3 ; p. ge, g -,. ...
e * . , - . , *'g,. . . .- . e ~ -i.
.,3U ' %Y f .
- g ** g '*
- k v ',-i$, .g>
$. k, Dh . % , *$
A 'e',-
[ -g *89' O.9.g *% 'e De,
,, g .,4 - 44 k. Y }.[ eV9g+@
gi
+ .8 * -
. 4e .f, e g ,
p.
< , n - . . . . e
. . ,
- M ) f ,'.P6
'k
- 9 --
.e J. N . ,,,.\* -h ,' S , I[ <' - a g4.4
. -} - F s' ;
+.e. . . -- ) .
g**g,
, 3 q- - & ' ,,g
-1 , nt
- s- < n.' %-_ % ! ,4 jO, , ,# \ /,-
h ^ i ,Oj=9 4 ** k kY*dM *"U"
- ", , Pf, .*
. g..c"
_ y ** h .
__^ .
pvl* D. g g.R. W ~.**lf,, tt W N h{f,@g 6%2QWQtR=n-w&.%' .. m%'.
, .. ,y,9,/3 , , ' ,, ; . ,-
-r ,f mt..**.- n
, ecw . my - ,.,j <
- 6 g, ,4y - _
..q , f . .e . ,. f .
.4.. . ,4- 6, F.s , ,
' s ' 4. -'* * * **:'J' W $ .~h: &&.$~ ?'";0!M* b %%Q ^ - l 7 * *k l Q th,-Q Q .}
$ 'V -%* f.&*'"*J\
2 - . s.
8
, .- - , , , .* y -= . -.- ' ..;., -. s , ..,g*ww g %+ . . s 1., ,
3
..g -. _ps; .a
.w e'." .. s.
, ,5 1.
. ~.
'*' 'M *. * -n *g; y . 4 s v as* s* n +
.e *f. ., .' .thar e
,r w . <
- r. w ., i *.-
,.,..e..- '. e-. y am ~ esua--w :, _ .,,-_. _ .
+~.,~r. w O, . , . '.w -
6 an,, . .s -
, ,,. .?
.) .. . . ) , ,- ,.
- %. . r e.s~* d., e- , **
e..f ., w ; ,. ,,- -
w 4.. - . - et , -. . -..s. ,
S* , ,, . ;s ,j
.e .s -(an.- . . >-
-. . . . s s e . , .. .
- e. - < 4 + .
., .~ . . ..., ., . . a-q , .-
. , ,, . g. ,, f . . / ..
i.e -e. .m. .- .g 4
,3 c. . . ,- ua,w m.- ..,w,
,- .c t. . . - -. . .
4'
- +'.....i a~jt.' v.
g ,M3.-.p,_ .. . . .
.s. . * , , ,,
~ ,* .<- s-q
. ; .< .. . - . 4,. . 3 + ,w. . ,. .
,. - f ., e
. W'
~^ . , W* yy*4,l. - - s
- { [,s - ..
O * *'*
- T ""'j >" ' . .s.- '~* -.r-
's '*F. e; vs:.". ..
ww',. w r _.
,~
- 1 *% 1 t,, * : ,,
% % ' 4 t,. '
. - .,~+ +
+ s v~. . . c r n; & .
"'*%v g3-. g ':*d ' : .
.., [ 7
,, 'j ;,#
g ky ' . 'g,.
,' A 'I4; .;. 4., {4' 5
[.'. ..
.i F"
.J O g **' - s-#T#
sava ,9 .
.M n'* #s 4 e *< , ,, ,g4 4 i .N .
h,.[ @ 't ^! O
{libhN v
M ss >
IM - .,
.7, ' , 'h f.*
p M O
" .u, ~. h. + -
~*c r.
.y".'+
7
=-
g
- b ,, ",
_E
? sL f,u, ?! Lel k g '
E L :.:.K J - '
- - ^ M' i u
t _ g
- b. ___,m _ _ __m _'_ ' . / e e .w ero f . .I . . * =* w. M e -
L- -
E
d
/5 F
~ >
g g . : s --
pp -
c- > , , ..- .y
. 3
$. . . * * ' b _e,,, e ,r,,. , n I
f* t x- , 3 - -
1 i
.w a
' s . ;
s, .y
+ ,
- % y , , '_ +
a ~. .
.J 6 ' ,
s s
f ,4 ,
I
.44 e.. i, e p
- b da 4 .
- y
_4 % b n -,, . 4., y
$9 6 e 4
..-r%sUy g '
- T
-e y, g *, g #
i* s .y* b l
,8
,- . .e #
/ +
- rc ""> ,
, vg.. .
j..
t -
6 c..< }'C - > , ,
[ : .
- a. -
..w ~ ,s . , .
~# ,,.-'W"y!\.
l pp S '. uggg' , ^, as.,+ ',, r- -
w.- . , . g , wr' ' - - b )I*s e' , Ve
- k "" * ,,.Vg. 4*j i pl
' b _
J g S 3 [. '"e, ,
-w * - : "-
, u. c g. . . , .
y e n' .5
, , -> r
{
- e , ,j yPN M ~. ; . .
- '" W
.. . q W"^$ (
u, m.c% - m.w .w,e.,w.m.a
- g *. # # ge ,--J.g ,,#,j *
~s g ,' ,, g
. . .& g
....v.w.' . .
'.g.
.g
.{ ,
._ g, -
d 1' .G.
- g i.,,,,. .[.g ._ g3* pS _ ,,,,' .
.. m , u. 4, .4 e , w w e
- p -- .. m e,u m v = -s . , .*v.-e,,=f**=. -
' ' r s v. , o ..6 ~ . . e = . */ - ,. c=2v **wrs * ..
( . / . ..v O ,s -
=. m.~~,r
, +- . a ~.wr ,
- Vr :..m p we..a-n . :.,
^ . 2.
. c- % ,kw. ~ .g~ ie.. m Q-- y* &-y w 4-
^
s .
- .~+~-
~
~ G h & Qanw J*p W &>d.& ? Y,.-~1,}*? N., ~' * -5.. $. ' 5 V"5 W ! S $ Q l*JJ.5 S.,~ & m+.r.AA
- g. ," ,.
- ' W
( . . ~.
. . . . < n s1 '... , . .- 4 .'.a. , ,
- e. +..-
g- g. _,,8* $p 5 " 4 8* '84 g j g [Th* 1 T .b S@ # -
ED . N~ * - d 0m Y e %- #~ -
,~9 i y
- q* , w a
, e
\ t
\/ 4\ ,e*, ].
- I ,
4
- C % .
b-9
"* # Y 4 e .
-,.' - .. w , .c. ,-
n m. , . .
y . (' a, (I ,$, -*t.J 7 .n4 A ,%, ' .
% -4
/ :., .
= sa.- , .-
a
,s=*
,*s
/ *g
-g;
- s
,,Q
- 3 > +.
v -. 7
'y m ..""e *#< -s
%+
j e 4,
,o p ...e ,ge ., ,e .s* _o. y
- - . - . . _ .. -F' l 9 Se p g "ha g fg d'
- 4 .. , ge . 4* s a ) , _ i' w .~... @ . - . h. . . , . % -
..x-
- n. , , .s, f...r .. ,
e r , - < nw . . .
.* \
.%. . -'g,yl
- -e*- :. A
_% . A L. .
je.w. .,'s gg .,5 .-..
,m a
, ~, p : 4 ;**l4" , 1
- g, , ,
+r 3s _
- 5 ,. .
- . .v.C v
"'. .c.
7
+ . ** w . qt, , . Qw . .,,s.,,,nea,
. ' - - - ^ e ,
[+.. mp
&.f. ,q . . y r
4
.x i ., c
.s~..=+x..
s
.. 4 4 E.*.ei=83Mg I
'O --
'k b g w .ta-ige ]
e-k N U . e,[ . e er 4.M pear s
-. - . 3Q ,. gy*.-=m..
- 4-
' - ~
s .' -% w,g , g'y ar. ,4,. . y . ]t g*; 4,4 g. ,
-,.. - , ,m .
,,3 . t , .: .
,,- m . La % sy ;.r', 5* _ l .' r um, .y, , -} l ,4} *
"q
. eq s
y,p, 4, gys .lQ j Q g ,,7 ,
l l
s,
. . .~ ~
f-
, ~
.{%
, 3 . 7 ,;
- 7=a,
< - ,4 a n.A. , ,
n m.::w e f, , , . ,
, *+
" 4 Q p g .
,, , . - ~ 45- % ,8** .
s .* , +- )t
- t. ! ,'t i e i C * ** . y ' . - . . , t
^4 s*
- q Q Q v t ? Q < 'rn ~.il. ~ ' o&' s .* - = --
'1.
. o .).3} ... ' ,, ..- - ,-
x, y,,. 'I.. ...,rwe3
.,, I, _.s.,.,.w - m a .v._.
- ,p +. .+s.q. 3. s .#. ,. .pr- . , -. .. ,,,f
. . ... . ,s., .. 3 .
- g. , . . j .*-
,,u,,,, 4. ,% . J'[# ,.y ., , ~ v.. s. tc sg -
t
. c ,w7 r
.#, 4 g .- a y' y, > V 4 %. j
.y i
. , , , y,u e.
a.
4 p i. p; M. '.y... geA..s T :t "sy.,p 'q,;N ,y
.: , p l c g,a % . .. .mp .,, f, '
- ~ v y,
r.
g.y rq;aQ . .. m.,
wtyp :> .
.n . . . . . .
7 I w .' Y' g
- +
- -Q f-f
- f
- I, , ' .g . h1h/herY hh l %[W
^
- . 0.N ~ w
- P Is*
q[ p -3 .
- t ~s ~.W.r('.,*:~ . . re
- t ~ V & % ~ :
N,b s .. - y r
J
% 4..~*T .\> S.,t ..'ts9 p%, .(.M W.,,. 9 .f %*. .L..M d.Li.?.4>..
. 'a', = $_".'~:..n..- , ; -. ~
~. '- -
n, . .
... %. % . P .'. -- - r* < r -
l
~ - s.. s.
"a.. .g.A h,tg v ; %seje l.
-="T&.p. c ,s c' ? '. ' . . ,~ s % .
, - .., '..w
. . - , w cg.s . es
- 9 . , g.. i g 3 w p. ,,
< . . . . . . . . ,y .
p "-#' ;* . - . vs* .,*C.ga ,, , s ie Q<t . , . 9g 9 .# g .m, g . ,p .
%p
}gQ *
- k[ , ., a .,.,(g*. ,.,, ,.
' m' " . g v .;#
I . *
.[- ,
g , .
, g
~._ . .- . . . .
(, , . j *g \c.e. .i -
4
-'s -e..
48 t,g,... g p..
ppi,* tQ *j s..t,,
1 .,.,-
. e.*
,ee ts a, e-5 a.p* w4 .Vt wr r Q' : . .
t.,),'
..y - 4 d .- s j , + . r n * * - "y M4. , e ', a. .,
, .- ,.1 . / "s . 5 -
1 c ... .
. ~, + .a y y.c 3 . _ __
.m y,. -
p Q. pag. w ,
w .- u: , .s
- 'm n ,.y.., .,
-y -aeq. ,' - -
- $ 8
.e g ,p. J*,-t> .p yq . , , .
a4 ,
- . e., sJ 6
k t- ,*w.,. .,-r -g., .' % ) 5 .
. . . . . . . . .. 7."d. ] 4. ' " . - .
, , . . . . . - ..- - =
F TT g.4 #'g . k*W( 'a.* %" *
.- 8
,,,,( .~p ~
0 A9 Y * # -* #
- -9 h .
44 i.,. .,
W***q g y g%M,.
.s=m.e *w. mp y o g.ah.* w ==N'WI"* i-Pg*%e e em _ t, , ,.
'd-
& 'g pm {.9e s'" *#
i, . . .
.-..f. ,,__ M .. e = .ks i
. w! . ..
I
.l'..~,,*1- . .w, - .
~ -
L.
,f 4; p
,44=,1.O4+ @ h W [ f.4 fee.#d L.'.I.[ [.A g*$ rA% # A ,' . '* ) , m *.,@ , ,, ,
h.'if.,m. #b" .
F&"' 'h g
.; M. swtu M. r ; ,y%e ,7.Q *,e^
,p -,4, .. i... - - +
_r , f ,em.
r w .n . .~ ,
m.oea u..My.a.y eyy .r a -m....
,s , y v~,> ,
Q,: . a ._ .
r .. s ,. . . .
-lm= f h,%.fY ll$hfff5$$hh;g.,r _j ;,,,0! 2 ,&,' ,y +:S f '.m?$$n?. k:ft + ~
. c.w wh n
k_ .f_ & T Y Q W & w&. _& h_ x t w; w$R D ,W W t iNWNQ?$:nw n ._ 1
. &n 2, m . .
~
O O O .
ErrEC-~ Or CE__ OXYGE\ C0 \ C E N ~~ RA~~::0 \
0\ 3E- AV::0R 07 -YJROGE\ G E \ E RA~~ E) l 3Y S0X:Uv/WA~~ER VA30R REAC~~::0\
< 5 % OXYGEN > 5 % OXYGE\
HYDROGEN ACCUMULATES HYDROGEN REACTS WITH OXYGEN IN CELL ATMOSPHERE NEAR Na/0 2 REACTION ZONE.
Ha0 REACTS WITH Naa 0 OR NaOH AEROSOL.
O
O O O .
r
- 300_
- S3 RAY
- VA3OR
- C0_UvsAR
- WA_L
- CO v 3::\ A~~::0 s S 3
s n. * . .. - .s .1 .,. - . . r* ,... .?.. , . i-. . ... ,- L v. - . .+. y?- ' -
.e: ., ;p .? ? : e .:. .% %. . :... -.s.
)8 s ~
- ~ - w,..#.
.a m.,ww. -- .. . . ;%n..n~ . n , ,~ ,r ~ w> e
. y y 9
3
. ~- . . . . .+,o ,,.c ;s.
., ., h
,m,..n" c ,'e mn e. .. f .
i (, '. * & ..*: r . .s
- w. .
- c ,
- * . Atr
- 7 &. &.g. 5.+ -
,,* W ..
r ; ;(
^'
.'; .. ~v r " % ,:' '@ ,YfV , n
- a, _s".~R.yy ty'l
,- q. 4%, 2.2. .. y, ,.,a,? *.:t,4 ; .K illlo. . D M.n,.n.., .W. ... i L : z.. m ~i&&. g.J_ _ &. w&)
,a , . .
g r. , a*.r-x, . .. . , . , . , . , , . . + . , ,. ,.y . ,, ~g. ..
- ) g p,. .ay. i.e,, , s. o . .y , . . , . , .. a..sw . . . .se4
. - ,,. . J A ,. .u (*e ir m , . y,. . ,,w.,f .
r J ,gj , . r**O ' . -P - ,- <- . ,
. ,,f'. .3 , g ,,jjj1, yj , . .. J. ', *9 .
- f. 'F*%b'b.h [N . +,,.. , i*,- 3P .n .; h.5i.* " p- , % i?.6 $ *i'O_ h t .I b.' L .! $ '
- -.pU.$.
n.e' *4-
.. s., eg
4
~
N'.b - ..
i'd
. N- .[. f;
- e.
., . , . , O,E, , ' .. . ' . " " ,
.. ~.., - .
.*% *E.-
s
. . . .. n. .. .
- m - .s . .
F. "g , . .egy op, *
..s.. n .. . . r w .,em s.m.Jvsm.t, =., o.e sg y,.qns W'w . w .~= < + r; ' , aw a. ' . - %a ~- +.1 % .4s j> - , t 7 -J ,o w e wd suea ,,-f. w. , -, - < 4,,,.e s. '. . .au %
. s ,y_
t
~ . . . . .
- - o .
__.r..g,1 gyg
,,.w' ..'..og.c. a,g I -
w.g g..
- M..m.f *% < #88 . . g a g,pby 9
-... 4. . d$> b. e#.44. g
- e. , ,J ,~l. A .*
- Mm M ks'y.sho- ... . . ..-..~,'t,."A.=.. . . w m" . h,4,.,.h_yl....%*. e. .,~. [p ,{ > p' 9e f,.
. F.M. ._ .
g, , . . .
4 ,.f.a .. . . ..,. -
.r.m y x, . s fl
~
- ?.'"-- '
ll
' ~
f* '
.D.:. ,,'m..,
- ~
3_."--
. W
~ o m. y
- =_ e. . .. . p..,.... ,. .- -...
-r e .---a,1,-q , , - ---' . .
- #,,cw. - , ; .f .,
- g -- c -,. e . , t. - ~ .
,. - - . - m . m ., ~ .. x. ~m.,. r ,. . ._ m %
.. . 4 ..-
i x m
.g a
1IC, %.
E '
r,<.+ <
6.. b. -'t 1
L ..L./, . . w.
,# m
~p g pq,-- . -a s
- ~
/j.
i g;,
(
. .-. .-A
, - . .~,
. ~
. -) c
.- % s(3 . . . , ,
- s. ,
4.
y, - r ,
- Q ] ,2 .s. .\. ..
}- - - . - - . . - - . - . -w. -4
- - - . w. .. . .. .= r
,,,.,m,<gA .,.* .
, . ..Ng ,
-v.- .'
-%vp - ., g h.
I
,, . , p.ywyrde, ,i.. s - m . .
r W. r
.s.
Ie.-= %8 .
- AIN*
. ... ,;. dga. == g
.g -
1-d ,
- ~
.=r
% & ,*A 7
w&qfg *_p ^O
- \ MO=- 4 O
- 9 k
- M*9' O M .
e...6.6 ., .8
.p.u.,..4'"-L e-8' g , .
, . ., n .
- .; . .. .t . .
_4
. ' J .I-- ,
, . _., .. ..~, ?
..M ,, w .w. . .
3.
- g. .,,,, .
r-%
r
- -. ~.- a. - p .- ,rs __ y ,;
,. ..v., .-g .$
g .. s-
-. .. 6 9 p 'l-
. .+ ~ ~c e . ..
. . . .-.e f . r utf, , *p. p- W
. -3 ! I O ,- = "44 9 r ,
-v. * ;
. ,. s
.4 d, a.,e > .- < . . . . - y , ,
i T
/,A-.,L,,,.,,.-
',. l
-e p ,
~.,v,.. . .
- p. ..-
wg ,. c-
- t ,>* - m-
..+-,A.
. u' op>, . . .,,c. , . .*
' a L
,- ,/ss ,. ; ;
wan *m sev. W. +:;.m .g .
..n<..,
.. s . .a. nw-
. +
n *O* t
- J.m . :-y;
- t. . .. z %4; 4x T N .".** b s <_[ % ,,p'
- n +2 ) . .w ' i , *
- .( . . , r'" . w'TEj ? *.f.WD'n%h .T
% ,* p y y n f % &. [,
s
, n.,.3,: 4 . g y.. _ ,. .-Q y,:) [. c } a,t 9.g.+.:-yggfygg t
- . < . . . . ~ . , . . , . . .y
,a ..m.
w
... m-e. ,..s. .oq,.
, .c , . . .t , , . --,.m. _. n, - ., ..., ,y , .
.,~,.3
~
1 m,w w
3 . ,/
- n w3-2. p ...
- .w-..y... g.,:.- . es de e w h
/f
- tfpiv,w- +qs_ y ~ g ', = ~ , 4
, ,; . . pg,y
- r.
,uj o+ _s ,s .,Q , y . _ _ ,
!" k - u- ~ 3 *,.t'. e'.gy o _
"r~-----.
- .m pm.s i x. . m.t g. w9. .n..;p. . < , . ..w~. . .w u . * . -.
. . - ~
.w
,w a ~.: p x .qn r..,s'~ ." - *w. -M.n ..*.wM
- . . ,. , . +. a
. ..- .c . .- ~. -
. , . . <v .
,(i. . .cx* h., n. ..h$3bI (f b MI feag c s* e N87 w t'- ' - * ,.9 *'e.*- # ; p .#.,. 44 ,,MS,. . y gh j . t( ', e w n e %. ,n % . . A ,' e - -
's,
~*'s d '*r *, i } '* WA,f
<<. ,. 1] g;,
g , ,. , , .
e.3 ,
- 4
- , ,~4b
~4 ' s[. _5 oC , , mM g
.a)
=
,,,d,. . . '
,1.,.. .. e
. , . .pj
- .. m., . ., , <
- j **q.A g , -e , , . . , , , . ,
- y. , ,; . . .
S
?.e**.'.- .%. m. p , , , ._- .. ,,, ,.,
,'"'s .'*..J g s
. y.%fe - '**"e- -.
<.'.I ..e.l. ; ~. y','v.
.s . we .. e
..s 3 .
s . . rA . , ,
's ..J. .
'. a.
. s p ,- sa. t A: :- M ' . . x,. - .y< ~.,.
sag.. .+.
,.' ., ,. ~. . .,'.g: g.,. -. , ,y
. q wi.c.,,. q ,v. .,, n.
, meiae. ,4 }
. ,, 'g ; ,4 av . .' + y . *g,3 l
/$ " e- .'%. . . so., . r .w s
s
- 4. s y sT y*W; *
- yc r 4e.-s.g, , . .s as .m. . + . 't, yJ -.s- + . .
.. . .m .
,. * *,. c.,y<.
f.y. ,,., ef.s,%
. 'N * ,sw.q_fi ,- -
4+.~kyyw-
., y.
- 5
--s
, . y p af. . g.
3
-,f m 4 .. .
. ,. ve'. M,,K,,, m.q.fg.,.c
. y.A.. 1 - p - gr.
,, e.
. ,:s 6 7 ,, g*,*g%*. Q .a. )
- g,,,**-, gg k 4 s- .I*'j_.=- )
- y-'.
- I -# *** b
< ,j, g g, ,
y-
=
W,-.4' e .{". , " , ,, .n .* ~. < - -
t d". i ,,. N S-*( , , 7, cape. g{ , , ,+ , .,pt,'4*f,y;y L.u ., .. . . . ,~l.
s'g**
o '
'v ,*,*,F-'a*
i ~. . ~>- .
.5 A.h p - .
- , ; . , , ., p,.
g
. s - .
.mg,. w..
.r 5
+. . .'. %...*' , .f s A, -,,. e- +v /'
t ,. - .
~ ',
.4.. .*
7,-w. .
i
' , ~
.h r ~ W - .A.'- V,- - ,
.t.".- '
- W *. 3 -- ~ 't' 6 4 'a M. < .'M egs'
' g 988$. E l ^ d e 9 , . g, e Q$ P$ M t4 -M4 . . , ,, e . 9& ,,~ gr W-&, , , ,e t
, a. ,
>+-..
-- -- ~
- a. A.
.. , ,b- ,, _ , ' , , .' . ] .ss .'':o" -
'.*.,...'.s'c.5+ ,n,
,., .._ *. ,.m.% . , m.m .m - , ,.- s - . - .- , - - - - . . 4 7, g +.,2 -.., .i..m I
, .+
h - '
pv.,yf E k ;
. . ..a.
. 3..' ,
mte w p.
E
, m, dt.
pg....,'...- h' ,, b 'd
.%.g
(( NN I "
. p [ 8 *, g
..m,..~m.,--..-.,-- .. . / .
- s. ,.. , . . ..
.-,ri , _. m .. ,,..... ; _ , s - . . . . . . ,
w'".h.p
- ~
. . . - ~ .. q g.s . + # .m. wig -s. * < a#% 4 . ..e. .f ai
.457 d t* 4Jvg a,su f- y, gan $
- , , g -- A. ,.w . .*it.te u's. W g e--
.- .g.,= g, .- 4 es. .ie.,, 9 .,
M,.g eie',.
m; g.% , _s.. , s,i , ,.
- '. M dh*[E t* '3 ' N '.Ee*[ . e,,, #->. s ' N .- e +"i fa~ ,. ,
- T.**TF*1 D( ' ~ 7 **I~ "*) bam ,., 'N.'f.C[**i,,, .2 l
- . +g .
, .' y..n...,.><s . . . .e.V..+..
.-3.. . .-
.e. . "Y ,, ._ a-. , , . 1 . . . .,
A ' ' <.1 y ? "7 Q ~* ' y ',4 y saw t- ;, p " 'l f .j e w i .s . ;"l * ,
- h 9n K/ . .% %,..jl T.e,*$n A*r W ~' ^
- ,, , d ,e .
E'
", w' "J,g 4. g' , hg* ,-!.?,".'h'A , , .* 1. 8,7 f, ,
,} b' **D- ..,M.7*t,TQ,M,d-.+a
,-~ ,e..- '.. . c%N.'. w-w -V.M.P .3- W & a f
e$
i
- g , .- . ~ ./. .
..'% .ep (.+a , s . -
'E- 9** g' ar* "'9. - * . .
W g s -' * - * *,84- . - - f %" =="E@% . *j
':'
- 7 * ).W"*Jf2 *f 4.*','.*'.*"A
',-* P f . - M",,~." ;N
- O.* ~
I" * * * '$ j, '
- k; 6.* % & *i
. .e -- ,
M ' ' cr. 4, b .# " -- # *.W C - %- h q,-h a
'* M- e
- J[ -- ae,
. p ay- ..r eg . e ,en.* .$,e- ,p.,. a.-.g.
1,
>.. + - . , , , *
.,=h, l . *)
, * %.e - *1 getree'p A.A.e. ., i. . w a.w anA . 47 "y.% . -4
- e 9 , , e o.wa ,g .4b7.. ' Jee, Je % #s gr. rv -.Fa,, . %
1 e= #,=.,4
-wwg ' *,(.
e ee,s._#,
.seg -
- , =v..g- 1 e s.e .
. ap . e .
g .,, .,w- , 1 iA .
- e. =
,e..e GF . , ,e..
,.dp el g,w W * '$ 2 r * * . _s *
- to* 1
% 'a#%s %' 1 ..
, . es
...es. w*,. .- , . .* . O .-a.s-s - - .e W ~ . **e.-. '
-y ' b*
,,y <.% .
t e-. ; . .
'<6 e .b
- d@ g -
,e,. # )=3 ' e # ' #,94 ,\ f g ." . ,ee *F(',,- , .g ~O + .9 8
v, be 4 )9 ,%'-
- s' * * . '
"i "*1,%.'f? 4~"u*}
p, 4
P . V.y ' .' @O ,. , g ap at,,." '
t
" . ,+ *, ". g. -, ,
- N'8 i4{'e . i 4 ' , , ,
n .- . .a go n. < . , 2.: qm g . af - . h"' , h . , > . ~
g
'95DMs,?.d h 4yv,oh;-' [ gM, . '*.f. '.
,l,#' f, w
~. ;m i, ,g~.,y ,.;% . *,".. 4 Y. z f.g. .'k
. W -
, * *~ .'.* , u .N. ,. . , -
- d %
- r. ** 2gy.
e
. , .* .g J - ,
% ,,..c-
,,y"", .
4 % $-
f ee 'e- p p* ,,M
.Y. s t , ,
'es -, - ,' .j . '
- 4
.i8y,.s , s
- . ,$ ' 4 j ,
o'
~
I '
W j fy. n , ,.y ? ? ~ ., . > 4 ,
l - a
, , g
.e
. e* Q g m .v i . 7*4
-t ' s' ,* . - .-- . -
ea , 3 ,
' not . y, - , ,
'. - - ,k "y,b '
- 8 S'. g ' . _ - .h . [ ' . .*
. _ _ m _ .__
s,
r M
, g #.g Nb .'- Mt N %.
m gg ,.,, *Nb., 4 q
4
,is , ' c w. ,
8 . ; -h 188 + .s".e.* .' .
. T,; ,,m. [.A8 d N.,~ ~e' Y.$k . ..x-D C -2$, M N.s.N,@* . , . j$bg'/@ ' 7#Y
.lTgNb,uh, f 4*~,
,4 .
- %a. v .- < .
+. . r N.g@4. ,Q ,,
% eml% ywy 91..gsW.
- %%ch.
' ' ' ,Q@WAhi,5
-m:wn w. .ww 4p. . . ,qQ;w;&, n.G WMg;;sk-Ag.fQ L j) t afMwk-%w5,%;.w;wc%: m& m;gYQ&;,4,u.4, m mu
- .,2 @ W q;;,b;3m
.;A.n a n%4
- g y .e-g .t ;; &. .
25; 9 -
- 3 O
wa. u. ,. .v m sew.m wm, 2e ~
"?y'.c.&,, ..,. . . w, % .~w, .
- ww.,..gm. . . . .
._.a. .
m,m _o-
.c..
e .,
. . . . , . .w %, m, y. %.
w3 %,, . m-w. . m.y_ ;. _m w, . -
r ' ybM'. w 4*k p. M I A M h b MM
- .N W N'MM[g. %..% 3;?[bidM t., e $ 8w- h) e-N s p Pv3.SW M 3-% V'M/ M..y f ;l( f.$ M*I s.4,$Q O e # -
e O
.w m.u p <r. : -, ': , g @ U TT-k:aW.
.Y $ rW T ,dc% .; $ P.7n W -M iQW g Ly 9v: .
.. ~ m M r. x m .., -2;,*3 M msn +n.n.assc-e,>. w am .aw.~.o. 7 p.m. c,i . -
,w.n u mm m m ..
nm
. ..c w n,..
U (M .s-g;.t ~ NtV. c,
, 'a.,..' ~ .M,, t _ ~ ww. s . ,
a,p . r
- 1, p.y.p-m.e q . %. . w+w, ,. . , w,y.
u
.a.
y .
s e .e 5 _. __
5 O '
.V < *~-~' ,* ' ,' u' -' ' ' - ' _
~~'++'v" > ' < .
.'- - ' r. e. A-* c. '.- .J'..,,'...
g ' w 2' + ^
s .
O
' ' - ~ aV C
. " - -c -2.~:
U - t . . s- ,v 4~.:
-,c g y ... . - e ,.y, .:yy . .. ,y, ., .(<
, r .
e i p
L e .-
=e4. : , 'y g w *- / '$ $
3,M,A M'i
^
a /p-MW7, (L o t: .
- , e: - ~J -p 5
,, p , T' - E f? ?
\e ~ - -
. n. .
. , .a
- w. .
h'..
} ,W"' '.$ Q ** & ?'e.#. *l$$5h.?,
,. $~.h?. . _.m . ~ ._ .M' { j 'e % M , '*.,[.r) s , , pY a
",} MN ^ k.'t&, ,t.,&,.g,, o g j %',*l s . .. ~ ~. . ... .p... .
.w %. +.
ee1 x x. ao m em Lao n .c s orv.ee :s,.nisM es n;~ . ~ .y -w.. .. . .. . u g e p * . n .m a ee~ s a em. was
., a p ,W n "TW
.,n, : .a n3w,2
\ l $ ,,. U h?, #
~W l s
,,r
.a u .a . - .
r n'
..C A % %. g. w I m,.
. {,- a y 's ; .., .
o, v . e ,,e ,? ,4. J. v.f aW- -
., r,- e s_. .,,
s, 2e. ,..u, y ,+ . .. , . v_.34 - , - < .
.- p. p 9
- c. m, g .-- mf-f...M. . .'n#
.,.y m
$( D M 'W 44
~
r
- g i.e. 4 - a O . 2 . - - vt -: , g s- F 3
3 Mr#PMWe$. 4 - s Lth $..JM.7M. . - r .g .gM,;.U.d.u ttNp
- p.,** *IWWM
, .. .- o NM. , i.k L u. f r
+
%N.@. m.,. N ...E
..m:n_. . . paeZ. ,. . r.,ep t s
'pa,%".%w .woma n ,y w w.--
.w a.$*x.ssgnm mem. n.p.e
. o e. .yy . y r.a, w w-. . , . 7..
x -. .
.... n .. .. . . . ,
g ..
)*
.e, 3 -.z.... g y <$.p-v,.k..h.D../.k.jk s
w h ,y
,.m n . .a,@. m.N y.
%H*
r m 1.. :, %.18% A".k e.w
, $[*h ' , " [3 . ~ bw,,w. .hY a.-y h , w;e' ..
k'mk*y' o.*I4.,.[g~,w., C 7.Mr[*$
.n
. m o . ~.W p, q,e -- - .g c.g.pu. .m .,4 4f..a g u.g sn < gw .u
. q. . , g . yp%.mc .p;is e z..g w.w.u. h,4 iM:'f fg x .h&.g.,SD
,y W. D n!
<TJW'gg@y %yP yWy-ragiMff;,_3 N5 % . 2 - G .Wk
& & % Whk A. N&e
*t W N '." M ** % ~ll4.-
.~.T*-Y" .
v e p# _?fd.' . # '
.y; .W v g q g g 4, r .3~Y %.'f%# d_h C h G.M
&.g Uhh @ku,,
q,. .e f 'f R . (j' f ~.MMA.WaeE::$MdUiWWu%vw.hn.i:: r. ,6 MM,ehors WM/5yf6M A. W @sMu'M F v MTV.M.w m4Fi.i.v T.M% m.c.a &%gEwe.mc
>mLm v.s ww km s.mwpg .
m w%as.,,. s w.; w m w.w.n.sM. . - m.nn w . m m.u w..n.*&n e y. m% p w: m t,w~z:..w. m
' ~
dagab.:,.$, M
.-A.h~ Jet'*M/ M L.i %J--N 5 d.w.7.'*.[ O M,I M iad h fr, M ..o'.- W $ -~. 'W.. . N *Cf'i J,</ .w.M M-f. E AM. 7,.*
4 , - Q . n '; N Q , s; *. w***.'r' <a& AQ..f m ,~.- .
~ . .
% s. he*,M*f.h,s*.,,.g-t
(.'*l %Q. * ',#) yew'.*<, [s a r tv
-" .* A an.
4 . . e e e + - -.hpmp
.s e.4
- e. '%
M ., . -
.r..+ - -
s
.+ 1,
. M
,-3 m e_
y 8 % ,Q' w;g, N j
? T t" 5
% <g h .
b@ he,* N w. @1, . NUM. *1***v., %.}Yh[4N> : .-d.g *J@ M k'p N#'$ a mM. p&.w m.M,w.,MI ge9$, s.,- ,M. 5. , $iNM.w..g*,N, #.e.v.s tw w..#.
. 3 f -
3 s s - - . , % . ,-- mm , .g 5 .
+ ,w ,s p .,
1,+ I'%L. h. \ 3
' Gb T r, WS C'Yy 4M5 O.? d C-U.'*
bin:h &.&e'%p*
- need A o.e r .c ate ~ w &'* ' . q[^ 5"" A.~; iffW.NQ Y tP Eb
%"*4W%u 'k[p. #.w.w.4. r.a.,w.'ym, m A u.w.M.:.5& M w .t7'ji>
w42 I* M h N p
%.'GM
& M*T""
- f % p#
M u.% *'"[ W %" 4
- Qg
'M r A.w . .
. g. - yW. c.m ;.,,, a c-- CA< .._m %d .y0 c m. e.w? d.? . Aeg, y- . ". - b Y ?14 - - _~
j.
,pw u, .a-. w .i.y,. .w i pos M.a 4
'y.1,- -.-m x
, , .v. ,-w4., .ys .( s,.,~. -v f.,n. s
. k.y 4.h.
r 40%.*
*. . ~.
- ys &, , .>.3..
v.. a , y gx 4-. . ~r ,&,, %' m. ~{ %. s
.)pw,, Q.
wm, u.ib, f% q
^
3
. p% us 2. .
' pp. i m ,. sMc .,
w.< 1 9 MM..'Nd Ah,,. ,/hwG m , y.imy.M.e"$' dk MLP 'k .A,'*%dAmAdh' S. ;,N; M M ,4
. . y . :Nd 5M w-
+ a.w .
5..w. ..a ~. .-.
. p. g,q.w y
, A .n ...n.
u
.- t
. .x & p',
. .g g
v ew,, n<. l &.vw.- wm w. +
.ny. ,. . . m . ..m, - , , ,s p u ,, s-t p,. +w7Has r
w m. h.s u 4 ~. . g.%% . , e&. .. p w..-.a.
.o* *
,s. w - . ~*w, 4 oy :
- q , .ms, Emi.
- . y. ..m*
.go u,. %
yn -..: v n,.m e. '.p .,w -x 3* ., m . ss,x %'. n.trof.o,,e. s, . .w[x,; v. w,ww : .'. ,rJ.t.c, ny -n.m%:<f , .. . .nv gy &myl;, wwws-n. ..u A . < a .. v4 ~g~ .
~ . u.a m v em. 3 pu.w,wy.m.. p 4 we yde us.H m 4 * . *4 w ..=e ewasseen we,.eenamr s.v .%n e.* w es n 4.. ,s
^
m o ,,
,,u... -y.p..
m,r a
. @.. nf m7 . .. ;. .. -a x,,f,-
.7."..,",J t W.- w e
.~. p ,,, / s.- ,,.g. s *wv4. c.. .d. 4j.g g g.y, .W, .~. ,,?
,,,,,y~,,.
<,-~ .m . . .m wa u M .,m m.,, w-
,s ~d'_,. ...wg, 4,. . . ..,,,-%.Si, .w.,.7, .
.a ;% ,,;o -p,.g,. p. . er
. po ,.me. ,,. ,r. ..
meD#* .W;,. . ., O,"'.
~c 's.
s.) . ' y e " e Q ham.. p- eng -**gd e.s* ..s
.e, e 4. A. ip tgs u, , y .d'-s- s ' A. ,,.-g.i a .%
p ia.* d*TN"*'W**"" k' * % F , o**** % , T' *" .* , % M* - -'7*/hr'M_ p ,9 & * *- -* mhF- " .n.s-s. -s m ,_ ****X er #. "i F',r.d" V*t 4 d d Q *F % D'e.ei h
's., . . , *- '
me .q4. ,4y~ = n. ,' ?' 'MeP4. v-*. y . ~~ r -2L vg yr;" as e Uy
-u of 5 a- h h .ad,NIV,fsc N .A,a.4.,,L.efid.F F-Q ._yrfet.*.Q. ..:././. .
~ - ;."&,.W.16C,A.. C u N.tga4cQJ
.. ,n,...
'*N,,.#E~
.m%
.1 M * ," % \ D*-'"%% ',, ' ? h *' D'"'** k * , 'FU *
- Y. "'/* O 1 M.#**.77' - /*74 5e
[.
.&.Ydh.,$. .4
- m. , 5
,. e
. * ,.v, A..gn.,
3 & f.hh w w ,.. -- . 4 ..;. . E 4.,. ,
. h. h
%,i f,f
- s.4.V.T*'hg75 4 p-e,a 7,-v. - ..% . w , .~,,,
* % , fh ,
wm ;3_. nw . -dw.a J u
; y
%C , v+ -. m'M M w. % . LNf.;',( d,a- n ; ' M e. O g - '
' c'i ** '
~, , " : r$aMC,@,7hm4N. 4QMW,. w*tM
- j. h,
, .o, 3 E ev' w w
.~ .
* ' - g s . ..- %.*y Jun > Wp.,.,4
. # *"~#
\h
. , N a' * * .w ., . s- p-
/
& .:. l. ;,.c.,-Q
~2
\. ~ ' ? *f ) L's ..+.s .. . . . M
,',w . . ~' . 5 2: k. 3. - i &* . * :'~ \ .) .' . . .. : ? . , 'g
~ cf.4 _,
g
,....u.
%. s . # .,. ,, . .,. ..,?
g g e % *~ =WN.=e e- ' mln- .o m. a. a. ~, ..,.n. s . %.,e
. w
- u
-.m,,,.
. -_ ,, _~,,.,,.,.-
a 4a- w
"' a n.m. . . b _w . . .,,
,%im-
- e. * % . gym,
.. g .,e 4 pee . ' w
[ '.% . p
. _ - - w.g..et..
4 -% .,n
- :s . . .oo _ p. ,
s m.a. W. =-..,. w w .4- w s. . . . Mkp-.gr. d. 4'
. . ~ .
3,r. m .~ ~ [D
' , . . .-- hm m .m vc ny w ~p h, S*gp ~ m p.g,-.IA .s ,Ar.cM -s iI --p g, : .-+~rs.-4 'b m% . .
b e ed ' Wm . , e C# ~*O 'daa
- a4
_*T'm a ' hbn'.% Mk.J.h,$pg g
-l E P
$. ' ' ' W $@ kA #.'M N MNAJ 'rWOTTFW t.M@.97#'* h LQhhh[,M.jf@ @ M 'G &fDW M ~
$.2 NW$t
- u. eii @wa?%h$s@l..,&s$$hWWWi$W.WWWWWWWWN sMimd 7 WygwmgW e W Wins .
3 W M. m 6 WWW m
<i .
kNh M M N $Ned!
O O O j S0)IUF IGNI ::0\ ~~ E 9 E R A~U RE l DEFINITION: THE MINIMUM TEMPERATURE AT WHICH A SELF-SUSTAINING TEMPERATURE RISE OCCURS N O T MPERATURE CC) FRESH POOL, UNDISTURBED 250 - 400 FRESH POOL, DISTURBED SURFACE ~200 EXTINGUISHED POOL AS LOW AS 25 SPRAY DROPS ~120 CFUNCTION OF DROP SIZE)
~
o o o AEROS0_ RE_ EASE FRACTION OF REACTED No RELEASED AS AEROSOL 300 _ 7:: RE 0.'5 - 0.3 S3 RAY ::RE
~'
.0
~
9
~
O O O v SO ] :: A\) \a < =::R E A E ROSO _ ~~0X::C::~~Y THRESHOLD LIMIT VALUE* FOR 8-HR EXPOSURE NaOH,Na2 Ox ,KOH 2 mg/m3 TIME-WEIGHTED AVERAGE EXCURSIONS TO 4 mg/m3 PERMITTED Na2 CO3 ,K 2CO3 NOT GIVEN: STATED TO BE A SLIGHT IRRITANT
- AMERICAN CONFERENCE OF GOVERNMENT INDUSTRIAL HYGIENISTS, 1981.
i W F
hEACTIONS DURIN(r ATMOSPRERIC DiFFUSl0N Waz o Na OH = Na,CO, Na, 0, few secoans 30 sec ro nr A.H. >30% FEw MINUTES
- n n n. > 407
- SLOW i V i
NahC03 Yn
a o o o ; V::S A_ RA\GE LR::\G A S0X: v 7:: R_r
~
t A EST ABI
- TEST AB2 :
ABI & AB2 BEST FIV . ; l === REFERENCE 14
~
b IM - 4 % i '
\ : -
N ! a : -
\ -
U -
~
*s g . \ . i
=> a .
D 1 0 ; l E 10 -. D
- 3. 6 C-0. M p -
- D = ME :
- C=g/m}RS -
.- a .
o 1 10~I l 1 10~1 100 10 l 3 l SUSPENDED MASS CONCENTRATION, g 10TAUm
O 0 6 l I Ef LIQUID METAL FIRES BRIEFING FOR eums
. ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS)
CRBRP SUBCOMMITTEE LIQUID METAL FIRE MITIGATION APPROACH PRESENTED BY C.J.BOASSO - SYSTEMS ENGINEERING WESTINGHOUSE CRBRP PROJECT JUNE 25,1982 6 82 2888 2
O O O LIQUID METAL A) FIRE MITIGATION APPROACH
- LIQUID METAL FIRES MITIGATED BY ,
CONTROLLED ENVIRONMENTS STRUCTURAL FEATURES ' SENSITIVE LEAK DETECTION SYSTEMS IN CONJUNCTION WITH OPERATOR ACTION
~
~
. CONTROLLED ENVIRONMENTAL l FEATURES l
- CELLS CONTAINING RADIOACTIVE LIQUID METAL ARE INERTED WITH NITROGEN TO MINIMlZE SODIUM BURNING
- ALL LIQUID METAL SYSTEMS SUPPLIED WITH INERT ARGON COVER GAS i
!L \ - STRUCTURAL FEATURES ! o STEEL CELL LINERS IN INERTED CELLS TO CONTAIN SPILL VOLUMES o INSULATION BEHIND CELL LINERS TO PROTECT j STRUCTURAL CONCRETE j I o CONTAINMENT / CONFINEMENT STRUCTURES TO CONTAIN o STE CATCH NSI AIR-FILLED CELLS TO CONTAIN j SPILL VOLUMES ) o STEEL CATCH PAN / FIRE SUPPRESSION DECK SYSTEM IN ' l AIR-FILLED CELLS TO MINIMlZE LONG TERM SODIUM BURNING o INSULATION UNDER CATCH PANS TO PROTECT STRUCTURAL CONCRETE l 6 42-2732-8-1 i
l 4 O O O i
~
i CRBRP INERTED CELL LINER SYSTEM (~ ! 1/4" GAP
- 'g c r l h6 'dY
! Pdo l i
#88O : STEAM VENT 4
~?p*,o.?i s.o,*6 c.g -lNSULATING CONCRETE
- ??.0,*dI o j NELSON STUD do
9,.
= 3/8" WELDED STEEL CELL LINER i
- i..hD@d.'
.. h ^
l 00 ~o ~~p 3
03:,0J
, < p, g. 4 ,g lNSULATING t#,i 2))Etf.&%Y#d.'*6.Si 2 9, ye CONCRETE c - .-)?
- i,%gog?$
$Du %id$bWE1?$Y""o$5$5hof s IQ n 4'u -x?S$
SUPPbRT BEAM j STRUCTURAL CONCRETE = $ ,
Il O O O
- AIR-FILLED CELL CATCH PAN SYSTEMS l
l
- OPEN CATCH PANS 1
i 1
- OPEN CATCH PANS WITH DRAINS TO CATCH PAN i
WITH FIRE SUPPRESSION DECK !
- CATCH PANS WITH FIRE SUPPRESSION DECKS l l 1
I , i I i
W W W CRBRP OPEN CATCH PAN FOR AIR-FILLED CELLS
~
STRUCTURAL CONCRETE STEEL GRATING o . Odd. , t asp -=p INSULATING _ _ PANEL 3/8 IN. WELDED STEEL AIR GAP ,
/ CATCH PAN PLATE VBWAV/ / A
#dd$ Pg*
STRUCTURAL CONCRETE W 682285 7
\ l i :
r " r O p_ E M R E I T ' L E FS V _ OY E - L T S a N _ N K I C
/ I
?
AE O RD - DN = - NO E P AS I - I - P PS N - E O H R N I C P T P A P H S\ S E RK AU C T PC PE C S A C UD S E R O m ' I F , _ t 0 1 2 3 7 2 2
~
8 2 f l11l '
O O O CRBRP CATCH PAN / FIRE SUPPRESSION ~ ~ DECK SYSTEM FOR AIR-FILLED CELLS STRUCTURAL CONCRETE
'" ====m STEEL GRATING
- a n - u na a a a = ====,
,o - l= = -
o
~l FIRE i
i- SODIUM FLOW UPPRESSION T DECK of~k n r(VENTn a on n q 'o nn i l '
,rg J n
A' i1 i: l t l i i t J INSULATING _ _ J.; g i
,1 1 J PANEL GAS E S' ' _ _ f _ , . I _I _ _ _ _f, ,_. , , _ _ _ , _ , _ _ _ _ _ _ _ _ _ _ _ _ _ , . , _ _ , _ _,__ , ,,
3/8 IN. WELDED STEEL ll' ll DRAINS 1-1/2 DIA. I AIR GAP j _
'gj [ (STAGGERED) / CATCH PAN PLATE
//
'Bb*b "fg! STRUCTURAL CONCRETE
LIQUID METAL FIRES j BRIEFING FOR REA TOR SA EGU RDS ACRS) CRBRP SUBCOMMITTEE I ! EVALUATION OF DESIGN BASIS SODIUM FIRES PRESENTED BY C.J.BOASSO 4 SYSTEMS ENGINEERING WESTINGHOUSE CRBRP PROJECT ) JUNE 25,1982
""1
EVALUATION OF DESIGN BASIS LIQUID METAL FIRES IN INERTED CELLS l l 2-82-2732-11 {
OVERVIEW OF LIQUID METAL FIRES IN INERTED CELLS
- o LIQUID METAL SPILL EVENTS EVALUATED IN ALL CRBRP INERTED CELLS
- 19 INERTED CELLS IN RCB
- 16 INERTED CELLS IN RSB 1 INERTED CELL IN SGB o SPILL CONDITIONS (TEMPERATURE, FLOWRATE, VOLUME?
SELECTED TO PROVIDE MAXIMUM CHALLENGE TO CELL STRUCTURES o SPRAY AND POOL BURNING CHARACTERISTICS EVALUATED o PRESSURE, TEMPERATURE, AND AEROSOL CONDITIONS INTEGRATED INTO BUILDINGS DESIGNS 2-82-2732-12
RANGE OF LIQUID METAL SPILL , CONDITIONS IN INERTED CELLS
- SPILL TEMPERATURES FROM 995 F TO 400 F
- SPILL RATES FROM 950 gpm TO 6 gpm
- SPILL VOLUMES FROM 35,000 GALLONS Il250,000 LBS) TO 800 GALLONS i:4,300 LBS?
l l l 0 6022888 6
O O O SPRAY FIRES o CONSERVATIVELY ASSUME THAT 100% OF DISCHARGED l SODIUM OR NaK IS CONVERTED TO DROPLETS l o DROPLETS REACT WITH AVAILABLE OXYGEN IN CELL o IN LOW N2%) CONCENTRATION OXYGEN ENVIRONMENT AND SODIUM TEMPERATURES ~1,000 F, COMBUSTION IS l NEGLIGIBLE COMPARED TO TRANSFER OF SENSIBLE HEAT l TO GAS ATMOSPHERE l o HEAT TRANSMITTED BY CONVECTION, CONDUCTION AND l RADIATION FROM BURNING DROPLETS TO GAS ! ENVIRONMENT TO CELL STRUCTURES l , l l
POOL FIRES o POOL FIRES OCCUR WHEN LIQUID SODIUM (>400 FJ ACCUMULATES ON CELL FLOORS. o BURNING OCCURS AT POOL SURFACE o HEAT TRANSFER VIA CONVECTION, CONDUCTION, AND l RADIATION FROM POOL SURFACE TO GAS ATMOSPHERE ! AND FROM THE GAS ATMOSPHERE TO CELL ' i STRUCTURES. HEAT TRANSFER VIA CONDUCTION FROM SODIUM POOL TO CELL STRUCTURES l o IN LOW (< 2E CONCENTRATION OXYGEN ENVIRONMENT AND SODIUM TEMPERATURES ~1,000 F, COMBUSTION IS , l i NEGLIGIBLE COMPARED TO TRANSFER OF SENSIBLE HEAT TO GAS ATMOSPHERE I
O O O SPRAY FIRES THERMAL CONSEQUENCES
~
- PEAK GAS PRESSURE AND TEMPERATURE OCCUR SHORTLY (<10 MINUTES) AFTER INITIATION OF SODIUM DISCHARGE
- INSIGNIFICANT BURNING DUE TO LOW (<2%) OXYGEN l CONCENTRATION
- SPRAY PHASE CONCLUDES AT END OF SODIUM !
DISCHARGE (<10 HOURS) i
~
PHTS CELL GAS TEMPERATURE (SPRAY PHASE) ' GAS TEMPERATURE ( F) 700 1 1 M - g .. ( 400 - . 300 - 200 - 100 1 i i 1 0 10000 20000 30000 40000 TIME (SEC.)
O O O PHTS CELL GAS PRESSURE (SPRAY PHASE) GAS PRESSURE (PSIG) 20 15 - 10 - 5 - 0 O 10000 20000 30000 40000 TIME (SEC.)
O O O
~
LONG TERM POOL THERMAL CONSEQUENCES IN INERTED CELLS
- THERMAL EVALUATIONS PERFORMED FOR EXTENDED TIME PERIOD (TENS-HUNDREDS OF HOURS)
- SPRAY END-POINT THERMAL CONDITIONS SERVE AS INITIAL BOUNDARY CONDITIONS FOR POOL EVALUATION
- ESSENTIALLY NO BURNING OCCURS
- CELL GAS PRESSURE AND TEMPERATURE CONDITIONS DECREASE '
l
O O O PHTS CELL FLOOR TEMPERATURE FLOOR TEMPERATURE ( F) 500 STEEL LINER 400 -
--- 1.5" INTO STRUCTURAL l CONCRETE l l
300 - 200 - 100 - 0 O 10 20 30 40 50 60 70 80 TIME (HRS.) 2 82 2732 23
l
~
AEROSOL GENERATION IN INERTED CELLS
- LOW K2% ) OXYGEN CONCENTRATION RESULTS IN SMALL K600 LBS) l QLIANTITIES OF SODIUM AEROSOLS
- AEROSOLS CONTAINED WITHIN INERTED CELL BOUNDARY
- HYPOTHESIZED RELEASE OF ALL RADIOACTIVE AEROSOLS RESULTING FROM POSTULATED SODIUM FIRE EVENT TO RCB OR RSB ATMOSPHERE RESULTS IN DOSES WITHIN 10 CFR 100 GUIDELINES
, g I
EVALUATION OF DESIGN BASIS ; LIQUID METAL FIRES IN AIR-FILLED CELLS 2-82-2732-26
O O 9
~
OVERVIEW OF LIQUID METAL IN AIR-FILLED CELLS
- SAME APPROACH AS FOR l INERTED CELLS. SPILL EVENTS EVALUATED FOR ALL AIR-FILLED CELLS l 6 CELLS IN RSB 17 CELLS IN SGB 3732-27-1
O O O
~
RANGE OF LIQUID METAL SPILL CONDITIONS IN AIR-FILLED CELLS ,
- SPILL TEMPERATURES FROM 940 F TO 400 F
- SPILL RATES FROM 1000 gpm TO 45 gpm l
l
- SPILL VOLUMES FROM 39,000 GALLONS ll302,000 LBS) TO 800 GALLONS ll4,300 LBS)
""~'
SPRAY FIRES o CONSERVATIVELY ASSUME THAT 100% OF DISCHARGED SODIUM OR NaK IS CONVERTED TO DROPLETS l o DROPLETS REACT WITH AVAILABLE OXYGEN IN CELL o IN HIGH (21E CONCENTRATION OXYGEN ENVIRONMENT AND SODIUM TEMPERATURES ~1000 F, COMBUSTION IS SIGNIFICANT IN TRANSFER OF HEAT TO GAS ATMOSPHERE
- HEAT TRANSMITTED BY CONVECTION, CONDUCTION AND RADIATION FROM BURNING DROPLETS TO GAS ENVIRONMENT TO CELL STRUCTURES
POOL FIRES
- POOL FIRES OCCUR WHEN LIQUID SODIUM (>400 F) ACCUMULATES IN IN CATCH PANS AND ON SURFACE OF FIRE SUPPRESSION DECK
!
- BURNING OCCURS AT POOL SURFACE AND ON SURFACE OF FIRE SUPPRESSION DECK
- HEAT TRANSFER VIA CONVECTION, CONDUCTION, AND RADIATION l FROM POOL SURFACE TO GAS ATMOSPHERE AND FROM GAS ATMOSPHERE TO CELL STRUCTURES. HEAT TRANSFERRED VIA CONDUCTION FROM SODIUM POOL TO CELL STRUCTURES
!
- IN HIGH (21%) CONCENTRATION OXYGEN ENVIRONMENT AND SODIUM j TEMPERATURES ~1000%F, COMBUSTION AT POOL SURFACE
! IS SIGNIFICANT IN TRANSFER OF HEAT TO GAS ATMOSPHERE.
- POOL BURNING SUPPRESSED BY FIRE SUPPRESSION DECK WITHIN 36 HOURS j!
l 1 i
l i O O O l SPRAY FIRES THERMAL CONSEQUENCES 4
- PEAK GAS PRESSURE AND TEMPERATURE OCCUR ,
! SHORTLY (~2 MINUTES? AFTER INITIATION OF SODIUM ! DISCHARGE.
- SIGNIFICANT BURNING WITH GENERATION OF DENSE AEROSOLS DUE TO HIGH OXYGEN CONCENTRATION.
l
- SPRAY PHASE CONCLUDES AT END OF SODIUM i
DISCHARGE (~6 HOURS) l 2-82-2732-31
~
POOL FIRE THERMAL i CONSEQUENCES
~
l !
- POOL FIRES OCCUR OVER EXTENDED TIME PERIOD (TENS OF HOURS). BURNING UNDER FIRE
! SUPPRESSION DECK SUPPRESSED WITHIN 36 HOURS. l!
- SPRAY BURNING EFFECTS INTEGRATED INTO l POOL FIRE ANALYSIS.
!
- SIGNIFICANT BURNING OCCURS (N4000 GALLONS I
OF DISCHARGED SODIUM BURNS IN SGBD
- CELL GAS PRESURE AND TEMPERATURE j CONDITIONS DECREASE.
1
.n mun ,
O O O CELL GAS TEMPERATURE IHTS PIPE LEAK , TEMPERATURE ( F) , 700 600 - 1 t 300 - ) 200 - i e' 100 -
''I ' ' 'I i' I I ' ' 8'l i
0 2 10 100 1000 10000 i TIME (s) i sar 2n2 m a
O o O PEAK FLOOR TEMPERATURE FOR !~ l IHTS PIPE LEAK l TEMPERATURE ( F) d = DISTANCE BELOW SURFACE 250 - d = 24 IN. i l 200 - 150 - 100 - I i ! 50 - t
' ' I i i i i i i i i i
- O 2 6 10 14 18 22 26 30 34 38 42 46 50 l TIME (HOURS)
O O O l AEROSOL GENERATION IN AIR-FILLED CELLS
- HIGH (21%) OXYGEN CONCENTRATION RESULTS IN LARGE (THOUSANDS OF LBS? QUANTITIES OF SODIUM AEROSOLS.
- AEROSOLS ARE VENTED FROM SGB (MAXIMUM OF 630 LBSD TO MAINTAIN PEAK PRESSURES WITHIN ACCEPTABLE LIMITS.
CRBRP AEROSOL MITIGATION SYSTEM VENT TO ATMOSPHERE h s
" g CLOSURE DAMPERS 2 PER STEAM M 4 (LOOP)
VENT
, , , ?> , i e 'Mi " " ' SuOxE
# -@ / /
INLET v (2 PER LOOP)
$$$>E" 2 SETS PER LOOP)
FIRE DAMPERS bb (2 PER EXHAUST LOOP) FIRE DAMPERS (2 PER LOOP)
;aG% q:
s' . IHTS PIPING REACTOR CONTAINMENT STEAM GENERATOR BUILDING BUILDING 12 Q12642 251 e ill ii , . , , .
O O O
SUMMARY
OF EVALUATIONS o PLANT CELL STRUCTURES DESIGNED TO ACCOMMODATE CONSERVATIVE SPECTRUM OF DESIGN BASIS LIQUID METAL SPILL EVENTS. o NO IMPACT ON CAPABILITY OF PLANT TO MAINTAIN SAFE SHUTDOWN CONDITION o NO IMPACT ON PUBLIC HEALTH AND S'AFETY AS A RESULT OF POSTULATED DESIGN BASIS LIQUID METAL SPILL EVENTS 2-82-2732-36
LIQUID METAL FIRES CONCLUSIONS
- PHENOMENOLOGY OF LIQUID' METAL FIRES HAS BEEN EXTENSIVELY INVESTIGATED AND IS WELL UNDERSTOOD.
- APPROPRIATE POSTULATED SPILL EVENTS HAVE BEEN SELECTED TO CHALLENGE PLANT STRUCTURES.
- CONSERVATIVE EVALUATIONS OF THESE SPILL EVENTS RESULT IN NO UNACCEPTABLE CONSEQUENCES.
l O}}